Optimized protocol for multiplex nested polymerase chain reaction to detect and differentiate Haemophilus parasuis, Streptococcus suis, and Mycoplasma hyorhinis in formalin-fixed, paraffin-embedded tissues from pigs with polyserositis

Abstract

An optimized protocol was developed for the simultaneous detection and differentiation of Haemophilus parasuis, Streptococcus suis, and Mycoplasma hyorhinis in formalin-fixed, paraffin-embedded (FFPE) tissues with multiplex nested polymerase chain reaction (PCR). This method also determines the prevalence of these bacteria in pigs with polyserositis. DNA extraction with a combination of a commercial reagent and proteinase K resulted in more frequent detection of the pathogens than DNA extraction with proteinase K alone. Among FFPE tissue samples from 312 cases of polyserositis in which at least 1 bacterial species was detected, multiplex nested PCR detected H. parasuis in 239 (77%), S. suis in 124 (40%), and M. hyorhinis in 40 (13%). The disease was caused by a single pathogen in 224 (72%) of the cases and multiple pathogens in 88 (28%). Among the pigs positive for H. parasuis, S. suis, and M. hyorhinis by multiplex nested PCR, the pathogen was isolated from only 11%, 35%, and 28%, respectively. Therefore, the PCR protocol developed in this study is a useful diagnostic method when samples are negative after isolation methods and even for samples in which only 1 pathogen was isolated.

Introduction

Fibrinous or fibrinopurulent polyserositis is an important source of production losses in the swine industry and a common cause of histopathological lesions in pigs (1). Most polyserositis cases result from infection by Haemophilus parasuis, Streptococcus suis, and Mycoplasma hyorhinis (2–4). Isolation of these organisms is difficult because of fastidious growth (5) and/or antibiotic treatment of the sick pig before laboratory diagnosis. Nonetheless, identification of the causative agent remains a critical step in choosing effective treatment and vaccination schemes for disease control even if the causative agent cannot be isolated.

Submission of formalin-fixed tissues to diagnostic laboratories, rather than submission of live sick pigs, is often preferred owing to the inconvenience and difficulty of live pig delivery. Therefore, a diagnostic method is needed to detect and differentiate the 3 likely pathogens in formalin-fixed tissues. Polymerase chain reaction (PCR) meets this requirement by amplifying pathogen-specific DNA from formalin-fixed, paraffin-embedded (FFPE) tissues, which leads to the identification of etiologic agents without isolation.

The objective of this study was to develop an optimized protocol for the simultaneous detection and differentiation of H. parasuis, S. suis, and M. hyorhinis in FFPE tissues by multiplex nested PCR and to determine the prevalence of these 3 organisms in polyserositis cases.

Materials and methods

Animals and samples

Tissue samples were obtained at necropsy from pigs submitted to the Department of Veterinary Pathology of Seoul National University’s College of Veterinary Medicine between January 2000 and December 2008. After 1 or 2 days’ fixation in 10% neutral buffered formalin, the samples were dehydrated through graded alcohols with a xylene step and embedded in paraffin wax for histopathological examination of sections 4-μm thick that were stained with hematoxylin and eosin. Samples from 384 pigs were selected for further study on the basis of the presence of polyserositis and microscopic findings of lesions such as fibrinous or fibrinopurulent pericarditis, peritonitis, pleuritis, perihepatitis, and perisplenitis.

Five colostrum-deprived pigs aged 35 d were used as sources of negative-control tissues. Lung sections from pigs naturally infected with Pasteurella multocida (strain SNUVP 4512) capsular type A (6) as well as pigs experimentally infected with Actinobacillus pleuropneumoniae serotypes 1 to 12 (7) or M. hyopneumoniae (8) were used as further control material. The lung sections were shown by method A (described in the next section) to be free of H. parasuis, S. suis, and M. hyorhinis.

DNA extraction

For each sample, a section 10-μm thick was prepared from FFPE tissue blocks showing severe fibrinous lesion, and excess paraffin was trimmed. The sections were placed in 1.5-mL sterile microcentrifuge tubes. The microtome blade, tweezers, and other equipment that could come into contact with the samples were carefully sterilized before each tissue block was processed.

