general:

Activators - protein-DNA interaction

MBV4230

Odd S. Gabrielsen

The sequence specific activators: transcription factors

Modular design with a minimum of two functional domains 1. DBD - DNA-binding domain 2. TAD - transactivation domain

DBD: several structural motifs classification into TF-families

TAD - a few different types Three classical categories

Acidic domains (Gal4p, steroid receptor) Glutamine-rich domains (Sp1) Proline- rich domains (CTF/NF1)

Mutational analyses - bulky hydrophobic more important than acidic

Unstructured in free state - 3D in contact with target?

Most TFs more complex Regulatory domains, ligand binding domains

etc

N

C

TAD

DBD

MBV4230

Odd S. Gabrielsen

TF classification based on structure of DBD

bHelix-Loop-Helix(Max)

Zinc finger

Leucine zipper(Gcn4p)

p53 DBD

NFB

STATdimer

Two levels of recognition1. Shape recognition

Anhelix fits into the major groove in B-DNA. This is used in most interactions

2. Chemical recognitionNegatively charged sugar-phosphate chain involved in electrostatic interactionsHydrogen-bonding is crucial for sequence recognition

MBV4230

Odd S. Gabrielsen

Alternative classification of TFs on the basis of their regulatory role Classification questions

Is the factor constitutive active or requires a signal for activation?

Does the factor, once synthesized, automatically enter the nucleus to act in transcription?

If the factor requires a signal to become active in transcriptional regulation, what is the nature of that signal?

Classification system I. Constitutive active nuclear factors II. Regulatory transcription factors

Developmental TFs Signal dependent

Steroid receptors Internal signals Cell surface receptor controlled

Nuclear Cytoplasmic

MBV4230

Odd S. Gabrielsen

Classification - regulatory function

Brivanlou and Darnell (2002) Science 295, 813 -

MBV4230

Odd S. Gabrielsen

Sequence specific DNA-binding- essential for activators TFs create nucleation sites in promoters for

activation complexes Sequence specific DNA-binding crucial role

Principles of sequence specific DNA-binding

MBV4230

Odd S. Gabrielsen

How is a sequence (cis-element) recognized from the outside?

Electrostaticinteraction

Hydrophobicinteraction

Hydrogen-bonds

Form/geometry

Shape recognition Chemical recognition

MBV4230

Odd S. Gabrielsen

Complementary forms

The dimension of anhelix fits the dimensions of the major groove in B-DNA

Side chains point outwards and are ideally positioned to engage in hydrogen bonds

MBV4230

Odd S. Gabrielsen

Direct reading of DNA-sequenceRecognition of form

The dimension of an -helix fits the dimensions of the major groove in B-DNA

Most common type of interaction

Usually multiple domains participate in recognition dimers of same motif tandem repeated motif Interaction of two different

motifs recognition: detailed fit of

complementary surfaces Hydration /water participates seq specific variation of DNA-

structure

MBV4230

Odd S. Gabrielsen

Example

Steroid receptor

MBV4230

Odd S. Gabrielsen

How is a sequence (cis-element) recognized from the outside?

Electrostaticinteraction

Hydrophobicinteraction

Hydrogen-bonds

Form/geometry

Shape recognition Chemical recognition

MBV4230

Odd S. Gabrielsen

Next level: chemical recognition - reading of sequence information

Negatively charged sugar-phosphate chain = basis for electrostatic interaction Equal everywhere - no sequence-recognition

Still a main contributor to the strength of binding

MBV4230

Odd S. Gabrielsen

How is a sequence (cis-element) recognized from the outside?

Electrostaticinteraction

Hydrophobicinteraction

Hydrogen-bonds

Form/geometry

Shape recognition Chemical recognition

MBV4230

Odd S. Gabrielsen

Recognition by Hydrogen bonding

A

D A Hydrogen-bonding is a

key element in sequence specific recognition

10-20 x in contact surface

Base pairing not exhausted in duplex DNA, free positions point outwards in the major groove

MBV4230

Odd S. Gabrielsen

Docked prot side chains exploit the H-bonding possibilities for interaction

Hydrogen-bonding is essential for sequence specific recognition 10-20 x in contact interphase

Most contacts in major groove

Purines most important

A Zif example

MBV4230

Odd S. Gabrielsen

Interaction: Protein side chain - DNA bp Close up

Amino acid side chains points outwards from the -helix and are optimally positioned for base-interaction

MBV4230

Odd S. Gabrielsen

How is a sequence (cis-element) recognized from the outside?

