INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Vertebrates contain a variety of endogenous retroviral sequences, which most likely stem from integrational events of exogenous retroviral sequences into the germ line of predecessor species. The germ line integration of a provirus enabled the vertical transfer of the proviral sequence to following generations, including newly arising species. Such integrational events are thought to have occurred frequently during evolution, and various families of endogenous retroviruses can be found in vertebrate genomes. Expression of stably integrated proviral sequences and subsequent retrotransposition often led to an increased copy number of proviral sequences in their host genomes. For instance, the human genome has been estimated to contain 0.6 to 1% endogenous retroviral sequences, termed human endogenous retroviruses (HERV), with some families having approximately 1,000 copies per haploid genome (reviewed in reference 29). However, a common feature of almost all HERV sequences is their inability to encode functional proteins due to multiple nonsense mutations and deletions.

The so-called HERV-K family (17) seems to be an important exception in that it contains open reading frames (ORFs) and may be involved in tumor development. Full-length ORFs for gag, prt, pol, and env have been isolated from the human genome (for a review, see reference 9). Several lines of evidence indicate that HERV-K sequences may be involved in the initiation or progression of germ cell tumors (GCT) in humans. (i) HERV-K Gag protein and gag-encoded retrovirus-like particles are present in GCT and derived cell lines, respectively (2, 22). (ii) HERV-K Gag and Env antibodies have been found in GCT patients at time of tumor detection (21, 22). (iii) HERV-K transcripts have been detected in GCT precursor lesions (carcinoma in situ) (4).

HERV-K proviral genomes integrated into the germ cell genome of an Old World primate predecessor species after the evolutionary separation from the New World primates (23). Consistently, HERV-K homologous genomes have been found in all Old World primates examined. The human haploid genome harbors approximately 25 to 30 HERV-K copies (16) with two types of proviruses, one possessing a 292-bp deletion within the pol-env boundary. The provirus which retains the 292-bp sequence is regarded as the more ancient variant (10, 17). This hypothesis is supported by (i) its capability for splicing of full-length transcripts into smaller mRNAs and (ii) the acquisition of specific features at the amino terminus of the HERV-K env gene and the encoded Env protein (11). These data suggest a mutational event during Old World primate evolution leading to two different HERV-K proviruses.

We recently identified eight human chromosomes containing at least one intact HERV-K gag ORF and three chromosomes harboring at least one intact HERV-K env ORF each, as determined by the protein truncation test (PTT) (12, 13, 20). The presence of multiple HERV-K genes possessing an ORF indicates that stably integrated retroviral sequences may provide biological benefits for their hosts. A protective function against exogenous retroviruses has been proposed. Recently, involvement of an endogenous retroviral gag sequence in resistance to murine leukemia viruses has been demonstrated in the mouse (1). Likewise, an endogenous retroviral env in the mouse Fv-4 gene is responsible for resistance mediated by receptor interference (5, 7, 8). Similar genes, e.g., ev3 and ev6, have been found in the chicken (19, 24). Most recently, the assistance of endogenous retroviral sequences in immunosuppression by the embryo has been considered by Villareal (28).

At present, there is only very limited information available on the status of HERV-K homologous sequences in Old World primate species other than humans. The finding of multiple HERV-K gag ORFs and env ORFs in humans prompted us to further examine the HERV-K homologs in other Old World primates. In this study, we analyzed several Old World primate species for full-length gag and env sequences as well as the presence of ORFs. Furthermore, we determined when the mutation which led to two HERV-K types occurred in Old World primate evolution. For simplification purposes, we will refer to all HERV-K homologous sequences detected in the various Old World primates as HERV-K sequences.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

DNA samples. Total genomic DNAs from Homo sapiens (human), Pan troglodytes (chimpanzee), Gorilla gorilla (gorilla), Pongo pygmaeus (orangutan), Thercopithecus gelada (gelada), Papio hamadryas (hamadryas), Macaca fascicularis (crab-eating macaque), Macaca mulatta (rhesus monkey), Ateles geoffroyi (black-handed spider monkey), and Lemur catta (ring-tailed lemur) were prepared according to standard procedures. Cercopithecus aethiops (African green monkey) DNA was prepared from an African green monkey kidney cell line (kindly provided by Dieter Neumann-Haefelin) by a method reported previously (15).

