Isolation and Characterization of an Infectious Replication-Competent Molecular Clone of Ecotropic Porcine Endogenous Retrovirus Class C

Abstract

Xenotransplantation of pig organs is complicated by the existence of polytropic replication-competent porcine endogenous retroviruses (PERV) capable of infecting human cells. The potential for recombination between ecotropic PERV-C and human-tropic PERV-A and PERV-B adds another level of infectious risk. Proviral PERV-C were characterized in MAX-T cells derived from d/d haplotype miniature swine. Three proviruses were cloned from a genomic library. Clone PERV-C(1312) generated infectious particles after transfection into porcine ST-IOWA cells. Electron microscopy revealed the same morphologies of virions in MAX-T cells and in ST-IOWA cells infected with cell-free PERV-C(1312) particles, indicating that MAX-T cells harbor one functional PERV-C provirus.

Pig-to-human xenotransplantation encompasses the potential transmission of, e.g., viruses present in the donor species (16, 17, 19). Two classes of infectious human-tropic replication-competent (HTRC) porcine endogenous retroviruses (PERV) (polytropic PERV-A and PERV-B) and one class of ecotropic PERV-C are known (1, 9, 12, 18). The HTRC PERV derived ex vivo from inbred miniature swine have been recombinants between PERV-A and PERV-C (11, 14, 20, 21). There is no evidence of inherited recombinations between PERV-A and PERV-C (15). The recombined env gene bore the receptor-binding domain of PERV-A on a PERV-C background (3, 11, 20). We used the cell line MAX-T (A. Saalmüller, Vienna, Austria) derived from d/d haplotype miniature swine (13) to isolate a replication-competent PERV-C.

A genomic bacteriophage λ library of MAX-T cells was generated using the λ Fix II/XhoI system (Stratagene, The Netherlands) and screened with a PERV-C env 281-bp probe. The sequence was PCR amplified on genomic MAX-T DNA with primers PERV-C FOR (5′-CTGACCTGGATTAGAACTGGAAG-3′) and PERV-C REV (5′-TATGTTAGAGGATGGTCCTGGTC-3′). One complete proviral clone [PERV-C(1312), nucleotide (nt) 1 to 8672] and two clones truncated due to the cloning strategy [PERV-C(1211), nt 1 to 7560; and PERV-C(6121), nt 1650 to 8672] were isolated (Fig. 1). The nucleic acid homologies between PERV-C(1312) and clone PERV-MSL (1) are 99.9% for gag, 99.9% for pol, and 99.7% for env (Table 1). The amino acid homologies are 99.8% for Gag, 99.8% for Pol, and 99.4% for Env, respectively (Table 1).

FIG. 1.

Alignment of three proviral PERV-C clones. Solid lines, inserts of three different bacteriophage λ clones; dashed lines, chromosomal flanking sequences; arrows, open reading frames (ORF) in the PERV-C sequences. For clone 6121, an asterisk indicates a stop codon at nt position 3854 in the pro-pol ORF, in correlation with PERV-C(1312) at nt position 5612. Numbers indicate the lengths of the bacteriophage inserts and of PERV-C sequences (numbers with asterisks). Genes and ORF are shown as open boxes. cap, transcriptional start site; PBS, primer binding site; SD, splice donor; SA, splice acceptor; SU/TM, (transmembrane) cleavage site; PPT, polypurine tract; p(A), poly(A) addition site; LTR, long terminal repeat.

TABLE 1.

Exchanges of nucleotide and amino acid sequences of reconstituted PERV-C provirus based on PERV-MSL compared with clone PERV-C(1312)

Nucleotide position	Nucleotide in:		Gene	Amino acid in:		Amino acid position
	PERV- MSL	PERV- C(1312)		PERV- MSL	PERV- C(1312)
1259	G	A	gag	V	I	67
3808	A	G	pol			391
4650	G	A	pol	G	E	672
5563	C	G	pol	H	Q	976
5842	T	C	pol			1069
6916	G	A	env	S	N	274
7022	A	G	env			308
7386	G	A	env	E	K	431
7755	G	A	env	E	K	554
7828	T	C	env	L	P	578

Electron microscopy revealed the morphology of gammaretroviruses in MAX-T cells (Fig. 2A and D) and in ST-IOWA cells (ATCC CRL 1746) infected with the molecular clone PERV-C(1312) (Fig. 2B and E). In contrast to PERV-A and PERV-B virions in infected 293 cells (Fig. 2C and F) (4, 21), PERV-C virions show a slightly different morphological structure for Env. While PERV-A/PERV-B particles appear smooth, PERV-C particles show spikes during budding of virions. Spikes might be readily lost after maturation, as is often the case with gamma type retroviruses (10). The PERV-C(1312) particles demonstrate the same structure as the particles released from MAX-T cells (Fig. 2A and D). Immunofluorescence analyses were done using two polyclonal antisera against the capsid protein of PERV (p27) (4) and the surface unit (SU) of the envelope protein of PERV-C. PERV-C Env antiserum was raised in rabbits against a synthetic peptide with amino acid sequence TGQRPPTQGPQPSSNI (Eurogentec, Belgium). The antisera react on MAX-T cells (Fig. 2G) and show no cross-reactivity on 293 PERV-B(33) cells (Fig. 2H). The p27 antiserum reacts on 293 PERV-B(33) (Fig. 2J) and on ST-IOWA cells infected with molecular clone PERV-C(1312) (Fig. 2I).

