Segregation of EAV-HP Ancient Endogenous Retroviruses within the Chicken Population

Abstract

Avian leukosis virus subgroup J (ALV-J), an exogenous avian retrovirus, is thought to have evolved by recombination with the highly identical env gene of the endogenous avian retrovirus EAV-HP. Embryonic expression of EAV-HP env has been suggested to be associated with the induction of immunological tolerance, a feature observed in a significant proportion of meat-type chickens infected with ALV-J. In support of this hypothesis, we demonstrate that EAV-HP loci, some of which could be associated with tolerance, are still segregating within the chicken population.

HPRS-103 is the prototype of the most recently emerged avian leukosis virus subgroup J (ALV-J). Since its first identification in the 1980s in the United Kingdom, ALV-J has become a worldwide problem affecting meat-type chickens (Gallus gallus), causing tumors of the myeloid lineage (1, 6, 7, 8). While the env genes of other ALV subgroups are closely related (with around 80% sequence identity), the ALV-J env is distinct (3). The HPRS-103 env gene demonstrates a closer relationship (>97% sequence identity) to a novel ancient avian endogenous retrovirus (ERV) designated EAV-HP, suggesting that ALV-J has evolved by recombination between EAV-HP transcripts and genomic RNA of an exogenous ALV (5, 14).

In ovo infection with ALV-J as well as with other subgroups, at 9 to 10 days of embryonic development (before the birds are immunocompetent), leads to tolerant infection with persistent viremia and no antibody response in both meat- and egg-type chickens (7). However, when infected posthatching, the two types of chickens show significant differences in responses to ALV-J infection. While the different lines of egg-type chickens infected with HPRS-103 at 1 day posthatch are able to clear the infection after a transient viremia by mounting a neutralizing antibody response, a number of infected meat-type chickens fail to induce a neutralizing antibody response, resulting in a persistent viremic infection (7). Although the numbers of birds developing tolerant infections can vary (7, 10), this phenomenon is consistently observed in different lines of meat-type chickens infected either by natural exposure or by experimental inoculation (8). Since the ALV-J envelope is highly identical to EAV-HP env sequences, it has been suggested that the specific tolerance against ALV-J infection is due to the embryonic expression of the env sequences of these endogenous elements (15). However, as all the infected birds do not develop a tolerant infection, the induction of tolerance could be associated with expression of some of the unique EAV-HP loci that might have a restricted distribution in the closed, but not fully inbred, population of commercial meat-type chicken lines.

EAV-HP proviruses have been shown to exist in the Gallus genus as approximately 10 to 15 copies per genome based on Southern blot hybridization with env or long terminal repeat (LTR) sequences (11, 14). All EAV-HP proviruses from chickens appear to be defective due to deletions of the pol gene; however, intact proviruses have been identified in Sonnerat's jungle fowl (Gallus sonneratii) (12). A single clone, designated ev/J clone 4-1, with the structure of a spliced env subgenomic transcript bounded by LTRs is the only chicken EAV-HP element described to date that possesses a complete env gene including the endoplasmic reticulum (ER) translocation signal coding sequence (9). Because the EAV-HP proviruses are ancient, these loci might have become fixed and occur in a homozygous state in the chicken genome. Therefore, it remains to be shown whether these loci are still segregating within the chicken population. In this study several chicken lines were analyzed by PCR to obtain evidence as to whether EAV-HP proviruses continue to segregate in the chicken population.

To examine the segregation of EAV-HP elements within the chicken genome, PCR was first used to examine the distribution of two previously isolated chicken EAV-HP loci, designated EAV-HP1 and EAV-HP2 (11). Genomic DNA from two egg-type chickens, line 0 and brown leghorn (BRL), and two meat-type chickens (lines 20 and 21), was PCR amplified as previously described (12) by using a strategy to specifically detect the EAV-HP1 or EAV-HP2 locus (Fig. 1). Genomes were analyzed by amplifying the EAV-HP1 or EAV-HP2 provirus left or right LTR with provirus primer EAVF or EAVR and a primer specific for the host DNA flanking the provirus (Table 1). PCR for the EAV-HP1 locus, which was originally isolated from an egg-type chicken (line N) genomic DNA library, produced products from only BRL chicken DNA (Fig. 2, top left panels). The EAV-HP2 provirus locus-specific primer pairs amplified products from line 21 and BRL chicken DNA (Fig. 2, top right panels).

FIG. 1.

