Intact EAV-HP Endogenous Retrovirus in Sonnerat's Jungle Fowl

Abstract

The EAV-HP group of chicken endogenous retrovirus elements was previously shown to be defective, with large deletions of the pol gene. In this report, we demonstrate that genomes of other Gallus species also maintain EAV-HP elements with similar deletions. The chicken EAV-HP1 locus was detected in both red (Gallus gallus gallus) and Sonnerat's (Gallus sonneratii) jungle fowl with identical integration sites, indicating that these elements had integrated before separation of the Gallus species. Furthermore, we demonstrate for the first time that the G. sonneratii genome carries EAV-HP elements with intact pol regions.

The EAV-HP (also designated ev/J), the most recently identified members of the endogenous avian retrovirus (EAV) family, are present in Gallus species as 10 to 15 copies per genome (12). Four different structures of EAV-HP proviruses have been identified in the chicken genome. Three of these 4-kbp-long provirus structures, designated types I, II, and III, show large deletions spanning the entire pol gene and parts of the gag and env regions. Each structure has a different gag-env deletion junction, with the type II and III provirus deletions extending an additional 57 and 86 bp, respectively, than the smallest (type I) provirus deletion size (9, 10). These proviruses are found at multiple loci, with the type I provirus appearing to be the most abundant, based on its frequency in clones randomly screened from separate chicken genomic DNA libraries (9, 10). A single clone, designated ev/J clone 4-1, forms the fourth type of proviral structure, comprising a cDNA of the env subgenomic transcript bounded by long terminal repeats (LTRs) (9). The EAV-HP proviruses are more than 97% identical to the newly emerged subgroup J avian leukosis virus (ALV-J) env gene and demonstrate >96% sequence identity to other EAV family members in the R and U5 regions, the 5′ untranslated region, and portions of the gag gene (6, 9, 10).

Because of their distribution in several Gallus species, the EAV family is considered an ancient group of retroviruses that integrated into a common ancestor. The genus Gallus is composed of four species: the Sonnerat's or grey jungle fowl (SJF; G. sonneratii), the green jungle fowl (GJF; G. varius), the Ceylonese jungle fowl (CJF; G. lafayettei), and the red jungle fowl (RJF; G. gallus), of which domesticated chickens are a subspecies (G. gallus domesticus) (4). EAV-HP proviruses have also been detected in both RJF and SJF by hybridization of env and LTR sequences and PCR amplification of the env region (10, 12). In this study, we have used a PCR approach to examine whether (i) the EAV-HP proviruses in RJF and SJF have the same pol region deletions as in chickens, (ii) the EAV-HP1 and EAV-HP2 loci, present in chickens, could be detected in the jungle fowls, and (iii) the genomic deletion and integration events occurred before or after the separation of Gallus species.

The EMBL accession numbers for the sequences described here are AJ292966 for clone EAV-JF1 and AJ292967 for EAV-JF2, which includes the assembled sequences of the 5′-end and 3′-end PCR products.

As a first step, PCR was carried out with primers for the gag and env regions (Fig. 1A), amplifying across the deletion junctions to detect proviruses with pol region deletions from chicken, RJF, and SJF DNA. Chicken and jungle fowl genomic DNAs isolated as previously described (10) were amplified with 0.5 μM each of primers H83REV (5′-TATTTCTTGCACCAACCTCCC-3′) and H8 (5′-TGGTGAATCCACAATATCTACGAC -3′) in PCR buffer (20 mM Tris-HCl [pH 8.4], 50 mM KCl, 0.25 mM deoxynucleoside triphosphate mix (dNTPs) 2 mM MgCl₂) with approximately 1 ng of template DNA using Taq Gold DNA polymerase (Bio/Gene Ltd., Kimbolton, Cambs, U.K.), with the following thermocycling program: one cycle of 94°C for 2 min, 29 cycles of 94°C for 30 s, 50°C for 15 s, and 72°C for 1 min, and one cycle of 72°C for 7 min. PCR produced two bands of similar size for SJF and line 0 chicken DNA and three bands of similar size for RJF and the remaining chicken lines (Fig. 1B). To determine the provirus deletion junctions, the PCR products were purified using the QIAquick gel extraction kit (Qiagen) and cloned into the pGEM-T vector (Promega) as described by the manufacturer. Sequencing was performed using an Applied Biosystems 377 automated sequencing system with the ABI Prism BigDye terminator reaction kit (Perkin-Elmer) and vector SP6 and T7 primers (Promega). The sequences of the jungle fowl gag-env junction PCR products were identical to the corresponding regions of the published sequences for the prototype clones EAV-HP1 (type I), ev/J clone 3A (type II), and ev/J clone 1C (type III) (data not shown). Detection of identical pol deletions in jungle fowl and chicken DNA strongly suggests that EAV-HP retroviruses integrated into a common ancestor.

