Eukaryotic transcription, post-transcriptional ctrl Flashcards
(22 cards)
Archaea and transcription
NB: exceptions to faithful genomic DNA copying- mature B and T cells (editing events), red blood cells (lose nucleus).
Aside on archaea: many in inhospitable env.s, exist widely in biosohere, ~30% world biomass. Morphology+ cell division aparatus similar to bacteria, circular genomes. No complex cytoskeleton/ predominant organelles.
Transcription: 5’-3’, read off template 3’-5’. Start site +1, upstream seq.s (-1 and back) have proximal promoter seq.s and distal enhancer elements. Promoters can also be In coding region (e.g., fo snRNAs).
DNA enters RNAP, strands melt, DNA and nascent RNA exit via different channels. Exp evidence-> transcription in higher euk happens in discrete sites (“transcription factories”) – need to release superhelices ahead+ behind RNAP with topoisomerases.
RNAPs and bacterial RNAP
RNAPs are widely conserved: bacterial has 2x alpha, beta+beta’+ associated sigma for sequence targeting. Euk and arch have more subunits, large subunits related to bacterial in seq, smaller subunit seq.s in euk related to arch. Arch have histone-like proteins-> nucleosome- like structures.
Bacterial RNAP: sigma factors bind RNAP-> “holoenzyme”. Bind DNA seq-specifically (target RNAP). Each sigma factor regulates a set (battery) of genes. E.g., sigma70- promoter elements from -10 to -35 of many genes. Others- global change e.g., during sporulation, heat shock, etc.
Euk RNAPs and accessory factors
Euk RNA Pols: NB mitochondria and chloroplasts have own RNAPs of bacterial/bacteriophage ancestry.
Pol I: rRNA precursors, nucleolus. rRNA=~ ½ cellular RNA. Genes generally in tandem arrays (~150-200 copies)
Pol II: protein-coding mRNA, some snRNAs (e.g., U2). Transcripts can be many 100s/kb. Highly regulated at initiation, also ctrl @ other steps. Transcripts modified by capping+ splicing (occasionally also for Pol I+III), PolyA (likely coupled with splicing, transcription+ mRNA export).
Pol III: small stable RNAs (t, some sn, 5S rRNA). Genes scattered (not arrays)
Accessory factors: recognise promoter DNA, regulate activity, elongate processively, transcribe accurately. General/basal factors incl TFIIB, TFIID, Pol. Usually specific to one Pol. Seq-specific DNA binding regulatory proteins ctrl rate of machinery assembly/activity.
Promoters and proteins, in vivo and in vitro transcription assays
Promoters and proteins: RNA Pol II requires general TFs (TFII) to initiate; promoters frequently have short+ conserved seq, e.g., TATA box (50% in human- AT rich octamer ~25bp upstream); downstream promoter (DPE) common when no TATA. Core Pol II promoter often includes TATA/Inr+DPE.
Biochemical (cell-free) transcription assays (ID basal/general TFs):
In vivo: tissue culture/animal tissues-> nuclear extract; add DNA template+ ribonucleotides (some radio); incubate (transcription complexes assemble, transcription); autoradiography. Can fractionate extract over column- hard to purify as present at low levels, affinity chromatography could help.
In vitro: (allow detailed mechanistic studies, but only DNA elements fairly close to start site have detectable activity, as DNA is naked, unlike in vivo, so chromatin effects not fully recapitulated- in some systems factors work over kbs). Complement with in vivo and structural studies.
TFIID, A and B
TFIID: nucleates assembly of other TFs, helps determine start site. Multi-protein complex, incl TBP that binds TATA in minor groove; highly conserved (C-terminal repeats of Human and Drosophila 87% identical), contains 2 direct repeats forming “molecular saddle”, bending DNA; “universal” TF (also needed for Pol I, III). TFIID also has TBP-associated factors (TAFs) that bind promoter seqs (esp important when no TATA), contacted by upstream seq-specific activators, some have histone-like folds. “Variant” TFIID complexes with different TAFs- may have tissue-specific characteristics.
