Occurrence, distribution and pattern analysis of methicillin resistant (MRSA) and methicillin sensitive (MSSA) Staphylococcus aureus on fomites in public facilities

ABSTRACT

Methicillin-resistant Staphylococcus aureus (MRSA) is a human pathogen incriminated as a causative agent of hospital nosocomial infections as well as a wide range of diseases in communities. This study was conducted to assess the occurrence and distribution of MRSA and methicillin-sensitive (MSSA) on different fomites in public facilities in northern Jordan and to determine their antibiograms, toxin genes profiles, as well as identify their genetic relatedness. A total of 2600 swabs were collected from 14 fomite surfaces in a variety of public facilities including hospitals, universities, schools, transportation sites, and market places. The identity of the 380 S. aureus isolates was confirmed. Among them, 158 (41.6%) were MRSA while the rest of the isolates, 222 (58.4%) were MSSA. MRSA isolates were recovered from all fomites sites. However, among the total collected samples, the percentages of MRSA in public facilities were significantly higher in hospitals and transportation fomites, while percentages of MRSA among fomites sites were higher in public reception sites, chairs, and toilet seats. Antibiotic resistance profiles indicated that 24.5% of the isolates were resistant to cefoxitin, oxacillin, and oxytetracycline. In contrast, only 3.95% were resistant to trimethoprim-sulfamethoxazole, and 15.3% were resistant to ciprofloxacin. Multidrug-resistant patterns were higher in MRSA than in MSSA isolates. There was no apparent difference in toxin gene profiles between MRSA and MSSA. Molecular analysis revealed 85 patterns and 16 clusters at a 9% mean similarity level. In conclusion, this study provides evidence for the potential of MRSA transmission via inanimate surfaces.

Introduction

Staphylococcus aureus is a human pathogen with a potential to cause a wide range of diseases in communities and hospitals [3].

As a commensal pathogen, S. aureus mainly colonizes the nasal cavity. In addition, it is found on the skin, oral cavity, upper respiratory tract, lower urogenital tract, gastrointestinal tract and perirectal area [13,14]. In fact, 25–30% of people are permanently colonized with S. aureus while about 60% of the population are transiently colonized with this pathogen [20].

Shortly after the introduction of penicillin, S. aureus developed resistance against it. Later in 1959, new beta-lactam semi-synthetic compounds called methicillin were developed to replace penicillin. Unfortunately, only 2 years after their introduction, some strains of S. aureus developed resistance to many beta-lactam antibiotics including methicillin, oxacillin, cefoxitin and others and thus, these strains were given the name methicillin-resistant Staphylococcus aureus (MRSA) [25]. Since then, MRSA has become a worldwide pathogen causing morbidity and mortality at rates higher than those caused by methicillin-susceptible S. aureus (MSSA) [19,43].

MRSA has been always considered as a hospital-associated pathogen, but because of its continuously changing epidemiology, it emerged in community settings causing infections in healthy individuals [13]. Thus, the name ‘community-associated MRSA’ (CA-MRSA) has emerged. The first case of CA-MRSA was reported in the United States in the 1990s, followed by other infections in 1999 [6,65]. MRSA commonly causes severe infectious diseases, including but not limited to food poisoning, pyogenic endocarditis, suppurative pneumonia, otitis media, osteomyelitis, and pyogenic infections of the skin, soft tissues and other diseases [3]. However, since the 1990s, genotypic differences between HA-MRSA and CA-MRSA begin to narrow down providing evidence that both CA-MRSA and HA-MRSA can each invade the other’s niche [65].

In the Mediterranean region, the highest proportions of MRSA were reported in hospitals in Jordan, Egypt and Cyprus [9]. This could be due to overcrowding and poor infection control policies in hospitals, and to uncontrolled use of antibiotics without prescription leading to the generation of populations of antibiotic resistance pathogens such as MRSA [9].

The survival of MRSA in the environment is considered a serious problem for the control of its infections [62]. Fomites in public facilities are vehicles for the transmission of MRSA in the environment and among humans. MRSA can be transmitted directly from person to person or indirectly from contaminated fomites to humans [57,70]. Because MRSA has the ability to survive well on public fomites for weeks, susceptible individuals in the community can be infected by contact with these fomites and transmit the infection to other people [70].

Molecular typing is essential for tracking outbreak agents, and for studying genetic relatedness and the distribution of agents causing these outbreaks such as S. aureus. Multiple methods are used for typing these pathogens. However, each of these methods has its own strengths and limitations. Pulsed-field gel electrophoresis (PFGE) is considered the ‘golden standard’ for molecular typing of S. aureus isolates and many other bacteria [49]. However, there are limitations for the use of PFGE, such as its complexity, demanding protocols, high cost and difficulty in interpreting the results especially if the experiment is repeated in multiple laboratories [16,63]. PCR-RFLP of the coagulase gene (coa) is one of the most reliable and easiest methods for typing S. aureus [23]. However, coagulase is not produced by all strains of S. aureus, thus excluding coagulase negative S. aureus strains from the analysis. Multi locus sequence typing (MLST) is a good method of tracing the spread of certain MRSA clones, however, it is an expensive method where the sequence of at least seven housekeeping genes should be determined to identify the clone of each isolate [18].

