Copyright is owned by the Author of the thesis. Permission is given for a copy to be downloaded by an individual for the purpose of research and private study only. The thesis may not be reproduced elsewhere without the permission of the Author. GENETIC IDENTIFICATION AND EVOLUTION OF EPICHLOE ENDOPHYTES A thesis presented in partial fulfilment of the requirements for the degree of Doctor ,of Philosophy III Molecular Genetics at Massey University, Palmerston North, New Zealand Christina Diane Moon 1999 1I ABSTRACT The Epichloe endophytes are a group of filamentous fungi that include both sexual (Epichloe) and asexual (Neotyphodium) species. As a group, they are genetically diverse and form mutualistic to antagonistic symbiotic associations with temperate grasses (subfamily Pooideae). In this study, a multi-locus microsatell ite-based peR fingerprinting assay was developed for the genetic identification of Epichloe endophytes, both in culture and in planta. Microsatellites were i solated from endophyte partial genomic libraries, or identified from existing endophyte DNA sequences, and peR assays that amplify these loci were developed. Multiplex assays were optimised, and fluorescently labelled primers were employed to allow precise sizing and automatic analysis of the peR products with a laser scanner and the appropriate software. A reference database of allele sizes has been establi shed for the panel of endophytes examined, and it has been shown that this assay is able to resolve endophyte groupings to the level of known isozyme phenotype groups. In a blind test the assay was used successfully to identify a set of endophytes in planta. The segregation of microsatellite alleles from an E. Jestucae sexual mating was also examined. This microsatel lite assay, in addition to B-tubulin and ribosomal RNA gene sequence analysis, was used to genetical ly characterise Neotyphodium-like endophytes from annual Lolium ryegrasses and Hordeum grasses. The endophytes examined were indeed found to be Neotyphodium with unique evolutionary origins. The common endophyte of annual ryegrasses was found to have a unique hybrid origin involving E. baconii and E. bromicola ancestry, and it is proposed that this taxonomic group is named LmTG- l . A second previously undescribed Neotyphodium endophyte was found in L. canariense, and this was shown to be an asexual derivative of E. typhina. The name proposed for this taxonomic group of endophytes is LcTG- l . The two Hordeum endophyte i solates, HaB and Hdl , were also both shown to have unique hybrid origins. HaB has E. elymi and E. amarillans ancestry, and Hdl has E. typhina and E. bromicola ancestry. The proposed names for the Neotyphodium taxonomic groups that contain these i solates are HdTG- l and HdTG-2 respectively. The revelation of interspecific hybrid origins for three more i i i Neotyphodium endophytes il lustrates the prevalence of hybridisation in the evolution of the Epichloe endophytes, and further contributes to our understanding of the evolution of thi s group. IV ACKNOWLEDGEMENTS I would first l ike to express my most heartfelt gratitude and appreciation to Professor Barry Scott, who has provided me with excellent guidance, support, and valuable opportunities throughout my degree. Thank you Barry! I am also extremely grateful to my second supervisor, Dr Brian Tapper, for valuable discussions and advice, and representation of Grasslands AgResearch, who have been most generous throughout the project. To Mike Christensen, for his inspirational enthusiasm, encouragement and valuable discussions - thank you! Thanks also for the provision of cultures and grass samples, especially for the annual ryegrass endophyte project. I also owe valuable thanks to Dr Chris Schardl for allowing me to study in his lab at the University of Kentucky, and exposing me to the world of Epichloe biology and phylogenetic analysis. I am indebted to the Schardl family for their exceptional hospitality during my stay in Lexington, thank you. I am very privileged to have worked with such a supportive group of dedicated and friendly people. To all the folks in the Scott Base lab over the years, a big thank you!!! Carolyn, Austen, Emily, Lisa, Xiuwen, Michelle, Renae, Raj , Lisbeth, Jo, Karyn, Eugenie, Mike, and the folks in the Bradshaw lab, thank you for your help and advice, and for creating such a supportive and social atmosphere which has been a pleasure to work in . I would also l ike to thank the folks in the Schardl lab - I had a most enjoyable stay, and look forward to returning soon. My thanks also to the staff of the old Dept. of Microbiology and Genetics, the Dept. of Plant Pathology at UK, and the Institute of Molecular B ioSciences, especially Lorraine Berry for the ABI analyses, and Dr Peter Lockhart for valuable constructive feedback on the phylogenetic work. To Wayne, Mum, Dad, Selena, and Deborah, thank you so much for your love and support over the past three and a bit years. I would especial ly l ike to thank Wayne for his computer expertise, which has been particularly valuable in these past months; and also to both Wayne and Selena for helping me compile this at the last minute!!! Many thanks to Michele, Lou, and Lisa for the good times, and welcome study distractions . v This work would not be possible without funding from Grasslands AgResearch. Thank you for generously funding thi s studentship, and giving me the opportunity to undertake thi s degree. My thanks also to the Institute of Molecular BioSciences, the Massey University Graduate Research Fund and Dr. Chris Schardl for additional funding to allow me to study at the University of Kentucky in 1998. TABLE OF CONTENTS ABSTRACT ACKNOWLEDGEMENTS TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES CHAPTER ONE - INTRODUCTION l. 1 THE LIFE CYCLES OF EPICHLOE ENDOPHYTES l.2 ANTIHERBIVORE ALKALOIDS 1.3 GRASSIENDOPHYTE SYMBIOSES 1.4 TAXONOMY AND EVOLUTION OF EPICHLOE ENDOPHYTES l.4. 1 Epichloe endophyte taxonomy l.4.2 Epichloe species 1 .4.3 Neotyphodium species 1.4.4 Interspecific hybridisation events 1.5 THE GRASS SUBFAMILY POOIDEAE 1.6 EPICHLOE ENDOPHYTE GENOMES 1.7 DNA FINGERPRINTING l.7. 1 Fungal identification 1.7.2 DNA fingerprinting strategies 1.7.3 Fungal microsatellite loci 1.8 OBJECTIVES VI i i IV VI XV XVI 1 2 4 6 8 8 1 1 15 16 18 20 22 22 23 27 28 vii CHAPTER TWO - MATERIALS AND METHODS 29 2. 1 BIOLOGICAL MATERIALS 30 2. 1 . 1 Endophyte Isolates 30 2 . 1 .2 Bacterial Strains 30 2 . l . 3 Cloning Vectors 30 2.2 GROWTH OF CULTURES 36 2.2. 1 Isolation of Endophyte from Host Tissue 36 2 .2 .2 Maintenance of Endophyte Cultures 36 2.2 .3 Endophyte Liquid Cultures 36 2.2.4 Escherichia coli Culture Conditions 37 2.2.5 Phage M13mp 19 Culture Conditions 37 2.3 MEDIA 37 2.3 . 1 Luria Broth Media 38 2 .3 .2 Potato Dextrose Media 38 2 .3 .3 SOC Medium 38 2 .3 .4 Top Agar 38 2 .3 .5 2 x YT medium 38 2 .3 .6 Media Supplements 39 2.4 COMMON BUFFERS, SOLUTIONS AND REAGENTS 40 2.4. 1 Common Stock Solutions 40 2.4.2 Acrylamide Mix 40 2.4.3 DIG Hybridisation and Detection Buffers 41 2.4.3 . 1 Standard Hybridisation Buffer 41 2.4.3 .2 Maleic Acid Buffer 4 1 2.4.3 .3 Washing Buffer 4 1 2.4.3 .4 Blocking Buffer 4 1 2.4.3 . 5 Detection Buffer 4 1 2.4.4 Ligase Reaction Buffer 4 1 vii i 2 .4.5 Lysozyme 42 2 .4.6 RNaseA (DNase Free) 42 2 .4.7 SDS Loading Dye 42 2.4.8 Southern Blotting Solutions 42 2.4. 8 . 1 Depurination Solution 42 2 .4 .8 .2 Denaturation Solution 42 2 .4 .8 .3 Neutralisation Solution 42 2.4.8 .4 20 x SSC 42 2.4.8 .5 2xSSC 43 2.4.9 TAE Buffer 43 2.4. 1 0 TBE Buffer 43 2.4. 1 1 TBE Sequencing Buffer 43 2.4. 1 2 TE Buffers 43 2.4. 1 3 Tris-Equil ibrated Phenol 43 2 .5 DNA ISOLATION AND PURIFICATION PROCEDURES 44 2.5 . 1 Large Scale Isolation of Endophyte Total DNA 44 2 .5 .2 Microwave Miniprep Isolation of Endophyte Total DNA 44 2.5 .3 Total DNA Isolation using a FastDNA Kit 45 2 .5 .4 Plant DNA Isolation using a CTAB Method 45 2 .5 .5 Plasmid DNA Extraction using a Rapid Boiling Method 46 2 .5 .6 Plasmid DNA Extraction using an Alkaline Lysis Method 46 2 .5 .7 Plasmid DNA Extraction using a Quantum Miniprep Kit 47 2 .5 .8 Single-Stranded M13mp 1 9 DNA Isolation 48 2 .5 .9 PCR Product Purification 48 2 .5 . 1 0 Isolation of DNA from Agarose 49 2.6 DNA QUANTIFICATION 49 2.6. 1 Fluorometric Quantification 49 2.6 .2 Mini Gel Method 50 2 .7 RESTRICTION ENDONUCLEASE DIGESTION OF DNA 50 ix 2 .8 AGAROSE GEL ELECTROPHORESIS 5 1 2 .8 . 1 Agarose Gels and DNA Size Markers 5 1 2 .8 .2 Mini Gel Electrophoresis 52 2 .8 .3 Large Gel Electrophoresis 52 2 .8 .4 Staining and Photographing Gels 52 2 .9 POLY ACR YLAMIDE GEL ELECTROPHORESIS (PAGE) 52 2 . 1 0 SOUTHERN BLOTTING 53 2. 1 1 SOLUTION HYBRIDISATION USING DIGOXYGENIN (DIG) 54 2. 1 1 . 1 DIG 3 '-End Labelling of Oligonucleotide Probes 54 2. 1 1 . 2 Hybridisation and Washing Conditions 55 2 . 1 1 .3 Chemiluminescent Detection of DIG-Labelled Probes 55 2 . 1 2 LIBRARY CONSTRUCTION AND SCREENING 56 2. 1 2. 1 Calf Alkaline Phosphatase (CAP) Treatment of Vector 56 2 . 12 .2 Preparation of Linkers 57 2 . 1 2 .3 Preparation of Linkered Insert DNA 57 2. 1 2.4 Ligation Reactions 58 2 . 1 2.5 Preparation of Electrocompetent XL- l Cells 58 2 . 12 .6 Transformation by Electroporation 58 2 . 1 2.7 Fil ter Lifts and Primary Library Screening 59 2 . 1 2.8 Second Round Library Screening 59 2 . 12 .9 Screening Positive Clones 60 2. 1 3 POLYMERASE CHAIN REACTION AMPLIFICATION 60 2. 1 3 . 1 Primers 6 1 2 . 1 3 .2 Optimisation of PCR Reaction Conditions 64 2. 1 3 .3 Randomly Amplified Polymorphic DNA PCR (RAPD-PCR) 64 2. 1 3 .4 MicrolMinisatell i te-Primed PCR (MP-PCR) 64 2. 1 3 .5 Microsatellite-Locus PCR (ML-PCR) 65 2 . 1 3 .6 Multiplex PCR 65 2. 1 3 .7 In Planta PCR 66 x 2. 1 3 . 8 B-Tubulin Gene Ampli fication 66 2 . 1 3 .9 rDNA-ITS Amplification 66 2 . 14 CLONING OF PCR PRODUCTS 67 2. 14 . 1 pGEM-T Cloning 67 2 . 14.2 Preparation of Inserts for Directional Cloning 68 2 . 14 .3 Directional Cloning Using pUC1 1 8 68 2 . 1 5 DNA SEQUENCING 69 2 . 1 5 . 1 AmpJiCycle Sequencing Kit 69 2 . 1 5 . 2 Dye Terminator Cycle Sequencing 69 2. 1 5 .3 Sequence Analysis 70 2 . 1 6 MICROSATELLITE ANALYSIS 70 2. 1 7 AUTOMATED GENETIC ANALYSIS 7 1 2 . 1 8 PHYLOGENETIC ANALYSIS 72 2 . 1 8 . 1 Sequence Alignment 72 2 . 1 8 .2 Phylogenetic Analysis by Maximum Parsimony 72 2. 1 8 .3 Phylogenetic Analysis by Neighbor-Joining 72 xi CHAPTER THREE - DEVELOPMENT AND APPLICATION OF AN IN PLANT A EPICHLOE ENDOPHYTE FINGERPRINTING ASSAY 74 3 . 1 INTRODUCTION 75 3 .2 RESULTS 76 3 .2 . 1 Pilot Study to Evaluate PCR-Based Fingerprinting Methods 76 3 .2 . 1 . 1 RAPD-PCR 76 3 .2 . 1 .2 MP-PCR 78 3 .2 . 1 .3 ML-PCR 78 3 .2 . 1 .4 Evaluation of Fingerprinting Methods 85 3 .2 .2 Development of a PCR-Based Microsatell i te Fingerprinting Assay 85 3 .2 .2 . 1 Detection of Microsatell i tes within Epichloe Endophyte Genomes 85 3 .2 .2 .2 Library Construction and Screening 87 3 .2 .2 .2 . 1 E8 Library 87 3 .2 .2 .2 .2 Lp 1 Library 87 3 .2 .2 .3 Analysis of Positive Clones 88 3 .2 .2 .4 Identification of Additional Epichloe Endophyte Microsatellite Loci 94 3 .2 .2 .5 ML-PCR Assay Design and Evaluation 94 3 .2 .2.5 . 1 Primer Design and Preliminary Amplifications 94 3 .2 .2 .5 .2 Amplification of Fungal Outgroups 99 3 .2 .2 .5 .3 In Planta Amplification 99 3 .2 .2 .5 .4 Screening Epichloe Endophyte Isolates Through ML-PCR Assays with High Resolution Separation of Alleles 103 3 .2 .2 .5 .5 Summary of Screening Epichloe Endophyte Isolates 1 24 3 .2 .2 .6 Development of Primer Set I 1 26 3 .2.2.6. 1 Redesigning Primers to Microsatellite Locus B9 1 26 3 .2.2 .6 .2 Multiplex PCR of Primer Set I 1 30 3 .2 .2 .7 Development of Primer Set II 1 34 3 .2 .2 .7 . 1 Screening Epichloe Species with ML-PCR Assays 1 34 3 . 2.2 .7 .2 Multiplex PCR of Primer Set II 3 .2 .2 .8 Automated Analysis of Fingerprints 3 .2 .2 .8 . 1 Comparison of Fingerprints Analysed on an ABI Prism 377 DNA Sequencer and an ABI Prism 3 10 Xll 143 147 Genetic Analyser 1 50 3 .2 .3 Application of the PCR-Based Microsatellite Fingerprinting Assays 1 53 3 .3 3 .3 . 1 3 .3 .2 3 .3 .3 3 .4 3 .2 .3 . 1 Microsatellite Fingerprinting of Epichloe Endophyte Isolates 1 53 3 .2 .3 .2 Application of Microsatellite Fingerprinting Assay in a B lind Test 1 53 3 .2 .3 .3 Analysis of a DNA Deletion in B l Amplified Alleles 1 60 3 .2 .3 .4 Inheritance of Microsatellite Alleles in an E. Jestucae Sexual Cross 1 65 DISCUSSION 168 An Assessment of the PCR-Based Fingerprinting Methods 168 Development of a PCR-Based Microsatellite Fingerprinting Assay 170 3 .3 .2 . 1 The Frequency of Microsatell ites in Fungi 1 70 3 .3 .2 .2 An Overview of the ML-PCR Assays 1 7 1 3 .3 .2 .3 Automated Fingerprint Analysis 1 73 Applications of the Microsatellite Fingerprinting Assay for Epichloe Endophytes 1 75 3 .3 .3 . 1 Evolutionary Relationships Between Endophyte Groups are Generally Supported by Microsatellite Genotype Data 1 75 3 .3 .3 . 1 . 1 The Evolutionary Relationships of Asexual Endophytes 1 75 3 .3 .3 . 1 .2 The Evolutionary Relationships of Sexual Endophytes 1 78 3 .3 .3 . 1 .3 The Distribution of a Deletion among B 1 Alleles - Impli cations for Endophyte Evolution 1 80 3 .3 .3 . 1 .4 Phylogenetic Inference using Microsatellite Data 1 8 1 3 .3 .3 .2 Endophyte Identification In Planta - A Trial Application 1 82 3 .3 .3 .3 Microsatel I i te Loci as Markers in Endophyte Genomes 1 83 CONCLUSION 1 84 X 111 CHAPTER FOUR - THE EVOLUTIONARY ORIGINS OF EPICHLOE 4 . 1 4 .2 4.3 4.4 ENDOPHYTES FROM ANNUAL LOLIUM AND HORDEUM GRASSES 1 85 INTRODUCTION 1 86 RESULTS 1 88 4.2. 1 Endophyte Isolation from Plant Tissue 1 88 4.2.2 Fungal and In Planta DNA Extraction 1 9 1 4.2 .3 rDNA-ITS Amplification and Sequencing 1 9 1 4 .2.4 tub2 Amplification and Sequencing 1 92 4.2.5 Phylogenetic Analysis of DNA Sequences 203 4.2.6 Microsatellite Analysis 206 4.2.7 Analysis of Endophyte Isolates from Hordeum Grasses 2 1 2 4.2 .7 . 1 Culture Descriptions 2 1 2 4.2 .7 .2 rDNA-ITS Analysis 2 1 2 4.2 .7 .3 tub2 Gene Analysis 220 4.2.7 .4 Microsatel lite Analysis 223 DISCUSSION 226 4.3 . 1 Two New Endophyte Taxonomic Groups are Identified from Annual Ryegrasses 226 4.3 .2 The Evolutionary Origin of LmTG- l 227 4 .3 .3 The Evolutionary Origin of LcTG- l 229 4.3.4 The Evolution of Lolium Endophytes 230 4 .3 .5 The Evolutionary Origins of Two Endophyte Isolates from Hordeum Grasses 233 4.3 .6 Interspecific Hybridisation in the Epichloe Endophyte Group 234 CONCLUSION 236 APPENDIX 1. Vector Maps APPENDIX 2. Manufacturer's References APPENDIX 3. Microsatel lite Electropherograms APPENDIX 4. Chi-Square (X2) Tests APPENDIX 5. GenBank Accession Numbers APPENDIX 6. DNA Sequence Alignments APPENDIX 7. Publication REFERENCES xiv 238 241 242 243 246 248 249 250 xv LIST OF TABLES Table 1 . 1 . Summary of the main nomenclature changes within the Epichloe endophytes 1 2 Table 1 .2. Relationships of Neotyphodium grass endophytes to Epichloe species 1 7 Table 1 .3 . Key features of tandemly repeated DNA classes 25 Table 2 . 1 . Endophyte and fungal isolates used i n this study 3 1 Table 2 .2 . Primer and probe sequences 62 Table 3 . 1 . Summary of polymorphism and size range of ML-PCR products amplified from Epichloe endophyte isolates with PAGE separation 1 25 Table 3 .2 . Summary of amplification of representative i solates of Epichloe spp. in B 1 , B2, B5 , B7, B8 , and B9 ML-PCR assays 142 Table 3 .3 . Allele sizes for microsatellite loci of Tf28 from electropherogram of Fig. 3 .29 1 49 Table 3 .4. Calculation of the average size differences of ML-PCR primer set I products when automatical ly analysed on an ABI 377 slab gel based system compared to an ABI 3 10 capil lary electrophoresis system 1 5 1 Table 3 .5 . Calculation of the average size differences of ML-PCR Primer Set I I products when automatically analysed on an ABI 377 slab gel based system compared to an ABI 3 10 capillary electrophoresi s system 152 Table 3 .6 . Allele sizes of Epichloe endophyte microsatel lite loci as determined by automatic analysis 1 54 Table 4. 1 . Microsatellite genotypes of annual ryegrass endophyte i solates 2 10 XVI LIST OF FIGURES Fig. l . 1 . The asexual and sexual l ife cycles of Epichloe Jestucae on F estuca rubra 3 Fig. 1 . 2. Classification of the family Clavicipitaceae 9 Fig. 1 .3 . Classification o f the genus Epichloe 1 0 Fig. 3 . 1 . RAPD-PCR profiles from a pilot range of endophyte isolates 77 Fig. 3 .2 . MP-PCR profi les from a pilot range of endophyte i solates 79 Fig. 3 .3 . MP-PCR profiles of endophyte i solates using the A2 primer 8 1 Fig. 3 .4 . MP-PCR profi les of endophyte i solates using the A3 primer 82 Fig. 3 .5 . ML-PCR amplification of a pilot range of endophyte i solates 83 Fig. 3 .6 . Hybridisation of microsatelli te probes to the Lp1 genome 86 Fig. 3 .7 . DNA sequences of the microsatellite loci used in this study 89 Fig. 3 . 8 . ML-PCR amplification with primers to microsatell ite loci from the E8 l ibrary 95 Fig. 3 .9 . ML-PCR amplification with primers to microsatell i te loci from the Lp1 l ibrary and HMG CoA reductase DNA sequence 97 Fig. 3 . 10 . Optimising the template DNA concentration for amplification of Epichloe microsatellite loci from endophyte-infected plant material 100 Fig. 3 . 1 l . Amplification of Epichloe microsatellite loci from endophyte-infected plant material 1 02 Fig. 3 . 1 2. P AGE separation of microsatellite alleles amplified from endophyte i solates with locus B l 104 Fig. 3 . 1 3 . PAGE separation of microsatellite alleles amplified from endophyte i solates with locus B2 106 Fig. 3 . 14 . PAGE separation of micro satellite alleles amplified from endophyte i solates with locus B4 Fig . 3 . 1 5 . PAGE separation of micro satellite alleles amplified from endophyte isolates with locus B5 Fig. 3 . 16 . PAGE separation of microsatellite alleles amplified from endophyte isolates with locus B6 Fig. 3 . 17 . PAGE separation of microsatellite alleles amplified from endophyte i solates with locus B7 Fig. 3 . 1 8 . PAGE separation of microsatellite alleles amplified from endophyte isolates with locus B8 Fig. 3 . 19 . PAGE separation of microsatellite alleles amplified from endophyte isolates with locus B9 Fig. 3 .20. P AGE separation of microsatellite alleles amplified from endophyte i solates with locus B 10 Fig. 3 .2 1 . PAGE separation of microsatellite aIleles amplified from endophyte isolates with locus B 1 1 Fig. 3 .22. Strategy for the dye-labelling of primers to microsatellite loci for Primer Set I Fig. 3 .23 . ML-PCR amplification of locus B9 using a new primer pair, B9. 1 and B9.4 Fig. 3 . 24. PAGE separation of microsatellite alleles amplified from endophyte xvii 1 08 1 10 1 1 2 1 14 1 16 1 17 1 19 1 2 1 1 27 1 29 isolates with locus B9, using primers B9. 1 and B9.4 1 3 1 Fig. 3 .25 . Multiplex amplification of microsatellite loci using Primer Set I 133 Fig. 3 . 26. ML-PCR screening of representative isolates of Epichloe spp. using primer pairs excluded from Primer Set I 135 xviii Fig. 3 .27. Strategy for the dye-labelling of primers to microsatel lite loci for Primer Set II 144 Fig. 3 .28 . Multiplex amplification o f microsatel lite loci using Primer Set I I 146 Fig. 3 .29 . Electropherogram of an endophyte microsatell i te fingerprint 148 Fig. 3 .30. Distribution of a 73 bp B 1 deletion product across Epichloe species 1 6 1 Fig. 3 .3 1 . Sequence alignment of B I -amplified ML-PCR products of Epichloe endophyte isolates 1 63 Fig. 3 .32. Segregation of B 10 and B 1 1 microsatel lite alleles in a sexual E. festucae cross 1 66 Fig. 4. 1 . Cultures of Neotyphodium endophytes isolated from annual ryegrass tissues 1 89 Fig. 4.2. rDNA- ITS amplification products from annual ryegrass endophyte isolates 1 93 Fig. 4.3 . Alignment of rDNA-ITS sequences from annual ryegrass endophyte i solates 1 94 Fig. 4.4. tub2 amplification products from annual ryegrass and Hordeum endophyte i solates 1 97 Fig. 4.5 . Alignment of the tub2 gene sequences from annual ryegrass endophyte i solates 1 99 Fig. 4.6. rDNA-ITS gene phylogeny based on Epichloe and annual ryegrass endophyte sequences 204 Fig. 4.7. tub2 gene phylogeny based on Epichloe and annual ryegrass endophyte sequences 207 Fig. 4.8 . Cultures of Hordeum endophyte isolates, HaB and Hdl 2 1 3 Fig. 4.9. Alignment of rDNA-ITS sequences from Hordeum endophyte i solates 2 1 5 Fig. 4. 10. rDNA-ITS gene phylogeny based on Epichloe and Hordeum endophyte isolate sequences xix 2 1 8 Fig. 4. 1 1 . Alignment of the tub2 gene sequences from Hordeum endophyte isolates 221 Fig. 4. 12 . tub2 gene phylogeny based on Epichloe and Hordeum endophyte i solate sequences Fig. 4. 1 3 . Possible scenario for the distribution o f Epichloe endophytes during the evolution of Lolium spp. 224 23 1 CHAPTER ONE INTRODUCTION 1.1 THE LIFE CYCLES OF EPICHLOE ENDOPHYTES 2 The Epichloe endophytes are a group of filamentous fungi comprised of the sexual Epichloe species and thei r asexual derivatives the Neotyphodium species, which were formerly classified as Acremonium species (Glenn et al. , 1 996; Schardl, 1 996a). This group of fungi , which hereafter may be refen'ed to as 'endophytes', grow systemically in the vegetative tissues of cool season grasses (subfamily Pooideae), where they are confined to the apoplast (intercellular spaces). Endophytes can be identified by the presence of characteristic septate, infrequently branched hyphae, running longitudinally in the plant tissue (Christensen et al. , 199 1 ) . In the apoplast the endophyte gains nutri tion, but it is not known whether it actively promotes leakage of nutrients from the plant cells, or lives on normal apoplastic substi tuents (Schardl, 1 996a), though the effect of infection on the host appears to be asymptomatic in naturally-occurring associations. The density of endophyte mycelium in the plant di splays an annual cycle. In spring, mycelium grows into the leaf sheaths, leaves and inflorescences, reaching a maximum density in late summer and then declining. In the winter it is confined to the plant crown where it is most concentrated in the meri stematic tissues (di Menna and WaIler, 1 986). Dissemination of Epichloe endophytes to new hosts can be achieved both sexually and asexually, resulting in dramatically di fferent effects on the host. Figure 1 . 1 illustrates these alternative life cycles of an E. Jestucae endophyte on Festuca rubra. The sexual cycle begins with growth of the fungus through the flowering tillers to the flag-leaf sheaths and immature inflorescences, where they emerge to form somatic reproductive structures called stromata. The stromata choke and sterili se the inflorescence. Spermatia, in the form of conidia, develop on the stroma and these are transferred to stroma of other infected hosts by symbiotic flies of the genus Botanophila (often referred to the genus Phorbia; Kohlmeyer and Kohlmeyer, 1 974) (Bultman et al. , 1 995; Schardl and Leuchtmann, 1 999). Epichloe species have a heterothallic mating system, and ascospores are formed following successful cross-fertilisation (White and Bultman, 1 987). These are ejected, and initiate new infections which proceed into florets to ultimately give rise to infected embryos and seed (Chung and Schardl, 1 997a). 3 Figure 1 . 1 . The asexual and sexual life cycles of Epichloe Jestucae on F estuca rubra ascospore tr�7;�� to � , " '-(J) pentheclum + g ' . ' rm eo ; ' n -- P , o ,; "' on ®;''J; -. Qj��:1° ,-"'; I ..--- ... \ 0 1O'.e',on o,ova� '\ q & d'����;lng J :1� /1 I grass panicle sexual life cycle mfectlon • ---..... mfectlon of of ovule �egetal!ve meristem & / leaf tissue \ f \ ! panicle t � asexual life cycle \ I I / of floral mens\em I / mycelium \ . _____ /1 ./ floral menstem penthecia on stroma \ ........ w/ / � f I � 1O""',on ':;:'00'" � ka'YogaC[)m � ) .;�' o � • �l o'ehoke . 0 , . ' \ ma�lng cOflldloma C> In the asexual cycle, highly convoluted hyphae grow intercelIularly in leaf sheaths, floral meri stems, and in the ovules of the florets such that the fungus is transmitted in the next generation of seed. In the sexual cycle, the fungus also grows asymptomatical ly and intercellularly within the host, but then emerges from the leaf sheath surrounding the immature host inflorescence, produces spermatia, and arrests inflorescence maturation. Fertil isation occurs by transfer of spermatia of opposite mating type. If the parents are conspecific (same mating population), pelithecia containing asci develop and filamentous ascospores are ejected. Germinating ascospores initiate cycles of asexual sporulation (conidiation) and are postulated to cause infection of host florets and ultimately of seed. Diagram adapted from Scott and Schardl ( 1993). 4 This horizontal mode of endophyte transmission is pathogenic to the host and is the causative process behind ryegrass choke by E. typhina (reviewed by Schardl, 1 996b). As i l lustrated in Figure 1 . 1 , E.festucae is also able to undergo clonal (vertical) propagation within the host, which is the only mode of transmission for the asexual Neotyphodium endophytes . This is accomplished by vegetative growth of the endophyte through the floral ti l lers into the host ovules. The endophyte is incorporated into the embryo and aleurone of the seed, where it is disseminated (Philipson and Christey, 1 986). This mode of transmission is highly efficient, where nearly 1 00% of seed from infected mother plants transmit the endophyte (Siegel et aI. , 1 984). Epichloe species can undergo both horizontal and vertical transmission simultaneously, though the degree to which each species does this varies greatly and i s discussed further in Section 1 .3. 1.2 ANTIHERBIVORE ALKALOIDS Endophyte-infected grasses are known to contain a number of biologically active alkaloids, which are responsible for many of the effects of endophyte presence on pasture growth and animal production. The main alkaloids responsible for these effects are classified into four chemical classes - ergopeptine alkaloids, lolines, lolitrems, and perarnine. All of these have activities against insects and most, or all, are toxic to mammals at some level (reviewed by Scott and Schardl, 1 993). In 1977, it was demonstrated that the toxicity syndromes of tall fescue grasses (Festuca arundinacea) were directly related to the endoPfyte cont�nt of these pastures (Bacon et aI. , 1 977). It was later shown that ergopephne alkaloids, such as ergovaline and ergotamine, are synthesised by the fun.gus and occur in many endophyte-plant interactions (Bacon et aI., 1 986; Rowan and Shaw, 1987). These compounds cause the classical symptoms of fescue toxicosis - elevated temperatures, reduced feeding, reduced fertility, reduced lactation, to fat necrosis, vasoconstriction, sti l lbirth, gangrene and f death, depending on dosage and environmental stresses (Raisbeck et aI. , 1 99 1 ). 5 Similarly, the nervous disorder, ryegrass staggers, was found to be associated with the incidence of endophyte infected perennial ryegrass (Lolium perenne) (Fletcher and Harvey, 1 98 1 ). It was later shown that lipophilic complex substituted-indole compounds, named lolitrems, produced by the endophyte-grass association produced these tremors in livestock (Gallagher et aI. , 1 98 1) . Lolitrem B is the main toxin implicated in ryegrass staggers (Gal lagher et aI. , 1 984; Raisbeck et aI. , 1 99 1 ). Both ryegrass staggers and fescue toxicosis are responsible for large economic losses to the sheep industry in New Zealand and the beef and cattle industry in the United States. Consequently, there is considerable interest in maximising the beneficial aspects of this symbiotic association for pastoral agriculture. Lolines, or saturated arninopyrrolizidines, are toxic to insects and may also play a direct role in fescue toxicosis of livestock. These compounds are the most abundant alkaloids in the associations between Neotyphodium coenophialum and tall fescue (Festuca arundinacea), with concentrations up to one thousand times greater than that of the other alkaloids (Dahlman et aI . , 1 99 1 ). The production of lolines was only recently found to be of endophyte origin (pers. comm. Heather Wilkinson, University of Kentucky). Peramine, a pyrrolopyrazine, is a major feeding deterrent to insects and has been shown to play a key role in protection of perennial ryegrass from Argentine stem weevil (Listronotis bonariensis), a major introduced pest in New Zealand (Prestidge et aI . , 1 985; Rowan and Gaynor, 1 986). The distribution of thes'e alkaloid compounds within the host does not necessarily correlate with the location of endophyte infection. Perarnine, ergo valine and the loline alkaloids (or a signal for their synthesis) appear to be translocated from the leaf sheath into the leaf blade where there is little fungal mycelium (Ball et aI . , 1 995; Bush et aI . , 1 982; Rowan and Shaw, 1 987). However, lolitrem B tends to remain concentrated at the leaf sheaf near the base of the plant (Ball et aI . , 1 995). It is obvious that the effects of these alkaloids are highly beneficial to the endophyte­ grass symbiosis, but are detrimental to livestock in agricultural grazing systems. 6 Therefore, 'elite' endophytes are highly sought after, which retain the benefits of the symbiosis without the detrimental effects to grazing livestock. The two main strategies of obtaining these 'eli te' endophytes is by selection of existing endophyte strains, or by genetic manipulation. It has been shown that natural strains of Neotyphodium differ in the spectrum of alkaloids they produce, and that expression is also dependent on the genotype of the host (Christensen et aI. , 1 993) . Selection of endophytes which display desirable combinations of alkaloids may be used to artificially inoculate grasses and replace existing inferior grass-endophyte associations in pasture. Alternatively 'elite' endophytes may be obtained by the genetic modification of existing ones. Currently, an area of intense research is in the identification of genes involved in alkaloid biosynthesis, and notable progress has been made for lolitrem B, where a cluster of genes involved in the biosynthesis of a related indole diterpenoid, paxil line, have been cloned from Penicillium paxilli (Young et aI . , 1 998; Young et aI. , 1 999). Genes involved in ergot alkaloid biosynthesis have also been cloned from related Clavicipitaceous fungi (Wang et aI . , 1 999), and progress has been made in identifying loline biosynthetic genes (pers. comm. Heather Wilkinson, University of Kentucky). Once these genes are isolated and characterised, modified endophytes can be engineered, disrupting synthesis of the detrimental alkaloids. Methods for the stable genetic transformation of Neotyphodium endophytes have already been successfully demonstrated, which include electroporation and polyethylene glycol (PEG) treatments (Murray et aI . , 1 992; Tsai et aI . , 1 992). Transformed endophytes can then be used to inoculate grass plants to form stable associations. 1.3 GRASSIENDOPHYTE SYMBIOSES The relationships between Epichloe endophytes and their hosts display extraordinary diversity, with symbioses between the two entities spanning the continuum from mutualism to antagonism, depending on the mode of dispersal of the endophyte (reviewed by Schardl, 1 996a). At one extreme, the mutualistic asexual endophytes exist as asymptomatic fitness enhancing symbionts with benefits to the host involving 7 resistance to herbivores (Bacon et aI., 1 977; Gallagher et aI., 1 984; Kimmons et aI., 1 990; Prestidge et aI., 1 985; Siegel et aI., 1 990), and drought tolerance (Arachevaleta et aI., 1 989), resulting in enhanced persistence and fecundity (Latch et aI., 1 985; Siegel et aI., 1 985). Benefits to the endophyte include provision of nutrients by the hosts, and dispersal by seedbome dissemination (Siegel et aI. , 1984). The most intensely documented of these symbioses are those of the agriculturally important pasture grasses - tal l fescue with Neotyphodium coenophialum, and perennial ryegrass with N. lolii, though Epichloe species may also form mutualistic associations. Strains of E. bromicola symbiotic with Bromus ramosus and B. benekenii never form stromata on their hosts, but their interferti l ity with E. bromicola strains that infect B. erectus have enabled these strains to be classified as Epichloe species (Leuchtmann and Schardl, 1 998). The survival of mutualistic endophytes is dependent on selection at the level of the symbiotic association, rather than the endophyte alone. At the other end of the spectrum, extreme cases of antagonism are displayed by obligately sexual endophytes. The relationship between E. typhina and perennial ryegrass i l lustrates this, where the sexual cycle results in choking of all inflorescences, rendering the host sterile. In a sense, these relationships are not strictly antagonistic as important mutualistic characteristics are also expressed before the sexual cycle of the endophyte is undertaken, such as antiherbivore activities and increased drought tolerance. Several Epichloe endophytes exhibit pleiotropic symbioses with their hosts, whereby both asymptomatic vertical and pathogenic horizontal transmission of the endophyte are undertaken simultaneously on separate flowering tillers of an infected plant. Pleiotropism is ill ustrated in Figure 1 . 1 by E. Jestucae infected F. rubra. The degree of mutualism and antagonism displayed varies, and depends not only on the genotypes of the endophytes and hosts involved, but also temporal, developmental and environmental factors. Examples of more extreme pleiotropic symbioses include that of E. typhina infected Poa nemoralis, which is highly antagonistic but occasionally produces rare seedbome infections, and E. brachyelytri infected Brachyelytrum erectum which very rarely expresses stromata and almost always develops normal inflorescences giving rise 8 to infected seeds (Schardl and Leuchtmann, 1 999) . The continuum from mutualism to antagonism displayed by the Epichloe endophytes make them an ideal system for studying the evolution of fungal mutualists (Tsai et al . , 1 994). 1.4 TAXONOMY AND EVOLUTION OF EPICHLOE ENDOPHYTES 1 .4.1 EPICHLOE ENDOPHYTE TAXONOMY The taxonomy of the Epichloe endophytes is an area which is complicated by taxonomic convention and changes in taxonomy over time. Sexual Epichloe species and their asexual derivatives, the Neotyphodium species, constitute a likely monophyletic group (Kuldau et al . , 1 997; White and Reddy, 1 998) but traditional taxonomic conventions do not classify these groups together. Instead, they have been separated at the level of fungal subdivision (Ascomycotina for Epichloe and Deuteromycotina for Neotyphodium) on the basis of the abi lity to undergo a sexual l ife cycle (Ainsworth et al . , 1 973). Molecular approaches to fungal taxonomy have brought about advocation for the abandonment of this type of dual naming system, in favour of a more natural one which reflects the evolutionary relationships of the organisms. Such a system has been adopted by Kendrick ( 1 992), which refers to fungal holomorphs that include both teleomorphic and anamorphic fungi . The cunent classification of the genus Epichloe from the level of Kingdom, according to Kendrick ( 1 992), and Olenn et al . ( 1996) i s given in Figures 1 .2 and 1 .3 respectively . Over time the Epichloe endophytes have undergone several taxonomic , and hence nomenclature changes as evolutionary relationships are being continually elucidated, revised, and updated. The most recent example of a major nomenclature change is from a molecular study of the form genus Acremonium (Olenn et al . , 1 996). Here, the Figure 1 .2 . Classification of the family Clavicipitaceae (Kendrick, 1 992) KINGDOM Animalia Plantae Eumycota -[ Protoctista Monera DIVISION Dlkaryomycota I Zygomycota SUBDIVISION CLASS Ascomycotlna L Ascomycetes Basidiomycotina Saccharomycetes * 15 of the 44 orders of Ascomycetes are shown here (Kendrick, 1992) ORDER FAMILY Taphrinales Dothideales Laboulbeniales Ophiostomatales Eurotiales Onygenales Clavicipitales ---- Clavicipitaceae Pezizales Erysiphales Leotiales Rhytismatales Diatrypales Sphaeriales Sordariales Elaphomycetales Figure 1 .3. Classification of the genus Epichloe (Glenn et al . , 1 996) FAMILY SUBFAMILY TRIBE GENUS Atkinsonella Oomycetoideae Clavicipiteae Balansia Clavicipitaceae --t-- Cordycipitoideae Balansieae Balansiopsis Clavicipitoideae Ustilagnoideae Epichloe Myriogenospora ...... o 1 1 Epichloe anamorphs were c lassified alongside various saprobic fungal species, but placed in section Albo-Ianosa (Morgan-Jones and Gams, 1982). Molecular data strongly supported the proposal of reclassification of the grass endophytes into the new fonn genus Neotyphodium, which enables the unique biological , morphological, and ecological characteristics of this group to be distinguished (Glenn et aI . , 1 996). Until endophyte species were identified and described, the epithet E. typhina had been used to broadly describe all stroma-forming endophytes of pooid grasses (Groppe et aI . , 1 995; Leuchtmann et aI . , 1994; White, 1 993), and even to describe the asexual endophyte of tall fescue, which is now known as N. coenophialum (Bacon et aI . , 1 977 ; Glenn et al. , 1 996). Thus, changes in taxonomy and nomenclature can be confusing, and one must be cautious and consider the date of publication of a piece of work and the state of the taxonomy of the organism at that time to determine which organism is actuall y being referred to. A summary of the different names used by the Epichloe endophytes is given in Table 1 . 1 . 1 .4.2 EPICHLOE SPECIES To date, ten specIes of Epichloe have been desclibed, primarily on the basis of interfertil ity groups and ascospore morphologies (White, 1 993; White, 1 994), but more recently with isozyme phenotype data and B-tubulin (tub2) gene DNA sequence data (Leuchtmann and Schardl , 1998; Schardl and Leuchtmann, 1 999). These include the native European species (mating populations, MP, in brackets): E. typhina sensu stricto (MP-I), E. clarkii White (MP-I), E. festucae Leuchtmann et al. (MP-H), E. baconii White (MP-V), E. bromicola Leuchtmann & Schardl (¥P-VI), and E. sylvatica Leuchtmann & Schardl (MP-VII); and native North American species: E. amarillans White (MP-IV), E. elymi Schardl & Leuchtmann (MP-Ill), E. glyceriae Schardl & Leuchtmann (MP-VIII), and E. brachyelytri Schardl & Leuchtmann (MP-IX) (Leuchtmann and Schardl, 1 998 ; Leuchtmann et aI . , 1 994; Schardl and Leuchtmann, 1 999; White, 1 993; White , 1 994) . Each species constitutes a single mating population, except the morphospecies E. typhina and E. clarkii, which tend to be interfertile in mating tests. Table 1 .1 . Summary of the main nomenclature changes within the Epichloe endophytes Name Reference Previous names Reference Epichloe typhina Epichloe amarillans White, 1994 Epichloe sp. MP IV Schardl, 1996 Epichloe typhina Epichloe baconii White, 1993 Epicldoe sp. MPV Schardl, 1996 Epichloe brachyelytri Schardl and Leuchtmann, Epichloe typhina 1999 Epichloe sp. MPIX Schardl, 1996 Epichloe typhina Epichloe bromicola Leuchtmann and Schardl, 1998 Epichloe sp. MPVI Schardl, 1996 Epichloe clarkii White, 1993 Epichloe typhina Epichloe elymi Schardl and Leuchtmann, Epichloe typhina Schardl, 1996 1999 Epichloe sp. MPIII Epichloe festucae Leuchtmann et aI., 1994 Epichloe typhina Schardl, 1996 Epichloe sp. MPII Reclassification basis mating tests size and density of perithecia size of asci and ascospores mating tests disarticulation of ascospores to form I-septum part spores mating tests tub2 data mating tests isozyme data tub2 data mating tests disarticulation of ascospores to form 3-6 septate part spores mating tests tub2 data mating tests morphology of fruiting structures size and articulation of ejected ..­N Table 1.1 . continued Name Reference Previous names Reference Reclassification basis Epichloe glyceriae Schardl and Leuchtmann, Epichloe typhina mating tests 1999 Epichloe sp. MPVIII Schardl, 1996 tub2 data Epichloe typhina mating tests Epichloe sylvatica Leuchtmann and Schardl, isozyme data 1998 Epichloe sp. MPVII Schardl, 1996 tub2 data Sphaeria typhina Persoon, 1798 Dothidia typhina 1823 Stromatosphaeria 1826 Epichloe typhina (Pers.:Fr) Tul. White, 1993 typhina Hypocrea typhina 1860 Epichloe typhina, MP I Schardl, 1996 Acremonium Morgan-Jones and Gams, 1982 Neotyphodium coenophialum Glenn et aI., 1996 coenophialum rDNA sequence data FaTG-1 Christensen et aI., 1993 Neotyphodium typhinum Glenn et aI., 1996 Acremonium typhinum Morgan-Jones and Gams, 1982 rDNA sequence data Acremonium loliae Latch et aI., 1984 Neotyphodium loW Glenn et aI., 1996 LpTG-l Christensen et aI., 1993 rDNA sequence data Acremoniuf1l lolii Neotyphodiuf1l chisosum Glenn et aI., 1996 Acremonium chisosum White and 1987a rDNA data Table 1 .1 . continued Name Reference Previous names Reference Reclassification basis Neotyphodium starrii Glenn et a!., 1996 Acremonium starrii White and Morgan-Jones, rDNA sequence data 1987b Neotyplzodium huerJanum Glenn et ai., 1996 Acremonium huerJanum White et a!., 1987 rDNA sequence data Acremonium uncinatum Gams et a!., 1990 Neotyphodium uncinatum Glenn et ai., 1996 rDNA sequence data FpTG-1 Christensen et a!., 1993 Neotyplzodium chilense Glenn et aI., 1996 Acremonium chilense Morgan-Jones et aI., 1990 rDNA sequence data isozyme data, alkaloid LpTG-2 Christensen et aI., 1993 Acremonium lolii Latch et ai., 1984 production, morphological characters, and benomyl resistance FaTG-2 Christensen et ai., 1993 A cremonium Morgan-Jones and Gams, as above coenophialum 1982 FaTG-3 Christensen et a!., 1993 Acremonium Morgan-Jones and Gams, as above coenophialum 1982 15 Epichloe clarkii differs from E. typhina only i n i ts host specificity for Holcus lanatus, and disarticulation of the ejected ascospores (White, 1 993). All mating populations, except MP-I, tend to have restricted host ranges within the Pooideae, with E. festucae strains symbiotic with Festuca spp. (tribe Poeae), E. elymi with Elymus spp. (Triticeae), E. amarillans with North American spp. of Sphenopholis and probably Agrostis (Aveneae), E. baconii with Eurasian Agrostis and Calamagrostis spp. (Aveneae), E. bromicola with Bromus spp. (Bromeae), E. sylvatica with Brachypodium sylvaticum (Brachypodieae), E. glyceriae with Glyceria striata (Meliceae), and E. brachyelytri with Brachyelytrum erectum (Brachyelytreae) (Schardl and Leuchtmann, 1 999). Mating population one has a rather broad host range which includes grasses of the tribes Poeae, A veneae and Brachypodieae (Schardl et al . , 1 997). Host specificity is a likely contributing factor for the speciation of Epichloe, but host flowering times and the symbiotic fly are also l ikely to have played key rol es (Schardl and Leuchtmann, 1 999) . 1 .4.3 NEOTYPHODIUM SPECIES The classification of asexual Neotyphodium species is not possible using a biological species concept, and depends on other criteria. Christensen et al . ( 1993) used a combination of cultural morphology, alkaloid production, grass host, and isozyme phenotypes to classify the Neotyphodium endophytes from perennial ryegrass, tall fescue, and meadow fescue. These endophytes were shown to have significant diversity and six taxonomic groups of endophytes were identified. These were later shown to conespond well with phylogenies delived from tub2 and nuclear ribosomal DNA internal transcribed spacer region (rDNA-ITS) sequence data (Schardl et al . , 1 994; Tsai et al. , 1 994). Meadow fescue contained one taxon, Festuca pratensis taxonomic grouping one (FpTG- l ) which met the original description of N. uncinatum. Perennial ryegrass contained two taxa, Lolium perenne taxonomic grouping one (LpTG- l ), which met the original description of N. lolii, and LpTG-2. Tall fescue had three taxa associated with it: 1 6 Festuca arundinacea taxonomic grouping one (FaTG- l ) , FaTG-2, and FaTG-3 . FaTG- 1 met the definition of N. coenophiaium, and so, was named this . As a result of Neotyphodium endophytes never showing any symptoms or signs of a fungal infection on grasses, they are difficult to detect, but several species have been formall y described, including: N. chisosum which infects 5tipa eminens (White and Morgan-lones, 1987a), N. starrii which infects Bromus anomalus and Festuca spp. (White and Morgan-lones, 1987b), N. huerfanum which infects F. arizonica (White et aI . , 1 987), and N. chilense which infects Dactylis glomerata pastures in Chile (Morgan­ lones et aI . , 1 990). The placement of N. chilense within the genus Neotyphodium is tenuous though, and requires vetification from genetic analysis (pers. comm. Mike Christensen, Grasslands AgResearch). Additional to the Neotyphodium spp. l isted above, many grass endophytes have been detected that are also likely to be Neotyphodium. These include endophytes of the hosts: Echinopogon ovafus, a native of Australasia (Miles et aI . , 1 998), annual Lolium ryegrasses (Latch et aI . , 1 988) , Poa amp la (Schardl, 1 996a) , Hordeum (pers. comm. Mike Christensen, Grasslands AgResearch). The apparent widespread distribution of Neotyphodium endophytes suggests that those listed above represent onl y a small fraction of this genus, and more species can be found with further systematic sampling of grasses. 