BioReg Workshop Assaying Protein:DNA Interactions

1
BioReg WorkshopAssaying ProteinDNA
Interactions
2
Immobilization/Migration Assays
Technique Overview
Electron Microscopy
Can Visualize

• Filter binding
• Southwestern
• MacKay
• DNA affinity
• Gel Shifts
• Gel Filtration
• Sucrose Gradient

• DS vs SS DNA DNA unwinding
• DNA wrapping around proteins
• Protein conformational changes

Crosslinking

• Affinity enhancement for EM or immobilization/
  migration assays
• Identifying specific contacts on protein or DNA

Protection Assays(positional information)
Atomic Level Analysis

• Protection
• Exo Protection
• Modif-Interference
• Genomic Footprint (in vivo)
• Indirect Endlabeling (in vivo)

• Crystal Structure
• NMR

In Vivo Hints

• One-hybrid
• Allele specific 2nd site suppression

3
Things to Consider When UsingThese Techniques

• What is being followed the protein or DNA?
• Which require pure proteins which can be done
  with crude extracts?
• Which require significant occupancy of the
  binding site?
• How fast is the probing of the complex relative
  to dissociation rates?

• In those involving crude, which are convenient
  for purifying a protein?
• How is the DNA prepared for each assay?
• How much protein is needed for full occupancy?
• How do you show specificity? How do you determine
  what is recognized?
• Roughly how tight do the binding interactions
  have to be (off rates)?
• What are the limitations of each approach?

• How do you measure binding affinity (what are
  rough Kds for tight, mod, and weak binding)?
• How do you measure association rate or
  dissociation rate?
• How do you detect and measure cooperativity (2
  dif proteins/1 protein)?
• How do you determine which part of a protein or
  which protein of a complex contacts the DNA?

• Which of these techniques is useful for showing
  protein induced DNA bending?
• Which of these techniques is useful for showing
  protein induced DNA looping?
• Which of these techniques is useful for showing
  protein induced DNA unwinding?

4
Filter Binding(DNA/pure/partial/rapid)

• Pure protein immobilization by adsorption to
  nitrocellulose filter
• Labeled DNA passes through unless bound to
  protein
• Useful for quantitating binding affinities and
  kinetics
• Note all proteins retain binding ability when
  adsorbed to filter

proteins protein-DNA complexes



free DNA

5
DNA Affinity Chromatography(protein/crude or
pure/partial/slow-but very high on rate)

• DNA immobilized at high density to matrix by
  adsorption or covalent linkage
• Bound proteins elute at higher salts
• At lower salts high DNA conc favors rapid
  reassociation whenever protein dissociates
• Useful purification step (proteins monitored by
  activity or western if Ab available)
• Specific binding proteins separated from
  nonspecific binding proteins either by loading
  column in competition with free nonspecific DNA
  or on the basis of higher salt elution required
  for specific binding proteins
• Variation bind proteins to biotin labeled DNA
  before immobilizing DNA to streptavidin matrix
  lose advantage of large excess of DNA

low salt load
high salt elution
6
McKay (DNA/crude/partial/slow)

• Immobilize protein by immunoprecipitation
• Labeled DNA bound to protein coIPs
• Specificity determined by preferential coIP of
  specific DNA out of mixture containing
  nonspecific DNA
• Variations cross-link before IP to counter
  problem of dissociation during washes detect DNA
  by PCR

pellet wash, elute with SDS





fragment with binding site




input DNA
IPd DNA
7
Southwestern(DNA/crude or pure/partial/slow)

• DNA binding domain of protein renatures enough on
  nitrocellulose filter (after SDS-PAGE and
  Western) to allow for specific recognition of
  labeled DNA
• Useful for identifying which protein in a crude
  extract or in a tightly bound complex is
  responsible for DNA binding (proteins that do not
  bind offer additional specificity control)
• Assumption a single polypeptide is sufficient
  for binding activity

probe with labeled fragment
autorad.
8
Gel Retention/Retardation/Mobility Shift
(DNA/crude or pure/partial/fast)

• Protein binding retards motility of labeled DNA
• Once complex has entered gel caging effect
  favors rapid reassociation whenever protein and
  DNA dissociate (effectively reducing off rate)
• Useful for following DNA binding activity during
  purification
• Has also been used to quantitate binding
  affinities and kinetics
• Multiple complexes can sometimes be seen as
  different shifted species
• Note not all complexes enjoy caging effect or
  are fully stable during time it takes to enter gel

