(Received for publication, September 18, 1996)
From the Departments of Biochemistry and
§ Dermatology, Asahikawa Medical College, Asahikawa
078, Japan
In an attempt to establish the stoichiometry of phosphorylation in the Ca2+-ATPase of sarcoplasmic reticulum (SR) vesicles, phosphorylation by ATP (or Pi) or labeling by fluorescein 5-isothiocyanate (FITC) was performed with the SR vesicles under the conditions in which almost all the phosphorylation sites or FITC binding sites are phosphorylated or labeled. The resulting vesicles were solubilized in lithium dodecyl sulfate and then the Ca2+-ATPase was purified by size exclusion high performance liquid chromatography. Peptide mapping and sequencing of the tryptic digest of the purified enzyme showed that Lys-515 of the Ca2+-ATPase was exclusively labeled with FITC, in agreement with the previously reported findings. The content of the phosphoenzyme from ATP (4.57 nmol/mg of Ca2+-ATPase protein) or from Pi (4.94 nmol/mg of Ca2+-ATPase protein) in the purified enzyme was approximately half the content of the FITC binding site (8.17-8.25 nmol/mg of Ca2+-ATPase protein) and also half the content of the Ca2+-ATPase molecule (9.06 nmol/mg of Ca2+-ATPase protein) calculated from its molecular mass (110,331 Da). These results show that there is one specific FITC binding site per molecule of the Ca2+-ATPase (in agreement with the previously reported findings) and that the stoichiometry of phosphorylation to FITC binding is approximately 0.5:1.0. All the above findings lead to the conclusion that only half of the Ca2+-ATPase molecules present in the SR vesicles can be phosphorylated. FITC binding completely inhibited the ATP-induced phosphorylation before the binding reached its maximum level. This finding indicates that FITC preferentially binds to a part of the Ca2+-ATPase molecules and that this binding is primarily responsible for the inhibition of phosphorylation, suggesting an intermolecular ATPase-ATPase interaction.
The SR1 Ca2+-ATPase
consists of a single 110-kDa polypeptide chain, of which the whole
amino acid sequence has been revealed (1). In the early step of the
catalytic cycle, the -phosphoryl group of Mg·ATP bound to the
Ca2+-activated enzyme is transferred to Asp-351 (2, 3) to
form an acid-stable EP intermediate (4, 5). EP
can also be formed from Pi by reversal of the late step of
the catalytic cycle (6, 7).
We have recently shown that, when SR vesicles are treated with fluoride
in the presence of Mg2+, two Mg2+ and four
F per phosphorylation site bind tightly to the catalytic
site of the Ca2+-ATPase (8). This finding and the related
kinetic data obtained previously (9) have suggested that only half of
the Ca2+-ATPase molecules present in the SR vesicles are
catalytically active (i.e. the active molecules can be
phosphorylated with ATP or Pi), the other half being
catalytically inactive (i.e. the inactive molecules cannot
be phosphorylated with ATP or Pi), and that one
Mg2+ and two F
bind equally to each of the
catalytically active and inactive molecules. In the subsequent study
(10), we have determined the total content of the sulfhydryl group of
the Ca2+-ATPase in the SR vesicles after exhaustive
reduction with dithiothreitol and found that the content of the
phosphorylation site in the SR vesicles is again half that of the
Ca2+-ATPase molecule calculated from this sulfhydryl
content.
It is now important to establish the stoichiometry of phosphorylation in the Ca2+-ATPase of SR vesicles, because the information on this stoichiometry is essential for the construction of the mechanistic model of this enzyme. A feasible approach to this problem is to determine the binding of FITC to this enzyme, because specific binding of FITC to Lys-515 at or near the ATP binding site of this enzyme is well established (11, 12).
