Distinct Antiretroviral Mechanisms Elicited by a Viral Mutagen.

5-aza-cytidine (5-aza-C) has been shown to be a potent human immunodeficiency virus type 1 (HIV-1) mutagen that induces G-to-C hypermutagenesis by incorporation of the reduced form (i.e., 5-aza-dC, 5-aza-dCTP). Evidence to date suggests that this lethal mutagenesis is the primary antiretroviral mechanism for 5-aza-C. To investigate the breadth of application of 5-aza-C as an antiretroviral mutagen, we have conducted a comparative, parallel analysis of the antiviral mechanism of 5-aza-C between HIV-1 and gammaretroviruses - i.e., murine leukemia virus (MuLV) and feline leukemia virus (FeLV). Intriguingly, in contrast to the hallmark G-to-C hypermutagenesis observed with HIV-1, MuLV and FeLV did not reveal the presence of a significant increase in mutational burden, particularly that of G-to-C transversion mutations. The effect of 5-aza-dCTP on DNA synthesis revealed that while HIV-1 RT was not inhibited by 5-aza-dCTP even at 100 µM, 5-aza-dCTP was incorporated and significantly inhibited MuLV RT, generating pause sites and reducing the fully extended product. 5-aza-dCTP was found to be incorporated into DNA by MuLV RT or HIV-1 RT, but only acted as a non-obligate chain terminator for MuLV RT. This biochemical data provides an independent line of experimental evidence in support of the conclusion that HIV-1 and MuLV have distinct primary mechanisms of antiretroviral action with 5-aza-C. Taken together, our data provides striking evidence that an antiretroviral mutagen can have strong potency via distinct mechanisms of action among closely related viruses, unlinking antiviral activity from antiviral mechanism of action.

Follow-Up of Viral Parameters in FeLV- or FIV-Naturally Infected Cats Treated Orally with Low Doses of Human Interferon Alpha.

Specific treatments for the long-life infections by feline leukemia virus (FeLV) and feline immunodeficiency virus (FIV) are either toxic, expensive or not too effective. Interferon α (IFN-α) is an immunomodulatory molecule which has been shown in vitro to decrease the release of infective particles. The aim of this study was to follow the progress of the clinical score and viral parameters of FeLV- and FIV-naturally infected privately owned cats treated with recombinant human IFN-α (rHuIFN-α, Roferon-A). Twenty-seven FeLV-infected cats (FeLV+) and 31 FIV-infected cats (FIV+) were enrolled in the study. Owners were instructed to orally administer 1 mL/day of 60 IU rHuIFN-α/mL in alternating weeks for four months. Blood samples were taken at the beginning of the study (M0), mid-treatment (M2), end of treatment (M4), and 6-10 months later (M10). Clinical status at these time points improved notably with rHuIFN-α treatment, regardless of the initial severity of the disease, an effect which lasted throughout the study in most animals (15 of the 16 FeLV+ symptomatic cats; 20 of the 22 FIV+ symptomatic cats) improved markedly their clinical situation. In FeLV+ cats plasma antigenemia (p27CA), reverse transcriptase (RT) activity, and proviral load decreased at M2 and M4 but increased again at M10 ("rebound effect"). The level of antigenemia or RT activity was below the detection limits in FIV+ cats, and the effect on proviral load was less marked than in FeLV+ cats. Taken together, these results indicate that rHuIFN-α is a good candidate for treating FeLV+ cats, but the "rebound effect" seen when treatment was discontinued suggests that additional studies should be conducted to clarify its effect on progression of the infection in cats.

Co-infection with feline retrovirus is related to changes in immunological parameters of cats with sporotrichosis.

