Thursday, 15 November 2012

Promoter recognition

The RNA polymerase II core promoter - the gateway to transcription by Juven-Gorshen, Hsu, Theisen, and Kadonaga and RNA polymerase II core promoter by Smale and Kadonaga
Focused versus dispersed core promoters
In focused core promoters there is either single transcription start site or a distinct cluster of short sites in a short region of several nucleotides. Most uk core promoters appear to be focused core promoters. In vertebrates only 1/3 core promoters are focused; the majority of genes contain dispersed core promoters.  A number of transcription start sites are distributed over a broad region that might range from 50-100 nts.

TATA box, BRE, Inr, MTE, DPE, and DCE are found in focused core promoters. These core promoter elements are not universal. Each is present in only a subset of core promoters.  TATA box and BRE are most ancient of core promoter motifs.

TATA box, BRE and cognate protein factors, TBP, TFIIB are conserved from archaea to humans. TATA  box is in plant promoters.

Dispersed core promoters are found in CpG islands in vertebrates and lack TATA, DEP and MTE.

Focused core promotesr are more ancient and used in a broader range of organisms than dispersed promoters.

Initiator (Inr)
Inr encompases the transcription start site.  Inr consensus is YYANWYY in humans and TCAKTY ni Drosophila. A in the middle is often the +1 start site in focused core promoters.  It is the most common sequence motif in focused core promoters. Inr is a recognition site for TFIID binding.  Sequence specificity of TFIID binding to Inr region of core promoter is identifcal to Inrs consensus sequence.

It can function independently of TATA box in an analysis of lymphocyte-specific terminal transferase promoter. Transcription from this promoter initiates at a single start site. But region between -25 and -30 is G/C-rich and is unimportant for promoter activity. Mutant analysis showed that sequence between -3 and +5 is necessary and sufficient for accurate transcription.

Tnr functions synergestically with TATA box when separated by 25 bp. When an Inr is inserted into a synthetic promoter downstream of 6 binding sites for transcription factor Sp1 (absence of TATA box), Inr supports high levels of transcription that initiate at a spec start site within Inr. When Inr is inserted at a diff location relative to Sp1 sites, RNA synthesis begins at nt dictated by Inr. When Inr is absent, transcription begins at heterogenous start sites at much lower frequencies.

Mammalian Inr consensus is suggested to be YR (R is +1 start site). Drosophila core promoters have a stricter consensus sequence of TCAGTY. Interestingly, transcription does not need to being at +1 nutcleotide for Inr to function.

TATA and Inr act independently when separated by more than 30 bp. When separated by 15 or 20 bp synergy is retained, but ocation of start site is dictated by location of TATA rather than that of Inr.

TFIID complex is essential for Inr activity. TBP is insufficient.  DNase I footprinting studies with purified human TFIID and simple synthetic promoters containing only TATA and Inr elements showed weak TFIID interactions at Inr.  When Inr was disrupted by point mutation, a much weaker footprint was detected that was confined to TATA box. Disruptin TATA box eliminated the entire footprint except an enhanced DNse I cleavage site at Inr. Cooperative binding of TFIID was disrupted by increasing or decrasing TATA- Inr spacing by 5 bp.

Inr is essential for stable TFIID binding.

Mammalian TFs evolved to function with a broader range of Inr sequences.

TATA box
TATA box has a consensus of TATAWAAR, where upstream T is most commonly at -31 or -30 relative to A+1 (pr G+1) in Inr. Mutations in TATA box reduced activity of cellular and viral promoters.  TATA box is uncommon in vertebrates, only 1/3 or less of vertebrate core promoters are focused and only a fraction of focused core promoters contain a TATA box.

However, in S. cerevisiae, TATA boxes was 40-120 bp from start site.

TATA box is recognised and bound by TBP, a subunit of TFIID complex in eukaryotes. TBF-associated factors, TAF, are components of TFIID complex.  TAFs contribute to core promoter recognition, kinase activity  ubiquitin activating and conjugating activity, coactivator functions.

DNA-binding region of TBP folds into a saddle structure. Saddle consists of 2 quasi-symmetrical domains. N-terminal domain contacts 3' half of a consensus TATA box. C-terminal domain contacts 5' half. The concave surface of the saddle consists of 10 β-sheets. 8 of 10 sheets contact minor groove of duplex DNA. TBP bindings to minor groove relies on hydrophobic interactions. TBP induces kinks in DNA at 5' and 3' ends of TATA box. It partially unwinds duplex due to insertion of phenylalanine residues. Distorted DNA structure is restricted to the region that is directly contacted by TBP.

Start site selection in TATA-containing promoters.  Mechanism that determines distance from TATA box to transcription start site has been studied.  By swapping basal factors purified from S. cerevisiae and S. pombe, TFIIB and RNAPII were found to dictate this distance. When S. cerevisiae TFIIB and RNAPII were combined with other S. pombe basal factors, transcription initiated 40-120 bp downstream of TATA box. When S. pombe TFIIB and RNAPII were combined with other S. cerevisiae basal factors, transcription initiated 25-30bp downstream of TATA.  TFIIB was implicated as mutations in its N-terminal charged cluster domain shifts the location of start sites by a few nts in yeast and mammalian promoters.

In TBP-DNA cocrystals, TBP binds in a polar manner to asymmetrical TATA sequences, TATAAAAG and TATAAA.

BRE (TFIIB recognition element) is a TFIIB-binding sequence. TFIIB can bind upstream or downstream of TATA box at the BREu (upstream BRE, immediately upstream) or BREd (downstream BRE).  BREu consensus is SSRGCC. GREd is immediately downstream of TATA box. It has a consensus of RTDKKKK. BREu and BREd can act in either  a positive of negative manner.

