Actin Activation of Heavy Meromyssin Adenosine Triphosphatase

SUMMARY Actin activation of the ATPase activity of heavy meromyosin, a tryptic digestion product of myosin, was studied over a wide range of actin and ATP concentrations with creatine kinase and creatine phosphate added to hold the ATP concentration constant. Double reciprocal plots of ATPase against ATP concentration at varied actin concentration and of ATPase against actin concentration at varied ATP concentration were linear, suggesting that the system obeys simple Michaelis kinetics. From extrapolation of these plots we found that not only does ATP increase the dissociation of actin from heavy meromyosin but, analogously, the Michaelis constant of the acto-heavy meromyosin ATPase is greater than that of the heavy meromyosin ATPase. However, whereas the dissociation of actin from the heavy meromyosin-ATP complex is markedly dependent on ionic strength, the Michaelis constant of the acto-heavy meromyosin ATPase shows no salt dependence whatever. These data have important implications for the kinetic model which we previously proposed for the actin-heavy meromyosin-ATP system.


SUMMARY
Actin activation of the ATPase activity of heavy meromyosin, a tryptic digestion product of myosin, was studied over a wide range of actin and ATP concentrations with creatine kinase and creatine phosphate added to hold the ATP concentration constant.
Double reciprocal plots of ATPase against ATP concentration at varied actin concentration and of ATPase against actin concentration at varied ATP concentration were linear, suggesting that the system obeys simple Michaelis kinetics. From extrapolation of these plots we found that not only does ATP increase the dissociation of actin from heavy meromyosin but, analogously, the Michaelis constant of the acto-heavy meromyosin ATPase is greater than that of the heavy meromyosin ATPase. However, whereas the dissociation of actin from the heavy meromyosin-ATP complex is markedly dependent on ionic strength, the Michaelis constant of the acto-heavy meromyosin ATPase shows no salt dependence whatever. These data have important implications for the kinetic model which we previously proposed for the actin-heavy meromyosin-ATP system.
It has long been known that actin activates the iLIg-i2'Wase activit,y of myoshl at low ionic strength, and there have been several studies on the ATP deliendence of the actomyosin ATl'ase (2)(3)(4). However, complete kinetic studies involving variation of both the ATP and actin concentrations have been prevented by the fact that actomyosin is precipitated at low ionic strength, which makes systematic variation of the free actin concentration essentially impossible.
Of rourse, this problem does not arise  (5), did study the dependence of the IIMM' A'l'l'ase on actin and AT1 concentration, although they did not attempt a detailed kinetic analysis.
We have previously shown that double reciljrocal plots of HMM ATPase against actin concentration are linear in the presence of excess ATP (6, 7), and we have now extended these studies to much lower ATP concentrations with the creatine kinase-creative phosphate system (EC 2.7.3 2) added to hold the ATP concentration constant.
We have found that, over a wide range of actin conrentration, double recil)rocal plots of ATPase against AT1 concentration are linear, and the system can therefore be analyzed in terms of the steady stat,e kinetics of a simple enzyme-substrate-modifier model as rcceutly described by London (8).

METlIODS
Protein Preparations-Actin and JTMi\l were prepared as wc described previously (6,9). Creatine kinase was prepared from rabbit muscle by the method of Kuby,Noda,and Lardy (10) with the addition of a final treatment with Norit-A charcoal to remove contamiuant nucleotide (11). All protein corlccrrtrations were determined as previously (9) by ultraviolet absorlrtion at 280 nip. The extinction coefficient used for creatine kiriase was 890 cmZ per g (12).
11 TPase '4 &Z&J-In all experiments the ATP concerrtration was maintained with the use of the creatine kinase-creatirre phosphate system. In this coupled system, the rate of ~2'1'1' hydrolysis is equal to the rate of breakdown of the creative phosphate, and this reaction in turn consumes approximately 0.3 mole of H+ per mole of creatine phosphate degraded at 1~1~1 7 ; therefore, we could measure the rate of the reaction with tlie pa-stat with HCl as the titrant.