PCR primers and conditions. Full-length HERV-K gag genes were amplified by PCR using primers T7gagFOR [5'TAATACGACTCACTATAGGAACAGACCACCATGGGGCAAACTAAAAGT3'; nucleotides (nt) 1111 to 1129 of the HERV-K10(+) sequence (17)] and gagREV (5'CAGGCAGTGGGCCATATAC3'; nt 3195 to 3177 in HERV-K10). HERV-K full-length env genes were amplified by using primers T7envFOR (5'GGATCCTAATACGACTCACTATAGGAACAGACCACCATGAACCCATCAGAGATGCA3'; nt 6450 to 6469) and envREV (5'AACAGAATCTCAAGGCAGAAGA3'; nt 8596 to 8617). Both forward primers have been modified with T7, spacer, and translation-initiating Kozak sequences at their 5' ends (6, 18). Underlined are the start codons of the HERV-K gag and env genes. PCR cycling conditions were as follows: 3 min at 94°C; 30 cycles of 50 s at 94°C, 50 s at 58°C, and 3 min at 72°C; 20 min at 72°C. Reaction mixtures contained 0.5 µM each primer, 200 µM deoxynucleoside triphosphates, and 2.5 U of Taq polymerase in standard PCR buffer (Pharmacia). PCR primers for the amplification of an HERV-K gag subregion mapped to nt 1626 to 1647 (gag-plus) and nt 2241 to 2261 (gag-minus) of the HERV-K10 sequence (17). An HERV-K pol-env region was amplified by using primers located at nt 6370 to 6388 (HK290FOR) and nt 7050 to 7067 (HK290REV) of the HERV-K10(+) genome (17). PCR conditions for the latter two reactions were as follows: 3 min at 94°C; 25 cycles of 45 s at 94°C, 45 s at 58°C, and 45 s at 72°C; 10 min at 72°C. Reaction conditions were as specified above except that 1.5 U of Taq polymerase was added.

Southern blot of PCR products. PCR products were diluted 1:200 and separated by standard Tris-acetate-EDTA gel electrophoresis. DNA was transferred onto a nylon membrane by saline transfer and UV cross-linked. For the detection of HERV-K gag-specific PCR products, a full-length gag sequence derived from plasmid pKG (16) served as the DNA probe. For the detection of HERV-K env-containing PCR products, a cloned PCR product ranging from nt 6370 to 6777 of the published HERV-K10(+) sequence was used (17). Detection of specific sequences was performed by using the ECL Random Prime Labelling and Detection system (Amersham) as recommended by the manufacturer. Final stringent washing conditions were in 0.5× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate) at 60°C for 30 min.

PTT. Full-length HERV-K gag and env genes amplified from genomic DNAs were precipitated by addition of 0.1 volume of 3 M sodium acetate (pH 5.5) and 3 volumes of ethanol and subsequent incubation for 1 h at 20°C. Precipitated DNAs were pelleted by centrifugation; each pellet was washed with 70% ethanol, vacuum dried, and resolved in 6.6 µl of H₂O. Resolved DNA was transcribed and translated for 90 min at 30°C in a TNT T7-coupled reticulocyte lysate system (Promega) in a final reaction volume reduced to 20 µl. Newly synthesized proteins were labeled by [³⁵S]methionine (ICN Biomedicals). After incubation, 2 volumes of sample buffer (65 mM Tris-HCl [pH 6.8], 3% [wt/vol] sodium dodecyl sulfate, 10% [vol/vol] 3-mercapto-1,2-propandiol, 0.025% [wt/vol] bromphenol blue, 5% [vol/vol] glycerin) was added, and proteins were denatured for 2 min at 100°C. Five microliters per sample was separated by standard polyacrylamide gel electrophoresis. Gels were subsequently fixed, washed, soaked for 55 min in 1 M sodium salicylate, vacuum dried at 80°C, and finally exposed at 70°C.

