chap4.tex

\chapter{Biochemical and structural support for a general model of promoter binding by TFIID}

The structural studies presented in Chapters 2 \& 3 have provided the foundation for understanding the nature of TFIID's interaction with promoter DNA. However, in order to arrive at a complete description of TFIID bound to promoter DNA, biochemical and structural studies were pursued in an attempt to propose a model of TFIID-TFIIA-SCP complex formation. To facilitate the biochemical studies, we engaged in a wonderful collaboration with James T. Kadonaga's laboratory at University of California - San Diego, where George A. Kassavetis performed all footprinting experiments shown in this chapter. The synergy between structural studies of TFIID-TFIIA-SCP and biochemical footprinting have helped construct a detailed model describing potential mechanisms for regulating human TFIID's interaction with promoter DNA.\\
\indent This chapter will address potential mechanisms of promoter binding by TFIID through a variety of experimental designs. First, detailed footprinting analysis using DNase I and MPE-Fe were used to probe the extent of contacts that TFIID makes with promoter DNA in a TFIIA-dependent and -independent fashion. These footprinting results will be extended by testing the structural effect that core promoter mutations have on TFIID's structure. After developing a detailed model of TFIID-TFIIA-DNA, the canonical conformation will be studied to investigate its DNA binding ability. Finally, cryo-EM structural analysis and footprinting will establish that the rearranged conformation is the TFIIB-binding conformation of TFIID, suggesting that the rearranged state is the conformation that loads RNAPII at the TSS. These results have been synthesized to propose a model that addresses the interplay between TFIID's conformational landscape and DNA binding.\\
  
\section{Detailed footprinting of TFIID-TFIIA-SCP}
\begin{figure}
\centering
\includegraphics[width=0.9\textwidth]{../Ch4_figs/Fig4.1.eps}
\caption[TFIID exhibits TFIIA-independent and TFIIA-dependent interactions with SCP DNA]{TFIID exhibits TFIIA-independent and TFIIA–dependent interactions with SCP DNA. DNase I (A) and MPE-Fe (B) footprinting of TFIID-SCP and TFIID-TFIIA-SCP. (C) B-DNA model of the SCP with positions of core promoter elements (top), DNase I, and MPE-Fe protection patterns (-66 to +55 shown) color-coded for TFIID-SCP and TFIID-TFIIA-SCP (four bottom rows).  White base pairs indicate protection (also marked by black lines), blue surfaces indicate partial digestion and red surfaces indicate complete digestion by DNase I, and pink indicates digestion by MPE-Fe.}
\label{fig:Fig4.1}
\end{figure}
Cryo-EM structural analysis of the TFIID-TFIIA-SCP suggested that the rearranged state interacted with the downstream DNA core promoter elements MTE and DPE using lobe C, whereas the TATA box and possibly the Inr are bound by lobe A. To test and extend this model, we performed footprinting analyses of the TFIID-SCP complex in the presence or absence of TFIIA.\\
\indent DNase I footprinting of TFIID-SCP showed an extended region of protection from -7 to +41, which is in agreement with previous footprinting data on TFIID-SCP (Figure~\ref{fig:Fig4.1}A) \cite{Juven-Gershon_1249}. The addition of TFIIA did not alter the downstream interactions of TFIID with the Inr, MTE, and DPE core promoter elements, but did, however, result in a substantial increase in the binding of TFIID to the TATA box and flanking DNA sequences from -38 to -20 (Figures~\ref{fig:Fig4.1}A, \ref{fig:Fig4.2}A \& B). DNase I footprinting analysis of both DNA strands in the absence or presence of TFIIA revealed distinct patterns of protection and cleavage throughout the core promoter region. The results indicate that the DNA sequence between the Inr and MTE/DPE sites exhibits a phasing in DNase I sensitivity in which one face of the SCP DNA between Inr and MTE-DPE is susceptible to DNase I cleavage, whereas the opposite face remains protected (Figure~\ref{fig:Fig4.1}C). Furthermore, the accessible face of the DNA exhibits DNase I hypersensitivity upon binding of TFIID (Figure~\ref{fig:Fig4.1}A). \\
\begin{figure}
\centering
\includegraphics[width=1\textwidth]{../Ch4_figs/Fig4.2.eps}
\caption[Line profiles for DNase I footprinting gels of TFIID-SCP and TFIID-TFIIA-SCP]{Line profiles for DNase I footprinting gels of TFIID-SCP and TFIID-TFIIA-SCP. (A) Line profile traces from DNase I footprinting gels for 5’-label upstream (A) and downstream (B). Bold base pair numbers indicate hypersensitive sites. Line profile traces from MPE-Fe footprinting gels for 5’-label upstream (C) and 5’-label downstream (D).}
