Interferon-gamma: A potent antiviral a g e n t targeting macrophages infected with LP-BM5 murine leukemia virus, the causative a g e n t of 'AIDS' in mice

Interferon-gamma: A potent antiviral agent targeting macrophages infected with LP-BM5 murine leukemia virus, the causative agent of 'AIDS' in mice. Can J Infect Dis 1992:3(Suppl B):ll5B-122B. Cells of the monocyte/ macrophage lineage (MM cells) are known to be infected by retroviruses. including the human immunodeficiency virus (HIV). without cytopathic changes and may serve as a persistent reservoir for the virus during the development of immunodeficiency disease. LP-BM5 murine leukemia virus (MuLV) infection ofC57BL/ 6 mice and cell lines has been used to optimize therapy directed against macrophages. Findings in this murine system may be applicable to HN infection in humans. The effect of recombinant murine interferon-gamma OFN-y) and 3' -azido-2',3' -dideoxythym.idine (AZlj as single agents or in combination was investigated in both LP-BM5 MuLV de novo infection and chronic infection of macrophages. Results indicate that the therapeutic effects of these single agents were dose-dependent and both agents were similarly effective in reducing the production of infectious virus determined by XC-plaque assay and by measurements of reverse transcriptase activity in culture super natants; and AZf and IFN-y reduced the production of virus proteins, quantified by laser densitometry of fluorographs from immunoprecipitated viral proteins using virus-specific antiserum. A combination oflFN-y and that AZf showed greater antiviral activity in both LP-BM5 MuLV de novo and chronic infection of macrophages than either agent alone, suggesting that IFN-y and AZf represent a potent combination of antiviral agents targeting macrophages. Further, since a lower concentration of each agent was required for efficacy in combination therapy. toxicity associated with single agent therapy may be avoided.


I N THE PAST F'EW YEARS. EVlDENCE HAS ACCUMULATED THAT
cells of the monocyte/macrophage lineage (MM cells) function as a persistent reservoir for retrovinlses and may be involved in the progression of retrovinls-induced acquired immune deficiency syndromes (AJDS) in humans and mice (MAIDS) by controlling production and dissemination of the virus ( l -4). Extensive research has identified MM cells as central modulators of the immune system due to their unique repertoire of functions as: antigen-presenting cells; producers of cytokines and other factors (interleukin-l [IL-l]. lL-6 , tumour necrosis factor-alpha, interferon-alpha [IFN-a]. granulocyte colony stimulating factor [G-CSF]. macrophage colony stimulating factor, prostaglandins, etc) ; phagocytes; and tumoricidal, cytotoxic cells (5)(6)(7)(8)(9)(10) .
The idea that alterations of MM cell functions occur after retrovirus infection and account in part for immune system suppression is controversial (ll -16) and needs further investigation. However, development of a therapy directed against retrovirus infection of macrophages may limit dissemination and production of the retrovirus and control the progression of the disease (ie, AJDS and MAIDS). The authors have focused their efforts on the development of a treatment regimen against retrovirus infection of macrophages by using LP-BM5 murine leukemia viruses (MuLVs) in cultured murine macrophages. LP-BM5 MuLV is a mixture of replication-competent ecotropic and mink cell focus-inducing (MCF) MuLVs, and replication-deficient BM5 MuLV (which causes a lymphoproliferative/immunosuppressive syndrome in mice [MAIDS! -similar to AJDS in humans) (17)(18)(19).
Effects of recombinant murine IFN-y and 3'-azido-2',3'-dideoxythymidine (AZT) as single agents or in combination were tested in both de novo and chronic infections of macrophages with LP-BM5 MuLV. Single agents such as AZT and other dideoxynucleosides (dideoxycytosine [DDC]. -adenosine [DDA) and -inosine [DDI)) inhibit human immunodeficiency virus (HIV) replication in lymphocytes (20,21) to a greater extent than in MM cells (22); this has been correlated to a diminished activity of dideoxynucleoside kinases in human MM cells which phosphorylate AZT and other dideoxynucleosides into active metabolites (22 ,23) . Doses of these agents which would partially overcome ll6B the block of phosphorylation cause severe hematopoietic toxicity (AZT). neurotoxicity (DDC) and nephrotoxicity (DDA) (24)(25)(26)(27). IFN -y has been recognized as a potent inducer of a variety of macrophage functions leading to an 'antiviral state of the cell' (28)(29)(30) , and it has been proposed that the decrease of lFN-y-producing T cells in AJDS leads to an impaired activation of macrophages (31). Therefore the authors initiated investigations using lFN-y as a therapy against retrovinls infection of macrophages (32,33).