1 .4.4 INTERSPECIFIC HYBRIDISATION EVENTS There appears to be at least two different pathways by which asexual Neotyphodium endophytes may evolve from their Epichloe predecessors (Schardl et aI . , 1 994) . The simple scenario whereby a pleiotropic Epichloe strain experiences a mutation that eliminates stroma expression, hence loss of sexual cycle, is postulated for N. loW from E.festucae (Leuchtmann, 1 994; Schardl, 1 996a; Schardl et aI . , 1 994) . More commonly, asexual endophytes appear to have evolved by a interspecific somatic hybridisation mechanism (Schardl , 1 996a). Interspecific hybrid endophytes have been identified by Tsai et al . ( 1 994), where a phylogenetic study of the genetic diversity of tall fescue endophytes, revealed that whereas each Epichloe and meadow fescue isolate studied had 1 7 onl y a single tub2 gene, most tal l fescue endophytes had two to three di stinct tub2 copies. From sequence data of the non-coding region of the gene, i t was proposed that the presence of the multiple copies of the gene is the consequence of multiple interspecific hybridisation events with Epichloe species (Tsai et al. , 1 994). Simi larly tub2, rDNA-ITS , pyr4 gene (which encodes orotidine-5 ' -monophosphate decarboxylase) restriction fragment length polymorphism, mitochondrial DNA, and isozyme analyses revealed that LpTG-2 endophytes of perennial ryegrass are the result of an interspecific hybridisation between E. typhina and an N. lolii mutualist (Schardl et aI . , 1 994). Further findings suggest that N. uncinatum is also likely to be a hybrid endophyte of E. typhina and E. bromicola Oligins (pers. comm. Christopher Schardl , University of Kentucky). It therefore appears that interspecific hybridisation is a relatively common event in the origin of asexual endophytes. Table 1 .2 shows a summary of the most l ikely ancestors of selected asexual Neotyphodium endophytes. Table 1 .2. Relationships of Neotyphodium grass endophytes to Epichloif species Grass host F estuca arundinacea F. arundinacea F. arundinacea F. pratensis Lolium perenne L. perenne Poa ampla Anamorphic specIes N. coenophialum FaTG-2 FaTG-3 N. uncinatum N. lolii LpTG-2 undescribed table adapted from Schardl ( l996a) Likely ancestors or c losest extant relatives N. uncinatum, E. baconii, E. Jestucae E. baconii, E. Jestucae E. baconii, E. typhina E. bromicola, E. typhina E. Jestucae N. lolii, E. typhina N. uncinatum, E. amarillans 1 8 Interspecific hyblidisation i s considered by some to be a mechanism to counteract Mull er's ratchet, whereby in asexual species the accumulation of deleterious mutations wil l cause loss of fitness without the conective influences of sexual recombination (Muller, 1 964). The favoured hypothesis for the mechanism of interspecific hybridisation involves the infection of a mutualistic endophyte inhabited host by spermatia or ascospores of an Epichloe species via the stigma. Fol lowing duel infection of the host, it is assumed that anastomosis (hyphal fusion) takes place, with subsequent fusion of nuclei (Tsai et aI . , 1 994) . Thereafter, redundant chromosomes or chromosomal segments may be lost without detriment to the endophyte resulting in heteroploidy or aneuploidy of the hybrid (Schardl et aI . , 1 994) . The formation of interspecific hybrid endophytes is assumed to be relatively rare, and has not been observed under experimental conditions, though some of the processes implicated in hybrid formation have been demonstrated. Infection of single host plants by mUltiple endophyte isolates has been studied, both naturally-occurring and artificially induced symbiota, and it was observed that i solates tend to segregate at the tj] ]er level (Meijer and Leuchtmann, 1 999; pers. comm. Mike Christensen, Grasslands AgResearch). Anastomosis of pairs of Epichloe spp. has also been successful ly demonstrated by complementation studies of nitrate non-uti l ising (nit) mutants, resulting in the production of heterokaryons (Chung and Schardl , 1 997b). 1.5 THE GRASS SUBFAMILY POOIDEAE The grass family Poaceae is the fourth largest of the flowering plant families, and includes all the major cereals such as wheat, maize, rice, barley and oats, as well sugar cane, sorghum, rye, bamboo, and mil let (Kel logg, 1998) . The Epichloe endophytes infect grasses of the Poaceae subfamily Pooideae, so a brief description of the organisation of this group is given, with particular detail to the grasses that host the endophytes species in this study. 1 9 Although the most recent classifications of Pooideae, based on morphological characters, describe seven to ten tribes, phylogenetic analyses using ndhF gene sequences suggest that the boundaries of this subfamil y contain twelve tribes: Poeae, A veneae, Tliticeae, Bromeae, Brachypodieae, Mel iceae, Stipeae, Lygeae, Nardeae, Diarrheneae, Brachyelytreae, and Phaenospermatae (Catahln et aI . , 1 997) . The Epichloe spp. that infect these tribes are described in Section 1 .4.2. A principal core clade within Pooideae includes the c losely related tribes: Poeae, Aveneae, Triticeae, and Bromeae. These four tribes have been referred to as the "core pooids" , and their monophyly and relationship within the Pooideae has been well supported with independent chloroplast DNA (cpDNA) restriction site data and rDNA-ITS sequences (Davis and Soreng, 1 993 ; Hsaio et aI . , 1 995). Poeae and A veneae form a sound monophyletic c lade as does Triticeae and Bromeae, and these two clades are sisters. Within the Poeael A veneae clade, A veneae appears to have given rise to Poeae (A veneae is paraphyletic to Poeae), whil st within the Triticeae/Bromeae c lade, each tribe is monophyletic and they are sister groups (Catah'in et aI . , 1 997). The important forage grasses in this study Festuca arundinacea (tall fescue), F. pratensis (meadow fescue) and Lolium perenne (perennial ryegrass), are members of the Poeae tribe. The genus Festuca contains over 400 species that are separated into two subgenera: subg. Schedonorus (broad-leaved fescues, including tall and meadow fescues) and subg. Festuca (fine-leaved fescues) (Charmet et aI . , 1 997). In contrast, Lolium contains only eight species (Charmet and Balfourier, 1 994) and phylogenetic analyses show that they appear to have recently evolved from Festuca subg. Schedonorus. Festuca pratensis appears to be the Festuca spp. most closely related to this genus (Charmet et aI . , 1 997). Elucidating the evolutionary relationships within the Festuca­ Lolium complex can be difficult as many species are pol yploid and have genomes that range from diploid, to decaploid states, and species divisions are l argely based on morphological characters and may not reflect infertil ity between groups. The diploid genus Lolium exemplifies a group of species that include several freely interfertile 'species', L. perenne, L. multiflorum, and L. rigidum. These species are all outbreeding, and are even able to hybridise with F. pratensis to form steri le hyblids (Borri l l , 1 976). 20 The remaining Lolium spp. , L. canariense, L. persicum, L. remotum, L. subulatum, and L. temulentum, are al l inbreeding biological species, and together with L. multiflorum and L. rigidum, share annual or short-lived life cycles in contrast to the perennial nature of L. perenne (Borril l , 1 976). The evolutionary relationships within the Pooideae are reasonably resolute at higher taxonomic levels , but the relationships between species are not full y resolved, and probably require data from additional phylogenetical ly informative loci . 1.6 EPICHLOE ENDOPHYTE GENOMES Our current understanding of the basic organisation of Epichloe genomes is rather sparse, and hampered by the notorious lack of reliable methods to analyse fungal chromosomes. Classic cytological methods do not resolve the chromosomes very well due to their small size, the asynchronous movement of chromatids during nuclear division , and the lack of techniques for the preparation of chromosome spreads (Taga and Murata, 1 994; Xu et al . , 1 995). Thus, contour-clamped homogeneous electric field (CHEF) gel-electrophoresis has commonly been used to analyse chromosomes from sexual and asexual-hybrid Epichloe endophytes (Murray et al . , 1 992). The conditions for CHEF separation can be difficult to optimise, and need to be sought for each species and taxonomic group of endophyte depending on the complexity of the genome and size range of i ts chromosomes (pers. comm. Austen Ganley, Massey University). Epichloe endophytes have also been shown to contain mitochondrial l inear p lasmids that exhibit mainly maternal transmission patterns (Chung et al . , 1 996), though the function and replicating mechanisms of these plasmids remain unknown (Schardl , 1 996b). Several genes have been characterised from the Epichloe endophytes, which have chiefly been targeted for phylogenetic purposes, or their role in the production of alkaloids. The B-tubulin gene (tub2), which encodes B-tubulin (a structural protein found in all eukaryotes which i s a main component of the microtubule apparatus), has been i solated and completely sequenced from E. typhina (Byrd et al . , 1 990). The non-coding regions 2 1 of this gene have proved highly informative i n phylogenetic studies (Schardl et aI . , 1 994; Tsai et aI . , 1 994). Another region of the genome which has been extensively used for Epichloe phylogenetic studies is the ribosomal RNA gene cluster (Glenn et aI . , 1 996; Schardl et aI . , 1 99 1 ; Schardl et aI . , 1 994; Schardl and Tsai , 1 992; Tsai et aI . , 1 994; White and Reddy, 1 998), and sets of 'universal' primers, based on sequences from various fungal species, have been designed for thi s region (White et aI . , 1 990). The internal transcribed spacer region (rDNA-ITS) is a popular target for analysis from the population to the genera levels , due to its small size and relatively high rate of sequence change (Bruns et aI . , 1 99 1 ) . The rRNA genes themselves can also be used for such analyses (Glenn et aI . , 1 996), but are more highly conserved and therefore less suited for resolving taxa that are very closely related. Interestingly, only a single rDNA-ITS sequence has been detected in all hybrid endophytes examined to date (Schardl et aI . , 1 994; Tsai et aI. , 1 994), though it has been shown that hybrid endophyte genomes contain several rDNA loci (pers. comm. Austen Ganley, Massey University). Concerted evolution, resulting from homogenisation of the rDNA repeat arrays, is l ikely to explain the absence of all parental rDNA-ITS sequences in the hybrid (Ganley and Scott, 1 998). Phylogenies are currently being established using other informative genes that code for transcription elongation factor alpha (TEF) , and actin, and these will further contribute to our understanding of how the endophytes have evolved (pers. comm. Christopher Schardl , University of Kentucky). Other genes that have been characterised from the Epichloe endophytes include pyr4, which encodes orotidine-5 ' -monophosphate decarboxylase from an LpTG-2 isolate (Collett et aI . , 1 995); Hmg from N. lolii, which encodes 3-hydroxy-3-methylglutaryl coenzyme A (HMG CoA) reductase (Dobson, 1 997); and a N. typhinum proteinase, At 1 (Reddy et aI. , 1 996). The pyr4 locus has been widely used as a selectable marker for gene manipulation in fungi , and was isolated to develop gene transfer and replacement systems in the Neotyphodium endophytes (Collett et aI . , 1 995) . HMG CoA reductase mediates the rate-determining step of cholesterol biosynthesis, and the product of this reaction, mevalonate, is a key intermediate in the isoprenoid pathway. The At1 gene was isolated as part of a characterisation of a highly abundant serine proteinase found in 22 N. typhinumlPoa amp la interactions. In culture this proteinase was found to be expressed by nutrient depletion (Lindstrom and Belanger, 1 994), and is homologous to proteases suspected to be virulence factors in fungal pathogens of insects, nematodes and other fungi (Reddy et aI. , 1 996) . The characteri sation of Epichloe genes is stil l very much in i ts infancy. From the focus of current research it is expected that many genes associated with the biosynthesis of lolitrems, ergot alkaloids, and lolines, wil l be found and characterised in the near future. 1.7 DNA FINGERPRINTING 1.7.1 FUNGAL IDENTIFICATION The identification of fungal types, and the analysis of the relationships between or within different species, populations, and strains of fungi is a central task of many mycologists. Traditional taxonomic identification of fungi is based on micro- and macromorphological characteristics, such as cultural morphologies including colony and colour characteristics on specific culture media, the size, shape, and development of sexual and asexual spores and spore-forming structures. The life history and physiological characteristics, such as nitrogen and carbon source utilisation, are also analysed. These techniques have been increasingly complemented by more modem molecular methods, which include the analysis of chemical constituents (eg. secondary compounds), and characterisation of protein and DNA macromolecules. For some time, isozymes were the molecular markers of choice for studies of genetic diversity, but more recently attention has focused on DNA as a rich source of informative polymorphisms. As each individual 's DNA sequence is unique, sequence information can be exploited for studies of genetic diversity and evolutionary relatedness between organisms. 23 1.7.2 DNA FINGERPRINTING STRATEGIES In the past two decades, a wide variety of techniques have been developed to analyse DNA sequence polymorphisms. These include the direct sequencing of DNA (Sanger et al . , 1 977), restriction fragment length polymorphism (RFLP) , c lassical 'hybridisation' ­ based fingerprinting (Jeffreys et al . , 1985), and polymerase chain reaction (PCR; Saiki et al . , 1 985) based fingerprinting strategies (reviewed by Weising et al . , 1 995b). The advent of PCR in 1985 (Saiki et al . , 1985) has opened up a new suite of rapid fingerprinting methodologies which require very little DNA. The polymerase chain reaction itself involves the in-vitro amplification of particular DNA sequences to high copy numbers using oligonucleotide primers (complementary to the ends of the sequence of interest), and thermostable DNA polymerases, in cycles of thermal denaturation, anneal ing and extension. The chosen plimers direct the specific amplification of the target product, and for DNA fingerprinting strategies, these may be arbitrary, semi specific, or specific. In 1990, several laboratories simultaneously introduced an amplification based fingerprinting strategy that made use of one or two short GC-rich primers of arbitrary sequence to generate PCR products from genomic DNA. This technique, which doesn't require any previous sequence information, was called randomly amplified polymorphic DNA (RAPD) analysis (Wil l iams et al . , 1990), arbitrari ly primed polymerase chain reaction (AP-PCR; Welsh and McClelland, 1 990), or DNA amplification fingerprinting (DAF; Caetano-Anolles et al . , 199 1 ). In general, the term RAPD has been adopted to describe the common characteristics of all these techniques. RAPD-PCR generates fingerprint-like patterns of variable complexity, and has been used by many research groups for identification and even phylogenetic purposes (Crowhurst et a l . , 1 99 1 ; Jungehlilsing and Tudzynski , 1 997; Liu et al . , 1995 ; Stammers et al . , 1995). The widespread popularity of RAPD-PCR may be attributed to its rapidity and ease of use, though it has been criticised for poor reproducibi lity, and the homology of amplified bands of the same size can be questionable (ChaImet et al . , 1 997). More recently, the amplified fragment length polymorphism (AFLP) system was devised, which combines 24 features of RFLP and peR technologies by generating fingerprint patterns from the selected amplification of subsets of genomic restriction fragments (Lin and Kuo, 1 995) . This method boasts high reproducibility, no need for prior sequence information, and identification of large numbers of polymorphisms; though high quality DNA preparations is required, and the procedure is relatively s lower than direct peR methods. The popularity of AFLP markers for genetic mapping and identification purposes i s steadil y increasing (Gibson e t al . , 1998; Zhu e t al . , 1 999), and they have been successful ly applied to Epichloe endophytes for l inkage analysis and strain identification purposes (pers. comm. Heather Wilkinson, University of Kentucky, and Andrew Griffiths, Grasslands AgResearch) . While AFLP technology was shown to be superior to other fingerprinting methods in bacterial strain identification (Koeleman et al . , 1 998), the routine identification of large numbers of samples could be impractical as AFLP is relatively labour intensive. Semispecific and specific-primer peR-based fingerprinting techniques exploit the presence and polymorphic nature of repetitive DNA elements in the genome. Repetitive DNA is ubiquitously found in all eukaryotic genomes, and may be c lassified as either interspersed or tandemly repeated. Interspersed repeats occur at multiple sites throughout the genome, and tandem repeats consist of motifs arranged in a head-to-tail fashion, sometimes up to several thousand at a time. It is thought that most tandem]y repeated DNA i s noncoding. Tandem repeats may be classified according to the length and copy number of the basic repeated element, as well as their genomic location. Satell ite DNA consists of very high repetitions (usuall y between 1000 and 100,000 copies) of a basic motif (usually of size 1 00 to 300 bp) , and was originall y described on the basis of this DNA fraction separating from bulk DNA in buoyant density gradient centrifugation. Sate l li tes occur at few genomic loci only. Minisatell ite DNA is a class of shorter motifs (usually 1 0 to 60 bp), and show a lower degree of tandem repetition at a given locus (Jeffreys et al . , 1985) . Minisatellite DNA occur at many loci in the genome. Microsatel lite DNA, also known as simple tandem repeats (STR) or simple sequence repeats (SSR; Tautz and Renz, 1984) are very short motifs (between 1 and about 10 bp (Sia et al . , 1 997») and have a comparatively low degree of repetition with tracts up to 25 100 bp. They are found evenly dispersed throughout eukaryotic genomes. Table 1 .3 summarises the key features of these tandemly repeated DNA elements. Table 1 .3. Key features of tandemly repeated DNA classes Repeat class satellite minisatellite micro satellite Motif size (bp) - 1 00 to 300 - 1 0 to 60 - 1 to 1 0 Degree of repetition very high moderate low Loci abundance very few many highly dispersed Polymorphism in the number of tandem repeats of microsatelli te tracts are thought to arise by slippage-like events, where misalignment of motifs results in the expansion or contraction of the locus during synthesis or repair (Levinson and Gutman, 1 987; Tautz, 1 989). Slippage synthesis has been demonstrated in vitro (SchlOtterer and Tautz, 1 992), and rates appear to depend on a number of factors including motif length, GC composition and the processivity of the polymerase used. This contrasts with the mechanisms that influence the hypervariabi lity of minisatell i te repeats, which are thought to include unequal crossover events and gene conversion (Charlesworth et aI . , 1 994). PCR-based fingerprinting with semispecific primers uti l ises the interspersed nature of mini- and microsatellite DNA, with primers complementary to these repetitive elements. Amplification patterns are obtained if pairs of such loci are inversely facing, and within amplifiable distance of one another, thus the inter-repeat region is amplified. Like RAPD-PCR, these mini- and microsatellite-primed PCR assays (MP-peR; Meyer et aI . , 1 993) do not require prior sequence knowledge. Fingerprint variation is due to the presence, position, and sequence of these microsatel lite loci , and this method has been used successfully in a growing number of applications, including the differentiation of 26 Cryptococcus neoformans strains (Meyer et aI . , 1 993), and Saccharomyces cerevisiae strains (Baleiras Couto et aI . , 1 996) . Vmiation in the number of repeating microsatel li te motifs at a locus can be detected by amplification across the entire locus using primers specific to the unique sequences flanking the locus. This technique is referred to as microsatel l i te locus PCR (ML-PCR). Microsatel li te loci represent almost optimal markers as they are polymorphic within populations, highly abundant and evenly distributed throughout eukaryotic genomes, inherited in a Mendelian co-dominant fashion, and are fast and easy to type with the option of automated analysis (Heame et aI . , 1 992; Weising et aI . , 1 995a). Microsatell ites also have very rapid evolutionary rates , but their util i ty in phylogenetic reconstruction has had limited success (Goldstein and Pollock, 1 997) . Microsatel lite markers have been shown to be extremely useful in studies of population and conservation genetics including species as diverse as house sparrows (Neumann and Wetton, 1 996) to sperm whales (Richard et aI . , 1 996). Genome mapping has also benefited greatly from the identification of microsate llite markers (Broun and Tanksley, 1 996; Morgante and Olivieri, 1 993 ; Wu and Tanksley, 1993). A major disadvantage in developing microsatel lite markers is that prior sequence information is required to design flanking primer sets. Whereas informative primer pairs may be derived from sequence database entries (Akkaya et aI . , 1 992), the usual strategy is to screen genomic l ibraries with appropriate probes, fol lowed by the sequencing of positive clones (FitzSimmons et aI . , 1 995; Menotti-Raymond and O'Brien, 1 995). Suitable primer sequences need to be identified and synthesi sed before PCR-assay optimisation can be performed. The requirement for cloning and sequencing makes the generation of PCR-based microsatel lite markers time-consuming and expensive, though increasingly more organisms are being studied in this manner. Several approaches have been developed to expedite microsatell i te isolation, including the construction of l ibraries that are enriched for microsatell ites (Edwards et aI. , 1 996; Fisher et aI. , 1 996; Karagyozov et aI . , 1 993 ; Ostrander et aI . , 1 992), and the screening of Southem blots of 27 RAPD profi les with microsatel l i te probes, thus omitting library construction altogether (Ender et aI . , 1 996) . 1 .7.3 FUNGAL MICROSATELLITE LOCI Very little information is known about the frequency and occurrence of microsatel l ite loci in fungal genomes. Extensive database research revealed that the relative abundance of different microsatell i te motifs in plants and animals differ considerably. For example, the (CA)n repeat is one of the most frequently occuning microsatell ites in humans and many mammals . In contrast, (A)n microsatell ites are the most abundant in plants, while (CA)n is comparatively rare (Lagercrantz et aI . , 1 993). Searches also indicated that microsatell ites are five times less abundant in the genomes of plants than in mammals . In fungi, database searches revealed that microsatelhtes of many motif types are widespread, and that (AT)n motifs are by far the predominant type (Groppe et aI . , 1 995). Several features of microsatel l ite loci make them very attractive as markers for the identification of Epichloe endophytes. The high specificity of the primers used would enable endophyte identification in planta, otherwise the endophyte would first need to be cultured out of the grass to be used in techniques such as RAPD-PCR or MP-PCR. Also, interspecific hybrid endophytes would be readi ly identifiable by the presence of multiple allele bands. In 1 995, a polymorphic (AAG)n microsatell ite locus from E. bromicola was isolated after i t had been identified from an analysis of RAPD patterns when assessing the genetic variabil i ty of a population of Bromus erectus endophytes (Groppe et aI . , 1 995) . This is thought to be the first microsatel l i te isolated from a fungal species, though the use of microsatell i te loci as markers in fungal genomes has not yet become as widespread as in plants and animals . 28 1.8 OBJECTIVES Given the importance of the Epichloe endophyte group, not onl y to agricultural pasture systems, but also to our understanding of the biology and evolution of grass-endophyte symbioses, it is imperative that endophyte isolates are able to be detected and accurately identified. As no single method was available to do this expeditiously, one of the objectives of this study was to develop a DNA-based fingerprinting assay for the rapid and accurate identification of Epichloe endophytes in planta. Initial l y a pilot study was carried out to eval uate several peR-based techniques for their abil ity to distinguish between endophyte isolates, and for their ease of use. From this study, the most promising technique was chosen for further development into a working assay, and used in various applications to identify endophyte i solates. In addition to developing the fingerprinting assay, the evolutionary origins of several putative Neotyphodium endophyte isolates from annual Lolium and Hordeum grasses were to be investigated. The objective here was to genetical l y characterise these poorly described isolates using the fingerprinting assay, as well as DNA sequences from the phylogenetical ly i nformative gene regions, tub2 and rDNA-ITS . CHAPTER TWO MATERIALS AND METHODS 2.1 BIOLOGICAL MATERIALS 2.1.1 ENDOPHYTE ISOLATES 30 A list of all endophyte isolates and fungal strains used in this study is gi ven in Table 2 . 1 . 2.1.2 BACTERIAL STRAINS Escherichia coli strain XL-I B lue was used for propagating recombinant c loning vectors (Bul lock et aI . , 1 987). The genotype of this strain is : supE44 hsdR1 7 recA 1 endAl gyrA46 thi relAl lac- F' [proAB+ lacIq lacZLlM1 5 Tn lO(TetR)] . 2.1 .3 CLONING VECTORS The cloning vectors used In this study are listed below, and maps are given In Appendix 1 . Vector Relevant characteristics Source or reference M13mp1 9 7.25 kb Norrander et aI . , 1 983 pUC l 1 8 3 .2 kb AmpR Vieira and Messing, 1 987 pGEM-T 3 .0 kb AmpR Promega Corp. 3 1 Table 2.1. Endophyte and fungal isolates used in this study Isolate Host species Source or reference Epichloe amarillans E52 Sphenopholis obtusata Schardl et aI., 1997 E57 Ag rostis hiemalis Schardl et aI., 1997 Epichloe baeonii E421 Agrostis eapillaris Schardl et aI., 1997 E424 Agrostis tenuis Schardl et aI., 1997 E I031 Calamagrostis villosa Schardl et aI., 1997 Epiehloe braehyelytri E1040 Braehyelytrum ereetum Schardl and Leuchtmann, 1999 E1045 Braehyelytrum ereetum Schardl and Leuchtmann, 1999 E1046 Braehyelytrum ereetltfll Schardl and Leuchtmann, 1999 Epichloe bromicola E501 BrOfllltS erectlts Leuchtmann and Schardl, 1998 E502 Bromus ereetlts Leuchtmann and Schardl , 1998 E798 Bromlts benekenii Leuchtmann and Schardl, 1998 E799 Bromus ramosus Leuchtmann and Schardl, 1998 Epichloe clarkii E422 Holcus lanatus Schardl et aI., 1997 E426 Holeus lanatus Schardl et aI., 1997 E427 Holeus lanatus Schardl et aI., 1997 Epiehloe elymi ES6 Elymus eanadensis Schardl and Leuchtmann, 1999 E184 Elymus virginiells Schardl and Leuchtmann, 1999 Epiehloe festueae Fr l Festuea rubra Leuchtmann, 1994 FrcS Festuca rubra subsp. cotnmutata M. Christensen Frc7 Festuca rubra subsp. commlltata M. Christensen Frr l Festuca rubra subsp. rubra M. Christensen 32 Table 2.1 . continued Isolate Host species Source or reference Epichloe Jestucae Fgl Festuca glauca M. Christensen Fl l Festuca longifolia Leuehtmann, 1994 E28 Festuca longifolia Sehard! et aI., 1997 E32 Festuca rubra subsp. commutata Sehard! et aI., 1997 E186 Festuca rubra subsp. commutata Leuehtmann and Sehardl, 1998 E189 Festuca rubra subsp. rubra Leuehtmann and Sehard!, 1998 E434 Festuca gigantea Sehardl et aI., 1997 Epiehloe glyeeriae E277 Glyceria striata Sehardl and Leuehtmann, 1999 E2772 Glyeeria striata Sehardl and Leuehtmann, 1999 Epichloe sylvatica E354 Brachypodiufll sylvaticum Leuehtmann and Schardl, 1998 E503 Brachypodium sylvaticum Leuchtmann and Schardl, 1998 Epichloe typhina E8 Lolium perenne Schardl et aI., 1997 E348 Phleum pratense Sehardl et aI., 1997 E358 Allthoxanthum odoratum Sehardl et aI., 1997 E425 Phleum pratense Schardl et aI., 1997 E428 Poa trivialis Schardl et aI., 1997 E431 Poa silvieola Schardl et al., 1997 E469 Dactylis glomerata Sehardl et aI., 1997 E470 Anthoxanthum odoratum Sehardl et aI ., 1997 E505 Brachypodium pinnatum Schardl et aI., 1997 E lO21 Poa nemoralis Leuchtmann and Sehardl, 1998 E1022 Poa nemoralis Leuchtmann and Schardl, 1998 E I032 Poa prate/1Sis C. Schardl E l O33 Poa pratensis C. Schardl 33 Table 2.1 . continued Isolate Host species Source or reference Neotyphodium coenophialum (= FaTG- I ) Tf5 Festuca arundinacea Christensen et aI., 1993 Tf27 Festuca arundinacea Christensen et aI., 1 993 Tf28 Festuca arundinacea Christensen et aI., 1993 Tf35 F estuca arundinacea M. Christensen AR542 F estuca arundinacea M. Christensen Neotyphodium lolii (= LpTG-l ) A9501 Lolium perenne L. Mikkelsen ARl Lolium perenne M. Christensen ARIWF Lolium perenne M. Christensen ARICY Lolium perenne M. Christensen CBS232.84 Lolium perenne L. Mikkelsen EndoSafe Lolium perenne M. Christensen Lp5 Lolium perenne Christensen et aI., 1993 Lp6 Lolium perenne Christensen et aI., 1993 Lp7 Lolium perenne Christensen et aI., 1993 Lp9 Lolium perenne Christensen et aI., 1 993 Lp 13 Lolium perenne Christensen et al., 1 993 Lp14 Lolium perenne Christensen et aI., 1993 Lp19 Lolium perenne Christensen et aI., 1993 Lp20 Lolium perenne Christensen et aI., 1993 Lp2 1 Lolium perelllle Christensen et aI., 1993 N. uncinatum FpTG-l) Fpl Festuca pratensis Christensen et aI., 1993 Fp2 Festuca pratensis Christensen et aI., 1993 Fp4 Festuca pratensis Christensen et aI., 1993 Table 2.1. continued Isolate Host species * Neotyphodium sp. (FaTG-2) Tf13 Tf15 Tf20 F estuca arundinacea Festuca arundinacea Festuca arundinacea Neotyphodium sp. (FaTG-3) Tf16 Tf18 AR572 AR577 F estuca arundinacea Festuca arundinacea F estuca arundinacea F estuca arundinacea Neotyphodium sp. (LpTG-2) Lp l Lp2 Neotyphodiul1l spp. AR l7 HaB Hdl Lc l Lc2 Lc3 Lc4 Lml Lm2 Lm3 Lps l Lps2 Lolium perenne Lolium perenne Lolium perenne Hordeum bogdanii Hordeum bogdanii Lolium canariense (020501) Lolium canariense (020502) Lolium canariense (020514) Lolium canariense (020516) Lolium multiflorum cv. Maverick Lolium multiflorulll cv. Maverick Lolium l1lultiflorum cv. Corvette (B3850) Loliulll persicum (39B) Lolium persicum (630001) Source or reference Christensen et aI., 1993 Christensen et aI., 1993 Christensen et aI., 1993 Christensen et aI., 1993 Christensen et aI., 1993 M. Christensen M. Christensen Christensen et aI., 1993 Christensen et aI., 1993 M. Christensen M. Christensen M. Christensen M. Christensen M. Christensen M. Christensen M. Christensen M. Christensen M. Christensen M. Christensen M. Christensen M. Christensen 34 Table 2.1. continued Isolate Neotyphodium spp. Lre l Lre2 Lre3 Lrrl Lrr2 Lrr3 Lrr4 Lro l Ls l Lt l Lt2 Host species Lolium remotum ( 1 5-4) Lolium remotum (40-C) Lolium remotum (42B) Lolium rigidum var. rigidul1l ( 16 . 1 ) Lolium rigidum var. rigidulll (9) Lolium rigidum var. rigidum (01 1460) Lolium rigidum var. rigidum Lolium rigidum var. rottboellioides (36A) Lolium subulatum ( 1 1-4) Lolium temulelltum (45B) Lolium temulelltum (6 1007) Claviceps purpurea Echinodothis tuberiformis 35 Source or reference M. Christensen M. Christensen M. Christensen M. Christensen M. Christensen M. Christensen M. Christensen M. Christensen M. Christensen M. Christensen M. Christensen C. Schardl C. Schardl The host seedline, if known, is given In brackets (Margot Forge Forage Germplasm Centre, Grasslands AgResearch). 2.2 GROWTH OF CULTURES 36 Standard aseptic techniques were employed when managing cultures and steri le glass and plasticware was used throughout. 2.2.1 ISOLATION OF ENDOPHYTE FROM HOST TISSUE Endophyte cultures were isolated from leaf sheath, nodal and immature inflorescence tissue by surface sterilising tissue briefly in 95% ethanol , 10% commercial bleach for 2 to 4 min, then rinsing in sterile water. Tissue was placed on stelile filter paper and leaf sheaths were transversely sectioned into 1 mm pieces. Tissue was transferred to PDA plates (Section 2 .3 .2) that were supplemented with tetracycline, or penici llin and streptomycin (Section 2.3 .6). Plates were sealed with parafilm, incubated at 22°C, and examined regularly under a stereo dissecting microscope for endophyte growth and contamination. 2.2.2 MAINTENANCE OF ENDOPHYTE CULTURES Endophyte cultures growing on PDA (Section 2.3 .2) were maintained by subculturing small blocks of mycelia from the periphery of a growing colony on to fresh PDA. Plates were sealed and incubated inverted at 22°C. 2.2.3 ENDOPHYTE LIQUID CULTURES Endophyte liquid cultures were obtained by grinding a 1 cm2 area of mycelium from the periphery of a colony in 500 III of PDB (Section 2 .3 .2) using a plastic grinder in a " microcentrifuge tube. Flasks of 1 25 ml capacity containing 30 ml PDB were inoculated with 1 00 II I of ground mycelium and incubated at 22°C on a shaking platform at 200 rpm for 7 to 14 days. 2.2.4 Escherichia coli CUI .. TURE CONDITIONS 37 E. coli XL- 1 cultures were plated on LB agar plates (Section 2 .3 . 1 ) with a glass spreader and incubated inverted overnight ( 1 6 to 20 hr) at 37°C. Liquid cultures were obtained by inoculation of 5 ml of LB (Section 2 .3 . 1 ) with a single colony, and overnight incubation at 37°C with shaking at 200 rpm. LB medium was supplemented (Section 2 .3 .6) as required. 2.2.5 PHAGE M13mp19 CULTURE CONDITIONS Plate cultures of M 1 3mp1 9 were obtained by mixing (in order given) 100 III of M 1 3mp19 transformed XL- 1 (Section 2 . 1 2.6), 200 III of overnight culture of XL- 1 (Section 2 .2 .4), 20 II I of X-gal (Section 2 .3 .6), and 20 III of IPTO (Section 2 .3 .6) in a culture tube. Molten top agar (3 ml ; Section 2.3 .4), equi librated to 50°C, was added to the tube, briefly vortexed and poured on to LB agar plates (Section 2 .3 . 1 ) supplemented with tetracycline (Section 2.3 .6). Top agar was allowed to set for 30 min, then plates were incubated inverted at 37°C for 16 to 20 hI'. Liquid cultures of M 1 3mp19 were obtained by adding 250 II I of an overnight culture of XL- l cells (Section 2 .2 .4) to 25 ml of 2 x YT medium (Section 2 .3 .5) and 1 .2 ml aliquots were dispensed into culture tubes. Plaques of interest were picked with a steri le toothpick and used to inoculate culture tubes. Tubes were shaken at 300 rpm for 5 to 6 hI' at 37°C. 2.3 MEDIA All media were prepared with Mil liQ water and steri l i sed at 1 2 1 °C for 1 5 min prior to use. Liquid media was cooled to room temperature before addition of antibiotic(s) and inoculation. Solid media was cool€d to approximately 50°C before addition of antibiotic(s) and pouring into steri le petri dishes. 2.3.1 LURIA BROTH MEDIA 38 Luria broth (LB) contained 10.0 g tryptone (Difco LaboratOlies), 5 .0 g yeast extract (Difco Laboratories), 5 .0 g NaCl per l i tre, and was adjusted to pH 7 .0 . LB agar was prepared by adding 1 5 g of agar (Life Technologies) per l itre of LB media. Where required, LB media was supplemented (Section 2.3 .6). 2.3.2 POTATO DEXTROSE MEDIA Potato dextrose broth (PDB) contained 24.0 g of dehydrated potato dextrose broth (Difco Laboratories) reconstituted in 1 l itre of Mil l iQ water and adj usted to pH 6 .5 . Potato dextrose agar (PDA) was prepared by addition of 1 5 g agar (Life Technologies) per l itre of PDB . Where required, PD media was supplemented (Section 2 .3 .6) . 2.3.3 SOC MEDIUM sac medium (Dower et aI . , 1988) contained 20.0 g tryptone (Difco Laboratories), 5.0 g yeast extract (Difco Laboratories), 0.6 g NaCI, 0.2 g KCI , 0 .95 g MgC}z, 2.5 g MgS04.7H20, and 3 .6 g glucose per l i tre; and adjusted to pH 7 .0. 2.3.4 TOP AGAR Top agar contained 10.0 g tryptone (Difco Laboratories), 8 g NaCI, and 8 g agar (Life Technologies) per l i tre. 2.3.5 2 x YT MEDIUM 2 x YT medium contained 1 6 g tryptone (Difco Laboratories), 1 0 g yeast extract (Difco Laboratories), and 5 .0 g NaCI per l i tre. The medium was adjusted to pH 7 .4 and stored in 25 ml aliquots. 2.3.6 MEDIA SUPPLEMENTS Media was supplemented at the following concentrations where required. Supplement Stock concentration Final concentration Ampicillin 100 mg/ml 1 00 llg/ml Penicillin 100 mg/ml 1 00 llg/ml Streptomycin 100 mg/ml 100 llg/ml Tetracycline 5 mg/ml 7 . 5 1lg/ml IPTG 200 mg/ml 24O Ilg/mJ X-gal 20 mg/ml 24 llg/ml 39 All supplements were resuspended in MiIliQ water and steri lised through a 0.22 Ilm filter, except tetracycline, which was dissolved in ethanol , and X-gal, which was dissolved in dimethylformamide. 2.4 COMMON BUFFERS, SOLUTIONS AND REAGENTS 40 All buffers, solutions and reagents were prepared with MilliQ water and stored at room temperature unless otherwise indicated. 2.4.1 COMMON STOCK SOLUTIONS Common stock solutions are listed below. Stock Ethanol Ethidium bromide MgCb NaCl Disodium EDTA (Na2EDTA) Sodium acetate Sodium dodecyl sulphate (SDS) Tris-HCI 2.4.2 ACRYLAMIDE MIX Concentration 70% (v/v) 10 mg/ml 0 .5 M I M O.S M 3 M 20% (w/v) I M pH 8.0 4.2 and 7.0 7 .6 and 8 .0 Acrylamide mix for polyacrylamide gels contained 288 g of urea, 34 .2 g of acrylamide and 1 . 8 g of bis-acrylamide made up to 500 ml with MilliQ water, and deionised with S g of amberlite MB-3 resin (Sigma) for at least 30 min with stirring. The solution was fi ltered through a porous sintered glass funnel and made up to a final volume of 600 ml with 60 mI of 10 x sequencing TBE (Section 2.4. 1 1 ) and Mil I iQ water. 2.4.3 DIG HYBRIDISATION AND DETECTION BUFFERS 4 1 The following set of buffers were used for the chemiluminescent detection of hybridised DIG-labelled probes, as described in Sections 2 . 1 1 .2 and 2. 1 1 .3 . The washing and blocking buffers (Sections 2.4.3 .3 and 2.4.3 .4) were made on the day of use. 2.4.3.1 Standard Hybridisation Buffer contained 5 x SSC (Section 2.4.8 .4), 0. 1 % (w/v) N-Iauroylsarcosine, 0.02% (w/v) SDS and 1 .0% (w/v) blocking reagent (Boehringer Mannheim GmbH). This buffer was autoclaved before use. 2.4.3.2 Maleic Acid Buffer contained 0 . 1 M maleic acid and 0. 1 5 M NaCI, adjusted to pH 7 .5 with 1 0 M NaOH and autoclaved. 2.4.3.3 Washing Buffer consisted of Maleic Acid Buffer (Section 2.4.3 .2) with addition of 0 .3% (v/v) Tween 20. 2.4.3.4 Blocking Buffer consisted of Maleic Acid Buffer (Section 2.4.3 .2) with addition of 1 .0% (w/v) blocking reagent (Boehringer Mannheim GmbH) di ssolved by gently heating in a microwave oven without boiling. 2.4.3.5 Detection Buffer contained 100 mM Tris-HCI (pH 9 .5) and 1 00 mM NaCL 2.4.4 LIGASE REACTION BUFFER Ligase reaction buffer (King and Blakesley, 1 986) at 5 x concentration contained 250 mM Tris-HCI (pH 7.6), 50 mM MgCh, 25% (w/v) polyethylene glycol 8000, 5 mM ATP, and 5 mM dithiothreitoI . Buffer was stored at -20°C. 2.4.5 LYSOZYME 42 Lysozyme was resuspended to a concentration of 1 0 mg/ml in 1 0 mM Tris-HCI (pH 8.0) , dispensed into small aliquots, and stored at -20°e. 2.4.6 RNaseA (DNase Free) RNaseA was prepared at a concentration of 1 0 mg/ml in 1 0 mM Tris-HCI (pH 7 .5) and 1 5 mM NaCl. The mixture was heated at 100°C for 1 5 min, allowed to cool slowly to room temperature, dispensed into small aliquots, and stored at -20°C. When used in an enzyme digestion reaction, RNase stock was diluted further to 0.2 mg/ml with MilliQ water. 2.4.7 SDS LOADING DYE SDS loading dye contained 1 % (w/v) SDS, 0.02% (w/v) bromophenol blue, 20% (w/v) sucrose and 5 mM Na2EDTA (pH 8 .0). 2.4.8 SOUTHERN BLOTTING SOLUTIONS The following solutions were used in the Southern blotting procedure described in Section 2 . 10 . 2.4.8.1 Depurination Solution consisted of 0.25 M HCl . 2.4.8.2 Denaturation Solution contained 0 .5 M NaOH and 0 .5 M NaCI . 2.4.8.3 Neutralisation Solution contained 2.0 M NaCI and 0 .5 M Tri s-HCI (pH 7.4). 2.4.8.4 20 x SSC contained 3 M NaCI and 0.3 M sodium citrate; pH 7.0 . 2.4.8.5 2 x SSC contained 0.3 M NaCI and 0.03 M sodium citrate; pH 7.0 . 2.4.9 T AE BUFFER 43 At 1 x concentration TAE buffer contained 40 mM Tris-acetate and 2 mM Na2EDTA; pH 8 .5 . 2.4.10 TBE BUFFER At 1 x concentration TBE buffer contained 89 mM Tri s, 89 mM boric acid and 2.5 mM Na2EDTA; pH 8 .2 . 2.4.1 1 TBE SEQUENCING BUFFER At 1 x concentration TBE sequencing buffer contained 1 34 mM Tris, 45 mM boric acid and 2.5 mM Na2EDTA; pH 8 . 8 . 2.4.12 TE BUFFERS TE ( 10 : 1 ) contained 10 mM Tris-HCl and 1 mM Na2EDTA; pH 8 .0 . TE ( 10 :0. 1 ) contained 10 mM Tris-HCI and 0. 1 mM Na2EDTA; pH 8 .0 . 2.4.13 TRIS-EQUILIBRA TED PHENOL Tris-equilibrated phenol was supplied by Amersham and supplemented with 0. 1 % (w/v) 8-hydroxyquinoline before storage at 4°C. 2.5 DNA ISOLATION AND PURIFICATION PROCEDURES 2.5.1 LARGE SCALE ISOLATION OF ENDOPHYTE TOTAL DNA 44 Large scale fungal genomIc DNA was prepared by a modification of the method described by Byrd et a1. ( 1990) . Mycelia from liquid cultures (Section 2.2 .3) was harvested by centrifugation, washed in MilliQ water and freeze dried. About 100 mg of freeze dried mycelia was ground to a fine powder under liquid nitrogen in a pre-chi lled mortar and pestle, and resuspended in 1 0 ml of extraction buffer ( 1 50 mM NazEDTA, 50 mM Tris-HCI (pH 8 .0), 1 % (w/v) sodium lauroyl sarcosine, and 2 mg/ml proteinase K). This mixture was incubated at 37°C for 20 min, then centrifuged for 10 min at 20,200 x g and 4°C to pellet the cellular debris . All subsequent centrifugation steps were carried out at this speed and temperature. The supematant was extracted successively with equal volumes of phenol, phenol/chloroform ( 1 : 1 ; v :v) and chloroform with mixing by inversion and centrifugation for 1 5 min between extractions. DNA was precipitated with an equal volume of isopropanol followed by chilling at -20°C for 1 5 min and centlifugation for 30 min . To remove polysaccharides the pelleted DNA was resuspended in 5 ml 1 M NaCI then centrifuged for 5 min. DNA was precipitated and pelleted as described, and the pellet was rinsed in 70% ethanol, air dried, and resuspended in 0.5 to 1 .0 ml steri le MilliQ water. 