Multiple-locus variable-number tandem-repeat analysis (MLVA) is one of the methods used for bacterial typing based on the detection of highly polymorphic variable numbers of tandem repeats (VNTR) [48].

VNTR analysis using multiplex PCR detects the polymorphism in seven individual genes (sspA, spa, clfA, clfB, sdrC, sdrD and sdrE) by the simultaneous amplification of highly polymorphic internal regions within these genes. The amplified regions are detected by gel electrophoresis and the banding patterns are used to infer the diversity and genetic relatedness among different S. aureus isolates [40,44,51].

Reports concerning prevalence rates and episodes of community-associated MRSA (CA-MRSA) in Jordan have been virtually absent which could be due to under reporting or lack of proper statistics. In hospitals though, MRSA nosocomial infections are creating many problems and proved to be a tough problem to solve. This study was undertaken to investigate the occurrence and distribution of S. aureus, in particular, MRSA, on fomites in public facilities including two hospitals in northern Jordan and to study their genetic diversity as well as to study their antibiograms and their toxin gene profiles.

Materials and methods

Materials

Peptone buffer solution, Baird–Parker agar (BPA) base supplemented with egg yolk-tellurite emulsion, trypticase soy broth, nutrient agar, nutrient broth, Muller-Hinton agar plates and antibiotic discs were purchased from Oxoid, UK. The wizard genomic DNA purification kits, PCR Master mix, nuclease free water, 50 bp DNA ladder, MgCl₂ and ethidium bromide were purchased from Promega, USA. DNA ladders were obtained from Fermentas, USA, agarose gel from Biobasic, Canada and 1 X TBE (Tris-borate-EDTA) buffer from Lonza, Belgium. All PCR primers were synthesized in the Princess Haya Biotechnology Center, Jordan. PCR reactions were run using a Mygenie 96 Thermal Block Cycler (Bioneer, Korea).10,11,28,64

Sample collections

A total of 2600 fomite’s swabs were collected in 2011 from fomites in 14 different sites in northern Jordan and were examined for the presence of S. aureus (Table 1). Samples were collected from two public hospitals (King Abdullah University Hospital, and Ramtha Hospital), 6 universities (Jordan University of Science and Technology, Jarash University, Irbid National University, Yarmouk University, Balqaa University and Al Albait University). In addition, swabs were taken from several governmental schools in Ramtha city, and from several ATMs in the city and from two different types of public transportation vehicles (buses and taxies).

Table 1.

Sources of swabs, number collected and percent distribution

Swab site	Number of collected swabs (%)						Total	%
	Hospitals1 (2)		Universities2 (6)	Schools (several)	Transportation Sites (2 types)	Markets (several)
Wash rooms3		92	149	54	-	-	295	11.34
Toilet seats		70	33	-	-	-	103	4
Toilet door handles		124	253	95	-	-	472	18.2
Other door handles		69	85	120	-	-	274	10.6
Toilet flushes		-	54	-	-	-	54	2.1
Elevators4		51	43	-	-	-	94	3.7
Student desks5		-	491	189	-	-	680	26.2
Hospital Receptions6		77	-	-	-	-	77	3
Chairs		60	-	-	-	-	60	2.3
Others7		31	-	-	-	-	31	1.2
Public Transportation8		-	-	-	337	-	337	13
Markets9		-	-	-	-	123	123	4.7
Total (%)		574(22.1)	1108(42.6)	458 (17.6)	337(13)	123(4.7)	2600	100

¹Two hospitals, ² 6 universities, ³wash rooms include all touched surfaces in the toilet except toilet seat, toilet door handles and toilet flushes, ⁴ elevator buttons outside and inside, ⁵desktop of the student desks, ⁶ reception counter tops in hospital wards, ⁷ others include coffee machines, ⁸ transprotaiton door handles and chairs in busses and taxies located in Ramtha and Iribd cities, ⁹markets include electronic ATMs, counter tops and carriages in Ramtha and Iribid cities.

Surfaces were sampled using sterile polyester swabs moistened in sterile 0.1% peptone buffer solution by moving the swab several times over the entire surface area being tested (size of the surfaced area vary depending on the swabbed surface). Swabs were maintained under aseptic conditions at 4° to 5°C and returned to the laboratory within 2 hours of sampling to be examined.

Enrichment and isolation of S. aureus

Swabs were enriched in trypticase soy broth and incubated for 24 hours at 37°C. After incubation, a loop of broth was streaked onto BPA base supplemented with egg yolk-tellurite emulsion and incubated for 24–48 hours at 37°C. Dark black shiny colonies that were surrounded by a clear zone were considered presumptive isolates of S. aureus. Presumptive S. aureus were propagated in nutrient broth and stored in a mixture of nutrient broth and glycerol at 80:20% at −40°C until used. Presumptive S. aureus isolates were confirmed by amplifying the nuc gene which is S. aureus specific.

DNA extraction and PCR amplification of nuc, coa, mecA, vanA and vanB genes

DNA was extracted from 10 ml broth cultures using the Promega wizard genomic DNA purification kit (Promega, USA) following the manufacturer’s instructions. Extracted DNA from all Staphylococcus isolates was tested for the presence of the following genes; nuc, coa, mecA, vanA and vanB genes using the primers listed in Table 2 and the PCR conditions as per the cited reference for each primer set.

