Molecular Epidemiological Status of Group B Streptococcus in Ile Ife South Western Nigeria

Studies in some sub-Saharan African countries like Zimbabwe, Malawi, Kenya and Gambia revealed that Group B Streptococcus (GBS) is emerging as the main cause of neonatal sepsis and meningitis. However, in Nigeria, information on GBS disease prevalence remains sparse. We sourced to isolate GBS from the rectovaginal and neonatal samples that were obtained from a tertiary hospital in a populated area of Osun state and give an updated information on the antibiotic susceptibility patterns, using demographic and clinical parameters. 170 samples were collected from consenting mothers and neonate from June, 2016 to January 2017. Ninety-Eight (98) GBS isolates were recovered from vaginal, rectal of the pregnant woman at the point of labour and Umbilical cord of the neonate within 24 h of birth. Cultures for the isolation and identification of Group B Streptococcus (GBS) were carried out using the CDC recommended microbiological methods. The Kirby-Bauer disk-diffusion method was mployed to determine antibiograms of GBS isolates in accordance with Clinical and Laboratory Standards institute (CLSI). The presence of resistant genes was examined using PCR. The prevalence rate of GBS maternal and neonatal colonization were 29.4% and 20.6% respectively while 4% of the colonized neonates had nosocomial GBS colonization. There was no significant association between GBS colonization status and age (p>0.05), parity (p>0.05), obstetric risk factors (p>0.05) and sex of neonate. One out of the 8 representatives of the multidrug resistant isolates harboured tetM gene while other resistant genes examined were negative in all MDR isolates. High prevalence of maternal and neonatal GBS colonization has been established among pregnant women and neonates in the study area. Nosocomial infection was implicated in GBS colonization among neonates. However further research is called for using larger sample size and multiple curve studies for adequate extrapolation into the general population.


Introduction
Group B Streptococcus (GBS) is the leading cause of neonatal infections in humans [1]. It is one of the main causes of infection in pregnant women with chorioamnionitis, endometritis, surgical wound infection as well as genital infections [2]. Approximately 10-30% of women of childbearing age carry GBS in the rectovaginal compartment [2]. A combination of phenotypic classification and molecular typing has been successfully used in epidemiological investigations of GBS to study clonal lineages associated with colonization or invasive disease [3,4]. A diverse range of molecular techniques have been applied to determine the population structure of GBS, including pulse field gel electrophoresis (PFGE), restriction digest pattern (RDP) and multilocus sequence typing (MLST) [5][6][7].
Multilocus sequence typing has several advantages over restriction digest based molecular typing techniques in that it uses standardized protocols, and provides specific data on single nucleotide changes rather than crude estimates of approximate fragment length. Furthermore, the MLST data from different laboratories can be stored and compared on an open access online database, and hence it has become the preferred method to compare the genetic relatedness of GBS isolates [8].

