Graduate students in my research group are members of either the Chemistry Department (M. S.) or the Biological Sciences Department (M. S., Ph. D.). Students may also carry out their research through the Center of Biomolecular Sciences. This is an interdisciplinary program of the Chemistry and Biological Sciences Departments, established in the Spring of 2000.
Capsule Overview of the Research Program My lab is interested in DNA-protein and protein-protein interactions, with the aim of better understanding the complex array of communications that take place during the build-up of the transcriptional machinery needed to initiate transcription. Initial steps in the transcription process involve the binding of one or more regulatory transcription factors &/or the basal TATA-binding factor (TBP) to specific sites in the regulatory or promoter regions of specific genes within DNA. After the initial binding of either of these "initiation" factors, it is thought that this permits more effective recruitment of the subsequent factors to the site. In addition, auxillary factors, called coactivators, can further facilitate the binding &/or the efficient assembly of the multiprotein complexes. The coactivators appear to have opportunities to enter the complex at multiple assembly/time points.
High Mobility Group Protein B 1/2 (HMGB1/2)
The HMG family of proteins is classified in three subgroups. HMGA (AT-hooks), HMGB (formerly HMG1 & -2) and the HMGN (formerly HMG14 & 17, that readily bind to nucleosomes) proteins. HMGB1 and -2 proteins are highly conserved (99% amongst mammals), ubiquitous in eukaryotes (absent in eubacteria & archaea) (10) and found in relatively high levels in many cell types. They are about 25 kDa and are highly charged throughout. There are two structural domains, called the A- and the B-boxes, which although they have different sequences, are structurally very similar. Each of the basic "HMG boxes" contains about 80 residues and each is highly positively charged. The HMGB boxes are contiguous to each other (A-B) starting at the N-terminus. The B-box is followed by a short positively charged region, with the last 30 residues in the acidic C-terminus of the protein made up entirely of aspartic or glutamic acid residues.
HMGB proteins bind non-specifically to the minor groove in DNA, produce significant bends in the DNA or bind preferentially to prebent DNA. The DNA bending produced by HMGB proteins is a result of the binding of an HMGB box, which folds into 3 a-helices forming an L-shaped structure, with hydrophobic residues inserted between the bases. Both HMGA and HMGB proteins are generally classified as "architectural" proteins because their binding to & bending of DNA has been shown to facilitate the facile assembly of complex arrays of proteins at a DNA site. The prototype for this, in the case of HMGA proteins, is the enhanceosome for the interferon-b (INF-b) gene promoter. HMGB1 is incorporated into an enhanceosome, targeting either the promoter or the enhancer for the Epstein-Barr (EB) gene, BHLF-1 gene. HMGB1 interaction involves either Epstein Barr ZEBRA & SP1 proteins or EB Rta dimer, respectively (11).
There is now evidence that HMGB1 is excreted from necrotic cells and triggers an inflammatory response, in contrast to apoptotic cells, in which HMGB1 is tightly attached to chromatin until cells are cleared (11, 12).
HMGB-knock-out mice (Hmgb1-/-) are born, although they have numerous immediate problems, including their smaller size, eyelid defects, glycogen utilization and glucose metabolism.
Recent findings indicate that HMGB1 is identical to amphoterin and sulfoglucuronyl binding protein-1 (SBP-1).
The Effect of HMGB1/2 on the Assembly of the Preinitiation Complex (PIC)(2-5)
We have examined the effect of HMGB1/2 and adenovirus E1A protein on the step-wise binding of TBP, TFIIB and TFIIA during the early stages of assembly of the transcriptional preinitiation complex (PIC).
HMGB1 binds to the hTBP/TATA complex to enhance the human TBP binding affinity by ca. 25X. This effect is primarily a result of an increased on-rate for TBP binding to. This effect is observed only with hTBP, but not with the TBP from yeast or Drosophila. We have shown that the binding interaction requires the Q-tract in the non-conserved N-terminus in hTBP, in addition to the acidic C-terminus in HMGB1. Of the numerous complexes that HMGB1 binds to and enhances complex formation, the HMGB1/TBP/TATA complex is the only complex in which HMGB1 has been clearly shown to be a stable component of the complex by EMSA supershift.
