Isolates collection

Non-repetitive clinical samples totaling 256 including purulent pus (n = 58), skin aspirates and effusions (n = 34), wounds (n = 55), otorrhea (n = 36), eye infection (n = 14), throat (n = 35) and endocervical (n = 24), collected between June 2017 and August 2018 from overall 12,654 outpatients. This included patients from neighbouring states attending three major health facilities which serve as referral clinics in southwest Nigeria. Ethical permission for the study was granted by the Federal Medical Centre Abeokuta Health Research Ethics Committees with protocol approval: FMCA/470/HREC/09/2017; NHREC/08/10-2015 with permission from other health facilities but data on their gender, age, disease conditions and subjects’ location of residence were not fully retrieved. Each sample were cultured for Staphylococci strains and phenotypically characterized on Baird-Parker agar and Mannitol salt agars, Gram stained for cellular morphology, tested for catalase and coagulase production as previously discussed [15].

Phenotypical beta-lactamase detection and antibiogram

Beta-lactamase production was assayed with modified starch-acidometric method [16] and Minimum inhibitory concentrations (MICs) for each antibiotic class against the strain was determined using micro-broth dilution assay [17] with 12 panel antibiotics consisting of tetracycline (TE), ceftazidime (CAZ), ciprofloxacin (CIP), gentamycin (GEN), amoxycillin-clavulanic acid (AMC), cefuroxime (CRO), ofloxacin (OFX), sulfamethoxazole (SXT), erythromycin (E), fosfomycin (FOX), vancomycin (VA) and Linezolid (LZD). Phenotypic resistance was interpreted according to CLSI guidelines [18]. Phenotypic screening for methicillin resistance was further determined by assessment of Staphylococcal growth on Mannitol salt agar and Mannitol salt agar supplemented with 4μg/ml Oxacillin as previously described [19]. Multi-antibiotic resistance index (MARI) was determined for each isolates.

Biofilm detection

Phenotypic assessment of biofilm production was done in micro-broth bioassay [20]. Briefly, overnight pure colonies were suspended in 200μl Brain Heart infusion (BHI) broth supplemented with 0.25% glucose and incubated at 37C for 24 hours. After incubation, the growth medium was aspirated away and the wells were washed thrice with Phosphate Buffer Saline (PBS). The wells were stained with 0.5% w/v Crystal Violet for 5 minutes and gently rinsed. Ethanol (70%) was added to dissolve the Crystal Violet and absorbance of the stained biofilm solution was measured with UV Spectrophotometer against absorbance of non-biofilm producing Staphylococcus aureus.

mecA and pvl genotyping

For detection of pathogenic MRSA, isolates were genotyped for mecA and pvl as described by Acevedo et al [21]. Briefly, DNA template was extracted using simple boiling method [22] and amplification was performed in separate reaction volume of 20 μl containing 0.8 μl of 10uM each primer of mec5 (AAAATCGATGGTAAAGGTTGGC) and mec6 (AGTTCTGCAGTACCGGATTTGC) and pvl gene with primers pvl-F (AATGAAATGTTTTTAGGCTCAAGACA) and pvl-R (TGGATAACACTGGCATTTTGTGA) with DNA template (1μl) and water (18 μl) for each reaction following previous described multiplex protocol [21]. Amplification reaction was carried out at initial denaturation of 94°C for 5 min, and followed by 30 cycles of denaturation 94°C for 45 s, annealing 60°C for 60 s, elongation 72°C for 1 min and final elongation 72°C for 5 min. Amplicon of 10 μl of each reaction was electrophoresed on 2% agarose TBE gel at 100V along with the marker. Bands of each PCR products were analyzed regarding their presence and size by using the positive control and the marker as references. MARI (multi-antibiotic resistance index) was determined by dividing the number of resisted antibiotics with total number of antibiotics used to which the organism was subjected. Relatedness of the MARI pattern, degree of biofilm production, beta-lactamase production and mecA detection among the identified strains from various sources were evaluated with dendrogram analysis constructed with DendroUPGMA algorithm.