Strategy for locus-specific PCR detection of the proviral or pre-integration state EAV-HP1 and EAV-HP2 loci. PCR was performed to detect the provirus left LTR with an EAV-HP-specific primer, EAVR, and primer F1REV or F1-2 for the host DNA flanking the EAV-HP1 or EAV-HP2 provirus, respectively. To detect the provirus right LTR, PCR was performed with the EAV-HP 3′-untranslated region primer EAVF and downstream host DNA primer H37REV or H37REV-2. For detection of the EAV-HP2 locus in the pre-integration state, the host DNA primers F1-2 and H37REV-2 were used together in a PCR. The EAV-HP1 pre-integration state locus was detected by PCR using host primers F1REV and T74, followed by amplification with primers F1REV and H37REV.

TABLE 1.

Oligonucleotide primers used for PCR amplification and probe labeling

Primer name	Provirus clonea	Position (nucleotides)	Sequence (5′–3′)
103ER	EAV-JF2	5508 to 5491	CACGTTTCCTGGTTGTTG
EAV-SD	ev/J 4-1	344 to 367	ATGGACCAAGTCATTAAGGAAATG
EAVFOR	EAV-HP1	3909 to 4044	GACGGGAGCTCTCGGCATAGGGAGGGGGAGATGTTG
EAVREV	EAV-HP1	241 to 205	TAAGTGAGCTCAAATGGCGTTTATTGCTATAGGCTACG
FIREVb	EAV-HP1	−209 to −190	TGGGTGCTGAGGAAGAAGAG
FIREV-2	EAV-HP2	19 to 38	TGTCATGAGCCCACTTCTCC
H37REVb	EAV-HP1	4351 to 4333	TCTTACTCAGGCTCAACTGC
H37REV-2b	EAV-HP2	4317 to 4336	GCAGATGACACCAAGCTGAG
T74b	EAV-HP1	5747 to 5727	CTTGTAATGAAACACCGGGG

^aAccession numbers for provirus clone sequences used for primer design are AJ292967 (EAV-JF1), AF125528 (ev/J 4-1), AJ238124 (EAV-HP1), and AJ238125 (EAV-HP2). 

^bPosition of primer in host DNA flanking sequence is from an unpublished sequence and is shown relative to the numbering of the corresponding published provirus clone sequence. 

FIG. 2.

Southern blot analyses of locus-specific PCRs for detection of the EAV-HP1 and EAV-HP2 proviral and pre-integration state loci in various chicken lines. For the EAV-HP1 proviral locus, PCR was performed on DNA from two meat-type chickens (lines 20 and 21) and two egg-type chickens (line 0 and BRL), with primer pairs EAVF and H37REV for the right LTR (top panel) and EAVR and F1REV for the left LTR (middle panel). For the EAV-HP2 proviral locus, primer pairs EAVR and H37REV-2 were used for the right LTR (top panel) and EAVR and F1-2 were used for the left LTR (middle panel). For the EAV-HP2 pre-integration state locus, DNA was amplified with primer pair F1-2 and H37REV-2 (lower panel). For the EAV-HP1 pre-integration state locus, DNA was amplified with primer pair F1REV and T74 followed by primer pair F1REV and H37REV (lower panel). The lower arrow indicates the pre-integration state locus PCR product, while the upper arrow points to an approximately 300 bp larger BRL PCR product. The EAV-HP1 and EAV-HP2 clones were used as positive controls, and water was used for negative controls. Blots were hybridized as previously described (12), with the corresponding positive control PCR products used as probes; pre-integration state locus PCR products were hybridized simultaneously with the proviral locus right and left LTR probes, which include the flanking host sequences.

To determine the zygosity of the locus, PCR was performed with primers for the host DNA sequences flanking both sides of the provirus locus (Fig. 1). The thermocycling program used a short extension time to allow amplification of only the small product expected from the pre-integration state locus, if present, but not the larger product expected with the additional 4.2-kbp provirus sequence. PCR with the two primers specific for the host sequence flanking the EAV-HP2 provirus produced products for all four chicken lines (Fig. 2, lower right panel), indicating that the line 21 and BRL chickens were hemizygous for this provirus. PCR detection of the pre-integration state locus from the remaining chicken lines further supported the absence of the provirus locus from these DNA samples.

For amplification of the EAV-HP1 pre-integration state locus, a hemi-nested PCR was required. Using the primers specific for the host sequences flanking the EAV-HP1 provirus, 360-bp PCR products (the expected size for the locus pre-integration state) were detected in the line 0, 20, and 21 chicken DNA samples, indicating that these chickens were homozygous for the pre-integration state locus (Fig. 2, lower left panel). The pre-integration state PCR products were also amplified from Sonnerat's and red jungle fowl DNA previously shown to be positive for the proviral locus (12), indicating the EAV-HP1 proviral locus was hemizygous in these birds and is still segregating within these Gallus species (data not shown). The PCR products detected from BRL DNA, however, were approximately 300 bp larger than those of the pre-integration state. Subsequent cloning and sequencing of line 0 and BRL PCR products confirmed that the line 0 PCR products represented the pre-integration state locus, while the BRL PCR products also included an EAV-HP solo LTR (data not shown).

