Transmission of porcineencephalomyocarditis virus (EMCV) to miceby transplanting EMCV-infected pig tissues

Introduction

Encephalomyocarditis virus (EMCV), a cardio-virus belonging to the picornavirus family, infectsmany animal species including pigs, non-humanprimates, cats, elephants, ungulates, and humans[1–8]. Among domestic animals, pigs are the mostsusceptible to EMCV infection. EMCV is endemicin pig herds worldwide, and infected animalsdisplay a wide range of clinical symptoms rangingfrom none in adult pigs to abortions in sows andacute fatal myocardial failure in young pigs

[3,9–13]. Instances of human infection with EMCVhave manifested as generalized febrile illness, butthe virus has also been isolated from patients withmore severe illnesses such as encephalitis andmeningitis [12,14].Research in the field of xenotransplantation is

increasingly focusing on the use of pigs as donorsof organs and tissues, due to the many advantagesof using these animals instead of primates as adonor species. Pigs are readily available and easyto breed, and they have some anatomical andphysiological similarities with humans. A major

Brewer L, Brown C, Murtaugh MP, Njenga MK. Transmission ofporcine encephalomyocarditis virus (EMCV) to mice by transplantingEMCV-infected pig tissues.Xenotransplantation 2003; 10: 569–576. � Blackwell Munksgaard, 2003

Abstract: We recently demonstrated that pigs infected with porcineencephalomyocarditis virus (EMCV) develop a persistent infection (up to90 days post-infection (PI)) in the heart and brain that is accompaniedby virus-induced pathologic changes, and that EMCV productivelyinfects human cardiomyocytes in vitro, suggesting that EMCV may posea risk to humans following transplantation of pig tissues to humans(Brewer et al. J Virol 2001; 75: 11621–11629). In this report, wedemonstrate that intra-abdominal transplantation of myocardial orpancreatic sections from acutely EMCV-infected pigs (2 days PI) intoeither non-mutant C57BL/6 or C57BL/6-RAG-1)/) mice that lack B orT lymphocytes, resulted in transmission of the virus and acute fataldisease in all mice. In recipient RAG-1)/) mice, fatal EMCV diseaseoccurred within 2 days post-transplantation, and it was accompanied byhigh virus titers in brain, heart, liver, spleen, kidneys and skeletalmuscle, whereas in non-mutant C57BL/6 mice, disease occurred 5 to6 days post-transplantation and was accompanied by lower virus titers.Transplantation of myocardial or pancreatic tissues from chronicallyEMCV-infected pigs (21 and 50 days PI) did not induce clinical disease,but resulted in detection of EMCV RNA in the brain of recipientRAG-1)/) mice, no viral RNA was detected in non-mutant C57BL/6mice. Intra-abdominal transplantation of uninfected porcine myocardialtissues into RAG-1)/) mice followed by intramuscular inoculation withEMCV induced acute clinical disease but did not result in transmissionof virus to the xenograft. These results show that EMCV can be effi-ciently transmitted from pig myocardial and pancreatic tissues to mice,providing a model of pig-to-human viral xenozoonosis that can be usedto develop and test prophylactic and therapeutic measures against suchinfection.

Laurie Brewer,1 Corrie Brown,2

Michael P. Murtaugh1 andM. Kariuki Njenga11Department of Veterinary Pathobiology, Universityof Minnesota, St Paul, MN, USA, and 2Departmentof Pathology, University of Georgia, Athens, GA, USA

Key words: porcine encephalomyocarditis virus –porcine myocardium – porcine pancreas – virustransmission – xenozoonosis

Address reprint requests to M. Kariuki Njenga,Department of Veterinary Pathobiology, University ofMinnesota, 1971 Commonwealth Avenue, St Paul,MN 55108, USA (E-mail: Njeng001@tc.umn.edu)

