Porcine Encephalomyocarditis Virus Persists in Pig Myocardium and Infects Human Myocardial Cells

Abstract

Recent advances toward using pig tissues in human transplantation have made it necessary to determine the risk of transmitting zoonotic viruses from pigs to humans or vice versa. We investigated the suitability of the porcine encephalomyocarditis virus (EMCV) model for such studies by determining its ability to persist in pigs, escape detection by routine serological methods, and infect human cells. Intraperitoneal inoculation of 5-week-old pigs with EMCV-30, a strain isolated from commercial pigs, resulted in acute cellular degeneration, infiltration of lymphocytes, and apoptosis in myocardium in 13 of 15 (86.7%) pigs during the acute phase of disease (3 to 21 days postinfection), followed by less-severe lymphocytic infiltration and apoptosis in 5 of 10 (50%) pigs during the chronic phase of the disease (day 45 to 90 postinfection). In the brain, lymphocytic infiltration, neuronal degeneration, and gliosis were observed in 26 to 33% of pigs in the acute phase of disease whereas perivascular cuffing was the predominant feature during chronic disease. EMCV antigens and RNA were demonstrated in the myocardium and brain during the chronic phase of disease. Analysis of 100 commercial pigs that were negative for EMCV antibodies identified two pig hearts positive for EMCV RNA. Porcine EMCV productively infected primary human cardiomyocytes as demonstrated by immunostaining using a monoclonal antibody specific for EMCV RNA polymerase, which is expressed only in productively infected cells, and by a one-step growth curve that showed production of 100 to 1,000 PFU of virus per cell within 6 h. The findings that porcine EMCV can persist in pig myocardium and can infect human myocardial cells make it an important infectious agent to screen for in pig-to-human cardiac transplants and a good model for xenozoonosis.

Encephalomyocarditis virus (EMCV) is a widely distributed picornavirus belonging to the Cardiovirus genus. The picornavirus infects many animal species including pigs (15), rodents (41), cattle, (35), elephants (11), raccoons (43), marsupials (30), and primates such as baboons, monkeys, chimpanzees, and humans (3, 14, 17, 30, 38, 41). Rats and mice are the natural hosts of the virus, passing the virus to other species through fecal-oral transmission. In rodents EMCV causes lesions in the heart, pancreas, central nervous system, and testes (4, 28). Pigs are the most commonly and severely infected domestic animals, as EMCV is endemic in many pig populations (2, 9). The virus causes acute myocarditis and sudden death in preweaned pigs, whereas transplacental infection of sows causes fetal mummification, abortion, stillbirth, and neonatal death (15). Infections in older pigs are asymptomatic. Even though no detailed pathogenetic studies have been performed to determine the porcine cells supporting EMCV replication and possible persistence, the heart, liver, and kidney have been shown to have higher EMCV titers than blood, suggesting that EMCV replicates in these organs (5).

Studies indicate that EMCV can cause interspecies infections, making it an important zoonotic agent (14, 18, 30, 31). For example, EMCV strains isolated from different species are antigenically similar, and isolates that have caused myocarditis and pancreatitis in pigs have been associated with rodent outbreaks (18, 31, 39). The few documented cases of EMCV infection in humans have been associated with fever, neck stiffness, lethargy, delirium, headaches, and vomiting (24). In Germany, strains of the virus have been isolated from children suffering from meningitis and encephalitis, although a causal relationship between EMCV and the symptoms was not demonstrated (10). In Australia, cases of human EMCV infection have been reported in New South Wales, an area with a high incidence of the pig disease (17). Although an EMCV outbreak in a United States zoo involving multiple animal species did not result in illness to humans, a zoo attendant who cared for EMCV-infected primates demonstrated an antiviral antibody titer of 1:1,280 (41).

Renewed interest in pig-to-human zoonotic viruses has arisen from advances in xenotransplantation as a means of overcoming the acute shortage of transplantation tissues and organs for humans. Porcine cells, tissues, and organs are the primary animal tissues being considered for human transplantation because of similar anatomical and physiological features in humans and pigs, ready availability of the species, and relative ease of breeding pigs. For example, porcine neuronal cells, hepatocytes, and pancreatic islet cells are in various stages of trials for transplantation into humans, and the results are encouraging (23, 27, 29, 32, 37). In patients with Parkinson's disease, a neurodegenerative disorder characterized by loss of neurons in the substantia nigra and a corresponding decrease in dopamine levels within the striatum, intracerebrally transplanted dopamine-producing pig neural cells have survived for as long as 7 months and formed extensive axonal connections with the human host neurons (23, 29). In diabetes mellitus, a disease for which islet transplants have the potential to become an effective treatment, transplantation of fetal porcine islet cell clusters under the kidney capsule of cynomolgus monkeys resulted in delayed rejection, holding promise that use of xenogeneic islet tissues in humans may be attempted soon (32, 34).

