Within-Host Diversity and Vertical Transmission of Group B Streptococcus Among Mother-infant Dyads in The Gambia

Introduction Understanding mother-to-infant transmission of Group B Streptococcus (GBS) is vital to the prevention and control of GBS disease. We investigated the transmission and phylogenetic relationships of mothers colonised by GBS and their infants in a peri-urban setting in The Gambia. Methods We collected nasopharyngeal swabs from 35 mother-infant dyads at weekly intervals from birth until six weeks post-partum. GBS was isolated by conventional microbiology techniques. Whole-genome sequencing was performed on GBS isolates from one mother-infant dyad (dyad 17). Results We recovered 85 GBS isolates from the 245 nasopharyngeal swabs. GBS was isolated from 16.33% and 18.37% of sampled mothers and infants, respectively. In 87% of cultured swabs, the culture status of an infant agreed with that of the mother (Kappa p-value <0.001). In dyad 17, phylogenetic analysis revealed within-host strain diversity in the mother and clone to her infant. Conclusion GBS colonisation in mothers presents a significant risk of colonisation in their infants. We confirm vertical transmission from mother to child in dyad 17, accompanied by within-host diversity.


Introduction
Streptococcus agalactiae, also known as Group B Streptococcus (GBS), is notable for causing meningitis, sepsis and death in young infants across the globe [1]. A recent global estimate reported that, in 2015, GBS caused 90,000 deaths in infants aged younger than three months of age [1]; Africa accounted for 54% of estimated cases and 65% of all foetal/infant deaths. Maternal colonisation of GBS is an important source of infection in new-borns-up to 70% of infants born to colonised mothers are likely to be colonised themselves [2]. However, the microbial population genetics of mother-infant transmission of GBS remains poorly understood.
Previous attempts to investigate the transmission of GBS have relied on mother-child colonisation status and typing methods with limited resolution, such as serotyping and multilocus sequence typing (MLST) [3][4][5]. However, whole-genome sequencing studies on other pathogens have shown that a cloud of genomic diversity-encompassing multiple singlenucleotide variants and arising from micro-evolutionary events-can exist within populations that appear homogenous by conventional typing approaches [6][7][8][9]. In particular, application of next-generation sequencing technology to multiple colonies from the same individual, whether from single or multiple samples, can shed light on the influence of within-host diversity on chains of transmission [6][7][8][9][10][11]. Intervention and prevention strategies targeted at GBS disease, such as maternal immunisation, would be enhanced by an improved understanding of the within-host diversity associated with this important pathogen.
Here, we combine conventional microbiologic investigations with whole-genome sequencing to elucidate mother-infant transmission of GBS among a cohort of mother-infant dyads in a peri-urban setting in The Gambia.

Study design and study participants
We carried out a prospective, longitudinal study of mother-infant dyads. Participants were recruited at the Fajikunda Health Centre, which is located in a peri-urban setting in the coastal region of The Gambia, about 16km from the capital, Banjul. The Fajikunda health centre and the Fajikunda community have been described elsewhere [12]. Mothers aged 18 years and older attending the study clinic were consented before delivery and assigned study numbers. The inclusion criteria included 1) willingness to deliver at the health centre; 2) absence of pregnancy-related complications; and 3) availability post-delivery for the duration of sampling (two months). Trained fieldworkers obtained written informed consent and administered questionnaires to study participants. History of antibiotic use two weeks prior to enrolment, age and area of settlement were recorded. Recruitment was done between August 2011 and November 2012. The mother-child dyads were swabbed consecutively at seven weekly time points (designated Week1-Week7) from birth till two months of age.

Ethical approval
The study was approved by the Joint Gambia Government and the Medical Research Council Unit The Gambia at London School of Hygiene and Tropical Medicine (MRCG at LSHTM) Ethics Committee (study number SCC 1173).

Sample collection
Nasopharyngeal swabs were collected from participants at birth and once weekly for the subsequent six weeks. Trained nurses collected the samples as described previously [12]. The swabs were immediately immersed into Skim-milk Tryptone Glucose Glycerol broth, kept in cooler boxes at 2-8 o C and transported to the MRCG at LSHTM's laboratories at Fajara within six hours. Upon receipt in the laboratory, the samples were homogenised by vortex for a minimum of 20 seconds and stored at -80 o C until further processing.

