Global Distribution of Invasive Serotype 35D Streptococcus pneumoniae Isolates following Introduction of 13-Valent Pneumococcal Conjugate Vaccine Stephanie W. Lo,a Rebecca A. Gladstone,a Andries J. van Tonder,a Paulina A. Hawkins,b,c Brenda Kwambana-Adams,d Jennifer E. Cornick,e,f Shabir A. Madhi,g,h Susan A. Nzenze,g,h Mignon du Plessis,i,j Rama Kandasamy,k Philip E. Carter,l Özgen Köseoglu Eser,m Pak Leung Ho,n Naima Elmdaghri,o,p Sadia Shakoor,q Stuart C. Clarke,r Martin Antonio,d,s,t Dean B. Everett,e,u Anne von Gottberg,i,j Keith P. Klugman,b Lesley McGee,c Robert F. Breiman,b,v Stephen D. Bentley,a The Global Pneumococcal Sequencing Consortium aInfection Genomics, The Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge, United Kingdom bHubert Department of Global Health, Rollins School of Public Health, Emory University, Atlanta, Georgia, USA cRespiratory Diseases Branch, Centers for Disease Control and Prevention, Atlanta, Georgia, USA dVaccines and Immunity Theme, Medical Research Council Unit The Gambia, Banjul, Fajara, The Gambia eMalawi Liverpool Wellcome Trust Clinical Research Programme, Blantyre, Malawi fInstitute of Infection & Global Health, University of Liverpool, Liverpool, United Kingdom gMedical Research Council: Respiratory and Meningeal Pathogens Research Unit, University of the Witwatersrand, Johannesburg, South Africa hDepartment of Science and Technology/National Research Foundation: Vaccine Preventable Diseases, University of the Witwatersrand, Johannesburg, South Africa iCentre for Respiratory Disease and Meningitis, National Institute for Communicable Diseases of the National Health Laboratory Service, Johannesburg, South Africa jSchool of Pathology, University of the Witwatersrand, Johannesburg, South Africa kOxford Vaccine Group, Department of Paediatrics, University of Oxford, and the NIHR Oxford Biomedical Research Centre, Oxford, United Kingdom lInstitute of Environmental Science and Research Limited, Kenepuru Science Centre, Porirua, New Zealand mHacettepe University Faculty of Medicine, Department of Medical Microbiology, Ankara, Turkey nDepartment of Microbiology and Carol Yu Centre for Infection, The University of Hong Kong, Queen Mary Hospital, Hong Kong, China oDepartment of Microbiology, Faculty of Medicine and Pharmacy, Hassan II University of Casablanca, Casablanca, Morocco pBacteriology-Virology and Hospital Hygiene Laboratory, University Hospital Centre Ibn Rochd, Casablanca, Morocco qDepartment of Pathology and Laboratory Medicine and Department of Paediatrics and Child Health, The Aga Khan University, Karachi, Pakistan rFaculty of Medicine and Institute of Life Sciences, University of Southampton, Southampton, United Kingdom sMicrobiology and Infection Unit, Warwick Medical School, Warwick, United Kingdom tLondon School of Hygiene and Tropical Medicine, London, United Kingdom uUniversity of Edinburgh, The Queens Medical Research Institute, Edinburgh, United Kingdom vEmory Global Health Institute, Emory University, Atlanta, Georgia, USA ABSTRACT A newly recognized pneumococcal serotype, 35D, which differs from the 35B polysaccharide in structure and serology by not binding to factor serum 35a, was recently reported. The genetic basis for this distinctive serology is due to the presence of an inactivating mutation in wciG, which encodes an O-acetyl- transferase responsible for O-acetylation of a galactofuranose. Here, we assessed the genomic data of a worldwide pneumococcal collection to identify serotype 35D iso- lates and understand their geographical distribution, genetic background, and inva- siveness potential. Of 21,980 pneumococcal isolates, 444 were originally typed as se- Received 9 February 2018 Returned for modification 27 February 2018 Accepted 24 April 2018 Accepted manuscript posted online 2 May 2018 Citation Lo SW, Gladstone RA, van Tonder AJ, Hawkins PA, Kwambana-Adams B, Cornick JE, Madhi SA, Nzenze SA, du Plessis M, Kandasamy R, Carter PE, Eser ÖK, Ho PL, Elmdaghri N, Shakoor S, Clarke SC, Antonio M, Everett DB, von Gottberg A, Klugman KP, McGee L, Breiman RF, Bentley SD, The Global Pneumococcal Sequencing Consortium. 2018. Global distribution of invasive serotype 35D Streptococcus pneumoniae isolates following introduction of 13-valent pneumococcal conjugate vaccine. J Clin Microbiol 56:e00228-18. https://doi.org/10.1128/ JCM.00228-18. Editor Daniel J. Diekema, University of Iowa College of Medicine Copyright © 2018 Lo et al. This is an open- access article distributed under the terms of the Creative Commons Attribution 4.0 International license. Address correspondence to Stephanie W. Lo, sl28@sanger.ac.uk, or Stephen D. Bentley, sdb@sanger.ac.uk. EPIDEMIOLOGY crossm July 2018 Volume 56 Issue 7 e00228-18 jcm.asm.org 1Journal of Clinical Microbiology D ow nl oa de d fr om h ttp s: //j ou rn al s. as m .o rg /jo ur na l/j cm o n 04 D ec em be r 20 24 b y 20 01 :4 53 0: 2: 20 3: ff ff :f ff f: ff ff :f fe 2. https://orcid.org/0000-0002-2182-0222 https://orcid.org/0000-0001-9186-0679 https://orcid.org/0000-0002-8811-1308 https://doi.org/10.1128/JCM.00228-18 https://doi.org/10.1128/JCM.00228-18 https://creativecommons.org/licenses/by/4.0/ https://creativecommons.org/licenses/by/4.0/ mailto:sl28@sanger.ac.uk mailto:sdb@sanger.ac.uk http://crossmark.crossref.org/dialog/?doi=10.1128/JCM.00228-18&domain=pdf&date_stamp=2018-5-2 http://jcm.asm.org rotype 35B by PneumoCaT. Analysis of the wciG gene revealed 23 isolates from carriage (n � 4) and disease (n � 19) with partial or complete loss-of-function