TFIIA: stabilises TFIID/DNA interaction by binding
TFIIB: Bridges TFIID and Pol II. Some seq preference for DNA binding (weakly conserved BRE element upstream of TATA). Help define start site. N-terminal region interacts Pol II+ upstream regulators. Rest of protein= 2 direct repeats.
PolII, TFIIF, H
Pol II and TFIIF: enter complex together. TFIIF= 2 proteins, weak seq similarity to sigma factors. TFIIE enters.
TFIIH: helicase activity, help promoter melting using ATP. DNA nt excision repair. NER targeted to template strand of transcribed regions. Kinase activity Pis Pol II CTD (largest subunit on Pol II, 52 repeats of consensus YSPTSPS in humans) conserved across euk. No CTD equivalent in prok/ arch/Pol I,III. May ctrl processivity+ link transcription to splicing, PolyA and export.
Ctrl Beyond transcription: RNA processing, transport and localisation, translation ctrl, mRNA degradation, protein activity.
Transcription activators and repressors
Activators/repressors bind specific promoter seqs, then regulate transcription via activation/repression domains, generally modular in function.
Direct recruitment of basal TFs (change rate of assembly+ activity of apparatus)
Indirect recruitment of basal TFs via co-activator/co-repressor (some serve as physical bridge activator to TF)
Change chromatin structure: Histone mod enzymes. Repressors can oppose recruitment (as above).
Can influence other events, e.g., Pol release. May allow faster, synchronous transcription ctrl and lower “background noise”. Pause site often downstream of DPE- paused Pol II tends to have Ser-5 on CTD Pi’d; when released into elongation, Ser-2 also Pi’d.
TAFs and TFIID, mediator complex and multiple TFs
TAFs in TFIID: TBP and TFIID work equally well in vitro in activator absence. TFIID > TBP in response to some activators (which interact with different TAFs-> combinatorial ctrl needed+ produced).
Other TFs contacted by other factors, e.g., VP16 (viral transcription activator) binds TFIIB, change conformation, aid Pol II recruitment.
Mediator complex: co-activator that brings together GTFs, RNAPII, TFs. Well-conserved ~20 polypeptide complex, functions overlapping TFIID. SAGA chromatin modifying complex that acetylates and de-ubiquitinates histones also conserved well, needed by many activators, has similar function.
Multiple TFs/promoter-> combinatorial response- “barcoding” of gene expression-> complexity despite only having ~1000 TFs in humans.
Modulating intrinsic activity of a factor
Modulate levels of the factor, e.g., HNF-1 directs liver-specific gene exp. Only found in liver- addition to, e.g., spleen results in liver-specific gene transcription
Modulate intrinsic activity:
* Ligand binding, e.g., steroid hormone receptors have DNA-binding Zn finger domain+ ligand-binding domain. Hormone membrane-permeable, binding of ligand causes conformational change that relieves repressive influence, allowing transcription. (e.g., GR binding glucocorticoid releases GR from HSP90 protein, exposes GR nuclear localisation seq and allows it to enter nucleus+ bind target genes)
E.g., thyroid-hormone receptor (TR)-RXR: if no ligand, complex binds “SMRT” HDAC, repressing transcription. Ligand binding TR-> SMRT dissociation, CBP/p300 recruitment, HAT activity.
Some other steroid hormone receptors- protein constitutively in nucleus, ligand binding activates it.
Modulating TFs through Pi
- Phosphorylation: e.g., p53 normally unstable. DNA damage-> p53 Pi’d by ATM and Chk2 kinases-> dissociation of negative inhibitor Mdm2 (usually targets p53 for degradation) so DNA damage-> p53 levels rise, p53 binds target genes (regulate apoptosis and cell cycle ctrl), activates transcription. Targets regulate apoptosis (Bax), cell cycle ctrl (p21) etc.
Chromatin structure and activity modulation
Chromatin structure: Dampen transcription by preventing protein binding. Histone N terminal domain mods alter compaction: HATs generally open chromatin (can be viewed as co-activators), HDACs undo this (co-repressors) by changing charge. Methyltransferases (SET, PRC2) used for genetic imprinting+ long-tern silencing but can also be activating, countered by demethylases. Chromatin remodelling complexes use ATP to mobilise nucleosomes, allowing them to slide around (SWI/SNF complex, ISWI complexes. Co-activators)- subunits of SWI/SNF mutated in >20% human cancers.e.g., RB+ E2F defective in retinoblastoma. RB inhibits E2F, recruits HDACs. As cells enter S-phase, RB Pi’d, dissociating it from E2F, which can now bind HATs, induce transcription of S-phase genes.
