Final Report Summary - PICENGINEERING (Basal factor engineering and synthetic nucleosomal DNA super-template construction for pre-initiation complex assembly)
Class II transcription initiation requires the assembly of a pre-initiation complex (PIC). PIC contains RNA polymerase II (RNAPII), general transcription factors (GTFs) including TFIIA, TFIIB and TFIID assembling on a DNA template containing core promoter elements. The DNA template is flanked by nucleosomes possessing modified histones which are hallmarks of active transcription. TFIID interacts with these epigenetic markers, specifically H3K4Me3 and H3K14Ac to install the PIC at the transcription start site (TSS). TFIID binding via its TATA binding protein (TBP) to the core promoter is stabilized by TFIIA. Once TFIID (TBP) is anchored onto the core promoter, TFIIB binds to this complex and recruits RNAPII. This cross-talk of TFIID with TFIIA, TFIIB and nucleosome at the TSS before RNAPII recruitment is still elusive, which leads to our objectives. (1) To obtain and solve the three dimensional (3D) structure of a sub-complex, QUART, containing TFIIA, TFIIB and TBP with dsDNA, signified as promoter recognition complex, the first step in RNAPII transcription initiation. (2) To construct and characterize a ‘PIC stage’ that constitutes nucleosome(s) with epigenetic markers, TFIID, TFIIA, TFIIB, and a DNA template containing core promoter elements like TATA box, Initiator (INR), downstream promoter element (DPE) and Motif Ten element (MTE).
(1) Efforts to crystallize the quaternary complex of QUART has so far been hampered by laborious protein purification procedures, exacerbated by the tendency of TFIIB to be present in sub-stoichiometric ratios in crystals obtained. Prompted by an elegant study of the TFIIIB system, we pursued a single-chain approach to potentially overcome these impediments. We reconfigured the connectivity of the three polypeptide components that make up TFIIA core and linked them covalently to give rise to a single-chain (scTFIIA, Figure 1a in additional sheet). Similarly, a fusion of TBP and TFIIB (scTB, Figure 2a in additional sheet) was also obtained. The fellow expressed scTFIIA and scTB in E.coli and purified these proteins in quality and quantity suitable for structural studies (Figure 1 and 2 in additional sheet). scTFIIA was linked to TBP (scTA) and the fellow could express scTA in both E.coli and in insect cells using the MultiBac system developed by the host lab in EMBL, Grenoble. However, scTA was not stable as purified. She also cloned and purified TFIIB to attempt a semi-classical approach to obtain QUART complex by mixing scTFIIA, TFIIB, TBP with dsDNA. For QUART crystallization, she produced several double stranded (ds) DNA from oligos and characterized using electron mobility shift assays (EMSA) for binding to scTB and scTFIIA to produce QUART (Figure 2d in additional sheet). dsDNAs exhibiting good binding efficiency were tried for crystallization using state-of-the-art high throughput crystallization robot (HTXlab) at EMBL, Grenoble. Several crystallization conditions from the crystallization robot produced crystals. The fellow manually reproduced these crystals suitable for diffraction and performed X-ray diffraction experiments in synchrotron beam lines in ESRF, Grenoble. Some crystals did not diffract x-rays. Such crystal conditions are under optimization to obtain diffraction. Crystals that diffracted were pursued for structure solving. However until now, the structures contained only scTFIIA and no other components of the complex. This high propensity of scTFIIA to crystallize became new impedance in obtaining the crystals of the quaternary complex. Nevertheless, the single chain approach had facilitated the crystallization of scTFIIA. The fellow solved the structure of TFIIA, in its apo form for the first time (Table 1 and Figure 3a). Interestingly, the crystal contacts were mediated primarily by the loops engineered to link the TFIIA domains (Figure 3b) and these loops are disordered in the ternary complex structure containing TFIIA, TBP and TATA DNA. To increase the propensity of obtaining QUART crystals, the fellow has screened several dsDNA (>25) with optimized sequence like BRE (TFIIB Response Element) so that all the protein components will bound to DNA with high affinity. She took another approach where TFIIA, TBP and TFIIB were tethered together as a single chain protein (scTAB) with 1:1:1 stoichiometry of each of the protein components. This construct would reduce the propensity of TFIIA from crystallizing and keep TFIIB intact in the quaternary complex.The fellow expressed scTAB using the state-of-the-art Insect cells - MultiBac expression system developed by the host lab. As scTAB was not highly soluble, she cloned and expressed it as a MBP tagged protein in insect cells (MBPscTAB, Figure 4a in additional sheet). She optimized the purification protocol and obtained sufficient quantity of MBPscTAB suitable for reconstitution of the ‘PIC stage’ and for structural studies (Figure 4b-4c). Using EMSA as a qualitative measure, she observed MBPscTAB binds to dsDNA as well (Figure 4d), albeit the absence of a clear band shift. MBPscTAB was mixed with several dsDNAs and subjected for crystallization trials in HTXlab, EMBL, Grenoble. Crystals obtained (Figure 4e) were directly checked for ‘in-plate’ diffraction in BM14 beamline in ESRF and as well manually reproduced and diffracted in more powerful beamlines in ESRF. But until now the crystals did not diffract X-rays. These crystal conditions are under optimization to obtain diffraction.
