Previous Page . Home . Public Message Forum . Send Private Email . Next Page

The Lancet on Xenotransplantation - Proceed with Care


The Lancet (29 August 1998) published three research articles on xenotransplantation, two of which reported no infection in actual cases of xenotransplantation, and one speculating on possible risks. The first two articles were based on the complete absence of pig endogenous retrovirus (PERV) infection in people who have received living pig tissue. The paper that discusses the risks of xenotransplantation is based on speculation and in-vitro observations, not on data showing in-vivo PERV infection in any species. Although there is little in this latter paper that applies to islet xenografts, it is important to note the main conclusions drawn by the researchers:

Our data are strong evidence for a viral risk but we do not know whether the titre will be high enough to bring about productive infection in vivo. Xenotransplant patients, however, will need long-term, intensive immunosuppression, increasing considerably the risk for viral xenozoonosis. Another disadvantage is that grafted organs will probably be obtained from pigs that are transgenic for human complement-regulators which may suppress complement-mediated virolysis,a natural protection against zoonosis.

There are some key points here which do not apply to islet xenografts, and which must be highlighted when discussing any risks associated with xenotransplantation in general:

  1. Islet xenograft recipients will not be immunosuppressed, and will retain a healthy and vigorous immune system.
  2. Islet xenografts will not come from transgenic pigs (as hyperacute rejection is not an issue), and so the added risk suggested in this paper does not exist.
  3. Petri dishes do not have immune systems, and that is the only place where PERV has been shown to infect human cells. Almost any viral or bacterial agent will prove fatal against cells that are not protected by an immune system.
Having said all that, it is important to hear the arguments being used by the anti-xenograft contingent, and to take comfort in the fact that they have failed to demonstrate any meaningful risk despite admirable perseverance. What can they say in the face of the recent findings that show no evidence of PERV infection in recipients of living pig tissue?

The arguments in this paper strengthen the case for islet xenografts being the best first candidate for xenotransplantation. If xenotransplantation proves successful in restoring normoglycemia to people with diabetes, the whole field will be advanced dramatically. Are you listening, Novartis?


Xenotransplantation Articles in The Lancet - 29 August 1998
Xenotransplantation and Risk of Infection
No evidence of infection with porcine endogenous retrovirus in recipients of porcine islet-cell xenografts
No evidence of pig DNA or retroviral infection in patients with short-term extracorporeal connection to pig kidneys
Expression of pig endogenous retrovirus by primary porcine endothelial cells and infection of human cells
No clear answers on safety of pigs as tissue donor source

Other Xenotransplantation Links
The Case for Islet Xenografts being the Best First Candidate for Xenotransplantation
The Xenotransplantation Debate - Science or Superstition?
Xenotransplantation is safe! CDC Report and Latest Xeno Events
BBC News Online -- Pig Viruses Don't Pass to Humans (August 8 1998)

The Journal
Info for Authors
Lancet Interactive
Early reports
Volume 352, Number 9129 
29 August 1998 

Use of animal organs for transplantation into humans-xenotransplantation-could end the worldwide organ shortage. If difficulties with rejection can be resolved, pigs would be the most practical species of animal to use: they are about the right size, are easy to raise, and, until recently, were thought to be fairly free of pathogens that could pose a threat to human beings. However, last year, Clive Patience and colleagues from the UK reported that a virus whose genes are found scattered throughout the pig genome-called the porcine endogenous retrovirus (PERV)-is shed by pig kidney cells and, in cell culture, can infect human cells. The discovery sparked fears that if pig organs were used for transplantation they could introduce new, possibly deadly disease-causing viruses into the human population.

In this week's Lancet, three research groups report new findings on PERV. In the first report, Dr Ulrich Martin and colleagues from Germany report that PERV is also produced by cells from pig aortas, livers, lung, and skin-all tissues that are likely to be used for transplants. The findings, say the researchers, suggest "a serious risk of retrovirus transfer after xenotransplantaion". In the second study, however, Walid Heneine and colleagues from the USA and Sweden report that they found no evidence of PERV infection in blood samples from ten Swedish diabetes patients who had received transplants of insulin-producing cells from pigs-even though the patients had been exposed to large numbers of pig cells (400 million to 2 billion) and had been treated with drugs that should have reduced their ability to resist PERV infection (immunosuppressive drugs). In the final paper, Clive Patience and colleagues from the UK and Sweden write that they could find no evidence of PERV infection in two kidney-failure patients who had their blood passed through pig kidneys. In his Commentary (p 666), Jonathan Stoye from London, UK, argues that only with limited clinical trials, which regulatory authorities are now moving towards permitting, "will it be possible to test long-term xenograft survival and function".

Back to Top

No evidence of infection with porcine endogenous retrovirus in recipients of porcine islet-cell xenografts 
Walid Heneine, Annika Tibell, William M Switzer, Paul Sandstrom, Guillermo Vazquez Rosales, Aprille Mathews, Olle Korsgren, Louisa E Chapman, Thomas M Folks, Carl G Groth

HIV and Retrovirology Branch, National Center for Infectious Diseases, Centers for Disease Control and Prevention, Atlanta, GA, USA (W Heneine PHD, W M Switzer MPH, P Sandstrom PHD, G V Rosales MD, A Mathews BS, L E Chapman MD, T M Folks PHD); Department of Transplantation Surgery, Karolinska Institute, Huddinge Hospital, Stockholm, Sweden (A Tibell MD, Prof C G Groth MD); and Department of Clinical Immunology, Academic Hospital, Uppsala University (O Korsgren PHD)

Correspondence to: Dr Walid Heneine, HIV and Retroviriology Branch, Centers for Disease Control and Prevention, 1600 Clifton Road, NE, mail stop G-19, Atlanta, GA 30333, USA (e-mail:

Patients and Methods


Background The study of whether porcine xenografts can lead to porcine endogenous retrovirus (PERV) infection of recipients is critical for evaluating the safety of pig-to-man xenotransplantation. PERV is carried in the pig germline, and all recipients of porcine tissues or organs will be exposed to the virus.

Methods We studied 10 diabetic patients who had received porcine fetal islets between 1990 and 1993, looking for evidence of PERV infection by using PCR serology, PCR, and reverse transcriptase assays. Prolonged xenograft survival (up to a year) was confirmed in five patients by porcine C-peptide excretion and detection of pig mitochondrial DNA (mtDNA) in serum.

Findings Despite the evidence for extended exposure to pig cells and despite concomitant immunosuppressive therapy, we were unable to detect markers of PERV infection in any patient. Screening for two PERV sequences in peripheral blood lymphocytes collected 4-7 years after the xenotransplantation was negative. Markers of PERV expression, including viral RNA and reverse transcriptase, were undetectable in sera from both early (day 3 to day 180) and late (4-7 years) time points. Western blot analysis for antibodies was consistently negative.

Interpretation These results suggested the absence of PERV infection in these patients. Also this study establishes a minimum standard for post-transplant surveillance of patients given porcine xenografts.

Lancet 1998; 352: 695-99

See Commentary


Pigs are under consideration as a source of organs and tissues for xenotransplants.1 Examples of porcine material already under study are fetal pig pancreatic islet cells and neural cells for insulin-dependent diabetes and refractory parkinsonism2,3 and extracorporeal pig liver perfusion as a bridging strategy to treat liver failure.4 The potential for infection of recipients with xenogeneic agents and the risk of transmission to the general population are major concerns,5 and the risk of xenogeneic infection may be significantly increased by the immunosuppressive regimens required. The risks can be reduced or eliminated by using specific-pathogen-free animal colonies but this approach will not work for the porcine endogenous retrovirus (PERV) because the genome of these viruses is in the germline of every pig. PERV particles are released spontaneously by cell lines originating from pig kidney, lymph node, testis, and fallopian tube.6-8 All PERVs originate from healthy tissues except ones from porcine lymphomas, and even there causation has not been established.6,7 PERV have about 60% sequence homology to the gibbon ape leukaemia and murine leukaemia C-type retroviruses.8-11

Retroviruses result in lifelong infection12 and reports that PERV from cell lines and porcine lymphocytes can infect human cells in vitro8,10 have prompted the US Food and Drug Administration to put porcine xenograft trials on hold until previously exposed patients are assessed for PERV infection and until prospective monitoring of xenograft recipients is established.