FIG. 2.

Ultrastructural morphology and immunostaining of PERV particles. Free mature (A, B, and C) and budding (D, E, and F) virus particles of PERV-C derived from MAX-T (A and D) and ST-IOWA (B and E) cells and particles from PERV-A/PERV-B-infected 293 cells (C and F) (12). Envelope spikes on the virion surface, especially on budding PERV-C particles, are indicated (arrows). The bar in panel A represents 100 nm. Indirect immunofluorescence analyses of MAX-T cells reacting with PERV-C Env antiserum (G), 293 cells infected with the molecular clone PERV-B(33)/ATG (PERV-C Env antiserum) (H), ST-IOWA cells infected with PERV-C(1312) (PERV Gag p27 antiserum) (I), and 293 cells infected with PERV-B(33)/ATG (PERV Gag p27 antiserum) (J) are shown. Bars in panels G through J represent 20 μm.

To study the capacity of PERV-C(1312), ST-IOWA cells were infected using cell-free supernatants of MAX-T cells and of ST-IOWA cells infected with the molecular clone PERV-C(1312) (Fig. 3A). To this end, cell-free supernatants from transfected ST-IOWA cells were transferred 24 days posttransfection to naïve ST-IOWA cells. Reverse transcriptase (RT) activity in ST-IOWA cell-free supernatants was detectable from day 15 postinfection (Fig. 3A). Noninfected ST-IOWA cells showed no RT activity (Fig. 3A). To test the RNA expression pattern, total RNA samples of 293FT, 293 PERV-B(33), MAX-T, ST-IOWA, and ST-IOWA cells infected with PERV-C(1312) were hybridized by Northern blot analysis with the PERV-C env probe. Full-length PERV-C RNA transcripts of approximately 8.3 kb and a spliced subgenomic PERV-C env RNA of approximately 3.1 kb are present in MAX-T cells and in infected ST-IOWA cells (Fig. 3B).

FIG. 3.

Infection studies and RNA expression of PERV-C. (A) RT assays with cell-free supernatant of ST-IOWA cells after incubation with cell-free supernatant of MAX-T cells containing 1,800 infectious mU/ml (crosses) and with cell-free supernatants of ST-IOWA cells. PERV-C(1312) (1 μg of plasmid DNA) was transfected into ST-IOWA cells. Supernatants of transfected cells (239 mU) were collected 24 days posttransfection and used for infection of naïve ST-IOWA cells (ST-IOWA/1312) (rectangles). Untransfected ST-IOWA cells show no RT activity (circles). mU/ml, milliunits per ml. (B) PERV-C RNA expression. The hybridization of 20 μg of total cellular RNA with a specific PERV-C env probe (upper panel) is shown. Lane 1, human 293FT; lane 2, 293 PERV-B(33); lane 3, MAX-T; lane 4, ST-IOWA; lane 5, ST-IOWA infected with PERV-C(1312). Lower panel, β-actin as a control for equivalent gel loading.

Conclusion.

This study presents the first description of a replication-competent PERV-C provirus. The novel proviral isolate represents an important tool for further characterization of functional PERV-C. The specificity of a PERV-C SU Env antiserum was revealed by immunofluorescence analyses (Fig. 2G).

Statistical analyses describe the positions of recombination events between the different PERV classes (7, 8). One recombinant, PERV-A/C, was characterized in detail (2, 6). Therefore, screening for the presence of replication-competent PERV-A, PERV-B, and PERV-C in different pig breeds is necessary to evaluate possible recombination events for intact and defective as well as for two defective retroviral sequences copackaged into PERV particles and resulting in a replication-competent virus with a putatively extended host cell tropism. At present, in genetic terms regarding PERV, the d/d haplotype miniature swine are the best origins for new pig breeds, encompassing a lower risk of recombination events because only HTRC PERV caused by recombination events have been detected (2, 11). Furthermore, the screening of pigs harboring PERV-C is important due to the close relationship of sequences in the SU Env C termini of PERV-A and PERV-C (5). Hence, there is a risk that a few point mutations in PERV-C SU Env lead to the infection of 293 cells and eventually enable PERV-C infections in human recipients.

Nucleotide sequence accession numbers.

The sequences for the proviral clones used in this study were deposited in GenBank under the following accession numbers: for PERV-C(1211), AM229311; for PERV-C(1312), AM229312; and for PERV-C(6121), AM229313. The sequences used for the homology studies are those for PERV-MSL (GenBank accession number AF038600) (1).