\label{fig:Fig4.2}
\end{figure}
\indent To obtain high resolution data on the interactions of TFIID with the core promoter, we carried out footprinting analyses with MPE-Fe, an intercalating agent that delivers Fe(II) for oxidation of the deoxyribose phosphate backbone of DNA and provides single bp resolution of protein-DNA contacts \cite{Hertzberg_3897,Papavassiliou_3156,Va_3928}. The MPE-Fe footprinting data on TFIID-SCP and TFIID-TFIIA-SCP revealed the extensive and continuous interaction of TFIID with DNA from the DPE through the Inr as well as the TFIIA-dependent protection of 8 bp (-31 to -24) of DNA encompassing the TATA box (Figures~\ref{fig:Fig4.1}B, \ref{fig:Fig4.2}C \& D). The strong stimulation of TFIID binding to the TATA box is consistent with the results of previous studies on the binding of TBP to DNA \cite{Geiger_2949,Kim_3416,Kim_3377,Nikolov_3177}. The region of DNase I protection observed on only one face of the helix between the MTE/DPE and Inr also shows continuous protection from MPE-Fe(II) cleavage on both DNA strands, which is likely due to the inhibition of the DNA unwinding, necessary for MPE intercalation, that would occur as a result of protein bound to one side of the helix \cite{Uchida_3659}. Thus, the DNase I and MPE-Fe footprinting results, summarized in Figure~\ref{fig:Fig4.1}C, provide insight into the TFIID-DNA contacts that complements the cryo-EM data and contributes to the placement of the core promoter DNA on the TFIID structure. \\
\begin{figure}
\centering
\includegraphics[width=1\textwidth]{../Ch4_figs/Fig4.4.eps}
\caption[Structural model of DNase I and MPE-Fe footprinting results within the ternary TFIID-TFIIA-SCP(-66) complex]{Structural model of DNase I and MPE-Fe footprinting results within the ternary TFIID-TFIIA-SCP(-66) complex. Promoter DNA for SCP(-66) docked into the TFIID-TFIIA-SCP(-66) map (shown in mesh). DNA models were taken from Figure~\ref{fig:Fig4.1}C for sequences from -66 to +45. Asterisk in (B) indicates DNase I hypersensitive site at +3. Black lines in (B) indicate regions of continuous protection along SCP helix. }
\label{fig:Fig4.4}
\end{figure}
\indent The footprinting data provide new structural insight into the model of DNA through TFIID-TFIIA-SCP. On each strand between the Inr and MTE, there is a distinct 10 bp phasing of DNase I hypersensitive sites at -2, +18, and +28 (with probe DNA 5’-labeled upstream) and at +4 and +13 (with probe DNA 5’-labeled downstream) (Figure~\ref{fig:Fig4.1}A).  Docking the DNase I footprinting model of protection (Figure~\ref{fig:Fig4.1}C) into the cryo-EM model of TFIID-TFIIA-SCP allows for one side of the DNA sequence to face away from the central cavity (DNase I sensitive sites) while the opposite side of the helix faces the inner cavity of TFIID (DNase I protected sites) (Figure~\ref{fig:Fig4.4}). Hence, these data suggest that TFIID presents the DNase I-accessible face of this region of the core promoter to the bulk solution for interactions with RNAPII and other factors involved in transcription initiation. \\
\indent As the DNA extends across the central channel towards lobe A, the footprinting data suggest that there are topological changes in the promoter DNA surrounding the TSS. For instance, the DNase I footprinting experiment shows a hypersensitive site at +3, indicating that the DNA has changed conformed for optimal cleavage by Dnase I only when bound to TFIID (Figures~\ref{fig:Fig4.1}A). Given that the strength of this hypersensitive site appears to correlate with the strength of transcription initiation from a given promoter \cite{Theisen_341}, it is interesting that this site is exposed within the central channel of TFIID, suggesting that TFIID-induced topological changes of DNA within the central channel may be relevant for later steps in transcription initiation. \\
\indent The modeled path of promoter DNA through TFIID-TFIIA-SCP suggested that, in addition to the kinking of the TATA box by TBP, there is an additional bend in promoter DNA (Figure~\ref{fig:Fig4.4}). While it is difficult to model the proposed DNA bend accurately at this resolution, the MPE-Fe footprinting indicate that the intervening DNA between TATA and Inr showed sensitivity to MPE-Fe, suggesting that the DNA helix is not distorted and there are not strong protein-DNA contacts. Therefore, while the DNA must bend through this part of the structure, it likely follows a gradual bending path as it enters into lobe A.\\

