ELIMINATION OF PORCINE ENDOGENOUS RETROVIRUSES FROM PIG CELLS

Recently, a group of scientists under the guidance of George Church at Harvard Medical School published in “Science” that it succeeded to inactivate 62 porcine endogenous retroviruses (PERVs) in the genome of a pig cell line [1]. At present, PERVs still represent the main microbiological risk for xenotransplantation. Virus-encoding gene segments (so-called proviruses) are integrated in the genome of all pigs, and can be expressed as infectious virus particles and can infect certain human cells in vitro (for review see [2]). They cannot be eliminated by breeding as it is possible for all other exogenous porcine micro-organisms such as hepatitis E virus or porcine herpesviruses.

Thus far, in vivo transmission has not been documented despite many studies on this topic. In the first clinical xenotransplantation trials including more than 200 patients, transplanting pig islet cells for the treatment of diabetes [3] or other cells (for review see [4]), no transmission of PERV was observed. In numerous pre-clinical trials transplanting pig cells or organs into non-human primates, no PERV transmission was documented, and there was no evidence for viral infection upon inoculation of highly concentrated preparations of virus particles with replication potential into small animals or non-human primates, with and without immunosuppression (for review see [2]).

In this context, it is certainly interesting to learn that feline leukaemia virus (FeLV), a retrovirus very closely related to PERV, did not infect human cat owners, despite scratches and bites, the easiest way to transmit retroviruses [5]. However, it is relevant for the safety evaluation of xenotransplantation that in almost all clinical xenotransplantation trials no immunosuppression was applied and only a small number of cells were transplanted. In animal transplantation and infection experiments, the recipient species of the pig material, for example non-human primates or rats, do not have a fully functional receptor for PERV [6,7]. And in the case of FeLV, most cat owners are assumed to be healthy with a good working immune system.

Therefore, the publication by Yang et al. [1] showing inactivation of multiple PERVs in the pig genome was applauded by many newspapers as a “breakthrough” in xenotransplantation including commentaries by experts. Prior to this study, several other strategies to prevent PERV transmission had been developed, including vaccination [8,9] and RNA interference [10–13]. As PERV-specific small inhibitory (si)RNAs only reduced the expression of PERV in the transgenic animals [10,12], it was questioned whether “gene editing” using nucleases cutting PERV-specific sequences can eliminate all PERVs integrated in the genome. These nucleases induce double-strand breaks in the virus sequence, which will be repaired by the cell. In most cases, this repair results in mistakes such as deletions or insertions and leads to an inactive gene.

A first attempt to eliminate PERVs from the genome by gene editing was made using the zinc finger nucleases (ZNF), and this approach unfortunately failed [14]. In this experiment, a PERVreleasing pig embryonic kidney cell line (PK15) and PERV-infected human tumour cells were used, and the ZFN was directed against a highly conserved region in the pol gene of PERV [14]. Pol encodes the reverse transcriptase, the key enzyme of retroviruses responsible for the transcription of the viral RNA genome into a DNA copy which will be integrated into the cell genome.

Obviously, multiple PERV-specific cuts by the ZFN destabilized the genome and the cells died [14]. Yang et al. [1] also used the PERV-producing PK15 porcine cell line, and also targeted a highly conserved sequence in pol, but used another nuclease for the specific cutting of the integrated PERV sequences, namely CRISPR/Cas9. Upon introduction of CRISPR/Cas9 by a transposon or lentiviral vector into PK15 cells, Yang et al. [1] succeeded in inactivation of 62 proviruses in the pig genome. Gene editing was reported to be associated with off-target effects, but the authors did not detect unexpected genomic rearrangements in the treated cells, indicating that the treatment did not cause genomic instability. This is of great importance for the generation of healthy pigs.