Received 5 June 2002;Accepted 7 November 2002

Xenotransplantation 2003: 10: 569–576Printed in UK. All rights reserved

Copyright � Blackwell Munksgaard 2003

XENOTRANSPLANTATION

569

concern, however, is that xenografts may transmitundetected viruses from pigs to humans (xenozoo-nosis). Much of the ongoing research to addressthis issue involves evaluating the risk of transfer-ring porcine endogenous retroviruses (PERVs)whose genome can be integrated into host cellularDNA, and on herpesviruses because of their abilityto establish latency. So far, there is no evidence ofinfection with PERVs or porcine herpesviruses inhuman patients who have undergone porcinexenoperfusion procedures or patients that havereceived porcine islet or neural cell transplants[15–17]. A recent study reported the transmis-sion of PERV and resultant infection in multipletissues of severe-combined immunodeficient(SCID) mice that had received pig islet celltransplants [18]. The risk of viral transmissionfrom whole organ xenotransplants, however,remains unknown.Another important but ignored concern in pig-

to-human xenotransplantation is that human vir-uses present in the recipient may be infectious tothe xenograft and result in damages that render ituseless to the recipient. Our goal is to develop amodel for viral xenozoonosis in order to morespecifically address pathogenesis of such infectionsand evaluate therapeutic and prophylactic approa-ches to combat the infections. We recently dem-onstrated that EMCV infection of 5-week-old pigsresulted in virus persistence in the heart and brainfor up to 90 days [19]. In addition, apoptosis andinflammatory changes consistent with active virusreplication were demonstrated in the chronic phaseof the disease. Most importantly, EMCV RNAwas detected in 2% of heart tissues obtained fromcommercial pigs that tested negative for EMCVantibodies using routine serologic techniques.EMCV-30, the porcine field strain used in thestudy, productively infected human cardiomyo-cytes, resulting in up to 1000 plaque-forming units(PFU) of virus per cell in 8 h. These resultssuggested that persistent EMCV infection in pigscould escape detection and pose a potential risk tohumans following xenotransplantation. The pur-pose of the current study was to determine whethertransplantation of EMCV-infected pig tissueswould result in transfer of the pig virus to therecipient host. Recombination-activating-gene-1deficient (RAG-1)/)) mice that lack T or Blymphocytes, and non-mutant mice of the samegenetic background (C57BL/6), received myocar-dial and pancreatic tissue transplants from EMCV-infected pigs, and were monitored for signs ofdisease and presence of virus products in thetissues. We report transmission of EMCV to micefrom pig tissue transplants.

Materials and methods

Mice

Six-week-old male non-mutant C57BL/6and C57BL/6-RAG-1-deficient mice (RAG-1)/))were purchased from Jackson Laboratories (BarHarbor, MA, USA) and maintained in animalfacilities at the University of Minnesota inaccordance with the Institutional Animal Careand Use Committee and National Institutes ofHealth guidelines.

Viral inoculum

EMCV-30, a porcine field strain of EMCV isolatedin Minnesota, was passaged on HeLa cells andpurified as described in the past [19]. The purifiedviral stock was titered by plaque assay on HeLacells and stored at )80 �C.

Pig infection and sample collection

Ten 5-week-old pigs were obtained from anEMCV-free swine facility (Midwest ResearchSwine, Gibbon, MN, USA) and housed in animalisolation units at the University of Minnesota inaccordance with the Institutional Animal Care andUse Committee and National Institutes of Healthguidelines. ELISA confirmed the pigs to be negat-ive for EMCV antibodies before infection. Eachanimal was injected intraperitoneally with 8 · 108

PFU of EMCV-30 and monitored daily for clinicalsigns of disease. One or two animals were sacrificedat days 2, 7, 21 and 50 post-infection (PI) byintravenous injection of sodium pentobarbital.A separately housed uninfected pig was used as anegative control. The heart and pancreas wereaseptically dissected out from each animal andthoroughly washed with phosphate-buffered saline(PBS) to remove residual blood. One cubic milli-meter sections were cut from each organ andplaced directly into fresh sterile PBS for immediatetransplantation, or in cryoprotectant mediumcontaining 60% RPMI-1640 media (Life Technol-ogies, Gaithersburg, MD, USA), 30% fetal bovineserum and 10% dimethylsulfoxide for storage.Tissues to be stored were cooled overnight at a rateof )1 �C/min to )80 �C, and placed in liquidnitrogen the following day [20–22]. Immediatelyprior to transplantation, the frozen tissues werethawed quickly at 37 �C and thoroughly washed insterile PBS. The heart or pancreas tissues trans-planted into mice from the day 2 (D2) pigcontained approximately 30–40 PFU of EMCV-30 per section, resulting in a total introduction of150 to 240 PFU of EMCV-30 per recipient mouse.

