Modulation of tumor necrosis factor and gamma interferon production by cocaine and morphine in aging mice infected with LP-BM5, a murine retrovirus.

The actions of retroviral infections, aging, and cocaine and morphine injection on cytokine production were investigated in C57BL/6 female mice. Retroviral infection with LP-BM5 murine leukemia virus was further developed as a model of murine acquired immunodeficiency syndrome (AIDS). The effects of cocaine and morphine on gamma-interferon (IFN) and tumor necrosis factor (TNF) production in vivo and with isolated spleen cells were measured by a sandwich enzyme-linked immunosorbent assay (ELISA) method. Serum IFN was generally not detected in any group except mice injected with saline and young mice infected with LP-BM5 virus. Splenocytes from mice with murine AIDS produced less IFN when stimulated in vitro by ConA. In aged mice, IFN production by spleen cells was severely suppressed by retroviral infection. Cocaine had a tendency to suppress IFN production by stimulated cells in vitro. Morphine tended to reduce IFN production by spleen cells from retrovirally infected animals. The serum TNF level in mice with murine AIDS was elevated creating higher levels in morphine and morphine plus cocaine treated uninfected mice while cocaine injection eliminated serum TNF. When stimulated in vitro by lipopolysaccharides (LPS), splenocytes from mice with murine AIDS also produced more TNF than uninfected controls. TNF production in vitro and in vivo was significantly increased by retroviral infection. Therefore, results indicate that cocaine and retroviral infection modulate TNF and IFN production.     

Cocaine injection and coxsackievirus B3 infection increase heart disease during murine AIDS

Coxsackievirus initiates myocarditis especially in the immunologically deficient or immature. To test whether Coxsackievirus B3 (CVB3) induced pronounced cardiomyopathy during severe immune dysfunction of murine AIDS, female C57BL/6 mice were infected with LP-BM5 retrovirus and superinfected with CVB3. Some were also injected daily with cocaine hydrochloride in 0.9% saline solution (30 mg/kg) intraperitoneally, because cocaine also suppresses cellular immune response. Heart tissue was analyzed histopathologically. Mice experiencing concurrent retrovirus and Coxsackievirus infection had a high degree of cardiac lesions consistent with myopathy compared with findings in uninfected animals (p <.05). Cocaine injection during murine retrovirus infection greatly exacerbated the pathogenesis of Coxsackievirus infection. C57BL/6 mice, essentially resistant to Coxsackievirus-induced cardiomyopathy, became susceptible during the immune dysfunction in murine AIDS. This suggests that retrovirus infection causes conditions favoring Coxsackie-induced cardiac lesions. Interleukin (IL)-2 and IL-4 expression by splenocytes from the dually infected retrovirus and Coxsackievirus group showed no significant differences when the animals were also cocaine treated. However tumor necrosis factor TNF-alpha production was significantly decreased in dually infected retrovirus + Coxsackievirus mice treated with cocaine, compared with findings in various controls (p <.05).     

Evaluation of a murine model of hepatic candidiasis.

A murine model of focal hepatic candidiasis which we suggest simulates certain conditions of this clinical variant of systemic candidiasis in leukemic patients is described. We have shown that outbred mice inoculated with Candida albicans by the oral-intragastric route as infants (6 days old) and then immunocompromised by cyclophosphamide and cortisone acetate treatment 2 weeks later demonstrate systemic spread of the opportunistic pathogen to the liver, lungs, spleen, and kidneys. Treatment with the immunosuppressive drugs cyclophosphamide and cortisone acetate resulted in alteration of the normal integrity of the mucosal epithelium of the gut as well as in granulocytopenia. Approximately 55% of the animals with C. albicans infections in the liver demonstrated hepatic abscesses. After these same infected, immunocompromised animals were treated with suboptimal dosages of antifungal agents (cilofungin or amphotericin B), either by intraperitoneal or subcutaneous (s.c.) routes, persistent hepatic abscesses were fewer in number and delimited by a distinct outer layer of host tissue but still contained large numbers of the viable pathogen. Blood cell counts indicated that these antifungal drug-treated animals had reestablished approximately the same number of leukocytes per microliter of blood as estimated prior to the immunocompromising drug treatment. Similar conditions in leukemic patients who were in remission and who were undergoing antifungal drug therapy for systemic candidiasis have been reported. Clearance of hepatic infections in mice was accomplished by using appropriate concentrations of amphotericin B administered by daily intraperitoneal or s.c. injection for 5 to 7 days or cilofungin by continuous s.c. infusion for 7 days. However, systemic antifungal therapy did not significantly reduce numbers of C. albicans cells in the stomach and esophagus. Persistent foci of gastrointestinal colonization by C. albicans, especially in the region of the cardial-atrium fold of the stomach of these mice, are reservoirs of the opportunistic pathogen from which reinfection may occur, leading to relapse of systemic candidiasis.     