A major concern in xenotransplantation is that disruption of anatomical barriers resulting in intimate contact between recipient and xenograft, combined with the routine immunosuppression of recipients, may facilitate interspecies transmission of xenogenic infectious agents to a substantially greater extent than would normal contact between humans and animals (13). To address this concern, researchers are investigating the risk posed by porcine viruses such as swine influenza virus, parainfluenza virus 1, EMCV, and retroviruses by determining whether the viruses can establish persistent infection (13). Cardioviruses can establish persistence via mechanisms that are not fully understood. For example, Theiler's murine encephalomyelitis virus (TMEV) can persist in the spinal cord white matter of susceptible mice for as long as 2.5 years, causing chronic myelin destruction (22). EMCV RNA has been detected in the myocardia of mice 90 days after the virus-mediated myocytolytic stage of the disease, and at 1 month after birth infectious virus can be recovered from piglets infected in utero (19, 21, 40). These findings suggest that EMCV can persist for some period after the acute disease. Detection of viral antibodies by virus neutralization or enzyme-linked immunosorbent assay (ELISA) and virus isolation are the routine tests used for the diagnosis of EMCV infection. We investigated the suitability of EMCV as a xenozoonosis model by determining whether the virus persists in pig tissues or infects human cells and by determining the potential for persistent EMCV to escape detection by routine screening methods. In situ hybridization, reverse transcription-PCR (RT-PCR), and immunohistochemistry were used to determine the persistence of EMCV in pig tissues, and morphometric analysis was used to characterize associated pathologic changes. The potential for seronegative pigs to harbor EMCV and the ability of porcine EMCV to infect human cells were determined.

MATERIALS AND METHODS

Virus.

The MN-30 strain of EMCV (EMCV-30), isolated from naturally infected pigs in Minnesota in 1987 and kindly provided by HanSoo Joo of the Department of Veterinary Pathobiology, University of Minnesota, was used in all experiments (16). The virus was propagated in HeLa cells. EMCV purification for ELISA was performed as described previously for TMEV (26). Briefly, supernatant from infected HeLa cells was clarified by adding IGEPAL CA-630 (Sigma) and centrifuging at 10,000 × g for 20 min. Sodium dodecyl sulfate (0.5%) was added to the lysate, and the solution was underlaid with 30% (wt/wt) sucrose and centrifuged at 77,000 × g for 3 h at 20°C. The pelleted virus was overlaid on a cesium chloride gradient (density, 1.2 to 1.4 g/ml), and centrifuged at 77,000 × g overnight at 5°C. The band containing purified virus was dialyzed in phosphate-buffered saline and stored at −80°C. Viral stocks were titered by plaque assay on HeLa cells and stored at −80°C.

Experimental infection of pigs and sample collection.

Twenty-five 5-week old pigs were obtained from an EMCV-free swine herd (Midwest Research Swine, Gibbon, Minn.) and placed in negative-pressure isolation units at the University of Minnesota animal facilities. ELISA confirmed the pigs to be negative for EMCV antibodies before infection. Animals were intraperitoneally inoculated with 2.9 × 10⁸ (PFU) of EMCV-30 in a 1-ml volume. Handling of animals, including feeding and euthanasia, was in conformity with the National Institutes of Health and University of Minnesota institutional animal care guidelines. Four to six pigs were euthanized at days 7, 21, 45, and 90 postinfection (p.i.) using pentobarbital sodium. Tissues from the brain, heart, kidney, liver, spleen, skeletal muscle, pancreas, and mesenteric lymph node tissues were collected for RNA isolation, histopathology, and immunohistochemistry. Tissues for RNA isolation and cryosectioning were snap-frozen in liquid nitrogen, whereas tissues for histopathology and in situ hybridization were fixed in 10% neutral buffered formalin and embedded in paraffin. Paraffin-embedded sections were cut at a 4 μm thickness and stained with hematoxylin and eosin for histopathologic analysis. Sera were collected from pigs before inoculation and at sacrifice for ELISA and virus neutralization.

Sera and heart tissues from commercial pigs.

A total of 100 hearts (10 samples per herd) were obtained from slaughter pigs originating from a central Minnesota pig farm and demonstrated to be EMCV free by virus neutralization. Heart tissues were analyzed for EMCV RNA by nested RT-PCR. Serum samples were also collected from these pigs and were analyzed for EMCV antibodies by virus neutralization and ELISA.

Nested RT-PCR.

