Comparison of reverse transcription polymerase chain reaction, immunohistochemistry, and in situ hybridization for the detection of porcine epidemic diarrhea virus in pigs

Abstract

Reverse transcription polymerase chain reaction (RT-PCR), immunohistochemistry, and in situ hybridization were compared for the detection of porcine epidemic diarrhea virus (PEDV). Fifteen piglets experimentally infected with PEDV were used in the study. In addition, 94 diarrheic piglets submitted to the Department of Veterinary Pathology in Seoul National University for diagnosis of PEDV infection were used to compare the 3 methods. Antigen and nucleic acid of PEDV were detected in 15/15, 13/15, and 14/15 of the intestinal and fecal samples from the PEDV-inoculated pigs by RT-PCR, immunohistochemistry, and in situ hybridization, respectively. The virus was isolated from 15/15 of the jejunal samples from the PEDV-inoculated pigs. Neither PEDV antigen nor PEDV nucleic acid was detected in the intestinal and fecal samples from mock-infected control pigs. Of the 94 samples, 63 were positive for PEDV by all 3 techniques. Six samples were positive for PEDV by immunohistochemistry and in situ hybridization. Three samples were positive for PEDV by in situ hybridization and RT-PCR. Seven samples were positive for PEDV by RT-PCR. Although RT-PCR identified the presence of PEDV more frequently than the other methods, when only formalin-fixed tissues are submitted, immunohistochemistry and in situ hybridization would be useful methods for the detection of PEDV antigen and nucleic acid.

Porcine epidemic diarrhea virus (PEDV) is a member of the genus Coronavirus, family Coronaviridae, order Nidovirales (1). PEDV causes a highly contagious enteric infection in swine (2). Clinical signs of PEDV infection include anorexia, vomiting, diarrhea, and dehydration. Morbidity and mortality in infected neonatal piglets less than 5 d old approaches 100% because of severe diarrhea and dehydration. However, mortality in infected piglets older than 10 d is less than 10% (3,4,5).

Rapid, reliable methods for the diagnosis of PEDV in swine are essential for veterinary diagnostic laboratories serving the swine industry in Korea, as PEDV infection is highly prevalent (6). Currently, diagnosis is based on characteristic clinical signs, gross and microscopic lesions, and results of fluorescent antibody (FA) examination of frozen tissue sections. Each diagnostic method has certain disadvantages. Clinical signs and gross lesions may be atypical. The sensitivity and specificity of histopathologic diagnosis has been limited in many cases because several enteric pathogens, such as transmissible gastroenteritis virus, rotavirus, Escherichia coli, and Isospora suis, cause lesions similar to those caused by PEDV (7,8).

Nucleic acid amplification techniques have the potential to revolutionize the diagnosis of infectious diseases and are already being used in diagnostic pathology. New approaches to the detection of PEDV are based on the detection of PEDV-specific nucleic acid by reverse transcription polymerase chain reaction (RT-PCR) and in situ hybridization (9,10). Because of the difficulty of isolating PEDV (11,12), these techniques promise to dramatically decrease the time required to diagnose PEDV-associated enteritis. However, studies that systematically evaluate the diagnostic reliability of these methods have not yet been reported. The objective of this study was to compare RT-PCR, immunohistochemistry, and in situ hybridization for the detection of PEDV in naturally and experimentally infected pigs.

A total of 94 diarrheic piglets, 3 to 10 d of age, submitted to the Department of Veterinary Pathology of Seoul National University from January 1998 to December 2000 were used in this study. Of these, 23 had been used in a previous study (6). All piglets were submitted alive. Intestinal and fecal samples were collected from each piglet at necropsy. Ribonucleic acid (RNA) was extracted from intestinal and fecal samples immediately. Intestinal samples were fixed in formalin.

Fifteen 3-day-old colostrum-deprived pigs were inoculated orally with 3 mL of tissue culture fluid containing 1 × 10⁴ tissue culture infective doses 50% (TCID₅₀)/mL of cell-adapted PEDV, strain 971496 (12). Ten control pigs were exposed in the same manner to uninoculated cell culture supernatant. Three PEDV-inoculated and 2 control pigs were euthanized at 24, 48, 60, 72, and 84 h postinoculation. Intestinal and fecal samples were collected from each pig at necropsy. Jejunal tissue from each infected and control pig was used for virus isolation, as described previously (3). All methods were approved by the Institutional Animal Care and Use Committee of Seoul National University.

