Extended analysis of the in vitro tropism of porcine endogenous retrovirus

Abstract

We previously reported that mitogenic activation of porcine peripheral blood mononuclear cells resulted in production of porcine endogenous retrovirus(es) (PERV[s]) capable of productively infecting human cells (C. Wilson et al., J. Virol. 72:3082-3087, 1998). We now extend that analysis to show that additional passage of isolated virus, named here PERV-NIH, through a human cell line yielded a viral population with a higher titer of infectious virus on human cells than the initial isolate. We show that in a single additional passage on a human cell line, the increase in infectivity for human cells is accounted for by selection against variants carrying pig-tropic envelope sequences (PERV-C) as well as by enrichment for replication-competent genomes. Sequence analysis of the envelope cDNA present in virions demonstrated that the envelope sequence of PERV-NIH is related to but distinct from previously reported PERV envelopes. The in vitro host range of PERV was studied in human primary cells and cell lines, as well as in cell lines from nonhuman primate and other species. This analysis reveals three patterns of susceptibility to infection among these host cells: (i) cells are resistant to infection in our assay; (ii) cells are infected by virus, as viral RNA is detected in the supernatant by reverse transcription-PCR, but the cells are not permissive to productive replication and spread; and (iii) cells are permissive to low-level productive replication. Certain cell lines were permissive for efficient productive infection and spread. These results may prove useful in designing appropriate animal models to assess the in vivo infectivity properties of PERV.

FIG. 1

Detection of PERV envelope classes by RT-PCR of virion RNA. Viral RNA was examined by RT-PCR from cell lysate derived from unactivated pPBMC (lane 1) or pPBMC activated with PHA and PMA (lane 2) (18) or pelleted virions of PERV-NIH-1° (lane 3) or PERV-NIH-2° (lane 4), using primers specific for each of the envelope classes indicated. Each primer-specific reaction was fractionated on a 1% agarose gel, transferred to nitrocellulose, and then hybridized to probes specific for each PERV envelope class listed on the left (see Materials and Methods). The hybridized products were visualized with a PhosphorImager.

FIG. 2

Comparison of amino acid residues encoded by clone 1.15-derived cDNA of 293/PERV-NIH-1° virions (PERV-1.15), PERV-A, and PERV-C. (A) Schematic representation of regions of the PERV-1.15 envelope homologous to PERV-A (hatched) and PERV-C (open) genes. (B) Deduced amino acids for the envelope surface glycoprotein (SU; amino acids 1 to 460) and transmembrane (TM; amino acids 461 to 659) regions of clone 1.15 derived from cDNA of PERV-NIH-1° virions (GenBank accession no. AF130444) compared to PERV-A and PERV-C. Identical amino acids are denoted by dots; gaps are denoted by dashes.

FIG. 3

RT activity observed in cocultures of 293/PERV-NIH-2° producer cells and human hematopoietic cell lines. Data points represent [³H]TTP incorporated in an RT assay measured in cell supernatants sampled at the times indicated after coculture of 293/PERV-NIH-2° producer cells with each of the cell lines indicated. Values of [³H]TTP incorporation for parallel mock-infected controls were subtracted from the values obtained in matched cocultures.