Table 2.

PCR Primers used to confirm the identity of the S. aureus isolates (nuc), presence of coa gene and the presence of antibiotic resistance genes (van A, vanB, mecA) as well as the toxin genes

Gene	Primers	Size of amplified Product (bp)	Reference
van A	5’CATGAATAGAATAAAAGTTGCAATA3’ R 5’ CCCCTTTAACGCTAATACGATCAA 3’	1030	[11]
van B	F 5’ GTGACAAACCGGAGGCGAGGA3’ R 5’ CCGCCATCCTCCTGCAAAAAA3’	433	[11]
mec A	F 5’ GCA ATC GCT AAA GAA CTA AG 3’ R 5’ GGG ACC AAC ATA ACC TAA TA3’	222	[55]
nuc	Pri-15’ GCGATTGATGGTGATACGGTT3’ Pri-2’AGCCAAGCCTTGACGAACTAAAGC3’	270	[10] [10]
Coa	F 5’ATAGAGATGCTGGTACAGG3’ R 5’ GCTTCCGATTGTTCGATGC3’	Variable	[23]
sea	SEA-1 AAAGTCCCGATCAATTTATGGCTA SEA-2 GTAATTAACCGAAGGTTCTGTAGA	219 bp	[64]
seb	SEB-1 TCGCATCAAACTGACAAACG SEB-2 GCAGGTACTCTATAAGTGCC	476 bp	[28]
sec	SEC-1 GACATAAAAGCTAGGAATTT SEC-2 AAATCGGATTAACATTATCC	257 bp	[28]
sed	SED-1 CTAGTTTGGTAATATCTCCT SED-2 TAATGCTATATCTTATAGGG	317 bp	[28]
see	SEE-1 TAGATAAAGTTAAAACAAGC SEE-2 TAACTTACCGTGGACCCTTC	169 bp	[28]
tst	TSST-1 ATGGCAGCATCAGCTTGATA TSST-2 TTTCCAATAACCACCCGTTT	350 bp	[28]
eta	ETA-1 CTAGTGCATTTGTTATTCAA ETA-2 TGCATTGACACCATAGTACT	119 bp	[28]
etb	ETB-1 ACGGCTATATACATTCAATT ETB-2 TCCATCGATAATATACCTAA	200 bp	[28]
pvl	luk-PV-1 ATCATTAGGTAAAATGTCTGGACATGATCCA luk-PV-2 GCATCAASTGTATTGGATAGCAAAAGC	433 bp	[32]
clfA	ClfA-F5’-GATTCTGACCCAGGTTCAGA ClfA-R5’-CTGTATCTGGTAATGGTTCTTT	Multiple	[50]
clfB	ClfB-F 5’-ATGGTGATTCAGCAGTAAATCC ClfB-R 5’-CATTATTTGGTGGTGTAACTCTT	Multiple	[50]
sdr	SdrCDE-F 5’-GTAACAATTACGGATCATGATG SdrCDE-R 5’-TACCTGTTTCTGGTAATGCTTT	Multiple	[50]
spa	Spa-F 5’-AGCACCAAAAGAGGAAGACAA Spa-R 5’-GTTTAACGACATGTACTCCGT	Multiple	[50]
ssp	SspA-F 5’-ATCMATTTYGCMAAYGATGACC SspA-R 5’-TTGTCTGAATTATTGTTATCGC	Multiple	[50]

PCR conditions used for these primers were as per the cited references

Antimicrobial sensitivity testing

Antimicrobial sensitivities of the 380 isolates were performed using the Kirby-Bauer disk diffusion procedure according to the 2010 CLSI guidelines [12] against six commonly used antibiotics representing 6 different antibiotic groups; [Oxicillin (OX, penicillins), Cefoxitin (FOX, cephalosporins), Vancomycin (VA, glycopeptides), Trimethoprim-Sulfamethoxazole (SXT, sulfonamides), Ciprofloxacin (CIP, quinolones) and Oxytetracycline (OT, tetracyclines)].

The inoculum was prepared by growing bacteria until the suspension reached a turbidity of 0.5 McFarland (1.5*10⁸ cfu/ml). Muller-Hinton agar plates (Oxoid, UK) were then inoculated using a sterile cotton swab. The plates were incubated for 16–18 h at 35°C after adding the antibiotic discs. The diameters of the inhibition zones were then measured. Results were reported as Sensitive (S) or Resistant (R) according to the 2010 CLSI guidelines [12].

Toxin genes profiling of the isolates

All the isolates were screened for the presence of nine toxin genes (sea, seb, sec, sed, see, tst, eta, etb and pvl) using the primers listed in Table 2. All PCR reactions were performed as uniplex. PCR mixtures contained 0.4 mM of each primer, 1X GoTaq® Green Master Mix (Promega) and nuclease-free water to bring the final volume to 25 μl. Finally, 1 μl of DNA template was then added to each reaction tube. PCR amplification conditions consisted of an initial denaturation at 94°C for 5 min followed by 35 cycles; denaturation at 94°C for 45s, annealing at 55°C for 45s and extension at 72°C for 45s. The final extension was done at 72°C for 7 min. The tubes were finally preserved at 4°C. PCR amplifications were performed by using Lab Cycler Thermocycler (SensoQuest). The PCR products were analyzed by gel electrophoresis on 2% agarose using SeaKem® LE Agarose (Lonza) in 1X AccuGENE® TBE Buffer (Lonza). The 50 bp Gene-Ruler™ DNA ladder (Fermentas) was included as the molecular marker. The agarose gels were run at 120 volts for 30 minutes.

