Nonconserved Tryptophan 38 of the Cell Surface Receptor for Subgroup J Avian Leukosis Virus Discriminates Sensitive from Resistant Avian Species

Abstract

Subgroup J avian leukosis virus (ALV-J) is unique among the avian sarcoma and leukosis viruses in using the multimembrane-spanning cell surface protein Na⁺/H⁺ exchanger type 1 (NHE1) as a receptor. The precise localization of amino acids critical for NHE1 receptor activity is key in understanding the virus-receptor interaction and potential interference with virus entry. Because no resistant chicken lines have been described until now, we compared the NHE1 amino acid sequences from permissive and resistant galliform species. In all resistant species, the deletion or substitution of W38 within the first extracellular loop was observed either alone or in the presence of other incidental amino acid changes. Using the ectopic expression of wild-type or mutated chicken NHE1 in resistant cells and infection with a reporter recombinant retrovirus of subgroup J specificity, we studied the effect of individual mutations on the NHE1 receptor capacity. We suggest that the absence of W38 abrogates binding of the subgroup J envelope glycoprotein to ALV-J-resistant cells. Altogether, we describe the functional importance of W38 for virus entry and conclude that natural polymorphisms in NHE1 can be a source of host resistance to ALV-J.

INTRODUCTION

The first step in retrovirus infection is attachment of the virus envelope glycoproteins to the specific cell surface receptor. Consequently, the susceptibility of each cell strictly depends on the expression and display of proper receptor molecules. This attachment, as well as the following phases of virus entry, requires a perfect match of receptors and envelope glycoproteins (1). Structural alterations within variable and hypervariable envelope glycoprotein regions easily abrogate the infectivity or even change the receptor usage and broaden the host range. This is best exemplified by avian sarcoma and leukosis viruses (ASLVs), a closely related group of retroviruses which evolved into five subgroups, A to E, that utilize three different receptors encoded by three genetic loci, tva, tvb, and tvc (1, 2). The tva locus encodes a protein belonging to the family of low-density lipoprotein receptors and determines susceptibility to the subgroup A ASLVs (3, 4). The tumor necrosis factor receptor-related protein encoded by the tvb locus confers susceptibility to subgroup B, D, and E ASLVs (5–7). Finally, subgroup C ASLVs utilize the Tvc protein of the butyrophilin family with two immunoglobulin-like domains (8). The complete resistance or decreased susceptibility of host chicken cells to a particular ASLV subgroup can then be caused by premature termination or a frameshift in the receptor-encoding loci (8–10), decreased receptor expression and display (11), and even single amino acid substitutions in the receptor sequence (9, 12).

Subgroup J avian leukosis virus (ALV-J), the prototype virus isolate of which is HPRS-103, is an independent envelope subgroup which does not interfere with subgroups A to E and cannot be neutralized by antisera raised against subgroups A to E (13). ALV-J was originally described as a nonacute virus causing myelocytomatosis in commercial meat-type lines in the United Kingdom (14) but rapidly evolved into a widespread pathogen inducing a broad spectrum of additional disorders, such as histiocytic sarcomatosis, hemangiomas, and erythroblastosis (see reference 15 for a review). Among a broad list of galliform species, only domestic chickens, jungle fowl, and turkeys are susceptible to ALV-J infection (16). The unique interference properties and host range of ALV-J are determined by the surface subunit of envelope glycoprotein gp85, which shares only 40% identity with the envelopes of subgroups A to E (17) and which has been generated by multiple recombinations between exogenous avian leukosis virus and chicken endogenous retrovirus (18).

The receptor for ALV-J was identified as chicken Na⁺/H⁺ exchanger type 1 (chNHE1), encoded by the tvj locus on chromosome 23 (19). Although ALV-J is highly diversified, chNHE1 binds all strains. This is evidenced by the engineered chicken cell line DF-1/J, where expression of the envelope glycoprotein from strain ADOL-Hc1 interferes with infection by six ALV-J isolates but not ASLV subgroups A to E (20). chNHE1 contains 12 predicted transmembrane segments, six extracellular loops, and a long C-terminal intracellular tail. The virus-binding site was predicted to reside in the N-terminal part of prominent extracellular loop 1 (ECL1), on the basis of the divergence of otherwise conserved amino acid sequences of sensitive chNHE1 and refractory human NHE1 (19).

Identification of the chNHE1 functional determinants responsible for ALV-J receptor activity is urgently needed because of the severe economic losses to the poultry industry caused by ALV-J (21). The aim is to find natural polymorphisms in chNHE1 as a source of the host resistance. Because no resistant chicken lines have been described until now, we started with a comparison of the NHE1 amino acid sequences from permissive and resistant galliform species. We concentrated on aromatic amino acid residues such as tyrosine and tryptophan, which were already identified to be critical determinants in subgroup B, C, and E ASLV receptors for the interaction with ASLV envelope glycoproteins (22–24). In our study, we found that W38 of chNHE1 discriminates sensitive from resistant tvj alleles in the order Galliformes, and we show that deletion or mutation of W38 abrogates the ALV-J receptor activity of wild-type (wt) chNHE1. We conclude that W38 of chNHE1 is critical for efficient ALV-J attachment and infection.