muta- tions, including mutations resulting in premature stop codons (n � 22) and an in- frame mutation (n � 1). These were selected for further analysis. The putative 35D isolates were geographically widespread, and 65.2% (15/23) of them was recovered after the introduction of pneumococcal conjugate vaccine 13 (PCV13). Compared with serotype 35B isolates, putative serotype 35D isolates have higher invasive dis- ease potentials based on odds ratios (OR) (11.58; 95% confidence interval[CI], 1.42 to 94.19 versus 0.61; 95% CI, 0.40 to 0.92) and a higher prevalence of macrolide resis- tance mediated by mefA (26.1% versus 7.6%; P � 0.009). Using the Quellung reac- tion, 50% (10/20) of viable isolates were identified as serotype 35D, 25% (5/20) as serotype 35B, and 25% (5/20) as a mixture of 35B/35D. The discrepancy between phenotype and genotype requires further investigation. These findings illustrated a global distribution of an invasive serotype, 35D, among young children post-PCV13 introduction and underlined the invasive potential conferred by the loss of O- acetylation in the pneumococcal capsule. KEYWORDS 35D, PCV, novel serotype, whole-genome sequencing Streptococcus pneumoniae (pneumococcus) is an important human pathogen that causes pneumonia, bacteremia, and meningitis. In 2015, �330,000 deaths globally in children of �5 years old were estimated to have been caused by pneumococci (1). The polysaccharide capsule of pneumococcus, which has almost 100 serological vari- ants (serotypes), is a major virulence factor (2, 3). Pneumococcal conjugate vaccines (PCVs) targeting up to 13 serotypes have gradually been introduced into 139 countries since the early 2000s (http://view-hub.org/viz/). Simultaneously, a proportional increase in nonvaccine serotypes, such as serotype 35B, has been reported in various countries (4). Recently, a serotype 35B variant, 35D, was identified in four pneumococcal isolates in Australia (5) and two in the United States (2, 6), all of which had an inactivating mutation in wciG, which encodes an O-acetyltransferase responsible for O-acetylation of a galactofuranose. Nuclear magnetic resonance (NMR) analysis on a single isolate representing this novel pneumococcal serotype verified that the serotype 35D capsule lacked O-acetylation but that it was otherwise identical to serotype 35B (2). Serologically, serotype 35D is distinct from serotype 35B by consistently not binding to factor serum 35a, but it displays variable reactivity to group 35 antiserum (2, 5, 6). WciG functionality has been shown to be the determinant of factor serum 35a recognition (2, 7). Presence and absence of O-acetylation is one of the mechanisms for generating diversity in capsular structure, as shown by other serotype pairs such as 9V/9A (O- acetylation mediated by WciE) (8), 11A/11E (WcjE) (8), 15B/15C (WciZ) (8), 33A/33F (WcjE) (9), and 35C/42 (WciG) (7). It is noteworthy that the O-acetyl group in the capsular repeat unit is important for innate immune recognition (10) and is the target of vaccine-elicited antibodies (11). Loss of O-acetylation in serotype 11E is predicted to assist pneumococci in evading host immune and vaccine response and has been suggested to occur during invasive disease after initial colonization with the serotype 11A strain expressing an O-acetylated form of capsule (12). The role of loss of O-acetylation in pneumococcal survival during invasion among the other serotype pairs has remained unknown due to the rarity of serotypes 9A, 33A, and 42 for comparisons, and by the difficulty in differentiation between serotype 15B and 15C. Although the serological profile and biochemical structure of serotype 35D have been described, there has not been an opportunity to comprehensively study this serotype across geographies and clinical considerations. Here, we assessed the genomic data on serotype 35D isolates from a worldwide pneumococcal collection to understand this serotype’s geographical distribution, genetic background and potential invasiveness. Lo et al. Journal of Clinical Microbiology July 2018 Volume 56 Issue 7 e00228-18 jcm.asm.org 2 D ow nl oa de d fr om h ttp s: //j ou rn al s. as m .o rg /jo ur na l/j cm o n 04 D ec em be r 20 24 b y 20 01 :4 53 0: 2: 20 3: ff ff :f ff f: ff ff :f fe 2. http://view-hub.org/viz/ http://jcm.asm.org MATERIALS AND METHODS We retrospectively determined serotypes of 21,980 assembled pneumococcal genomes from the Global Pneumococcal Sequencing (GPS) project (n � 16,575; May 2017; http://www.pneumogen.net/ gps/) and a compiled data set (n � 5,405) by van Tonder et al. (13). DNA extraction was performed on a pure overnight culture derived from a single colony. Sequencing was performed on the Illumina HiSeq platform to produce paired-end reads of either 75 (in 2010 and 2011), 100 (in 2013 and 2014), or 125 (in 2015 and 2016) bp in length. In silico serotype was determined using the whole-genome sequence (WGS)-based serotyping method PneumoCaT (14). As the current version of PneumoCaT does not distinguish serotype 35D from serotype 35B, all samples that were initially typed as serotype 35B were included in this study. To differentiate these two serotypes, nucleotide sequences of wciG were extracted from the assembled genome sequences and aligned to a reference sequence of 35B wciG (GenBank accession number KX021817) described by Geno et al. (2) using CLUSTALW (15). Nonsense and frameshift mutations that led to premature stop codons and in-frame insertions/deletions in wciG were predicted to result in complete