TEXT

Top Abstract Text References

Foamy viruses (FVs) are members of the Spumavirus genus of the family Retroviridae and are widely distributed among mammals, e.g., monkeys, cows, and cats (10, 25). Although humans are not natural hosts, accidental infection of animal caretakers with various FV strains has been reported elsewhere (9, 21). Ex vivo propagation of FV isolates is accompanied by a characteristic cytopathic effect (CPE). Due to the formation of multinucleated giant cells with cytoplasmic vacuoles, the infected cell culture has a peculiar foamy appearance (10). In contrast to this dramatic cytopathicity, no clinical symptoms or histological alterations are observed within infected hosts, although FV infection persists and proviral DNA is readily detectable in all organs (6a, 10). This suggests that FV replication is strictly controlled in vivo presumably by immunological mechanisms.

The immune response during FV infections is not well characterized. Titers of neutralizing antibodies are low (15) and thus might not play a dominant role in restricting FV replication. Extensive studies of other persistent virus infections like those with the human immunodeficiency virus and hepatitis B virus (HBV) have demonstrated that noncytolytic antiviral responses mediated by cellular soluble factors could play a crucial role in suppressing virus propagation (8, 11). Those factors are mainly produced by activated lymphocytes. We therefore have studied the impact of supernatants from activated human peripheral blood lymphocytes (PBL) on FV infection of various cell lines in vitro. Here it is shown that various stimuli can activate PBL to produce a very potent soluble FV-inhibitory activity. It acts in a species-specific manner and can impede FV isolation from activated PBL when autologous coculture systems are used. Most of the inhibitory activity can be ascribed to gamma interferon (IFN-), because it can be almost completely suppressed by a specific monoclonal antibody. Thus, it appears that IFN- could play an important role in controlling FV replication to nonpathogenic levels in vivo.

PBL from normal blood donors were prepared by Ficoll Paque 400 (Pharmacia, Freiburg, Germany) gradient centrifugation of buffy coats followed by adherence to plastic for 2 h at 37°C. The cells were cultivated in the presence of various stimuli in RPMI 1640 medium containing 10% fetal calf serum and antibiotics. Stimulation was for 20 h with 10 µg of phytohemagglutinin (PHA; Difco Laboratories, Detroit, Mich.) per ml, 5 µg of concanavalin A (ConA; ICN Biomedicals GmbH, Eschwege, Germany) per ml, 10⁸ M phorbol 12-myristate 13-acetate (PMA; Sigma, Deisenhofen, Germany), or 200 ng of anti-CD3 monoclonal antibody (Dako GmbH, Hamburg, Germany) per ml. Cell-free supernatants were immediately used or aliquoted and kept at 80°C until use.

The supernatant of activated human PBL is able to inhibit FV-induced CPE in cells of human origin but not in mouse, monkey, or hamster cells. Different cell types including Alpha-1 human diploid fibroblasts (15), NIH 3T3 murine fibroblasts, and BHK-21 baby hamster kidney cells were infected with simian foamy virus strain Ka (22) at a multiplicity of infection of 0.1. Three hours postinfection (p.i.), the cells were washed and kept in culture with either fresh medium or the conditioned medium of activated PBL. After 4 days, infected cells were fixed with cold methanol and analyzed for the presence of FV-specific nuclear antigens by indirect immunofluorescence (IF) with simian foamy virus-positive monkey sera and fluorescein-conjugated rabbit anti-human immunoglobulins (Dako GmbH) as previously described (15, 21). As shown in Fig. 1, the supernatant of PHA-activated PBL was able to completely inhibit FV infection of Alpha-1 human fibroblasts but showed no anti-FV effect on cells of nonhuman origin. In addition, due to FV-induced CPE, all cell monolayers except for the Alpha-1 fibroblasts treated with the supernatant detached within 14 days p.i. from the plastic surface of the culture flasks (data not shown). Likewise, when the supernatant was added 3 h p.i. and washed off 24 h later, the drastic inhibitory activity was observed at day 14. Thus, the soluble suppressive activity was species specific and long lasting.

