The CAR1 Gene Encoding a Cellular Receptor Specific for Subgroup B and D Avian Leukosis Viruses Maps to the Chicken tvb Locus

Abstract

Host susceptibility to subgroup B, D, and E avian leukosis viruses (ALV) is determined by specific alleles of the chicken tvb locus. Recently, a chicken gene that encodes a cellular receptor, designated CAR1, specific for subgroups B and D ALV was cloned, and it was proposed that this gene was the s3 allele of tvb (J. Brojatsch, J. Naughton, M. M. Rolls, K. Zingler, and J. A. T. Young, Cell 87:845–855, 1996). We now report that in a backcross derived from an F₁ (Jungle Fowl × White Leghorn [WL]) male mated with inbred WL females, the cloned ALV receptor gene cosegregated with two markers linked to tvb. The two markers used were a tvb^s1-specific antigen recognized by the chicken R2 alloantiserum and restriction fragment length polymorphisms associated with the expressed sequence tag com152e. With all three markers, no crossovers were observed among 52 backcross progeny tested and LOD linkage scores of 15.7 were obtained. These data demonstrate that CAR1 is the subgroup B and D ALV susceptibility gene located at tvb^s3.

Based on host range and viral interference patterns, avian leukosis viruses (ALV) isolated from chickens are classified into six major subgroups (A to E and J). Host susceptibility to infection by members of subgroups ALV-B, ALV-D, and ALV-E is governed by the autosomal tvb locus, thought to encode cellular receptors for these viruses (for a review, see reference 12). Functionally distinct alleles of this chicken locus have been identified: tvb^s1 permits infection by these three viral subgroups; tvb^s3 permits infection only by ALV-B and ALV-D; and the recessive tvb^r allele does not permit entry by any of these viruses (for a review, see reference 12). We recently identified CAR1, a chicken gene that encodes a tumor necrosis factor receptor-related protein, which is a cellular receptor specific for ALV-B and ALV-D (3). We have also identified SEAR, a protein encoded by the apparent turkey homolog of CAR1, which is a cellular receptor for ALV-E (1). Based upon the properties of CAR1, we proposed that it was the s3 allele of tvb (3).

To determine whether CAR1 maps to tvb, we initially attempted to identify specific restriction fragment length polymorphisms (RFLPs) that could be used to follow the segregation of distinct alleles of the cloned gene in a cross between chickens with different tvb genotypes. Genomic DNA from line 6₃ (tvb^s1 homozygotes) and from line 7₂ (tvb^r homozygotes) (2) was subjected to Southern blot analysis (10) with a radioactively labeled CAR1-specific DNA fragment as a probe. With a number of enzymes tested, this probe detected a unique DNA restriction fragment (Fig. 1), indicating that there are no other chicken genes highly related to CAR1. However, the CAR1 alleles from both chicken lines were highly conserved, as judged by the similarly sized DNA restriction fragments (Fig. 1). Indeed, when a panel of more than 50 independent restriction enzymes was used, identical patterns of DNA restriction fragments that cross-hybridized with the probe were detected with both types of CAR1 allele (data not shown). Therefore, RFLP analysis proved not to be a useful method for mapping CAR1.

FIG. 1.

Southern blot analysis of CAR1 in two chicken lines with different tvb genotypes. Samples of 5 μg of genomic DNA from chickens of line 6₃ (homozygous for tvb^s1; labeled as s1) and of line 7₂ (homozygous for tvb^r; labeled as r) were digested with AflIII, BclI, Bsu36I, and HincII, electrophoresed on a 1% agarose gel, and transferred to a nylon membrane (Hybond; Amersham) (10). The membrane was then probed at 65°C with two radioactively labeled EcoRI restriction DNA fragments (1.5 and 1.9 kb in size) derived from a genomic clone of CAR1 (3), by using previously described hybridization and washing conditions (4).

Instead, a PCR-based method was used for mapping (described below) that distinguished between alleles of CAR1 from Jungle Fowl (JF) and White Leghorn (WL) chickens. JF and WL chicken lines are known to differ at tvb because the R2 alloantiserum, which recognizes a tvb^s1-specific antigen, bound to cells from JF but not WL chickens (2). This alloantiserum can agglutinate erythrocytes, apparently by binding to the putative receptor encoded by tvb^s1 in a complex with endogenous subgroup E viral envelope glycoprotein (2). Therefore, R2-specific agglutination served as a useful marker to follow the segregation of the tvb^s1 allele from JF chickens among the East Lansing (Mich.) reference population, representing 52 F₂ progeny of a backcross between a male (JF × WL) and WL females (5).