Feline sporotrichosis due to Sporothrix brasiliensis is frequently severe and often correlated to zoonotic transmission. Feline Immunodeficiency Virus (FIV) and Feline Leukemia Virus (FeLV) cause immunodeficiency in cats; no association has been identified with critical cases of sporotrichosis. Moreover, the cytokine profile in Sporothrix-infected cats and a potential impact of retrovirus co-infections on their immunity is unknown. This study assessed immunological parameters in cats with sporotrichosis with and without FIV or FeLV co-infection. FeLV infection was detected by antigen ELISA and by provirus PCR. FIV infection was investigated through ELISA and Western blot. Cytokine transcription (IFN-γ, IL-4, IL-5, IL-6, IL-10, IL-12, TNF-α) was quantified using RT-qPCR and lymphocyte subpopulations (CD4, CD8, CD5 and CD21) were assessed by flow cytometry. Thirty cats with sporotrichosis were recruited to the study, including three FIV-positive and five FeLV-positive (progressive infection) cats. One cat with regressive FeLV infection was excluded from statistics. In comparison to retrovirus-negative cats, FIV-positive cats and FeLV-positive cats had higher IL-10 levels, FeLV-positive cats had lower IL-4 levels and FIV-positive cats had lower IL-12 levels and a lower CD4+/CD8+ ratio. Remarkably, all cats with poor general condition were FeLV (progressive infection) or FIV-positive, but the retrovirus status was not associated with the sporotrichosis treatment length or outcome. The immunological changes and the more severe clinical presentation observed in cats with retrovirus co-infections encourage future prospective studies that address the impact of these changes on prognostic determinants of feline sporotrichosis and the development of new therapy strategies that control disease spread.

Relevance of feline interferon omega for clinical improvement and reduction of concurrent viral excretion in retrovirus infected cats from a rescue shelter.

Feline Immnunodeficiency (FIV) and Feline Leukemia (FeLV) viruses are common infectious agents in stray cats and shelter environments. Recombinant feline interferon-ω (rFeIFNω) has shown an antiviral action not only against FIV and FeLV but also against herpesvirus (FHV-1) and calicivirus (FCV). Sixteen naturally infected FIV/FeLV cats were followed during rFeIFNω therapy in order to monitor clinical signs and to correlate with excretion of concomitant viruses (FCV, FHV-1, feline coronavirus (FCoV) and parvovirus (FPV)). Cats were submitted to clinical evaluations and concomitant virus excretion assessement. Comparing D0-D65, 10/16 cats improved clinical scores. Of the 10 cats positive for FHV-1 on D0, 4 were negative and 6 reduced viral loads. Of the 11 FCoV positive cats, 9 reduced viral loads. The 13 FCV positive cats and the FPV positive cat were negative on D65. In conclusion, rFeIFNω improves clinical signs and reduces concurrent viral excretion in naturally infected retroviral cats.

Discovery of drugs that possess activity against feline leukemia virus.

Feline leukemia virus (FeLV) is a gammaretrovirus that is a significant cause of neoplastic-related disorders affecting cats worldwide. Treatment options for FeLV are limited, associated with serious side effects, and can be cost-prohibitive. The development of drugs used to treat a related retrovirus, human immunodeficiency virus type 1 (HIV-1), has been rapid, leading to the approval of five drug classes. Although structural differences affect the susceptibility of gammaretroviruses to anti-HIV drugs, the similarities in mechanism of replication suggest that some anti-HIV-1 drugs may also inhibit FeLV. This study demonstrates the anti-FeLV activity of four drugs approved by the US FDA (Food and Drug Administration) at non-toxic concentrations. Of these, tenofovir and raltegravir are anti-HIV-1 drugs, while decitabine and gemcitabine are approved to treat myelodysplastic syndromes and pancreatic cancer, respectively, but also have anti-HIV-1 activity in cell culture. Our results indicate that these drugs may be useful for FeLV treatment and should be investigated for mechanism of action and suitability for veterinary use.

In vitro inhibition of feline leukaemia virus infection by synthetic peptides derived from the transmembrane domain.

BACKGROUND: The feline leukaemia virus (FeLV) is a gammaretrovirus commonly affecting cats. Infection with this virus often leads to fatal outcomes and, so far, no cure is available for this disease. Synthetic peptides with structures mimicking the transmembrane protein of the viral surface proteins hold the potential to effectively interfere with viral entry by hampering the fusion of viral and host cell membranes and constitute a novel approach for the treatment of infections with retroviruses. We identified and synthetically produced 11 FeLV peptides and evaluated their potential to block FeLV infection in vitro. METHODS: Cell cultures were exposed to FeLV subgroup A prior to the addition of the peptides. The inhibitory effect of the peptides was assessed by measuring FeLV gag protein in the supernatant of peptide versus mock-treated cell cultures using an ELISA. RESULTS: A peptide (EPK364) of 37 amino acids in length, with sequence homology to the HIV fusion inhibitor T-20, significantly suppressed viral replication by 88%, whereas no effects were found for shorter peptides. Two structurally modified variants of EPK364 also inhibited viral replication by up to 58% (EPK397) and 27% (EPK398). CONCLUSIONS: Our data support the identification of synthetic FeLV peptides that have the potential for a curative short-term therapy of viraemic cats. In addition, these peptides might become an important tool in xenotransplantation, where endogenous gammaretroviruses of the donor species might be able to infect the host.