Cells and viruses:
The C-III macrophage cell line u sed in this study was originally isolated from the adherent population of cells derived from bone marrow of a C57BL/6 mouse. The C-Ill macrophages were cloned by a series of limiting dilutions in the presence of recombinant human macrophage CSF (Genzyme Corp, Massachusetts). More than 80% of the cells stained positive for nonspecific esterase by histochemistry and more than 90% expressed MAC-l antigen on the cell surface as determined by fluorescence activated cell sorter (FACS) analysis. Electron microscopy of C-TIJ cells is consistent with typical morphology of macrophages. Uninfected and LP-BM5 MuLV-infected macrophages were cultured in Dulbecco's modified eagle medium (DMEM) (Gibco Lab Inc, New York) . supplemented wilh 5% fetal calf serum (FCS) (Hyclone Lab Inc, Utah) and antibiotics (5000 U/mL penicillin, 5 11g/mL streptomycin) at 37°C and 5% carbon dioxide in air atmosphere.
The mixture of LP-BM5 MuLV containing replication-competent ecotropic MuLV, MCF-MULV and replication-deficient BM5 MuLV was propagated in SCl cells, a fibroblastoid cell line susceptible to LP-BM5 MuLV infection. LP-BM5 MuLV containing culture supernatant from infected SC-1 cells was used to infect macrophage C-III cells. LP-BM5 MuLV-infected macrophage C-IIl cells produce infectious virus particles as determined by XC-plaque assay, measurements of lhe activity of reverse transcriptase (RT) in supernatants of infected macrophages and electron microscopy. In addition, both LP-BM5 MuLV-infected macrophage C-lll cells and culture supernatant caused MAIDS in 100% of intraperitoneally inoculated mice.  . . -...

) Toxicity and antiviraL activity of 3'-azido-Z.3'-dideoxythymidine (AZr) and interferon-gamma (IFN-"() in LP-BM5 murine leukemia virus (MuLV) de novo infected and chronically infected cultured murine macrophages. Toxicity was determined by FHJ-thymidine incorporation and antiviral activity (elficacy) by reverse transcriptase assay for AZr {A) and IFN-"( {B). Each symbol represents mean± SD of duplicates from three separate determinations. Curves were obtained by jitLing the data to the Hill equation using Adapt software (37)
XC-plaque assay: Detection of ecotropic MuLV was essentially performed according to the XC-plaque technique described by Rowe and colleagues (34) -carried out with 10 4 cells per well in six-well plates. Syncytia formation was quantified by phase contrast microscopy and using a video camera supported colony counter system (Artek counter 982, Dynatech Lab, Virginia) . Bolli methods gave similar counts of syncytia. Reverse transcriptase assay and ~H)-thymidine incorporation: The activity of Rf in supematants of LP-BM5 MuLV de novo and chronically infected macrophages was determined via a microtitre assay, modified from the procedures described by Gregersen et al (35) and Somogyi and co-workers (36 ...

