Use of CD134 As a Primary Receptor by the Feline Immunodeficiency Virus

See allHide authors and affiliations

Science  20 Feb 2004:
Vol. 303, Issue 5661, pp. 1192-1195
DOI: 10.1126/science.1092124


Feline immunodeficiency virus (FIV) induces a disease similar to acquired immunodeficiency syndrome (AIDS) in cats, yet in contrast to human immunodeficiency virus (HIV), CD4 is not the viral receptor. We identified a primary receptor for FIV as CD134 (OX40), a T cell activation antigen and costimulatory molecule. CD134 expression promotes viral binding and renders cells permissive for viral entry, productive infection, and syncytium formation. Infection is CXCR4-dependent, analogous to infection with X4 strains of HIV. Thus, despite the evolutionary divergence of the feline and human lentiviruses, both viruses use receptors that target the virus to a subset of cells that are pivotal to the acquired immune response.

The primary event in the process of viral entry into a target cell is the interaction between the virus and its cellular receptor, and the specificity of this interaction determines both viral cell tropism and pathogenicity. For the primate lentiviruses, the viral receptor is CD4, targeting the virus to helper T cells, resulting in their depletion and the eventual development of acquired immunodeficiency syndrome (AIDS) (1). However, CD4 expression alone is insufficient to confer susceptibility to infection with human immunodeficiency virus (HIV), which also depends on the expression of coreceptors, principally the chemokine receptors CXCR4 and CCR5 (2). The virus attaches via a high-affinity interaction with CD4, resulting in a conformational change in the envelope glycoprotein (Env) and exposing the binding site for the chemokine receptor (3). This then triggers a further conformational change that exposes the fusion domain of the viral transmembrane protein gp41 and enables fusion of the viral and cellular membranes (4).

The feline immunodeficiency virus (FIV) is unique among the nonprimate lentiviruses in that in its natural host species, the domestic cat, it induces a disease similar to AIDS, characterized by a progressive depletion of CD4+ T lymphocytes (5). Consequently, an immunodeficiency syndrome develops that is characterized by wasting, neurological manifestations, chronic stomatitis and gingivitis, and an increased incidence of lymphoma (6). However, CD4 is not the primary receptor for FIV, suggesting that the virus may bind to an alternative molecule in order to target CD4+ T cells (7). Partial elucidation of the mechanism of infection with FIV was provided by the discovery that primary isolates and laboratory strains of virus use CXCR4 as a cofactor for infection (8, 9). However CXCR4 expression alone was insufficient to confer susceptibility to infection with primary isolates of FIV, indicating the existence of an as yet unidentified primary receptor for the virus.

To identify the primary receptor for FIV, we generated a cDNA library in the pMX retroviral vector from an interleukin-2 (IL-2)–dependent CD4+ feline T cell line (MYA-1), which is highly susceptible to infection with FIV (10). After transduction of P3U1 murine myeloma cells with the library, cells capable of binding FIV were identified by “panning” with FIV-coated plates (11). Positive colonies were expanded, genomic DNA was prepared, and the cDNA insert was amplified from the retroviral vector by polymerase chain reaction. Of three cDNAs cloned from the library, only one showed reproducible binding to FIV when reexpressed in P3U1 cells and was characterized as the feline homolog of CD134 (DNA Data Bank of Japan accession no. AB128982). CD134 was first described as MRC OX-40, an antigen expressed on activated rat CD4+ T lymphocytes and a member of the tumor necrosis factor–nerve growth factor receptor family (12, 13). The feline CD134 cDNA identified here predicts a 270–amino acid protein with strong identity to human CD134 (Fig. 1A). To confirm that the feline CD134 protein binds FIV, we expressed the cDNA in the human cell line Plat-E and the murine cell line NIH-3T3. Specific FIV binding was detected on Plat-E and NIH 3T3 cells transduced with CD134 but not with cells expressing CD2 from the same vector (Fig. 1B).

Fig. 1.

Molecular cloning and characterization of feline CD134. (A) Alignment of feline and human CD134. Homologous amino acids are shaded light gray; identical residues are shaded dark gray. TM, transmembrane region. (B) Binding of FIV to CD134-expressing cells. Cell cultures transduced with CD134 or CD2 were incubated with culture supernatant from FIV-infected (thick line) or uninfected (thin line) cells, followed by FIV Env detection by flow cytometry.