Brewer et al.

570

Transplantation tissues from D7, D21, or D50 pigswere negative for infectious virus but positive forEMCV RNA.

Transplantation of pig tissues into mice

All surgeries were performed under the guidance ofthe University of Minnesota Research AnimalResources. Non-mutant C57BL/6 or RAG-1)/)

mice were tranquilized with acepromazine (0.01 to0.014 mg/g of body weight) and anesthesia inducedwith sodium pentobarbital (0.019 mg/g bodyweight). A midline incision was made through theskin and the body wall of the ventral abdomen andthe peritoneal surface exposed. Six porcine myo-cardial or pancreatic sections (1 mm3 each) weresutured onto the peritoneal surface lateral to theincision approximately 3 mm apart using 5-O silksuture [23]. The body wall was closed in a simpleinterrupted pattern, and the skin closed using 3MVetbondTM skin adhesive (3M Animal CareProducts, St. Paul, MN, USA). All animals wereplaced on a covered heating pad at low settingovernight for recovery, and treated with 2-mg/mltetracycline in drinking water for 7 days. The micewere sacrificed at the onset of clinical signs ofEMCV disease (See Results section) or after14 days if no clinical signs developed. Euthanasiawas carried out using 6-mg/g body weight ofsodium pentobarbital injected intraperitoneally.The heart, spleen, kidneys, liver, brain and skeletalmuscle sections were collected from each animal.One half of each organ was immediately frozen at)80 �C for virus detection using plaque assay orreverse transcription-polymerase chain reaction(RT-PCR) analysis. The remaining half of theorgans were fixed in formaldehyde, embedded inparaffin and sectioned at a 4-lm thickness for insitu hybridization, or stained with hemotoxylin andeosin for histopathologic analysis.

Viral RNA detection

EMCV RNA was detected in mouse tissues usingnested RT-PCR and in situ hybridization asdescribed previously [19]. Briefly, RNA was extrac-ted from mouse tissues using TRIzol (Life Tech-nologies, Gaithersburg, MD, USA), and reversetranscribed using the Superscript II kit (LifeTechnologies). Primary PCR was performed usingprimer pairs specific to the VP1 and VP2 viralcapsid protein genes of EMCV, followed by asecond round of PCR using nested primers for thesame genes. In situ hybridization was performedon 4-lm thick paraffin-embedded mouse tissuesections using a 35S-dATP labeled VP2 specific

probe. Sections were deparaffinized, digested withproteinase K and treated with triethanolamineprior to pre-hybridization for 4 h at 37 �C.Hybridization was carried out overnight at 37 �C,followed by washes in reducing buffer at 55 �C andexposure in NTB2 film emulsion (Eastman KodakCo., Rochester, NY, USA) for 5 days at 4 �C.

Virus detection by plaque assay

Infectious virus was detected in mouse tissuesusing plaque assay as described previously [24].Briefly, harvested tissues were weighed and homo-genized in RPMI-1640 media, sonicated for two1-min cycles to release virus, and centrifuged at3000 rpm for 10 min. Supernatants were diluted10-fold (10)1 to 10)6) in RPMI-1640, and 200 llof each dilution added to duplicate wells ofconfluent HeLa cells in 12-well plates. After virusattachment, infected cells were overlaid with pre-warmed 0.4% agarose and incubated at 37 �C for30 h for plaque development. Wells were fixed in60% ethanol, 20% acetic acid, and 10% formal-dehyde for 2 h, stained with 1% crystal violet andthe plaques in each dilution counted. The averagenumber of plaques formed was calculated fromduplicate wells per sample, multiplied by thedilution factors and expressed as PFU per gramof tissue.

Histopathologic analysis

Recipient mouse hearts, brains, kidneys, livers,spleens, and skeletal muscles were examined forhistopathologic changes.

