(Received for publication, February 11, 1997, and in revised form, April 3, 1997)
From the Department of Physiology and Biophysics, School of Medicine, Case Western Reserve University, Cleveland, Ohio 44106-4970
Myosin heavy chain kinase A (MHCK A) participates in the regulation of cytoskeletal myosin assembly in Dictyostelium, driving filament disassembly via phosphorylation of sites in the myosin tail. MHCK A contains an amino-terminal coiled-coil domain, a novel central catalytic domain, and a carboxyl-terminal domain containing a 7-fold WD repeat motif. We have overexpressed MHCK A truncation constructs to clarify the roles of each of these domains. Recombinant full-length MHCK A, MHCK A lacking the predicted coiled-coil domain, and MHCK A lacking the WD repeat domain were expressed at high levels in Dictyostelium cells lacking endogenous MHCK A. Biochemical analysis of the purified proteins demonstrates that the putative coiled-coil domain is responsible for the oligomerization of the MHCK A holoenzyme. Removal of the WD repeat domain had no effect on catalytic activity toward a synthetic peptide, but did result in a 95% loss of protein kinase activity when native myosin filaments were used as the substrate. Cellular analysis confirms that the same severe loss of activity against myosin occurs in vivo when the WD repeat domain is eliminated. These results suggest that the WD repeat domain of MHCK A serves to target this enzyme to its physiological substrate.
Conventional myosin (myosin II) is involved in a wide range of contractile events in eukaryotic cells. In Dictyostelium, genetic and cellular analyses have demonstrated roles for myosin in the maintenance of cortical tension, cytokinesis, morphogenesis, capping of receptors, and cell locomotion (1-5). Although the roles of myosin II appear similar in many cell types, the in vivo mechanisms regulating myosin II assembly and activity in nonmuscle cells are not well understood.
Assembly of Dictyostelium myosin II bipolar filaments can be
regulated by phosphorylation on the myosin II heavy chain
(MHC).1 Two distinct MHC kinases (MHCKs)
have been purified and cloned from Dictyostelium. A 130-kDa
MHCK (MHCK A), discussed below, is expressed during both growth and
development. An 84-kDa MHCK that is expressed only during development
appears to contain two distinct catalytic domains, one related to
protein kinase C (6-8) and one related to diacylglycerol kinases (9).
Both MHCKs are capable of phosphorylating threonine residues on the
myosin tail and can drive myosin bipolar filament disassembly in
vitro. A third kinase, related to MHCK A, has recently been cloned
from Dictyostelium (10), but is has not yet been established
whether this protein regulates myosin assembly in vivo. The
in vitro target sites of MHCK A have been mapped to
threonines 1823, 1833, and 2029 in the tail region of the MHC (11, 12).
The physiological significance of these sites was demonstrated by
mutating the sites either to alanine (3X ALA myosin) to create a
nonphosphorylatable MHC or to aspartic acid (3X ASP myosin) to mimic
phosphorylated MHC (13). In vivo, cells expressing 3X ALA
myosin display severe myosin II overassembly in the cytoskeleton,
whereas cells expressing 3X ASP myosin display severely reduced myosin
II assembly in the cytoskeleton, resulting in a block in cytokinesis
and development. The 3X ASP cells are also unable to complete
development. In all tested assays, the 3X ASP cells are phenotypically
identical to MHC null cells (mhc). These
studies indicate that the MHCK A target sites play a critical role in
regulated myosin II assembly in vitro and that filament
assembly is required for myosin function in vivo.
Subsequent cellular analysis of Dictyostelium MHCK A null
cells (mhck A) and overexpressing cell lines
(MHCK A+) indicated that MHCK A regulates the cytoskeletal
myosin II assembly level during both growth and development (14). The
mhck A
cells are viable, but display partial
overassembly of myosin II in the cytoskeleton. MHCK A+
cells display reduced myosin II assembly in the cytoskeleton and
reduced efficiency of growth in suspension and fail to complete development. The MHCK A+ cell phenotype resembles those of
the mhc
cells and the 3X ASP cell lines,
indicating that overexpressed MHCK A can drive myosin filament
disassembly in vivo.