FIG. 1.

PCR analysis of the gag-env deletion junctions of the EAV-HP proviruses from chickens and jungle fowl. (A) Diagram indicating the positions of primers in the gag region (H83REV) and in the env region (H8) flanking the deletion junctions of the 4-kbp EAV-HP provirus. (B) Ethidium bromide-stained PCR products amplified with the H83REV and H8 primers from two layer-type chicken lines, line 0 and brown leghorn (BRL), two meat-type chicken lines, 20 and 21, and two jungle fowl species, RJF and SJF. The EAV-HP1 clone and water were used as positive and negative controls, respectively. PCR products are indicated as I, II, and III, corresponding to the 4-kbp provirus types described in the text.

In order to establish that these proviruses integrated into the genome of an ancestral Gallus species and to rule out the possibility that defective EAV-HP proviruses with similar structures entered the Gallus genomes by separate infections, locus-specific PCR analysis was performed on chicken and jungle fowl DNA. Identification of loci shared by different species would indicate that an integration event occurred before separation of the species, since germ line integrations are random and infrequent (7). PCR was carried out as described above using a primer specific for EAV-HP with another one specific for the flanking host DNA for the two chicken loci EAV-HP1 and EAV-HP2 (Fig. 2A) (10). PCR produced multiple bands for all DNA reactions (not shown) since an annealing temperature of 50°C was used with the primer pairs (T_m, >57.3°C) to ensure amplification of loci from divergent jungle fowl DNA that may have a base pair substitution in the primer annealing site. In order to confirm the specificity of the PCR products, DNA gels were blotted and hybridized under high-stringency conditions (10) with probes derived from the corresponding purified PCR products of the EAV-HP1 and EAV-HP2 clones. The locus-specific reactions were carried out across both the left and right LTRs so that agreement would provide confirmation of the results.

FIG. 2.

Southern blot analyses of EAV-HP1 and EAV-HP2 locus-specific PCR products from chicken and jungle fowl DNA. (A) Diagram indicating the relative positions of primers used for EAV-HP locus-specific PCR analysis. PCR was performed on chicken and jungle fowl DNA with reverse primer EAVR (5′-TAAGTGAGCTCAAATGGCGTTTATTGCTATAGGCTACG-3′), complementary to the EAV-HP LTR, and a forward, host DNA-specific primer, F1REV (5′-TGGGTGCTGAGGAAGAAGAG -3′) for the EAV-HP1 locus or F1-2 (5′-TGTCATGAGCCCACTTCTCC-3′) for the EAV-HP2 locus, for amplification of the left LTR and upstream host sequence. The right LTR of the two EAV-HP loci was amplified with forward primer EAVF (5′-GACGGGAGCTCTCGGCATAGGGAGGGGGAGATGTTG-3′) and reverse, host-specific primer H37REV (5′-TCTTACTCAGGCTCAACTGC-3′) for EAV-HP1 or H37REV-2 (5′-GCAGATGACACCAAGCTGAG-3′) for EAV-HP2. (B) PCR was performed on chicken and jungle fowl DNAs, with EAV-HP1 and EAV-HP2 clones used as positive controls and water used as a negative control. PCR products were separated on a 2% agarose gel, blotted, and hybridized with the corresponding purified labeled PCR product of the positive control reactions.

