the binding may be more complex or if multiple ITC experiments
reproduce systematic deviations from a simpler model. Some programs also report fitted values of K d , ΔH, and n with unrealistic
levels of precision that are based purely on the NLLS curve fitting
exercise. There are many other sources of error that contribute to
uncertainty in fitted values from individual experiments (discussed
above) and so fitted values and errors should be quoted with a more
generous margin than indicated by the fit or preferably determined
rigorously by repeats of the experiment using freshly prepared
materials and based on new concentration determinations.
ITC data can be exported from the fitting programs (see Note
6) and used to generate a cumulative sum of total heat produced
during the titration for plotting in alternate graphing packages.
This can be fit with binding models such as one would use for
changes in a spectroscopic property and graphed to generate more
familiar looking binding hyperbola. Processing data in this way was
a requirement of employing more complex binding models in the
early era of ITC experimentation, although the software supplied
with instruments has improved greatly in recent years. Thus in
parallel with the development of ITC as a technique, a number of
alternate fitting packages have been developed as academic and
commercial exercises. These provide increasingly complex binding
models, global fitting multiple experiments to common binding
parameters and even fitting the shape of individual ITC injection
profiles to gain kinetic information [5–10]. Reviewing these
options is beyond the scope of this practical introduction. However, following Occam’s principle, it is not advisable to introduce
additional complexity when fitting a single data set unless supported by other experimental evidence. If reliable and robust ITC
data that does not fit a simple model are obtained as a matter of
routine, then global fitting can help to constrain the additional
binding parameters that are required (see Chapter 2).
3.6 Titrations Either
Way Round: Varying
Ligand or Protein
The choice in an ITC binding experiment of which component to
load into the cell and which to titrate from the syringe may be
dictated by considerations such as the availability of materials or
their solubility. A typical starting concentration in a small cell
volume instrument such as the iTC200 is 20–40 μM. This can be
adjusted up or down depending on the K d and/or the signal
amplitude (magnitude of ΔH). It is typical that the ligand in the
syringe is added to a higher final molar concentration in order to
fully saturate the binding partner in the cell, thus requiring more
material. Also the syringe volume is typically a factor of ~5 or so
smaller than the cell volume so to achieve the final excess concentration, it requires that the ligand be 10–20 times the concentration
of the cell material. It is therefore important that the ligand can be
concentrated to these levels without aggregating or undergoing
other types of specific self-association.
Isothermal Titration Calorimetry
147
reproduce systematic deviations from a simpler model. Some programs also report fitted values of K d , ΔH, and n with unrealistic
levels of precision that are based purely on the NLLS curve fitting
exercise. There are many other sources of error that contribute to
uncertainty in fitted values from individual experiments (discussed
above) and so fitted values and errors should be quoted with a more
generous margin than indicated by the fit or preferably determined
rigorously by repeats of the experiment using freshly prepared
materials and based on new concentration determinations.
ITC data can be exported from the fitting programs (see Note
6) and used to generate a cumulative sum of total heat produced
during the titration for plotting in alternate graphing packages.
This can be fit with binding models such as one would use for
changes in a spectroscopic property and graphed to generate more
familiar looking binding hyperbola. Processing data in this way was
a requirement of employing more complex binding models in the
early era of ITC experimentation, although the software supplied
with instruments has improved greatly in recent years. Thus in
parallel with the development of ITC as a technique, a number of
alternate fitting packages have been developed as academic and
commercial exercises. These provide increasingly complex binding
models, global fitting multiple experiments to common binding
parameters and even fitting the shape of individual ITC injection
profiles to gain kinetic information [5–10]. Reviewing these
options is beyond the scope of this practical introduction. However, following Occam’s principle, it is not advisable to introduce
additional complexity when fitting a single data set unless supported by other experimental evidence. If reliable and robust ITC
data that does not fit a simple model are obtained as a matter of
routine, then global fitting can help to constrain the additional
binding parameters that are required (see Chapter 2).
3.6 Titrations Either
Way Round: Varying
Ligand or Protein
The choice in an ITC binding experiment of which component to
load into the cell and which to titrate from the syringe may be
dictated by considerations such as the availability of materials or
their solubility. A typical starting concentration in a small cell
volume instrument such as the iTC200 is 20–40 μM. This can be
adjusted up or down depending on the K d and/or the signal
amplitude (magnitude of ΔH). It is typical that the ligand in the
syringe is added to a higher final molar concentration in order to
fully saturate the binding partner in the cell, thus requiring more
material. Also the syringe volume is typically a factor of ~5 or so
smaller than the cell volume so to achieve the final excess concentration, it requires that the ligand be 10–20 times the concentration
of the cell material. It is therefore important that the ligand can be
concentrated to these levels without aggregating or undergoing
other types of specific self-association.
Isothermal Titration Calorimetry
147