(2) We proposed earlier that the assembly of all the components on the DNA template will be addressed by cross-linking the protein components to DNA flanked by a -1 and +1 nucleosome. This method involves protein-DNA crosslinking via AHA (Methionine analogue) residues on the proteins. To achieve this, all intrinsic methionine residues of the protein will be first substituted with non-methionine residues. Second, based on existing structural knowledge, the residues spatially close to promoter DNA will be substituted to methionine which will then be modified to AHA residues and chemically cross-linked to DNA. The fellow had obtained constructs of scTFIIA, TBP, TFIIB and scTB devoid of methionines and two site specific methionine mutants for scTFIIA (scTFIIA_N57M, scTFIIA_S179M). To verify the integrity of the structure after mutation, she had crystallized and solved structures of some of these mutant proteins. During this course, new developments in this area of research in our collaborators laboratory have provided a rather simplistic model for PIC assembly. According to this new study, TFIID from HeLa extracts binds to mono nucleosome with H3K4Me3 modification in a TFIIA dependent manner (Figure 5a). Together with our progress in the production of MBPscTAB protein (see above), this study has simplified our proposed approach for PIC assembly, ie, (i) ‘the stage’ requires only one nucleosome with histone modifications rather than two flanking nucleosomes and (ii) the stage can be prepared by ‘in-vitro’ reconstitution with MBPscTAB in addition to the proposed method of cross-linking as illustrated in Figure 5b and 5c. The DNA template that will be used in ‘in-vitro’ reconstitution is designed based on the established super core promoter. It contains the core promoter elements namely TATA box, Initiator (INR), downstream promoter element (DPE) and Motif Ten element (MTE) and a 601 Widom sequence tethered downstream of INR (Figure 6). The fellow sub-cloned this DNA sequence (Kpn-NucI-BamH) into PUC vector as 32 repeats to facilitate its large scale production for structural studies. She had produced this DNA template in milligram quantities to support the future ‘in-vitro’ reconstitution experiments of the ‘PIC stage’.
Main scientific results are:
1. Single chain protein MBPscTAB (scTAB as an MBP tagged protein) has been produced in sufficient quantity and quality for the ‘PIC stage’ reconstitution and the crystallization experiments with dsDNA. MBPscTAB interaction with dsDNA was confirmed by EMSA.
2. MBPscTAB complexed with a dsDNA have provided crystals, which requires further optimization.
3. The DNA template containing TFIID recognition elements including TATA box, INR, DPE and MTE followed by 601 Widom sequence for the nucleosome was cloned and produced in large scale quantities for the ‘PIC stage’ reconstitution.
4. The three-dimensional structure of TFIIA has been determined in its apo-form. The crystallization of the protein is facilitated by the loops engineered between the domains.
5. Mutants of single chain proteins scTFIIA and scTB and of TFIIB and TBP devoid of methionines have been cloned, expressed and purified and for some, structure was solved using X-ray diffraction to confirm their structural integrity. Methionine mutants of scTFIIA (scTFIIA_N57M, scTFIIA_S179M) have been produced.
In short, the work has been productive towards the goals of the proposed project. As new research developments on the knowledge of TFIID-nucleosome interaction evolved during the course of the study and the purification of MBP-tagged scTAB construct, an ‘in vitro’ reconstitution method emerged as a complementary method to the proposed cross-linking approach. These two methods were pursued in parallel. QUART crystals were obtained with MBPscTAB and efforts are in progress towards obtaining diffracting crystals.