Sequencing. PCR products of interest were TA cloned into pMOSBlue T vector (Amersham). Inserts of positive clones were sequenced by using vector-specific T7 and U19 primers in an Applied Biosystems model 373A automated DNA sequencer as instructed by the manufacturer. Sequencing data were compared to published sequence entries, using World Wide Web services provided by the Baylor College of Medicine. Sequences were aligned by using the Clustal W algorithm (25).

Nucleotide sequence accession numbers. The sequences reported in this study have been deposited at GenBank under accession no. AF018153, AF018154, and AF018155.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Amplification of an HERV-K pol-env subregion. We used a PCR primer pair (HK290FOR-HK290REV) encompassing the position of the 292-bp sequence in the HERV-K pol-env boundary which distinguishes between HERV-K type 1 and type 2 genomes. We obtained a 699-bp product characteristic for HERV-K type 2 genomes and a 407-bp product representing type 1 genomes, as predicted by the HERV-K10(+) sequence (17). Specificity of PCR products was verified by Southern blotting. PCR products representing HERV-K type 2 genomes were found in samples from all Old World primate species, whereas the mutated variant was found only in the hominid human, chimpanzee, gorilla, and orangutan samples (Fig. 1). Based on these results, the mutational event leading to type 1 genomes occurred after the evolutionary split of Old World primates and hominids and before the evolutionary expansion of hominids. No PCR products were obtained from black-handed spider monkey and ring-tailed lemur (Fig. 1). Likewise, no PCR products were obtained from these species in all other PCR experiments.

View larger version (38K): [in this window] [in a new window]   FIG. 1.   Amplification of an HERV-K pol-env region from different primate species. The PCR primers were designed to differentiate between HERV-K type 1 and type 2 genomes as indicated by different sizes of PCR products. PCR products were separated in 1.5% agarose gels, blotted, and hybridized with an HERV-K env DNA probe. Lanes: Hsa, H. sapiens; Ptr, P. troglodytes; Ggo, G. gorilla; Ppy, P. pygmaeus; Tge, T. gelada; Pha, P. hamadryas; Mfa, M. fascicularis; Mmu, M. mulatta; Cae, C. aethiops; Age, A. geoffroyi; Lca, L. catta; N, PCR control without DNA. Additional signals located between signals representing type 1 and type 2 very likely stem from the formation of heteroduplexed DNA and are discussed in more detail in reference 12.

Amplification of full-length HERV-K env sequences. We PCR amplified full-length env genes and subsequently verified the PCR products by Southern blot hybridization. The forward primer (T7envFOR) for the amplification of entire env genes was located at the start codon for the env gene, and the reverse primer (envREV) was located approximately 50 bp downstream from the presumed env stop codon, leading to a PCR product of approximately 2.2 kb (11, 12). Since the PCR primers encompass the 292-bp insertion/deletion site, they also permit differentiation between HERV-K type 1 and type 2 genomes. Full-length HERV-K env sequences with the 292-bp sequence were amplified from all Old World primate species. env genes lacking the 292-bp sequence were found only in human, chimpanzee, gorilla, and orangutan DNAs (Fig. 2). Notably, signal intensities from orangutan and crab-eating macaque DNAs were fainter, possibly indicating fewer HERV-K env copies or sequence divergencies in these species. Human DNA additionally produced env-specific bands at 800 and 950 bp, the latter of which was also seen as a fainter signal in gorilla and orangutan DNAs. Chimpanzee DNA furthermore resulted in a 750-bp product which was not detected by the DNA probe, however, and therefore may be an unspecific product. These results indicate that full-length HERV-K env genes are present in all Old World primate species examined. Importantly, the presence of HERV-K type 1 genomes exclusively in the hominid species was confirmed by PCR with a different primer pair.