View larger version (104K): [in this window] [in a new window]   FIG. 1.   The supernatant from activated human PBL inhibits the CPE of FV on human fibroblasts but not on cells of other species. Shown are results of IF of FV-infected Alpha-1 human fibroblasts, NIH 3T3 mouse fibroblasts, and BHK-21 baby hamster kidney cells in the presence (+) or in the absence () of PBL-PHA supernatant.

The activation of PBL by other mitogens or anti-CD3 antibodies also led to the production of the soluble suppressive activity, inhibiting the FV-induced CPE on Alpha-1 fibroblasts by more than 90% (Fig. 2A). The percentage of inhibition is given by the percent difference between the number of infected cells without and those with supernatant. The inhibitory activity was still apparent when the supernatant was added before infection (24, 16, and 3 h) or shortly after infection (+3 h). However, the addition 24 or 32 h p.i. showed a significantly reduced inhibitory activity (<20% [Fig. 2B]). This observation seems to indicate that the inhibitory effect does not interfere with virus adsorption.

View larger version (28K): [in this window] [in a new window]   FIG. 2.   (A) Suppressive effect of the supernatant from differently stimulated PBL. (B) Suppressive effect of the supernatant from PHA-stimulated PBL on FV-infected human fibroblasts at various times around FV infection.

FV isolation from accidentally infected humans has been successful albeit with low efficiency. In some of the attempts, PHA-activated PBL have been cocultivated with human diploid fibroblasts. Due to the species-specific nature of the suppressive activity of activated PBL, it was possible that virus isolation was hampered in such an autologous coculture system. Therefore, we have compared this procedure with virus isolation with heterologous hamster cells. PBL of an FV-infected individual were kept in culture in complete RPMI medium and stimulated for 3 days with 10 µg of PHA per ml and 100 U of interleukin-2 (Sigma) per ml as described elsewhere (22). Five million stimulated lymphocytes were then cocultivated with 5 × 10⁵ Alpha-1 human fibroblasts or BHK-21 cells. Cells were kept in culture and subcultivated until typical FV CPE was observed. At every subcultivation, IF was performed in addition. As shown in Table 1, BHK-21 cells showed typical FV nuclear fluorescence as early as 4 days after beginning of cocultivation whereas Alpha-1 cells were first positive after 14 days. Accordingly, the first FV-specific CPE could be observed in BHK-21 cells on day 6 of cocultivation but in Alpha-1 fibroblasts only after 16 days. Thus indeed, the use of FV-permissive nonhuman cells in the coculture system significantly accelerated virus isolation.

View this table: [in this window] [in a new window]   TABLE 1.   Enhancement of FV isolation by cocultivation of human PBL with heterologous permissive cells or by addition of anti-IFN- antibodies to a homologous coculture system

We have speculated that IFNs might be a component of the FV-suppressive activity. IFNs play a central role in the resistance of mammalian hosts to pathogens (2). They act in a species-specific manner (1), are produced by PBL upon activation (26), and are known to inhibit FV (18, 19). In vivo, as demonstrated with IFN receptor knockout mice, the unresponsiveness to IFN makes the animals highly susceptible to virus infections including those with vaccinia virus, lymphocytic choriomeningitis virus, and coronavirus (13, 20, 23). To evaluate the role of IFN- in the supernatant of PHA-activated PBL, the supernatant was first added 3 h p.i. at different dilutions to FV-infected Alpha-1 fibroblasts and found to inhibit FV infection in a dose-dependent manner (Fig. 3A). The highly suppressive 1/100 dilution was then preincubated with monoclonal antibodies against human IFN- (kind gift of L. Ozmen, Hoffmann La Roche, Basel, Switzerland) for 1 h at room temperature and added to the infected cell culture at the same time point. A drastic decrease of the suppressive activity was observed (Fig. 3C). Treatment of Alpha-1 human fibroblasts with recombinant human IFN- (kind gift of P. Stäheli, University Freiburg, Freiburg, Germany) led to strong but not complete inhibition of FV infection even when 500 U of IFN- per ml was added (Fig. 3B). The IFN- effect could be completely inhibited by anti-IFN- antibodies (Fig. 3D). Thus, IFN- is a major, but probably not the only, component released by activated PBL able to inhibit FV replication. As measured by an enzyme-linked immunosorbent assay (R&D Systems GmbH, Wiesbaden, Germany), around 1 to 2 ng of IFN-/ml was produced by PHA-activated PBL after 20 h of stimulation. The over 50% suppressive activity at the 1/800 dilution indicates that the FV-inhibitory activity is remarkably potent.