Two procedures were used to extract DNA from the FFPE tissue. In method A, the tissue sections were deparaffinized with toluene for 10 min, then washed twice with 100% ethanol to remove the solvent. The ethanol was evaporated under a vacuum for 10 min. To isolate the genomic DNA 500 μL of digestion buffer [10 mM Tris (pH 8.5), 1 mM ethylene diamine tetraacetic acid, and 0.5% Tween-20] containing 200 μg/mL of proteinase K was added to the dried samples. The resuspended tissues were incubated overnight at 56°C. Next the proteinase K was inactivated at 100°C for 8 min. The genomic DNA was extracted with the use of a commercial reagent (Trisol LS, Gibco BRL, Grand Island, New York, USA) and then precipitated in 50% ethanol. The final ethanol-washed DNA pellet was air-dried and then dissolved in 30 μL of diethylpyrocarbonate-treated water. In method B, the tissue sections were deparaffinized with toluene for 10 min, then washed twice with 100% ethanol to remove the solvent. The ethanol was evaporated under a vacuum for 10 min. To isolate the genomic DNA 500 μL of digestion buffer [10 mM Tris (pH 8.5), 1 mM ethylene diamine tetraacetic acid, and 0.5% Tween-20] containing 200 μg/mL of proteinase K was added to the dried samples. The resuspended tissues were incubated overnight at 56°C. Next the proteinase K was inactivated at 100°C for 8 min. The DNA was then extracted from the digested samples with the standard phenol–chloroform– isoamyl alcohol procedure and precipitated in ethanol (9).

Primers

The PCR primers were designed on the basis of multiple sequence alignments of the 16S small subunit ribosomal RNA genes from H. parasuis and S. suis and the operon sequences encoding the p37, p29, and p69 proteins of M. hyorhinis with MegaAlign (DNAStar, Madison, Wisconsin, USA) and Oligo 4.0 (Molecular Biology Insights, Cascade, Colorado, USA) software programs (Table I). This operon is believed to operate analogously as extracytoplasmic binding lipoprotein (10). The primers were determined by BLAST 2.2.22+ (http://blast.ncbi.nlm.nih.gov/Blast.cgi) of the National Center for Biotechnology Information (NCBI), US National Library of Medicine, to be highly specific for H. parasuis, S. suis, and M. hyorhinis. No other sequences of Haemophilus, Streptococcus, and mycoplasmas completely matched the designed primers.

Table I.

Primers for amplifying DNA of Haemophilus parasuis, Streptococcus suis, and Mycoplasma hyorhinis by polymerase chain reaction (PCR)

Bacterium	Primer	Sequences (5′–3′)	Location	Size (bp)
H. parasuis	Outer	F-GTGATGAGGAAGGGTGGTGT	393–412	316
		R-AGCGTCAGTATTTTCCCAAG	689–708
	Inner	F-AGAACATTACATTGACGTTAGTC	419–441	197
		R-TAAAATACTCTAGCAACCCAGTA	592–615
S. suis	Outer	F-AACGCTGAAGTCTGGTGCTT	38–57	228
		R-TGTATCGATGCCTTGGTGAG	246–265
	Inner	F-CTTGCACTAGACGGATGA	55–72	114
		R-TGCGGTAAATACTGTTATGC	149–168
M. hyorhinis	Outer	F-CATATGGCCCACTTTTAGGG	1878–1897	366
		R-GGGATTGAAGTGGTTGTCTG	2226–2245
	Inner	F-TGACAATTTCCAAAAAGAGA	1918–1937	236
		R-AAAAAGGATTGTTCCTTCAA	2134–2153

bp — base pairs; F — forward; R — reverse.

Polymerase chain reaction

For outer PCR 10 μL of extracted DNA was used as the template, and for nested PCR 10 μL of the amplified product was used. For outer PCR the amplification was done in a 50-μL reaction mixture containing 1.5 mM MgCl₂, 1× PCR buffer (Perkin Elmer, Foster City, California, USA), 0.2 mM of each deoxynucleotide triphosphate, 200 nM of each primer, and 2.5 U of Taq DNA polymerase (Perkin Elmer). Both reactions were run in a thermocycler (GeneAmp PCR System 9600, PerkinElmer/Cetus, Norwalk, Connecticut, USA) under the following conditions: 45 cycles of denaturation at 95°C for 30 s, primer annealing at 57°C for 45 s, and extension at 68°C for 30 s. The reaction was stopped with a final extension at 68°C for 2 min. For nested PCR the amplification was done in 50 μL of the same reaction mixture as used for the outer PCR. Both reactions were run in the same thermocycler under the following conditions: 40 cycles of denaturation at 95°C for 30 s, primer annealing at 55°C for 45 s, and extension at 68°C for 30 s. The reaction was stopped with a final extension at 68°C for 2 min.

