We have studied the abilities of different transactivation domains to stimulate the initiation and elongation (postinitiation) steps of RNA polymerase II transcription in vivo. Nuclear run-on and RNase protection analyses revealed three classes of activation domains: Sp1 and CTF stimulated initiation (type I); human immunodeficiency virus type 1 Tat fused to a DNA binding domain stimulated predominantly elongation (type IIA); and VP16, p53, and E2F1 stimulated both initiation and elongation (type IIB). A quadruple point mutation of VP16 converted it from a type IIB to a type I activator. Type I and type IIA activators synergized with one another but not with type IIB activators. This observation implies that synergy can result from the concerted action of factors stimulating two different steps in transcription: initiation and elongation. The functional differences between activators may be explained by the different contacts they make with general transcription factors. In support of this idea, we found a correlation between the abilities of activators, including Tat, to stimulate elongation and their abilities to bind TFIIH.
The human host cell factor (HCF) is expressed in a variety of adult and fetal tissues, and its gene is conserved in animals as diverse as mammals and insects. However, its only known function is to stabilize the herpes simplex virus virion transactivator VP16 in a complex with the cellular POU domain protein Oct-1 and cis-acting regulatory elements in promoters of immediate-early viral genes. To identify a cellular function for HCF, we used the yeast two-hybrid system to identify a cellular ligand for HCF. This protein, Luman, appears to be a cyclic AMP response element (CRE)-binding protein/activating transcription factor 1 protein of the basic leucine zipper superfamily. It binds CREs in vitro and activates CRE-containing promoters when transfected into COS7 cells. This activation of transcription was synergistically enhanced by the presence of CCAAT/enhancer-binding protein elements and inhibited by AP-1 elements in the promoter. In addition to a basic DNA binding domain, Luman possesses an unusually long leucine zipper and an acidic amino-terminal activation domain. These features in Luman are also present in what appear to be homologs in the mouse, Drosophila melanogaster, and Caenorhabditis elegans. Luman and VP16 appear to have similar mechanisms for binding HCF, as in vitro each competitively inhibited the binding of the other to HCF. In transfected cells, however, while VP16 strongly inhibited the ability of GAL-Luman to activate transcription from a GAL4 upstream activation sequence-containing promoter, Luman was unable to inhibit the activity of GAL-VP16. Luman appears to be a ubiquitous transcription factor, and its mRNA was detected in all human adult and fetal tissues examined. The possible role of HCF in regulating the function of this ubiquitous transcription factor is discussed.
Modification of proteins by small ubiquitin-like modifier (SUMO) plays an important role in the function, compartmentalization, and stability of target proteins, contributing to the regulation of diverse processes. SUMO-1 modification can be regulated not only at the level of conjugation; it may also be reversed by a class of proteases known as the SUMO-specific proteases. However, current understanding of the regulation, specificity, and function of these proteases remains limited. In this study, we characterize aspects of the compartmentalization and proteolytic activity of the mammalian SUMO-specific protease, SENP1, providing insight into its function and regulation. We demonstrate the presence of a single nonconsensus nuclear localization signal within the N terminus of the protein, the mutation of which results in pronounced cytoplasmic accumulation in contrast to the nuclear accumulation of the parental protein. In addition, we observe that the N terminus of the protein may be essential for the correct regulation of the protease, since expression of the core domain alone results in limited expression and loss of SUMO-1, indicative of constitutive catalytic activity. Consistent with the prediction that the protease is a member of the cysteine family of proteases, we mutated a key cysteine residue and observed that expression of this catalytic mutant had a dominant negative phenotype, resulting in the accumulation of high molecular weight SUMO-1 conjugates. Furthermore, we demonstrate that SENP1 may itself be a target for SUMO-1 modification occurring at a nonconsensus site. Finally, we demonstrate that SENP1 localization is influenced by expression and localization of SUMO-1-conjugated target proteins within the cell.