R2-specific agglutination of cells reflects a tvb^s1-encoded cell polymorphism related to the R blood group antigen that has been mapped to East Lansing linkage group E38 (3b, 11). The E38 linkage group also contains an expressed sequence tag, com152e, that was first isolated from a chicken T-cell library (11). SacI RFLPs can distinguish between alleles of com152e in JF and WL chickens (11). RFLP analysis of the 52 backcross progeny revealed that the JF-specific tvb^s1 allele (detected by R2 agglutination) always cosegregated with the JF-specific com152e allele (Table 1).

TABLE 1.

Cosegregation of the JF allele of CAR1 with two independent markers of tvb^a

Progeny no.	CAR1 allele	R2 agnb	com152e allele		Progeny no.	CAR1 allele	R2 agn	com152e allele
1	JF	+	JF		27	JF	+	JF
2	WL	−	WL		28	WL	−	WL
3	WL	−	WL		29	WL	−	WL
4	WL	−	WL		30	WL	−	WL
5	WL	−	WL		31	JF	+	JF
6	JF	+	JF		32	JF	+	JF
7	JF	+	JF		33	JF	+	JF
8	JF	+	JF		34	WL	−	WL
9	JF	+	JF		35	JF	+	JF
10	WL	−	WL		36	JF	+	JF
11	WL	−	WL		37	JF	+	JF
12	WL	−	WL		38	JF	+	JF
13	WL	−	WL		39	WL	−	WL
14	WL	−	WL		40	WL	−	WL
15	JF	+	JF		41	JF	+	JF
16	JF	+	JF		42	WL	−	WL
17	JF	+	JF		43	WL	−	WL
18	WL	−	WL		44	WL	−	WL
19	JF	+	JF		45	WL	−	WL
20	WL	−	WL		46	WL	−	WL
21	JF	+	JF		47	JF	+	JF
22	JF	+	JF		48	JF	+	JF
23	JF	+	JF		49	WL	−	WL
24	JF	+	JF		50	WL	−	WL
25	WL	−	WL		51	WL	−	WL
26	WL	−	WL		52	JF	+	JF

^aProgeny were typed for the JF allele of CAR1 as described in the legend to Fig. 3. The presence of the JF allele of tvb in these birds was confirmed by using the tvb^s1-specific R2 antiserum, which agglutinates erythrocytes of JF, but not WL, chickens (2). The presence of the JF-specific allele of com152e was shown by RFLP analysis, which detected two independent SacI restriction enzyme genomic DNA fragments that are found specifically in JF but not WL chickens (11). 

^bagn, agglutination. 

To determine whether CAR1 mapped to the E38 linkage group, and thus to tvb, we attempted to identify nucleotide differences that distinguish the JF-specific and WL-specific alleles of this gene. DNA fragments containing an intron of CAR1, located between nucleotides (nt) 732 and 733 of the CAR1 cDNA clone (3, 3a), were isolated from WL and JF genomic samples by using a PCR-based approach that employed exon-specific oligonucleotide primers (Fig. 2). PCR products were cloned into the PCR 2.1 vector by using the TA cloning protocol (Invitrogen), and these DNA samples were sequenced by the dideoxy chain termination method by using an ABI model 373A automatic DNA sequencer.

FIG. 2.

Two nucleotide differences distinguish an intron of the JF allele of CAR1 from that of the WL allele. Outer oligonucleotide primers located in two separate exons of CAR1 were initially used to amplify a 525-bp intron-containing DNA fragment for comparative DNA sequence analysis of JF and WL alleles of the chicken gene. The DNA sequence of the JF allele is shown with exon and intron sequences represented by large and small capital letters, respectively. The locations of two nucleotide differences between the JF and WL alleles (described in the text) are indicated by underlined bold letters. Oligonucleotide primers located within the intron were used as a second-round primer pair to amplify a 209-bp DNA fragment specifically from the JF allele using a 3′ end mismatch PCR (9). The PCR conditions used for specific amplification of the JF allele involved denaturation, annealing, and extension at 94, 50, and 72°C, respectively.