\section{TFIID interacts with diverse promoter architectures through the rearranged conformation}
\begin{figure}
\centering
\includegraphics[width=0.9\textwidth]{../Ch4_figs/Fig4.7.eps}
\caption[Core promoter architecture dictates TFIIA-dependent and TFIIA-independent interactions of TFIID with core promoter DNA]{Core promoter architecture dictates TFIIA-dependent and TFIIA-independent interactions of TFIID with core promoter DNA. DNase I footprinting on wild-type and mutant SCP DNA sequences. 5'-labeled downstream probes were analyzed for DNase I protection in the presence or absence of TFIIA for wild type, mutant TATA (mTATA), Inr (mInr), and MTE/DPE (mMTE/DPE).}
\label{fig:Fig4.7}
\end{figure}
\begin{figure}
\centering
\includegraphics[width=0.8\textwidth]{../Ch4_figs/Fig4.8.eps}
\caption[Line profiles for DNase I footprinting on mTATA promoter]{Line profiles for DNase I footprinting on mTATA promoter.  (A) Line profiles for TFIID (blue) and TFIID-TFIIA (red) on wild-type SCP.  (B) Line profiles for TFIID (blue) and TFIID-TFIIA (red) on mTATA promoter.}
\label{fig:Fig4.8}
\end{figure}
While the SCP DNA has served as an important tool for structurally dissecting the structure of TFIID-TFIIA-SCP, the presence of four consensus promoter motifs within the SCP represents a non-physiological arrangement, since most human promoters contain only one or two consensus motifs \cite{Juven-Gershon_468}. To explore the effect of core promoter architecture on TFIID-promoter interactions, DNase I footprinting experiments were performed with ‘mutant’ SCP DNA constructs in the presence or absence of TFIIA. Mutation of the TATA box within the SCP sequence (mTATA) resulted in a wild-type interaction with the promoter DNA from the Inr to the DPE, as seen previously (Figure~\ref{fig:Fig4.7}) \cite{Juven-Gershon_1249}. In addition, the inclusion of TFIIA resulted in a weak but detectable footprint over the mutant TATA box (Figure~\ref{fig:Fig4.8}). The strong resemblance between the the protection patterns of the downstream Inr-MTE/DPE region from mTATA and the wild-type SCP suggested that TFIID is bound to the mTATA sequence in the rearranged conformation. To test this hypothesis, we collected cryo-EM data and visualized a sample of TFIID-TFIIA-SCP(mTATA) (Figure~\ref{fig:Fig4.6}). This experiment revealed that TFIID binds to the mTATA promoter in a nearly identical conformation as that observed with the wild-type SCP sequence. Thus, the combined footprinting and EM data indicate that the rearranged state of TFIID serves as the predominant DNA binding conformation for the SCP and mTATA promoter architecture.\\
\indent The conformation of promoter-bound TFIID was further addressed by analysis of promoters that contain mutations in the Inr (mInr) or MTE/DPE (mMTE/DPE) promoter motifs (Figure~\ref{fig:Fig4.7}). TFIID did not interact appreciably with either the mInr promoter or the mMTE/DPE promoter in the absence of TFIIA, as seen previously \cite{Juven-Gershon_1249}. However, the addition of TFIIA resulted in strong binding of TFIID to the TATA box as well as to sequences from the Inr through the DPE regions. In the presence of TFIIA, the overall patterns of protection observed with the mInr and mMTE/DPE promoters are similar to those seen with the wild-type SCP. It thus appears likely that TFIID-TFIIA binds to the mInr and mMTE/DPE promoters in the rearranged conformation.  Hence, the footprinting and EM data both suggest that TFIID binds to the wild-type and mutant SCPs in the newly discovered rearranged conformation.\\
\begin{figure}
\centering
\includegraphics[width=0.8\textwidth]{../Ch4_figs/Fig4.6.eps}
\caption[TFIID-TFIIA interacts with SCP(mTATA) within the rearranged conformation]{TFIID-TFIIA interacts with SCP(mTATA) within the rearranged conformation.  (A) 2D reference-free class averages calculated from cryo-EM data of TFIID-TFIIA-SCP(mTATA) are shown alongside 2D reference-free class averages from TFIID-TFIIA-SCP. (B) 3D models of the rearranged conformation for TFIID-TFIIA-SCP and TFIID-TFIIA-SCP(mTATA) at 34\AA. }
\label{fig:Fig4.6}
\end{figure}
\indent With the 'wild-type' SCP as well as with the three 'mutant' (mTATA, mInr, mMTE/DPE) versions of the SCP, we observed that TFIIA stimulates the binding of TFIID to the TATA box region (Figure~\ref{fig:Fig4.7}). This effect is consistent with the well-established TFIIA-mediated enhancement of TBP binding to the TATA box \cite{Thomas_1201}. With the mTATA promoter, the primary interaction of TFIID with the DNA is via the Inr, MTE, and DPE motifs, and a weak stimulation by TFIIA of the binding of TFIID to the mutant TATA box region is also observed. With the mInr and mMTE/DPE promoters, it seems likely that TFIIA stimulates the binding of TBP to the TATA box and that the remainder of the TFIID complex then interacts with the Inr through the DPE region of the core promoter, irrespective of the presence of consensus Inr or MTE/DPE elements. These findings may be analogous to the previously observed stimulation of the binding of partially-purified TFIID to the downstream promoter region of the adenovirus major late promoter (which lacks MTE/DPE motifs) by the upstream stimulatory factor, USF \cite{Sawadogo_3840,Va_3783}. In this light, it is possible that other sequence-specific activators, as well as coactivators, may function in a related manner to stabilize TFIID on promoter DNA and thus promote the formation of the rearranged conformation. 


