The Yin and Yang of P-TEFb regulation: implications for human immunodeficiency virus gene expression and global control of cell growth and differentiation

Abstract

The positive transcription elongation factor b (P-TEFb) stimulates transcriptional elongation by phosphorylating the carboxy-terminal domain of RNA polymerase II and antagonizing the effects of negative elongation factors. Not only is P-TEFb essential for transcription of the vast majority of cellular genes, but it is also a critical host cellular cofactor for the expression of the human immunodeficiency virus (HIV) type 1 genome. Given its important role in globally affecting transcription, P-TEFb's activity is dynamically controlled by both positive and negative regulators in order to achieve a functional equilibrium in sync with the overall transcriptional demand as well as the proliferative state of cells. Notably, this equilibrium can be shifted toward either the active or inactive state in response to diverse physiological stimuli that can ultimately affect the cellular decision between growth and differentiation. In this review, we examine the mechanisms by which the recently identified positive (the bromodomain protein Brd4) and negative (the noncoding 7SK small nuclear RNA and the HEXIM1 protein) regulators of P-TEFb affect the P-TEFb-dependent transcriptional elongation. We also discuss the consequences of perturbations of the dynamic associations of these regulators with P-TEFb in relation to the pathogenesis and progression of several major human diseases, such as cardiac hypertrophy, breast cancer, and HIV infection.

FIG. 1.

P-TEFb phosphorylates the Pol II CTD and negative elongation factors to stimulate processive elongation. At the start of the transcription cycle, Pol II with the hypophosphorylated CTD is assembled into the preinitiation complex (PIC) at the promoter. To facilitate promoter clearance and jump-start initiation, Ser5 residues of the CTD heptapeptide repeats are phosphorylated by CDK7 (a component of the TFIIH complex). However, shortly after initiation, the progression of Pol II is stalled by the concerted actions of two negative elongation factors, DSIF and NELF. This checkpoint facilitates the recruitment of capping enzymes to ensure proper capping of the nascent pre-mRNA. To overcome this checkpoint, P-TEFb, which is recruited by Brd4 to transcription template in the vicinity of the stalled Pol II, phosphorylates DISF, NELF, and the CTD repeats at the Ser2 positions. These phosphorylation events promote the dissociation of NELF and convert DSIF into a positive elongation factor, thereby allowing the Pol II to engage in productive elongation to produce full-length transcripts.

FIG. 2.

P-TEFb is essential for Tat transactivation of HIV-1 transcription. Shortly after transcription is initiated from the HIV-1 promoter, the progression of Pol II is stalled by the concerted actions of negative elongation factors DSIF and NELF. For Pol II to escape from this promoter-proximal pausing, the HIV-1-encoded Tat protein binds to host cellular P-TEFb and recruits it to the stalled Pol II through forming a stable ternary complex involving the TAR RNA stem-loop structure located near the 5′ end of the nascent viral transcript. Subsequently, P-TEFb phosphorylates the Pol II CTD as well as the negative elongation factors to stimulate processive elongation.

FIG. 3.

Domain structure of HEXIM1. The N-terminal domain of HEXIM1 functions as a self-inhibiting regulatory domain. The centrally located arginine-rich motif (ARM) that largely overlaps with the nuclear localization signal (NLS) also serves as the 7SK RNA-binding domain. The 7SK-binding domain in HEXIM1 has a bipartite structure, with the first half showing a near-perfect match with the TAR RNA-binding motif found in HIV-1 Tat. The P-TEFb-binding/inhibiting function of HEXIM1 resides in the HEXIM1 C-terminal domain. Within this domain, the coiled-coil region mediates the dimerization of HEXIM1, whereas a region enriched in acidic residues is proposed to interact with the adjacent basic residues in the ARM/NLS domain, thus preventing the premature binding of HEXIM1 to P-TEFb in the absence of the 7SK snRNA (2). All numbers refer to amino acid positions in HEXIM1.

FIG. 4.

Architectural resemblance between the Tat-TAR-P-TEFb and HEXIM1-7SK-P-TEFb ribonucleoprotein complexes. (A) Comparison of nucleotide sequences between the G302-C324 apical region of the 3′ hairpin of 7SK RNA and the bulge-loop region of the HIV-1 TAR RNA (positions +20 to +42). The conserved nucleotides are shaded in red. Reprinted from reference 22 with permission.) (B) The Tat-TAR-P-TEFb and HEXIM1-7SK-P-TEFb complexes display many similar features and may share a common evolutionary origin. In addition to the sequence similarity between the HIV-1 TAR RNA and the 3′ hairpin of the 7SK RNA, both sequences also make direct contacts with CycT1. Moreover, both Tat and HEXIM1 directly interact with P-TEFb through binding to a region near the cyclin box in CycT1. Lastly, both HEXIM1 and Tat utilize a highly homologous and functionally interchangeable arginine-rich motif (ARM) for binding to their respective RNA partners. Given the multiple similarities shared between the two P-TEFb-containing ribonucleoprotein complexes, Tat has been shown to efficiently compete with HEXIM1 for binding to the same P-TEFb molecule. Tat can also disrupt the 7SK snRNP in an effort to convert the inactive P-TEFb into the active Tat/TAR-bound form for stimulating HIV-1 transcription. The secondary structure of 7SK is a modified version based on two previous reports (77, 92). For simplicity, only one copy each of HEXIM1, CDK9, and CycT1 is shown, although two copies of each are believed to interact with one 7SK molecule to form one HEXIM1-7SK-P-TEFb snRNP.

FIG. 5.

P-TEFb is maintained in a functional equilibrium by dynamic associations with its positive and negative regulators. In the nucleus, a major portion of P-TEFb is sequestered into the 7SK/HEXIM1 snRNP, where P-TEFb's kinase activity is inhibited by HEXIM1 in a 7SK-dependent manner. Because P-TEFb within the 7SK snRNP is unable to phosphorylate the Pol II CTD or associate with promoters, Pol II goes into the abortive elongation mode. Treatment of HeLa cells with stress-inducing agents or cardiac myocytes with hypertrophic signals can cause a rapid disruption of the 7SK snRNP and quantitative conversion of the released P-TEFb into the Brd4-bound form. This results in the increased recruitment of P-TEFb by Brd4 to transcriptional templates and stimulation of productive elongation by Pol II. On the other hand, when murine erythroleukemia cells are induced to differentiate by the treatment with HMBA, the P-TEFb equilibrium is shifted to the inactive, HEXIM1/7SK-bound state. Thus, the dynamic associations of P-TEFb with its positive and negative regulators are kept under tight cellular control in response to ever-changing transcriptional demand in the cell. Since HEXIM1 is known to display an antigrowth effect in a number of cell types, whereas Brd4 is progrowth during mouse development, their targeting of the general transcription factor P-TEFb is expected to affect the global control of cell growth and differentiation. For simplicity, only a monomer each of P-TEFb and HEXIM1 is depicted in the 7SK snRNP.