Clonality and Antimicrobial Susceptibility of Staphylococcus aureus and Methicillin-Resistant S. aureus Isolates from Food Animals and Other Animals

Abstract

Out of 3,081 animals studied, 24.9% of pigs, 4.7% of chickens, 6.3% of dogs, 10.5% of cats, and 7.1% of rodents were Staphylococcus aureus positive. Prevalence of methicillin-resistant S. aureus (MRSA) was high in pigs (animals, 21.3%; batches, 46.5%), with all MRSA isolates and most methicillin-sensitive S. aureus isolates belonging to clonal complex 9 (CC9) and being multidrug resistant. The predominant S. aureus CCs among dog and cat isolates were similar. Among rodent isolates, CC398 predominated, with spa t034 the most frequent spa type detected.

TEXT

Staphylococcus aureus is a major pathogen in human and veterinary medicine (3, 4, 14). Since 2005, there has been a marked increase in methicillin-resistant S. aureus (MRSA) among farm animals, especially pigs (14). These MRSA isolates belong to specific clones (sequence type 9 [ST9] and ST398) and are often resistant to other non-β-lactam antibiotics (14). However, limited studies have been performed to compare the clonal structures of methicillin-sensitive S. aureus (MSSA) and MRSA populations in food animals and other animals. Here, we investigated the prevalence of MRSA and MSSA among various animals and the strains were characterized by molecular methods. From September 2008 to August 2011, nasal or tracheal swabs were obtained from animals in a central slaughterhouse (cattle and pigs), wet markets (chickens), and urban areas (stray dogs, stray cats, and wild rodents) (6). chromID MRSA (bioMérieux, France) and mannitol salt agar plates were used for the recovery of S. aureus isolates, following an overnight broth enrichment step (5). The bacterium was identified as S. aureus by PCR methods based on an S. aureus chromosomal fragment (sau) (11). The PCR assay could distinguish S. aureus from other closely related species (11). Isolates from cats and dogs were further tested by a multiplex PCR method based on the thermonuclease (nuc) gene, allowing separation from the newly described Staphylococcus pseudointermedius (16). The disc diffusion method was used for susceptibility testing according to the CLSI (2). The isolates were characterized by spa typing, multilocus sequence typing (MLST), and staphylococcal cassette chromosome mec (SCCmec) typing as described previously (5, 8). PCR assays were used to detect the macrolide, lincosamide, and streptogramin B (MLS) resistance determinants (ermA, ermB, ermC, and mefA and mefE) and two virulence genes (pvl and arcA) (5, 7, 8).

In total, 3,081 animals, including 609 cats, 660 chickens, 589 dogs, 310 cattle, 305 pigs, and 608 rodents (281 Rattus norvegicus, 22 Rattus rattus, 151 Rattus andamanensis, 100 Niviventer fulvescens, and 54 unidentified species), were cultured (Table 1). Overall, 24.9% of pigs, 4.7% of chickens, 6.3% of dogs, 10.5% of cats, and 7.1% of rodents were S. aureus positive. A total of 254 S. aureus isolates, including 188 MSSA and 66 MRSA isolates, were recovered from 252 animals (Table 2). All but one of the isolates from chickens and pigs were resistant to three or more non-β-lactam drugs. In contrast, most of the isolates from dogs, cats, and rodents were fully susceptible. Overall, 20.3% (22/108) and 73.1% (79/108) erythromycin-resistant isolates were positive for the ermB and ermC genes, respectively (Table 2). No isolates were found to have the ermA or mefA and mefE genes.

Table 1.

Prevalence of S. aureus by animal group

Animal group	% positive (no. of animals or batches positive/total no.)
	Dog	Cat	Rodent	Chicken	Pig	Cattle
Animals with S. aureus	6.3 (37/589)	10.5 (64/609)	7.1 (43/608)b	4.7 (31/660)	24.9 (76/305)	0.3 (1/310)
Animals with MRSA	0 (0/589)	0 (0/609)	0 (0/608)	0.2 (1/660)	21.3 (65/305)	0 (0/310)
Batches with S. aureusa	24.1 (32/133)	33.8 (47/139)	57.8 (26/45)	36.4 (12/33)	53.5 (46/86)	3.2 (1/31)
Batches with MRSAa	0 (0/133)	0 (0/139)	0 (0/45)	3.0 (1/33)	46.5 (40/86)	0 (0/31)

^aAnimals were randomly sampled in batches as follows: chickens (20 animals per batch), cattle (10 animals per batch), pigs (2 to 7 animals per batch), stray cats (1 to 10 animals batch), stray dogs (1 to 10 animals per batch), and rodents from urban areas (2 to 23 animals per batch). For cats, dogs, and rodents, the batches refer to animals held in the same area of a holding facility.

^bOf the 43 S. aureus isolates from rodents, 40 were recovered from R. norvegicus and one each was recovered from R. andamanensis and N. fulvescens. The species source for one rodent S. aureus isolate was unknown.

Table 2.

Occurrence of resistance among S. aureus isolates in different animals^a

Animal type	No. of isolates	% nonsusceptible							MLS phenotype (genotype)b
		CHL	CIP	SXT	ERY	CLI	GEN	TET
Dog with MSSA	37	2.7	2.7	0	8.1	5.4	0	16.2	2 cMLS (2 ermB), 1 M (1 none)
Cat with MSSA	64	0	4.7	0	4.7	4.7	7.8	15.6	3 cMLS (3 ermB)
Rodent with MSSA	43	0	2.3	0	0	2.3	0	0
Chicken with MSSA	30	20	96.7	23.1	96.7	96.7	40	100	29 cMLS (13 ermB, 15 ermC, 1 none)
Pig with MSSA	13	84.6	53.8	23.1	92.3	100	100	92.3	12 cMLS (3 ermB, 5 ermC, 4 none)
Pig with MRSA	65	89.2	100	23.1	92.3	100	90.8	98.5	60 cMLS (1 ermB and 59 ermC)

^aCHL, chloramphenicol; CIP, ciprofloxacin; SXT, cotrimoxazole; ERY, erythromycin; CLI, clindamycin; GEN, gentamicin; TET, tetracycline; MLS, macrolide, lincosamide, and streptogramin B antibiotics; M phenotype, only ERY resistant; L phenotype, only CLI resistant; and cMLS, constitutive CLI resistance.