\section{DNA binding within the canonical conformation}

While the rearranged conformation is the predominant form of TFIID bound to promoter DNA, we next wanted to investigate if the canonical state is capable of interacting with promoter DNA. Specifically, we wanted to know if the canonical state was 'inhibited' for binding to the MTE/DPE, given the close proximity of lobe A and the MTE/DPE binding sites within the canonical conformation. This investigation into DNA binding by the canonical state is similar in analysis to the localization of TFIIA within lobe A (Figure~\ref{fig:Fig4.5}A), where 2D reference-free class averages of the canonical state from cryo-EM of TFIID-TFIIA-SCP(gold) were generated and analyzed. \\
\indent After collecting larger datasets for both +45 and TATA-Nanogold labeled samples, class averages of the canonical conformation for TFIID-TFIIA-SCP showed that the promoter DNA binds in a manner similar to the rearranged conformation. This conclusion comes from comparison of the low defocus thresholded average from +45 Nanogold (Figure~\ref{fig:Fig4.5}B, right) with the high defocus class averages (Figure~\ref{fig:Fig4.5}B, left). The presence of additional density extending away from lobe C in the high defocus average and the terminal localization of +45-Nanogold indicates that SCP DNA is bound. Furthermore, considering that the +45 is 15 bps away from lobe C within the rearranged conformation, the presence of a similar density extending away from lobe C in the canonical state suggests that the promoter DNA bound is in a near identical manner (Figure~\ref{fig:Fig4.5}B). Unfortunately, the high degree of flexibility of lobe A did not allow careful comparison of lobe A's position and DNA binding, preventing conclusions from being drawn regarding potential inhibition of MTE/DPE binding by lobe A. However, despite this limitation, analysis of the canonical conformation from TFIID-TFIIA-SCP(+45 gold) indicated that the canonical conformation is capable of binding MTE/DPE DNA sequences within the same configuration as the rearranged state.\\
\indent Since lobe C interacts with the MTE/DPE within the canonical state, we hypothesized that if lobe A engages the TATA box, it should occur in a position located away from lobe C. To test this, canonical class averages from TFIID-TFIIA-SCP(TATA gold) were analyzed for Nanogold localization (Figure~\ref{fig:Fig4.5}C). Unlike the stable binding of MTE/DPE to lobe C, the Nanogold signal for TATA gold was diffusely localized along the trajectory of lobe A's rearrangement. Furthermore, considering the small cluster of peaks at a position opposite the central channel from lobe C, DNA binding in the canonical state appears to position the TATA box for binding by TBP within lobe A of the rearranged conformation (Figure~\ref{fig:Fig4.5}C). \\
\indent Analysis of the highly flexible canonical state using Nanogold labels for TFIIA and DNA has revealed insight into the dynamic process of DNA binding by TFIID (Figure~\ref{fig:Fig4.5}).  While it is difficult to know the position of lobe A within these Nanogold labeled class averages, the Nanogold labeling approach has provided key localizations in an otherwise flexible conformation. Despite the flexibility of lobe A, these labeling data indicate that the canonical conformation is competent for binding to promoter DNA. Lobe C binds to the MTE/DPE sequence in a similar fashion as the rearranged conformation, suggesting that its intrinsic DNA binding activity is preserved within the canonical state. Additionally, as suggested earlier, lobe A appears to exist as a modular TATA binding component of TFIID, where the TATA gold label localized to locations along the path of lobe A's rearrangement.\\ 
\begin{figure}
\centering
\includegraphics[width=1\textwidth]{../Ch4_figs/Fig4.5.eps}
\caption[Lobe C interacts with MTE/DPE motifs within the canonical conformation]{Lobe C interacts with MTE/DPE motifs within the canonical conformation. 2D reference-free class averages for the canonical conformation from cryo-EM data of TFIID-TFIIA-SCP with Nanogold labeled TFIIA (A), +45 (B), and TATA (C). Left of the averages is a model of the canonical state indicating location of Nanogold label. (D) Model of DNA path and TFIIA within the canonical conformation based upon gold labeling.}
\label{fig:Fig4.5}
\end{figure}

