Antimicrobial Resistance Profile and Genotypic Characteristics of Streptococcus suis Capsular Type 2 Isolated from Clinical Carrier Sows and Diseased Pigs in China.

Streptococcus suis serotype 2 is an important zoonotic pathogen. Antimicrobial resistance phenotypes and genotypic characterizations of S. suis 2 from carrier sows and diseased pigs remain largely unknown. In this study, 96 swine S. suis type 2, 62 from healthy sows and 34 from diseased pigs, were analyzed. High frequency of tetracycline resistance was observed, followed by sulfonamides. The lowest resistance of S. suis 2 for β-lactams supports their use as the primary antibiotics to treat the infection of serotype 2. In contrast, 35 of 37 S. suis 2 with MLSB phenotypes were isolated from healthy sows, mostly encoded by the ermB and/or the mefA genes. Significantly lower frequency of mrp+/epf+/sly+ was observed among serotype 2 from healthy sows compared to those from diseased pigs. Furthermore, isolates from diseased pigs showed more homogeneously genetic patterns, with most of them clustered in pulsotypes A and E. The data indicate the genetic complexity of S. suis 2 between herds and a close linkage among isolates from healthy sows and diseased pigs. Moreover, many factors, such as extensive use of tetracycline or diffusion of Tn916 with tetM, might have favored for the pathogenicity and widespread dissemination of S. suis serotype 2.

1. Introduction

Streptococcus suis is an important swine pathogen leading to big loss in pig production worldwide [1]. Among 35 serotypes currently identified, serotype 2 has gained more attention for the high prevalence and mortality rates in swine and human and is considered as an emerging zoonotic agent [2]. Furthermore, epidemiological surveillance has confirmed that S. suis type 2 can transmit from carrier pigs to humans [3, 4]. Clinically healthy carrier sows, harboring S. suis type 2, are considered as the major source of infection for their offspring [5]. Since the infection of S. suis type 2 commonly occurred among suckling and weaned piglets, it is essential to investigate the association of antimicrobial resistance profile and genotypic characteristicsof isolates from carrier sows and diseased pigs.

In the absence of effective vaccines to fight against S. suis, antimicrobial agents have become increasingly important in treating and controlling the infection of S. suis type 2. Of these, β-lactams, tetracyclines, sulphonamides, and macrolides are the most common antimicrobials used for the prevention and treatment of streptococcal infection in pig production. But inappropriate use of antibiotic has led to the development of resistance of S. suis to these drugs worldwide [6-9]. Furthermore, coresistance to tetracyclines and macrolides/lincosamides in human S. suis isolates was observed, and the relevance of Tn916-like conjugative transposon in coresistant mechanisms and clone diffusion have been studied [9, 10]. However, to date, few reports about coresistance to these three classes of antibiotics and the Tn916 family were found in swine S. suis serotype 2.

In recent years, research on S. suis type 2 has mainly concentrated on its potential virulence factors and pathogenic mechanisms. Many factors, including polysaccharide capsule (cps), muramidase-released protein (mrp), the extracellular protein factor (epf), the suilysin (sly), glyceraldehyde-3-phosphate dehydrogenase (gdh), and a fibronectin/fibrinogen-binding protein (fbp), were found to be associated with virulence of S. suis 2 [11-14]. Of these, mrp, epf, and sly were considered as the most relevant factors to the pathogenesis of S. suis by many researchers [15-17]. Recently, other new putative virulence factors, such as surface-associated subtilisin-like serine protease (SspA), factor H-binding protein (fhb), have also been identified [18, 19]. However, our knowledge about pathogenesis of serotype 2 is still limited despite the increasing number of studies.

To deeply understand S. suis type 2 infection, molecular typing methods are usually applied to identify individual isolates and establish genotypic characterization. Currently, many typing methods, such as randomly amplified polymorphic DNA (RAPD) [20], ribotyping [21], multilocus sequences typing (MLST) [22, 23], and genome sequencing [24], have been used to define the diversity of S. suis and to distinguish virulent from nonvirulent isolates. Of these, pulsed-field gel electrophoresis (PFGE) is one of the most powerful molecular typing methods. Many researches had been done to compare differences between S. suis from different animals, and genetic diversity was found on the basis of isolates of the different pathogenic serotypes [25-27]. However, a thorough characterization of serotype 2 isolates from healthy sows and diseased pigs has not thus far been reported.

