Previously nitric oxide (NO) was postulated to have a negative effect on HIV-1 replication through a cGMP-independent route 1. However, it was not characterized as to how this cGMP-independent effect manifested mechanistically. On the other hand, it is well-accepted that a major intracellular signaling pathway for NO is through a cytosolic-guanylate cyclase linked cGMP-dependent protein kinase, PKG, pathway 2. cGMP/PKG has been shown to activate abundantly both CREB 3 and NF-κB 45. Interestingly, to our knowledge, no systematic investigation of cGMP/PKG's activity on the HIV-1 LTR has been reported to date.

To address how cGMP/PKG might influence HIV-1 LTR-directed expression, we transfected HeLa cells with a LTR-luciferase reporter with or without a Tat-plasmid 6789- and assayed for expression with or without simultaneously co-transfecting a PKG1β-expression vector 10- (Figure 1). We found that PKG1β-alone activated reporter expression by approximately 4 fold (also see Figure 2B) while Tat-alone activated expression by >45 fold (Figure 1A). Co-transfection of LTR-luciferase + Tat + PKG1β activated expression cumulatively to >175 fold (Figure 1A). These results are consistent with PKG1β inducing LTR-driven expression whether in the absence or presence of Tat.

To verify more physiologically the above LTR-reporter assay, we next checked the effect of PKG1β on an HIV-1 infectious molecular clone, pNL 4-3 (Figure 1B). Here, increasing amounts of PKG1β-plasmid were transfected into cells with a constant level of pNL 4-3. Viral proteins expressed from the HIV-1 molecular clone were then measured. Using HIV-specific hyper-immune sera in Western blots, we observed that PKG1β increased pNL 4-3 expression in a dose dependent manner (Figure 1B, lanes 2 – 6). To investigate whether the intact function of PKG1β was needed for this activity, we created two loss-of-function PKG1β deletion mutants (Figure 2A). Both deletion mutants were unable to activate either a LTR-reporter (Figure 2B) or an HIV-1 infectious molecular clone (Figure 2C).

Optimal PKG activity is dependent on activation by cGMP 11. While over expression of exogenously transfected PKG offered significantly measureable effects (Figures 1, 2), we wished to understand next how cell endogenous PKG might act mechanistically in response to cGMP treatment. Elsewhere, it was reported that NF-κB p65, p52, and p50 are substrate proteins activated by PKG-mediated phosphorylation. Because expression of the HIV-1 LTR is regulated by NF-κB 1213, we asked if cGMP activated NF-κB in our experimental.

To assess if NF-κB was activated by cGMP, we assayed for enhanced nuclear localization of NF-κB protein p65, which is one measure of activation 1415. We treated cells with a cGMP agonist, 8-pCPT-cGMP, and compared results to cells treated with a known NF-κB activator, tumor necrosis factor alpha, TNFα. When cytosolic and nuclear p65 proteins were assayed, we observed that 8-pCPT-cGMP behaved quantitatively very similarly to TNFα in inducing increased NF-κB p65-translocation into the nucleus (Figure 3A). This result supports that PKG-activation of HIV-1 (Figures 1, 2) acts through an NF-κB-mediated process. To confirm that PKG1β can directly activate NF-κβ activity, we transfected a NF-κβ luciferase reporter into cells with or without a co-transfected PKG1β expression plasmid (Figure 3B). Results from this assay showed that NF-κβ-dependent luciferase activity was indeed elevated by PKG1β.

A more direct verification of cGMP/PKG's role in HIV-1 replication can be established using chemical agonists and antagonists of this pathway seeking for effects on viral infection. Because the NO/cGMP signaling pathway has been reported to function potently in cells of macrophage lineage, we performed infection of monocyte derived macrophages (MDMs) using HIV-1 strain AD8 1617. MDMs infected with AD8 were treated individually with three different drugs. We employed two cGMP agonists (8-pCPT-cGMP, and Sp-8-pCPT-cGMP) and one cGMP antagonist (Rp-8-pCPT-cGMPs) and monitored their impact on HIV-1 replication. Informatively, in two separate experiments, both agonist drugs increased virus replication over control-treated infection, while the cGMP-antagonist drug decreased (or did not affect) virus replication (Figure 4).

Here, we report evidence that both in the absence and presence of Tat the cGMP/PKG pathway can serve to modulate HIV-1 expression/replication. Understanding how HIV-1 LTR expression is affected by ambient cellular pathways 181920 may help to address potential approaches for treating latent HIV-1 infection 21. The current findings may be important because cGMP is a ubiquitous second messenger that affects multiple cellular pathways in most, if not all, cells. Accordingly, cGMP-influenced pathways are likely to interdigitate with some of the signaling routes utilized by HIV-1 in infected cells 22. Additionally, because many cGMP chemical agonists and antagonists are available 2324, practical chemotherapeutic interventions in these pathways (if they should be useful for anti-viral purposes) could be amenable.

The author(s) declare that they have no competing interests.

JHL and VY carried out the experiments. JHL, VY, and KTJ conceived of the study and wrote the manuscript.