MATERIALS AND METHODS

Construction of subgroup J reporter vector and virus propagation.

An RCAS(A)GFP subgroup A retrovirus vector (25, 26) transducing the green fluorescent protein (GFP) reporter gene was digested by KpnI and StuI (both New England BioLabs), and the 1,853-bp fragment containing the 3′ end of pol and the whole env gene was discarded. The HPRS-103 ALV-J molecular clone was kindly obtained from V. Nair (Institute for Animal Health, Compton, United Kingdom), and the 2,120-bp KpnI-BsaBI (New England BioLabs) fragment containing the 3′ end of pol, the whole env gene, the redundant transmembrane region, and the direct repeat was inserted into the RCAS backbone (Fig. 1). Thus, the resulting vector, RCAS(J)GFP, produces GFP-transducing virus of subgroup J in DF-1 cells, which are free of ALV-related endogenous retroviral loci. Plasmid DNA of RCAS(J)GFP was transfected into DF-1 cells using the XtremeGENE transfection reagent (Roche), and virus stocks were harvested on day 9 or 10 posttransfection (p.t.). The cell supernatants were cleared of debris by centrifugation at 2,000 × g for 10 min at 10°C, and aliquoted viral stocks were stored at −80°C. The virus titer was determined by terminal dilution and subsequent infection of DF-1 cells and reached 10⁶ infection units (IU) per ml.

Fig 1.

Schematic diagrams of the parental viruses RCAS(A)GFP and HPRS-103 and the chimeric reporter virus RCAS(J)GFP. Restriction sites used for the env gene substitution are indicated. LTR, long terminal repeats; rTM, redundant transmembrane domain; DR, direct repeat; GFP, green fluorescent protein; SA, acceptor for the splicing of the GFP mRNA.

Preparation of embryo fibroblasts and cell culture.

Chicken embryo fibroblasts (CEFs) were prepared from 10 10-day-old embryos from the inbred lines L15 and H6, from outbred population BL, and individually from four embryos of Indian red jungle fowl (Gallus gallus murghi). L15, H6, and BL chickens were maintained at the Institute of Molecular Genetics, Prague, Czech Republic (27), and embryonated eggs of G. g. murghi were obtained from the zoological garden Ohrada u Hluboké, Czech Republic. The procedure was described previously (25). Embryo fibroblasts of other galliform species were used for comparison: guinea fowl (Numida meleagris), turkey (Meleagris gallopavo), Japanese quail (Coturnix japonica), chukar (Alectoris chukar), gray partridge (Perdix perdix), common pheasant (Phasianus colchicus), and Reeve's pheasant (Syrmaticus reevesi). Embryonated eggs were obtained from the University of Veterinary and Pharmaceutical Science Brno (guinea fowl, turkey, chukar, and Japanese quail), the Židlochovice Forest Enterprise (common pheasant and Reeve's pheasant), or a private breeder (gray partridge), and embryo fibroblasts were prepared in the middle of incubation, i.e., on day 8 (Japanese quail), 11 (common pheasant and Reeve's pheasant), 12 (gray partridge and chukar), or 14 (guinea fowl and turkey), in the same way as CEFs. Embryo fibroblasts of domestic duck (Anas platyrhynchos), semi-inbred line Khaki Campbell, maintained at the Institute of Molecular Genetics, Prague (28), were prepared from 12-day-old embryos and used as an outgroup in infection experiments. All embryo fibroblasts, as well as chicken permanent cell line DF-1 (29) and the transformed QT6 quail cell lines (30), were grown in a mixture of 2 parts Dulbecco's modified Eagle's medium and 1 part F-12 medium supplemented with 8% fetal calf serum, 2% chicken serum, and 1× antibiotic-antimycotic solution (Sigma) in a 5% CO₂ atmosphere at 40°C.

RCAS(J)GFP virus spread assayed by flow cytometry.

Avian embryo fibroblasts or QT6 cells were seeded at a density of 5 × 10⁴ per well in a 24-well plate and infected with 5 × 10⁵ IU of RCAS(J)GFP virus at 24 h after seeding. The virus was applied in 0.25 ml medium for 1 h. The percentage of GFP-positive cells was quantitated by fluorescence-activated cell sorting (FACS) using an LSRII analyzer (Becton, Dickinson) on days 1, 2, 3, and 4 postinfection (p.i.). The cells were trypsinized and washed in phosphate-buffered saline (PBS) before the analysis.

Amplification and analysis of the tvj alleles from various avian species.