In this study, phosphorylation by ATP (or Pi) or labeling by FITC has been performed with the SR vesicles under the conditions in which almost all the phosphorylation sites or FITC binding sites are phosphorylated or labeled. The results show that there is one specific FITC binding site per molecule of the Ca2+-ATPase (in agreement with the previously reported findings (12)) and that the stoichiometry of phosphorylation to FITC binding is approximately 0.5:1.0. These findings lead to the conclusion that only half of the Ca2+-ATPase molecules present in the SR vesicles can be phosphorylated.
SR vesicles were prepared from
rabbit skeletal muscle as described previously (8) and stored in 0.3 M sucrose, 0.1 mM CaCl2, 0.1 M KCl, and 5 mM MOPS/Tris (pH 7.0) at
80 °C.
FITC was freshly dissolved in N,N-dimethylformamide. The SR vesicles (0.5 mg of protein/ml) were labeled with 15 µM FITC at 25 °C in the dark in 2 mM EGTA, 5 mM MgCl2, 0.1 M KCl, 1% (v/v) N,N-dimethylformamide, and 50 mM Tris/HCl (pH 8.0), unless otherwise stated. The FITC-labeled vesicles were isolated by either of the following methods. Method A: the labeling was stopped by adding 4.5 volumes of an ice-cold solution containing 12.2 mM ATP, 2 mM EGTA, 5 mM MgCl2, 0.1 M KCl, and 122 mM MOPS/Tris (pH 7.0). The resulting suspension was centrifuged at 411,000 × g and 0 °C for 5 min. Method B: the labeling was stopped at 4 °C by adding 200 mM ATP and 200 mM MES/Tris (pH 6.0) to give 10 and 36 mM, respectively. The resulting suspension was applied at 4 °C to a coarse Sephadex G-50 column preequilibrated with 0.5 mM EGTA, 0.1 M KCl, 0.3 M sucrose, and 2.5 mM MOPS/Tris (pH 7.0). The FITC-labeled vesicles in the effluent were collected by centrifugation as in Method A.
Size Exclusion HPLC of SR Vesicles Labeled with FITC and Solubilized in LDSThe FITC-labeled SR vesicles (5 mg of protein/ml) were solubilized at 4 °C in a buffer containing 2% (w/v) LDS, 0.1 mM CaCl2, 5 mM MgCl2, 100 mM Li2SO4, and 20 mM sodium phosphate (pH 7.0). After centrifugation at 0 °C, 0.3-0.4 ml of the supernatant was subjected to size exclusion HPLC at room temperature by the use of a TSK SWXL guard column (0.6 × 4 cm, Tosoh, Japan), a TSKgel G3000SWXL column (0.78 × 30 cm, Tosoh, Japan), and a TSKgel G3000SW column (0.75 × 30 cm, Tosoh, Japan) that were connected in series. The elution was performed at a flow rate of 0.5 ml/min (unless otherwise stated) with the above solubilization buffer. The absorbance (280 nm) and fluorescence (excitation at 485 nm and emission at 525 nm) were monitored with a detector UV-M (Pharmacia Biotech Inc.) and a fluorescence detector F-1000 (Hitachi, Japan), respectively. The fractions of 0.25 ml were collected.
Determination of Bound FITCThe contents of FITC in the above fractions were determined from the absorbance at 496 nm in 8% (w/v) SDS and 200 mM Tris/HCl (pH 8.8) by using an extinction coefficient of 80,000, which was widely used in other laboratories (12-16).