Suspensions of fecal samples were prepared by vortexing 2 g of feces with 2 mL of 0.1 M phosphate-buffered saline (PBS; pH 7.2). The suspensions were clarified at 9000 × g for 10 min at 4°C. Five hundred μL of the supernatant was mixed with 500 μL of Trizol LS reagent (Gibco BRL, Grand Island, New York, USA) for 10 min, and then 300 μL of chloroform was added for an additional 10 min at 4°C. After centrifugation at 4000 × g for 15 min, isopropanol precipitation was carried out overnight at 4°C. The precipitated RNA was pelleted by centrifugation for 10 min at 14 000 × g, washed twice with 70% ethanol, and then dissolved in 30 μL of diethylpyrocarbonate-treated water.

Primers spanning the membrane protein sequence (GenBank accession number Z24733) of PEDV were used in this study, as described previously (10). The forward (PEDVF01) and reverse (PEDVR02) primers were 5′-GGGCGCCTGTATAGAGTTTA-3′(nucleotides 927 to 946) and 5′-AGACCACCAAGAATGTGTCC-3′ (nucleotides 1319 to 1338), respectively.

For RT-PCR, the commercial one-step RT-PCR system kit (Promega Biotech, Madison, Wisconsin, USA) was used. Two μL of extracted RNA was added to each mixture containing a final concentration of 1 mM of magnesium sulfate, 5 μL of AMV/Tfl 5X reaction buffer, 0.2 mM of deoxynucleotide triphosphate (dNTP), 2.5 U/μL of avian myeloblastosis virus reverse transcriptase, 2.5 U/μL of Tfl DNA polymerase, 0.25 μM of each primer, and 11.5 μL of sterile nuclease-free water, for a total volume of 25 μL per sample. All reagents and mixtures were kept on ice and then placed in the thermal cycler for 1 cycle at 42°C for 45 min and then 94°C for 2 min. The amplifications were then performed for 40 cycles at 94°C for 30 s, 55°C for 30 s, and 72°C for 30 s, followed by a 10-minute extension period at 72°C. Positive and negative controls were used with each RT-PCR reaction.

Tissues were processed for histologic examination by routine methods, and 4-μm sections were placed on poly-L-lysine-coated glass microscope slides. Tissues were deparaffinized with xylene, rehydrated with graded alcohols, and then air dried. Endogenous peroxidase was quenched with absolute methanol containing 1% hydrogen peroxide for 1 h. All slides were then treated with 0.01% protease XXIV (Sigma Chemical Company, St. Louis, Missouri, USA) in PBS (0.1 M, pH 7.4) for 2 min at room temperature. All slides were incubated with 10% normal horse serum (NHS) (Vector Laboratories, Burlingame, California, USA) and 3% bovine serum albumin mixture (ratio 1:20) in PBS (0.1 M, pH 7.4) for 1 h at room temperature to saturate non-specific protein-binding sites. C9-9-9 monoclonal antibody, which reacted with the 30-kD membrane protein of PEDV by Western blotting analysis (13), was diluted in PBS (0.1 M, pH 7.4) containing 2% NHS and 0.3% Triton-X100 (Sigma Chemical Company) at 1:100. The slides were incubated overnight at 4°C in a humidity chamber.

Following 3 washes with PBS (0.1 M, pH 7.4), sections were flooded and incubated for 1 to 2 h at room temperature with biotinylated horse anti-mouse IgG (Vector Laboratories) diluted 1:400 in PBS (0.1 M, pH 7.4) containing 2% NHS and 0.3% Triton-X100. The slides were washed in PBS (0.1 M, pH 7.4), followed by incubation for 30 min in avidin-biotin complex (ABC) solution (Vector Laboratories), prepared according to the manufacturer's instructions. After washing in PBS (0.1 M, pH 7.4), the final reaction product was produced by immersing the sections in a solution of 0.01% hydrogen peroxide and 0.05% 3,3′-diaminobenzidine tetrahydrochloride (Vector Laboratories) in PBS (0.1 M, pH 7.4) for 10 min. The sections were lightly counterstained with Mayer's hematoxylin, dehydrated with graded concentrations of ethanol and xylene, and mounted.