ARCHIVES OF MEDICINE ISSN 1989-5216
Multilocus sequence typing has been successfully used to characterize the clonal diversity of a number of human bacterial pathogens including Neisseria meningitides and S. pneumoniae [9,10]. Multilocus sequence typing is based on the DNA sequencing of seven housekeeping GBS genes, which include alcohol dehydrogenase (adhP), phenylalanyl tRNA synthetase (PheS), amino acid transporter (atr), glutamine synthetase (glnA), serine dehydratase (sdhA), glucose kinase (glcK) and transkelotase (tkt) [7]. Based on the Multi-Locus Sequence Typing (MLST) data, a number of different sequence types (STs) have been described in maternal and neonatal colonizing GBS isolates.
Despite variability in the proportionality of clonal lineages, molecular characterization of GBS colonizing isolates has shown a limited number of major sequence types as associated with colonization regardless of geographical and epidemiological distinction [11,12]. There are four major MLST defined genetic lineages (ST-1, ST17, ST-19 and ST23), with ST-1 and ST-19 being responsible for over 80% of sequence types in colonizing isolates [7]. The sequence polymorphisms at each locus are classified and each allele is given a number. The combination of alleles found in each isolate defines the sequence type, and sequence types having allelic profiles that share at least five or six identical alleles are grouped within a particular clonal complex (CC) [13].
About 60% of all sepsis in preterm neonates is caused by GBS and sepsis is the second most common cause of death in preterm babies (behind lethal malformations) [14]. Prior to 2001, 15 to 50% of neonates systemically infected with GBS die annually.
Notably, epidemiological data from the USA consistently mentions a higher incidence of invasive GBS disease in black Americans compared to their white counterparts, although reductions in invasive GBS disease incidence have been observed in both groups [15]. In a report by some workers, only a modest decline in overall incidence was observed in the USA state of Minnesota between 2000 and 2010. However, there have been an increase in incidence since 2010 [16].
A study from Africa that characterized isolates collected from Senegal and Central African Republic showed genotypic overlap between the two regions. In Senegal, CC-1 (32.0%) was significantly higher compared to the Central African Republic (10.2%). In contrast, CC-19 (43.2%) was significantly more common in the Central African Republic compared to Senegal (12.0%; P<0.0001). Despite some differences in the population of colonizing isolates circulating in Senegal and Central African Republic, there was some similarity between the two countries, such as the detection of a rarely described CC-26, in 20.0% and 11.4% of isolates from Senegal and Central African Republic, respectively. It has generally been reported that the prevalence of CC-17, which is mostly associated with invasive disease, remains modest in GBS isolates from colonized pregnant women. In contrast, a study by Davies et al. Found no significance difference in the relative frequency of CC-17 when comparing invasive and colonizing isolates (P=0.80) [17]. The higher proportion of CC-17 in colonizing isolates from a Canadian study could be attributed to only serotype III isolates having been characterized. The association between serotype-III and CC-17 has been widely documented by others [18,19].
One million children die each year in low-income countries in the first 4 weeks of life because of neonatal sepsis, accounting for about 23% of the total number of child birth. In 2015, Nigeria recorded a neonatal death rate of 35 per 1000 live birth (CDC, 2016), sepsis accounting for majority of them. In sub-Sahara African, epidemiological data on maternal group B streptococcus carriage is scarce but necessary to design and implement prevention strategies. Unfortunately, this menace killing our neonates, had been debunked mythically by many communities/ tribes and in Yoruba language it is termed 'ABIKU'. Scientifically, neonatal death can be traced to a cause. Illiteracy and poverty level are the major factors responsible for this and data scarcity in this part of world.
The global burden of invasive GBS disease in young infants was summarized in a systematic review, which included 56 studies over the period 2000 and 2011; the majority of which were from European and American countries [20]. This review highlighted striking variability in incidence of invasive GBS disease between and within regions. In a meta-analysis of invasive GBS disease in infants less than 90 days of age, the incidence (per 1,000 live births) was reported as 0.57 (range: 0.00-2.60) in Europe, and 0.67 (range: 0.25-2.13) in the Americas; as low as 0.02 (range: 0.00-0.14) in Asia; and highest in Africa as 1.21 (range: 0.24-1.97).
The low incidences of invasive GBS disease reported across Asia may be an underestimate because a large number (>70%) of deliveries occur outside the health-care settings, possibly resulting in an ascertainment bias with many of the EOGND cases missed at birth [21]. It is likely that the incidence in Africa may also be underestimated. Poor access to microbiology laboratories to confirm invasive GBS disease may also contribute to underestimating the incidence [22]. In a separate review of studies conducted only in low-middle income countries, high incidences were reported in Africa, with South Africa having the highest reported incidence (3.06 cases per 1,000 live births) [23].
A few hospital studies have identified regional GBS colonization rates for pregnant women ranging from 12 to 25.8% [24,25]. In 1993, a small study of 162 pregnant women attending a public antenatal clinic in Toowoomba, Queensland, Canada reported a GBS colonization rate of 16.7%. Based on USA data, McLaughlin and Crowther proposed that 10 to 30% of pregnant women in Australia could be colonized with GBS of the Lower genital tract [26].
Studies in some sub-Saharan African countries like Nairobi, Ghana, Gambia and Cameroon revealed that Group B Streptococcus (GBS) is emerging as the main cause of neonatal sepsis and meningitis. However, in Nigeria, information on GBS disease prevalence remains sparse, hence this study.

Study population/Area
The study was conducted from June 2016 to January 2017, at a tertiary hospital (Obafemi Awolowo University Teaching Hospitals Complex, OAUTHC) which provides health care to the majority of indigenous Africans living in Ile-Ife, Osun State and its environs.

Inclusion and Exclusion Criteria
Inclusion criteria for maternal participants • Pregnant women delivering at OAUTHC or referred to this facility from other private facilities or traditional birth facilities.
• Able to understand and comply with planned study procedures.
• Signed written informed consent.