Gamma interferon is not essential in host defense against disseminated candidiasis in mice.

In vitro studies have suggested a role for interferon gamma (IFN-gamma) in host defense against disseminated candidiasis, but in vivo studies are inconclusive. We utilized homozygous IFN-gamma knockout (GKO) mice to determine if the cytokine is essential in host defense against this disease. Genotypes of mice were determined by PCR with specific primers for the normal or disrupted IFN-gamma gene. The GKO status of the mice was confirmed by an enzyme-linked immunosorbent assay, which showed no detectable IFN-gamma produced by their splenocytes stimulated by concanavalin A. To test the susceptibility of GKO mice to candidiasis, the animals were infected either intravenously (i.v.) or intragastrically (i.g.) with Candida albicans. GKO mice infected i.v. survived as long as wild-type (WT) mice and showed no difference in Candida CFU counts in liver, spleen, or kidneys compared to those for WT mice. When animals were given Candida i.g., at 3 h or at 10 or 21 days after infection, there was no dissemination of Candida to the lung, liver, spleen, or kidneys for either GKO or WT mice. There was no difference in Candida CFU counts recovered from the stomach or intestines between GKO and WT mice. Histological examination of the stomach cardial-atrium fold, where the fungus was located, showed that GKO mice did not have evidence of more tissue damage or fungal invasion than WT mice. Finally, the jejunum for both types of mice showed no evidence of tissue damage or fungal invasion. These studies indicate that IFN-gamma is not essential in host defense against C. albicans that originates from a mucosal site or that is given directly into the bloodstream in a mouse model.     

Cryptococcus neoformans infection in mice previously infected with LP-BM5 MuLV, the agent of murine AIDS (MAIDS)

We studied susceptibility to experimental systemic cryptococcosis in mice previously infected with the retroviral complex LP-BM5 (responsible for murine AIDS). LP-BM5 was inoculated to C57B1/6 mice by intravenous (i.v.) injection 8 weeks before an i.v. challenge with 4×l0³ CFU of Cryptococcus neoformans. Uninfected and singly infected mice were used as controls. LP-BM5 infection did not result in a significant increase in fungal burdens in the lungs or brains of co-infected animals compared to mice infected with C. neoformans alone. However, mortality was enhanced in the co-infected animals. The kinetics of splenocyte subsets differed in co-infected mice and LP-BM5-infected mice; the increase in CD4⁺, CD8⁺ and Ly5⁺ cells was only moderate in the former. Cytokine production by concanavalin A (Con A)-stimulated splenocytes from co-infected mice showed a marked decrease in the Thl response (IFN-γ, IL-2) and an increase in the Th2 response (IL-4, IL-10). Furthermore, cryptococcosis altered the course of MAIDS, inhibiting splenomegaly. This effect was not related to a decrease in ecotropic virus titres in the spleen or to improved in vitro responsiveness of spleen cells to Con A. The marked decrease in IFN-γ production in co-infected animals could partly explain the inhibition of LP-BM5-induced splenomegaly. This model of murine retroviral infection does not seem to be suitable for studying cryptococcosis in immunosuppressed animals, but remains valuable for investigating in vivo interactions between two pathogens.     

Interactions between murine AIDS (MAIDS) and toxoplasmosis in co-infected mice.