We prepared total RNA from QT6 cells and cultured embryo fibroblasts from the panel of galliform species using the TRI Reagent (Sigma). cDNA was reverse transcribed from 1 μg total RNA with Moloney murine leukemia virus (M-MLV) reverse transcriptase (RT) and oligo(dT)₁₅ primers (both from Promega). We amplified the left part of the NHE1 coding sequence, which contains all predicted transmembrane domains and extra- and intracellular loops (19), using the forward primer chTVJ1L (5′-CTTCCCTGGGCTCTGCTG-3′, nucleotides 25 to 42 downstream of the ATG initiation codon) and reverse primer chTVJR6 (5′-CAGGAACTGCGTGTGGATCTC-3′, complementary to nucleotides 1,537 to 1,557 of chNHE1). PCR conditions were the following: 98°C for 180 s, 35 cycles of 98°C for 10 s, 67.6°C for 30 s, and 72°C for 100 s, and terminal extension at 72°C for 7 min with Taq polymerase (TaKaRa). The resulting PCR product of 1,533 bp in length was treated with ExoSAP-IT (USB) and then directly sequenced from the chTVJ5′RACE1 primer (5′-TCATCAGGCAGGCCAGCAGGAT-3′, complementary to nucleotides 301 to 322) using a BigDye Terminator v3.1 cycle sequencing kit (PE Applied Biosystems).

Construction of the chNHE1 expression vector, mutagenesis, and transfection experiments.

One microgram of total RNA prepared from L15 CEFs was reverse transcribed with an AccuScript High Fidelity first-strand cDNA synthesis kit (Agilent Technologies). The full-length coding sequence of chNHE1 was amplified from the resulting cDNA by Phusion Hot Start DNA polymerase (Finzymes) using primers TVJ5′UTRXhoI (5′-TACTCGAGATCTCCTCGCAGCGTCTCTG-3′, matching nucleotides 44 to 25 upstream of the ATG) and TVJ3′UTRXhoI (5′-TACTCGAGAAACCAGAGGAGGGGCC-3′, complementary to nucleotides 2,426 to 2,445 downstream of ATG). Cycling conditions were the following: 98°C for 30 s, 30 cycles of 98°C for 7 s, 67.5°C for 30 s, and 72°C for 100 s, and terminal extension at 72°C for 7 min. The resulting PCR product of 2,489 bp in length was cloned into the pGEM-T Easy vector (Promega), and the entire chNHE1 open reading frame was cut using XhoI (New England BioLabs) and ligated into the XhoI-linearized pVitrotdTomato vector under the control of the hFerL promoter. The resulting expression construct was dubbed pVitrotdT-tvj. The pVitrotdTomato vector was previously created (31) from the pVitro expression plasmid (InvivoGene) by replacement of the GFP coding sequence with the tdTomato fluorescence marker in the first expression cassette to allow tracking of efficiently transfected cells.

Mutagenesis was performed in a subgenic fragment of chNHE1 using a Transformer site-directed mutagenesis kit (Clontech). The 1,533-bp PCR product amplified by primers chTVJ1L and chTVJR6 was cut with ApaI and BamHI (both from New England BioLabs), and the resulting cassette of 244 bp containing the whole ECL1 sequence was cloned into ApaI-BamHI-linearized vector pcDNA3 (Invitrogen). This vector was further shortened by 1.4 kbp, eliminating the SexAI-BstZ17I (both enzymes were from New England BioLabs) fragment with the neoR and simian virus 40 regulatory sequences. For the mutagenesis and selection procedures with the ScaI/HindIII selection primer, see reference 32. The changes ΔW38, W38G, W38E, and P52H were introduced using the following respective primers: ΔW38mut (5′-TCCGAGCCCACCGAGCAGCCATGGGGAGAGCCC-3′), which creates a de novo diagnostic NcoI site; W38Gmut (5′-TCCGAGCCCACCGGTGAGCAGCCGTGG-3′), which creates a de novo AgeI site; W38Emut (5′-TCCGAGCCCACCGAGGAGCAGCCATGGGGAGAGCCC-3′), which creates a de novo NcoI site; and P52Hmut (5′-ATCACCGCCGCCCACCTGGCTACGGCCCAGGAG-3′), which destroys the MscI site.

For the transfer of mutagenized ApaI-BamHI cassettes into the expression construct pVitrotdT-tvj, the pTagBFP-N vector (Evrogen) was adapted by eliminating the ApaI and BamHI sites by mung bean nuclease blunting and self-ligation of the linearized plasmid. The XhoI fragment containing the entire chNHE1 open reading frame was then ligated into the pTagBFP-N XhoI site, and the mutagenized ApaI-BamHI cassettes were cloned into the resulting adapter vector, replacing the wild-type ApaI-BamHI sequences. Mutant tvj coding sequences were then transferred as XhoI fragments into the pVitrotdT-tvj expression vector.

The expression constructs with wt and mutated tvj were transfected into QT6 cells. A total of 5 × 10⁴ cells were seeded onto a 24-well plate in the cultivation medium, and transfection was done 12 h later by Ca²⁺ phosphate precipitation of 2 μg of plasmid DNA. At 5 h after the transfection, the precipitate was washed, and at 24 h after the transfection, the cells were used for RCAS(J)GFP infection and virus spread assay. The tdTomato fluorescent protein was used as a marker of successfully transfected cells, and the infection efficiency was expressed as the percentage of GFP-positive cells among the tdTomato-positive cells.