As in our previous work (6, 9), we allowed the reaction to continue to completion to detcrmirre the total amount of HCl consumed during hydrolysis of a kno~vn total amount of creatine phosphate, and from this value WC determined the exact equivalence between creatine phnsl)h;rt,c breakdown and HCI added.
Since the measured ATl'ase rates were always coristant throughout the reaction eveu when the creatine phoslrhate 1. Dependence of the HMM ATPase on added ATP concentration at fixed actin concentrations.
Conditions: 2.5 rn~ MgCln, 2.5 rn~ creatine phosphate, 1 mg of creatine kinase per ml, 0.036 rng of ITMM per ml.  had dropped far below its Michaelis constant, it is clear that the rate of the over-all coupled reaction was not limited by the rate of the creatinc kinase reaction; i.e. the creatine kinase concentration was sufficient to maintain essentially all of the free nucleotide as ATP.
Furthermore, since creatine kinase has a relatively high turnover rate per mole of enzyme and a relatively weak affinity for ATP (13), the complexes of ADP and ATP with creatine kinase represented a negligible fraction of the total nucleotide even at the lowest ATP concen tration used.

RESULTS
In Fig. 1, the measured ATPase rate of a mixture of actin and HMM is plotted against the added ATP concentration.
The resulting curve appears to obey a simple Michaelis hyperbolic relation except that a small but significant ATPase activity is observed even when no ATP is added. This occurs because, at very low ATP concentrations and relatively high actin con centrations, the small amount of free nucleotide inevitably present in the actin solution makes a significant contribution to the total substrate in the reaction mixture.
Not only is a small portion of the actin-bound ADP always released into solution, but it is also impossible to remove all traces of the free ATP which is added during preparation of the F-actin solution, and these small amounts of contaminant nucleotide are, of course, maintained as ATP by the creatine kinase-phosphocreatine system.
Before further kinetic studies could be carried out, it was necessary to determine the contaminant ATP concentration at each actin concentration so that we could know the true substrate concentration in the reaction mixtures. This was done by the following kinetic method, which is based on the assumption that the system does indeed obey Michaelis kinetics.
The Michaelis equation can be rearranged to give Equation 1, in which 2, = the ATPase rate; V',,, = the ATPase rate at infinite ATP concentration with the given actin concentration; K' = the apparent Michaelis constant; and the total ATP concentration is given by the sum of X', the added ATP concentration, and X", the contaminant ATP concentration.
First, the contaminant ATP concentration was very roughly estimated by extrapolating plots like those in Fig. 1 to their intercepts on the abscissa.
These values were used to calculate the approximate total ATP concentrations so that Lineweaver-Burk double reciprocal plots of ATPase rate against ATP concentration could be constructed for determination of V'lr,&X.
Since these plots were most accurate at high ATP concentration, a precise value of V',,, could be obtained by extrapolation to the ordinate.
With this value of V',,,, the data could then be plotted as X' against v/(V',, -V) according to Equation I and extrapolated to the value of -8" given by the intercept on the ordinate.
As illustrated in Fig. 2 (1, 14), we could only know the free RTP concentration with an accuracy of about *0.2 pM. Therefore, for the double reciprocal plots which follow, data were not taken at ATP concentrations below about 1.5 PM.
Double reciprocal plots based on the data shown in Fig. 2, as well as data obtained at two intermediate actin concentrations, are shown in Fig. 3a. As can be seen, at all four actin concentrations the plots were linear over a range of ATP concentration from 1.5 PM to 50 PM. Furthermore, as shown in Fig. 4, Hill plots of the data are also linear with a slope equal to 1. It would therefore appear that, over a wide range of ATP and actin concentrations, the acto-HMM ATPase system obeys simple Michaelis-Menten kinetics. In addition to the plots of l/v against l/ATP at varied actin concentration shown in Fig. 3~2, a complete kinetic analysis requires plots of l/v against l/actin at various fixed ATP concentrations.
Since the contaminant ATP concentration was different at different actin concentrations, it was difficult to vary the actin concentration at constant ATP concentration.
Therefore, data were taken from the plots of l/v against l/ATP shown in Fig. 3a to construct the plots of l/v against l/actin which are shown in Fig. 3b. As can be seen, these plots were also linear over the range of ATP concentration tested.