Pig tissues collected at days 7, 21, 45, and 90 p.i., and heart tissues obtained from commercial pigs, were analyzed for EMCV RNA by nested RT-PCR. One gram of brain, heart, liver, kidney, spleen, or skeletal muscle was homogenized in TRIzol (Life Technologies, Gaithersburg, Md.), followed by chloroform extraction of total RNA. Five micrograms of total RNA was reverse transcribed using an oligo(dT) primer and the Superscript II reverse transcription kit (Life Technologies) before the outer and nesting PCRs using primer pairs specific for the VP1 or VP2 genes (6) (GenBank accession no. M81861) were performed. The VP1 primers used for the outer PCR were CGAACTCAGTGATACTGACCCCTG for the 5′ primer and CCAGCTCTCGGGGTCATATCAATC for the 3′ primer, whereas those for the nesting reaction were GTCTGACAGAAATTTGGGGCAATG for the 5′ and GTCAGGCTTTGTGCCAGCAAAGAAC for the 3′ primer. The VP2 primers for the outer PCR were CAGTGGGCCGTCTTGTCGGTTATG for the 5′ and CCTCAAGATCCACTGTGGTGTTAG for the 3′ end, whereas nesting primers were GGCATCATGTGCTGACACTGCTTCAG for the 5′ and CCACTGCCAGAAGTTCTGATGGTC for the 3′ end. All primers are listed in 5′-to-3′ orientation. The outer PCR was performed on 1 μl of the template cDNA by adding 0.2 mM deoxynucleoside triphosphates, 2 mM magnesium chloride, 10 pmol of each primer, and 1 U of Taq polymerase (Life Technologies). For the nested PCR, 2.5 μl of the first-round PCR reaction mixture was used. PCR products were analyzed using agarose gel electrophoresis.

In situ hybridization.

To detect the presence of viral genome in pig tissues, in situ hybridization was performed using a 309-bp VP2-specific probe as described previously for TMEV (25). Briefly, a VP2 cDNA was subcloned from a full-length EMCV cDNA (pEC9 clone) into plasmid pUC 18 using BamHI and EcoRI restriction sites, and the cDNA probe was prepared by digesting the VP2 plasmid with NcoI and KpnI restriction enzymes (12). The probe was labeled with [³⁵S]dATP using the Random Primers DNA Labeling System (Life Technologies) and was purified using G-50 Sephadex Quick Spin columns (Roche, Indianapolis, Ind.). To prepare tissue samples, paraffin-embedded sections were deparaffinized using xylene whereas cryostat sections were fixed in 0.5% paraformaldehyde–0.5% glutaraldehyde–0.02 M disodium phosphate–0.08 M sodium phosphate–0.002% calcium chloride–1% dimethyl sulfoxide–1.6% glucose. Sections were digested with 10 μg of proteinase K/ml in phosphate-buffered saline for 30 min at 37°C and then treated with 0.1 M triethanolamine containing acetic anhydride. The sections were prehybridized in a buffer containing deionized formamide, Denhardt's solution, sodium chloride, salmon sperm DNA, yeast total RNA, and yeast tRNA for 4 h at room temperature before hybridization with ³⁵S-labeled 309-bp VP2 probe. Hybridization was performed overnight at 37°C, followed by extensive washes in reducing buffer at 55°C. Air-dried slides were immersed in an NTB2 film emulsion (Eastman Kodak Co., Rochester, N.Y.) and exposed at 4°C for 5 days.

Histopathologic analysis.

Blinded histologic analysis was performed on the hearts, brains, livers, kidneys, spleens, skeletal muscles, pancreases, and mesenteric lymph nodes of pigs infected with EMCV for 7, 21, 45, and 90 days. Each section was given a score between 0 and 4 based on specific criteria. Sections with no pathologic changes were given a score of 0, and sections demonstrating only rare foci of degeneration or inflammatory cell infiltrate were given a score of 1. Sections with multiple foci of moderate cellular degeneration, increased inflammatory cell infiltrate, and moderate fibroblast infiltration were given a score of 2, whereas those with severe degeneration and inflammatory infiltration but no necrosis were given a score of 3. Tissues with areas of acute degeneration, necrosis, severe inflammatory cell infiltrate, mineralization, and extensive fibroplasia were given a score of 4.

Immunohistochemistry.

To detect the presence of replicating virus in pig tissues and primary human cells, a mouse monoclonal antibody specific for the EMCV RNA polymerase (3D protein) was used (7). Sections of frozen heart tissue (thickness, 6 μm) were fixed in cold acetone for 20 min and blocked in 10% normal goat serum for 30 min. An anti-polymerase antibody was added, followed by peroxidase-conjugated goat anti-mouse immunoglobulin G (IgG) (Roche). Addition of 3,3′-diaminobenzidine substrate (Vector Laboratories, Burlingame, Calif.) for 10 min localized the EMCV antigens. Slides were counterstained with Mayer's hematoxylin and were examined by light microscopy.

Detection of apoptosis.

