Improved health and survival of FIV-infected cats is associated with the presence of autoantibodies to the primary receptor, CD134

Abstract

We analyzed antibody responses in sera from feline immunodeficiency virus (FIV)-infected and uninfected cats. A strong antiviral response to the viral surface glycoprotein (SU) was noted in both natural and experimental infections. In addition, 143 of 226 FIV-infected animals (63%) also expressed antibodies to the primary binding receptor, CD134, whereas cats infected with other feline RNA viruses, including calicivirus, coronavirus, herpesvirus, and feline leukemia virus, did not. Both affinity-purified anti-CD134 and anti-SU antibodies blocked FIV infection ex vivo. FACS analyses revealed that the anti-CD134 antibodies bound to a cryptic epitope on the receptor that was only exposed when SU bound to CD134. Anti-CD134 binding caused displacement of SU from the surface of the cell and inhibition of infection. The presence of antibodies to CD134 correlated with lower virus loads and a better overall health status in FIV⁺ cats, whereas anti-SU antibodies were present independent of health status. The findings are consistent with a role for antireceptor antibodies in protection from virus spread and disease progression.

Feline immunodeficiency virus (FIV) is a lentivirus associated with an AIDS-like syndrome in the domestic cat (1). Like HIV, FIV can be transmitted via mucosal exposure, blood transfer, and vertically via prenatal and postnatal routes (2–5). Progression of FIV disease follows a pattern typical of that observed with primate lentiviruses, starting with a relatively short (weeks) acute phase denoted by increasing viral loads, febrile episodes, weight loss, lymphadenopathy, and neutropenia. Many of these early symptoms resolve as the animals proceed into an oft-protracted (years) latent phase denoted by relatively strong antiviral immune responses, lower viral titers, a gradual decline in CD4⁺ cells, and minimal clinical symptoms. The terminal phase of infection is marked by immunologic decompensation, exacerbation of plasma viral load, and clinical symptoms of immunodeficiency with opportunistic infections (1, 6). Lymphoid tissue alterations are consistent with those in HIV and SIV infections and include thymic depletion, lymphoid hyperplasia, plasmacytosis, and terminal lymphoid depletion (6–12). Neurologic manifestations are also evident (13–18), including delayed auditory evoked and visual evoked potential changes (15, 16) and marked alterations in sleep patterns (18, 19). As the disease progresses, decline in the number of CD4⁺ T cells continues, with an ultimate increase in viral load in the later stages of the disease (3, 20, 21). Animals, if not euthanized, generally die of opportunistic infections or lymphomas (22).

As outlined above, there are many parallels between the outcomes of infection with FIV in cats and infection of humans by HIV, despite the differences in genomic sequences and certain viral genes. The primary target cell populations are the same, and thus both viruses are confronted with similar host obstacles for replication and for dealing with innate and adaptive immune responses. The two lentiviruses have evolved along unique pathways that have led to development of alternative mechanisms to deal with certain aspects of replication, including transcriptional transactivation, uracil misincorporation, and cell entry. Of direct relevance to the present study, both HIV and FIV have adopted a “2 receptor” mechanism to optimize entry into the cell. The first interaction is with a primary binding receptor, which in turn enhances interaction with the entry receptor. HIV uses CD4 as a primary binding receptor for high-affinity interaction with its entry receptor, whereas FIV accomplishes the same goals using CD134 as the primary binding receptor (23–25). However, both FIV and certain HIV share the utilization of the chemokine receptor CXCR4 as an entry receptor (26, 27). Parallels in the role of the binding receptor in facilitating interaction with CXCR4 in the 2 systems are striking and imply a strong selection for similar mechanisms of entry and evasion of immune surveillance.

An important advantage to studies involving controlled infections in the cat is the ability to assess the temporal humoral and cellular responses that occur as the infection progresses. In the present study, we evaluated the rate of appearance, degree, and specificity of the humoral response to FIV infection in a large (>600) collection of cat sera accumulated over the past 30 years. The results demonstrate a strong antiviral humoral response to multiple viral antigens, with appearance of expected neutralizing antibody responses to the major viral surface glycoprotein (SU). However, in addition to antiviral antibodies, we also found antibody responses to the primary binding receptor for FIV, CD134, which was included in the screen as an additional nonviral antigen. Importantly, affinity-purified anti-CD134 antibodies blocked FIV infection ex vivo at titers approaching neutralization titers of anti-SU antibodies affinity purified from the same animals. Furthermore, anti-CD134 titers correlated with lower viral loads and improved health status of infected cats.