We coinfected C57B1/6 mice with LP-BM5 murine leukaemia viruses, responsible for murine AIDS (MAIDS), and an avirulent strain of Toxoplasma gondii. Virus-infected mice were infected perorally on day 30 with 10 cysts of T. gondii, and T. gondii-infected mice were challenged with LP-BM5 on day 20, 30 or 60 after parasite inoculation. Uninfected and singly infected mice were used as controls. The kinetics of parasite burden in blood, lungs and brain, together with blood lymphocyte subsets, and spleen and lymph node weights, were serially determined in each group of mice. The kinetics of parasite counts in mice infected by LP-BM5 then by T. gondii were similar to those in mice infected by T. gondii only, except for lung counts, which reached higher values than in animals infected with T. gondii alone, then fell and re-increased until the end of the experiment. The only significant change in parasite burdens when mice were first infected by T. gondii and then by LP-BM5, compared with T. gondii controls, was an increase in lung counts in mice challenged with LP-BM5 20 days after T. gondii inoculation. Whatever the schedule of co-infection, the kinetics of lymphocyte subsets in co-infected mice differed from those in T. gondii- or LP-BM5-infected mice; in dually infected mice CD4+ and CD8+ cell counts were intermediate between values in mice singly infected by the parasite or the virus. Enlargement of spleen and lymph nodes, which is a major criterion of MAIDS progression, was significantly less marked in co-infected mice than in mice infected with LP-BM5 alone. These data point to cross-regulation of T. gondii and LP-BM5 infections, which results in increased susceptibility to T. gondii, and may alter the progression of MAIDS.     

Tsl and LP-BM5: a comparison of two murine retrovirus models for HIV

The ts1 murine leukemia virus produces an immunodeficiency state in mice that parallels human immunodeficiency virus (HIV) infection in humans. Other murine leukemia viruses, such as LP-BM5 used in the murine acquired immune deficiency virus (MAIDS) model, have been studied extensively as a small animal model for HIV research, but lack many key similarities to HIV. Mice infected with ts1, however, utilize CD4 target cells for infection, undergo neuronal loss and demyelination, and develop clinical immunodeficiency. These features make this retrovirus in many ways an ideal candidate for a small animal model for HIV research. In this review article, the early development, the molecular and clinical pathogenesis of both the ts1 mutant of the Moloney murine leukemia virus and LP-BM5 are examined. Based on an extensive evaluation of the literature on LP-BM5 and ts1, it is concluded that the ts1 virus may serve as a better animal model to human retrovirus infection.     

CD40-deficient mice infected with the defective murine leukemia virus LP-BM5def do not develop murine AIDS but produce IgE and IgG1 in vivo

CD40-deficient mice, when inoculated with the LP-BM5def murine retorvirus, become infected and show virus expression similar to wild-type mice. However, unlike the wild-type mice, CD40-deficient mice do not develop symptoms of immunodeficiency, lymphoproliferative disease and the typical histological changes in the lymphoid tissue. These results show that the CD40-CD40 ligand (CD40L) interaction in vivo is essential for anergy induction and the subsequent development of immunodeficiency and pathologic expansion of lymphocytes. Infected CD40-deficient mice and their littermates express a similar pattern of cytokine mRNA, which is not biased towards a Th2 phenotype. Nevertheless, hypergammaglobulinemia is induced in infected wild-type and CD40-deficient mice. Surprisingly, murine AIDS infection even induces IgE production in CD40-deficient mice in vivo. Our data demonstrate that antibody class switch to IgE and IgG1 can be induced by a retroviral infection in vivo even in the absence of CD40-CD40L interaction and an apparent switch to a Th2 cytokine production.     

Opportunistic infections and retrovirus-induced immunodeficiency: studies of acute and chronic infections with Toxoplasma gondii in mice infected with LP-BM5 murine leukemia viruses.

Mice infected with LP-BM5 murine leukemia viruses develop a syndrome, termed mouse AIDS (MAIDS), characterized by increasingly severe immunodeficiency and progressive lymphoproliferation. Virus-infected mice were examined for the ability to resist acute infection and to control chronic infection with the protozoan Toxoplasma gondii, a major opportunistic pathogen of individuals infected with human immunodeficiency virus. Mice infected with the retroviruses for 2 or 4 weeks responded normally to challenge with the parasite, but mice inoculated with the protozoan 8 or 12 weeks after viral infection died with acute disease due to T. gondii. Increased sensitivity to acute infection was associated with a reduced ability to produce gamma interferon (IFN-gamma) and with established changes in CD4+ T-cell function. Mice latently infected with T. gondii and then inoculated with the retrovirus mixture were found to reactivate the parasite infection, with 30 to 40% of dually infected animals dying between 5 and 16 weeks after viral infection. Reactivation was associated with reduced proliferation and impaired production of IFN-gamma in response to stimulation with soluble T. gondii antigens or to concanavalin A. Continuing resistance to lethal reactivation in the remaining mice was shown to require CD8+ T cells and expression of IFN-gamma. In addition, it was found that chronic infection with T. gondii altered the course of MAIDS by inhibiting the progression of splenomegaly and immunodeficiency and reducing the expression of both the helper and etiologic defective viruses. These results support previous studies which indicate that infection with T. gondii is controlled by synergistic interactions between CD4+ and CD8+ T cells, the functions of which are progressively impaired during the course of MAIDS.     

