ANALYTICALBIOCHEMISTRY
192, 112-116
(1991)
A Malachite Green Calorimetric Assay for Protein Phosphatase Activity Taxiarchis P. Geladopoulos, Theodore G. Sotiroudis,l and Athanasios Institute of Biological Research,The National Hellenic ResearchFoundation, 48 Vassi.!eos Constantinou Avenue, Athens 116 35, Greece
Received
May
E. Evangelopoulos
31,199O
A simple and sensitive calorimetric assay for protein phosphatase activity based on the determination of released P, by an improved malachite green procedure (A. A. Baykov, 0. A. Evtushenko, and S. M. Avaeva, 1988, Anal. Biochem. 171,266-270) is described. Proteins must be removed or stabilized prior to Pi determination with 0.25 N sulfuric acid or 3% (w/w) perchloric acid. Alternatively, to avoid possible acid hydrolysis of phosphate groups from organic compounds during deproteinization, the protein present in the phosphatase assay mixture can be stabilized with sodium dodecyl sulfate. In this case, the excess detergent is subsequently removed by precipitation with KC1 because it colors with the malachite green reagent. The above procedure was applied to the determination of phosphorylase phosphatase activity in bovine brain extracts and the results are comparable to those obtained with the radioisotopic phosphatase assay. 0 1991 Academic Press, Inc.
by the reduction of phosphomolybdate, although simple, lacks sensitivity (3,4). Highly sensitive assays for Pi have relied upon formation of a colored complex between the acidified molybdate and the dye malachite green (5). The sensitivity of this procedure is 30 times greater than the widely used Fiske-SubbaRow method (4,5). Recently a modified malachite green procedure has been reported by Baykov et al. (6). In this method all necessary reagents are combined in one concentrated solution making the assay more sensitive and convenient (6). In the experiments presented in this paper the malachite green method for Pi determination of Baykov et al. (6) was adapted to measure protein phosphatase activity after precipitation or stabilization of the protein present in the assay mixture. MATERIALS
AND
METHODS
Chemicals The reversible phosphorylation of proteins catalyzed by protein kinases and protein phosphatases is now recognized to be a major process for regulating cellular functions (1,2). A number of methods are available for measuring Pi release in protein phosphatase assays (3). The most sensitive technique involves acid precipitation of a protein substrate containing bound [32P]phosphate and determination of the released radioactivity in the supernatant. This procedure is handicapped by restrictions for working with radioactivity and it cannot distinguish between the release of acid-soluble 32P-labeled peptides through proteolysis and the release of free [3”P]Pi by the phosphatase. On the other hand in nonradioisotopic phosphatase assays, the measurement of released Pi, using the formation of molybdenum blue
’ To whom
correspondence
should
be addressed.
Malachite green (C.I. 42,000), ammonium heptamolybdate, perchloric acid 70%, and sulfuric acid suprapur, were obtained from Merck. Tween 20 and SDS’ were products of Serva and Sigma respectively. [T-~‘P]ATP (3 Ci/mmol) was purchased from Radiochemical Centre, Amersham. Protein Preparations Bovine albumin (pure), protamine sulfate, and casein (vitamin free) were obtained from Serva. Phosvitin, histone type II-AS from calf thymus, trypsin twice crystallized, and alkaline phosphatase from bovine intestinal mucosa (type VII-G) were products of Sigma. Bovine albumin, phosvitin, and human serum were purified by gel filtration on a Sephadex G-25 column. Crystalline rabbit skeletal muscle phosphorylase b * Abbreviation
used:
SDS,
sodium
dodecyl
sulfate.
112 All
Copyright 0 1991 rights of reproduction
0003-2697/91 $3.00 by Academic Press, Inc. in any form reserved.
