mechanism (Scheme 2) or a fast conformational change followed by
binding (Scheme 3). “Fast” in this respect refers to a conformational
change occurring on a timescale significantly shorter than that for the
association reaction under the chosen ligand concentration (see Note
14). The data do not rule out more complex mechanisms, for example,
there could be several conformational changes (induced fit or conformational selection) occurring on faster timescales.
3.6.3 Case Study 3:
ACTR and NCBD, Multistep
Binding with Several
Kinetic Phases
ACTR and NCBD are interaction domains from two different
transcriptional coactivators. Both are regarded as disordered
although NCBD has a hydrophobic core and folds into a
dynamic structure. ACTR and NCBD interact in a complicated
coupled binding and folding reaction with several kinetic phases.
Here, we will consider two kinetic phases detected in experiments in the presence of high ionic strength. The concentration
of ACTR was varied (1–20 μM) at constant NCBD (1 μM). The
kinetic traces were double exponential as assessed from the residuals (Fig. 3).
One of the kinetic phases increased linearly with [ACTR] while
the other one appeared rather constant. The data were fitted to
models describing a three-state binding: induced fit and conformational selection, respectively [16]. It is clear that both models
describe the data equally well. The kinetic traces can also be
fitted directly to each model using a numerical global fitting
approach. The advantage of this is that the kinetic amplitudes
can be taken into consideration. However, in this case it did not
resolve the mechanism since we cannot assign a fluorescence
signal to the intermediate state. Thus, it could not be concluded
from the data whether the conformational change that produces
the observed slow kinetic phase occurs before or after binding.
What about the kinetic trick of varying NCBD at constant ACTR
as used in the study of N TAIL and XD binding? Unfortunately,
this experiment was also inconclusive because the kinetic traces
were not clearly double exponential but something in between,
as illustrated in Fig. 1c. Comparison of k obs values determined at
excess NCBD and ACTR, respectively, suggested that curve
fitting to a single exponential resulted in a weighted average
k obs value from the two kinetic phases. Thus, the combined
observations are at best “more consistent” with an induced fit
model [23].
Conclusions: Double exponential binding kinetics proves that the
interaction involves (at least) two distinct events (three state). If one
kinetic phase increases linearly with the concentration of the varied
protein and the other one appears hyperbolic, the experiment is
122
Elin Karlsson and Per Jemth
binding (Scheme 3). “Fast” in this respect refers to a conformational
change occurring on a timescale significantly shorter than that for the
association reaction under the chosen ligand concentration (see Note
14). The data do not rule out more complex mechanisms, for example,
there could be several conformational changes (induced fit or conformational selection) occurring on faster timescales.
3.6.3 Case Study 3:
ACTR and NCBD, Multistep
Binding with Several
Kinetic Phases
ACTR and NCBD are interaction domains from two different
transcriptional coactivators. Both are regarded as disordered
although NCBD has a hydrophobic core and folds into a
dynamic structure. ACTR and NCBD interact in a complicated
coupled binding and folding reaction with several kinetic phases.
Here, we will consider two kinetic phases detected in experiments in the presence of high ionic strength. The concentration
of ACTR was varied (1–20 μM) at constant NCBD (1 μM). The
kinetic traces were double exponential as assessed from the residuals (Fig. 3).
One of the kinetic phases increased linearly with [ACTR] while
the other one appeared rather constant. The data were fitted to
models describing a three-state binding: induced fit and conformational selection, respectively [16]. It is clear that both models
describe the data equally well. The kinetic traces can also be
fitted directly to each model using a numerical global fitting
approach. The advantage of this is that the kinetic amplitudes
can be taken into consideration. However, in this case it did not
resolve the mechanism since we cannot assign a fluorescence
signal to the intermediate state. Thus, it could not be concluded
from the data whether the conformational change that produces
the observed slow kinetic phase occurs before or after binding.
What about the kinetic trick of varying NCBD at constant ACTR
as used in the study of N TAIL and XD binding? Unfortunately,
this experiment was also inconclusive because the kinetic traces
were not clearly double exponential but something in between,
as illustrated in Fig. 1c. Comparison of k obs values determined at
excess NCBD and ACTR, respectively, suggested that curve
fitting to a single exponential resulted in a weighted average
k obs value from the two kinetic phases. Thus, the combined
observations are at best “more consistent” with an induced fit
model [23].
Conclusions: Double exponential binding kinetics proves that the
interaction involves (at least) two distinct events (three state). If one
kinetic phase increases linearly with the concentration of the varied
protein and the other one appears hyperbolic, the experiment is
122
Elin Karlsson and Per Jemth