The innate immunity is imperative during the early phase of host defense against various infections before an antigen-specific adaptive immune response is induced. Given that the majority of CD4 T cells in vivo are in a resting stage, APOBEC3G could function as a highly effective barrier to prevent extensive replication of HIV-1 and subsequently restrict its massive cytopathic effect on CD4 T cells. Unfortunately, this barrier is not yet reliable enough. It has been known that in cell cultures, HIV-1 could initially infect resting primary CD4 T cells and stay in a preintegration latency condition. Work from Greene's lab has indicated that APOBEC3G may play a major role in restricting HIV-1 replication in resting CD4 T cells (5
). It seems that APOBEC3G, at its native concentrations in resting CD4 T cells, only transiently blocks the reverse transcription of HIV-1 and cannot completely eradicate the viruses, resulting in a preintegration latency. When resting CD4 cells are activated, a large amount of these viruses continue to complete their life cycle and generate progeny viruses, indicating that APOBEC3G in resting CD4 T cells could just transiently block reverse transcription and viral replication. Our data have demonstrated that IFN-α is able to enhance the anti-HIV-1 activity of APOBEC3G by increasing its concentration and keeping its association with an LMM. As very few or no viruses can be recovered after activation of the infected CD4 T cells (Fig. ), IFN-α-enhanced APOBEC3G can significantly inhibit reverse transcription and irreversibly inactivate HIV-1 viruses in the preintegration stage. This effect is in contrast to the transient restriction mediated by APOBEC3G at its normal concentration. These differences are of great in vivo relevance. HIV-1 could infect resting CD4 T cells and then be restricted by basic APOBEC3G in vivo. However, the infected resting CD4 T cells could be activated by antigen or cytokine stimulation at any moment. Upon activation, APOBEC3G will be associated with a high molecular mass rather than an LMM, and its inhibitory effect upon HIV-1 could be decreased. Therefore, the viruses blocked at the preintegration stage will continue their life cycle. However, if the IFN-α/APOBEC3G pathway is activated in resting CD4 T cells, the viruses blocked at the preintegration period will be inactivated and no live viruses will be produced upon activation. In summary, our work has demonstrated that interferon, a well-known antiviral innate immunity system, can potently regulate the expression and distribution of APOBEC3G in CD4 T cells. Similar results have been reported for macrophages and hepatocytes (21
). Therefore, the connection between two important antiviral innate immunity systems, the interferon system and APOBEC3G and its family members, has been identified.
We have simultaneously examined the pattern of the APOBEC3G-associated complex in the activated cells treated with IFN-α, along with several other treatments. As shown in Fig. , the majority of APOBEC3G is associated with an LMM in the activated CD4 T cells treated with IFN-α. It should be emphasized, however, that the total concentration of APOBEC3G in the activated CD4 T cells is significantly decreased by IFN-α (Fig. ). Therefore, the antiviral effect of LMM-associated APOBEC3G could be quite limited in the activated CD4 T cells. Conversely, as APOBEC3G has already been associated with an LMM in activated CD4 T cells treated with IFN-α, the poor replication of HIV-1 in the activated phase of CD4 T cells which are initially treated with IFN-α at their resting phase could not be due to the possible “locked” LMM-associated APOBEC3G which occurs during the resting stage of cells. Instead, it is more likely due to the fact that the viral infectivity is significantly inactivated in the resting stage by the treatment of IFN-α. The single-round infection experiments with a reporter gene and intracellular reverse transcription have further supported this hypothesis (Fig. and ). Furthermore, the inhibitory effect of IFN-α delivered transiently or continuously to the activated CD4 T cells upon HIV-1 replication is much less than the inhibitory effect of IFN-α delivered transiently to the initially quiescent CD4 T cells upon HIV-1 replication (Fig. ), which also supports this hypothesis.
Previous studies have already demonstrated that IFN-α has anti-HIV-1 activity for which several mechanisms have been proposed. It can inhibit the process of reverse transcription at an early stage, restrict the generation of viral particles but not viral proteins in chronically infected cells, down-regulate viral protein synthesis, or suppress the HIV-1 LTR promoter (6
). It is noteworthy that the addition of IFN-α into cell cultures of resting primary T lymphocytes before mitogen stimulation could have a much stronger inhibitory effect on reverse transcription (26
), which is consistent with our observations (Fig. ). Moreover, IFN-α has been used in clinical trials to treat HIV-1-infected individuals. However, findings from early clinical trials in the pre-highly active antiretroviral therapy (HAART) era have shown that IFN-α plus reverse transcriptase inhibitor(s) is effective but unable to completely control HIV-1 replication (1
). It should be emphasized, however, that HIV-1 extensively replicates in the replicating CD4 T cells and quickly kills them in the pre-HAART era. As IFN-α does not enhance the concentration of APOBEC3G in replicating CD4 T cells, APOBEC3G could not play a leading role for IFN-α to restrict HIV-1 replication in replicating CD4 T cells. In this report, we have demonstrated that APOBEC3G is IFN-α inducible and makes important contributions for the anti-HIV-1 activity of IFN-α in resting CD4 T cells. As APOBEC3G potently inhibits the reverse transcription of HIV-1 in resting T lymphocytes, IFN-enhanced APOBEC3G expression could mediate the potent inhibitory effect upon reverse transcription by IFN-α. Importantly, because the inactivation of HIV-1 in resting CD4 T cells by the IFN-α/APOBEC3G pathway is so remarkable, it could lead to a novel therapeutic strategy for IFN-α to treat HIV-1-infected individuals. Intriguingly, a very preliminary study has shown that IFN-α in combination with HAART could be more effective in decreasing viral RNA in blood plasma and PBMC-associated viral RNA and DNA than HAART alone (10
It is notable that the concentrations of IFN-α observed in these experiments are relatively high compared to the concentrations in the blood plasma of individuals administrated IFN-α (24
). However, these experiments are “proof-of-concept” experiments. In the coming experiments, we are going to investigate the effect of IFN-α at “realistic” concentrations upon intracellular APOBEC3G and anti-HIV-1 activity. As the first step of a series of in vivo experiments, it is interesting to investigate whether IFN-α which is administrated with a regular dose could enhance the expression of APOBEC3G in resting CD4 T cells.
In the post-HAART era, the replication of HIV-1 in replicating CD4 T cells is significantly prevented. Latent HIV-1 infection is the major obstacle in eradicating residual HIV-1 viruses. It has been demonstrated that resting CD4 T cells are the major reservoirs for HIV-1 viruses. HIV-1 may maintain latency in these cells by postintegration latency in the memory CD4 T cells or by a “cryptic” low-level replication (12
). There have been many attempts to eradicate residual HIV-1. It is likely that APOBEC3G at its normal concentration in resting CD4 T cells cannot block this replication. Based on our findings, we propose to prevent residual HIV-1 replication by strengthening APOBEC3G-mediated intracellular innate immunity. If IFN-α is regularly administrated, or alternatively, secreted by interferon-producing cells (28
) stimulated with certain cytokines, up-regulation of APOBEC3G in resting CD4 T cells could block HIV-1 viruses to complete the “cryptic” low-level replication or abort the established latency, at least the preintegration latency. Our finding could start a “new thinking” to reevaluate IFN-α in HIV/AIDS clinics, especially when we do not have any other reliable way to control the “cryptic” replication at present.