The amplified products were visualized by standard gel electrophoresis of 10 μL of the final reaction mixture on 2% agarose gel. Amplified DNA fragments of specific sizes were located by ultraviolet fluorescence after staining with ethidium bromide. Their lengths were verified by a digested lambda DNA standard run simultaneously.

The PCRs were done in triplicate. Control DNA from reference strains of H. parasuis (SNUVP 27032), S. suis (SNUVP 650098), and M. hyorhinis (SNUVP 49217) were included in each reaction.

DNA sequencing of PCR products

The PCR products were purified by means of a 30-kD cutoff membrane ultrafiltration filter. The nucleotide sequences of the purified products were determined by BigDye chemistry with the ABI Prism Sequencer (Applied Biosystems, Foster City, California, USA).

Specificity and sensitivity assays

To determine the specificity of the outer and nested PCRs, both sets of primers were used to test independently A. pleuropneumoniae (strain SNUVP 231), A. suis (SNUVP 8723), P. multocida (SNUVP 4512), Escherichia coli (SNUVP 9835), Bordetella bronchiseptica (SNUVP 2395), Salmonella ser. Typhimurium (SNUVP 3329), Salmonella ser. Choleraesuis (SNUVP 9843), Streptococcus porcinus (SNUVP 7219), M. hyopneumoniae (ATCC 25934 and 3 field strains), M. hyosynoviae (ATCC 27095), M. arginini (ATCC 23243), and M. flocculare (ATCC 27716). The 3 field strains of M. hyopneumoniae were isolated from postweaning pigs with severe swine enzootic pneumonia.

To determine the sensitivity of the outer and nested PCRs, 10-fold serial dilutions of extracted genomic DNA of H. parasuis, S. suis, and M. hyorhinis were measured spectrophotometrically at 260 nm. Five H. parasuis strains (serotypes 2, 4, 5, and 13, and nontypeable) isolated from postweaning pigs with pericarditis and eight S. suis strains [serotypes 2, 3, 4, 8, 16, 22, and 33, and nontypeable (11)] were used to evaluate specificity and sensitivity. Serotyping of the H. parasuis was done by Dr. Conny Turni, Animal Research Institute, Brisbane, Australia.

Culture and isolation

The studied pathogens were isolated from 384 pigs (124 live and 260 dead). Most of the dead pigs had been submitted to the Department of Veterinary Pathology within 24 h after death. For isolation of H. parasuis, specimens were cultured on chocolate agar and blood agar plates with a nurse strain of Staphylococcus aureus. After 48 h of incubation in 5% CO₂ at 37°C, colonies were removed for biochemical identification as previous described (12).

For isolation of S. suis, swabs were cultured on sheep blood agar at 37°C in 5% CO₂ for 24 to 48 h. Colonies of catalase-negative, gram-positive cocci exhibiting α-hemolysis were subcultured onto sheep blood agar for 24 h in 5% CO₂ at 37°C. All isolates were tested conventionally for arginine hydrolysis, production of acetoin (by the Voges–Proskauer test), and production of acid from various carbohydrates (inulin, salicin, trehalose, lactose, sucrose, sorbitol, mannitol, and glycerol). The strains were also tested for their ability to grow in the presence of 6.5% NaCl.

For isolation of M. hyorhinis, all tissues with serositic lesions were ground in 5 mL of BHL broth medium (13,14) and Friis medium. The suspension was centrifuged at 15 000 ×g for 25 min and the sediment resuspended in 5 mL of BHL broth medium (or Friis medium) and filtered through a 0.45-μm filter. The filtrate was then incubated at 37°C until the phenol red indicator became a distinctive yellow (pH approximately 6.8) or for up to 30 d. Isolates were confirmed as being M. hyorhinis by previously described PCR methods (15).

Statistical analysis

Chi-squared tests were used to assess the prevalence of H. parasuis, S. suis, and M. hyorhinis in preweaning, postweaning, and growing pigs. Statistical analysis was conducted with SPSS software (SPSS, Chicago, Illinois, USA). Results were considered statistically significant when P < 0.05.

Results

The results obtained with multiplex outer and nested PCR were consistent in at least 3 independent repetitions. The samples that were positive by multiplex outer PCR were always positive by multiplex nested PCR, regardless of the DNA extraction method. Multiplex nested PCR (Figure 1) detected H. parasuis, S. suis, and M. hyorhinis more frequently than multiplex outer PCR, regardless of the DNA extraction method, and DNA extraction by method A detected the 3 pathogens more frequently than DNA extraction by method B with both types of PCR (Table II).