Results

ELISA assays were performed on sera from cats experimentally infected with the Petaluma strain of FIV to compare the relative antibody response against a number of viral proteins and peptide regions of Env, but also a soluble form of the primary binding receptor for the virus, CD134 (Table 1). As expected, no responses to the viral proteins were observed in 23 uninfected control cats, and all 24 infected cats made robust antibody responses to SU. Of the 24 infected animals, 17 expressed antibodies reactive with a peptide corresponding to the V3 loop, previously shown to be a target for neutralizing monoclonal antibodies (28, 29). Eight cats had antibodies that recognized a short 12-aa epitope (N212) within the V3 loop that is a binding site for CD134-dependent neutralizing mouse monoclonal antibodies (29). Moderate to strong antibody responses to an immunodominant peptide region (30) of the transmembrane protein (TM) were noted in all 24 infected cats, consistent with the immunogenic nature of this region of TM reported in previous studies (30). A surprising result was that 21 of the 24 FIV-infected cats had antibodies reactive to CD134, whereas only 1 of 23 control cats showed mild reactivity to the CD134 preparation (Table 1). These cats had been infected for 30 months when tested and were deemed sufficiently healthy to receive a challenge dose of feline herpesvirus (FHV) in a follow-up study (31).

Table 1.

Antibody reactivities in control and FIV-infected cats

Parameter	N212	Tm‡	V3	SU	CD134
Uninfected control (n = 23)	0*	0	0	0	0
FIV-Petaluma-infected (n = 24)	8	24	22	24	21
Uninfected
SPF (n = 107)	nd	nd	nd	0	0
Other viruses+ (FHV, FIP, FeLV) (n = 225)	nd	nd	nd	6	1
FIV infected
Pet cats (n = 108)	nd	nd	nd	108	61
Experimental (n = 118)	nd	nd	nd	118	82
Pet cats of known health status†
Asymptomatic (n = 20)	nd	nd	nd	20	16‡
Moderate (n = 33)	nd	nd	nd	33	22
Severe (n = 23)	nd	nd	nd	23	9

N212, synthetic peptide encompassing the CD 134-dependent neutralizing antibody epitope on SU (29); TM, synthetic peptide corresponding to an immunodominant region of the transmembrane protein (30); V3, synthetic peptide corresponding to the V3 region of FIV SU (29); SU, purified SU-Fc immunoadhesin glycoprotein corresponding to FIV-34TF10 SU; CD134, purified soluble feline CD134-Fc immunoadhesin (33); SPF, specific pathogen-free; nd, not determined.

*Arbitrary cutoffs for OD values: negative = 0–0.199; positive = >0.2.

^†Health status: asymptomatic: FIV positive but showing no adverse signs; moderate: non–life-threatening but persistent symptoms, such as mild gingivitis, stomatitis, upper respiratory infection or cystitis, weight loss (<15% body weight), and transient lymphadenopathy; severe: life-threatening or fatal diseases, including emaciation (>25% body weight), nonresolvable or recurrent and protracted infections, nonregenerative anemia, lymphomas.

^‡P values were determined for the presence of anti-CD134 antibodies and health status of the of the cats using Fisher's exact test. For cats demonstrating asymptomatic vs. severe disease, P = 0.0124; moderate vs. severe disease, P = 0.0574. The difference between asymptomatic vs. moderate disease was not significant (P = 0.3590).

ELISA assays were subsequently performed on >600 cat serum samples, accumulated over the past 30 years from studies of feral and pet cats in and around the Seattle, Washington and San Diego, California areas, as well as from several controlled experimental infection studies (Table 1). Of the cats included in the study, 332 were either uninfected specific pathogen-free cats; cats infected with other viral pathogens, including FHV, feline coronavirus (FCoV/FIPV), or feline leukemia virus (FeLV); or pet cats with various maladies but previously tested as FIV negative. Another 226 cats were either experimentally infected with different strains of FIV or presented at veterinary clinics with FIV infections (32). Only 1 pet cat of the 332 cats reported as FIV negative expressed antibodies reactive to CD134. In contrast, 143 of the 226 confirmed FIV-infected cats (63%) had antibodies to CD134 (Table 1).