Simple Shift
ex Multiple Shifts
shifted DNA

free DNA

extract amount
9
Gel Retardation with Antibody SupershiftUse to
identify proteins in gel-shift complex
- Ab
Ab

supershifted DNA
shifted DNA
free DNA
10
Gel Retardation with Protein Induced DNA
BendingIf bending is induced, circularly
permuted fragments will gel shift
differently.The closer the binding site is to
the fragment center the greater the shift.
shifted DNA
free DNA
11
Gel Filtration (protein/pure/partial/slow)

• Large DNA molecules elute in void volume with any
  bound proteins
• Free proteins elute in included volume

DNA Bound Protein
Free Protein
Protein
Fraction
Vo
Included Volume
DNA DNA-protein VOID VOLUME
Free protein INCLUDED VOLUME
12
Sucrose Gradient(DNA/crude or partial/partial/slo
w)

• DNA coated with proteins sediments faster than
  naked DNA
• Useful for very large protein DNA complexes or
  extensive coating of DNA (e.g. chromatin
  association)

protein-coated DNA
naked DNA
DNA
increasing density
Fraction dripped from tube bottom
Increasing Size
Sucrose SedimentationGradient
13
Protection Assays

• Gives positional information about the
  accessibility of a sequence to DNA
  modifying/nicking/cutting probes
• Position is defined with respect to a fixed
  reference site defined by a restriction cut or a
  primer and so that
• Fragment LENGTH Modification/Nick/Cut POSITION
• Fragment length assessed at single nucleotide
  resolution by running on sequencing gel
• For this length/position correlation to occur and
  to be informative
• 1) The DNA template must be labeled at only one
  end and on only one strand.
• 2) modification/nicks/cuts must occur at most
  once per DNA molecule
• 3) Inherent susceptibility of the DNA to the
  probe should be relatively independent of
  sequence (and hence of position)
• 4) Protein binding should cause significant
  decrease in probe susceptiblity of DNA
• Uses determine where a protein contacts DNA
• measure binding affinities and kinetics
• monitor multiple sites on a DNA molecule and
  quantitating cooperativity
• Note full to near full occupancy needed to see
  protection (exception modif interference)
• may not see protection if protein-DNA
  interaction is dynamic

14
Standard Protection(DNA/crude or pure/full/fast)

• Probes DNase I
• Dimethyl Sulfate (G,A modified, convert to nick
  with piperidine)
• Copper Phenanthroline (attack on sugar)
• Iron-EDTA (hydroxy radical attack on sugar)

A
B
SEQ Gel
X
X
X
X
15
Interference(DNA/usually pure/partial/fast)

• Modifying Probes Dimethyl Sulfate (G, A) DEPC
  (A) KMnO4 (T) Ethylnitrosourea (phosphate)
  Formic Acid (depurination) Hydrazine
  (depyrimidation)
• Only DNA in Protein-DNA complex is examined
  (isolated by preparative gel shift)
• DNA molecules modified at positions that
  interfere with protein binding will be excluded
• Modification is converted to nick after complexes
  are isolated

bound DNA




Interference



free DNA






16
Genomic Footprinting(DNA/in vivo/full/fast)

• Probes Dimethyl Sulfate (permeable through live
  cells)
• DNAseI (have to lyse cells)
• Analysis by
• Primer extension with 5 end-labeled primer nick
  causes runoff at that position (linear
  amplification from multiple rounds in
  thermocycler)
• PCR nicked fragments details more complicated
  and not discussed (exponential amplification from
  multiple rounds in thermocycler)

protection
DNaseI nick
primer extension to nick
17
DNA-Protein Crosslinking-Identifying Protein
Contacts by Label Transfer(protein/pure/partial/f
ast)

• Crosslinking nucleotide is incorporated in DNA
  very close to a radiolabeled nucleotide
• Crosslinking initiated by UV photoactivation
  (usually inefficient)
• DNA is exhaustively digested with DNaseI leaving
  only crosslinked and neighboring nucleotides
  associated with protein (and thereby labeling it)
• Useful for identifying which protein in a complex
  or which segment of a protein (after proteolytic
  cleavage) is near a specific nucleotide position
• Crosslinking probe T (has inherent ability to
  crosslink proteins when exposed to UV)
• BrdU (significantly enhanced crosslinking over
  T)
• Photoreactive crosslinking arm attached to
  thiophosphate nucleotide

Br
UV

DNaseI
18
DNA-Protein Crosslinking -Identifying Protein
Contacts by Peptide Transfer(DNA/pure/partial/fas
t)

• Protein becomes UV crosslinked to nucleotides
  (particularly Ts) in DNA
• Protease digestion leaves nucleotide covalently
  linked to peptide
• Primer extension is blocked at modified nucleotide

