(Received for publication, October 1, 1996, and in revised form, December 19, 1996)
From the Institute of Molecular Biology, Austrian Academy of Sciences, Billrothstrasse 11, A-5020 Salzburg, Austria
A myofibrillar form of smooth muscle myosin light chain phosphatase (MLCPase) was purified from turkey gizzard myofibrils, and it was found to be closely associated with the myosin light chain kinase (MLCKase). For this reason we have named this phosphatase the kinase- and myosin-associated protein phosphatase (KAMPPase). Subunits of the KAMPPase could be identified during the first ion exchange chromatography step. After further purification on calmodulin (CaM) and on thiophosphorylated regulatory myosin light chain affinity columns we obtained either a homogenous preparation of a 37-kDa catalytic (PC) subunit or a mixture of the PC subunit and variable amounts of a 67-kDa targeting (PT) subunit. The PT subunit bound the PC subunit to CaM affinity columns in a Ca2+-independent manner; thus, elution of the subunits required only high salt concentration. Specificity of interaction between these subunits was shown by the following observations: 1) activity of isolated PC subunit, but not of the PTC holoenzyme, was stimulated 10-20-fold after preincubation with 5-50 µM of CoCl2; 2) the pH activity profile of the PC subunit was modified by the PT subunit (the specific activity of the PTC holoenzyme was higher at neutral pH and lower at alkaline pH); and 3) affinity of the holoenzyme for unphosphorylated myosin was 3-fold higher, and for phosphorylated myosin it was 2-fold lower, in comparison with that of the purified PC subunit. KAMPPase was inhibited by okadaic acid (Ki = 250 nM), microcystin-LR (50 nM) and calyculin A (1.5 µM) but not by arachidonic acid or the heat-stable inhibitor (I-2), which suggested that this is a type PP1 or PP2A protein phosphatase.
Contraction of smooth muscle is regulated by a myosin-linked regulatory system that involves phosphorylation of the 20-kDa light chain of myosin (ReLC)1 by myosin light chain kinase (MLCKase) (1). This phosphorylation is a prerequisite for the actin-activated MgATPase activity of myosin, which is considered essential for the initiation of smooth muscle contraction (2, 3). Furthermore, phosphorylation of the ReLC influences the assembly and stability of myosin filaments (4, 5), which must also be related to tension development in vivo. Consistent with such a regulatory system, it has also been shown that MLCKase and its activator, calmodulin (CaM), are tightly bound to filamentous myosin (6), but their affinity for monomeric or folded myosin is low (7). However, it should be noted that the high affinity binding and high content of MLCKase and CaM in filamentous myosin are not generally recognized (see Ref. 8).
Relaxation of smooth muscle is associated with dephosphorylation of myosin (for review, see Refs. 3, 9, and 10), and therefore it was clear very early that another regulatory enzyme, a myosin light chain phosphatase (MLCPase), must be responsible for relaxation in vivo. Several phosphatases have been isolated from smooth muscle, which could act to dephosphorylate myosin or its ReLC. Many of these enzymes seem to be cytosolic, because they have been purified from whole smooth muscle extracts. For example, only two of the four protein phosphatases identified by Pato et al. (11, 12) in the cytosol of turkey gizzard exhibited high activity toward native myosin and bound tightly to myosin, suggesting that they may be localized on the thick filaments. The properties of these and other phosphatases with respect to the effects of divalent cations and substrate specificity have been investigated. The emphasis in all of these studies has been to establish the differences among the various phosphatases, although from the introduced classification (13) and subunit composition, the similarities were also apparent. This aspect of the MLCPase identification has not been investigated. As a result, the phosphatase specifically responsible for the dephosphorylation of myosin in vivo has not yet been identified.
It might be expected that a myofibrillar phosphatase that is tightly associated with the contractile apparatus, such as the enzyme copurified with MLCKase (14), would be the most likely to serve as the phosphatase responsible for the dephosphorylation of myosin in vivo. In the present report, we characterize a holoenzyme of this MLCPase, which, as we show, is bound to calmodulin in a Ca2+-independent manner. Formation of a complex between the holoenzyme and MLCKase is addressed in the accompanying report (15).