View larger version (51K): [in this window] [in a new window]   FIG. 2.   Amplification of full-length HERV-K env genes from different primate species. PCR products were verified by separation in a 1.5% agarose gel, blotting, and hybridization with an HERV-K env DNA probe. Lanes are as described in the legend to Fig. 1, except Ate (A. geoffroyi).

PTT for HERV-K env ORFs. We subsequently determined the presence of HERV-K env ORFs within the primate species. The modification of the forward primer enabled the PCR products which represent full-length HERV-K env genes to be transcribed and translated in vitro (20). Full-length HERV-K env genes encode proteins of approximately 76 kDa (9). Proteins of this size were identified in human, chimpanzee, and gorilla samples, as demonstrated by the PTT using env genes from these species (Fig. 3). Since orangutan and the nonhominid species encoded proteins of significantly lower molecular masses, HERV-K env ORFs appear to be restricted to human, chimpanzee, and gorilla DNAs.

View larger version (89K): [in this window] [in a new window]   FIG. 3.   HERV-K Env proteins encoded by different Old World primate species as revealed by the PTT. Proteins were translated from primate species with full-length HERV-K env genes and separated in standard 10% polyacrylamide gels. Molecular masses of marker proteins are given in kilodaltons. Lanes: Hsa, H. sapiens; Ptr, P. troglodytes; Ggo, G. gorilla; Ppy, P. pygmaeus; Tge, T. gelada; Pha, P. hamadryas; Mfa, M. fascicularis; Mmu, M. mulatta; Cae, C. aethiops; Lu, 61-kDa luciferase control.

Amplification of full-length HERV-K gag sequences. The presence of full-length HERV-K gag sequences within the primate species was analyzed as described for the HERV-K env sequences. Entire HERV-K gag genes were amplified by PCR using a forward primer (T7gagFOR) located at the putative start codon of the gag reading frame and a reverse primer (gagREV) located about 80 bp downstream from the presumable stop (13), resulting in PCR products of 2.1 kb (17). As for HERV-K env, the specificity of the PCR products was verified by Southern blot hybridization. PCR products of the expected sizes were amplified from human, chimpanzee, gorilla, and orangutan DNAs. However, PCR products from the lower Old World primates gelada, hamadryas, crab-eating macaque, rhesus monkey, and African green monkey were slightly larger than expected (approximately 2.2 kb) (Fig. 4). Hence, full-length HERV-K gag genes could be identified in all primate species examined, but lower Old World primate species contain a different, larger type of HERV-K gag.

View larger version (64K): [in this window] [in a new window]   FIG. 4.   Full-length HERV-K gag genes in various primate species. (A) PCR products of full-length HERV-K gag were separated in 1.5% agarose gels, blotted, and hybridized with an HERV-K gag-prt DNA probe. (B) PCR products of full-length HERV-K gag identified in an ethidium bromide-stained 1.5% agarose gel. Lanes are as in Fig. 1; sizes of marker bands are given in base pairs.