\section{TFIID-TFIIA-TFIIB-SCP adopts the rearranged conformation on promoter DNA}

The cryo-EM analysis presented thus far indicates that the predominant DNA binding form of TFIID is the rearranged conformation, suggesting that the remaining GTFs may bind the rearranged state for RNAPII loading. To test this hypothesis, cryo-EM samples were prepared with TFIID-TFIIA-TFIIB-SCP(-66) and analyzed (Figure~\ref{fig:Fig4.9}A). 2D reference-free class averages showed that the there were a variety of orientations containing DNA bound to the rearranged conformation (Figure~\ref{fig:Fig4.9}B), suggesting that the rearranged conformation is recognized and bound by TFIIB. \\
\begin{figure}
\centering
\includegraphics[width=1\textwidth]{../Ch4_figs/Fig4.9.eps}
\caption[Cryo-EM of TFIID-TFIIA-TFIIB-SCP(-66)]{Cryo-EM of TFIID-TFIIA-TFIIB-SCP(-66). Representative micrograph (A) and 2D reference-free class averages (B).  Scale bar is 200 nm in (A) and 200\AA\ in (B).}
\label{fig:Fig4.9}
\end{figure}
\indent To verify the binding of TFIIB to the rearranged state of TFIID-TFIIA-SCP, DNase I and MPE-Fe footprinting experiments were performed on TFIID-TFIIA-TFIIB-SCP. The footprinting results show that TFIIB extends the footprint surrounding the TATA, while the downstream contacts along the Inr, MTE, and DPE remain unchanged (Figure~\ref{fig:Fig4.11}A \& B). The DNase I patterns of protection indicate that an additional 8 - 10 bps are protected on either side of the TATA box, increasing the footprint around the TATA box from -45 to -18 (Figure~\ref{fig:Fig4.11}A). This shows that nearly 30 bps of promoter DNA upstream of the TSS are sequestered within the TFIID-TFIIA-TFIIB-SCP complex. Furthermore, given that the remaining downstream contacts with promoter DNA are unaffected by TFIIB, these data suggest that TFIIB preferentially interacts with the rearranged conformation. Considering that the TATA box was localized to lobe A, this result indicates that TFIIB localizes to lobe A within the cryo-EM structure of TFIID-TFIIA-TFIIB-SCP.\\
\begin{figure}
\centering
\includegraphics[width=.7\textwidth]{../Ch4_figs/Fig4.11.eps}
\caption[TFIIB interacts with TFIID-TFIIA-SCP within the rearranged conformation on promoter DNA]{TFIIB interacts with TFIID-TFIIA-SCP within the rearranged conformation on promoter DNA. DNase I (A) and MPE-Fe (B) footprinting of TFIID-TFIIA-TFIIB-SCP on 5'-labeled downstream DNA. }
\label{fig:Fig4.11}
\end{figure}
\indent In addition to the DNase I footprinting experiments, MPE-Fe was used in order to provide high-resolution information on the protein-DNA contacts introduced by TFIIB within the TFIID-TFIIA-TFIIB-SCP complex. This analysis revealed that TFIIB makes intimate contacts with the flanking DNA sequences around the TATA box from -34 to -19 (Figure~\ref{fig:Fig4.11}B). These results are consistent with the presence of a downstream BRE motif within the SCP sequence \cite{Juven-Gershon_1249}, a region of the promoter that is capable of engaging in sequence-specific contacts with TFIIB \cite{Deng_2005,Tsai_2000}. Interestingly, however, there are upstream contacts from TATA box, even though the SCP sequence does not contain a functional upstream BRE motif \cite{Juven-Gershon_1249}. Thus, TFIIB makes specific contacts with the upstream and downstream DNA sequences surrounding the TATA box, potentially stabilizing the TFIID-TFIIA-TFIIB-SCP complex due to an extended footprint on the DNA. \\
\begin{figure}
\centering
\includegraphics[width=.7\textwidth]{../Ch4_figs/4.12.eps}
\caption[TFIIB does not induce strong TATA box protection within TFIID-TFIIA-TFIIB-SCP(mTATA)]{TFIIB does not induce strong TATA box protection within TFIID-TFIIA-TFIIB-SCP(mTATA). DNase I footprinting of TFIID-TFIIA-TFIIB on SCP(mTATA). \emph{W} and \emph{M} indicate wild-type SCP and SCP(mTATA) promoters, respectively. }
\label{fig:Fig4.12}
\end{figure}
\indent From the high affinity binding of TFIID-TFIIA-TFIIB-SCP to the TATA box and flanking sequences, we hypothesized that the this ternary complex would make stronger contacts with the SCP(mTATA) sequences. To test this, we performed DNase I footprinting of TFIID-TFIIA-TFIIB on wild-type and mutant TATA box promoters (Figure~\ref{fig:Fig4.12}). The footprinting results revealed that mutation of the TATA box disrupts protein-DNA contacts along the TATA box and downstream BRE motifs in addition to the contacts upstream of the TATA box. Since TFIIB is unable to bind downstream BRE motifs in the absence TBP-TATA contacts (Figure~\ref{fig:Fig4.11}A), the footprinting on SCP(mTATA) suggests a cooperative activity of TFIIB-TBP-TATA in order for high affinity binding to the TATA box and downstream BRE motifs. This explanation is consistent with the crystal structure of TBP-TFIIB-TATA, where the TFIIB simultaneously contacts TBP and the downstream BRE \cite{Tsai_2000}. It should be noted, however, that there is slight protection of the TATA sequence, comparable to that previously observed (Figure~\ref{fig:Fig4.8}), indicating that TBP may be within close proximity of the mutated TATA sequence. These footprinting data suggest that, within the context of a mutant TATA box, TFIIB does not stimulate stronger TATA box binding nor does it interact with the downstream BRE element.\\
\indent The data above indicate that TFIIB binds the rearranged conformation in the presence of a functional TATA box and, given the extended footprinting surrounding the TATA box, the binding of TFIIB to TFIID-TFIIA-SCP may stabilize the ternary complex to enable high resolution structural studies using cryo-EM. Therefore, 3D refinements were performed on a sample of TFIID-TFIIA-TFIIB-SCP(-66) using previously obtained models corresponding to the canonical and rearranged conformations. After collecting and analyzing 56,000 cryo-EM particles of TFIID-TFIIA-TFIIB-SCP, the resulting 3D refined model for the rearranged state achieved a resolution of 36\AA, comparable to that obtained previously for TFIID-TFIIA-SCP. To investigate if the binding of TFIIB to the ternary complex altered structural contacts between components within TFIID-TFIIA-SCP, the structure is shown alongside the previously obtained TFIID-TFIIA-SCP(-66) (Figure~\ref{fig:IIB_cryo}). The overall structural features of these two models are nearly the same at this resolution, indicating that TFIIB does not introduce global changes in the structure of TFIID-TFIIA-SCP upon binding. \\
\indent Despite these overall similarities, there are subtle changes in the DNA density within the TFIID-TFIIA-TFIIB-SCP(-66) structure. First, serving as a positive control, this sample was prepared with SCP(-66) to test the presence and location of the upstream DNA density. Since the initial 3D model used for refinement did \emph{not} contain density corresponding to the upstream DNA, the presence of additional density exiting lobe A in a near-identical location suggests that position of upstream DNA does not change relative to the TATA box within TFIID-TFIIA-TFIIB-SCP(-66) (Figure~\ref{fig:IIB_cryo}). Interestingly, however, the shape of the DNA density appears to be narrower than the previously determined structure. This may be due to the increased rigidity of upstream DNA due to the TFIIB-DNA contacts surrounding the TATA box (Figure~\ref{fig:Fig4.11}A). Unfortunately, even though there were changes in the upstream DNA, we were unable to detect any differences within lobe A that was consistent in size with TFIIB (30 kDa). For clarity, the previously proposed atomic model of promoter DNA, TBP, TFIIA, and TFIIB were docked into the TFIID-TFIIA-TFIIB-SCP(-66) structure to show the proposed TFIIB binding site within lobe A. While TFIIB likely interacts with the rearranged conformation, the low resolution of the structure prevents further conclusions to be drawn regarding structural consequences of TFIIB binding. \\
\begin{figure}
\centering
\includegraphics[width=.8\textwidth]{../Ch4_figs/IID_IIA_IIB_SCP_fig.eps}
\caption[3D reconstruction of the rearranged conformation of TFIID-TFIIA-TFIIB-SCP(-66) reveals similar topology as TFIID-TFIIA-SCP(-66)]{3D reconstruction of the rearranged conformation of TFIID-TFIIA-TFIIB-SCP(-66) reveals similar topology as TFIID-TFIIA-SCP(-66). 3D models for rearranged conformation of TFIID-TFIIA-TFIIB-SCP(-66) (A) and TFIID-TFIIA-SCP (B) at 36\AA. (C) Docked DNA model for -66 to +45 with TBP-TFIIB-TFIIA crystal structure model (D) from Figure ~\ref{fig:Fig3.18}. }
\label{fig:IIB_cryo}
\end{figure}