[-32P]ATP was
purchased from DuPont NEN and diluted with cold ATP of special grade
(catalog number 519 987; Boehringer Mannheim) to the required specific
activity and concentration (16). Q-Sepharose CL-6B, Sephacryl S-100 HR,
and CNBr-activated Sepharose 4B-CL were obtained from Pharmacia Fine
Chemicals (Uppsala, Sweden). Electrophoresis reagents were purchased
from Bio-Rad. The other chemicals were of analytical grade and, unless
otherwise stated, were purchased from E. Merck (Darmstadt, Germany) or
Fluka (Buchs, Switzerland).
Common inhibitors of protein phosphatases such as microcystin-LR and calyculin A were purchased from LC Service Corp. (Woburn, MA). Okadaic acid was obtained from Boehringer Mannheim. The heat-stable inhibitor (I-2) was a gift of Dr. S. Shenolikar from the Department of Pharmacology, Duke University Medical Center (Durham, NC).
The buffer (AA) used throughout this study had the following composition: 60 mM KCl, 2 mM MgCl2, 0.5 mM dithiothreitol, and 10 mM imidazole with pH adjusted to 7.5 at 4 °C. It was used, with the necessary additions, during protein purification, activity measurements, binding experiments, and all other procedures.
Protein PreparationsTurkey gizzard MLCKase (14), myosin
(17), CaM (18), and ReLC (19) were prepared as described previously.
32P labeling of the isolated ReLC, used as the substrate,
was carried out by adding 25 nM of purified MLCKase and 35 nM of CaM to the light chain at a 500 µM
concentration. After 90 min, the phosphorylation reaction was
terminated by placing the tube in boiling water for 6 min, and the
radioactive ATP was removed by exhaustive dialysis against our AA
buffer. The 32P-labeled preparations of ReLC were stored at
30 °C, and before use they were usually diluted 3-10-fold with
the AA buffer to obtain the required level of cpm.
Before any type of measurements, purified KAMPPase preparations were first diluted, to the required concentrations, with AA buffer containing 0.1% of bovine serum albumin added as an inert protein carrier. Phosphatase activity was measured by the release of 32P from 32P-ReLC used as a substrate as described previously (14). The assays were carried out at 25 °C in the AA buffer, and concentrations of the substrate (32P-ReLC) were in the range of 50-150 µM.
MLCKase assays were carried out with [-32P]ATP using
intact myosin or its isolated ReLC as substrates as described
previously (16, 18). Details of specific assays are given in the
corresponding figure legends.
The affinity of KAMPPase for myosin filaments was determined by using the sedimentation method (20). In our case, increasing amounts of myosin were added to 50 µl of a KAMPPase preparation, and, after centrifugation, the KAMPPase activity in the supernatants was measured. Concentration of the phosphatase was chosen to ensure that not more than 30% of the substrate was dephosphorylated in the controls (no myosin present). The reciprocal of the fraction of free KAMPPase was plotted against the myosin concentration. This was a little more convenient than the original procedure.
Ligand Coupling and Other Standard ProceduresThe coupling
of the ligands to CNBr-activated Sepharose 4B was done as recommended
by the manufacturer (Pharmacia, Uppsala, Sweden). The unphosphorylated
ReLC was coupled to the Sepharose, and this was thiophosphorylated for
60-90 min by adding 0.25 mM [-S]ATP (Boehringer
Mannheim) and 20-40 nM of CaM·MLCKase. The gels were
packed into a column approximately 0.9 × 60 cm. Stability of the
column was such that it could be used for four to six applications with
2-3 rethiophosphorylations within a few months.
SDS-polyacrylamide gel electrophoresis was performed in 8-18% gradient acrylamide minislab gels according to the procedure of Matsudaira and Burgess (21) in the buffer system of Laemmli (22) with some improvements described recently in detail by Sobieszek (23). Protein concentration was determined by the methods of Gornall et al. (24) and Bradford (25) by using Bio-Rad protein standard. Unless otherwise stated, all purification steps were carried out on ice or in the cold room (4 °C).
The 40-55% ammonium sulfate
fraction obtained from the myofibrillar MLCKase and MLCPase extract
was purified by chromatography on a Q-Sepharose column rather than on
previously used DEAE-Sepharose 6B-CL. The advantage of the Q-Sepharose
column was that tropomyosin was separated from the MLCPase. As shown in
Fig. 1A, tropomyosin coeluted with the
MLCKase and not with the MLCPase activity peak as it does from
DEAE-Sepharose column. Activity measurements showed that, for both
columns, most of the kinase activity could be separated from the
phosphatase, but the reverse was not true; the phosphatase was always
heavily contaminated by the kinase. The first peak from the Q-Sepharose
column contained essentially homogenous MLCKase (with the residual
tropomyosin), while the phosphatase constituted only a small portion of
the total protein in the second peak (Fig. 1B).