2.5.2 MICROWA VE MINIPREP ISOLATION OF ENDOPHYTE TOTAL DNA Total genomic DNA from very small amounts of fresh mycelium was extracted using a modified microwave miniprep method based on that of Goodwin and Lee ( 1993). Around 3 to 15 mg mycelium was ground in 100 1-11 of lysis buffer (50 mM Tri s-HCI (pH 7.2), 50 mM NazEDTA, 3% (w/v) SDS, and 1 % (v/v) 2-mercaptoethanol) in a microcentrifuge tube with a plastic grinder. Sample tubes were left open and loosely covered in plastic food wrap, and heated in a microwave oven (The Genius 650 W, National) on full power for a series of 1 5 sec, 10 sec and 5 sec bursts to prevent boiling. Immediately after heating, 300 1-1 1 lysis buffer was added to each sample, briefly mixed 45 and incubated at 80°C for 10 min. Sequential extractions with equal volumes of phenol, phenol :chloroform ( 1 : 1 ; v :v) and chloroform were performed with vortexing and centrifugation at 1 5 ,000 x g for 1 5 min between extractions. The aqueous phase was removed and DNA precipitated by addition of 1 6 11 1 3 M sodium acetate (pH 7.0) and 0 .54 volumes of i sopropanol. DNA was pelleted by centrifugation at 1 5 ,000 x g for 1 5 min, briefly washed in 70% ethanol, then air dried. Pellets were resuspended in 30 to 50 III of MilI iQ water, depending on the initial mass of mycelium used. 2.5.3 TOT AL DNA ISOLATION USING A F ASTDNA KIT DNA suitable for PCR amplification was extracted from fungal culture and plant tissue samples using the FastDNA Kit (Bio 10 1 Inc . ) . To a lysing matrix tube containing a 6 mm ceramic sphere and garnet matrix , 10 to 200 mg of tissue was added followed by a 6 mm ceramic cylinder. To this, 800 III CLS-VF and 200 III PPS were added and the sample processed in a FastPrep shaker (Bio 10 1 Inc. ) for 30 sec at speed 4.5 . Samples were centrifuged for 5 min at 1 5 ,000 x g and the available supematant was transferred to a new microcentlifuge tube. The supematant was often contaminated with fine tissue debris , so tubes were spun for an additional 5 min at 15 ,000 x g, and 600 II I of the clear supematant was transferred to a fresh tube containing 600 III binding matrix . To allow DNA binding, samples were gently mixed and incubated at room temperature for 5 min, then spun for 1 min and the supematant discarded. The pellet was washed by gently resuspending in 500 II I SEWS-M, then tubes were spun for 1 min and the supematant discarded. A further 10 sec spin was performed to allow removal of residual liquid with a small bore tip. DNA was eluted from the binding matrix by addition of 100 III DES followed by a 2 to 3 min incubation at room temperature. Samples were centlifuged for 1 min at 1 5 ,000 x g and the DNA containing supematant transferred to a new tube. 2.5.4 PLANT DNA ISOLATION USING A CTAB METHOD Total genomic DNA was isolated from endophyte-infected and endophyte-free plant tissue using the hexadecyltrimethylammonium bromide (CTAB) based method of Doyle 46 and Doyle ( 1990). Fresh grass pseudostem tissue, 0 .5 to 1 .0 g, was ground to a fine powder in a chilled mortar and pestle with liquid nitrogen, transferred to 5 .0 to 7 .5 ml CTAB i solation buffer (2% (w/v) CTAB, 1 .4 M NaCl, 0.2% (v/v) 2-mercaptoethanol, 20 mM Na2EDTA, and 100 mM Tris-HCI (pH 8 .0» , that was preheated to 60°C, and swirled gently to mix. The sample was incubated at 60°C for 30 min, then extracted once with chloroform-isoamyl alcohol (24: 1 ; v :v) . Tubes were centrifuged at 1 ,600 x g for 5 min, the aqueous phase removed and a 0.67 volume of cold isopropanol was added to precipitate the DNA. Tubes were centrifuged at 500 x g for 1 to 2 min, the supernatant gently poured off, and the pellet washed for 20 min in 10 to 20 ml of wash buffer (76% (v/v) ethanol and 10 mM ammonium acetate). Samples were centrifuged at 1 ,600 x g for 1 0 min, the supernatant poured off and the DNA pellet air dried. DNA was resuspended in 1 ml TE ( 1 0: 1 ; Section 2.4. 1 2) . 2.5.5 PLASMID DNA EXTRACTION USING A RAPID BOILING METHOD Plasmid DNA from E. coli was isolated by the method of Holmes and Quigley ( 198 1 ). Cells from 1 .5 ml of an overnight culture (Section 2 .2 .4) were pelleted by centrifugation and thoroughl y resuspended in 350 II I STET buffer (8% (w/v) sucrose, 5% (v/v) Triton X- lOO, 50 mM Na2EDTA (pH 8 .0), and 50 mM Tris-HCI (pH 8 .0» . To this, 25 II I lysozyme (Section 2.4.5) was added, the solution vortexed and heated in a boiling water bath for 40 sec, immediately followed by centlifugation for 1 0 min at 1 5 ,000 x g. The gelatinous pellet was removed with a toothpick and DNA was precipitated by addition of an equal volume of isopropanol and cooling at -20°C for 30 min. DNA was pelleted by centrifugation for 10 min at 1 5,000 x g, washed in 70% ethanol, air dried and resuspended in 50 III Mill iQ water. 2.5.6 PLASMID DNA EXTRACTION USING AN ALKALINE LYSIS METHOD Plasmid DNA for sequencing was prepared by the alkaline lysis mini prep method of Sambrook et al . ( 1989). Cells from 1 .5 ml of an overnight culture of E. coli 47 (Section 2 .2 .4) were pelleted by centrifugation for 3 min at 1 5 ,000 x g (all centrifugation steps were performed at this speed) and resuspended in 1 00 II I of solution I (50 mM glucose, 25 mM Tris-HC1 , and 10 mM Na2EDTA; pH 8 .0), followed by 200 II I of solution II (0.2 M NaOH and 1% (w/v) SDS). The contents of the tube were mixed thoroughly by inversion, then 1 50 II I solution In (29.44 g potassium acetate and 1 1 . 5 ml glacial acetic acid per 1 00 ml) was added and the mixture vortexed with the tube inverted. The solution was incubated on ice for 3 to 5 min and centrifuged for 5 min. To the supernatant, an equal volume of phenol (Section 2.4. 1 3) was added, then vortexed and centrifuged for 5 min. The aqueous phase was extracted with an equal volume of phenol/chloroform ( 1 : 1 ; v :v) , followed by an equal volume of chloroform. DNA was precipitated with two volumes of 95% ethanol, centrifuged for 5 min , washed with 70% ethanol , then air dried and resuspended in 50 III of TE ( 10:0 . 1 ; Section 2.4. 1 2) . 2.5.7 PLASMID DNA EXTRACTION USING A QUANTUM MINIPREP KIT Plasmid DNA required for sequencing or vector preparation was extracted usmg a Quantum Prep Plasmid Miniprep Kit (Bio-Rad). Cells from 1 .5 to 3 .0 ml of an overnight culture of E. coli (Section 2 .2 .4) were pelleted by centrifugation for 3 min at 1 5 ,000 x g (all centrifugation steps were all performed at this speed) and resuspended in 200 III of cell resuspension solution. Cell lysis solution, 250 Ill , was added and the tube inverted 1 0 times to lyse the cells, then 250 III of neutralisation solution was added and the tube inverted a further 1 0 times to mix . Cellular debris was pelleted by centrifugation for 5 min and 200 II I matrix solution was added to the supernatant and mixed. The suspension was poured into a spin fi lter that had been inserted into a microcentrifuge tube, spun for 30 sec, and the filtrate discarded. The matrix was washed by addition of 500 II I wash solution to the spin filter, fol lowed by centrifugation for 30 sec, and the supernatant discarded. A second wash was performed by addition of 500 II I wash solution followed by centrifugation for 2 min to remove all traces of ethanol . DNA was eluted by addition of 100 II I MilliQ water to the spin column in a new microcentrifuge tube and centrifuged for 30 sec. 2.5.8 SINGLE-STRANDED M 13mp19 DNA ISOLATION 48 Single-stranded M13mp19 DNA suitable for sequencing was prepared from phage liquid cultures (Section 2.2 .5) . Cells were pelleted by centIifugation for 5 min at 1 5 ,000 x g and 0 .8 ml of the supernatant was transferred to a new tube. The remainder of the l iquid was aspirated off and the cells retained to extract the replicative form of the phage by the rapid boil method (Section 2 .5 .5) . To the phage supernatant, 200 III of a solution of 20% PEG 8000 in 2.5 M NaCl was added and incubated at room temperature for 1 5 min. Samples were centrifuged for 5 min at 1 5 ,000 x g and the supernatant aspirated off to leave the phage pellet. Addition of 100 II I TE ( 10:0 . 1 ; Section 2.4. 1 2) and 50 II I phenol to the pellet was fol lowed by vortexing for 1 0 sec, standing for 5 min and vortexing for 1 0 sec. Samples were centrifuged for 3 min at 1 5 ,000 x g and the aqueous phase removed and sequentially extracted with equal volumes of phenol : chloroform ( 1 : 1 , v :v) and chloroform. DNA was precipitated by addition of 1 0 III 3 M sodium acetate, 250 II I 9 5% ethanol and chilled overnight at -20°C. DNA was pelleted by centrifugation for 5 min at 1 5 ,000 x g, washed in 95% ethanol, air dried and resuspended in 30 III TE ( 10:0. 1 ; Section 2.4. 1 2) . 2.5.9 PCR PRODUCT PURIFICATION Polymerase chain reaction (PCR) amplification products were purified from salts, free nucleotides, polymerases, and primers using the PCR Clean Up Kit (Boehringer Mannheim GmbH). PCR reactions were extracted with an equal volume of chloroform and centrifuged for 5 min at 1 5 ,000 x g. The aqueous phase was extracted and made up to 1 00 III with TE ( 10:0. 1 ; Section 2.4. 1 2). Nucleic acid binding buffer, 400 Il l , was added followed by 10 III silica suspension. The mixture was incubated for 10 min at room temperature with vortexing every 2 to 3 min , then the matrix was pelleted by a 30 sec centrifugation and the supernatant discarded. The pellet was resuspended in 400 III nucleic acid binding buffer, vortexed, pelleted, and the supernatant discarded. The pellet was washed by resuspension in 400 II I washing buffer, centrifugation and the supernatant discarded. This step was repeated once, then the pellet air dried at room 49 temperature for 1 5 min. DNA was eluted in 20 to 50 II I TE ( 10:0. 1 ; Section 2.4. 1 2) or MilliQ water at pH 8.0 to 8 . 5 for 1 0 min, then centrifuged for 30 sec and the DNA containing solution recovered. 2.5.10 ISOLATION OF DNA FROM AGAROSE DNA was recovered from SeaPlaque agarose gels , 0 .7% to 1 .5% (w/v) in TAE electrophoresis buffer (Section 2 .4.9), by phenol freeze extraction (ThUling et aI . , 1 975) . Following electrophoresis and ethidium bromide staining (Section 2 .8 .4), fragments of interest were excised under long wave UV light (wavelength 366 nm) and excess agarose was trimmed away. The agarose was melted at 65°C, mixed thoroughly with an equal volume of tris-equil ibrated phenol (Section 2.4. l 3) and frozen at -20°C for at least 2 hr. The tube was centrifuged for 10 min at 1 5 ,000 x g at 4°C, the aqueous phase recovered and extracted with an equal volume of chloroform. DNA was precipitated by addition of 1 1 10 volume of 3 M sodium acetate, 2 .5 volumes of 95% ethanol and chilled at -200C for at least 2 hr. DNA was pelleted by centrifugation for 1 5 min at 1 5 ,000 x g, washed in 95% ethanol , air dried and resuspended in TE ( 10:0. 1 ; Section 2.4. 1 2) or MilliQ water. 2.6 DNA QUANTIFICATION DNA preparations were quantified using one or both of the methods described. 2.6.1 FLUOROMETRIC QUANTIFICATION DNA was quantified on a Hoefer Scientific TKO 100 Fluorometer using a dye solution containing 1 x TNE buffer ( 1 0 mM Tris-HCl , 1 mM Na2EDTA, and 100 mM NaCl ; pH 7 .4) and 0. 1 Ilg/ml Hoechst 33258 dye. The instrument was calibrated by adjusting the zero dial until a '0' reading was obtained from 2 ml dye only, and adjusting the scale dial until a ' 100' reading was consistently obtained when 2 II I of 1 00 ng/Il l calf thymus DNA standard was added to the dye. Using these settings, the concentration of the DNA 50 sample to be quantified was detennined by adding 2 III of DNA (at a suitable dilution in water if necessary) to 2 ml of TNE dye. The resulting reading was recorded as the DNA concentration in ng/Il l. 2.6.2 MINI GEL METHOD A 1 or 2 II I aliquot of DNA to be quantified was electrophoresed (Section 2 .8 .2) on an agarose mini gel alongside a series of standard DNA solutions of known concentration. In general, A DNA standards were used when quantifying genomic DNA, pUC l 1 8 DNA standards were used when quantifying plasmid DNA, and a low DNA mass ladder (Life Technologies) was used when quantifying PCR products. After staining with ethidium bromide and photographing (Section 2 .8 .4), the DNA concentration was estimated from the relative intensity of the sample of interest to those of the standard concentration samples. 2.7 RESTRICTION ENDONUCLEASE DIGESTION OF DNA Restriction endonuclease digestion reactions were set up on ice using the commercial buffer supplied by the manufacturer and a 2 to 1 0-fold excess of restriction enzyme. Digestion reactions were incubated at the temperature recommended by the manufacturer, for 2 to 3 hr for plasmid DNA or 3 hr to overnight for genomic DNA. A small aliquot of digested DNA was checked on a minigel (Section 2 . 8 .2) to ensure the digestion had gone to completion. If the reaction was incomplete, more enzyme was added and the reaction incubated for an additional hour or two. Once complete, a 1 /5 volume of SDS loading dye (Section 2.4.7) was added to stop the reaction if the DNA was to be electrophoresed, otherwise the reaction was stored at -20°C until required. 2.8 AGAROSE GEL ELECTROPHORESIS 2.8.1 AGAROSE GELS AND DNA SIZE MARKERS 5 1 The concentration of agarose used was detennined by the desired range of fragment separation as outlined in the table below. Agarose was added to 1 x TAE or 1 x TBE buffer (Sections 2.4.9 and 2.4. 10) and melted by microwaving or pressure cooking. Molten agarose was equilibrated to 50°C before pouring. All gels were run in 1 x TBE buffer except when purifying DNA from agarose, a low melting point agarose in 1 x TAE buffer was used. DNA size markers were selected as outlined below depending on the range of fragment separation required, and estimates of band size were detennined from the relative migration of the subject bands compared to the standard size markers. DIG-labelled DNA size markers (Boehringer Mannheim GmbH) were run if the gel was to be blotted and probed using the DIG system (Section 2. 1 1 ) . Agarose grade % (w/v) DNA standard size Size range of of agarose markers separation Molecular Biology (BRL) 0.7 A HindUI 2 to 25 kb Molecular Biology (BRL) 1 .0 A HindIII 0.5 to 1 0 kb Molecular Biology (BRL) 1 .2 A HindUI 0.4 to 5 kb NuSieve (FMC) 2.0 1 kb ladder 0.3 to 1 .5 kb NuSieve (FMC) 3 .0 50 or 100 bp ladder 0. 1 to 1 .0 kb SeaPlaque (FMC) 0.7 A HindUI 0 .5 to 25 kb SeaPlaque (FMC) l .25 A HindIII 0. 1 5 to 6 kb 2.8.2 MINI GEL ELECTROPHORESIS 52 Horizontal agarose mini gels, of approximately 25 ml volume, were run at 60 to 1 00 V in 1 x TAE or 1 x TBE buffer (Sections 2.4.9 and 2.4 . 10) . Gels were run until the SDS loading dye (Section 2.4.7) had migrated approximately % of the way down the gel . 2.8.3 LARGE GEL ELECTROPHORESIS Large gels, of approximately 1 50 ml volume, were run in a BioRad Sub-Cell apparatus at 30 V for about 1 8 hr, or until the SDS loading dye (Section 2.4.7) had migrated at least 1 50 mm from the wells . 2.8.4 STAINING AND PHOTOGRAPHING GELS After electrophoresis, gels were stained in 1 I1g/ml ethidium bromide for 1 0 to 20 min for a mini gel , or 30 min for a large gel. Gels were destained in water and bands were visualised on a UV transilluminator and photographed with either Polaroid 667 film (Polaroid Corp.) or an Alpha Innotech gel documentation system. 2.9 POLYACRYLAMIDE GEL ELECTROPHORESIS (PAGE) Denaturing polyacrylamide gel electrophoresis (PAGE) was used to separate DNA fragments of small size including radio-labelled sequencing reactions (Section 2 . 1 5 . 1 ) and radio-labelled microsatellite locus PCR products (Section 2 . 1 3 .5) . PAGE gels (30 x 40 cm) were poured between glass plates with 0.4 mm spacers using 60 ml acrylamide mix (Section 2.4.2) containing 36 11 1 of N, N, N', N ;-tetramethylethylenediamine (TEMED) and 360 III of freshly prepared 10% (w/v) ammonium persulphate. Gels were run on a model S2 sequencing gel electrophoresis apparatus (Life Technologies), and were pre-run for 30 min at 1 ,500 V in 1 x sequencing TBE (Section 2.4. 1 1 ) . Samples with stop solution added were denatured at 95°C for 2 min and quenched on ice prior to 53 loading 2 to 3 II I onto the gel . Sequencing samples were run for either a short run of 2 hr (until the first blue dye front ran off) or a long run of about 6 hr (when 3 dye fronts had run off). The run time for microsatellite PCR products was adjusted to allow maximal separation of products and depended on the expected DNA fragment size for the microsatellite locus being run. Once run, gels were transferred to blotting paper, dried on a model 583 gel dryer (Bio-Rad) for 2 hr at 80°C, then exposed to X-ray film (Fuji Photo Film Co.) overnight or longer if necessary. 2.10 SOUTHERN BLOTTING The blotting method used is based on that of Southern ( 1 975). DNA to be transferred was separated on a large agarose gel (Section 2 .8 .3) alongside DIG-labelled DNA standard markers (Boehringer Mannheim GmbH), stained, destained, and photographed as described in Section 2 .8 .4. The gel was immersed and gently agitated in depUlination solution (Section 2.4. 8 . 1 ) for 1 5 min, denaturation solution (Section 2 .4 .8 .2) for 30 min, and neutrali sation solution (Section 2.4.8 .3) for 30 min. A final wash was carried out in 2 x SSC (Section 2.4.8 .5) for 2 min. The blotting apparatus was assembled on a plastic trough with wells at each end. Two sheets of Whatman 3MM chromatography paper, cut to the width of the trough and just longer than the length of the apparatus, were soaked in 20 x SSC (Section 2.4.8 .4) then placed in the trough so that the ends of the paper were submerged in the wells to act as a wick. The wells were fil led with 20 x SSc. A piece of plastic food wrap was placed over the entire trough and a 'window' was cut from the centre that was 2 mm smaller on each side than the size of the gel . The treated gel was placed over the window and a piece of positively charged nylon membrane (Boehringer Mannheim GmbH), cut to 2 mm larger than the gel and presoaked in 2 x SSC, was placed on the gel . Four sheets of 3MM paper were cut to the size of the gel . Two of these were presoaked in 2 x SSC and laid on the nylon membrane fol lowed by two dry sheets. A stack of paper towels, approximately 100 mm high, was laid on top of the 3MM paper fol lowed by a large 54 plastic tray containing a weight of approximately 500 g to keep the stack flat. Following overnight DNA transfer, the apparatus was disassembled and the membrane washed in 2 x SSC for 5 min, then air dried. DNA was cross-linked by placing the membrane face down on a UV transilluminator and exposed to UV light for 2 min . 2.11 SOLUTION HYBRIDISATION USING DIGOXIGENIN (DIG) The nonradioactive DIG system (Boehringer Mannheim GmbH) was used for all hybridisation work and procedures were based on The DIG System User's Guide for Filter Hybridisation manual (Boehringer Mannheim, 1 995). 2.11 . 1 DIG 3'-END LABELLING OF OLIGONUCLEOTIDE PROBES Oligonucleotide probes, listed in Table 2.2, were 3 '-end labelled with DIG using the following protocol . In a microcentrifuge tube on ice, 4 }!1 of 5 x reaction buffer ( 1 M potassium cacodylate, 1 25 mM Tris-HCI, 1 .25 mg/ml bovine serum albumin ; pH 6.6) , 4 } l l of 25 mM cobalt chloride, 1 00 pmol of oligonucleotide, 1 }l l of 1 mM digoxigenin­ I I -ddUTP, 1 U of terminal transferase, and MilliQ water to a total volume of 20 }!l was added and gently mixed. The reaction was incubated at 37°C for 1 5 min and placed on ice where the reaction was stopped by addition of 1 }l l 200 mM Na2EDTA (pH 8 .0). The labelled oligonucleotide probe was stored at -20°C. A control oligonucleotide, C l , was included in each batch of labelling reactions to monitor the label ling of sample oligonucleotides. To check the yield of labelled probe, a side by side comparison of the DIG-labelled sample oligonucleotide with a commercially prepared DIG-labelled oligonucleotide control (Boehringer Mannheim GmbH) was performed. A five-fold dilution series in MilliQ water of the sample and control probes was made, including concentrations 50 fmoll}l l , 10 fmol/}!l , 2 fmol/}l l , 0.4 fmol/}lJ , and 0.08 fmol/}ll . Aliquots of 1 }ll from each dilution were spotted in rows onto positively charged nylon membrane (Boehringer 55 Mannheim GmbH), allowed to air dry, and cross-linked (Section 2 . 1 0). Chemiluminescent detection was performed on the membrane as described in Section 2 . 1 1 .3 and spot intensities were compared between the labelled control and sample dilutions to estimate the concentration of the sample probes . 2.1 1.2 HYBRIDISATION AND WASHING CONDITIONS Hybridisations were carried out in roller bottles in a Bachofer hybridisation oven for Southern blots (Section 2 . 1 0) , and in small plastic pots in a shaking waterbath for plaque filter lifts (Section 2 . 1 2.7) . The membranes were pre-hybridised for an hour in DIG standard hybridisation buffer (Section 2.4.3 . 1 ) and the hybridisation temperature used was 60°C for (CA)1 5 and (GA)1 5 dinucleotide probes, or 50°C for (CAA)lO, (GAA)lO, and (ATC)lO trinucleotide probes. Hybridisations were performed using 2 pmol/ml of each DIG 3 '-end labelled oligonucleotide probe (Section 2 . 1 1 . 1 ) in DIG standard hybridisation buffer at the pre-hybridisation temperature for 2 to 6 hr. Following hybridisation, blots were washed twice in 0. 1 x SSC for 10 min each at room temperature. A final stringency wash was performed for 2 min using a combination of temperature and salt concentration, which for low stringency conditions, was approximately 1 0 to 20°C below the Tm for the probe (calculated using the formula of Bolton and McCarthy, 1 962) and for high stringency conditions was approximately 0 to 5°C below the T m. Hybridised probes were then detected by chemiluminescent detection as described in the fol lowing section. 2.1 1.3 CHEMILUMINESCENT DETECTION OF DIG-LABELLED PROBES Chemiluminescent detection was carried out at room temperature in clean plastic trays with gentle agitation to ensure the membranes were completely covered with solution. At no time were membranes allowed to air dry. After stringency washes the membranes were washed for 1 min in washing buffer (Section 2.4.3 .3) , then blocked in blocking buffer (Section 2.4.3 .4) for 30 min. Anti-Digoxigenin-AP antibody (Boehringer Mannheim GmbH) was diluted 1 : 10,000 in blocking buffer and membranes were 56 incubated in this for 30 min, then washed twice in washing buffer for 1 5 min each. Membranes were equilibrated in detection buffer (Section 2.4.3 .5) for 2 min, drained briefly and transfened face up to an opened out clear plastic bag. CSPD® (Boehringer Mannheim GmbH) was diluted 1 : 100 in detection buffer and pipetted over membranes. The bag was closed to create a liquid seal around the membranes and incubated for 5 min. The membranes were drained briefly, transfened to a clean plastic bag, sealed and incubated at 37°C for 1 5 min to activate the alkaline phosphatase. The membranes were then exposed to X-ray film (Fuji Photo Film Co.) for 30 to 1 20 min and developed. 2.12 LIBRARY CONSTRUCTION AND SCREENING Partial l ibraries of size-selected DNA fragments from endophyte strains ES (E. typhina) and Lpl (LpTG-2) were constructed in the vector M13mp 1 9 (Section 2. l .3) and screened using methods described by Sambrook et al . ( 1989). 2.12.1 CALF ALKALINE PHOSPHATASE (CAP) TREATMENT OF VECTOR Digested vectors were treated with calf alkaline phosphatase (CAP) to prevent self ligation of vector in subsequent ligation reactions . Vector DNA, 5 )lg , was digested to completion in an excess of the appropriate restriction enzyme (Section 2.7). The restriction enzyme was inactivated by heating for 10 min at 65°C and 0 .5 U of CAP was added and incubated at 37°C for 30 min. Na2EDTA (pH S .O) to a final concentration of 5 mM, SDS to give a final concentration of 0.5%, and proteinase K to a final concentration of 50 )lglml were added, and the mixture was incubated at 56°C for 30 min . The DNA was purified by sequential extractions with equal volumes of phenol , phenol/chloroform ( 1 : 1 , v :v) , and chloroform with centrifugation for 5 min at 1 5 ,000 x g between steps. DNA was then precipitated by addition of 1 1 10 volume of 3 M sodium acetate (pH 7 .0) and 2 .5 volumes of 95% ethanol , chilling at -20°C for 30 min and centrifugation for 10 min at 1 5 ,000 x g. The pellet was washed in 70% ethanol, air dried, resuspended in MilliQ water to a concentration of 20 ngf)l l , and stored at 4°C. 2.12.2 PREPARA TION OF LINKERS 57 BamHI linkers were constructed from the self annealing of 5 1-chemically phosphorylated 5 1-pCGGGATCCCG oligonucleotides (Life Technologies) that were based on a sequence by Stratagene. In a microcentrifuge tube, 2 III of 1 0 x annealing buffer ( 1 00 mM Tris-HCl (pH 7.6), 1 0 mM Na2EDTA, and 1 M NaCI), 4 nmol of oligonucleotide, and MilliQ water to a total volume of 20 II I was mixed. The tube was incubated in a small waterbath at 65°C for 10 min , then the waterbath turned off and allowed to cool slowly to room temperature. The annealed Iinkers, at a final concentration of 1 00 pmol/ll l , were stored at _20DC. 2.12.3 PREPARATION OF LINKERED INSERT DNA Genomic inserts were prepared by the digestion of genomic DNA using combinations of the blunt end 4 bp cutters AluI, RaeIII, and ThaI, as well as BamHI (Section 2.7) . Digests were separated on SeaPlaque gels in T AE buffer (Section 2.8) , and fragments in the size range 200 to 500 bp (E8 digests), or 1 00 to 1 000 bp (Lp1 digests) were excised and purified from agarose (Section 2 .5 . 10) . Linkers were ligated to the blunt ends of the inserts to facilitate c loning, and the ligation reaction was set up on ice as follows: 100 to 500 ng of blunt end DNA fragments, 1 to 2 ug of phosphorylated l inkers (Section 2 . 1 2 .2) , 4 III of 5 x ligation buffer (Section 2 .4 .4), 0 .8 U of T4 DNA ligase (New England Biolabs), and MilliQ water to a total volume of 20 Ill . Ligation was canied out for 1 6 to 20 hr at 4°C and the ligase was inactivated by heating the reaction to 65°C for 1 5 min . Half microlitre aliquots of the ligation reaction were taken before addition of the ligase and after the ligation reaction and run on a mini gel (Section 2 .8 .2) to check for ligation. If fragments appeared ligated, 10 III of the appropliate 10 x restriction enzyme buffer, 20 to 50 U of restriction enzyme and Mil l iQ water to a total volume of 1 00 II I was added to the inactivated ligation reaction and incubated for at least 4 hr at 37°C. A fUl1her 1 0 U of restriction enzyme was added and the reaction incubated for another hour, then Na2EDTA (pH 8 .0) was added to a final concentration of 1 0 mM to stop the reaction. DNA was extracted with an equal volume of phenollchloroform ( 1 : 1 , v : v) and 58 centrifuged for 2 min at 1 5 ,000 x g and the aqueous phase extracted. The linkered DNA fragments were purified from the digested linkers by gel purification in SeaPlaque agarose gels in TAE buffer as outlined in Section 2 .5 . 10 . 2.12.4 LIGATION REACTIONS Ligation reactions were performed in 1 0 or 20 fl l volumes. The quantities of reagents for a 20 fl l reaction are as follows: 4 fl l of 5 x l igation buffer (Section 2 .4 .4), 20 ng of vector, 40 U of T4 DNA ligase, and a 2 to 5 molar excess of insert vector DNA. The ligation mixture was incubated for at least 1 6 hI' at 4°C. To check that l igation had occurred, two 2 fll samples of the ligation mixture were removed, one before adding the ligase and the other after ligation, mixed with SDS loading dye (Section 2.4.7) and separated on a mini gel (Section 2 . 8 .2) . 2.12.5 PREPARATION OF ELECTROCOMPETENT XL-1 CELLS One l itre of LB (Section 2 .3 . 1 ) was inoculated 1 1 100 with an overnight culture of XL- 1 cells (Section 2 .2 .4) and grown at 37°C with vigorous shaking to mid-log phase (A6oo of 0.5 to 1 .0, approximately 3 hr) . The cells were chilled on ice for 20 min and harvested by centrifugation at 4,000 x g for 1 0 min at 4°C. The cells were resuspended twice in ice­ cold water ( 1 litre, then 500 ml ,), fol lowed by resuspension in 20 ml ice-cold 1 0% (v/v) glycerol, centrifuging as above between steps. The cells were resuspended in 4 ml ice­ cold 1 0% (v/v) glycerol and stored in 40 fl l aliquots at -70°C. 2.12.6 TRANSFORMATION BY ELECTROPORA TION Aliquots, 40 fll , of electrocompetent cells (Section 2 . 1 2.5) were thawed gently at room temperature, placed on ice and 1 to 2 fl l of DNA to be transformed was added. The tube contents were gently mixed, and tubes were left on ice for 1 min. The Gene Pulser (Bio­ Rad) was set to 25 flF and 2.5 kV, and the pulse controller set to 200 ohms resistance in parallel with the sample chamber. The mixture of DNA and cells was transferred to an 59 ice cold 0 .2 cm electroporation cuvette and pulsed at the above settings . The time constant was checked and, if this was between 4 and 5, the cells were immediately resuspended in 1 ml of SOC medium (Section 2 .3 .3) . Transformed cells were incubated at 37°C for an hour to aid the recovery of the E. coli. A positive control of 2 ng pUC l 1 8, and a negative control of water only was always employed with each set of transformations. Cells were plated (Sections 2 .2 .4 and 2 .2 .5) at suitable dilutions onto selective LB plates (Section 2 .3 . 1 ) . 2.12.7 FILTER LIFTS AND PRIMARY LIBRARY SCREENING Recombinant M13mp19 phage were plated at approximately 1 0,000 plaques per plate to form the primary library. Filter lifts were taken from the library by marking nylon filters (Hybond N+, Amersham) asymmetlically and placing on the phage plates for 1 min . The marks were transferred to the plates for later alignment. Filters were placed on blotting paper to air dry and DNA was cross-linked by placing face down on a UV transilluminator and exposing for 2 min. To eliminate false positives, two l i fts were taken from each plate where the second lift was left on the plate for 2 min . Hyblidi sation was performed as described in Section 2 . 1 1 . Positive plaques, identified by alignment of the plate with positive hybridisation signals from the duplicate plaque lifts, were picked with a sterile cut-off 1 ml tip. The agar plug was ejected into 1 ml of LB medium (Section 2 .3 . 1 ) and stored at 4°C overnight to elute the phage from the agar. 2.12.8 SECOND ROUND LIBRARY SCREENING Phage from first round positive plaques (Section 2 . 1 2.7) were diluted to 1 0-6 to 1 0-8 in TE buffer ( 10: 1 ; Section 2.4. 1 2) and 100 III plated as previously described (Section 2.2 .5) . Plates containing between 30 and 300 plaques were lifted and hybridised as previously described (Section 2 . 1 2.7) . Positive plaques from the second round of library screening were picked with a toothpick and used to inoculate l iquid cultures (Section 2.2 .5) . 60 2.12.9 SCREENING POSITIVE CLONES Unique clones were identified by digesting plasmid DNA, isolated from positive plaques using a rapid boiling method (Section 2 .5 .5) , with BamHI (Section 2 .7) to excise the insert fragments. Digests were separated on mini gels (Section 2 . 8 .2) and unique clones were identified based on the size of their inserts. If many clones had inserts of the same size, they were analysed using a 'single tract sequencing' approach, where sequencing reactions were carried out using only a single terminating dideoxynucleotide. This was carried out as described in Section 2 . 1 5 . 1 but reactions contained the fol lowing quantities: 2 to 20 fmol of template DNA, 0 .8 II I 10 x cycling mix , 4 pmol of sequencing primer, 0.2 11 1 of [a_33p] dCTP (�2,500 Ci/mmol, Amersham), 2 11 1 termination mix and MilliQ water to bring the volume to 8 Il l . Unique clones were then ful ly sequenced (Section 2 . 1 5) . 2.13 POLYMERASE CHAIN REACTION AMPLIFICATION PCR reactions were set up in a dedicated 'clean' area that was free of PCR product and genomic DNA handling. Reactions were typical ly performed in 25 III volumes, though half and double volumes were sometimes used depending on the application. For n PCR reactions, a cocktail of n+ 1 reactions was prepared on ice using aerosol barrier tips to prevent contamination. The cocktail was dispensed into 0 .2 ml tubes on ice and template DNA added. Positive and negative (water) control samples were always included with each set of reactions, and reactions were amplified in a Corbett PC-960 or FTS-960, or a GeneAmp 2400 (Perkin-Elmer Corp.) model thermocycler. Reactions were checked on a mini gel (Section 2 .8 .2). 6 1 2.13.1 PRIMERS Oligonucleotide primers were manufactured by Life Technologies and rehydrated to a final concentration of 1 00 JlM in autoclaved MilliQ water, except RAPD primers, which were kindly provided by Ross Crowhurst, HortResearch, Auckland, New Zealand. Primers were diluted to a working concentration of 5 JlM and stored at -20°C. The sequences of all primers and probes used in this study are li sted in Table 2 .2 . Table 2.2. Primer and probe sequences Name A i A2 A3 A4 AS A6 A7 AS A9 AlO B1.1 B1.1-FAM B1.2 B2. I-HEX B2.1 B2.2 B3.1 B3.2 B4.1 B4.1-FAM B4.2 BS. I-TET BS.2 B6.1 B6.1-FAM B6.2 B7.1 B7.2 Sequence (S' - 3') (GTG)s (GACA)4 GAG GGT GGY GGY TCT (T A) 1 5 (TAA) IO (CA) lS (GA)ls (CAA) 10 (GAA) 10 (ATC) 10 CGC ACA ATA CGT CAG CTA GGA ATG fam - CGC ACA ATA CGT CAG CTA GGA ATG CCT GAA TCA ACT TTG CT A TCA GGC hex - TCC GT A ATG CTC TGC TTT GGC AGG TCC GT A ATG CTC TGC TTT GGC AGG CTG CTG ACA GT A GGT GGA TGA TGG TGA TGT CCG TTC rCG TGG CAT TCG CGG TGT AGA AGG ACC TGC AGT TTG TGG ACT CGA CTT GCC CTC TCT CAG fam - TGG ACT CGA CTT GCC CTC TCT CAG TGC GAG CAG CGT TTG CGT GTG CGT tet - CAA CTC CAA CAA ACT CAA CCA GCG GGA CTC GTG CAA AGC TTC GGA TGG GGC ATG GT A TGG GCA ATG AGT GTC fam - GGC ATG GT A TGG GCA ATG AGT GTC CAT CAT CGA TGT TTT GT A CTG TGG AAG TTG AAT T AA CT A GGG GCT AGG TT A ATT CGC T AC CCT CTC TTT TGC 62 Application * MP-PCR MP-PCR MP-PCR microsatellite probe microsatellite probe microsatellite probe microsatellite probe microsatellite probe microsatellite probe microsatellite probe ML-PCR, Bl ML-PCR, Bl ML-PCR, Bl ML-PCR, B2 ML-PCR, B2 ML-PCR, B2 ML-PCR, B3 ML-PCR, B3 ML-PCR, B4 ML-PCR, B4 ML-PCR, B4 ML-PCR, BS ML-PCR, BS ML-PCR, B6 ML-PCR, B6 ML-PCR, B6 ML-PCR, B7 ML-PCR, B7 63 Table 2.2. continued Name Sequence (5' - 3') Application . B8. 1 CT A ACA CGC TCA CGG GAG ATT GCC ML-PCR, B8 B8.2 GGC GAG GTG CAA CCC TGT T AA TGG ML-PCR, B8 B9. 1 AAT CGT TGT GCG AGC CAT TCT GGC ML-PCR, B9 B9. 1-TET tet - AAT CGT TGT GCG AGC CAT TCT GGC ML-PCR, B9 B9.2 TCC A TC TCC GCA ATC TGC ATG TCC ML-PCR, B9 B9A GCC CCG TCA TGC ATT ATC TCC TTG ML-PCR, B9 B lO. l CGC TCA GGG CT A CAT ACA CCA TGG ML-PCR, B lO B lO. I -TET tet - CGC TCA GGG CT A CAT ACA CCA TGG ML-PCR, B lO B lO.2 CTC ATC GAG T AA CGC AGG CGA CG ML-PCR, B lO B I U-HEX tet - CAT GGA TGG ACA AGA GAT TGC ACG ML-PCR, B l l B l l .2 CTG CT A CAA TTC TGT CAA GCT TCG ML-PCR, B l l B 1 1 04 TTC ACT GCT ACA ATT CTG TCC AGC ML-PCR, B l l GT02 TGG TGG GTC C RAPD-PCR RC05 AGG AGA TAC C RAPD-PCR RC08 GGA TGT CGA A RAPD-PCR RC09 GAT AAC GCA C RAPD-PCR ITS4 TCC TCC GCT TAT TGA TAT GC rDNA-ITS amplification ITS5 GGA AGT AAA AGT CGT AAC AAG G rDNA-ITS amplification M 13 fwd GCC AGG GTT TTC CCA GTC ACG A sequencing M l3 rev GAG CGG AT A ACA ATT TCA CAC AGG sequencing T l . l GAG AAA ATG CGT GAG ATT GT tub2 gene amplification T 1 .2 TGG TCA ACC AGC TCA GCA CC tub2 gene amplification T l . l-BamHl CGG GAT CCG AGA AAA TGC GTG AGA TTG T T l . l primer with BamHI site at 5' end for cloning T 1 .2-EcoR l CGG AAT TCT GGT CAA CCA GCT CAG CAC C T 1 .2 primer with EcoRI site at 5' end for cloning MP-PCR is micro/minisatellite primed PCR, ML-PCR is microsatellite locus PCR, and RAPD-PCR is randomly amplified polymorphic DNA PCR. 2.13.2 OPTIMISATION OF PCR REACTION CONDITIONS 64 In attempts to optimise PCR reaction conditions, annealing temperatures were varied between 50°C and 65°C, the Mg2+ concentration was varied between 1 and 5 mM, the polymerase concentration was varied between 5 and 80 mU/ill , and hotstart PCR reactions with AmpliWax beads (Perkin-Elmer Corp.) was tried. Products were evaluated on agarose mini gels (Section 2 .8 .2). 2.13.3 RANDOMLY AMPLIFIED POLYMORPHIC DNA PCR (RAPD-PCR) The amplification conditions for RAPD-PCR reactions were based on those of Williams et al . ( 1990) using modifications as described by Crowhurst et al . ( 1 99 1 ) . Reactions contained 1 x Taq polymerase buffer, 50 IlM of each dNTP, 0.04 U/1l 1 of Taq polymerase, 200 nM of primer (GT02, RC05, RC08, or RC09; Table 2 .2) and 40 pg/Il l of fungal genomic DNA. Reactions were amplified using the fol lowing programme: 95°C for 3 min, then 40 cycles of 95°C for 1 min, 35°C for 1 min, and 72°C for 3 min. 2.13.4 MICROIMINISATELLITE-PRIMED PCR (MP-PCR) Primers of microsatellite sequence, or the M13 core mini satellite sequence (primer A3) were used to amplify sections of the genome that were fl anked by the inverted plimer sequences, producing characteristic fingerprint patterns. This method was based on the technique described by Meyer et al. ( 1 993). MP-PCR reactions contained 1 x Taq polymerase buffer, 50 IlM of each dNTP, 0.04 U/1l 1 of Taq polymerase, 200 nM of primer (A I , A2, or A3 ; Table 2.2) and 40 pg/Il l of fungal genomic DNA, and reactions were amplified using the following programme: 40 cycles of 94°C for 1 min, 55°C for 1 min, and 72°C for 1 min, followed by a final extension at 72°C for 1 0 min. 2.13.5 MICROSATELLITE LOCUS PCR (ML-PCR) 65 Microsatellite loci were amplified usmg primers designed to the umque sequences flanking the tandem repeat region. ML-PCR reactions contained 1 x Taq polymerase buffer, 50 IlM of each dNTP, 0.04 U/1l 1 of Taq polymerase, 200 nM of each microsatellite locus PCR primer (Table 2 .2) and 40 pg/Ill of fungal genomic DNA, and were amplified using the fol lowing programme: 30 cycles of 94°C for 1 min, 65°C for 2 min, and 72°C for 1 min, fol lowed by a final extension at 72°C for 1 0 min . If products were to be radio-labelled for separation on PAGE gels (Section 2.9), [a_33p] dCTP was added to the reaction mixture to a final concentration of 0. 1 IlCi/lll. For automated analysis of microsatellite products (Section 2 . 1 7), primers that were 5 '-end labelled with the phosphoramidite dyes 6-carboxyfluorescein (FAM), 4, 7, 2', T-tetrachloro-6- carboxyfluorescein (TET) , or 4, 7, 2', 4', 5 ' , T-hexachloro-6-carboxyfluorescein (HEX) were used in the amplification. Specifically, primers B l . 1 , B4. 1 , and B6. 1 were labelled with FAM; B5 . l , B9. l , and B lO. l were labelled with TET; and B2. 1 and B l l . l were labelled with HEX. 2.13.6 MUL TIPLEX PCR Conditions for the multiplex PCR of microsatellite loci were the same as those used for a single locus amplification (Section 2 . l 3 .5) , but with additional primers added at a final concentration of 200 nM each. Microsatellite loci were grouped such that the size ranges of alleles from loci labelled with the same dyes did not overlap, thus allowing automated analysis of the reaction in a single run (Section 2 . 1 6) . Two sets of multiplex primers were established including Primer Set I, which contained primers B4. 1 -F AM, B4.2, B6. l -FAM, B6.2, B9. 1 -TET, B9.4, B lO. l -TET, B l O.2, B l l . l -HEX, and B l l .4, and Primer Set Il, which contained plimers B l . I -FAM, B l .2, B2. 1 -HEX, B2.2, B5 . 1 -TET, B5.2 , B6. 1 -FAM, B6.2, B9. l -TET, and B9.2. Where amplifications were poor, a problem particularly noticeable for in planta amplification, the five locus multiplex reactions were partitioned into separate reactions by either pooling loci of the same dye­ label, or dividing the reactions into their single locus constituents. 66 2.13.7 IN PLANTA PCR Amplification of fungal microsatel lites from the total genomIc DNA of endophyte infected grass ti ssue (Sections 2 .5 .3 and 2.5 .4) was carried out using the standard ML-PCR reactions conditions (Section 2 . 1 3 .5) , except the DNA concentration was increased to 2 ng/Il l . 2.13.8 6-TUBULIN GENE AMPLIFICATION The 5 ' portion of the B-tubulin gene, including IVS 1 , IVS2, and IVS3 , were PCR amplified using primers (Tl . 1 and T 1 .2 ; Table 2 .2) complementary to conserved coding sequences as previously described (Byrd et aI . , 1 990). Reactions contained the fol lowing: 1 x Taq polymerase buffer, 50 llM of each dNTP, 0.04 U/1l 1 of Taq polymerase, 200 nM of each primer and 40 pg/Il l of fungal genomic DNA. The thermocycler was programmed as follows: an initial denaturation at 94°C for 105 sec ; 35 cycles of 94°C for 45 sec, 60°C for 45 sec, and 72°C for 60 sec ; followed by a final extension at 72°C for 10 min. To faci litate the cloning of tub2 PCR products , pnmers were redesigned to include BamHI and EcoRI restriction enzyme recognition sequences at the 5 '-ends. These primers, T l . 1 -BamHI and T l .2-EcoRI (Table 2 .2) were used with the same amplification conditions as above. 2.13.9 rDNA-ITS AMPLIFICATION PCR amplification of the nuclear ribosomal DNA regIOn spanning the internal transcribed spacers (rDNA-ITS), and the 5 .8S rRNA gene was performed using primers ITS5 and ITS4 (Table 2.2), which are complementary to the 1 6S and 28S rRNA genes near the ITS1 and ITS2 borders respectively (White et aI . , 1 990). Initial reactions were performed in 25 III volumes containing 1 x Taq polymerase buffer, 50 IlM of each dNTP, 0.04 U/1l1 of Taq polymerase, 200 nM of each primer and 5 II I of 200 Pg/Ill fungal 67 genomic DNA. The thennocycler was programmed for an initial denaturation at 95°C for 3 min, then 30 cycles of 95°C for 30 sec, 55°C for 30 sec , and 72°C for 60 sec, followed by a final extension at 72°C for 10 min. Products were evaluated on agarose mini gels (Section 2.8 .2) and if the yield was too low to sequence, a second round of amplification was performed using similar conditions to those previously described with the fol lowing modifications . For increased fidelity of DNA synthesis , Pwo DNA polymerase (Boehringer Mannheim GmbH) was used instead of Taq, and the buffer system changed accordingly. Initial amplification products were diluted 1 / 10 to 111000 for use as templates and the number of cycles in the thermocycling programme was reduced to twenty. 2.14 CLONING OF PCR PRODUCTS 2.14.1 pGEM-T CLONING PCR amplification by Taq polymerase often results in the addition of single deoxyadenosine residues to the 3 ' ends of the amplicon, therefore the pGEM-T vector (Promega Corp . ; Section 2 . 1 .3), which contains 3' thymidine overhangs at the insertion site, was be used to clone such fragments. peR products were purified by standard phenollchlorofonn and ethanol precipitation procedures, or the PCR clean up kit (Section 2 .5 .9) . Ligation reactions were carried out in 1 0 p I volumes containing 25 ng of pOEM-T vector, 6 ng of PCR product, 1 x T4 DNA ligase buffer, 3 U of T4 DNA ligase, and MilliQ water, and were incubated and evaluated under the conditions described in Section 2 . 1 2.4. Recombinant products were transfonned into E. coli XL- 1 cells by electroporation (Section 2 . 1 2 .6) and plated on to LB plates (Section 2 .3 . 1 ) supplemented with ampicillin, IPTO, and X-gal (Section 2 .3 .6). White colonies were screened for unique c lones as described in Section 2 . 1 2 .9, though in this case Sal! was used to excise the insert fragments. Unique clones were ful ly sequenced using both M l 3 fwd and M13 rev primers that are complementary to sequences flanking the insertion site of the vector. 2.14.2 PREPARATION OF INSERTS FOR DIRECTIONAL CLONING 68 PCR tub2 products amplified with T l . I -BamHI and T 1 .2-EcoRI (Section 2 . 1 3 .8) were extracted with an equal volume of phenol , fol lowed by an equal volume of chloroform to inactivate Taq polymerase, and digested with BamHI and EcoRI in a 100 II I volume (Section 2.7) . Products were then purified from the excised fragment ends using the PCR clean up kit (Section 2 .5 .9) and the concentrations were estimated using the mini gel method (Section 2 .8 .2). 2.14.3 DIRECTIONAL CLONING USING pUC118 The pUC 1 1 8 vector (Section 2 . 1 .3) was prepared for directional BamHI and EcoRI cloning by restriction endonuclease digestion with these two enzymes (Section 2.7) . The vector was purified on a SeaPlaque mini gel in T AE buffer (Section 2 . 8 .2), extracted from agarose (Section 2 .5 . 10) , and resuspended at a final concentration of 1 0 ng/Il l . The prepared PCR insert fragments (Section 2 . 14.2) were ligated to the vector (Section 2 . 1 2 .4), transformed into E. coli XL- l cel ls by electroporation (Section 2 . 1 2.6) and plated on to LB plates (Section 2 .3 . 1 ) supplemented with ampicillin, IPTO, and X-gal (Section 2 .3 .6). White colonies were screened for unique clones as described in Section 2 . 1 2.9 , and by BamHIIEcoRI double digests to confirm the presence of the insert. Unique clones were ful ly sequenced using both the T 1 . 1 -BamHI and T 1 .2-EcoRI pnmers. 