Real-time PCR analysis of NHE1 transcription.

Total RNAs were extracted from nontransfected and pVitrotdT-tvj-transfected QT6 cells by the RNAzol RT (Molecular Research Center) according to the manufacturer's instructions. One microgram of total RNA was treated with DNase for 15 min, and reverse transcription was performed in a total volume of 50 μl with M-MLV reverse transcriptase (Gibco) and random hexamers. For quantitative real-time PCR of NHE1 transcripts, triplicate samples of 1 μl of the resulting cDNA were analyzed with a Mesa Green quantitative PCR MasterMix Plus for SYBR assay kit (Eurogentec) and a CFX96 system for real-time PCR detection (Bio-Rad). Primers tvj-E2-FW (5′-CCTATCTCTCTGCTGAGATCTTCC-3′) and tvj-E3-RV (5′-GAGGAAGTACTTGATGGTGGTGTG-3′) amplified chicken NHE1 exons 2 and 3, primers tvj-E10-FW (5′-TTGCAGAACGCCTACCTGA-3′) and tvj-E11-RV (5′-AGGTTCTCCGAGTCCGGTTT-3′) amplified exons 10 and 11. As an internal control, chicken GAPDH (glyceraldehyde-3-phosphate dehydrogenase) was amplified using primers chGAPDH-FW (5′-CATCGTGCACCACCAACTG-3′) and chGAPDH-RV (5′-CGCTGGGATGATGTTCTGG-3′). The volume of the reaction mixtures was 20 μl with a 250 nM final concentration of each primer, and cycling conditions were 95°C for 15 s, 62°C for 20 s, and 72°C for 20 s for 40 cycles. The specificity of the PCR products was confirmed by melting curve analysis and by sequencing the PCR products. To assess the amount of contaminating exogenous DNA, either genomic or plasmid, we included reactions run in the absence of reverse transcriptase as negative controls. Amplification efficiencies were 1.82 and 1.84 for the two tvj fragments and 1.85 for GAPDH.

Cell surface expression of NHE1 receptors.

The cell surface proteins were biotinylated and isolated using a Pierce cell surface protein isolation kit (Thermo Scientific) according to the manufacturer's instruction. The 4E9 monoclonal antibody (Santa Cruz) specific for a conserved epitope within the cytoplasmic tail of the porcine NHE1 was used for Western blot analysis of the cell surface fraction of wt or mutated NHE1 receptor. The secondary antibody used was horseradish peroxidase-linked goat anti-mouse IgG antibody (Cell Signaling). The results were visualized using 20× LumiGlo reagent and a 20× peroxide chemiluminescent detection system (Cell Signaling).

Construction of SU(J)-rIgG and production of immunoadhesin.

The recombinant fusion immunoadhesin SU(J)-rabbit IgG (rIgG) was constructed by replacing the SU sequence of the A subgroup env in the immunoadhesin SU(A)-rIgG with the SU sequence of the env gene of HPRS-103. The SU(A)-rIgG sequence was obtained from the RCAS(BP)B SUATEVrIgG vector (a kind gift from Mark J. Federspiel, Molecular Medicine Program, Mayo Clinic, Rochester, MN) as a ClaI fragment and cloned into the adapter plasmid Cla12 (25). The resulting construct, Cla12-SU(A)rIgG, was amplified in dcm-negative SCS110 bacteria, and the SU sequence was removed by MscI-EcoRI cleavage. The SU sequence of HPRS-103 was amplified by RT-PCR from DF-1 cells transfected with the molecular clone HPRS-103. At two passages after transfection, total RNA was prepared from DF-1 cells using the RNAzol reagent (Molecular Research Center) and cDNA was synthesized by use of a AccuScript high fidelity first-strand cDNA synthesis kit (Agilent Technologies). Primers IAJF (5′-CACTGTGGCCAATGGAAGCCGGTCATAAAGGCATTT-3′, complementary to the HPRS-103 env signal peptide immediately upstream of the splice site) and IAJR (5′-ACTCGGAATTCCTTGCCAAGCCCTGTCCCCACA-3′, complementary to the conservative 3′ part of the SU sequence) amplified the SU sequence of HPRS-103 as a 1,006-bp fragment, and the MscI and EcoRI sites were added at the 5′ and 3′ ends, respectively. The resulting PCR product was cloned into the pGEM-T Easy vector (Promega), sequenced, and used to replace the SU(A) sequence in Cla12-SU(A)rIgG by use of the MscI and EcoRI restriction sites. The SU(J)-rIgG immunoadhesin sequence was then transferred into the RCAS(BP)B replication-competent vector (25) as the ClaI fragment. The final vector, RCAS(BP)B-SU(J)rIgG, was then transfected into DF-1 cells and the immunoadhesin was harvested as a filtered supernatant (pore size, 0.45 μm) three passages after transfection.