Multiple-Locus Variable Number Tandem Repeat Analysis (VNTR)

A set of PCR primers (Table 2) was used to amplify the hypervariable regions of spa, sspA, clfA, clfF, sdr. The sdr locus comprises two or three closely linked and tandemly arrayed open reading frames containing sdrC, sdrD, and sdrE, which encode sdr proteins [50]. Amplified DNA bands were visualized using a UV transilluminator, photographed, scanned and documented.

Computer analysis of VNTR data

Banding VNTR patterns of the 380 S. aureus isolates were scored using Quantity one software (Biorad, USA), with the data coded as a factor of 1 or 0, representing the presence or absence of bands, respectively. A similarity matrix among S. aureus was produced using Jaccard coefficient as:

S_j = n_xy/n_x + n_y. Where n_xy is the number of shared fragments between two samples; n_x and n_y are total number of non-repeating fragments observed in samples x and y, respectively. A dendrogram showing the genetic relatedness among the isolates was constructed from the resulting data using SPSS, version 16.0 (SPSS Inc.1989–2007). The cutoff for the dendrogram was selected based on the average of mean similarity matrix (9%). Any band change was classified as a distinct MLVA pattern [50].

Results

Isolation, primary identification and confirmation of S. aureus

Among the 2600 tested fomite’s swabs, presumptive S. aureus colonies were recovered from 475 swab samples on BPA. When these colonies were further tested for coagulase activity, 451 showed a positive reaction, representing 17.34% of total collected samples. However, for a final confirmation of the S. aureus identity, the 451 isolates were tested using primers for the S. aureus specific-thermonuclease gene (nuc) and 380 (14.6%) isolates were confirmed as S. aureus.

PCR confirmation for the presence of mecA, vanA and vanB genes in S. aureus isolates

The presence of mecA gene was used as indicator of the number of MRSA among the confirmed isolates. Testing with PCR revealed that 158 (41.6 %) of the isolates harbored the mecA gene and thus were described as MRSA, while the other 222 (58.42%) tested negative and were described as MSSA. The results showed that MRSA isolates were recovered from all swab sites and public facilities. However, among the total 2600 collected samples, the percentages of MRSA in public facilities were higher in hospital fomites then transportation sites followed by university fomites and schools and least in market samples. According to fomite types, the percentages of MRSA were significantly higher on toilet door handles, student seats and wash rooms in comparison with other fomite types (p > 0.001) (Table 3A). In contrast, percent MSSA of the total isolates were almost the same around 9 for all locations except for transportation were percent MSSA was only 3.2% (Table 3B). Regarding the distribution of MSSA isolates among different fomites, the student disks contained the highest percentage followed by toilet door handles, door handles, markets, washrooms and the others (Table 3B).

Table 3A.

Distribution of MRSA isolates on fomites of the five public facilities identified in the total collected samples (2600]

		Isolates confirmed as MRSA
Swap site	Total No. of collected samples	Hospitals (n = 574)	Universities (n = 1108)	Schools (n = 458)	Transportation sites (n = 337)	Markets (n = 123)	Total No. of MRSA isolates
Wash roomsa	295	2	10	5		-	20
Toilet seats	103	5	9	-	-	-	14
Toilet door handles	472	17	14	6	-	-	37
Door handles	274	10	1	2	-	-	13
Elevators	94	2	2	-	-	-	4
Student disks	680	-	15	9	-	-	24
Toilet flushes	54	-	5	-	-	-	5
Hospital receptions	77	10	-	-	-	-	10
Chairs	60	9	-	-		-	9
Markets and Othersb,c	154	2	-		-	4	6
Publicd trans	337	-	-	-	19	-	18
Total	2600	57	56	22	19	4	158
% of total samples		10	5	4.8	5.6	2.5
% of MRSA		36.1	35.4	14	12	2.5

^aWash room include all touched surfaces in the toilet except toilet seats, toilet door handles and toilet flushes,

^bOther include hand prints and coffee machines

^cMarkets include electronic exchangers, counter tops and carriages; ^d transprotation n: total collected samples of each public facility.

Table 3B.

Distribution of MSSA isolates on fomites of the five public facilities identified in the total collected samples (2600)

		Isolates confirmed as MSSA
Swap site	Total No. of collected samples	Hospitals (n = 574)	Universities (n = 1108)	Schools (n = 458)	Transportation sites (n = 337)	Markets (n = 123)	Total No. of MSSA isolates
Wash roomsa	295	8	10	3		-	21
Toilet seats	103	9	3	-	-	-	12
Toilet door handles	472	10	29	9	-	-	48
Door handles	274	9	6	17	-	-	32
Elevators	94	3	2	-	-	-	5
Student disks	680	-	43	12	-	-	55
Toilet flushes	54	1	5	-	-	-	6
Hospital receptions	77	5	-	-	-	-	5
Chairs	60	1	-	-		-	1
Markets and Othersb,c	154	3	-		-	23	26
Public transd	337	-	-	-	11	-	11
Total	2600	49	98	41	11	23	222
% of total samples		8.5	8.8	8.9	3.2	8.7
% of MSSA		22.1	44.1	18.4	5	10.4

^aWash rooms include all touched surfaces in the toilet except toilet seats, toilet door handles and toilet flushes

^bOther include hand prints and coffee machines,

^cMarkets include electronic exchangers, counter tops and carriages; ^dtransprotation n: total collected samples of each public facility.