Results

Clinical disease in EMCV-infected pigs

Two of the 10 EMCV-infected pigs (20%) diedsuddenly of acute cardiac failure on day 3 PI. Onepig became moribund on day 2 PI and waseuthanized for graft tissue collection (D2 pig).Two of the remaining pigs were sacrificed at day 7,two at day 21, and one at day 50 PI for transplanttissue collection (D7, D21, and D50 pigs, respect-ively). None of the animals sacrificed at days 7, 21,or 50 developed any clinical signs of EMCVinfection. Two of the infected pigs were not usedfor transplantation. The myocardial and pancre-atic tissues collected for transplantation from theD2 pig demonstrated high amounts of infectiousvirus, while the xenograft tissues collected fromD7, D21, and D50 pigs were positive for EMCVRNA but not infectious virus.

Pig-to-mouse EMCV xenozoonosis

571

Non-mutant C57BL/6 and RAG-1)/) mice that received acutelyinfected pig tissues developed acute encephalitis

All non-mutant C57BL/6 mice that received myo-cardial (n ¼ 4) or pancreatic (n ¼ 2) tissue graftsfrom the D2 pig developed clinical signs of EMCVencephalitis 5 to 6 days after transplantation. Incontrast, all RAG-1)/) mice that received myocar-dial (n ¼ 3) or pancreatic (n ¼ 2) grafts from theD2 pig developed clinical signs of encephalitiswithin 2 days after transplantation. Sick micebecame moribund and were sacrificed 1 to 3 daysafter developing clinical signs. The clinical signs inboth non-mutant and RAG-1)/) mice includedhunched posture, ruffled fur, lethargy, anorexia,and hind limb paresis. The mice that receivedmyocardial or pancreatic grafts from D7, D21 orD50 pigs (five C57BL/6 and 10 RAG-1)/) mice)did not develop clinical signs of EMCV infectionand were euthanized 14 days after transplantationfor tissue collection. A negative control RAG-1)/)

mouse that received myocardial and pancreatictissues from an uninfected pig showed no clinicalsigns and was euthanized on day14 post-transplan-tation. At necropsy, the engrafted pig heart andpancreas tissues were collected and examined forevidence of viability. The xenografts were surroun-ded by extensive abdominal adhesions consisting offibrin deposits on the abdominal wall, xenograft,and viscera in over half of the recipient animals,including the negative control. Histological exam-ination of xenografts collected 14 days after trans-plantation showed normal morphology of themyocardial and pancreatic cells with mild toextensive infiltration of mononuclear cells, partic-ularly along the edges of the graft. Approximately30% of the xenografts had necrotic centers,perhaps resulting from inadequate perfusion. As

positive controls, two RAG-1)/) mice and twonon-mutant C57BL/6 mice inoculated with 100PFU per mouse of EMCV-30 developed clinicalsigns within 6 to 8 days PI that were identical tothe clinical signs observed in the D2 recipient mice.

Histopathologic changes in the brain and myocardium of recipientmice

Histopathologic changes were observed in thebrain and heart of D2 recipient mice, whereasminimal pathologic changes were observed in D7,D21, and D50 recipient mice. In D2 recipientRAG-1)/) mice, lesions observed in the brainconsisted of malacia and degenerative neuronswith minimal infiltration of inflammatory cells(Fig. 1A). Several subtle myocardial lesions con-sisting of increased eosinophilia, possibly indica-ting degeneration was also observed. D2 recipientnon-mutant C57BL/6 mice sacrificed after day 5post-transplant showed regions of extensive mono-nuclear cell infiltration in both the brain (Fig. 1B)and myocardium (not shown). Brain lesions wereobserved in the cerebellum, hippocampus, andcerebral cortex in four of five RAG-1)/) and five ofsix non-mutant C57BL/6 mice that receivedD2-infected xenografts. Myocardial lesions inclu-ded foci of degeneration, necrosis and lymphocyteinfiltration; and were more severe in non-mutantC57BL/6 mice as compared with RAG-1)/) mice,which do not produce B or T lymphocytes. Similarlesions containing neuronal and myocardial degen-erative changes were observed in the non-mutantC57BL/6 and RAG-1)/) control mice that wereinoculated with 100 PFU EMCV-30. No lesionswere observed in kidneys, liver, spleen or skeletalmuscle of recipient mice, or in tissues of thenegative control mouse.