SU(J)-rIgG immunoadhesin binding to the chicken, chukar, and quail cells was assayed as follows. About 2 × 10⁶ chukar embryo fibroblasts and DF-1 and QT6 cells were harvested by use of a nonenzymatic cell dissociation solution (Sigma), washed in PBS supplemented with 2% calf serum (PBS-CS) on ice, centrifuged for 4 min at 300 × g, and resuspended in 100 μl of PBS-CS. One milliliter of immunoadhesin supernatant was added, and the mixture was incubated on ice for 1 h. After three washes with PBS-CS, the goat anti-rabbit IgG linked to Alexa Fluor 488 (Invitrogen) was diluted 1:1,000 in PBS supplemented with 4% calf serum, and washed cells were incubated in 700 μl of diluted antibody on ice for 30 min. After incubation and three washes in PBS-CS, cells were resuspended in 150 μl PBS-CS, and the fluorescence was quantitated by FACS using an LSRII analyzer (Becton, Dickinson).

RESULTS

Construction of subgroup J reporter virus.

Infection of cells with ALV-J has been studied in vitro mostly using the prototype HPRS-103 strain or its pseudotypes (16, 19). In order to simplify the procedure and improve the quantitative assessment of ALV-J infection, we constructed a GFP-transducing reporter virus of subgroup J specificity based on the RCAS retrovirus vector (25, 26). We replaced the env gene of RCAS(A)GFP with the complete env gene from the cloned HPRS-103, including the redundant part of the transmembrane domain and direct repeat (33). The resulting replication-competent virus, RCAS(J)GFP, produced in DF-1 cells, reached a titer of 10⁶ IU per ml. We checked the subgroup specificity of the RCAS(J)GFP virus by infection of chicken DF-1 cells and quail QT6 cells. The virus-infected DF-1 cells and the proportion of GFP-positive cells gradually increased over time, evidencing virus replication. In contrast, we observed less than 0.05% GFP-positive cells in infected as well as noninfected QT6 cells, which could be attributed to the background autofluorescence (data not shown).

Host range of RCAS(J)GFP virus in avian species.

Using the reporter virus RCAS(J)GFP, we evaluated the susceptibility to ALV-J of a panel of eight galliform species and domestic ducks as an outgroup. Embryo fibroblasts were infected with RCAS(J)GFP, and virus spread was followed as the percentage of GFP-positive cells quantified by FACS. Our results (Table 1) corroborate the data obtained previously (16) with HPRS-103-pseudotyped RSV. We observed that only domestic chicken, red jungle fowl, and turkey were susceptible; other species did not display any RCAS(J)GFP infection even after 4 days p.i. We tested a panel of eight inbred lines of domestic chickens (only data for the H6 and L15 lines, in addition to data for the outbred BL chicken, are shown in Table 1), and we noted slight differences in the time course of infection among various lines. This suggests that there might be differences in susceptibility to the virus, with the H6 inbred line being the least susceptible and L15 being the most susceptible.

Table 1.

Time course of RCAS(J)GFP infection in embryo fibroblasts from susceptible and resistant avian species

Species/inbred line	% GFP-positive cells on the following day p.i.a:
	1	2	3	4
Chicken
L15	4.1/32.5	18.3/88.8	52.8/95.1	69.8/97.6
H6	2.9/26.4	6.9/50.5	9.0/56.1	11.3/61.4
BL	6.1/42.7	21.8/86.1	36.9/90.9	43.9/91.2
Red jungle fowl	7.7/63.6	16.9/90.7	18.1/93.7	20.6/96.9
Turkey	2.9/18.0	11.4/64.4	16.6/84.4	20.1/90.6
Common pheasant	<0.05b	<0.05	<0.05	<0.05
Reeve's pheasant	<0.05	<0.05	<0.05	<0.05
Japanese quail	<0.05	<0.05	<0.05	<0.05
Gray partridge	<0.05	<0.05	<0.05	<0.05
Chukar	<0.05	<0.05	<0.05	<0.05
Guinea fowl	<0.05	<0.05	<0.05	<0.05
Khaki Campbell duck	<0.05	<0.05	<0.05	<0.05

^aThe embryo fibroblasts were infected at multiplicities of infection of 1 and 10. Unless indicated otherwise, data are given as the percentage at a multiplicity of infection of 1/percentage at a multiplicity of infection of 10.

^bA value of 0.05% for GFP-positive cells represents the natural autofluorescence of mock-infected cells.

Polymorphisms in NHE1 discriminating between susceptible and resistant species.