Use of recombinant interferon omega in feline retrovirosis: from theory to practice.

Type-I interferons (IFNs) are cytokines that have non-specific antiviral activity, participating mostly in innate defense mechanisms. Their administration has been proposed to treat several viral and immunomediated diseases as an immunomodulatory therapy. Due to its availability, recombinant human interferon-alpha (rHuIFN-α) has been studied in relation to feline retrovirosis, both in vitro and in vivo. However, IFNs are species-specific and antibodies have been shown to develop in response to the high rHuIFN-α doses necessary for an effective therapy. A recombinant feline IFN has been developed, which has been characterized as interferon-omega (rFeIFN-ω), designed to overcome these problems. Nonetheless, very few studies have been undertaken to evaluate its efficacy in cats naturally infected with FIV or FeLV. In an initial study, we here demonstrated that rFeIFN-ω can dramatically improve the clinical condition of infected cats, and induce improvement of hematologic parameters. Minor changes or no change was observed for hypergammaglobulinemia, CD4/CD8 ratio, proviral load, viremia and RT activity, suggesting that the overall effect of IFN was on innate immunity. More studies are needed in order to better understand its in vivo mechanisms.

Inhibition of Feline leukemia virus replication by the integrase inhibitor Raltegravir.

The oncogenic gammaretrovirus Feline leukemia virus (FeLV) has been the leading cause of death among domestic cats until the introduction of efficient diagnostics and vaccines in the late 1980s. So far, no efficient treatment for viremic animals is available. Hence, use of the FeLV model to evaluate antiretroviral therapies applied to HIV is a timely task. The efficacy of the integrase inhibitor Raltegravir, which is widely used for the treatment of HIV in humans, has been assessed in vitro for the FeLV-A/Glasgow-1 strain. EC(50) values for FeLV-A inhibition in feline cell lines are in the range of that observed for HIV and xenotropic murine leukemia virus-related gammaretrovirus. Therefore, Raltegravir may be a potential therapeutical agent for felids with progressive FeLV infection.

Single-dose pharmacokinetics and genotoxicity of metronidazole in cats.

Single-dose pharmacokinetics and genotoxicity of metronidazole in cats were evaluated. Cats received either 5mg/kg metronidazole intravenously, or 20mg/kg metronidazole benzoate (12.4mg/kg metronidazole base) orally in a single dose. Serial plasma samples were collected and assayed for metronidazole using high pressure liquid chromatography (HPLC). Genotoxicity was assessed in vitro in feline peripheral blood mononuclear cells (PBMC) and a feline T-cell lymphoma line incubated with metronidazole, and in vivo in PBMC collected before, during and 7 days after oral metronidazole, by use of the COMET assay. Systemic absorption of metronidazole was variable (mean=65+/-28%) with a peak of 8.84+/-5.4microg/ml at 3.6+/-2.9h. The terminal half-life was 5.34h from the intravenous dose and 5.16h from the oral dose. Systemic clearance was low (mean=91.57ml/h/kg [1.53ml/kg/min]), and the apparent volume of distribution (steady state) was 0.650+/-0.254l/kg. Genotoxicity was detected at all concentrations of metronidazole in feline PBMC and the T-cell lymphoma line in vitro. Genotoxicity was also observed in PBMC collected from cats after 7 days of oral metronidazole but resolved within 6 days of discontinuing metronidazole.

Effect of type I interferons on the expression of feline leukaemia virus.

The efficacy of interferons (IFNs), used empirically to treat retrovirus-infected cats has been shown in vivo, but the direct effect on infected cells is largely unknown. Ten-fold serial dilutions of three recombinant IFNs available for therapy, human IFNalpha(2a), IFNalpha(A/D) and feline IFNomega were added to the chronically FeLV-infected cell line FL74. IFNs did not apparently affect viral protein expression. However, reverse transcriptase activity (RT), directly proportional to the amount of infectious free virions, decreased with increasing concentrations of IFN and longer treatment times. The induction of apoptosis by IFN was suspected. Results of its evaluation by annexin V-Fluos staining showed that IFNs decreased the viability of treated FeLV-infected cells, and increased apoptosis, but not of non-infected cells. According to the IC(50), rHuIFNalpha(A/D) appeared to be the most effective IFN in inhibiting RT.

Effect of 17beta-estradiol and progesterone on the expression of FeLV in chronically infected cells.