Molecular analysis of the MHCK A sequence (15) indicates that it has an
amino-terminal domain that is predicted to have an -helical
coiled-coil structure, a catalytic domain that is not related to
conventional protein kinases, and a carboxyl-terminal domain that
contains the 7-fold WD repeat motif (16, 17) characteristic of
-subunits of heterotrimeric G proteins. MHCK A is activated 50-fold
by autophosphorylation (18), and autophosphorylation is increased by
polyanions such as DNA, heparin, phosphatidylserine, and
phosphatidylinositol (19).
It has been demonstrated recently that the MHCK A catalytic domain is a prototype for a completely novel family of protein kinases unrelated to the conventional eukaryotic protein kinase superfamily (10, 15, 20-22). This novel group includes mammalian elongation factor-2 kinase, also known as calcium/calmodulin-dependent protein kinase III. Given its novel structure and catalytic domain, we have performed a structure-function dissection of MHCK A to identify the biochemical and physiological roles of the domains that flank the catalytic domain. The results reported here indicate that the amino-terminal coiled-coil domain is responsible for MHCK A oligomerization and that the WD repeats are important for full activity against native myosin II.
Cells were cultured axenically as described previously (14). All cell lines described here are derivatives of the parental line JH10 (23). For developmental studies, cells were harvested in log phase growth and plated on starvation plates containing 20 mM MES, 0.2 mM CaCl2, 2 mM MgSO4, and 1.5% agar, pH 6.8.
Plasmid Constructs and TransformationsDNA manipulations
were done according to standard procedures (24). The plasmid pLMHCK has
been previously described (14). This construct expresses residues
8-1146 of the mhck A cDNA from the extrachromosomal
vector pLittle, with constitutive expression driven by an actin-15
promoter. pLMHCK and other constructs were transfected into mhck
A as described previously (14). Initial selections
for these transfections were done using 10 µg/ml G418. Once
transfected cell lines were established, G418 was raised to 50 µg/ml
for cell line maintenance. Truncation constructs were generated as
modifications of pLMHCK, in the same vector. In each case, inserts were
fused to an amino-terminal 6-histidine tag derived from the vector
pDXA-HC (25). The
WD-MHCK construct also bears a carboxyl-terminal 6-histidine tag. Neither 6-histidine tag proved reliable for
purification from Dictyostelium. Nickel chelation
chromatography was therefore not used for the purification reported
here. The insert for pL
Coil-MHCK expresses residues 499-1146 of
mhck A and relies upon the native MHCK A stop codon for
translation termination. The insert for pL
WD-MHCK expresses residues
8-844 of mhck A and includes 19 residues of vector-derived
coding region at the carboxyl terminus.
MHCK A was
purified from Dictyostelium MHCK A+ cells as
described previously (26) with modifications described below. The location of MHCK A activity in column eluants was determined using a
peptide phosphorylation assay. Assays for MHCK A activity were performed at 23 °C in assay buffer containing 10 mM TES,
pH 7.0, 2 mM MgCl, 1 mM DTT, 500 µM [-32P]ATP (100-500 cpm/pmol), and 30 µM peptide MH-1. The previously described peptide MH-1
(RKKFGESEKTKTKEFL) (18), which corresponds to amino acids
2020-2035 in the Dictyostelium myosin II heavy chain, was
dissolved in 10 mM TES, pH 7.0. Protein purification column
elution fractions were assayed by the addition of 2 µl of each
fraction to 48 µl of assay buffer. The tube was vortexed and
incubated for 2 min at 25 °C. The reaction was terminated by
spotting a 25-µl aliquot onto phosphocellulose P-81 paper and washing
as described previously (18). Dictyostelium myosin II used
in this study was stored in 10 mM TES, pH 7.0, 2 mM EDTA, 0.5 mM DTT, and 20% sucrose.
Cells were harvested from HL5 suspension culture at late log phase
(typically 4-6 liters). Cells were resuspended in buffer (2 ml/g of
cells) containing 10 mM TES, 25 mM KCl, 30%
sucrose, 2 mM EDTA, 2 mM EGTA, 1 mM
DTT, 0.5 mg/ml leupeptin, 0.2 mg/ml phenylmethylsulfonyl fluoride, 0.05 mg/ml
N-p-tosyl-L-lysine
chloromethyl ketone, 0.1 mg/ml L-1-tosylamido-2-phenylethyl chloromethyl ketone, 1 mM benzamidine, and 1 mM
Pefabloc. The cell suspension was sonicated for 3 min on ice using a
Branson Sonifier 450 sonicator, and lysis was assessed by light
microscopy.