In order to identify the amino acid residues critical for the NHE1-EnvJ interaction, we compared the sequence of NHE1 from various galliform species and looked for polymorphisms constantly present in the resistant species. There are two large extracellular loops, ECL1 and ECL5, within NHE1, and ECL1 is highly divergent, with only 39% identity between chNHE1 and human NHE1 (19). We concentrated on the left part of ECL1, which is almost completely divergent between chicken and human. The a1ignment of the NHE1 amino acid sequence from various galliform species (Fig. 2) points to a single site discriminating between susceptible and resistant species. First, W38 was either deleted in chukar and pheasants or changed in quail and guinea fowl. NHE1 from chukar displayed a single amino acid deletion, ΔW38, whereas other resistant species bore deletions or changes of several amino acids immediately around or in a broader vicinity of W38. Other amino acid changes scattered along ECL1 of NHE1 either were present in a single resistant species, e.g., P52H in chukar, or were not common in all resistant species (K64R in quail and pheasants, P77T in quail, H66P/Q in guinea fowl and pheasants, and G85S in quail). Some amino acid changes were present both in resistant species and in susceptible turkeys and cannot explain the resistance to the J subgroup of ALV without any contribution of other substitutions (D29N, AE72, and 73SD in pheasants and turkeys and P75S/A in chukar, guinea fowl, pheasants, and turkeys). NHE1 from chukar was the most similar to chNHE1, with only three single amino acid changes, ΔW38, P52H, and P75S. Of these changes, P52H was unique to chukar and cannot explain the resistance in the other species, and P75S was shared with susceptible turkeys.

Fig 2.

NHE1 alleles from various galliform species contain polymorphisms within the first extracellular loop (ECL1). The deduced amino acid sequences of TM1 and ECL1 corresponding to chNHE1 amino acids 23 to 104 are aligned and compared. The susceptibility or resistance of galliform species under study is denoted (+) or (−), respectively. The border between ECL1 and putative transmembrane domains TM1 and TM2 is shown. The W38 and P52 amino acid residues examined are denoted by vertical arrows. Amino acids matching the consensus sequence are on a gray background.

Apart from the ECL1 sequence, the NHE1 amino acid sequence is highly conserved. In our panel of galliform species, we found only a few substitutions, mostly in transmembrane regions and intracellular loops. Only three amino acid substitutions were found in extracellular loops: the S210T substitution was found in guinea fowl, I284M was found in chukar, guinea fowl, quail, and turkey, and I285V was found in chukar, quail, and turkey (data not shown). None of these changes discriminated the ALV-J-susceptible and -resistant species. Altogether, W38 seems to be crucial for susceptibility to ALV-J.

W38 of chNHE1 is a critical amino acid for ALV-J entry.

We tested the importance of W38 for subgroup J ALV entry by expression of wild-type and mutated tvj cDNA in resistant QT6 cells of Japanese quail and analysis of subsequent RCAS(J)GFP infection. We prepared an expression vector encoding the entire chNHE1 and a series of mutant chNHE1 sequences bearing single amino acid changes within the ECL1. The sequence of the chNHE1 ΔW38 mutant simulates the deletion observed in NHE1 of chukar, and the sequences of the W38E and W38G mutants simulate the effect of the W38 substitutions found in nonconserved clusters of NHE1 proteins from Japanese quail and guinea fowl, respectively (Fig. 2). Furthermore, the sequence of the P52H mutant simulates the unique substitution of P52 observed in NHE1 of chukar. The efficiency of transfection with individual expression constructs was normalized by the coexpression of fluorescent tdTomato from the same construct. Transfection of QT6 cells with a construct expressing wt chNHE1 conferred sensitivity to infection with RCAS(J)GFP (Fig. 3). We observed 36.4% GFP-positive cells among the transfected tdTomato-positive cells on the second day after infection. The P52H mutation conferred almost the same sensitivity to RCAS(J)GFP, which indicates that the P52H substitution does not abrogate the receptor activity of chNHE1 and is not attributable to the resistant phenotype of Japanese quail. In contrast, expression of the ΔW38 mutation of chNHE1 did not confer susceptibility to the ALV-J infection and the mutant did not produce any GFP-positive cells, similar to control transfection of the empty vector into QT6 cells. We conclude that W38 is a critical amino acid residue necessary for the receptor function of chNHE1, and its deletion observed in chukar explains the resistance to ALV-J in the respective birds. Interestingly, the W38G and W38E mutants of chNHE1 displayed an intermediate phenotype, achieving receptor activity 8% and 5.7% of the receptor activity of wt chNHE1, respectively (Fig. 3A). Therefore, the W38 substitutions do not completely abrogate the receptor activity of chNHE1, and the resistance of Japanese quail and guinea fowl cells probably requires the presence of the additional mutations observed in NHE1 in these species (Fig. 2).

Fig 3.

Expression of the cloned wt chNHE1 and its ΔW38, W38E, W38G, and P52H mutants confers various levels of susceptibility to ALV-J in QT6 cells. QT6 cells were transfected with constructs expressing the wt and mutant forms of chNHE1 and infected with RCAS(J)GFP on the next day. The percentage of GFP-positive cells was analyzed by flow cytometry 2 days postinfection. (A) Susceptibility to RCAS(J)GFP is given as the percentage of GFP-positive cells among the efficiently transfected tdTomato-positive cells. Results were calculated as the average of three parallel wells. (B) Examples of FACS histograms showing the percentages of GFP-positive cells after transfection of QT6 cells with wt chNHE1, the P52H, ΔW38, W38E, and W38G mutants, and empty vector and after infection with RCAS(J)GFP.