Figure 1.

Results of agarose gel electrophoresis of products of DNA extracted from formalin-fixed, paraffin-embedded (FFPE) tissue samples collected from 348 pigs with polyserositis and subjected to multiplex nested polymerase chain reaction. Left to right: M — 100-base pair DNA ladder; lane 1 — negative control; lane 2 — positive control of Haemophilus parasuis; lane 3 — positive control of Streptococcus suis; lane 4 — positive control of Mycoplasma hyorhinis; lane 5 — FFPE sample positive for H. parasuis and S. suis; lane 6 — FFPE sample positive for H. parasuis and M. hyorhinis; lane 7 — FFPE sample positive for S. suis and M. hyorhinis; lane 8 — FFPE sample positive for H. parasuis, S. suis, and M. hyorhinis.

Table II.

Frequency of successful multiplex PCR with 2 methods of DNA extraction from formalin-fixed, paraffin-embedded tissue samples and of isolation of the 3 organisms from 348 pigs with polyserositis

Organism	Multiplex PCR; Number of pigs				Isolation; Number of pigs
	Method Aa		Method Bb
	Outer	Nested	Outer	Nested
H. parasuis	82	239	69	220	27
S. suis	34	124	23	114	43
M. hyorhinis	14	40	11	38	11

^aDNA was extracted with a commercial reagent and proteinase K.

^bDNA was extracted with proteinase K alone.

Single nested PCR detected 0.1 pg of H. parasuis DNA, 1 pg of S. suis DNA, and 4.8 pg of M. hyorhinis DNA. Single outer PCR detected 100 pg of DNA from each of the 3 pathogens. Neither the outer nor the inner primers cross-reacted with any of the bacteria tested in this study. The multiplex nested PCR products of H. parasuis, S. suis, and M. hyorhinis were identical to the corresponding sequences in NCBI’s GenBank database (EF396300.1, AB071348.1, and X14140.1).

Among the 384 samples tested during the study, at least 1 bacterial species was detected in 312. Multiplex nested PCR detected H. parasuis in 239 (77%) of the 312, S. suis in 124 (40%), and M. hyorhinis in 40 (13%).

Among the 239 pigs positive for H. parasuis by multiplex nested PCR, the organism was isolated from 27 (11%), 22 of which tested positive only for H. parasuis by multiplex nested PCR; the other 5 pigs tested positive for both H. parasuis and S. suis by multiplex nested PCR. Conversely, H. parasuis was not isolated from samples negative for H. parasuis by multiplex nested PCR (Table II).

Among the 124 pigs positive for S. suis by multiplex nested PCR, the organism was isolated from 43 (35%), 36 of which tested positive only for S. suis by multiplex nested PCR; the other 7 pigs tested positive for both S. suis and H. parasuis. Conversely, S. suis was not isolated from samples negative for S. suis by multiplex nested PCR (Table II).

Among the 40 pigs positive for M. hyorhinis by multiplex nested PCR, the organism was isolated from 11 (28%), 8 of which tested positive for both M. hyorhinis and H. parasuis by multiplex nested PCR; the other 3 tested positive for both M. hyorhinis and S. suis by multiplex nested PCR. Conversely, M. hyorhinis was not isolated from samples negative for M. hyorhinis by multiplex nested PCR (Table II).

A single bacterial species was detected by multiplex nested PCR in 224 cases (Table III), and multiple species were detected in the remaining 88 cases (Table IV). Of the 312 pigs 58 were preweaning (1 to 4 wk old), 169 were postweaning (5 to 10 wk old), and 85 were growing (11 to 15 wk old). Infection with H. parasuis was more prevalent in the postweaning (P = 0.029) and growing (P < 0.001) pigs than in the preweaning pigs. In contrast, S. suis infection was more prevalent in the preweaning pigs than in the postweaning (P = 0.029) and growing (P < 0.001) pigs.

Table III.

Prevalence of single infection with 1 of the 3 organisms in the same group of pigs as determined by multiplex nested PCR

		Organism
Animal group	Number of pigs	H. parasuis	S. suis	M. hyorhinis
Preweaning	46	25	21	0
Postweaning	122	88	34	0
Growing	56	48	6	2
Total	224	161	61	2

Table IV.