The extract was centrifuged for 30 min at 45,000 rpm in a Beckman Ti-45
rotor. The supernatant from the spin was decanted and mixed with packed
DE52 cellulose (2 ml of supernatant/ml of cellulose) and rotated at
4 °C. After 60 min, the DE52 cellulose was pelleted by
centrifugation, and the supernatant was decanted. The DE52 supernatant
was mixed with packed phosphocellulose (3 ml of supernatant/ml of
phosphocellulose) and rotated at 4 °C for 60 min. The
phosphocellulose was collected by centrifugation and washed two times
with buffer A (10 mM TES, pH 7.0, 50 mM KCl, 20% sucrose, and 1 mM DTT). The resin was poured into a
3 × 20-cm column and washed with buffer A until the absorbance
(A280) of the eluant reached base line. MHCK A
was eluted from phosphocellulose by applying buffer A adjusted to 300 mM KCl. The resulting activity pool was applied to a
hydroxylapatite column and washed until the A280
of the eluant reached base line. Kinase activity was eluted with 200 mM KPO4, 20% sucrose, and 1 mM
DTT, pH 7.0. This pool was loaded onto a 3 × 100-cm Sephacryl
S-300 gel filtration column equilibrated with 10 mM TES,
0.3 M KCl, 20% (w/v) sucrose, 1 mM EDTA, and 1 mM DTT, pH 7.0. WD-MHCK A was purified as described above except an aminohexyl-Sepharose column step was added after gel
filtration. The peak of activity off of gel filtration was diluted to
50 mM KCl; loaded onto a 1-ml aminohexyl-Sepharose column;
and eluted with 300 mM KCl, 1 mM DTT, and 10 mM TES, pH 7.0.
Coil-MHCK A was purified as described
above except that the peak of activity off of gel filtration was
diluted to 50 mM KCl and loaded onto a 5-ml Mono-S column.
Elution was performed with a 50-400 mM KCl gradient in 1 mM DTT and 10 mM TES, pH 7.0.
Coil-MHCK A
eluted at ~150 mM KCl. Protein concentrations were determined throughout the purification using the method of Bradford (32) with bovine serum albumin as the standard. Purified MHCK A,
Coil-MHCK A, and
WD-MHCK A were concentrated by binding to a
small volume of hydroxylapatite resin and eluting with 200 mM KPO4, 100 mM KCl, 20% sucrose,
and 1 mM DTT. Eluants were immediately divided into
100-µl aliquots and frozen in liquid nitrogen until use. Final
protein concentrations of purified protein kinases were determined by
performing densitometry on Coomassie Blue-stained SDS-polyacrylamide
gels using bovine serum albumin to generate a standard curve.
Total
cell extracts from each of the cell lines were prepared for Western
blot analysis as described previously (14). Quantitation of expression
of MHCK A+, Coil-MHCK A+, and
WD-MHCK
A+ in each cell line was performed by Western blot analysis
using affinity-purified anti-MHCK A polyclonal antibodies and
125I-labeled goat anti-rabbit secondary antibody. The
relative level of expression was determined by performing parallel
Western blotting of cell extracts and a series of known amounts of each
purified kinase construct to generate a standard linear curve for each cell line. All signals were then quantitated by a PhosphorImager. Calculation of the cellular concentrations of each protein was performed by making the estimate that a single Dictyostelium
cell is a sphere 5 µm in diameter. The concentration of myosin in
Dictyostelium was determined by densitometry on
SDS-polyacrylamide gels of cell extracts using bovine serum albumin to
generate a standard linear curve.