We have studied 10 Swedish patients, transplanted with porcine fetal pancreatic islets between 1990 and 1993, for evidence of infection with PERV. These patients have been exposed to a large number of pig cells and xenograft survival has been prolonged. All these patients had a rise in antipig antibody titre within a month of transplantation.13,14

 Patients and Methods


10 patients (mean age 40) with insulin-dependent diabetes of mean duration 30 years and end-stage diabetic nephropathy underwent transplantation with fetal porcine pancreatic islet-like cell clusters (ICC) procured from nontransgenic Swedish cross-bred pigs that had been extensively screened for conventional microbes.2,15 Patients were given about 400 million to 2 billion cells (roughly 2000 per ICC). Maintenance immunosuppression was with cyclosporin, prednisolone, and azathioprine (prednisolone and azathioprine only in one) and at the time of the xenotransplantation, five patients were given adjunctive rabbit-antithymocyte globulin and the other five were given 15-deoxyspergualin.

PCR for proviral PERV DNA

Cryopreserved peripheral blood lymphocytes (PBLs) were lysed with proteinase K and the quality of the lysates was checked by PCR with ß-globin primers.16 Lysates were then tested by PCR for proviral sequences from the gag and pol regions of PERV. Primers and probes were based on conserved sequences identified in our laboratory from a PERV clone derived from Shimozuma-1 pig cells7 or reported from other porcine cells.8,10,11 Wobble bases and inosines were used to accommodate nucleotide variability. A 212 bp sequence is amplified by pol primers* PK15GF2 and PK15GR2, and a 187-bp gag sequence is amplified by PRETF1 and PRETR1.

Lysate containing DNA from 150 000 PBLs was subjected to 35 cycles of amplification followed by Southern blot hybridisation to 32P end-labelled internal oligoprobes PK15GP1 and PRETP2 for the pol and gag sequences, respectively. Negative controls included water and uninfected human PBL lysate. Pig PK15 cell lysates were 10-fold serially diluted with uninfected human PBL lysate to 1·5 and 0·15 PK15 cells as positive controls. The PCR reaction with the DNA equivalent of 0·15 PK15 cell represents the detection limit of the assay.

*Sequences obtainable from WH.

PERV RNA in sera

Nucleic acids were extracted from serum.17 Some was used for PCR analysis of porcine mitochondrial DNA (see below) and the rest was digested with 10 units of RNase-free DNase-I (Boehringer-Mannheim) for 1 h at 37oC in DNase buffer, followed by inactivation by boiling for 5 min. RNA extracts corresponding to 200 µL serum (for samples collected between 4-7 years post-transplant) or 50 µL (for samples collected during the first year) were tested for PERV gag RNA by reverse transcriptase (RT) PCR. Control PCR reactions with no RT excluded the possibility of residual contamination with PERV genomic DNA. Reverse transcription was primed with PRETR1 in the presence of 50 U murine leukaemia virus RT. After incubation at 37oC for 2 h, RT was inactivated and the reaction was amplified in the presence of PRETF1. The PERV gag amplicon was detected by probe PRETP2. Positive controls included DNA-free RNA extracts from tissue culture supernatants from the PK15 or Shimozuma-1 cell lines. PCR controls for DNAse digestion included DNase treated and untreated DNA from pig PF15 cells.

Detection of porcine mtDNA

A 255 bp sequence from porcine mtDNA was amplified in DNA extracts of 25 µL of serum by the oligoprimers* PMTF2 and PMTR2 followed by Southern blot hybridisation to the 32P-labelled internal oligoprobe* PMTP1. Sera from unexposed individuals were used as negative controls. Sera from pigs and DNA lysates of PK15 cells or culture fluids were used as positive controls.

Screening for RT activity in serum

We detected RT using an ultrasensitive PCR assay (Amp-RT)18-20 in duplicate on ultracentrifuged pellets equivalent to 10 µL serum, followed by Southern blot hybridisation. Positive controls were culture supernatants containing PERV (PK-15 cells) and HIV-1, and negative controls included water and a human serum that is antibody negative to all known human retroviruses. Control sera from 35 HIV-1 seropositive patients who had CD4 counts above 500/µL and from 15 US blood donors were also tested by Amp-RT. 12 serum samples from pigs were also analysed for RT activity.

Serological screening for antibodies to PERV

Whole-cell lysates derived from human kidney 293 cells infected with PERV-PK15 were used as a source of PERV antigen.8 Blots were reacted for 3 h at either a 1:50 dilution of patient sera, or a 1:100 dilution of control antisera followed by a 1:7000 dilution of protein A/G horseradish peroxidase for 1·5 h. Blots were visualised by chemiluminescence (ECL western blotting reagents, Amersham). Based on the cross-reactivity between the gag antigens of PERV and simian sarcoma-associated virus (SSAV, a retrovirus highly related to gibbon ape leukaemia virus). A goat anti-SSAV p29 antiserum was used as a positive control.21 This antiserum shows strong reactivity to the PERV p30 found in 293 PERV-PK cells and no reactivity to uninfected 293 cells (see below). 


Evidence for the viability and insulin-producing capacity of the transplanted islet cells is published elsewhere.2,22 Follow-up has been for 4·5-7·5 years. During the first year no patient was admitted to hospital for febrile disease. One patient with chronic asthma was later admitted several times for pneumonia. Six patients have been treated for infectious diabetic ulcers with concomitant local infections. There have been several instances of lower-urinary-tract infection, and one patient was treated for klebsiella septicaemia 4 years after transplantation. Patients XIT2 and XIT1 died of myocardial infarction 2·5 and 5 years after the xenotransplantation. Patient XIT4 lost a renal graft in chronic rejection after 12 years (and 6 years after the xenoislet transplantation). No patient had signs of lympoproliferative or neurological disease of the kind associated with C-type retroviruses in man or animals.2,3

Patient Pig mtDNA*           Transplant characteristics(dagger)  
  2-3 days 2 wk 3 wk 6 mo 1 yr 4-7 yr(ddouble agger) Evidence of Site ICCs (1000s)
              xenoislet survival    
XIT1 + + + + - - C-peptide+ IP 390
XIT2 + - - - NA NA   IP 520
XIT3 - NA - - NA -   IP 460
XIT4 + - - - NA -   IP 410
XIT5 + - - - NA -   IP 330
XIT6 - - - - NA - C-peptide+ IP 520
XIT7 + - - + + - C-peptide+ IP 800
XIT8 + + - + - - C-peptide+ IP 1020
XIT9 - - - - NA -   RC 200
XIT10 - + - + NA NA Biopsy+ RC 410
*PCR results at time after xenotransplantation; NA=samples not available.          
(dagger)IP=intraportal; RC=renal capsule; ICC=islet-like cell clusters; C-peptide+=urinary excretion of porcine C-peptide detected; biopsy+=detection of pig cells under renal capsule in biopsy 3 weeks post-tranplant.
(ddouble agger)4-7 year results are for samples collected in both April and August, 1997.          
Table 1: PCR analysis of pig mtDNA sequences in sera from 10 diabetic patients given pig islet cells

PBLs collected from patients XIT2 and XIT10 at one time point and from the other eight patients at two or three time points 32-86 months post-transplantation were all negative for both gag and pol PERV sequences (table 2, figure 1).
Figure 1: PCR analysis of PBL from pig xenograft recipients for PERV gag and pol sequences

Lanes 1-8=samples collected in August, 1997, from recipients XIT 1 and 3-9; lanes 9-11=PERV-uninfected cellular DNA controls; lanes 12, 13=negative water controls; lanes 14, 15=positive DNA controls from porcine PK15 cell lysates representing DNA equivalents of 1·5 and 0·15 cell, respectively. 

All patients' serum samples collected between 3 days and 7 years after the xenotransplant were negative for PERV RNA.