Cells undergoing apoptosis were identified using a commercial kit according to the manufacturer's instructions (Roche Diagnostics GmbH, Mannheim, Germany). Frozen heart sections were fixed in 4% paraformaldehyde before quenching with 3% H₂O₂ in methanol for 10 min. Terminal deoxynucleotidyl transferase (derived from calf thymus) was added to incorporate the fluorescein-labeled nucleotides at DNA strand breaks, followed by horseradish peroxidase-labeled sheep anti-fluorescein Fab fragment to detect the incorporations. Apoptotic cells were localized by addition of 3,3′-diaminobenzidine substrate and brown color development. Slides were lightly counterstained with hematoxylin and were analyzed by light microscopy.

Virus neutralization.

Sera were serially diluted from 1:8 to 1:2,048 in microtiter plates (dilutions run in triplicate); then 1,000 50% tissue culture infective doses of EMCV-30 was added to each dilution and incubated at 37°C for 1 h. Baby hamster kidney cells were added to each well, and the neutralizing antibody titer was read as the highest serum dilution at which a confluent cell monolayer was observed. A neutralizing serum titer of 1:16 dilution or higher was considered definitive for EMCV infection, with a sensitivity of 93.9% and specificity of 100% (42).

Anti-EMCV IgG ELISA.

EMCV-specific IgGs were detected by ELISA using purified EMCV as an antigen as described previously for TMEV (26). Plates were coated with UV-inactivated purified EMCV at a concentration of 0.5 μg/well in 0.1 M sodium carbonate buffer (pH 9.5) and blocked with 1% bovine serum albumin. Fourfold dilutions of serum from 1:40 to 1:128,000 made in 0.2% bovine serum albumin were added to the wells, followed by biotinylated goat anti-swine IgG (Kirkegaard and Perry, Gaithersburg, Md.). Each serum dilution was run in triplicate. Virus-specific IgG binding was detected using a 1:2,000 dilution of alkaline phosphatase-conjugated streptavidin (Jackson ImmunoResearch, West Grove, Pa.) followed by substrate development with p-nitrophenyl phosphate (Sigma Diagnostics) in buffer containing 0.1 M carbonate and 1 mM magnesium chloride at pH 9.5. Color intensity was read at 405 nm using an ELISA plate reader (Molecular Devices, Sunnyvale, Calif.).

Human cell culture and infection.

To determine the susceptibility of human cells to porcine EMCV, primary human cells were obtained from various sources and inoculated with EMCV-30. Cryopreserved human renal epithelial cells, fetal aortic endothelial cells, cardiomyocytes, bone marrow progenitor mononuclear cells, and peripheral blood mononuclear cells were obtained from Clonetics, a subsidiary of BioWhittaker (Walkersville, Md.). The human cardiomyocytes were isolated from normal human hearts and confirmed to be positive for MF-20 myosin (>95% of cells) and sarcometric actin (95 to 100% of cells) but negative for smooth muscle α-actin. All cell types from Clonetics were cultured in the media and under the conditions recommended by the manufacturer. Primary human hepatocytes derived from liver biopsies were generously provided by Stephen Strom, University of Pittsburg. Human neuroblastoma cells were obtained from the American Type Culture Collection (CHP-212) and cultured in a 1:1 mixture of minimum essential medium and Ham's F-12 medium with 10% fetal bovine serum. All cell types were directly inoculated with EMCV-30 or passaged once before inoculation. Virus inoculation was performed in T-75 flasks with 3 to 5 PFU of EMCV-30 per cell for 7 h at 37°C. After infection, cells were mounted onto glass slides and used for in situ hybridization or immunohistochemistry.

RESULTS

Clinical and histopathologic changes are associated with primary myocardial infection.

Of the 25 pigs infected with EMCV, 4 (16%) died at day 3 p.i. from acute myocarditis. These pigs showed extensive lysis of sarcoplasm, cellular degeneration, and early mineralization in the myocardium. Congestion of the lungs and liver were also noted, but no abnormalities in the brain or any other organs were detected at day 3 p.i. The other 21 pigs (84%) did not develop any clinical illness throughout the 90-day experimental period. At days 7, 21, 45, and 90 p.i., heart, brain, spleen, liver, skeletal muscle, kidney, pancreas, and mesenteric lymph node tissues were collected and processed for histopathologic analysis. The most severely affected organs were the heart and brain. During the acute phase of the disease (days 3 to 21 p.i.), the most common gross abnormalities were multiple foci of pale myocardial lesions, observed in all pigs (n = 15) sacrificed during this period. Histologically, pigs sacrificed at 7 and 21 days p.i. showed severe myocardial lesions including multiple foci of degeneration and necrosis with lysis of sarcoplasm and early mineralization (Fig. 1A and 2). In some cases, a lymphocytic inflammatory infiltrate was present (Fig. 1). In the chronic phase of the disease (days 45 and 90 p.i.), multiple discrete nodules of myocardial mineralization were observed grossly in 3 of 10 (30%) pigs. Approximately 63% of the pigs sacrificed at days 45 and 90 p.i. showed discrete foci of apoptotic cells and inflammatory cell infiltration in the myocardium (Fig. 1C and 3), suggesting an active infectious and/or inflammatory process in the chronic phase of the disease. Compared to heart tissues from acutely infected pigs, myocardial tissues from chronically infected pigs had smaller and fewer areas of inflammation and discrete foci of fibrosis and repair. Cells undergoing apoptosis were demonstrated in myocardial tissues during both the acute and chronic phases of the disease; however, apoptotic cells were most numerous at day 7 and were fewer but consistently present at days 45 and 90 p.i. (Fig. 3).