TFIIB interacts with DNA in a sequence specific manner.  Archaeal TBP and TFB (archaeal homolog of TFIIB) bound cooperatively to T6 promoter when TATA box and upstream element were present. In presence of TFB, strong preference for purines 3 and 6 bp upstream of TATA, with weaker nt preferences at other positions. Recognition is BRE is mediated by a HTH motif at C-terminus of TFIIB. This motif is missing in yeast and plants. BRE may not contribute to gene regulation in these.

BRE represses transcription in vitro with crude nuclear extracts and in vivo in transfection assays. Repression of TFIIb-BRE was relieved when transcriptional activators bound distal sites. Tramscriptional activation increased.

CpG islands
CpG dinucleotide, a DNA methyltransferase substrate is under-represented in genomes of many vertebrates, 5-methylcytosine deaminates to form thymine, which is not repaired by DNA repair enzymes. 0.5-2kbp stretches of DNA possess a high density of CpG islands.  Human genome has 29 000 islands.

In mammals CpG islands are associated with half of promoters for protein-coding genes.

They lack consensus or near-consensus TATa boxes, DPE elements or Inbr. Have multiple transcription start sites that span 100 bp or more.  Mutations in vicinity of start site can cause use of alternative start sites but promoter strength is often unaffected.

Has multiple binding sites of Sp1. Transcription start sites are 40-80bp downstream of Sp1 sites. Sp1 may direct basal machinery to form a preinitiation complex.

DPE (downstream core promoter elements) is a downstream TFIID recognition sequence important for basal transcription activity. IT is required for binding of purified TFIID to a subset of TATA-less promoters. TFIID binds cooperatively to DPE and Inr.

DPE is conserved from Drosophila to humans. It is +28 to +33 relative to A+1 in Inr. DPE consensus is RGWYVT in Drosophila.

DPE functions cooperatively with Inr. Spacing between Inr and DPE is critical for optimal transcription. Single nt increase of decrease in spacing reduces TFIID binding and transcriptional activity.

Photocrosslinking studies show DPE is proximity to TFIID subunits TAF6 (TAFII60) and TAF9 (TAFII40) which contain histone fold motifs and are related to histones Hr and H3. It is possible that TAF6-TAF9 subunits of TFIID interfact with DPE similar to binding of histones H3-H4 to DNA in nucleosomes.

Similarities and differences between DPE and TATA box. Both are recognition sites for binding of TFIID. On the other hand, TATA box but not DPE can function independently of Inr. If TATA-dependent promoter is inactivated by mutation of TATA motif, core promoter activity can be restored by adding DEP at its downstream position.

An activity stimulates DPE-dependent transcription and repressed TATA-dependent transcription  This acivity is mediated by NC2/Dr1-Drap1, a repressor of TATA-dependent transcription.

MTE(motif ten element) has a consensus of CSARCSSAAC from +18 to +27 relative to A+1 in Inr in Drosophila.  Mutations of nt from +18 to +22 can abolish MTE actibity in vitro and in  cultured cells.

MTE functions cooperatively with Inr with strict Inr-MTE spacing requirement.  MTE addition compesnates for loss of basal transcription activity that occurs on TATA box or DPE mutation.

MTE shows synergy with TATA and DPE.  Synergy between MTE and other core promoter motifs inspired design of a Super Core Promoter.

DCE and XCPE1 motifs
DCE (downstream core element) has been found in human beta-globin promoter and characterised in adenovirus major late promoter. Occurs often with TATA box.  Appears to be distrinct from DPE.

DCE consists of 3 subelements: S1, CTTC from +6 to +11; SII, CTGT from +16 to +21; and SIII, AGC from +30 to +34.

XCPE1 (X core promoter element 1) motif is from -8 to +2 relative to +1 start site. It is in 1% of human core promoters, most of which are TATA-less.  Shows little activity by itself.  Acts with sequence-specific activators.

Binding of TBP to TATA box nucleates formation of pre-initiation complex (PIC) consisting of RNAPII, TFIIA, B, D, E, F amd H.  Tsai and Sigler (2000) crystallised a human TFIIBc-TBPc complex bound to an idealised and extended adenovirus major late promoter (MLP).

TFIIB makes spec base contacts in major groove of BRE and minor groove immediately downstream of TATA box. Upstream interactions are mediated by HTH motif. There are 5 ternary complexes linked.

Side chains of Lys 189 and Arg 193 of TFIIBc H bond with phosphate backbone at upstream end of TATA box and immediately downstream if ut, These bridging polar interactions between TFIIBc and promoter require deformation of TATa box. May stabilise ternary complex and reinforce TBP-induced deformation of DNA.

Human TFIIBc interacts with major groove upstream of TATA box through HTH motif. In this structure, Val283 makes vdW contacts with C5-C6 edge of C(-34'), 3 bp upstream of TATA box, in all 5 copies of ternary complex. A G:C bp 3 bp upstream of TATA box is conserved.  Mutagenesis indicates Cal 283 confers specificity for a G:C bp at position -34 of BRE. (Langrange 1998).

Side chain of mutational sensitive ARg 286 makes water-mediated contacts with G(-38) in 4 copies, G (-37) in 2 copies, and vdW with phosphoribose backbone of G(-38) in all 5 copies of complex.

A Role for TBP dimerisation in preventing unregulated gene expression by Jackson-Fisher
TBP dimerisation
TBP negatively autoregulates accessibility to promoter DNA in yeast through dimerisation. In absence of DNA, carboxy-terminal domain of TBP crystallised as a dimer, in which DNA-binding and dimerisation surfaces overlapped.

Jackson-Fisher et al (1999) applied site-directed mutagenesis at interfacial residues N69, V71 and V161. 2 alleles targeted at each site: Bulky charged substitution (Arg, expected to be severe) and a smaller potentially less severe mutation (Glu or Ser). A pulldown assay quantitated the binding of mutant or wild-type TBPs to immobilised GST-TBPs, which contains a wt yeast TBP C-terminal domain. Wt TBP retained on resin. Mutants was retained over a range of efficiencies. H bonds and hydrophobic interactions stabilised dimer structure.