Tryptic Digestion of the FITC-labeled Ca2+-ATPase and Reversed Phase HPLCThe fraction at the peak of protein concentration in the size exclusion HPLC was applied twice to centrifuge columns (5-ml disposable syringe filled with coarse Sephadex G-50) preequilibrated with 1 mM CaCl2 and 10 mM sodium phosphate (pH 7.0) at 4 °C according to Penefsky (17) in order to remove excess dodecyl sulfate. The FITC-labeled Ca2+-ATPase thus obtained was digested with trypsin (1.4:1 of trypsin to ATPase by weight) at 37 °C for 10 min in 1 mM CaCl2 and 10 mM sodium phosphate (pH 7.0). Reaction was stopped by addition of a 2-fold excess of trypsin inhibitor. The supernatant obtained by centrifugation was subjected to reversed phase HPLC, which was performed at a flow rate of 1 ml/min by using a C2/C18 SuperPac Pep-S column (5 µm, 0.4 × 25 cm, Pharmacia Biotech Inc.). The absorbance was monitored at 214 nm with a detector SPD-10AV (Shimadzu, Japan) and the fluorescence monitored in the same way as in the size exclusion HPLC.
Sequencing and Amino Acid Analysis of Isolated PeptidesSequencing of isolated peptides was performed with a protein sequenator 477A (PE Applied Biosystems) connected with a phenylthiohydantoin-derivative analyzer 120A (PE Applied Biosystems). Amino acid analysis was performed by the use of an amino acid analyzer 835-10 (Hitachi, Japan) after hydrolysis of the peptides in 6 N HCl at 110 °C for 24 h.
Phosphorylation of SR Vesicles with ATP or PiPhosphorylation was performed under the conditions
of Barrabin et al. (18), in which almost all the
phosphorylation sites of the Ca2+-ATPase can be
phosphorylated with ATP or Pi. When EP was
formed from ATP, the SR vesicles (1 mg of protein/ml) were
phosphorylated with 0.1 mM [-32P]ATP at
25 °C for 5 s in 10 mM CaCl2, 5 mM MgCl2, 80 mM KCl, and 30 mM Tris/HCl (pH 7.5). The reaction was quenched with 9 volumes of ice-cold 5.6% (w/v) trichloroacetic acid containing 0.56 mM ATP and 5.6 mM Pi. The resulting
mixture was centrifuged at 0 °C, and the pellet was washed at
4 °C once with distilled water (unless otherwise stated). When
EP was formed from Pi, the SR vesicles (1 mg of
protein/ml) were phosphorylated with 2 mM 32Pi at 25 °C for 10 min in 1 mM
EGTA, 20 mM MgCl2, 40% (v/v)
Me2SO, and 30 mM MES/Tris (pH 6.0). The
reaction was quenched with 9 volumes of ice-cold 5.6% (w/v)
trichloroacetic acid containing 5.6 mM PPi and
111 mM Pi. The pellet was washed at 4 °C
twice (unless otherwise stated) with 4% (v/v) perchloric acid
containing 20 mM Pi and 5 mM
PPi and then once with distilled water.
The phosphorylated SR vesicles (5 mg of protein/ml) were solubilized at 4 °C in a buffer containing 2% (w/v) LDS, 100 mM Li2SO4, and 50 mM lithium acetate (pH 4.5). After centrifugation at 0 °C, 0.3-0.4 ml of the supernatant was subjected to the size exclusion HPLC described above at room temperature. The elution was performed at a flow rate of 0.5 ml/min with the above acidic solubilization buffer (pH 4.5), and the absorbance was monitored at 280 nm. The fractions of 0.25 ml were collected. The radioactivity of each fraction was measured with Cerenkov radiation using a liquid scintillation counter. The content of EP was obtained from the radioactivity and protein concentration.
Miscellaneous MethodsFITC, trypsin-treated with
L-1-tosylamide-2-phenylethyl chloromethyl ketone, and
soybean trypsin inhibitor (type I-S) were purchased from Sigma.
[-32P]ATP was prepared by the method of Post and Sen
(19). 32Pi was purified according to Kanazawa
and Boyer (6). Protein concentrations were determined by the method of
Lowry et al. (20) with bovine serum albumin as a
standard.