FIG. 1.

EMCV-induced pathologic changes in the heart and brain. Five-week-old pigs were intraperitoneally inoculated with 2.9 × 10⁸ PFU of EMCV-30, and histopathologic changes were, analyzed at days 7, 21, 45, and 90 p.i. Micrographs show acute inflammatory and degenerative changes in the myocardium at 7 days p.i. (A), an inflammatory and fibrotic myocardial lesion at 21 days p.i. (B), myocardial lymphocyte infiltration at 90 days p.i. (C), and perivascular cuffing in the cerebral cortex at 90 days p.i. (D). Heart and brain sections were embedded in paraffin, and 4-μm-thick sections were stained with hematoxylin and eosin. Magnification, ×200 for panels A through C and ×400 for panel D.

FIG. 2.

Scatter plot showing pathology scores of heart tissues from EMCV-infected pigs. Five-week-old pigs were intraperitoneally inoculated with 2.9 × 10⁸ PFU of EMCV-30, and histopathologic changes were analyzed at 3 (n = 4), 7 (n = 5), 21(n = 6), 45 (n = 5), and 90 (n = 5) days p.i. Each heart tissue section was given a score between 0 (no pathology) and 4 (necrosis on days 3 and 7, or fibrosis on days 21, 45, and 90) based on the extent of pathology as described in Materials and Methods. Solid lines represent the mean score at each time point (3 at day 3, 2.7 at day 7, 1.6 at day 21, 1.4 at day 45, and 1.2 at day 90 p.i.). The data indicate that EMCV, causes acute damage to the heart tissue, but in some pigs it can establish persistent infection.

FIG. 3.

EMCV-induced apoptosis in pig myocardium. Frozen sections of the heart were immunostained to identify cells undergoing apoptosis using the in situ end-labeling peroxidase-based detection system. Extensive apoptosis (brown-stained cells indicated by arrows) was observed at day 7 p.i., and fewer apoptotic cells were detectable at days 45 and 90 p.i. Uninfected myocardial sections were negative. Sections were lightly counterstained with hematoxylin. Magnification, ×100.

Histopathologic changes in the brain included infiltration of inflammatory cells in the meninges and perivascular cuffing in the cerebral cortex and hippocampi in 5 of 15 (33.3%) pigs examined between days 3 and 21 p.i. In addition, there was neuronal degeneration in four (26.3%) and foci of gliosis in two (13.3%) pigs. In the chronic phase, perivascular cuffing was observed in the cerebral cortex, medulla, and cerebellum in 3 of 10 (30%) pigs (Fig. 1D). Spleens had mild lymphocytic hyperplasia at day 7 p.i. but no histopathologic changes at any other time point. There were no pathologic changes detected in the liver, pancreas, kidney, skeletal muscle, or mesenteric lymph nodes. The presence of pathologic changes in the myocardium and brain in the chronic stages of EMCV infection suggested that viral persistence either directly induced tissue damage or stimulated immune effector functions that caused tissue damage.

Presence of EMCV antigens and RNA in pig myocardium 90 days p.i.

EMCV persistence was analyzed using in situ hybridization and nested RT-PCR on the heart, brain, liver, kidney, spleen, skeletal muscle, pancreas, and mesenteric lymph nodes. In situ hybridization localized EMCV RNA in heart, brain, spleen, kidney, and skeletal muscle at days 3, 7, and 21 p.i., but in myocardium and brain only in the chronic phase of the disease (Fig. 4A and B). More importantly, EMCV antigens were localized in the myocardium using anti-EMCV RNA polymerase 90 days after infection (Fig. 5C). Hybridization and immunostaining performed on myocardial tissues from uninfected control pigs were negative for viral RNA and antigens (data not shown). RT-PCR analysis of tissues from pigs that died from acute cardiac failure at day 3 p.i. showed large amounts of viral RNA in all tissues. Viral VP1 and VP2 RNA were easily demonstrated by gel electrophoresis after primary PCR, whereas at other time points (days 7, 21, 45, and 90) nested PCR was required to produce visible electrophoresis bands. This indicated a greater viral load in tissues at the early stage of the disease (day 3 p.i.). At days 7 and 21 p.i., EMCV RNA was detected in heart and spleen tissues (7 of 10 pigs), whereas in the chronic stages of the disease (days 45 and 90 p.i.), EMCV RNA was most commonly detected in brain (7 of 10 pigs), heart (6 of 10 pigs), and skeletal muscle (6 of 10 pigs) as shown in Table 1 and Fig. 5. Uninfected control pig tissues did not generate PCR products. The observation that 12 of 16 hearts (75%) were positive for EMCV RNA between days 21 and 90 p.i. reinforced pathology data suggesting that the heart is the primary site of EMCV persistence in pigs (Table 1).