Electrostaticinteraction

Hydrophobicinteraction

Hydrogen-bonds

Form/geometry

Shape recognition Chemical recognition

MBV4230

Odd S. Gabrielsen

Hydrophobic contact points

Ile

Homeodomains

MBV4230

Odd S. Gabrielsen

The Homeodomain-family: common DBD-structure

Homeotic genes - biology Regulation of Drosophila development Striking phenotypes of mutants – body-

parts move Control genetic developmental program

Homeobox / homeodomain Conserved DNA-sequence “homeobox” in a

large number of genes Encode a 60 aa “homeodomain” A stably folded structure that binds DNA Similarity with prokaryotic helix-turn-

helix 3D-structure determined for several HDs

Drosophila Antennapedia HD (NMR) Drosophila Engrailed HD-DNA complex

(crystal) Yeast MAT2

MBV4230

Odd S. Gabrielsen

Homeodomain-family: common DBD-structure Major groove contact via a 3 -helix structure

helix 3 enters major groove (“recognition helix”)

helix 1+2 antiparallel across helix 3

16 -helical aa conserved 9 in hydrophobic core some in DNA-contact interphase (common docking mechanism?)

MBV4230

Odd S. Gabrielsen

Engrailed

MBV4230

Odd S. Gabrielsen

Homeodomain-family: common DBD-structure

Minor groove contacted via N-terminal flexible arm R3 and R5 in engrailed and R7 in MAT2 contact AT in minor groove

R5 conserved in 97% of HDs Deletions and mutants impair DNA-binding

Loop between helix 1 and 2 determines Ubx versus Antp function Close to DNA exposed for protein-protein interaction

MBV4230

Odd S. Gabrielsen

HD-paradox: what determines sequence specificity? Drosophila Ultrabithorax (Ubx), Antennapedia (Antp), Deformed

(Dfd) and Sex combs reduced (Scr): closely similar HD, biological role very different

Minor differences in DNA-binding in vitro TAAT-motif bound by most HD-factors contrast between promiscuity in vitro and specific effects in vivo

Swaps reveal that surprisingly much of the specificity is determined by the N-terminal arm which contacts the minor groove Swaps: Antp with Scr-type N-term arm shows Scr-type specificity in

vivo Swaps: Dfd with Ubx-type N-term arm shows Ubx-type specificity in

vivo N-terminal arm more divergent than the rest of HD

R5 and R7 (contacting DNA) are present in both Ubx, Antp, Dfd, and Scr

Other tail aa diverge much more

MBV4230

Odd S. Gabrielsen

Solutions of the paradox

Conformational effects mediated by N-term arm Even if the -helical HDs are very similar, a much

larger diversity is found in the N-terminal arms that contact the minor groove

Protein-protein interaction with other TFs through the N-terminal arm - enhanced affinity/specificity - the basis of combinatorial control MAT2 interaction with MCM1 - cooperative

interactions Ultrabithorax- Extradenticle in Drosophila Hox-Pbx1 in mammals

MBV4230

Odd S. Gabrielsen

Combinatorial TFs give enhanced specificity TFs encoded by the the

homeotic (Hox) genes govern the choice between alternative developmental pathways along the anterior–posterior axis.

Hox proteins, such as Drosophila Ultrabithorax, have low DNA-binding specificity by themselves but gain affinity and specificity when they bind together with the homeoprotein Extradenticle (or Pbx1 in mammals).