In a previous study, it was found that even though more male cats were infected by feline leukaemia virus (FeLV), females seemed to progress easier to overt disease. To study the effect of female hormones, 17beta-estradiol and progesterone were added in different concentrations (10(-3) M to 10(-12) M) to a culture of persistently FeLV-infected cells. The effect of both hormones was very similar. After 24 h the cell viability was very low at 10(-3) M and 10(-4) M but similar to controls at the remaining concentrations. Liberation of viral particles was estimated by the reverse transcriptase activity (RT), which was the lowest also at 10(-3) M and 10(-4) M. However, low viability could not account for this low RT, as when cells were lysed with lysis buffer RT was high. Thus, cells were dying without freeing viral particles, suggestive of apoptosis. This possibility was confirmed by staining hormone-treated cells with annexin V and propidium iodide. The FeLV antigen p27 measured in the cultures had a maximum at 10(-3) M and 10(-4) M, higher than controls and lysed cells, so the presence of p27 in the supernatant was not only due to cell lysis but a consequence of hormone effect. In conclusion, 17beta-estradiol and progesterone induce death of FeLV-infected cells at high concentrations, probably through a process of apoptosis, which might limit the spread of the infection, as infective viral particles would be hampered from budding.

Antiretroviral efficacy of a 98% solution of glycerol or ethylene oxide for inactivation of feline leukemia virus in bone.

OBJECTIVE: To determine whether infectious retrovirus was inactivated in bones from FeLV-infected cats after ethylene oxide (ETO) sterilization or preservation in a 98% solution of glycerol in an in vitro cell culture system. SAMPLE POPULATION: Metatarsal bones obtained from 5 FeLV-infected cats and cultured with feline fibroblast cells. PROCEDURE: Metatarsal bones were treated with 100% ETO, a 98% solution of glycerol, or left untreated. Twenty-five flasks of feline fibroblast cells were assigned to 5 groups: negative control, positive control, ETO-treated bone, glycerol-treated bone, and untreated bone with 5 replicates/group for 4 passages. Media and cell samples were harvested from every flask at each passage to measure FeLV p27 antigen and the number of copies of provirus per 100 ng of DNA, respectively. RESULTS: All negative control and ETO-treated group replicates were negative for FeLV p27 antigen and provirus throughout the study. All positive control group replicates were positive for FeLV p27 antigen and provirus at passages 1 to 4. Untreated bone group replicates were positive for FeLV p27 antigen at passages 3 and 4 and provirus beginning at passage 2. Glycerol-treated group replicates had delayed cell replication and were negative for FeLV p27 antigen and provirus at passages 1 to 4 and 2 to 4, respectively. CONCLUSIONS AND CLINICAL RELEVANCE: Ethylene oxide sterilization of bone from FeLV-infected cats appeared to abrogate transmission of infectious retrovirus and effectively sterilized bone allografts. Impact for Human Medicine-Additional studies to confirm effectiveness of ETO treatment of allograft tissues for prevention of pathogen transmission via transplantation are warranted.

A biomimetic approach to the synthesis of an antiviral marine steroidal orthoester.

Orthoesterol B, a marine natural product exhibiting antiviral activities, contains a [3.2.1]-bicyclic orthobutyrate bridging the steroid side chain and ring D. A biosynthetic reaction was proposed by which rearrangement of an epoxy ester results in the formation of the orthoester moiety. Steroidal model compounds incorporating 16-butyroxy and 20,22-epoxy groups were synthesized from tigogenin and were shown to rearrange to orthoesters under mild acidic catalysis.

Effects of a 98% solution of glycerol or sterilization with ethylene oxide on FeLV in bone allografts and effects on bone incorporation of allografts in cats.