Significantly, this peak always included CaM. The separation of
tropomyosin from the MLCPase is important because the viscosity of
tropomyosin could impair further purification of the enzyme. The
purification factor for the ionic exchange column was about 800-fold
(80% yield), making it possible to detect, on SDS-PAGE, the bands that
correlated with the MLCPase activity (Fig. 1A; i.e. PT and PC).
In the initial attempts, the Q-Sepharose fractions that contained the
highest MLCPase activity were subjected to a standard purification on a
SP-ReLC affinity column. Unexpectedly, the phosphatase that eluted
at 180 or 360 mM salt was composed of three subunits with
molecular masses of 37, 67, and 130 kDa (Fig. 2,
lane a). From the absence of the 130-kDa band in the low
ionic strength washes from this column and the presence of high MLCKase activity in the eluate, the 130-kDa band was identified as MLCKase. Further attempts to purify this kinase/phosphatase fraction on hydroxylapatite and/or ion exchange chromatography produced no separation of these enzymes.
Because of the high affinity of the kinase for CaM and the independence
of the phosphatase activity from this activator, the alternative
purification step used was CaM affinity chromatography. As shown in
Fig. 3, however, the MLCKase still contained some MLCPase activity, and the expected complete separation of these two
enzymes was again not achieved. It was clear from this and from other
attempts of the type shown in Fig. 3, that the MLCKase always and to a
variable extent, copurified with the phosphatase. The opposite was not
true, and the "breakthrough" phosphatase that was not bound by the
CaM affinity column was typically free of the kinase activity. The
fractions bound by the column and eluted at high salt concentration
always contained both MLCPase and MLCKase activities. Moreover, the
same was true for the standard MLCKase preparations
(e.g. see Fig. 3) eluted from the CaM affinity column in the
absence of Ca2+. Because the apparent very tight
association of the MLCPase activity with the kinase also was observed
with MLCKase bound to myosin filaments (see below), we refer to this
MLCPase as the kinase- and myosin-associated protein phosphatase
(KAMPPase). The properties of the complex formed between the two key
regulatory enzymes of smooth muscle contraction is the subject of the
companion report (15). The present article focuses on the
characteristics of the KAMPPase holoenzyme (PTC) and its isolated
catalytic subunit (PC).
Catalytic Subunit and KAMPPase Holoenzyme
The breakthrough fractions from the CaM affinity column of the Q-Sepharose KAMPPase always contained a relatively high level of phosphatase activity. The amounts of the bound and unbound activities were variable and depended on the amount of proteolytic degradation. This, in turn, introduced some variability in appearance of the KAMPPase on SDS-PAGE during purification.
The breakthrough phosphatase was subjected to three purification steps,
which included a long SP-ReLC affinity column (Fig. 4A) and a short DEAE-Sepharose column (Fig.
4B) followed by a 0.6 × 110-cm Sephacryl S-100HR gel
filtration column. If a relatively higher degree of proteolysis had
occurred, this three-step column procedure resulted in a homogenous
KAMPPase preparation, which exhibited only one subunit of 37 kDa (Fig.
2, lanes b and c). With less proteolytic
degradation, the analogous preparation contained an additional subunit
of 67 kDa (Fig. 2, lane f). The relative ratio of the 37- and 67-kDa subunits, therefore, varied from preparation to preparation
and depended on the extent of proteolytic degradation that occurred.
Unless otherwise indicated, the complex composed of these two subunits
was used throughout the experiments described below. We refer to it as
the KAMPPase holoenzyme or simply KAMPPase.
The purified 37-kDa subunit contained all of the KAMPPase activity and was therefore identified as the catalytic (PC) subunit of the KAMPPase. It eluted as a single peak at a position corresponding to ~67 kDa when run on a calibrated Sephacryl S-100 HR or AcA54 column equilibrated with AA buffer, containing or not containing 0.3 M NaCl. The purified PC subunit showed a closely spaced doublet on SDS-PAGE (Fig. 2, lane b), which would be consistent with the dimer formation. However, a single band of this size was observed in the initial purification steps (Fig. 1A). The doublet could then result from a slight proteolytic degradation of the subunit.