Identification of an HERV-K gag subregion different in lower Old World primate species. Using a PCR primer pair (gag-plus-gag-minus) located within the central portion of the HERV-K10 gag gene, as expected, the primer pair resulted in a 635-bp product in human DNA (14, 17). Products of the same size were found in chimpanzee, gorilla, and orangutan DNAs. However, PCR products from gelada, hamadryas, crab-eating macaque, rhesus monkey, and African green monkey DNAs had a size of approximately 740 bp. Hybridization with the HERV-K gag DNA probe confirmed the gag specificity of both PCR products. Orangutan DNA also displayed hybridization signals of 740 bp, detectable by the DNA probe (Fig. 5A). For gorilla and chimpanzee DNAs, a weak 740-bp PCR product was detectable by an ethidium bromide-stained agarose gel (Fig. 5B) but not by hybridization with the gag DNA probe. The 740-bp signal could not be detected in human DNA by ethidium bromide staining or by Southern blotting. Southern blot hybridization revealed an additional 590-bp PCR product in gelada, hamadryas, and African green monkey DNAs. Specifically in hamadryas DNA, ethidium bromide staining revealed bands of approximately 950 and 440 bp, both of which did not hybridize with the gag probe. Taken together, we identified a central region in the HERV-K gag gene containing a deletion in all hominid species examined. The undeleted sequence has been found to be still present in orangutan DNA.

View larger version (58K): [in this window] [in a new window]   FIG. 5.   PCR products of the central HERV-K gag regions from hominids and lower Old World primates. (A) PCR products were hybridized by Southern blotting using an HERV-K gag-prt DNA probe. (B) PCR products obtained from lower Old World primates were identified in an ethidium bromide-stained agarose gel. Lanes are as in Fig. 1; sizes of marker bands are given in base pairs.

Sequencing of HERV-K gag genes from lower Old World primates. To further analyze the sequences in the central region of HERV-K gag genes in lower Old World primate species, we selected one clone derived from African green monkey and two clones obtained from crab-eating macaque for sequencing. Sequences were compared to the corresponding HERV-K10 gag region (17) by using the Clustal W algorithm (25). Besides several smaller insertions/deletions, the three sequences were highly homologous (Fig. 6). A common feature of all three sequences was the presence of nearly identical 96-bp insertions which display no homologies to other HERV-K regions. The insertions displayed a high homology to H. sapiens CpG DNA clone 29b5 (accession no. Z58084) [P(2) = 1.6e27]. Further comparisons showed that clone 29b5 displays homologies (approximately 82% in the 291-bp overlap) to HERV-K gag which extend upstream from the gag-plus forward primer used to amplify the central gag region.

View larger version (85K): [in this window] [in a new window]   FIG. 6.   Clustal W alignment of sequences isolated from two lower Old World primate species and the corresponding HERV-K10 gag region (17). Cae is a sequence derived from C. aethiops (GenBank accession no. AF018153); Mfa-1 (AF018154) and Mfa-2 (AF018155) are two clones derived from M. fascicularis.

PTT for HERV-K gag ORFs. Full-length HERV-K gag ORFs encode a Gag protein of approximately 73 kDa (13, 16). HERV-K gag-positive primate species were analyzed for HERV-K gag ORFs by translating the 5'-modified PCR products in the PTT. Autoradiography shows proteins of 73 kDa in human, gorilla, and orangutan samples (Fig. 7). The chimpanzee sample did not produce a protein of that size even after longer exposure but produced only smaller proteins. HERV-K gag genes derived from gelada, rhesus monkey, African green monkey, hamadryas, and crab-eating macaque encoded proteins having molecular masses clearly larger than 73 kDa. The African green monkey sample displayed only a faint signal at 73 kDa which was not reproducible in the print. Our studies indicate HERV-K gag ORFs in all Old World primate species except P. troglodytes.

View larger version (82K): [in this window] [in a new window]   FIG. 7.   HERV-K Gag proteins derived from different Old World primate species as demonstrated by the PTT. Proteins were translated from primate species with full-length HERV-K gag genes and separated in 10% polyacrylamide gels. Lanes are as in Fig. 3; sizes of marker proteins are given in kilodaltons.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

HERVs of the HERV-K family integrated into the germ cell genome of a predecessor Old World primate species after the evolutionary separation from New World primates. HERV-K homologs are present in all Old World primate species examined. The two HERV-K variants observed in humans most likely resulted from a mutational event during Old World primate evolution. The mutational event occurred immediately after the Hominidae tribe separated from the lower Old World primates, as indicated by the presence of HERV-K genomes with the 292-bp deletion exclusively in hominid species. This finding is confirmed by the use of two different primer combinations, both of which demonstrate the lack of mutated HERV-K genomes in the nonhominid species. These independent results render the possibility of a PCR artifact highly unlikely.