Figure 1. Mutations in the Crystallographic TBP Dimer Interface Affect Dimerization(A and B) Space-filling model of the core yeast TBP/TATA complex and TBP dimers, respectively ( [7] and [24]).(C) Ribbon diagram of TBP showing the location of V161, N69, and V71 in a space-filling representation.(D) A ball and stick representation of the crystallographic dimer interface between one monomer and amino acids V161, N69, and V71 on the other monomer. A dashed line indicates an intermolecular hydrogen bond.(E) GST-TBP(181C) (10 nM) was immobilized on glutathione agarose and incubated with the indicated yeast TBP mutants (100 nM). The resins were washed several times, and the amount of bound TBP was determined by Western blotting. Assuming a 1:1 stoichiometry in the “heterodimer,” a maximum of 10 nM TBP is expected to be retained. For wild-type TBP, retention of 10%–20% of this level was observed, which is expected given the competition from homodimerization and the extensive washing steps involved.
TBP dimer instability correlates with activator-dependent transcription
If dimer association is rate limiting for gene expression, then dimerisation-defective mutants should stimulate expression of a basal promoter.  "Basal" is level of trasncription in the absence of activator function. Residual levels of monomer TBP would cause low level of basal transcription with wt TBP. This model predicts that TBP over-expression would largely increase dimer pool and marginally increase monomer pool.

If dimer dissociation is rate limiting without activator function, and activators accelerate rate-limiting steps, then activated transcription may not be positively affected by dimer instability.  Activators might discoate dimers. An activator might facilitate dimer dissociation through interactions with TBP, TAFs, chromatin remodelling complexes or components of Pol II holoenzyme.

At repressed promoters, dimer dissocation may not be rate limiting. Disscoation of a repressor might be primary step in initiating transcription complex assembly.

None of mutants could support cell viability at 22, 30 or 37C. Shuffling in wt TBP supported cell growth.

In vivo assay for basal transcription was adopted. Uses truncated ADH1 promoter fused to a lacZ promoter gene. Gal4-binding sites are upstream of this promoter. To maintain basal level expression, where Gal4 is repressed, cells were grown in glucose.

For activated trasncription, a fusion of entire PGK promoter to lacZ was used.  Reporter constructs were on high copy plasmids.  Test TBP was terminally HA-tagged and expressed from a low-copy plasmid using TBP promoter (low expression) or PGK promoter (high expression). TBP expressed under control of promoters is targeted to nucleus.

Correlation between decreased dimer stability in vitro and elevated basal levels of β-galactosidase activity in vivo at low and high levels of test TBP.  This trend was not apparent with activated transcription.

Table 1. In Vivo Transcription and TBP Levels in the Presence of TBP Mutants
Promoter of lacZ Reporter
DimerTBPBasal (low)Basal (high)Activated (high)Chromosomal Basal (high)Relative TBP levels (low)(high)
Dimerization data (column 1) represent relative binding to GST-TBP(181C) as determined in Figure 1E and are an average of six sets of data. See Figure 5 for standard errors. β-galactosidase activity (columns 3–6) is relative to wild-type test TBP in column 3 (1.0 = 3 Miller units using the high-sensitivity substrate CPRG) and represents an average of at least three independent determinations. All data were obtained in the linear range of the assay. The null allele expresses only the first 81 amino acids of TBP. “Test” HA-tagged TBP levels were determined by quantitative Western blotting with TBP antibodies as described in Figure 5B. TBP levels are relative to the endogenous (untagged) TBP (1 = 17,000 ± 2,000 molecules per cell). The standard error for all data is presented graphically in Figure 5E. Expression of the TBPEBmutants did not cause a decrease in the expression of TAFII145 (data not shown). “Low” in columns 3 and 7 indicates that the test TBP was driven by the TBP promoter; “high” in columns 4–6 and 8 indicates that the test TBP was driven by the PGK promoter.ND, not determined.
Strains harboring TBP(N69S) expressed at high levels grew very slowly.

Dimer dissociation is not rate limiting under these conditions.

Figure 2. A Model for How TBP Dimers Might Prevent Activator-Independent Transcription(A) Dimerization prevents TBP from binding to DNA. The little that does escape might give rise to low levels of basal transcription, occurring in the absence of a functional activator.(B) TBP dimer dissociation and DNA binding are facilitated either directly or indirectly by a functional promoter-bound activator, giving rise to high levels of transcription.(C) Dimer dissociation might not be rate limiting at repressed promoters

TBP dimer stability correlates with steady-state levels of TBP in vivo
To investigate if increase in basal transcription is caused by greatly increased expression of TBP mutants, the steady-state level of TBP was examined under the control of either TBP (low) or PGK (high) promoter. N-terminal tags on test TBPs cause slower mobilities in SDS PAGE, distinguishing it from endogenous TBP.

Under TBP promoter, expression level of mutants were below that of uniformly expressed endogenous wt TBP.  Highest expression levels were wt TBP and V71E, expressed at 40% level of endogenous TBP. Mutants (V161R and N69R) with highest β-galactosidase activity were present at lowesr levels. Elevated basal lacZ expression cannot be explained by increased expression of TBPEB mutants.

When test TBPs were expressed at high levels under PGK promoter, greater fluctuation in their steady-state levels whereas endogenous TBP levels remained uniform.  Wt TBP was expressed 14-fold higher levels than endogenous wt TBP. V161R and N69R were expressed at less than half level of endogenous TBP.  Strong correlation between dimer stability and in vivo levels of TBP. This suggests that TBP dimerisation stabilises TBP against degradation.

Strong correlation between dimer instability and specific activity of TBP. Higher dimer stability, lower specific activity of TBP. TBP dimerisation may be barrier to expression of uninduced genes.