Post-transplantation Time of sampling PERV sequences*
    gag pol
32-60 mo April, 1995 0/9 0/9
32-60 mo April, 1997 0/8 0/8
32-60 mo April, 1997 0/8 0/8
*Number of positive samples/total tested; sample from XIT6 not available in April, 1995 and XIT2 and XIT10 not available in April and August, 1997
Table 2: PCR analysis of PERV proviral sequences in lymphocytes from 10 diabetic patients given pig islet-cell xenografts between June, 1990 and April, 1993

Sera were also tested for porcine mtDNA, a pig-specific cellular marker that is more sensitive than single-copy genomic sequences. Pig mtDNA was detectable from day 3 in six patients and up to one year in patient XIT7 (table 1). Among patients tested 4-7 years after the xenotranplant, none had detectable pig mtDNA (figure 2). The pig mtDNA PCR signals were strongest at day 3, consistent with higher levels of pig-source cells or cellular products in patients' blood during this early period. The prevalence of detectable pig mtDNA in sera was also highest at 3 days in six of 10 patients and decreased gradually. However, the presence of pig mtDNA in sera of four patients 6 months after transplant argues for successful persistence of pig cells. The inconsistent detection at intermediate time points in three patients probably reflects the low fluctuating levels of porcine mitochondria in the small volumes (25 µL) of sera tested. Transplant technique may also have an influence: porcine mtDNA was detectable at 3 days in six of eight patients who received intraportal transplants but in neither of the two with pig cells implanted under the renal capsule. Monitoring urinary excretion of porcine C-peptide, as a marker of xenograft persistence, was generally concordant with the mtDNA findings. Urinary porcine C-petide was detected for more than 6 months but not in any of the five who were persistently porcine mtDNA-negative after day 3.
Figure 2: PCR analysis of pig xenograft recipients' sera for PERV RNA and pig mtDNA sequences

(A) Lanes 1-5=patient XIT1 at days 3, 14, 26, 194, 478 post-xenograft; lanes 6-9=patient XIT2 at days 3, 17, 24, and 178 post-xenograft; lane 10=human control serum; lane 11=pig serum; lanes 12-14=mtDNA from PK15 tissue culture supernatant diluted 10, 100, and 1000 fold in medium; lane 15=uninfected culture medium control; lane 16=human cellular DNA negative control; lane 17=pig DNA positive control from PK15 cells; lane 18=water as negative reaction control; lane 19=PCR seneitivity control, pig DNA corresponding to 0·015 PK15 cell lysate. 

(B) RT-PCR results of PERV gag RNA sequences in presence of RT. Lanes as in (A) except that lanes 12-14=PERV RNA from PK15 tissue culture supernatant diluted 10, 100, and 1000 fold medium, respectively; lanes 16, 17=DNase treatment control, DNA from 0·15 PK15 cell treated with DNase (lane 16) and DNase-untreated (lane 17); lanes 18, 19=water as negative control. All negative except for positive control in lane 17. RT-PCR results of reaction in (B) in absence of RT not shown. 

RT activity, the presence of which indicates retroviral expression, was not detected in 16 samples obtained from eight patients 4-7 years after the transplant. Eight samples from patients XIT7 and XIT8, 3, 13, 24 days, and about 6 months post-transplant, were also RT-negative (data not shown). Amp-RT testing of 15 HTLV/HIV seronegative controls was negative. RT activity was detected in 75% of sera from HIV-1-infected controls.

Antibodies to p30 PERV protein were not detected in any serum collected around 6 months post-transplantation from 10 patients. Additional samples collected from eight patients from two time points 4-7 years after transplantation were also seronegative. Sera from two pigs also tested negative, confirming immunological tolerance to PERV proteins, as expected with an endogenous virus (figure 3).
Figure 3: Representative western immunoblot analysis of antibodies to PERV in xenograft recipients, non-transplant controls, and pigs

Lanes 1 and 2 are blots from uninfected HK293 cells reacted with pre-immune serum and goat anti-p29 protein of simian sarcoma associated virus (SSAV) serum, respectively. All other lanes represent blots from PERV-infected HK293 cells reacted with: lane 3, pre-immune control serum; lanes 4 and 28, anti-SSAV p29 immune serum; lanes 5-21, sera from porcine xenograft recipients taken 4-7 years post-transplant; lanes 22-25, control sera from unexposed human blood donors; lanes 26 and 27, pig control sera. 

Nine of 12 serum samples from pigs were PERV RNA-positive (figure 2, lane 11). Contamination with genomic DNA was excluded. RT activity was detected in viral pellets of these nine sera but not in the three sera in which PERV RNA sequences were not detected. 


This study is a first step in assessing cross-species transmission of PERV via xenografts from pigs. We sought several markers of persistent PERV infection in sera and cells from these ten patients and used highly sensitive PCR assays. The consistently negative serological results are not likely to be due to compromised humoral immunity because all patients mounted increased titres of antipig antibodies within weeks of transplantation.13,14 PERV can infect human cells, including those of T and B cell lymphocyte origin,8-10 and evidence of extended (more than 6 months) persistence of the pig cells, which all harbour potentially infectious PERV provirus, was found in five patients. Our negative data, with prolonged exposure of immunocompromised patients to large numbers of transplanted porcine cells, thus contribute to the assessment of the ability of PERV to establish persistent infection in man.

The potential for exposure to PERV from the xenograft may be highest during the first 6 months. Viraemia after exposure to other retroviruses (eg, HIV-1) is commonly seen during this time.24-26 However, we found no evidence of PERV RNA in patients' sera during this period, arguing against transient or abortive primary productive PERV infections.

Detection of PERV DNA transcripts in cellular RNA from pig tissues does not show that such expression yields detectable cell-free virus production in serum.8,9 Our findings of both PERV RNA and RT activity in sera from 75% of tested pigs demonstrate productive release of PERV virions into the serum, and imply that any PERV infection in man could also be associated with detectable cell-free virus in serum. These findings highlight the diagnostic importance of using serum PERV RNA and RT levels as markers of PERV expression.

The rapid clearance of porcine cells from five patients may explain the failure of PERV to establish infection. However, excretion of porcine C-peptide by four patients up to 450 days after the transplantation combined with the finding of porcine mitochondrial sequences in serum for 6 months to one year in four patients argue for extended graft survival in at least five patients. Another factor that might influence the potential for PERV to transmit from xenograft to man is a generally low PERV infectivity for human cells, and little or no PERV expression by porcine ICCs.8 We have started to analyse the kinetics of PERV expression in ICCs and preliminary results confirm both PERV RNA and RT activity in the supernatants of cultured fetal pig ICCs after 2 and 4 days in culture. However, the ability of this ICC-derived PERV to infect human cells has not been studied yet. Host factors, including complement-mediated viral lysis, may also protect against PERV infection. Pig-cell-derived PERV are inactivated and lysed by human sera.8 The xenografts were not procured from transgenic pigs carrying human complement-inhibition factors2 and no attempt was made to remove preformed xenoantibodies13,14 or to block complement activation. Complement-based inactivation of PERV released from the pig xenograft should not, therefore, have been compromised.

The absence of detectable RT is consistent with the absence of PERV RNA and, since RT is a generic marker for retroviruses, failure to detect it in these patients' sera points to the absence of any other, unrecognised, retrovirus of porcine origin. Our data cannot definitively exclude infection with retroviruses; nevertheless, the observations are reassuring. The risk that any xenograft recipient will become infected with PERV is likely to be a function of several factors associated with the source animal, xenotransplant technique, and the recipient's characteristics, so defining the risk will be complex. Nor can the results of any individual study be generalised to other types of exposure. However, this is an important initial evaluation of the risk for PERV infection after exposure to cellular xenografts from non-transgenic pigs. Furthermore, this study sets a standard for post-transplantation laboratory surveillance of PERV infection.


Walid Heneine, William Switzer, Paul Sandstrom, Guillermo Vazquez Rosales, Aprille Mathews, Louisa Chapman, and Thomas M Folks were involved in the laboratory study design, execution and analysis. Annika Tibell, Olle Korsgren, and Carl Groth were involved in the clinical management of the xenograft participants and provision of samples. All investigators contributed to the writing of the paper.