Expression of MHCK A Constructs
A schematic representation of full-length MHCK A and constructs in
which the amino-terminal coiled-coil (Coil-MHCK A) or the
carboxyl-terminal WD repeat (
WD-MHCK A) domains were truncated are
shown in Fig. 1. The three MHCK A constructs were
overexpressed in the Dictyostelium mhck A
cell
line by fusing each cDNA segment to an actin-15 promoter in an
extrachromosomal vector. Once established, transfected cell lines were
maintained in a 50 µg/ml concentration of the selective antibiotic
G418. This elevated G418 concentration resulted in significantly higher
kinase expression (3-4-fold) relative to the previously described
overexpression of full-length MHCK A (14). The expression of each
kinase was quantified by phosphoimaging of Western blots performed with
MHCK A polyclonal antibodies and 125I-labeled anti-rabbit
secondary antibody using each of the three purified proteins (described
below) to generate a standard curve for each corresponding cell line,
including wild-type cells. Calculations were performed using this
Western blot analysis to generate estimates of the in vivo
expression levels of each recombinant kinase construct. Based upon this
analysis, recombinant full-length MHCK A is expressed at 27 µM,
Coil-MHCK A is expressed at 46 µM,
and
WD-MHCK A is expressed at 90 µM. MHCK A expression
in the parental cell line JH10 was calculated to be 0.3 µM. These expression levels represent 90- and 153-fold
overexpression of recombinant full-length MHCK A and
Coil-MHCK A,
respectively, and 300-fold overexpression of
WD-MHCK A relative to
that of MHCK A in the parental cell line (Fig. 1). By comparison, we
calculate myosin II in Dictyostelium to be ~6
µM.
Cytokinesis
Previous studies have demonstrated that myosin II is essential for
cytokinesis as assayed by growth of cells in suspension cultures (1).
mhc cells are unable to form a contractile
ring, which results in the cells becoming large and multinucleated
without any increase in cell number (1, 27). Similar phenotypes have
been observed in 3X ASP cells and in MHCK A+ cells (13,
14). We compared growth rates of MHCK A truncation cell lines in
suspension cultures as an in vivo test of myosin function
(Fig. 2a). The mhck
A
cell line, which was previously shown to grow at a
similar rate compared with control cells (14), was used as a control
(filled circles). Cells that overexpress full-length MHCK A
(filled squares) and
Coil-MHCK A (open
squares) in the mhck A
cell line were
unable to grow in suspension. These MHCK A+ and
Coil-MHCK A+ cells increased in size, were
multinucleated, and eventually lysed under suspension culture
conditions. In contrast, these cells grew at relatively normal rates
when maintained as attached cells in plastic Petri dishes.
Interestingly, when the WD repeats of MHCK A were removed (
WD-MHCK
A+) (open circles), these cells grew at a
similar rate and final density as mhck A
cells, indicating that myosin function was not impaired during cytokinesis. The
WD-MHCK A+ cells also displayed normal
cell size during growth in suspension (data not shown).
Development
Previous studies have shown that mhc
cells, 3X ASP cells, and MHCK A+ cells are unable to
complete development (13, 14, 27). We therefore used development as an
in vivo test for myosin function in the MHCK A truncation
cell lines (Fig. 2b). MHCK A+ cells and
Coil-MHCK A+ cells arrested at mound stage and were
unable to complete development, whereas control (mhck
A
) and
WD-MHCK A+ cells completed
development and formed spores at normal rates.
Overexpression of MHCK A and Coil-MHCK A causes defects in myosin
filament-dependent processes as assayed in vivo
by cytokinesis and development, suggesting that these MHCK A constructs
drive myosin filament disassembly in vivo. In contrast, no
defects in cytokinesis or development were observed in the
WD-MHCK A
cell line. Interestingly,
WD-MHCK A had the highest expression
compared with the other MHCK A-expressing cell lines (Fig. 1), with
300-fold higher expression than wild-type cells. Lack of phenotypic
defects in
WD-MHCK A+ cells therefore is not due to
lower expression of the kinase. As reported previously (14), elevated
MHC phosphorylation in crude cell lysates was observed in MHCK
A+ cells. Similar behavior was observed in
Coil-MHCK
A+ cells, whereas no detectable increase in MHC
phosphorylation was observed in lysates of
WD-MHCK A+
cells (data not shown).
Elimination of Myosin-based Defects with 3X ALA Myosin
Further analysis was performed to confirm that the defects in
cytokinesis and development in MHCK A+ and Coil-MHCK
A+ cell lines were due to phosphorylation of myosin II and
disassembly of filaments. Cell lines were constructed in which either
wild-type myosin or 3X ALA myosin was expressed in an
mhc
background. The 3X ALA myosin bears
mutations in the in vitro mapped target sites of MHCK A in
the myosin II tail. This 3X ALA myosin is predicted to be resistant to
disassembly in vivo by overexpressed MHCK A if the in
vitro mapped target sites are also the physiologically relevant
in vivo target sites. Clonal populations of these cells
expressing either wild-type or 3X ALA MHC were transfected with
cDNA encoding full-length mhck A or
coil-mhck A to test whether the 3X ALA myosin mutation relieves the defects induced by MHCK A or
Coil-MHCK A overexpression. When cDNA
expressing MHCK A or
Coil-MHCK A was transfected into cells
expressing wild-type myosin, they failed to grow in suspension (Fig.