69 2.15 DNA SEQUENCING 2.15.1 AMPLICYCLE SEQUENCING KIT Purified DNA templates suitable for sequencing were sequenced using the AmpliCycle sequencing kit (Perkin-Elmer Corp.) . A reaction was set up on ice containing 10 to 100 fmol of template DNA, 4 II I 1 0 x cycling mix , 20 pmol of sequencing plimer, 1 III of [a_33p] dCTP (-2,500 Ci/mmol, Amersham), and MiIliQ water to make a total volume of 30 Ill . Six microlitre aliquots of this mixture was added to 2 III each of the four dideoxy termination mixes in 0.2 ml microamp tubes and placed in a Corbett thermocycler set to 94°C. The following programme was used: 94°C for 60 sec, fol lowed by 25 cyc les of 94°C for 30 sec, 50°C for 30 sec, and 72°C for 60 sec . Stop solution, 4 Ill , was added to each reaction after thermocycling, and reactions were stored at -20°C. Reactions were separated on polyacrylamide gels as described in Section 2 .9 . 2.15.2 DYE TERMINATOR CYCLE SEQUENCING DNA sequences to be analysed on an ABI Prism 377 DNA Sequencer or ABI Prism 3 10 Genetic Analyser (Section 2. 1 7) were sequenced using the ABI Prism Dye, ABI Prism dRhodamine, or ABI Prism BigDye terminator cyc le sequencing ready reaction kits (PE Applied Biosystems). Cycle sequencing reactions were set up on ice and contained: 8 .0 II I of terminator ready reaction mix , 3 .2 pmol of primer, 50 to 200 fmol of purified DNA template, and MilliQ water to a total volume of 20 Il l . Reactions were placed in a thermocycler set to 96°C. The fol lowing programme was used: 96°C for 1 0 sec , 50°C for 5 sec, and 60°C for 4 min for 25 cycles, with rapid thermal ramping between steps. Amplification products were precipitated by addition of 2 .0 II I of 3 M sodium acetate (pH 4.6) and 50 III of 95% ethanol, mixing, and chilling on ice for 1 0 min. Tubes were centrifuged for 30 min at 1 5 ,000 x g and the supematant discarded. The pellet was rinsed in 70% ethanol , briefly vOltexed, and centrifuged for 1 5 min at 1 5 ,000 x g. The supematant was discarded and the pellet air dried at 37°C for 30 min . Samples to be run on an ABI Prism 377 DNA Sequencer were resuspended in 4 III of formamide dye (83% 70 deionised formamide and 17% of 25 mM Na2EDT A (pH 8 .0) and 50 mg/ml of blue dextran) and stored at -20°C, and samples to be run on an ABI Prism 3 1 0 Genetic Analyser were resuspended in 1 2 to 25 fl l of template suppression reagent (TSR; PE Applied Biosystems) and stored at 4°C until run (Section 2 . 17) . 2.15.3 SEQUENCE ANALYSIS DNA sequence data was analysed with the Wisconsin Package version 9 . 1 , Genetics Computer Group (GCG), Madison, Wisc. The programs used were SEQED, GETSEQ, GAP, and PILEUP (Section 2 . 1 8 . 1 ) . Basic local alignment search tool (BLAST) searches were performed using the Basic BLAST search option of BLAST 2.0 on the internet at the National Center for Biotechnology Information (NCBI) website (http://www.ncbi .nih.gov). DNA sequences were deposited into GenBank using the BankIt option on the NCBI website. 2.16 MICROSATELLITE ANALYSIS Microsatellite locus PCR products (Section 2 . l 3 .5) were analysed in several ways, depending on the requirement, and the stage of development of the microsatellite-based fingerprinting assay. To check the presence and concentration of PCR products , and to estimate the degree of polymorphism at a microsatellite locus, products were run on agarose mini gels (Section 2 .8 .2). High resolution separation of products was achieved by radio-labelling products (Section 2. 1 3 .5 ) followed by PAGE separation and autoradiography (Section 2.9). The control sequence from the AmpliCycle sequencing kit (Section 2 . 1 5 . 1 ) was loaded at regular intervals across the gel , and used as a DNA size marker. The sizes of microsatellite PCR products were preCisely detelmined from the automated analysis of dye-labelled fragments (Section 2 . 1 3 .5) using ABI Prism 377 or ABI Prism 3 10 analysers (PE Applied Biosystems; Section 2 . 1 7). 7 1 2.17 AUTOMATED GENETIC ANALYSIS Automated analysi s of dye-labelled sequencmg reactions (Section 2 . 1 S . 1 ) and dye­ label led microsatellite locus PCR products (Section 2 . 1 3 .S) were pelformed on an ABI Prism 377 DNA Sequencer or an ABI Prism 3 10 Genetic Analyser (PE Applied Biosystems). The ABI Plism 377 is a slab gel based system where samples were separated on a 4.2S% denaturing polyacrylamide gel . In contrast, the ABI Plism 3 10 is based on capillary electrophoresis in a performance optimised polymer (POP) matlix where samples were loaded by autoinjection and individually run. Sequencing reactions prepared for analysis on an ABI Prism 377 were resuspended in formamide dye (Section 2 . 1 5 .2), denatured at 95°C for 2 min, quenched on ice, and 0.7S to 1 .S II I were loaded onto the gel . Sequencing reactions to be analysed on an ABI Prism 3 1 0 were resuspended in TSR (Section 2. 1 S .2), denatured at 9SoC for 2 min, quenched on ice and autoinjected into a 6 1 cm capi llary with POP-6 polymer (PE Applied Biosystems). Genetic analysers were run according to the manufacturers standard protocols and data was analysed using DNA Sequencing Analysis Software Version 3.3 . Dye-labelled microsatelli te products ( l .5 Il l) to be analysed on an ABI Pri sm 377, were mixed with 2 .S II I of formamide, O.S II I of 5% blue dextran , and 0 .5 II I of OS-500 TAMRA (Perkin-Elmer Corp.), an internal size standard, denatured at 9SoC for 2 min, quenched on ice and 2 III loaded onto the gel . If microsatel1 ite products were to be analysed on an ABI Prism 3 10, 1 II I of diluted products was mixed with 12 . 5 II I of deionised formamide and 0.5 fl l OS-SOO TAMRA and samples were denatured for 5 min at 9SoC, then quenched on ice. Samples were run through a 40 cm capillary with POP-4 polymer (PE Applied Biosystems). Oenetic analysers were run according to the manufacturers standard protocols and data was analysed using GeneScan Analysis 2 . 1 software. 72 2.18 PHYLOGENETIC ANALYSIS 2.18.1 SEQUENCE ALIGNMENT DNA sequences were aligned using the PILEUP program of the Wisconsin Package Version 9. 1 (Section 2 . 1 5 .3 ) and parameters were empiricall y adjusted to a gap penalty of one and gap extension penalty of zero, as described by Schardl and Leuchtmann ( 1999). Alignments were checked by eye for ambiguities and adjusted if necessary. 2.18.2 PHYLOGENETIC ANALYSIS BY MAXIMUM PARSIMONY Maximum parsimony (MP) trees were inferred from aligned sequences (Section 2 . 1 8 . 1 ) using a branch-and-bound search algorithm for unordered and unweighted characters as implemented in PAUP* Version 4.64 (Swofford, 1998). If the computing time of the branch-and-bound search exceeded 48 hours, the heUl1stic search algorithm was used instead. Alignment gaps were treated as missing information, and all missing and unambiguous characters were excluded from analysis . Tree roots were inferred by mid­ point rooting. One hundred bootstrap trees were generated under the maximum parsimony criteria, using the heuristic search option. Bootstrap values for the majority rule consensus tree have been shown. 2.18.3 PHYLOGENETIC ANALYSIS BY NEIGHBOR-JOINING Neighbor-joining (NJ) trees were generated using the PAUP* Version 4.64 software package. Alignment gaps were treated as missing information, and all missing and unambiguous characters were excluded from analysis. Since the calculation of genetic distances between sequences should only be made on sites in the data that are free to vary, an estimate of the proportion of unvaried character sites in the data was made using the maximum likelihood approach (Lockhart et al. , 1 996). Trees were generated using both Jukes-Cantor (one-parameter) and Kimura two-parameter distance mat11ces, with the estimated proportion of unval1ed sites removed in proportion to the empirical base 73 frequencies estimated from the constant sites. Bootstrap consensus trees of 100 replicates were performed using the NJ option. CHAPTER THREE DEVELOPMENT AND APPLICATION OF AN IN PLANTA EPICHLOE ENDOPHYTE FINGERPRINTING ASSAY 3.1 INTRODUCTION 75 As a group, the Epichloe endophytes show considerable variation in morphology and physiology in culture (Christensen and Latch, 199 1 ; Christensen et aI . , 1 99 1 ; Christensen et aI. , 1 993). While these properties are useful in the identification of endophytes, used alone they are of limited value in defining distinct taxonomic groupings. However, when used in combination with molecular methods such as isozyme (Leuchtmann and Clay, 1 990) and gene sequence (Schardl et aI . , 1 99 1 ) analysis , distinct taxonomic groupings can be defined and the phylogeny established (Christensen et al., 1 993 ; Schardl et aI . , 1 994; Tsai e t al . , 1 994). Both isozyme and DNA sequence analysis are lengthy processes for routine strain identification and require that the endophyte be first isolated from the grass host, a process which can take up to several weeks. Several methods are available for the in planta detection of endophytes. These include histological staining (Latch et al . , 1984) , enzyme-linked immunosorbent assay (ELISA) (Reddick and Collins, 1988) , and tissue print-immunoblot (Gwinn et aI . , 1 99 1 ) . More recently, PCR based methods, including both RAPD and microsatellite locus analysis , have been used for endophyte detection and quantification both in culture (Groppe e t aI . , 1 995) and i n planta (Doss and WeIty, 1 995 ; Groppe and BoIler, 1 997). PCR methods provide fast, sensitive and specific amplification of target DNA sequences in complex DNA samples, and thus are particularly amenable to endophyte detection in the plant background. The objective of this study was to develop a PCR-based fingerprinting assay for the rapid and accurate identification of Epichloe endophytes in planta. Initiall y a pilot study was carried out to evaluate several PCR-based techniques for their ability to distinguish between endophyte isolates, and for their ease of use. The most promising technique was further developed into a working assay and used in various applications to identify endophyte i solates . 3.2 RESULTS 3.2.1 PILOT STUDY TO EVALUATE PCR-BASED FINGERPRINTING METHODS 76 A pilot study was conducted to evaluate three PCR-based fingerprinting methods, RAPD-PCR (Section 2. 1 3 .3), MP-PCR (Section 2. 1 3.4) and ML-PCR (Section 2. 1 3 .5) , for their ease of use and ability to discriminate between isolates. A panel of twelve Epichloe endophyte i solates (Section 2 . 1 . 1 ) , representing LpTG-2 (Lp l ), Neotyphod£um lolii (LpS, Lp7, Lp 13 , and Lp 1 9), FaTG-2 (Tfl 3), FaTG-3 (Tfl 8), N. coenophialum (Tf27 and Tf28), N. uncinatum (Fp2), Epichloe typhina (E8), and E. Jestucae (Fr 1 ), were typed using each of the fingerprinting methods and comparisons of the methods were made. 3.2.1.1 RAPD-PCR DNA samples from the panel of twelve Epichloe endophyte isolates (Section 3 .2 . 1 ) were amplified by RAPD-PCR (Section 2 . 1 3 .3) using primers RCOS, RC08 , and RC09 (Table 2 .2) and separated on 1 .75% agarose gels . These results are shown in Fig. 3 . l . Overall , the fingerprints had very few diagnostic bands in order to identify endophyte i solates, even at the species or taxonomic group level . The non-amplification of samples, such as in lanes 2 , 3, and 1 1 of Fig. 3 . 1B , was frequently observed and was inconsistent between repeated runs. Simi larly, fingerprint patterns were not consistently reproducible between runs , and contamination of the PCR negative control was occasionally ex peri enced. 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 M A 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 M B 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 M c Fig. 3 . l . RAPD-peR profiles from a range of Epichloe endophyte i solates - 1 .6 kb - 1 .0 kb - 0.5 kb 3.0 kb 2.0 kb 1 .6 kb 1 .0 kb 77 RAPD profiles following amplification with primers Re05 (panel A), Re08 (panel B), and Re09 (panel e), and separated on l .75% agarose gels . Reactions contained genomic DNA from isolates Lp1 (lane 1 ) , Lp5 (lane 2), Lp7 (lane 3) , Lpl 3 (lane 4), Lp 1 9 (lane 5), Tfl3 (lane 6), Tfl 8 (lane 7), Tf27 (lane 8) , Tf28 (lane 9) , Fp2 (lane 1 0) , E8 (lane 1 1 ) , and Frl (lane 1 2) . Lane 1 3 is a peR negative control without DNA and lane M contains 1 kb ladder. 78 3.2.1 .2 MP-PCR Fingerprint patterns from the panel of twelve endophyte i solates (Section 3 .2 . 1 ), generated by MP-PCR (Section 2 . 1 3 .4) using primers A I , A2, and A3 (Table 2 .2), are shown in Fig. 3 .2 . Fingerprints amplified using primer A l (Fig. 3 .2A) generated patterns of 2 to 9 bands between 0 .5 and 1 .5 kb in size. High background smearing, and the tight clustering of bands made these fingerprints difficult to analyse. In contrast, amplification using primers A2 and A3 produced clear banding patterns of 7 to 20 bands in the size range 0.2 to 3 .5 kb. Species and taxonomic groups of Epichloe endophytes were able to be clearly distinguished, and intraspecific differences were detectable within N. lolii and N. coenophialum. As MP-PCR fingerprinting with the A2 and A3 primers looked promising for strain differentiation, a more extensive range of endophyte isolates were fingerprinted using the A2 (Fig. 3 .3) and the A3 (Fig. 3 .4) primers. Common banding patterns were clearly identified for each endophyte species, or taxonomic group, and minor intraspecific differences were detected as observed previously. The reproducibility of fingerprints generated using this method was good. 3.2.1.3 ML-PCR Primers B 1 . 1 and B 1 .2 (Table 2 .2), which amplify an (AAG)n microsatellite (Groppe et a1 . , 1 995), were used in ML-PCR amplification (Section 2 . 1 3 .5) of DNA samples from the panel of twelve endophyte isolates (Section 3 .2 . 1 ) . Amplification products, separated on a 3% NuSieve agarose gel (Fig 3 .5A), were obtained for all i solates except Tf1 8 and E8, and were of the expected fragment size of around 300 bp (Groppe et aI . , 1 995). As the variation in fragment length was expected to be as li ttle as 3 bp, products were radiolabelled with [a_33p] dCTP (Section 2. 1 3 . 15 ) and separated on a denaturing polyacrylamide gel fol lowed by autoradiography (Section 2 .9 ; Fig. 3 . 5B) . Alleles were detected from all samples, and multiple bands were detected from Lpl , Tf1 3 , Tf27 , Tf28, and Fp2, which are endophytes known to have interspecific hybrid origins . Each allele occurred as a doublet of two bands, and this was presumed to be each labelled strand of the product which had separated under the denaturing conditions. 79 Fig. 3 .2 . MP-peR profi les from a range of Epichloe endophyte i solates MP-peR profi les fol lowing amplification with primers A l (panel A), A2 (panel B) , and A3 (panel e), and separated on 1 .75% agarose gel s. Reactions contained genomic DNA from isolates Lpl (lane 1 ) , Lp5 (lane 2), Lp7 (lane 3), Lp1 3 (lane 4), LpI 9 (lane 5) , Tf13 (lane 6), Tfl 8 (lane 7), Tf27 (lane 8) , Tf28 (lane 9), Fp2 (lane 1 0) , E8 (lane 1 1 ) , and Frl (lane 1 2). Lane 1 3 is a peR negative control without DNA and lane M contains 1 kb ladder. 80 M 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 M A - 1 .6 kb - 1 .0 kb - 0.5 kb M 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 M B M 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 M c 3.0 kb - 2.0 kb - 1 .6 kb - M 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 M 1 .0 kb _ M 19 20 2 1 22 23 24 25 26 27 28 29 30 31 32 33 34 35 M 3.0 kb - 2.0 kb - 1 .6 kb - 1 .0 kb - Fig. 3 .3 . MP-PCR profiles of isolates using the A2 primer 8 1 Lanes (bold) contain PCR reactions with genomic DNA from isolates: 1 , Lpl ; 2, Lp2; 3, Lp5 ; 4, Lp6; 5, Lp13 ; 6, Lp19 ; 7, Lp20; 8, Lp21 ; 9, Lp7; 10, Lp14 ; 1 1 , ARl ; 12, ARIWF; 13, ARICY; 14, Tf5 ; 15, Tf27 ; 16, Tf28; 17, Tfl3 ; 18, Tfl 5 ; 19, Tf20; 20, Tfl6 ; 21 , Tfl8 ; 22, Fpl ; 23, Fp2 ; 24, Fp4; 25, E8; 26, Frl ; 27, Frc5 ; 28, Frc7 ; 29, FIT1 ; 30, Fgl ; 31, Fl 1 ; 32, HaB; 33, Hdl ; and 34, EndoSafe. Lane 35 contains PCR negative control , and M contains 1 kb ladder. MP-PCR profiles were resolved on 1 .25% agarose gels . 3.0 kb - 2.0 kb - 1 .6 kb - 1 .0 kb - 0.5 kb - 2.0 kb 1 .6 kb 1 .0 kb 0.5 kb M 1 2 3 4 5 6 7 8 9 1 0 1 1 12 1 3 14 15 16 1 7 18 M M 1 9 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 M Fig. 3 .4 . MP-PCR profiles of isolates using the A3 primer 82 Lanes (bold) contain PCR reactions with genomic DNA from i solates : 1 , Lpl ; 2, Lp2; 3, Lp5 ; 4, Lp6; 5, Lp13 ; 6, Lpl 9; 7, Lp20; 8, Lp2 1 ; 9, Lp7 ; 10, Lpl4 ; 11 , ARl ; 12, ARl WF; 13, ARlCY; 14, Tf5 ; 15 , Tf27; 16 , Tf28 ; 17, Tfl3 ; 18 , Tfl5 ; 19, Tf20; 20, Tfl6 ; 21, Tfl 8 ; 22, Fp1 ; 23, Fp2; 24, Fp4; 25, E8; 26, Frl ; 27, Frc5 ; 28, Frc7; 29, FITl ; 30, Fgl ; 31, Fl l ; 32, HaB; 33, Hd1 ; and 34, EndoSafe. Lane 35 is a PCR negative control without DNA and M contains 1 kb ladder. MP-PCR profiles were resolved on l .25% agarose gels . 83 Fig. 3 .5 . ML-PCR amplification of a range of Epichloe endophyte isolates Products were amplified with primers B 1 . 1 and B 1 .2 , and resolved on a 3% NuSieve agarose gel (panel A), and a denaturing 6% polyacrylamide gel (panel B). Reactions contained genomic DNA from isolates Lpl (lane 1 ), Lp5 (lane 2), Lp7 (lane 3), Lp1 3 (lane 4), Lp1 9 (lane 5), Tf1 3 (lane 6), Tf1 8 (lane 7), Tf27 (lane 8) , Tf28 (lane 9), Fp2 (lane 10) , E8 (lane 1 1 ) , and Frl (lane 1 2) . On panel A, lane 1 3 is a PCR negative control without DNA and lane M contains 1 00 bp ladder. A B 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 M 2 3 4 5 6 7 8 9 1 0 1 1 1 2 84 - 400 bp - 300 bp - 200 bp 294 bp 85 3.2.1 .4 Evaluation of Fingerprinting Methods On the basis of the above results, ML-PCR was chosen over RAPD and MP-PCR for further development of an endophyte identification assay. This is further discussed in Section 3 .3 . 1 . 3.2.2 DEVELOPMENT OF A PCR-BASED MICROSATELLITE FINGERPRINTING ASSAY 3.2.2.1 Detection of Microsatellites within Epichloe Endophyte Genomes To determine the presence and relative abundance of microsatellites in the genomes of endophyte i solates E8 and Lpl (Table 2 . 1 ), genomic DNA was digested to completion with combinations of the 4 bp blunt-end cutting restriction enzymes AluI, Haem, and ThaI, as well as BamHI (Section 2.7), and Southern blots (Section 2 . 1 0) of these digests were probed for the presence of microsatellite sequences (Section 2 . 1 1 ) . These digests resulted in large populations of small DNA fragments of a size suitable for the construction of small insert libraries (Fig. 3 .6A). Hybridisation of Lp1 digests with DIG-labelled microsatellite probes, A6 to A lO (Table 2.2) , under conditions of low stringency (Section 2 . 1 1 ) revealed up to 25 bands for dinucleotide repeat probes (Fig. 3 .6B), and up to 20 bands for the trinucleotide repeat probes (Fig. 3 .6C). This represents an average frequency of one (CA)1l or (GA)1l microsatellite per 2 Mb of DNA, or one (CAA)Il' (GAA)Il' or (ATG)1l microsatellite per 2 .5 Mb of DNA, assuming a genome size of 50 Mb for Lp1 (Murray et al. , 1 992). Genomic digests of strain E8 were also screened with di- and trinucleotide repeat probes and similar results were obtained (results not shown). A 23.1 - 6.6 - 4.4 - 2.3 - 2.0 - M 1 2 3 4 5 6 7 8 M B 23.1 6.6 - M 1 2 3 4 5 6 7 8 M Fig. 3 .6 . Hybridisation of microsatellite probes to the Lpl genome c 23.1 - 6.6 - 4.4 - 2.3 - 2.0 - M 1 2 3 4 5 6 7 8 M (A) Agarose ( 1 .25%) gel of Lpl genomic DNA ( 1 !J.g1lane) digested with AluI ( lane 1 ); ThaI (lane 2); Haem (lane 3) ; BamHI (lane 4); AluI, ThaI, and BamHI (lane 5) ; AluI, HaeIII, and BamHI ( lane 6) ; ThaI, HaellI, and BamHI (lane 7); and AluI, ThaI, Haem, and BamHI (lane 8) . Lane M contains HindUI digested lambda DNA, and band sizes are given in kb. (B) and (C) show autoradiographs of a Southern blot of the gel of panel A that has been probed at low stringency with pools of DIG-Iabeled oligonucleotides (CA)15 and (GA)15 (panel B), and (CAA)IO, (GAA)IO, and (ATC)lO (panel C) , followed by chemiluminescent detection. 3.2.2.2 Library Construction and Screening S7 Partial genomic librmies were constructed in the vector M 13mp 1 9 using size-fractionated DNA inserts from strains ES and Lp1 . Genomic DNA was digested to completion with the same combinations of 4-bp cutting restriction enzymes and BamIll as described previously (Section 3 .2 .2 . 1 ) , and fragments were separated on 1 .25% SeaPlaque agarose gels in TAE buffer (Section 2 .S) . Fragments of 200 to 500 bp from ES, and 100 to 1000 bp from Lp l , were excised and purified from the agarose (Section 2 .5 . 1 0). Insert DNA was prepared using BamHI linkers (Section 2 . 1 2.3), and inserts were ligated to CAP-treated BamHI digested M13mp19 (Section 2 . 12 . 1 ) . Libraries were transformed into E. coli strain XL- 1 Blue (Section 2 . 1 2.6), ti tred, then plated at a density of approximately l O,OOO white plaques per plate. 3.2.2.2.1 E8 Library Approximately 1 lO,000 plaques of the ES library were screened by plaque hybridisation with a mixture of DIG-labelled dinucleotide probes, at a low stringency. Under these conditions, 1 % of the plaques hybridised and 5 1 of these were taken through to the second round of screening. When the same fi lters were screened with a mixture of DIG­ labelled trinucleotide probes under conditions of high stlingency, 0. 1 % of the plaques hybridised and 30 of these were taken through a second round of screening. Second round screening confirmed 35 plaques identified with the dinucleotide probe mixture as positive, and 22 with the trinucleotide probe mixture. 3.2.2.2.2 Lp1 Library Screening of 1 20,000 plaques of the Lp! library was carried out under conditions of low stringency for both pools of di- and tri- nucleotide probes. Using these conditions 0.4% of the plaques hybridised with each set of probes, and 24 and 1 2 plaques respectively were taken through to second round of screening. Second round screening confirmed 22 88 plaques identified with the dinucleotide probe mixture as positive, and nine with the trinucleotide probe mixture. 3.2.2.3 Analysis of Positive Clones The second round positive plaques were selected and amplified in E. coli and both single and double-stranded M13 DNA was i solated from each clone (Section 2 .5). The double­ stranded DNA was digested with BamHI and separated by agarose gel electrophoresis to determine the insert size. Many of the inserts appeared to be the same size, and therefore were possibly clones of one another. This was confirmed by single dideoxy/deoxynucleotide sequencing reactions of single-stranded DNA templates (Section 2 . 1 2.9). Unique clones identified by this method were then completely sequenced using the M13 forward primer (Section 2. 1 5) . Ten different microsatellite types were identified; six from E8 (B2 to B6, and B 12) and four from Lp l (B7 to B lO). The microsatellite sequences identified are listed in Fig. 3 .7 , and with the exception of B 12, have been deposited into GenBank with accession numbers AF063085 to AF063094. Of these microsatellites, five (B2, B3, B5 , B8 , and B 1 0) contained perfect repeats, and the remaining loci (B4, B6, B7, B9, and B 12) contained imperfect repeats . Additional single nucleotide microsatellite loci were identified in c lones B3 , B8 , and B9. For B8, the single nucleotide locus was immediately adjacent to the dinucleotide repeat locus as a compound microsatellite. For B5 and B9 the repeat motifs identified were only partiall y complementary to the probes used (tri- and di- nucleotide mixtures respectively), suggesting that they were identified as positives because of the low sttingency hybridisation conditions employed in the screening process. 89 Fig. 3.7. DNA sequences of the microsatellite loci used in this study The DNA sequences of the microsatellite loci used in this study are given below with microsatel lite sequences highlighted i n green, and primer binding sites indicated i n red. For loci which were isolated by library screening, sequence data from the entire insert, excluding l inkers, is given. Microsatellite locus B 1 (370 bp) isolated from Epichloe bromicola by Groppe et al . ( 1 995). GenBank accession number X9 1 79 1 . 1 AATGCCGCAGCACTTTCCCAAGCCCGCACAATACGTCAGCTAGGAATGTTGCATACTTGA B1 . 1 => 6 1 TTTCCATCCATCTGGCTCTAGACCTGATGACAGATTCCATGGTCGAACAAACGCGATCGC 1 2 1 CCCAAGCTCGTTTTATGTCGACGAAAAGAAGAAAAGAAACTATGCAACGTCTGAAGAGTT 1 8 1 GAAGAAGAAGAAGAAGAAGAAGAAGAAGAAGAAGAAGAAGAAGAAGAAGAAGAAAGA 2 4 1 GACGGCCCAACATCAAGATTCATTCGTCAACCTATAGCCCTGGTCCAGACCAAATGGGTT 3 0 1 CTCTCTCATATACCTACTAACATCCGCCTGATAGCAAAGTTGATTCAGGTAATCATGGAC <= B 1 .2 3 6 1 CTGCGGCATT Microsatel lite locus B2 clone (2 1 5 bp) isolated from E8 l ibrary with dinucleotide probe mixture. GenBank accession number AF063085. 1 TGGCAATAATACATAATGTAATCCGTAATGCTCTGCTTTGGCAGGATGCGAATCGTCTTG B2. 1 => 6 1 CCTGAAGCTGCCTTCTGCGGACCGTTGCTCCCTACATATTACGTGCACACACACACACAC 1 2 1 ACACACACACACACACACACACGTAATATGTCAGCATCCACCATCATCCACCTACTGTCA <= B2.2 1 8 1 GCAGCCAGAAGATAAGATAGACGCTGCTCATTGCG 90 Microsatellite locus B3 clone ( 1 84 bp) i solated from E8 library with dinucleotide probe mixture. GenBank accession number AF063086. 1 CGGTTCGGTACGCCCCCCCCCCCCGGTCTTTTTCTGCCTCGCTTGCCAAACCCAAACATG 6 1 TTGAAAATATGCTGTCCGTTCTCGTGGCATTCGGTCGGGGGGTCTTTTTTCGAGAGAGAG 83. 1 => 1 2 1 AGAGAGCCGGAACGAGCCGGGGATACACGATTCTACAAATCAAACTGCAGGTCCTTCTAC <= 83.2 1 8 1 ACCG Microsatellite locus B4 clone ( 1 27 bp) isolated from E8 library with dinucleotide probe mixture. GenBank accession number AF063087. 1 GTGGACTCGACTTGCCCTCTCTCAGCCTCAGCAATCATTGTTGCCATCATTCGAACTCAC 84. 1 => 6 1 AAAACGCACA AGCGCACAC CGCACAC CACACACAAAACGCACACGCAAACGCTGC 1 2 1 TCGCACG <= 84 .2 M icrosatellite locus B5 clone (22 1 bp) isolated from E8 library with trinucleotide probe mixture. GenBank accession number AF063088. 1 TTCCAACTCCAACAAACTCAACCAGCGCATCTCATCACATCTCATCACATCTCATCACAT 85. 1 => 6 1 CTCATCACATCTCATCACAATGCTTTCCATCCGAAGCTTTGCACGAGTCCGCACCCCGGG <= 85.2 1 2 1 CTGTATCTCGCTCCACCGCTTCAATGCGATCTAGCGTCGCCGCGCCGGCACCCTCATCAA 1 8 1 GACATCAGCCGTGAGCGCGCTGCGGCAATCCACCGGTGCAG 9 1 Microsatellite locus B6 clone (206 bp) isolated from E8 l ibrary with trinucleotide probe mixture. GenBank accession number AF063089. 1 GTGGAATATTCATCTGCATAATTTTTGGCATGGTATGGGCAATGAGTGTCTTTGAAATCT 86. 1 => 6 1 AGACGGCGCTGGCTGCGACTGCCGGTGTTGACCTGGCAAAGTATGCAATGACAAACGCAA 1 2 1 GATCCACAGCAACCCATCACATCAACCACCATCTCTTCATCATCACCATCATCATCCACA 1 8 1 GTACAAAACATCGCAGCAGTAAGAGG <= 86.2 Microsatellite locus B7 clone ( 1 7 8 bp) isolated from Lp 1 library with dinucleotide probe mixture. GenBank accession number AF063090. 1 CTATCGATCTGCATGATATTAATGATAATCTTGTAGTTAAATTCGTGGTTATCTTTGATT 6 1 ATAAACTTGAATTAACTAGGGGCTAGGAAAGAGAGAGAGAAGAGAGGGAAAGTCTAGATG 87. 1 => 1 2 1 CTAGGAAAGTAGGCGAAACCCGTATATAAGGCAAAAGAGAGGGTAGCGAATTAATATA <= 87.2 Microsatellite locus B 8 clone (235 bp) isolated from Lp l l ibrary with dinucleotide probe mixture. GenBank accession number AF06309 1 . 1 CTCTAACACGCTCAAGGGAGATTGCCTAAAGCATCCCATCCACAACGACACGGGCATTCA 88. 1 => 6 1 CACACACACACACACTTTTTTCATCGTCACTCCAACTCCCCGTCCACGCATCCTTCCCGC 1 2 1 CCATTAACAGGGTTGCACCTCGCCACCTTGCCCCCCCCCAGCTCAACGACGTACTTGCAC <= 88.2 1 8 1 CCCCCGTACCGCTCGCTCGCACCCTCGACAAAGAAGTCGCTCGACACAATCTGCG 92 Microsatellite locus B9 clone (50 1 bp) isolated from Lp l library with dinucleotide probe mixture. GenBank accession number AF063092. 1 CTTCATGGAAGCAAAAAAATGATTTCCTAGAATCAGAGTGGCTAGCAGGGTTACACACAA 6 1 TCCACAATCGTTGTGCGAGCCATTCTGGCATTCCAAGTCTACCCGACAGACTGGGGGAGG 89. 1 => 1 2 1 GGGGGGGTCAACAAGGGATGCATCTGTGACTCGCGTGCGTGAGAAGATTGATGCGAGAGG 1 8 1 AGAGCGAGGAGAGGACATGGACATGCAGATTGCGGAGATGGACGGGGTCAAGCTGATGTA <= 89.2 2 4 1 CTGTAACATGCATCACGCGAGCAACCACCCACGAGCTTTAACACGCTCAAGGAGATAATG <= 89.4 3 0 1 CATGACGGGGCGCCGAAGCAGCGCCATTGCGGGAGTGAATTGTGGCTGGTTGGCTGCATT 3 6 1 GTGCGGCATTGTCAGCCTGGCTTCTGTGAGCAGACACGTCTGATGGTGTTGTAAACATTC 4 2 1 CCCAGACCATCATTTGTATTTATCATTTCATGCAACATGCAACATGCAACATGCAACATG 4 8 1 CAACAACCACCTACAAGTTGG Microsatellite locus B 1 0 clone ( 1 80 bp) isolated from Lp 1 l ibrary with trinucleotide probe mixture. GenBank accession number AF063093 . 1 CCGCTCAGGGCTACATACACCATGGTCAACATCACCAGCAACAGCCGCAGCAGCCGCAAC 8 1 0. 1 => 6 1 AGCAGCAGCCGCATCATCATCAACAACAACAACAACCACGACAGCGACCCCAAAACTGGA 1 2 1 ACATGTATCCTGGCATGTACTATGACAGTAACCAAA CGTCGCCTGCGTTACTCGATGAG <= 81 0.2 93 Microsatellite locus B l l ( 1 79 bp) from Lp 1 9 HMG CoA 3' untranscribed sequence. GenBank accession number AF063094. The primer binding site for B 1 1 .4 is underlined. 1 TTCCATGGATGGACAAGAGATTGCACGAGTCAAGGCGCTGGTGTGAGGGACAGTTACAGA B1 1 . 1 => 6 1 TGGACAGACAGACAGACAGACAGACAGACAGACAGACAGACAGACAGACAGACAGACAGA 1 2 1 CAGACAGACAGACATGTACAGGAGATATCGAAGCTGGACAGAATTGTAGCAGTGAAACA <= B1 1 .2 <= B1 1 .4 Microsatel lite locus B 1 2 clone ( 1 7 1 bp) isolated from E8 l ibrary with dinucleotide probe mixture. 1 CCTGTGGCGGGATTTATTTCCTCTCGTCATCAGGTTGTTTGCTTTGTGCTTGGATTTCGG 6 1 CGGGCTTTGTTTGCTTTGCGAACCTGTAGAATGTTTGCTATCAGGTTGGGTAGTCGAGAA 1 2 1 TGAAGGGGCGGCGAGAACTGATGAGAGATGAGAGAGATGAGAGAGCGAACG 3.2.2.4 Identification of Additional Epichloe Endophyte Microsatellite Loci 94 Two additional endophyte microsatellite sequences have been used in this study and are included in Fig . 3 .7 . An (AAG)n microsatellite, referred to here as B 1 , was identified by Groppe et al . ( 1995) in an Epichloe bromicola isolate from Bromus erectus. A second locus, B l l , was serendipitously identified 3' of the untranscribed region of the 3- hydroxyl-3-methylglutaryl Coenzyme A (HMG CoA) reductase gene from N. lolii strain Lp 1 9 (Dobson, 1 997). 3.2.2.5 ML-PCR Assay Design and Evaluation 3.2.2.5.1 Primer Design and Preliminary Amplijications To establish PCR assays which amplify the microsatellite loci identified, primers were designed by eye to the sequences flanking each of the microsatel lites (Fig. 3 .7 and Table 2 .2). A particular consideration in the design of the primers was for all to have a similar length and GC content to facilitate potential multiplexing of all loci . Primer sequences to the B 1 locus were those used previously by Groppe et al . ( 1 995). Unfortunately primers could not be designed to locus B 1 2 as insufficient flanking sequence to one side of the microsatellite was isolated, therefore further analysis of this locus was abandoned. Initial evaluation of primer pairs (Bx . lIBx.2 for each locus, Bx) was performed by screening a panel of twelve isolates, representative of seven taxonomic groups of endophytes, using a standard set of conditions (Figs. 3 . 8 and 3 .9). All loci , apart from, B3 , B7 and B l l gave specific products with high yield using these primers. No products were amplified from the B3 locus using standard conditions (Fig. 3 . 8B) , so the annealing temperature was reduced to 6 1°C, 58°C, and 55°C. This resulted in the amplification of both specific and nonspecific bands of increasing yield (results not shown). Despite attempts to fUlther optimise the amplification conditions (Section 2 . 1 3 .2), the nonspecific bands could not be eliminated. Given the difficulty of 95 Fig. 3 .S . ML-peR amplification with primers to microsatellite loci from the ES library Products were amplified from loci B2 (panel A), B3 (panel B), B4 (panel e), B5 (panel D), and B6 (panel E) and resolved on 3% NuSieve gels . peR reactions contained genomic DNA from i solates : Lp l (lane 1 ) , Lp5 (lane 2), Lp7 (lane 3), Lp1 3 (lane 4), Lpl 9 ( lane 5), Tfl 3 (lane 6), Tfl S (lane 7), Tf27 (lane S), Tf2S (lane 9), Fp2 (lane 10) , ES (lane 1 1 ) , and Frl (lane 1 2) . Lane 13 is a peR negative control without DNA and lane M contains 1 kb ladder (panels A, B , e, and E) or 50 bp ladder (panel D). A B c o E 220 bp - 1 54 bp - 220 bp 1 54 bp 220 bp - 201 bp - 1 54 bp - 1 50 bp - 1 00 bp - 50 bp - 220 bp - 201 bp - 1 54 bp - M 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 M 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 M 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 M 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 M 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 96 97 Fig. 3 .9 . ML-peR amplification with primers to microsatell ite loci from the Lp I library and HMG eoA reductase DNA sequence Products were amplified from loci B7 (panel A), B8 (panel B) , B9 (panel e), B lO (panel D), and B I l (using primers B I l . I and B I l .4; panel E) and resolved on 3% NuSieve gels . peR reactions contained genomic DNA from isolates: LpI (lane 1) , Lp5 (lane 2), Lp7 (lane 3), Lp1 3 (lane 4), LpI9 (lane 5) , Tf1 3 (lane 6), Tf1 8 (lane 7), Tf27 (lane 8), Tf28 (lane 9), Fp2 (lane 1 0) , E8 (lane 1 1 ) , and Fr1 (lane 1 2) . Lane 1 3 i s a peR negative control without DNA and lane M contains 50 bp ladder. 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 M A 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 M B 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 M c 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 M D 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 M E 200 bp 1 50 bp - 250 bp - 1 50 bp - 1 00 bp 250 bp 200 bp 1 50 bp 1 00 bp 250 bp 200 bp 1 50 bp - 250 bp - 200 bp - 1 50 bp 50 bp 98 99 redesigning primers to this locus because of the lack of available flanking sequences, further analysis of locus B3 was discontinued. Amplification of locus B7 only yielded a faint product from Tf1 8 (Fig. 3 .9A). Attempts to optimise the yield were not successful so any further work that was perfOImed using this locus was under standard conditions. Amplification of the B 1 1 locus with primer pair B 1 1 . 1 and B 1 1 .2 produced weak specific bands and high yields of primer dimers (results not shown). Further inspection of the primer sequences revealed that there was 3 '-end complementmity. Consequently primer B 1 1 .4 was designed 4 bp downstream from the primer B 1 1 .2 . The new plimer combination of B 1 1 . 1 coupled with B 1 1 .4 yielded specific products of good concentration under standard amplification conditions (Fig. 3 .9E). 3.2.2.5.2 Amplification of Fungal Outgroups To test the specificity of the ML-PCR assays for Epichloe endophytes, total DNA from two outgroups from the family Clavicipitaceae, including Echinodothis tuberiformis and CZaviceps purpurea were also amplified under the standard conditions for each ML-PCR assay, but no products were obtained with any of these primer sets (results not shown) . 3.2.2.5.3 In Planta Amplification To determine the specificity of the microsatellite locus primers for detecting Epichloe endophyte DNA in planta, total DNA extracted from endophyte-infected grass tissue using the CTAB method (Section 2 .5 .4) was used as a template in PCR reactions. The results for ARl -infected samples using primers to locus B 1 are shown in Fig. 3 . 1 0A. A product of approximately 298 bp was amplified from genomic DNA isolated from endophyte-infected plant material (lanes 2 , 4, 6, and 8) but no product was detected in endophyte-free material (lanes 1 , 3 , 5, and 7). The DNA sequence of this product was identical to that from AR1 alone confirming the specificity of the assay for detecting M A 506 bp - 298 bp - 201 bp - M B 200 bp - 1 50 bp - 1 00 bp - 2 3 2 3 4 5 6 7 4 5 6 8 9 1 0 1 00 Fig. 3 . 1 0. Optimising the template DNA concentration for amplification of Epichloe microsatellite loci from endophyte-infected plant material Agarose (3% NuSieve) gel s of PCR products amplified from genomic DNA isolated from endophyte-infected and endophyte-free grass tissue by using primers to microsatelli te loci B 1 (panel A) and B l O (panel B) . (A) PCR products generated with DNA prepared from AR1 -infected L. perenne x L. multiflorum hybrid ryegrass (lanes 2, 4, 6 , and 8), and endophyte-free hybrid ryegrass (lanes 1 , 3 , 5 , and 7 ) by the CTAB method at total DNA concentrations of 2 ng/IlI (lanes 1 and 2), 400 Pg/Il l (lanes 3 and 4), 200 Pg/Ill (lanes 5 and 6), 40 pg/Ill (lanes 7 and 8); and 40 pg/Ill of AR1 fungal genomic DNA (lane 9), and with no DNA (lane 1 0). Lane M, 1 kb ladder. CB) PCR products generated with DNA prepared from AR1 -infected L. perenne x L. multiflorum hybrid ryegrass by the FastDNA Kit H minipreparation method at total DNA concentrations of 2 ng/IlI (lane 1 ), 400 Pg/Il l (lane 2), 80 pg/Il l (lane 3), 1 6 pg/Ill (lane 4); and 40 Pg/Il l of ARI fungal genomic DNA (lane 5) , and with no DNA (lane 6). Lane M, 50 bp ladder. 10 1 endophyte sequences i n a template mixture that i s predominantly plant in origin . The best yield of product (lane 2) was obtained using a template concentration (2 ng//..tl) that was fifty-fold higher than that required to amplify the corresponding fungal DNA (lane 9); presumably reflecting the relative abundance of fungal DNA to plant DNA in this particular association. To improve the rapidity with which in planta detection of endophyte strains could be carried out, microsatell ite loci were amplified using DNA extracted from plant material using the FastDNA Kit H miniprep method (Section 2.5 .3) . The results for AR1 -infected L. perenne x L. multiflorum hybrid rye grass material amplified with primers to locus B 10 are shown in Fig. 3 . lOB . A product of the size ( 178 bp) characteristic o f this locus was obtained, but only when DNA concentrations in the range of 1 6 to 400 pg/Ill were used (lanes 2 to 4). At a DNA concentration of 2 ng/lll (lane 1 ) , which was shown to be optimal for DNA prepared by the eT AB method (Fig. 3 . l OA), no product was generated. This was presumably a result of the presence of inhibitors in DNA prepared by the FastDNA method. To test that endophyte amplification in planta occurred at each microsatel lite locus (B 1 , B4 to B6, and B 8 to B l l ) , genomic DNA from AR1 - infected L. perenne x L. multiflorum hybrid ryegrass, Fl l -infected tal l fescue cv. Grasslands Roa, and Tfl 5- infected tall fescue cv. Kentucky 3 1 (results not shown) , was amplified for each locus under standard conditions. DNA from the corresponding uninfected grasses along with fungal genomic DNA were also amplified. The results of amplification for the B4 and B6 loci are shown in Fig. 3 . 1 1 . For all microsatellite loci , products of the sizes characteristic of the corresponding endophyte alleles were amplified from the total DNA of in planta samples and fungal genomic DNA, and no products were detected from plant DNA only. A 250 bp - 1 50 bp - B 250 bp - 1 50 bp - M 2 3 4 5 6 7 M 2 3 4 5 6 7 1 02 Fig. 3 . 1 1 . Amplification of Epichloe microsatel li te loci from endophyte-infected plant material . Agarose (3% NuSieve) gels show microsatel lite peR products amplified using primers to microsatell i te loci B4 (panel A) and B6 (panel B) with template DNA from: endophyte-free hybrid ryegrass, 2 ng/)ll ( lane 1 ) , ARl -infected L. perenne x L. mult{florum hybrid ryegrass, 2 ng/)ll ( lane 2), ARl fungal genomic DNA, 40 Pg/Ill ( lane 3), endophyte-free tall fescue, 2 ng/)ll ( lane 4), Fl l -infected tall fescue, 2 ng/)ll ( lane 5), and Fl l fungal genomic DNA, 40 pg/Ill ( lane 6). Lane M, 50 bp ladder. 3.2.2.5.4 1 03 Screening Epichloil Endophyte Isolates Through ML-PCR Assays with High Resolution Separation of Alleles Having established standard conditions for PCR amplification of Epichloe endophyte microsatellite loci , 35 endophyte i solates were screened with primers to loci B 1 , B2 and B4 to B 1 1 . The isolates included representatives of the asexual taxonomic groups LpTG-2, N. lolii, N. coerwphialum, FaTG-2, FaTG-3 , and N. uncinatum, though sexual E. typhina and E. Jestucae isolates and two undescribed endophytes from Hordeum bogdanii were also screened (Table 2 . 1 ) . Products were labelled by [a_33p] dCTP incorporation (Section 2 . 1 3 .5), and separated on denaturing polyacrylamide gels to determine the size and size variation of the single stranded products generated at each locus (Section 2 .9). Figures 3 . 12 to 3 . 2 1 show autoradiographs of the gels separating the endophyte isolate alleles amplified in each ML-PCR assay. Isolates sharing l ike-banding patterns were readily identifiable and fingerprinting trends between different taxonomic groups were readil y recognised. The absolute size of individual alleles was difficult to determine as: (i) in most cases, each microsatellite allele was represented by several bands, and (ii) it was difficult to size bands run some distance from the sequencing ladder, due to the variation in electrophoretic mobility across the gel. Multiple bands can be attributed to a combination of separation of the product into single strands under denaturing electrophoretic conditions, slippage products generated during amplification, and the ability of Taq polymerase to add an additional deoxyadenylate to the end of products. Typically, alleles were observed to migrate in a c luster of four bands and this was interpreted as the presence of the two single strands, both with and without the extra deoxyadenylate added. Despite these limitations it was often possible to establish allele groupings, including the identification of hybrids which contain multiple alleles. w » � � I I t M fi; , x Lp1 Lp1 x Lp2 Lp2 x Lp5 Lp5 x Lp6 Lp6 x Lp1 3 Lp1 3 x Lp1 9 Lp1 9 x Lp20 Lp20 x Lp21 Lp21 x Lp? Lp? x Lp1 4 Lp1 4 x AR1 AR1 x AR1WF AR1 W F x AR1 CY AR1 CY x Tf5 Tf5 x l Tf2? Tf2? x l Tf28 Tf28 x Tf1 3 Tf1 3 x Tf1 5 Tf1 5 x Tf20 Tf20 Tf1 6 Tf1 6 Tf1 8 Tf1 8 I x Fp1 Fp1 I x Fp2 Fp2 x Fp4 Fp4 E8 E8 x Fr1 Fr1 x Frc5 Frc5 x Frc? Frc? x Frr1 Frr1 Fg1 Fg1 m x x x FI1 FI1 HaS HaS Hd1 Hd1 x EndoSafe EndoSafe M vOT Fig. 3 . 1 2. PAGE separation of micro satellite alleles amplified from endophyte isolates with locus B 1 (A) Autoradiograph of [a_33p] radiolabeI led ML-PCR products amplified from fungal genomic DNA and separated on a denaturing 6 % polyacrylamide gel . (B) Schematic diagram showing an interpretation of the position of bands observed in panel A. Band estimates are represented by x. 0 ...... » (j) 10 CO � � I I M Lp1 Lp2 Lp5 Lp6 Lp1 3 Lp1 9 Lp20 Lp21 Lp7 Lp1 4 Lp9 AR1 AR1 WF AR1 CY EndoSafe Tf5 Tf27 Tf28 M % ' Tf1 3 Tf1 5 Tf20 Tf1 6 Tf1 8 Fp1 Fp2 Fp4 HaS Hd1 E8 Fr1 Frc5 Frc7 Frr1 Fg1 FI1 neg clrl M 90 1 ID IJ.. >- ro S (J) C') (]) 0 N '-' o --.l <0 --' » -I:> 0 ;::t ;::t I I , M $ >< >< Lp1 Lp1 >< >< Lp2 Lp2 >< Lp5 Lp5 >< Lp6 Lp6 >< Lp1 3 Lp1 3 >< Lp1 9 Lp1 9 >< Lp20 Lp20 >< Lp21 Lp21 >< Lp7 Lp7 >< Lp1 4 Lp 1 4 >< Lp9 Lp9 >< AR1 AR1 >< AR1 W F A R 1 WF >< AR 1 CY A R 1 CY >< EndoSafe EndoSafe >< Tf5 Tf5 >< Tf27 Tf27 >< Tf28 Tf28 , M � >< Tf1 3 Tf1 3 >< Tf1 5 Tf1 5 >< Tf20 Tf20 >< Tf1 6 Tf1 6 >< Tf1 8 Tf1 8 >< Fp1 Fp1 >< Fp2 Fp2 >< Fp4 Fp4 HaS HaS >< Hd1 Hd1 >< E8 E8 >< Fr1 Fr1 >< Frc5 Frc5 >< Frc7 Frc7 >< Frr1 Frr1 >< Fg1 Fg1 >< FI1 FI 1 neg clrl �� M 80T Fig. 3 . 14. PAGE separation of microsatellite alleles amplified from endophyte isolates with locus B4 (A) Autoradiograph of [a_33p] radiolabelled ML-PCR products amplified from fungal genomic DNA and separated on a denaturing 6 % polyacrylamide gel . (B) Schematic diagram showing an interpretation of the posi tion of bands observed in panel A. Band estimates are represented by x. <0 ):> 0 ;:l. I M )( Lp1 Lp1 )( Lp2 Lp2 )( Lp5 Lp5 )( Lp6 Lp6 Lp1 3 Lp1 3 Lp1 9 Lp1 9 Lp20 Lp20 Lp21 Lp21 Lp7 Lp7 Lp 1 4 Lp1 4 Lp9 Lp9 AR1 AR1 AR 1 W F A R 1 WF AR1 CY AR1 CY EndoSafe EndoSafe )( Tf5 Tf5 )( Tf27 Tf2? )( Tf28 Tf28 Tf1 3 Tf1 3 Tf1 5 Tf1 5 Tf20 Tf20 Tf1 6 Tf1 6 Tf1 8 Tf1 8 '. M )( Fp1 Fp1 )( Fp2 Fp2 )( Fp4 Fp4 HaS HaS )( Hd1 Hd1 )( E8 E8 Fr1 Fr1 Frc5 Frc5 Frc? Frc? Frr1 Frr1 Fg1 Fg1 FI1 FI1 M O I T Fig. 3 . 1 5 . PAGE separation of microsatellite alleles amplified from endophyte i solates with locus B5 (A) Autoradiograph of [a}3p] radiolabel led ML-PCR products amplified from fungal genomic DNA and separated on a denaturing 6% polyacrylamide gel . (B) Schematic diagram showing an interpretation of the position of bands observed in panel A. Band estimates are represented by x. >-' >-' >-' CD ...... » (Xl (0 I\) ()1 ::l. ::l. I I M )( )( Lp1 Lp1 )( )( Lp2 Lp2 )( Lp5 Lp5 )( Lp6 Lp6 )( Lp1 3 Lp13 )( Lp1 9 Lp1 9 )( Lp20 Lp20 )( Lp21 Lp21 )( Lp7 Lp7 )( Lp1 4 Lp1 4 )( Lp9 Lp9 )( AA1 AA1 )( AA1WF AA1WF )( AA1 CY AR1 CY )( EndoSafe EndoSafe )( Tf5 Tf5 )( Tf27 Tf27 )( Tf28 Tf28 M )( Tf1 3 Tf1 3 )( Tf1 5 Tf1 5 )( Tf20 Tf20 )( Tf1 6 Tf1 6 )( Tf1 8 Tf1 8 )( Fp1 Fp1 )( Fp2 Fp2 )( Fp4 Fp4 )( HaS HaS )( Hd1 Hd1 )( E8 E8 )( Fr1 Fr1 )( Frc5 Frc5 )( Frc7 Frc7 )( Frr1 Frr1 )( Fg1 Fg1 )( FI1 F I1 M Z I T Fig. 3 . 