Structural studies of eukaryotic transcription

ABSTRACT
Cellular differentiation, development and homeostasis depend on regulation of gene expression, which is
largely focused on the DNA transcription initiation process. During transcription initiation, Mediator, a large
multi-protein complex conserved throughout eukaryotes, conveys regulatory signals to RNA polymerase II
(RNAPII), the enzyme responsible for transcription of all protein-coding genes. Mediator includes 25-30
different polypeptides (depending on the specific organism) organized into Head, Middle and Tail modules,
plus a dissociable kinase module (CKM) that includes the Cdk8 kinase, the only catalytically-active Mediator
subunit. Mediator conformational rearrangements that stabilize preinitiation complex (PIC) components have
explained the effect of Mediator on basal transcription. However, conformational rearrangements alone are
insufficient to explain the response of Mediator to transcription factors (TFs) that enables transcription
activation and repression. Here we propose biochemical, functional and cryo-EM studies of mammalian
Mediator (mMED) that build on our previous work and explore the significance of mMED’s antagonistic
interaction with the CKM and with MED26, a metazoan-specific, dissociable mMED subunit closely linked to
modulation of mMED–RNAPII interaction. The CKM and MED26 interact with Mediator around a Head-Middle
module interface (the CTD-binding gap) where RNAPII interaction is initiated by binding of the carboxy-
terminal domain of the largest RNAPII subunit (the CTD). CKM-bound (CKM-mMED) and MED26-bound
(MED26-mMED) forms of mMED were independently identified by various research groups shortly after
Mediator’s discovery and we propose to test a mMED activation mechanism based on CKM-mMED to MED26-
mMED interconversion that we hypothesize controls mMED interaction with RNAPII and PIC formation.
In Aim1 we will Investigate the connection between CKM – MED26 antagonism and Mediator activation.
We posit that the crux of the mMED activation mechanism is control of the initial CTD-dependent mMED–
RNAPII interaction by antagonistic effects of the CKM (limits RNAPII interaction) and MED26 (required for
RNAPII interaction) at the CTD-binding gap. We will use in vitro and in vivo approaches including biochemical,
functional and genomic analyses to understand modulation of mMED interaction with RNAPII and its effects on
transcription initiation. These studies will test a proposed activation mechanism that would explain the
significance of mMED subpopulations with opposite functional effects and test whether interconversion
between mMED forms can explain mMED activation.
In Aim 2 we will determine the structural underpinnings that enable regulation of Mediator-RNAPII
interaction. We postulate that TF targeting of CKM-mMED and subsequent effects on MED26 and CTD
interaction at the CTD-binding gap are enabled by mMED structural rearrangements or changes in mMED
conformational dynamics. Structural analysis of well-defined intermediate steps will reveal how mMED’s
structure enables activation. We will use cryo-EM to determine near-atomic resolution maps of various
intermediates and use state-of-the-art image analysis approaches to understand their conformational and
interaction dynamics. These studies will reveal structural factors that underpin primary (initial CTD-dependent
interaction with RNAPII) and subsequent (further modulation of RNAPII interaction and PIC assembly) aspects
of the mMED activation mechanism.
In Aim 3 we will investigate long-range structural rearrangements in mMED that enable mMED activation
by TF binding to Tail module subunits. We believe that changes in the composition, structure or
conformational dynamics of the Tail module triggered by interaction with TFs can be communicated along the
mMED structure to the CTD-binding gap, allowing mMED to respond to a variety of regulatory signals through
the same fundamental activation mechanism. We will use cryo-EM, image analysis and biochemistry to
understand how signals from TF binding to various mMED Tail subunits converge to control interaction of
mMED with RNAPII. The studies described in this aim will test the generality of the proposed mMED activation
mechanism and reveal how the mMED structure can integrate signals from a variety of TFs that target different
mMED subunits.
The studies proposed in this application are both conceptually and technically innovative. We will test a novel
model for regulation of activated transcription initiation based on interconversion between “repressed” and
“activated” forms of mMED by applying a multi-disciplinary approach combining state-of-the-art molecular
biology, genomics, bioinformatics and cryo-EM/image analysis techniques to understand early steps of
mammalian PIC assembly and transcription initiation. We will combine in vitro studies allowing examination of
individual aspects of the proposed mechanism with in vivo studies to verify biological significance and
functi