The aim of the study is to investigate antimicrobial resistance phenotypes, genotypes, and genetic patterns of S. suis type 2 from clinical carrier sows and diseased pigs. Furthermore, the phenotypic and genotypic characterizations of these isolates were compared. To the best of our knowledge, this is the first integrative report about resistance profile and genetic diversity of S. suis serotype 2 from clinically healthy sows and diseased pigs.

2. Materials and Methods

2.1. S. suis Type 2 Isolates

A total of 96 swine S. suis serotype 2 were included in this study. 62 isolates were recovered from tonsils of clinically healthy sows of 15 epidemiologically unrelated farms in 10 regions (Jiangsu, Sichuan, Guangdong, Guangxi, Anhui, Henan, Hebei, Jiangxi, Shandong, and Beijing) from March 2005 to November 2012. Of the 34 S. suis serotype 2 obtained from diseased pigs, 31 outbreak isolates were from 6 epidemiologically unrelated farms in Sichuan, Jiangsu, Anhui, Henan, and Beijing during 2005–2008, and three historical isolates (C55604, C55609, and C55612) were provided by China Veterinary Culture Collection Center (CVCC). All isolates were determined as S. suis types 2 by biochemical characteristics (API 20 strep, bioMérieux SA, France) and sera agglutination reaction (special antisera provided by the Statens Serum Institut, Copenhagen, Denmark) and further confirmed as serotype 2 by positive PCR for the genes coding for the 16S rRNA of S. suis and for the capsule of S. suis serotype 2 (cps2J) [23].

2.2. Antimicrobial Susceptibility

Antimicrobial susceptibility testing was performed using the standard broth microdilution method [28]. The following antimicrobial agents, the representatives of commonly used drug classes in China, were tested (with dilution ranges in parentheses): penicillin G (0.06–8 µg/mL), ampicillin (0.12–16 µg/mL), ceftiofur sodium (0.12–16 µg/mL), enrofloxacin (0.015–4 µg/mL), clindamycin (0.03–4 µg/mL), erythromycin (0.06–8 µg/mL), tilmicosin (0.5–64 µg/mL), chloramphenicol (1–128 µg/mL), tiamulin fumarate (0.25–32 µg/mL), tetracycline (0.06–8 µg/mL), sulfisoxazole (16–512 µg/mL), and trimethoprim/sulfamethoxazole (0.5/9.5–16/304 µg/mL). Streptococcus pneumoniae ATCC 49619 was used as the quality control strain. The isolate was defined as an MLSB-phenotype that was found to be resistant to erythromycin, lincomycin, and tilmicosin.

2.3. Genotyping

PCR virulence genotyping of all isolates was performed for the mrp, epf, sly, orf2, fbp, and gdh genes. Moreover, 85 tetracycline-resistant (tetr) isolates were also analyzed for the presence of tetracycline-resistant genes (tetK, tetL, tetM, and tetO), macrolide-resistant genes (ermB, ermA, and mefA), lincosamide-resistant gene (lnuB), and Tn916-like transposon family (intTn and xis), respectively. Positive and negative controls were included with each PCR assay. Target genes and the corresponding primer sequences were listed in Table 1. Confirmation of the amplicons was determined by DNA sequencing and the results were analyzed using BLAST software (http://www.ncbi.nlm.nih.gov).

Table 1

Target genes and PCR primers used in this study.