Competition by TATA DNA
When yeast TBP was preincubated with increasing concentrations of TATA DNA, mutant TATA, or GC-rich DNA, only TATA efficiently inhibited crosslinking of TBP into dimers. When pure TATA and TBP are equimolar and at equilibrium, TBP prefers binding TATA to binding to itself 

General role for TBP dimerisation in preventing unregulated gene expression

TBP dimers might be general TBP status before promoter binding. Dimers block promoter binding and expression of uninduced genes. TBP dimersiation occludes its TFIIB-binding surface and may prevent TFIIB from associating with TBP until TBP is bound to DNA.

TBP Antagonists
Crystal structure of Negative Cofactor 2 recognising the TBP-DNA transcription complex by Kamada (2001)
Positive and negative cofactors interact with general transcription machinery. NC2 is the best characterised negative cofactor.  It inhibited transcription initiation by RNAPII by direct interactions with TBP-DNA complex in human tumour cells.

NC2 consists of 2 subunits, α: 22kDa and β: 20kDa.  They are conserved among eukaryotes.  AA sequences of α and β are related to histones H2A and H2B. They heterodimerise through N-terminal core histone-like regions.  In vitro, NC2 recognises TBP-DNA complex and inhibits incorporation of TFIIB and positive cofactor TFIIA, blocking PIC assembly.

NC2 was thought to be general negative regulator of transcription initiation.  NC2 is suggested to establish appropriate balance between positive and negative cofactors.  Inactivation of NC2 α in yeast shows limited positive effects. It has a positive role in transcription initiation controlled by DPE in certain D. melanogaster gene promoters.  It has been associated with PICs assembled on actively transcribing promoters in yeast.

Kamada determined X-ray structure of human NC2 recognising a preformed binary complex of human TBP bound to a TATA.  N-terminal portions of NC2α and β resemble H2A and H2B.

NC2α and β  form a heterodimer that binds underside of TBP-DNA. This allows C-terminus of NC2β to spec contact upper surface of saddle and block TFIIB entry. NC2 heterodimer acts as a molecular clamp, gripping upper and lower surfaces of TBP-DNA complex. NC2 and TFIIB cannot bind at the same time.

NC2 heterodimer binds major groove face of TATA element beneath molecular saddle. It makes protein-protein contacts with N-terminal stirrup of TBP.  Remainder of NCβ traverses up and over the saddle.

α helix H4 interacts with DNA backbone  α helix H5 interacts with upper surface of TBP C-terminal domain.  N-terminal half of F4 is stabilised by interactions with NC2α and within NC2β.  It is likely that H5 and remainded of H4 are random coil in absence of TBP-DNA.

NC2 and H2A/B show similar DNA binding properties.  The calculated electrostatic potential of upper surface of histonelike portion of NC2 heterodimre is highly basic. Positive electrostatic potential allows favourable polar interactions with negatively charged nucleic acid backbones bound beneath molecular saddle. NC2α makes salt bridges and water-mediated H bonds with backbone of top DNA strand upstream of TATA and with backbone of bottom DNA strand.  NC2β interacts with backbone of top DNA strand downstream of TATA element.  In TATA box, Lys 29 of NC2α projects into minor groove.

2 C-terminal α helices of NC2β support TBP and TFIID recognition.  Remained of NC2 molecular clamp consists of C-terminal α helices H4 and H5. Polar penultimate helix, H4 makes backbone contacts with both DNA strands on 3' side of TATA.  C-terminal helix, H5 lies on upper surface of TBP.

Residues 115-133 of H5 of NC2β participate in salt bridges, H bonds, and vdW contacts with TBP.

Structure of ternary complex makes it likely that C-terminal half of H4 and all of H5 are random coils in solution. It is suggested that NC2 binding occurs in 2 stages.Core histonelike heterodimer interacts with DNA on underside of TBP-DNA via electrostatic interactions. Structural consequences of target induced disorder-to-order transitions in C terminus of NC2β dictates directional NC2 binding.

It is presumed that Lys95, Arg 101 and Lys 102 contribute to this process by contacting backbones of both DNA strands downstream of TATA,. facilitating random coil to helix transition, giving rise to H4.

Transcriptional regulation by NC2. TBP-induced bend in core promoter is unaffected by NC2 binding. It seems to be recognised by core histone-like portion of NC2 heterodimer.  NC2 recognises TBP-TaTA complex, making protein-protien contacts with both TBP domains.  Steric clash between H4 of NC2 β and TFIIB.  NC2 blocks TFIA binding by steric hindrance. NC2 represses transcription initiation by acting as a molecular clamp that recognises TBP-DNA and physically blocks binding of TFIIB and TFIIA.

H5 supports molecular recognition of upper surface of TBP. H4 steric interference inhibits TFIIB binding and transcription initiation.

Transcriptional activators could bind certain promoters and target H4 and/or H5 of NC2 β. Disrupting interactions between NC2 and TBP-TATA could allow TFIIB to replace NC2 heterodimer.  This selectively relives repression of only those genes  Transcriptional activators and/or positive cofactors could bind to and alter conformations of  H4 and/or H5 of NC2 β. This allows productive TFIIB binding to an NC2-TBP-promoter complex. Histone-like portion of NC2 may remain bound to core promoter.

Kadonaga showed that NC2 can inhibit assembly of functional PIC at TATA and activate DPE-dependent transcription on separate DNA templates.  Activating transcription from a DPE-dependent promoter requires only H2A/H2B-like heterodimeric portion of NC2.

Albert 2010 Genome Biology
Albert et al (2010) did a genome-wide analysis on promoter association of human TFIIB and NC2 and correlated with gene expression and core promoter architecture.  TFIIB/NC2 ratio increased towards most highly expressed genes.  In 81% of TFIIB-dominated genes but in only 38% of NC2-dominated genes, at least one core promoter motif was present. TATA was most strongly enriched motif.  TFIIB strongly selects for TATA.