All isolates as part of the pathogenicity characterization were separately tested for their susceptibility to vancomycin by PCR amplification of the vanA and vanB genes. None of the isolates was positive for these genes and therefore, was considered vancomycin susceptible isolates.

Antibiotic susceptibility testing of S. aureus isolates using the disc diffusion method

Antibiotic susceptibility results indicated that 93 isolates (24.5%) were resistant to cefoxitin, oxacillin and oxytetracycline. In contrast, only 15 isolates (3.95%) were resistant to trimethoprim-sulfamethoxazole, and 58 isolates (15.3 %) were resistant to ciprofloxacin. In addition, none of the S. aureus isolates exhibited resistance to vancomycin.

Based on the detection of the mecA gene by PCR, 158 of the 380 S. aureus isolates were classified as MRSA and 222 as MSSA. Therefore, the antibiotic susceptibility for each group was evaluated separately. Figure 1. depicts the percentages of the MRSA and MSSA isolates that were resistant to the 6 different classes of antibiotics.

Figure 1.

The resistance of MRSA (n = 158) and MSSA (n = 222) isolates to 6 antibiotic groups

An isolate is considered multidrug resistant (MDR) if it exhibited resistance to 3 or more antibiotic classes [8]. By evaluating the resistance profiles of MRSA and MSSA isolates, 27.2% (n = 43) of the MRSA isolates exhibited multi-drug resistance patterns, while only 6.4 % (n = 14) of the MSSA isolates exhibited multidrug resistance. However, neither MRSA nor MSSA isolates exhibited resistance to more than four antibiotic classes, (Figure 2).

Figure 2.

The percentages of MRSA and MSSA isolates that were resistant to different antibiotic groups

Toxin gene profiling of the isolates

All the 380 isolates were tested for the presence of nine Staphylococcus enterotoxin genes. Approximately 73% (n = 277) of both MRSA and MSSA isolates contained at least one toxin gene, while 27% (n = 103) of the isolates did not harbor any of the tested toxin genes. The highest prevalence was for pvl, with 248 (65.3%) of the isolates containing the gene. In addition, 68 (17.8%) and 30 (7.9%) of the isolates were positive for the tst and sea genes, respectively. Other toxin genes where less prevalent such as sec and eta, with each being detected in 16 isolates. Interestingly, etb and sed genes were absent from all the isolates. However, when comparing toxin gene profiles between MRSA and MSSA, there were no major differences among them except for the pvl toxin gene, which was more prevalent in MSSA isolates (71.2%) compared to MRSA isolates (57%) (Figure 3).

Figure 3.

The percent distribution of toxin genes present in MRSA and MSSA isolates

When looking at isolates harboring different numbers of toxin genes, we observed only small differences between MRSA and MSSA isolates. For example, 32.3% of MRSA isolates did not harbor any toxin gene compared to 23.4% of the MSSA isolates. Almost 55% of MSSA isolates harbored only one toxin gene, compared to 47.5% for MRSA. Both groups have similar percentages of two toxin genes. Interestingly, 7.6 % of the MSSA isolates harbored three genes compared to only 2.5% of the MRSA isolates. Only one MRSA isolate exhibited four toxin genes and one exhibited five toxin genes while none of the MSSA isolates exhibited more than three toxin genes, (Figure 4).

Figure 4.

Numbers of toxin genes per MRSA and MSSA isolates

MLVA of S. aureus isolates

Multiple locus variable number tandem repeat analysis (MLVA) was performed to study genetic relatedness among the S. aureus isolates. Patterns were determined based on visual analysis of the band scores of PCR products for VNTR patterns, using the criteria established by [50], in which any band change was classified as a distinct VNTR pattern (Figure 5). MLVA analysis of the 380 isolates revealed 85 distinct patterns for the isolates. Some of these patterns, encompassed a very high number of isolates, with one pattern contained 73 isolates from different locations. Other pattern containing 23 isolates, while another two patterns contained 20 isolates each. Among the other patterns, 210 isolates were distributed among 45 patterns, with 2–12 isolates in each pattern, while the last 34 patterns, each had a single isolate.

Figure 5.

Agarose gel electrophoresis (2%) showing PCR amplification product of the variable number tandem repeats of S. aureus. Lanes: M, 50 bp DNA ladder; (1–15), different S. aureus isolates producing different VNTR patterns; -ve, negative control (water)

The 85 patterns were used to generate a dendrogram to study the hierarchical clustering of the isolates based on the Jaccard coefficient and UPGMA clustering analysis. A cutoff value of a 9% similarity was obtained based on calculating the similarity indices between the isolates. The dendrogram revealed 16 clusters (1–16), as depicted in Figure 6.

Figure 6.