FIG. 4.

Localization of EMCV RNA and antigens in the hearts and brains of chronically EMCV-infected pigs. EMCV RNA was localized by in situ hybridization using a 309-bp ³⁵S-labeled VP2-specific probe in the myocardium (A) and brain (B) of a pig infected for 90 days. Black grains (arrows) indicate viral RNA-positive cells. (C) Demonstration of EMCV antigens (brown staining, indicated by arrows) in the myocardium of a 90-day-infected pig by immunohistochemistry using a monoclonal antibody specific for EMCV RNA polymerase. Magnification, ×268 for panels A and B and ×134 for panel C.

FIG. 5.

Detection of EMCV RNA in pig tissues by RT-PCR. Five-week-old pigs were intraperitoneally inoculated with 2.9 × 10⁸ PFU of EMCV-30, and the brain, heart, kidney, liver, spleen, and skeletal muscle were tested by nested RT-PCR for EMCV RNA using VP1- or VP2-specific primer sets at days 7, 21, 45, and 90 p.i. (A) Agarose DNA gel showing RT-PCR products specific for VP1 (436 bp) and VP2 (390 bp) in the heart (lanes 2), liver (lanes 4), spleen (lanes 5), and skeletal muscle (lanes 6) of a pig infected for 90 days. The brain was positive with VP1 primers (lane 1) but negative with VP2 primers (lane not shown), whereas the kidney was negative with both VP1 and VP2 primers. (B) Presence of EMCV RNA in hearts of seronegative commercial pigs. Hearts were obtained at the time of slaughter and tested by nested RT-PCR. The gel shows that pig 67 was positive for VP1 RNA whereas the rest of the pigs (animals 61 to 66 and 68 to 70) were negative. Two of the 100 pig hearts analyzed (from 10 different herds) were positive.

TABLE 1.

Number of EMCV-RNA positive pigs following EMCV infection^a

Day p.i.	No. of pigs positive for EMCV RNA in:
	Brain	Heart	Spleen	Liver	Kidney	Skeletal muscle
7	1	5	4	1	0	0
21	4	6	3	3	3	4
45	3	4	2	1	2	3
90	4	2	3	2	3	3
Total no. positive (%)	12 (57.1)	17 (81)	12 (57.1)	7 (33.3)	7 (33.3)	10 (47.6)

^aFive-week-old EMCV-negative pigs were intraperitoneally injected with EMCV, and the tissues shown were analyzed for viral genome by nested RT-PCR using EMCV VP1 or VP2 primers. A total of 21 pigs were analyzed on days 7, 21, 45, and 90 p.i., 4 to 6 pigs per time point. 

Pig tissues can transmit EMCV to immunocompromised animals.

To determine the ability of infected pig tissues to transmit EMCV to a susceptible host, homogenized pig myocardial tissues harvested 7, 21, 45, or 90 days p.i. were intraperitoneally inoculated into recombination activation gene 2-deficient (RAG2^−/−) mice. RAG2^−/− mice lack B or T lymphocytes and are susceptible to acute viral infections. Myocardial tissues were homogenized to a 10% suspension in RPMI-1640 medium and passed through a 0.2-μm-pore-size filter. Mice were intraperitoneally injected with 1 μl of the homogenate and monitored for clinical signs of myocarditis and encephalitis. As a positive control, RAG2^−/− mice were inoculated with 10 PFU of EMCV-30 per mouse. The RAG2^−/− mice inoculated with 10 PFU of EMCV-30 (n = 2) died 35 to 38 days later, whereas mice inoculated with myocardial homogenate from day 7-infected pigs (n = 4) died 27 to 54 days after inoculation as a result of EMCV-induced disease. RAG2^−/− mice inoculated with homogenate from pigs infected for 21 (n = 4), 45 (n = 4), or 90 (n = 4) days showed no clinical signs of EMCV disease for 60 days.

EMCV RNA can be detected in seronegative pigs.