CaM Affinity-bound PhosphataseAfter extensive washing at low ionic strength, and before EGTA elution, the CaM affinity column was eluted at 180 or 360 mM salt concentration, both in the presence of 0.2 mM calcium (Fig. 3). The composition of the eluted peak was complex and, as before, as a result of proteolysis, the band patterns differed from preparation to preparation. Nevertheless, the 37- and 67-kDa subunits were always present as the main protein bands (Fig. 2, lanes d and e).
Further purification of the CaM affinity-bound 37-67-kDa complex on the short DEAE-Sepharose column produced a relatively homogenous preparation of the PTC holoenzyme (Fig. 2, lane f). The two subunits comigrated during the elution, and their positions correlated well with the KAMPPase activity. The same was true during the subsequent purification by gel filtration (see above), which served to remove some minor impurities, e.g. tropomyosin (Fig. 2, compare lanes f and g). No difference was noted between this KAMPPase and the one purified from the breakthrough fraction, provided that both contained the PT subunit.
Localization of Myofibrillar MLCPase on MyosinWe have
previously shown (6) that MLCKase and CaM are very tightly bound to
filamentous smooth muscle myosin. As a result, the addition of
Ca2+ and ATP to purified myosin leads to a very rapid
phosphorylation of the ReLC. In the same experiments, we also showed
that the MLCPase was also associated with the same type of
(filamentous) myosin preparation. As shown in Fig.
5A, after initial rapid phosphorylation by
the kinase, the myosin became dephosphorylated by the endogenous phosphatase. Both enzymes were simultaneously active, and the resulting
phosphorylation rate and extent depended only on the kinase and the
phosphatase activities associated with the myosin. However, as soon as
ATP was depleted, the kinase could not act, and the phosphatase action
became apparent (Fig. 5A). For longer incubation times, this
resulted in biphasic phosphorylation curves (see Ref. 15).
Correspondingly, the duration and the levels of phosphorylation could
be increased by adding phosphatase inhibitor (microcystin-LR) or
decreased after adding purified KAMPPase (Fig. 5A) or the PC
subunit. This demonstrates clearly that the KAMPPase was very active
toward phosphorylated filamentous myosin and could be responsible for
the dephosphorylation of myosin filaments in vivo.
At the myosin concentrations (30-50 µM) used in these experiments, most of the ATP hydrolysis was due to myosin ATPase. This is not the case in similar experiments with MLCKase autophosphorylation, where only the kinase is present (26). By adding different concentrations of the KAMPPase (or the PC subunit) to the kinase, we could decrease both the autophosphorylation level and the initial rate (Fig. 5B). In contrast, the addition of microcystin-LR resulted in 2-fold (or greater) increase in the autophosphorylation even for homogenous MLCKase preparations (Fig. 5B). Thus, our KAMPPase is also very active toward MLCKase autophosphorylation sites.
Divalent Cation Requirements and Inhibition by Common Protein Phosphatase InhibitorsThe activity of the isolated PC subunit
was stimulated by preincubation in micromolar concentrations of
CoCl2. The 10-20-fold stimulation was observed with the
two substrates used and depended more on the preincubation time than on
the Co2+ ion concentration. It was observed only for the
purified PC subunit and not for the PTC holoenzyme (Fig.
6). In agreement with a recent observation by Okubo
et al. (27), the activity of the phosphatase copurified with
myosin was also stimulated by CoCl2. The stimulation was,
however, rather variable and consistently lower than that observed for
the purified PC subunit.
The effects of other divalent cations on the activity of the PC subunit were also investigated (data not shown). As expected for a type PP1 phosphatase, other cations were not required for activity, although some stimulation was observed. The stimulation occurred in the 0.5-2.5 mM concentration range and amounted to not more than a 1.5-fold increase for Mn2+, Mg2+, and Ca2+. Heavy metal cations such as Ni2+ inhibited the activity at approximately the same concentration range, while a 50% inhibition by Cd2+ required only 50-100 µM of CdCl2.
Unexpectedly, the heat-stable protein inhibitor (I-2) at concentrations
of up to 5 µg/ml had no effect on the activity of the PTC holoenzyme
(Fig. 7A) or the PC subunit. Okadaic acid, a
potent and specific inhibitor of the type PP1 and PP2 protein phosphatases, inhibited the activity of both the PTC holoenzyme and the
PC subunit (Fig. 7A) with an IC50 value of about
250 nM, which is 25 times higher than that characteristic
for a type 1 protein phosphatase (28, 29).