Genotyping and clonal diversity of spa types

Extracted genomic DNA obtained from overnight culture, was typed for S. aureus protein A (spa gene). PCR assay was performed in constituted reaction mixture of 2x MyTaq HS Mix (10μL), containing spa primers; spa1095F (5’-AGACGATCCTCCGGTGAGC-3’), spa1517R (5’-GCTTTTGCAATGTCATTTACTG-3’) of 5μL each and 1μL template DNA through 30 cycles of denaturation at 94°C for 30 seconds, annealing at 60°C for 30seconds and elongation at 72°C for 30 seconds, with final extension at 72°C for 5minutes [23,24]. DNA of S. aureus DSM 1104L strain served as a positive and distilled water as negative control. Quality of amplicon products was examined on electrophoresed agarose gel and positive strains were purified with GFX PCR DNA and Gel Band Purification Kit (GE Healthcare). Purified PCR products were sequenced with forward primer spa1095F using BigDye 3.1 terminator sequencing and analyzed on ABI Genetic Analyzer 3500Dx (Applied Biosystems, CA, USA). Categorisation of spa types was carried out with Based Upon Repeat Pattern (BURP) algorithm of the Ridom Staph Type software version 1.4 (RidomGmbH, Sedanstr, Germany) to cluster all spa types in the database according to spa clonal complexes [25]. Clonal diversity of Nigerian spa types with other meta-spa sequences were analysed with MEGA software (version 6.0).

Virulence and resistance genotyping

Encoded S. aureus strains with mecA and pvl were further genotyped with StaphyType DNA microarray (Alere Technologies GmbH, Jena, Germany) for other virulence genes. Approximately 170 distinct genes and their allelic variants were targeted for PCR amplification and hybridization on Microtiter strip-mounted DNA microarrays following manufacturer’s instruction and the image of the array was recorded and analysed using a designated reader and software (Arraymate, Iconoclust, Alere Technologies) [26].

Geospatial analysis

Geographical coordinates of individual subjects with staphylococcal infection were identified and recorded with differential global positioning system (GIS) and interpolated for analysis in ArcGIS 10.5.1 programme with respect to land division according to boundary marks in southwest Nigeria [27].

Data analysis

To identify variables and risk factors that could influence staphylococcal infection rate among dependent variables (age, gender and clinical samples), univariate logistic regression analyses was performed to calculate the odds ratio and corresponding 95% confidence intervals. Resistant rates was analysed with radar plots while Median and 75^th percentile resistance were evaluated with Boxplot analysis. Significance of resistance level of staphylococci strains was determined with chi-square and staphylococcal infectivity was calculated with multiple comparison using Kruskal-wallis test.

Risk factor for staphylococcal infection and phenotypic resistance pattern

Staphylococcal infection was significant among the ages (p<0.05, OR[CI] = 0.021[0.545–1.914]) while higher prevalence rate of MSSA (53.0%) and MRSA (3.0%) infection were recorded among female and male subjects respectively. MSSA (37.9%) and MRSA (1.5%) infection rates were significant in wound infection as other clinical conditions presented by the subjects (p<0.05) while MSSA (42.2%) were observed in other conditions (eye, throat and endocervical infection) as shown in Table 1 and S1 Table.

Table 1

Univariate analysis of staphylococcal infections.

Characteristic	MSSA n(%)	MRSA n(%)	OR(CI)	P value
Age (yrs) (Median age: 36.5)	63(24.6)	3(1.2)	0.021(0.545–1.914)	0.004
Gender
Female	35(53.0)	1(1.5)	1.021(0.374–1.785)	0.013
Male	31(47.0)	2(3.0)
Clinical samples
Otitis media	16(24.2)	1(1.5)
Wound infection	25(37.9)	1(1.5)
Purulent pus	12(18.2)	0(0.0)	0.434(0.569–4.183)	0.039
Aspirate effusions	13(19.7)	1(1.5)
*Other infections	108(42.2))	0(0.0)

(P<0.05 significant

*other infection include eye infection, throat and endocervical collections, n, number;%, percentage rate).