To assess whether CD134 acts as a functional viral receptor, the feline large granular lymphoma cell line MCC (these cells are refractory to infection with primary strains of FIV) was stably transduced with a retroviral vector bearing the CD134 cDNA. Surface CD134 expression was detected by flow cytometry (Fig. 2A), whereas a ∼45-kD species was detected by immunoblotting (Fig. 2B). CD134 expression rendered MCC cells permissive for infection with the five primary strains of FIV tested (Fig. 2C), whereas cells transduced with vector only remained non-permissive. To quantify viral entry into the CD134-expressing cells, we prepared HIV pseudotypes bearing FIV Env proteins and carrying a luciferase gene. Each successful entry event would trigger luciferase production in the target cells. MCC-CD134 cells were permissive for infection with pseudotypes bearing the Env proteins of either the GL8 or TM2 strains of FIV (Fig. 2D), confirming that CD134 expression overcame the block to viral entry into these cells. CD134 expression also relieved the block to infection in the feline fibroblast line AH927, whereas a control vector encoding the feline leukemia virus (FeLV)–C receptor (FeLV-CR) had no effect (fig. S1).

Fig. 2.

CD134 expression renders feline cells permissive to infection with FIV. (A) Flow cytometric analysis of CD134 expression on MCC cells. Control antibody (solid histogram) and anti-CD134 (Ber-ACT35, open histogram) are shown. The percentage positive is shown. (B) Immunoblot analysis with rabbit antibody to rat OX40. MCC cells transduced with vector only (CON) were compared with cells transduced with the CD134 expression vector. A ∼45-kD species was detected in CD134-transduced cells (arrowhead). (C) Feline (f) CD134 expression renders MCC cells permissive of productive infection with primary strains of FIV. MCC-CON or MCC-CD134 cells were infected with the GL8, F0425, F0827, F0795, and F0556 primary isolates of FIV. Reverse transcriptase activity in the culture supernatants was measured at 1, 4, and 6 days after infection. (D) Luciferase activity in MCC-CON or MCC-CD134 cells infected with HIV (FIV) luciferase pseudotypes bearing the GL8 or TM2 Envs (mean ± SE, n = 4 replicates).

Syncytium formation occurs when Env expressed on the surface of a cell comes into contact with an adjacent cell expressing the viral receptor and coreceptor (4). Thus, transfection of cells expressing the viral receptor and coreceptor with env expression vectors should trigger syncytium formation. Accordingly, transfection of AH927-CD134 cells with constructs bearing the env genes from primary isolates (Fig. 3, E to G) led to the development of syncytia, whereas no syncytia were observed in the control cultures (Fig. 3, A to C) expressing the FeLV-CR. In contrast, FIV-PET env [derived from a FIV strain analogous to CD4-independent strains of HIV (14)] induced syncytium formation in both the CD134 (Fig. 3H) and FeLV-CR–expressing (Fig. 3D) cells, consistent with this Env interacting with CXCR4 in the absence of a primary receptor (15).

Fig. 3.

CD134 expression renders AH927 cells permissive for syncytia formation mediated by FIV Env. FeLV-CR–expressing cells (A to D) or CD134-expressing cells (E to H) were transfected with pVR1012 expressing the GL8 [(A) and (E)], F0425 [(B) and (F)], F0827 [(C) and (G)], or PET [(D) and (H)] env genes.

Human cells are refractory to infection with primary strains of FIV (16). We therefore examined the species specificity of the interaction between FIV and CD134. The human cell line HeLa was stably transduced with either feline or human CD134, and surface expression of CD134 was confirmed by flow cytometry (Fig. 4B). Although ectopic expression of feline CD134 rendered HeLa cells permissive for infection with primary FIV strain GL8 or TM2 pseudotypes (Fig. 4A), human CD134 expression had little effect. In contrast, the susceptibility of the feline and human CD134-expressing cells to HIV pseudotypes with a vesicular stomatitis virus glycoprotein (VSVG) envelope was broadly similar to that of the control HeLa cells. Similarly, feline, but not human, CD134 expression permitted infection with HIV pseudotypes bearing FIV envelopes and encoding green fluorescent protein (GFP) (Fig. 4B). These results suggest that the specificity of the interaction between the virus and its primary receptor may be a major determinant of the species specificity of FIV.