NC2 is more frequent on genes with multiple start sites lacking defined core promoter elements.

TFIIB/NC2 ratios are influenced by activators and core promoter elements.
Transcription complexes were assembled on a Gal4-responsive heterologous promoter template with wt and mutant TATA box in presence and absence of activator Gal4-VP!6. Activator enhanced TFIIB binding. NC2 did not respond to activator. Positive activator effect on TFIIB was stronger for mutant TATA than wt.

PIC (or TBP-TFIIB) association correlates with TATA or is independent of core elements altogether. NC2 association is largely independent of underlying core promoter structure.

NC2 occupancy shows a positive correlation of binding with TATA presence. It becomes negative relative to competing TFIIB.  TFIIB/NC2 ratios increase in strongly expressed 5% of B cell genes. Indicates negative role for NC2 at strongly expressed genes carrying intact core promoters.

TATA is positively correlated with TFIIB binding.  Genes with high TFIIB/NC2 ratio often carry G/C-rich regions.

High TFIIB/NC2 ratios select for promoters with focused start sites and conserved core elements. High NC2/TFIIB ratios correlate to multiple start sites promoters lacking defined core elements.

The TATA box regulates TBP dynamics in vivo by Tora
Tora proposes that TATA box role in RNAPII transcription assists rapid TBP dissociation from a subset of highly regulated PolII promoters. DNA bending on TBP binding coul release TBP rapidly from TATA by regulators. Eg NC2 and ATPases BTAF1 and Mot1p.

Regulation of TBP activity
NC2 binding to TBP-TATA blocks incorporation of basal trascription factors TFIIA and TFIIB intoPIC.  NC2 association induces structural changes in TBP-TATA. This allows DNA sliding and recognition of nonTATA sequences by TBP.

Mot1p in yeast cells is an ATP-dependent inhibitor of TBP-TATA complex formation.  MOT1 regulates pol II transcription posiitvely and negatively.  Mot1p-TBP has a high affinity for DNA but relaxed specificity for TATA.  NC2-TBP lacks prference for TATA sequences.  Mot1p and Nc2 collaborate to regulate TBP function in pol II transcription.  (See 30 and 39).

TATA plays a role in rapid TBP dissociation in vivo
DNA bending is suggested to play a role in functional distinctio of TBP between TATA-containing and TATA-less promoters. It is assumed that in vivo binding of TATA box by TBP or TFIID creates a bent DNA conformation.  Mot1p and Nc2 action would release strained conformation. This assists in rapid dissociation of TBP from TATA.

TATA would contribute to rapid dissociation of TBP and possibly PIC.  Single TATA box mutations reduce TBP-induced bending.  These promoters may not release TBP as rapidly as TATA=containing promoters on action of Mot1p and NC2.

Figure 2. Bending of the TATA-box acts as a ‘spring’ for rapid release of TBP. (a) Binding and rapid release of TBP from TATA-box containing promoters (i). Upon binding to TATA the DNA adopts a bent conformation (ii). NC2 approaches the TBP–TATA complex from the underside and BTAF1 (Mot1p) from the top (iii). BTAF1 (Mot1p) and NC2 are directly involved in dissociation of the protein–DNA complex (indicated by dashed arrows). The bent conformation of the TATA box contributes to the rapid release of TBP (iv). (b) TATA-less promoters carrying broad and dispersed initiation sites. TBP can bind to different positions on these promoters, but for clarity, only a single TBP is indicated. TBP binding does not result in bending of promoter DNA (i, ii). Therefore, release of TBP by the combined action of BTAF1 (Mot1p) and NC2 is less rapid (iii, iv). It is important to stress that although the bent arrows emanating from the DNA indicate transcription start sites, they do not reflect transcriptional activity.

Non-optimal TATA elements exhibit diverse mechanistic consequences
Certain off-consensus TATA elements form poor binding sites for TBP. This interferes with complex formation with TFIIA and /or TFIIB. Recruitment step of TBP affected.

Other motifs (CATAAA and TATAAG) do not affect initial formation of TBP, TFIIA-TBP or TFIIB-TBP complexes. But unable to form stable TFIIA-TBP-DNA complex in vitro.

Single Cytosine of Guanine Replacements in TATAAA element diminish gene expression in vivo
Less induction of β-galactosidase for mutants when induced by Gal4.

Internal base substitutions in TATA sequence can disrupt TBP binding.
Each of the C and G series elements was tested for ability to form TBP-DNA complex.  Abilities to form TBP-DNA complexes varied for each.  Some were disrupted.

Higher Order complex formation also depends on sequence of core element.
TFIIA-TBP-DNA complex formation was decreased.  Similar trend for TFIIB-TBP-DNA complex formation.

TFIIA-TBP-DNA complex is destabilised on TATAAG element
Electrophoretic mobility shift assays measured reltaive binding and stability of TFIIA-TBP complex on TATAAA and ATAAG. TFII-TBP was stable on TATAAA. TFIIA-TBP on TATAAG showed great loss of complex.

TFIIB-TBP-DNA behaved similarly on both TATAAA and TATAAG. TBP binds TATAAA amd TATAAG in a manner that allows stable association of TFIIB-TBP higher order complex and difference in stability of TFIIA-TBP-DNA is specific to complex containing TFIIA.

Stability of TFIIB-TBP-DNA complex depends strongly on sequence of TATA element
Fusion of TBP and TFIIA results in an increase in expression from CATAAA and TATAAG in vivo.
Used yeast strains expressing TFIIa-TBP and TFIIB-TBP as fusion molecules. TFIIA-TBP fusion yeast strains containing CATAAA and TATAAG showed dramatic increase in activity to 30% of TATAAA activity. This increase is specific to TFIIA. No such increase in TFIIB-TBP fusion strain.

Unexpected roles for core promoter recognition factors in cell-type-specific transcription and gene regulation by Goodrich (2010)
TFIID subunits function in most RNAPII promoters in higher eukaryotes.