1 6. PAGE separation of microsatel lite alleles amplified from endophyte i solates with locus B6 (A) Autoradiograph of [a}3p] radio labelled ML-PCR products amplified from fungal genomic DNA and separated on a denaturing 6% polyacrylamide gel . (B) Schematic diagram showing an interpretation of the position of bands observed in panel A . Band estimates are represented by x. --I <0 .... » VJ .p.. I\) :::l. :::l. :::l. I I I M Lp1 Lp2 Lp5 Lp6 Lp1 3 Lp1 9 Lp20 Lp21 Lp7 Lp1 4 Lp9 AR1 AR1WF AR1 CY EndoSafe Tf5 Tf27 Tf28 M Tf1 3 Tf1 5 Tf20 Tf1 6 Tf1 8 Fp1 Fp2 Fp4 HaS Hd1 E8 Fr1 Frc5 Frc7 Frr1 Fg1 FI1 t M vI I CD U. >- '@ 3: 0 (j) '" (j) 0 - '- � $ Cl) C'l (J) 0 � '- � $ () Cl) C'l (J) 0 N '- co 3: () (j) C') Ol 0 � '- '@ 3 () (/) C') CJ) 0 N "- M Lp1 Lp2 Lp5 Lp6 Lp1 3 Lp1 9 Lp20 Lp21 Lp? Lp1 4 Lp9 AR1 AR1WF AR1 CY EndoSafe Tf5 Tf2? Tf28 M Tf1 3 Tf1 5 Tf20 Tf1 6 Tf1 8 Fp1 Fp2 Fp4 HaS Hd1 E8 Fr1 Frc5 Frc? Frr1 Fg1 FI1 M 1'er >< >< >< >< >< >< >< >< >< >< >< >< >< >< >< >< >< >< >< >< >< >< >< >< >< >< >< >< ZZl >< >< >< >< >< >< >< >< >< >< >< >< >< DJ >< >< Lp1 Lp2 LpS Lp6 Lp1 3 Lp1 9 Lp20 Lp21 Lp7 Lp1 4 Lp9 AR1 AR1 WF AR1 CY EndoSafe TfS Tf27 Tf28 Tf1 3 Tf1 S Tf20 Tf1 6 Tf1 8 Fp1 Fp2 Fp4 HaS Hd1 E8 Fr1 FrcS Frc? Frr1 Fg1 FI1 Fig. 3 .2 1 . PAGE separation of microsatellite alleles amplified from endophyte i solates with locus B l l (A) Autoradiograph of [a_33p] radiolabelled ML-PCR products amplified from fungal genomic DNA and separated on a denaturing 6% polyacrylamide gel. (B) Schematic diagram showing an interpretation of the position of bands observed in panel A. Band estimates are represented by x. 124 3.2.2.5.5 Summary of Screening Epichloi/ Endophyte Isolates For polyacrylamide gels (Figs. 3 . 1 2A to 3 .2 1A), an interpretation of how the different allele groupings were made are compiled in schematic diagrams (Fi gs . 3 . 1 2B to 3 .2 I B). A summary of these i s given in Table 3 . 1 . Screening with microsatel lite locus B 2 (Fig. 3 . 1 3), only LpTG-2, FaTG-3, HaB, Hd1 , and E. typhina amplified. Banding patterns were difficult to interpret as the usual tight cluster of four bands were not present, but the different taxonomic groups within thi s group were di stinguishable from one another. Primers to the BS locus amplified LpTG-2, N. uncinatum, Hdl , and E. typhina (Fig. 3 . l 5) . Very faint amplification of N. lolii (loA) and N. coenophialum samples was visible. Of these samples, only two alle1es were identified, so it was not a very informative locus . The B7 locus was amplified in LpTG-2, FaTG-3 , and E. typhina (Fig . 3 . 1 7) . Only two different fingerprints were seen, which grouped LpTG-2 with E. typhina. The B8 locus was amplified in LpTG-2, N. lolii, FaTG-2 and E. festucae isolates Fgl and FI l , though the different taxonomic groups were not completely distinguishable from one another. Primers to the remaining six loci (B I , B4, B6, B9, B lO, and B l l ) amplified almost all of the isolates tested and, between them, were able to distinguish isolates to at least their i sozyme phenotype grouping (as described by Christensen et al . , 1 993). Locus B l l was particularly informative, with 23 different fingerprints distinguishable across the 35 i solates surveyed. This was the only locus that was able to distinguish most of the different i sozyme phenotypes within the N. lolii group. Banding patterns from the B lO locus were difficult to interpret as there were multiple overlapping al lele clusters present and it was difficult to determine whether alleles were homologous. Nonetheless, loci B9 and B lO proved to be palticularly informative in identifying known hybrid endophytes, such as the tall fescue endophytes, that are characterised by the presence of multiple alleles . Multiple products were amplified from these same loci in N. uncinatum, providing further evidence for a hybrid Oligin for thi s group. Table 3.1. Summary of polymorphism and size range of ML-PCR products amplified from Epichloe endophyte isolates with PAGE separation ML-PCR Expected product assay size (bp) B l 325 B2 163 B4 124 B5 107 B6 173 B7 111 B8 142 B9 c 157 B9 d 246 B IO 179 B I l 173 a For 35 endophyte isolates screened b Lp9 was not included in this screen C With primer pair B9. l and B9.2 d With primer pair B9.1 and B9.4 Isolate of original microsatellite NF 62A E8 E8 E8 E8 Lp l Lp l Lpl Lpl Lp l Lpl9 Size range of No. of fingerprint products (bp) patterns a 300 - 340 9 b 95 - 170 4 105 - 125 7 70 - 110 2 150 200 10 70 - 120 4 130 - 150 5 145 - 190 12 235 - 280 7 145 - 200 16 115 240 23 1 25 3.2.2.6 Development of Primer Set I 1 26 Given that primers to microsate Il i te loci B4, B6, B9, B 1 0, and B 1 1 amplified almost all of the 35 i solates tested, these primer sets were selected for the development of an automated DNA fingerprinting system uti l ising fluorescently labelled primers and the laser detection technology associated with the ABI automatic sequenceI'. The size range of alleles generated for each of these microsatellite loci are summarised in Table 3 . 1 . With reference to these size ranges, a primer dye-labelling strategy that utilises the three available dye phosphoramidites: 6-FAM (blue), TET (green), and HEX (yellow), was formulated to allow the simultaneous size analysis of the products of all five microsatellite loci on a single lane of the automatic sequencer (Fig. 3 .22). The Bx. 1 primer of each primer pair was arbitrari ly chosen to be 5 '-end labelled with the dye phosphoramidite. As microsatelli te B 1 1 had such a wide range of PCR product sizes, i t was assigned its own dye, HEX. Primers to loci B4 and B6 were both labelled with 6-FAM. As the allele size ranges from both B9 ( 145 to 1 90 bp) and B lO ( 145 to 200 bp) overlapped, both loci could not be labelled with the remaining dye, TET. Therefore a new primer to the B9 locus (B9A) was designed to pair with B9 . 1 , to increase the size range of alleles to 235 to 280 bp. A consequence of this change was a reduction in the level of polymorphism at the B9 locus (described in Section 3 .2 .2 .6 . 1 ), but sufficiently informative fingerprints were generated at the remaining four loci to resolve the endophyte groups. The primers: Bx . l IBx.2 for B4, B6, and B 1 0, and Bx. l/BxA for B9 and B 1 1 , were therefore grouped together, and are hereby referred to as Primer Set 1 . 3.2.2.6. 1 Redesigning Primers to Microsatellite Locus B9 A new pnmer, B9A, was designed 89 bp downstream of B9.2, to pair with B9. 1 (Fig. 3 .7) . Standard amplification conditions were used to amplify genomic DNA from the original panel of 1 2 endophyte isolates (Section 3 .2. 1 ) using the B9. 1 /B9A primer combination. Good amplification was obtained (Fig. 3 . 23) though Tfl 8 , Fp2, and E8 fai led to amplify, and multiple alleles were not detected from the tal l fescue endophytes, contrary to amplification with the B9. 1 /B9.2 pair (Fig. 3 .9C) . The panel of 35 endophyte 1 27 Fig. 3 .22. Strategy for the dye-labelling of primers to microsatelI ite loci for Primer Set I The strategy shows how the products for microsatell i te loci B4, B6, B9, B lO, and B l l can be analysed in a single lane of an automatic genetic analyser. Double headed arrows indicate the observed size range of microsatellite peR products, as determined by separation on polyacrylamide gels, for the endophyte isolates analysed. For each locus, Bx, primers Bx . 1 and Bx.2 were used, except loci B9 and B l l , where B9. 1 and B9.4, and B I l . I and B I 1 .4 were used, respectively. 1 28 allele size (bp) 280 260 89 240 - 220 - 200 81 1 1 80 - 86 81 0 1 60 140 - 1 20 - � 1 00 6-FAM TET HEX dye phosphoramidite label 350 bp 250 bp 1 50 bp M 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 Fig. 3 .23 . ML-PCR amplification of locus B9 using a new plimer pair, B9. l and B9.4 1 29 PCR reactions contained genomic DNA from isolates : Lp1 (l ane 1 ), Lp5 (lane 2), Lp7 (lane 3), Lp1 3 (lane 4), Lp 19 ( lane 5), Tfl 3 (lane 6), Tfl 8 (lane 7), Tf27 (lane 8 ), Tf28 ( lane 9), Fp2 ( lane 1 0) , E8 ( lane 1 1 ), and Frl (lane 1 2) . Lane 1 3 i s a PCR negative control without DNA and l ane M contains 50 bp ladder. Products were separated on a 3% NuSieve gel . 1 30 isolates were screened through the B9. lIB9.4 assay, with polyacrylamide gel separation as previously described (Section 3 .2 .2 .5 .4) and an autoradiograph of this gel i s shown in Figure 3 .24. In this screening N. uncinatum and isolates Tf16 , HaB, and E8 failed to amplify, and very weak amplification was detected from Hdl and Tfl 8 . Also, multiple alle les were not amplified from hybrid endophytes , though very good amplification was obtained from all other i solates . 3.2.2.6.2 Multiplex peR of Primer Set I Primers to the five informative microsatel l i te loci identified above (B4, B6, B9, B 10, and B l l ) were tested in a single multiplex reaction (Section 2 . 1 3 .6) for their abil ity to amplify all loci , thus economising on reagents, time, and l abour. A comparison of the reaction products generated for the five-locus multiplex reaction and for individual locus amplifications, using template DNA from endophytes Tf1 3 and Frl , i s shown in Figure 3 .25 . All products amplified in single-locus reactions were present in the multiplex reaction. This was confirmed by calTying out the same reactions in the presence of [a_33p] dCTP fol lowed by separation of the products on polyacrylamide gels and detection by autoradiography. The products of the single-locus ampl ifications matched exactly the size of the products of the multiplex reaction (results not shown). The success of the multiplex reaction was however dependent of the quality of the template DNA used. Multiplex PCR of DNA prepared from plant matelial using the CT AB method (Section 2.5 .4) gave consistently stronger signals than that prepared using the FastDNA Kit (Section 2 .5 .3) ; presumably a consequence of the presence of inhibitors in the latter. To increase the signal strength of the weaker products from sub-optimal amplifications, alternative PCR strategies were undertaken including single-dye, and single-locus PCR reactions with subsequent pooling of products . Single-dye multiplex PCR reactions ampli fied loci , whose primers were labelled with the same dye, in a single reaction. Thus three reactions, one for each dye, were peJformed for each sample. This approach reduced the complexity of the multiplex reaction and allowed good amplification of M >< Lp1 Lp1 >< Lp2 Lp2 >< Lp5 Lp5 >< Lp6 Lp6 >< Lp1 3 Lp1 3 >< Lp1 9 Lp1 9 >< Lp20 Lp20 >< Lp21 Lp21 >< Lp7 Lp7 >< Lp 1 4 Lp1 4 >< Lp9 Lp9 >< AR1 AR1 >< AR1 W F AR1 W F >< AR1 CY AR1 CY >< EndoSafe EndoSafe >< Tf5 Tf5 >< Tf27 Tf27 >< Tf28 Tf28 M >< Tf1 3 Tf1 3 >< Tf1 5 Tf1 S >< Tf20 Tf20 Tf1 6 Tf1 6 >< Tf1 8 Tf1 8 Fp1 Fp1 Fp2 Fp2 Fp4 Fp4 HaS HaS >< Hd1 Hd1 E8 E8 >< Fr1 Fr1 >< FrcS FrcS >< Frc7 Frc7 >< Frr1 Frr1 >< Fg1 Fg1 >< FI1 F I1 M In Fig. 3 . 24. PAGE separation of microsatelli te alleles amplified from endophyte i solates with locus B9 , using primers B9. 1 and B9.4 (A) Autoradiograph of [a_33p] radiolabelled ML-PCR products amplified from fungal genomic DNA and separated on a denaturing 6% polyacrylamide gel . (B) Schematic diagram showing an interpretation of the position of bands observed in panel A. Band estimates are represented by x. -W Iv 350 bp - 250 bp - 200 bp - 1 50 bp - 1 00 bp - M 2 3 4 5 6 7 8 9 1 0 1 1 1 2 Fig. 3 .25 . Multiplex amplification of microsatellite loci using Primer Set I 133 Products were amplified by either single locus or multiplex PCR for endophyte i solates Tfl 3 (lanes 1-6), and Fr! (lanes 7-1 2). PCR reactions contained the primer combinations: Primer Set I (lanes 1 and 7), B4. 1 and B4.2 (lanes 2 and 8), B6. I and B6.2 (lanes 3 and 9), B9 . I and B9.4 (lanes 4 and 10), B IO. l and B IO .2 (lanes 5 and 1 1 ), and B l l . 1 and B l l .4 (lanes 6 and 1 2) . Lane M contained 50 bp ladder. Products were separated on a 3% NuSieve agarose gel . 1 34 samples which did not amplify optimally in the five locus multiplex. In extreme cases single loci could be amplified separately, then pooled for subsequent analysis . 3.2.2.7 Development of Primer Set II On the basis of the high amount of allele length polymorphism observed when representative isolates of each of the ten Epichloe spp. were screened through ML-PCR assays that were not included in Primer Set I, a second set of dye-labelled primers, Primer Set H, was developed. Primer Set H further extends the analysis of microsatell i te genotypes in Epichloe endophytes and amplifies microsatellite l oci B l , B2, BS , B6, and B9. 3.2.2. 7.1 Screening Epichloi/ Species with ML-PCR Assays Following Primer Set I development, fungal genomic DNA from 4 1 i solates representing each Epichloe species, were made available (c. L. Schardl) . These DNA samples were prepared using the method of Byrd et a1 . ( 1990). Isolates were screened through ML-PCR assays using standard amplification conditions (Section 2 . 1 3 .S) which amplified the loci excluded from Primer Set I, ie . B l , B2, B5, B7, B8 , and locus B9 using primers B9. l and B9.2. Amplification products were separated on agarose gels (Section 2 .8), these are shown in Figure 3 .26. No products were detected from Epichloe spp. when screened with microsatellite locus B8 (Fig. 3 .26E), and only E. clarkii and several E. typhina isolates (E8 , E348, E42S, E428, E43 1 , and E470) amplified weakly with primers to the B7 locus (Fig. 3 .26D). The remaining four loci amplified a considerable number of the isolates tested, and the results of these are summarised in Table 3 .2 . Locus B l amplified al l isolates except E. typhina, E. clarkii and E. baconii i solates E421 and E424 (Fig. 3 .26A); locus B2 amplified E. typhina, E. clarkii, E. elymi, E. bromicola, and E. sylvatica (Fig. 3 . 26B) ; locus BS amplified E. typhina, E. clarkii, and E. sylvatica (Fig. 3 .26C); and B9 amplified all i solates except E. clarkii, E. amarillans, E. glyceriae and E. typhina isolates E8, E348, E3S 8, E42S, E428 , E43 1 , E469, E470, ESOS, and E. sylvatica i solate ES03 (Fig. 3 .26B) . Although the 1 35 Fig. 3 .26. ML-PCR screening of representative i solates of Epichloe spp. using plimer pairs excluded from Plimer Set I Agarose (3%) gels showing PCR products amplified with plimers to microsateIlite loci B 1 (Fig. 3 .26A), B2 (Fig. 3 .26B), B5 (Fig 3 .26C), B7 (Fig 3 .26D), B 8 (Fig 3 .26E), and B9 (Fig 3 .26F). Lanes contain reactions with genomic DNA from i solates: ( 1 ) E8, (2) E348 , (3) E358 , (4) E425 , (5) E428 , (6) E43 1 , (7) E469, (8) E470, (9) E505, ( 10) E 102 1 , ( 1 1 ) E 1022, ( 1 2) E 1032, ( 1 3) E 1033, ( 1 4) E2466, ( 1 5) E422, ( 1 6) E426, ( 17 ) E427, ( 1 8) E28 , ( 1 9) E32, (20) E 1 86, (2 1 ) E 1 89, (22) E434, (25) E56, (26) E 1 84, (27) E52, (28) E57, (29) E42 1 , (30) E424, (3 1 ) E l03 1 , (32) E50 1 , (33) E502, (34) E798, (35) E799, (36) E354, (37) E503, (38) E277, (39) E2772, (40) E 1040, (4 1 ) E 1045, (42) E 1046, (43) E 1035.30, (23 and 44) PCR positive control. Lanes 24 and 45 are PCR negative controls without DNA and lane M contains 1 23 bp ladder. The positive control genomic DNA used in amplifications of loci B l , B8, and B9 is from isolate Lp6, and for amplifications B2, B5, and B7 is from i solate E8. M 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 M M 1 3 1 4 1 5 1 6 1 7 1 8 1 9 20 2 1 22 23 24 M M 25 26 27 28 29 30 31 32 33 34 35 36 M M 37 38 39 40 4 1 42 43 44 45 M 61 5 bp 369 bp 246 bp 1 23 bp 1 36 - 369 bp - 246 bp - 1 23 bp - 615 bp - 369 bp - 246 bp - 1 23 bp - 6 1 5 bp - 369 bp - 246 bp 1 23 bp Fig. 3 .26A. ML-PCR screening of representative i solates of Epichloe spp. with primers to the B 1 locus. M 2 3 4 5 6 7 8 9 1 0 1 1 1 2 M M 1 3 1 4 1 5 1 6 1 7 1 8 1 9 20 2 1 22 23 24 M M 25 26 27 28 29 30 31 32 33 34 35 36 M M 37 38 39 40 4 1 42 43 44 45 M 369 bp 246 bp 1 23 bp 369 bp - 246 bp - 1 23 bp - 369 bp - 246 bp - 1 23 bp 369 bp 246 bp 1 23 bp 1 37 Fig. 3 .26B . ML-PCR screening of representative isolates of Epichloe spp. with primers to the B2 locus. 1 38 M 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 M 1 23 bp M 1 3 1 4 1 5 1 6 1 7 1 8 1 9 20 2 1 22 23 24 M - 1 23 bp M 25 26 27 28 29 30 31 32 33 34 35 36 M - 1 23 bp M 37 38 39 40 4 1 42 43 44 45 M - 1 23 bp Fig. 3 .26C. ML-PCR screening of representative i solates of Epichloe spp. with primers to the B5 locus. M 2 3 4 5 6 7 8 9 1 0 1 1 1 2 M M 1 3 1 4 1 5 1 6 1 7 1 8 1 9 20 2 1 22 23 24 M M 25 26 27 28 29 30 31 32 33 34 35 36 M M 37 38 39 40 4 1 42 43 44 45 M 1 23 bp 1 39 - 1 23 bp Fig. 3 .26D. ML-PCR screening of representative i solates of Epichloe spp. with primers to the B7 locus. M 2 3 4 5 6 7 8 9 1 0 1 1 1 2 M M 1 3 1 4 1 5 1 6 1 7 1 8 1 9 20 2 1 22 23 24 M M 25 26 27 28 29 30 31 32 33 34 35 36 M M 37 38 39 40 41 42 43 44 45 M 369 bp - 246 bp 1 23 bp 140 - 369 bp - 246 bp - 1 23 bp - 369 bp - 246 bp - 1 23 bp - 369 bp - 246 bp - 1 23 bp Fig. 3 .26E. ML-PCR screening of representative i solates of Epichloe spp. with ptimers to the B8 locus. M 2 3 4 5 6 7 8 9 1 0 1 1 1 2 M M 1 3 1 4 1 5 1 6 1 7 1 8 1 9 20 2 1 22 23 24 M M 25 26 27 28 29 30 31 32 33 34 35 36 M M 37 38 39 40 4 1 42 43 44 45 M 369 bp 246 bp 1 23 bp - 369 bp - 246 bp 1 23 bp 369 bp 246 bp - 1 23 bp - 369 bp - 246 bp 1 23 bp 1 4 1 Fig. 3 .26F. ML-PCR screening o f representative isolates of Epichloe spp. with primers to the B9 locus, B9. 1 and B9.2 142 Table 3.2. Summary of amplification of representative isolates of Epichloif spp. in Bl, B2, BS, B7, BS, and B9 ML-PCR assays Epichloe spp. Mating Isolates amplified population a B l B2 B5 B7 B8 B9 E. typhina X j j some X some E clarkii X j j j X X Efestucae II j X X X X j E. elymi HI j j X X X j E. amarillans IV j X X X X X E baconii V some X X X X j E. bromicola VI j j X unsure X j E sylvatica VII j j j unsure X some E glyceriae VIII j X X unsure X X E. brachyelytri IX j X X unsure X j a Mating population designations as defined by Schardl et a\. (1997). 143 PCR products in these screenings were not resolved on polyacrylamide gels, allele size polymorphism was clearly observed. An unexpected result was the presence of two bands when the B2 locus of E. typhina isolate E428 was amplified (lane 5, Fig. 3 .26B) . This may be the result of gene duplication, or the remnant of an interspecific hybridisation event. 3.2.2. 7.2 Multiplex peR of Primer Set II As primers to microsatel1ite loci B 1 , B2, B5 , and B9 amplified a wide range of Epichloe i solates, these primer sets were selected for the development of a second automated DNA fingerprinting system. Given the size ranges of alleles, which are summmised for each of these microsatellite loci in Table 3 . 1 , a second primer dye-labelling strategy was developed to allow the simultaneous size analysis of the products of these loci on a single lane of the automatic sequencer (Fig. 3 .27). As microsatellite B2 generated the greatest size range of PCR products, it was assigned its own dye, HEX. Primer B9. 1 was already labelled with TET from Primer Set I, and this primer coupled with B9.2, generated products in the size range 1 45 to 1 90 bp. The size range of products of B5 (70 to 1 10 bp) do not overlap with those of B9, so this locus was also labelled with TET. B l was labelled with 6-FAM, and 6-FAM labelled B6 was also included as a control between Primer Set I and Primer Set H. As with Primer Set I (Section 3 .2.2.6 .2), the ability of the plimers in Primer Set H to amplify al l microsatellite products in a single multiplex peR reaction was tested. A comparison of the reaction products generated for the five-locus multiplex reaction and for individual locus amplifications, using template DNA from endophytes E8, E 1D2 1 , and E354, i s shown in Figure 3 .28 . All products amplified i n single-locus reactions were present in the multiplex reaction. Further research with regard to the multiplex reaction of Primer Set II was not performed as it was expected that any problems would be overcome as discussed in Section 3 .2.2.6 .2 . 144 Fig. 3 .27. Strategy for the dye-labelling of primers to microsatell ite loci for Primer Set IT The strategy shows how the products for microsatelli te loci B 1 , B2, B5 , B6, and B9 can be analysed in a single lane of an automatic genetic analyser. Double headed arrows indicate the observed size range of microsatel lite peR products, as determined by separation on polyacrylamide gels, for the endophyte i solates analysed. For each locus, Bx, primers Bx. 1 and Bx.2 were used. 145 allele size (bp) 340 - 320 - 300 - 280 260 240 220 - 200 - 1 80 86 89 1 60 82 1 40 120 - 1 00 80 6-FAM TET HEX dye phosphoramidite label 146 M 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 369 bp - 246 bp - 1 23 bp - Fig. 3 .2S . Multiplex amplification of microsatellite loci using Primer Set II Products were amplified by either single locus or multiplex PCR for endophyte isolates ES (lanes 1-6), E 102 1 (lanes 7-1 2) , and E3S4 (lanes 1 3-1 S). PCR reactions contained the primer combinations: Primer Set II (lanes 1 , 7, and 1 3), B l . l and B l .2 (lanes 2, 8, and 1 4) , B2 . 1 and B2.2 (lanes 3 , 9, and I S) , BS . l and BS .2 (lane 4, 10 , and 1 6) , B6. 1 and B6.2 (lanes 5 , 1 1 , and 1 7), and B9. 1 and B9.2 (lanes 6, 1 2, and I S). Lane M contains 1 23 bp ladder. Products were resolved on a 3% NuSieve agarose gel . 3.2.2.8 Automated Analysis of Fingerprints 147 A limitation of the radiolabelling method described in Section 3 .2.2 .5 .4 is the inability to accurately determine both the size and the number of alleles present in a given sample. This problem was particularly acute in samples from hybrid endophytes with multiple alIeles. To overcome these problems, samples were amplified using fluorescently labelled primers and the products analysed on ABI Prism 377 and 3 10 analysers using GeneScan Analysis 2 . 1 software (Section 2 . 17) . The electropherogram for one such analysis, that for Tf28 amplified with Primer Set I and analysed on an ABI Prism 377 DNA Sequencer, is shown in Figure 3 .29 and the allele sizes, in nucleotide units (ntu), for thi s sample are shown in Table 3 .3 . The peaks associated with each microsatellite locus were identified by reference to the dye label and the expected size range of products at each locus (Figs. 3 . 24 and 3 .27) . Many peaks occurred as a doublet, one nucleotide apart, corresponding to the labelled strand of the peR with and without the extra deoxyadenylate added at the 3 ' -end by Taq polymerase. In constructing a database of allele sizes for each endophyte strain, we have entered the size of the product without the additional deoxyadenylate. Where only a single peak was observed, the value for this was used. The allele sizes determined by this method were generally in good agreement with those generated by radiolabelled peR (Table 3 . 1 ), though the ease of size determination was far greater for the automated multi-dye colour system. In addition, automated analysis allows precise size determination of mUltiple products, including similar sized products from different loci , in a single run. To check the reproducibility of allele size-calling by the automatic sequencer, dilutions of various fluorescently-Iabelled peR products of Tfl 3 were submitted for analysi s . Allele sizes were found to be most consistent when the peak heights of the samples were less than 2000 fluorescent units (fu) . Working within thi s range resulted in allele size variation of up to 0.4 nucleotide units. 1 48 DNA fragment size (ntu) 80 100 1 20 140 1 60 1 80 200 220 240 260 280 300 , , , , , , , , , , , S- 1600 '!::.- "'iii 1 200 c Cl 'Ui C 800 (]) u (/) � 400 0 :;) u:: 0 I I I 84 81 1 89 Fig. 3 .29. Electropherogram of an endophyte microsatellite fingerprint. Genomic DNA from Tf28 was amplified with fluorescently-labelled Primer Set I, and the products were separated by PAGE. GeneScan 2. 1 software was used to generate the electropherogram, and the sizes of the products were determined by reference to GS-500 T AMRA internal size standards. Peaks in red, blue, green, and black correspond to the size standards, 6-FAM (B4 and B6), TET (B9 and B l O) , and HEX ( B l l )-labelled products, respectively. Table 3.3. Allele sizes for microsatellite loci of Tf28 from electropherogram of Fig. 3.29 Colour Dye label Peak size Peak height Deduced microsateIlite (ntu) (fu) locus blue 6-FAM 100.00 4726 B4 blue 6-FAM 193.75 348 B6 green TET 175 .14 909 B lO green TET 176.15 382 B IO green TET 180.93 2054 B lO green TET 181.93 990 B lO green TET 183.81 1191 B IO green TET 184.81 632 B lO green TET 272.23 407 B9 green TET 273.22 447 B9 yellow HEX 128.65 1275 B l l yellow HEX 192.65 273 B l l yellow HEX 193.64 127 B l l red TAMRA 75.00 292 OS-500 T AMRA red TAMRA 100.00 113 OS-500 T AMRA red TAMRA 139.00 298 OS-500 T AMRA red TAMRA 150.00 296 OS-500 T AMRA red TAMRA 160.00 301 OS-500 T AMRA red TAMRA 200.00 304 OS-500 T AMRA red TAMRA 250.00 337 OS-500 T AMRA red TAMRA 300.00 366 OS-500 T AMRA 149 3.2.2.8. 1 Comparison of Fingerprints Analysed on an ABI Prism 377 DNA Sequencer and an ABI Prism 310 Genetic Analyser 1 50 DNA fragment size estimation using ABI Prism analysers depends on the e lectrophoretic mobi l i ty of the single-stranded DNA fragment relative to the electrophoretic mobilities of a set of standard fragments of known sizes. The use of different separation matrices can alter the electrophoretic mobility of a DNA molecule in a sequence-specific manner, thus resulting in different size-calling (Rosenblum et aI . , 1 997). Indeed, di screpancies in size­ calling of up to 6 ntu were observed between identical samples which had been analysed on both the slab gel based ABI Prism 377 DNA Sequencer and the capi l lary polymer based ABI Prism 3 10 Genetic Analyser (Section 2 . 1 7). To quantify the extent of the deviation precisely at each microsatell i te locus, the sizing of fingerprint patterns determined on both models of ABI Prism analyser was compared. Specifically, the fingerprints analysed were from i solates Lp6, Fpl , Tfl 5 , E348, E32, E56, and E798 amplified with Primer Set I ; and i solates E102 1 , E422, E32, E354, and E 1040 amplified with Primer Set II (Section 2 . 1 3). These results and the calculated offsets are shown in Tables 3.4 and 3 . 5 . Generall y the sizes determined on the ABI Prism 377 were larger than those on the ABI Prism 3 10 for all loci except B 1 . The offsets ranged between -0.8 to +5.9 ntu, with loci B4 and B5 showing the greatest differences with average offsets of 4.3 ntu and 5 .0 ntu respectively. It was hoped that precise universal offsets might be calculated at each locus, so that data generated on either model of ABI Prism analyser could be readily compared in a common database, but the range of offset sizes at many of the loci were wider than expected. For aU loci except B4, the offset range was less than ±0.9 ntu from the average offset, but for locus B4 the offset range was ± 1 . 8 ntu. Despite these wide ranges, the calculated average offsets have sti ll been used in this study to make comparisons between the genotypes analysed by either ABI system. In the analysis of human and l ivestock microsatell ite genotyping, it i s common practise to compare data generated on either model of ABI Prism by correcting the data by a locus specifi c offset, but in these cases only single species are compared, and these are assumed to have conserved sequences Table 3.4. Calculation of the average size differences of ML-PCR Primer Set I products when automatically analysed on an ABI Prism 377 slab gel system compared to an ABI Prism 310 capillary electrophoresis system. Allele size (ntu) Isolate B4 B6 B9 B l O B11 377 310 offset 377 310 offset 377 310 offset 377 310 offset 377 310 offset Lp6 100.1 95.0 5 .1 187.6 185.4 2.2 247.3 245.0 2.2 178.1 175. 8 2.3 177.0 176.4 0 .6 84.8 87.1 2.3 160.7 157.8 2.9 Fp l 178.7 177.1 1.6 120.7 118 .5 2.2 104.4 98.9 5 .5 195 .8 192.9 2.9 172.1 169. 8 2.3 128.8 126.6 2.2 Tf l 5 101.0 95.1 5.9 186.6 185.5 1.1 247.5 245.1 2.4 178.0 175.7 2 .3 165.2 164.4 0.8 E348 128.3 124.6 3 .7 l71.8 l70.5 1.3 192.4 190.8 1.6 128 .2 127.0 1.2 E32 99.4 95.5 3.9 187.6 186.1 1.5 271.9 269.6 2 .3 171.8 170.5 1.3 165.3 164.5 0 .8 E56 83 .0 79.1 3 .9 178.8 177.4 1.4 195.0 192.8 2.2 128.3 126.1 2.2 E798 84.3 80.0 4.3 178.7 177.2 1.5 186.3 184.0 2.3 112.7 110.5 2 .2 average offset 4.3 1.5 2.3 2 .2 1.5 offset range 3 .6 1.1 0.2 1.6 1.4 n 8 7 3 9 8 Table 3.5. Calculation of the average size differences of ML-PCR Primer Set II products when automatically analysed on an ABI Prism 377 slab gel system compared to an ABI Prism 310 capillary electrophoresis system Allele size (ntu) Isolate B 1 B2 B5 B9 377 3 10 offset 377 3 1 0 offset 377 3 10 offset 377 3 10 offset E 102 1 147.5 143.8 3 .7 75. 1 70.7 4.4 180. 1 1 79 . 1 1 .0 E422 160.9 158.5 2.4 85.8 80. 1 5.7 E32 304.7 305.2 -0.5 1 82.9 1 82.2 0.7 E354 302.5 225.9 22 1 .8 4. 1 75. 1 70.2 4.9 180.2 179.3 0.9 E 1040 3 14. 1 3 14.9 -0.8 174. 1 173 . 1 1 .0 average offset -0.7 3.4 5.0 0.9 offset range 0.3 1 .7 1 .3 0.3 n 2 3 3 4 153 flanking the microsatellite. As the Epichloe endophytes represent a wide assemblage of species, the sequence differences in the flanking regions affect the migration and hence allele sizing on both systems, thus resulting in the wide range of offsets observed. For more accurate offset calculation, sequence specific offsets would need to be determined at each locus for each species individuall y (pers. comm. Mark Holloway, PE Applied Biosystems). 3.2.3 APPLICATION OF THE PCR-BASED MICROSATELLITE FINGERPRINTING ASSAYS 3.2.3.1 Microsatellite Fingerprinting of Epichloe Endophyte Isolates Having establi shed optimum conditions for automated analysis (Section 3 .2 .2 .8) , fungal genomic DNA from 80 Epichloe endophyte i solates were amplified with both Primer Sets I and II in separate multiplex reactions. The dye-labelled PCR products were automatical ly analysed on either ABI Prism model 377 or 3 10 genetic analysers (Section 2 . 1 7), and the sizes of the products were entered into a database. A summary of the genotypes from these isolates is shown in Table 3 .6. Microsatell ite electropherogram files from all samples are avai lable on the CD-ROM with detai l s in Appendix 3 . 3.2.3.2 Application of Microsatellite Fingerprinting Assay in a Blind Test To test the robustness of the microsatel l i te fingerprinting assay for identifying endophyte strains in planta, twelve samples of plant material from endophyte-infected perennial ryegrass, tall fescue and meadow fescue were submitted blind by an independent researcher at AgResearch (M. J. Christensen). DNA was isolated from these samples using the FastDNA Kit H (Section 2.5 .3) and amplified with Primer Set I in single-dye PCR reactions (Section 2. 1 3 .6). The products from each reaction were pooled and submitted for automatic analysis on an ABI Prism 377 genetic analyser (Section 2. 1 7) . AlIeles for each sample were identified, and genotypes were compared with those entered Table 3.6. Allele sizes of Epichloe endophyte microsatellite loci determined by ABI Prism automatic analysis. Allele size (ntu) Taxonomic group Isolate Primer Set I Primer Set I I B4 B6 B9 B l O B l l B l B 2 B 5 B 9 1 00.0 1 72.4 1 69 . 3 + 1 72.2 LpTG-2 Lp l 247.4+ 1 1 9 . 8 2 9 3 . 7 1 60.8 1 05 . 2 1 1 9 . 8 1 87.6+ 1 78 . 1 + 1 87 . 6 1 00.0 1 72 . 4 1 69 . 3 1 72 . 2 Lp2 247.3+ 1 1 9 . 6 2 9 3 . 7 1 60.8 1 05 . 4 1 20.0 1 87 . 5 1 78 . 1 + 1 87 .4 LpTG- I Lp5 1 00. 0 1 87.7 247. 3+ 1 78. 1 + 1 77.0+ 293.7 1 5 8. 4 Lp6 1 00. 1 1 87.6 247.3+ 1 78 . 0+ 1 76.9+ 293.7 1 5 8.2 Lp 1 3 1 00.0 1 87.7 247.4+ 1 7 8. 0+ 1 77 . 0+ 293 . 7 1 5 8.2 Lp 1 9 1 00 . 0 1 87.6 247.4+ 1 7 8.0 1 80.8 293.7 1 58 . 3 Lp20 1 00.0 1 87 . 6 247.4+ 1 7 8 . 1 + 1 80.9 2 9 3 . 7 1 58 . 3 Lp2 1 1 00.0 1 87 . 6 247 .4+ 1 7 8 . 1 1 80.9 293.7 1 58.2 Lp7 1 00 . 0 1 87.6+ 247.4+ 1 78 . 1 1 88 . 9 293 . 7 1 5 8.2 Lp l 4 1 00.0 1 86.6+ 27 1 .2+ 1 6 3 . 5 1 3 2 . 8 293.7 1 8 1 .9 Lp9 1 00. 1 1 87.6 248.3+ 1 7 8 .0+ 1 8 8.8+ 293.7 1 59.2 A R l 1 00.0 1 87.6 246.4+ 1 78.0+ 1 49 . 8 2 9 3 . 7 1 57 . 3 A R I WF 1 00.0 1 87.6 246.4+ 1 7 8 . 1 + 1 49.8+ 293.7 1 57 . 3 A R I CY 100.0 1 87.5 246.3+ 1 78.0+ 1 49 . 8 293.7 1 57.4 EndoSafe 1 00.0 1 87 . 7 247.4+ 1 7 8 .0+ 24 1 .4+ 2 9 3 . 7 1 58.0+ ,..... VI � Table 3.6. continued Allele size (ntu) Taxonomic group Isolate Primer Set I Primer Set I I B 4 B 6 B 9 B l O B l l B l B2 B5 B9 LpTG- l cont. A950 1 1 00.3 1 87.7+ 247.5+ 1 78 . 1 + 1 77.0+ 293.7 1 5 8 . 2 CBS232.84 1 00.2 1 87.6 247.3+ 1 78. 1 + 1 7 6.9+ 2 9 3 . 7 1 58 . 3 J 60.7+ 1 00.0 1 49 . 9 1 80.0 FaTG- l Tf5 1 93 . 7 272. 1 + 1 7 2.2+ 293.6 74.8+ 1 03.9 1 92.7+ 1 82 . 9 1 86.7+ 1 55.0+ 1 00.0 1 72.9+ 1 80.0 Tf27 1 96.8 272.2+ 1 63 . 5+ 293.6 75.0+ 1 03.2 1 97.0+ 1 83 .0 1 80.9+ 175. 1 + 1 28 . 7 1 80.0 Tf28 1 00.0 1 93 . 8 272.2+ 1 80.9+ 293.7 74.8+ 1 92.7+ 1 82 . 9 1 83 . 8+ 1 55.0+ 99.9 1 73 . 0+ 1 80.0 Tf35 1 96 . 9 272.2 1 63.5+ 293.6 7 5 . 1 1 03.7 1 9 7.0+ 1 82 . 9 1 8 1 . 0+ 1 00.3 1 63 . 5+ 1 8 1.1 1 80 . 0 AR542 1 93 . 8 272.2+ 293.6 7 5 . 0 1 04.0 1 72 . 3 + 1 92.7+ 1 83 . 0 1 72 . 1 + 1 28 . 7+ 1 8 1 . 8 FaTG-2 Tfl 3 1 00.5 1 8 7.6 271. 3+ 293.7 1 86.8+ 1 4 1 .0 1 86.0 1 7 2 . 1 + 1 28 . 8+ 1 58 . 3 Tf1 5 1 0 1 .0 1 86.6+ 247.5+ 293.6 1 78 . 0+ 1 65 . 2+ 1 8 7 . 1 ...... IJI VI Table 3.6. continued Allele size (ntu) Taxonomic group Isolate Primer Set I Primer Set Il B 4 B 6 B9 B l O B 1 1 B l B 2 B 5 B 9 1 72 . 1 + 1 28 . 6 1 58 .2 FaTG-2 cont. Tf20 1 00. 3 1 87 . 6 247.4+ 2 93 . 7 1 78 . 0+ 1 84.8+ 1 87 . 1 1 72.2+ FaTG-3 Tf1 6 1 20.4 1 7 1 .6+ 1 28 . 6 2 9 5 . 6 1 44.3 1 8 6. 2 1 80.9+ 1 72.2+ Tf18 1 20.3 1 72.6 1 28.6 2 9 1 . 8 1 44.4 1 86.0 1 80.9+ 1 72 . 2 AR572 1 20.3 1 72.6+ 1 28.6 2 9 1 .7 1 44.4 1 86. 1 1 8 1 . 0+ 1 72. 1 + AR577 1 1 9 . 9 1 7 1 . 8+ 1 28 .6 295.6 1 44.4 1 86.0 1 80.9+ 84.8 1 60.7+ 1 62 . 8+ FpTG- I Fp l 1 78.7 1 20.7 2 9 8 . 5 75. 1 1 04.4 1 95 . 8+ 1 79.3+ 84.7 1 60.6 1 63 . 0 Fp2 1 78 . 7 1 20.7 298.4 7 5 . 1 1 03 . 7 1 95.7+ 1 78 . 9 84.7 1 72. 1 + 1 63 . 2+ Fp4 1 78.7 1 20.7 295.6 74.4+ 1 03 . 9 1 92.7 1 79.7 E. typhina E8 1 20 . 5 1 72.5+ 1 7 8 . 1 2 9 1 . 8 1 62.7 1 04.4 E348* 1 24.6 1 70.5 1 90.8 1 27 . 0 1 47.9+ 90.7 E358 * 1 1 6 0 1 70.5 1 82.0 1 06.8 1 54.3+ 80.2+ E425* 1 1 5 . 8 1 79.0 1 30.6+ 80.2+ ..... VI 0\ Table 3.6. continued Allele size (ntu) Taxonomic group Isolate Primer Set I Primer Set I f B 4 B 6 B 9 B l O B 1 1 B I B 2 B 5 B 9 1 78 . 3 E. typhina cont. E428* 1 1 5 . 9 1 60.6+ 80.0+ 1 84.6 E43 1 * 1 1 5 .9 1 95 . 9 1 30.6+ 80.2+ E470* 1 1 6.0 1 70.2 1 82.0 1 52. 1+ 80. 1 + E505* 1 04.2 1 75 . 4 1 39.0+ E 1 02 1 * 9 9 . 6 1 79.0 1 43 . 8+ 70.7+ 179 . 1 + E 1 022* 99.5 1 78 . 5 143.8+ 70.7+ 1 79 . 1 + E 1 03 2 * 1 02 . 3 1 84.0 1 52.0+ 7 0 . 8+ 1 65 . 0+ E I 03 3 * 1 02.4 1 83 . 9 1 52 . 0+ 70.7+ 1 64.9+ E2466* 1 1 6.0 1 70.2 1 75.9 1 5 6.5+ 80.3+ E. clarkii E422* 1 1 6.0 1 84 . 6 1 5 8.5+ 80.0+ E426* 1 1 6. 1 1 84.5 1 5 8.5+ 8 1 .0+ E427* 1 1 6.0 1 85 . 3 1 60.5+ 80.0+ E. feslucae E28* 95.4 1 86.0 269.6 1 76.3 1 1 4.4 3 0 1 .4+ 1 82 . 4 E32* 95.5 1 86.0 269 . 6 1 70. 5 1 64.5 305.2 1 82 . 2+ E1 86* 9 5 . 5 1 85 . 8 269 . 8 1 76. 1 1 47.2 3 0 1 .4+ 1 82 . 3 E 1 89 * 9 5 . 4 1 86.0 269.7 1 70.2 1 64.4 3 04.3+ 1 82. 1 E434* 9 5 .4 1 85 . 7 270.0 1 76.0 1 47.2 304.4+ 1 82.2 Fr ! 1 00.0 1 86.6 272. 1 + 1 72. 1 + 1 65 . 3 304.9 1 82.9 Frc5 1 00.5 1 92 . 5+ 273.2+ 1 78 . 0+ 1 69. 1 + 3 0 1 . 8 1 84.0 ..... Ul -.l Table 3.6. continued Allele size (ntu) Taxonomic group Isolate Primer Set I Primer Set I I B 4 B 6 B 9 B I 0 B l l B l B 2 B S B 9 E. festucae cont. Frc7 1 00.0 1 87.7 272. 1 + 1 7 8.0+ 1 49.9 3 0 1 . 8 1 8 1 .9 FIT 1 1 00.0 1 75.7 272. 1 + 1 66.3+ 1 1 6. 8 307.7 1 82 . 6 Fg l 1 00.0 1 87.7 279.0+ 1 95.7+ 1 28 .7 296.4 1 89 . 7 296.4 1 82 . 9 FI l 1 00 . 0 1 87 . S 272.2+ 1 78.0+ 1 1 6.8+ 302.0 1 90. 1 E. elymi E56* 7 9 . 1 1 77.4 1 92 . 8 1 26. 1 286.7 1 70 . 2 E 1 84* 7 9 . 3 1 7 7 . 1 1 92 . 6 1 26 . 5 285.7+ 1 70. J E. amarillans ES2* 1 77 . 2 1 67 . 6 1 3 8 . 3 297.7 E57* 1 80.2 1 8 1 . 3 1 38 . 2 298.6 E. baconii E42 1 * J 69 . 5 1 3 1 .0 1 82 . 9 + E424* 1 70.0 1 3 1 .0 1 78 . 8 + E 1 03 J * 209. 1 1 67 . 5 1 47 . 1 288.7+ 1 6 1 . 8 + E. bromicola E50 1 * 1 7 7 . 2 1 89 . 7 1 1 0 . 8 3 1 0.2 1 62 . 7 + E502* 1 77. 3 1 89 . 6 1 1 0. 7 292.4 1 62.7+ E798* 80.0 1 77 . 2 1 84.0 1 1 0 . 5 2 9 1 .S+ 1 6 1 .6+ E799* 79.9 1 77 . 3 1 84.0 1 1 0.5 292.4 1 6 1 .4+ E. sylvatica E354* 1 07 . 9 1 77 . 0 1 70 . 3 22 1 . 8 + 70.2+ 1 79 . 3 + E503* 99.2 1 72 . 9 1 37 . 1 + 70.2+ ....... VI 00 Table 3.6. continued Allele size (ntu) Taxonomic group Isolate Primer Set I Primer Set II B4 B6 B9 B W B l l B 1 B 2 B 5 B 9 E. glyceriae E277* 78. 1 1 54.7 1 1 8.4 294.9+ E2772 * 78.2 1 54 . 6 1 1 8.4 294.9+ E. brachyelytri E I 040* 90.5 1 66.8 1 34.2 3 1 4. 9 1 7 3 . 1 E I 04S * 90.4 1 66. 7 1 34. 1 305.9 1 73 . 2+ E I 046* 90.5 1 66. 8 1 34.2 30S.8 1 73 .2+ 1 84.0 1 28.6 287.2 1 62. 1 Neotyphodium spp. HaB 8 3 . 8 1 7 8 . 8 73. 1 1 95 . 7 1 4 1 .0 293.6 1 78.0 1 6 1 .0 1 72. 1 Hd l 1 1 1 . 5 1 1 2.9 309.4 1 55 . 3 74.8 1 73 . 7 1 9 2 . 7 A R 1 7 1 00. 1 1 86 . 6+ 278.9+ 1 83 . 8+ 1 53 . 8 296.4 1 89.9 + indicates that a peak, approximately 1 ntu larger than that given in the table, was also detected * indicates that the automatic analysis of this fingerprint was performed on an ABI Prism 3 10 genetic analyser, whereas all other fingerprints were analysed on an ABI Prism 377 - indicates that no product was detected at thi s l ocus 160 in the database. In al l cases a con"ect match was found with a taxonomic grouping and i sozyme phenotype (if available) that had been previously determined for that i solate (Lp l , Lp5 , Lp 14, ARl , Tfl 3 , Tfl5 , Tf27, AR542, Tfl6, Fl l , Fpl , and AR17) using a combination of morphological , biochemical , and genetic criteria (results not shown) . 3.2.3.3 Analysis of a DNA Deletion in B l Amplified Alleles As reported in Section 3 .2 .2 .7 . 1 , representative isolates of all ten Epichloe spp. were screened through ML-PCR assays of microsatellite loci B l , B2, B5 , B7, B8, and B9. For a representative sample of isolates that did not amplify using the standard ML-PCR conditions, further work was done by adjusting annealing temperature from 65°C, to 60°C and 55°C (Section 2 . 1 3 .2), to lower the annealing stringency. Products were only obtained from isolates when the annealing temperature was reduced to 55°C for the B 1 assay. Therefore, all i solates that did not amplify at 65°C for locus B 1 (Fig . 3 .26A), E. typhina, E. clarkii, E. baconii, as well as E. sylvatica, were screened with the annealing temperature at 55°C, and these products are shown in Figure 3 .30A. Several i solates (E358, E425, E428, E43 1 , E470, E505, E2466, E354, and E503) had B l alleles of around 240 bp, a result that was also observed for the amplification of E. sylvatica isolates under standard PCR conditions (Fig . 3 .26A). This i s significantly smaller than the product size expected at the B 1 locus (around 300 bp), and the difference is greater than one would expect solely from variation in the number of microsatellite repeats . Therefore, the Bl products of E354, E503 , E358 , E502, E8 , and Lp6 were fully sequenced (Section 2 . 1 5), and aligned (Section 2. 1 8. 1 ) with the Oliginal B l sequence from E. bromicola (Groppe et aI . , 1 995), to characterise the size reduction. The sequence alignment, shown in Figure 3 . 3 1 , reveals a 73 bp deletion relative to the original E. bromicola i solate NF62A, for isolates E354, E503, and E358 . I t was therefore assumed that all B 1 alleles around 240 bp long contained the 73 bp deletion. The origin and evolutionary distribution of tbe deletion product was analysed by mapping the occurrence of deletion to a phylogeny of European Epichloe spp. which was based on tub2 sequence data (Figure 3 .30B). The deletion was confined to a clade 1 6 1 Fig. 3 . 30. Distribution of a 73 bp B 1 deletion product across Epichloe species (A) Screening ML-PCR B 1 products from Epichloe typhina, E. clarkii, E. baconii, and E. sylvatica isolates for deletion product. Agarose gel (3%) separation of Epichloe isolates o amplified in the B 1 ML-PCR assay with the annealing temperature reduced to 55C. Reactions contained genomic DNA from isolates: E8 (lane 1 ), E348 (lane 2), E358 (lane 3), E425 (lane 4), E428 (lane 5), E43 1 (lane 6), E469 (lane 7), E470 (lane 8), E505 (lane 9), E 1 02 1 (lane 1 0), E 1 022 (lane 1 1 ) , E 1 032 (lane 1 2), E 1 033 (lane 1 3), E2466 (lane 1 4), E422 (lane 1 5), E426 (lane 1 6), E427 (lane 1 7), E42 1 (lane 1 8), E424 (lane 1 9), E354 (lane 20), E503 (lane 2 1 ), and Lp6 (lane 22). Lane 23 contained PCR negative control and M contains 1 23 bp ladder. (B) Distribution of the 73 bp B 1 deletion product with respect to the evolution of European Epichloe species. A strict consensus cladogram from the six most parsimonious trees derived from a branch-and-bound search on tub2 intron sequences (Leuchtmann and Schardl, 1 998) is shown, with numbers on branches indicating the percentage of trees possessing the branch in 450 bootstrap replications. Isolates have been marked with a d if it contains the 73 bp deletion, f if it does not, and n if no B 1 product was amplified. All other isolates were not tested. A 369 bp - 246 bp - 1 23 bp B M 2 3 4 5 6 7 8 9 1 0 1 1 1 2 M M 1 3 1 4 1 5 1 6 1 7 1 8 1 9 20 21 22 23 M 69 97 95 E1021 f E 1022 f E430 E8 f r - - - - - - · : E422 n : : E426 d .. - - I : !'��� 2_ J E2461 E425 d E429 E431 d E469 E348 f E470 d E358 d E428 d E2466 d E505 d E503 d ] E354 d E 189 f ] E434 f E28 f E248 J E421 n E424 n E1031 f E. clarkii MPI E. typhina MPI E. sylvatica MPVII E. festucae MPII E. baconii MPV E501 f E502 f J E- MPI E798 f E799 f Claviceps fusiformis 1 62 Fig. 3 . 3 1 . Sequence alignment of B I -amplified ML-PCR products of Epichloe endophyte isolates 1 63 The DNA sequences of B 1 amplified products from E. typhina i solates E8 and E358, E. bromicola isolates E502 and NF62A, E. sylvatica isolates E354 and E503, and N. lolii isolate Lp6 were aligned using the PILEUP programme of the Wisconsin Package with the gap penalty set to one and the gap extension penalty set to zero. Gaps C . ) were introduced to facilitate sequence alignment. The CAAG)n microsatellite sequence i s highlighted in blue. 1 E3 54 CAATA . CGTC E503 CAATA . CGTC E3 5 8 CAATA . CGTC E502 CAATA . CGTC NF62A CAATA . CGTC Lp6 CAATATCGTC E8 CAATA . CGTC 5 1 E3 54 .. .. .. .. .. .. .. .. .. .. E503 .. .. .. .. .. .. .. .. .. .. E3 5 8 .. .. .. .. .. .. .. .. .. .. E502 GCTCTAGACC NF62A GCTCTAGACC Lp6 GCTCTTTATC E8 GCTCTTGATC 1 0 1 E3 54 . . . . . . . . . T E503 . . . . . . . . . T E3 5 8 . . . . . . . . . T E502 AGCTCGTTTT NF6 2A AGCTCGTTTT Lp6 AGCTCGTTTT E8 AGCTCGTTTT 1 5 1 E3 5 4 AGAGTAAACC E503 AGAGTAAAAC E3 5 8 AGAGTGAAAC E502 AGAGTTAAAG NF62A AGAGTTAAAG Lp6 AGAGTAA . . . E8 AGAGTAA . . . AGCTAGGAAT AGCTAGGAAT AGCTAGGAAT AGCTAGGAAT AGCTAGGAAT AGCTAGGAAT AGCTAGGAAT .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. TGATGACAGA TGATGACAGA TGATGACAGA TGATGACAGA GTTGCATACT GTTGCATACT GTCGCGTACT GTTGCATACT GTTGCATACT GTTGCATACT GTCGCATACT .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. TTCCATGGTC TTCCATGGTC TTCCATCGTC TTCCATCGTC TGA . . . . . . . TGA . . . . . . . TGA . . . . . . . TGAT . TT . . C TGAT . TT . . C TGAT . TT . . C TGATATTATC .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. GAACCAACGC GAACAAACGC GAACCAACGC GAACCAACGC ATGTCCACGA AAAGAAGAAA AGACAATATG ATGTCCACGA AAAGAAGAAA AGACACTATG ATGTCCGCGA AAAGAAGGAA AGAAACTATG ATGTCGACGA AAAGAAGAAA AGAAACTATG ATGTCGACGA AAAGAAGAAA AGAAACTATG ATGTCCACGA AAAGAAGAAG AGAAACTATG ATGTCCATGA AAAGAAAAAA AGAAACTATG AAGAACAAGA AGAAGAAGAA G . . . . . . . . . AAGAAGAAGA AGAAGAAGAA GAAGAAGAAG AAGAAGAAGA A . . . . . . . . . .. .. .. .. .. .. .. .. .. .. AAGAAGAAGA AGAAGAAGAA G . . . . . . . . . AAGAAGAAGA AGAAGAAGAA GAAGAAGAAG 5 0 .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. CATCCATCTG CATCCATCTG CATCCGTCTG CATCG . . . TG 1 0 0 .. .. .. .. .. .. .. .. .. .. .. .. .. " .. " .. .. .. .. .. .. .. .. .. .. .. .. .. .. AATCGCCCCA GATCGCCCCA GATCGCCCCA AATCTCCCCA 1 5 0 CAACGTCTGA CAACGTCTGA CATCATCTGA CAACGTCTGA CAACGTCTGA CAACGTCTGA CAACGTCTGA 2 0 0 .. .. .. .. .. .. .. .. .. .. AAGAAGAAGA . AGAAAAATG .. .. .. .. .. .. .. .. .. .. AAGAAGAAGA .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . . . . TAAA A AGGAAGAAGA .