OBJECTIVES: To compare virucidal effects and bone incorporation properties of cortical bone allografts transplanted into specific-pathogen-free (SPF) cats. Allografts consisted of untreated bone from a SPF cat (negative-control group) and bone from 5 FeLV-infected cats that was subjected to sterilization with ethylene oxide (ETO), preservation with glycerol, or no treatment (positive-control group). SAMPLE POPULATION: Bones from the aforementioned groups and twenty 8-week-old SPF cats (5 cats/group) implanted with an allograft from 1 of the aforementioned groups. PROCEDURE: After implantation, blood samples were collected weekly to monitor FeLV p27 antigen and antibody titers. Quantification of FeLV provirus was performed on blood samples at weeks 0, 4, and 8 and donor bone samples at time of implantation. Cats were euthanatized 8 weeks after transplantation, and graft sites were evaluated. RESULTS: All results for negative-control cats were negative. All ETO group cats had negative results for antigen and provirus in blood, whereas 1 cat had a low antibody titer. Although 3 ETO-treated allografts were positive for provirus, the DNA appeared denatured. One cat in the glycerol group had positive results for all tests in blood samples. All glycerol-preserved allografts were positive when tested for provirus. All results for positive-control group cats were positive. Differences in incorporation of bone grafts were not observed. CONCLUSIONS AND CLINICAL RELEVANCE: Glycerol preservation of FeLV-infected bone allografts did not eliminate transmission of retrovirus to recipients. In contrast, ETO sterilization appeared to denature DNA and prevent infection. Treatments did not affect incorporation of bone grafts in young cats.

Bone marrow necrosis in a cat infected with feline leukemia virus.

A one-year old castrated male cat was admitted to the hospital with vomiting and diarrhea. Laboratory examination revealed pancytopenia and positive for FeLV antigen. A bone marrow examination indicated necrosis of the nucleated cells. Based on these findings, the cat was diagnosed as bone marrow necrosis. Pancytopenia was effectively treated with corticosteroids. Re-examination of the bone marrow confirmed a recovery of normal hematopoietic cells with a infiltration of many macrophages. It is strongly suspected that the bone marrow necrosis in this case could be associated with a bone marrow suppression due to FeLV infection.

The role of interferon-gamma in antiretroviral activity of methionine enkephalin and AZT in a murine cell culture.

The ability of spleen cells treated with methionine enkephalin (Met-ENK) in the presence of 3'-azido-3'-deoxythymidine (AZT) to produce cytokines and inhibit Friend leukemia virus (FLV) replication in Mus dunni cell cultures was investigated. In the presence of murine spleen cells, combination treatments using AZT plus Met-ENK or concanavalin A reduced FLV replication by 63% and 84%, respectively, as compared with 47% for AZT alone. When interleukin (IL)-2, IL-4 and interferon (IFN gamma) levels were measured in FLV-infected cell cultures, both AZT and Met-ENK treatments induced a higher production of IFN gamma and a slight increase in IL-2 and IL-4, as compared with either treatment alone. Subsequent treatment of FLV-infected cells with concanavalin A-stimulated cell supernatants, containing approximately 10 U/ml each of IFN gamma and IL-2, resulted in inhibition of viral replication. Thus, in the absence of spleen cells, IFN gamma was added to cell cultures to determine whether this cytokine contributed to combination antiviral effects. Results show that addition of IFN gamma alone results in a slight suppression of FLV expression, whereas treatment with both AZT and IFN gamma inhibits FLV replication significantly. Subsequently, addition of anti-IFN gamma antibody to cell cultures treated with Met-ENK blocked antiviral effects due to this neuropeptide. Thus anti-FLV effects of spleen cells treated with Met-ENK in combination with AZT are mediated to a large degree by IFN gamma.

Therapeutic effects of diethylcarbamazine and 3'-azido-3'-deoxythymidine on feline leukemia virus lymphoma formation.

Twenty-four specific pathogen-free kittens were infected with the Rickard strain of feline leukemia virus (FeLVR). The kittens were divided into four equal groups and were orally administered either a high dose of diethylcarbamazine (DECH, 12 mg kg-1), a low dose of diethylcarbamazine (DECL, 3 mg kg-1), 3'-azido-3'-deoxythymidine (AZT, 15 mg kg-1, b.i.d.), or a placebo (250 mg granular dextrose) daily for 10 weeks. Blood was collected at 2-week intervals for complete blood counts (CBC) and flow cytometric analysis (FACS) of peripheral blood lymphocytes (PBL). Plasma was assayed for antibodies to FeLV gp70 and for FeLV p27 antigen using ELISA assays. For FACS analysis, lymphocytes were incubated with monoclonal antibodies to feline Pan T, CD8+, CD4+, and B cell (Anti-Ig) antigens. In the placebo treated cats, FeLVR infection caused an early (2 weeks p.i.) and persistent decrease in leukocyte numbers attributable primarily to a decrease in neutrophil numbers and a secondary lesser decrease in B and CD4+ lymphocyte numbers. The DEC-treated groups showed a delayed but similar leukopenia by 4 weeks p.i. The lymphopenia in the DEC groups (primarily B cells and CD4+ cells) was reversed by 10 weeks p.i., but the neutropenia persisted. AZT treatment inhibited FeLVR-induced lymphopenia but did not prevent a reduction in neutrophil numbers. A marked p27 antigenemia that peaked at 4 weeks p.i. was noted in the placebo treated cats and in most cats (11/12) treated with either dose of DEC. However, AZT significantly inhibited the p27 antigenemia and all cats were negative for p27 antigen between 6 and 10 weeks of treatment. In general, placebo treated cats as well as DECH and DECL cats had low levels of antibody to gp70 throughout the study, suggesting FeLVR-induced immunosuppression. In contrast, significantly higher titers of anti-gp70 antibodies were seen in AZT-treated cats at 6 weeks p.i., and were maintained throughout treatment. Eighteen month survival rates provide efficacy data for AZT as well as both DEC treatment groups. While all placebo treated cats were euthanized by 52 weeks p.i. due to FeLV associated lymphomas with a mean survival time of 35.5 weeks p.i., median survival time of the AZT treated group was > or = 102 weeks p.i., while that of the DECH and DECL groups was 69.7 and 72 weeks p.i., respectively. Thus, DEC as well as AZT therapy delays the development of lymphomas associated with FeLV infection and significantly improves survival.