Gene target(s)	Primer sequence (5′-3′)	Amplicon size (bp)	Reference
Macrolide/Lincosamide resistance genes
ermB	GAAAAGGTACTCAACCAAATA	639	[29]
	AGTAACGGTACTTAAATTGTTTAC
ermA	GAAGTTTAGCTTTCCTAA	395	[29]
	GCTTCAGCACCTGTCTTAATTGAT
mefA	AGTATCATTAATCACTAGTGC	346	[29]
	TTCTTCTGGTACTAAAAGTGG
lnuB	CCTACCTATTGTTTGTGGAA	944	[29]
	ATAACGTTACTCTCCTATTC
Tetracycline resistance genes
tetK	TATTTTGGCTTTGTATTCTTTCAT	1159	[26]
	GCTATACCTGTTCCCTCTGATAA
tetL	ATAAATTGTTTCGGGTCGGTAAT	1077	[26]
	AACCAGCCAACTAATGACAATGAT
tetO	AACTTAGGCATTCTGGCTCAC	519	[26]
	TCCCACTGTTCCATATCGTCA
tetM	GAACTCGAACAAGAGGAAAGC	740	[26]
	ATGGAAGCCCAGAAAGGAT
IntTn and xis genes
intTn	GGTCTTCGTATTTCAGAGTTTGG	473	[30]
	GTTGCATGTGCGTAATAGTTCAG
xis	AAGCAGACTGACATTCCTA	193	[30]
	GCGTCCAATGTATCTATAA
Virulence-associated factors
mrp	ATTGCTCCACAAGAGGATGG	188	[15]
	TGAGCTTTACCTGAAGCGGT
epf	CGCAGACAACGAAAGATTGA	744	[15]
	AAGAATGTCTTTGGCGATGG
sly	GCTTGACTTACGAGCCACAA	248	[15]
	CCGCGCAATACTGATAAGC
fbp	GACGGATCCTTTTTACATCACATGACGG	247	this study
	CCGTCGACGTATTTCCGCAGAATCAT
orf2	CAAGTGTATGTGGATGGG	860	this study
	ATCCAGTTGACACGTGCA
gdh	GGCGCCGAATTCGTCGACATTTAGCAATTTTTGCG	1039	this study
	CGCCGCGGATCCGTAGTTAAAGTTGGTATTAAC

2.4. PFGE Analysis

All S. suis serotype 2 isolates were typed using pulsed-field gel electrophoresis (PFGE) as described previously [31, 32] with minor modifications. Briefly, cell culture was suspended in cell suspension buffer (100 mM Tris : 100 mM EDTA, pH8.0) to 10 McFarland standards. Proteinase K was then added at final concentration of 0.5 mg/mL and mixed with equal volume of molten 1% Seakem Gold Agarose before loading into the plug mold. After solidification, the agarose plugs were submerged in cell lysis buffer (CLB, 50 mM Tris : 500 mM EDTA, pH8.0 + 1% sarkosyl) with lysozyme (1 mg/mL) before incubation at 37°C for 12 h and then CLB with proteinase K (0.5 mg/mL) was added and cultured at 54°C for 2 h with vigorous shaking. After washing, the plugs were sliced and then digested in fresh restriction buffer with the enzyme SmaI (50 U/µL) at 25°C for 12 h. The electrophoresis was performed with CHEF-DR III system (Bio-Rad) at 14°C under the electric field strength 6 v/cm for 19 h with pulse time ramping from 2.2 s to 63.8 s.

Salmonella enterica serovar Braenderup H9812 restricted with XbaI was used for molecular weight and size determinations. Similarities between restriction endonuclease digestion profiles were analyzed using BioNumerics software (Applied Maths, Kortrijk, Belgium) with Dice coefficients and clustering by an unweighted paired group with arithmetic averaging. The dendrogram of PFGE patterns of isolates tested was drawn with a 1.5% position tolerance and 1% optimization. And the cluster cutoff was set at an 85% similarity level. The different PFGE fingerprints were assigned as different uppercase letters.

2.5. Statistical Analysis

SPSS for Windows, version 16.0, was used for statistical analysis. The frequencies of antibiotic resistance, resistant genes, and virulence-associated factors were compared between the isolates from healthy carrier sows and from diseased pigs. The chi-square (or Pearson chi-square) and Fisher exact tests were used when appropriate. Differences were considered significant when two-sided P value was less than 0.05.