Supported in part by the Emerging Infectious Diseases Fellowship Program administered by the US Centers for Disease Control and Prevention and the Association of State and Territorial Public Health Laboratory Directors (ASTPHLD) by the Åke Wiberg Foundation, the Martin Rind Foundation, the Lars-Erik Gelin Memorial Foundation, and the Swedish Medical Research Council. We thank I Suzuka (Tsukuba, Japan) for Shimozuma-1 pig cells and R A Weiss (London, UK) for human kidney 293 cells infected with PERV-PK15. 


1 Dunning JJ, White DJ, Wallwork J. The rationale for xenotransplantation as a solution to the donor organ shortage. Pathol Biol 1994; 42: 231-35.

2 Groth CG, Korsgren O, Tibell A, J, et al. Transplantation of porcine fetal pancreas to diabetic patients. Lancet 1994; 344: 1402-04.

3 Deacon T, Schumacher J, Dinsmore J, et al. Histological evidence of fetal pig neural cell survival after transplantation into a patient with Parkinson's disease. Nat Med 1997; 3: 350-53.

4 Chari RS, Collins BH, Magee JC, et al. Treatment of hepatic failure with ex vivo pig-liver perfusion followed by liver transplantation. N Engl J Med 1994; 331: 234-37.

5 Chapman LE, Folks TM, Salomon DR, Patterson AP, Eggerman TE, Noguichi PD. Xenotransplantation and xenogenic infections. N Engl J Med 1995; 333: 1498-501.

6 Lieber MM, Sherr CJ, Benvensiste RE, Todaro GJ. Biologic and immunologic properties of porcine type C viruses. Virology 1975; 66: 616-19.

7 Suzuka I, Shimizu N, Sekiguchi K, Hoshino H, Kodama M, Shimotohno K. Molecular cloning of unintegrated closed circular DNA of porcine retrovirus. FEBA 1986; 198: 339-43.

8 Patience C, Takeuchi Y, Weiss RA. Infection of human cells by an endogenous retrovirus of pigs. Nat Med 1997; 3: 282-86.

9 Le Tissier P, Stoye JP, Yasuhiro Y, Patience C, Weiss RA. Two sets of human-tropic pig retrovirus. Nature 1997; 389: 681-82.

10 Wilson CA, Wong S, Muller J, Davidson CE, Rose TM, Burd P. Type C retrovirus released from porcine primary peripheral blood mononuclear cells infects human cells. J Virol 1998; 72: 3082-87.

11 Akiyoshi DE, Denaro M, Zhu H, Greenstein JL, Banerjee P, Fishman JA. Identification of a full length cDNA for an endogenous retrovirus of miniature swine. J Virol 1998; 72: 4503-07.

12 Coffin JM. Retroviridae and their replication. In: Fields BN, Knipe DM, Chanock RM, et al, eds. Fields virology, 2nd edn. New York: Raven, 1990: 1437-500.

13 Satake M, et al. Kinetics and character of xenoantibody formation in diabetic patients transplanted with fetal porcine islet cell clusters. Xenotranplantation 1994; 1: 24.

14 Galili U, Tibell A, Samuelsson B, Rydberg L, Groth CG. Increased anti-gal activity in diabetic patients transplanted with fetal porcine islet cell clusters. Transplantation 1995; 59: 1549-56.

15 Bjöersdorff A, Korsgen O, Feinstein R, et al. Microbiological characterization of porcine fetal islet-like cell clusters for intended clinical xenografting. Xenotransplantation 1995; 2: 26-31.

16 Saiki RK, Bugawan TL, Horn GT, Mullis KB, Erlich HA. Analysis of enzymatically amplified B-globin and HLA-DQa DNA with allele-specific oligonucleotide probes. Nature 1986; 243: 163-66.

17 Mulder J, McKinney N, Christopherson C, Sninsky J, Greenfield L, Kwok S. Rapid and simple PCR assay for quantitation of human immunodeficiency virus type 1 RNA in plasma: application to acute retroviral infection. J Clin Microbiol 1994; 32: 292-300.

18 Heneine W, Yamamoto S, Switzer WM, Folks TM. Detection of reverse transcriptase by a highly sensitive assay in sera from individuals infected with the human immunodeficiency virus type 1. J Infect Dis 1995; 171: 1210-16.

19 Garcia Lerma JG, Yamamoto S, Gomez-Cano M, et al. Measurement of human immunodeficiency virus type 1 plasma virus load based on reverse transcriptase (RT). J Infect Dis 1998; 177: 1221-29.

20 Yamamoto S, Folks TM, Heneine W. Highly sensitive qualitative and quantitative detection of reverse transcriptase activity. J Virol Methods 1996; 61: 135-43.

21 Sherr CJ, Fedele LA, Benventiste RE, Todaro GJ. Interspecies antigenic determinants of the reverse transcriptases and p30 proteins of mammalian type C viruses. J Virol 1975; 15: 1440-48.

22 Tibell A, Reinholt FP, Korsgren O, et al. Morphological identification of porcine islet cells three weeks after tranplantation to a diabetic patient. Transplant Proc 1994; 26: 1121.

23 Kaplan JE, Khabbaz RF. The epidemiology of human T lymphotropic virus types I and II. Rev Med Virol 1993; 3: 137-48.

24 Busch MP, Satten GA. Time course of viremia between exposure and seroconversion in health care workers with occupationally acquired infection with human immunodeficiency virus. Am J Med 1997; 102: 117-24.

25 Watson A, Ranchalis J, Travis B, et al. Plasma viremia in macaques infected with simian immunodeficiency virus. J Virol 1997; 71: 284-90.

26 Lutz H, Pedersen NC, Theilen GH. Course of feline leukaemia virus infection and its detection by enzyme-linked immunosorbent assay and monoclonal antibodies. Am J Vet Res 1983; 44: 2054-59. 

Back to Top

No evidence of pig DNA or retroviral infection in patients with short-term extracorporeal connection to pig kidneys 
Clive Patience, Gillian S Patton, Yasuhiro Takeuchi, Robin A Weiss, Myra O McClure, Lennart Rydberg, Michael E Breimer

Section of Virology, Institute of Cancer Research, Chester Beatty Laboratories, London, UK (C Patience PhD, Y Takeuchi PhD, Prof R A Weiss FRS); Department of GU Medicine and Communicable Diseases, Imperial College School of Medicine at St Mary's, London (G S Patton BSc, M O McClure PhD); and Departments of Clinical Chemistry, Transfusion Medicine, and Surgery (L Rydberg MD), Göteborg University, Sahlgrenska Hospital, Göteborg, Sweden (L Rydberg MD, M E Breimer MD)

Correspondence to: Dr Clive Patience, Section of Virology, Institute of Cancer Research, Chester Beatty Laboratories, 237 Fulham Road, London SW3 6JB, UK (e-mail

Patients and methods


Background The xenotransplantation of organs and tissues, in particular those from pigs, is viewed as a means to alleviate the shortage of human donor organs and cells available for transplantation and also as a therapy for other diseases. The potential microbiological hazards of xenotransplantation have recently attracted much attention. One concern is over pig endogenous retroviruses (PERV). Until the possible consequences of infection by PERV are better understood it is unlikely that a significant number of porcine xenotransplants will proceed. However, a small number of patients have already been treated with or exposed to living porcine cells or tissue, and investigation of these patients may provide valuable information.

Methods We took serial blood samples from two renal dialysis patients whose circulation had been linked extracorporeally to pig kidneys and tested them for pig DNA and PERV DNA by nested PCR. The patients' plasma was also tested for neutralising antibodies to two anthropotropic PERV strains.

Findings Having established that the nested PCRs could detect single molecules of target sequence, we analysed DNA isolated from patients' peripheral blood mononuclear cells. We found no evidence of pig or PERV DNA in either patient, even in samples taken as early as 6 h after the perfusion. Furthermore, we found no evidence of seroconversion for PERV-specific antibodies.

Interpretation The absence of porcine cells in the circulation of both patients, even in the samples taken soon after the perfusion experiment, suggests that any porcine cells dislodged from the kidney became rapidly sequestered from the circulation. Since cell-to-cell contact increases the efficency of infection of PERV this removal of porcine cells may increase the risk of transmission of PERV to the xenograft recipient. We did not, however, detect indications of infection by PERV by PCR or neutralisation assay. The genetic and serological methods described here will be useful for detection of possible PERV infection in other patients.