The ev/J clone 4-1 is the only EAV-HP provirus identified to date from the chicken genome carrying a complete env gene (9, 12), making it a candidate as the source of env sequences involved in the generation of the ALV-J subgroup and the associated tolerance. A previous attempt to PCR amplify the 5′ ends of proviruses with intact env genes failed to amplify this provirus from a meat-type chicken (line 21), suggesting that this provirus might be absent from this chicken line and, consequently, segregating within the chicken population (12). To specifically amplify only provirus DNA with the structure of the spliced env subgenomic transcript, PCR was performed with primer EAV-SD (Table 1), which includes the splice donor sequence and the first six bases of the splice acceptor site, along with the env primer 103ER (Fig. 3A). DNA from the four chicken lines was amplified alongside DNA from red jungle fowl and Sonnerat's jungle fowl, previously found to be positive and negative, respectively, for this provirus structure (12). Proviruses with the spliced transcript structure were detected in line 0 and line 20 chicken DNA but not in the line 21 and BRL chicken DNA tested (Fig. 3B). A probe derived from a previously described intact EAV-HP clone (EAV-JF2) from Sonnerat's jungle fowl (12), including the last 186 bp of pol and env sequences up to the 103ER primer sequence, detected a single band by Southern blot hybridization of line 0 chicken DNA (data not shown), indicating that the spliced provirus was present in the line 0 genome as a single locus. Since the spliced provirus structure is a unique and important locus due to its complete env gene sequence, the frequency of the clone 4-1 locus was examined further in two outbred meat-type chicken lines, designated line 21 and line 12, from two different breeding companies. Of the 103 line 21 birds tested by PCR as described above, 57 were found to be positive for the spliced provirus while all 21 of the line 12 birds tested had the locus.

FIG. 3.

PCR analysis of the distribution of the ev/J clone 4-1 provirus in Gallus DNA. (A) Diagram showing the positions of primers EAV-SD, spanning the gag and env genes across the splice junction, and 103ER in the env gene used to specifically detect the spliced subgenomic transcript provirus structure. (B) Ethidium bromide-stained gel of PCR products of two meat-type chickens (lines 20 and 21), two egg-type chickens (lines 0 and BRL) and the two jungle fowl species, red jungle fowl (RJF) and Sonnerat's jungle fowl (SJF).

The presence of this spliced provirus alone, however, appears to have no direct correlation with ALV-J tolerance, since it was detected in the egg-type chicken line 0, which does not demonstrate tolerance to posthatch infection by ALV-J (1). Disease resistance and susceptibility are strongly influenced by other factors, such as the major histocompatibility complex (MHC) of the chicken (2). Since the meat-type chickens are outbred and have diverse MHCs, the immunological tolerance in these breeds may correlate with the presence of a specific EAV-HP locus, such as the 4-1 provirus, only in birds with a specific MHC haplotype that can allow presentation of peptides encoded by the env gene. Research to date on ALV-J has primarily involved the use of highly inbred egg-type chickens that do not exhibit the immunological tolerance observed in the meat-type chickens. The finding that EAV-HP loci are segregating reveals the importance of focusing future studies on a careful characterization of the genetics of tolerant birds in order to determine the cause of ALV-J tolerance induction.

It is possible that one or more of the EAV-HP proviruses might have contributed to the generation of the ALV-J genome, making their continued presence in commercial breeding flocks relevant for the possibility of future recombination events with exogenous ALVs, allowing the re-emergence of J subgroup viruses. EAV-HP transcripts have been detected by RT-PCR in chicken embryonic tissues, indicating that these elements are expressed (5). As the packaging efficiency of spliced env transcripts is not very high (4), it is most likely that the sequences that gave rise to ALV-J were derived from a provirus locus with an intact env splice acceptor signal and ER translocation signal.

The results presented here significantly establish that the EAV-HP loci of the chicken have not become fixed within the population, unlike the more ancient retroviral elements, such as the human ERVs (13). Continued segregation of EAV-HP loci demonstrates the feasibility of using locus-specific PCR in a selective breeding program to eliminate important loci from commercial flocks that have consequences on host immunity to ALV-J or which encode functional genes with potential for recombination with exogenous ALVs. Since many of the EAV-HP proviruses appear to be defective and unable to encode the envelope glycoproteins, it will now be important to isolate loci that have these complete sequences. The locus represented by the ev/J clone 4-1 is such a target that can be eliminated by using the PCR detection presented in this study.