Fig. 4.

CD134-mediated infection is species specific and CXCR4-dependent. (A) HeLa cells transduced with retroviral vectors bearing feline or human (h) CD134 cDNAs or a vector-only control were infected with HIV (FIV) (GL8 or TM2) or HIV (VSV) luciferase pseudotypes (mean ± SE, n = 4 replicates). (B) CD134-dependent infection of HeLa cells expressing feline CD134 by HIV (FIV) pseudotypes bearing a GFP marker gene. CD134 and GFP expression were analyzed by flow cytometry. (C) HeLa cells expressing feline CD134 were infected with HIV (FIV) (GL8 or TM2) or HIV (VSV) luciferase pseudotypes in the presence of CXCR4-antagonist AMD3100 (0, 10, 100, or 1000 ng/ml) (mean ± SE, n = 4 replicates).

Because infection with both primary and cell culture–adapted strains of FIV is CXCR4-dependent (8, 9), we determined whether CD134-dependent infection required coexpression of CXCR4. HeLa cells express high levels of CXCR4, and human CXCR4 is an efficient coreceptor for FIV (17). Infection of HeLa-CD134 cells with GL8 and TM2 pseudotypes was inhibited efficiently by the selective CXCR4 antagonist AMD3100, whereas infection with control HIV (VSV) pseudotypes was not affected (Fig. 4C), confirming an absolute requirement for both receptor and coreceptor.

Further evidence for a role for CD134 in viral infection was provided by downregulation of CD134 from the surface of FIV-infected cells (fig. S2). Moreover, CD134-expressing cells expressed levels of CXCR4 similar to those expressed by control cells, indicating that CD134 did not mediate its effect on FIV infection by modulating CXCR4 expression (fig. S3).

We have shown that CD134 functions as a primary receptor for an immunodeficiencycausing lentivirus. CD134 expression is largely restricted to CD4+ T lymphocytes (12, 1820); however, in humans and mice, CD134 is also expressed at lower levels on activated CD8+ T cells (18, 21), macrophages, and activated B cells (22). CD4+ T cells are the primary target for FIV in early infection, whereas in chronic infection CD8+ T cells and B cells are also infected (23, 24). The tropism of FIV in vivo therefore appears to be consistent with the predicted expression of CD134. In addition, the viral coreceptor CXCR4 is expressed widely in the cat [in activated T cells, B cells, and monocytes (25)], and because some primary and cell culture–adapted strains of FIV can infect via CXCR4 alone (26) (CD134-independent infection), the broadening of cell tropism of the virus in chronic infection may represent a shift toward CD134-independent infection.

Signaling through CD134 plays a crucial role in the survival and proliferation of CD4+ T cells that have encountered antigen (20). By targeting CD134-expressing cells, FIV would selectively deplete a subset of CD4+ T cells that is integral to the development of antigen-specific T cell responses. In contrast, by using CD4 as a primary receptor, HIV has the potential to infect all CD4+ T cells and induce a more profound immune defect. However, the cell tropism of HIV is restricted by the expression of the viral coreceptor which, for the majority of strains that are transmitted, is CCR5. CCR5 expression on CD4+ T cells is restricted to an effector/memory T cell subset (27, 28). Thus, despite the use of distinct primary binding receptors, both the human and feline viruses selectively impair antigen-specific helper T cell responses.

Vaccination may lead to enhancement of infection in the feline model of AIDS (29). Our new data may shed new light on the mechanism of enhancement, because CD134 is a T cell activation antigen, with expression in vivo restricted predominately to CD4+ T cells. Vaccination may induce an expansion of a population of cells expressing the viral receptor, so that if sterilizing immunity is not achieved, vaccination may prove counterproductive.

That two lentiviruses with host species as divergent as human beings and the domestic cat should use distinct primary receptors to target similar T cell subsets underlines the central role of CD4+ T lymphocyte infection in the pathogenesis of AIDS. Whether the feline and human lentiviruses evolved from a common ancestor, such as a CD4- or CD134-independent virus, is an intriguing question regarding the development of viral virulence, and this study represents a first step toward providing a solution.

Supporting Online Material

Materials and Methods

SOM Text

Figs. S1 to S3


References and Notes

View Abstract

Navigate This Article