\section{A general model of regulated DNA binding by TFIID}
\begin{figure}
\centering
\includegraphics[width=0.8\textwidth]{../Ch4_figs/Fig4.13.eps}
\caption[Model for SCP DNA binding by TFIID-TFIIA]{Model for SCP DNA binding by TFIID-TFIIA. TFIID undergoes conformational changes between canonical (A) and rearranged (B) states. The addition of TFIIA stabilizes the canonical conformation (C). Alternatively, the addition of SCP DNA leads to binding of the Inr-MTE/DPE to the rearranged conformation (D).  The combined presence of TFIIA and SCP DNA leads to a stabilization of the rearranged conformation (G).  While most particles adopt the rearranged state for TFIID-TFIIA-SCP, there is a small subset that is bound to SCP DNA and TFIIA within the canonical state (E). This state converts to the rearranged state through a likely intermediate state (F). The rearranged TFIID-TFIIA-SCP conformation can then be bound by TFIIB to load RNAPII.}
\label{fig:Fig4.13}
\end{figure}

The structural work presented within this chapter has served to test and extend a model of promoter binding by TFIID. In order to describe TFIID's interaction with promoter DNA, the model must take into account the interplay between TFIID's conformational dynamics and promoter binding. Therefore, the following models have been proposed in an attempt to summarize the structural data presented throughout Chapters 2, 3 \& 4 (Figures~\ref{fig:Fig4.13} - \ref{fig:Fig4.15}). While the rearranged state serves as the predominant high affinity DNA binding conformation for TFIID, the structural data presented here address alternative structural states that TFIID can adopt during the process of DNA binding. The proposed models show the interplay between core promoter architecture and TFIIA, where the arrangements of core promoter motifs dictate the mode of DNA binding by TFIID: SCP (Figure~\ref{fig:Fig4.13}), SCP(mTATA) (Figure~\ref{fig:Fig4.14}), and SCP(mMTE/DPE) (Figure~\ref{fig:Fig4.15}). Broadly, we believe these models serve as a framework for understanding TFIID's pleiotropic interactions with promoter DNA in an activator-dependent manner.\\

\subsection{TFIID-TFIIA-SCP}

As previously suggested, TFIID's conformational state appears to be intimately connected to its promoter binding properties. Before binding DNA, however, TFIID already exhibits an unprecedented degree of conformational dynamics, where lobe A reorganizes between two distinct conformations: the canonical and rearranged states (Figures~\ref{fig:Fig4.13}A \& B). Measurements of lobe A position from both negative stain and cryo-EM sample preparations revealed that lobe A undergoes a structural transition between these two states. Surprisingly, these results indicated that approximately 50\% of the TFIID molecules adopted the rearranged state before the addition of activators or promoter DNA  (Figure~\ref{fig:Fig2.4}). Therefore, within our model, lobe A coexists equally between the canonical (Figure~\ref{fig:Fig4.13}A) and rearranged states (Figure~\ref{fig:Fig4.13}B). In order to explain the lobe A's pendulum-like motion, we have modeled lobe A to contain an unstructured linker domain that remains stably attached to the BC core throughout the process of lobe A's reorganization. \\
\indent From this existing equilibrium between the canonical and rearranged states, we propose that there are two alternative paths towards the high-affinity rearranged state bound to DNA. One path involves the binding of SCP DNA to the rearranged state of TFIID in the absence of TFIIA (Figure~\ref{fig:Fig4.13}D). This was indicated by the identical footprinting patterns of the Inr-MTE/DPE within TFIID-SCP and TFIID-TFIIA-SCP. While the rearranged state is competent for DNA binding, the presence of SCP DNA alone is not sufficient to dramatically alter the conformational distribution of TFIID (Figure~\ref{fig:Fig2.6}B). Therefore, while the conformational dynamics of lobe A remain unchanged, the presence of high affinity Inr and MTE/DPE motifs within the SCP allows TFIID to interact with the DNA within the rearranged conformation.  Notably, TBP remains inhibited in the absence of TFIIA (Figure~\ref{fig:Fig4.1}), which is likely due to the inhibitory N-terminal domain of TAF1 \cite{Bagby_2202,Geiger_2949,Liu_2574}. \\
\indent Given the ability of TFIIA to stimulate high affinity TFIID-DNA complexes, we hypothesized that TFIIA should stabilize TFIID within the rearranged state. Surprisingly, TFIIA stabilizes TFIID within the canonical state (Figure~\ref{fig:Fig2.6}C). While we do not understand the functional role of this canonical state-stabilization, we propose that this conformation primes TFIID for TBP-TATA interactions through the release of the inhibitory N-terminal domain of TAF1 (Figure~\ref{fig:Fig4.13}C). \\
\indent These two parallel branches of TFIID's conformational dynamics converge when TFIID is in the presence of TFIIA and SCP DNA, where the rearranged conformation becomes preferentially stabilized. Given that a majority (60\%) of the particles for TFIID-TFIIA-SCP adopted the rearranged state (Figure~\ref{fig:Fig2.6}D) and were bound to promoter DNA (Figure~\ref{fig:Fig2.7}A), we modeled the conformational landscape of TFIID-TFIIA-SCP to facilitate the formation of the rearranged state (Figure~\ref{fig:Fig4.13}G). We believe that the transition from Figure~\ref{fig:Fig4.13}D to Figure~\ref{fig:Fig4.13}G involves the stabilization of TBP-TATA interactions through a TFIIA-mediated release of TBP inhibition. This is supported by the strong TATA box protection observed only within the ternary TFIID-TFIIA-SCP complex. \\
\indent On the other hand, we believe that there are at least two structural transitions that are necessary to facilitate the reorganization from the canonical to rearranged state. Since TFIIA localizes to lobe A (Figure~\ref{fig:Fig4.5}A), we predicted that the TATA box would be bound by lobe A within the canonical state.  This hypothesis was tested through Nanogold localization of SCP(+45 gold), where the SCP DNA was oriented in a manner that the MTE and DPE motifs were bound by lobe C (Figure~\ref{fig:Fig4.5}B). This binding geometry was incompatible with TATA box binding by lobe A in the canonical state due to the rigidity of the DNA. This was shown through SCP(TATA gold) localization in a distribution of states \emph{away} from lobe A in the canonical state (Figure~\ref{fig:Fig4.5}C). Therefore, we have modeled the canonical state for TFIID-TFIIA-SCP to contain lobe C interacting with the MTE/DPE in a similar manner state as the rearranged state (Figure~\ref{fig:Fig4.13}E). We believe that this conformation is primed for rearrangement due to the TFIIA-mediate release of TBP inhibition, allowing lobe A to bind the TATA box within the rearranged state. Given the flexible nature of lobe A's attachment to the BC core, the structural reorganization between the canonical and rearranged states likely involves an intermediate state, where DNA is bound to lobe C and TFIIA is within lobe A (Figure~\ref{fig:Fig4.13}F). The culmination of these processes results in the net-stabilization of TFIID-TFIIA-SCP within the rearranged state (Figure~\ref{fig:Fig4.13}G).\\ 