Inclusion criteria for infants
• Infants ≤ 24 hours of age

Exclusion Criteria
• Refusal for study participation

Methods for Recruitment of Case subjects
The mother was approached for consent and enrolment of her child and herself on arrival for delivery in the labour ward or theatre in case of caesarean section. The Spouses were approached for the consent in cases where the women were inattentive because of labour pain, for the consent. Appointments were booked with some mothers during ante-natal meetings, expected date of delivery (EDD) was taken and phone calls by the nurse/mid wife on duty were put through when such mothers were in labour.

Sample Collection
Rectovaginal swab from 170 consenting pregnant women and umbilical cord swab samples from corresponding neonates(s) were collected using a sterile cotton-tipped applicator. The method of swabbing was otherwise consistent during the study period. Both vaginal and rectal specimens were collected using micropoint diagnostics tipped swabs. The rectal swab was inserted approximately 2 cm pass the anal verge and rotated against the rectal mucosa. The vaginal swab was inserted approximately 2 cm pass the introitus towards the lower vagina mucosal wall and rotated. Additionally, a swab of the baby's umbilical cord was collected also by inserting the swab 2 cm into the cord immediately after it was cut. All three swabs were inserted into the Stuart transport medium without charcoal and transported to the department of Microbiology laboratory for processing. Swabs collected during normal working hours, on weekends or public holidays were processed within 24 h. Where this was not possible, samples were stored at 2-8°C and processed during laboratory hours the next day.

Isolation of the Organisms
Maternal lower vaginal, rectal swabs and neonatal umbilical swabs were inoculated into Lim broth: Todd-Hewitt broth (2 ml) supplemented with 10 mg/l colistin and 15 mg/l nalidixic acid. The broth was incubated overnight at 37°C for 18-24 h for growth. When growth (turbidity) was noticed, samples were then streaked unto freshly prepared sterile enriched plate media (5% blood agar) and incubated for 24-48 h at 37°C in 5% CO 2 in anaeropack jar (2.5 liter, order no. 50-25, product of Mitsubishi Gas Chemical Company Co. Inc., Japan).
Suspected isolates based on morphology on blood agar were again streaked on GBS brilliance agar (oxoid, UK) as it is more sensitive in detecting GBS from rectal swabs and vaginal swabs. Furthermore, GBS brilliance agar (oxoid, UK) has a higher specificity and detects non-hemolytic GBS as well [27].
The already prepared GBS brilliance agar (oxoid, UK) plates were incubated at 37°C for 18-24 h in aerobic conditions and examined for growth of mauve GBS-like colony morphologies. GBS like colonies were identified, they were subjected to further confirmatory tests, such as the catalase test, the Christie Atkinson Munch-Petersen (CAMP) test and Confirmed GBS isolates were stored at 4°C.

Identification of Isolates
Preliminary identification of bacterial isolates was performed using colonial and morphological characteristics. Isolates were further characterized by physiological characteristics through biochemical reactions. Bacterial isolates were identified using the Bergey's Manual of Determinative Bacteriology. Colonial appearance of isolates on culture media and their relative size, colour, texture, elevation, edge and shape were observed. Gram stain, microscopy analysis, catalase and CAMP tests were conducted to confirm Streptococcus agalactiae.

DNA Extraction
A colony of GBS bacteria harbouring the DNA of interest was picked and an overnight broth of the colony was prepared at 37°C for 24 h. The tube containing an overnight broth culture of GBS isolates were vortexed at high speed to re-suspend the cells. One milliliter of the vortexed broth culture of GBS isolate was then transferred into an already labeled eppendorf tube and centrifuged at 13,000 rpm for 10 min. The supernatant in the tube was then discarded and blotted on paper towel. After this, 50 µl of sterile distilled water was added and vortexed to homogenize the pellets. The tube was then boiled at 100°C for 10 min. After boiling, the tube was again vortexed and centrifuged at 13,000 rpm for 10 mins.

Polymerase Chain Reaction (PCR) Amplification and Detection of PCR Products
Isolates were screened for the detection of resistance genes (tetM, tetO and linB) using Polymerase Chain Reaction (PCR). The selection of the isolates was based on the antibiotic profiles. Primers designed as indicated below were synthesized at Inqaba-Biotechnical industries, Pretoria, South Africa.

ARCHIVES OF MEDICINE ISSN 1989-5216
genes for tetracycline resistance, tetM, tetO and one gene for clindamycin resistance linB using a set of specific primers as previously described [28,29]. Primer were tested for specificity by BLAST search on NCBI website.