loss of function and reduction of function of the WciG protein, respectively. Isolates with these mutations were in silico typed as serotype 35D, and their phenotypic serotype were determined by the Quellung reaction, tested on an overnight culture derived from a single colony (16). Phylogenetic analysis was performed on all serotype 35B and 35D isolates by constructing a maximum likelihood tree using RAxML v.8.2.X (17) based on single-nucleotide polymorphism sites extracted from a core gene alignment with Roary v.3.6.1 (18). An empirical odds ratio for invasive disease due to serotype 35B and 35D was calculated based on a pneumococcal collection of 3,333 randomly selected carriage (n � 1,260) and disease (n � 2,073) isolates from children aged �2 years old, collected during the pre-PCV (n � 1,691), post-PCV7 (n � 678), and post-PCV13 (n � 964) eras using a previously described method (19). For each country, the random selection was carried out from a collection of disease isolates collected via laboratory-based surveillance and carriage isolates collected via cohort studies using the following criteria: 50% of the isolates represented the pre-PCV period (�1 year before) and 50% the post-PCV period (�2 years after primary and �1 after subsequent PCVs). The randomly selected collection in this study included 67 different serotypes plus nontypeable pneumococci. These isolates were collected in South Africa (carriage n � 721, disease n � 1,047), Malawi (carriage n � 336, disease n � 60), and the Gambia (carriage n � 1,016, disease n � 153). Isolates from other locations in the GPS data set were either not randomly selected or consisted of only disease or only carriage isolates and thus could not be used to calculate odds ratios. Susceptibility to chloramphenicol, co-trimoxazole, erythro- mycin, penicillin, and tetracycline were predicted by the identification of resistant determinants in the assembled genomes using previously described pipelines (20–22). The epidemiological and phylogenetic data can be interactively visualized and analyzed online by using the Microreact tool (https://microreact .org/project/GPS_serotype_35B_35D). RESULTS AND DISCUSSION Of 21,980 assembled pneumococcal genomes from the Global Pneumococcal Se- quencing (GPS) project (n � 16,575; May 2017) and a compiled data set (n � 5,405) by van Tonder et al. (13), 444 isolates from disease (n � 173), carriage (n � 270), and an unknown source (n � 1) were originally typed as serotype 35B by PneumoCaT (5). The wciG alignment revealed that 78.6% (349/444) of isolates were identical to the serotype 35B reference, 8.3% (37/444) had silent mutations, 7.9% (35/444) had missense muta- tions, 3.4% (15/444) had frameshift mutations, 1.6% (7/444) had nonsense mutations, and 0.2% (1/444) had an in-frame insertion. All frameshift mutations led to a premature stop codon that disrupted the coding region of wciG. Given that the latter three types of mutations lead to reduced function or a complete loss of function of WciG, the 23 isolates were designated serotype 35D (Table 1). The Quellung reaction of 20 viable isolates showed that 50% (10/20) were serologically typed as serotype 35D, 25% (5/20) as serotype 35B, and 25% (5/20) as a mixture of serotype 35B and 35D (Table 2). In all discrepant cases, we examined the cps locus sequences in an attempt to identify any gene loss and mixed wciG alleles. The cps locus region shared the same capsular genes with the serotype 35D reference (GenBank accession number KY084476), and the mutations in wciG were supported by at least 42� depth of reads (median, 80�; range, 42� to 143�) with 100% consistency. The discrepancy between phenotype and genotype could be due to (i) our inability to capture the serotype diversity in a clinical sample, since the bacterial cultures subjected to DNA extraction and Quellung testing were derived from a single colony that could be different between experiments, and (ii) the possible interconvertibility between serotype 35B and 35D during bacterial culture in vitro. In all five isolates that were both positive and negative to antisera fs35a under one microscope (Table 2), the mutations in wciG were either a 1-bp insertion or deletion that occurred after a 6- to 7-bp homopolymer, highlighting the possibility of intercon- version between serotype 35B and 35D during DNA replication. Metagenomic analysis Epidemiology of Serotype 35D Streptococcus pneumoniae Journal of Clinical Microbiology July 2018 Volume 56 Issue 7 e00228-18 jcm.asm.org 3 D ow nl oa de d fr om h ttp s: //j ou rn al s. as m .o rg /jo ur na l/j cm o n 04 D ec em be r 20 24 b y 20 01 :4 53 0: 2: 20 3: ff ff :f ff f: ff ff :f fe 2. http://www.pneumogen.net/gps/ http://www.pneumogen.net/gps/ https://www.ncbi.nlm.nih.gov/nuccore/KX021817 https://microreact.org/project/GPS_serotype_35B_35D https://microreact.org/project/GPS_serotype_35B_35D https://www.ncbi.nlm.nih.gov/nuccore/KY084476 http://jcm.asm.org of clinical samples to snapshot the serotype diversity and investigation into the interconvertibility of serotype 35B and 35D will potentially explain the discrepancy between the phenotypes and genotypes observed in this study. Considering the limitation of this study and our recent understanding of the genetic basis that differ- entiates serotype 35B and 35D (2, 6, 7), the nonsilent mutations detected in wciG in this study strongly indicated the presence of serotype 35D pneumococci in the sample. Thus, the 23 in silico serotype 35D isolates were selected for further analysis. The