The W38 deletion does not affect the expression or cell surface display of chNHE1.

To exclude the possibility that the resistance of avian species with mutated W38 of NHE1 is caused by decreased display of the Tvj receptor on the surface of target cells, we compared the expression of the wt and ΔW38 NHE1 forms at the level of transcription and in the cytoplasmic protein fraction. The real-time RT-PCR detected a 2-fold higher level of ΔW38 NHE1 mRNA over that of wt NHE1 mRNA expressed from pVitrotdT-tvj in QT6 cells (Fig. 4A). The transfected QT6 cells were then used for separation of the cell surface protein fraction from the whole-cell lysate by biotinylation of intact cells and streptavidin affinity pulldown of labeled proteins. The NHE1-specific antibody detected approximately the same amount of the wt and ΔW38 forms both in the whole-cell lysate and in the biotinylated cell surface fraction (Fig. 4B). This clearly indicates that the W38 deletion does not affect the cell surface expression of the NHE1 protein. The endogenous NHE1 expression is very low, as the signal in mock-transfected QT6 cells (Fig. 4B) or in nontransfected chicken embryo fibroblasts (not shown) was hardly visible.

Fig 4.

Expression and cell surface display of the wt and ΔW38 forms of chNHE1. (A) QT6 cells were transfected with pVitrotdT-tvj plasmids bearing the wt and ΔW38 forms of chNHE1. The levels of wt and DW38 chNHE1 transcripts were estimated by qRT-PCR and are shown as relative levels of expression, where the level of expression by the ΔW38 mutant represents 100%. The endogenous expression of quail NHE1 is shown in mock-transfected QT6 cells. Data were calculated as the average ± SD from three triplicates. (B) QT6 cells transfected with pVitrotdT-tvj plasmids bearing the wt and ΔW38 forms of chNHE1 were surface biotinylated, and both cell surface protein fractions and control whole-cell lysates were subjected to SDS-PAGE and Western blot analysis using an anti-NHE1 monoclonal antibody. Coomassie blue-stained gels are shown as a loading control. The cell surface fraction was ca. 10 times less concentrated than the whole-cell lysates. The endogenous expression of quail NHE1 is shown in the whole-cell lysate of mock-transfected cells.

The chicken wt NHE1 form but not the chukar or quail NHE1 form specifically interacts with ALV-J Env.

Mutations in chNHE1 can either abrogate virus envelope-receptor binding or prevent the envelope from fusing with the target cell. In order to extend our analysis of ALV-J-susceptible and -resistant NHE1 forms, we constructed chimeric immunoadhesin SU(J)-rIgG composed of the subgroup J-specific SU subunit linked to the constant fragment (Fc) of rabbit IgG. This immunoadhesin was produced in the RCAS system (18) and allowed to bind cells either permissive (chicken) or resistant (chukar and quail) to ALV-J infection. Binding of SU(J)-rIgG was detected by fluorescein-conjugated anti-rabbit antibody. Whereas chicken DF-1 cells incubated with SU(J)-rIgG displayed a positive peak against the negative control, QT6 cells and chukar embryo fibroblasts did not bind SU(J)-rIgG (Fig. 5). We therefore conclude that W38 mutations in NHE1 are defective at the level of envelope-receptor binding.

Fig 5.

Specific binding of SU(J)-rIgG immunoadhesin to ALV-J-susceptible but not ALV-J-resistant cells. Chicken DF-1 cells susceptible to ALV-J infection and chukar embryo fibroblasts and QT6 cells, both resistant to ALV-J infection, were incubated with SU(J)-rIgG immunoadhesin and then Alexa Fluor 488-conjugated goat anti-rabbit secondary antibody. Binding of the immunoadhesin was measured by flow cytometry (black line). The results for the negative controls incubated without SU(J)-rIgG are given as shaded areas.

DISCUSSION

In the present study, we have demonstrated that a single amino acid deletion or substitution of W38 in ECL1 of chNHE1 abrogates or at least strongly decreases its capacity to be a receptor for ALV-J. Additionally, we have found corresponding deletions or substitutions in a panel of ALV-J-resistant avian species, either alone or in combination with other alterations of the ECL1. This altogether suggests that W38 of chNHE1 is a critical determinant of Tvj receptor function. We provide here another example of an aromatic residue critically required for the retrovirus-receptor interaction (23, 24, 34–36). Furthermore, our model generated for this study, the replication-competent RCAS(J)GFP reporter vector, may be a useful tool for testing for sensitivity to ALV-J.