It was not possible to determine whether the sets of lines in Fig. 3, a or b, were parallel or whether they intersected somewhere in the lower left quadrant; however, it is clear that in neither Fig. 3a nor Fig. 3b do the plots intersect on the ordinate or in the upper left quadrant, a finding which is important for our later discussion of the velocity equations applicable to this system.
In effect, the ordinate intercepts of the plots in Fig. 3a give a measure of the dependence of the ATPase rate on actin concentration at infinite ATP and, analogously, the ordinate intercepts of the plots in Fig. 3b give a measure of the dependence of the ATPase rate on ATP concentration at infinite actin. These intercepts are plotted in double reciprocal form in Fig. 5, a and  turns out to 1~ about 4 pnloles 1)er mg-rnin, in agreernellt with the value for TiTlliLy that we ljreviously obtained from 1 /v against 1 /actin plots at high corlcelltl,:\tiolls of ;\Tl' (6). The interceljt on the abscissa of Fig. 5b gives t,he ,\Iichaelis constant of the acto-HP\IhI ATl'ase and, analogously, the intercept 011 the abscissa of Fig. 5n gives the dissociation rolnstant of actin from H!VILLI at infinite 11Tl concentration, i.e. of actin from the IIXIlWATl' complex.
Assuming a nlolccular weight of 47,000 for actin, the latter dissociation constant is found from Fig. 5a to be about 13 pM. This is a smaller dissociation constant than we found in our previous st,udies (6), but this was expected because the ionic strength was lower in the present work. Nevertheless, the dissociating effect of LiTP 011 acto-HMM is still evident if this value is compared with the dissociation constant of about 2 ~IV obtained by Young in t,he absence of hT1' (I 5), although it should be noted that Young's dissociation constant was determined as a lower temperature and a higher ionic strength than we used here.
The Michaelis constant of the a&o-H&W AT&se was found from the abscissa intercept of Fig. 5b to be about 6 pM, which is larger than t'he reported values of 0.5 to 1 FM for the Michaelis constant of the HMM ATPase (lr 14). It would therefore appear that, just as ATP increases the dissociation constant of I  I  I  I  I  I   I  I  I  I  I  -2 Conditions: same as in Fig. 3  actin from HMM, so too dots actin increase the Michaelis COW st'ant of the HMM ATPase. We next investigated the relative effects of ionic strength on the dissociation of actin from III\IlWATP and on the Michaelis constant of the acto-HMM ATPass. Fig. 6, a and b, shows plots of 1 /f) against l/ATP at varied actin and l/v against l/actin at varied XTP, respectively, at an ionic strength about twice that used in Fig. 3, and here again both sets of plots are linear. The plots in Figs. 3 and 6b have markedly different slopes (note difference in abscissa scale), as is expected from our previous fillding that the binding of actin to ITMM-ATI' shows a marked salt dependence (6) ; however. the plots of 1 /v against l/AT1 in Figs. 3a and 6a have essentially identical slopes, which suggests that the Michaelis co&ant of the acto-HMM ATPase shows little or no salt dependence.
These conchsions are confirmed in Fig. 7, a and b, which shows reciprocal plot's of the ordinate intercepts of Fig. 6, a and b, respectively, as well as similar data at three other ionic strengths.
In two of the other plots, the ionic strength was altered by varying the KC1 concentration, \vhile ill the third the creatine phosphate concen- T~YZ ordinate intercepts of Fig. 7, a and A, show that ionic strength has essentially no effect on V,,,nX, while the abscissa intercepts of Fig. 7a show t'he marked effect of ionic strength on the dissociation constant of actirr from the HM_\I-ATP complex-both of these effects being in agreement with out previous findings from double reciprocal plots of ATI'ase against actin at high 9Tl' concentration (6). Howevrr, the abscissa intercepts of Fig. 76 show, as we suggested above, that ionic strength has no significant effect on the Michaelis constant of the acto-HMM ATPase.
There is thus a clear difference in the effects of ionic strength on the dissociation constant of actill from ETMM-ATP and on the ;\Iichaelis constant of the ncto-I-InIkI XTl'ase.
