Resolution of Streptococcus suis Serotypes 1/2 versus 2 and 1 versus 14 by PCR-Restriction Fragment Length Polymorphism Method

Streptococcus suis is an important pathogen of pigs but is also transmissible to humans, with potentially fatal consequences. Among 29 serotypes currently recognized, some are clinically and epidemiologically more important than others. This is particularly true for serotypes 2 and 14, which have a large impact on pig production and also on human health. Conventional PCR-based serotyping cannot distinguish between serotype 1/2 and serotype 2 or between serotype 1 and serotype 14.

ABSTRACT

Streptococcus suis is an important pathogen of pigs but is also transmissible to humans, with potentially fatal consequences. Among 29 serotypes currently recognized, some are clinically and epidemiologically more important than others. This is particularly true for serotypes 2 and 14, which have a large impact on pig production and also on human health. Conventional PCR-based serotyping cannot distinguish between serotype 1/2 and serotype 2 or between serotype 1 and serotype 14. Although serotype 1/2 and serotype 2 have a very similar cps locus, they differ in a single-nucleotide substitution at nucleotide position 483 of the cpsK gene. Similarly, serotypes 1 and 14 have a very similar cps locus but also differ in the same nucleotide substitution of the cpsK gene. Fortunately, this cpsK 483G→C/T substitution can be detected by BstNI restriction endonuclease. A PCR-restriction fragment length polymorphism (RFLP) detection method amplifying a fragment of the cpsK gene digested by BstNI restriction endonuclease was developed and tested in reference strains of these serotypes and also in field isolates.

INTRODUCTION

Streptococcus suis is a facultative anaerobic Gram-positive bacterium and is the causative agent of important diseases, particularly in pigs (1, 2), but it can also cause infections in humans and other animal species. The majority of human cases have been identified in Asia (90%), of which Vietnam, Thailand, and China alone accounted for 83.6% of all cases worldwide (3, 4).

Currently, 29 serotypes of S. suis are distinguished, whereas the previously described serotypes 20, 22, 26, 32, 33, and 34 are referred to as S. suis-like (5, 6). Information about the serotype is very important from an epidemiological point of view, as some serotypes have the ability to cause more severe disease than others, and only some serotypes (especially serotypes 2 and 14) cause human disease (4).

Methods traditionally used for serotyping are the agglutination and coagglutination tests using serotype-specific antisera (7, 8). Currently, PCR methods based on the detection of genes encoding capsular polysaccharide (CPS) production of individual serotypes have been developed (5, 9, 10). Unfortunately, PCR-based methods cannot distinguish serotype 2 from serotype 1/2 or serotype 1 from serotype 14 due to the similar gene content of their cps loci (5, 9). However, a single-nucleotide G→C/T substitution at nucleotide position 483 in the cpsK gene (11) causes preferential addition of galactose (in the case of serotypes 2 and 14) or N-acetylgalactosamine residues (in the case of serotypes 1/2 and 1) to the capsular polysaccharide repeating units (12). Identification of this mutation by reading the cpsK sequence or by a recently published method based on high-resolution melting analysis (13) enables specific identification of these serotypes. Here, we present a simple and low-cost method of detecting cpsK gene polymorphism based on PCR-restriction fragment length polymorphism (PCR-RFLP).

MATERIALS AND METHODS

Reference serotype strains were used as controls (14, 15). Another 214 field isolates were recovered from clinically ill pigs on farms in the Czech Republic. S. suis identification from each sample was confirmed by biochemical testing and by matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS; Bruker Daltonik, Bremen, Germany).

All samples were tested serologically using antisera prepared by rabbit immunization with reference S. suis and S. suis-like serotype strains 1 to 34 from the S. suis strain collection at the University of Montreal (Quebec, Canada), according to a methodology originally described for serotyping Actinobacillus pleuropneumoniae (16) and S. suis isolates (8).

All samples were further tested by a multiplex PCR method detecting capsular polysaccharide synthesis genes (10). Each PCR consisted of a particular primer mix and PCR master mix (HotStarTaq Plus mastermix kit, Qiagen, Hilden, Germany). DNA for PCR was isolated by a rapid boiling method (17). PCR products were resolved on 2% agarose gel stained with ethidium bromide and visualized under UV light. Serotypes were determined according to the DNA band position (10). The multilocus sequence type (MLST) was determined according to https://pubmlst.org/ssuis/ (18).