To further test the specificity of the anti-SU and anti-CD134 responses, pooled serum samples were prepared from FIV-positive cats that did or did not have anti-CD134 antibodies. Affinity columns were then prepared that contained either covalently bound SU or CD134 proteins that had been expressed and purified as immunoadhesins (containing a human Fc tag) in CHO cells. The pooled cat sera were then sequentially passed over the CD134 column, followed by passage over the SU column. Antibodies bound to each column were then eluted and specificity assessed by Western blot analysis (Fig. 1). Antibodies recovered from the anti-CD134 positive serum pool using CD134 beads as an adsorbent reacted by Western blot to CD134 but not SU (Fig. 1, left 2 lanes), and antibodies recovered from the serum pool using the SU beads reacted to SU but not CD134 (Fig. 1, right 2 lanes). The serum pool from FIV-positive cats lacking anti-CD134 yielded only antibodies to SU, and no reactivity was noted to the human Fc tag.

Fig. 1.

Western blot against SU and CD134 using affinity-purified anti-SU and anti-CD134 from FIV-infected cats. A pool (≈50 mL) was made of sera from FIV-infected cats and passed over an affinity column of immobilized soluble CD134. Bound antibodies were eluted as detailed in Materials and Methods and concentrated for further analyses, including Western blots shown here. The serum pool was subsequently passed over a column bearing immobilized FIV SU, and bound antibodies were eluted and concentrated for further use. The 2 left lanes show Western blots using the anti-CD134 affinity-purified antibody used against SU and CD134, respectively; the 2 right lanes are Western blots against the same antigens using affinity purified cat anti-SU. No cross-reactivity of these 2 antibody pools is evident.

The affinity-purified antibodies to CD134, as well as the 2 preparations of anti-SU from anti-CD134⁺ and anti-CD134⁻ serum pools, were then tested for ability to interfere with virus infection ex vivo (Fig. 2A). Anti-SU antibodies from both serum pools blocked FIV-PPR infection (only anti-SU from anti-CD134⁺ cat pool shown) with >80% inhibition down to 3 μg/mL in culture supernatant. Of significance, the anti-CD134 antibodies also prevented virus infection. Titration experiments (Fig. 2B) comparing the ability of the anti-SU and anti-CD134 affinity-purified antibodies to block FIV infection demonstrated 50% tissue culture infectious dose (TCID₅₀) values of 1.8 μg/mL and 0.2 μg/mL for anti-CD134 and anti-SU antibodies, respectively.

Fig. 2.

Blocking/neutralization of FIV infection by cat anti-CD134 and anti-SU antibodies. (A) Infections using FIV-PPR on 104-C1 T cells were performed in the continued presence or absence of either 12.5 μg/mL cat anti-CD134 antibody or cat anti-SU antibody, prepared as detailed in Materials and Methods. Progression of the infection was monitored by measuring RT activity in the culture supernatants over time, as detailed in Materials and Methods. (B) Titration experiments to determine the relative TCID₅₀ of anti-CD134 and anti-SU affinity-purified antibodies.

FACS analyses were performed to characterize and compare anti-SU and anti-CD134 affinity-purified cat antibody binding to cells in the presence or absence of bound SU (Fig. 3). The 104-C1 cell line (CD134^hi, CXCR4^lo) was used as a target for measure of CD134 binding (24, 33). Remarkably, the affinity-purified cat anti-CD134 antibodies bound to cell-surface CD134, but only after binding of SU to the cells, indicating that the antireceptor antibodies recognized a discrete epitope on cell-associated CD134 that remained cryptic until SU bound to the receptor (Fig. 3 A–C). This epitope must be exposed on the soluble CD134 used to purify antibodies from the serum pool. Premixing of anti-CD134 antibody with SU did not influence binding of the cat anti-CD134 to cell-associated CD134, indicating that SU does not interact directly with the cat anti-CD134 antibody.

Fig. 3.

FACS analyses on live CD134⁺ 104-C1 cells using anti-CD134 antibody in the presence and absence of soluble SU-Fc adhesin. (A–C) Detection of anti-CD134 antibody binding using anti-cat antibody. (D–F) Detection of SU-Fc binding using anti-Fc antibody. (A and D) anti-CD134 antibody only; (B and E) SU-Fc only; (C and F) anti-CD134 and SU-Fc. The anti-CD134 antibody does not recognize CD134 on the cells (B) until SU is bound to CD134 (C); when anti-CD134 antibody binds to the CD134-SU complex, SU is released from the cell (compare E and F).