RFIID consists of TBP and 13 or 14 TAF subunits. It binds core promoter DNA through multiple subunits (TBP, TAF1, 2, 6, 9).

TBP subunit binds TATA bozes. Several TAFs bind promoter elements downstream of TATA box. TAF1 and TAF2 bind initiator element, which spans TSS.  TAF6 and TAF9 bind the downstream promoter element. TAF1 is in close proximity to downstream core leement when TFIID is bound to promoters.

Some TAF subunits are also targets of transcriptional activators. This allows TFIID to integrate signals from activators to core promoter.

Multiple subunits of the transcription factor IID (TFIID) complex bind core promoter elements12. TATA-box-binding protein (TBP) binds TATA boxes. TBP-associated factor 1 (TAF1) and TAF2 bind the initiator element (Inr). TAF6 and TAF9 bind the downstream promoter element (DPE).
Pathology of TBP mutations
Spinocerebellar atazia 17 (SCA17) and Huntington's disease-like 4 (HDL-4) by Stevanin
Autosomal dominant cerebellar ataxis are characterised by progressive cerebellar ataxia. Uncoordinated movements, unsteady gait, dysarthia.

15 genes and mutations have been identified. Translated (CAG)n/polyglutamine repeat expansions are responsible for disease caused by 7 of these genes, SCA 1-3, 6, 7, 17 and DRPLA.

A rare neurodegenerative disorder belonging to the polygluatminopathy group.  Can lead to death. Number of glutamines observed in pathological proteins varies from 21 to more than 400. Phenotype usually manifests above a repeat number between 35 and 40.

SCA17 was first reported in a complex neurological disorder with cerebellar ataxia. Patient carried 63 trinucleotide repeats in gene encoding TBP on chromosome 6q27.

Size and structure of normal repeat in TBP gene
N-terminus of TBP contains a long stretch of glutamines. Repeat is impure nd is encoded by 3 CAG stretches, interrupted by 1 to 3 CAA codons. Qstretch is polymorphic in normal populations. Normal range is between 25-42 residues.  Most alleles contain 32-39 repeats.

SCA17, a novel autosomal dominant cerebellar ataxia caused by an expanded polyQ in TBP by Nakamura
IC2 antibody screening was used to identify polyQ tract in lymphoblastoid cell lines.  Antibody recognises proteins of largely expanded polyQ tracts. A polyQ protein of 49 kDa was found in one proband.  Its MW is similar to TBP. Might be mutant TBP, as the monoclonal IC2 Ab was originally raised against TBP.

Using primer pairs flanking CAG/CAA repeat of TBP gene it was found that proband and his affected sibling had same repeat expansions. Western blot confirmed identity of 49kDa band using monoclonal anti-TBP antibody.

The CAA/CAG repeat numbers of TBP genes were determined in two affected siblings and their mother.  CAA/CAG repeat in healthy mother was 37/39. In affected siblings was heterozygous, 37/55 and 39/55.

Number of CAG/CAA repeats in TBP gene ranged from 29 to 42 in healthy control chromosomes. A TBP gene having a CAG/CAA number in excess of 43 is pathological.

Immunocytochemical study of post mortem brain tissues
Immunocytochemical examination of post mortem brain tissues from a patient who had expanded polQ repeats in his TBP. Shrinkage and moderate loss of small neurons with gliosis in caudate nucleus and putamen.

Immunocytochemical analysis was performed with anti-ubiquitin and anti-TBP antibodies. Showed neuronal intranuclear inclusion bodies. Most neuronal nuclei were diffusely stained with 1C2-Ab. None were stained in healthy control brains. Neuronal inclusion bodies may provide marker of disease process in many polyQ diseases.

Polyglutamine domain modulates the TBP-TFIIB interaction: implications for its normal function and neurodegeneration by Friedman
PolyQ regions are present in many eukaryotic TFs inc vertebrate TBPs. Length of polyQ varies among vertebrates species. It is higher in mammals. Friedman (2007) showed that expanded polyQ tract in TBP influences its dimerisation and interaction with TFIIB.

They generated transgenic SCA17 mice expressing mutant TBP. cDNAs encoded human TBP with polyQ tracts of different length. PolyQ-expanded TBPs were preferentially detected with polyQ-spec antibody 1C2.

Immunofluorescent staining showed that transfected TBPs localised mainly to nucleous.  TBPs with pathologic polyQ tracts formed nuclear aggregates.

Neurological phenotype and neuropathology of SCA17 mice
Abnormal posture in TBP-105Q mice. All SCA17 mice had reduced lifespan.

Neurodegeneration in transgenic SCA17 mice
Immunogold labelling showed that TBP-105 sformed nuclear inclusions in cerebellar granule neurons. Degenerating Purkinje cells and degenerating axons.

Alteration of TBP function caused by polyQ expansion.
TBP forms transcriptionally incompetent dimers to prevent unregulated gene expression. Length of polyQ domain was inversely related to relative amount of TBP dimers. Dimer-to-monomer ratio was lower for polyQ-expanded TBPs than normal TBP in transfected HEK293 cells and HT22 hippocampal cells.

They generated polyQ-expanded TBPs with an internal deletion spanning part of carboxy-terminal dimerisation domain. Double mutants could not dimerise but formed aggregates.

After cotransfection of TBP-31Q ant TBP-71G, a heterodimer was detected on western blots between 2 homodimers. Incrasing expression of TBP-71Q depleted level of TBP-31homodimer.  PolyQ formation can reduced formation of TBP homodimerisation in cis and in trans. Decreased dimerisation by mutant TBP may precede and facilitate its aggregation.

Enhanced interaction of polyQ-expanded TBP with TFIIB
Almost all TBP inclusions in TBP-105Q cerebellar sections contained TFIIB. GST-TFIIB pulled down more TBP-71Q than TBP-31Q. TFIIB pulled down more solutble TBP-105Q than TBP-13Q.