Fig. 1 shows the size exclusion
HPLC of the SR vesicles which were labeled with FITC for 200 min (FITC
labeling reached almost the highest level; see the inset of
Fig. 5) and solubilized in LDS. The major peak of the absorbance at the
retention time of 12 min (Fig. 1A) was highly enriched with
110-kDa chains derived from the Ca2+-ATPase, as revealed by
SDS-polyacrylamide gel electrophoresis (not shown). Several minor peaks
of lower molecular mass proteins eluted after the major peak may be
mostly assigned to the well documented Ca2+-binding
proteins (21). The content of FITC per mg of protein was constant
throughout six successive fractions in the major peak of proteins (Fig.
1B). A small amount of residual unreacted FITC was eluted at
the retention time of 26 min.
The fraction containing 8.87 nmol of FITC/mg of protein at the peak of protein concentration was applied to the second size exclusion HPLC with the same column system as in the first size exclusion HPLC. The elution profile at 280 nm showed a single symmetric peak (not shown). The content of FITC was constant throughout seven successive fractions in this peak, being 8.43-8.96 nmol of FITC/mg of protein. This agreement in the FITC content between the first and second size exclusion HPLC indicates that the Ca2+-ATPase protein eluted in the major peak in the first size exclusion HPLC is sufficiently pure for the determination of FITC bound to the Ca2+-ATPase.
Reversed Phase HPLC of the Tryptic Digest of FITC-labeled Ca2+-ATPaseThe FITC-labeled Ca2+-ATPase
purified by the first size exclusion HPLC was digested with trypsin.
After centrifugation of the digest, 90% of bound FITC was recovered in
the supernatant. This supernatant was applied to reversed phase HPLC
(Fig. 2). The fluorescence profile showed one minor peak
(Peak 1) and the other major peak (Peak 2).
Sequence Analysis
Peak 1 and Peak 2 in the reversed phase HPLC were further purified repeatedly by reversed phase HPLC. The fluorescence profiles in these repeated HPLC showed that fluorescent components of Peak 1 and Peak 2 were both homogeneous (not shown). Sequencing of purified Peak 1 gave no phenylthiohydantoin-derivatives. Amino acid analysis of this peak gave no amino acids, being consistent with the results of sequencing. Sequencing of purified Peak 2 gave a single sequence from Met-512 to Arg-524 of the Ca2+-ATPase (1), but the Lys-515 residue was missing. This indicates that Lys-515 was labeled with FITC. Amino acid analysis of this peak showed the composition expected from the above sequence.
Size Exclusion HPLC of SR Vesicles Phosphorylated and Solubilized in LDSThe SR vesicles phosphorylated with
[-32P]ATP and solubilized in LDS were subjected to
size exclusion HPLC (Fig. 3), in which an acidic
solubilization-elution buffer (pH 4.5) was used to stabilize EP and to facilitate its quantitative recovery. The major
peak of the absorbance at the retention time of 24 min was highly
enriched with 110-kDa chains derived from the Ca2+-ATPase,
as revealed by SDS-polyacrylamide gel electrophoresis (not shown). The
content of EP per mg of protein was constant throughout five
successive fractions in this peak. Some radioactivities eluted at the
retention time of 50 min are most likely attributed to residual
[
-32P]ATP and 32Pi.
In a separate experiment, after quenching of phosphorylation the sample
was washed four times with 4% (v/v) perchloric acid containing 20 mM Pi and 5 mM PPi in order
to remove residual [-32P]ATP and
32Pi. Then, the sample was solubilized and
subjected to size exclusion HPLC in the same way as above. The elution
profiles of proteins and radioactivities were almost the same as those
in Fig. 3, but neither [
-32P]ATP nor
32Pi was detected (not shown). The
EP contents per mg of protein in five successive fractions
in the major peak of proteins were almost the same as those given in
Fig. 3. These results show that EP did not decompose to an
appreciable extent during the purification process.