Fig. 1. Histopathologic changes and EMCV RNA localization in the brain of RAG-1)/) and non-mutant C57BL/6 mice thatreceived myocardial tissue transplants from a D2 EMCV-infected pig. Pieces of porcine myocardium or pancreas collected fromEMCV-infected pigs were transplanted into the abdominal cavity of mice. Paraffin-embedded brain sections were stained withhematoxylin and eosin to show; (A) focal areas of malacia, degenerative/necrotic and swollen axons (arrows) and hemorrhage in thecerebrum of a RAG-1)/) mouse sacrificed 4 days after receiving myocardium from a 2-day EMCV-infected pig, and (B) extensiveinflammatory cell infiltration and hemorrhage in the cerebral cortex of a non-mutant C57BL/6 mouse sacrificed 5 days after receivingmyocardial sections from a 2-day EMCV-infected pig. (C) In situ localization of EMCV RNA (black grains) in a brain section fromthe RAG-1)/) mouse shown in (A), using an 35S-labeled probe corresponding to the VP2 region of EMCV.

Brewer et al.

572

Infectious EMCV isolated from recipient mice

Infectious EMCV was detected by plaque assay inbrain, heart, kidney, liver and skeletal muscle ofboth non-mutant C57BL/6 (n ¼ 4) and RAG-1)/)

(n ¼ 4) mice that received D2 pig myocardial orpancreas tissues. All RAG-1)/) mouse tissues hadinfectious virus titers higher than 5log10 PFU/gramof tissue (Fig. 2A). In contrast, tissues from non-mutant C57BL/6 mice had lower or no infectiousvirus (Fig. 2B). The brain and heart demonstratedthe highest viral titers in either mouse strain,

confirming that these organs are primary sites ofEMCV replication in both pigs and mice.

EMCV-30 RNA detected in RAG-1)/) but not in non-mutantC57BL/6 mice transplanted with tissues from chronically infectedpigs

Primers corresponding to the VP1 and VP2 genesof EMCV were used in nested RT-PCR to detectpresence of viral RNA in recipient mouse tissues[19]. All RAG-1)/) mice (n ¼ 5) that received D2pig xenografts demonstrated EMCV RNA in theheart, brain, liver, kidney, and skeletal muscle(Fig. 3A), whereas viral RNA was less dissemin-ated (found in three of five tissues per mouse) in thenon-mutant C57BL/6 (n ¼ 6) mice that receivedD2 pig xenografts. RAG-1)/) mice transplantedwith myocardial or pancreas tissues from D7 pigs(n ¼ 3) demonstrated viral RNA in the heart,kidney, and liver; whereas C57BL/6 mice trans-planted with the same tissues (n ¼ 3) demonstratedviral RNA only in the heart and brain, which arethe primary sites of viral replication. RAG-1)/)

mice that received D21 (n ¼ 3) and D50 (n ¼ 3)myocardial pig tissues showed viral RNA only inthe brain (Fig. 3B–D) whereas EMCV RNA wasnot detected in non-mutant C57BL/6 mice thatreceived porcine tissue grafts from D21 pigs. Inaddition, no viral RNA was detected in tissues inthe RAG-1)/) mouse that received uninfected pigmyocardial or pancreatic tissue grafts.Using in situ hybridization, EMCV RNA was

detected in multiple tissues of RAG-1)/) mice thatreceived myocardial and pancreas tissues from theD2 EMCV-infected pig. Multiple areas of concen-trated silver grains were observed particularly inthe brain (Fig. 1C), indicating extensive viralreplication and dissemination from the site oftransplantation. EMCV RNA was not detected intissues from RAG-1)/) mice transplanted with D7,D21 and D50 pig tissues.

Infection of recipient mice does not transmit infectious virus tointra-abdominal xenografts

To investigate whether virus-infected transplantrecipients posed a major threat to xenografts,porcine myocardial tissue grafts were transplantedintra-abdominally into RAG-1)/) (n ¼ 2) or non-mutant C57BL/6 (n ¼ 2) mice. Two days aftertransplant, recipient mice were intramuscularlyinoculated (thigh muscle) with 100 PFU ofEMCV-30. All mice developed EMCV diseaseand were moribund 6 to 7 days after infection.The xenografts had no infectious EMCV, suggest-ing that graft-to-recipient transmission is more