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . . . . . . ACCA 1 64 2 0 1 2 5 0 E3 54 .. . .. .. .. .. .. . . . · .AAA GACG GCCCAACATC AAGATTCATT CGTCAACCTA E503 AGAAGAAG . . · .AAA GACG GCCCAACATC AAGATTCATT CGTCAACCTA E3 5 8 AGAAGAAG . . · . AAAAGACG GCCCAACATC AAGATTCATT CGTCAACCTA E502 .. .. .. .. .. .. . .. .. .. . AAAGAGACG GCCCAACATC AAGATTCATT CGTCAACCTA NF62A AGAAGAAGAA GAAAGAGACG GCCCAACATC AAGATTCATT CGTCAACCTA Lp6 AGAAGAAGAA · . AAAAGACG GCCCAACATC AAGATTCATT CGCCAACCTA E8 AGAAGAAGAA GAAA . AGACG GCCCAACATC AAGATTCATT CGTCAACCTA 2 5 1 3 0 0 E3 54 TAGCCTTGAT CCAGGCCAAA TGGGTTCTCT CTCATATACC TACTAACATC E503 TAGCCTTGAT CCAGGCCAAA TGGGTTCTCT CTCATATACC TACTAACATC E3 5 8 TAGCCTTGAT CTAGGGCAAA TGGGTTCTCT CTCATATACC TACTAACATC E502 TAGCCCTGGT CCAGACCAAA TGGGTTCTCT CTCATATACC TACTAACATC NF62A TAGCCCTGGT CCAGACCAAA TGGGTTCTCT CTCATATACC TACTAACATC Lp6 TAGCCTTGAT CCAGACCAAA TGGGTTCTCT CTCATATACC TACTAACATC E8 TAGCCTTGAT CCAGGGCAAA TGGGTTCTCT CTCATATACC TACTAACATC 3 0 1 3 2 5 E3 54 GGCCTGATAG CAAAGTTGAT TCAGG E503 GGCCTGATAG CAANN TGAT TCAGG E3 5 8 GGCCTGATAG CAAAGTTGAT TCAGG E502 CGCCTGATAG CAAAGTTGAT TCAGG NF62A CGCCTGATAG CAAAGTTGAT TCAGG Lp6 CGCCTGATAG CAAAGTTGAT TCAGG E8 GGCCTGATAG CAAAGTTGAT TCAGG 1 65 containing E. typhina, E. clarkii, and E. sylvatica, and more specifical ly, was found in both of the E. sylvatica isolates tested, two of the three E. clarkii isolates tested, and dispersed throughout E. typhina. 3.2.3.4 Inheritance of Microsatellite Alle)es in an E. jestucae Sexual Cross Fungal genomic DNA from 42 randomly selected progeny of a sexual cross between E. festucae isolates E 1 89 and E434, were made available for the analysis of the segregation of microsatel1i te alleles (H. H. Wilkinson). In this cross, E 1 89 (Primer Set I genotype: 95 .4 (B4), 1 86.0 (B6), 269.7 (B9), l 70.2 (B I0) , and 1 64.4 ntu (B 1 1 ) was used as the stromal parent, and E434 (Primer Set I genotype: 95 .4 (B4), 1 85 .7 (B6), 270.0 (B9), 1 76.0 (B lO) , and 147 .2 ntu (B l l )) was used as the spermatial parent in an artificial1y induced mating. Genomic DNA from each progeny, prepared using the method of Byrd et al . ( 1990), was amplified with Primer Set I and analysed on an ABI Prism 3 10 Genetic Analyser. As expected, the progeny contained B4, B6, and B9 alleles identical to the parental types. However, the genotypes at loci B 1 0 and B 1 1 were variable, and all four expected genotypes at these loci were observed, as summarised in Figure 3 .32 . Nine progeny had identical genotypes to parental i solate E 1 89 , and 1 9 progeny had identical genotypes to parental i solate E434. Both possible recombinant genotypes were represented by seven progeny each. To show that the inheritance of the microsatellite genotypes conformed to Mendel's laws of segregation and independent assortment, data was tested with the chi-square (X2) test for goodness-of-fi t between the observed and expected results. Mendel 's first law , the principle of segregation, states that the two members of a gene pair (alleles) segregate from each other in the formation of gametes (Russell , 1 990). Under the null hypothesis that microsatellite alleles segregate at the B 10 and B 1 1 loci , it is expected that the allele frequencies at each locus conform to a 1 : 1 ratio in the progeny, P generation genotypes F 1 generation frequencies of genotypes x 1 47 1 64 subtotal E 1 89 1 70 1 64 1 70 7 9 1 6 x E434 1 76 1 47 1 76 1 9 7 26 1 66 subtotal 26 1 6 42 Fig. 3 . 3 2 . Segregation of B 10 and B 1 1 rnicrosatellite alleles in a sexual E. festucae cross E. festucae isolate E 1 89 (stromal parent) was crossed with E434 (spermatial parent) and 42 progeny were randomly selected. Parental and progeny isolates were rnicrosatellite fingerprinted using Primer Set I, and allele polymorphisms were observed at loci B l Q and B 1 1 . Genotypes are shown with B 1 0 allele sizes are indicated in blue and B 1 1 allele sizes are indicated in green. Allele sizes are given in nucleotide units (ntu) . Frequency of progeny are given in box below. 1 67 which represents frequencies of 2 1 each. The observed allele frequencies at each locus are 1 6 and 26 (Fig. 3 .32), but thi s difference is not considered statistically significant at the 5% level using the chi-square test (see Appendix 4). Thus the null hypothesis that the microsatellite alleles are segregating in accordance to Mendel's first law is supported. Mendel 's second law, the principle of independent assortment, states that the genes for different traits assort independently of one another dming gametogenesis (Russel l , 1 990) . Under the null hypothesis that microsatellite alleles of loci B 1 0 and B 1 1 assort independently of one another, the expected frequencies of the four possible progeny genotypes in the haploid dihyblid cross should fit a 1 : 1 : 1 : 1 ratio, which is a frequency of 1 0. 5 each. The observed data (Fig. 3 .32) clearly differ from these expected values, and this difference is stati stically significant at the 5% level using the chi-square test (see Appendix 4). Thus the null hypothesis is rejected. The likely explanation for non­ independent assortment is that loci B 10 and B 1 1 are l inked in E. Jestucae. If this is indeed the case, the genetic map distance between these loci is equal to the percentage of recombinant genotypes, which i s 33.3 map units (mu). 3.3 DISCUSSION 1 68 The molecular c loning of a set of microsatelli te loci from Epichloe grass endophytes and the development of an automated PCR fingerprinting method based on sequences at those loci , that distinguishes the different groups of isolates to the level of their i sozyme phenotype grouping is described. Surprisingly, there are few reports to date on the cloning of microsatellites from fungi (Geistlinger et aL, 1 997; Groppe et aI. , 1 995; Longato and Bonfante, 1 997) despite their widespread use for kinship, population structure and mapping studies in plants and animals, though their popularity has been gaining in the past year (Bart-Delabesse et aI . , 1998; McKay et aI . , 1 999). We are aware of only one other such study that uti l ises microsatell ite markers as the basis of a fungal typing method with automated analysis, which i s for Aspergillus fumigatus i solate identification (Bart-Delabesse et a1 . , 1 998). The assay described in this study has been used in several applications, including the identification of endophytes in infected grass tissue, and tracking the inheritance of microsatell ite alle1es in an E. festucae genetic cross . The genotype data generated from the assay has also been shown to support evolutionary relationships between Epichloe endophyte species and taxonomic groups generated by molecular methods, and therefore may be used to supplement such studies. 3.3.1 AN ASSESSMENT OF THE peR-BASED FINGERPRINTING METHODS The ML-PCR method offers several advantages over RAPD and MP-PCR for endophyte identification, and it is the method of choice for further assay development. The main benefit of ML-PCR is that the high specificity of the primers gives the assay potential to be used on in planta samples, therefore eliminating the need to initially i solate the endophyte from the grass , as required with the other two methods. Another appeal of the ML-PCR method is that the identity of the amplified bands is known, and the 169 microsatelli te locus copy number can be readil y determined. This information is particularly useful for Epichloe endophyte i solates where many of the asexual endophytes are of interspecific hybrid origin, and the relationship between the hybrid and the parents can be seen . A prime example is i l lustrated by ML-PCR amplification of the B 1 microsatellite locus in Fig. 3 .SB, where Lp 1 of the taxonomic group LpTG-2 shows two bands, whose sizes correspond to those of its parents, N. lolii and E. typhina. In contrast, RAPD-PCR and MP-PCR generates fingerprint patterns where the identity of the bands are unknown. Analysis of fingerprints of this type is laborious, and requires that standard samples are included in each run to make scoring consistent between runs. A concern in developing the ML-PCR assay is whether there would be enough polymorphism at microsatellite loci to identify i solates to the required level. Fingerprints produced by MP-PCR were able to distinguish isolates intraspecifically, and the assay was reasonably reproducible. By contrast, RAPD-PCR did not show such high variabil ity in fingerprint patterns, the reproducibility between runs was not rel iable, and contamination of runs was occasionally experienced which is probably a consequence of using arbitrary primers at such low annealing specifici ty . Overall , RAPD-PCR did not appear to be a rel iable method for fingerprinting endophytes so further development of this method was discontinued. Consequently, it was decided to further develop the ML-PCR assay to include several informative polymorphic microsatellite loci . It was hoped that these loci could be amplified in a multiplex reaction using dye-labelled primers so that fingerprints could be analysed precisely using laser detection technology, such as that associated with the Applied Biosystems automatic sequencers. 3.3.2 DEVELOPMENT OF A PCR-BASED MICROSATELLITE FINGERPRINTING ASSAY 3.3.2.1 The Frequency of Microsatellites in Fungi 1 70 Based on Southern analysis, an average frequency of one (CA)n or (GA)n locus per 2 Mb of DNA, and one (CAA)n, (GAA)n, or (ATC)n locus per 2 .5 Mb of the Lp l genome was found. By comparison with similar estimates for plants, (CA)n and (GA)n frequencies range from 1 1270 kb in wheat (Roder et aI . , 1 995), to 1 1 1 .2 Mb in tomato (Broun and Tanksley, 1 996). The frequency of the same classes of microsatellites in mammals is even greater (Lagercrantz et aI . , 1 993). However, these figures should be regarded as a minimum, since a single hybridising band in a Southern may correspond to more than one locus and, l ike the alternative approach of determining the number of positives in a library screen, they will also vary depending on the stringency of the hybridisation used. The effect of the latter was evident here in that not all of the positive clones identified had microsatellite sequences that were perfect matches with the probes used. For example, the B5 locus (CATCTCATCA)s and the B9 locus (GAGAG)zC(GAGGA)z only partially match the sequences of the tri- and di- nucleotide repeat probes that were used for the isolation of these sequences. It would also appear that the partial libraries made in M13mp19 were not as representative of the diversity of loci as the Southern blots had indicated. For example, the Southern analysis of Lp1 genomic DNA indicated that there were at least 20 (CA)n and (GA)11 microsatel1 ites in the size range of 100 - 1000 bp, used to construct the library, yet only three different microsatellite clone types were identified from the 24 c lones screened. This was probably a consequence of preferential l igation of the smaller inserts into the M13mp l9 vector, as a1l clones that were analysed had inserts of between 100 and 300 bp in size. In retrospect, small insert library construction using a random DNA fractionation method, such as shearing or sonication, would be favoured to generate inserts. Therefore, most, if not all of the genome would be represented in the small fragment size range. More accurate determinations of both the frequency and 1 7 1 diversity of microsatellites i n fungi and other organisms wi I 1 emerge from analysis of DNA sequence data generated from either partial or complete genome sequencing initiatives, such as that for Saccharomyces cerevisiae (Field and Wills , 1998; Richard and Dujon, 1 996). A search by Groppe et al . ( 1 995) of fungal DNA sequences held in the GenBank and EMBL databases for all possible mono-, di-, and trinucleotide motifs (>20 bp) showed that the predominant microsatellite type found in fungi is (AT)n; representing 42% of the microsatellites identified. The other most frequently occurring microsatellite types were in decreasing order: (A)n, (AAT)n, (AAC)n, (AAG)n, (ATC)n, (AC)n, and (AG)n. Screening for the abundant (AT)n repeat class was not carried out in this study since these sequences are self- complementary, making it very difficult to i solate such sequences by solution hybridisation methodology (Kostia et aI . , 1 995 ; Lagercrantz et aI . , 1 993). 3.3.2.2 An Overview of the ML-PCR Assays As shown in Figure 3 .7 the ten microsatellite loci isolated contain a variety of sequence motifs ranging from repeats of dinucleotides to decanucleotides . To determine the degree of length polymorphism associated with these microsatell ites, and those isolated by Groppe et al. ( 1 995) and Dobson ( 1997), PCR assays were initially carried out with DNA from a collection of Epichloe endophytes, predominantly representative of asexual isolates from the forage grasses: perennial ryegrass, tall fescue, and meadow fescue. Further screening of representative isolates of the ten Epichloe spp. were carried out, and in total, fingerprinting assays util i sing eight of the loci in two sets of primers, designated Primer Sets I and II, were devised. The degree of length polymorphism at each locus was estimated by the number of alleles observed, and varied between eight and 25 alleles at each locus for the 80 isolates listed in Table 3 .6 . Allele numbers were recorded after the approximate standardisation of alleles for the ABI 377 DNA SequenceI', using the average offsets calculated in Tables 3 .4 and 3.5 to convert data that was analysed on an ABI 3 1O. The most 172 polymorphic loci were B I0 and B l l with 25 alleles each, followed by B l and B9 (using primer pair B9. 1 and B9.2) with 20 all eles each, B2 and B6 with 14 alleles each, B4 with 1 2 alleles, and lastly B5 and B9 (using primer pair B9. 1 and B9.4) with eight alleles each. The size ranges of alleles observed in the SO isolates varied between 29 ntu (B 1 ) and 1 3 3 ntu (B l l ) with an average range of 5 6 ntu. No products were amplified from i solates in several cases and this may be due to either: the absence of the microsatellite locus or at least one of i ts priming sites, or sufficient sequence divergence at the priming site so that the primer cannot anneal at the given stringency. Nonetheless, the non­ amplification of alleles provides useful insights into the evolutionary relationships between groups of endophytes, and this is discussed further in Section 3 .3 .3 . 1 . Locus B l l , serendipitously identified downstream of the HMG CoA reductase gene from strain Lp1 9 (Dobson, 1 997), together with locus B lO, are the most polymorphic microsatellite loci in this study. As HMG CoA reductase is an essential enzyme in the isoprenoid metabolism pathway, the transcription level of the gene encoding it i s relatively high. Simple repeat sequence tracts are destabilised by high rates of transcription (Wierdl et aI. , 1 996), so this could partiall y explain the high degree of polymorphism observed at the B 1 1 locus. There is also good evidence to supp0I1 the notion that B lO and B l l are l inked loci (Section 3 .2 .3 .4), thus B I0 may also be found in an area of high transcriptional activity. The proximity to highly transcribed regions of the genome of the remaining microsatellite loci in this study are not known, though tract stability i s also dependent on its length and 'purity'. Thi s i s presumably because the opportunity for a stable misalignment is lower in shorter tracts, and tracts interspersed with non-micro satellite sequences (Wierdl et aI. , 1 996). Apart from locus B l l « GACA) 1 8 in Lp I 9) , which is both relatively long and continuous, there are no obvious relationships between the length and continuity of the microsatelli te tract, and the degree of polymorphism at the other loci . In fact, most loci contain imperfect repeats (B4, B6, B7, B9, and B IO) or are composed of compound microsatellites (BS, B9, and B IO) in the i solates they were identi fied from. 173 An interesting observation that was made while screening with locus B 1 1 is that alleles amplified from Epichloe spp. were generally shorter than those amplified from Neotyphodium i solates. The relevance of this is not known, and such a pattem was not seen in the other loci . No microsatellite PCR products were obtained with any of the primer pairs with the outgroups Claviceps purpurea and Echinodothis tuber�formis, which are both from the same family as the Epichloe endophytes (Clavicipitaceae). This shows that these primers are highly specific for the Epichloe endophytes, demonstrating that the assay is not influenced by the presence of unrelated fungal endophytes or contaminating fungal species in the plant tissue. 3.3.2.3 Automated Fingerprint Analysis While the microsatellite allele data generated by manual analysis of autoradiographs of 33p labelled products was largely concordant with that generated by automated analysis of fluorescently label led products, the latter method was superior in many respects. Automated analysis allows precise PCR product size estimation over a wide size range in a single run, with a sensitivity that reduces the amount of product required for analysis . This becomes particularly important in the detection of endophytes in planta, where the fungal biomass can be very low. Dye-labelled primers and products are able to be stored for long periods of time without signal decay and are safer to use than radioisotopes. Automated analysis is also faster since data is collected while the products are being electophoresed, thus eliminating the need for gel drying, film exposure, and development. The complexity of the banding pattem from a single product is reduced since only one strand of the product is l abelled; thi s greatly simplifies the analysis of hybrid endophytes which have multiple alleles of simil ar size for some loci . The three­ colour dye system allows products from different loci to be identified simultaneously, even if their peaks overlap, as shown in Figure 3 .29. This cannot be done with manual methods, and separate gels are often required to analyse individual loci. Finally, the data are generated as the estimated size of the DNA fragment, enabling comparisons to be 174 readil y made between l aboratories, a task which i s difficult to achieve with other typing methods, such as restriction fragment length polymorphism (RFLP) or RAPD-PCR analysi s . Thus, microsatellite fingerprints may be generated at independent laboratories and submitted into a common database. However, as the estimated size of the fragment is dependent on its electrophoretic mobility, using a different separating matrix may alter the mobil i ty of the DNA fragment, thus resulting in different size calling. Indeed this was the case when identical endophyte fingerprints were analysed with separation in both a non-crosslinked network polymer and a denaturing polyacrylamide gel. Unfortunately universal offsets could not be precisely determined for the endophytes as a group as they are such a diverse assemblage, though the average offsets calculated in this study may be used as an approximate guideline. Thus, when analysing fingerprints it is important to specify which model of genetic analyser was used when reporting its genotype. A problem that was encountered when analysing fingerprints from the automated system was the inconsistent presence of split peak signals, 1 nt apart; this was a consequence of the inherent property of Taq polymerase to occasionally add a deoxyadenylate to the 3 ' terminus of the PCR product. The primer dependence of this property of Taq polymerase (Brownstein et aI. , 1 996) was reflected in the frequent OCCUlTence of spl i t peaks for amplification at loci B2, B5 , B9, B 10, and B 1 1 . Single peaks predominated at loci B 1 , B4, and B6 and were assumed to be the products without the addition of a deoxyadenylate. Methods that would potentially eliminate this problem would be the use of a DNA polymerase such as Pwo or Tfl (which amplify only fragments with blunt ends), enzymatic removal of the overhang by T4 DNA polymerase (Ginot et aI . , 1 996), or use of the 'PIGtail ing' method (Brownstein et aI. , 1 996). Another difficulty when amplifying loci using a multiplex assay is that alleles can be amplified with widely different efficiencies. When examined on an automatic analyser, particularly strong signal peaks would often appear with false subpeaks of different colours directly beneath the main peak; this situation was often observed when analysing products from multiplex amplification with Primer Set I, B4 allele(s) amplified strongly and green and yellow subpeaks were often displayed under the main blue peak l 75 (Fig. 3 .29). The subpeaks, which are an artefact of the dye-detection system and the algorithms in the analytical software, were ignored when analysing the endophyte genotypes. A more serious problem was that strong signals were occasionally displayed as having flat peaks, thus fragment sizing would be imprecisely determined on the basis of the midpoint of the plateau. These issues can be resolved by amplifying products at each locus individually , fol lowed by pooling products in proportions which would give them approximately the same amounts, or rerunning dilutions of the multiplexed sample so that the concentrated product is within the peak height range for accurate size calling. 3.3.3 APPLICATIONS OF THE MICROSATELLITE BASED FINGERPRINTING ASSAY FOR EPICHLOE ENDOPHYTES 3.3.3.1 Evolutionary Relationships Between Endophyte Groups are Generally Supported by Microsatellite Genotype Data As outlined in the introduction, the evolutionary relationships between Epichloe endophyte species and taxonomic groups are complex, and there is very good evidence from molecular phylogenetic studies that the asexual Neotyphodium spp. have evolved from the sexual Epichloe spp. (Schardl, 1 996a). The microsatellite genotype data shown in Table 3 .6 shows support for many of these relationships. 3.3.3. 1.1 The Evolutionary Relationships of Asexual Endophytes The asexual endophytes from perennial ryegrass studied by Christensen et al . ( 1 993) fall into two taxonomic groups; LpTG- l (= N. lolii) and LpTG-2, with the former being further resolved by i sozyme analysis into six different phenotype groupings. While most of the microsatellite loci were unable to resolve the N. lolii isolates to the level of the i sozyme phenotype, locus B l l , was particularly informative, with six different alleles being amplified. The other loci were less informative in resolving this relatively homogeneous group, though isolates Lp 14 (loE) and Lp9 (loF), were frequently shown to 176 be different from the other N. lolii isolates. An interesting finding was that the endophyte i solate used in Grasslands Pacific EndoSafe, hereafter referred to as EndoSafe, shares the same al1eles as N. lolii isolates of loA, loB and 10D phenotypes at all loci except B 1 1 . This provides overwhelming support for the classification of EndoSafe as an N. lolii i solate, whereas originally Grasslands Pacific EndoSafe was thought to have been inoculated with the LpTG-2 isolate, Lpl . From molecular phylogenetic studies, the most closely related sexual ancestors of the N. lolii group are the E. Jestucae (Leuchtmann, 1 994; Schardl et aI . , 1 994) . Locus B4 is particularly informative in demonstrating this evolutionary connection with both groups sharing a common allele. LpTG-2 i s represented in this study by just two isolates, Lpl and Lp2 (Christensen et aI . , 1 993) and both are known to be interspecific hybrids between E. typhina and N. lolii (Schardl et aI . , 1 994). The hybrid nature of these isolates is supported here in that the two different alleles found in E. typhina isolate E8 and N. lolii, for loci B4 and B6 were both amplified from the LpTG-2 isolates. Alleles amplified in N. lolii i solates at loci B 1 , B9, and B I0 were also present in LpTG-2. A third type of asexual endophyte found within the perennial ryegrass , is represented by i solate AR17 , has E. Jestucae-like morphological characteristics, though stroma production on infected plants has not been observed (pers. comm. Mike Christensen, Grasslands AgResearch). Mating attempts using AR17 (conidiating) cultures as spermatial material to inoculate stroma of E. Jestucae isolates have also been unsuccessful (pers. comm. Christopher Schardl, University of Kentucky). However, the microsatell ite genotype of AR17 i s highly similar to those of E. Jestucae i solates, particularly Fg l , where six of the nine alleles analysed are shared between these i solates (see Table 3 .6). This level of variability is smaller than that observed between E. Jestucae i solates themselves. The endophyte isolates from tall fescue have been resolved into three taxonomic groups (Christensen et aI . , 1 993) and all are known to be interspecific hybrids (Tsai et aI . , 1 994). The proposed evolutionary origins of these hybrids i s complex in that multiple 177 hybridisation events with Epichloe species would be required to account for the diversity of tub2 and other genes that are found in this group (Tsai et aI . , 1 994). Loci B9, B 10, and B 1 1 are particularly supportive of a hybrid origin for this group of isolates as multiple alleles are present for all strains at locus B IO, and for FaTG- l and FaTG-2 at loci B l l and B9 (using primers B9. 1 and B9.2) . Typically, sexual species and the haploid N. lolii group have single alleles at these loci . Although the high degree of polymorphism observed at these loci obscures c lear evolutionary origins of these hybrid groups, many common alleles were found between the tall fescue isolates and their proposed ancestors (Schardl , 1 996a), which include N. uncinatum. E. Jestucae and E. baconii for FaTG- l ; E. Jestucae and E. baconii for FaTG-2; and E. typhina and E. baconii for FaTG-3 . Only single alleles were amplified from N. coenophialum isolates at the B l locus, though in a study by Groppe and BoIler, DNA samples from N. coenophialum isolates were amplified using a reduced annealing temperature (50°C) and two bands were amplified from each isolate (Groppe and Boller, 1 997). Of these bands, one was of constant size (approximately 300 bp), but the other was highly polymorphic and its size ranged between 400 and 600 bp. It will be interesting to sequence these larger alleles to determine whether the size increase is due solely to microsatel lite expansion. The isolates from meadow fescue used in this study were readily identified from the other endophytes by their distinctive microsatellite genotypes. While N. uncinatum has only a single E. typhina-l ike copy of the �-tubulin gene (tub2), the phylogeny generated from sequences for this gene contradicts that for the rDNA-ITS, suggesting that N. uncinatum i s also a hybrid (Schardl, 1 996a), a concept which i s also supported by the presence of multiple bands from isozyme analysis (Leuchtmann, 1 994). Further evidence from transcription elongation factor alpha (TEF) sequence data suggests that N. uncinatum is the product of interspecific hybridisation between E. bromicola and E. typhina, the E. typhina isolates being c losely related to those from the grass host Poa nemoralis (pers. comm. Christopher Schardl, University of Kentucky). This relationship is strongly supported by the sharing of many microsatellite alleles between these groups, 178 particularly at loci B4 and B9, where the alleles found In E. bromicola and two E. typhina i solates E102 1 and E1022, from P. nemoralis, are both observed in the N. uncinatum i solates . Furthermore, alleles amplified in E. bromicola at loci B6 and B 10, and in E. typhina (isolates E 102 1 and E 1022) at locus BS , are all found in N. uncinatum. The final group of asexual endophytes used in the development of this microsatellite assay were two rather poorly described i solates, HaB and Hd1 , from Hordeum grasses. The microsatellite genotypes of these i solates enables them to be readil y distinguished from the other endophytes and from each other. From phylogenetic analysis of tub2 and rDNA-ITS sequences (Section 4.2 .7), HaB and Hdl are thought to have interspecific hybrid origins involving E. amarillalls and E. elymi, and E. typhina and E. bromicola respectively. The presence of multiple alleles at microsatell ite loci B l , B9, B IO, and B l l for HaB, and B6 and B 1 0 for Hdl , supports the hybrid origins of these i solates, but direct correlation of these alle1es to those of their putative ancestral species were not apparent . Common alle1es were observed between Hdl and E. bromicola at loci B l l , and Hd1 and E. typhina at loci B2 and BS, but no common alleles were observed between HaB and its putative ancestors. A number of common al le1es were observed between HaB and Hdl and many of the asexual endophytes (particularly HaB with N. uncinatum) but it i s unclear whether these have common ancestry. 3.3.3. 1.2 The Evolutionary Relationships of Sexual Endophytes Epichloe typhina and E. Jestucae were the only species of sexual endophytes where a substantial number of i solates were available to study. These i solates, on average, showed a higher degree of polymorphism than the groups of asexual isolates, which would be expected for an outcrossing sexual species. However, it is interesting to note that the stability of microsatell ite repeat tracts i s similar both during meiosis and mitosis (Strand et aI . , 1993), therefore the observed variation in polymorphism is likely to be due to non-random sampling of the i solates, rather than a genuine difference in amount of microsatel lite polymorphisms generated between sexual and asexual groups. 179 An unexpected result was the amplification of multiple alleles at the B2 and B 10 loci of E. typhina isolate E428 . The multiple alleles at the B2 locus were observed when E428 DNA was amplified with primers B2. 1 and B2.2 followed by separation on agarose gels (Fig. 3 .26B), though only one allele was detected when amplified with Primer Set II fol lowed by automated analysis . Two all eles were also amplified at the B 1 locus of E. festucae isolate FI l . These multiple alleles may be due to gene duplication events or aberrant chromosome segregation. Alternatively, these i solates may be the product of a hybridisation event, resulting in heteroploidy or aneuploidy, but thi s i s highly unlikely as multiple alleles were observed at relatively few microsatellite loci . Leuchtmann has also observed occasional double-banded electromorphs during alIozyme analysis of Epichloe species, and he considers that these are due to gene duplication events (Leuchtmann and Clay, 1 990; Leuchtmann et at, 1 994). Fundamental relationships between Epichloe spp. were supported by microsatellite data, particularly with regard to the actual amplification of alleles from i solates. A tub2 gene phylogeny of isolates representing all ten Epichloe spp. (Schardl et aI . , 1 997) separates the species into two main elades containing E. typhina, E. clarkii, and E. sylvatica in one, and the remaining species in the other. This elementary division is supported by amplification of loci B 1 , B2, B5, and B l l (with the occasional exception), where out of all Epichloe spp. alleles were only amplified from isolates of E. typhina, E. clarkii, and E. sylvatica with loci B2 and B5. In contrast, alleles from loci B 1 and B 1 1 were I amplified in all isolates except E. typhina, E. clarkii, and E. sylvatica (though E. baconii i solates, E42 1 and E424, did not amplify with the B 1 plimers either). The major differences between E. typhina and E. clarkii are that the ascospores of the latter species undergo disarticulation within the ascus to form uniquely shaped part­ spores, and E. clarkii also appears to be confined to the host genus Holcus (White, 1 993). Despite their species status, E. typhina and E. darkii isolates tend to be interferti le in artificial matings, and are grouped together in the same mating population (MP I ; Leuchtmann and Schardl, 1 998). Their c lose evolutionary relationship i s reflected in the 1 80 similarities of their microsatellite genotypes, in particular the genotypes of E. typhina i solate E428 and E. clarkii i solate E422 are almost identical (Table 3 .6) . Isolates of E. brachyelytri form a discrete basal clade in Epichloe phylogenies based on tub2 sequence data (Schardl and Leuchtmann, 1 999) . Morphologically they are also distinguished by a flattened ascus cap which is unique to thi s species of Epichloe. So far, all microsatellite alleles amplified from E. brachyelytri i solates are unique to this species. 3.3.3.1.3 The Distribution of a Deletion among B1 Alleles - Implications for Endophyte Evolution A 73 bp deletion was discovered in the flanking regions of B 1 alleles amplified from E. typhina, E. clarkii, and E. sylvatica isolates . Specifical ly, the deletion was found in both of the E. sylvatica i solates tested, two of the three E. clarkii isolates tested, and dispersed throughout E. typhina (Section 3 .2 .3 .3) . This finding raises questions of how such a conserved deletion has become so widespread throughout the three species. Three scenarios for the distribution of the deletion are considered. The deletion may have arose long ago in a progenitor to the three Epichloe species, and was passed on to each one by vertical transmission. Alternatively, it may have arisen in one species after speciation, and has transferred to the other species by interspecific mating or hybridisation events. A final possibility is that the deleted segment is some transposable genetic element (TOE) which has exci sed from some isolates. By process of elimination, the first scenario appears to be the most likely of the three to explain the widespread distribution of the deletion. If the deletion was a recent event and originated in one species with subsequent spreading to the others by interspecific mating or hybridisation, it would be expected that the sequences flanking the deletion would be quite conserved, but a reasonable degree of sequence variation was observed in these regions. Also, if the deletion had spread via interspecific hybridisation, i t would be expected that multiple alleles would be apparent at other genetic loci , though this is rarely observed in Epichloe species. It is therefore unlikely that the deletion has spread 1 8 1 this way. The possible scenario that the deleted region i s some TOE which has excised also appears unlikely. Integration of TOEs often results in short direct repeats of the target sites flanking the elements and these duplications may be retained after the TOE has excised. Short direct repeats are not apparent in the deleted or non deleted B 1 alleles sequenced. Moreover, the sizes of TOEs are often two to three times the order of magnitude of the deletion reported in this study. The deletion is therefore l ikely to have originated some time ago in a progenitor to the three Epichloe species, and has distributed to each species by vertical transmission. This hypothesis i s supported by the numerous sequence differences in the flanking regions of deleted alleles. If the deletion event occurred a long time ago, this provide the opportunity to accumulate a reasonable number of mutations in these regions. The fact that the deletion is not found in all B 1 alleles of E. typhina, E. clarkii, and E. sylvatica i solates may reflect that the deletion has not been fixed in these populations. In an independent study by Oroppe and BoIler where a range of Epichloe i solates were amplified with the B 1 . 1 and B 1 .2 primers, only alleles of approximately 300 bp in size were amplified from isolates of E. typhina and E. clarkii at an annealing temperature of 50°C (Oroppe and Boller, 1 997). However, smaller alleles of approximately 250 bp were amplified with these primers from the Tridens flavus endophyte Balansia epichloe (Oroppe and Boiler, 1 997). It will be interesting to sequence these alleJes to find whether they contain the same 73 bp deletion as found in this study. 3.3.3. 1.4 Phylogenetic Inference llsing Microsatellite Data Although many evolutionary relationships between endophyte species and taxonomic groups are supported by the sharing of common microsatell ite alleles (Sections 3 .3 .3 . 1 . 1 and 3 .3 .3 . 1 . 2), phylogenetic inference using microsatell ite allele length data has not been approached in this study. Though several genetic distance statistics have been formulated for microsatel li te data to infer phylogenies, these have had mixed success in resolving evolutionary relationships between groups in various applications (Ooldstein et aI . , 1995 ; Kimmel et aI . , 1 996; Paetkau et aI . , 1 997). The mutational models used in 1 82 computer simulations are generally based on stepwise mutation models as expected for DNA slippage synthesis, though these tend to oversimplify the actual dynamics of mutations at microsatel li te loci (Paetkau et aI . , 1 997) . For example, these models do not necessarily account for: non-'perfectness' of microsatellite tracts; compound microsatellite loci ; and mutation of the flanking sequences, particularly insertions and deletions which affect allele length. There is also debate as to whether the rate and direction of mutation can vary between closely related species (Ellegren et aI . , 1 995 ; Rubinsztein et aI . , 1 995), and it has been suggested that the usefulness of microsatell i te data for inferring phylogeny may be limited to a short time perspective and to relatively small populations (Nauta and Weissing, 1 996). Given the doubts regarding the reliabi lity of these methods, in this study evolutionary analysi s has been restricted to basic observations of common allele sharing between groups. 3.3.3.2 Endophyte Identification In Planta - A Trial Application This assay has been designed for the rapid, l arge-scale identification of endophytes in the field situation, and this has been accomplished by: in planta endophyte detection, the use of rapid miniprep DNA extraction methods, peR based microsatelli te amplification for fingerprint generation, and the automated analysis of fingerprints . To test this assay in the field situation, a blind test was carried out to identify the endophyte i solates infecting twelve samples of grass tissues (Section 3 .2 .3 .2). In all cases a correct match was made with a taxonomic grouping and isozyme phenotype (if available) that had previously been determined for that isolate. The simple and robust procedures employed in the assay enables results to be obtained in as little as 24 hours. Although identification cannot be resolved to the level of a specific isolate, it can be deduced to a group of c losely related isolates that share the same microsatellite genotype. Moreover, isolates in the reference database that have different genotypes can be eliminated as a positive match. Plant DNA extracts notoriously contain compounds which are inhibitory to peR, thus a safeguard against false-negative peR results (Section 3 .2 .2 .5 .3) could be employed in 1 83 conjunction with this assay. Such an approach was taken by Groppe and BoIler ( 1 997) who developed a multiplex reaction which amplified both plant and fungal sequences to ensure that reaction conditions were suitable for amplification. 3.3.3.3 Microsatellite Loci as Markers in Endophyte Genomes Microsatellite loci are rapidly becoming popular genetic markers in the mapping of animal and plant genomes (Broun and Tanksley, 1 996; Buchanan et aI . , 1 994; RodeI' et aI . , 1 995 ; Wu and Tanksley, 1 993), though their application to fungal genomes has not yet been approached. As physical markers in Epichloe endophyte genomes, microsatell i te loci could prove useful in documenting the chromosomal changes which have occurred during the evolution of this group, particularly in the case of interspecific hybrid endophytes. These groups, which are often heteroploid or aneuploid (Leuchtmann and Clay, 1 990), are thought to have undergone nuclear fusion followed by a subsequent loss of redundant chromosomes or chromosomal segments (Tsai et aI . , 1 994) . Microsatellite markers may also be genetical ly linked to loci for traits such as alkaloid production and sexual mating types. The microsatel li te loci from this study were to be physical ly mapped to the endophyte genomes, but this was not achieved due to time and financial constraints . Currently, the only information regarding the genomic location of microsatel l i te loci used in this study, is from B 10 and B 1 1 , whose alleles both show 1 : 1 segregation in E. festucae crosses (Section 3 .2 .3 .4). Since this is the segregation ratio expected for nuclear genes, these loci are probably located in the nucleus, not the mitochondria. Locus B 1 1 is linked to the gene encoding HMG CoA in Lp 1 9 (Dobson, 1 997), and it is also linked to B lO in E. festucae (Section 3 .2 .3 .4). It is possible that these three loci are linked in many Epichloe endophyte i solates. 3.4 CONCLUSION 1 84 A multi locus microsatellite-based peR fingerprinting assay for the identification of Epichloe endophytes , both in culture and in planta, has been developed and successfully used to identify endophyte i solates from infected grass tissues. This assay should therefore be a useful tool for grass breeders to rapidly determine both the identity of known endophytes in pasture to the i sozyme phenotype level, and also identify novel endophytes which are potential sources for new low alkaloid phenotypes. This assay should also be of value to other endophyte researchers as a general fingerprinting tool, and as an adjunct to the study of evolution of the endophytes, particularly with regard to the many hybrid asexual Epichloe endophytes . CHAPTER FOUR THE EVOLUTIONARY ORIGINS OF EPICHLOE ENDOPHYTES FROM ANNUAL LOLIUM AND HORDE UM GRASSES 4.1 INTRODUCTION 1 86 The genus Lolium contains eight species of ryegrasses (Charmet et al., 1 997), including the agriculturally important perennial ryegrass (L. perenne) and Ital ian ryegrass (L. multiflorum). Perennial ryegrass is the most widely used of these grasses for forage purposes, and i t has been extensively surveyed for endophytes where i t i s found to host several taxonomic groups of Epichloe endophyte. These include Neotyphodium lolii (Latch et aI . , 1984), Epichloe typhina (Schardl et aI . , 1 99 1 ), an asexual hybrid of these two endophytes that has been designated as LpTG-2 (Christensen et aI . , 1 993), and an E. Jestucae-like endophyte (pers. comm. Mike Christensen, Grasslands AgResearch). In contrast, l i ttle research on endophytes of annual ryegrasses has been undertaken. Latch et al . ( 1 988) had surveyed six species of annual and one species of hybrid ryegrass for endophytes and reported the occurrence of a seedbome endophyte that is basally confined in the vegetative tillers of these grasses. The growth habit of this endophyte, both in planta and in culture, and the results of serological tests (Carter et aI. , 1 997) suggest these endophytes are related to Epichloe and Neotyphodium species, and as such, have so far been described as Neotyphodium-like endophytes (Latch et aI., 1 988). Very l ittle is known about the taxonomy and evolution of these endophytes, and this i s possibly due to the difficulties in isolating and obtaining sustainable cultures for analysis, as reported by Latch et al. ( 1 988) . Also, these grasses are not as widely used in pasture systems as perennial ryegrass and are not associated with selious livestock disorders . However, loline alkaloids have been detected in endophyte-infected annual ryegrasses (TePaske et aI . , 1 993), and the presence of endophyte does deter Argentine stem weevil (Listronotus bonariensis) feeding on young seedlings (Latch et aI . , 1988) . The aim of this study was to genetical ly characterise the Neotyphodium-like endophytes from all seven species of annual ryegrasses, and determine their evolutionary Oligins from tub2 and rDNA-ITS gene sequence data and microsatellite fingerprint profiles. Additionally, two endophyte isolates, denoted HaB and Hd1 , from the grass genus Hordeum were available in the Scott Laboratory fungal culture collection. Hordeum, a member of the grass tribe Triticeae, comprises about 45 species and subspecies 1 87 distributed across temperate regions of the world (De Bustos et aI . , 1 999), and includes cultivated barley, H. vulgare subsp. vulgare. Extensive studies of endophytic fungi within the genus Hordeum have not been conducted, but barley is not known to contain Epichloe endophytes. Little background information is available regarding HaB and Hdl , though it is believed they were found in seed from Hordeum spp . , in particular Hdl is probably from H. bogdanii, collected on an expedition in China in the 1980's. The seed have since been kept in the Margot Forde Forage Germplasm Centre at Grasslands AgResearch (pers. comm. Mike Christensen, Grasslands AgResearch). The evolutionary origins of HaB and Hdl were also to be investigated, using an approach simi lar to that for the genetic characterisation of the endophytes of annual ryegrasses. 