COLORIMETRIC
ASSAY
FOR
PROTEIN
was prepared as described previously (7). [32P]Phosphorylase a was prepared as in (8). Prior to use phosphorylase b or a was chromatographed on a Sephadex G-25 column equilibrated with 50 mM Tris-HCI, pH 7.0, 1 mM EDTA, 10 mM 2-mercaptoethanol. Brain homogenate was prepared as follows: Bovine brain was homogenized in 3 vol of 50 mM Tris-HCl buffer, pH 7.4, containing 250 mM sucrose, 15 mM 2-mercaptoethanol, 0.1 mM EDTA, and 0.1 mM phenylmethylsulfonyl fluoride, with a Potter-Elvehjem homogenizer. The supernatant fluid obtained after centrifugation of the homogenate at 17,OOOg for 40 min was finally passed through a Sephadex G-25 column, equilibrated with the homogenization buffer. Casein peptides were prepared by incubating casein (20 mg/ml) with trypsin (2 mg/ml) for 2 h at 22°C in 5 mM Tris-HCl buffer pH 7.0 (1 ml). The reaction mixture was heated for 2 min at lOO”C, centrifuged at 14,000g for 5 min, and chromatographed on a Sephadex G-25 column (22 X 1.6 cm) equilibrated with the reaction buffer. The main phosphopeptide peak eluted from the column was used as a substrate of alkaline phosphatase. Color Reagent The malachite green reagent was prepared as described by Baykov et al. (6) except that the final reagent solution contained 3% (w/v) ammonium molybdate. Phosphate
Determination
The color reagent (50 ~1) was mixed with 200 ~1 of the solution to be analyzed, the mixture was allowed to stand for 10 min at 22°C and the absorbance at 630 nm was measured in lo-mm glass microcuvettes with a Hitachi U-2000 spectrophotometer. Blanks contained water (or the corresponding buffer) and the color reagent. Deproteinization
and Protein
Stabilization
Deproteinization with sulfuric or perchloric acid was performed by adding I vol of 2.25 N H,SO, or 70% (w/w) HClO, to 8 or 22 vol of protein or phosphatase reaction mixture, respectively. After standing for 10 min at O”C, the mixture was centrifuged for 3 min in an Eppendorf (type 54148) centrifuge and the supernatant was removed for phosphate determination. Deproteinization was also performed by heating the protein solution at 100°C for 2 min and removal of the precipitate was as described above. SDS-stabilization of the protein against acid precipitation was performed as follows: 50 ~1 of 10% (w/v) SDS was mixed with 150 ~1 of protein solution and, after 10 min incubation at 22”C, 133 ~1 of 2.5 M KC1 was added to the mixture. After standing for 1 h at 22°C the precipitate produced was removed by two successive centrifugations in the Eppendorf centrifuge
PHOSPHATASE
0
113
ACTIVITY
10 Time
(mlnj
30
FIG. 1. Effect of H,SO, and ammonium molybdate concentration on the time course of color development in presence of 9 gM Pi. Final concentrations in the assay: (A) 1.5% (w/v) and (B) 3% (w/v) ammonium molybdate; (0) 0.97 N and (0), 1.16 N H,SO,.
and the supernatant was further used for phosphate determination. Other Procedures Standard phosphorylase phosphatase activity measurements were performed by determining the release of trichloroacetic acid-soluble radioactivity from 32P-labeled phosphorylase a (8). Protein concentration was measured by a modified Lowry procedure (9). Total phosphate content of phosvitin or casein peptides was determined using the ashing procedure of Ames (10) and the malachite green assay of Pi (6). RESULTS
Kinetics of Color Development A prerequisite for the malachite green procedure for protein phosphatase activity determinations is to eliminate the interference due to precipitation of proteins in the acid environment of the color reagent. Preliminary experiments have shown that such interference may be effectively prevented by treatment with H,SO, or HClO, (final concentrations, 0.25 N and 3% (w/w), respectively). However, the increase of acidity in the final Pi assay mixture drastically reduces the rate of color development. As shown in Fig. 1, increase in the acidity of the standard assay mixture of Baykov et al. (6) (0.97 N H,SO, and 1.5% (w/v) ammonium molybdate) to 1.16 N H,SO, reduces the rate of color development (half-time of the kinetics 1.5 and 17 min at 0.97 N and 1.16 N H,SO,, respectively), in accordance with previous findings demonstrating a dependence of color formation on the final acidity (11). In our effort to overcome the retardation effect of overacidification we found that in-
114
GELADOPOULOS, TABLE
The Efficacy of Deproteinization Procedures in Removing Protein chite Green Assay
SOTIROUDIS,
1
or Protein Stabilization Interference with the MalaTreatment
Protein
No.”
