Science Spotlight

Mapping essential interfaces in Transcription Factor H

Model of budding yeast TFIIH.
Image provided by Dr. Hahn.

The information stored in genes must be "read" and then translated into proteins that carry out the essential functions of the cell.  During the process of transcription, transcription factors recognize genes in the DNA and then recruit RNA polymerase to make a copy of the gene, generating a messenger RNA (mRNA).  mRNAs are then taken to protein factories called ribosomes where they are translated into the proteins they encode.  The transcription factor complex TFIIH is one of several general transcription factor complexes that, together with RNA polymerase II, form what is called the pre-initiation complex (PIC).   The PIC forms at the transcription start site (TSS) where it unwinds and positions the DNA in the RNA polymerase II active site to begin the process of transcription.  TFIIH, despite its essential function, has been the least understood of the RNA polymerase II-associated transcription factors.  Previously, researchers in the Hahn Laboratory (Basic Sciences Division) and the Ranish Laboratory at Institute for Systems Biology (ISB, Seattle) built a model of the structure of human and yeast TFIIH by a combination of protein cross-linking-mass spectrometry, modeling, and electron microscopy. In their recent publication in Molecular and Cellular Biology, scientists in the Hahn Laboratory determined which protein interfaces were essential for viability in budding yeast and uncovered unexpected functions of several subunits.

To test whether specific regions of proteins are required for the essential function of TFIIH in transcription initiation, the scientists made use of the "awesome power of yeast genetics" by generating different truncation and deletion mutants of TFIIH components and performing a plasmid shuffle assay.  The purpose of the assay is to determine whether a mutant version of an essential gene can function as the only copy of the gene in the cell.  If it disrupts the essential function of the gene, the yeast cells will die in certain growth media.  In addition to the genetic assay, scientists in the Hahn Lab immunoprecipitated each mutant protein and examined whether other TFIIH components co-purified, to determine if formation of the complex was disrupted by the mutations. 
The Tfb2 "Hub" domain binds Tfb5 and Ssl2. This study showed that deletion of the interacting domains is lethal and the mutant proteins are unable to make the expected interactions. Surprisingly, deletions of the HEAT domains in Tfb2 were also lethal and immunoprecipitation of these Tfb2 mutant proteins revealed that the co-immunoprecipitation (co-IP) of Tfb3 and Rad3 was lost. No direct interactions have previously been detected between Tfb2 and Tfb3 however this study found that deleting several different regions of Tfb2 resulted in the loss of interaction with Tfb3. 
Another surprising observation was that deletion of the Tfb3 "Latch" region did not cause a growth defect. This region interacts with many other TFIIH components based on the structural data from a previous study.  Despite the ability of cells to survive without this region, deleting parts of the Latch disrupted the association of Tfb3 with all other subunits of TFIIH except for the kinase Kin28. These results suggest that tight association of the Latch region of Tfb3 with the rest of TFIIH besides Kin28 is not required for TFIIH function in cells. In contrast to this, deletion of the Ring domain in the N-terminus of Tfb3 caused cell death and also disrupted TFIIH interactions except for binding to Kin28. This bolsters previous studies that have found that the Ring domain is important for function.
Rad3 is a helicase, an enzyme that unwinds DNA, and it contains the Arch and FeS domains, which are proposed to form a tunnel that binds single-stranded DNA. To map which specific amino acid residues in this protein contact residues in other proteins in TFIIH within cells, the researchers used a site-specific in vivo crosslinking strategy.  Based on known structural data, they replaced selected surface-exposed residues of Rad3 with artificial amino acids which could be crosslinked to other residues upon exposure to UV light. This uncovered that Rad3 does indeed bind Tfb3 in cells but also revealed that there appears to be more than one binding interface. Tfb3 crosslinked to opposite faces of the Arch in Rad3. 

Dr. Hahn noted that the findings from their study "will be a valuable guide for future structural studies.  Rather than trying to obtain high resolution structures of the complete TFIIH complex, TFIIH subdomains defined by our work could be used. These structures could then be pieced together to give a picture for the complete TFIIH complex." 

Additionally, this study highlights how hypotheses generated from in vitro data need to be confirmed by in vivo studies. More often than not, the in vivo results surprise and challenge scientists. 

Warfield L, Luo J, Ranish J, Hahn S.  2016.  "Function of conserved topological regions within the S. cerevisiae basal transcription factor TFIIH." Molecular and Cellular Biology.  [Epub ahead of print]

This research was supported by the National Institutes of Health.