Prevalence of multiple infection with 2 or 3 of the organisms in the same group of pigs as determined by multiplex nested PCR

		Organisms
Animal group	Number of pigs	H. parasuis and S. suis	H. parasuis and M. hyorhinis	S. suis and M. hyorhinis	H. parasuis, S. suis, and M. hyorhinis
Preweaning	12	10	1	1	0
Postweaning	47	26	14	7	0
Growing	29	14	10	2	3
Total	88	50	25	10	3

Discussion

The results of this study demonstrate that H. parasuis, S. suis, and M. hyorhinis DNA can be detected and differentiated in FFPE tissues. Among the 312 cases of polyserositis in which at least 1 bacterial species was detected, multiplex nested PCR detected H. parasuis in 239 (77%), S. suis in 124 (40%), and M. hyorhinis in 40 (13%). Although the PCR technique might be more sensitive in detecting the 3 bacterial pathogens if fresh tissue samples rather than FFPE tissue samples were used, most Korean swine practitioners prefer to ship formalin-fixed tissues for diagnosis to avoid maintenance of cool temperature during transportation.

In this study the 3 organisms were detected more frequently by PCR than by culture and isolation. Although PCR detection would avoid antimicrobial susceptibility testing, under field conditions animals showing clinical signs suggestive of H. parasuis and S. suis infection are generally treated with antibiotics, which makes H. parasuis isolation difficult; attempts are usually unsuccessful. Isolation of these organisms is also highly dependent on the skill of the testers, the medium used, the tissue taken, and other factors. Although it is ideal to identify the organisms by PCR and isolation with the use of fresh tissues, PCR with FFPE samples is an alternative when fresh tissue is not available or retrospective studies are conducted to correlate histologic lesions with clinical signs.

Since H. parasuis, S. suis, and M. hyorhinis are commonly isolated from normal healthy pigs (1,16,17), detection of this organism by PCR only may not provide a definite diagnosis of polyserositis. Therefore, PCR results must be interpreted in conjunction with histopathologic findings: characteristic lesions such as fibrinous or fibrinopurulent inflammation would be expected. Alternatively, in situ hybridization may be used to avoid misinterpretation of PCR results (18). In situ hybridization provides results within the histologic architecture, so detection of causative organisms and histopathologic evaluation may be done simultaneously in the same tissue section. However, the greater technical complexity of in situ hybridization largely restricts the technique to diagnostic laboratories.

The ability to simultaneously amplify and differentiate DNA of these organisms from FFPE tissues by PCR has a profound impact on diagnostic pathology, especially when the causative organisms are difficult to isolate and the histopathological lesions are similar. For long-term preservation FFPE tissue preparation is the standard method, and FFPE samples are often the only samples available in pathology laboratory archives, as fresh tissue and serum from suspected cases are rarely saved for extended periods. Formaldehyde, the effective component of formalin, induces cross-links between proteins and DNA or RNA (19), hindering molecular amplification. Furthermore, DNA degrades in archival FFPE tissue, which results in significantly reduced amplification at high fragment lengths (20). Hence, relatively short amplified fragments of DNA (approximately 100 to 300 base pairs) may be useful in detecting bacterial pathogens in archival FFPE tissue.

The multiplex PCR developed in this study can detect different serotypes of H. parasuis and S. suis as well as the operon encoding the p37, p27, and p69 proteins of M. hyorhinis. This operon is a well-conserved region, and the amino acid sequences are 99% to 100% homologous (GenBank accession nos. X14140.1, CP002669.1, M37339.1, and CP002170.1). Hence, this method is able to detect genotypically diverse populations of H. parasuis, S. suis, and M. hyorhinis in swine.

When DNA of several microbial species is present in a sample, the sensitivity of multiplex PCR usually drops, and the DNA of the less prevalent species can even be missed. A major disadvantage of multiplex PCR is a lack of differentiation among the serovars or serotypes of H. parasuis and S. suis. For example, at least 15 serovars of H. parasuis (based on immunodiffusion) are recognized (21), and at least 35 capsular types of S. suis have been identified (22,23). Current bacterins for H. parasuis and S. suis provide only serovarand serotype-specific protection (24 and M.G., unpublished observations, 2010). Therefore, serotype identification is a critical component of strategies for the control of H. parasuis and S. suis infection. In addition, very few laboratories in the world are able to serotype H. parasuis isolates. Further study is needed to develop PCR protocols for the differentiation of serovars or serotypes of H. parasuis and S. suis in FFPE tissues.