Study design and sample collection
This was a cross-sectional study done using 43 GBS isolates from 420 vaginal samples that were collected from pregnant mothers who attended Antenatal Care (ANC) at the Harare and Chitungwiza Central hospitals in Zimbabwe. The sample collection was conducted between October 2016 to November 2016. The patients’ age, sex, HIV status were some of the demographic data collected together with the samples. A consecutive sampling method was used to select the participants, then two vaginal swabs were collected from each consenting participant by trained research nurses.
Inclusion and exclusion criteria
Patients who were on treatment for any STI and those who had received antibiotic treatment one month prior to recruitment were excluded from the study. All pregnant women who did not give their consent as well as those who were less than 13 weeks pregnant and less than 18 years of age were also excluded from the study.
Sample Size Determination
Sample size was calculated using a single proportion estimate with specified precision. A two sided significance level alpha was set at 5%, power of 80% and a prevalence of GBS in Zimbabwe of 31.6% according to Moyo et al, (2000) (29). The sample size was based on an estimated proportion of 0.33, confidence level of 0.95 and a desired precision of estimate of 0.005. The estimated sample size became 340 plus 5% attrition giving a total minimum sample size of 357
Isolation and Identification of Streptococcus agalactiae
A total of 43 suspected GBS isolates were obtained from the vaginal swabs. These isolates were subcultured by streaking onto 5% sheep blood agar plate (Columbia CNA Agar) (Acumedia, USA). The plates were incubated at 37°C for 18–24 hours in aerobic conditions. The suspected β-hemolytic colonies were identified as GBS using standard microbiological S. agalactiae identification tests such as the Gram stain (gram positive cocci), the Catalase test (catalase negative), Bile Aesculin test, CAMP (Christie, Atkins, Munch, Petersen) test which were confirmed using the Streptex grouping latex reagent (Remel, UK). The isolates identified as GBS strains were then stored in Mueller Hinton broth along with 20% glycerol at -80°C. The negative culture results were issued after 72hrs.
Antimicrobial Susceptibility Testing
The following 9 antimicrobial discs (Oxoid, Basingstoke, UK) and concentrations were selected: Penicillin G (10units), Ampicillin (10µg), Vancomycin (30µg), Cefazolin (30µg), Ceftriaxone (30µg), Chloramphenicol (30µg), Tetracyline (30µg), Clindamycin (2µg) and Erythromycin (15µg). Antimicrobial susceptibility testing (AST) of GBS was done on 5% sheep blood containing Mueller-Hinton agar using the modified Kirby Bauer methods as recommended by the Clinical Laboratory Standard Institute (CLSI 2017) (30). Bacterial inocula were prepared by suspending 3-4 freshly grown GBS colonies in 3-5ml sterile physiological saline and turbidity was adjusted to 0.5 McFarland standard used as a reference to adjust the bacterial suspension. A sterile cotton swab was used to inoculate and excess fluid was removed by pressing and rotating the swab against the side of the tube above the level of the suspension. It was then swabbed over the entire surface of the agar. Then, antibiotic impregnated paper disks were placed on the plate and incubated in 5% CO2 at 37ºC for 24hrs. Incubated, plates were examined to ensure the growth was confluent or near confluent before zone diameters were measured. The minimum inhibitory concentrations (MICs) of ampicillin, chloramphenicol and vancomycin were also confirmed using E-test strips (BioMerieux, Basingstoke, UK). The MICs for three antimicrobials were defined as the lowest concentration of antibiotic that completely inhibited bacterial growth. The susceptibilities of GBS to antibiotics were categorized as sensitive, intermediate and resistant based on the CLSI 2017 guidelines (30).
Quality control
Streptococcus pneumoniae (ATCC 49619), Streptococcus agalactiae (ATCC 12386), Escherichia coli (ATCC 25922) and Staphylococcus aureus (ATCC 29213) were used as quality control in this study.
DNA Isolation and Quantification
After subculturing the stored isolates, DNA was extracted using the boiling method. A pipette was used to place 600µl of sterile Tris-EDTA (TE) buffer pH 8.0 (Amresco, USA) into a 1.5ml sterile Eppendorf tube. A sterilized loop was used to emulsify the cultured colonies in TE and this mixture was vortexed for 30-45secs before placing the tightly closed Eppendorf tubes into a boiling water bath for 15 minutes. A -20ºC freezer was then used to freeze the boiled mixture for 10 minutes. DNA samples were stored at -20ºC in Tris-EDTA buffer at pH 8.0 to minimize the degradation, before being tested for quality and quantity. The preparation of ×1 Tris-Borate electrophoresis (TBE) buffer, 1% agarose gel and molecular weight markers (1kb and 50bp) were all done according to manufacturer’s instructions.
