UC Berkeley

This dissertation is mainly focused on developing therapeutic strategies to eradicate the human immunodeficiency virus-1 (HIV-1) from infected individuals. Although HIV-1 could be suppressed in infected individuals through combined anti-retroviral therapy (cART), a latent reservoir remains in every infected individual. The reservoir is mainly composed of latently infected resting CD4+ T cells. Since the exceedingly long half-life of the resting CD4+ T cells and the unpredictability of when the latent HIV-1 inside them would become active, life-long medication is needed to prevent the resurgence of HIV/AIDS. The following researches are for seeking and characterizing novel drug targets, testing/repurposing drugs to reactivate latent HIVs, exposing them for recognition and clearance by host immune system, thus purging HIV from infected individuals.

The first part of this dissertation proves that RNA polymerase II (Pol II) transcription elongation is a key step during the reversal of HIV latency. Using the CRISPR/Cas9 technique and complementation assays, I find the human transcription elongation factors ELL2 and AFF1 play predominate roles in HIV-1 transcription in CD4+ T cells. I further discover that, compared with their orthologs, ELL2 and AFF1 constitute only a minor subset of the Super Elongation Complexes (SECs), but are the preferred functional partners for the HIV Tat protein. Through artificially elevating the levels of ELL2 and AFF1 in the cells, latent HIV could be efficiently reactivated. In summary, these results shed light on the mechanisms of the elongation step in HIV transcription and lay the ground work for future researches to develop novel avenues to stimulate the HIV transcription elongation and reactivate the latent viruses.

The second part of this dissertation is a follow up of one of my earlier studies that identified that the bromodomain protein Brd4 promotes HIV latency by binding to the viral LTR to inhibit Tat-induced transcription. Here, I discover that the LTR of latent HIV has low acetylated histone H3 (AcH3) but high AcH4 content, which recruits Brd4 to inhibit Tat-transactivation. Furthermore, I find the lysine acetyltransferase KAT5 but not the paralogs KAT7 and KAT8 promotes HIV latency through acetylating H4 on the provirus. Antagonizing KAT5 removes AcH4 and Brd4 from HIV LTR, enhances loading of the Super Elongation Complex, and interferes with the establishment of latency. Thus, the KAT5-AcH4-Brd4 axis is a key regulator of HIV latency and a potential therapeutic target for eradicating latent HIV reservoirs.

The third part of this dissertation uses the CRISPR-inhibition (CRISPRi) technology to extensively screen the functions of all the 20,000 human genes in maintaining HIV latency. The result not only includes several known genes, but also discovers several so far unreported genes playing vital roles in HIV latency. Through verification in cell line models, I found inhibition of these genes could significantly reactivate latent HIV. Excitingly, several hits in the screen are subunits of proteasomes, and there are FDA-approved drugs targeting proteasomes. To test the effect of these drugs on reactivating latent HIV, I isolated primary CD4+ T cells from 13 ART-suppressed HIV-1 infected individual and used the drugs at different concentrations and in combination with other drugs to treat the cells. The results demonstrate that the FDA-approved proteasome inhibitors could indeed significantly enhance the reversal of HIV latency, without inducing global T cell activation or proliferation.

In summary, this dissertation proves that enhancing the activity of the human transcription elongation machinery is an effective avenue to reverse HIV-1 latency. The insights gained in this dissertation could potentially benefit future therapeutic intervention to eradicate HIV/AIDS.