118B
tively weighted regression analysis using Adapt software (37). [ 14 C]leucine incorporation into macrophages: Macrophages were seeded out in microtitre plates, incubated with DMEM plus 5% (volume/volume) FCS containing 0, 1, 10, 50, 100, 500, 1000, 5000 or 10,000 U/mL IFN-y. Cells were allowed to grow for two days and the culture medium, containing the different concentrations of IFN-y, was changed every day. On the third day, cells were washed twice with leucine-and antibiotic-free modified DMEM and incubated for 4 h with leucine-and antibiotic-free, modified DMEM containing 5 )..l.Ci/mL [ 14 C]leucine (DuPont) and the respective concentration of lFN-y . Cells were lysed with distilled water and proteins were harvested on glass-fibre filters using a PHD cell harvester (Cambridge Technology Inc). Protein-incorporated [ 14 C]leucine was quantitated by liquid scintillation counting. Metabolic labelling of macrophages: De novo and LP-BM5 MuLV chronically infected macrophages in sixwell plates were grown for three days with or without the addition of IFN, AZf or a combination. The subconfluent monolayers of LP-BM5 MuLV de novo and chronically infected macrophages were washed twice with methionine-free modified DMEM and then incubated for 12 h with methionine-free modified DMEM containing 50 )..l.Ci/mL [ 35 S]methionine (DuPont). 5% FCS and the same addition of IFN-y and AZf as before. The supematant was removed and cell monolayers were washed twice with HEPES buffered saline (0.9% sodium chloride, 50 mM N-2 -hydroxyethylpiperazine-N' -2ethane-sulphonic acid, pH 7.4) prior to the addition of 0.5 mL per well ice-cold extraction buffer A containing 5 mM Tris-hydrochloride, 1 mM EDTA, 0.4 M potassium chloride, 1o/o Triton X-100, 1 mM phenylmethylsulphonylfluoride (Boehringer Mannheim, Indiana) and 1 mM tosylphenylchloroketone (Boehringer Mannheim). Cell extracts were used for immunoprecipitation. Immunoprecipitation of viral proteins: Preparation of the cell extracts and immunoprecipitation with polyclonal, monospecific goat antisera to purified retrovirus protein p30gag was performed as described by Bilello et a! (38) . To compensate for differences in protein synthesis of the differently treated macrophages, each immunoprecipitation was done with approximately 2.5x10 7 dpm of cell extract. The volume of each sample was adjusted to 1 mL with extraction buffer B (extraction buffer A without potassium chloride). Cell extracts were precipitated with normal goat serum prior to immunoprecipitation with virus-specific goat antiserum, containing ap30 antibodies. Staphylococcus aureus immunoabsorbant (Life Tech Inc, Maryland) was used to collect antigen-antibody complexes by centrifugation. Pellets were washed twice with a solution containing 20 mMTris-hydrochloride (pH 7.4). 100 mM sodium chloride, 1 mM EDTA, and 0 .5% Nonidet P-40 (Sigma Chemical Co, Missouri) and once with the same solution plus USE 2.5 mM potassium hydrochloride. Samples were prepared for sodium dodecylsulphate (SDS) polyacrylamide (12.5%) slab gel electrophoresis (PAGE) and SDS-PAGE was performed according to the method ofLaemmli (39). Rainbow markers (Amersham Corp. Illinois) were used to estimate molecular weight of the proteins. Fluorography and quantification of labelled protein bands: SDS-polyacrylamide slab gels were fixed for 30 mins with a solution of propanol-2:acetic acid:water (25: 10:65) soaked in Amplify (Amersham Corp) for another 30 ruins and dried at 70°C under vacuum. The dried gels were exposed to Chronex 4 x-ray fllms at -70°C. The amount of viral proteins present in the bands of fluorographs was quantified via an integrating laser densitometer (Ultroscan XL; Pharmacia LKB Biotechnology, New Jersey) and GelScanXL 2 .1 software.