Culture and isolation of GBS
Aliquots of the nasopharyngeal swab suspensions were enriched in Todd-Hewitt Broth with 5 % Yeast Extract (Oxoid, Basingstoke, UK) and rabbit serum (B&K Universal Ltd, Grimston, East Yorkshire, UK) prior to culture, as described previously [13]. Next, the nasopharyngeal swab suspensions were plated on 5% Sheep blood agar supplemented with crystal violet.
GBS isolates were identified using conventional methods. Briefly, candidate bacterial colonies demonstrating beta-haemolytic activity on blood agar, bacitracin negative, catalase negative, Christie, Atkins, Munch and Peterson (CAMP) test positive were confirmed as GBS using the Streptex® Streptococcal grouping kit (Remel & Oxoid, Thermo Fisher Scientific, Basingstoke, UK). Single colonies were purified and stored from each nasopharyngeal swab culture. All strains were stored in 20% glycerol broth at -80 o C until further processing.

Genomic DNA extraction and sequencing
GBS isolates from one informative mother-child dyad (dyad 17) was chosen for wholegenome sequencing. To provide a context for cluster analysis of the isolates, we included an isolate each from Mother 04 (Mother4_Week4), Child 19 (Child19_Week4) and Child 07 (Child07_Week4). Selected GBS isolates were streaked on 5% sheep blood agar plates and incubated overnight in ambient air. Following overnight incubation, a single colony from each plate was picked, inoculated into 1ml Luria-Bertani broth and incubated for 4 hours at

Data Management and Statistical analysis
Data were managed on an SQL integrated database with a Microsoft Access front-end linked to the demographic, clinical and laboratory databases. We used Stata version 12 (Stata Corp, College Station, TX) and R software (https://www.r-project.org) packages for the statistical analyses. Descriptive statistics were computed including frequency counts (n) and percentages (%), factor variables and mean (SD) or median (IQR) (normally or non-normally distributed continuous variables, respectively). To assess the agreement of GBS carriage between the mother and child pairs, McNemmar's test and Weighted kappa coefficients for correlation of two raters (irr package) were applied. P values of less than 5% were considered significant to reject the null hypothesis.

GBS colonisation
We recruited 35 mother-infant dyads (17 male infants; 18 females) from villages around the Fajikunda Health Centre in The Gambia (Table 1) Table   2). In 87.35% (214/245) of cultured swabs, the culture status of an infant (i.e. positive or negative for GBS) agreed with that of the mother at a given time point (Kappa p-value <0.001) ( Table 3). GBS positivity among the mother-infant dyads across the six weeks of sampling is shown in Figure 1.
Four mother-infant dyads had GBS isolated from both mother and child at birth (dyads 15,

16, 17 and 19). Of the remaining four instances where an isolate was recovered at birth, two were in mothers (mothers 4 and 5) only and the other two in infants only (infants 18 and 21).
In seven out of nine instances where GBS was isolated post-birth, the infant was found to have GBS when the mother was negative or before the isolation of GBS from the mother (dyads 7, 9, 10, 11, 13, 32 and 49). In the remaining instances, GBS was first isolated from both mother and infant at week 2 (dyad 6) or from the mother two weeks prior to isolation from the infant (dyad 20).
Only dyad 17 had an isolate recovered at each weekly sampling visit. Eleven genomes in all were recovered from dyad 17; however, one isolate was lost to contamination, leaving ten genomes for genomic analysis. All ten isolates from dyad 17 belonged to ST196 and were assigned by in-silico capsular typing to serotype II. GBS isolates Child07_Week3 and Mother04_Week4 were both assigned to ST10 and serotype II; Child19_Week4 was assigned to serotype V and ST26.

Phylogenomic diversity among isolates from dyad 17
A phylogenetic tree constructed from core-genome SNPs shows that strains from mother and infant 17 cluster together and are separated from the SGBS021 genomic reference and our other three GBS genomes (Figure 2). Genomes from dyad 17 were at most 7 SNPs apart, whereas the next closely related genome was 1,2447 SNPs apart (Table 4). GBS genomes recovered from mother 17 at weeks one and week four and all the genomes recovered from infant 17 were identical (Table 4). However, we observed a cloud of minor SNP variants within the isolates from mother 17, encompassing four genotypes. Genotype-1 (from Week1 and Week4) differed from the genotype-2 recovered at week two by one SNP. Genotype-2 differed from genotype-3 recovered at week three by seven SNPs; genotype-3 was one SNP apart from genotype-4 recovered at Week5.