PCR for the EAV-HP1 locus, originally isolated from a line N chicken genomic DNA library, produced products from both jungle fowl DNA reactions (Fig. 2B). The presence of the chicken EAV-HP1 provirus locus in both jungle fowl species provides direct evidence of integration of this EAV-HP provirus before speciation. The EAV-HP2 locus-specific PCR amplified products from the chicken DNA but not the jungle fowl DNA samples. The absence of this provirus may be the result of segregation of this locus within the Gallus population, leading to its loss from the contemporary jungle fowl populations. Conversely, this locus may be the result of reintegration of an endogenous retrovirus from another locus following the domestication of chickens during a subsequent exogenous retrovirus infection.

EAV-HP proviruses with intact pol sequences encoding reverse transcriptase (RT) and integrase (IN) and the splice acceptor at the pol-env junction have not yet been identified, possibly due to the random selection of genomic library clones and the preferential amplification of small PCR products from deleted proviruses. A PCR approach using one primer from the LTR and another from the env region that is deleted in the predominant EAV-HP provirus types was employed to selectively amplify proviruses with intact pol genes in chicken and jungle fowl DNA (Fig. 3A). The jungle fowl species were examined for intact proviruses, since they have not undergone any selective breeding, as is the case for chickens, and may contain provirus sequences that have otherwise been eliminated from commercial chicken flocks. DNA samples from line 21 chickens, RJF, and SJF were amplified with primers EVJFOR (5′-TTCGTGATTGGAGGAAACACTTG-3′) and 103ER (5′-CACGTTTCCTGGTTGTTG-3′) with the thermocycling program of one cycle at 94°C for 2 min, 29 cycles at 94°C for 30 s, 50°C for 15 s, and 72°C for 4 min, and one cycle at 72°C for 7 min. A 568-bp product was produced from RJF (Fig. 3B, bottom arrowhead). Cloning and sequencing established that this PCR product was amplified from the corresponding region of the type IV provirus structure described for ev/J clone 4-1 (9). Since we failed to amplify this provirus from a line 21 chicken after repeated attempts, we conclude that this provirus was absent in this line of chicken and the inability to amplify the provirus was not due to the failure of PCR. Remarkably, an approximately 5-kbp PCR product was amplified from SJF which was the correct size for an intact EAV-HP provirus containing the pol gene (Fig. 3B, top arrowhead).

FIG. 3.

PCR amplification of 5-kbp EAV-HP provirus products with putative pol region sequences. (A) Schematic diagram showing the positions of oligonucleotide primers EVJFOR and 103ER used for PCR relative to a complete provirus. The sequence recognized by 103ER is deleted from the 4-kbp provirus types, indicated as chicken EAV-HP (ev/J), but is present in the type IV ev/J clone 4-1 sequence. Ψ, packaging signal; SD, splice donor; SA, splice acceptor. (B) Ethidium bromide-stained agarose gel of separated PCR products amplified from line 21 chicken, RJF, and SJF DNA. The 0.6-kbp product from RJF was amplified from the type IV provirus, and a 5-kbp product with putative pol sequences was amplified only from SJF DNA.

PCR was repeated using the Expand high-fidelity PCR system (Boehringer Mannheim) for cloning the 5-kbp SJF EAV-HP PCR product into the pGEM-T vector. Two cloned PCR products, designated EAV-JF1 and EAV-JF2, were isolated and sequenced. The sequences of the SJF EAV-HP products were compared with published avian leukosis and sarcoma virus (ALSV) sequences to determine the structure of the proviruses from which they were amplified. The two clones had complete open reading frames encoding the gag-pro-pol proteins. The EAV-HP sequences encoding the terminal three protease residues and the first RT residue (ACAAAUUUAUA) were identical to the ALSV gag-pol frameshift site sequence (5). While the EAV-JF1 clone encoded potentially functional Gag, RT, and IN, the EAV-JF2 clone showed a single-base-pair deletion causing a frameshift in the gag gene leading to truncation of the deduced Gag polyprotein. A comparison between the two EAV-HP clones and modern ALSVs demonstrated approximately 61% nucleotide sequence identity in the pol gene. The deduced amino acid sequence demonstrated approximately 62% identity and 70% similarity with the RT and IN of ALV-J (1) and 62% identity and 71% similarity to the RT and IN of the Prague C strain of Rous sarcoma virus (11). Highly conserved regions such as the RT catalytic site showed stretches of high sequence identity. At the pol-env boundary, both the EAV-JF1 and EAV-JF2 clones demonstrated the intact EAV-HP splice acceptor site for processing of the env subgenomic transcript.