When the SR vesicles phosphorylated with 32Pi were solubilized and subjected to size exclusion HPLC, the elution profiles of proteins and radioactivities were very similar to those in Fig. 3 (not shown). The EP content per mg of protein was again constant throughout five successive fractions in the major peak of proteins. A small amount of 32Pi was eluted at the retention time of 50 min, but it disappeared when the sample was washed twice more with the perchloric acid after quenching of phosphorylation. The EP content per mg of protein in the major peak of proteins was unaffected by these repeated washings. These results again show that EP did not decompose during the purification process.
Comparison between the Content of Phosphorylation Site and the Maximum Content of Bound FITCThe results presented so far
indicate that the size exclusion HPLC provides a reliable means for
purification of the FITC-labeled Ca2+-ATPase and
EP. Accordingly, FITC binding and phosphorylation were
determined repeatedly by this purification method. FITC labeling was
performed for 240 min in which it reached the highest level (see the
inset of Fig. 5). Phosphorylation was performed with [-32P]ATP or 32Pi under the
conditions in which almost all the phosphorylation sites are
phosphorylated (18). As summarized in Table I, the content of EP was approximately half the content of bound
FITC.
|
In order to assess
how much binding of FITC is specific to the ATP binding site,
inhibition of FITC binding by Mg·ATP was investigated. The SR
vesicles were labeled with FITC for 20 min in the presence of different
concentrations of Mg·ATP, and the labeled Ca2+-ATPase was
purified. The content of bound FITC in the purified enzyme decreased
with increasing concentration of Mg·ATP and reached a constant low
level at 30-60 mM Mg·ATP (Fig. 4). When
the SR vesicles were labeled with FITC in the presence of 60 mM Mg·ATP, Peak 2 in reversed phase HPLC of the tryptic
digest (cf. Fig. 2A) disappeared almost entirely
(not shown). This indicates that labeling of Lys-515 with FITC was
almost completely blocked by 60 mM Mg·ATP. On the other
hand, the area of Peak 1 was unaffected by this Mg·ATP and also
unaffected by extending the labeling period from 20 to 240 min in the
absence and presence of 60 mM Mg·ATP. By the repeated
determinations of FITC binding in the presence of 60 mM
Mg·ATP, the amount of this Mg·ATP-insensitive part of bound FITC
was estimated to be 0.49-0.57 nmol/mg of protein.
Inhibition of ATP-induced Phosphorylation by FITC Binding
In
order to examine the inhibitory effect of FITC binding on the
ATP-induced phosphorylation, the SR vesicles were labeled with FITC for
various periods and then phosphorylated with [-32P]ATP
under the conditions in which all the active phosphorylation sites
remaining are phosphorylated. The Ca2+-ATPase of these
vesicles was purified, and the contents of bound FITC and EP
in this purified enzyme were determined (Fig. 5). FITC
binding occurred rapidly in the initial 2-min period, gradually proceeded during the further incubation and then reached the highest level at 240 min (Fig. 5, inset). On the other hand, the
capacity for phosphorylation decreased rapidly after the start of FITC labeling and disappeared entirely within 4 min. The relation between the content of bound FITC and that of EP formed from ATP was
not linear, showing a concave curve (Fig. 5, main figure). FITC binding equivalent to about 7 nmol/mg of protein was required for complete inhibition of the ATP-induced phosphorylation. This value was appreciably higher than the total content of the phosphorylation site,
but it was definitely lower than the maximum content of bound FITC
(Table I).