UV
or
19

• How would you demonstrate the following
  conclusions
• (in addition to ways discussed in the lecture,
  could you apply other techniques to show the same
  point?)
• Ku binds blunt DNA ends
• SSBP is cooperative
• Beta subunit of Pol III is tightly bound to
  circular but not linear DNA
• Gamma complex of Pol III can load beta onto a
  primer template or nicked DNA
• Gamma complex of Pol III can also unload beta
  from primer template or nicked DNA
• Core Pol III can prevent gamma complex from
  unloading beta from a primer-template (not from
  nicked DNA) but cannot prevent gamma from loading
  beta in the first place
• RFC (clamp loader) binds to a structure (primer
  template junction) not a specific sequence
• Largest subunit of the three protein RPA complex
  is responsible for the single-stranded DNA
  binding activity
• oriC DNA wraps around approximately 30 dnaA
  proteins bound to the 9-mer elements
• Nature of proteins assembled at yeast origins
  change during cell cycle (how would you try to
  identify the proteins responsible for the
  change?)
• Which subunits of ORC contact the origin
  consensus sequence
• CDC45 and the MCM proteins leave the origin and
  move down the DNA with the fork when elongation
  begins
• Both early and late origins assemble pre-RCs
  during the M and G1 transition (can you say these
  pre-RCs are truly identical?)

20

• SV40 DNA replicated in vitro in the presence of
  CAF-1 and H3-H4 tetramer is preferentially coated
  with tetramers (as compared to unreplicated DNA)
• Sir proteins coat the ends of chromosomes from
  the telomeres to several kb in toward the
  centromere
• mutS recognizes and binds to mismatches (how
  would you determine if there is any mismatch
  preference?)
• mutS, mutL, and mutH may form a DNA loop
• E. coli RNAP binds -45 to 20 (what does it
  actually recognize?)
• Sigma factor is actually responsible for E. coli
  RNAP holoenzyme recognizing and binding -35 and
  -10 boxes
• E. coli RNAP undergoes a conformational change
  when it shifts from open to closed complex
• Sigma factor is not released from core RNAP and
  the DNA template until roughly 10 nt are
  polymerized (and core clears the promoter)
• lac repressor operators have different inherent
  binding affinities lac repressor binds
  cooperatively to these sites
• CAP adjusts position of MalT binding to promoter
  of MalT regulon
• RNAP changes its interaction with the DNA
  template as it marches down the DNA
• LEF-1 (or HU, or TBP) induces DNA bending when it
  binds its site (how would you determine the angle
  of the bend?)

21
A Note on Measuring Binding Equilibria
Kdis PD/PD where P is free protein
conc, D is free DNA conc, and PD is conc of
protein-DNA complex To measure Kdis you need to
determine values for P, D, and PDYou
control and hence know Pt and Dt in the
experimentYou also know the following
relationships Pt P PD Dt D
PD Most binding assays allow measure of
D/PD ratioe.g., protection on footprint
shifted up stuck to filter Knowing the value
of Dt you can use Dt D PD and the
measured ratio D/PD to figure out D and
PD Now that you know PD you can calculate
P from P Pt - PD However . . .
22
In practice, Kdis can only be accurately
determined if Dt is not greater than Kdis
Otherwise, if Dt is greater than Kdis, then the
vast majority of any protein added will bind to
the DNA (until the amount of protein reaches the
amount of DNA). This can be inferred from the
math. When the DNA is half occupied, D will be
close in magnitude to Dt and thus greater than
Kdis. In order for D x P/PD to equal
Kdis, P/PD must be less than 1. Thus, Pt
will be roughly equal to PD, and P Pt -
PD will be the small difference of two much
larger numbers and will not be accurately
determined. The best situation is if Dt is
significantly smaller than Kdis so that P/P-D
is large and P is approximately equal to
Pt You might think, then, that all you need to
do is make sure Dt is much smaller than most
Kdis you will be measuring (say 1pM). However,
for each assay there is a practical limitation of
how low Dt can get based on how hot the DNA can
be labeled and how much total counts are needed
in the assay. For footprinting, Dt is often on
the order of 0.1 to 1nM.
23
Corollary Occupancy is not automatically
determined by stoichiometry
Not all protein added to reaction will be bound
to DNA Percent association is determined by both
DNA concentration and the value of Kdis Only when
DNAtgtgtKdis, will occupancy be determined by
stoichiometry, i.e. protein binding versus total
protein concentration will be roughly
linear. Under these circumstances, in order for
the product of P/P-D x D to be equal Kdis,
P/P-D must be a small number, i.e. most of
the protein that is added to the reaction will be
bound to the DNA (the high DNA concentration is
pushing the equilibria toward association)