Lancet 1998; 352: 699-701

See Commentary


The potential risk associated with xenografts is the simultaneous transfer of microorganisms which might infect the recipient. Endogenous retroviruses are a special risk because they are part of the normal genetic material of all mammalian cells. Our report1 describing the infection of human cells by a pig endogenous retrovirus (PERV) has focused attention on the fact that careful monitoring of porcine xenotransplant recipients will be necessary. To detect infection by PERV three strategies can be used. Isolation of viable PERV is laborious and insensitive. PCR amplification of PERV DNA is rapid but it is necessary to differentiate between PERV sequences present as part of the genome of porcine cells and those present in the genome of the recipient's cells as a result of virus replication. One approach is to use control PCRs for cellular genes other than PERV;2 identification of non-viral xenograft sequences in a DNA sample would suggest that xenograft cells rather than infected recipient cells are the source of the PERV sequences, whereas detection of viral sequences only would indicate infection by PERV. The third approach is to monitor seroconversion to PERV.

We report sensitive techniques to detect porcine DNA and PERV genomes in two dialysis patients who had short-term extracorporeal vascular connection to pig kidneys.3,4

*Sequences obtainable from CP.

Three porcine sequences were targeted for PCR amplification: the mitochondrial cytochrome oxidase subunit II (COII) gene, the ß-globin gene, and the protease gene of PERV. We also report on an assay for neutralising antibodies to two distinct amphotropic PERV isolates5 in the patients' sera. 

 Patients and methods

Both patients had chronic glomerulonephritis. Patient 1, a 44-year-old man, had a remaining non-functional kidney and was in haemodialysis without immunosuppression. Patient 2 was a 47-year-old man in haemodialysis and in good condition. Both patients were treated by daily filtration plasmapheresis for 3 days to deplete anti-xenoantibodies that react to pig antigens before the perfusion procedure in the absence of immunosuppressive drugs or steroids.3 About 2·5 L of blood passed through the pig kidney over 65 min (patient 1) and 15 min (patient 2).3

PCR primers* were designed to be specific for porcine sequences. Thermal cycling conditions were: 92°C for 4 min, 1 cycle; 94°C for 1 min, 55°C for 1 min 10 s, 72°C for 1 min, 30 cycles; and 72°C for 5 min, 1 cycle (Robocycler, Stratagene). First round (outer) reactions were set up in 50 µL volumes with 1 µL being transferred into the second round (inner) reactions. All DNA samples were tested in triplicate.

To search for neutralising antibodies to PERV envelope proteins, murine leukaemia virus (MLV) vectors bearing PERV env were used. Open-reading frames for PERV-A and PERV-B5 were individually cloned and transfected into TELCeB6 cells, which produce a MLV core containing a LacZ vector genome.6 The human TELCeB6 packaging cell line ensured that Gal(alpha)1-3Gal epitopes, which might have contributed to particle neutralisation, were absent from the vector. Vector particles were harvested in serum-free Opti-MEM and incubated with fresh or heat-inactivated samples of patient's plasma in a 1:1 mixture at 37°C for 1 h. The virus/plasma mixture was then titrated on mink Mv-1-Lu cells.

All DNA samples and PCR reagents were kept separated and in laboratories where porcine DNA and cells had not been handled. Water control amplifications were done in all assays and were always negative. We also differentiated between amplicons arising from genomic sequences and positive control DNAs by sequencing cloned PCR products of the three genes and linearising them with restriction enzymes within the region amplified by the inner primer pair. Nucleotides were then removed from the plasmid ends by nuclease digestion and the deleted plasmids were religated and transformed into bacteria to yield PCR products smaller than those from genomc DNA (figure, upper part).
Representative PCR results

Upper: Comparison of PCR products from genomic and deleted ((Delta)) plasmid controls. Amplifications from DNA of 50 cell equivalents or 50 copies of plasmid DNA. P=PERV protease; C=cytochrome oxidase; ß=ß-globin; M=marker (sizes in base-pairs). All water controls negative. 

Lower: sensitivity of nested PCRs by endpoint dilution of porcine cells. PCRs for deleted control plasmids done on three separate occasions. All water controls negative. Numbers are porcine cells per reaction. 


The primers for porcine sequences did not amplify sequences from human, baboon, or cynomolgus monkey DNA. The deleted clones were also used to determine assay sensitivity. Endpoint dilution of the plasmids in human genomic DNA (105 cell equivalents per reaction) indicated that all assays could detect single molecules of target sequence, and titrations confirmed that the sensitivity was similar for genomic length sequences and for the deleted plasmid constructs.

To determine the number of porcine cells needed in the patient cell preparation to produce a positive result, porcine PBMC were titrated in 2×106 human PBMC, and 105 cell equivalents were analysed by PCR after DNA extraction. The PCR for COII could detect a single porcine cell; the PCRs for ß-globin and PERV required about 20 and 2 cells, respectively (figure, lower part). The results from these controls are similar in sensitivity to those of Heneine and Switzer.2

DNA was obtained from PBMC samples taken 6 h up to 36 months after the extracorporeal perfusion.3 DNA was prepared from 2×106 PHA-stimulated PBMC which had been cultured for up to 4 days. The DNA of 105 cell equivalents was analysed by PCR. Porcine DNA was not detected in the circulation of either patient by any primer pair, even in samples taken only 6 h after dialysis (table). Neither patient's plasma from the perfusion experiment nor control human plasma affected the titre of the PERV A or B vector particles, indicating the absence of neutralising antibodies to these PERV envelopes.

Time post PCR for     Anti-PERV
dialysis PERV COII ß-globin antibodies*
Pre-dialysis . . . . . . --
6 h . . . . . . NT
7 days . . . . . . NT
21/24 mo . . . . . . NT
33/36 mo . . . . . . --
In all cases positive controls indicated that PCR assays could detect near single copy number of target sequence. . .=negative PCR.
*"--"titre after plasma treatment was the range of 70-130% of the titre after treatment with control plasma of culture medium for both PERV-A and PERV-B vectors; NT=not tested.
Analysis of patient PBMC DNAs for porcine sequences and patient plasma for anti-PERV antibodies


If, in a patient with 8 L of blood containing 5×109 leucocytes/L, pig cells are flushed from a xenograft and evenly distributed in the patient's circulation, a positive PCR for porcine mtDNA should be achieved if only 40000 porcine cells, equivalent to only 8 µL of pig blood, were flushed from the kidney. Probably more than 8 µL of pig blood did enter the patients' blood--because both of them mounted a significant humoral response to pig antigens 2-5 weeks post-dialysis.4 The absence of porcine cells from the circulation of both patients suggests that any porcine cells, probably passenger lymphocytes, that were dislodged from the kidney were rapidly sequestered from the circulation. This has been reported in a pig-to-baboon xenotransplantation model.7 Such a filtration would probably result in close contact between xenograft and recipient cells and thus aid the transfer of PERV to a patient's cells.1 The depletion of xenoantibodies by plasmapheresis pretreatment3,4 may also increase the probability of PERV transmission to xenograft recipients.

Our nested PCR systems, which are sensitive enough to pick up single molecules of target porcine sequences, did not detect such sequences in the circulation of two patients who had been temporarily connected to porcine kidneys. Nor did we find evidence of PERV infection either by PCR (although in vitro PBMC are not readily infected1) or by seroconversion. Negative findings on just two patients must be interpreted with caution and, if xenotransplantation is to proceed, it will be important to monitor other exposed individuals. The methods described here will be useful for that purpose.


Clive Patience designed and performed the molecular techniques. Gillian Patton isolated patients' PBMC and standardised all DNA samples for molecular analysis. Robin Weiss, Yasuhiro Takeuchi, and Myra McClure were involved with the design and analysis of the study. Michael Breimer and Lennart Rydberg peformed the operative procedures and archived the patients' samples.


We thank patients for providing serial samples and L H Breimer for suggesting the collaboration between the Swedish and British laboratories. This study was funded in part by the British Medical Research Council, the Jefferiss Trust, and the Swedish Medical Research Council (grants 6521 and 11612). 