Fig. 2. Isolation of infectious EMCV-30 from tissues of non-mutant C57BL/6 and RAG-1)/) mice that received myocardialor pancreatic xenografts from EMCV-infected pigs. (A) Higherthan 5log10 PFU of EMCV-30 per gram of tissue were isolatedfrom brain, liver, kidney, heart, and skeletal muscle of RAG-1)/) mice (n ¼ 4) that received myocardial or pancreatic sec-tions from a pig infected with EMCV for 2 days (D2 pig). (B)In contrast, lower amounts of virus were isolated from thesame organs in non-mutant C57BL/6 mice (n ¼ 4). The highestamount of virus was isolated from the brain and heart tissuesof either mouse strain. No infectious virus was isolated frommice transplanted with tissues from D7, D21, or D50 pigs.

Pig-to-mouse EMCV xenozoonosis

573

efficient than recipient-to-graft in our modelsystem.

Discussion

Our results demonstrate graft-to-recipient virustransmission in a pig-to-mouse xenotransplantationmodel. This model appears ideal for investigatingpotential changes in the pathogenesis of xenozoo-notic viral infections, and for testing therapeutic andprophylactic drugs and vaccines produced to pre-vent and combat such opportunistic infections.Previously, we demonstrated that porcine EMCVcan persist in pig myocardium and brain for up to90 days after infection, and that the virus canproductively infect primary humanmyocardial cells[19]. In this report, we demonstrate that transplan-tation of myocardial or pancreatic tissues from

acutely EMCV-infected pigs (2 days PI) into non-mutant and RAG-1)/) mice of the same geneticbackground resulted in transmission of virus to therecipient animals and acute fatal EMCV-induceddisease in all animals. Clinical disease was firstobservedwithin 2 days inRAG-1)/)mice,whereas 5to 6 days were required for disease development innon-mutant C57BL/6 mice. Of importance was thefact that transplantation of myocardial or pancre-atic tissues from D21 to D50 infected pigs intoRAG-1)/) mice, while not inducing clinical disease,resulted in viral RNA transmission to the brain ofseveral recipient mice, whereas no viral RNA wasdetected in non-mutant C57BL/6 mice that receivedtissues from the same infected pigs. This suggeststhat RAG-1)/) mice may be ideal in investigatingtransmission of low levels of virus typically seen inchronic or persistent infections. The decision to

Fig. 3. Detection of EMCV RNA by RT-PCR in RAG-1)/) mice that received myocardial tissue transplants from EMCV-infectedpigs. Brain, heart, kidney, liver, and skeletal muscle from mice transplanted with tissues from pigs infected with EMCV-30 for 2, 7,21, or 50 days were tested by nested RT-PCR using VP1 or VP2-specific primer sets as described previously [19]. The agarose DNAgel shows VP1 results (436 bp) of tissues from RAG-1)/) mice transplanted with; (A) D2 pig myocardial tissues showing presence ofEMCV RNA in the brain (B), heart (H), kidney (K), liver (L), and skeletal muscle (SM); (B) D7 pig myocardial tissues showingpresence of EMCV RNA in the heart and kidney but negative in the brain, liver and skeletal muscle; (C) D21 pig myocardial tissuesshowing presence of EMCV RNA only in the brain; and (D) D50 pig myocardial tissues showing presence of EMCV RNA only inthe brain.

Brewer et al.

574

terminate the experiment at 14 days after transplan-tation was aimed at identifying the early patterns ofvirus dissemination and histopathologic changeswhich may be lost in a longer experiment. Whetherthe 14-day period allowed after transplantation wassufficient to enable low levels of virus in persistentlyinfected graft tissues to disseminate and induceclinical disease and pathologic changes in recipientmice remains unanswered.The pattern of EMCV dissemination in recipient

mice and in the mice directly inoculated with 100PFU of EMCV-30 was similar, with the highestamount of virus replication and associated patho-logic changes localized in the brain and myocar-dium. A similar pattern was observed in pigsexperimentally infected with EMCV-30; EMCVRNA was detected in brain, heart, liver, spleen,kidney and skeletal muscle but viral antigens andpathologic changes were detected only in the brainand myocardium [19,25]. An interesting but highlyreproducible observation was the different clinicaldisease phenotypes between pigs and mice. EMCV-infected pigs exhibited primarily an acute myocar-ditiswith 10%to20%dying suddenlywithin 3 days,whereasEMCV-infectednon-mutantC57BL/6miceand RAG-1)/)mice developed acute encephalitis. Itis intriguing to speculate on the disease phenotypeEMCV-30 would produce in non-human primates,as this may provide valuable information aboutEMCV pathogenesis in humans.Pig myocardial tissues were selected for trans-

