The factors involved in the accurate transcription of eukaryotic protein-coding genes can be classified into two groups. First, general transcription factors (GTFs) are necessary and can be sufficient for accurate transcription initiation in vitro. The GTF TFIID is composed of the TATA-box–binding protein (TBP) and approximately 14 TBP-associated factors (TAFs). We found, unexpectedly, that human embryonic stem cells (hESCs) contain only six TAFs. Altering the composition of hESC TAFs resulted in misregulated expression of pluripotency genes and induction of differentiation. Thus, the selective expression and use of TAFs underlies the ability of hESCs to self-renew.
Transcriptional activity is greatly stimulated by the second class of factors, promoter-specific activators. Activators are thought to work, in part, by contacting components of the general transcription machinery, although the identification of direct targets of activators has been a major challenge. Work from our lab has suggested that the Tra1 subunit of the SAGA (Spt-Ada-Gcn5-acetyltransferase) complex is the target of the model yeast activator Gal4. To provide support for this model, we derived and characterized mutant tra1 alleles that are selectively defective for interaction with Gal4. These Tra1 mutants are not recruited by Gal4 to the promoter and cannot support Gal4-directed transcription, confirming the essentiality of the Gal4-Tra1 interaction.
We have also been using genome-wide RNA interference (RNAi) screens to identify factors and pathways involved in transcriptional regulation of tumor suppressor genes (TSGs) and other cancer-related proteins. For example, we used RNAi screening to identify 28 cofactors through which a RAS oncoprotein directs transcriptional silencing of TSGs. Using RNAi-based epistasis analysis we ordered these factors into a pathway that is initiated by the sequence-specific DNA-binding protein ZFP354B and culminates in recruitment of the DNA methyltransferase DNMT1. In another screen, we found that transcriptional silencing of the TSG RASSF1A requires the homeobox protein HOXB3. HOXB3 binds to the DNA methyltransferase DNMT3B gene and increases its expression. DNMT3B, in turn, is recruited to the RASSF1A promoter, resulting in RASSF1A silencing.
Activating transcription factor 5 (ATF5), an anti-apoptotic transcription factor originally identified in our laboratory, is highly expressed in several cancers, particularly malignant glioma, where it promotes cell survival. Through an RNAi screen, we identified transcriptional regulators of ATF5, which revealed an essential survival pathway in malignant glioma that is initiated by RAS/mitogen-activated protein kinase (MAPK) signaling. Accordingly, the RAF inhibitor sorafenib suppressed ATF5 expression in glioma stem cells and inhibited malignant glioma growth in cell culture and mouse xenografts. Our results provided the rationale for an early-phase clinical trial combining sorafenib with irradiation.
Eukaryotic gene expression is also regulated at the level of pre-mRNA splicing. We have a long-standing interest in the mechanisms involved in pre-mRNA splicing, with an emphasis on the role of splicing factors that act early during spliceosome assembly. U2AF35, a splicing factor originally identified in our laboratory, functions by binding to the 3' splice site and initiating spliceosome assembly. Recently, cancer genome-sequencing studies have identified recurrent mutations in U2AF35 in several hematopoietic malignancies and lung cancer. We are investigating the basis by which oncogenic U2AF35 promotes transformation.
Another factor of interest is Urp, a U2AF35-related protein previously implicated in splicing of the major class of U2-type introns. We found that Urp is also required for splicing of the minor class of U12-type introns. We demonstrated that Urp facilitates distinct steps of U2- and U12-type intron splicing through recognition of a common splicing element, the 3' splice site.
We are also performing RNAi screens to identify factors that are involved in various types of alternative splicing. One such screen has uncovered an unanticipated role for mRNA 3'-end formation factors in global regulation of alternative internal exon usage.
Cancer Molecular Biology
We are taking a variety of experimental approaches to address questions in cancer biology, identify genes that promote or prevent cancer, and delineate cancer-relevant regulatory pathways. The pathways and components revealed by these screens provide potential new targets for therapeutic intervention. For example, we have performed an RNAi-based synthetic interaction screen and identified a regulatory pathway that is preferentially important for proliferation of p53-deficient cancer cells and is thus a potential therapeutic target.
To discover new TSGs, we developed a functional genomics approach in which immortalized but nontumorigenic cells were stably transduced with large-scale, short hairpin RNA (shRNA) pools and tested for tumor formation in mice. Using this approach, we identified 24 human lung squamous cell carcinoma (hLSCC) TSGs. Increased fibroblast growth factor receptor (FGFR) signaling is a common aberration in hLSCC. Remarkably, we found that many of the TSGs encode repressors of FGFR signaling. Our results indicate that increased FGFR signaling promotes tumorigenesis in many hLSCCs that lack FGFR1 amplification or activating mutations.
Certain oncogenes that promote solid tumors, such as RAS and BRAF, can induce senescence in primary cells; subsequent inactivation of a TSG allows the cell to bypass senescence and form a tumor. We have found that leukemogenic fusion proteins, such as BCR-ABL, also can induce senescence in hematopoietic progenitors, suggesting that leukemias harbor genetic or epigenetic alterations that inactivate "leukemia suppressor genes." We are performing RNAi screens to identify such genes.
Some TSGs encode secreted proteins, which have substantial therapeutic potential because they can be readily delivered to block tumor growth. We have used both RNAi screening and bioinformatic approaches to identify secreted tumor suppressors. We are collaborating with Guangping Gao (University of Massachusetts Medical School) to show that adeno-associated virus (AAV)-mediated delivery of secreted tumor suppressors can inhibit tumor growth in xenografted mice.
Finally, we are also using functional genomic approaches to study the mechanism of resistance to anticancer drugs. For example, in collaboration with Daniel Bolon (University of Massachusetts Medical School), we developed a systematic saturation mutagenesis approach to search for second-site mutations within oncogenic BRAF, BRAF(V600E), that could confer resistance to the inhibitor vemurafenib. Using this method, we identified a novel vemurafenib-resistant BRAF(V600E) mutant, which arises by a single-nucleotide substitution mutation. We are also performing RNAi screens to elucidate mechanisms of resistance to imatinib, a BCR-ABL inhibitor, that do not arise by secondary mutations in BCR-ABL.
Grants from the National Institutes of Health provided partial support for these projects.
As of October 8, 2013