mutation patterns of wciG among the in silico serotype 35D isolates were diverse. The wciG mutation patterns in the 23 serotype 35D isolates were different from those of the 6 serotype 35D isolates reported previously (2, 5, 6). In total, there were 20 mutation patterns observed in 29 serotype 35D isolates from 10 countries across four continents (Table 1). The most common naturally deficient WciG was due to 86_87insG, which occurred within a 6-bp homopolymeric stretch of guanine. It was first observed in an isolate from Malawi in 2006, prior to the introduction of PCV7, and was also found in isolates from Senegal in 2011, South Africa and the United States in 2012, and New Zealand in 2015. Isolates with this mutation were sporadically distributed on the phylogenetic tree (Fig. 1), suggesting that the mutations had arisen independently on multiple occasions. The convergence of mutations may imply that this site is a muta- tional hot spot. The majority of serotype 35D isolates belonged to clonal complex 558 (CC558) (n � 9), CC198 (n � 6), and CC156 (n � 5), which were primarily associated with serotype 35B (6, 24, 25). The CC558 and CC156 lineages accounted for most of the increase in serotype 35B isolates after the introduction of PCV13 in the United States (6), while TABLE 1 Genetic diversity of inactivating mutations in wciG of 29 serotype 35D S. pneumoniae isolates from the Global Pneumococcal Sequencing (GPS) project (n � 23) and previous studies (n � 6) Type of mutation (n) or wciG nucleotide mutation n Clonal complex and or sequence type (n) Isolation: Reference or sourceGeographical location(s) (n) Yr(s) (n) Site(s)e (n) Frameshift mutation (18)a 86_87insG 6 CC156 (2), CC558 (2), CC198 (1), CC9813 (1) Malawi (2), New Zealand (1), Senegal (1), South Africa (1), United States (1) 2006 (1), 2011 (1), 2012 (2), 2015 (2) CSF (3), blood (2), joint pus (1) GPS 914_929del_16bp 2 CC558 South Africa, United States 2012 (1), 2013 (1) CSF (1), blood (1) GPS 162_163insT 2 CC558 United States 2004 (1), 2007 (1) Nasopharynx (2) GPSd 92_93insC 1 CC198 The Gambia 2013 Blood GPS 705_706insT 1 CC156 Malawi 2015 CSF GPS 86delG 1 CC156 Cameroon 2012 CSF GPS 312delA 1 CC198 The Gambia 2009 Nasopharynx GPS 382_385_del_4bp 1 CC9813 South Africa 2012 CSF GPS 306_307insA 1 CC198 Australia 2016 Unknown 5 36delA 1 CC558 Australia 2015 Unknown 5 663_696del_34bp 1 CC452 Australia 2016 Unknown 5 In-frame deletion/insertion (3) 792_968del_177bpb 1 CC156 USA 2015 Blood (2) 6 755_808del_54bpb 1 CC558 Australia 2016 Unknown 5 523_524ins_15bp 1 CC558 USA 2009 Blood GPS Nonsense mutation (7) C220T 2 CC156, ST373 Nepal, South Africa 2013 (1), 2014 (1) CSF (1), nasopharynx (1) GPS T732G 2 CC198 The Gambia 2014 (2) CSF (1), blood (1) GPS C104A 1 CC558 USA 2012 Blood GPS C323A 1 CC558 USA 2012 Blood GPS T434G 1 CC198 The Gambia 2009 Lung aspirate GPS Missense mutations (1) G533A, G679Ac 1 Unknown USA Unknown Unknown (2) aAll frameshift mutations resulted in a premature stop codon. bThe in-frame deletion rendered WciG, an acetyltransferase, nonfunctional. It was evidenced by the serological profiles reported by Chochua et al. (6) and Staples et al. (5). cThe resulting amino acid changes were R178K and A227T. The substitution led to a nonfunctional WciG, confirmed by serological test and NMR spectroscopic analysis. dThese two isolates were reported in a previous study by Croucher et al. (23) and in silico serotype was updated as serotype 35D in this study. eCSF, cerebrospinal fluid. Lo et al. Journal of Clinical Microbiology July 2018 Volume 56 Issue 7 e00228-18 jcm.asm.org 4 D ow nl oa de d fr om h ttp s: //j ou rn al s. as m .o rg /jo ur na l/j cm o n 04 D ec em be r 20 24 b y 20 01 :4 53 0: 2: 20 3: ff ff :f ff f: ff ff :f fe 2. http://jcm.asm.org TA B LE 2 Se ro lo gi ca l p ro fil es of 29 se ro ty p e 35 D S. pn eu m on ia e is ol at es fr om th e G lo b al Pn eu m oc oc ca l Se qu en ci ng (G PS ) p ro je ct (n � 23 ) an d p re vi ou s st ud ie s (n � 6) te st ed b y Q ue llu ng re ac tio n St ra in n am e C ou n tr y C C Y r w ci G m ut at io n (s )f Po ol G Ty p e G ro up 35 A n ti se ru m Ph en ot yp ic se ro ty p e Re fe re n ce or so ur ce 29 42 fs 35 a fs 35 b fs 35 c fs 29 b fs 42 a 34 31 -0 6 U SA N /A N /A G 53 3A ,G 67 9A � N D N D � � � � � � 35 D 2 16 S4 71 A us tr al ia C C 19 8 20 16 30 6_ 30 7i ns A � � � � � � � � � 35 D 5 SA M D U -0 00 05 30 5 A us tr al ia C C 55 8 20 15 36 de lA � � � � � � � � � 35 D 5 16 S4 9 A us tr al ia C C 45 2 20 16 66 3_ 69 6d el _3 4b p � � � � � � � � � 35 D 5 16 S3 5 A us tr al ia C C 55 8 20 16 75 5_ 80 8d el _5 4b p � � � � � � � � � 35 D 5 20 15 28 77 U SA C C 15 6 20 15 79 2_ 96 8d el _1 77 b p � N D N D � � � � � � 35 D 6 C H 20 75 U SA C C 55 8 20 07 16 2_ 16 3i ns T � � � � � � � � � 35 B G PS e 30 25 U SA C C 55 8 20 04 16 2_ 16 3i ns T � � � � � � � � � 35 B G PS e G PS _U S_ 20 10 20 99 45 _R 1 U SA C C 55 8 20 09 52 3_ 52 4i ns _1 5b p � � � � � � � � � 35 B G PS G PS _G M _1 13 0 Th e G am b ia C C 19 8 20 14 T7 31 G (L 24 4* ) � � � � � � � � � 35 B G PS G PS _G M _1 14 8 Th e G am b ia C C 19 8 20 14 T7 31 G (L 24 4* ) � � � � � � � � � 35 B G PS G PS _Z A _2 37 0 So ut h A fr ic a C C 98 13 20 12 38 2_ 38 5d el A TA T � � � � � � � � � 35 D G PS G PS _Z A _2 63 6 So ut h A fr ic a C C 55 8 20 13 91 4_ 92 9d el _1 6b p � � � � b � � � � � 35 D G PS 20 12 21 55 93 U SA C C 55 8 20 12 91 4_ 92 9d el _1 6b p � � � � � � � � � 35 D G PS 20 12 21 56 08 U SA C C 55 8 20 12 C 10 4A (S 35 *) � � � � � � � � � 35 D G PS G PS _Z A _2 