By using an anti-Fc antibody, the binding of the SU-Fc immunoadhesin could be monitored within the same experiment (Fig. 3 D–F). The Fc tag is of human origin, and the antibody conjugate used for detection in FACS analyses is highly specific for human Fc and does not react with cat antibodies. Thus, binding of both SU and cat anti-CD134 antibodies could be monitored in the same experiment. SU bound strongly to cell-associated CD134 (Fig. 3E), but was displaced (≈60%) from the cell surface concomitant with the binding of anti-CD134 (Fig. 3F). This occurred regardless of whether the cat anti-CD134 was added before or after SU addition to the cells, indicating that binding of the cat anti-CD134 antibody caused dissociation of SU bound to CD134 instead of directly blocking SU binding to cell-surface CD134 (i.e., SU binds to cell-surface CD134, exposing an epitope for interaction with cat anti-CD134 antibody). The binding of anti-CD134 antibody then displaces bound SU.

Binding of the affinity-purified cat anti-SU antibodies to cells was also assessed. The 3201 cell line (CD134^nil, CXCR4^hi) was used in FACS analyses to assess the influence of affinity-purified antibodies on SU-CXCR4 interactions (Fig. 4A) in comparison with binding to CD134 on 104-C1 cells (Fig. 4B). SU strongly binds to CXCR4 on 3201 cells, and binding is inhibited by the CXCR4 antagonist AMD3100 (24, 33). The affinity-purified cat anti-SU antibodies bound to SU that had been prebound to CXCR4 on the cell surface (Fig. 4A, SU plus cells). Preincubation of SU with anti-SU diminished SU binding to CXCR4, indicating the presence of a subset of antibodies that block SU association with CXCR4. These antibodies likely facilitate the virus neutralization by this antibody preparation (Fig. 2). Consistent with the lack of CD134 expression on these cells, the affinity-purified cat anti-CD134 failed to bind in FACS analysis, regardless of the presence of SU on the cell surface, and had no influence on SU binding to CXCR4 (Fig. 4A). In contrast, the anti-SU antibodies recognized SU bound to CD134 on 104-C1 cells (Fig. 4B), but binding of SU to CD134 was not diminished by pretreatment of SU with anti-SU, indicating a lack of antibodies to SU at the CD134 interaction site. Again, as shown in the Fig. 3, anti-CD134 antibody caused the release of SU from CD134 (Fig. 4B).

Fig. 4.

Influence of anti-CD134 and anti-SU antibodies on SU-Fc binding to CXCR4. (A) Live 3201 cells (CD134^nil, CXCR4^hi) were used as targets for SU-Fc binding in FACS analyses in the presence and absence of affinity-purified antibodies, using anti-human Fc antibody to detect SU-Fc binding. The cat anti-SU antibody partially blocked SU binding to CXCR4, but the anti-CD134 antibody had no influence on binding to the chemokine receptor. (B) Identical analyses performed using the CD134^hi, CXCR4^lo 104-C1 cells. The anti-SU antibody had negligible influence on SU binding to CD134. As shown in Fig. 3, the anti-CD134 antibody caused release of SU bound to CD134.

Among cats of known health status, the highest frequency of sera positive for CD134⁺ antibodies were from FIV-infected cats that were asymptomatic, clinically healthy, and considered long-term survivors (Table 1). These cats served as the source for preparation of affinity-purified anti-CD134 antibody. In contrast, the CD134⁻ pooled serum was derived from cats that were clinically ill and/or that had required euthanasia. Analyses by Fischer's exact test indicated that differences were statistically significant as to anti-CD134 antibody expression between both the asymptomatic and moderate groups relative to the severely ill group, but not between the asymptomatic and moderate groups (Table 1).

