Term 2 Lecture 1: Binding Interactions Flashcards
Binding interactions
1)Structural changes: lock & key, induced fit model, ligand induced confirmational changes
2) 1:1 ligand binding equation and it’s assumptions: equilibrium constant, dissociation constant
3) More complex binding situations: interactions between binding sites: cooperative allostery, regulation, Hill/Adair/ NWC models e.g. 02 binding by myoglobin & haemoglobin
Proteins of a typical genome
Can be simplified into the following classes:
- catalysts (enzymes)
- transporters (usually across membranes)
- Regulators ( of gene expression and activity of other proteins)
- signalling activity (ligands, receptors, binding proteins)
- structural (and storage)
- movement (motors)
Binding
- the reversible interaction of 2 molecules aka association
- Reverse of binding is dissociation - where 2 or more interacting molecules separate
- binding can occur between a protein and a small molecule e.g. enzyme+substrate
- or between a protein and another bio macromolecule e.g. a protein binding to DNA to regulate gene expression
- or between 2 or more proteins e.g. protein protease inhibitor+digestive enzyme to inhibit it.
- binding involves non cov interactions as in protein structure but can also involve transient formation of chemical bonds
Binding related definitions
Ligand - molecule that binds to protein, usually at a specific site
Receptor- binding partner in context of signal transduction
Specific binding - binding partner only interacts with specific ligands (the norm in bio systems)
Non specific/ background binding - protein interacting w/wrong molecule (uncommon) it’s a binding partner that interacts with a wide range of ligands w/out preference
Complex- ligand+binding partner combination
Strength of binding - loose term for amount of energy required to separate ligand from partner once complexed
Models to explain specificity of binding between a protein and it’s ligand(s)
Emil Fischer (1890) “Lock & key” model for substrate binding by enzyme. Shape of active site complementary to the substrate
Dan Koshland (1957) “induced fit” model for substrate binding by enzyme. The shape of the active site changes on binding the substrate to give an exact “fit” - requires the protein to be flexible in shape. Proteins are dynamic so induced fit is considered normal
Protein binding sites can be selected to “fit” almost any ligand inducing DNA & RNA e.g. TATA box binding protein bends the DNA
Binding sites can involve residues
Residues from diff parts of the polypep chain - including those not adjacent e.g. 9LY2
3 D structure determines binding strength/specificity of the protein
Proteins can act as ligands as well as receptors
E.g. bovine pancreatic trypsin inhibitor (BPTI) acts as a ligand towards the digestive pancreatic enzyme trypsin with a loop region that blocks trypsins active site
Binding of ligands can cause confirmational changes in proteins and modulate their functions
E.g. Changes in confirmation of signalling protein Calmodulin (a sensor) which transmits Ca signals to target proteins such as Ca²+/ Calmodulin dependent-protein Kinase ll
Calmodulin can have:
A: no bound Ca (unstructured central region)
B: bound Ca (central region helical)
C: interaction with target & Ca bound (central region collapses brining outer domains together around ligands)
Ligand binding can alter functional properties of proteins
E.g. ligand gated ion channels open when the relevant ligand binds to them
E.g. binding of ADP & 1,3 biphosphoglycerate to phosphoglycerate kinase causes the binding site to “close up” on the bound ligands, causing overall protein structure to compress vertically - often termed “hinge bending” Energy necessary to change the confirmation comes from the favourable interaction of the ligand with the binding site
Measuring dissociation constant for protein-ligand interactions - importance of graphs/equations
A+B <-> C
Equilibrium constant Keq for binding of a ligand (B) to a macromolecule (A) to produce a complex (C)
Keq= [C]/[A][B] so that [C]=Keq[A][B]
Also the dissociation constant
Kd=[A][B]/[C]
(Note that concentrations are reached at equilibrium and are diff from initial concentrations when A & B are mixed)
So we should actually write [B] as [B]free to be clear. Also [B]total= [B]free+[C]
& [A] total=[A]free + [C]
Define fractional saturation (Y) of the macromolecule (A) with ligand (B)
Y= [C]/{[A]+[C]}
I.e. Y ranges between 0 and 1 for all binding sites on A being empty to all sites on A being full
This can be rearranged as:
Y= Keq[A][B]/{A+Keq[A][B]}
And then
Y=Keq[B]/{1+Keq[B]} = [B]/{1/Keq+[B]}
=[B]/{Kd+[B]}
Therefore plotting Y versus [B] will give a hyperbola as expected with a half-saturation being reached at:
[B]=1/Keq=Kd
And the curve approximately a value of 1
(100% saturation of binding site on A)
This is the same mathematical form as the Michaelis Mentin enzyme kinetics equation and a rearrangement of Henderson Hasselbach equation
Measuring dissociation constants for protein-ligand interactions
Y=[B]/{Kd+[B]}
By carrying out the initial experiment at a range of different concentrations Kd can be estimated from the resulting hyperbolic plots or better by plotting a reciprocal plot of 1/fraction vs. 1/ligand or by computer
Data are often presented as semilogarithmic plots where Kd can be directly visualised from the midpoint of the curve - this plot spreads the data better.
Typical values for Kd for ligands binding to proteins lie in the range of 10-⁴ to 10-¹⁰ M and are usually similar to ligand concentration in Vivo. Kd for biotin binding in avidin protein is ~10-¹⁵ M ( about the strongest non-cov bond known)
Example: binding of O2 to myoglobin
(can be measured by spectrometer)
Haem group (heme) is held in the protein by a coordinate bond to a histidine side chain allowing O2 to bind to the “free” 6th position on the Fe atom.
Binding causes haem group to flatten and shift the position of the histidine. The absorbance spectrum of haem changes with increasing levels of O2 binding. Plotting the change in absorbance Vs O2 concentration will give a hyperbolic binding curve.
For Y(O2) Vs [O2] where Kd corresponds to [O2] and Y(O2)= 0.5.
Reaction in which O2 binds to Mb can be written Mb + O2 <-> MbO2
Keq=[MbO2]/[Mb][O2] = [C]/[A][B]
Kd=[Mb][O2]/[MbO2] = [A][B]/[C]
Fraction of myoglobin molecules containing bound O2 is:
Y(O2)=[MbO2]/[Mb]+[MbO2]= [C]/[A]+[C]
From our definition of Kd:
YO2= [O2]/Kd+[O2]
If the protein has multiple ligand binding sites that don’t interact:
If there are “q” sites and they are identical and do not interact then:
Y= q[B]/{Kd+[B]}
Giving the same hyperbola as before.
If there are “q” sites and they are not identical and do not interact then we get a superposition of binding hyperbolas
E.g. if there are 2 nonidentical non-interacting sites for ligand B then:
Y=[B]/{Kd1+[B]}+[B]/{Kd2+[B]}
If the Kd are v. diff from each other (more than 10 fold) then the binding curve has a clear”bump” but if the Kd are not that diff then curve becomes more like a hyperbola and is difficult to analyse without additional info
What if multiple binding sites do interact?
This happens with most proteins and leads to cooperativity and allostery
Allostery means other space in Greek or action at a distance