RESULTS
The antiviral activity of AZf and IFN-y as single agents and in combination was correlated to the reduction of RT activity in culture supernatants of LP-BM5 MuLV de novo infected and chronically infected macrophages. Both agents were similarly effective antivirals at concentrations where toxicity, measured as inhibition of [ 3 H)-thymidine incorporation, was low (Figure 1). IFN-y and AZf were more effective antivirals against the de novo infection of macrophages than in chronic infection. Treatment of cells with combinations of IFN-y (10 or 100 U/ mL) and various concentrations of AZf resulted in greater reduction of RT activity tha11 for either agent alone. In addition, no increased cytotoxicity was observed with combination regimens in the de novo infection of macrophages with LP-BM5 MuLV (Figure 2). suggesting that a combination of AZf and IFN-y is more effective than either agent alone. Statistical analysis of the single-agent treatment regimen was carried out by nonlinear reiteratively weighted regression analysis (37) and concentrations of EC5o and IC5o were calculated for IFN-y and AZf (Table 1).
Essentially the same results were obtained when the production of infectious virus was quantitated by XCplaque technique in the presence of single agents (AZf, IFN-y) or their combination ( Table 2).
Little is known about the mechanism by which IFN-y decreases the production of infectious virus in macrophages. The authors examined this mechanism by investigating the effect of IFN-y on production of viral core proteins in LP-BM5 MuLV de novo and chronically infected macrophages by immunoprecipitation of [ 35 S]methionine metabolically labelled virus proteins with virus-specific antiserum (Figure 3). Quantitation of the 30 kD virus-specific bands (p30gag) by laser densitometry indicated that: cell-derived virus protein p30gag was reduced by IFN-y in a concentration-dependent fashion (Table 3)    Values are mean contentrations at which IC50 and EC50 were estimated using Adapt software (37)    cultured cells presented above. However, the reduction in synthesis of viral proteins in LP-BM5 MuLV de novo and chronically infected macrophages by IFN-y was of lower magnitude than the decreases seen in RT activity or in XC-plaques. De novo protein synthesis measured as [ 14 C)leucine incorporation inC-III macrophages was slightly stimulated by IFN-y at concentrations below 100 U/mL ( Figure 4) and, thus, it is likely that IFN-y inhibits specifically the translation of virus proteins.

DISCUSSION
Biological response modifiers (BRMs) are effective in the treatment of cancer, immunocompromized individuals and patients with hematological disorders. Most BRMs act locally at very low concentrations and are selective for specific cells in a defined microenvironment. Therapeutic approaches using systemic application of BRMs often are followed by toxic side effects due to the overall activation of target cells. IFN-y has been recognized as the major product. from T lymphocytes and is known to induce a variety of processes in macrophages leading to functional activation of the cells (30)(31)(32). With regard to the effects of IFN-y on macrophages, the authors addressed the following questions: can macrophages be protected from retroviral infection? and does IFN-y have an effect on virus expression in chronically infected cells using the LP-BM5 MuLV model? Results indicate that IFN-y is more effective in reducing production of infectious virus from de novo infected macrophages than from chronically infected macrophages, an observation supported by the reduc-USE tion of Rr activity in s u pernatants ofLP-BM5 MuLV-infected macrophages, by decreased formation of syncytia in an XC-plaque assay and by a diminished expression of viral core protein p3~ag (and envelope protein gp70env. unpublished data) by IFN-y. Quantitative differences exist between the effects of IFN-y on the production of infectious virus and expression of virus protein. The authors speculate that IFN -y interferes with assembly and release of the virus, and leads to production of noninfectious virus particles in macrophages and 'trapping of virions' on the cell membrane, similar to using IFN-13 in MCF-infected mouse fibroblasts (38,40). Since there is a reduction of virus polyprot.ein precursors (Prl80gag-pol, Pr67gag), IFN-y alters translation of viral mRNA and may effect the level of transcription. The inhibitory effect of IFN-y on production of virus protein can not be explained by ACKNOWLEDGEMENTS: We thank Dr JA Bilello for kindly providing the virus specific antisera and for helpful discussions and Dr Janet W Hartley of the National Institute of Allergy and Infectious Diseases (NIAID) for the macrophage used to clone Clll. AZT was generously provided by DrS Nusinoff-Lehrmann of Burroughs Wellcome Co. Recombinant murine IFN-y was obtained from Genzyme Corp. Massachusetts, and goat antisera against viral protein p3ollag was kindly provided by Dr J Bilello, Department of Microbiology and Immunology. University of Maryland, Baltimore, Maryland. This work was supported by NIAID Contract #NOIA72666. macrophages at concentrations under 100 U/mL and was only moderately inhibited at higher concentrations of IFN-y.
In conclusion, the present results indicate that IFN-y is a potent BRM with antiviral activity against retrovirus infection of murine macrophages. These findings are in agreement with the recently reported protective effect of IFN-y against HIV infection of macrophages (33). The enhanced inhibitory effect ofiFN-y and AZT on both LP-BM5 MuLV de novo and chronic infection of macrophages suggests that this combination warrants further testing in vitro and in vivo using combinations of IFN-y and AZT with concentrations below the EC5o of each agent to elucidate further the interaction between these two antiviral agents.