PCR products representing the 3′ end of an intact provirus were amplified using primer EAV-IN1 (5′-TTCCCGCCCCAAATTAAGAC-3′) from the EAV-HP IN sequence and primer EAVR for the right LTR. A cloned PCR product that was 100% identical to the EAV-JF2 clone in the 700-bp overlapping region was chosen for sequence assembly using Staden version 2000.0 (2, 13). The intact SJF EAV-HP provirus had a 6.8-kbp genome with complete gag, pol, and env genes. As in the case of the gag and pol genes of clone EAV-JF2, the deduced amino acid sequence encoded by the env gene of this provirus indicated that it was also nonfunctional due to the presence of a stop codon in the surface protein coding sequence. The EAV-HP clones reported here would be useful for the reconstitution of competent retrovirus for studying the pathology and tropism of this ancient retrovirus and generating new tools such as antibodies to further define the repertoire of EAV-HP elements resident in commercial chicken flocks. The assembled full-length sequence is an important contribution to the growing database of ancient avian retrovirus sequences as a complete EAV-HP reference provirus that would be essential to describe the evolution of modern ALVs.

Since PCR failed to amplify products representing the 5′ end of intact EAV-HP proviruses from line 21 chickens and RJF, Southern blot analysis was conducted to determine if these proviruses were also present in chicken and RJF DNA. A 275-bp sequence was amplified from the pol region with primers RT1 (5′-CTTTTCGCTTGCTGCATGAC-3′) and RT3 (5′-TTGACAAATGGTGGGGGAG-3′) to be used as a probe for Southern blot analysis as described above. The pol fragment had approximately 68% nucleotide sequence identity with the corresponding ALV-J region. Using high-stringency conditions to prevent cross-hybridization with endogenous ALV-E loci, three EAV-HP proviruses with pol region sequences were detected by Southern blot analysis in SJF (Fig. 4). No proviruses were detected with the pol probe from RJF or the chicken lines examined, demonstrating conclusively that these intact elements are not present in the G. gallus genomes examined. As in the case of EAV-HP2 locus, the absence of these intact elements from RJF and chickens could be due to segregation of these loci in the Gallus ancestral population and their subsequent loss from the evolutionary line leading to domesticated chickens. Although it is possible that these elements integrated after the separation of SJF, it is unlikely, because the more complete form must have preceded the deleted provirus structures, and the high sequence identity between the deleted and intact EAV-HP proviruses (>97%) suggests that these viruses have not diverged over a great length of time. Previous studies have demonstrated that line 0 chicken embryo fibroblasts, which are free of ALV, produce retrovirus-like particles with associated RT activity in culture supernatants (3, 8). The demonstration that all line 0 chicken EAV-HP proviruses have their pol genes deleted shows that these elements are not likely to contribute to the RT activity. However, it remains to be resolved whether EAV-HP elements produce structural proteins that may yet play a role in the formation of retrovirus-like particles associated with the RT activity.

FIG. 4.

Detection of EAV-HP RT-encoding sequences in SJF. Southern blot hybridization was performed on EcoRI-digested genomic DNA from three chicken lines, line 0, line 21, and brown leghorn (BRL), and two jungle fowl species, RJF and SJF, by hybridization under high-stringency conditions with the 275-bp RT-specific probe. Three bands hybridized only from the SJF DNA.