3a, filled and open
squares, respectively), as expected based upon results presented above. This defect in cytokinesis caused by overexpression of MHCK A or
Coil-MHCK A was eliminated in cells expressing 3X ALA myosin (Fig.
3a, filled and open triangles,
respectively). Overexpression of MHCK A or
Coil-MHCK A in wild-type
myosin cells caused the cells to become large and multinucleated in
suspension culture, whereas 3X ALA myosin cell lines expressing these
kinase constructs remained small and displayed no increase in size
(Fig. 3b) in suspension culture. In addition, wild-type MHC
cells that overexpress either full-length MHCK A or
Coil-MHCK A
constructs were not able to complete development and arrest at the
mound stage, whereas 3X ALA MHC cells that overexpress either kinase
construct were able to complete development (data not shown). These
results are consistent with the hypothesis that full-length MHCK A and
Coil-MHCK A hyperphosphorylate myosin on residues 1823, 1833, and
2029 in vivo (or a subset of these residues), resulting in
myosin disassembly and consequently an mhc
phenotype.
Purification of Full-length MHCK A, Coil-MHCK A, and
WD-MHCK
A
To determine the activity of each construct, the proteins were
purified from Dictyostelium cultures. The previously
reported purification method for MHCK A (26) was modified to allow
purification of each MHCK A construct using four to five column steps.
A summary of the purification of MHCK A constructs is found in Table
I, and details on the purification of each MHCK A
construct are given under "Experimental Procedures." The
purification table is presented as a guide to show relative
purification during the purification process. A detailed comparison of
the activities of the proteins under equivalent conditions is presented
below. Fig. 4 shows purified MHCK A, WD-MHCK A, and
Coil-MHCK A after SDS-polyacrylamide gel electrophoresis and
Coomassie Blue staining (panel a) or Western blot analysis
(panel b). MHCK A and
Coil-MHCK A ran at 130 and 73 kDa,
respectively.
WD-MHCK A ran as a doublet at 96 and 90 kDa. This
could represent autophosphorylation heterogeneity or a partial
proteolytic clip. Both forms were recognized by polyclonal antisera,
and both became radioactive in an autophosphorylation test performed
with [
-32P]ATP.
|
MHCK A Oligomerization and Rotary Shadowing
Several lines of evidence indicate that MHCK A is an oligomeric
protein kinase, including gel filtration analysis of native protein
(8), cross-linking studies (15), and coiled-coil predictive algorithms
(15). Gel filtration chromatography performed as the final or
penultimate step in each purification allowed us to test the hypothesis
that the amino-terminal domain of MHCK A (residues 1-500) is
responsible for oligomerization. The gel filtration analysis of
full-length MHCK A and truncation constructs is presented in Fig.
5. Full-length MHCK A migrated at 130 kDa on
SDS-polyacrylamide gel electrophoresis, but under native conditions, gel-filtered in the void volume with an estimated molecular mass >1000
kDa. This is consistent with the 130-kDa protein being oligomeric. Rod-shaped protein domains such as coiled-coil helices gel filter with
larger apparent mass that globular protein of equivalent size, so the
number of MHCK A monomers in each oligomer cannot be estimated
accurately from this elution behavior.
Deletion of the carboxyl-terminal WD repeats (WD-MHCK A) resulted in
gel filtration at a slightly smaller apparent mass than native MHCK A,
but still >670 kDa, which is much larger than its predicted molecular
size of 96 kDa. Elimination of the amino-terminal 500 amino acids
caused
Coil-MHCK A to gel filter with an apparent molecular mass of
80 kDa, which is consistent with the predicted mass of 73 kDa. These
results indicate that the amino-terminal 500 residues of MHCK A, which
are predicted to form a coiled-coil structure, are responsible for MHCK
A oligomerization. The computer algorithm SCORER, created by Woolfson
and Alber (28), predicts whether a sequence is expected to form a two-
or three-stranded coiled-coil. According to this algorithm, MHCK A is
predicted to form a three-stranded coiled-coil (data not shown).