55 9 So ut h A fr ic a C C 15 6 20 13 C 22 0T (Q 74 *) � � � � � � � � � 35 D G PS G PS _N P_ 72 42 N ep al Si ng le to nd 20 14 C 22 0T (Q 74 *) � � N D � � � � � � 35 D G PS 20 12 22 06 13 U SA C C 55 8 20 12 C 32 3A (S 10 8* ) � � � � � � � � � 35 D G PS 20 13 20 87 23 U SA C C 55 8 20 12 86 _8 7i ns G � � � � � � � � � 35 D G PS G PS _M W _D 38 25 3_ R1 M al aw i C C 15 6 20 06 86 _8 7i ns G � � � � � � � � � 35 D G PS G PS _M W _B KR 60 9 M al aw i C C 15 6 20 15 86 _8 7i ns G � � � � � � � � � 35 D G PS PI 01 67 Se ne ga l C C 19 8 20 11 86 _8 7i ns G � � � � � b � � � � 35 B/ D G PS G PS _N Z_ 15 SP 07 20 N ew Ze al an d C C 55 8 20 13 86 _8 7i ns G � � N D � � c � � � � 35 B/ D G PS G PS _Z A _2 48 7 So ut h A fr ic a C C 98 13 20 12 86 _8 7i ns G � � � � � b � � � � 35 B/ D G PS G PS _M W _B KR 5W C M al aw i C C 15 6 20 15 70 5_ 70 6i ns T � � � � b � b � � � � 35 B/ D G PS PI 02 58 C am er oo n C C 15 6 20 12 86 de lG � � � � � b � � � � 35 B/ D G PS G PS _G M _0 28 2 Th e G am b ia C C 19 8 20 13 92 _9 3i ns C N D a N D N D N D N D N D N D N D N D N D G PS G PS _G M _0 60 0 Th e G am b ia C C 19 8 20 09 31 2d el A N D N D N D N D N D N D N D N D N D N D G PS G PS _G M _0 32 0 Th e G am b ia C C 19 8 20 09 T4 34 G (L 14 5* ) N D N D N D N D N D N D N D N D N D N D G PS a N D ,d at a no t av ai la b le . b U nd er th e m ic ro sc op e, ce lls th at w er e de riv ed fr om a si ng le -c ol on y ov er ni gh t cu lt ur e sh ow ed b ot h p os iti ve an d ne ga tiv e to th e an tis er a te st ed . c T hi s is ol at e w as te st ed in tw o di ff er en t la b or at or ie s an d ex hi b ite d as b ot h p os iti ve to an tis er um fs 35 a in on e la b or at or y an d ne ga tiv e in an ot he r. d Is ol at e G PS _N P_ 72 42 b el on g to ST 37 3, a si ng le to n th at do es no t b el on g to an y cl on al co m p le x. e T he se tw o is ol at es w er e re p or te d in a p re vi ou s st ud y b y C ro uc he r et al .( 23 ) an d in si lic o se ro ty p e w as up da te d as se ro ty p e 35 D in th is st ud y. f * ,s to p co do n. Epidemiology of Serotype 35D Streptococcus pneumoniae Journal of Clinical Microbiology July 2018 Volume 56 Issue 7 e00228-18 jcm.asm.org 5 D ow nl oa de d fr om h ttp s: //j ou rn al s. as m .o rg /jo ur na l/j cm o n 04 D ec em be r 20 24 b y 20 01 :4 53 0: 2: 20 3: ff ff :f ff f: ff ff :f fe 2. http://jcm.asm.org CC198 is the major serotype 35B lineage in the Gambia (unpublished data). Based on a high-resolution single-nucleotide polymorphic tree (Fig. 1), serotype 35D pneumo- cocci emerged among closely related serotype 35B isolates within different clusters. Together with the unrelated mutations observed in wciG, this strongly indicated that serotype 35B is the progenitor of serotype 35D. Compared with serotype 35B isolates, serotype 35D isolates were more likely to be recovered from sterile anatomical sites, including cerebrospinal fluid (CSF; n � 9), blood (n � 8), lung aspirate (n � 1), and joint aspirate (n � 1), than among carriage isolates (n � 4) (82.6% [19/23] versus 36.7% [154/420]; P � 0.001 by Fisher’s exact test). Based on a larger pneumococcal collection (n � 3,333) randomly selected from the GPS project database, the empirical odds ratio (OR) for invasive disease due to serotype 35D is 11.58 (95% confidence interval, 1.42 to 94.19), whereas the OR for serotype 35B is 0.61 (95% CI, 0.40 to 0.92). The increased invasive capacity in serotype 35D strains could be a result of evasion of the immune response targeting the capsule O-acetyl group. The observation in serotype 35B/35D coincides with a previous study on serotype 11A/11E, FIG 1 Maximum likelihood phylogenetic tree was constructed using 56,848 single-nucleotide polymor- phisms (SNPs) extracted from a 1.02-Mb codon alignment of 1,141 core genes from 444 serotype 35B and 35D S. pneumoniae isolates. The tree is colored according to the geographic location of each sample’s isolation. This analysis used an unrelated nontypeable isolate as the outgroup on which to root the tree. Clonal complex (CC) and mutations in wciG are shown to the right of the tree. Singleton sequence types and minor CCs with �5 isolates in this study are indicated in pink and gray, respectively. Lo et al. Journal of Clinical Microbiology July 2018 Volume 56 Issue 7 e00228-18 jcm.asm.org 6 D ow nl oa de d fr om h ttp s: //j ou rn al s. as m .o rg /jo ur na l/j cm o n 04 D ec em be r 20 24 b y 20 01 :4 53 0: 2: 20 3: ff ff :f ff f: ff ff :f fe 2. http://jcm.asm.org in which serotype 11E strains with a loss or reduced amount of acetylation in the capsule were found to be significantly associated with invasive pneumococcal disease (12, 26). The emergence of serotype 35D is likely explained by Calix et al.’s hypothesis (12) that pneumococcal capsule structure undergoes microevolution during progres- sion from carriage to infection in response to divergent selection pressure in early mucosal colonization compared to later in a sterile site. This model of microevolution needs to be further investigated by characterizing the serotype dynamic over the development of invasive disease in vivo. Compared with the pre-PCV era, the prevalence of serotype 35D has not increased more than serotype 35B after the introduction of PCV13. (OR, 12.36; 95% CI, 1.5 to 100.6 versus OR, 3.54; 95% CI, 2.4 to 5.4; Table 3) in the randomly selected pneumococcal collection. A large proportion of 35D isolates (65.2%, 15/23) were collected after the rollout of PCV13. The post-PCV introduction isolates were all invasive isolates and