MBV4230

Odd S. Gabrielsen

N-tail in protein-protein interaction- adopt different conformations

Mat-2/Mcm-1HD

HD

Conformation determinedby prot-prot interaction

MBV4230

Odd S. Gabrielsen

It works impressively well

Hox genes

POU family

MBV4230

Odd S. Gabrielsen

POU-family: common DBD-structure

The POU-name : Pit-1 pituitary specific TF Oct-1 and Oct-2 lymphoid TFs Unc86 TF that regulates neuronal development in C.elegans

A bipartite160 aa homeodomain-related DBD a POU-type HD subdomain (C-terminally located) et POU-specific subdomain (N-terminally located) Coupled by a variable linker (15-30 aa)

POU is a structurally bipartite motif that arose by the fusion of genes encoding two different types of DNA-binding domain.

MBV4230

Odd S. Gabrielsen

POU: Two independent subdomains

POUHD subdomain 60 aa closely similar to the classical

HD Only weakly DNA-binding by itself (<HD) contacts 3´-half site (Oct-1: ATGCAAAT) docking similar to engrailed. Antp etc Main contribution to non-specific

backbone contacts

POUspec subdomain 75 aa POU-specific domain enhances DNA-affinity 1000x contacts 5´-half site (Oct-1: ATGCAAAT) contacts opposite side of DNA relative

to HD structure similar to prokaryotic - and

434-repressors The two-part DNA-binding domain

partially encircles the DNA.

MBV4230

Odd S. Gabrielsen

Flexible DNA-recognition

POU-domains have intrinsic conformational flexibility and this feature

appears to confer functional diversity in DNA-recognition

The subdomains are able to assume a variety of conformations, dependent on the DNA element.

ZNFs:zinc finger families

MBV4230

Odd S. Gabrielsen

Zinc finger proteins

Zinc finger proteins were first discovered as transcription factors.

Zinc finger proteins are among the most abundant proteins in eukaryotic genomes.

Their functions are extraordinarily diverse include DNA recognition, RNA packaging, transcriptional activation, regulation of apoptosis, protein folding and assembly, and lipid binding.

Zinc finger structures are as diverse as their functions.

MBV4230

Odd S. Gabrielsen

Examples

C2H2-type Ziffra Sp1 (3.fngr)

C4-type Ziffra GATA-1

LIM-domain type Ziffra ACRP

PKC-type Zif

Zn++

The C2H2 subfamily

MBV4230

Odd S. Gabrielsen

Zn++ Zn++ Zn++ Zn++Zn++ Zn++Zn++ Zn++

Classical TFIIIA-related zinc fingers: n x [Zn-C2H2]

History: Xenopus TFIIIA the first isolated and cloned eukaryotic TF Function: activation of 5S RNA transcription (RNAPIII) Rich source : accumulated in immature Xenopus oocyttes as “storage particles” = TFIIIA+5S RNA (≈ 15% of total soluble protein)

Purified 1980, cloned in 1984 Mr= 38 600, 344 aa

Primary structure TFIIIA Composed of repeats: 9x 30aa minidomains + 70aa unique region C-trm

Each minidomain conserved pattern of 2Cys+2His Hypothesis: each minidomain structured around a coordinated zinc ion (confirmed later)

MBV4230

Odd S. Gabrielsen

Zinc finger proteins

Finger-like in 2D Not in 3D

MBV4230

Odd S. Gabrielsen

Common features of TFIIIA-related zinc fingers Consensus for each finger: FXCX2-5CX3FX5FX2HX2-5H

Number of fingers in related factors varies: 2-37 Number of members exceptionally high

S.cerevisiae genome: 34 C2H2 zinc fingers C.elegans genome 68 C2H2 zinc fingers Drosophila genome 234 C2H2 zinc fingers Humane genome 564 C2H2 zinc fingers, (135 C3HC4 zinc finger)

We now recognize the classical C2H2 zinc finger as the first member of a rapidly expanding family of zinc-binding modules.