Statistical Analysis
Significant differences and relationship between various data obtained were compared using SPSS 20 version.

Results
Ninety eight (98)  On demographic parameters (   (Table  3). Twenty two (22) from 41 positive women had fever during pregnancy while the remaining 19 did not. 10 from 27 positive neonates came from mothers who had fever during pregnancy while the remaining 17 didn't.
35 from 170 neonatal umbilical cord specimens were positive representing 20.6% as shown in Table 3.
The test of association (using the Pearson chi-square) between demographic and clinical parameters versus the GBS status of the subjects is shown in

Molecular Detection of tetM, tetO and linB Resistance Genes in GBS Isolates
The agarose gel electrophoresis of tetM is shown in Figure 1.

Figure 1
Total The route of GBS infection and occupation of the parents had a significant association with p value of <0.05, 25.6% of the skilled mothers and 25.5% of skilled fathers (Doctors, lawyers, civil servants etc.) tested positive to GBS with samples collected from the rectal. 25% in the category "others" (students, traders and businessmen/women) also tested positive to GBS with rectal samples. This two groups also had the highest transmission rate to neonate with 20% and 25% respectively. The Overall GBS colonization seem to be more significant with the semi-skilled (artisans) and the "others" group mothers with 33.3% each and their husbands in skilled (civil servants etc.) and "others" groups. In this study, previous UTI and fever during pregnancy had significant association with vaginal GBS colonization (p values <0.05) with 26.6% and 29.3% positive to GBS respectively, although the causal organism in previous UT infection was not recorded. Patients who delivered through caesarean section had more rectal GBS colonization but less vaginal colonization than SVD. 25.5% of GBS colonized neonates were delivered through CS compared with 19% of neonates from SVD.
Mullaney reported GBS as a nosocomial infection which correlates with this study. Nine (9) from the thirty-five (35) GBS colonized neonates had no link to the maternal GBS status, which suggests the neonates GBS status could have been acquired from the hospital environment [33].
Interesting, eight (8) from the nine (9) were delivered through SVD which suggest the labour ward as the link between this GBS infected neonates, which can be the origin of late-onset GBS disease.
Hanley, reported low birth weight in neonates with GBS colonized mothers, however, in this study birth weight and anthropometric parameter at birth had significant association with GBS colonization. As high as 45.3% of GBS colonized mothers and 35.8% of GBS colonized neonates had neonatal birth weight of 2.00-2.99, recording p values <0.05 in both cases. 58.3% of GBS colonized mothers and 50% GBS colonized neonate had their full length 50-59, taking a very significant percentage of the GBS colonized mother and neonate. More than 50% of both GBS colonized mother and neonate had Head circumference between 33 and 36.
This study established that GBS had a prevalence of 29.4% among pregnant women and 20.6% among neonates in Nigeria, which is in conjunction with research works from other Africa countries. Pregnant women have been found to be asymptomatic to GBS colonization, however, it has been implicated as a causal organism for sepsis in neonates which is the leading cause of neonatal death worldwide.
The resistant genes amplified (tetO and LinB) were absent in the isolates, suggesting the absence of these genes in the environs.

Recommendation
Data from this study provides important epidemiological information on GBS colonization among pregnant women and their neonates in Nigeria and one of the first studies to report GBS prevalence in neonates and vertical transmission (maternal) in Ile-Ife and Nigeria at large. The study reveals high carriage rate of GBS among pregnant women compare to some previous studies in the country. More than one-fourth of pregnant women, one-fifth of neonates born within 24 h harbored GBS. Our study highlighted that all age groups among pregnant women in Ile-Ife is potentially at risk of GBS colonization and pregnant women <20 years are at the greatest risk of GBS colonization with up to one in three, neonates are at risk of nosocomial GBS colonization and those born with SVD are at greater risk. Neonates born with a negative GBS status can acquire it in the hospital. Nosocomial being responsible for about 5% of GBS colonization among neonates.
Further research is called for using larger sample size and multiple curve studies for adequate extrapolation into the general population. This is also an eye opener for health policy makers in Nigeria to look into maternal and neonatal GBS colonization for possible review in National Health Policy. Reductions in GBS infant and neonatal disease will occur when a screening based consensus protocol is implemented. Screening provides clinicians with a maternal GBS colonization results which provides a basis for early maternal preventative treatment options. When nonscreening based prevention protocols are implemented, the