View larger version (58K): [in this window] [in a new window]   FIG. 3.   (A and B) Dose-dependent inhibitory effect of the supernatant from PHA-activated PBL (A) or recombinant human IFN- (B) on FV CPE in human fibroblasts. (C and D) Neutralization of the suppressive activity of PBL-PHA supernatant (1:100 dilution) (C) and IFN- (100 U/ml) (D) by specific anti-IFN- monoclonal antibodies. As a specificity control for anti-IFN- monoclonal antibodies, control mouse immunoglobulin G of the same isotype was used.

The strong species-specific FV-inhibitory effect of supernatants from activated PBL has an immediate practical implication: attempts to isolate FV should preferentially be carried out in heterologous coculture systems. In fact, this is concordant with old protocols in which heterologous long-term cocultivation has proved sensitive for the rescue from animal cells of endogenous viruses otherwise very difficult to isolate (17, 24). Alternatively, to overcome the risk of heterologous species selection, homologous coculture systems in the presence of anti-IFN- neutralizing antibodies could be used. Indeed, cocultivation of human PBL and Alpha-1 human fibroblasts with 20 µg of monoclonal anti-IFN- antibodies per ml significantly enhanced the efficiency of FV isolation (Table 1). The same treatment did not affect virus isolation with heterologous BHK-21 cells.

IFN- is a major component of the detected FV-suppressive activity. While the exact mechanism is as yet unknown, IFN- may exert its antiviral effect via the induction of responsive genes. At least three such gene products are likely candidates for FV suppression: the double-stranded RNA activated protein kinase (p68 kinase), 2'-5' oligoadenylate synthetase, and double-stranded RNA-specific adenosine deaminase (2). For the first two, direct antiviral activity was demonstrated against encephalomyocarditis virus and vaccinia virus (12) and against picornavirus (4) and encephalomyocarditis virus (5), respectively, while the latter might produce hypermutated viral mRNAs whose translation could lead to nonfunctional proteins (3). Comparable experiments with FV-permissive cells that overexpress the individual candidate genes should help to clarify their potential role in FV suppression.

Although IFN- is the major component of the FV-suppressive activity, anti-IFN- antibodies cannot entirely neutralize this activity, and even very high doses of IFN- (500 U/ml) do not completely block FV replication, as does the supernatant from PBL-PHA. This would suggest that IFN- might act in concert with other soluble factors not yet identified. A recently published example of such a synergy is the noncytopathic inactivation of HBV in the HBV transgenic mouse model by the combined action of IFN- and tumor necrosis factor alpha (8). There, virus clearance was achieved by elimination of viral nucleocapsids and replicative DNA intermediates, as well as destabilization of the viral RNA. The two cytokines were induced either by HBV itself (8) or by an inflammatory immune response to an unrelated virus (7). In this context, some resemblance between the replication strategies of HBV and FV is worth mentioning (27), which might extend to similar ways of virus suppression by the host defense system.

Whether IFN- also plays a protective role against lytic replication of FV in vivo still remains to be elucidated. FVs have been shown not to induce IFNs in infected cell cultures (19). Nevertheless, as suggested by several examples, it may well be that indirect mechanisms would induce sufficient amounts of IFN- to be protective (6, 7, 14). Appropriate in vivo studies appear to be of crucial importance. The development of a small animal model, preferably the mouse, and the use of different knockout mice will be essential in solving the discrepancy of high FV cytopathicity ex vivo and its asymptomatic persistence in vivo.