We were able to assess the relationship between health status and relative viral load and the presence or absence of anti-CD134 antibodies in a retrospective analysis of experimentally infected cats (34). Cats were divided into 3 groups according to relative viral load (Fig. S1), and sera collected throughout the course of the experiments from each group were assessed for anti-CD134 and anti-SU antibody titers (Fig. 5). Of the 20 FIV-infected cats in the study, 8 cats fell into the high viral load group, which was associated with rapid disease onset and high mortality (Fig. 5 A and D); 6 cats exhibited moderate viral loads (Fig. 5 B and E); and 6 cats had low relative viral loads (Fig. 5 C and F). Cats in the latter 2 groups survived the experimental virus challenge. All 3 groups demonstrated strong anti-SU responses over time, typically peaking at approximately 6–8 weeks after infection. In contrast, the high viral load animals showed no anti-CD134 antibody response (Fig. 5D); 1 of 6 cats in the intermediate viral load group showed an anti-CD134 antibody response that peaked at 12 weeks after infection (Fig. 5E). Consistent with the trends noted in the larger cat study (Table 1), 4 of 6 cats in the low virus load group demonstrated strong anti-CD134 responses (Fig. 5F).

Fig. 5.

Anti-SU and anti-CD134 responses over time as a function of viral load in cats experimentally infected with FIV-C. Animals were divided into 3 groups according to relative peak viral load (Fig. S1), then serum samples were analyzed by ELISA for either anti-SU or anti-CD134 antibodies throughout the course of the experiment. The CD134 antibody response is associated with cats having low viral load. (A–C) Anti-SU antibody; (D–F) anti-CD134 antibody.

Discussion

The findings of the present study demonstrate that a majority of cats infected with FIV mount an antibody-mediated immune response against the primary binding receptor, CD134. Furthermore, the response is highly specific in that exposure of the reactive epitope(s) on CD134 is dependent on initial binding of the SU viral glycoprotein to the receptor. SU is not part of the epitope, and binding of the antibody to CD134 results in release of SU from the cell surface. We considered the possibility that the cat anti-CD134 antibodies might be anti-idiotypic antibodies generated against anti-SU antibodies. However, the findings are inconsistent with the notion that these antibodies mimic the SU–CD134 interaction site, the target that would generate a receptor mimic. The antibodies do not directly block SU binding to CD134. Instead, the results imply that 2 conformational changes occur in cell-surface CD134; the first upon binding of SU, which exposes the cryptic epitope recognized by the cat anti-CD134 antibody. Binding of anti-CD134 in turn induces a second conformational change that causes release of SU from the receptor. This induced release of SU explains the important observation that the cat anti-CD134 antibodies block virus infection ex vivo at concentrations approaching that required for direct neutralization by anti-SU antibodies from the same animals, but not by direct blocking of SU (virus) binding. A subset of anti-SU antibodies from the same animals also neutralize virus infection, but likely by the more traditional mechanism of direct blocking of receptor interactions. Indeed, FACS analyses indicate that these antibodies primarily act by blocking CXCR4 entry receptor interactions (Fig. 4A) and fail to block interaction with CD134 (Fig. 4B). Coupled with the observation that the cats expressing anti-CD134 antibodies are clinically more healthy, survive the infection better, and have lower viral loads, the findings suggest that the antireceptor antibody response contributes to controlling and/or reducing viral infection.

A majority of the FIV⁺ cat sera tested were collected from the Seattle and San Diego areas, but approximately 15% samples were obtained from more diverse regions (United Kingdom, Massachusetts, New York, Los Angeles); it is noteworthy that all sera reacted in a similar manner to the SU-derived peptides, SU constructs, and purified CD134, and this finding shows that FIV variation according to geographic boundaries had minimal influence on the results. Antibodies to Gag and Pol proteins were also noted in virtually all FIV-infected cats, independent of disease status and anti-CD134 antibody titers.

One concern might be that induction of autoimmune responses such as to CD134 during FIV infection might have deleterious effects on the host. Given the polyclonal B cell hyperplasia that characterizes FIV infections (35), there may be a panoply of other antibody responses occurring that are not seen, owing to the limited target repertoire used in most analyses. However, in regard to antibodies to CD134, the cryptic nature of the epitope may protect the animal from adverse responses to the presence of these antibodies. As with FIV infection of cats, HIV infection of humans is associated with a polyclonal B cell hyperplasia. Part of the resultant humoral antibody repertoire involves autoreactive antibodies, including antibodies to the primary binding receptor, CD4 (26, 36–38), as well as the main entry receptor, CCR5 (39). Early concerns regarding anti-CD4 antibodies centered around the potential for autoimmune syndromes, particularly in patients receiving therapeutic doses of recombinant CD4 as a potential treatment for infection (36). No adverse effects were noted, however, and no reports emerged to suggest that anti-CD4 antibodies interfered with or reduced HIV infection ex vivo. However, assays using purified anti-CD4 to determine whether these antibodies interfered with infection were not performed. Chams et al. (36) alluded to the possibility of adverse effects from anti-CD4 antibodies, but as with the present study, the receptor epitope was not exposed on live cells, leaving them to conclude that anti-T cell responses were likely not an issue. Although it was suggested that reactivity might depend on SU binding to CD4, they did not indicate that the analysis was performed. Anti-CCR5 antibodies have also been noted in association with HIV infection (39) and are thought to result in lower levels of surface CCR5 expression, with concomitant lower viral loads. Induction of anti-CCR5 autoantibodies has been proposed as an antiviral strategy for HIV-1 (40).