were recovered in six countries (Cameroon, Malawi, New Zealand, South Africa, the Gambia, and the United States), highlighting that this invasive serotype is present in the residual pneumococcal population worldwide and could potentially be an example of serotype replacement. Among the 23 serotype 35D isolates, 87.0% (20/23) had at least one resistance determinant conferring resistance to commonly used antibiotics, including penicillin (65.2%, 15/23), erythromycin (30.4%, 7/23), co-trimoxazole (21.7%, 5/23), and tetracy- cline (4.3%, 1/23). Similar to the previous studies on serotype 35B (6, 24), the penicillin- resistant isolates in this study were predominantly CC558 (60.0%, 9/15), followed by CC156 (35.7%, 5/15) and a singleton of sequence type 73 (ST373) (6.7%, 1/15). Macro- lide resistance mediated by mefA was significantly higher in serotype 35D isolates than in serotype 35B isolates (Table 4). Five of six serotype 35D isolates harboring mefA were from the United States, where macrolides are recommended for use as an empirical therapy for pneumonia in children (27–29); they all belonged to CC558, a major contributor to penicillin resistance in the United States after introduction of PCV13 (24). Unlike the highly invasive but usually antibiotic-susceptible serotype 1, pneumococci expressing serotype 35B (lower-invasive capsule) are more likely to be commensal in the nasopharynx, which could allow them to acquire antibiotic resistance determinants TABLE 3 The prevalence of serotype 35B and 35D S. pneumoniae from South Africa (n � 1,768), the Gambia (n � 1,169) and Malawi (n � 396) in each vaccine period Vaccine perioda No. of isolates (%) for serotype: Odds ratio (95% confidence interval) for serotype: serotype 35B serotype 35D 35B 35D Pre-PCV (n � 1691) 36 (2.12) 1 (0.06) Baseline Baseline Post-PCV7 (n � 678) 12 (1.77) 0 0.83 (0.4–1.6) Post-PCV13 (n � 964) 69 (7.16) 7 (0.73) 3.54 (2.4–5.4)b 12.36 (1.5–100.6)b aBased on the year of PCV introduction, we grouped each year of collection into three categories, as follows: pre-PCV period (years when no conjugated vaccine was used and the year of PCV7 introduction); post-PCV7 (the second year of PCV7 introduction until the year when a higher-valency PCV was introduced); and post- PCV13 (the second year of PCV13 introduction until the end of the study year). PCV7 was introduced in South Africa and the Gambia in 2009; PCV13 was introduced in South Africa, the Gambia, and Malawi in 2011. bP value � 0.05. TABLE 4 Antimicrobial resistant determinants in serotype 35B and 35D S. pneumoniae isolates from the Global Pneumococcal Sequencing (GPS) project Antibiotic resistance determinant(s) No. of isolates (%) for serotype: P value35B (n � 421) 35D (n � 23) ermB 3 (0.7) 1 (4.3) 0.192 mefA 32 (7.6) 6 (26.1) 0.009 tetM 36 (8.6) 1 (4.3) 0.710 folA I100L and folP insertion 140 (33.3) 5 (21.7) 0.361 Epidemiology of Serotype 35D Streptococcus pneumoniae Journal of Clinical Microbiology July 2018 Volume 56 Issue 7 e00228-18 jcm.asm.org 7 D ow nl oa de d fr om h ttp s: //j ou rn al s. as m .o rg /jo ur na l/j cm o n 04 D ec em be r 20 24 b y 20 01 :4 53 0: 2: 20 3: ff ff :f ff f: ff ff :f fe 2. http://jcm.asm.org via horizontal gene transfer from other nasopharyngeal bacteria; a subsequent switch to serotype 35D (high-invasive capsule) would then transform the antibiotic-resistant strain into a more virulent form. The limitation of this study is that the carriage and disease isolates included for calculating the invasiveness index were sampled in different cities in each country; all isolates included were collected between 2007 and 2015 from children aged �2 years old. Ideally, the carriage and disease isolates should be geography-, time-, and age- matched. In this instance, we calculated ORs for invasiveness separately for each country. The ORs for invasive disease due to serotype 35B and 35D in the Gambia were 0.37 (95% CI, 0.09 to 1.56) and 20.3 (95% CI, 2.10 to 196.42), respectively. The ORs could not be calculated for invasive disease, as all serotype 35D isolates in South Africa and Malawi were from disease. The ORs for disease due to 35B in South Africa and Malawi were 0.68 (95% CI, 0.40 to 1.16) and 0.72 (95% CI, 0.11 to 2.15), respectively. The ORs by country were consistent with the ORs calculated from the combined data sets of all three countries. Another limitation was that the effects of an in-frame insertion of 15 bp and the missense mutations in wciG on the protein function have not been evaluated. Removing these samples from all comparisons of serotype 35B and 35D did not alter the conclusions drawn from the statistical analyses. This study highlighted the global distribution of an invasive serotype, 35D, among young children in the post-PCV13 era and underlined the invasive potential conferred by the loss of O-acetylation in the pneumococcal capsule. ACKNOWLEDGMENTS This work is funded by the Bill and Melinda Gates Foundation (grant OPP1034556), the Wellcome Trust Sanger Institute, and the Centers for Disease Control and Preven- tion (Atlanta, GA). We thank the Centre for Genomic Pathogen Surveillance (CGPS), Sanger Institute for their excellent visualization tool, Microreact. We gratefully acknowledge Bernard W. Beall and Cynthia Whitney for their careful and critical reading of our manuscript and insightful comments. We appreciate Stephen I. Pelton from Boston University School of Medicine and Public Health, who kindly provided us with two putative serotype 35D isolates. We thank Belabbes Houria, Idrissa Diawara, Khalid Zerouali, Mohamed Ben- bachir, and Houria Belabbes from Hassan II University of Casablanca and University Hospital Centre Ibn Rochd, Morocco, and Furqan Kabir and Shahida Qureshi from Aga Khan University, Pakistan, for sample and metadata collection. We also thank Julie Morgan from the Institute of Environmental Science and Research Limited, New Zea- land, and Olga Hattingh from the National Institute for Communicable Disease, South Africa, for performing the Quellung test. We declare that we have no additional conflicts of interest. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention. REFERENCES 1. Wahl B, O’Brien KL, Greenbaum A, Liu L, Chu Y, Black R, Majumder A, Lukšić I, Nair H, McAllister D, Campbell H, Rudan I, Knoll M. 2016. Global burden of Streptococcus pneumoniae in children younger than 5 years in the pneumococcal conjugate vaccines (PCV) era: 2000 –2015. Interna- tional Symposium on Pneumococci and Pneumococcal Diseases, Glasgow, Scotland. 2. Geno KA, Saad JS, Nahm MH. 2017. Discovery of novel pneumococcal serotype 35D, a natural WciG-deficient variant of serotype 35B. J Clin Microbiol 55:1416 –1425. https://doi.org/10.1128/JCM.00054-17. 3. Mostowy RJ, Croucher NJ, De Maio N, Chewapreecha C, Salter SJ, Turner P, Aanensen DM, Bentley SD, Didelot X, Fraser C. 2017. Pneumococcal capsule synthesis locus cps as evolutionary hotspot with potential to generate novel serotypes by recombination. Mol Biol Evol 34:2537–2554. https://doi.org/10.1093/molbev/msx173. 4. Balsells E, Guillot L, Nair H, Kyaw MH. 2017. Serotype distribution of Streptococcus pneumoniae causing invasive disease in children in the post-PCV era: a systematic review and meta-analysis. PLoS One 12: e0177113. https://doi.org/10.1371/journal.pone.0177113. 5. Staples M, Graham RM, Hicks V, Strachan J, Goncalves da Silva A, Peverall J, Wicks V, Jennison AV. 2017. Discovery of Streptococcus pneumoniae serogroup 35 variants in Australian patients. Clin Microbiol Infect 23: 476 – 479. https://doi.org/10.1016/j.cmi.2016.12.029. 6. Chochua S, Metcalf BJ, Li Z, Walker H, Tran T, McGee L, Beall B. 2017. Invasive serotype 35B pneumococci including an expanding serotype switch lineage, United States, 2015–2016. Emerg Infect Dis 23:922–930. https://doi.org/10.3201/eid2306.170071. 7. Geno KA, Bush CA, Wang M, Jin C, Nahm MH, Yang J. 2017. WciG O-acetyltransferase functionality differentiates pneumococcal serotypes 35C and 42. J Clin Microbiol 55:2775–2784. https://doi.org/10.1128/JCM .00822-17. 8. Bentley SD, Aanensen DM, Mavroidi A, Saunders D, Rabbinowitsch E, Collins M, Donohoe K, Harris D, Murphy L, Quail MA, Samuel G, Skovsted IC, Kaltoft Lo et al. Journal of Clinical Microbiology July 2018 Volume 56 Issue 7 e00228-18 jcm.asm.org 8 D ow nl oa de d fr om h ttp s: //j ou rn al s. as m .o rg /jo ur na l/j cm o n 04 D ec em be r 20 24 b y 20 01 :4 53 0: 2: 20 3: ff ff :f ff f: ff ff :f fe 2. https://doi.org/10.1128/JCM.00054-17 https://doi.org/10.1093/molbev/msx173 https://doi.org/10.1371/journal.pone.0177113 https://doi.org/10.1016/j.cmi.2016.12.029 https://doi.org/10.3201/eid2306.170071 https://doi.org/10.1128/JCM.00822-17 https://doi.org/10.1128/JCM.00822-17 http://jcm.asm.org MS, Barrell B, Reeves PR, Parkhill J, Spratt BG. 2006. Genetic analysis of the capsular biosynthetic locus from all 90 pneumococcal serotypes. PLoS Genet 2:e31. https://doi.org/10.1371/journal.pgen.0020031. 9. Spencer BL, Saad JS, Shenoy AT, Orihuela CJ, Nahm MH. 2017. Position of O-acetylation within the capsular repeat unit impacts the biological properties of pneumococcal serotypes 33A and 33F. Infect Immun 85: e00132-17. https://doi.org/10.1128/IAI.00132-17. 10. Brady AM, Calix JJ, Yu J, Geno KA, Cutter GR, Nahm MH. 2014. Low invasiveness of pneumococcal serotype 11A is linked to ficolin-2 recog- nition of O-acetylated capsule epitopes and lectin complement pathway activation. J Infect Dis 210:1155–1165. https://doi.org/10.1093/infdis/ jiu195. 11. Rajam G, Carlone GM, Romero-Steiner S. 2007. Functional antibodies to the O-acetylated pneumococcal serotype 15B capsular polysaccharide have low cross-reactivities with serotype 15C. Clin Vaccine Immunol 14:1223–1227. https://doi.org/10.1128/CVI.00184-07. 12. Calix JJ, Dagan R, Pelton SI, Porat N, Nahm MH. 2012. Differential occurrence of Streptococcus pneumoniae serotype 11E between asymp- tomatic carriage and invasive pneumococcal disease isolates reflects a unique model of pathogen microevolution. Clin Infect Dis 54:794 –799. https://doi.org/10.1093/cid/cir953. 13. van Tonder AJ, Bray JE, Quirk SJ, Haraldsson G, Jolley KA, Maiden MC, Hoffmann S, Bentley SD, Haraldsson A, Erlendsdottir H, Kristinsson KG, Brueggemann AB. 2016. Putatively novel serotypes and the potential for reduced vaccine effectiveness: capsular locus diversity revealed among 5405 pneumococcal genomes. Microb Genom 2:000090. https://doi.org/ 10.1099/mgen.0.000090. 14. Kapatai G, Sheppard CL, Al-Shahib A, Litt DJ, Underwood AP, Harrison TG, Fry NK. 2016. Whole genome sequencing of Streptococcus pneumoniae: development, evaluation and verification of targets for serogroup and serotype prediction using an automated pipeline. PeerJ 4:e2477. https://doi.org/10.7717/peerj.2477. 15. Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA, McWilliam H, Valentin F, Wallace IM, Wilm A, Lopez R, Thompson JD, Gibson TJ, Higgins DG. 2007. Clustal W and Clustal X version 2.0. Bioinformatics 23:2947–2948. https://doi.org/10.1093/bioinformatics/ btm404. 16. Statens Serum Institut. 