Full-length MHCK A was visualized by platinum rotary shadowing (Fig.
6). The majority of the rotary-shadowed images appeared globular or aggregated, as in the lower right panel of Fig.
6. However, a number of images were also observed that appeared to reveal a globular domain associated with an extended rod-like structure
(all other panels in Fig. 6). Measurements of these images indicated an
average length for the rod-like segments of ~50 nm. The average
calculated pitch of coiled-coil helices from proteins such as myosin
tails is 6.7 residues/nm. If MHCK A residues 100-500 were contained
entirely in a coiled-coil structure (15), a rod domain of ~60 nm
would be predicted. This value is agreement with the ~50-nm length of
the rod structures observed by rotary shadowing. The apparent globular
segment of the images was always observed at only one end of the rod
segment, suggesting that MHCK A complexes form via parallel
oligomerization rather antiparallel oligomerization.
Biochemical Characterization of MHCK A Constructs
AutophosphorylationPrevious studies showed that MHCK A is
activated by autophosphorylation, incorporating up to 10 mol of
phosphate/mol of enzyme. Incorporation of the first 3 mol of phosphate
was shown to be sufficient for maximum activation (18).
Autophosphorylation behavior was assessed for MHCK A truncation
constructs. Full-length MHCK A, Coil-MHCK A, and
WD-MHCK A all
autophosphorylated to a similar range of 8-10 mol of phosphate/mol of
enzyme (Fig. 7a). The autophosphorylation
rates of each construct were independent of concentration, suggesting
that autophosphorylation is an intramolecular process (data not
shown).
Activity of MHCK A Constructs toward Peptide
MHCK A has been
shown to phosphorylate the myosin tail on threonines 1823, 1833, and
2029. We used a 16-residue peptide (MH-1) corresponding to the sequence
around Thr-2029 in the myosin tail to test the enzymatic activity of
each construct. Fig. 7b shows phosphorylation of the peptide
by purified autophosphorylated MHCK A, Coil-MHCK A, and
WD-MHCK
A. The phosphorylation of the peptide obeys Michaelis-Menten kinetics;
full-length recombinant MHCK A (circles) had a
Km of 92 µM with a
Vmax of 3.6 µmol × min
1 × mg
1 for the peptide, in close agreement with previous
values reported for native MHCK A by Medley et al. (18)
(Km = 105 µM and
Vmax = 2.2 µmol × min
1 × mg
1). As indicated by the kinetic parameters (Table
II), both truncation constructs displayed relatively
similar activity against peptide MH-1 as compared with full-length MHCK
A.
|
Autophosphorylated MHCK A, Coil-MHCK A, and
WD-MHCK A
were assayed for their ability to phosphorylate wild-type
Dictyostelium myosin II (Fig. 7c).
Coil-MHCK A
was able to phosphorylate myosin II at a rate similar to that of
full-length MHCK A. In contrast,
WD-MHCK A showed a substantial
decrease in activity against myosin as compared with MHCK A or
Coil-MHCK. While it is difficult to obtain accurate kinetic
constants for myosin II because it is filamentous under the assay
conditions, the specific activities of
WD-MHCK A measured with
several concentrations of myosin were consistently 20-fold lower than
those of MHCK A and
Coil-MHCK A. These results are consistent with
the in vivo results and suggest that the WD repeat domain of
MHCK A is required for efficient substrate recognition of myosin II
both in vitro and in vivo.
In vitro and in vivo analyses of MHCK A truncation constructs have been used to identify the roles of the two domains that flank the catalytic domain of MHCK A. The earlier demonstration that full-length MHCK A overexpression leads to disassembly of cytoskeletal myosin II in vivo (14) provided a useful assay for in vivo activity of the expressed constructs.
Gel filtration analysis of the purified constructs demonstrated a loss
of oligomerization in the Coil-MHCK A construct. An extended
rod-like domain was also observed in the rotary-shadowed images of
native MHCK A. These results provide experimental evidence that the
segment of MHCK A spanning residues 1-500 contains the oligomerization
functions and that the oligomer formation probably occurs via formation
of a parallel coiled-coil rod structure.