MBV4230

Odd S. Gabrielsen

3D structure of the classical C2H2-type of zinc fingers

Each finger = a minidomain with -structure each finger an independent module Several fingers linked together by flexible

linkers First 3D structure: the 3-finger Zif268 (mouse)

DNA interaction in Zif268 major groove contact through -helix in recognition of base triplets aa in three positions responsible for sequence

recognition: -1, 3 and 6 (rel. til -helix) Simple one-to-one pattern (contact aa - baser)

can a recognition code be defined ?? DNA interaction in GLI and TTK differs

different phosphate contact distortion of DNA finger 1 without DNA contact

DNA

MBV4230

Odd S. Gabrielsen

The Zif268 prototype

Finger 2 from Zif268 including the two

cysteine side chains and two histidine side chains that coordinate the zinc ion

DNA-recognition residues indicated by the

numbers identifying their position relative to the start of the recognition helix

MBV4230

Odd S. Gabrielsen

Three fingers in Zif268

Zif268 - first multi-finger structure

recognition of base triplets

Finger 1

Finger 2

Finger 3

LINKER

DNA

MBV4230

Odd S. Gabrielsen

Recognition code?

The DNA sequence of the Zif268 site is color coded to indicate base contacts made by each finger.

MBV4230

Odd S. Gabrielsen

Structure of the six-finger TFIIIA–DNA complex

In a multi-finger protein some fingers contact base pairs and some will not, but rather function as bridges Fingers 1–2–3, separated by

typical linkers, wrap smoothly around the major groove like those of Zif268

In contrast, fingers 4–5–6 form an open, extended structure running along one side of the DNA. Of these, only finger 5 makes contacts with bases in the major groove. The flanking fingers, 4 and 6, appear to serve primarily as spacer elements.

Nuclear receptors 2xC4

MBV4230

Odd S. Gabrielsen

Nuclear receptors: 2x[Zn-C4]:

Large family where DBD binds two Zn++ through a tetraedrical pattern of Cys

conserved DBD 70-80 aa Protein structure

Two “zinc fingers” constitute one separate domain Two -helices with C3-Zn-C4 N-terminally These perpendicular on the top of each other with

hydrophobic interactions Mediates trx response to complex extra cellular signals Evolutionary coupled to multi cellular organisms

Yeast = 0 but C.elegans 233 or 1.5% of genes !! Sequence prediction: 90% with nuclear receptor DBD has

potential ligand-BD Implies that lipophilic signal molecules have been important

to establish communication between cells

MBV4230

Odd S. Gabrielsen

DNA-binding by nuclear receptors

MBV4230

Odd S. Gabrielsen

Nuclear receptors - DNA interaction

3D Prot-DNA structure glucocorticoid receptor + estrogen receptor

Dimer in complex (monomer in solution) DNA interaction

First “finger” binds DNA Second “finger” involved in dimerization Binds to neighboring “major grooves” on same

side of DNA Extensive phosphate contact and recognition

helix docked into the groove specificity determined by 3 aa (E2, G3, A6)

in recognition helix Structured dimer interphase formed upon DNA-

binding

GATA factors

MBV4230

Odd S. Gabrielsen

GATA-factors: 1x [Zn-C4]:

Small family Prototype erythroid TF: GATA-1 (2 fingers)

C-terminal finger – DNA binding N-terminal finger – protein interactions?

From fungi to humans Structure ≈ 1.finger in nuclear receptors Hydrophobic DNA interphase Evolutionary implications

Early duplication of primitive finger divergent functions developed in NR

Gal4p factors

MBV4230

Odd S. Gabrielsen

GAL4-related factors: 1 x [Zn2-C6]:

GAL4-DBD = 28aa cys-rich domain binds 2 Zn++ + 26aa C-terminal domain involv. in dimerization

Cys-rich domain consensus: CX2CX6CX6CX2CX6C A Zn-Cys cluster with shared Cys (1. and 4.) Two short -helices with C-Zn-C N-terminal

MBV4230

Odd S. Gabrielsen

GAL4-related factors: 1 x [Zn2-C6]:

Dimerization domain Monomer in solution, dimer in DNA-complex In solution only Cys-rich motif structured In complex forms two extended helix-strand motives Amfipathic helices form a dimer-interphase in the complex

DNA interaction contacts CGG-triplets in major groove C-terminal of 1. -helix contacts bases Phosphate contact via helix-strand motif Coiled-coil dimer-interphase at right angle to DNA (≈bZIP) Linker determines spacing of CGG-triplet: 11bp in GAL4, 6bp

in PPR1