3. Results

3.1. Antimicrobial Susceptibility Testing and Detection of Resistance Genes

The collections of 96 S. suis 2 were tested for susceptibility to 12 antimicrobials (Table 2). High frequency of resistance was observed for tetracycline, followed by sulfonamides. 57 and 46 isolates from healthy sows and 28 and 18 from diseased pigs were resistant to tetracycline and sulfisoxazole, respectively. These two antimicrobial agents had MIC50 values (>8, 512 µg/mL, resp.) equal to or higher than the highest concentration tested. The lowest resistant rates of S. suis 2 for β-lactams were found, and all isolates were susceptible to ampicillin. Data also suggested the high incidence rates of resistance for macrolides and lincosamides in the isolates from healthy carrier sows. Among 85 tetr isolates, 37 had MLSB resistance phenotypes, 35 from carrier sows and the remaining two from diseased pigs. No inducible resistance pattern was discovered. A Significantly higher occurrence of MLSB resistance was observed in S. suis 2 from carrier sows than those from diseased pigs (P < 0.005).

Table 2

Antimicrobial resistance profile of S. suis serotype 2 from clinically healthy carrier sows and diseased pigs.

Antimicrobials	MIC breakpoint*   (µg/mL)			Healthy carrier sows (n = 62)		Diseased pigs (n = 34)		Total (n = 96)
	S	I	R	MIC range	n ∗∗	MIC range	n ∗∗	MIC range	n ∗∗
Penicillin	0.12	0.25–2	4	≤0.06–>8	2	≤0.06–1	0	≤0.06–>8	2
Ampicillin	0.25	0.5–4	8	≤0.12–2	0	≤0.12–0.5	0	≤0.12–2	0
Erythromycin	0.25	0.5	1	0.12–>8	35	≤0.06–4	2	≤0.06–>8	37
Clindamycin	0.5	1-2	4	0.06–>4	35	0.06–>4	2	0.06–>4	37
Enrofloxacin	0.25	0.5–1	2	0.12–>4	6	0.06–1	0	0.06–>4	6
Tetracycline	2	4	8	0.5–>8	57	0.25–>8	28	0.25–>8	85
Ceftiofur	2	4	8	≤0.12–>16	3	≤0.12–2	0	≤0.12–>16	3
Tiamulin	16	—	32	≤0.25–>32	7	≤0.25–16	0	≤0.25–>32	7
Tilmicosin	16	—	32	1–>64	35	≤0.5–64	2	≤0.5–>64	37
Chloramphenicol	4	8	16	≤1–32	3	≤1–8	0	≤1–32	3
Sulfisoxazole	256	—	512	32–>512	46	32–>512	18	32–>512	64
Trimethoprim/sulfamethoxazole	2/38	—	4/76	≤0.5/9.5–>16/304	13	≤0.5/9.5–16/304	2	≤0.5/9.5–>16/304	15

∗MIC breakpoints were taken from Clinical and Laboratory Standards Institute standards (CLSI).

S: susceptible; I: intermediate; R: resistant.

**Number of resistance isolates.

Antimicrobial resistance patterns and resistant determinants for tetracyclines, macrolides/lincosamides were analyzed in Table 3. No detection of tetK and tetL genes was observed among tetr S. suis 2. The tetM gene was found among 77 and the tetO gene among 51 of 85 tetr S. suis type 2. None of erythromycin- and clindamycin-resistant isolates carried ermA or lnuB. 35 S. suis 2 with MLSB-phenotype were shown to be ermB and tetM positive and 18 mefA positive. Neither ermB nor mefA was detected in the non-MLSB  tetr isolates. Presence of the ermB gene was strongly associated with MLSB-phenotype of S. suis 2. Significant carrier difference of the tetO gene, but not tetM, was observed between erythromycin-resistant (32/37) and erythromycin-susceptible (19/48) isolates.

Table 3

Distribution of antimicrobial resistance patterns, resistant determinants, virulence factors, and pulsotypes of S. Suis type 2 isolates.