To compare serological and multiplex PCR serotyping data, whole-genome sequences of all 35 reference serotypes and an additional 35 field isolates were acquired. The DNA of each sample was isolated by the QIAamp DNA minikit (Qiagen, Hilden, Germany). The sequencing library was prepared with the Nextera XT DNA library preparation kit (Illumina, San Diego, CA) and sequenced on a NextSeq 500 instrument (Illumina, San Diego, CA). Paired-end 2 × 150-bp reads were processed by the Tormes pipeline (19). The serotype of each sample was deduced from the assembled raw genome by comparing cps genes with cps genes corresponding to the known serotypes (5) using the BLAST tool (20). Serotypes 1/2 and 2 and serotypes 1 and 14 were further distinguished according to presence of substitution in nucleotide position 483 of the cpsK gene (11) by reading the cpsK sequence.

All field isolates belonging to serotypes 1, 2, 1/2, or 14 identified by the multiplex PCR method were further tested by both PCR-RFLP identifying substitution on nucleotide 483 of the cpsK gene (Table 1) and by Sanger sequencing (Eurofins, Ebersberg, Germany) of cpsK PCR products. A fragment of the cpsK gene was amplified with forward primer cpsK-F 5′-GTTGCTGGTTATGATAGGGTAG-3′ (11), reverse primer cpsK-R2 5′-AAGCTTCTTTTGCTGTTTGCTC-3′ (this study), and Taq-based PCR PPP mastermix (TopBio, Prague, Czech Republic) in a total volume of 10 μl. Cycling conditions were as follows: initial denaturation at 95°C for 5 min, followed by 30 cycles of 95°C for 15 s, 53°C for 30 s, and 72°C for 30 s. The final extension was 72°C for 10 min. A 5-μl aliquot of the resultant 486-bp product was digested with 10 U of BstNI restriction endonuclease (NEB, Ipswich, MA) in a total volume of 20 μl. The resulting products were resolved on 2% agarose gel stained with SimplySafe (EURx Ltd., Gdansk, Poland). In the case of the presence of nucleotide G at nucleotide position 483 of the cpsK gene (serotypes 2 and 14), corresponding to nucleotide 139 of the PCR product, the PCR product could be cleaved by BstNI into two fragments of 139 and 347 bp, respectively (11). In the case of nucleotide C or T at nucleotide position 483 of the cpsK gene (serotypes 1/2 and 1), corresponding to nucleotide 139 of the PCR product (11), the PCR product could not be cleaved, and this resulted in a single 486-bp band (Table 1 and Fig. 1).

TABLE 1.

Schematic representation of the workflow

Step	Workflow
1	Streptococcus suis multiplex PCR for cps gene typing
2	Undifferentiated serotype 1/2 or 2		Undifferentiated serotype 1 or 14		Other serotypes detected by multiplex PCR
3	PCR-RFLP		PCR-RFLP
4	486-bp band	139-bp and 347-bp bands	486-bp band	139-bp and 347-bp bands
5	Serotype 1/2	Serotype 2	Serotype 1	Serotype 14

FIG 1.

Representative picture of cpsK PCR-RFLP patterns. M, DNA marker, with the number representing fragment length in bp. Samples are indicated as follows: A1, serotype 1/2, 483T; A2, serotype 1, 483T; A3, serotype 2, 483G; B3, serotype 14, 483G; D12, 13/18/Ji/1670 = serotype 1/2, 483C; NTC, no-template control.

RESULTS

Determination of serotype by whole-genome sequencing (WGS) of reference serotypes and 35 field isolates was in accordance with serological and multiplex PCR serotyping. Based on this knowledge, an additional 179 field isolates were tested just by serological and multiplex PCR serotyping. All field isolates identified as serotype 1/2 or 2 and 1 or 14 were further tested by the above-described PCR-RFLP and by Sanger sequencing of cpsK PCR product. Out of 214 field isolates from the Czech Republic, seventeen were of serotype 2 (multilocus sequence type ST1 and ST28), fourteen were of serotype 1/2 (mainly ST28 but also ST1 and ST7), six were of serotype 1 (ST1 and ST11) and two of serotype 14 (ST13 and ST17). WGS data and/or Sanger sequencing of cpsK PCR product confirmed nucleotide G at 483 position of the cpsK gene for serotype 2 and 14 field isolates, while serotype 1/2 and 1 field isolates had C or T at this position (Table 2). All PCR-RFLP results were in 100% accordance with the sequence data.

TABLE 2.