100°C
Albumin Histone II-AS Protamine Phosphorylase b Phosvitin Brain homogenate Human serum
+ +b -
-
+
-
+
+
+e
-
+
-
+
+
+ +
+ +
fb
-
WO,
HClO,
+
SDS/KC1
+
+
Note. Proteins and protein mixtures (2 mg/ml) in 10 mM Tris-HCl buffer pH 7.0, were treated with acid or SDS/KC1 as described under Materials and Methods or heated for 2 min at 100°C. After centrifugation the supernatants (200 ~1) were mixed with the malachite green reagent (50 ~1) and examined for possible interference. (+) No interference;
(-)
formation
of precipitate.
a The proteins were added directly to the assay mixture. b The assay was not influenced by increasing protein concentration up to 6 mg/ml. ’ Protamine interferes at concentrations higher than 1 mg/ml.
creasing the final ammonium molybdate concentration up to 3% (w/v) highly accelerates color formation, so that the reaction is almost complete within 1 min, even in the presence of excess H,SO, (Fig. 1). Similar kinetic results were obtained when perchloric acid (final concentration, 0.28 N) was used for acidification (not shown). At this point it must be added that the doubling of the ammonium molybdate concentration over the formulation of Baykov et al. (6) does not affect the stability characteristics of the original. At 10 PM Pi, the color was stable for more than 30 h (not shown).
Deproteinization or Protein Stabilization in Acid Solution Prior to the Malachite Green Phosphate Assay
AND EVANGELOPOULOS
lybdate largely eliminates the inhibition of color formation (Fig. 2). All proteins examined (Table 1) were either removed and/or stabilized after treatment with H,SO, (0.25 N), which is a component of the color reagent, or with HClO, (3% w/w), so that no interference with subsequent Pi determination was observed. In the case of glycogen phosphorylase, only HCiO, was effective. Taking into account that there is always the possibility of acid hydrolysis of labile phosphate during deproteinization with strong acids (13), we tried to establish a procedure for protein removal or stabilization at neutral pH. In this context, we examined the possibility of protein removal by heat treatment at neutral pH, but as shown in Table 1, only glycogen phosphorylase can be efficiently removed under these conditions. In contrast, when we first denature the proteins with SDS at neutral pH and then remove the excess detergent by precipitation with KC1 (SDS gives intense color reaction with malachite green reagent (14)), proteins do not interfere with Pi determination. In spite of that, phosvitin, a highly phosphorylated protein, could not be stabilized after treatment with SDS/KC1 (Table 1). Standard Phosphate Curves after Deproteinization Protein Stabilization
or
Although proteins may be removed or stabilized by treatment with strong acids or SDS/KC1 at neutral pH prior to Pi determination, it is possible that some Pi may be carried away with the precipitated proteins or that the solubilized proteins interfere with the malachite green Pi assay. Standard plots obtained with bovine albumin, phosvitin, and brain homogenate after treatment with H,SO, (Fig. 3) or with glycogen phosphorylase b and brain homogenate after treatment with SDS/KC1 (Fig. 4) were essentially identical with those obtained in absence of protein. Moreover, Pi standard curves obtained with glycogen phosphorylase b (1 mgl ml) and brain homogenate (0.5 mg/ml) after treatment with HClO, (3% w/w) were again linear up to 4 nmol Pi,
A number of proteins and proteins containing biological materials were used as standards in order to establish methods suitable for deproteinization or protein stabilization in acid solutions prior to the malachite green Pi assay. As shown in Table 1, bovine albumin, Histone II-AS, and human serum can be used in the final Pi assay mixture without interference; a mixture of casein peptides was also soluble in the final assay system, up to of 0.1 mg/ml, without affecting color formation (not shown). All the other proteins or protein mixtures examined interfere precipitation. The popular
because of acid-induced protein precipitants, perchlo-
ric acid and trichloroacetic acid, were observed to provide interference (12) but, although trichloroacetic acid always gives color by itself (not shown), in the case of HClO,
an increased
concentration
of ammonium
mo-
HC104
(%
w/w)
FIG. 2. Effect of HClO, concentration on the color formation by 7.8 pM Pi in presence of 1.5% (w/v) (0) or 3% (w/v) (0) ammonium molybdate. The mixtures were allowed to stand for 10 min before absorption measurement.