Origin of isolates	Resistant patterna	Number of isolates	Number of isolates  with resistant determinantsb					Number of isolates with virulence genesc (%)			PFGE subtypes  (number of isolates)
			intTn	ermB	mefA	tetM	tetO	mrp	epf	sly
Healthy sows  (n = 62)	EryrTilrClirTetr	35	0	33	16	33	30	16	33	26	C (5), D (8), F (4), G (10), H (8)
	ErysTilsClisTetr	22	12	0	0	20	12	20	20	17	B (2), C (8), G (8), E (4)
	ErysTilsClisTets	5	/	/	/	/	/	5	5	4	G (3), H (2)
	Total	62	12	33	16	53	42	41	58	47	B (2), C (13), D (8), E (4), F (4), G (21), H (10)
Diseased pigs  (n = 34)	EryrTilrClirTetr	2	0	2	2	2	2	2	2	2	A (2)
	ErysTilsClisTetr	26	26	0	0	22	7	26	24	26	A (10), E (14), G (2)
	ErysTilsClisTets	6	/	/	/	/	/	6	6	6	A (2), E (4)
	Total	34	26	2	2	24	9	34	32	34	A (14), E (18), G (2)

aEryr: erythromycin resistant; Tilr: tilmicosin resistant; Clir: clindamycin resistant; Tetr: tetracycline resistant; Erys: erythromycin susceptible; Tils: tilmicosin susceptible; Clis: clindamycin susceptible; Tets: tetracycline susceptible.

bThe ermA, lnuB, tetK, tetL, and xis genes were not detected in the isolates included in this study.

cAll isolates carried the fbp and orf2 genes and none harbored the gdh gene.

/: Not detected.

None of 85 tetr S. suis serotype 2 was positive for the xis gene. 38 tetr isolates (26 from diseased pigs and the remaining 12 from healthy sows) carried the intTn gene. Unexpectedly, all isolates (38/96) with intTn gene were resistant to tetracycline only, while they were susceptible to macrolides/lincosamides. And higher frequency of intTn with tetM was observed in isolates from diseased pigs (22/34) compared to those from healthy carrier sows (12/62).

3.2. Virulence-Associated Genes Analysis

96 serotype 2 isolates were cloned and screened for the presence of the mrp, epf, sly, orf2, fbp, and gdh genes. The distribution of virulence-associated genes was reported in Table 3. All isolates were positive for the virulence genes coding for fbp and orf2 and negative for gdh. 41, 58, and 47 of 62 healthy sows isolates harbored mrp, epf, and sly, respectively. In contrast, the mrp, epf, and sly genes were detected in all S. suis 2 recovered from diseased pigs but two epf-negative isolates.

Six kinds of virulence genotypes were obtained in S. suis capsular type 2 from carrier sows, with high frequency of mrp+/epf+/sly+ (30/62), mrp−/epf+/sly+ (15/62), and mrp+/epf+/sly− (9/62) (Table 4). All S. suis type 2 from diseased pigs had the virulence genotype of mrp+/epf+/sly+ with two exceptions of mrp+/epf−/sly+. Significantly lower carrier rate of mrp+/epf+/sly+ genotype was observed among isolates from healthy sows compared to those from diseased pigs (P < 0.005).

Table 4

Virulence genotypes of 96 S. suis capsular type 2.

Virulence genotypes*	Number of S. suis capsular type 2
	Healthy sows	Diseased pigs	Total
mrp+/epf+/sly+	30	32	62
mrp+/epf−/sly+	2	2	4
mrp−/epf+/sly+	15	0	15
mrp−/epf+/sly−	4	0	4
mrp+/epf+/sly−	9	0	9
mrp−/epf−/sly−	2	0	2
Total	62	34	96

∗All isolates had the genotype of fbp+/orf2+/gdh−.

3.3. PFGE Typing

On the basis of an investigation of 96 S.  suis type 2, PFGE typing produced 15 different fingerprints, which were grouped into types A to H (Figure 1). Pulsotypes C1–C3, E1–E4, G1-G2, and H1-H2 were considered to be respectively related, with more than 85% similarity.

Figure 1

Genetic relationship of representative bands of S. suis type 2 isolated from clinically healthy sows and diseased pigs. Dendrogram showed the genetic relatedness of S. suis type 2 pulsotypes A to H2. The cluster cutoff (red line) was set at an 85% similarity level. Numbers at the upper left indicate percent similarity.