A comparison of DNA sequences revealed two nucleotide substitutions between the introns of the JF and WL alleles of CAR1. A C residue and a G residue in the WL allele were replaced by a G residue (nt 285) (Fig. 2) and an A residue (nt 379) (Fig. 2), respectively, in the JF allele (nt 379) (Fig. 2). Consequently, a reverse oligonucleotide primer mismatched at the 3′ end (nt 285) (Fig. 2) with respect to the WL allele and another forward primer were used to generate a 209-bp DNA product that was amplified specifically from the intron of the JF allele (Fig. 2). Under the conditions used, the 209-bp DNA product was obtained only from JF and not WL chickens (Fig. 3, lanes J and W, respectively), and therefore the presence of this fragment could serve as a molecular marker to follow the segregation of the JF allele of CAR1 among the F₂ backcross progeny.

FIG. 3.

Segregation of the JF allele of CAR1 among the 52 backcross progeny of the East Lansing reference population. PCR analysis was performed as described in the legend to Fig. 2 with genomic DNA samples obtained from the backcross progeny and from JF (J) and WL (W) chickens. The DNA samples were subjected to agarose gel electrophoresis, transferred to a nylon membrane (Hybond; Amersham) (10), and probed with a radioactively labeled 718-bp BglII-SalI DNA fragment of the CAR1 genomic DNA clone (3), prepared by random hexamer labeling (7). The conditions used for DNA hybridization at 65°C were described previously (3); the membranes were washed at 65°C with 2× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate)–0.1% sodium dodecyl sulfate and exposed to Kodak XAR-5 film at −80°C.

To follow the segregation of the JF allele of CAR1, DNA from the 52 backcross progeny was analyzed by the PCR approach. Southern blot analysis with a radiolabeled, intron-containing DNA fragment derived from the BK9 CAR1 genomic DNA clone (3) identified those samples that contained the 209-bp DNA fragment (Fig. 3). In addition, fragments of this size derived from representative backcross progeny were subjected to DNA sequencing to validate the identity of the segregating allele (data not shown). These studies revealed that 25 of the 52 progeny contained the JF allele of CAR1 (samples 1, 6 to 9, 15 to 17, 19, 21 to 24, 27, 31 to 33, 35 to 38, 41, 47, 48, and 52) (Fig. 3). Although the signal observed with samples 7 and 9 appears weak, the presence of the 209-bp DNA fragment in these samples was confirmed independently by ethidium bromide staining of the PCR products following agarose gel electrophoresis (data not shown). Some of the amplified samples (e.g., samples 38 to 45) contained larger DNA fragments of approximately 500 bp that hybridized with the CAR1 intron probe (Fig. 3). These fragments were presumably derived from either the JF or WL alleles of CAR1 during the first step of PCR, which employed the outer oligonucleotide primer pair (Fig. 2).

Among these progeny, the JF alleles of CAR1, com152e, and tvb (defined by R2 agglutination) always cosegregated (Table 1). The lack of recombination between CAR1 and tvb^s1 among progeny tested and the high LOD score (15.7) obtained with the Map Manager program have indicated that CAR1 and tvb^s1 either are allelic or occur within 0.5 Mb. In the East Lansing linkage map, 1 centimorgan (1% recombination) represents a genetic distance of about 0.5 Mb (8). Because only three markers thus far have been mapped to linkage group E38, it appears likely that E38 represents one of the 30 chicken microchromosomes. Taken together, these data demonstrate that CAR1 maps to the chicken tvb locus and provide compelling evidence that this cloned gene is the s3 allele of tvb.

We are now attempting to isolate and characterize the single-copy CAR1 gene at tvb^s1 (Fig. 1). This gene is predicted to encode a receptor for ALV-E in addition to those for ALV-B and ALV-D. Once this receptor has been identified, a comparison of its amino acid sequence with those of CAR1 (3) and SEAR (1) should help delineate ALV-B/D/E entry determinants of chicken Tvb proteins. A detailed analysis of these receptors should also help us understand how ALV-B(D)-receptor interactions apparently lead to cell death following infection, whereas ALV-E-receptor interactions do not (3, 6, 13, 14).