2015. Streptococcus pneumoniae: textbook in diagnosis, serotyping, virulence factors and enzyme-linked immunosor- bent assay (ELISA) for measuring pneumococcal antibodies. Statens Serum Institut, Copenhagen, Denmark. 17. Liu K, Linder CR, Warnow T. 2011. RAxML and FastTree: comparing two methods for large-scale maximum likelihood phylogeny estimation. PLoS One 6:e27731. https://doi.org/10.1371/journal.pone.0027731. 18. Page AJ, Cummins CA, Hunt M, Wong VK, Reuter S, Holden MT, Fookes M, Falush D, Keane JA, Parkhill J. 2015. Roary: rapid large-scale pro- karyote pan genome analysis. Bioinformatics 31:3691–3693. https://doi .org/10.1093/bioinformatics/btv421. 19. Brueggemann AB, Griffiths DT, Meats E, Peto T, Crook DW, Spratt BG. 2003. Clonal relationships between invasive and carriage streptococcus pneumoniae and serotype- and clone-specific differences in invasive disease potential. J Infect Dis 187:1424 –1432. https://doi.org/10.1086/ 374624. 20. Metcalf BJ, Chochua S, Gertz RE, Li Z, Walker H, Tran T, Hawkins PA, Glennen A, Lynfield R, Li Y, McGee L, Beall B. 2016. Using whole genome sequencing to identify resistance determinants and predict antimicro- bial resistance phenotypes for year 2015 invasive pneumococcal disease isolates recovered in the United States. Clin Microbiol Infect 22: 1002.e1–1002.e8. https://doi.org/10.1016/j.cmi.2016.08.001. 21. Li Y, Metcalf BJ, Chochua S, Li Z, Gertz RE, Jr, Walker H, Hawkins PA, Tran T, Whitney CG, McGee L, Beall BW. 2016. Penicillin-binding protein transpeptidase signatures for tracking and predicting �-lactam resis- tance levels in Streptococcus pneumoniae. mBio 7:e00756-16. https://doi .org/10.1128/mBio.00756-16. 22. Li Y, Metcalf BJ, Chochua S, Li Z, Gertz RE, Jr, Walker H, Hawkins PA, Tran T, McGee L, Beall BW, Active Bacterial Core surveillance t. 2017. Valida- tion of beta-lactam minimum inhibitory concentration predictions for pneumococcal isolates with newly encountered penicillin binding pro- tein (PBP) sequences. BMC Genomics 18:621. https://doi.org/10.1186/ s12864-017-4017-7. 23. Croucher NJ, Finkelstein JA, Pelton SI, Mitchell PK, Lee GM, Parkhill J, Bentley SD, Hanage WP, Lipsitch M. 2013. Population genomics of post-vaccine changes in pneumococcal epidemiology. Nat Genet 45: 656 – 663. https://doi.org/10.1038/ng.2625. 24. Metcalf BJ, Gertz RE, Jr, Gladstone RA, Walker H, Sherwood LK, Jackson D, Li Z, Law C, Hawkins PA, Chochua S, Sheth M, Rayamajhi N, Bentley SD, Kim L, Whitney CG, McGee L, Beall B; Active Bacterial Core surveil- lance team. 2016. Strain features and distributions in pneumococci from children with invasive disease before and after 13-valent conjugate vaccine implementation in the USA. Clin Microbiol Infect 22: 60.e9 – 60.e29. https://doi.org/10.1016/j.cmi.2015.08.027. 25. Beall B, McEllistrem MC, Gertz RE, Jr, Boxrud DJ, Besser JM, Harrison LH, Jorgensen JH, Whitney CG; Active Bacterial Core Surveillance/Emerging Infections Program Network. 2002. Emergence of a novel penicillin- nonsusceptible, invasive serotype 35B clone of Streptococcus pneu- moniae within the United States. J Infect Dis 186:118 –122. https://doi .org/10.1086/341072. 26. Calix JJ, Brady AM, Du VY, Saad JS, Nahm MH. 2014. Spectrum of pneumococcal serotype 11A variants results from incomplete loss of capsule O-acetylation. J Clin Microbiol 52:758 –765. https://doi.org/10 .1128/JCM.02695-13. 27. Brogan TV, Hall M, Williams DJ, Neuman MI, Grijalva CG, Farris RW, Shah SS. 2012. Variability in processes of care and outcomes among children hospitalized with community-acquired pneumonia. Pediatr Infect Dis J 31:1036 –1041. 28. Hersh AL, Shapiro DJ, Pavia AT, Shah SS. 2011. Antibiotic prescribing in ambulatory pediatrics in the United States. Pediatrics 128:1053–1061. https://doi.org/10.1542/peds.2011-1337. 29. Kronman MP, Hersh AL, Feng R, Huang YS, Lee GE, Shah SS. 2011. Ambulatory visit rates and antibiotic prescribing for children with pneu- monia, 1994-2007. Pediatrics 127:411– 418. https://doi.org/10.1542/peds .2010-2008. Epidemiology of Serotype 35D Streptococcus pneumoniae Journal of Clinical Microbiology July 2018 Volume 56 Issue 7 e00228-18 jcm.asm.org 9 D ow nl oa de d fr om h ttp s: //j ou rn al s. as m .o rg /jo ur na l/j cm o n 04 D ec em be r 20 24 b y 20 01 :4 53 0: 2: 20 3: ff ff :f ff f: ff ff :f fe 2. https://doi.org/10.1371/journal.pgen.0020031 https://doi.org/10.1128/IAI.00132-17 https://doi.org/10.1093/infdis/jiu195 https://doi.org/10.1093/infdis/jiu195 https://doi.org/10.1128/CVI.00184-07 https://doi.org/10.1093/cid/cir953 https://doi.org/10.1099/mgen.0.000090 https://doi.org/10.1099/mgen.0.000090 https://doi.org/10.7717/peerj.2477 https://doi.org/10.1093/bioinformatics/btm404 https://doi.org/10.1093/bioinformatics/btm404 https://doi.org/10.1371/journal.pone.0027731 https://doi.org/10.1093/bioinformatics/btv421 https://doi.org/10.1093/bioinformatics/btv421 https://doi.org/10.1086/374624 https://doi.org/10.1086/374624 https://doi.org/10.1016/j.cmi.2016.08.001 https://doi.org/10.1128/mBio.00756-16 https://doi.org/10.1128/mBio.00756-16 https://doi.org/10.1186/s12864-017-4017-7 https://doi.org/10.1186/s12864-017-4017-7 https://doi.org/10.1038/ng.2625 https://doi.org/10.1016/j.cmi.2015.08.027 https://doi.org/10.1086/341072 https://doi.org/10.1086/341072 https://doi.org/10.1128/JCM.02695-13 https://doi.org/10.1128/JCM.02695-13 https://doi.org/10.1542/peds.2011-1337 https://doi.org/10.1542/peds.2010-2008 https://doi.org/10.1542/peds.2010-2008 http://jcm.asm.org MATERIALS AND METHODS RESULTS AND DISCUSSION ACKNOWLEDGMENTS REFERENCES