In contrast, significant difference was observed between the isolates from diseased pigs and healthy sows. All isolates from diseased pigs were classified as pulsotypes A, E, and G (Tables 3 and 5). Pulsotypes A and E predominated in diseased pigs and were detected in 32 of 34 S. suis 2 isolates. For isolates from clinically healthy sows, pulsotype G (21/64) was the most frequently observed one, followed by types C (13/64), H (10/64), D (8/64), E (4/64), F (4/64), and B (2/64). Interestingly, the same pulsotype, E or G, was found in S. suis 2 both from healthy sows and from diseased pigs. For example, isolates C55604 (from historically diseased pigs) and one isolate (farm GD-2, from healthy sows) with 100% homogeneity were present in pulsotype E2 (Table 5).

Table 5

PFGE patterns of 96 S. suis serotype 2 isolates included in this study.

Origin	PFGE pattern (n #)	Number of isolates	Farm number##	Isolated years
Healthy sows	B (2)	2	AH-2	2011
	C1 (4)	2	GD-1	2009
		2	GX-1	2007
	C2 (5)	3	SD-1	2006
		2	HB-1	2007
	C3 (4)	4	GD-1	2006, 2009
	D (8)	2	BJ-2	2008
		6	HN-3	2009, 2010
	E1 (3)	3	GD-2	2009
	E2 (1)	1	GD-2	2007
	F (4)	4	HB-2	2007
	G1 (13)	7	HN-1	2006
		1	JS-2	2009
		5	SD-2	2011
	G2 (8)	8	GX-2	2010, 2012
	H1 (8)	8	JX-1	2007, 2009
	H2 (2)	2	SC-3	2007
Diseased pigs	A (14)	8	SC-1	2005
		2	SC-2	2005
		4	JS-1	2005
	E1 (2)	2	C55609, C55612	/
	E2 (1)	1	C55604	/
	E3 (12)	12	AH-1	2007, 2008
	E4 (3)	3	BJ-1	2005
	G1 (2)	2	HN-2	2006

#Number of isolates with the same PFGE pattern.

##Farm number was named as capital letters (abbreviation of the province/region)—serial number. AH: Anhui; GD: Guangdong; GX: Guangxi; SD: Shandong; HB: Hebei; BJ: Beijing; HN: Henan; SC: Sichuan; JS: Jiangsu; JX: Jiangxi. C55604, C55609, and C55612 were provided by CVCC.

/: isolated time was not provided.

All S. suis serotype 2 from the same farm, including those isolated in different years, shared the pulsotype with more than 85% similarity; most of them had the identical PFGE patterns. 6 isolates from farm GD-1 in 2006 and 2009 were assigned to PFGE type C, and the similar results were observed for pulsotypes D, G2, H1, and E3. Complex relationship among S. suis 2 isolates from different farms was discovered. Three isolates from Beijing had the pulsotype (E4) different from Sichuan and Jiangsu isolates (pulsotype A) even if these S. suis serotype 2 were isolated during the largest outbreak of human S. suis 2 infection occurring in 2005. However, some pulsotypes were more frequently isolated and exhibited a wide distribution over herds compared to others. For instance, 13 isolates from different farms in Henan, Jiangsu, and Shandong provinces were classified as PFGE subtype G1, and similar results were obtained for patterns C1, C2, and D. Furthermore, S. suis serotype 2 with types G1, C1, C2, and D were resistant to tetracycline and positive for the tetM gene, although four different virulence genotypes, mrp+/epf+/sly+, mrp−/epf+/sly+, mrp+/epf−/sly+, and mrp−/epf+/sly−, were involved (not shown in tables).