In future, we will exploit extensive crystallization trials focusing on MBPscTAB-dsDNA complex to obtain diffracting crystals of QUART complex. For the ‘PIC stage’, the ‘in-vitro’ reconstitution approach will be focused as it seems optimistically feasible in a limited period. Production of TFIID suitable for ‘PIC stage’ construction is under progress in the host laboratory. MBPscTAB will then be integrated into TFIID. Histone octamer with H3K4Me3 will be produced and reconstituted onto Kpn-NucI-BamH DNA template via 601 widom sequence with the help of Tim Bartke, Oxford, who has extensive experience in this field. Reconstituted nucleosome with H3K4Me3 modification will be complexed with MBPscTAB-TFIID using the core promoter elements on the DNA template upstream of 601 widom sequence and the epigenetic markers, thus producing the ‘PIC stage’. The supra-molecular architecture of the ‘PIC stage’ will be studied by state-of-the-art multi-resolution structural approach, including EM, SAXS and X-ray crystallography. The study will yield novel, crucial and timely scientific insight into committed human transcription mechanism. If QUART crystals are obtained, its long-awaited 3D structure will provide various insights on the molecular organization of three GTFs on the TATA box and reveal the network of interactions involved in template recognition in RNAPII transcription initiation.
(1) Efforts to crystallize the quaternary complex of QUART has so far been hampered by laborious protein purification procedures, exacerbated by the tendency of TFIIB to be present in sub-stoichiometric ratios in crystals obtained. Prompted by an elegant study of the TFIIIB system, we pursued a single-chain approach to potentially overcome these impediments. We reconfigured the connectivity of the three polypeptide components that make up TFIIA core and linked them covalently to give rise to a single-chain (scTFIIA, Figure 1a in additional sheet). Similarly, a fusion of TBP and TFIIB (scTB, Figure 2a in additional sheet) was also obtained. The fellow expressed scTFIIA and scTB in E.coli and purified these proteins in quality and quantity suitable for structural studies (Figure 1 and 2 in additional sheet). scTFIIA was linked to TBP (scTA) and the fellow could express scTA in both E.coli and in insect cells using the MultiBac system developed by the host lab in EMBL, Grenoble. However, scTA was not stable as purified. She also cloned and purified TFIIB to attempt a semi-classical approach to obtain QUART complex by mixing scTFIIA, TFIIB, TBP with dsDNA. For QUART crystallization, she produced several double stranded (ds) DNA from oligos and characterized using electron mobility shift assays (EMSA) for binding to scTB and scTFIIA to produce QUART (Figure 2d in additional sheet). dsDNAs exhibiting good binding efficiency were tried for crystallization using state-of-the-art high throughput crystallization robot (HTXlab) at EMBL, Grenoble. Several crystallization conditions from the crystallization robot produced crystals. The fellow manually reproduced these crystals suitable for diffraction and performed X-ray diffraction experiments in synchrotron beam lines in ESRF, Grenoble. Some crystals did not diffract x-rays. Such crystal conditions are under optimization to obtain diffraction. Crystals that diffracted were pursued for structure solving. However until now, the structures contained only scTFIIA and no other components of the complex. This high propensity of scTFIIA to crystallize became new impedance in obtaining the crystals of the quaternary complex. Nevertheless, the single chain approach had facilitated the crystallization of scTFIIA. The fellow solved the structure of TFIIA, in its apo form for the first time (Table 1 and Figure 3a). Interestingly, the crystal contacts were mediated primarily by the loops engineered to link the TFIIA domains (Figure 3b) and these loops are disordered in the ternary complex structure containing TFIIA, TBP and TATA DNA. To increase the propensity of obtaining QUART crystals, the fellow has screened several dsDNA (>25) with optimized sequence like BRE (TFIIB Response Element) so that all the protein components will bound to DNA with high affinity. She took another approach where TFIIA, TBP and TFIIB were tethered together as a single chain protein (scTAB) with 1:1:1 stoichiometry of each of the protein components. This construct would reduce the propensity of TFIIA from crystallizing and keep TFIIB intact in the quaternary complex.The fellow expressed scTAB using the state-of-the-art Insect cells - MultiBac expression system developed by the host lab. As scTAB was not highly soluble, she cloned and expressed it as a MBP tagged protein in insect cells (MBPscTAB, Figure 4a in additional sheet). She optimized the purification protocol and obtained sufficient quantity of MBPscTAB suitable for reconstitution of the ‘PIC stage’ and for structural studies (Figure 4b-4c). Using EMSA as a qualitative measure, she observed MBPscTAB binds to dsDNA as well (Figure 4d), albeit the absence of a clear band shift. MBPscTAB was mixed with several dsDNAs and subjected for crystallization trials in HTXlab, EMBL, Grenoble. Crystals obtained (Figure 4e) were directly checked for ‘in-plate’ diffraction in BM14 beamline in ESRF and as well manually reproduced and diffracted in more powerful beamlines in ESRF. But until now the crystals did not diffract X-rays. These crystal conditions are under optimization to obtain diffraction.