DNA foot printing+ ChIP seq provide evidence for protein binding/association (see BMB)
Regulatory networks and evolution
Some cases involve hierarchical transcriptional cascade, e.g., TF MyoD expression can drive mammalian cells to differentiate along muscle lineage, 4TFs sufficient to reprogramme somatic-> iPS (Yamanaka). Often regulation more so a network.
Cis elements can evolve faster than proteins, have important role in evolution (Tc1 mice have human chr 21 added (Down’s)- for a set of TFs in the liver, DNA-binding pattern is as in humans- suggests DNA seq wins over proteins in regulation)- still unclear if TF structure changes to recognise different elements or cis elements change to alter TF binding, though.
Also far-off enhancers that loop-round to promoters (->enhanceosome- arrangement of factors. Billboard: mostly inf=dependent elements that loosely collaborate) (esp important in higher eukaryotes) plays important role, as DNA looping also regulated by “insulators” like CTCF:
CTCF transcription factor binds DNA elements, can homodimerize. Interacts w/ cohesins, CTCF binding sites can reinforce looping. Binding is regulated (e.g., by chromatin state). CTCF can insulate genes from enhancer elements. Exp to study looping include HiC+ ChIA-PET.
CHIA-PET
CHIA-PET Chromatin Interaction Analysis by Paired-End Tag Sequencing combines Chip-based methods and 3C. resolves the issues of non-specific interaction noise found in ChIP-Seq by sonicating the ChIP fragments to separate random attachments from specific interaction complexes.
Crosslink protein+ DNA, sonicate to break up chromatin+ reduce non-specific interactions. Purify w/ antibody. Split into 2 aliquots+ add different linkers to each; elute and ligate at low concentration (proximity ligation); digest with restriction enzymes to create PETs with 2 of the same linker or different linkers. Sequence each type and analyse/map computationally.
Footprinting and ChIP seq
Foot-printing: 32P label 1 strand of dsDNA, divide reaction mix-> 2 tubes. Add binding protein to 1. Limited nuclease digestion (DNAse I, 1 cut) both. Analyse sizes of resulting cleaved ssDNA by autoradiography. Compare 2 reactions on denaturing gel (find protected binding site).
ChIP(seq): provide direct evidence for binding sites. Crosslink chromatin+ DNA binding proteins in live cells (formaldehyde). Shear DNA, purify TF with associated DNA (antibody on beads). Remove TF by digestion. Amplify DNA- can couple to microarray/seq analyses.
Hi-C and 3C
Hi-C: crosslink+ cut chromatin, repair and biotinylate ends, ligate, shear DNA + pull down biotinylated DNA, seq (Illumina) to find TADs. Produce Hi-C maps. Bold diagonal can be used as proxy reference for 100% interactions. Spots mark base of loops, dark dashes mark where one end of loop can slide in and out, altering loop length. H3K4 Me1 (for enhancers) peaks in TADs, Me3 line up with gene in TAD peak.
3C: in vivo, formaldehyde crosslink DNA, apply antibody+ purify on column, elute w/antibody, protease. Ligate and de-crosslink, biotin tagging the connection between the two previously separate stands. Enrich w/ streptomycin (recognises biotin), ligate on adaptors, seq.
Promoter capture Hi-C (3C-based) focuses on interactions at promoters (many short distance interactions+ some w/ distant enhancers).
Position-effect variation
Position-effect variation (PEV): chromosomal domains can inherit their chromatin state through mitosis, esp constitutive heterochromatin, e.g., @ centromeres. Boundaries eu/heterochromatin not precise-> seen in flies: if white gene translocated close to boundary, silences in some cells-> white patches in otherwise red (wt) eye, suggests that once White gene exp/not exp, daughters maintain this state. Initial boundary setting random, then maintained. Genetic screens for PEV modifiers-> chromatin regulators e.g., Su-Var genes.