4. Discussion

4.1. Resistant Phenotypes and Genotypes of S. suis 2

The lowest resistance of S. suis serotype 2 for β-lactams was in accordance with other discoveries [7, 8], supporting their use as the primary drugs to treat the infection of swine S. suis serotype 2. The resistance to tetracyclines in S. suis has become a major worldwide problem, closely related to the widespread use of tetracycline in swine production. And tetracycline-resistance has been considered to be an important cofactor in the selection of resistance to macrolides/lincosamides [9]. In this study, 85 of the 96 isolates were resistant to tetracycline, 37 of which were coresistant to macrolides and lincosamides antibiotics. This indicated less frequent coresistance of S. suis 2 to tetracyclines and macrolides/lincosamides. And the similar results were also observed by other researchers [6, 10, 33]. Furthermore, 35 of 37 S. suis 2 with MLSB-phenotype were isolated from healthy sows, indicating the presence of selective pressure of antimicrobial agents since tilmicosin and tylosin were widely used as swine feed additives in China. In contrast, most of tetr isolates from diseased pigs in backyard without feed additives were susceptible to macrolides/lincosamides.

Tetracyclines resistance in streptococci is mediated by ribosomal protection proteins or efflux proteins, encoded mainly by the tet genes [34]. Neither tetK nor tetL was detected in this study, which was consistent with other recent analyses [10]. The tetM and tetO genes, both coding for ribosomal protection protein, were widespread in tetr S. suis, and higher carrier rate of the tetM gene than the tetO gene was also observed in other studies [9, 35].

Resistance to macrolide of streptococcal clinical isolates is commonly encoded by ribosomal methylase (erm) genes and efflux (mef) genes [29]. The ermB gene was found in all but two erythromycin-resistant isolates, confirming its frequency in S. suis type 2 in China [9]. High number of MLSB-resistant isolates with the ermB gene is in agreement with other research findings [30, 36]. To our knowledge, no large scale survey concerning the distribution of the mefA gene in S. suis type 2 has been described in the literature. Wierzbowski et al. found that the mefA gene conferred low-level resistance of Streptococcus pneumoniae to macrolides only (M phenotype) [36]. In this study, about half of S. suis 2 with MLSB resistance harbored the mefA gene. In the view of the fact that all mefA-positive isolates showed MLSB-phenotype and harbored the ermB gene, mef-mediated resistance can be obscured by the effects of the erm gene in phenotypic tests.

4.2. Tn916-Like Transposon Family

Tn916 is one of the most extensively studied conjugative transposons in gram-positive bacteria. The integrase intTn gene is responsible for transposition, and the excisase xis gene may increase the frequency of excision but is not required [37]. In this present study, it is interesting that the intTn gene was detected only in erythromycin- and clindamycin-susceptible isolates and no xis genes were detected, indicating the absence of relatedness between presence of Tn916-like conjugative transposon and macrolides/lincosamides resistance phenotypes of S. suis type 2.

It is worth noting that the tet genes are often carried by Tn916-like conjugative transposon and erythromycin resistance genes are also carried on the same element, which contributes to the coresistance of streptococci to tetracyclines and macrolides. Previous studies have investigated the association between tetM/tetO and ermB/mefA in S. pneumoniae or S. pyogenes, and conjugal transfer experiments demonstrated that tetM/tetO and ermB/mefA were consistently cotransferred by Tn916-Tn1545-like transposons [38, 39]. However, the intTn gene was not detected among 35 MLSB isolates with tetM and ermB in this study, and therefore, elements other than the Tn916 family might be associated with coexistence of these two genes in S. suis serotype 2 with MLSB phenotype.

The tetM and intTn genes, the markers for the Tn916 family of elements [40], were harbored by 22 of 36 isolates from diseased pigs in the study. Ye et al. also found that tetM was associated with Tn916 in S. suis type 2 from human outbreak [9]. Thus, the presence in S. suis 2 of elements related to Tn916 with tetM could play an important role in the pathogenicity of this bacterial pathogen. Further studies are necessary to monitor the spread of these elements in S. suis serotype 2 circulating in environments.