If mutant TBP abnormally binds TFIIB to affect gene transcription, TFIIB overexpression should reduce mutant TBP-induced toxicity  TFIIB was overexpressed in cultured cerebellar granule cells from TBP-105Q mice.  GFP transfection visualised integrity of neurites of individual neurons. More neurons expressing only GFP showed disrupted neurites than those cotransfected with TFIIB. TFIIB overexpression had protective effect.

Polyglutamine expansion reduces the association of TBP with DNA and induces DNA binding-independent neurotoxicity by Friedman
PolyQ expansion reduces in vitro binding of TBP to TATA box DNA.

N-terminal TBP fragments harbour the expanded polyQ tract. It lacks an intact C-terminal DNA-binding domain. They are present in transgenic SCA17 mouse brains.

polyQ-expanded TBP cannot bind DNA. It formed nuclear inclusions and caused severe neurological phenotype in transgenic mice.

PolyQ-expanded TBP is inhibitory to TATA_dependent transcription when it cannot bind DNA productively. It can induce neurotoxicity independent of DNA binding.

PolyQ expansion inhibits TBP binding to DNA
TBP N-terminus is normally antagonists to formation of stable TBP-TATA box complex. They compared in vitro interaction of normal and polyQ-expanded TBP with TATA box DNA. Recombinant TBPs containing either 31 or 71 Qs were incubated with a radiolabelled probe containing TATA box. Mutant TBP shifted less probe than normal TBP.

Truncated, polyQ-expanded TBP forms nuclear aggregates that sequester TFIIB
Double immunolabelling showed that truncated TBPs with expanded polyQ sequestered endogenous TFIIB in nuclear aggergates. Truncated polyQ-expanded TBP could still interact aberrantly with TFIIB.

As expanded polyQ can inhibit association of TBP with TATA, it is possible that interaction of TBP C-terminus with DNA may be regulated by length of N-terminal polyQ domain.

It is expected that expanded polyQ suppresses gene expression. Full-length mutant TBP stimulates a TATA-dependent transcription reporter in culture cells. Reducing intrinsic DNA binding ability of TBP may not be sufficient to inhibit TBP-mediated transcription.

Truncated polyQ-expanded TBP which is unable to bind DNA is more toxic than full-lenth polQ TBP. Truncatedd TBP inhibited TATA-dependent transcriptional reporter activity. Level of inhibition increased with polyQ tract length.

Truncated polyQ TBP formed more nuclear aggregates than fulllength mutant TBP in transfected cells. As aggregates formed by truncated mutant TBP can also sequester TFIIB, it is likely that mutant TBP becomes more toxic when it cannot bind DNA productively but can still associate with other TFs. Continued interaction with TFIIB, which is strengthened by poly-expansion, might undermine assembly of PIC and inhibit transcription activity.

Friedman proposes a model for molecular pathogenesis of SCA17. When TBP is soluble, a long polyQ tract may antogonise dimerisation, a mech of negative autoregulation.  It may enhance interactions with other TFs eg TFIIB. This facilitates TBP recruitment to promoter DNA. Enahnced protein-protein interations compensate for reduced DNA binding ability of TBP. However, polyQ causes TBP to misofld and aggregate over time. TBP aggregates presumably cannot bind promoter DNA but can still interact with TFIIB and/or other TFs. This could negatively affect gene transcription by reducing promoter occupancy of affected proteins.
Model for mutant TBP-mediated transcriptional dysregulation in SCA17. Increased interaction of soluble polyQ-expanded TBP with TFIIB and/or other transcription factors may allow for recruitment of the former to certain TATA-containing promoters. Because this soluble form of TBP is not inherently defective, its recruitment can stimulate TATA-dependent transcriptional activity (upper panel). However, mutant TBP can no longer productively interact with promoter DNA after proteolytic processing and/or misfolding. TBP aggregates can sequester particular transcription factors, such as TFIIB, and thereby reduce their availability at certain promoters. Decreased transcriptional activity is the likely consequence of aberrant interaction of misfolded TBP with basal transcription factors or activator proteins (lower panel).

TBP, a polyglutamine tract containing protein, accumulates in Alzheimer's disease by Reid
AD is characterised by protein cleavage, protein misfolding and protein accumulation. Results in abnormal deposition of βA plaques and NFTs throughout brain. Major component ofNFT is filamentous aggregates of hyper-phosphorylation Tau protein.

Aberrant TBP associates with neurofibrillary structures in AD brain.  Polyclonal TBP antibody showed disease specific staining in cortical regions of MTG sections.

Insoluble βA was detected in all AD samples and at low levels in 3 of controls. Insoluble Tau was detected in disease samples.

TBP and Tau colocalise in most tangles studied.

Western analysis showed a rapidly migrating form of TBP at high levels in some disease brains. This may be truncated form of TBP. As the epitope of pTBP is at N-terminus, C-terminal cleavage might occur. Smaller product might however be caused by conformational change causing rapid migration through gel.

A splice variant of TBP encoding polyQ-containing N-terminal domain that accumulates in Alzheimer's disease by Reid
Search of human EST databases identified TBP splice variants with C-terminal truncated ORFs. A variant ORF encodes N-terminal domain of TBP including polyQ tract and terminates at an alternative stop codon conserved in most tetrapods. This splice variant encodes TBP fragment observed in disease brain.

Vertebrate TBP N-terminal domain is encoded by exons 2 and 3 including polyQ domain. Identified splice transcript encodes an ORF which skips exon3 3 splice acceptor site. Terminates at variant stop site in intron 3 immediately following splice acceptor site.