4.2 RESULTS 4.2.1 ENDOPHYTE ISOLATION FROM PLANT TISSUE 1 8 8 Isolations were attempted from plants of four populations of L. canariense Steud. , two populations of L. multiflorum Lam. cv. Maverick, one population of L. persicum Boiss & Hoh. , two populations of L. remotum Schrank. , three populations of L. rigidum Gaud. var. rigidum sens. Ten'ell , one population of L. rigidum Gaud. var. rottboellioides Heldr. ex Boiss, two populations of L. subulatum Vis. , and four populations of L. temulentum L. All plants used in these isolations had been confirmed as containing hyphae similar in appearance to that of the Neotyphodium-like endophyte by examining ani line blue­ stained tissue, removed from the basal region of leaves, under a compound microscope. Isolations were carried out in early and late spring following the initiation of elongation of reproductive tillers . The use of elevated nodes and associated basal leaf sheaths, and in the case of the uppermost nodes, immature inflorescences, was to reduce fungal and bacterial contamination that frequently occurs when isolations are attempted from the basal leaf sheath of vegetative til lers. Endophytes were isolated as described in Section 2 .2 . 1 . Colonies consistent with those of the Neotyphodium-like endophyte previously described by Latch et al . ( 1 988) were obtained from a few tissue pieces from two populations of L. multiflorum cv. Maverick (Lml and Lm2), three of L. rigidum var. rigidum (Lrr l , Lrr2 and Lrr4), and one of L. rigidum var. rottboellioides (Lro l ) . These colonies were very slow to emerge from excised incubated tissue with slow subsequent growth, waxy, convoluted, white to tan coloured and lacking conidia. Examples of such colonies i solated from L. multiflorum cv. Maverick (Lm 1 ) and L. rigidum var. rigidum (Lrrl ) are shown in Figs. 4. lA and 4 . 1B . From one population of L. canariense (Lc4), colonies developed after just one to two weeks incubation. These sterile colonies differed from those previously described for the endophyte of annual rye grasses as they were dark and feIted in appearance, and they 1 89 Fig. 4. 1 . Cultures of Neotyphodium endophytes isolated from annual ryegrass tissues Endophytes were isolated from plant tissue and grown on ABPDA media. (A) Colony of Lm! endophyte i solated from L. multiflorum cv. Maverick seed material after 14 weeks incubation. (B) Colony of LITl isolated from L. rigidum var. rigidum immature inflorescence tissue after 1 6 weeks incubation. (C) The variable growth forms of a 4 week old colony of Lc4 from L. canariense which had been subcultured. Scale bars are 1 mm long. 190 A B c 1 9 1 were able to be repeatedly subcultured (Fig. 4 . 1 C). Re-examination of plants from this population revealed hyphae throughout the leaf sheath, rather than being confined to the most basal tissue. Further i solations from the sheath and blades of vegetative leaves gave rise to similar colonies. Surface steril ised tissue pieces from the remainder of plants failed to give rise to colonies, although trace hyphal growth was observed from plants of two L. temulentum populations. Many of the excised tissue pieces became dark and mushy after a few weeks incubation. Fungal contaminants, including Acremonium spp . , arose from a number of tissue pieces, but these were not included in this study. 4.2.2 FUNGAL AND IN PLANTA DNA EXTRACTION Total genomic DNA was isolated from fresh fungal endophyte cultures Lc4, Lml , Lm2, Lrr l , Lrr2, and Lro l using a microwave miniprep method (Section 2 .5 .2) . The DNA yields from the pure cultures were too low to visualise on agarose gels, and fluorometer readings were too low to give accurate readings, however when extracted DNA was used directly in PCR reactions, amplification yields were generally satisfactory. The small amount of biomass and consequent low yields of DNA from each annual ryegrass endophyte l imited the number of PCR reactions that could be done. Total genomic DNA from fresh nodal tissue of endophyte-infected grasses, where endophyte i s known to be highly abundant (Latch et aI . , 1 988) , was isolated using the FastDNA Kit (Section 2 .5 .3) . High yields of DNA were obtained from endophyte infected grasses, about 3 to 1 0 Ilg per preparation, which was more than sufficient for microsatel l i te fingerprinting. 4.2.3 rDNA-ITS AMPLIFICATION AND SEQUENCING PCR amplification of the nuclear ribosomal DNA region spanning the internal transcribed spacers (rDNA-ITS), and the 5 .8S rRNA gene (Section 2 . 1 3 .9) was amplified 1 92 with primers ITS5 and ITS4, using DNA prepared from annual ryegrass endophyte cultures (Section 4.2.2) . As shown in Figure 4.2, the amplification products were resolved as single bands of approximately 600 bp when separated on 2% agarose minigels (Section 2.8) . If the peR product yields were very low, as in lanes 1 , 2, 5 , and 6, products were diluted and re-amplified with Pwo DNA polymerase (Section 2. 1 3 .9) to obtain quantities sufficient for sequencing. peR products were excised from the gel and purified (Section 2 .5 . 10), and both strands were sequenced using each of the ITS5 and ITS4 primers (Section 2 . 1 5 .2). The complete sequences of the rDNA-ITS regions were obtained from the endophyte i solates : Lc4, Lm1 , Lm2, Lrr 1 , Lrr2, and Lro l . Figure 4.3 shows an alignment of these sequences against that of N. coenophialum isolate e 1 9 (Schardl et aI . , 1 99 1 ) , generated using the PILEUP program of the Wisconsin Package (Section 2. 1 8 . 1 ). The alignment revealed that the rDNA-ITS sequences of Lm1 , Lm2, Lrr 1 , and Lr0 1 were identical, and that Lrr2 differed from this by a 2 nt indel (AT) in the ITS2 region (positions 445 and 446). The rDNA-ITS sequence of Lc4 had 97% identity to those of the other aligned annual rye grass endophyte i solates , and all sequence differences were located in the ITS] and ITS2 regions. Database similarity searching of these sequences by BLAST (Section 2. 1 5 .3) identified only sequences from the Epichloe and Neotyphodium genera. 4.2.4 tub2 AMPLIFICATION AND SEQUENCING The 5' portions of the B-tubulin (tub2) genes, including introns IVS 1 , IVS2, and IVS3, were peR amplified with primers T 1 . 1 and T 1 .2 (Section 2 . 1 3 .8) using DNA prepared from Lc4, Lm1 , Lm2, LIT 1 , Lrr2, and Lro1 endophyte cultures (Section 4.2.2). An agarose gel of the tub2 amplification products from various annual ryegrass endophyte isolates is shown in Figure 4.4. Amplification of Lc4 (lane 1) yielded a single product that gave unambiguous results when sequenced directly (Section 2. 1 5). BLAST similarity searching of this sequence revealed that it had high similarity to the tub2 gene from N. uncinatum and the tub2-3 copy from N. coenophialum (Tsai et aI . , 1 994). Fig. 4.2. rDNA- ITS amplification products from annual ryegrass endophyte i solates 900 bp - 800 bp - 700 bp - 600 bp - 500 bp - M 2 3 4 5 6 7 8 193 Products were amplified with primers ITS5 and ITS4 and resolved on a 2% agarose gel . Reactions contained genomic DNA from endophyte i solates Lrr1 (lane 1 ) , Lrr2 (lane 2), Lro 1 (lane 3), Lro 1 (lane 4), Lrr1 (lane 5), Lm2 (lane 6), and positive control isolate FlT1 (lane 7). Lane 8 contains a peR negative control without DNA, and the numbers on the left of the figure indicates the sizes in bp of the 1 00 bp l adder in lane M. 1 94 Fig. 4 .3 . Alignment of rDNA-ITS sequences from annual ryegrass endophyte isolates . The DNA sequences of the rDNA-ITS regions from annual ryegrass endophyte isolates Lc4, Lm 1 , Lm2, Lrr l , Lrr2, and Lro 1 , and N. coenophialum isolate e 1 9 were aligned using the PILEUP programme of the Wisconsin Package with the gap penalty set to one and the gap extension penalty set to zero. Gaps ( . ) were introduced to facilitate sequence alignment. The sequenced region from e 1 9 encodes ITS 1 (positions I - 1 69), 5 . 8S rRNA (positions 1 70 - 337, shown italicised), and ITS2 (positions 338 - 509) (Schardl et aI. , 1 99 1 ) . Nucleotides that differ from the reference sequence, e 19 , are highlighted i n green. 1 5 0 e 1 9 CGAGTTTACA CTCCCAAACC CCTGTGAACC T . TACCTTTA CTGTTGCCTC Lc4 CGAGTTTACA CTCCCAAACC CCTGTGGACC TATACCCTCA CTGTTGCCTC Lrr2 CGAGTTTTCA CTCCCAAACC CCTGTGGACT TATACCTTTA CCGTTGCCTC Lrr1 CGAGTTTTCA CTCCCAAACC CCTGTGGACT TATACCTTTA CCGTTGCCTC Lro1 CGAGTTTTCA CTCCCAAACC CCTGTGGACT TATACCTTTA CCGTTGCCTC Lm2 CGAGTTTTCA CTCCCAAACC CCTGTGGACT TATACCTTTA CCGTTGCCTC Lm1 CGAGTTTTCA CTCCCAAACC CCTGTGGACT TATACCTTTA CCGTTGCCTC 5 1 1 0 0 e 1 9 GGCGGGCACG GCCGCGGACG CCCCCTCGCG GGGGCACCGG GGCCGGGCGC Lc4 GGCGGGCACG GCCGCGGACG CCCCCTCGCG GGGGCGCCGG GGCCi\GGCGC Lrr2 GGCGGGCACG GCCGCGGACG CCCCCTCGCG GGGGCACCGG GGCCl,GGCGC Lrr1 GGCGGGCACG GCCGCGGACG CCCCCTCGCG GGGGCACCGG GGCCAGGCGC Lro1 GGCGGGCACG GCCGCGGACG CCCCCTCGCG GGGGCACCGG GGCCi\GGCGC Lm2 GGCGGGCACG GCCGCGGACG CCCCCTCGCG GGGGCACCGG GGCCAGGCGC Lm1 GGCGGGCACG GCCGCGGACG CCCCCTCGCG GGGGCACCGG GGCCAGGCGC 1 0 1 1 5 0 e 1 9 CCGCCGGAGG ACCCAAACCC TTCTGTATTT TTCTTACGCA TGTCTGAGTG Lc4 CCGCCGGAGG ACCCAAACCC TTCTGTA . TT TTCTTGCGCA TGTCTGAGTG Lrr2 CCGCCGGAGG ACCCAAACCC TTCTGTA TT TTCTTACGCA TGTCTGAGTG Lrr1 CCGCCGGAGG ACCCAAACCC TTCTGTA , TT TTCTTACGCA TGTCTGAGTG Lrol CCGCCGGAGG ACCCAAACCC TTCTGTA TT TTCTTACGCA TGTCTGAGTG Lm2 CCGCCGGAGG ACCCAAACCC TTCTGTA TT TTCTTACGCA TGTCTGAGTG Lm1 CCGCCGGAGG ACCCAAACCC TTCTGTA , TT TTCTTACGCA TGTCTGAGTG 195 1 5 1 2 0 0 e 1 9 GATTTAATAT CAAATGAATC AAAACTTTCA ACAACGGATC TCTTGGTTCT Lc4 GATTTA CAAATGAATC AAAACTTTCA ACAACGGATC TCTTGGTTCT Lrr2 GATTTAATAT CAAATGAATC AAAACTTTCA ACAACGGATC TCTTGGTTCT Lrr1 GATTTAATAT CAAATGAATC AAAACTTTCA ACAACGGATC TCTTGGTTCT Lro1 GATTTAATAT CAAATGAATC AAAACTTTCA ACAACGGATC TCTTGGTTCT Lm2 GATTTAATAT CAAATGAATC AAAACTTTCA ACAACGGATC TCTTGGTTCT Lm1 GATTTAATAT CAAATGAATC AAAACTTTCA ACAACGGATC TCTTGGTTCT 2 0 1 2 5 0 e 1 9 GGCATCGATG AAGAACGCAG CGAAATGCGA TAAGTAATGT GAATTGCAGA Lc4 GGCATCGATG AAGAACGCAG CGAAATGCGA TAAGTAATGT GAATTGCAGA Lrr2 GGCATCGATG AAGAACGCAG CGAAATGCGA TAAGTAATGT GAATTGCAGA Lrr1 GGCATCGATG AAGAACGCAG CGAAATGCGA TAAGTAATGT GAATTGCAGA Lro1 GGCATCGATG AAGAACGCAG CGAAATGCGA TAAGTAATGT GAATTGCAGA Lm2 GGCATCGATG AAGAACGCAG CGAAATGCGA TAAGTAATGT GAATTGCAGA Lm1 GGCATCGATG AAGAACGCAG CGAAATGCGA TAAGTAATGT GAATTGCAGA 2 5 1 3 0 0 e 1 9 ATTCAGTGAA TCA TCGAA TC TTTGAACGCA CATTGCGCCC GCCAGTATTC Lc4 ATTCAGTGAA TCA TCGAA TC TTTGAACGCA CATTGCGCCC GCCAGTATTC Lrr2 ATTCAGTGAA TCA TCGAA TC TTTGAACGCA CATTGCGCCC GCCAGTATTC Lrr1 ATTCAGTGAA TCA TCGAA TC TTTGAACGCA CATTGCGCCC GCCAGTATTC Lro1 ATTCAGTGAA TCATCGAATC TTTGAACGCA CATTGCGCCC GCCAGTATTC Lm2 ATTCAGTGAA TCA TCGAA TC TTTGAACGCA CATTGCGCCC GCCAGTATTC Lm1 A TTCAGTGAA TCA TCGAA TC TTTGAACGCA CATTGCGCCC GCCAGTATTC 3 0 1 3 5 0 e 1 9 TGGCGGGCAT GCCTGTTCGA GCGTCATTTC AACCCTCAAG CCCGCTGCGC Lc4 TGGCGGGCAT GCCTGTTCGA GCGTCATTTC AACCCTCAAG CCCGCTGCGT Lrr2 TGGCGGGCAT GCCTGTTCGA GCGTCATTTC AACCCTCAAG CCCGCTGCGC Lrr1 TGGCGGGCAT GCCTGTTCGA GCGTCATTTC AACCCTCAAG CCCGCTGCGC Lro1 TGGCGGGCAT GCCTGTTCGA GCGTCATTTC AACCCTCAAG CCCGCTGCGC Lm2 TGGCGGGCAT GCCTGTTCGA GCGTCATTTC AACCCTCAAG CCCGCTGCGC Lm1 TGGCGGGCAT GCCTGTTCGA GCGTCATTTC AACCCTCAAG CCCGCTGCGC 3 5 1 4 0 0 e 1 9 GCTTGCTGTT GGGGACCGGC TCACCCGCCT CGCGGCGGCG GCCGCCCCCG Lc4 GCTTGGTGTT GGGGACCGGC CCCCCCGCCT CGCGGCGGCG GCCGCCCCTG Lrr2 GCTTGGTGTT GGGGACCGGC CCGCCCGCCT CGCGGCGGCG GCCGCCCCCG Lrr1 GCTTGGTGTT GGGGACCGGC CCGCCCGCCT CGCGGCGGCG GCCGCCCCCG Lro1 GCTTGCTGTT GGGGACCGGC CCCCCCGCCT CGCGGCGGCG GCCGCCCCCG Lm2 GCTTGGTGTT GGGGACCGGC CCCCCCGCCT CGCGGCGGCG GCCGCCCCCG Lm1 GCTTGCTGTT GGGGACCGGC CCGCCCGCCT CGCGGCGGCG GCCGCCCCCG 1 96 4 0 1 4 5 0 e 1 9 AAATGAATCG GCGGTCTCGT CGCAAGCCTC CTTTGCGTAG TAGC . . ACAC Lc4 AAATGAATTG GCGGTCTCGT CGC . AGCGTC CTTTGCGTAG TAGC . . ACAC Lrr2 AAATGAATTG GCGGTCTCGT CGC . GGCCTC CTTTGCGTAG TAACATATAC Lrr1 AAATGAATTG GCGGTCTCGT CGC . GGCCTC CTTTGCGTAG TAAC . . ATAC Lro1 AAATGAATTG GCGGTCTCGT CGC . GGCCTC CTTTGCGTAG TAAC . . ATAC Lm2 AAATGAATTG GCGGTCTCGT CGC . GGCCTC CTTTGCGTAG TAAC . . ATAC Lm1 AAATGAATTG GCGGTCTCGT CGC . GGCCTC CTTTGCGTAG T C . . ATAC 4 5 1 5 0 0 e 19 CACCTCGCAA CCGGGAGCGC GGCGCGGCCA CAGCCGTAAA ACGCCCAACT Lc4 CACCTCGCAA CCGGGAGCGC GGCGCGGCCA CTGCCGTAAA ACGCCCAAC . Lrr2 CACCTCGCAA CCGGGAGCGC GGCGCGGCCA CTGCCGTAAA ACGCCCAAC . Lrr1 CACCTCGCAA CCGGGAGCGC GGCGCGGCCA CTGCCGTAAA ACGCCCAAC . Lro1 CACCTCGCAA CCGGGAGCGC GGCGCGGCCA CTGCCGTAAA ACGCCCAAC . Lm2 CACCTCGCAA CCGGGAGCGC GGCGCGGCCA CTGCCGTAAA ACGCCCAAC . Lm1 CACCTCGCAA CCGGGAGCGC GGCGCGGCCA CTGCCGTAAA ACGCCCAAC . 5 0 1 e 1 9 TTCTCCAAG Lc4 TTCTCCAAG Lrr2 TTCTCCAAG Lrr1 TTCTCCAAG Lro1 TTCTCCAAG Lm2 TTCTCCAAG Lm1 TTCTCCAAG Fig. 4.4. tub2 amplification products from annual ryegrass and Hordeum endophyte isolates 900 bp 800 bp - 700 bp - 600 bp - M 2 3 4 5 6 7 M 1 97 Products were amplified with primers T 1 . 1 -BamHI and T 1 .2-EcoRI and resolved on a 2% agarose gel . Reactions contained genomic DNA from endophyte i solates Lc4 (lane 1 ) , Lrrl (lane 2), Lrr4 (lane 3), Lro 1 (lane 4), Lm2 (lane 5), HaB (lane 6) and Hd! (lane 7). The numbers on the left of the figure indicates the sizes in bp of the 100 bp ladder in lane M. 198 In contrast, the tub2 PCR products from the remammg annual ryegrass endophytes yielded multiple bands between approximately 700 and 800 bp, examples of these are shown in Figure 4.4, lanes 2 to 5 . These bands could not be separated, so had to be cloned for sequencing. The tub2 PCR products from Lml were c loned into the pGEM-T vector (Section 2 . 14 . 1 ) , but extremely low numbers of white colonies, five per transformation, were produced using thi s method despite several attempts to optimise the ligation conditions. To increase the cloning efficiency, a new strategy involving sticky­ end ligation reactions was undertaken. Plimers to the tub2 gene were redesigned to affix BamHI and EcoRI restriction enzyme recognition sequences to the 5 '-ends of T 1 . 1 and T 1 .2 respectively. These primers, T 1 . 1 -BamHI and T 1 .2-EcoRI (Table 2), were used with the same amplification conditions (Section 2 . 1 3 .8) , and the resulting products were prepared for cloning by cleaving with BamHI and EcoRI to produce sticky ends (Section 2 . 14.2) . Products were then directionally cloned into EcoRI and BamHI cleaved pUC l 1 8, and about 24 white clones from each transformation were screened as described previously (Section 2 . 1 2.9). All unique clone types were fully sequenced (Section 2 . 1 5) . Three to five unique clone types were identified from each endophyte isolate but analysi s of sequence alignments showed that many of these were recombinant products, which were probably generated during PCR. The 'true' tub2 genes were identified based on both their similarity to extant Epichloe and Neotyphodium tub2 genes, and the relative frequencies of the different clone types as recombinant genes typically accounted for about one sixth of the clones screened. Overall it was found that each endophyte isolate: Lm 1 , Lm2, Lrr 1 , Lrr2, and Lrol had two distinct copies of tub2 genes, and copies were labelled tub2-a and tub2-b. BLAST similarity searches (Section 2 . 1 5 .3) revealed that these copies showed high similarity to tub2 genes of E. baconii and E. bromicola respectively. A PILEUP generated alignment (Section 2 . 1 8 . 1 ) of the tub2 DNA sequences from each of the annual ryegrass endophyte isolates analysed, and N. coenophialum isolate e 1 9 tub2-2 copy (Tsai e t aI . , 1 994) i s shown i n Figure 4 .5 . The E. baconii-l ike genes, Lml 1 99 Fig. 4 .5 . Alignment of the tub2 gene sequences from annual ryegrass endophyte isolates The DNA sequences of the tub2 genes from annual rye grass endophyte isolates Lc4, Lml , Lm2, Lrr l , Lrr2, and Lro l , and N. coenophialum isolate e 1 9 (tub2-2 copy) were aligned using the PILEUP programme of the Wisconsin Package, with the gap penalty set to one and the gap extension penalty set to zero. Gaps (.) were introduced to facilitate sequence alignment. The sequenced region from e 1 9 encodes IVS 1 (positions 1 - 206), exon 2 (positions 207 - 23 1 ), NS2 (positions 232 - 304), ex on 3 (positions 305 - 428), IVS3 (positions 429 - 536), and ex on 4 (positions 537 - 542) (Tsai et aI. , 1 994). Exon positions are shown italicised, and nucleotides that differ from the reference sequence, e 19, are highlighted in green. 1 5 0 e1 9 t ub2-2 AAGTTCAACC TCTCTGTTTG TCTTGGGGAC CCCCTCCTCG ACGCGTTCCG Lc4 t ub2 GAGTTCAACC TCTCTGTTTG TCTTGGGGAC CCCCTCCTCG ACGCGTTCCG Lml t ub2-a AAGTTCAAAC TCTCTGTCTG TCTTGGA�;CC CCTCTCCTCG ACGCGTTCCG Lm2 t ub2-a AAGTTCAA,;C TCTCTGTCTG TCTTGGAA(�C CCTCTCCTCG ACGCGTTCCG Lrol t ub2-a AAGTTCAA1\C TCTCTGTCTG TCTTGGl',ACC CCTCTCCTCG ACGCGTTCCG Lrr2 t ub2-a AAGTTCAAA , , CTCTGTCTG TCTTGGAl\('C CCTCTCCTCG ACGCGTTCCG Lrrl t ub2-a AAGTTCAA.AC TCTCTGT��TG TCTTGGl,i\CC CCTCTCCTCG ACGCGTTCCG Lml t ub2 -b AAGTTCAACC TCTCTGTTTG TCTTGGGGAC CCCCTCCTCG ACGCGTTCCG Lm2 tub2-b AAGTTCAACC TCTCTGTTTG TCTTGGGGAC CCCCTCCTCG ACGCGTTCCG Lrol tub2-b AAGTTCAACC TCTCTGTTTG TCTTGGGGAC CCCCTCCTCG ACGCGTTCCG Lrrl tub2-b AAGTTCAACC TCTCTGTTTG TCTTGGGGAC CCCCTCCTCG ACGCGTTCCG Lrr2 t ub2-b AAGTTCAACC TCTCTGTTTG TCTTGGGGAC CCCCTCCTCG ACGCGTTCCG 5 1 1 0 0 e19 t ub2-2 GTGTTGAGCC CCTGATTTCG TACCCCGCCG AGCCCGGCCA CGAAGTGCAC Lc4 tub2 GTGTCGAGCC CCTGATTTCG TACCCCGCCG AGCCCGGCCA CGACGTGCAC Lml t ub2-a GTGTTGAGCC CCTGATTTCG TACCCC, 'CCG AGCCCGGCCA CGACGTGCAC Lm2 t ub2-a GTGTTGAGCC CCTGATTTCG TACCCC 'CCG AGCCCGGCCA CGAt 'GTGCAC Lrol t ub2-a GTGTTGAGCC CCTGATTTCG TACCCCCCCG AGCCCGGCCA CGACGTGCAC Lrr2 tub2-a GTGTTGAGCC CCTGATTTCG TACCCC; :CCG AGCCCGGCCA CGACGTGCAC Lrrl tub2-a GTGTTGAGCC CCTGATTTCG TACCCC( 'CCG AGCCCGGCCA CGACGTGCAC Lml tub2-b GTGTTGAGCC CCTGATTTCG TACCCCGCCG AGCCCGGCCA CGACGTGCAC Lm2 t ub2-b GTGTTGAGCC CCTGATTTCG TACCCCGCCG AGCCCGGCCA CGACGTGCAC Lrol t ub2-b GTGTTGAGCC CCTGATTTCG TACCCCGCCG AGCCCGGCCA CGACGTGCAC Lrrl t ub2 -b GTGTTGAGCC CCTGATTTCG TACCCCGCCG AACCCGGCCA CGACGTGCAC Lrr2 tub2-b GTGTTGAGCC CCTGATTTCG TACCCCGCCG AACCCGGCCA CGACGTGCAC 200 1 0 1 1 5 0 e1 9 tub2-2 GCCCAACGAA CAGTCGTGAT GAGAGGCGGA CCGAGAC . . . AAAATTAATG Lc4 tub2 GCCCAACGGA CAGTCGTGAT GAGAGGCGGA CCGAGAC . . . AACATCATTG Lml tub2-a GCCCAACGAA CAGTCGTGAT GAGAGGCGGA TCGAGAC . . . AAAATTAATG Lm2 tub2-a GCCCAACGAA CAGTCGTGAT GAGAGGCGGA TCGAGAC . . . AAAATTAATG Lrol tub2-a GCCCAACGAA CAGTCGTGAT GAGAGGCGGA TCGAGAC . . . AAAATTAATG Lrr2 tub2-a GCCCAACGAA CAGTCGTGAT GAGAGGCGGA TCGAGAC . . . AAAATTAATG Lrrl tub2-a GCCCAACGAA CAGTCGTGAT GAGAGGCGGA TCGAGAC . . . AAAATTAATG Lml tub2-b GCCCAATGAA CAGTCGTGAT GAGAGGCGGA CCGAGACAAA AAAAATAATG Lm2 tub2-b GCCCAATGAA CAGTCGTGAT GAGAGGCGGA CCGAGACAAA AAAAATAATG Lrol tub2-b GCCCAATGAA CAGTCGTGAT GAGAGGCGGA CCGAGACAAA AAAAATAATG Lrrl tub2-b GCCCAATGAA CAGTCGTGAT GAGAGGCGGA CCGAGACAAA AAAAATAATG Lrr2 tub2-b GCCCAATGAA CAGTCGTGAT GAGAGGCGGA CCGAGACAAA AAAAATAATG 1 5 1 2 0 0 e1 9 tub2-2 AATGCGGTAT TCGAGAACTG TAGCTGACCT G . TTCTTCCC CTCTTTTCCC Lc4 tub2 AATGCGGTAT TCGAGAACTG TAGCTGACCT TTTTCTTT . . . . . . . . . . CC Lml tub2-a AATGCGGTAT TCGAGAACTG TAGCTGACCT GTTTCTTTCC CTCTTTTCCC Lm2 tub2-a AATGCGGTAT TCGAGAACTG TAGCTGACCT GTTTCTTTCC CTCTTTTCCC Lrol tub2-a AATGCGGTAT TCGAGAACTG TAGCTGACCT GTTTCTTTCC CTCTTTTCCC Lrr2 tub2-a AATGCGGTAT TCGAGAACTG TAGCTGACCT GTTTCTTTCC CTCTTTTCCC Lrrl tub2-a AATGCGGTAT TCGAGAACTG TAGCTGACCT GTTTCTTTCC CTCTTTTCCC Lml tub2-b ATTGCTGTAT TCGAGAACTG TAGCTGACCT . TTTC . . . . . . . CTTTT . . C Lm2 tub2-b ATTGCTGTAT TCGAGAACTG TAGCTGACCT . TTTC . . . . . . . . CTTTT . C Lrol tub2-b ATTGCTGTAT TCGAGAACTG TAGCTGACCT . TTTC . . . . . . . . CTTTT . . C Lrrl tub2-b ATTGCTGTAT TCGAGAACTG TAGCTGACCT . TTTC . . . . . . . . CTTTT . . C Lrr2 tub2-b ATTGCTGTAT TCGAGAACTG TAGCTGACCT . TTTC . . . . . . . . CTTTT . . C 2 0 1 2 5 0 e1 9 tub2-2 CTCTAGGTTC ATCTTCAAAC CGGTCAGTGC GTAAGTGACA AATTCGCCGA Lc4 tub2 CTCTAGGTTC ATCTTCAAAC CGGTCAGTGC GTAAGTGACA AATCTGCCGA Lml tub2-a CTCTAGGTTC ATCTTCAAAC CGGTCAGTGC GTAAGTGACA AA . TCGCCGA Lm2 tub2-a CTCTAGGTTC ATCTTCAAAC CGGTCAGTGC GTAAGTGACA AA . TCGCCGA Lrol tub2-a CTCTAGGTTC ATCTTCAAAC CGGTCAGTGC GTAAGTGACA AA . TCGCCGA Lrr2 tub2-a CTCTAGGTTC ATCTTCAAAC CGGTCAGTGC GTAAGTGACA AA . TCGCCGA Lrrl tub2-a CTCTAGGTTC ATCTTCAAAC CGG TCAG TGC GTAAGTGACA AA . TCGCCGA Lml tub2-b CTCTAGGTTC ATCTTCAAAC CGGTCAGTGC GTAAGTGACA AATCCGCCGA Lm2 tub2-b CTCTAGGTTC ATCTTCAAAC CGGTCAGTGC GTAAGTGACA AATCCGCCGA Lrol tub2 -b CTCTAGGTTC ATCTTCAAAC CGGTCAGTGC GTAAGTGACA AATCCGCCGA Lrrl tub2 -b CTCTAGGTTC ATCTTCAAAC CGGTCAGTGC GTAAGTGACA AATCCGCCGA Lrr2 tub2-b CTCTAGGTTC ATCTTCAAAC CGGTCAGTGC GTAAGTGACA AATCCGCCGA 2 5 1 3 0 0 e1 9 tub2-2 CCTCGAACGA CAGGCACAAA CA . . GCATGA AAAA . CTCAC ATTCATTTGG Lc4 tub2 CCTCGAACGA CAGGCACAAA TA . . ACATGA AAAA . CTCAC ATTGATTTGG Lml tub2-a CCTCGAAC CAGGCACAAA CA . . GCATGA AAAAACTCAC ATTCATTTGA Lm2 tub2-a CCTCGAACAA CAGGCACAAA CA . . GCATGA AAAAACTCAC ATTCATTTGA Lrol tub2-a CCTCGAAC CAGGCACAAA CA . . GCATGA AAAAACTCAC ATTCATTTGA Lrr2 tub2-a CCTCGAAC CAGGCACAAA CA . . GCATGA AAAAACTCAC ATTCATTTGG Lrrl tub2 -a CCTCGAACAA CAGGCACAAA CA . . GCATGA AAAATCTCAC ATTCATTTGG Lml tub2-b CCTCGAACGA CAGGCACAAA CAACACATGA AAAA . CTCAC ATTTCTTTGG Lm2 tub2-b CCTCGAACGA CAGGCACAAA CAACACATGA AAAA . CTCAC ATTTCTTTGG Lrol tub2-b CCTCGAACGA CAGGCACAAA C CACATGA AAAA . CTCAC ATTTCTTTGG Lrrl tub2-b CCTCGAACGA CAGGCACAAA C CACATGA AAAA . CTCAC ATTTCTTTGG Lrr2 tub2-b CCTCGAACGA CAGGCACAAA C CACATGA AAA . CTCAC ATTTCTTTGG 20 1 3 0 1 3 5 0 e19 tub2-2 GCAGGGTAAC CAAATTGGTG CTGCTTTCTG GCAGACCATC TCTGGCGAGC Lc4 tub2 GCAGGGTAAC CAAATTGGTG CTGCTTTCTG GCAGACCATC TCTGGCGAGC Lml tub2-a GCAGGGTAAC CAAATTGGTG CTGCTTTCTG GCAGACCATC TCTGGCGAGC Lm2 tub2-a GCAGGGTAAC CAAATTGGTG CTGCTTTCTG GCAGACCATC TCTGGCGAGC Lrol tub2-a GCAGGGTAAC CTAATTGGTG CTGCTTTCTG GCAGACCATC TCTGGCGAGC Lrr2 tub2-a GCAGGGTAAC CAAATTGGTG CTGCTTTCTG GCAGACCATC TCTGGCGAGC Lrrl tub2-a GCAGGGTAAC CAAATTGGTG CTGCTTTCTG GCAGACCATC TCTGGCGAGC Lml tub2-b GCAGGGTAAC CAAATTGGTG CTGCTTTCTG GCAGACCATC TCTGGCGAGC Lm2 tub2-b GCAGGGTAAC CAAATTGGTG CTGCTTTCTG GCAGACCATC TCTGGCGAGC Lrol tub2-b GCAGGGTAAC CAAATTGGTG CTGCTTTCTG GCAGACCATC TCTGGCGAGC Lrrl tub2-b GCAGGGTAAC CAAATTGGTG CTGCTTTCTG GCAGACCATC TCTGGCGAGC Lrr2 tub2-b GCAGGGTAAC CAAATTGGTG CTGCTTTCTG GCAGACCATC TCTGGCGAGC 3 5 1 4 0 0 e19 tub2-2 ACGGCCTCGA CAGCAATGGT GTGTACAATG GTACCTCCGA GCTCCAGCTC Lc4 tub2 ACGGCCTCGA CAGCAATGGT GTGTACAATG GTACCTCCGA GCTCCAGCTC Lml tub2-a ACGGCCTCGA CAGCAATGGT GTGTACAATG GTACCTCCGA GCTCCAGCTC Lm2 tub2-a ACGGCCTCGA CAGCAATGGT GTGTACAATG GTACCTCCGA GCTCCAGCTC Lrol tub2-a ACGGCCTCGA CAGCAATGGT GTGTACAATG GTACCTCCGA GCTCCAGCTC Lrr2 tub2-a ACGGCCTCGA CAGCAATGGT GTGTACAATG GTACCTCCGA GCTCCAGCTC Lrrl tub2-a ACGGCCTCGA CAGCAATGGT GTGTACAATG GTACCTCCGA GCTCCAGCTC Lml tub2-b ACGGTCTCGA CAGCAATGGT GTGTACAACG GTACCTCCGA GCTCCAGCTG Lm2 tub2-b ACGGTCTCGA CAGCAATGGT GTGTACAACG GTACCTCCGA GCTCCAGCTG Lrol tub2-b ACGGTCTCGA CAGCAATGGT GTGTACAACG GTACCTCCGA GCTCCAGCTG Lrrl tub2-b ACGGTCTCGA CAGCAATGGT GTGTACAACG GTACCTCCGA GCTCCAGCTG Lrr2 tub2-b GCGGTCTCGA CAGCAATGGT GTGTACAACG GTACCTCCGA GCTCCAGCTG 4 0 1 4 5 0 e 19 tub2-2 GAGCGTATGA GTGTCTACTT CAACGAGGTA AGTCTTCATA ATCTAAAGTC Lc4 tub2 GAGCGTATGA GTGTCTACTT CAACGAGG . . . . . � . . . . . . . . . . CAAGTC Lml tub2-a GAGCGTATGA GTGTCTACTT CAACGAGGTA AGTCTTCATA ATCTAAAGTC Lm2 tub2-a GAGCGTATGA GTGTCTACTT CAACGAGGTA AGTCTTCATA ATCTAAAGTC Lrol tub2-a GAGCGTATGA GTGTCTACTT CAACGAGGTA AGTCTTCATA ATCTAAAGTC Lrr2 tub2-a GAGCGTATGA GTGTCTACTT CAACGAGGTA AGTCTTCATA ATCTAAAGTC Lrrl tub2-a GAGCGTATGA GTGTCTACTT CAACGAGGTA AGTCTTCATA ATCTAAAGTC Lml tub2-b GAGCGTATGA GTGTCTACTT CAACGAGGTA AGTCTTCATA ATCTAAAGTC Lm2 tub2-b GAGCGTATGA GTGTCTACTT CAACGAGGTA AGTCTTCATA ATCTAAAGTC Lrol tub2-b GAGCGTATGA GTGTCTACTT CAACGAGGTA AGTCTTCATA ATCTAAAGTC Lrrl tub2 -b GAGCGTATGA GTGTCTACTT CAACGAGGTA AGTCTTCATA ATCTAAAGTC Lrr2 tub2-b GAGCGTATGA GTGTCTACTT CAACGAGGTA AGTCTTCATA ATCTAAAGTC 4 5 1 5 0 0 e1 9 tub2-2 TCCATTGAGC TACATACATA CCGCCCCGGA GATGGGACGG AAAGAGAACG Lc4 tub2 TCCATTGAGC . . . . TACATA CCGCCCTG GATGGGACGG AAAAAGAAGG Lml tub2-a TCCGTTGAGC TACATACATA CCGCCCCGGA GATGAGACGG AAAGAGAACG Lm2 tub2-a TCCGTTGAGC TACATACATA CCGCCCCGGA GATGAGACGG AAAGAGAACG Lrol tub2-a TCCGTTGAGC TACATACATA CCGCCCCGGA GATGAGACGG AAAGAGAACG Lrr2 tub2-a TCCGTTGAGC TACATACATA CCGCCCCGGA GATGAGACGG AAAGAGAACG Lrrl tub2-a TCCGTTGAGC TACATACATA CCGCCCCGGA GATG CGG AAAGAGAACG Lml tub2-b TCCATTGAGC . . . . TACATA CCGCCCTGGA GATGAGACGG AAAGAGATCG Lm2 tub2-b TCCATTGAGC . . . . TACATA CCGCCCTGGA GATGAGACGG AAAGAGATCG Lrol tub2 -b TCCATTGAGC . . . . TACATA CCGCCCTGGA GATGAGACGG AAAGAGATCG Lrrl tub2-b TCCATTGAGC . . . . TACATA CCGCCCTGGA GATGAGACGG AAAGAGATCG Lrr2 tub2-b TCCATTGAGC . . . . TACATA CCGCCCTGGA GATGAGACGG AAAGAGATCG 202 5 0 1 5 4 2 e1 9 tub2-2 AAAGAAAAAG TGTTATCATG CTCATCCATG TGACAGGCTT CT Lc4 tub2 AAAGAAAAAG TGTTATCATG CT TCTATG TGACAGGCTT CT Lml tub2-a AAAGAAAAAG TGTTATCATG CTCATCCATG TGACAGGCTT CT Lm2 tub2-a AAAGAAAAAG TGTTATCATG CTCATCCATG TGACAGGCTT CT Lrol tub2-a AAAGAAAAAG TGTTATCATG CTCATCCATG TGACAGGCTT CT Lrr2 tub2-a AAAGAAAAAG TGTTATCATG CTCATCCATG TGACAGGCTT CT Lrrl tub2 -a AAAGAAAAAG TGTTATCATG CTCATCCATG TGACAGGCTT CT Lml tub2-b · . ATAAAAAG TGTG CATG CT TCTATG TGACAGGCTT CT Lm2 tub2-b · . ATAAAAAG TGTG CATG CT TCTATG TGACAGGCTT CT Lrol tub2-b · . ATAAAA G TGTG CATG CT TCTATG TGACAGGCTT CT Lrrl tub2-b · . ATAAAAAG TGTG CATG CT TCTATG TGACAGGCTT CT Lrr2 tub2-b · . ATAAAAAG TGTG CATG CT TCTATG TGACAGGCTT CT 203 tub2-a and Lm2 tub2-a were identical , as were the E. bromicola- like genes Lml tub2-b, Lm2 tub2-b, and Lro l tub2-b. 4.2.5 PHYLOGENETIC ANALYSIS OF DNA SEQUENCES To determine the most closely related genes to those found in the endophytes of annual ryegrasses, phylogenetic analyses were canied out using tub2 and rDNA-ITS sequences from these i solates, as well as representative isolates from Epichloe species and closely related Neotyphodium i solates from perennial ryegrass, tall fescue and meadow fescue (GenBank accession numbers given in Appendix 5 ; Leuchtmann and Schardl , 1998 ; Schardl et aI . , 1 99 1 ; Schardl and Leuchtmann, 1 999; Schardl et aI . , 1 997; Schardl et aI . , 1 994; Tsai e t aL, 1 994) . Sequences were aligned using PILEUP (Section 2 . 1 8 . 1 ), and gene trees using both distance and maximum parsimony (MP) criteria were generated from sequence alignments using PAUP* Version 4 .64 (Sections 2 . 1 8 .2 to 2 . 1 8 .3) . Sequence alignment files are given in Appendix 6 . For neighbor-joining (NJ) analysis of the rDNA-ITS sequences, 89.33% constant sites were observed in the data, and 8 1 .63% of the sites were estimated as invariable, and excluded from analysis . Similarly, 79.73% constant sites were observed in the tub2 alignment, and 64.47% of the sites were estimated as invariable and excluded from neighbor joining analyses (Section 2 . 1 8 .3) . The support for c lades within the generated trees was verified by bootstrap analysis as described (Section 2 . l 8) . Four most parsimonious trees, of length 55 steps, were obtained from MP analysis of the rDNA-ITS sequence alignment (Section 2. 1 8 .2), and the strict consensus of these i s shown in Figure 4.6. Bootstrap analysis, using 100 replicates, was performed to detelmine the support for clades within this topology. The basic topology of this tree was similar to the one by Schardl et a1 . ( 1997), and likewise, boots trap analysis indicated a lack of support for many of the internal branches. This is likely to be due to homoplasy and/or the smaller number of informative characters compared to tub2 sequences (Schardl et aI . , 1 997). However, bootstrap replication values indicate there is good 204 Fig. 4.6. rDNA-ITS gene phylogeny based on Epichloe and annual ryegrass endophyte sequences The strict consensus tree of the four most parsimonious gene trees identified by a branch­ and-bound search is shown. The total length of each of the maximum parsimony trees was 55 steps. The rDNA-ITS sequences are from representative Epichloe endophytes (GenBank accession numbers given in Appendix 5) and annual ryegrass endophyte i solates (shown underlined). The percentages (>50%) that each clade was represented in 100 bootstrap replications are indicated on their respective branches. The left-hand edge of the tree was inferred by midpoint rooting. 78 76 73 88 94 53 61 E. typhina E430 LcTG-1 Lc4 E. typhina E2461 E. typhina E8 E. g/yceriae E277 FaTG-3 Tf1 8 E. sy/vatica E503 E. brachye/ytri E 1 045 LmTG-1 Lm1 LmTG-1 Lm2 LmTG- 1 Lro1 LmTG-1 Lrr1 LmTG-1 Lrr2 N. uncinatum e1 66 E. e/ymi E56 E. bromicola E502 E. festucae E28 N. lolii Lp5 FaTG-2 e41 N. coenophia/um e 1 9 E. baconii E248 E. amarillans E 1 1 4 205 206 support for the clusters of interest within the tree. Specifically, there was 78% support for the clade grouping Lc4 with E. typhina isolate E430, and 88% support for the c lustering of Lml , Lm2, Lrr l , Lrr2, and Lro l with N. uncinatum isolate e 1 66. The topologies of the neighbor-joining trees, constructed using both the Kimura 2-parameter and Jukes-Cantor distance matrices (Section 2 . 1 8 .3), were very simi lar to each other, but differed from the MP consensus tree in the branching order of E56 and the annual ryegrass endophyte clade (results not shown). Forty most parsimonious trees of length 1 3 1 steps were obtained from MP analysis of the tub2 sequence alignment (Section 2 . 1 8 .2) by a heuristic search, and the strict consensus of these i s shown in Figure 4 .7 . The basic topology of this tree conforms closely to that of all Epichloe mating populations generated by Schardl et al . ( 1 997). One hundred bootstrap replicates were performed and high bootstrap replication percentages were obtained for many of the tree's internal nodes, indicating that many patterns in the data supported these clades . Furthermore, the tree topologies obtained from both NJ methods employed (Section 2 . 1 8 .3) were almost identical to the MP trees (results not shown) . The tub2-a copies from Lm1 , Lm2, Lrr 1 , Lrr2, and Lro1 grouped with the E. baconii-like tub2-4 genes from all tall fescue endophytes. The tub2-b copies grouped with the E. bromicola i solates; the bootstrap percentages for these clades were 99% and 100% respectively . The tub2 gene from Lc4 grouped with that of E. typhina isolate E102 1 , the tub2-3 gene copy from N. coenophialum isolate e 1 9, and the single tub2 copy of N. uncinatum isolate e 166. The bootstrap value for this clade is 84%, and pairwise comparisons of the tub2 sequences within the clade show that there is only one nucleotide substitution difference_between E102 1 and each of the other isolates. 4.2.6 MICROSA TELLITE ANALYSIS To rapidly determine the genetic diversity of endophyte i solates from all seven annual ryegrass species, including those that cultures were not available from, microsatellite fingerprints using Primer Set I (Section 3 .2 .2.6) were amplified either from fungal genomic DNA, or from in planta DNA preparations. In planta amplification often Fig. 4.7. tub2 gene phylogeny based on Epichloe and annual ryegrass endophyte sequences 207 The strict consensus tree of the 40 most parsimonious trees identified by a heuristic search is shown. The total length of each of the maximum parsimony trees was 1 3 1 steps. The tub2 sequences of introns 1 - 3 are from representative Epichloe endophytes (GenBank accession numbers given in Appendix 5) and annual ryegrass endophyte isolates (shown underlined). The percentages (>50%) that each clade was represented in 100 bootstrap replications are indicated on their respective branches. The left-hand edge of the tree was inferred by midpoint rooting. The topologies of the distance-based neighbor-joining trees, using both Jukes-Cantor and Kimura 2-parameter distance matrices, closely matched the MP trees. 84 69 99 61 90 96 99 93 55 76 79 90 1 00 82 1 00 68 98 96 59 58 64 56 72 N. coenophialum e 19. tub2-3 N. uncinatum e 166 E typhina E l02 1 LcTG- l Lc4 E typhina E430 E sylvatica E503 E typhina E505 E. sylvatica E354 E typhina E348 E. typhina E469 E typhina E2466 E. typhina E8 FaTG-3 T11 8. tub2-5 E. typhina E425 FaTG-2 e41 . tub2-4 FaTG-3 Tf1 8, tub2-4 N. coenophialum e19 , tub2-4 Lm TG-l Lm 1, tub2-a LmTG-l Lm2, tub2-a LmTG-l Lro 1 , tub2-a LmTG-l Lrr2, tub2-a LmTG-l Lrrl, tub2-a E baconii E248 FaTG-2 e41 , tub2-2 N. coenophialum e19, tub2-2 E. festucae E 1 89 E. festucae E32 N. lolii Lp5 E baconii E l 03 1 E. amarillans E52 E amarilians E57 LmTG-l Lml . tub2-b LmTG- l Lm2. tub2-b LmTG-l Lrol , tub2-b LmTG-l Lrrl, tub2-b LmTG-l Lrr2, tub2-b E. bromicola E502 E bromicola E798 E bromicola E50l E glyceriae E277 E. glyceriae E2772 E elymi E 184 E. elymi E56 E. brachyefytri E l 040 208 209 yielded very low amounts of product, so a second round of amplification was performed on a 1 1 100 dilution of the peR reaction before automated analysis on an ABI Prism 377 DNA Sequencer (Section 2 . 16) . The microsatellite genotypes of the annual ryegrass endophyte i solates are listed in Table 4 . 1 . Al l isolates , except Lc4, have very similar genotypes, with amplification at four of the five microsatellite loci . No amplification products were obtained from locus B9. AlleIes of size 84/85 , 179, and 109 ntu were amplified at loci B4, B6, and B 1 1 respectively, and a second all ele was occasional ly amplified at locus B4, which was about 1 30 ntu. Microsatellite locus B 10 was highly polymorphic with a total of five alleles observed at this locus ( 1 52 , 1 55 , 1 58 , 1 6 1 and 1 66 ntu). Two alleles were consistently amplified at B 10, suggestive of a hybrid origin for these endophytes. Several of the B 10 alleles were also observed in other closely related endophytes. For example, N. uncinatum also has a B4 allele of 85 ntu, B6 of 1 79 ntu, B l O of 1 6 1 ntu and no amplification at locus B9. E. bromicola shared the same B4, B6 and B9 alleles with the common endophyte, but there were no E. baconii alleles in common. Additionally, isolates of N. coenophialum have been observed to contain B 10 of 1 6 1 ntu and E. Jestucae B 10 had 1 66 ntu. Only two microsatellite loci amplified from isolate Lc4. These were B4 and B l O, whose allele sizes were 1 04 and 1 8 1 ntu respectively. This genotype is identical to that of E. typhina strains E 102 1 and E1022 from Poa nemoralis (Table 3 .4). Lc4 also shares alleles with N. uncinatum (locus B4 contains 1 04 ntu allele) and FaTG-3 (locus B 10 contains 1 8 1 ntu allele). Three other seedlines of endophyte infected L. canariense tissue were also examined and it is interesting to note that the microsatel l ite profiles of these endophytes, Lc 1 , Lc2, and Lc3, were similar to those of the rest of the annual ryegrass endophytes (Table 4. 1 ). 2 10 Table 4.1 . Microsatellite genotypes o f annual rye grass endophyte isolates Isolate Tissue t Allele size (ntu) B4 B6 B lO B l l 151.6* Lc l IP 84.3 178.8 108.8* 166 .2* 151.9* Lc2 IP 84.7* 178.8* *109.9 *167.4 84.2* 151.6* Lc3 IP 178.8 108.9* 130.4* 166.2* Lc4 C 104.4 180.8* 151.8 Lm1 C 85.0 178.8 109.0 166.4 151.9* Lm2 C *86.5 178.7 108.8* 166.4* 151.6* Lm3 IP 84.3* 178.7 109.0* 166.2* 84.1 151.6* Lps l IP 178.8 108.9* 130.5 166.2* 151 .6* Lps2 IP 84.4 178.8 108.9* 166.2* 154.6* Lre1 IP 85.4* 178.6* *109.9 *161.7 151.6* Lre2 IP 84.2 178.8 108.9* 166.2* 84.3* 151.7* Lre3 IF 178.8 108.9* 130.3* 166.3* 154.7* LIT 1 C 84.8 178.7 108.9* 160.6* *155 .7 Lrr1 IP 86.0 178.7* *109.9 *161.7 84.4* 151.7* Lrr2 IP 178.7 108.9* 130.8* 157.7* 84.3* 151.7* Lrr3 IP 178.8 108.8* 131.2 166.2* 151.8* Lrr4 C 85.7 178.8* 108.9* 166 .4* 2 1 1 Table 4.1. continued Isolate Tissue t Allele size (ntu) B4 B6 B lO Bll *152.9 Lro1 C 86.8 178 .8* *109.8 *167.5 84.3* 151.6* Ls1 IP 178.7 108.9* 131.3 157.5* *85.3 151.9* L t l IP 179.0* *110.1 130.6 166.6* 84.7 151.8 * Lt2 IP *179.9 *109.9 130.8 * 166.6* t Fingerprints were amplified fj'om DNA prepared fi'om in planta (IP), or fungal culture (C) t issue * an asterisk shown after the allele indicates the presence of a band -1 ntu larger than that shown, and an asterisk shown before the allele indicates that it is assumed that this is the +A band only present alleles in boldface indicate that an extremely large peak height was observed when this allele was automatically analysed, and size-calling may not be accurate 2 1 2 4.2.7 ANAL YSIS OF ENDOPHYTE ISOLATES FROM HORDEUM GRASSES Two poorly described endophyte i solates, HaB and Hdl , from Hordeum grasses were available in the Scott Laboratory fungal culture collection . These isolates were recovered from -70°C storage as agar blocks in 50% glycerol , and maintained on PDA plates (Section 2.2 .2) for further study. 4.2.7.1 Culture Descriptions Figure 4 .8 shows four week old colonies of HaB (Fig. 4. 8A) and Hdl (Fig . 4 .8B) growing on PDA medium (Section 2 .2.2). HaB formed white raised colonies with abundant cottony mycelia, which were aggregated into erect tufts at the centres. Colonies have superficial margins. HaB is a fast growing isolate, with an average radial growth rate of 0.57 mm d-I over a four week period. Moderate conidia production was observed. In contrast, Hdl formed flat white felted colonies with superficial margins. The average radial growth rate was 0.64 mm d- I over a four week period, and conidia production was abundant. For all genetic analyses that fol low, DNA from these i solates was prepared from l iquid cultures of fungal mycelia, using the large scale i solation procedure (Section 2 .5 . 1 ) . 4.2.7.2 rDNA-ITS Analysis The rDNA-ITS regions of HaB and Hdl were amplified from fungal genomic DNA (Section 2 .5 . 1 ) as described in Section 4.2.3 . Amplification products were resolved as single bands of around 600 bp, simi lar to those shown in Figure 4.2, when separated on 2% agarose gels. Products were purified and both strands sequenced (Section 2 . 1 5), the sequences for these and the rDNA-ITS region from N. coenophialum i solate e 1 9 are shown in an alignment in Figure 4 .9. The HaB and Hdl rDNA-ITS sequences have 96.7% identity to one another. BLAST database similarity searches (Section 2. 1 5 .3) were conducted and the HaB sequence identified rDNA-ITS sequences from 2 1 3 Fig. 4.8 . Cultures o f Hordeum endophyte i solates, HaB and Hdl Four week old cultures of HaB and Hdl that were grown on PDA at 22°C are shown. Scale bar is 10 mm long. 2 14 A B 2 1 5 Fig. 4.9. Alignment of rDNA-ITS sequences from Hordeum endophyte isolates The DNA sequences of the rDNA-ITS regions from Hordeum endophyte isolates HaB and Hd 1 , and N. coenophialum isolate e 1 9 were aligned using the PILEUP programme of the Wisconsin Package with the gap penalty set to one and the gap extension penalty set to zero. Gaps (.) were introduced to facilitate sequence alignment. The sequenced region from e 1 9 encodes ITS l (positions 1 - 1 70), 5 .8S rRNA (positions 1 7 1 - 338, shown italicised), and ITS2 (positions 339 - 508) (Schardl et aI . , 1 99 1 ). Nucleotides that differ from the reference sequence, e 1 9, are highlighted in green. 1 5 0 e 1 9 CGAGTTTACA CTCCC . AAAC CCCTGTGAAC CT . TACCTTT ACTGTTGCCT Hd1 CGAGTTTACA CTCCC . AAAC CCCTGTGGAC CTATACCTTT ACTGTTGCCT HaB CGAGTTTACA CTCCC]\AAAC CCCTGTGCAC CT,4.TACCTCT ACTGTTGCCT 5 1 1 0 0 e 1 9 CGGCGGGCAC GGCCGCGGAC GCCCCCTCGC GGGGGCACCG GGGCCGGGCG Hd1 CGGCGGGCAC GGCCGCGGAC GCCCCCTCGC GGGGGCACCG GGGCCAGGCG HaB CGGCGGGCAC GGCCGCGGAC GCCCCCTCGC GGGGGCGCCG GGGCTAGGCG 1 0 1 1 5 0 e 1 9 CCCGCCGGAG GACCCAAACC CTTCTGTATT TTTCTTACGC ATGTCTGAGT Hd1 CCCGCCGGAG GACCCAAACC CTTCTGTA . T TTTCTTACGC ATGTCTGAGT HaB CCCGCCGGGG GACCCAAACC CTTCTGTA . T TTTCTTACGC ATGTCTGAGT 1 5 1 2 0 0 e 1 9 GGATTTAATA TCAAATGAAT CAAAACTTTC AACAACGGAT CTCTTGGTTC Hd1 GGATT . TA , CAAATGAAT CAAAACTTTC AACAACGGAT CTCTTGGTTC HaB GGATTGAATA TCGAATGAAT CAAAACTTTC AACAACGGAT CTCTTGGTTC 2 0 1 2 5 0 e 1 9 TGGCATCGAT GAAGAACGCA GCGAAATGCG ATAAGTAATG TGAATTGCAG Hd1 TGGCATCGAT GAAGAACGCA GCGAAATGCG ATAAGTAATG TGAATTGCAG HaB TGGCATCGAT GAAGAACGCA GCGAAATGCG ATAAGTAATG TGAATTGCAG 2 51 3 0 0 e 1 9 AATTCAGTGA ATCATCGAAT CTTTGAACGC ACATTGCGCC CGCCAGTATT Hd1 AATTCAGTGA ATCATCGAAT CTTTGAACGC ACATTGCGCC CGCCAGTATT HaB AATTCAGTGA ATCATCGAAT CTTTGAACGC ACATTGCGCC CGCCAGTATT 2 1 6 3 0 1 3 5 0 e 1 9 CTGGCGGGCA TGCCTGTTCG AGCGTCATTT CAACCCTCAA GCCCGCTGCG Hd1 CTGGCGGGCA TGCCTGTTCG AGCGTCATTT CAACCCTCAA GCCCGCTGCG HaB CTGGCGGGCA TGCCTGTTCG AGCGTCATTT CAACCCTCAA GCCCGCTGCG 3 5 1 4 0 0 e 1 9 CGCTTGCTGT TGGGGACCGG CTCACCCGCC TCGCGGCGGC GGCCGCCCCC Hd1 TGCTTGGTGT TGGGGACCGG CCAGCCCGCC TCGCGGCGGC GGCCGCCCCT HaB CGCTTGGTGT TGGGGACCGG CTCGCCCGCC TCGCGGCGGC GGCCGCCCCT 4 0 1 4 5 0 e 1 9 GAAATGAATC GGCGGTCTCG TCGCAAGCCT CCTTTGCGTA GTAGCACACC Hd1 GAAATGAATT GGCGGTCTCG CCGC . AGCCT TCTTTGCGTA GTAACACACC HaB GAAATGAATC GGCGGTCTCG TCGC . AGCCT CCTTTGCGTA GTAACACACC 4 5 1 5 0 0 e 1 9 ACCTCGCAAC CGGGAGCGCG GCGCGGCCAC AGCCGTAAAA CGCCCAACTT Hd1 ACCTCGCAAC CAGGAGCGCG GCGCGGCCAC TGCCGTAAAA CGCCCAAC . T HaB ACCTCGCAAC CGGGAGCGCG GCGCGG . CAC TGCCGTAAAA CGCCCAAC . T 5 0 1 e 1 9 TCTCCAAG Hd1 TCTCCAAG HaB TCTCCAAG 2 17 E. amarillans, and Hdl identified rDNA-ITS sequences from E. typhina and E. glyceriae in the first five entries of each search. As with the rDNA-ITS phylogenetic analysis of the annual ryegrass endophytes (Section 4.2.5), to determine the most c losely related genes to those of the Hordeum endophyte isolates, phylogenetic analyses were performed using rDNA-ITS sequences from HaB and Hdl , as well as the same representative i solates from Epichloe and Neotyphodium i solates that were used in Figure 4.6. Sequences were aligned (Appendix 6), and gene trees were constructed using both distance and MP criteria. For NJ analyses, 89 .83% constant sites were observed in the data, and 83.72% of the sites were estimated as invariable and excluded (Section 2 . 1 8 .3) . The support for clades within the generated trees was determined by bootstrap analysis (Section 2 . 1 8) . Sixty-nine most parsimonious trees of length 54 steps were obtained from MP analysis of the rDNA-ITS sequence alignment using a branch-and-bound search , and the strict consensus of these is shown in Figure 4. 10 . The basic topology of this tree is unresolved, unlike the one by Schardl et al . ( 1 997) . Bootstrap analysis indicated a lack of support for many of the branches, and hence the overall topology of the tree. Again, this is likely to be due to homoplasy and/or the smaller number of informative characters compared to tub2 sequences, however bootstrap replication values indicate there is reasonable support for the clusters of interest within the tree. Specifical ly, there was 72% support for the clade grouping Hdl with E. typhina and E. glyceriae isolates, and 99% support for the clustering of HaB with E. amarillans isolate E1 l4. The topologies of the NJ trees, constructed using both the Kimura 2-parameter and Jukes-Cantor distance matrices (Section 2 . 1 8 .3) , were almost identical , and closely matched the NJ trees generated from the analysis of annual ryegrass endophyte rDNA-ITS data (Section 4.2 .5) , though minor differences in the branching order of E1045 and E56 were apparent (results not shown). 2 1 8 Fig . 4 . 10 . rDNA-ITS gene phylogeny based on Epichloe and Hordeum endophyte i solate sequences The strict consensus tree of the 69 most parsimonious trees identified by a branch-and­ bound search is shown. The total length of each of the maximum parsimony trees was 54 steps. The rDNA-ITS sequences are from representative Epichloe endophytes (GenBank accession numbers given in Appendix 5) and Hordeum endophyte i solates (shown underlined) . The percentages (>50%) that each clade was represented in 100 bootstrap replications are indicated on their respective branches. The left-hand edge of the tree was inferred by midpoint rooting. 