Our localization of the critical receptor determinant within ECL1 is not surprising. Due to its length, ECL1 is probably prominent over the membrane-bound NHE1 structure and accessible for the attachment of the virus envelope. Comparison of the chicken and human NHE1 amino acid sequences pointed to the divergent N-terminal part of ECL1 and the conserved remainder of both proteins (19). Among the galliforms analyzed here, the divergences between susceptible and refractory species concentrated in a narrow region around W38. This clustering resembles amino acid substitutions in ECL4 of the XPR1 receptor invariantly found in multiple rodent species resistant to various xenotropic viruses (37). Interestingly, this region and the critical aromatic residue are close to the putative border between transmembrane domain 1 (TM1) and ECL1 and not on the top of ECL1. ALV-J might have evolved to avoid the receptor moieties critical for normal cellular function and the sites of interaction with natural binding partners. This is, however, speculative, as the exact structure of NHE1 and interactions of its extracellular domains are matters of discussion (38).

NHE1 is a Ca²⁺-controlled Na⁺/H⁺ exchanger that regulates intracellular pH and cell volume. It is ubiquitously expressed (39) together with multiple members of the NHE family, and the profound effects of its hypo- or overexpression suggest that it is a housekeeping gene essential for basic cellular functions. NHE1-null fibroblasts display reduced adhesion, loss of polarity, and greatly diminished motility and chemotaxis (40). In contrast, increased tumor growth and tumor cell invasion were observed in cells overexpressing NHE1 (41). In our transfection experiments, we also noted minor cytotoxicity in quail QT6 cells, which might have slightly decreased the percentage of RCAS(J)GFP-infected cells (Fig. 3). The lower sensitivity of cells induced by ectopic receptor expression in comparison with the natural sensitivity conferred by wt endogenous receptor was also observed in a tvc transfection experiment (8). Furthermore, the RCAS vectors of the J subgroup are less efficient than those of other subgroups in virus production and spread, reaching a titer of 10⁶ IU per ml (this study) or 10² IU per ml (33).

The deletion of W38 completely abrogates receptor activity and explains the resistance of chukar to the J subgroup of ALV; however, alleles with W38 replaced by G or E conferred weak susceptibility to the virus when overexpressed in our virus entry assay (Fig. 3A and B). W38 substitution, in contrast to W38 deletion, might keep the receptor conformation at least partially suitable for ALV-J binding. The strong resistance of guinea fowl and Japanese quail can then be explained in one of two ways. First, additional mutations actually observed around the substituted W38 could potentiate the effect of substituted W38. This could be tested by ectopic expression of combined mutations simulating the alleles from the resistant species. Alternatively, W38G and W38E substitutions could be sufficient for resistance at the normal level of expression but not under conditions of overexpression from the transfected construct. The problem of overexpression might be solved by a knock-in of W38 substitutions into the endogenous tvj locus in DT40 cells. W38 may be a part of the binding site and play a role in virus attachment to the receptor. On the other hand, the general importance of aromatic amino acids, and particularly tryptophan, for receptor functioning suggests that W38 contributes to a common conformational event leading to the virus-cell fusion process. Further deletion/mutation analysis and parallel examination of immunoadhesin-receptor binding and virus entry should be used to solve this question.

From the point of view of virus-host coevolution, it is tempting to speculate that the strong resistance to the J subgroup of ALV in most galliform species has been exerted by the selection pressure from an unknown virus. ALV-J arose by recombination of an exogenous virus with an endogenous counterpart of the ev/J family (18). This family is insertionally polymorphic, and at least some members contain transcribed gag and pol open reading frames (42). The most homologous ev/J 4.1 endogenous sequence encodes a complete envelope glycoprotein which pseudotypes murine leukemia virions and efficiently interferes with ALV-J (43). Endogenous alpharetroviruses have been reported in chicken, turkey, and grouse (Bonasa umbellus) but not in zebra finch (Taenopygia guttata) (44), and it is probable that all galliforms contain ev/J sequences, which might have driven the selection for resistant receptor alleles. The receptor-mediated interference between exogenous and endogenous retroviruses is a phenomenon also observed in the chicken RAV-0 (5, 6) and in murine ecotropic (45) and xenotropic (46) retroviruses. The Tvb^S3 allele, which confers resistance to the E subgroup of ASLV (6, 47), and the XPR1 alleles, which confer resistance to murine xenotropic retroviruses (28), are believed to be products of the selection pressure imposed by such endogenous retroviruses. Nonspecific restriction mechanisms should also be taken into account in the case of cross-species retrovirus infection. We have preliminarily infected chukar embryo fibroblasts with A subgroup virus RCASBP(A)GFP and observed virus spread (data not shown), which demonstrates the absence of strong general retrovirus restriction in this species.

ALV-J causes enormous economic losses in the poultry industry (21). Identification of the chNHE1 receptor determinants responsible for the sensitivity to ALV-J can help in finding natural polymorphisms in chNHE1 as a potential source of host resistance. If the chNHE1 allele with a W38 deletion or substitution does not exist in chicken populations, it might be prepared artificially by the use of current genome editing tools, such as sequence-specific transcription activator-like effectors or ZnF nucleases, and introduced into commercial breeds.