\subsection{TFIID-TFIIA-SCP(mTATA)}

\begin{figure}
\centering
\includegraphics[width=0.8\textwidth]{../Ch4_figs/Fig4.14.eps}
\caption[Model for SCP(mTATA) DNA binding by TFIID-TFIIA]{Model for SCP(mTATA) DNA binding by TFIID-TFIIA. TFIID undergoes conformational changes between canonical (A) and rearranged (B) states. The addition of SCP(mTATA) interacts with the rearranged conformation (C) that is further stabilized by TFIIA (D). Note that the DNA likely makes weak contacts with the surface of TFIID near TBP without TBP engagement of DNA. The rearranged TFIID-TFIIA-SCP(mTATA) conformation should then be bound by TFIIB, although it is difficult to know where the binding site of TFIIB at this time (E).}
\label{fig:Fig4.14}
\end{figure}

The model presented for the conformational landscape of TFIID-TFIIA-SCP describes at least two structural pathways sampled by TFIID during the binding of SCP DNA. While these studies of the SCP have provide the framework for understanding the relationship between TFIID's conformational dynamics and promoter binding, the majority of promoters across the human genome likely contain Inr-MTE/DPE or TATA-Inr promoters \cite{Juven-Gershon_468}. To address TFIID's promoter selectively, we performed footprinting and cryo-EM visualization of TFIID bound to SCP(mTATA), an SCP construct lacking a functional TATA box. DNase I footprinting showed that TFIID retains its intrinsic affinity for Inr-MTE/DPE in the absence of TFIIA (Figure~\ref{fig:Fig4.7}). Since the pattern of protection from the Inr to the MTE/DPE for TFIID-SCP(mTATA) is nearly identical to TFIID-SCP and TFIID-TFIIA-SCP, we can infer that TFIID binds the SCP(mTATA) sequence within the rearranged conformation (Figure~\ref{fig:Fig4.14}C). As seen previously for wild-type SCP, the addition of TFIIA does not change Inr-MTE/DPE contacts. However, when the TATA box is mutated, strong DNA contacts along the TATA box are disrupted while retaining low levels of DNA protection within the mutant TATA box sequence (Figures~\ref{fig:Fig4.7} \& \ref{fig:Fig4.8}). These footprinting data indicate that TFIID-TFIIA-SCP(mTATA) interacts with DNA through the rearranged conformation, but that there are minimal interactions between TFIID-TFIIA and the mutant TATA box sequence (Figure~\ref{fig:Fig4.14}D).   \\
\indent We propose that the absence of a functional TATA box severely reduces the role of the canonical state for DNA binding (Figure~\ref{fig:Fig4.14}). For wild-type SCP, we previously proposed that TFIIA binding to lobe A within the canonical state activates TBP for TATA box binding. Due to the lack of a functional TATA box and the minimal DNase I protection for TFIID-TFIIA-SCP(mTATA), we believe that TFIIA-mediated binding of the SCP(mTATA) to lobe A does not facilitate high-affinity interactions in the same way that it stimulates TATA box binding for the wild-type SCP.\\
\indent The well-characterized interactions between TFIIB, TBP, and DNA suggested that the addition of TFIIB may stimulate near wild-type binding of TFIID-TFIIA to the mutated TATA box sequence. While there may be TFIIB-TBP interactions present within TFIID-TFIIA-TFIIB-SCP(mTATA), there was only minimal protection of the mutant TATA box sequence, similar to that observed previously for TFIID-TFIIA-SCP(mTATA) (Figure~\ref{fig:Fig4.12}). While the downstream interactions from the Inr to the MTE/DPE remain unchanged, the lack of any TFIIB-induced DNA protection near the TATA box suggests that either TFIIB is not bound to the TFIID-TFIIA-SCP(mTATA) complex or that TFIIB is anchored to the complex through protein-protein interactions with TBP. The extensive literature of TFIIB-TBP-TATA would suggest that these interactions would be maintained, even in the absence of consensus TATA box motifs \cite{Rhee_24}. Our results indicate TFIIB is not sufficient to induce DNase I protected sites within TFIID-TFIIA-TFIIB-SCP(mTATA), warranting further study of transcription initiation from non-TATA containing promoters.\\