A recent study of SIV infection of macaques suggests that at least in an experimental infection, autoantibodies to several self-antigens are present, and it is suggested that these antibodies may contribute to CD4 cell depletion (41). The study showed a correlation between CD4 cell decline and presence of antibodies on the surface of both CD4⁺ and CD8⁺ T cells. It was not determined whether any of the observed antibodies were to either virus binding or entry receptors, and the response seems to be greater than that observed in naturally infected cats or reported results with HIV-infected humans. Further examination using a battery of potential antibody targets is, however, warranted.

Our data provide direct evidence that FIV entry into host cells can be prevented by 2 distinct antibody-mediated mechanisms: (i) classic direct virus neutralization by anti-SU antibodies to appropriate envelope epitopes, and (ii) by autoimmune antibodies to the primary virus receptor, CD134, which are induced by FIV infection and which bind to the “altered” cell membrane–bound receptor, thereby competitively inhibiting binding of virus. These 2 types of antibody responses are not dissimilar in their efficacy (as measured by antibody concentrations required to effect their distinct functions). It seems, however, that only the presence of persistent high titers of autoantibody to CD134 correlate with lower virus loads in plasma and also with a more favorable long-term health status.

Materials and Methods

Cell Lines, Reagents, and Virus.

The primary feline T cell line, 104-C1, was isolated by limited dilution cloning from feline PBMCs and has the receptor expression phenotype, CD134^high, CXCR4^low. The cells were maintained in RPMI MEDIUM 1640 supplemented with 10% FBS (Innovative Research), 2 mM L-glutamine (Sigma), 1 mM sodium pyruvate (Sigma), 10 mM Hepes (Sigma), MEM-vitamins (Mediatech), nonessential amino acids (Sigma), β-mercaptoethanol (Gibco-BRL), 2.5 μg/mL Con A(Sigma), and 50 U/mL of human recombinant IL-2 (Hoffman-La Roche) per mL, and 50 μg/mL gentamicin (Gemini Bioproducts) per mL. The feline lymphoma cell line, 3201, was obtained from W. Hardy (Sloan-Kettering Memorial Hospital) and has the receptor expression phenotype CXCR4^high, CD134^nil (24). The cells were cultured in the same medium as 104-C1 cells but do not require Con A and IL-2. GFox cells are CXCR4⁺ Crandell feline kidney cells that have been stably transfected with the primary binding receptors feline CD134 and are thus productively infectable by field strains of FIV (24). FIV-PPR is a molecularly cloned clade A isolate (42).

Cat Serum Samples.

Sera from cats were collected between 1978 and 2008 and were stored continuously at −80 °C. Sera sources included private multicat households with ongoing endemic natural FeLV infections that were monitored for up to a decade and retrospectively were found to be coinfected with FIV (22, 43); a randomized survey of >650 pet cats that visited participating Seattle veterinary clinics for any reason during a 2-month period in 1988 and that were retrospectively tested for FeLV and FIV infections (32); and sick cats that presented at the Pacific Northwest Research Foundation Retrovirus Clinic with FeLV, FIV, or FIPV/FCoV-related diseases (22, 43). Experimentally infected, specific pathogen-free cat samples included sequential sera from FIV Petaluma strain–infected cats, some of which were subsequently infected with FHV (31); FIPV/FCoV-infected cats, some of which were proven persistent coronavirus secretors; cats infected with the Glasgow 8 strain of FIV and studied over a 5-year period (44); and cats infected with FIV clade C over a 5-month period (34).

Infection and Neutralization Assays.

