Biomedical Sciences Graduate Program | University of California, San Francisco

We are interested in the molecular biology of severe combined immunodeficiencies, be they genetic (BLS) or acquired (AIDS). These investigations focused on many aspects of transcription, cellular signaling, and retroviral replication. Fundamental insights with important consequences for eukaryotic biology have been forthcoming.

AIDS is caused by the human immunodeficiency virus (HIV). Transmitted via secretions or intravenously, HIV enters cells of the immune system and replicates at a furious rate. Originally, a vigorous immune response suppresses the growth of the virus. However, the recruitment of T helper cells is inadequate for cytolytic T cells to kill all virally infected cells. We found that this impairment is due to the inhibition of antigen processing and presentation via the major histocompatibility complex class II (MHC II) molecules on the surface of infected macrophages. The viral transactivator Tat competes with the class II transactivator (CIITA) for the binding of their essential coactivator, the positive transcription elongation factor b (P-TEFb) and blocks the transcription of MHC II genes.

Additionally, we were among the first to demonstrate that HIV establishes a latent reservoir in the infected host. These infected cells harbor proviruses that are transcriptionally inactive. RNA polymerase II (RNAPII) is arrested slightly downstream of the viral promoter. Copied, short transcripts form the transactivation response (TAR) RNA structure that binds Tat. Thus, upon cellular activation that synthesizes endogenous Tat or the addition of exogenous Tat, the replication of HIV commences and fully competent viruses infect new cells. This latent reservoir of HIV remains an insurmountable problem in the eradication of the virus from the body. Now, we know how this proviral latency is established and how it can be overcome. In the mouse, my studies revealed that Tat can be secreted from the b-cells of the pancreas, is distributed throughout the organism without adverse effects and wakes up silent proviruses in latently infected cells.

Studies on Tat led me to discover the regulation of elongation of transcription. Although this type of transcriptional control is used extensively in the prokaryotic world, before HIV, it was ignored in the eukaryotic world. First, we demonstrated that Tat only affects rates of elongation of transcription. Moreover, it modified RNAPII that has already transcribed its substrate, TAR. Further studies revealed that Tat accesses two important cyclin-dependent kinases, TFIIH and P-TEFb. The first contains cyclin H and Cdk7, which are required for promoter clearance. The second contains cyclins T1, T2 or K and Cdk9, which are required for the elongation of transcription. Moreover, P-TEFb phosphorylates the C-terminal domain (CTD) of RNAPII. Other important players are negative transcription elongation factors (N-TEF), which consist of DRB-sensitivity inducing (DSIF) and negative elongation (NELF) factors. DSIF and NELF interact with each other. In the case of HIV, NELF binds to the double stranded stem in TAR. Tat and cyclin T1 bind to the 5’ bulge and central loop in TAR.

We found this sequence of events. In the absence of Tat, RNAPII is assembled on the HIV promoter and copies TAR. N-TEF then aborts further elongation. Upon the synthesis of Tat, the complex between P-TEFb and Tat binds to TAR, phosphorylates N-TEF and CTD, removes general transcription and negative elongation factors, upon which RNAPII copies the gene. Now, we also know the structure of the complex between P-TEFb, Tat and TAR.

One major question remained. If Tat is required for the copying of the viral genome, then how is Tat made in the first place. We found that NF-kB was required for this effect. This activator binds to the HIV long terminal repeat (HIV LTR) and stimulates the elongation of viral transcription. It is composed of p50 and p65 (RelA). We found that RelA also binds to P-TEFb, so that elongation of viral transcription can occur in the absence of Tat. However, Tat is much more efficient, and RNA presentation of P-TEFb is preferred to DNA presentation. This finding explains the initiation of viral replication in cells and points the way to the abrogation of the latent reservoir of HIV.

These studies also led me to define how eukaryotic transcriptional enhancers function. Earlier studies defined enhancers as being dependent on promoter elements, but independent of distance, position and orientation with respect to the start site of transcription. Moreover, enhancers require CTD to function. We found that all these effects occurred via P-TEFb, i.e. activators on enhancers recruit P-TEFb, which binds and phosphorylates CTD from great distances, leading to the elongation of transcription. Thus, promoters recruit the preinitiation complex (PIC) and enhancers modify RNAPII for elongation. Other examples of proteins that recruit P-TEFb to enhancers are cMyc, b-catenin, and CIITA.

So, the simple view of transcription assembles PIC on promoters, activates RNAPII by enhancers and disrupts this switch by silencers. There is another way to block the elongation of transcription and that is by a potent, noncompetitive inhibitor of Cdk9. We were very fortunate to discover flavopiridol,which is used to treat cancer. Flavopiridol inhibits the progression through the cell cycle, but not by inhibiting Cdk4 and Cdk6. Rather, flavopiridol blocks the transcription of cyclin D1, which is the cyclin partner of these kinases. It blocks Cdk9 noncompetitively, as well as Tat transactivation and HIV replication at low nanomolar concentrations. Higher concentrations of flavopiridol also inhibit NF-kB and CIITA, thus blunting inflammation and autoimmunity. Morever, as flavopiridol targets a cellular protein, the virus cannot mutate to escape its effects.

Studies on the viral accessory protein Nef revealed steps in the HIV replicative cycle that deal with the production and budding of new virions. We found that Nef binds to the phosphoinositol 3-kinase (PI-3K) signalosome, which is composed of PI-3K, the guanine nucleotide exchange factor Vav, small GTPases Cdc42/Rac1 and p21 activated kinase PAK. This activation results in cytoskeletal rearrangements and downstream effector functions. They provide a better milieu for the production of HIV as well as aggregate lipid rafts from which new virions bud. Indeed, HIV produced in the presence of Nef contains more lipid and is more infectious.

The other interaction is with the catalytic subunit of the unversal proton pump, the vacuolar membrane ATPase (V-ATPase). This binding is important for the proper trafficking and processing of viral structural proteins. In addition, we believe that V-ATPase is incorporated into virions, where the acid pH is required for optimal uncoating in the target cell. NIH has supported studies on Nef from SIV.

Finally, we study BLS, a congenital SCID. We defined not only an interesting variant of this disease but also lesions in several transcription factors that lead to this clinical picture. For example, four genes that are defective in BLS were characterized in detail. The fifth gene affects chromatin remodeling and is being identified. Additionally, we were able to create a super CIITA that activates constitutively antigen processing and presentation. This master regulator is key for my study of the immune response! Moreover, we can turn on and off immune responses at will. This technology led to a new model of human rheumatoid arthritis in the mouse. Similar strategies should reproduce any human autoimmunity in the rodent! They should also prove invaluable for the immunotherapy of cancer and the creation of better adjuvants for vaccination.