A 412-bp cDNA probe for the viral RNA encoding the membrane protein of PEDV was generated by RT-PCR (10). The RT-PCR products were purified (Wizard PCR Preps; Promega Biotech) and labeled by random priming with digoxigenin-dUTP with a commercial kit (Boehringer Mannheim, Indianapolis, Indiana, USA), according to the manufacturer's instructions.

Sections were dewaxed in xylene and rehydrated in PBS (0.01 M, pH 7.4) for 5 min. De-proteinization was carried out in 0.2 N HCl for 20 min at room temperature. Tissues were then digested at 37°C for 20 min in 100 μg/mL of proteinase K (Gibco BRL) in PBS (pH 7.4, 0.01 M) and fixed in 4% paraformaldehyde in PBS for 10 min. After 2 rinses with PBS, the slides were acetylated in 300 mL of 0.1 mM triethanolamine-HCl buffer (pH 8.0) to which 0.75 mL of acetic anhydride (0.25%) had been added. After 5 min, an additional 0.75 mL of acetic anhydride was added, and 5 min later the slides were rinsed in 2X saline sodium citrate (SSC) (1X SSC: 50 mM NaCl and 15 mM sodium citrate, pH 7.0). The slides were allowed to equilibrate for 60 min in a standard hybridization buffer that consisted of 5X SSC with 50% deionized formamide, 10X 2% buffered blocking solution (Boehringer Mannheim), 0.1% N-laurylsarcosine, and 0.02% sodium dodecyl sulfate.

Hybridization was done overnight at 45°C. The digoxigenin-labeled probe (0.1 ng/μL) was diluted in 300 μL of the standard hybridization buffer, heated for 10 min at 95°C on a heating block, and quenched on ice before being applied to the tissue sections. Approximately 50 ng of the digoxigenin-labeled probe was added to the standard hybridization buffer (50 μL), which was then layered over the section. Fluid was held in place by a coverslip, and the edges were sealed with rubber cement. After overnight hybridization, the sections were thoroughly washed twice in 4X SSC for 5 min at room temperature, twice in 2X SSC for 10 min at 37°C, twice in 0.2X SSC for 5 min at room temperature, and once in maleic acid buffer (100 mM maleic acid and 150 mM NaCl, pH 7.5) for 5 min at room temperature.

For detection of hybridization, sections were incubated with anti-digoxigenin conjugated with alkaline phosphatase (Boehringer Mannheim) diluted 1:250 in 0.1 M Tris HCl (pH 7.4), 0.15 M NaCl with 1% blocking reagent (Boehringer Mannheim). After 3 washes in buffer, substrate consisting of nitroblue tetrazolium and 5-bromocresyl-3-indolylphosphate was layered over the sections. Color was allowed to develop for 5 to 8 h in the dark, and the reaction was stopped by dipping the slides briefly in triethylenediaminetetraacetic acid (EDTA) buffer (10 mM Tris HCl and 1 mM EDTA, pH 8.0). The sections were counterstained with 0.5% methyl green, and the slides were then washed with distilled water for 1 min, allowed to dry completely, dipped into the absolute xylene, and coverslipped with Canada balsam mounting medium (Hayashi Pure Chemical Industries Ltd., Osaka, Japan).

Table I summarizes the number of pigs in which PEDV was detected at each postinoculation time point for each treatment group. The sensitivity and specificity of PEDV RT-PCR were calculated from the data in Table I. The sensitivity was determined by dividing the number of PEDV-positive samples by the number of PEDV-inoculated pigs (15/15 = 100%). The specificity was determined by dividing the number of control samples that were negative for PEDV by the number of mock-infected control pigs (10/10 = 100%).

Table I.

With RT-PCR, PEDV nucleic acid was detected in all (15/15) of the fecal samples from the PEDV-inoculated pigs. The virus was isolated from all (15/15) of the jejunal samples from the cell-adapted PEDV-inoculated pigs. Antigen and nucleic acid were detected in 13/15 and 14/15 of the jejunal and ileal samples from the PEDV-inoculated pigs by immunohistochemistry and in situ hybridization, respectively. In immunohistochemistry and in situ hybridization (Figure 1), positive enterocytes were distributed over the tip and along the sides of atrophied or fused villi in the jejunal and ileal samples. Positive cells had dark brown (immunohistochemistry) or black (in situ hybridization) reaction product in the cytoplasm of the infected cells without background staining. Neither PEDV antigen nor PEDV nucleic acid was detected in the intestinal and fecal samples from mock-infected control pigs.