Vitamin E supplementation modulates cytokine production by thymocytes during murine AIDS

Female C57BL/6 mice were infected with LP-BM5 retrovirus, causing murine AIDS, which is functionally similar to human AIDS. Retrovirus infection targeted the thymus, producing altered T cell differentiation via the dysregulation of thymocyte cytokine production. Human AIDS causes vitamin deficiencies, therefore the effects of dietary vitamin E supplementation were determined on the kinetics of cytokine production by concanavalin A-stimulated thymocytes in uninfected normal mice and mice with murine AIDS. Dietary supplementation, with a 15-fold increase in vitamin E (160 IU/1) in the liquid diet (National Research Council), modulated interleukin-2 (IL) production in both uninfected mice and retrovirus-infected mice. Vitamin E significantly reduced the level of IL-4 secretion in the uninfected mice at 4 and 8 weeks, but not at 12 and 16 weeks. It also significantly reduced IL-4 production, elevated by retrovirus infection. Vitamin E significantly reduced IL-6, and interferon-γ production increased in murine AIDS. The effects of dietary vitamin E on concanavalin A-induced proliferation of thymocytes were consistent with the finding of changes in IL-2 secretion. No effects of dietary vitamin E on thymus weight were observed in uninfected or retrovirus-infected mice, whereas vitamin E significantly increased serum and thymic vitamin E concentration, which had been reduced by retrovirus infection. These data indicate that dietary vitamin E supplementation can modulate cytokine production by thymocytes, affecting T cell differentiation, especially during retrovirus-induced immune dysfunction.     

Repeated cycles of alternate administration of fludarabine and Zidovudine plus Didanosine inhibits murine AIDS and reduces proviral DNA content in lymph nodes to undetectable levels

The results of the combined use of Fludarabine, an anticancer agent that may be able to target latently infected cells, and conventional antiretroviral therapy (AZT+DDI) in a murine model of AIDS, i.e., LP-BM5 infection, are reported. Eighty percent of infected mice, treated with four cycles of alternate administration of Fludarabine and AZT+DDI, showed undetectable levels of proviral DNA in lymph nodes. After 8 weeks of treatment interruption, the infected/treated animals, although still alive at a time when all untreated animals had succumbed to the infection, showed disease progression and reappearance of proviral DNA in lymph nodes. The retrospective analysis of proviral DNA content in spleen and bone marrow at the end of the fourth cycle of treatment revealed a low but detectable amount of BM5d proviral DNA. We thus concluded that the spleen and bone marrow may be less sensitive to lympholitic drugs and therefore act as viral reservoirs in LP-BM5 infection. This study suggests that optimized protocols of alternate administration of cytolytic and antiretroviral drugs may represent a useful strategy to eradicate retroviral infections.     

ANTI-INFLAMMATORY EFFECTS OF THEOPHYLLINE: MODULATION OF IMMUNE FUNCTIONS DURING MURINE LEUKEMIA VIRUS INFECTION

LP-BM5 murine leukemia virus induces immune dysfunction leading to B cell leukemia and murine AIDS with cytokine dsyregulation. Theophylline induces apoptosis of leukemia cells in humans. Therefore the effects of theophylline on immune dysfunction in a murine model of leukemia were investigated. C57BL/6 mice consumed drinking water containing 0.3% theophylline beginning 2 weeks after murine retrovirus infection for 4 months. Theophylline largely prevented the retrovirus induced splenomagaly, lymphodenopathy, reduction in B and T cell proliferation, and suppression of Th1 cytokines (IL-2) secretion. It also suppressed Th2 cytokine (IL-4, TNF-alpha, and IL-10) production, which was otherwise stimulated by retrovirus infection. These data suggest that immune dysfunction, induced by murine retrovirus infection, was largely prevented by theophylline treatment.     