The foregoing data are considered in lehtion to the kinetic model which we have previously applied to this system (6, 7). This model is shown below, together with definitions of certain constants which we will find useful: MS k, M + products In this model, ,I represents actin, df represents HMM, and S represents the substrate, ATP. Depending on the relative values of t,he rate constants in the above kinetic model, many different rate equations based on the model can be written.
Only a few of these equations, however, are consistent with linear l/v against l/S plots, and since u-e find linear l/v against I./S plots for our system, these are the only equations with which we need be concerned.
In a recent paper, London (8) has listed the 11 velocity equations based on the above type of kinetic model which give linear plots of l/v against l/X at various concentrations of modifier (actin). Of these equations, some can be ruled out because they require that the plots of I/V against l/S at different actin concentrations intersect at a single point on the ordinate or in the upper left quadrant, which clearly does not occur with our system. Others can be ruled out either because they require actin to inhibit rather than activate the IIMM ATPase, or because they reyuire that Krl or Kz be infinite, which would contradict the obvious fact that HMM can bind to actin in the absence of nucleotide and can hydrolyze nucleotide in the absence of actin.
Finally, several equations can be A ctin A ctivation of Heavy Merom yosin A TPase Vol. 245, Nb. 9 ruled out because they are not consistent with our finding that plots of l/v against l/actin at varied ATP concentrations are linear. It therefore turns out that, in all likelihood, Equation 2 below is the only velocity equation which is consistent both with our data and with the above kinetic model.
v Our experiments were done at actin concentrations comparable to KS, the dissociation constant of actin from the HMXATP complex (abscissa intercept of Fig. 5a or 7~). On the other hand, kg, which is the same as V,,, in our plots, is some 200-fold larger than ks, the ATPase activity of IIMM alone (6). Therefore, ksKS << &A, and Equation 2 is consistent with our finding of linear plots of l/v against 1 /actin.
It also predicts that all of the plots of l/v against l/S at different actin concentrations will intersect each other at a single point in the lower left quadrant of the graph, as will all of the plots of l/v against l/actin. Our data are not accurate enough at low ATP and low actin concentration to permit these extrapolations; however, in preliminary experiments with ITP rather than ATP as the substrate, it does indeed appear that the plots of 1 /v against 1 /ITP intersect each other at a single point, as do the plots of 1 /v against 1 /actin in the presence of ITP when the plots are corrected for the appreciable ITPase activity of HMM alone. 2 If we make the one further assumption that the rates of the reactions involving the binding of ATP or actin to HMM alone are not negligible in our steady state system, then Equation 3 below, which London (8) has described as an analogue of detailed balance, follows directly from Equation 2 and should hold true for our system.3

Ka
K: -= 7 Kz KI (3) The first point to be noted about this equation is that if the binding of ATP to HMM weakens the binding of actin, i.e. if KS > Rz, then it is required that the Michaelis constant of the acto-HMM ATPase (K',) be greater than the Michaelis constant of the HMM ATPase (K',).
In fact, as we noted earlier, the data from Fig. 5, a and b, indicate that this is indeed the case, at least qualitatively, for our system. Equation 3 also makes another prediction. Since we found, as shown in Fig. 7b, that the Michaelis constant for the acto-HMM ATPase shows no dependence on ionic strength, and since the Michaelis constant of the HMM ATPase also shows no salt dependence (14), it follows that the left-hand side of Equation 3 must also be independent of ionic strength.
Therefore, since we showed in Fig. 7a that the dissociation constant of actin from HMM-ATP shows a marked salt dependence, it follows from Equation 3 that the dissociation constant of actin from HMM in the absence of ATP must show the same marked salt dependence.
It is not known at the present time whethei the binding of actin to HMM does indeed show a marked dependence on ionic strength.
If it does not, however, then it follows either that one or more reactions in the above kinetic model occur at negligible rates, or that this simple kinetic model is not sufficient to describe our system and a more complex model is required.
Application of this test of our kinetic analysis will have to await an accurate independent measure of the binding between actin and HXIM in the absence of nucleotide at varied ionic strength.
,4cknowledgmentsWe wish to thank Miss Alice Harlow for skillful technical assistance and Dr. Wayne P. London for many helpful discussions during preparation of the manuscript.