We have also tested the inhibition of the KAMPPase by other potent protein phosphatase inhibitors commonly used for smooth muscle fiber experiments (Ref. 30; see also Ref. 29). One of them, arachidonic acid, had no effect on the KAMPPase activity (Fig. 7B). With calyculin A, the inhibition occurred in the 1-3 µM concentration range (Fig. 7A). The most effective inhibitor was microcystin-LR (Fig. 7A), with full inhibition obtained at approximately 0.250 µM. Thus, okadaic acid, calyculin A, and microcystin-LR appear to be the only inhibitors suitable for in vivo experiments on smooth muscle fibers or muscle strips.
Targeting Subunit and KAMPPase ActivityThe effect of the PT
subunits on the specific activity of the PC subunit was not accurately
established because the PT subunit could not be completely purified
without proteolytic degradation or contamination with the phosphatase.
However, indirect estimations indicated a 3-4-fold stimulation of the
PC subunit after the addition of partially purified PT. This
stimulation depended on ionic conditions and pH. As shown in Fig.
8, the pH activity profile of the PC subunit was
completely different from that of the PTC holoenzyme. Significantly,
the difference amounted to a shift of the activity optimum from basic
pH values to neutral, physiologically relevant values in the case of
PTC. The increase in the activity at basic pH was also not observed for
the endogenous myosin phosphatase. Thus, in this respect, the
endogenous myosin phosphatase was similar to the PTC holoenzyme and not
to the isolated PC subunit.
The effect of the PT targeting subunit on the binding of the KAMPPase
to myosin was also investigated. We measured the affinities of the PC
subunit and the PTC holoenzyme for myosin filaments assembled from
unphosphorylated and thiophosphorylated smooth muscle myosin. The
affinity of the PC subunit for unphosphorylated myosin was moderate
(Kd = 15 µM) but increased about 3-fold in the presence of the PT subunit (Fig. 9). The
opposite effect was observed with thiophosphorylated myosin. The
relatively high affinity of the PC subunit (Kd = 4 µM) was 2-fold lower in comparison with that of the PTC
holoenzyme (Fig. 9). Although these effects were not large, they
indicate that the regulatory role of the PT subunits also includes
modification of the KAMPPase affinity for the in vivo
substrate (filamentous myosin).
The identification and partial purification of a myofibrillar smooth muscle myosin MLCPase was first described by Sobieszek and Barylko (14), although a similar phosphatase was noted earlier (31). Until recently, much effort has been made to identify and purify cytoplasmic type smooth muscle myosin phosphatases, and less attention has been paid to the myofibrillar MLCPases. From its close association with MLCKase, we suggested that myofibrillar MLCPase is more likely to be responsible for the relaxation of smooth muscle.
The present report describes the further purification and characterization of turkey gizzard myofibrillar MLCPase denoted here as KAMPPase. The KAMPPase was purified as a PTC holoenzyme composed of the two subunits or as a large multienzyme complex, which included MLCKase and CaM (15). In addition, a homogenous preparation of the PC catalytic subunit was obtained when the targeting PT subunit was degraded during purification. The phosphatase must be considered as being myofibrillar in origin because the starting material (myofibrils) was obtained by extensive fragmentation and washing of gizzard muscle, as originally described by Sobieszek and Bremel (32). The myofibrils are practically depleted of cytosolic proteins, and only tightly bound enzymes (including the phosphatase considered here) are extracted with the AA buffer containing 25 mM MgCl2. As originally indicated (33), adding MgCl2 to the extraction medium represents a significant improvement of the purification procedure and is generally used in extraction of the MLCKase (e.g. see Refs. 31 and 34).
The procedure used to purify the KAMPPase included a new modified
affinity chromatography approach on CaM and on SP-ReLC affinity
columns. The fragmentation and extensive washing prior to KAMPPase
extraction (see Ref. 32) made it possible to identify the KAMPPase
bands on SDS-PAGE of fractions obtained during the first purification
step on the ion exchange column. The use of CaM affinity chromatography
to purify MLCPase has not been reported previously and is a novel
aspect of this study.