Discussion
Our results highlight the utility of nasopharyngeal sampling in the investigation of GBS colonisation and transmission. We found a nasopharyngeal colonisation prevalence of 16.33% in mothers and 18.37% in their infants. These results are consistent with a recent meta-analysis of global maternal GBS colonisation, which reported a prevalence of 15% regardless of sampling site [19], and with the neonatal prevalence rate we previously reported in a two-month-old infant cohort [13].
We have documented vertical transmission followed by persistent colonisation of an ST196 GBS strain in a mother-infant dyad. ST196 is a successful coloniser [3] as well as an essential cause of invasive disease in adults [20]. This clone has also been linked to transmission between humans and cattle [21]. Several studies have associated ST196 with serotype IV [22]. We are aware of only one other study which reported one strain out of 102 prospectively collected strains to be ST196 with a serotype II capsular backbone [23].
Diversity among the maternal ST196 genomes tended to increase across time, with microevolution producing a cloud of closely related genetic variants. Four distinct GBS genotypes were recovered from mother 17; however, all the GBS genomes from infant 17 were identical to each other and to the maternal GBS genomes recovered from weeks one and four, suggesting that a single genotype was transmitted to the infant at birth and persisted during the sampling window.
Of nine instances where GBs was detected post-birth, seven were found in the infants alone or prior to detection in their mothers. This finding may be due to the insensitivity of the culture method, or it may affirm the role of non-maternal sources in neonatal GBS infection.
The importance of horizontal transmission-from healthcare providers, family and friendsin the investigation of GBS infection in new-borns has been documented [24][25][26]. Enhanced surveillance of the burden of horizontal transmission is warranted to inform early intervention and prevention strategies.

Conclusion
Although we sequenced a limited number of GBS genomes, our findings provide valuable insight into the dynamics of transmission in our setting. Future longitudinal studies incorporating large-scale whole-genome sequencing of multiple isolates from mother-infant dyads will further enhance our understanding of the full scope of the genomic diversity of colonising GBS.   Figure 1: Temporal colonisation of GBS among the Study participants 12 V1, birth; V2-V7, weeks 1 to 6. The lab IDs of mother-infant pairs are shown on the 13 secondary y-axis (4, 5, 17 etc). Red, mother; light sea green, baby. 14 Of six infants who were colonised by GBS at birth, four were born to colonised mothers 15 (dyads 15, 16, 17 and 19). Of the GBS colonisations detected post-birth, seven were found in 16 infants alone or prior to the isolation of GBS from their mothers (dyads 7, 9, 10, 11, 13, 32 17 and 49). In the remaining instances, GBS was first isolated from both mother and infant at the 18 third timepoint (dyad 6, week 2) or from the mother prior to isolation from the infant (dyad 19 20). 20 21 22 Figure 2A&B: Phylogenetic relationship between GBS isolates from mother-infant dyad 23 17 and three other genomes (2 mothers and two infants) 24 The genomes are labelled with the dyad number from which they were derived. Mother17 & 25 Child17 represent genomes from dyad17. Figure 2A shows the genomes from dyad 17 cluster 26 together and are several 1000s of SNPS apart from the genomic reference and the other 27 genomes, indicating that the genomes from mother 17 and infant 17 from the same genetic 28 pool (See table 6). The genomes derived from infant 17 were identical to each other and to 29 the maternally-derived genomes at weeks one (Mother17_Week1) and four 30 (Mother17_Week4). Also, the non-dyad 17 genomes have at least 46 SNPs between them 31 (see Table 4). Branches longer than 0.2 were shortened (denoted by broken lines). 32 Figure 2B shows the phylogenetic relationship between genomes within dyad 17. Not shown 33 here are the number of SNPS between the respective genomes within dyad 17. Among the 34 genomes from infant 17, Child17_Week5 and Child17_Week6 are the closest, separated by 35 3SNPS. Of the maternal genomes, Mother17_Week1 and Mother17_Week2 are the closest, 36 with 5 SNPs separating them. The highest pairwise distance between the genomes within 37 dyad 17 is 25SNPs (between Mother17_Week3 and Mother17_Week6). 38