Figure 1. Piglet jejunum. Nucleic acid of porcine epidemic diarrhea virus is present in the epithelium. In situ hybridization; cDNA, nitroblue tetrazolium/5-bromocresyl-3-indolylphosphate, methyl green counterstain.

A total of 94 samples were tested by all 3 methods in this study (Table II). Sixty-three (67%) of the samples were positive and 15 (16%) were negative for PEDV by all 3 methods, showing an agreement of 83% among the techniques. Of these 94 samples, 79 (83%) were positive and 15 (16%) were negative for PEDV by at least 1 of the 3 methods. Among the 79 samples known to be positive for PEDV by at least 1 of the 3 methods, 69 (87%) were positive for PEDV by immunohistochemistry, 72 (91%) were positive by in situ hybridization, and 73 (92%) were positive by RT-PCR.

Table II.

This study compared 3 independent techniques for specific identification of PEDV using the same specimens. The RT-PCR test for the detection of PEDV in fecal samples correlated well with detection of PEDV by immunohistochemistry and in situ hybridization in formalin-fixed, paraffin-embedded intestinal samples from the same piglets. The overall correlation among the 3 methods was 83%. High agreement provides confidence that any one of the techniques may be used alone to yield an accurate diagnosis of PEDV infection.

Immunohistochemistry and in situ hybridization were useful for confirming a diagnosis of PEDV-induced enteritis when only formalin-fixed tissues were available. Formalin fixation allows veterinary practitioners to ship clinical samples for PEDV diagnosis in a well-preserved, noninfectious state. Viral nucleic acid was detected by in situ hybridization in some tissue sections in the absence of viral antigen. The reason for this discrepancy is not known. One possibility is that PEDV infection is non-productive; thus, there is no viral antigen protein to be detected by immunohistochemical methods. Another possibility is that formalin fixation denatures viral antigens. Such denaturation can lead to false- negative results (14,15). Immunohistochemistry and in situ hybridization provide cellular detail and histologic architecture, so that small numbers of PEDV-infected cells and lesions may be studied in the same section. Additionally, the 2 methods may be useful for performing retrospective studies when fresh tissue is no longer available.

The RT-PCR used in this study was demonstrated to be an easy, rapid, specific, and sensitive test for the detection of PEDV from fecal samples. It provides a practical method for the rapid diagnosis of clinical samples. In contrast, a previous study indicated that RT-PCR was unsuitable for the detection of PEDV (16). The assay could be affected by the presence of inhibitors in the intestinal and fecal samples (16). Although the mechanism for this inhibition is unknown, intestinal and fecal samples contain variable amounts of bilirubin and bile salts, which inhibit the activity of the thermostable DNA polymerase (17,18). Using optimal conditions and different reagents circumvented the problems of inhibition and allowed development of a sensitive and specific technique. The assay is effective in detecting PEDV in fecal samples from live pigs (10); hence, killing is not necessary. Because of the close correlation with other diagnostic techniques, the RT-PCR used in this study should be considered a useful and practical tool for the detection of PEDV infection in fecal samples.

Diagnosis by transmission electron microscopy (TEM) is often inconclusive because Coronavirus particles are difficult to identify if the spikes are lost or not clearly visible (2). Transmissible gastroenteritis virus and PEDV have similar size and morphology (3,19). An enzyme-linked immunosorbent assay (ELISA) can also detect PEDV antigen in fecal samples. However, viral antigen was inconsistently detected by ELISA in fecal samples (20). Recently, the reliability of a new ELISA method for detecting PEDV antigen in intestinal contents was confirmed (16). Further study is needed to compare these 3 methods and the new ELISA method for the detection of PEDV in clinical samples.

Rapid and accurate diagnosis and treatment are crucial to reducing mortality and morbidity from PEDV-induced enteritis in piglets. Because the differential diagnosis of PEDV-associated enteritis is broad (18), the ability to diagnose this enteritis promptly would greatly facilitate the management of swine producers. The present study demonstrated that the selection of diagnostic technique depends on the types of sample provided. When only fecal samples are submitted to laboratories, RT-PCR is recommended for the diagnosis of PEDV infection. When only formalin-fixed intestinal samples are submitted, immunohistochemistry and in situ hybridization are more suitable and sensitive.