Peptide mapping (Fig. 2A) and sequencing (see "Results") of the tryptic digest of the purified FITC-labeled Ca2+-ATPase show that Lys-515 of the enzyme is exclusively labeled with FITC. This is in agreement with the findings reported by Mitchinson et al. (12). The content of bound FITC given in Table I includes a small Mg·ATP-insensitive part of bound FITC, which is derived from Peak 1 containing no peptides. The origin of this part is obscure. By subtracting the amount of this Mg·ATP-insensitive part (0.49-0.57 nmol/mg of protein, see "Results") from the total content of bound FITC (8.74 nmol/mg of protein, Table I), the maximum content of FITC bound to Lys-515 is found to be 8.17-8.25 nmol/mg of protein. This value is in fair agreement with the theoretical value (9.06 nmol/mg of Ca2+-ATPase protein) calculated from the molecular mass (110,331 Da) of this ATPase chain (1). These findings indicate that there is one specific FITC binding site per molecule of the Ca2+-ATPase in the SR vesicles, being consistent with the previously reported results (12). Since the content of the phosphorylation site is 4.57-4.94 nmol/mg of protein (Table I), the stoichiometry of phosphorylation to FITC binding in the Ca2+-ATPase of SR vesicles is approximately 0.5:1.0. These findings lead to the conclusion that only half of the Ca2+-ATPase molecules present in the SR vesicles can be phosphorylated. This conclusion is consistent with our recent findings (8, 10) (see Introduction).
The 0.5:1.0 stoichiometry of phosphorylation to FITC binding also appears to be in harmony with the data obtained previously with SR vesicles by Mitchinson et al. (12), although they did not actually determine the content of the phosphorylation site in the SR vesicles. In those experiments Mitchinson et al. used the SR vesicles ("R1-washed") which were prepared by the method of MacLennan et al. (22) and obtained the maximum incorporation of 4.5 nmol of FITC/mg of vesicle protein. The EP level in the SR vesicles used in their experiments is expected to be as low as 1.82 nmol/mg of vesicle protein as shown by MacLennan et al. (22) (see Table II in Ref. 22).2 Therefore, it is likely that in their experiments the content of the phosphorylation site has been approximately half the maximum level of FITC binding.
It is improbable that the observed lack of phosphorylation in half of
the Ca2+-ATPase molecules is due to a possible
denaturation, because FITC shows a highly specific reactivity toward
Lys-515 in almost all the Ca2+-ATPase molecules and because
FITC labeling of Lys-515 in almost all the Ca2+-ATPase
molecules is blocked by Mg·ATP (Fig. 4 and see "Results"). This
conclusion is supported by the earlier findings to show that an
inactive portion of the Ca2+-ATPase in preparations of SR
vesicles does not bind FITC (15) and that the high specific reactivity
of Lys-515 toward FITC requires special steric conditions only given in
the native conformation of the active site (23, 24). Intactness of all
the Ca2+-ATPase molecules present in the SR vesicles used
is further supported by our previous observations (8) that all the
Ca2+-ATPase molecules in the SR vesicles are equally
available for tight binding of Mg2+ and F and
that this tight binding is entirely prevented by Ca2+
binding to the high-affinity Ca2+ sites. The lack of
phosphorylation in half of the Ca2+-ATPase molecules may be
possibly due to an intermolecular ATPase-ATPase interaction in the SR
vesicles.
FITC binding completely inhibits phosphorylation before the binding reaches its maximum level (Fig. 5). This finding indicates that FITC preferentially binds to a part (possibly half) of the Ca2+-ATPase molecules in the SR vesicles and that this binding is primarily responsible for the inhibition of phosphorylation. The above preferential binding is again likely due to an intermolecular ATPase-ATPase interaction. It seems possible that the intermolecular ATPase-ATPase interaction is partially disrupted by detergents, because it was reported previously (12, 14, 15) that 0.89-0.97 mol of bound FITC/mol of Ca2+-ATPase molecule is required for complete inhibition of the detergent-purified Ca2+-ATPase.
Recently, Liu et al. (25) have presented the data suggesting
that half of the -subunits of membrane-bound
Na+,K+-ATPase are dormant (i.e. the
half cannot be phosphorylated). It is possible that the half-of-the
sites reactivity in the membrane-bound Na+,K+-ATPase corresponds to that indicated
with the Ca2+-ATPase of SR vesicles in the present
study.
S. N. is grateful to Prof. Hajime Iizuka (Department of Dermatology, Asahikawa Medical College) for his continued encouragement during this work.