Polycomb
Polycomb: gene exp via TFs/enhancers can be coupled to cellular memory, maintains cell identity. Spatially-restricted exp of TF set-> body plan info. Memory/gene exp maintenance through many mitoses by complex chromatin modification/maintenance-based mechanisms involving Trithorax+ Polycomb pathways.
TFs action through enhancers dependent of cell localisation in body. Polycomb acts through Polycomb Response Element (PRE). Enhancer provides spatial info (needed in early development), PRE provides memory (maintains memory even in TF absence). (NB Hox gene exp in respective segments can be visualised w/FISH)
Methylation and its heritability
Methylation: often lost through (+has role in) evolution. Genome defence, regulation. Some methylation easily maintained, e.g., CpG, i.e., DNA context methylation symmetric, both strands retain memory through replication. In vertebrates, maintenance by DNMT1, de novo methylation by DNMT3 family. Can be actively removed by oxidation by TET family.
Methylation heritability: If chromatin passes through female line, female eggs will be methylated even if chromatin paternal. If pass via male line, all germline cell unmethylated, so all unmethylated in males+ vice versa in females. When genes differentially imprinted, embryo survival may depend on whether functional allele provided by parent with unmethylated gene (so imprinted gene heterozygote survival depends on direction of cross). Imprinted genes occur in clusters, may depend on local ctrl site where de novo methylation occurs, unless prevented,
e.g., IGF2- need 1 copy to be meth and other not. Maternal methylated, enhancers downstream of H19 activates IGF2 exp. Paternal unmethylated, CTCF binds, enhancers activate H19 exp instead. IGF2 disregulation-> disease states like cancer.
miRNAs
miRNAs: bioinformatics-> up to 30% human genes regulated by 100s of miRNAs in humans (each miRNA potentially targets 100s/transcripts, which in turn may contain sites for 2+ miRNAs). Most miRNAs TBD. Mostly transcribed by Pol II. Often tissue-specific exp. Primary transcripts processed by Drocha and Dicer RNAses-> ~22nt short ssmRNAs, bound by Ago proteins. In animals, bind target by base-pairing (allow mismatches). Inhibit translation, destabilise target. Mismatch tolerance and 3D structures-> more diverse response, not as good for experimentation.
siRNAs and piRNAs
siRNAs/RNAi: Template usually mRNA in cytoplasm, but nascent RNA can be target of nuclear RNAi pathway. Long dsRNA-> ~23nt siRNAs by Dicer, bind Agos. If siRNA base-pairs target, Ago cuts RNA @ siRNA binding site, degraded. siRNA+ Ago=RISC. Knockdown experimental tool, new therapeutic approach (if introduce siRNA, bypassing Dicer, no interferon response. siRNA source usually pathogenic. HCV treatment 2013). Important to antiviral immunity, destroying genome of RNA viruses during replication.
piRNAs: exp in germline. Target RNA from transposable elements. Lack-> infertility (“guardians of the genome”). Provide memory of transposable elements (comparable to immunity). Bind specialised Agos (Piwi), act in cytoplasm (like RNAi/miRNA) AND nucleus (post and co-transcriptional regulation.
siRNA/piRNA and chromatin: nuclear RNAi and piRNA pathways can link RNAi to transcriptional repression via chromatin mods.
Long non-coding RNAs: more diverse roles. Structural RNAs, roles in transcription (transcriptional enhancers), chromatin regulation, etc. Transcription (antisense) itself affects local chromatin structure. NB that ~80% genome transcribed, only 2% coding (noise). E.g., XX cells have Barr body (XX individuals inactivate 1 X by means of it)- RNA code changes chromatin landscape, shuts down chromosome.
Splicing and sequestration
Aside: defining a gene, consider polycistronic mRNAs, prokaryotic polycistrons, unicellular euks that transcribe most of the genome as a polycistron, trans-splicing ENA from different genomic locations, enhancers and promoters.
NB: sometimes, mRNA sequestered for later activation/translation, e.g., ends of nerve axons (far from nucleus) for quick responses