1 Patience C, Takeuchi Y, Weiss RA. Infection of human cells by an endogenous retrovirus of pigs. Nat Med 1997; 3: 282-86.

2 Heneine W, Switzer WM. Highly sensitive and specific polymerase chain reaction assays for detection of baboon and pig cells following xenotransplantation in humans. Transplantation 1996; 62: 1360-62.

3 Breimer ME, Björck E, Svalander CT, et al. Extracorporeal ("ex vivo") connection of pig kidneys to humans I: clinical data and studies of platelet destruction. Xenotransplantation 1996; 3: 328-39.

4 Rydberg L, Björck S, Hallberg E, et al. Extracorporeal ("ex vivo") connection of pig kidneys to humans II: the anti-pig antibody response. Xenotransplantation 1996; 3: 340-53.

5 Le Tissier P, Stoye JP, Takeuchi Y, Patience C, Weiss RA. Two sets of human-tropic pig retrovirus. Nature 1997; 389: 681-82.

6 Takeuchi Y, Patience C, Magre S, Banerges PT, le Tissier P, Stoye JP, Weiss RA. Host-range and interference studies on three classes of pig endogenous retrovirus. J Virol (in press).

7 Hoopes CW, Platt JL. A molecular epidemiological probe for pig microchimerism. Transplantation 1997; 64: 347-50. 

Back to Top

Expression of pig endogenous retrovirus by primary porcine endothelial cells and infection of human cells 
Ulrich Martin, Verena Kiessig, Jürgen H Blusch, Axel Haverich, Klaus von der Helm, Tanja Herden, Gustav Steinhoff

Leibniz Research Laboratories for Biotechnology and Artificial Organs, Hannover Medical School, Hannover, Germany (U Martin PhD, V Kiessig, A Haverich MD, T Herden, G Steinhoff MD); Department of Thoracic and Cardiovascular Surgery, Hannover Medical School (A Haverich, G Steinhoff); and Pettenkofer-Institute, Munich (J H Blusch PhD, Prof K von der Helm MD)

Correspondence to: Dr Ulrich Martin, Leibniz Research Laboratories for Biotechnology and Artificial Organs, Hannover Medical School, Podbielskistr 380, D-30659 Hannover, Germany (e-mail:

Materials and methods
Hot Paper Discussion


Background The risk of interspecies transmission of retroviruses during xenotransplantation is suggested by reports of pig endogenous retrovirus (PERV) released from porcine cell lines productively infecting human cell lines in vitro and of infectious PERV being released from pig peripheral blood mononuclear cells after mitogenic stimulation. Endothelial cells are the main interface between a xenograft and the recipient's leucocytes and tissues.

Methods We have analysed pig primary aortic endothelial cells (PAEC) together with other transplantation-relevant porcine cells and tissues for expression of PERV mRNA. Release of virus particles by PAEC was monitored by reverse transcriptase (RT) activity in the medium of cultured PAEC. Infectivity for human cells was tested by co-cultivation of irradiated PAEC with the human embryonal kidney cell line HEK293 and looking for virus release from the human cells.

Findings PAECs, hepatocytes, lung, and skin from a variety of pig strains and breeds expressed PERV mRNA. PAEC released infectious particles. Co-cultivation of PAEC and HEK293 led to productive infection of the human cells and expression of PERV types A and B.

Interpretation Release of infectious virus from PAEC occurred without mitogenic stimulation, suggesting a serious risk of retrovirus transfer after xenotransplantation.

Lancet 1998; 352: 692-94

See Commentary


A shortage of human organs for transplantation has led to attempts at xenotransplantation. Primate donors carry the risk of transmission of potentially pathogenic retroviruses. Pigs have been considered suitable and until recently were thought to pose a low risk of transmission of retroviruses. However, Patience et al1 demonstrated the release of pig endogenous retrovirus (PERV) particles from porcine kidney-cell lines and their infectivity for human cell lines. At least two infectious variants (PERV A and B), differing in their envelope proteins, have been detected in pig heart, spleen, and kidney.2 Wilson et al extended this finding to pig peripheral blood monocytes (PBMC), which, after mitogenic stimulation, produced virus infectious for human embryonal kidney (HEK) 293 and hela cell lines.3 These findings have provoked serious debate about the safety of xenografts.4-8 Since porcine aortic endothelial cells (PAEC) will be the main interface between the xenograft and the recipient's leucocytes and tissues after transplantation, we decided to find out whether these cells can release PERV particles and whether pigs of different breeds or origin express endogenous retroviruses to a similar extent. 

 Materials and methods

Culture of endothelial cells

Endothelial cells were isolated by incubating the inner layer of pig aortas or human veins with 0·2% collagenase type A for 20 min at 37°C. Detached cells were washed twice in phosphate-buffered saline and cultured in either M199 (Biochrom, Berlin) plus 20% fetal calf serum (FCS) (Life Technologies, Eggenstein) and EC growth factor (Boehringer Mannheim) or EGM 2 (Clonetics) plus 15% FCS. Cells were harvested by incubation with 0·05 mmol/L trypsin/0·02 mol/L EDTA. HEK293 cells and the co-cultivated cells were cultured in M199 or RPMI1640 (Biochrom) including 15% FCS. PK15 cells were cultivated in Dulbecco's modified Eagle medium (DMEM) and Ham's F12 1:1 (Life Technologies) plus 10% FCS.

PCR detection of PERV

106 cells were lysed in 100 µL of 200 µg/mL proteinase K in PCR buffer for 3 h at 56°C, followed by 10 min inactivation at 95°C. 3·5 µL of this crude extract served as template for PCR with PERV pol, env A, and env B specific primers.1,2 ß-globin-specific controls were used.9,10 ß-actin specific primers* were used for internal positive controls. PCR sensitivity was determined by mixing different quantities of lysed PK15 cells with 106 lysed HEK 293 cells.

*Sequences obtainable from UM.

DNA sequence analysis

Direct sequencing on both strands of purified PCR products was carried out on an ABI 377 DNA sequencer using TaqFS-DyeDeoxyTerminator sequencing technology (Perkin Elmer/Applied Biosystems, Weiterstadt). Sequences were corrected with the SeqEd program (Applied Biosystems) and compared with sequences from the database.

Detection of PERV-specific mRNA

Total RNA was prepared using TriReagent (Molecular Research Center, Cincinatti). Contaminating DNA was digested by DNase for 15 min at 37°C, followed by a phenol/chloroform-extraction. After precipitation, 2 µg RNA was used for random-primed cDNA synthesis with avian myeloblastosis virus (AMV)-RT (Boehringer Mannheim). PERV pol and env specific PCRs were then done. GAPDH specific primers* were used for internal positive controls.

RT activity in cell culture supernatant

We used published methods1,11 with some modifications. Retroviruses were isolated by ultracentrifugation of 3 mL of 0·45 µm filtered supernatant. The pellet was resuspended in 20 µL 1% NP40 and 5 µL was then assayed by RT-PCR. After reverse transcription (sequence of RT-primer as in Silver et al11) and denaturation the BMV-RNA was digested by RNase H for 15 min at 37°C, followed by PCR-based cDNA amplification.11 18 µL of PCR product was loaded onto 2% agarose. After separation the DNA was blotted in 20×SSC on a nylon membrane (Hybond N, Amersham), denatured, and cross-linked by ultraviolet irradiation.

An internal BMV-specific oligonucleotide* was labelled (DIG-oligonucleotide tailing kit, Boehringer Mannheim). The blot was hybridised in 5×SSC, 0·1% N-laurylsarcosine, 0·02% SDS, 1% blocking reagent (Boehringer Mannheim), 0·1 mg/mL Poly(A) (Boehringer Mannheim) at 50°C overnight with the digoxigenin (DIG)-tailed oligonucleotide, washed twice for 5 min at room temperature and twice for 15 min at 62°C in 2×SSC. Chemoluminescence was detected with the DIG luminescence kit (Boehringer Mannheim).