plantation because of the high demand for hearttransplants in humans, and because the myocar-dium is a primary site of EMCV replication andpersistence [19]. Pancreatic tissues were graftedbecause porcine islets are under consideration asa potential treatment for diabetes mellitus inhumans. Few xenotransplantation studies havebeen conducted to date, and a recently proposedclinical trial in the Cook Islands that would studythe effects of transplanting encapsulated pig pan-creatic islet cells into patients with type I diabeteshas met with criticism because of the risk oftransferring porcine viruses, especially endogenousretroviruses [26]. The continuing concern regardingvirus transmission despite the fact that all patientswho have been therapeutically treated with pigtissues or cells have shown no evidence of infectionwith porcine endogenous viruses is particularlyinstructive [15]. It indicates that no amount ofnegative findings can fully assure the publicregarding the risk of pathogen transmission. Thenext step may be for researchers to accept thepossibility that transmission of pig viruses tohumans will occur once the transplantation of pigtissues becomes routine, and therefore begin

research to develop therapeutic and prophylacticstrategies to control such outcomes.Equally important is addressing the issue of

whether human viruses can be transferred to aporcine xenograft resulting in damages that wouldrender the graft ineffective. Coxsackie B virus, ahuman picornavirus related to EMCV, has beenshown to infect porcine cells in vitro [27]. Otherhuman viruses capable of infecting porcine cellsinclude herpes simplex virus and human cytomega-lovirus [28,29]; in addition, human gastroenteritisvirus has been shown to infect pigs, resulting inillness [30]. The risk to xenograft survival in humanrecipients carrying these viruses may be a concernthat will also need to be addressed before pig tohuman transplantation can become a reality. Inour studies, infection of recipient mice did nottransmit EMCV to intra-abdominal xenografts,suggesting that transmission of virus from recipientto graft may not be as efficient as from graft torecipient. However, in whole pig organ transplan-tation into humans, there would be more intimatecontact between recipient and graft involvinganastomosis of blood vessels, which would signi-ficantly increase the risk of recipient-to-graft virustransmission.

Acknowledgments

This work was supported by the National Insti-tutes of Health grant HL04369-01. We thankJeremy Alley, Cristina Marques, HumphreyLwamba, and Rebecca LaRue for their assistance.

References

1. Knowles NJ, Dickinson ND, Wilson G et al. Molecularanalysis of encephalomyocarditis viruses isolated frompigs and rodents in Italy. Virus Res 1998; 57: 53.

2. Seaman JT, Boulton JG, Carrigan MJ. Encephalo-myocarditis virus disease of pigs associated with a plagueof rodents. Aust Vet J 1986; 63: 292.

3. Kim HS, Christianson WT, Joo HS. Characterization ofencephalomyocarditis virus isolated from aborted swinefetuses. Am J Vet Res 1991; 52: 1649.

4. Hirasawa K, Han JS, Takeda M, Itagaki S, Doi K.Encephalomyocarditis (EMC) virus-induced myocarditisby different virus variants and mouse strains. J Vet MedSci 1992; 54: 1125.

5. Helwig FC, Schmidt EH. A filter-passing agent produ-cing interstitial myocarditis in anthropoid apes and smallanimals. Science 1945; 102: 31.

6. Reddacliff LA, Kirkland PD, Hartley WJ, Reece

RL. Encephalomyocarditis virus infections in an Austra-lian zoo. J Zoo Wildl Med 1997; 28: 153.

7. Wells SK, Gutter AE. Encephalomyocarditis virus:epizootic in a zoological collection. J Zoo Wildl Med 1989;20: 291.

Pig-to-mouse EMCV xenozoonosis

575

8. Verlinde JD, van Tangeren HE. Human infections withviruses of the Columbia-SK group. Arch Ges Virusforsch1953; 5: 217.