Elimination of this domain did not appear to significantly impair
in vitro kinase activity against native myosin II or peptide MH-1. Although Coil-MHCK A displayed a slightly elevated
Km for peptide MH-1, its Vmax
and Kcat values were similar to those of
full-length MHCK A. The autophosphorylation stoichiometry in this
construct also appeared similar to that in full-length MHCK A. Furthermore,
Coil-MHCK A appeared active in vivo, as
assessed by its ability to impede myosin-based contractile events
in vivo in the presence of wild-type myosin II. It can be
speculated that the coiled-coil region of MHCK A could have other
functions, such as annealing with bipolar myosin filaments or serving
as a pseudosubstrate or autoinhibitory domain. Preliminary results in
our laboratory have not found evidence to support either of these
theories, however.
Expression of either Coil-MHCK A or full-length MHCK A completely
eliminated myosin contractile function in vivo in cells containing wild-type MHC. However, when these kinase constructs were
overexpressed in cells containing 3X ALA MHC, no inhibition of
myosin-based contractile function was observed. The relatively normal
phenotype of these cells provides the first demonstration that the
in vitro mapped target sites of MHCK A on the myosin tail
are also the physiologically relevant in vivo target sites of MHCK A. The properties of these cell lines furthermore confirm the
hypothesis that defects observed when MHCK A is overexpressed in
wild-type MHC cells are due solely to the activity of the kinase against myosin II, as opposed to other in vivo substrates.
All the phenotypic defects observed when MHCK A is overexpressed can be
attributed to its activity toward residues 1823, 1833, and 2029 in the
myosin II tail.
While overexpressing full-length MHCK A and Coil-MHCK A resulted in
striking phenotypes, the elimination of amino acids 842-1146, corresponding to the WD repeat domain, resulted in no increase in MHC
phosphorylation in Triton extracts and no impairment of myosin II
function in vivo during development or growth in suspension. This phenotype is interesting since
WD-MHCK A is expressed at a
3-fold higher level than full-length MHCK A in MHCK A+
cells. Purified
WD-MHCK A autophosphorylated to a similar extent as
MHCK A and
Coil-MHCK A. More important,
WD-MHCK A also displayed similar kinetics as MHCK A and
Coil-MHCK A toward the MH-1 peptide substrate, which corresponds to the target site at residue 2029 on the
myosin II tail.
Full catalytic activity against peptide MH-1 by WD-MHCK A indicates
that the peptide substrate-binding regions of MHCK A and all core
catalytic functions are intact in this construct. Despite this, a 95%
drop in activity against native myosin II was observed in
vitro, and the overexpressed kinase did not interfere with myosin
II function in vivo despite its high expression level. These
results support the hypothesis that the WD repeat domain of MHCK A
plays a critical role in substrate recognition with the native myosin
II substrate. Attempts to measure the Km of each
kinase construct for native myosin II are complicated by the
filamentous nature of the substrate under assay conditions. Other
methods are currently being explored to determine directly whether the
WD repeats of MHCK A act as a myosin II-binding domain.
-Subunits of heterotrimeric GTP-binding proteins serve as the
prototype for members of the WD repeat family of proteins (16, 17).
This prototype class consists of an amino-terminal segment of ~50
residues, which is followed by seven repeats of ~40 residues, with
each repeat containing a conserved "WD" motif. Although the number
of WD repeats varies in different members of the WD repeat family of
proteins, most members probably fold into a structure similar to the
seven-blade "
-propeller" structure identified for the human G
protein
-subunit via x-ray crystallography (29, 30).
Although -subunits of GTP-binding proteins have an established role
in binding to and activating effector targets, the roles of WD repeat
domains in other classes of WD repeat-containing proteins are not well
characterized. To our knowledge, the work presented here is the first
biochemical evidence for a WD repeat domain serving to target an
attached catalytic domain to its substrate.
Neer et al. (16, 17) have suggested that the conserved residues of the WD repeats provide a rigid scaffold and that the variable residues between may specify the interactions with different proteins. A more detailed analysis of the MHCK A WD repeat domain will be necessary to determine which residues dictate the specificity toward myosin II filaments.
We thank Dr. Graham Côté for advice and helpful discussions and Dr. Roger Craig for rotary shadowing MHCK A.