More than 52.5% resistant rates to tetracycline, sulphamethaxazole, gentamycin and amoxicillin were recorded among S. aureus from different sources as shown in Fig 1. The Box plot further reveal the overall antibiotic resistance rates of staphylococcal strains in different clinical conditions with significant estimated median resistance in all the presenting clinical disease conditions (p = 0.001) excluding susceptible strains isolated from eye, throat and endocervical samples but strains from aspirates and otitis had close median resistance rates (p = 0.056). Highest percentile (75^th) and median resistance were observed in wound strains than others. Overall resistance rate of 78.8% to AMC at MIC₉₀ (128μg/ml) was recorded for strains obtained from aspirates, while strains recovered from pus, ear and wound infections showed more than 30% resistance at MIC₅₀ (8–16μg/ml). S. aureus strains (59.3%) recovered from pus and aspirate were resistant to CRO at MIC₉₀ (64μg/ml) and MIC₅₀ (8μg/ml), respectively. More than 40% of S. aureus strains obtained from aspirate had high resistance to TET (MIC₉₀ and MIC₅₀ at 64μg/ml), while strains recovered from pus were resistant to GN (MIC₉₀ and MIC₅₀ at 128μg/ml and 16μg//ml), respectively (Table 2).

Fig 1

Radar plot of antibiotic resistance profile of Staphylococus aureus strains obtained from different infections and Box plot showing median distribution of antibiotic resistance pattern.

Table 2

Phenotypic resistant of S. aureus strains from various infection sources to antibiotics.

Antibiotics Agents	Range	Break point of resistance	Wound (n = 26)		Ear (n = 18)		Pus (n = 13)		Aspirate (n = 9)		Percentage of resistance (%)
		MIC (μg/ml)
			MIC50	MIC90	MIC50	MIC90	MIC50	MIC90	MIC50	MIC90
TE	0.25–128	16	8	32	4	64	4	64	8	64	43.0
CAZ	0.1–64	4	4	64	4	32	16	64	2	32	36.5
CIP	0.12–16	4	4	16	1	64	2	64	1	32	38.9
GEN	0.03–2.0	1	2	16	1	64	2	128	16	64	40.2
AMC	0.25–64	16	2	32	2	32	1	32	8	128	78.8
CRO	0.1–64	4	8	32	2	32	2	64	4	64	59.3
OFX	0.12–64	4	4	16	1	64	1	32	8	32	35.6
SXT	0.5–128	32	4	32	16	128	16	128	16	64	41.7
E	0.5–64	32	1	16	4	16	4	64	16	128	34.0
FOX	0.1–64	4	2	16	2	32	8	128	8	64	46.5
LZD	0.1–64	2	1	16	1	32	4	64	2	32	34.2
VA	0.1–64	4	1	8	4	32	16	32	8	4	30.1

Notes: N, number of isolates; N, Number of samples; TE, Tetracycline; CAZ, Ceftazidime; CIP, Ciprofloxacin; GEN, Gentamycin; AMC, Amoxycillin-clavulanic acid; CRO, Cefuroxime; OFX, Ofloxacin; SXT, sulfamethoxazole; E, Erythromycin; FOX, fosfomycin; LZD, Linezolid; VA, Vancomycin, MIC; Minimum inhibitory concentration.

Resistance relatedness of extra-intestinal S. aureus strains

Only three S. aureus strains recovered from aspirate, otitis media and wound expressed mecA gene (Fig 2), several strains clustered into group C with similar MARI of more than 0.50, characterized with biofilm and high beta-lactamase production. More than 0.83 MARI were observed among the strains that clustered into group A with high number of strains producing beta-lactamase and strong biofilm, but only one strain of MARI 0.92 clustered to Group D. In all, only 23/66 strains were biofilm producer but differed in level of production.

Fig 2

Antibiotic resistance relatedness of the recovered Staphylococci strains with high multi-antibiotic resistance index (MARI), biofilm and beta-lactamase production and mecA genotype.

Clonal diversity of identified spa types

Heterogeneous spa types from extra-intestinal staphylococci strains clustered meta-spa types into six separate clades, of which spa t442 (from wound and pus), t657 (wound), t091 (otitis media) and t657 (otitis media and wound) clustered into clade F1 with other spa types from blood stream and soft tissue infection (red rectangular). High phylogenetic relatedness of spa sequences of livestock-associated S. aureus strains (bovine milk-MH675788.1, MG821315.1 and MH675814.1) clustered with the human strains (Fig 3).