In vitro infection of HEK293 cells

2×105 PEAC were X-irradiated (100 Gy), mixed with 105 non-irradiated HEK293 cells, and passaged every 3-5 days. PERV infection and expression were monitored 20 or more days after the beginning of the cocultivation by PCR and RT-PCR as described above. 


To test a representative variety of transplantation-relevant porcine cells and tissues for PERV RNA expression, total RNA was prepared from primary endothelial cells, hepatocytes, lung, and skin from minipigs, Yucatan micropigs, and German landbreed pigs obtained from 12 sites in Germany, Denmark, Russia, and France. PERV pol RNA was detected in all samples (figure 1, upper line). Porcine PK15 and human endothelial cells served as positive and negative controls. To exclude contamination of genomic DNA, we did all RT-PCRs in parallel without added RT and detected no signals (figure 1, middle line).
Figure 1: Expression of PERV mRNA by PAEC and porcine hepatocytes, lung, and skin

Upper: PERV expression analysed by pol specific RT-PCR. 

Middle: Internal controls without RT excluded contamination with genomic DNA. 

Lower: GAPDH-specific RT-PCR was positive control. 

One representative RT-PCR result shown for eight landbreed pigs from Germany. HuEC=human endothelial cells. 

We tested 10 preparations of PAEC from different pig strains and breeds after at least three passages for release of PERV into the medium of cultured cells. RT-activity was found in all PAEC supernatants tested (figure 2). Human primary-endothelial-cell controls did not release RT activity.
Figure 2: RT activity in culture supernatant of PAEC

Human endothelial cells (HuEC) as negative control. One representative RT-PCR result shown for six landbreed pigs from Germany. 

After cocultivation of X-irradiated PAEC with HEK293 cells PCR of the HEK293 cell DNA yielded a strong signal for PERV sequences (figure 3). HEK293 cells served as negative control. Direct sequencing of the env PCR products revealed type A and type B variants, identical to those published by Le Tissier et al.2 The nucleotide sequence for PERV pol-specific PCR fragments was also identical to published sequences.
Figure 3: PERV DNA sequences in human HEK293 cells after cocultivation with PAEC

Transmission of PERV detected by PCR specific for PERV pol, env A, env B, and porcine ß-globin. PAEC alone and HEK 293 cells alone are controls; ß-actin is internal control. 

Contamination of pig genomic DNA from pig cells remaining alive throughout the cocultivation was excluded by negative PCR with porcine specific ß-globin (figure 3) and (alpha)-galactosyltransferase specific primers10 (not shown) PERV pol-specific PCR could detect one PK15 cell in 105 HEK 293 cells; with porcine ß-globin and (alpha)-galactosyltransferase specific primers, one PK15 cell could be detected in 104 HEK293 cells (not shown). PERV-specific sequences could also be detected with newly designed PERV LTR/leader primers with 5-10 times lower sensitivity than the porcine ß-globin specific primers (not shown).

We also examined PERV RNA expression in cocultured HEK293 cells by RT-PCR and detected PERV type A and B mRNA (figure 4). Internal controls without RT excluded contamination with pig genomic DNA. Sequence analysis of the pol cDNA fragment yielded a sequence identical to the published pol sequence.1 Strong RT activity in the supernatant of the cultured cells (figure 4) confirmed productive infection of HEK293 cells.
Figure 4: Productive infection of HEK293 after cocultivation with PAEC

Left: PERV expression by RT-PCR specific for PERV pol, env A, and env B. Internal controls without RT (second row) excluded contamination with genomic DNA. GAPDH specific RT-PCR as positive control. (a) HEK293 cells only, (b) HEK293/PAEC cocultivation. 

Right: RT activity in supernatant of HEK293 cells subsequent to cocultivation with PAEC (b). Uninfected HEK293 cells as negative control (a). 


The promise of pig xenografts suffered a disappointment when PERV were found to infect human cell lines. Transfer of infectious endogenous virus from porcine to human cell lines in vitro does not, however, closely mirror the situation in vivo. A more recent publication on primary pig PBMCs (inevitably present in xenografts) revealed the release of PERV, after mitogenic stimulation, that were infectious for human cell lines. A closer simulation of xenotransplantation is to monitor vascular endothelial cells, the interface between a xenograft and the recipient's cells, and this is what we have done. We found that pig endothelial cells, without mitogenic stimulation, expressed PERV RNA and released viral particles. Minute stimulation by mitogenic factors in the fetal calf serum cannot be excluded in cell culture. These particles were productively infectious for the human cell line HEK293. Cultured hepatocytes, lung, and skin were also positive for PERV mRNA, confirming and extending the recent report of PERV expression in porcine kidney, heart, and spleen.2

In our cocultivation experiments, contamination with pig cells, still viable despite the X-irradiation, would have caused false-positive PERV results but we excluded this by control PCRs with pig-specific ß-globin and (alpha)-galactosyltransferase primers. Quantitative comparison of the PERV amplification products with the results of endpoint dilution experiments (PERV pol/LTR-leader vs porcine D-galactosyltransferase and ß-globin specific PCR) ruled out residual contamination by porcine DNA or persistent PAEC in the human cells.

Our results confirm recent reports of the activation of infectious endogenous pig virus and extend them to pig endothelial cells, a constitutive part of all vascularised xenografts. The high number of PERV gene copies (about 10-40) in the porcine genome2,3 will make the task of producing PERV-free swine difficult or impossible. Our data are strong evidence for a viral risk but we do not know whether the titre will be high enough to bring about productive infection in vivo. Xenotransplant patients, however, will need long-term, intensive immunosuppression, increasing considerably the risk for viral xenozoonosis. Another disadvantage is that grafted organs will probably be obtained from pigs that are transgenic for human complement-regulators which may suppress complement-mediated virolysis,12 a natural protection against zoonosis. If transmission of PERV does occur clinically, two outcomes are possible--either the retrovirus is pathogenic but is restricted to the recipient or the virus (pathogenic or not) could infect people not participating in the xenotransplantation process. Recombination with human retroviral elements might enhance the risk. Further studies, in suitable primate models, are thus necessary to assess the risk of PERV infection in vivo after xenotransplantation.


U Martin designed the study, did RT assays, and wrote the paper.
V Kiessig did PCR assays and cell cultures, J H Blusch did PCRs and sequencing, A Haverich, K von der Helm, and G Steinhoff designed the study and wrote the paper, and T Herden did cell cultures.


Irradiation of porcine cells was done in the Department of Strahlentherapie und spezielle Onkologie, MHH. We thank R A Weiss and C Patience for valuable information and helpful discussion and Jean Deinhardt for her help in reading this manuscript. This work was supported by the Deutsche Forschungsgemeinschaft (Hav 1306/3-1). 


1 Patience C, Takeuchi Y, Weiss RA. Infection of human cells by an endogenous retrovirus of pigs. Nat Med 1997; 3: 282-86.

2 Le Tissier P, Stoye JP, Takeuchi Y, Patience C, Weiss RA. Two sets of human-tropic pig retrovirus. Nature 1997; 389: 681-82.

3 Wilson CA, Wong S, Muller J, Davidson CE, Rose TM, Burd P. Type C retrovirus released from porcine primary peripheral blood mononuclear cells infects human cells. J Virol 1998; 72: 3082-87.

4 Bach FH, Fineberg HV. Call for moratorium on xenotransplants. Nature 1998; 391: 326.

5 Vogel G. No moratorium on clinical trials. Science 1998; 279: 648.

6 Weiss RA. Transgenic pigs and virus adaptation. Nature 1998; 391: 327-28.

7 Editorial. Halt the xeno-bandwagon. Nature 1998; 391: 309.

8 Editorial. Does biomedical research need another moratorium? Nat Med 1998; 4: 131.

9 Heneine W, Switzer HM. Highly sensitive and specific polymerase chain reaction assays for detection of baboon and pig cells following xenotransplantation in humans. Transplantation 1996; 62: 1360-62.

10 Hoopes CW, Platt JL. A molecular epidemiological probe for pig microchimerism. Transplantation 1997; 64: 347-50.

11 Silver J, Maudru T, Fujita K, Repaske R. An RT-PCR assay for the enzyme activity of reverse transcriptase capable of detecting single virions Nucleic Acids Res 1993; 21: 3593-594.