The resultant complex formed in the presence of HMGB1, TFIIB and TBP/TATA is condition dependent. The evidence indicates that either a TFIIB/TBP/TATA or a HMGB1/TFIIB/TBP/TATA complex can be formed. Competition between TFIIA and HMGB1 binding to TBP/TATA clearly shows that they compete for overlapping sites on the TBP/TATA complex. TFIIA can readily dissociate HMGB1 from a preformed HMGB1/TBP/TATA complex to form a TFIIA/TBP/TATA complex. When all factors are reacted simultaneously, the TFIIA/TBP/TATA complex is formed, further supporting previous evidence that one of the functions of TFIIA is to dissociate repressor factors from PIC during its assembly.
The Effect of HMGB1/2 on Regulating Regulatory Factors
My lab, and a number of others, has shown that the presence of HMGB1/2 enhances the binding affinity of a number of transcription factors (TFs) to their recognition sites. We have classified these HMGB1-targeted TFs as the “HMGB-sensitive TFs family”. This family includes the human basal transcription factor, hTBP, in addition to activators, such as the steroid hormone receptors, HOX9 homeodomain proteins, rel family proteins, POU-domain transactivators, p53, p73, USF (1), herpes simplex virus ICP4 and Epstein Barr viral activators, ZEBRA and Rta. With the exception of ICP4, HMGB1 facilitates stronger binding of these factors to their recognition site. With all but hTBP, this increased binding affinity correlating with increased transcriptional activity. At least in the case of PR, there is evidence that it is the A- &/or the B-box of HMGB1 that is directly involved in the interactions.
Edward’s lab has shown that HMGB1/2 enhances the binding of steroid hormone receptors, such as progesterone, estrogen, glucocorticoid and androgen receptors, but not class II, such as the thyroid and vitamin D3, and the retinoic acid receptors.
Since HMGB1 binding to TBP/TATA represses transcription, while HMGB1 interaction enhances transcription by binding to regulatory factors, HMGB1 is a representative of a context-dependent regulator of transcription.
Estrogen Receptors (ERa & ERb) (6-8)
Nuclear hormone receptors make up an enormous group of ligand-activated transcription factors. The steroid hormone receptors (SHRs) make up Class 1 and include the glucocorticoid receptor (GR), mineralocorticoid receptors (MR), progesterone receptors (PR), androgen receptors (AR) and estrogen receptors (ER). Class II receptors include those for retinoic acid (RAR), retinoic X (RXR), vitamin D (VDR), thyroid (TR) and peroxisome proliferator activated receptor (PPAR), amongst others. So-called orphan receptors make up the last class and were so named because their endogenous activating ligand was originally unknown or has been recently been defined. The SHRs vary in size, but can be viewed as containing three functional domains.
These domains carryout the functions of DNA binding, nuclear localization, receptor dimerization, ligand binding and transcriptional modulation. The A/B domains are relatively nonconserved across the nuclear receptors, being poorly conserved even between the two ER isoforms, ER
a and ERb. This domain makes up the N-terminus (NTD) and is associated with hormone-independent activation, with an activating function (AF1) residing in this region (6,7) that appears to be cell and promoter specific (9). The C-domain, called the DNA-binding domain (DBD), is essential for the receptor binding to its DNA recognition site, which is called the estrogen response element (ERE). The 84 residue ERDBD is entirely conserved between chicken and humans, with the smaller, core ERDBD (66 residues), considered the smallest unit that exhibits DNA binding. Additional residues that are C-terminal to the DBD (the C-terminal extension or CTE) contribute to additional binding affinity in class II nuclear receptors, with increasing evidence suggesting that this may also be the case for the SHRs. The DBD contains two zinc ions, each coordinated to four cysteines, providing zinc binding motifs, with the unit being essential for proper protein folding and DNA binding. Residues in the first of the two a–helices in the DBD (in the P-box) interact directly with the bases in the major groove of the DNA to form the ER/ERE interaction.
