HIV-1 IN catalyzes the integration of the viral DNA into the cellular genome in a preintegration complex composed of cellular and viral proteins. The dynamic of these interactions regulates IN activities but also non catalytic activities such as nuclear import and tethering of the complex to the integration site. Studies have provided important insights into the post-translational modifications (PTM) of IN, which regulate its multifaceted functions. Yet, how IN functions are fine-tuned by these PTM is not determined. In a previous work, we identified GCN2 as a cellular partner of IN via a yeast two-hybrid system. GCN2 is a cellular protein kinase involved in stress response along with PERK, PKR and HRI. Different stresses have been shown to activate GCN2 such as amino acid starvation and UV irradiation. Upon autophosphorylation (p-GCN2), p-GCN2 phosphorylates eIF2α, which results in the control of translation. GCN2 has been implicated in human burdens such as cancer and Alzheimer’s disease and in addition, we showed that GCN2 is activated during HIV-1 infection (Cosnefroy et al. 2013). In vitro, we used recombinant enzymes to show that GCN2 efficiently phosphorylates IN from HIV-1 but also from other retroviruses, such as MLV and ASV (ANRS 2013-2016). In cells, depletion of GCN2 strikingly increased infectivity of HIV-1, enabling us to establish a link between GCN2 and retroviral replication. In agreement, we showed that infectivity of HIV-1 was also increased in the context of viruses harboring IN mutations unable to sustain phosphorylation by GCN2. Although reverse transcription was not affected, integration was increased in these mutant viruses. Further in vitro analysis demonstrated the formation of a complex between GCN2 and LEDGF/p75, a partner of HIV-1 IN. GCN2 does not phosphorylate LEDGF/p75 alone but the phosphorylation of the latter requires the presence of IN, suggesting a tripartite complex. Mass spectrometry analyses and use of deletion mutants pinpointed to the central domain of LEDGF as the target of the IN-GCN2 complex.
Similarly, we studied the capability of GCN2 to phosphorylate BRD4, the tethering factor of MLV IN. Interestingly, the phosphorylation of the tethering factor by GCN2 is possible only when used with its own IN, from HIV for LEDGF and from MLV for BRD4. Thus, the data indicate that retroviral IN could serve as a docking platform for GCN2 that in turn phosphorylates proteins within the complex. While GCN2 is becoming an attractive target in the field of cancer and neurodegenerative diseases, our data indicate that it could also be the case for HIV therapy. As such, our project will focus on understanding the regulation operated by GCN2.