(2) We proposed earlier that the assembly of all the components on the DNA template will be addressed by cross-linking the protein components to DNA flanked by a -1 and +1 nucleosome. This method involves protein-DNA crosslinking via AHA (Methionine analogue) residues on the proteins. To achieve this, all intrinsic methionine residues of the protein will be first substituted with non-methionine residues. Second, based on existing structural knowledge, the residues spatially close to promoter DNA will be substituted to methionine which will then be modified to AHA residues and chemically cross-linked to DNA. The fellow had obtained constructs of scTFIIA, TBP, TFIIB and scTB devoid of methionines and two site specific methionine mutants for scTFIIA (scTFIIA_N57M, scTFIIA_S179M). To verify the integrity of the structure after mutation, she had crystallized and solved structures of some of these mutant proteins. During this course, new developments in this area of research in our collaborators laboratory have provided a rather simplistic model for PIC assembly. According to this new study, TFIID from HeLa extracts binds to mono nucleosome with H3K4Me3 modification in a TFIIA dependent manner (Figure 5a). Together with our progress in the production of MBPscTAB protein (see above), this study has simplified our proposed approach for PIC assembly, ie, (i) ‘the stage’ requires only one nucleosome with histone modifications rather than two flanking nucleosomes and (ii) the stage can be prepared by ‘in-vitro’ reconstitution with MBPscTAB in addition to the proposed method of cross-linking as illustrated in Figure 5b and 5c. The DNA template that will be used in ‘in-vitro’ reconstitution is designed based on the established super core promoter. It contains the core promoter elements namely TATA box, Initiator (INR), downstream promoter element (DPE) and Motif Ten element (MTE) and a 601 Widom sequence tethered downstream of INR (Figure 6). The fellow sub-cloned this DNA sequence (Kpn-NucI-BamH) into PUC vector as 32 repeats to facilitate its large scale production for structural studies. She had produced this DNA template in milligram quantities to support the future ‘in-vitro’ reconstitution experiments of the ‘PIC stage’.
Main scientific results are:
1. Single chain protein MBPscTAB (scTAB as an MBP tagged protein) has been produced in sufficient quantity and quality for the ‘PIC stage’ reconstitution and the crystallization experiments with dsDNA. MBPscTAB interaction with dsDNA was confirmed by EMSA.
2. MBPscTAB complexed with a dsDNA have provided crystals, which requires further optimization.
3. The DNA template containing TFIID recognition elements including TATA box, INR, DPE and MTE followed by 601 Widom sequence for the nucleosome was cloned and produced in large scale quantities for the ‘PIC stage’ reconstitution.
4. The three-dimensional structure of TFIIA has been determined in its apo-form. The crystallization of the protein is facilitated by the loops engineered between the domains.
5. Mutants of single chain proteins scTFIIA and scTB and of TFIIB and TBP devoid of methionines have been cloned, expressed and purified and for some, structure was solved using X-ray diffraction to confirm their structural integrity. Methionine mutants of scTFIIA (scTFIIA_N57M, scTFIIA_S179M) have been produced.
In short, the work has been productive towards the goals of the proposed project. As new research developments on the knowledge of TFIID-nucleosome interaction evolved during the course of the study and the purification of MBP-tagged scTAB construct, an ‘in vitro’ reconstitution method emerged as a complementary method to the proposed cross-linking approach. These two methods were pursued in parallel. QUART crystals were obtained with MBPscTAB and efforts are in progress towards obtaining diffracting crystals.
In future, we will exploit extensive crystallization trials focusing on MBPscTAB-dsDNA complex to obtain diffracting crystals of QUART complex. For the ‘PIC stage’, the ‘in-vitro’ reconstitution approach will be focused as it seems optimistically feasible in a limited period. Production of TFIID suitable for ‘PIC stage’ construction is under progress in the host laboratory. MBPscTAB will then be integrated into TFIID. Histone octamer with H3K4Me3 will be produced and reconstituted onto Kpn-NucI-BamH DNA template via 601 widom sequence with the help of Tim Bartke, Oxford, who has extensive experience in this field. Reconstituted nucleosome with H3K4Me3 modification will be complexed with MBPscTAB-TFIID using the core promoter elements on the DNA template upstream of 601 widom sequence and the epigenetic markers, thus producing the ‘PIC stage’. The supra-molecular architecture of the ‘PIC stage’ will be studied by state-of-the-art multi-resolution structural approach, including EM, SAXS and X-ray crystallography. The study will yield novel, crucial and timely scientific insight into committed human transcription mechanism. If QUART crystals are obtained, its long-awaited 3D structure will provide various insights on the molecular organization of three GTFs on the TATA box and reveal the network of interactions involved in template recognition in RNAPII transcription initiation.