MLVA dendrogram of the isolates generated by the UPGMA algorithm. Isolate clusters were delineated with 9% similarity cutoff value generating 16 different clusters

The majority of the isolates were clustered within five major clusters with 97, 56, 48, 45 and 40 isolates falling in clusters 7, 11, 6, 2 and 9, respectively. The other 94 isolates were distributed among the remaining 11 clusters as shown in Table 4.

Table 4.

Distribution of patterns, total MRSA, CA-MRSA, HA-MRSA and multi-resistant MRSA isolates among the generated clusters

Phylogenetic Cluster	No. of isolates per cluster	No. of patterns	MRSA In each cluster (%)	No. of MRSA (No. of Multi-drug resistant MRSA)a
				CA-MRSAb	HA-MRSAc
1	27	8	13 (48%)	10 (5)	3 (2)
2	45	16	13 (29%)	8 (3)	5 (3)
3	2	2	0	0	0
4	8	3	6 (75%)	1 (1)	5 (0)
5	14	7	7 (50%)	2 (1)	5 (4)
6	48	8	16 (33%)	15 (7)	1 (1)
7	97	11	19 (20%)	8 (4)	11 (4)
8	11	5	2 (18%)	1 (0)	1 (0)
9	40	5	33 (82.5%)	16 (10)	17 (9)
10	3	1	0	0	0
11	56	10	34 (60.7%)	21 (6)	13 (2)
12	8	2	2 (25%)	2 (0)	0
13	2	1	2 (100%)	2 (0)	0
14	8	2	3 (37.5%)	2 (2)	1 (1)
15	8	2	8 (100%)	8(2)	0
16	3	2	0	0	0
	380	85	158	96	62

^aMRSA isolate resistant to at least 3 different classes of antibiotics; ^bCA: community acquired;

^cHA: hospital acquired; No: number. Clusters 3, 10 and 16 did not contain any MRSA.

The distribution of MRSA and MSSA isolates among the clusters revealed that the majority of isolates belonging to clusters 9, 4, 11 were MRSA, which constituted approximately 83%, 75%, and 61% of isolates, respectively. In addition, all isolates belonged to clusters 13 and 15 were MRSA (100%). In contrast, MRSA isolates were completely absent in clusters 3, 10 and 16. In addition, it was observed that MRSA and MSSA were almost evenly distributed among clusters 1 and 5. On the other hand, several clusters (2, 6, 7, 8, 12, and 14) contained a mixture of both MRSA and MSSA (Table 4).

The distribution of CA-MRSA and HA-MRSA among clusters varied. For example, all MRSA isolates contained in clusters 12, 13, and 15 were CA-MRSA, while 94% and 77% of those in clusters 6 and 1 were also CA-MRSA, respectively. On the other hand, in cluster 5, 71% of MRSA isolates were of HA-MRSA origin while clusters 2, 7, 9 and 11 contained both CA-MRSA and HA-MRSA isolates with no specific distribution (Table 4).

Discussion

A total of 380 S. aureus isolates were recovered from 2600 swabs used for sampling different fomites. The sites of the samples were chosen carefully to represent public places with abundance of highly touched surfaces. In addition, these places may serve as a reservoir for S. aureus and a source of its infections at the same time. They were taken from several places in northern Jordan, which is heavily populated and thus represents other parts of Jordan like the capital. Nonetheless, it would have been better if samples were taken from places representing the whole country to better reflect and probably generalize the results.

To determine the prevalence of MRSA among these isolates, amplification of mecA gene using PCR was chosen due to its consistency and specificity. PCR results revealed that 158 isolates (41.6 % of S. aureus) were MRSA, while 222 isolates (58.4 %) were MSSA. MRSA was isolated from the all sampled fomites without any exception at a rate of about 6% of the total samples. These results are consistent with the previously reported ability of MRSA to withstand numerous environmental stresses. In addition, our findings highlighted the role of highly touched surfaces as reservoirs and vehicles for transmission of this pathogen between the environment and humans [27].

In this study, the frequency of MRSA in samples collected from five public facilities varied significantly among hospitals, 57 (36.1%), universities, 56 (35.4%), schools, 22 (14%), transportation, 19 (12%), and markets, 4 (2.5%). The high prevalence of MRSA in community and hospital fomites, explains the increasing rates of community-associated MRSA (CA-MRSA) among healthy individuals. The results obtained in this study are in accordance with many results reported in other studies, which also reported isolation of MRSA from fomites in different community and hospital settings [4, 39, 41, 47, 56, 59, 60, 69, 72]. For a full review of MRSA in fomites, please refer to [27].

All isolates were vancomycin susceptible, as none of the isolates harbored the vanA and vanB genes when tested with PCR. Further, all the isolates tested negative were tested using vancomycin disks to ascertain the susceptibility of the isolates to vancomycin. This result is very comforting as it assures us that Northern Jordan might be free of VRSA. However, many global studies reported the isolation of intermediate VRSA. For example, [24], reported that none of the MRSA isolates studied expressed full vancomycin resistance according to CLSI breakpoints, but they exhibited intermediate resistance. Other researchers reported the presence of VRSA. For example, [61], reported the presence of VRSA in Northern India. Out of confirmed 783 S. aureus, they found two isolates resistant to vancomycin and teicoplanin. Similarly, in Iran, [52], recovered one vancomycin resistant isolate from clinical samples. Furthermore, a recent study from Egypt reported that 17.4% of confirmed MRSA isolates exhibited the vanA gene. Phenotypically, 21 (21.7%) of MRSA isolates were confirmed VRSA by broth macrodilution [38].