The experimentally infected pigs rapidly developed EMCV-specific neutralizing IgGs within 7 days (Fig. 6A), reaching peak levels by day 21 p.i. before dropping to low levels by day 45 p.i. (Fig. 6B). All sera from pigs infected for 7, 21, 45, or 90 days neutralized EMCV. We envisioned that persistent EMCV may escape detection by routine serological methods such as virus neutralization and ELISA. To investigate this possibility, tissues from 10 pig herds obtained from a farm that has been EMCV free were examined. Ten pig hearts and serum samples from the same pigs were analyzed for each herd. Sera were tested for EMCV-specific IgG and neutralizing antibodies. All 100 pigs were negative by virus neutralization (no neutralization at a 1:16 dilution) and ELISA (Fig. 6C). However, EMCV RNA was reproducibly demonstrated by RT-PCR in 2 of the 100 heart tissues tested (Fig. 5B), both of which had no detectable EMCV antibodies by virus neutralization (at a 1:16 dilution) or ELISA (absorbance reading below 0.2). To eliminate the possibility of contamination, PCR tests were conducted in a room dedicated to the test, and each test included positive- and negative-control samples.

FIG. 6.

Detection of EMCV-specific IgG in sera from experimentally infected and commercial pigs. EMCV-specific IgGs were determined by ELISA. (A) Virus-specific IgG levels in sera of pigs infected for 7, 21, 45, and 90 days. Data are means (± standard errors) of A₄₀₅ readings from four to five serum samples at each time point, performed at serum dilutions between 1:125 and 1:128,000. (B) Profile of virus-specific IgG levels during acute (day 7 to 21) and chronic (day 45 to 90) infection determined at a 1:500 serum dilution, showing that antibody levels peaked at day 21 p.i. before decreasing to low levels at days 45 and 90 p.i. (C) Analysis of EMCV-specific IgGs in sera from 10 commercial pig herds obtained from an EMCV-free farm over a 10-month period. Each herd is represented by the mean (± standard error) A₄₀₅ reading of serum samples from 10 pigs. Sera from pigs experimentally infected with EMCV for 21 days were used as a positive control.

EMCV can infect human myocardial cells.

The ability of porcine EMCV to infect human cells was assessed by inoculating primary human cardiomyocytes, renal epithelial cells, bone marrow progenitor cells, aortic endothelial cells, peripheral blood mononuclear cells, and hepatocytes with 3 to 5 PFU per cell of EMCV-30. Cells were harvested 7 and 16 h after inoculation and were subjected to immunohistochemistry and in situ hybridization to localize viral antigens and RNA, respectively. Of importance was the immunostaining using a monoclonal antibody specific for EMCV RNA polymerase (3D protein), because the polymerase is detected only during productive virus infection in susceptible cells. Human cardiomyocytes demonstrated high immunoreactivity with anti-EMCV polymerase antibody (Fig. 7B), whereas uninfected cells were negative (Fig. 7A). Ninety-five percent of the EMCV-inoculated cardiomyocytes were positive for viral polymerase, and the reactivity was always in the cytoplasm (Fig 7B), confirming the cytoplasmic restriction of EMCV replication. Productive infection of the cardiomycytes was further confirmed by the large amount of viral RNA localized in inoculated cells (Fig. 7D) compared to uninfected cardiomyocytes hybridized for the virus (Fig. 7C). More than 95% of the cardiomyocytes infected for 16 h underwent cytolysis, and the remaining viable cells (5%) were positive for EMCV antigens and RNA. Hepatocytes, renal cells, aortic endothelial cells, bone marrow progenitor cells, peripheral blood mononuclear cells, and neuroblastoma cells were negative for EMCV polymerase antigens. These results indicated that human cardiomyocytes are susceptible to porcine EMCV infection.

FIG. 7.

EMCV antigens and RNA in primary human myocardial cells. Primary human cardiomyocytes obtained from Clonetics were inoculated with 3 to 5 PFU of EMCV-30 per cell, for 7 h. Cells were harvested and immunostained using a anti-EMCV RNA polymerase (3D protein) monoclonal antibody. Infected human cardiomyocytes were positive, as shown by the brown staining in the cytoplasm (B), whereas uninfected cells were negative (A). In addition, in situ hybridization using a 309-bp ³⁵S-labeled probe specific for the VP2 gene of EMCV was used to localize viral RNA in the infected cells. Large amounts of viral RNA were observed in human cardiomyocytes (black grains) infected with EMCV (D), whereas uninfected cells hybridized with the VP2 cDNA probe were negative (C). Cells were lightly counterstained with hematoxylin. Magnification, ×400.

To determine whether porcine EMCV can productively infect human cardiomyocytes, confluent primary human cardiomyocytes were grown in 12-well tissue culture plates and inoculated with 10 PFU of EMCV-30 per cell for 1, 2, 4, 6, 8, or 16 h before harvesting. Infected cells were washed to remove unattached virus, and four wells were harvested at each time point using a cell scraper. The cells were freeze-thawed and sonicated to release intracellular virus and were centrifuged to remove cellular debris, and serial 10-fold dilutions were added to a confluent HeLa cell monolayer for a plaque assay as described previously for TMEV (25). The results show a typical picornaviral growth curve with a 4-h lag (latent or eclipse) phase followed by a 2-h exponential (log) phase characterized by production of 100 to 1,000 PFU of virus per cell (Fig. 8). The EMCV growth curve shows that human cardiomyocytes can be efficiently and productively infected by porcine EMCV.