Description of strains examined

Isolate	Serotype detected bya :		cpsK sequenceb	Sequencing methodc	Date of isolation	Species	Age	Tissue	MLSTd
	Multi-PCR	PCR-RFLP
S1 reference strain	1 or 14	1	ggtggcctgTaataaac	WGS					13
S1/2 reference strain	2 or 1/2	1/2	ggtggcctgTaataaac	WGS					56
S2 reference strain	2 or 1/2	2	ggtggcctgGaataaac	WGS					1
S14 reference strain	1 or 14	14	ggtggcctgGaataaac	WGS					6
3/18/Ji/376	2 or 1/2	2	ggtggcctgGaataaac	WGS, S	16 January 2018	Pig	25 days	Lung	28
10/18/Ji/830	2 or 1/2	2	ggtggcctgGaataaac	WGS, S	26 January 2018	Pig		Lung	28
11/18/Ji/830	2 or 1/2	2	ggtggcctgGaataaac	WGS, S	26 January 2018	Pig		Spleen	28
13/18/Ji/1670	2 or 1/2	1/2	ggtggcctgCaataaac	WGS, S	7 February 2018	Pig	40 days	Lung	28
36/18/Ji/3895	2 or 1/2	1/2	ggtggcctgTaataaac	S	20 March 2018	Pig	14 days	Brain	28
55/18/Ji/6714	1 or 14	1	ggtggcctgTaataaac	S	9 May 2018	Pig	7 days	Brain	1
62/18/Ji/8065	2 or 1/2	1/2	ggtggcctgTaataaac	S	6 June 2018	Pig	28 days	Spleen	7
63/18/Ji/7649	2 or 1/2	1/2	ggtggcctgCaataaac	S	6 June 2018	Pig		Nasal swab	28
74/18/Ji/10270	1 or 14	1	ggtggcctgTaataaac	S	25 June 2018	Pig	10 days	Brain	11
75/18/Ji/10270	1 or 14	1	ggtggcctgTaataaac	S	22 June 2018	Pig	10 days	Joint	11
78/18/Ji/12432	2 or 1/2	2	ggtggcctgGaataaac	S	9 August 2018	Pig	75 days	Lung	1
94/18/Ji/13557	2 or 1/2	2	ggtggcctgGaataaac	S	5 September 2018	Pig		Lung	28
104/18/Ji/15350	2 or 1/2	1/2	ggtggcctgTaataaac	S	10 October 2018	Pig	14 days	Lymph node	1
105/18/Ji/15351	2 or 1/2	1/2	ggtggcctgCaataaac	S	10 October 2018	Pig	20 days	Pericardium	28
106/18/Ji/15351	2 or 1/2	1/2	ggtggcctgCaataaac	S	10 October 2018	Pig	20 days	Spleen	28
107/18/Ji/15351	2 or 1/2	1/2	ggtggcctgCaataaac	S	10 October 2018	Pig	20 days	Lung	28
119/18/Ji/16786	2 or 1/2	1/2	ggtggcctgCaataaac	S	8 November 2018	Pig	20 days	Joint	28
120/18/Ji/16789	2 or 1/2	1/2	ggtggcctgCaataaac	S	8 November 2018	Pig	20 days	Lung	28
124/18/Ji/18103	2 or 1/2	1/2	ggtggcctgTaataaac	S	5 December 2018	Pig	20 days	Nasal Swab	28
125/18/Ji/18566	2 or 1/2	2	ggtggcctgGaataaac	S	13 December 2018	Pig	Weaned piglet	Pericardium	1
126/18/Ji/18567	2 or 1/2	2	ggtggcctgGaataaac	S	13 December 2018	Pig	Weaned piglet	Brain	1
131/18/Ji/18762	1 or 14	1	ggtggcctgTaataaac	S	19 December 2018	Pig	20 days	Brain	1
132/18/Ji/18762	1 or 14	1	ggtggcctgTaataaac	S	19 December 2018	Pig	20 days	Lung	1
141/19/JI/1352	2 or 1/2	2	ggtggcctgGaataaac	S	22 January 2019	Pig	68 days	Lymph node	1
142/19/JI/1353	2 or 1/2	2	ggtggcctgGaataaac	S	22 January 2019	Dog		Brain	1
143/19/JI/1352	2 or 1/2	2	ggtggcctgGaataaac	S	22 January 2019	Pig	68 days	Brain	1
144/19/JI/1625	1 or 14	14	ggtggcctgGaataaac	S	24 January 2019	Pig	Fattening pig	Heart	13
145/19/JI/1634	1 or 14	14	ggtggcctgGaataaac	S	25 January 2019	Pig	Weaned piglet	Brain	17
149/19/JI/2993	1 or 14	1	ggtggcctgTaataaac	S	29 January 2019	Pig	19 days	Lymph node	1
156/19/JI/3589	2 or 1/2	2	ggtggcctgGaataaac	S	19 February 2019	Pig	8 weeks	Lung	28
159/19/JI/3959	2 or 1/2	2	ggtggcctgGaataaac	S	28 February2019	Pig		Lung	28
163/19/JI/4739	2 or 1/2	2	ggtggcctgGaataaac	S	11 March 2019	Pig	Fattening pig	Liver	28
187/19/JI/10440	2 or 1/2	2	ggtggcctgGaataaac	S	19 June 2019	Pig	Fattening pig	Brain	1
188/19/JI/10440	2 or 1/2	2	ggtggcctgGaataaac	S	21 June 2019	Pig	Fattening pig	Intestine	1
189/19/JI/10440	2 or 1/2	2	ggtggcctgGaataaac	S	24 June 2019	Pig	Fattening pig	Lymph node	1
203/19/JI/11813	2 or 1/2	1/2	ggtggcctgCaataaac	S	18 July 2019	Pig	Weaned piglet	Lung	28
211/19/JI/14336	2 or 1/2	1/2	ggtggcctgTaataaac	S	28 August 2019	Chicken	1 day	Yolk sac	28
213/19/JI/14852	2 or 1/2	1/2	ggtggcctgTaataaac	S	4 September 2019	Pig	40 days	Heart	28
214/19/JI/14881	2 or 1/2	2	ggtggcctgGaataaac	S	4 September 2019	Pig	55 days	Lung	28