In our study, all the 380 isolates were tested for their susceptibility against six different antibiotic groups. Differences in resistance profiles of antibiotic between MRSA and MSSA were observed. For example, percentages of isolates with multi-drug resistance were higher in MRSA (27.6%) than those of MSSA (6.4%) (Figure 2). Similar results were reported by [33], who found a significant difference in antibiotic resistance to many antibiotics between MRSA and MSSA isolates.

General antibiograms also showed a large variation between MRSA and MSSA, with the majority of MSSA isolates being susceptible to all antibiotics (ranging from 83.3% for ciprofloxacin to 97.7% for trimethoprim-sulfamethoxazole). These results can be explained by the presence of SCCmec elements in MRSA isolates, which act as a trap for additional resistance genes for different antibiotics such as erythromycin, tetracycline and kanamycin. This could also happen because of transposons and plasmids integration into this cassette [46,47,53]. Similar reports from studies in Malaysia, India and Singapore indicated that MRSA isolates exhibited higher multidrug-resistance patterns than MSSA isolates [54,58,66]. Other studies reported the presence of non-multidrug resistant MRSA strains but often they are associated with CA-MRSA rather than HA-MRSA [21,42].

The resistance of MRSA isolates to trimethoprim-sulfamethoxazole and ciprofloxacin was low, with the value being 6.3% and 10.75%, respectively. This low resistance clearly contradict the numbers reported by [2],who reported that 77.8% of S. aureus isolated from fomites were resistant to Trimethoprim-sulfamethoxazole and had a remarkable resistance to ciprofloxacin. Similarly, [22] reported 87.5 % resistance of MRSA isolates from bacteremia patients to ciprofloxacin. Nevertheless, our results appeared to be in agreement with results obtained by [31] who reported 0 % resistance of MRSA isolates from community acquired soft and skin infections to trimethoprim-sulfamethoxazole. Apparently, the large variation in these numbers could be due to differences in the methods and frequency of prescribing antibiotics.

When tested separately, 41.7% and 42.2% of MRSA isolates were also resistant to cefoxitin and oxacillin, respectively, while only 19 (8.6%) of MSSA isolates were resistant to the two antibiotics (Figure 1). These results indicate the inability of these two antibiotics to accurately identify MRSA isolates in comparison to the PCR method. [55], highlighted the inability of phenotypic tests to give accurate results in identifying methicillin resistant S. aureus. The mecA-PCR positive S. aureus isolates that are susceptible to oxacillin might have heterogeneous phenotypic expression of methicillin resistance. Further, this phenomenon may be a result of nonfunctional regulatory genes such as mecRI, that is responsible for mecA gene expression [5,30,34], as well as the presence of the nonfunctional femAB operon, since inactivation of this operon has been shown to restore sensitivity to methicillin in MRSA [45].

S. aureus produces a number of toxins including Staphylococcal enterotoxins (SEs), toxic shock syndrome toxin (TSST), exfoliative toxins (ETs) and Panton Valentine Leucocidin (pvl). In this study, the toxin gene profiling of both MRSA and MSSA revealed no major differences between the two groups (figures 3 &4). The very low toxin profiles could be due to the fact that all the isolates were collected from fomites and not from patients. Interestingly, the prevalence of the SEs was very low in both groups with the most prevalent toxin gene being the sea gene, with 8% of both MRSA and MSSA carrying this gene, while none of the isolates harbored the sed gene. These results are in accordance with previous findings by [1], who reported a very low prevalence (sea, 1.2%; seb-sec 1.2%; sed, 0%) of SEs in S. aureus isolated from poultry, as well as low prevalence of the exfoliative toxins [eta and etb 0%). In addition, [1], reported a low prevalence [5.4%] of the TSST toxin. In contrast, a study conducted by [46], reported that 15% of S. aureus harbored the sed gene. Neither any of the isolates in their study harbored the tst gene nor any of the ET genes. In another study conducted, the sed gene was found in 74.5% of 19 MRSA isolates of the USA 100 clone that was isolated from patients with the invasive disease [46]. The origin of the isolates seems to affect the toxin profiles as these isolates were obtained from retail meat samples, or from clinical samples.

The pvl toxin is associated with S. aureus soft tissue infections and severe necrotizing pneumonia implying that it is frequently isolated from clinical S. aureus isolates. In this study, the pvl gene was prevalent in 57% of the MRSA and 71% of the MSSA isolates, although none of the samples was obtained from patients. This result is similar to those reported by [67], who showed that 75.5% of their S. aureus isolates that were recovered from Holstein milk harbored the pvl toxin gene. In contrast, these results were different from those reported by [7], who reported only 18% prevalence of pvl gene among S. aureus isolates recovered from pediatric patients, and a prevalence of tst gene of 17%, which is similar to the prevalence of tst obtained in this study (17.9%). These reports highlight the relation between the toxin profiles and the source of the isolates (clinical vs nonclinical).

