Mechanistic Investigations of Transcriptional Activator Function for the Design of Synthetic Replacements.

Abstract

Transcriptional activators play a critical role in regulating gene expression, precisely controlling the transcriptional response of their cognate genes in a signal-responsive fashion. Transcriptional activation occurs when the activator localizes to a specific DNA sequence and facilitates the assembly of the transcriptional machinery (RNA polymerase II and associated transcription factors) at a gene. Malfunctioning transcriptional activators are associated with a variety of human diseases; greater than 50% of all cancers, for example, are associated with mutations in the transcriptional activator p53. Thus, the development of activator artificial transcriptional factors (ATFs) that functionally replicate their natural counterparts is emerging as a potential therapeutic strategy. One of the biggest hurdles to this goal is the lack of knowledge about the binding interactions utilized by natural activators to up-regulate gene expression. The major goal of this work is to delineate the features of activator binding interactions in order to develop useful activator ATFs. Natural transcriptional activators exhibit a multi-partner binding profile in vitro and there is also evidence of this in vivo, although the identities of the binding partners are unknown. To investigate the feasibility that a single binding interaction could lead to transcription function, peptide ligands for the postulated activator target Med15(Gal11) were identified through a binding screen. Satisfyingly, localizing these ligands to a promoter in S. cerevisiae results in transcription activation that is Med15(Gal11)-specific. Activator ATFs constructed with these ligands were not, however, as active as natural activators. Activator ATFs with enhanced function could be created by incorporating ligands that were able to interact with more than one partner. Ligands that interact with both Med15(Gal11) and the SAGA component Tra1 upregulated transcription to higher levels than those targeting Med15(Gal11) alone. In addition, the incorporation of a masking interaction into the activator ATFs led to a profoundly positive impact on function. Finally, towards identifying the functionally relevant binding partners of transcriptional activators, a nonsense suppression strategy was adapted for incorporation of photoactivatable, crosslinking amino acids into natural transcriptional activators in S. cerevisiae. Crosslinking experiments with Gal4 thus modified revealed at least three binding partners that will be the subject of further study.