72 77 99 95 62 E. glyceriae E277 HdTG-2 Hd1 E. typhina E8 E. typhina E2461 FaTG-3 Tf1 8 E. sylvatica E503 E. typhina E430 E. brachyelytri E1 045 N. uncinatum e1 66 E. bromicola E502 E. amarilfans E 1 1 4 HdTG-1 HaS E. festucae E28 N. lolii Lp5 FaTG-2 e41 N. coenophialum e 1 9 E. baconii E248 E. elymi E56 2 1 9 220 4.2.7.3 tub2 Gene Analysis The tub2 genes of HaB and Hdl were amplified from fungal genomic DNA as desctibed in Section 4.2.4. Like the annual ryegrass endophytes , the Hordeum endophytes also yielded multiple bands, as shown in lanes 6 and 7 of Figure 4.4. These products were cloned and sequenced using the pue l 1 8 cloning method implemented for the annual ryegrass endophyte tub2 analysis (Section 4.2.4) . Sequence analysis of the various clones identified two 'true' tub2 genes from each Hordeum endophyte i solate, and these gene copies were designated tub2-c, tub2-d from HaB and tub2-e and tub2J from Hdl . The sequence of the 620 bp fragment from Hd1 (lane 7 , Fig. 4.4) showed no significant similarity to entries in the database by BLAST searches (Section 2 . 15 .3) , so it was presumed that this was a spurious amplification product and was not included in further analysis . BLAST database searches of the tub2-c copy from HaB identified tub2 genes from E. elymi (the first three entries), the tub2-d copy from HaB identified tub2 genes from E. amarillans (the first three entries), the tub2-e copy from Hdl identified tub2 genes from E. typhina (the first six entries), the tub2J copy from Hdl identified tub2 genes from E. bromicola and E. elymi (the first four entries). The tub2 gene copies from HaB and Hdl were aligned with the tub2-2 gene copy from N. coerLOphialum isolate e 1 9 and this alignment i s shown Figure 4. 1 1 The most closely related tub2 genes to those of the Hordeum endophyte i solates were determined by phylogenetic analyses using the tub2 sequences from HaB and Hd1 , as well as the same representative i solates from E.pichloe and Neotyphodium isolates that were used in the tub2 analysis of Figure 4 .7 . Sequences were aligned and gene trees, using both distance and MP, were generated. For NJ analyses, 79. 8 1 % constant sites were observed in the data, and 62. 89% of the sites were estimated as invariable and excluded (Section 2 . 1 8 .3) . The support for clades within the generated trees was verified by boots trap analysis (Section 2 . 1 8) . 22 1 Fig. 4 . 1 1 . Alignment of the tub2 gene sequences from Hordeum endophyte isolates The DNA sequences of the tub2 genes from the Hordeum endophyte isolates HaB and Hd l , and N. coenophialum isolate e 1 9 (tub2-2 copy) were aligned using the PILEUP programme of the Wisconsin Package, with the gap penalty set to one and the gap extension penalty set to zero. Gaps ( . ) were i ntroduced to faci litate sequence alignment. The sequenced region from e 1 9 encodes IVS 1 (positions 1 - 209), exon 2 (positions 2 1 0 - 234), IVS2 (positions 235 - 306), exon 3 (positions 307 - 430), IVS3 (positions 43 1 - 543), and exon 4 (positions 544 - 549) (Tsai et aI ., 1 994). Exon positions are shown italicised, and nucleotides that differ from the reference sequence, e 1 9, are highlighted i n green. Ambiguous (G, A, T, or C) nucleotide positions are denoted b y N. e1 9 t ub2-2 HaB tub2-d HaB tub2-c Hdl tub2-f Hdl tub2-e e 1 9 t ub2-2 HaB tub2-d HaB tub2-c Hdl tub2-f Hd1 t ub2-e e 1 9 t ub2-2 HaB tub2 -d HaB tub2-c Hd1 t ub2-f Hd1 tub2-e e1 9 t ub2-2 HaB tub2-d HaB tub2-c Hd1 t ub2-f Hd1 tub2-e e 1 9 t ub2-2 HaB tub2-d HaB tub2-c Hd1 tub2-f Hd1 t ub2-e 1 5 0 AAGTTCAACC TCTCTGTTTG TCTTGGGGAC CCCCTCCTCG ACGCGTTCCG NNNNNNNNCC TCTCTGTTTG TCTTGGGGAC CCCCTCCTCG ACGCGTTCCG NCGTTTAACC �;CTCTGTTTG TCTTGGGGAC CCCCTCCTCG ACGCGTTCCG NNNNNNNNNN NNNNNNTG TCTTGGGGAC CCCCTCCTCG ACGCGTTCCI\ NNNNNNNNNN NNTCTG TTG TCTTGGGGAC CCCCTTCTCG ACGCGTTCC!, 5 1 GTGTTGAGCC CCTGATTTCG TACCCCGCCG AGCCCGGCCA GTGTTGAGCC CCTGATTTCG TACCCCGCCG AGCCCGGCCA GTGTTGAACC CCTGATTTCG TACCCCGCCG AGCCCGGCCA GTGTTGAGCC CCTGATTTCG TACCCCGCCG AGCCCGGCCA GTGTTGAGCC CCTGATTTCG TACCCCGNCG AGCCCGGNCA 1 0 1 GCCCAACGAA C . AGTCGTGA TGAGAGGCGG ACCGAGAC . . GCCCAACGAA C . AGTCGTGA TGAGAGGCGG ACCGAGAC . . GCCCAATGGA CAAGTCGTGA TGAGAGGCGG ACCGAGAC . . GCCCAA ;'GGA C . AGTCGTGA TGAGAGGCGG ACCGAGACAA GCCCAACGGA C ?\AGTTGTGA TGAGAGGCGG ACCGAGAC . . 1 5 1 TGAATGCGGT ATTCGAGAAC TG . TAGCTGA CCTG . TTCTT ('GAATGCGGT ATTCGAGAAC TG . TAGCTGA CCTG" 'TTCTT TGATTGCGGT ATTCGAGAAC TG'TTAGCTGA CCTT T'TTC . TGATTGCGGT ATTCGAGAAC TG . T( 'GCTGA CAT ;'TTl\C , , TGAA TGCGGT ATTCGA;\AAC CG . TAGCTGA CCTYTTT , , , 2 0 1 CCCCTCTAGG TTCATCTTCA AACCGGTCAG TGCGTAA . . G CCCCTCTAGG TTCATCTTCA AACCGGTCAG TGCGTAA . . G CCCCTCTAGG TTCATCTTCA AACCGGTCAG TGCGTAA . . G CCCCTCTAGG TTCATCTTCA AACCGGTCAG TGCGTAAc 'rG ' i '( ;CCTCTAGG TTCATCTTCA AACCGG:'CAG TGCGTAA . . G 1 0 0 CGAAGTGCAC CGACGTGCAC CGACGTGCAC CGA( 'GTGCAC CGACGTGCAC 1 5 0 · . AAAATTAA · . AAAATCAA 1\;\AAAA TT AA 7\}\AAAA TT AA · . AACATTGA 2 0 0 CCCCTCTTTT 'T CCCTCTTTT ' 0 " . TT . TT , T;ITT 25 0 TGACAAATTC TGACAAATCC TGACAAAT('C TGACAAATCC TGAAAAATCC 222 2 5 1 3 0 0 e19 tub2-2 GCCGACCTCG AACGACAGGC ACAAACAGCA TGAAAAACTC ACATTCATTT HaB tub2-d GCCGACCTCG AACGACAGGC ACAAACAGCA TGAAAAACTC ACATTCATTT HaB tub2-c GCCGACCTAG AACGACGGGC ACAAAT CA TGAAAAACTC ACATTTATTT Hd1 tub2-f GCCGACCTCG AACGACAGGC ACAGAT CA TGAAAAACTC ACATTTCTTT Hd1 tub2-e GCCGACCTCG AACGACAGGC ACAGAT CA TGAAAAACTC ACATTGATTT 3 0 1 3 5 0 e19 tub2-2 GGGCAGGGTA A CCAAATTGG TGCTGCTTTC TGGCAGACCA TCTCTGGCGA HaB tub2-d GGGCAGGGTA ACCAAATTGG TGCTGCTTTC TGGCAGACCA TCTCTGGCGA HaB tub2-c GGGCAGGGTA ACCAAATTGG TGCTGCTTTC TGGCAGACCA TCTCTGGCGA Hd1 tub2-f GGGCAGGGTA ACCAAATTGG TGCTGCTTTC TGGCAGACCA TCTCTGGCGA Hd1 tub2-e GGGCAGGGTA ACCAAATTGG TGCTGCTTTC TGGCAGACCA TCTCTGGCGA 3 5 1 4 0 0 e 1 9 tub2-2 GCACGGCCTC GACAGCAATG GTGTGTACAA TGGTACCTCC GAGCTCCAGC HaB tub2-d GCACGGCCTC GACAGCAATG GTGTGTACAA TGGTACCTCC GAGCTCCAGC HaB tub2-c GCACGGCCTC GACAGCAATG GTGTGTACAA TGGTACCTCC GAGCTCCAGC Hd1 tub2-f GCACGGTCTC GACAGCAATG GTGTGTACAA CGGTACCTCC GAGCTCCAGC Hd1 tub2-e GCACGGCCTC GACAGCAATG GTGTGTACAA TGGTACCTTC GAGCTCCAGC 4 0 1 4 5 0 e19 tub2-2 TCGAGCGTAT GAGTGTCTAC TTCAACGAGG TAAGTCTTCA TAATCT . . . A HaB tub2-d TTGAGCGTAT GAGTGTCTAC TTCAACGAGG TAAGTCTTCA TAATCT . . . A HaB tub2-c TGGAGCGTAT GAGTGTCTAC TTCAACGAGG TATGTCTTCA TAATCT . . . A Hd1 tub2-f TGGAGCGTAT GAGTGTCTAC TTCAACGAGG TAAGTCTTCA TAATTT . . . A Hd1 tub2-e TCGAGCGTAT GAGTGTCTAC TTCAACGAGG CAAGTCTTCA TAATCTA 4 5 1 5 0 0 e1 9 tub2-2 AAGTCTCCAT TGAGCTACAT ACATACCGCC CCGGAGATGG GACGGAAAGA HaB tub2-d AAGTCTCCAT TGAGCTACAT ACATACCGCC CCGGAGATGA GACGGAAAGA HaB tub2-c AAGTCTCCAT TGAGC . . . . T ACATACCGCC CTGGAGATGA GACGGAAAGA Hd1 tub2-f AAGTCTCCAT TGAGC . . . . T ACATACCGCC CTGGAGATGA GACGGAAAGA Hdl tub2-e AAGTCTCCAT TGAGC . . . T ACATACCGCC CTGGAGATGG GACGGGAAAA 5 0 1 5 4 9 e 1 9 tub2-2 GAACGAAAGA AAAAGTGTTA T . . CATGCTC ATCCATGTGA CAGGCTTCT HaB tub2-d GAACGAGAGA AAA GTGTCA T . . CATGCTC ATCCATGTGA CAGGCTTCT HaB tub2-c GAACGAAAGA AAAAGTGTCA TCATATGCTA ATCTATGTGA CAGGCTTCT Hd1 tub2-f GAACG . . ATG AAAAGTGT . . TAACATGCTA ATCTATGTGA CAGGCTTCT Hd1 tub2-e GAAGCGGAGA AAAAGTATTA T . . CATGCTA ATCTATGTGA CAGGCTTCT 223 Twenty-four most parsimonious trees of length 1 23 steps were obtained from MP analysis of the tub2 gene sequence alignment (Section 2. 1 8 .2) using a branch-and-bound search, and the strict consensus of these is shown in Figure 4. 1 2. The basic topology of this tree conforms closely to that of all Epichloe mating populations generated by Schardl et al. ( 1 997) . One hundred bootstrap replications were performed (Section 2 . 1 8) and high boots trap replication percentages were obtained for most of the tree's internal nodes, indicating that many patterns in the data supported these clades . Furthermore, the tree topologies obtained from both NJ methods employed (Section 2 . 1 8 .3) were highly simi lar to the MP trees (results not shown). The tub2-c gene copy from HaB grouped with high confidence with the E. elvmi i solates E 1 84 and E56 (96% of the bootstrap trees), and the tub2-d gene copy from HaB grouped together with the E. amarillans i solate E52 in 80% of the bootstrap trees. The Hdl tub2-e copy grouped with E. typhina gene from isolates E348 and E469 (64% of bootstrap trees), and the tub2J copy grouped with the three E. bromicola i solates in 96% of the bootstrap trees . 4.2.7.4 Microsatellite Analysis Microsatellite fingerprints of HaB and Bdl were generated from fungal genomic DNA with Primer Sets I and II (Sections 2. 1 3 and 2 . 1 6) , and their genotypes are l isted in Table 3.6. Multiple alleles were observed at several loci, which supports the hybrid origins of both of these i solates, and common alleles shared with other endophytes are discussed in Section 3 .3 .3 . 1 . 1 . Fig . 4 . 1 2. tub2 gene phylogeny based on Epichloe and Hordeum endophyte i solate sequences 224 The strict consensus tree of the 24 most parsimonious trees identified by a branch-and­ bound search is shown. The total length of each of the maximum parsimony trees was 123 steps . The tub2 sequences of introns 1 - 3 are from representative Epichloe endophytes (GenBank accession numbers given in Appendix 5) and Hordeum endophyte isolates (shown underlined) . The percentages (>50%) that each clade was represented in 1 00 bootstrap replications are indicated on their respective branches. The left-hand edge of the tree was inferred by midpoint rooting. The topologies of the distance-based neighbor-joining trees, using both hIkes-Cantor and Kimura 2-parameter distance matrices, closely matched the MP trees. 68 93 64 90 66 96 71 73 58 97 86 91 95 80 59 71 96 96 62 96 78 1 00 I I I I I L E. typhina E8 FaTG-3 Tf1 8, tub2-5 E. typhina E425 E. typhina E348 E. typhina E469 HdTG-2 Hd1, tub2-e E. typhina E2466 E. syJvatica E503 E. typhina E505 E. sylvatica E354 N. coenophialum e 1 9, tub2-3 N. uncinatum e 1 66 E. typhina E 1 02 1 E. typhina E430 E. brachyelytri E 1 040 FaTG-2 e41 , tub2-4 FaTG-3 Tf1 8, tub2-4 N. coenophialum e 1 9, tub2-4 E. baconii E248 FaTG-2 e41 , tub2-2 N. coenophialum e 1 9, tub2-2 E. festucae E 1 89 E. festucae E32 N. /olii Lp5 E. amariflans E52 HdTG-1 HaB, tub2-d E. amarillans E57 E. baconit' E1 031 E. bromicola E502 E. bromicola E798 E. bromicola E501 HdTG-2 Hd1. tub2-f E. elymi E 1 84 E. elymi E56 HdTG-1 HaB, tub2-c E. g/yceriae E277 E. g/yceriae E2772 225 4.3 DISCUSSION 226 4.3.1 TWO NEW ENDOPHYTE TAXONOMIC GROUPS ARE IDENTIFIED FROM ANNUAL RYEGRASSES This study has confirmed through genetic analysis that the seed-borne endophytic fungus of Lolium annual rye grass species, which was isolated and described by Latch et al . ( 1988), i s a single distinct Neotyphodium species. We have surveyed all seven species of annual ryegrasses and have found this common endophyte in each. A second previously un described Neotyphodium endophyte was found in L. canariense. The distinctive in planta distribution of the common endophyte (Latch et aI . , 1 988) was conserved across all seven host grass species, where mycelia was confined to the basal regions of vegetative tillers, and was more extensively distributed in flowering ti llers. Cultures were isolated from L. mult(florum, L. rigidum var. rigidum and L. rigidum var. rottboellioides, but the extremely slow growth rate of this fungus made isolation difficult, so despite attempts, cultures were not obtained from the other annual ryegrass species. Nonetheless, these strains were genetical ly characterised by in planta microsatellite fingerprinting, and the results show that the degree of fingerprint variabi li ty is of a level ," similar or less than that displayed by established taxonomic grouping, or species of Neotyphodium (Moon et aI . , 1 999) . Phylogenetic analysis of the rDNA-ITS and tub2 gene sequences reveal unique evolutionary origins for this group of endophytes, and it i s proposed that this group be named Lolium mult�florum taxonomic group one (LmTG- l ), after the convention used by Christensen et al . ( 1 993), and by reference to the endophyte of L. mult(florum, as this was the first grass to be isolated from in this study. The second distinctive type of endophyte found in this study, i solate Lc4, was readily i solated from one of the four seedlines of L. canariense examined, and to our knowledge, represents the first report of such an endophyte. . The in planta distribution of Lc4 was more extensive than LmTG- l , and the cultural characteristics were clearly different, both in the rate of growth and the colony appearance. The genetic profi le of Lc4 for 227 rDNA-ITS, tub2, and microsatellite data is unique, suggesting that this i s a new species of an asexual Epichloe endophyte. It i s therefore proposed that this endophyte be named Loliurn canariense taxonomic group one (LcTG- 1 ) . It should be noted that neither of the two endophyte taxonomic groups described in this study, LmTG- 1 or LcTG-1 , have been found in other grasses, including closely related perennial ryegrass , meadow fescue or tall fescue. In addition to Epichloe endophytes , annual rye grasses have been reported to contain Acrernoniurn-like seedbome fungal endophytes, which are also basall y confined in the plant, and produce rod shaped conidia (Naffaa et aI . , 1 998). These fast growing endophytes are morphologically similar to A. chilense, though we did not observe such endophytes in thi s study. More specificall y, we did not have evidence to suggest the Acrernoniurn endophytes encountered in this study were growing as seedbome endophytes as they were not i solated from immature inflorescences of any of the annual ryegrass species. A. chilense is a seed transmitted endophyte of Dactylis glornerata (Morgan-Jones et aI . , 1 990), and though i t was renamed Neotyphodium chilense by Glenn et al. ( 1996), genetic and morphological data suggest that it is a true Acremoniurn species. The position of this endophyte within the Neotyphodiurn genus is currently under review (pers. comm. Mike Christensen, Grasslands AgResearch). 4.3.2 THE EVOLUTIONARY ORIGIN OF LmTG-l LmTG-l contained multiple copies of tub2 genes and multiple alleles at microsatell ite loci , a result suggestive o� an interspecific hybrid oligin. From phylogenetic analysis of the tub2 genes, it i s proposed that the parental endophytes include E. baconii and E. brornicola, or c lose relatives . Thi s result is also supported by the LmTG- l rDNA-ITS sequence similarities to N. uncinatum (which has E. typhina and E. brornicola ancestry), and the common microsatelli te alleles of LmTG- l endophytes with extant E. bromicola strains. This genetic data confirms that LmTG- l is a Neotyphodium of unique hybrid origin, and does not group with N. lolii, LpTG-2 or other currently recognised endophyte species. 228 An interesting question regarding the evolution of LmTG- l endophytes is how an endophyte with ancestors from E. baconii and E. bromicola origins , could have arisen in a Poeae host, given that the parents inhabit grasses of host tribes A veneae and Bromeae respectively. Higher order grass phylogeny shows that within the subfamily Pooideae, four tlibes constitute a core monophyletic group - Bromeae, Triticeae, A veneae and Poeae, and these have been referred to as the "core" pooids (Davis and Soreng, 1 993), where Poeae and A veneae form a monophyletic group, the A veneae being paraphyletic to Poeae (Catalan et al . , 1 997), and Bromeae and Triticeae form sister tribes within a monophyletic group (Catalan et al . , 1 997; Hsaio et al . , 1 995) . The most l ikely scenario for the evolution of LmTG-1 is that an ancestral E. baconii existed before the Poeae diverged from Aveneae, and was passed on to both tribes. If an E. baconii infected Poeae host, ancestral to the Lolium ryegrasses, was infected by an E. bromicola spore and hybridisation followed, this could have given rise to LmTG- l . Support for this scenario comes from the fact that E baconii-like tub2 gene copies (tub2-4) have also been found in al l three taxonomic groups of Neotyphodium from tall fescue (Tsai et al . , 1 994), a Poeae grass. These tub2-4 genes group closely with the E. baconii-like tub2 copies from the LmTG-1 endophytes, and they are somewhat diverged from extant E. baconii sequences, so the sequence differences may reflect changes that have accumulated since the Poeae/Aveneae divergence. If the proposed evolutionary scenario was true, it would be likely that E. baconii-like species are present in Poeae grasses, but possibly have not been sampled yet. An alternative hypothesis for the evolution of LmTG- l is that if E. baconii is actually confined to A veneae hosts, an ascospore infection of this endophyte into the Lolium ancestor may have established a rare stable symbiosis long enough for subsequent infection by an E. bromicola spore. Tall fescue endophyte may have arisen through infection and hybridisation of similar genotypes of E. baconii, hence the similarity in extant Neotyphodium endophyte E. baconii-like tub2 sequences. Another option to this hypothesis i s that the initial E. baconii infection could have infected a FestucalLolium ancestor, and was thus passed on to both tall fescue and rye grass endophytes. This scenario would explain the apparent absence of E. baconii-like strains throughout the 229 Poeae. However, i t is doubtful whether the same endophyte isolate could be compatible with such diverged hosts, as would be required with this scenario (Christensen et aI . , 1 997). 4.3.3 THE EVOLUTIONARY ORIGIN OF LcTG-l On the basis of genetic similarity, i t is proposed that LcTG- l is an asexual derivative of an E. typhina genotype closely related to the extant E I02 1 from Poa nemoralis. The microsatellite genotypes of these two endophytes are identical , and there is only an single nucleotide difference between the tub2 sequences. There was no rDNA-ITS sequence available from EI02 1 , but the sequence from a closely related E. typhina isolate, E430, showed high similarity to that of LcTG- l . Only single copies of microsatellite alleles, tub2 genes, and rDNA-ITS sequences were found in the LcTG-l endophyte, thus giving no evidence for interspecific hybridisation. Neotyphodium lolii is also thought to have evolved directly from a sexual Epichloe sp. (E.festucae; Leuchtmann, 1 994; Schardl, 1 996a; Schardl et aI . , 1 994), and there are interesting morphological similarities between N. lolii and LcTG- l . The colony morphologies of both endophytes are often highly unstable and heterogeneous when isolated directly from plant tissue, and both have rare or apparent lack of conidia production (Christensen and Latch, 1 99 1 ; Christensen et aI . , 1 993). It is interesting to note that unlike the majority of E. typhina i solates examined, which choke all infected flowering tillers, the E. typhina i solates from Poa nemoralis such as E102 1 , have been observed to from rare seed-borne infections, though they also normally choke all infected flowering tillers (pers. comm. Christopher Schardl , University of Kentucky). E 102 1 -like tub2 genes have now been found in three Neotyphodium taxonomic groups, including N. coenophiaium, N. uncinatum and LcTG-1 . Whether there is a relationship between the abi lity of E. typhina to become seed-borne, and the contribution of this or related isolates to taxonomic groups of Neotyphodium is not known at this time. 4.3.4 THE EVOLUTION OF LOL/UM ENDOPHYTES 230 Assuming a general mode of coevolution by common descent of grass hosts and their endophytes (Schardl et aI . , 1 997), a possible scenario for the evolution of Epichloe endophytes within Lolium ryegrasses is proposed. Several Lolium and Festuca phylogenies have been established based on RAPD profiles, rDNA-ITS sequences and chloroplast (cpDNA) restriction site data (Charmet et aI . , 1 997; Stammers et aI . , 1 995), and it is estimated that the diploid genus Lolium i s of very recent origin, and diverged from the F. pratensis lineage about 2 Myr ago. Differentiation into several Lolium species is estimated to have started only 1 Myr ago (Channet et aI . , 1 997). Because of such recent evolution, the order of speciation is not resolved with certainty, and differs depending on the type of data used. It i s generally observed that the interferti le outcrossing species, L. perenne, L. mult{florum, and L. rigidum, group together and are differentiated from the self-fertili sing species (Bennett, 1 997 ; Charmet et aI . , 1 997). The Lolium ryegrasses are very closely related to diploid F. pratensis, and this i s reflected in the ability of the outbreeding Lolium species to readily hybridise with this grass (Borril l , 1 976). The reproductive cycle of L. canariense is unusual as this grass i s generall y self­ ferti l ising, but i t also di splays moderate levels of outbreeding (Charmet and Balfourier, 1 994). L. canariense also has unique genetic affinities as rDNA-ITS and i sozyme data group it together with F. pratensis, though cpDNA analysis groups it with annual Lolium ryegrasses (Charmet and Balfourier, 1 994; Charmet et aI . , 1 997). This suggests L. canariense could be a hybtid grass of a matemal annual Lolium with F. pratensis parentage. Hexaploid tall fescue, F. arundinacea, is also closely related to the Lolium grasses, and from analysis of cpDNA and rDNA-ITS data it appears to be the result of hybridisation between a matemal tetraploid F. glaucesens with F. pratensis (Charmet et aI . , 1 997). A proposed scenario for the evolution of Lolium endophytes within the Lolium genus is shown in Figure 4. 1 3 . This is based on a UPGMA dendrogram of Festuca and Lolium species generated from cpDNA restriction site data (Charmet et aI . , 1 997), where all extant Lolium species fonn a monophyletic group that has descended from a common ancestor. LmTG- 1 + E8- l ike LmTG- l Fig. 4. 1 3 . Possible scenario for the distribution of Epichloe endophytes during the evolution of Lolium spp. L. perenne N. lolii. E. typhina, LpTG-2, and E. Jestucae-l i ke A schematic diagram of Lolium speciation is shown, with a scenario proposed to explain the distribution of Epichloe endophytes (blue) within this host genus (green). Plus (+, red) or minus (-, purple) indicates the gain or loss, respectively, of Epichloe endophytes from the host lineage. The branch lengths and branching order shown have been simplified for diagrammatic purposes. t 232 LmTG- l was found to be distributed throughout the annual Lolium species, and i t is l ikely that this endophyte was formed, or at least was present, in the ancestral Lolium grass. L. canariense, which is endemic to the Canary Islands, was found to host both LmTG- l and LcTG-l endophytes, so it appears that after the L. canariense l ineage had diverged from the other ryegrass species, it became infected with LcTG- l . A possible mechanism for infection is transfer through the stigma of a spore from an E. typhina genotype related to E102 1 , and subsequent infection of the seed. On the contrary, L. perenne i s not known to contain LmTG-I or LcTG- 1 endophytes . Rather, the most common endophyte is N. lolii, though L. perenne occasionall y hosts other endophytes including E. typhina, LpTG-2, and E. jestucae-like endophytes . A plausible explanation for the origin of N. lolii i s that the newly speciated L. perenne contained LmTG- l , but i t subsequently became infected with E. jestucae, which evolved into N. lolii. I t i s thought that the Lolium genus originated from a rye grass with c lose affinity to the extant annual L. rigidum (Charmet and Balfourier, 1 994). Therefore, assuming the l ife cycle of L. perenne evolved from an annual to a perennial cycle, the more extensive in planta distribution of N. lolii and subsequently higher production of protective alkaloid compounds gave the N. lolii symbioses a selective advantage over the LmTG-1 symbioses, and LmTG-1 was subsequently lost. In addition to N. lolii, E. typhina i s also found in L. perenne, and this has probably come from a successful spore infection from an E8-hke E. typhina i solate. Subsequent hybridisation between these two groups formed LpTG-2 (Schardl et aI . , 1 994). The order of events along the lineage of L. perenne shown in Figure 4. 1 3 are not certain, but it is accepted that N. lolii and E. typhina were establi shed before LpTG-2 was formed (Schardl et aI . , 1 994). Recently, a rare E. jestucae-like endophyte was found in L. perenne, which is thought to be a recent introduction as its morphology is very similar to E. jestucae, lending credence to the possibi li ty that E. jestucae infection gave rise to N. lolii. The commonality of genotypes shared between LmTG- 1 , LcTG- 1 and N. uncinatum, which appears to have hybrid origins involving E1021 -like E. typhina and E. bromicola strains (pers. comm. Christopher Schardl, University of Kentucky); and the close relationships between their hosts suggest that these endophytes may have arose in related 233 events. A possibility is that the common ancestor of F. pratensis and Lolium contained a hybrid endophyte involving an E1021 -like E. typhina, E. baconii, and E. bromicola parentage. The loss of the appropriate genes in each host l ineage could explain the distribution of genes that are seen in these endophytes today. This scenario would appear unlikely though as it would involve the apparent non random loss of genes from the original hybrid. For example, in the extreme case of LcTG-I , affinities to only one of the three parents is evident. The simplest explanation for the evolution of these endophytes is the independent evolution of each endophyte type within its host lineage. Direct evolutionary relationships between the endophyte of ryegrass , meadow fescue and tall fescue are not obvious from tub2 data and many scenarios could be postulated to explain the distributions of these genotypes. Further sequence data from other phylogenetically informative loci , such as the TEF genes, may help to elucidate these relationships. 4.3.5 THE EVOLUTIONARY ORIGINS OF TWO ENDOPHYTE ISOLATES FROM HORDEUM GRASSES The results of this study strongly support the unique hybrid origins of each of the two previously undescribed Hordewn endophyte i solates , HaB and Hdl . This conclusion is based principal ly on data from phylogenetic analysis of rDNA-ITS and tub2 DNA sequences, but also from microsatellite genotypes. Two copies of tub2 genes are present in HaB, and their sequences suggest that HaB has descended from E. amarillans and E. elymi ancestry. The rDNA-ITS sequence from HaB shows high simi larity to those of E. amarillans isolates, providing further support for this hypothesis. As E. elymi is presently found in Elymus grasses, which are also of the tribe Triticeae, it is most l ikely that E. elymi, or an E. elymi-like endophyte is the maternal parent of HaB. E. amarillans infects grasses of the more distantly related tribe, A veneae, so presumably a spore from E. amarillans was transmitted to an E. elymi infected Hordeum host for hybridisation to ensue. Analysis of microsatellite genotypes reveal that there are no common alleles between HaB and its putative parents, and this presumably reflects changes accumulated over a considerable length of time since the hybridisation event. 234 Two copies of tub2 genes are also found in Hd1 , and these identify c losely to the tub2 genes of extant endophyte strains from E. typhina and E. bromicola. Interestingly, it i s this combination of species that are presumed t o be the progenitors of N. uncinatum (pers. comm. Christopher Schardl, University of Kentucky), though the hybridisation event that gave rise to Hdl is probably independent of the one that produced N. uncinatum, since the E. typhina-like tub2 genes of Hd1 identify more closely to those of E. typhina strains from Phleum and Dactylis hosts, rather than that from N. uncinatum (Fig. 1 2) . It is again assumed that the maternal parent i s the one that i s currently found infecting hosts most closely related to Hordeum, so this i s likely to be E. bromicola. E. bromicola i s fairly closely related to E. elymi, and the tribes of their host grasses, Bromeae and Triticeae respectively, share a c lose evolutionary relationship (Catahln et aI . , 1 997). Hdl was therefore probably derived from a hybridisation event involving an E. typhina spore infecting an E. bromicola-infected Hordeum grass. There are a few common microsatelli te alleles shared between Hdl and extant E. bromicola and E. typhina i solates to provide further support for this relationship. On the basis of their unique genetic identities, and hence, unique evolutionary origins , i t i s proposed that the Neotyphodium taxonomic groups that HaB and Hdl belong, are named Hordeum taxonomic groups one and two (HdTG- l and HdTG-2) respectively. 4.3.6 INTERSPECIFIC HYBRIDISATION IN THE EPICHLOE ENDOPHYTE GROUP Hybridisation events, as speculated for the evolution of LmTG- 1 , HaB, and Hd1 are relatively common within the Neotyphodium endophytes ego LpTG-2 (Schardl et aI . , 1 994), all tall fescue endophytes (Tsai e t aI . , 1 994), and N. uncinatum (pers. comm. Christopher Schardl, University of Kentucky); and might counteract Muller's ratchet, which states that within an asexual species the inevitable accumulation of deleterious mutations would cause a loss of fitness without the correcting influences of sexual recombination (Muller, 1 964). In the case of seedborne asexual grass endophytes, 235 interspecific hybridisation can increase the fitness of the endophyte in several ways. The resulting heteroploid genome carries duplicate copies of genes, therefore to disrupt the function of a particular gene, multiple mutations would be required to disable each of its gene copies. Hybridisation may also introduce or augment protective characteristics of the endophyte such as the protective alkaloids produced by endophytes. New alkaloid types may be produced in the hybrid, and/or the production levels of existing alkaloids may be increased. 4.4 CONCLUSION 236 The Neotyphodium-l ike endophytes of annual ryegrasses, described by Latch et al . ( 1 988), were found to have a unique hybrid O1igin involving Epichloe baconii and E. bromicola ancestry. These endophytes form a new Neotyphodium taxonomic group and it is proposed that this is named LmTG- l . A second previously undescribed Neotyphodium endophyte was found in Lolium canariense, and this was shown to be an asexual derivative of E. typhina. The name proposed for this taxonomic group of Neotyphodium endophytes is LcTG- I . The two Hordeum endophyte isolates, HaB and Hdl , were also both shown to have unique hybrid origins. HaB has E. elymi and E. amarillans ancestry, and Hdl has E. typhina and E. bromicola ancestry. The proposed names for the Neotyphodium taxonomic groups that contain these isolates are HdTG- l and HdTG-2 respectively. The revelation of interspecific hybrid origins for three more Neotyphodium endophytes il lustrates the prevalence of hybridisation in the evolution of the Epichloe endophytes, and further contributes to our understanding of the evolution of thi s group. APPENDICES APPENDIX 1 VECTOR MAPS 6529 Finl 6508 Mstll I 6493 Gsul 6468 HglEIl �\ 6431 Bg/l \ \\ 6425 FSPI� , 6405 Pvul-- , " 6293 Gdill_ laoZ' Polyclonlng slte 6036 Gdill 6001 Ahall, Nllrl:::::::: 5914 AVIlII __ 5825 AVIlI __ 5716 Otalll '---5643 Bllnll _ ""- 5613 Nllel , Cfrl� 6882 GIIlI � SU1 6899 \ BglII 6935 ,\ I 7000 Xmn l 357 1000 Bacteriophage M 1 3mp1 9 Bsml 1746 (7.25 kb) 2000 4000 Polycloning site 10 1 1 12 13 14 15 16 1 7 1 8 � � N � � � � � � � � � � � � � � � � � � � � � � - � � � � = � � � � � � � = � � = � Se, TeG Hindlll ALl M13mpl9 1 1 1 <--I __ �I I 1 <--I __ -' L-_---'I I I I I Sphl Pstl Sail Xbal BamHI 5mB I Kpnl Accl Hincll Xmal Sacl 238 Asn Se, Leu Ala AAT TCA CTG Gee EcoAI Restriction map of the bacteriophage vector M13mp 1 9 (Norrander et al . , 1 983) . The figure shows the sites at which a number of restriction enzymes cleave the double­ stranded replicative form of M l 3mp l 9. The HindIII site of the polycloning site lies immediately downstream from Plac' Figure adapted from Sambrook et al . ( 1 989). Polycloning site 2674 Eco0109 2622 Aatll 2501 Sspl Thr Met lie Thr Asn Ser Ser Set ATG ACC ATG ATT ACG AAT TCG AGC TCG A1.2 I I EcaRI pUC 1 18 259 Mstl pUC1 1 8 (3.2 kb) \ 8g11 1830 Avail 1838 Va! Pro Gly Asp GT A CCC GGG GAT I l........-.-J Kpnl Smal BamHI Xmal 10 Pro Leu Glu Ser CCT CT A GAG TCG I Xbal Sail Accl Hincll 1 1 Thr ACC I I 1 2 13 Cys Arg TGC AGG I PSII 14 His CAT Sphl 1000 1 5 1 6 1 7 16 239 Ala Ser Leu Ala Lau Ala GCA AGe TTG GCA CTG GCC I HindUI Restriction map of the p lasmid vector pUC 1 1 8 (Vieira and Messing, 1 987). The figure shows the sites at which a number of restriction enzymes cleave this vector. The EcoRl site of the polycloning site lies immediately downstream from P'ac. Figure adapted from Sambrook et al. ( 1 989). Xmn 1 1 994 A l.3 pGEM-T pGEM®-T ( 3003 bp) T7 ! 1 start Apa I 1 4 Aat 1 1 20 Sph I 26 BstZ 1 31 Nco I 37 Sac 1 1 46 Spe I 55 Not I 62 BstZ I 62 Pst l 73 SaI l 75 Nde I 82 Sac ! 94 BstX I 1 03 Nsi I 1 1 2 i-----t 1 26 t SP6 240 Map of the PCR product cloning vector, pGEM-T (Promega Corp.) . The figure shows the sites at which a number of restriction enzymes cleave this vector. The binding sites of the Ml3 forward sequencing primer is 2944 to 2960 bp, and the M 1 3 reverse sequencing primer is 1 6 1 to 177 bp. Figure adapted from Promega Corp. APPENDIX 2 MANUFACTURER'S REFERENCES References for the manufacturers used in this study are listed below. Company Alpha Innotech Corporation Amersham BiolO l Inc . Bio-Rad Boehringer Mannheim GmbH Claris Corp. Corbett Research Di fco Laboratories FMC Bioproducts Fuji Photo Film Co. Hoeffer Scientific Instruments Life Technologies New England Biolabs PE Applied Biosystems Perkin-Elmer Corp. Polaroid Corp. Promega Corp. Savant Instruments Inc . Sigma Chemical Co. Stratagene Address San Leandro, CA., USA Buckinghamshire, UK La Jolla, CA. , USA Richmond, CA., USA Mannheim, Germany Santa Clara, CA., USA Mortlake, Australia Detroit, MI., USA Rockland, MA. , USA Tokyo, Japan San Francisco, CA., USA Gaithersberg, MD., USA Beverly, MA. , USA Foster City, CA., USA Foster City, CA., USA Cambridge, MA. , USA Madison, WI. , USA Farmingdale, NY., USA St Louis, MO., USA La Jolla, CA. , USA 241 242 APPENDIX 3 MICROSATELLITE ELECTROPHEROGRAMS Electropherogram fi les of the microsatellite fingerprints, analysed on ABI Prism model 3 10 or 377 genetic analysers with Genescan 2 . 1 software, are supplied on CD-ROM accompanying this thesis . These fi les require GeneScan software to be opened. For file identification, the file name format includes : • the endophyte i solate name, fol lowed by • the primer set used in the amplification: " 1 " or "2" , for Primers Sets I or Il, respectively, and • the model of ABI genetic analyser used: "3 10" or "377" For example, the file "Tf27 _2.3 10" refers to the analysis of DNA from endophyte i solate Tf27 , amplified by Primer Set Il, and analysed on an ABI Prism 3 1 0 genetic analyser. 243 APPENDIX 4 CHI-SQUARE (X2) TESTS Chi-square (X2) tests to examine for goodness-of-fit between observed and expected frequency data were used to test the hypotheses in Section 3 .2 .3 .4. A4.1 MENDEL'S FIRST LAW To test Mendel's first law, the principle of segregation, the potential of the observed data to fit the expected Mendelian ratio of 1 : 1 was examined. A4.1 . 1 Test One Hypotheses: Ho = Microsatellite alleles at the B lO locus segregate at a ratio of 1 : 1 Ha = Microsatellite alleles at the B lO locus do not segregate at a ratio of 1 : 1 Chi-square test: B lO genotypes Observed frequency Expected frequency (0 - E)2/E (0) (E) 1 70 1 6 2 1 1 . 1 90 1 76 26 2 1 1 . 1 90 Sum 42 42 X2 = 2 .380 Conclusion: For X2 = 2.380 at 1 degree of freedom (df), (0. 1 0 < P < 0.20), so Ho is accepted at the 5% significance level . 244 A4.1.2 Test Two Hypotheses: Ho = Microsatellite alleles at the B 1 1 locus segregate at a ratio of 1 : 1 Ha = Microsatellite alleles at the B 1 1 locus do not segregate at a ratio of 1 : 1 Chi-square test: B 1 1 genotypes Observed frequency Expected frequency (0 - E)2/E (0) (E) 147 26 2 1 1 . 1 90 164 16 2 1 1 . 1 90 Sum 42 42 X2 = 2 .380 Conclusion: For X2 = 2 .380 at 1 degree of freedom (df), (0. 10 < P < 0.20), so Ho is accepted at the 5% significance level . A4.2 MENDEL'S SECOND LAW 245 To test Mendel's second law, the principle of independent assortment, the potential of the observed data to fit the expected Mendelian ratio of 1 : 1 : 1 : 1 for haploid organisms was examined. Hypotheses: Ho = Microsatellite alleles at B 10 and B 1 1 loci assort independently of one another Ha = Microsatellite alleles at B 10 and B 1 1 loci do not assort independently of one another Chi-square test: Genotypes Observed frequency Expected frequency (0 - E)2/E (B lO/B l l ) (0) (E) 1 701147 7 1 0. 5 1 . 1 67 170/ 164 9 10 .5 0.2 14 1 76/147 1 9 1 0.5 6.8 8 1 1 76/ 164 7 1 0.5 1 . 1 67 Sum 42 42 X2 = 9.429 Conclusion: For X2 = 9.429 at 1 degree of freedom (df), (0.0 1 < P < 0.00 1 ), so Ho is rejected at the 5% significance level . 246 APPENDIX 5 GENBANK ACCESSION NUMBERS The GenBank accession numbers for all sequences used for phylogenetic analyses in this thesis (Sections 4.2 .5 , 4.2.7 .2, and 4.2.7.3) are listed below. Epichloe species Isolate Gene GenBank accession E. amarillans E52 tub2 L06958 E. amarillans E57 tub2 L06959 E. baconii E248 tub2 L0696 1 E. baconii E103 1 tub2 L78270 E. brachyelytri E1040 tub2 L7827 1 E. bromicola E501 tub2 L78289 E. bromicola E502 tub2 L78290 E. bromicola E798 tub2 AF062430 E. elymi E56 tub2 L06962 E. elymi E 1 84 tub2 L78273 E. festucae E32 tub2 L06957 E· festucae E189 tub2 L06955 E. glyceriae E277 tub2 L78275 E. glyceriae E2772 tub2 L78276 E. sylvatica E354 tub2 L78278 E. sylvatica E503 tub2 L7829 1 E. typhina E8 tub2 X526 16 E. typhina E348 tub2 L78277 E. typhina E425 tub2 L78280 E. typhina E430 tub2 L78284 E. typhina E469 tub2 L78287 E. typhina E505 tub2 L78292 E. typhina E102 1 tub2 AF062429 E. typhina E2466 tub2 L78274 247 APPENDIX 5 continued Epichloe species Isolate Gene GenBank accession N. coenophialum e 1 9 tub2 (tub2-2) L0695 1 N. coenophialum e 1 9 tub2 (tub2-3) L06964 N. coenophialum e 1 9 tub2 (tub2-4) X56847 N. lolii Lp5 tub2 sequence identical to E32 and E 1 89 (Schardl et aI . , 1 994) N. uncinatum e 1 66 tub2 (tub2-3) sequence identical to e 1 9 tub2-3 (Tsai et ai. , 1 994) FaTG-2 e41 tub2 (tub2-2) L06952 FaTG-2 e41 tub2 (tub2-4) L06963 FaTG-3 Tfl 8 tub2 (tub2-4) sequence identical to e41 tub2-4 (Tsai et aI . , 1 994) FaTG-3 Tfl 8 tub2 (tub2-5) L020308 E. amarillans E1 14 rDNA-ITS L07 1 29 E. baconii E248 rDNA-ITS L07 1 38 E. brachyelytri E1045 rDNA-ITS L78296 E. bromicola E502 rDNA-ITS L78295 E. elymi E56 rDNA-ITS L07 1 3 1 E. festucae E28 rDNA-ITS L07 139 E. glyceriae E277 rDNA-ITS L78302 E. sylvatica E503 rDNA-ITS L78304 E. typhina E8 rDNA-ITS L07 132 E. typhina E430 rDNA-ITS L78298 E. typhina E246 1 rDNA-ITS L20306 N. coenophialum e 1 9 rDNA-ITS U68 1 1 5 N. lolii Lp5 rDNA-ITS L07 1 30 N. uncinatum e 166 rDNA-ITS L07 1 35 FaTG-2 e41 rDNA-ITS L07 140 248 APPENDIX 6 DNA SEQUENCE ALIGNMENTS The DNA sequence alignment fi les, used for phylogenetic analysis with PAUP* software (Sections 4.2.5 , 4 .2 .7 .2, and 4.2.7.3), are supplied on the CD-ROM accompanying this thesis . "ITS Lolium align" and "tub2 Lolium align" are alignments of the annual ryegrass endophytes with representative Epichloe endophyte rDNA-ITS and tub2 gene sequences respectively. Similarly, "ITS Hordeum align " and "tub2 Hordeum align " show the gene sequences from the Hordeum isolates, HaB and Hdl , aligned with the same representative Epichloe isolates . 249 APPENDIX 7 PUBLICA TION The following publication was based on work from this thesis : Moon, C. D. , Tapper, B . A . , and Scott, B . ( 1 999). Identification of Epichloe endophytes in planta by a microsatellite-based peR fingerptinting assay with automated analysis . Applied and Environmental Microbiology 65(3): 1 268-1 279. REFERENCES 25 1 REFERENCES Ainsworth, G. C . , Sparrow, F. K, and Sussman, A. S . ( 1 973). "The Fungi - An Advanced Treatise." first ed. , IV A Academic Press, New York. Akkaya, M. S . , Bhagwat, A. A. , and Cregan, P. B. ( 1 992). Length polymorphisms of simple sequence repeat DNA in soybean. Genetics 132: 1 13 1 - 1 1 39 . Arachevaleta, M. , Bacon, C. W., Hoveland, C. S . , and Radcliffe, D. E . ( 1989). Effect of the tall fescue endophyte on plant response to environmental stress. Agronomy Journal 81 : 83-90. Bacon, C. W. , Lyons, P. c., Porter, J . K, and Robbins, J. D. ( 1986). Ergot toxicity from endophyte-infected grasses: A review. Agronomy Journal 78: 106- 1 16 . Bacon, C. W. , Porter, J . K, Robbins, 1. D. , and Luttrel l , E. S . ( 1 977). Epichloe typhina from toxic tall fescue grasses. Applied and Environmental Microbiology 34(5) : 576- 58 1 . Baleiras Couto, M. M., Eij sma, B . , Hofstra, H. , Huis in't Veld, J . H . J . , and Van Der Vossen, J. M. B. M. ( 1996). Evaluation of molecular typing techniques to assign genetic diversity among Saccharomyces cerevisiae strains. Applied and Environmental Microbiology 62(1 ) : 4 1 -46. Ball , O. J .-P. , Prestidge, R. A., and Sprosen, J. M. ( 1 995). Interrelationships between Acremonium lolii, peramine, and lolitrem B in perennial ryegrass. Applied and Environmental Microbiology 61(4): 1 527- 1 533. Bart-Delabesse, E. , Humbelt, H.-F., Delabesse, E., and Bretagne, S . ( 1998). Microsatellite markers for typing Aspergillus fumigatus isolates. Journal of Clinical Microbiology 36(9): 24l3 -24 1 8 . Bennett, S . J . ( 1 997). A phenetic analysis and lateral key of the genus Lolium (Gramineae). Genetic Resources and Crop Evolution 44: 63-72. Boehringer Mannheim. ( 1 995). "The DIG system user's guide for fi lter hybridisation.". Biochemica (R. van Miltenburg, B. Rliger, S . Grlinewald-Janho, M. Leons , and C. Schroder, Eds. ), Mannheim. 252 Bolton, E. T. , and McCarthy, B. J. ( 1962). A general method for the i solation of RNA complementary to DNA Proceedings of the National Academy of Science of the United States of America 48: 1 390. Borrill, M. ( 1 976). Temperate grasses. 1 st ed. In "Evolution of Crop Plants" (N. W. Simmonds, Ed.), pp. 1 37- 142. Longman Scientific & Technical, Edinburgh. Broun, P. , and Tanksley, S. D. ( 1 996). Characterization and genetic mapping of s imple repeat sequences in the tomato genome. Molecular and General Genetics 250: 39-49. Brownstein, M. 1 . , Carpten, J. D., and Smith, J. R. ( 1 996). Modulation of non-templated nucleotide addition by Taq DNA polymerase: Primer modifications that facilitate genotyping. BioTechniques 20: 1 004- 1010. Bruns, T. D . , White, T. J . , and Taylor, J. W. ( 199 1) . Fungal Molecular Systematics. Annual Review of Ecology and Systematics 22: 525-564. Buchanan, F. C . , Adams, L. J . , Littlejohn, R . P . , Maddox , J . F. , and Crawford, A. M. ( 1 994) . Determination of evolutionary relationships among sheep breeds using microsatellites. Genomics 22: 397-403. Bullock, W. 0., Fernandez, J. M., and Short, J . M. ( 1987). XLI -Blue: a high efficiency plasmid transforming recA Escherichia coli strain with beta-galactosidase selection. Biotechniques 5: 376-378. Bultman, T. L., White, J . F. , Jr. , Bowdish, T. 1 . , Welch, A M., and Johnston, J. ( 1 995). MutuaIistic transfer of Epichloe spermatia by Phorbia flies. Mycologia 87(2): 1 82- 1 89 . Bush, L. P. , Cornelius, P. L. , Buckner, R. c. , Varney, D. R, Chapman, R A, Burrus IT, P. B . , Kennedy, C. W. , Jones, T. A, and Saunders, M. J. ( 1982). Association of N­ acetyl loline and N-formyl loline with Epichloe typhina in tall fescue. Crop Science 22: 94 1 -942. Byrd, A D., Schardl, C. L. , Songlin, P. J . , Mogen, K. L. , and Siegel , M. R ( 1990). The B-tubulin gene of Epichloe typhina from perennial ryegrass (Lolium perenne). Current Genetics 18: 347-354. Caetano-Anolles, G., Bassam, B. J., and Gresshoff, P. M. ( 199 1 ) . DNA amplification fingerprinting using very short arbitrary oligonucleotide primers. Biotechnology 9: 553-557. 253 Carter, J . D., Bal l , O. J .-P. , Gwinn, K D. , and Fribourg, H. A. ( 1 997). Immunological detection of the Neotyphodium-like endophyte of annual ryegrass. In "NeotyphodiumlGrass Interactions" (C. W. Bacon, and N. S . Hil l , Eds.) , pp. 247-249. Plenum Press, New York. Catal