In contrast, other researchers reported that there is no obvious trend in the toxin profiles among Staphylococcus isolates, [71] stated that the geographic distribution of super antigenic toxin genes of S. aureus strains shows large variations with the origin of the sample possibly being responsible for the difference in toxin genes profiles for Staphylococcus isolates [71].

VNTR analysis was performed on the 380 S. aureus isolates to study their genetic relatedness and distribution based on the variability of the number of the tandem repeats in seven loci (sspA, spa, sdrC, sdrD, sdrE, clfA, and clfB) amplified by multiplex PCR. Using the criteria established by [50], (where any band change was classified as a distinct MLVA pattern), VNTR produced 85 distinct patterns. This result indicates that there was a high degree of genetic variation among the 380 S. aureus isolates in this study, in spite of all samples being collected from one geographical region, Northern Jordan. However, the number of VNTR patterns appeared to be small when compared to other studies conducted by [50], who reported that 26 MLVA patterns were produced from only 34 human MRSA isolates. Another study conducted by [37],reported 40 MLVA patterns from only 59 S. aureus isolates. This genetic variation among S. aureus isolates in these studies probably was affected by the source of the isolates, which were recovered from a variety of human infections that were collected from different geographical regions to evaluate the reliability of MLVA as a typing method for S. aureus and to compare it with PFGE. Similarly, [36], reported a high heterogeneity among their isolates in which 37 MLVA patterns were observed among only 47 MSSA and MRSA isolates recovered from six hospitals in Croatia and Poland between 2004 and 2006. Geographical regions from which these isolates were collected may have influenced this high degree of heterogeneity. However, [29], obtained 69 MLVA patterns from 202 S. aureus (mostly methicillin-resistant) isolates recovered from 29 Polish hospitals, which were similar to the results obtained in the current study.

Five clusters, (3, 10, 13, 15 and 16) in this study were able to discriminate between MRSA and MSSA isolates, as clusters 13 and 15 only contained MRSA, while clusters 3, 10 and 16 contained only MSSA isolates. This indicates the limited ability of MLVA to segregate between MRSA and MSSA isolates. Similarly, [35,36] also observed a different distribution of MRSA and MSSA isolates in the MLVA clusters.

However, the other 11 clusters in our study contained percentages of both MRSA and MSSA isolates indicating that MRSA isolates in these 11 clusters shared genotypes with MSSA isolates. Among these 11 clusters, almost all MRSA and MSSA isolates in clusters 4, 8, 11, and 12 were none multi-drug resistant. This finding demonstrates that both MSSA and MRSA in these clusters had the same genetic background, suggesting that several MSSA genotypes may have acquired SCCmec. In addition, the results suggest that the dissemination of MRSA in Northern Jordan region was not only caused by the spread of MRSA clones. Instead, several prevalent clones of MSSA seem to have acquired the Staphylococcal cassette chromosome mec by horizontal transfer from MRSA isolates, thus, these MSSA clones that acquired the Staphylococcal cassette chromosome mec continued to transfer it horizontally to other MSSA isolates with a similar genetic background. This is in accordance with the findings of [26], who showed based on PFGE typing and ribotyping, that the majority (68%) of the studied 299 MSSA isolates had the same genetic background with MRSA strains and that the majority of these isolates shared genotypes with none multi-drug resistant MRSA, including CA-MRSA. Another study by [17], based on typing of S. aureus by MLST, showed that the major MRSA clones have frequently evolved from successful epidemic MSSA clones. Furthermore, the horizontal transfer of SCCmec from MRSA to MSSA has been shown to occur in a hospitalized patient during antibiotic treatment [68]. Further, the loss of SCCmec from the chromosome of MRSA strain has been observed in vivo [15]. This indicates that MRSA could be converted to MSSA spontaneously especially in CA-MRSA strains which harbor a small size SCCmec type IV and two functional recombinase genes, leading to its extensive movement and the repeated integration into MSSA isolates [53].

These results show that the location of MRSA is continuously changing. This indicates that MRSA strains from the hospital may be identified in the community and vice versa, because of the diffused boundaries that exist between the community and the hospital settings [53].

In conclusion, this study provides evidence for the potential of MRSA transmission via inanimate surfaces and underscores the need for good hygiene practices as well as the need for establishing effective infection control measures in health-care institutions. About half of MRSA isolates in this study were susceptible to cefoxitin and oxacillin, indicating that these isolates contained the nonfunctional mecA gene. MRSA isolates exhibited multi-drug-resistant patterns higher than MSSA isolates. Antibiotic susceptibility results suggest that trimethprim-sulfamethoxazole would be the best treatment for infections caused by MRSA in Jordan. There were no apparent differences in toxin profiles between MRSA and MSSA. Typing by VNTR revealed that the studied isolates were genetically diverse and comprised of a highly heterogeneous population. VNTR revealed that the majority of MSSA isolates shared the same genetic background with MRSA, suggesting that the dissemination of MRSA in Northern Jordan region is not only due to clonal spread of MRSA, but also due to horizontal transfer of the Staphylococcal cassette chromosome mecA from MRSA to MSSA.

Funding Statement

This work was supported by the Deanship of Research at Jordan University of Science and Technology [190/2009].

Ethics approval and consent to participate

There were no experiments on animals or humans or collected human or animal samples.

Competing interests

Authors declare no competing interest.

Disclosure statement

No potential conflict of interest was reported by the authors.