The KAMPPase holoenzyme obtained after two affinity columns exhibited, on SDS-PAGE, the two subunits of 37 and 67 kDa. In agreement with previous reports (13, 28, 35, 36), the 37-kDa protein was identified as the catalytic subunit. This subunit possessed all of the phosphatase activity. The role of the 67-kDa subunit was established as the CaM-targeting one because it was responsible for binding the PC subunit to the CaM affinity gel; the most illustrative data is presented in the companion paper (15).
As expected from its tight association with MLCKase, the PC subunit and the PTC holoenzyme effectively dephosphorylated not only intact myosin but also the kinase itself. The dephosphorylation rates for the three substrates investigated were similar and, for example, intact myosin was dephosphorylated only 1.5-fold faster than the isolated ReLC. The corresponding Km values were also similar (3-10 µM). We demonstrated that a similar phosphatase was tightly associated with purified MLCKase and could not be completely removed even after extensive attempts to separate the two. This endogenous phosphatase effectively dephosphorylated MLCKase autophosphorylation sites but not the well characterized sites phosphorylated by cAMP-dependent protein kinase or Ca2+/CaM-dependent multifunctional kinase II (26). We believe that the latter endogenous phosphatase as well as the one associated with filamentous myosin are the same or similar enzymes, possessing the same catalytic subunit.
Association with MyosinBecause the affinity of the PC
subunit for myosin was only moderate (Kd 15 µM), it would appear that the tight association of
endogenous MLCPase with myosin requires the presence of another
subunit. Our results suggest that the PT targeting subunit could
fulfill this role, since the other components needed for the
interaction (CaM or CaM·MLCKase complex) are known to be firmly
incorporated into myosin filaments (see Ref. 6). Recently, Mitsui
et al. (37) reported the purification of a smooth muscle
phosphatase (myosin-associated protein phosphatase I) from chicken
gizzard myosin. Since we have also concluded that our KAMPPase is
associated with myosin, it is likely that these two phosphatases are
similar if not identical. Both enzymes are tightly associated with
filamentous myosin, exhibit similar size during gel filtration on
Sephacryl S-100HR columns, have catalytic subunits of approximately the
same size, and have a similar elution position (170 kDa) from a gel
filtration column. A targeting subunit of 67 kDa, however, was not
identified for the myosin-associated protein phosphatase I phosphatase
(37). The 3-fold increase in the affinity of the PC subunit for myosin
in the presence of the PT subunit supports our hypothesis for the
existence of and the role for the PT subunit. However, an independent
demonstration of the presence of this targeting subunit on myosin
filaments is still lacking. The observed CaM-dependent but
Ca2+-independent targeting by this subunit is also
consistent with the independent binding of CaM and the kinase to myosin
filaments (6).
During the last
10-15 years, several cytosolic MLCPases have been purified from smooth
muscle extracts, and many of them appear to be similar to our KAMPPase.
Its relation to the only other myofibrillar phosphatase of Alessi
et al. (38) is discussed in the companion paper (15), and it
is only briefly considered here (see below). A purified aortic smooth
muscle myosin phosphatase, composed of two subunits of 67 and 37 kDa
(39), is structurally similar to the myosin phosphatase from cardiac
muscle (40). Another phosphatase, which is composed of three subunits
(67, 54, and 34 kDa), was purified from chicken gizzards by Onishi et al. (41) and by Pato and Adelstein (42). They both
dephosphorylated isolated ReLC, but only the former
dephosphorylates myosin. Nevertheless, they may represent the same
enzyme, since both were purified by SP-ReLC affinity
chromatography. The same comment could be applied to the MLCPases
purified by Di Salvo et al. (43). Taken together, it appears
that most smooth muscle myosin phosphatases have a catalytic subunit of
about 37 kDa. Similarly, it can be concluded that a subunit of 60-67
kDa is a constant feature of this type of phosphatase and that the band
of 54-58 kDa sometimes observed may represent a proteolytic fragment
of this subunit.
During the course of the present study, two papers were published that describe another myofibrillar type of MLCPase from gizzard muscle (Ref. 38; see also Ref. 44). The holoenzyme of that enzyme is composed of a regulatory subunit of 130 kDa, a catalytic subunit of 37 kDa, and an unidentified component of 20 kDa. Differences in the methods of preparing myofibrils and in extraction conditions make it difficult to determine the relationship of that enzyme to ours. However, the absence of the 67-kDa targeting subunit indicates that the MLCPase described by Alessi et al. (38) and by Shirazi et al. (44) is different from the enzyme we have isolated.
We thank Professor Malvin Stromer for corrections to the manuscript.