^aMulti-PCR, multiplex PCR for detecting capsular polysaccharide synthesis genes (10).

^bNucleotide position 483 of the cpsK gene is indicated in uppercase, and the BstNI recognition site is underlined.

^cWGS, whole-genome sequence; S, Sanger sequence of cpsK PCR product.

^dMLST, multilocus sequence type.

Digestion of the cpsK PCR product of reference strains and isolate number 13/18/Ji/1670 with BstNI and subsequent electrophoresis on agarose gel led to the presence of 486-bp undigested PCR product in the case of serotypes 1/2 and 1 (samples were as follows: A1, serotype 1/2, 483T; A2, serotype 1, 483T; D12, 13/18/Ji/1670 = serotype 1/2, 483C) and two bands of 139 bp and 347 bp in the case of serotype 2 and 14 (samples were as follows: A3, serotype 2, 483G; B3, serotype 14, 483G) (Fig. 1).

DISCUSSION

Identification of S. suis serotypes by serological methods is still considered the standard, but the development of multiplex PCR typing methods detecting particular genes in the cps locus of the known S. suis and S. suis-like serotypes made serotyping more convenient and enabled resolution of dubious results occasionally obtained by serological typing (12). Unfortunately, serotypes 1/2 and 2 share highly similar genes in the cps cluster. The same applies to serotypes 1 and 14, and thus detection of the serotype-specific cps gene cannot distinguish these serotypes (5, 9). However, the discovery of a causal 483G→C/T mutation in the cpsK gene (11) enables clear identification of these serotypes.

Detection of cpsK 483G→C/T gene polymorphism is possible by Sanger sequencing of cpsK PCR product or by whole-genome sequencing. Although this approach is accurate, it is time consuming and probably not available for all diagnostic laboratories. Recently, a method based on high-resolution melting analysis of cpsK PCR product was reported (13). This method could be quick and reliable in laboratories equipped with a quantitative PCR (qPCR) apparatus. Here, we describe a PCR-RFLP-based method detecting cpsK 483G→C/T substitution by BstNI restriction endonuclease. The presence of one 486-bp band, in the case of serotypes 1/2 and 1, or of two bands of 139 bp and 347 bp, in the case of serotype 2 and 14, can easily be resolved by agarose gel electrophoresis. PCR-RFLP results were 100% consistent with WGS and/or Sanger sequences of the tested 43 strains of serotype 1/2, 1, 2, and 14. As additional information in this report, although we found no exclusive association between a particular serotype and sequence type (ST), some ST data are interesting. As expected for European strains, potentially zoonotic ST1 serotype 2 strains were identified (3). However, the finding of serotype 1/2 ST1 or ST7 strains was unexpected. A serotype 2 ST7 strain was responsible for the deadly outbreak in humans in China in 2005 (21). More recently, serotype 14 ST7 strains were also described in China (22). This is the first report on serotype 1/2 ST7 strains, and the zoonotic potential of such strains should be further studied.

In conclusion, the PCR-RFLP method described here is relatively inexpensive, rapid, and simple but still highly reliable, and it requires just basic laboratory equipment.