17. When loading the reaction solutions into the syringes, it is
important to avoid introducing bubbles into the syringe. Bubbles can scatter light and/or can compress upon the motor
drive firing thus preventing the total reaction volume from
being mixed and causing the liquid column after the flow was
stopped to move, ultimately creating noise in the measurements. Therefore, large bubbles should be removed by displacing the reaction solution from the stopped-flow syringe back
into the disposable syringe trapping the bubbles in the disposable syringe. Additionally, it is important to move the reaction
solution between the stopped-flow and disposable syringes
several times before starting the experiment to make sure the
reactant solutions are well mixed (e.g., after centrifuging to
remove precipitates in the previous step).
18. Ideally, the fluorescence signal should plateau upon reaching
equilibrium, and thus it is important to measure until a stable
signal is reached (plateau). This will improve the fitting step.
Furthermore, most of the data points do not need to be in the
plateau and if they are, the time recorded for each shot can be
reduced so that more of the data points fall within the section
of the signal that shows the fastest change. It is recommended
to do single shots until the optimal recording time and point
distribution is determined for each experiment (for fast reactions it is often beneficial to use exponential data distributions
with increasing separation between each recorded point).
19. When fitting an equation to the data set, it is important to
check that the resulting fit adequately describes the signal. As
general rule, this entails that the experimental data is equally
distributed above and below the fit curve. A good fit should
have equivalent data points on either side of the line
corresponding to the fit and no systematic deviation should
be observed (e.g., the first part of the data is below and the later
part above the fit). If not, additional exponential (or linear)
terms can be added to the used equation. However, additional
exponential terms indicate additional kinetic steps
(or processes) in the reaction and must be reflected in the
kinetic mechanism. It may be necessary to titrate one of the
reaction components to determine which of the rates is concentration dependent, i.e., the rate of binding. Fundamentally,
a protein–ligand binding event should be a one-step process
and fit with a one-exponential equation; however, it is possible
that additional kinetic steps can be observed in these pre-steady
state measurements as a result of a change in either fluorophore’s position and/or local environment (see Subheading 3).
For example, in Shields et al., we observed a second phase in
the nucleotide binding experiments that, through subsequent
titration of the Mant-nucleotide, was shown to be
Fluorescence-Based Equilibrium and Pre-Steady State Methods
285
important to avoid introducing bubbles into the syringe. Bubbles can scatter light and/or can compress upon the motor
drive firing thus preventing the total reaction volume from
being mixed and causing the liquid column after the flow was
stopped to move, ultimately creating noise in the measurements. Therefore, large bubbles should be removed by displacing the reaction solution from the stopped-flow syringe back
into the disposable syringe trapping the bubbles in the disposable syringe. Additionally, it is important to move the reaction
solution between the stopped-flow and disposable syringes
several times before starting the experiment to make sure the
reactant solutions are well mixed (e.g., after centrifuging to
remove precipitates in the previous step).
18. Ideally, the fluorescence signal should plateau upon reaching
equilibrium, and thus it is important to measure until a stable
signal is reached (plateau). This will improve the fitting step.
Furthermore, most of the data points do not need to be in the
plateau and if they are, the time recorded for each shot can be
reduced so that more of the data points fall within the section
of the signal that shows the fastest change. It is recommended
to do single shots until the optimal recording time and point
distribution is determined for each experiment (for fast reactions it is often beneficial to use exponential data distributions
with increasing separation between each recorded point).
19. When fitting an equation to the data set, it is important to
check that the resulting fit adequately describes the signal. As
general rule, this entails that the experimental data is equally
distributed above and below the fit curve. A good fit should
have equivalent data points on either side of the line
corresponding to the fit and no systematic deviation should
be observed (e.g., the first part of the data is below and the later
part above the fit). If not, additional exponential (or linear)
terms can be added to the used equation. However, additional
exponential terms indicate additional kinetic steps
(or processes) in the reaction and must be reflected in the
kinetic mechanism. It may be necessary to titrate one of the
reaction components to determine which of the rates is concentration dependent, i.e., the rate of binding. Fundamentally,
a protein–ligand binding event should be a one-step process
and fit with a one-exponential equation; however, it is possible
that additional kinetic steps can be observed in these pre-steady
state measurements as a result of a change in either fluorophore’s position and/or local environment (see Subheading 3).
For example, in Shields et al., we observed a second phase in
the nucleotide binding experiments that, through subsequent
titration of the Mant-nucleotide, was shown to be
Fluorescence-Based Equilibrium and Pre-Steady State Methods
285