4.3. Virulence-Associated Factors

To further characterize the molecular features of the isolates from the diseased pigs and healthy sows, six virulence-associated genes were detected. The gdh gene was not detected in any isolates, revealing that presence of this gene may be not necessary for these isolates included in this study. Different results from other studies indicated that carriage of the gdh gene was associated with multilocus sequence type (ST) or origins of isolates [41]. Moreover, the fbp gene was present in all S. suis type 2, and similar results were also observed by de Greeff et al. [12], who suggested that the fbp gene is present among most serotypes except for serotypes 32 and 34. Thus, virulence difference between the isolates from clinically healthy sows and those from diseased pigs may lie in the frequency of the mrp, epf, and sly genes, since all S. suis serotype 2 analyzed in the present study also carried the orf2 gene. Interestingly, more than 60% isolates from clinically healthy sows harbored the mrp, epf, and sly genes. The higher carrier rate of virulent genotype of mrp+/epf+/sly+ in S. suis type 2 from diseased pigs compared to healthy sows showed that these three genes together may contribute to differentiating the virulence of serotype 2, which is in accordance with results of other epidemiological reports [9, 17]. However, since 30 of 62 isolates from clinical carrier sows were also genotyped mrp+/epf+/sly+ in this study and similar result from healthy pigs was obtained by other researchers [15], it is necessary to perform further studies to specify the virulence of serotype 2 mrp+/epf+/sly+ isolates.

4.4. PFGE Subtyping

Many studies showed that PFGE could effectively detect relationship between genetic background, virulence traits, and epidemiologic implication of many bacterial pathogens [27, 31]. In this study, all isolates were typed by PFGE and clustered in 8 pulsotypes. S. suis serotype 2 from diseased pigs showed more homogeneously genetic patterns than those from healthy sows, with most of them clustering in pulsotypes A and E. In contrast, the majority of isolates from healthy sows clustered in the patterns G, C, and H, presenting a high level of divergence.

Despite the genetic diversity observed, four PFGE profiles, G, E, A, and C (accounted for 72 of 96 isolates), were more frequently observed than other patterns. These four prevalent PFGE profiles were isolated from diseased pigs and healthy sows from 9 provinces in different years (Table 5), indicating their widespread distribution in Chinese swine population. Of these, PFGE pattern G1 was unique and predominant among S. suis serotype 2 isolates from four epidemiologically unrelated herds, suggesting the existence of a prevalent clone. In consideration of the characteristic of their tetM-positive tetracycline resistance, diffusion has probably provided considerable advantages by the use of antimicrobial agents in different farms or horizontal acquiring of genetic elements, such as Tn916 with tetM [9, 42]. In addition, S. suis serotype 2 with pulsotype G1 showed dissimilarity at virulence genotypes, including mrp+/epf+/sly+, mrp+/epf−/sly+, and mrp−/epf+/sly+, which demonstrates that factors other than antimicrobial susceptibility might have favored for its diffusion.

The identical PFGE pattern (C3, D, G2, H1, or E3) was detected for the isolates from the same farm in different years (Table 5), and the persistent dissemination of S. suis serotype 2 clone within the herd of swine could be confirmed [26]. Furthermore, pulsotype E1 (E2 or G1) was found among the isolates from historical diseased pigs and from healthy sows. In our opinion, this can be explained with two main reasons. Firstly, after a long-term adaption to the healthy sows, S. suis 2 gradually lost virulence and finally became avirulent. Secondly, as a kind of conditional pathogenic bacteria, the possibility that healthy carrier sows harbor S. suis 2 is capable of causing disease under specific circumstances cannot be ruled out [5], suggesting a close linkage of S. suis serotype 2 from healthy sows and diseased pigs.

Taken together, the present study is the first systematic description of resistant phenotypes and genetic genotypes of S. suis type 2 isolated from clinically healthy sows and diseased pigs in China. Significant differences of MLSB resistance phenotype, virulence-associated genotypes, and PFGE pulsotypes were observed between isolates from clinical carrier sows and those from diseased pigs. The results indicate that β-lactams are still the primary drugs to treat the infection of swine S. suis serotype 2. The unique and predominant PFGE types within and between herds show persistent dissemination of S. suis 2 and a close linkage among isolates from healthy sows and diseased pigs. Moreover, our data also support the contention that extensive use of tetracycline and horizontal acquiring of genetic element, Tn916 with tetM, could act as a selective factor for the pathogenicity and widespread diffusion of serotype 2 [9, 42].