Recently, we determined the chromosomal distribution of HERV-K type 1 and type 2 genomes, finding almost equal numbers of the two HERV-K types within the human genome (12). On the assumption that the two types of HERV-K genomes had the same transcriptional activity, the 292-bp insertion seems to have had no effect on their probability for retrotransposition, and both variants contributed equally to the amplification of HERV-K sequences observed in hominids (23).

At least three HERV-K env genes within the human genome have been shown to possess an ORF (12). Despite the presence of HERV-K env full-length genes in all Old World primate species, we found env genes displaying an ORF exclusively in human, chimpanzee, and gorilla DNAs; the remaining species displayed only proteins of clearly lower molecular weights. However, it cannot be excluded that env reading frames in these species encode Env proteins restricted to the amino-terminal env portion.

The physiological role of HERV-K Env proteins in their hosts remains elusive. It is well known from studies of other mammal species that proteins encoded by endogenous retroviral env sequences cause resistance against exogenous retroviruses by blocking of receptor-mediated virus uptake (5, 7, 8, 19, 24). It is conceivable that HERV-K env-encoded proteins provide similar functions in those species possessing ORFs for Env protein. There are several lines of evidence clearly indicating Env expression. In addition to env mRNAs, Env-directed antibodies have been found in GCT patients. However, no Env protein has been detected (21, 26). This may in part be explained by a developmentally regulated HERV-K expression.

As for HERV-K gag sequences, our recent results demonstrated the presence of multiple gag ORFs within the human genome (13), indicating a physiological function of HERV-K gag. As shown in this study, full-length HERV-K gag genes are present in all Old World primates. All species except P. troglodytes possess ORFs for Gag protein. A remarkable finding in this context is the presence of partly deleted gag sequences in the hominid species. From our results, the recently reported amplification of HERV-K genomes within the hominid species (23) can be attributed exclusively to the expansion of HERV-K genomes with mutated gag genes. In contrast, ancient gag sequences, as they are present in lower Old World primates, were not conserved during hominid evolution. The GenBank entry showing homologies to an ancient gag gene possibly represents a vestige of such a more ancient HERV-K genome.

The 96-bp deletion within the gag gene in principle does not affect the ORF but may have caused a modification. It is legimitate to speculate that the modified Gag proteins no longer have the original functions and may have acquired novel functions providing a biological benefit for the relevant hominid species. The recent isolation of the mouse Fv-1 gene (1) has given insight into the resistance against exogenous retroviruses mediated by endogenous retroviral gag sequences, and there are some similarities between the Fv-1 gene product and HERV-K Gag. (i) Both proteins resemble endogenous retroviral Gag proteins. (ii) The Fv-1 Gag portion is likely not functional in the original way, leading to a blocking effect (3). Mutated HERV-K gag may likewise have lost some original function but may still be able to bind Gag proteins of exogenous retroviruses, as Fv-1 does (3). HERV-K gag-encoded retrovirus-like particles (2) demonstrate that HERV-K Gag still has binding capacity, at least for itself. (iii) Very low Fv-1 expression sufficiently blocks infection (3). Very low HERV-K expression has also been found in various normal human tissues (27).

Whether the putative function of HERV-K Gag is not required in chimpanzee and this primate therefore lacks gag ORFs remains elusive. Since, as for Fv-1, the sequence similarities between endogenous and exogenous retroviral genes might not be very high (1), it may be difficult to find exogenous retroviruses in primates that are influenced by HERV-K gag sequences. Our study provides first insights into the distribution of intact HERV-K genes in Old World primates other than humans and will help to clarify a putative function of HERV-K sequences in Old World primates.