\subsection{TFIID-TFIIA-SCP(mMTE/DPE)} 

\begin{figure}
\centering
\includegraphics[width=0.8\textwidth]{../Ch4_figs/Fig4.15.eps}
\caption[Model for SCP(mMTE/DPE) DNA binding by TFIID-TFIIA]{Model for SCP(mMTE/DPE) DNA binding by TFIID-TFIIA. TFIID undergoes conformational changes between canonical (A) and rearranged (B) states. The addition of TFIIA stabilizes the canonical conformation (C). The addition of SCP(mMPE/DPE) leads to the formation of the rearranged conformation (F) through the proposed intermediates of (D) and (E). The rearranged TFIID-TFIIA-SCP(mMTE/DPE) conformation can then be bound by TFIIB to load RNAPII (G).}
\label{fig:Fig4.15}
\end{figure}

We further tested our model by studying the structural effects of a TATA-Inr promoter within the context of TFIID in the presence or absence of TFIIA. Surprisingly, our results showed that mutation of the MTE/DPE motifs abolished DNA binding of TFIID to promoter DNA for TFIID-SCP(mMTE/DPE) (Figure~\ref{fig:Fig4.7}). This indicates that TFIID's interactions with promoter DNA through the rearranged conformation in the absence of TFIIA was dependent on functional MTE/DPE motifs. The addition of TFIIA to TFIID-SCP(mMTE/DPE) resulted in a strong protection of the TATA box in conjunction with near wild-type binding to the mutated MTE/DPE sequences (Figure~\ref{fig:Fig4.7}). Considering that the protection patterns of TFIID-TFIIA-SCP(mMTE/DPE) strongly resembled wild-type interactions, we propose that TFIID-TFIIA-SCP(mMTE/DPE) adopts the rearranged conformation. However, unlike previous models, the footprinting data suggest that TFIIA-induced binding of TBP to the TATA box facilitates the reorganization of lobe A into the rearranged state. Therefore, we propose that TFIIA, bound to lobe A within the canonical state (Figure~\ref{fig:Fig4.15}C), stimulates TATA box binding by TBP (Figure~\ref{fig:Fig4.15}D) and thereby facilitates the movement of lobe A into the rearranged state (Figure~\ref{fig:Fig4.15}E \& F). \\
\indent In addition to the results obtained from footprinting on SCP(mMTE/DPE), similar results were obtained for a mutant Inr (Figure~\ref{fig:Fig4.7}), suggesting that TFIIA-mediated stimulation of TATA box binding is sufficient to drive rearrangement of lobe A in the absence of functional Inr or MTE/DPE motifs. We believe that these results are generally applicable for TFIID's interaction with low affinity promoter DNA sequences, where the binding of additional factors, anchoring TFIID to the promoter DNA, will stimulate DNA binding through an increase in TFIID's avidity for these low affinity sites. Given TFIID's ability to interact with a variety of upstream activators \cite{Goodrich_503,Levine_1710}, in addition to histone modifications (e.g. \cite{Jacobson_2000,Vermeulen_2007}), we believe that the combined effect of TFIID's low affinity interactions with binding partners and DNA sequences is to stabilize TFIID bound to promoter DNA within the rearranged conformation.\\

\subsection{Implications for RNAPII loading at the core promoter}

\begin{figure}
\centering
\includegraphics[width=0.8\textwidth]{../Ch4_figs/BentDNA_mdoels.eps}
\caption[TFIID may facilitate transcription initiation through topological changes in promoter DNA]{TFIID may facilitate transcription initiation through topological changes in promoter DNA. Model of promoter DNA path through TFIID-TFIIA-SCP(-66) for core promoter element location (A), DNase I footprinting (B), and MPE-Fe footprinting (C).  (D) DNA path through an elongating RNAPII (adapted from PDB 3PO3). (E) Corresponding view from TFIID-TFIIA-SCP(-66) model.}
\label{fig:BentDNA}
\end{figure}

Our studies suggest that TFIID introduces specific changes in the structure and topology of promoter DNA that may contribute to its role in transcription initiation.  The cryo-EM structure of TFIID-TFIIA-SCP provides insight into a previously observed DNase I hypersensitive at +3, localizing this cleavage site to DNA spanning the central channel (Figure~\ref{fig:Fig4.4}, denoted with *).  This hypersensitive site has been observed in a wide range of core promoters upon binding of purified TFIID from Drosophila \cite{Burke_3081,Kutach_2124,Lim_1522,Parry_382,Purnell_3259,Theisen_341} as well as human sources \cite{Chi_3023,Juven-Gershon_1249,Kaufmann_3320,Lieberman_2671,Oelgeschlager_2880}.  The strength of this hypersensitivity at +3 is positively correlated with the strength of \emph{in vitro} transcription initiation, suggesting that the topological changes that TFIID induces at this site may be important for subsequent steps in transcription initiation.  \\
\indent The path of the upstream promoter DNA presented in this model has some common features with a recently published model \cite{Treutlein_2} of the formation of a yeast open RNAPII promoter complex (Figure~\ref{fig:BentDNA}).  Through single-molecule F\"{o}rster resonance energy transfer experiments of an assembled open RNAPII complex that included TBP, TFIIB, TFIIF, RNAPII, and DNA, the authors concluded that TBP, TFIIB and upstream DNA move with respect to RNAPII during the transition from the closed to open promoter complex \cite{Treutlein_2}.  The TBP-TFIIB-DNA complex moves from a position proximal to the wall of RNAPII to a position directly above the clamp. There is an intriguing similarity between the DNA structure modeled by these authors \cite{Treutlein_2} and the path of the DNA that we describe and place within the context of TBP, TFIIA and TFIIB (Figure~\ref{fig:BentDNA}). Based on this similarity it is tempting to speculate that the \emph{in vitro} and \emph{in vivo} strength of the SCP sequence relates to the re-positioning of TFIID subunits and DNA into an optimal configuration for RNAPII loading.