^bPCR results for 108 ERY-resistant isolates with the indicated phenotype. Values show the number of isolates. One MSSA isolate from cattle and one MRSA isolate from a chicken were not included in the table. The cattle MSSA isolate was susceptible to all antibiotics. The chicken MRSA isolate was susceptible to SXT but resistant to CHL, CIP, GEN, TET, ERY, and CLI (cMLS and ermC positive).

Seventy-two unique spa types were identified, including 68 and 5 spa types among the MSSA and MRSA isolates, respectively. Figure 1 shows that 78.9% (194/246) of the isolates were clustered into the following seven spa clonal complexes (CCs): spa CC899 (65 MRSA isolates and 6 MSSA isolates), spa CC034 (35 MSSA isolates), spa CC189 (27 MSSA isolates), spa CC002 (22 MSSA isolates), spa CC091 (15 MSSA isolates), spa CC701 (15 MSSA isolates), and spa CC084 (9 MSSA isolates). All porcine MRSA isolates and the only chicken MRSA isolate clustered into spa CC899 (Fig. 1). Representative isolates for each spa type were tested by MLST. The results showed that all spa types within spa CC899 belong to sequence type 9 (ST9), which is within CC9. The porcine MSSA isolates with spa type 899 (t899) were also found to belong to ST9. Among the chicken isolates, the two predominating spa types were t002 (ST5/CC5) and t034 (ST398/CC398 and ST2169/CC398). Isolates from dogs and cats shared similar spa type distributions, with the major spa types (MLST) being t189 (ST188/CC188), t701 (ST6/CC6), t091 (ST7/CC7 and ST943/CC7), and t084 (ST15/CC15). Overall, 54.1% (20/37) and 68.8% (44/64) of the isolates from dogs and cats belonged to the four spa CCs, including spa CC189, spa CC701, spa CC091, and spa CC084. The 43 rodent isolates were found to belong to 21 unique spa types, of which 24 (55.8%) isolates belonged to spa CC034. Multiplex PCR showed that MRSA isolates from pigs had SCCmec type IVb (n = 62) or type V (n = 3). The only MRSA from chickens had SCCmec type IV. The presence of the pvl and arcA genes was sought in 96 isolates (66 MRSA isolates and 30 MSSA isolates, which were chosen to represent the major spa types), and all isolates were PCR negative.

Fig 1.

Minimum-spanning tree based on spa sequence data. A total of 246 Staphylococcus aureus isolates (65 methicillin-resistant isolates and 181 methicillin-sensitive isolates) with 65 unique spa types were clustered by BioNumerics software (version 6.6) using the default parameters. The sources of the isolates are indicated by the following colors: green, pigs; red, cats; purple, rodents; yellow, dogs; turquoise, chickens. The numbers of isolates from each animal source were as follows: dogs (n = 36), cats (n = 63), rodents (n = 42), chicken (n = 29), and pigs (n = 76). Seven isolates with spa types of fewer than 5 repeats, including t2922 (MRSA, n = 1), t111 (MSSA, n = 2), t4070 (MSSA, n = 1), t605 (MSSA, n = 1), t7295 (MSSA, n = 1), and t8615 (MSSA, n = 1), and one spa PCR negative isolate were excluded. The size of each node (i.e., circle) is drawn in proportion to the number of isolates for each spa type. The values between the nodes are the distance matrices.

This study demonstrated that ST9 MRSA is the major livestock-associated MRSA (LA-MRSA) in pigs. In China, the predominant spa type in this lineage was reported to be t899, while those identified in Thailand, Malaysia, and The Netherlands were t337, t4358, and t1430, respectively (9, 12, 13). Our data showed that the predominant MSSA isolates carried by pigs were also spa t899, and they have multidrug resistance profiles similar to their MRSA counterparts. The findings indicate that this LA-MRSA might have emerged through the introduction of a mecA-carrying element into the ST9/t899 MSSA lineage. Despite the prevalence of MRSA ST9 among pigs in Asia, infections caused by this clone in animals and humans could not be found in the literature. As LA-MRSA has the potential to colonize humans and its virulence may change over time (14), ongoing surveillance is needed to detect changes in epidemiology. In pigs and chickens, the S. aureus populations are far more clonally conserved than those in the other animal species and are often multidrug resistant. We did not collect any information about antibiotic usage on the farms. Nonetheless, the observation might be the result of farm overcrowding and antibiotic selection, facilitating the spread of multidrug-resistant clones (1). Our data on the S. aureus population structure corroborates previous observations, based primarily on animal-associated MRSA, that there are host-specific subsets of S. aureus typically associated with different animal species (10, 17). Remarkably, the t034/CC398 clone was dominant among isolates from rodents (mainly R. norvegicus), which revealed that urban rodents could be an important reservoir of CC398 isolates. Dogs are not natural hosts of S. aureus (15); therefore, the S. aureus isolates detected in this study may be accidental hosts of S. aureus clones circulating in the community. In conclusion, this study revealed that there were distinctive distributions of major spa and MLST types among S. aureus populations and that only limited numbers were associated with multidrug resistance phenotypes.