FIG. 8.

Growth curve demonstrating productive infection of primary human cardiomyocytes by porcine EMCV. Confluent primary human cardiomyocytes in 12-well tissue culture plates were inoculated with 10 PFU of EMCV-30 per cell for 1, 2, 4, 6, 8, or 16 h before harvesting. Infected cells were washed to remove unattached virus, and four wells were harvested at each time point for a plaque assay (3.1 × 10⁵ cells/well). Cells were freeze-thawed and sonicated to release intracellular virus and clarified by centrifugation, and serial 10-fold dilutions were added to confluent HeLa cells for plaque development. Results are expressed as the mean (± standard error) number of plaques per 6.2 × 10⁵ cells (from two wells) of four samples per time point. They show a 4-h lag phase followed by a 2-h exponential phase characterized by rapid production of infectious EMCV particles. The limit of detection for the test was 2 PFU per 6.2 × 10⁵ cells.

DISCUSSION

Previous studies have suggested that EMCV can be detected beyond the acute myocarditis and encephalitis phase of the disease in pig tissues (2, 19). Piglets infected with EMCV for 23 days and subsequently treated with an immunosuppressive drug developed myocardial pathologic and enzymatic abnormalities and transmitted the virus to susceptible contact pigs (2). However, a comprehensive study to determine the duration, location in tissues, and form of porcine EMCV persistence in pig tissues had not been performed. The importance of such an undertaking, and of determining the potential of persistent porcine EMCV to infect humans, has been rekindled by advances in pig-to-human xenotransplantation. Our studies demonstrate, for the first time, EMCV-induced pathologic changes (lymphocyte infiltration, myocardial degeneration, and apoptosis), and EMCV antigens and RNA in porcine myocardium for as long as 90 days after infection, suggesting the presence of infectious viral particles and/or active inflammatory processes in the chronic stages of infection. In addition, EMCV RNA, but not pathologic changes, was detected in the spleen, kidney, and skeletal muscles in the chronic stages (days 45 and 90 p.i.) of infection. Clearly, EMCV induced the most extensive pathologic changes in the myocardium early in the disease process (within 7 days), and approximately 40% of these pigs recovered with no detectable myocardial damage by day 21 after infection, as shown in Fig. 2. However, more than 60% of the pigs had myocardial pathologic changes, some of them severe (pathology scores of 3 and 4), in the chronic phase of the disease. Given that our attempts to isolate infectious virus in the chronic phase of the disease were unsuccessful, one may conclude that the persistent viral products detected (antigens and RNA) were not products of an infectious virus present in these tissues. However, it should be noted that both the viral antigens and pathologic changes, including lymphocyte infiltration and apoptosis, were detected in myocardial cells throughout the chronic disease. Most likely, the inability to detect infectious virus was due to a combination of insensitivity of the techniques used (plaque assay and mortality in RAG2-deficient mice), and inadequate sampling because only small portions of the pig heart could be processed for virus isolation.

We also investigated the hypothesis that persistent EMCV in pigs may escape detection by the existing serologic methods to pose a risk in xenotransplantation. In agreement with this hypothesis, EMCV RNA was demonstrated in 2 out of 100 pig hearts obtained from commercial pigs at the time of slaughter. These findings are important because the success of xenotransplantation is predicated on efficient and error-free screening of donor animal tissues for potential infectious agents. The findings point to a need to develop a rapid bedside RT-PCR test for screening pig tissues harvested for transplantation against zoonotic porcine virus genes.

EMCV-30 productively infected primary human cardiomyocytes, indicating that persistent EMCV in pig tissues can pose a risk to humans. More than 95% of the EMCV-infected cardiomyocytes underwent cytolysis between 7 and 16 h after inoculation, whereas the remaining 5% of viable cells still supported EMCV replication. Combined with the pathologic findings for the infected pigs, these data indicate that EMCV is capable of inducing cytodestructive changes in both human and pig myocardial cells. The demonstration of EMCV persistence and pathologic changes in chronic infection, and the ability of the virus to productively infect human cells, makes EMCV an ideal model of characterizing the risk of transplanting virus-infected pig tissues into humans. A recent study has shown showed that the presence of picornavirus, adenovirus, parvovirus, and herpesvirus genomes within transplanted hearts results in increased graft loss in children (33). Cardiac transplants comprise a large portion of organ replacement procedures performed in the United States, and many human patients die while waiting for donors. Mechanical heart devices, used mostly for bridge-to-transplant therapy, have not been successful for long-term use, leaving animals as an important alternative source of healthy hearts (1, 8, 20). Identifying potential zoonotic infectious viruses and characterizing the risk they pose to human transplant recipients is important in advancing xenotransplantation research.