The PCR process was done in four stages, i.e. amplification of the housekeeping gene atr, the mobile genetic element IS1548, the antibiotic resistance genes (ermB, ermTR, mefA, linB, tetM and tetO) and the virulence genes (hly, bca, bac, rib and scpB).
PCR Reactions Mixtures and Conditions
Amplification of all the genes was done using standard PCR. For each DNA sample a 25μl reaction mixture was made which consisted of 15.875μl of nuclease-free water (Amresco, USA), 2.5μl of ×10 PCR buffer (New England BioLabs, UK), 0.5μl of dNTPs Solution Mix, 10mM (New England BioLabs, UK), 0.125μl Taq DNA polymerase, 5U/μl (New England BioLabs, UK), 5μl of DNA template and 1μl of the 10µM primer (0.5μl of forward primer and 0.5μl of reverse primer) (Inqaba biotech, South Africa). The primers used in our study are listed (see Additional file 2: Table S1).
Nuclease free water was used as a negative control. A Labnet Multigene OptiMax (Labnet International Inc, USA) was used for the PCR thermal cycling conditions with an initial denaturation step at 94ºC for 1 min, 35 cycles {denaturation 94ºC for 1 min, annealing at 55ºC for 1 min, extension 72ºC for 1 min} and a final elongation step at 72ºC for 10 min. Ten microliters of the amplified products were mixed with 5µl of gel loading dye (6×) then run along a 1% ethidium bromide (Amresco, USA) (EtBr final conc. of 0.5µl/ml) stained agarose gel with a 1kb and 50bp DNA ladder (New England BioLabs, UK) in 1× TBE buffer (Amresco, USA) for 1hr 30mins at 100V (Biobase, China) and then viewed using a Wealtec KETA UV Transilluminator (Wealtec Corp, USA).
The contents of each 25μl reaction mixture is summarized in Table 1. Optimization of the PCRs was performed by adjusting the DNA template concentration and the annealing time.
Table 1: Final primer concentrations and volumes used in the PCRs
Components
|
25µl Reaction
|
Final Conc.
|
10× PCR Standard Taq reaction buffer
|
2.5 µl
|
1×
|
10mM dNTPs Solution Mix
|
0.5µl
|
200µM
|
10µM Forward Primer
|
0.5µl
|
0.2µM
|
10µM Reverse Primer
|
0.5µl
|
0.2µM
|
Template GBS DNA
|
5µl
|
variable
|
Taq DNA Polymerase
|
0.125µl
|
1.25U/50µl PCR
|
Nuclease Free Water
|
15.875µl
|
-
|
Statistical analysis
All data entry and analysis was done using IBM Statistical Package for the Social Sciences (SPSS) version 24.0 software. Prevalence figures were calculated for the total sample population. Tests for normality were done such as the Shapiro-Wilks test and Skewness-Kurtosis All Normality Test. P value less than 0.05 was considered statistically significant.
Sample Characteristics
Tetracyline Resistance Genes
A Shapiro-Wilks test (p < 0.05) and a visual inspection of their histograms, normal Q-Q plots and box plots showed that tetracycline resistance was not normally distributed for tetM and tetO with a skewness of 6.481 (S.E = 0.365) and a kurtosis of 42 (S.E = 6.717) for both tetM and tetO.
Erythromycin Resistance Genes
A Shapiro-Wilks test (p < 0.05) and a visual inspection of their histograms, normal Q-Q plots and box plots showed that erythromycin resistance was not normally distributed for ermTR, mefA and ermB, with a skewness of 2 (S.E = 1.014) and a kurtosis of 4 (S.E = 2.679) for ermTR, a skewness of 0.816 (S.E = 0.369) and a kurtosis of -1.405 (S.E = 0.724) for mefA and a skewness of 0.321 (S.E = 0.409) and a kurtosis of -2.023 (S.E = 0.793) for ermB.
Clindamycin resistance and the linB gene could not be tested for normality because linB’s absence was constant and that showed a complete absence of correlation which is represented by 0. The VFs could not also be tested for normality, because they did not have an established dependent and independent variable, however a correlation did exist. Since the data that was tested is not normal, we used non-parametric test, such as the Spearman rank correlation test (see Additional file 2: Table S5). A p-value of 0.05 was used.