Fig. 1. Schematic diagram of the human full-length TBP and splice variant TBPv3 transcripts. The predominant TBP isoform is encoded by eight exons and is 339 aa in length, assuming that the polymorphic polyglutamine (polyQ) region has 38 glutamine residues. A black box denotes the N-terminal polyQ tract and arrows represent the C-terminal domain direct repeat. The location of a putative nuclear localisation signal (NLS) in the TBP mRNA is indicated by a cross-striped box. TBPex3 encodes a C-terminal truncated isoform of 165 amino acids. Exons and introns are not drawn to scale.

Fig. 2. Vertebrate TBP exon three splice acceptor sites and predicted truncated proteins. (A) Nucleotide and deduced amino acid sequences of conserved the 3′ splice acceptor sites of TBP exon three in vertebrate genomes in the current public database. Exonic nucleotides are shown in upper case and intronic nucleotides in lower case. The splice acceptor sites are underlined and amino acids encoded by potential intron-skipping transcripts are shown in bold text. (B) Alignment of deduced amino acid sequences of TBP variant TBPv3 from human, mouse and bovine (verified) and other vertebrates (predicted). Sequence accession numbers/loci: Homo sapiens (human), NM_003194; Pan troglodytes(chimp), NM_001104607; Equus caballus (horse), LOC100049846; Macaca mulatta (macaque), LOC696258; Canis familiaris (dog), LOC611193; Bos taurus (bovine), NM_001075742; Rattus norvegicus (rat), NM_001004198;Ornithorhynchus anatinus (platypus), LOC100091633; Mus musculus (mouse), NM_013684; Monodelphis domestica(opossum) LOC100028736; Gallus gallus (chicken), NM_205103; Danio rerio (zebrafish), NM_200096.
Human TBPex3 variant is transcribed in diverse adult tissues.

TBPex3 encodes a protein with low electrophoretic mobility. Western analysis shows that MW for full-length uman TBP with 37 repeat polyQ tract is 37.5kDa. Under SDS PAGE it travels at 42 kDa. TBPex has a predicted MW of 17.9kDa, but travels at 29kDa.

TBPEx protein accumulates in AD.  It was detected at higher levels in AD middle temporal gyrus tissue than in control middle MTG tissue.

Cellular distribution and localisation of TBPex3 protein.
Immunocytochemistry showed diffuse nuclear TBP labelling in all full-length TBP transfected cells. TBPex3 expression resulted in extranuclear and perinuclear inclusions. In many cases cell nucleus was distorted.

This indicates that TBPex3 has a greater tendency to aggregate in vivo and that TBPex3 can cross nuclear envelope and pass through nuclear pore.

Rational design of a super core promoter that enhances gene expression by Juven-Gershon
Super core promoter 1 (SPC1) directs high amounts of transcription by RNAPII in metazoans.  SPC1 contains TATA box, Inr, MTE and DPE in a single promoter.  It is stronger than CMV IE1 and AdML core promoters in vitro and invvo. Each of 4 promoter motifs needed for full SCP1 activity.

Strong synergy between MTE and TATA and between MTE and DPE.

SPC1 comprises sequences form -36 to +45 relative to +1 Adenine in Inr.

TATA box is from CMV IE1 core promoter. A cp,[psote Onr based on sequences on AdML and D. melanogaster G retrotransposon core promoters. MTE from D. melanogaster Tolllo core promoter. DPE from Drosophila G core promoter. BRE was not included as it has been reported to have positive and negative effects on transcription.

Figure 1. SCP1 is stronger than CMV and AdML core promoters in vitro.
(a) Diagram of SCP1. KrD. melanogaster Krüppel gene; G, Drosophila G retrotransposon. The nucleotide sequence is given in Methods. (b) SCP1, CMV and AdML core promoters (each of which contains their respective sequences from -36 to +45 relative to the +1 start site cloned into pUC119) were subjected to in vitro transcription analysis (in duplicate reactions) with a standard HeLa transcription extract21. The resulting transcripts were detected by primer extension. To test whether transcription was catalyzed by RNA polymerase II, alpha-amanitin (4 mug/ml) was included, as indicated. (c) Transcription reactions were performed as in b with the indicated amounts of template DNA in a volume of 50 mul.
SCP1 is stronger than CMV core promoter or AdML core promoter in vitro. Transcription from SCP1 is inhibited by α-amanatin so transcription from SCP1 is catalysed by RNAPII.

SCP1 is transcribed more efficiently than other core promoters. Rate of PIC assembly was higher in SPC1 than in other core promoters. Higher template usage (40%) in SCP1 than in CMV (15%) or AdML (6%) core promoters.

Purified TFIID binds to SCP1 with high affinity
DNase I footprinting showed TFIID binds SCP1 with at least 10-fold higher affinity than to CMV or AdML core promoters. Unlike TBP, TFIID binds TATA box weakly. Mutations in TATA box, Inr, MTE or DPE in SCP1 variants decreased binding of TFIID. All core promoter motifs contribute to TFIID interaction with SCP1.

SCP1 is a strong core promoter in vivo
SCP1, CMV and AdML core promoters were subcloned into a promoterless, enhancerless vector with a leciferase (luc) reporter gene. Constructs were transfected into HeLaS3 cells. Luciferase activity was measured. SCP1 showed several-fold higher levels of luciferase than either the CMV or AdML core promoters.

SCP1 increases level of enhancer-driven transcription
SCP1, CMV and AdML core promoters were inserted intopGL3-enhancer luc reporter vector which contains SV40 enhancer   SCP1 is 3fold more actibe than CMV or AdML core promoters with SV40 enhancer.

SCP1 directs accurate initiation of transcription in vivo
HeLa cells were transfected with pGL3-enhancer constructs containing SPC1, CMW and AdML core promoters. RNA was isolated from cells. RNA was subjected to primer extension analysis. Cells were co-transfected with a β-galatosidase (lacZ) reporter plasmid. SCP1 mediates transcription from A+1 site in Inr motif. Amount of RNA producd in SCP1 was higher than from CMV or AdML core promoters.

No comments:

Post a Comment