9. Billinis C, Paschaleri-Papadopoulou E, Psychas Vet al. Persistence of encephalomyocarditis virus (EMCV)infection in piglets. Vet Microbiol 1999; 70: 171.

10. Christianson WT, Kim HS, Joo HS, Barnes DM.Reproductive and neonatal losses associated with possibleencephalomyocarditis virus infection in pigs. Vet Rec1990; 126: 54.

11. Koenen F, de Clercq K, Lefebvre J, Strobbe R.Reproductive failure in sows following experimentalinfection with a Belgian EMCV isolate. Vet Microbial1994; 39: 111.

12. Murnane TG. Encephalomyocarditis. In: Beran GW,sect. ed. CRC Handbook Series in Zoonoses, Section B:Viral Zoonoses, Vol. 2. Boca Raton: CRC Press, 1981.

13. Sanger DV, Rowlands DJ, Brown F. Antibodies in serafrom apparently normal pigs. Vet Rec 1977; 100: 240.

14. Gajdusek C. Encephalomyocarditis infection in child-hood. Pediatrics 1955; 16: 819.

15. Herring C, Cunningham DA, Whitman AJ, Fernan-

dez-Suarez XM, Langford GA. Monitoring xeno-transplant recipients for infection by PERV. Clin Biochem2001; 34: 23.

16. Tucker AW, Galbraith D, McEwan P, Onions D.Evaluation of porcine cytomegalovirus as a potentialzoonotic agent in xenotransplantation. Transpl Proc 1999;31: 915.

17. Gunzburg WH, Salmons B. Xenotransplantation: is therisk of viral infection as great as we thought? Mol MedToday 2000; 6: 199.

18. van der Laan JW, Lockey C, Griffeth BC et al.Infection by porcine endogenous retrovirus after isletxenotransplantation in SCID mice. Nature 2000; 407: 90.

19. Brewer LA, Lwamba HCM, Murtaugh MP et al.Porcine encephalomyocarditis virus persists in pig myo-cardium and infects human myocardial cells. J Virol 2001;75: 11621.

20. Walgenbach S, Rosniatowski R, Bittinger F et al.Modified cryopreservation and xenotransplantation ofhuman parathyroid tissue. Langenbeck’s Arch Surg 1999;384: 277.

21. Sorio C, Bonora A, Orlandini S et al. Successful xeno-grafting of cryopreserved primary pancreatic cancers. VirchArch 2001; 438: 154.

22. Nakanishi R, Hashimoto M, Muranaka H, Yasumoto

K. Effect of cryopreservation period on rat tracheal allo-grafts. J Heart Lung Trans 2001; 20: 1010.

23. van den Broecke R, Liu J, Handyside A et al. Folliculargrowth in fresh and cryopreserved human ovarian corticalgrafts transplanted to immunodeficient mice. Eur J ObsGyn Repr Biol 2001; 97: 193.

24. Njenga MK, Asakura K, Hunter SF et al. The immunesystem preferentially clears Theiler’s virus from the graymatter of the central nervous system. J Virol 1997; 71:8592.

25. Billinis C, Paschaleri-Papadopoulou E, Psychas Vet al. Persistence of encephalomyocarditis virus (EMCV)infection in piglets. Vet Microbiol 1999; 70: 171.

26. Archer K, McLellan F. Controversy surrounds pro-posed xenotransplant trial. Lancet 2002; 359: 949.

27. Spiller OB, Goodfellow IG, Evans DJ, Hinchliffe SJ,Morgan BP. Coxsackie B viruses that use human DAF asa receptor infect pig cells via pig CAR and do not use pigDAF. J Gen Virol 2002; 83: 45.

28. Leventhal JR, Martinez W, Roszkowski J et al. Por-cine cells are susceptible to infection by human alphaherpesviruses. Transpl Proc 2001; 33: 691.

29. Degre M, Ranneberg-Nilsen T, Beck S, Rollag H,Fiane AE. Human cytomegalovirus productively infectsporcine endothelial cells in vitro. Transplantation 2001; 72:1334.

30. Middleton PJ, Petric M, Szymanski MT. Propaga-tion of infantile gastroenteritis virus (orbi-group) inconventional and germfree piglets. Infect Immun 1975;12: 1276.

Brewer et al.

576