entropy-driven reaction with a positive unfavorable enthalpy. The
interaction with the 42 bp DNA gives a stoichiometry of 0.47 Æ 0.1
molecule of 42 bp DNA bound to one Ku in good agreement with
the expected positioning of two Ku molecules on such DNA. The
interaction between and the 42 bp DNA shows a K d of
2.4 Æ 0.5 nM and a reaction also driven by the entropy. The heat
exchanges measured for the two interactions are very small (maximum value of heat exchange ~0.02 μcal/s for Ku–DNA (18 bp))
and highlight the high sensitivity of the VP-ITC.
3.2 Methods for MST
We describe here general guidelines to measure interactions using
MST and we illustrate two experimental setups with a labeled DNA
and non-labeled Ku or a labeled Ku and non-labeled DNA. We will
call receptor the fluorescently labeled molecules and ligand the
non-labeled one hereafter. The first step of an MST experiment is
to find a suitable combination of the receptor concentration
(10–100 nM are usually sufficient) and excitation LED power.
The LED can be adjusted to 1–100%. For the binding experiment
itself, you will need ~200 μl at twice the concentration you have
established at this step (for one replicate measurement). To get an
accurate, well-defined K d , it is important to use a high enough
concentration of the ligand to reach saturation. Saturation is a
hallmark of ligand-specific molecular interaction: if the interaction
cannot be saturated, i.e., the signal does not change anymore by
increasing the ligand concentration, it is not a specific interaction,
but a nonspecific adsorption effect. To reach saturation, it is recommended that the highest concentration of ligand is equal or higher
than 20-fold the expected K d value.
In a preliminary assay, we compare the fluorescence intensity of
the receptor after switching on the IR laser LED (MST power) in
the absence and in the presence of a high enough concentration of
the ligand (respective to the expected K d ). We perform this experiment with at least four capillaries with and without ligand. The
signal is defined as the response amplitude F norm . It is the ratio
between fluorescence intensity after turning on the IR laser LED
(usually measured at the 2-second or 5-second mark, called “MST
on time”) divided by the fluorescence intensity measured before
turning on the IR laser LED. It is expressed as ‰. The noise floor is
the error (standard deviation) between replicates without ligand. If
it is above 8‰, it may reflect some protein aggregation that needs
optimization. This also determines what should be considered
binding (or not) of the ligand: binding is only confirmed, if the
change observed upon ligand binding is threefold larger than your
noise level established at this step. For example, for a typical noise
136
Amandine Gontier et al.
interaction with the 42 bp DNA gives a stoichiometry of 0.47 Æ 0.1
molecule of 42 bp DNA bound to one Ku in good agreement with
the expected positioning of two Ku molecules on such DNA. The
interaction between and the 42 bp DNA shows a K d of
2.4 Æ 0.5 nM and a reaction also driven by the entropy. The heat
exchanges measured for the two interactions are very small (maximum value of heat exchange ~0.02 μcal/s for Ku–DNA (18 bp))
and highlight the high sensitivity of the VP-ITC.
3.2 Methods for MST
We describe here general guidelines to measure interactions using
MST and we illustrate two experimental setups with a labeled DNA
and non-labeled Ku or a labeled Ku and non-labeled DNA. We will
call receptor the fluorescently labeled molecules and ligand the
non-labeled one hereafter. The first step of an MST experiment is
to find a suitable combination of the receptor concentration
(10–100 nM are usually sufficient) and excitation LED power.
The LED can be adjusted to 1–100%. For the binding experiment
itself, you will need ~200 μl at twice the concentration you have
established at this step (for one replicate measurement). To get an
accurate, well-defined K d , it is important to use a high enough
concentration of the ligand to reach saturation. Saturation is a
hallmark of ligand-specific molecular interaction: if the interaction
cannot be saturated, i.e., the signal does not change anymore by
increasing the ligand concentration, it is not a specific interaction,
but a nonspecific adsorption effect. To reach saturation, it is recommended that the highest concentration of ligand is equal or higher
than 20-fold the expected K d value.
In a preliminary assay, we compare the fluorescence intensity of
the receptor after switching on the IR laser LED (MST power) in
the absence and in the presence of a high enough concentration of
the ligand (respective to the expected K d ). We perform this experiment with at least four capillaries with and without ligand. The
signal is defined as the response amplitude F norm . It is the ratio
between fluorescence intensity after turning on the IR laser LED
(usually measured at the 2-second or 5-second mark, called “MST
on time”) divided by the fluorescence intensity measured before
turning on the IR laser LED. It is expressed as ‰. The noise floor is
the error (standard deviation) between replicates without ligand. If
it is above 8‰, it may reflect some protein aggregation that needs
optimization. This also determines what should be considered
binding (or not) of the ligand: binding is only confirmed, if the
change observed upon ligand binding is threefold larger than your
noise level established at this step. For example, for a typical noise
136
Amandine Gontier et al.