All nuclear receptors, except for the orphan receptors, bind to their recognition sequences as either homodimers (class I) or as heterodimers (class II). The SHRs bind as homodimers to a palindromic or imperfect palindromic sequence in which each half-site is composed of 5 or 6 bps, separated by 3 bps, with the composition of the latter being unimportant. The palindromic arrangement is also referred to as an inverted repeat (IR) and the ERE sequence for ER is 5’-AGGTCAnnnTGACCT-3’, while the recognition sequence for all other SHRs is 5’-AGAACAnnnTGTTCT-3’. Notice that the only differences occur in the central 2 bps in each half-site (shown in underlined bold type). With a spacer of 3 bps between each half-site, the center-to-center distance between the half-sites is 10 bps & 34A or essentially one complete turn of the helix. This places each half-site on the same side of the helix and is thought to maximize or optimize the protein-protein interactions between the SHR dimers.
Class II nuclear receptors differ in their binding selectivity in a number of aspects. They bind to the sequence 5’-AGGTCA-3’, but with various spacing between these half-sites, in addition to binding as hetero- or homodimers.
The current paradigm regarding selective binding for nuclear hormone receptors to their recognition elements is that they recognize 1) DNA sequence, 2) the spacing between the half-sites and 3) the orientation (inverted repeat, direct repeats or everted repeats) of the half-sites. This would appear to explain the different selectivity between the SHRs, ER and GR, but not between different SHRs, GR, PR, AR & MR since each of these receptors recognize the same recognition sequence (5’-AGAACAnnnTGTTCT-3’), usually referred to as the GRE.
The residues in ER that appear to be essential for selective interaction with ERE are Glu25, Gly26 and Ala29 and reside in the a-helix within P-box of the ERDBD. The corresponding residues in the GRDBD are Gly, Ser and Val.
The amino acid sequence for the 84-residue ERDBD that is conserved from chicken to humans is shown below (8, 13; residue 64 (L/E) differs in the two reports), with the additional residues, starting from residue 73, corresponding to the C-terminal extension (CTE). The residues in the ERb sequence that are not identical to those in ERa are shown below the ERDBD & CTE sequence for ERa. Note the highly conserved nature in the DBD sequence, which is contrasted by the very limited conservation in the CTE regions.
Residue 7 (*) is the first residue in the “core” DBD, with actual residue numbers for full length ER being 185-251. 66 amino acid residues are the “core” DBD
The 84 amino acid residue fragment is conserved from chicken to humans
Zn binds to the cysteines shown in bold type.
The underlined sequences are the
a–helical regions.The three enlarged and bold residues are those that are essential for ER binding selectivity to ERE.
66 amino acid residues are the “core” DBD
84 amino acid residue fragment is conserved from chicken to humans
Graphic of ER DBD binding to the ERE. Each ER DBD monomer uses its recognition a–helix to bind in the major groove of the ERE. The Zn ions are shown as silver spheres, with an ER-ER (protein-protein) interaction shown utilizing residues in the D-box. The graphic is from U. Illinois University.
Effect of HMGB1/2 on ER Binding to ERE half-sites (HEREs) (5)
In progress
References
1. Marmillot & Scovell (1997) Biochim. Biophys. Acta. 1395, 228-236.
2. Lu, Peterson, Dasgupta & Scovell (2000) J. Biol. Chem. 275, 35006-35012
3. Das & Scovell (2001) J. Biol. Chem. 276, 32597-32605.
4. Dasgupta & Scovell (2003) Biochim. Biophys. Acta 1627, 101-110.
5. Das, Peterson & Scovell (2004) Mol. Endocrinol., in press, Nov., 2004.
6. Mangelsdorf et al (1995) Cell 83, 835-839.
7. Ruff et al (2000) Breast Cancer Res. 2, 359-359.
8. Sanchez et al (2002) Bio Essays 24, 244-254.
9. Truss & Beato (199) Endocrine Revs 14, 459-479.
10. Agresti & Bianchi (2003) Curr. Opin. Gen. & Dev. 13, 170-178.
11. Mitsouras et al (2002) Mol. Cell. Biol. 22, 4390-4402.
12. Scaffidi et al (2002) Nature 418. 191-195.
13. Muller et al (2001) EMBO J. 20, 4337-4340.
14. Schwabe et al (1993) Structure 1, 187-204.
15. Aranda & Pascual (2001) Physiol. Rev. 81, 1269-304.
16. Nilsson et al (2001) Physiol. Rev. 81, 1535-1565.
17. Klinge (2001) Nucleic acids Res. 29, 2905-2919.