ANTI-INFLAMMATORY EFFECTS OF THEOPHYLLINE: MODULATION OF IMMUNE FUNCTIONS DURING MURINE LEUKEMIA VIRUS INFECTION

LP-BM5 murine leukemia virus induces immune dysfunction leading to B cell leukemia and murine AIDS with cytokine dsyregulation. Theophylline induces apoptosis of leukemia cells in humans. Therefore the effects of theophylline on immune dysfunction in a murine model of leukemia were investigated. C57BL/6 mice consumed drinking water containing 0.3% theophylline beginning 2 weeks after murine retrovirus infection for 4 months. Theophylline largely prevented the retrovirus induced splenomagaly, lymphodenopathy, reduction in B and T cell proliferation, and suppression of Th1 cytokines (IL-2) secretion. It also suppressed Th2 cytokine (IL-4, TNF-alpha, and IL-10) production, which was otherwise stimulated by retrovirus infection. These data suggest that immune dysfunction, induced by murine retrovirus infection, was largely prevented by theophylline treatment.     

Prevention of immune dysfunction and vitamin E loss by dehydroepiandrosterone and melatonin supplementation during murine retrovirus infection

Female C57BL/6 mice infected with the LP-BM5 leukaemia retrovirus developed murine acquired immune-deficiency syndrome (AIDS). Dehydroepiandrosterone (DHEA) and melatonin (MLT) modify immune dysfunction and prevent lipid peroxidation. We investigated whether DHEA and MLT could prevent immune dysfunction, excessive lipid peroxidation, and tissue vitamin E loss induced by retrovirus infection. Retrovirus infection inhibited the release of T helper 1 (Th1) cytokines, stimulated secretion of Th2 cytokines, increased hepatic lipid peroxidation, and induced vitamin E deficiency. Treatment with DHEA or MLT alone, as well as together, largely prevented the reduction of B- and T-cell proliferation as well as of Th1 cytokine secretion caused by retrovirus infection. Supplementation also suppressed the elevated production of Th2 cytokines stimulated by retrovirus infection. DHEA and MLT simultaneously reduced hepatic lipid peroxidation and prevented vitamin E loss. The use of DHEA plus MLT was more effective in preventing retrovirus-induced immune dysfunction than either DHEA or MLT alone. These results suggest that supplementation with DHEA and MLT may prevent cytokine dysregulation, lipid oxidation and tissue vitamin E loss induced by retrovirus infection. Similarly, hormone supplementation also modified immune function and increased tissue vitamin E levels in uninfected mice.     

Avirulence of Candida albicans CaHK1 Mutants in a Murine Model of Hematogenously Disseminated Candidiasis

Deletion of both alleles of the Candida albicans CaHK1 gene, which causes cells to flocculate when grown at pH 7.5, a pH comparable to that of mammalian blood, abolishes the ability of the yeast to establish a successful infection in a murine model of hematogenously disseminated candidiasis. Within 72 h all mice inoculated with the parental C. albicans strain had died. The mice infected with either the heterozygote or revertant strain, either of which harbors only one functional CaHK1 allele, also succumbed to the infection, although survivors were observed for up to 16 days postinfection. However, mice inoculated with the Deltacahk1 null strain survived for the course of the infection. These results indicate that CaHK1 is required for the virulence of C. albicans in a murine model of hematogenously disseminated candidiasis. In contrast, CaHK1 is not required for the virulence of C. albicans in a rat model of vaginal candidiasis.     

Analysis of the Capacity to Produce IL-3 in Murine AIDS

Adult C57BL/6 mice infected with LP-BM5 murine leukaemia virus represent a model of murine AIDS (MAIDS). In this study we have analysed the capacity of CD4⁺ T cells from infected mice to produce IL-3 following stimulation with ConA for 24–72 h. In contrast to the position with IL-2, the production of which is markedly impaired during LP-BM5 infection, similar levels of IL-3 were measured in culture supernatants of splenocytes from infected and uninfected mice harvested at 24 h of stimulation. Forty eight and 72 h of ConA stimulation led to increasing levels of IL-3 being measured in cultures from uninfected mice, whilst in cultures from infected animals, IL-3 levels remained stagnant. Similar results were obtained 4, 8 and 13 weeks post-infection. In view of the fact that parallel experiments revealed markedly impaired proliferative responses to ConA during MAIDS, we conclude that IL-3 production is basically intact at the cellular level in T cells during MAIDS; but when in a situation requiring clonal expansion of the activated T cells, IL-3 production will be inhibited owing to the impaired capacity for proliferation.     

Effects of preinduced Candida-specific systemic cell-mediated immunity on experimental vaginal candidiasis.