Differential antiviral activities and intracellular metabolism of 3'-azido-3'-deoxythymidine and 2',3'-dideoxyinosine in human cells.

Dideoxynucleosides such as 3'-azido-3'-deoxythymidine (AZT) and 2',3'-dideoxyinosine (ddI) can effectively inhibit the replication of human immunodeficiency virus (HIV) in T lymphoid cells. There is evidence that HIV can infect and replicate in other cells including monocytoid cells and macrophages. The present study compared the antiretroviral activities of ddI and AZT in three lineages of human cells, i.e., MOLT4 (T lymphocytoid, CD4+), U937 (monocytoid, CD4+), and HT1080 (fibroblastoid, CD4-) cells. Feline leukemia virus, a retrovirus that causes immunodeficiency in cats, was used to infect the cells. The drug concentrations needed to reduce the viral p27 antigen titers in cell lysates by 50% (IC50s) were determined. The data show that AZT and ddI inhibited viral replication in all three cell lines. The IC50s of AZT were 0.02, 1.75, and 2.31 microM in MOLT4, HT1080, and U937 cells, respectively. For ddI, the IC50s were 4.31, 9.52, and 43.5 microM, respectively. These data indicate differential antiviral activities of ddI and AZT in the different cells with the following rank order of drug sensitivity: MOLT4 > HT1080 > U937. A study of the intracellular metabolism of [3H]AZT and [3H]ddI shows that the antiretroviral activities of AZT and ddI in the three cell lines correlated with the levels of their intracellular triphosphate metabolites.

Macrophage tropism of feline leukemia virus (FeLV) of subgroup-C and increased production of tumor necrosis factor-alpha by FeLV-infected macrophages.

Erythroid aplasia is induced in cats by feline leukemia virus (FeLV) of subgroup C but not by FeLV of subgroup A. In an investigation of the role of macrophages in FeLV-C-induced diseases, the concentrations of FeLV and tumor necrosis factor-alpha (TNF-alpha) were compared between feline peritoneal macrophages incubated with FeLV of subgroup A or C. FeLV of both subgroups infected macrophages, but expression of FeLV-C was 21-fold higher than FeLV-A in peritoneal macrophages (P = .004). The supernatants of FeLV-C-inoculated macrophage cultures contained significantly higher levels of TNF-alpha (70 +/- 14 U/mL) at 72 hours postincubation compared with FeLV-A-inoculated (38 +/- 8 U/mL) and uninoculated (31 +/- 8 U/mL) cultures. Moreover, a positive correlation was shown between cell-associated FeLV surface glycoprotein gp70 and TNF-alpha expression in FeLV-C-infected macrophages by immunofluorescence (r = .6; P = .001), measured with a computer-assisted, laser-based digital imaging system. The addition of TNF-alpha to a uniform population of FeLV-infected cells (feline embryonic fibroblasts) caused an enhancement of viral expression (P < .05). These results indicate that FeLV-C has tropism for macrophages, FeLV expression is positively correlated with TNF-alpha expression in macrophages, and TNF-alpha enhances FeLV replication in fibroblasts. We suggest that FeLV-C infection of macrophages and secretion of TNF-alpha may be important in hematopoietic suppression in FeLV-C-infected cats.