12 Saifuddin M, Parker CJ, Peeples ME, et al. Role of virion-associated glycosylphosphatidylinositol-linked proteins CD55 and CD59 in complement resistance of cell line-derived and primary isolates of HIV-1. J Exp Med 1995; 182: 501-09. 

Back to Top

No clear answers on safety of pigs as tissue donor source 
Xenotransplantation shows great promise of providing a virtually limitless supply of cells, tissues, and organs for a variety of therapeutic procedures.1 Provided that ethical concerns can be met, there are three major challenges that are crucial for the ultimate success of xenotransplantation. First, immunological rejection must be overcome. Second, there must be no fundamental physiological incompatabilities between xenograft and recipient. Third, the transmission of animal pathogens to the human recipients and from them to the human population at large must be prevented. The risk of zoonoses arising as a result of xenotransplantation has been much discussed over the past 4 years.1,2 Three Early Reports in this issue of The Lancet address this topic, providing data that all sides in the debate will take as supporting their position.

Experience with allotransplantation has provided ample evidence that donor organs can be a source of infection.3 This point has focused attention on the microbiological status of potential source animals, which is one reason for favouring pigs over primates for this role.4 However, even the advanced techniques developed to breed high-health-status pigs will not eliminate pathogens such as those capable of infecting fetuses congenitally or transmitted in the germline. Although congenitally acquired pathogens might prove important, it is the latter group of agents, known as the endogenous retroviruses, that are believed to pose the greatest threat of infectious disease to xenotransplantation. These viruses consist of retroviral elements that are inherited in a mendelian fashion after infection of germ cells during evolution.5 All vetebrates contain thousands of these elements. Most of the elements, apparently including all those present in human beings, seem defective, but some, including proviruses present in mice, cats, and chickens, will give rise to infectious retroviruses. These viruses can, but need not, cause "spontaneous" neoplasms in their host species.

Extrapolating from these observations, several researchers have suggested that porcine endogenous retroviruses (PERVs) might be an obstacle in xenotransplantation.6-8 To constitute a serious threat there would probably have to be a chain of events (panel). There is now substantial evidence for the first three events in this chain.9-12 Two sets of pig retrovirus, PERV-A and PERV-B, capable of in-vitro replication in certain human cell lines, have been described. They are widely distributed in different pig breeds and expressed in different tissues, including spleen, kidney, and heart. The paper by Ulrich Martin and colleagues adds further weight to these conclusions. These investigators add four more cell or tissue types (aortic endothelial cells, hepatocytes, skin, and lung) to the list of PERV mRNA expressors and also show that primary aortic endothelial cells from several different pig breeds produce infectious PERV-A and PERV-B. Briefly described in the paper by Walid Heneine and colleagues are preliminary data indicating that fetal porcine islet cells also express PERVs. Because of the way these elements are inherited in pigs, they will be virtually impossible to eliminate from source herds. Taken together, these data show that most if not all transplanted porcine tissues will express one or more retroviruses capable of infecting human cells.


Does the presence of potentially infective retroviruses in procine tissue mean that infection of human beings exposed to porcine cells will inevitably follow? Several groups are investigating this question; the papers by Heneine and Clive Patience and their colleagues represent the first published reports of such studies. Both groups have developed sensitive methods for detecting PERV infection in vivo by use of PCR and assays for seroconversion. Heneine and colleagues examined ten diabetic patients who received porcine fetal islets. In no case was there any evidence for PERV infection, even though the xenograft survived a long time in five of the patients. Patience and colleagues examined two renal-dialysis patients with short-term extracorporeal vascular connnection to pig kidneys. Neither showed any sign of viral infection. Since rates of infection are likely to depend on a variety of factors--including the pig source, whether or not the pig carries transgenes designed to regulate complement activation,13 the nature of the xenograft, and the degree of immunosuppression--care must be taken not to draw too broad a set of conclusions from these two studies. Nevertheless, it seems reasonable to conclude that the PERVs will not show the very high levels of transmission associated with some viruses. Some unpublished studies involving other examples of exposure to pig tissues seem to be reaching the same conclusion. These studies ought to be made available for public scrutiny in the near future.

So far no conclusions can be drawn about the potential pathogenicity of PERVs if cross-species infection occurs. It has been suggested that trials in non-human primates might shed some light on this question; unfortunately recent results indicate that PERV-A and PERV-B will not recognise receptors on cells from non-human primates thereby providing evidence against the value of such studies.14

What is the significance of the findings reported today for the future of xenotransplantation? Some people will undoubtedly be concerned by the results of Martin and colleagues. They will argue that, so long as porcine xenografts express retroviruses capable of infecting human cells, clinical trials should not be allowed to proceed because any risk posed by these elements is unacceptably large in the absence of information about the pathogenic potential of the PERVs. Others will be reassured by the results of Heneine and Patience. They will argue that any such risks are worth taking, given the promise of xenotransplantation and the apparent low levels of infectivity of the PERVs.

The regulatory climate is moving toward permitting limited clinical trials in the near future, especially in the USA. Only with such trials will it be possible to test long-term xenograft survival and function. However, xenotransplantation is a term covering a wide range of different procedures; precautions that may be appropriate in one case may not be sufficient in another, and consequences that would be perfectly acceptable for individual patients would be unthinkable if they were to affect the general population. The clinical trials must therefore be conducted in a deliberate, stepwise fashion, using the recently developed assay methods to monitor PERV infection and to assess whether transmission to contacts can occur. These trials must involve close liaison between surgeons, immunologists, and virologists, and include lifetime monitoring by regulatory authorities.

Jonathan Stoye

Division of Virology, National Institute for Medical Research, London NW7 1AA, UK

1 Advisory Group on the Ethics of Transplantation. Animal tissues into humans. Norwich, UK: Stationery Office, 1997: 1-258.

2 Bach FH, Fishman JA, Daniels N, et al. Uncertainty in xenotransplantation: individual benefit versus collective risk. Nat Med 1998; 4: 141-44.

3 Michaels MG, Simmons RL. Xenotransplant-associated zoonoses. Transplantation 1994; 57: 1-7.

4 Fishman JA. Miniature swine as organ donors for man: strategies for prevention of xenotransplant-associated infections. Xenotransplantation 1994; 1: 47-57.

5 Boeke JD, Stoye JP. Retrotransposons, endogenous retroviruses, and the evolution of retroelements. In: Coffin JM, Hughes SH, Varmus HE, eds. Retroviruses. Cold Spring Harbor: Cold Spring Harbor Press, 1997: 343-435.

6 Smith DM. Endogenous retroviruses in xenografts. N Engl J Med 1993; 328: 142-43.

7 Stoye JP, Coffin JM. The dangers of xenotransplantation. Nat Med 1995; 1: 1100.

8 Chapman LE, Folks TM, Salomon DR, Patterson AP, Eggerman TE, Noguchi PD. Xenotransplantation and xenogeneic infections. N Engl J Med 1995; 333: 1498-501.

9 Patience C, Takeuchi Y, Weiss RA. Infection of human cells by an endogenous retrovirus of pigs. Nat Med 1997; 3: 282-86.

10 Wilson CA, Wong S, Muller J, Davidson CE, Rose TM, Burd P. Type C retrovirus released from porcine primary peripheral blood mononuclear cells infects human cells. J Virol 1998; 72: 3082-87.

11 Le Tissier P, Stoye JP, Takeuchi Y, Patience C, Weiss RA. Two sets of human-tropic pig retrovirus. Nature 1997; 389: 681-82.

12 Akiyoshi DE, Denaro M, Zhu H, Greenstein JL, Banerjee P, Fishman JA. Identification of a full-length cDNA for an endogenous retrovirus of miniature swine. J Virol 1998; 72: 4503-07.

13 Weiss RA. Transgenic pigs and virus adaptation. Nature 1998; 391: 327-28.

14 Takeuchi Y, Patience C, Magre et al. Host-range and interferon studies on three classes of pig endogenous retrovirus. J Virol (in press). 

Back to Top

Previous Page . Home . Public Message Forum . Send Private Email . Next Page