Fig 3

Neighbor-joining tree showing the phylo-diversity of Staphylococci characterized by heterogeneous spa types.

Encoded agr and focal dissemination

Clonal strains pvl+ MSSA-CC1 (4.6%) obtained from wound samples, of spa t1839, majorly encoded exfoliative toxin (etD, etB), proteases (aur, slpA sspB, sspE, sspP) and resistant determinants; bla (beta lactamase repressor (inhibitor) and beta-lactamase regulatory protein); fosB (Metallothiol transferase); sdrM (tet efflux protein) and Q2YUB3 (Multidrug resistance transporter) (Table 3), expressed agrI functionality. Obviously, 13.6% pvl+ MSSA belonging to clonal lineage CC5 from pus, wound and abscess harboured numerous heterogeneous spa types with functional agrII encoding enterotoxin sea, sec, sed, sej, sel, ser), leukocidins (LukF-PV, lukD, lukE) and proteases (aur, slpA sspB, sspE, sspP). In addition, agrII was also recorded in 4.6% pvl+ MRSA-CC7 strains of spa t091 characterized with LukF-PV, lukD, lukE, proteases and aphA3, (3,5-aminoglycoside phosphotransferase encoding neomycin/ kanamycin resistance); sat (Streptothricine-acetyltransferase); tetK (Tetracycline resistance markers); msr (A) (Macrolide efflux); and mph(C), (lysylphosphatidyl-glycerol synthetase). Most MSSA strains were observed to be prevalent at urban communities showing focal dissemination to other nearest suburbs while identified MRSA was observed to be spreading together with other MSSA strains (Fig 4).

Fig 4

Geospatial mapping and focal dissemination of multi-antibiotic resistance Staphylococcus aureus pathotypes characterized with heterogenous spa genes distributed in various communities divided according to boundary marks in southwest Nigeria.

Table 3

Functional agr, clonal types and gene determinants in MSSA and MRSA pvl positive strains.

Agr types	Strains (%)	Sources	Clonal Complex	spa types	Virulence determinants	Antibiotic resistance genes
agrI	MSSA (4.6)	wound	CC1(ST772,ST573)	t1839	sea, lukD, lukE, sak, chp, scn, etD, etB, aur, slpA sspA, sspB,sspP	Q2YUB3, fosB, sdrM, bla, dfrG,tetK
agrII	MSSA (13.6)	Pus, wounds, abscess	CC5(ST5, ST73, ST492, ST1447)	t002, t010, t053, t067, t088, t179, t214, t242, t442, t509, t688, t1062, t1265, t6709	Sea, sec, sed, sej, sel, ser, LukF-PV, lukD, lukE, scn, aur, splA sspA, sspB,sspP	fosB, msr (A), bla mph(C),aphA3, sat,fosB, sdrM, Q7A4X2
agrII	MRSA (4.6)	wound	CC7(ST789)	t091	lukD, lukE, sak, scn, aur, splA, slpE, sspA, sspB,sspP	bla, fosB,aacA-aphD, aphA3,sat,tetK,sdrM, ccrC, aacA-aphD,aphA3,

Note: Enterotoxin genes (sea, sec, sed, sej, sel, ser); Leukocidins (LukF-PV, lukD, lukE); exfoliative toxin (etD, etB); Proteases(aur, slpA), bla (beta lactamase repressor (inhibitor) and beta-lactamase regulatory protein); fosB (Metallothiol transferase); aacA-aphD (Bifunctional enzyme Aac/Aph; gentamicin, tobramycin resistance); aphA3, (3,5-aminoglycoside phosphotransferase, neo-/ kanamycin resistance); sat (Streptothricine-acetyltransferase); tetK (Tetracycline resistance markers); sdrM (Multidrug efflux protein, tetEfflux); msr (A) (Macrolide efflux); mph(C) (Probable lysylphosphatidyl-glycerol synthetase); Q7A4X2 (Putative protein); Q2YUB3 (Multidrug resistance transporter).