104-C1 cells (10⁵ per well) were seeded in a 96-well microtiter plate and incubated with 2.5 μg/mL of cat anti-CD134 or cat anti-SU affinity-purified IgG for 30 min at room temperature. Cells were then infected with 4 μL of virus stock for 2 h at room temperature before the viral supernatant was removed and replaced with fresh media. Virus production was measured over time using a micro-RT assay. Briefly, cell-free supernatants (50 μL) were incubated with 10 μL of lysis buffer (0.75 M KCl, 20 mM DTT, and 0.5% Triton X-100) for 10 min. Enzyme mixture (40 μL) containing 100 mM Tris·HCl (pH 8.1), 10 mM MgCl₂, 2.5 μg poly(rA)-poly(dT)_12–18 (Amersham Biosciences), and 2.5 μCi [³H]dTTP (Dupont) was then added, followed by a 2-h incubation at 37 °C. RT activity was measured as described in ref. 33.

For titration of affinity-purified cat anti-CD134 and anti-SU antibodies, GFox cells (2 × 10⁴ per well) were seeded in 48-well microtiter plates and incubated with 12 μg/mL, 6 μg/mL, or 3 μg/mL of either anti-CD134 or anti-SU for 30 min, followed by addition of 500 TCID₅₀ FIV-PPR to each well. The cells plus virus were spinoculated (45) for 2 h at room temperature at 1,000 × g. The viral supernatant was then removed and replaced with fresh media with the appropriate concentration of antibody. Virus production was then measured over time using the micro-RT assay described above, starting from the third day after infection. Fresh media plus diluted antibody were added after sampling each time point. Assays were also performed using an initial single exposure of antibody for comparison with results whereby antibody was continually present in the culture supernatant. Spinoculations were performed as detailed above, then the viral supernatant was removed, and the cells were washed once with Earle's balanced salt solution before fresh media were added. No additional antibodies were added after the initial exposure.

Flow Cytometry.

Binding of cat antibodies to 3201 (CXCR4 binding) or 104-C1 cells (CD134 binding) was analyzed by flow cytometry. Cells (3 × 10⁵) were incubated with 300 ng gp95-Fc, prepared as described in ref. 24, for 30 min at room temperature. The cells were then washed twice with 2% FBS diluted in PBS to remove unbound gp95. A final concentration of 12.5 μg/mL of cat anti-CD134 or cat anti-SU was added to the cells and incubated for 45 min at room temperature before being washed twice as above. A 1:100 dilution of GPB2–2B1, a monoclonal mouse anti-cat IgG antibody (Custom Monoclonals) was incubated with the cells for 45 min, then the cells were washed twice with PBS. A 1:1,000 dilution of a phycoerythrin-conjugated goat anti-mouse antibody (Cappel) was incubated with the cells for 45 min, at room temperature, in the dark. The cells were then washed twice with PBS and analyzed by flow cytometry on a FACScan2 (Beckton Dickinson) using the FLOWJO (Tree Star) software program.

Affinity Purification of Antibodies.

FIV-34TF10 SU-Fc and soluble feline CD134 were expressed and purified from CHO cells as described in ref. 33. The proteins were then coupled at ≈700 μg/mL to Actigel ALD superflow support matrix (Sterogen) according to the manufacturer's protocol and washed extensively with PBS. Two serum pools (50 mL each) were prepared from FIV-infected cats that were either positive or negative for the presence of anti-CD134 antibodies. Each serum pool was then passed over separate CD134 columns (2-mL bed volume), and after extensive washing with PBS, bound antibodies were eluted using citrate elution buffer (pH 1.2). The eluted antibodies were adjusted to pH 8 with 2 M Tris buffer, then concentrated using a centrifuge concentrator (Millipore). The serum pools were subsequently passed over SU columns, with washing, elution, and concentration of bound anti-SU antibodies as described above for anti-CD134 antibodies.

ELISA for Titration of Cat Antibodies to SU, CD134, and V3-Related Synthetic Peptides.

Immulon 2 plates were used throughout (Thermo). To coat, SU and CD134 were diluted in sodium carbonate buffer (0.1 M, pH 9.6) at 1 μg/mL (100 μL per well); the V3-related peptides (some of which were semisoluble) were coated at 2 μg per well in carbonate buffer plus 10% DMSO (Sigma, catalog no. D-2650). Sera were tested at 1:100 dilution in B3T as described in ref. 28.