It has been postulated that systemic cell-mediated immunity (CMI) is an important host defense factor against recurrent vaginal infections caused by Candida albicans. Using an estrogen-dependent murine model of vaginal candidiasis, we have previously shown that mice inoculated vaginally with C. albicans acquire a persistent vaginal infection and develop Candida-specific Th1-type systemic CMI. In the present study, experimental vaginitis was monitored in the presence of preinduced systemic Candida-specific CMI. Mice immunized systemically with C. albicans culture filtrate antigens (CaCF) in complete Freund's adjuvant (CFA) had Th1-type reactivity similar to that of vaginally infected mice. CaCF given to mice intravenously induced Candida-specific suppressor T (Ts) cells. Mice preimmunized with CaCF-CFA and given a vaginal inoculum of C. albicans had positive delayed-type hypersensitivity (DTH) reactivity from the time of vaginal inoculation through 4 weeks. Conversely, mice infected in the presence of Ts cells had significantly reduced DTH responses throughout the 4-week period in comparison with naive infected mice. However, the presence of Th1-type Candida-specific DTH cells or Ts cells, either induced in mice prior to vaginal inoculation or adoptively transferred at the time of inoculation, had no effect on the vaginal Candida burden through 4 weeks of infection. A similar lack of effects was obtained in animals with lower Candida population levels resulting from a reduction in or absence of exogenous estrogen. These results suggest that systemic Th1-type CMI demonstrable with CaCF is unrelated to protective events at the level of the vaginal mucosa.     

Local Production of Chemokines during Experimental Vaginal Candidiasis

Recurrent vulvovaginal candidiasis, caused by Candida albicans, is a significant problem in women of childbearing age. Although cell-mediated immunity (CMI) due to T cells and cytokines is the predominant host defense mechanism against C. albicans at mucosal tissue sites, host defense mechanisms against C. albicans at the vaginal mucosa are poorly understood. Based on an estrogen-dependent murine model of vaginal candidiasis, our data suggest that systemic CMI is ineffective against C. albicans vaginal infections. Thus, we have postulated that local immune mechanisms are critical for protection against infection. In the present study, the kinetic production of chemokines normally associated with the chemotaxis of T cells, macrophages (RANTES, MIP-1alpha, MCP-1), and polymorphonuclear neutrophils (MIP-2) was examined following intravaginal inoculation of C. albicans in estrogen-treated or untreated mice. Results showed significant increases in MCP-1 protein and mRNA in vaginal tissue of infected mice as early as 2 and 4 days postinoculation, respectively, that continued through a 21-day observation period, irrespective of estrogen status. No significant changes were observed with RANTES, MIP-1alpha, or MIP-2, although relatively high constitutive levels of RANTES mRNA and MIP-2 protein were observed. Furthermore, intravaginal immunoneutralization of MCP-1 with anti-MCP-1 antibodies resulted in a significant increase in vaginal fungal burden early during infection, suggesting that MCP-1 plays some role in reducing the fungal burden during vaginal infection. However, the lack of changes in leukocyte profiles in vaginal lavage fluids collected from infected versus uninfected mice suggests that MCP-1 functions to control vaginal C. albicans titers in a manner independent of cellular chemotactic activity.     

Follicular dendritic cell function and murine AIDS.

Infection of mice with LP-BM5 elicits an immunodeficiency state referred to as murine acquired immune deficiency syndrome (MAIDS). Shortly after infection, retrovirus particles become associated with follicular dendritic cells (FDC) and this study was undertaken to determine whether retroviruses alter FDC functions. The FDC functions examined included the ability to: (1) retain antigen (Ag) trapped prior to infection; (2) trap new Ag after infection; (3) maintain specific IgG responses; and (4) provide co-stimulatory signals to B cells. Mice were infected with LP-BM5 and the ability of their FDC to trap and retain 125I-Ag (HSA) was assessed. Serum anti-HSA levels were monitored and FDC co-stimulatory activity was indicated by increased B-cell proliferation. HSA trapped on FDC prior to infection began to disappear by 3 weeks and was practically gone by 6 weeks. Serum anti-HSA titres were maintained normally for about 3 weeks after infection and then declined precipitously. The ability of FDC to trap new Ag began to disappear around the second and third week of infection and was markedly depressed by the fourth week. However, FDC recovered from infected mice retained their ability to co-stimulate anti-mu- and interleukin-4 (IL-4)-activated B cells throughout a 5-week period. In short, the ability of FDC to trap and retain specific Ag and maintain specific antibody levels was markedly depressed after retrovirus infection. However, FDC from infected mice continued to provide co-stimulatory signals and these signals may contribute to the lymphadenopathy and splenomegaly characteristic of MAIDS.