Calcium Regulation of Myosin-I Tension Sensing
John H. Lewis, Michael J. Greenberg, Joseph M. Laakso, Henry Shuman, and E.
Pennsylvania Muscle Institute and Department of Physiology, Perelman School of Medicine at the
University of Pennsylvania, Philadelphia, Pennsylvania
Calcium sensitivity of the key steps in the ATPase cycle of myo1b from ensemble-level
transient kinetic experiments
We determined the effect of calcium on the rates of the individual steps of the myo1bIQ
ATPase cycle in the presence of 100 µM Ca2+F at 37 °C (Scheme 1; Table 1). For comparison,
we also provide the rate constants for myo1bIQ in the absence of free calcium, which we
published previously (Figure 2; Table 1; (1)). To correlate biochemical kinetic rates with
mechanical transitions observed in the optical trap, we determined the effect of calcium on key
steps on the ATPase cycle at 20 °C (Tables 1-2; Figure S2).
Pyrene-actin fluorescence was used to measure the rate of ATP binding and the
population of the weakly-bound states. Strong binding of myo1bIQ to pyrene-actin quenches the
pyrene fluorescence ~75%, while ATP binding to myo1bIQ results in the population of detached
or weakly bound myosin states and the restoration of pyrene fluorescence (A* in Scheme S1).
The fluorescence transients acquired at different ATP concentrations were best fit by the sum of
two exponential rates, with a fast phase (kf) hyperbolically related to the ATP concentration (0 -
2000 µM) and a slow phase (ks) independent of ATP at concentrations > 500 µM (Figure 2). The
data were modeled according to Scheme S1:
where nucleotide-free myo1b exists in two states (2), one that binds nucleotide (AM’), and
another that is nucleotide inaccessible (AM). The ATP dependence of kf can be described by:
][ / 1
where K1’ is a rapid equilibrium and k+2’ is a rate-limiting isomerization to the high fluorescence
AM.ATP state. At both 37 °C and 20 °C, K1’ and k+2’ are only moderately affected by calcium,
with k+2’ increased ~ 2-fold by calcium (Table 1, Figure 2) for myo1bIQ. Interestingly, a similar
2-fold change in k+2' upon the addition of calcium is seen in myo1ba at 20 °C. However, the
affinity of myo1ba for ATP (K1') is 1.4 times lower than in myo1bIQ. This lower ATP affinity
leads to a slower second-order rate constant for ATP binding in myo1ba than myo1bIQ. At
physiological ATP concentrations in the cell (>1 mM), the differences between ATP binding
affinities in myo1ba and in myo1bIQ are not consequential since this concentration of ATP is well
above the dissociation constant for ATP (1/K1') for both isoforms.
The rate of ks reports the rate of isomerization of AM to AM’ (k+α) at ATP concentrations
> 500 µM, and the ratio of the amplitudes determines the equilibrium constant, Kα (1, 2). At 37
°C, the rate of going from AM to AM’ (k+α) is relatively unaffected by the introduction of
calcium, whereas the reverse rate, k-α, is increased 1.9-fold (Table 1). At 20 °C, both myo1ba
and myo1bIQ show similar fold increases in k-α. However, the forward rate, k+α, is increased
approximately 3-fold upon the addition of calcium.
The binding of mantATP to myo1bIQ does not result in a fluorescence change upon direct
excitation of the mant fluorophore, but a fluorescence change is observed when the mantATP
fluorescence is excited via energy transfer from neighboring tryptophans (1). The rate of this
change reports the second-order rate of mantATP binding at concentrations < 50 µM, and reports
the rate of the conformational change that accompanies ATP hydrolysis (k3app = k+3 + k-3) at
higher mantATP concentrations (Scheme S2).
The maximum rates in the presence and absence of calcium at 37 °C are 41 ± 4.9 s-1 and 39 ± 3.2
s-1, respectively (Figure S1, Table 1; (1)), indicating that calcium binding does not affect ATP
Fluorescently labeled phosphate-binding protein was used to measure the rate of
phosphate release (k+4’; (1, 3, 4)) at 37 °C. Phosphate release is typically measured in
sequential-mix experiments where myosin and ATP are mixed and aged to form the M.ADP.Pi
state, followed by mixing with various actin concentrations. However, pre-equilibration of
myo1bIQ and 100 µM free calcium at 37° C in the absence of actin resulted in occasional protein
aggregation. Therefore, single-mix, single-turnover experiments were performed by mixing 1
µM ATP with pre-mixed 3 µM myo1bIQ, 100 µM free calcium, and 0 - 80 µM actin. Time
courses of the phosphate-binding transients contained a small lag phase due to ATP binding and
hydrolysis (Scheme 1). The transients were well fit by a double exponential function, where one
rate emerged as actin dependent (Figure 2 c,d). The actin concentration independent rate was
found to be consistent with the rate of ATP binding at 1 µM ATP (~3.3 s-1) in the presence of
100 µM free calcium for all actin concentrations (data not shown). The absence of a linear
component in the transient confirms that the myosin underwent a single turnover. The
experimental reaction mechanism can be simplified to Scheme S3 since the M.D.Pi complex is
formed much more rapidly relative to the rate of phosphate release (1).
Assuming a rapid equilibrium for actin binding, K9, the intrinsic rate k+4’ could then be
determined according to equation S2:
The rate of phosphate release, k+4’ at 37 °C, was found to be 2.8 times faster in the presence of
calcium than in its absence (Table 1, Figure 2). Furthermore, the affinity of the M.ADP.Pi
complex for actin, K9, was increased 11-fold.
The rate of ADP release (k+5’) was determined by ATP-induced dissociation of myo1bIQ
from pyrene-actin. When actomyo1bIQ active-sites are saturated with ADP, ATP binding is rate-
limited by the slow dissociation of ADP (k+5’), as described in Scheme S4.