COLORIMETRIC
ASSAY
FOR
PROTEIN
PHOSPHATASE
0
nmol
curves in presence of protein solutions treated FIG. 3. Pi standard with H&SO,. Phosphate solutions in 10 mM Tris-HCl buffer, pH 7.0, containing proteins were treated with H,SO, and assayed as described under Materials and Methods. (0) Without protein; (0) albumin, 2 mg/ml; (Al brain homogenate, 0.5 mg/ml; (0) phosvitin, 0.2 mg/ml.
although in this case the assay values showed a difference of about 5% compared with those in absence of proteins (not shown). Application of the Method Determination
to Protein
10
20
Time
Pi
Phosphatase
Based on the previous results, the malachite green micromethod (with or without deproteinization) was applied to the determination of protein phosphatase activities. We used as substrates, phosvitin, a highly phosphorylated naturally occuring protein (3), and glycogen phosphorylase a, a physiological substrate for type 1 and type 2A protein phosphatases (15). A commercial preparation of alkaline phosphatase and a freshly prepared bovine brain extract were used as sources of protein phosphatase activity. As shown in Fig. 5, using the H,SO,-treatment procedure, we were able to follow the alkaline phosphatase-catalyzed release of Pi from phosvitin (present at PM concentration of protein bound phosphate). On the other hand, using a mixture of casein peptides as substrate we could determine Pi released by the phosphatase, directly, in one step, without
Pi
FIG. 4. P, standard curves in presence of protein solutions treated with SDS/KCl. Phosphate solutions in 10 mM Tris-HCl buffer, pH 7.0, containing proteins were treated with SDS/KC1 and assayed as described under Materials and Methods. (A) Without protein; (0) phosphorylase b, 1.5 mg/ml; (0) brain homogenate, 1.2 mg/ml.
30
40
50
(min)
FIG. 5. Time course of the release of Pi from phosvitin by alkaline phosphatase. Alkaline phosphatase (0.1 pg/mll was incubated with phosvitin (0.1 mg/ml) in 40 mM Tris-HCl buffer, pH 8.2, containing 1 mM MgCl,, at 30°C (final volume, 1.5 ml). At various time intervals aliquots of the assay mixture were treated with HeSO,, centrifuged, and assayed for Pi as described under Materials and Methods. The phosphate released at 50 min represents 7% of total phosphate present in the final assay system.
any previous treatment of the enzyme assay mixture (not shown). Furthermore, using the calorimetric Pi microassay, it was possible to determine phosphorylase phosphatase activity in brain homogenates. The comparison of the time course of dephosphorylation determined colorimetrically with that monitored by following the release of [32P]Pi, clearly shows that the results obtained by both methods are almost the same (Fig. 6). To our knowledge this is the first nonradioactive procedure applied for monitoring the enzymatic release of Pi from phosphorylase a, a large homodimer with only one phosphorylated residue in each polypeptide chain (97,400 Da) (16). Concerning the phosphorylase phosphatase assay based on the determination of phosphorylase a catalytic activity, it must be emphasized that this
Time
nmol
115
ACTIVITY
(min)
FIG. 6. Kinetics of Pi released from phosphorylase a catalyzed by phosphorylase phosphatase present in brain homogenate. 32P-labeled phosphorylase a (2 mg/mll was incubated at 30°C with brain homogenate (0.33 mg/mll in 50 mM Tris-HCl buffer, pH 7.0, containing 0.7 mM ethylene glycol bis(@aminoethyl ether) N,N’-tetraacetic acid, 0.3 mM EDTA, 1.3 mM MnCl,, 2 mM caffeine, 80 mM NaCI, 40 mM sucrose, 0.1 mg/ml albumin, and 15 mM 2-mercaptoethanol (final volume, 2 ml). At various time intervals aliquots of the assay mixture were withdrawn and, after protein precipitation with trichloroacetic acid or heating at 100°C. P, released in the supernatant was determined by Cerenkov counting (a) or malachite green assay (0) as described under Materials and Methods.
116
GELADOPOULOS,
SOTIROUDIS,
coupled enzymatic procedure may give quite different courses. This is due to the fact that variations of the conditions for the phosphorylase a assay may result in formation of partially phosphorylated intermediates (17). DISCUSSION Dephosphorylation of phosphopeptides and proteins is most commonly monitored by following the release of [32P]Pi. To avoid interference of soluble phosphopeptides derived by proteolysis of an original protein substrate, or in order to use a peptide as substrate of a phosphatase, specific isolation of Pi or separation of Pi from the phosphopeptide is needed. On the other hand, calorimetric procedures used previously for measuring Pi release in protein phosphatase assays suffered from a lack of sensitivity, so that the presence of high concentrations of a multiphosphorylated protein substrate (casein or phosvitin) in the assay mixture was necessary (18,19). Using the most sensitive calorimetric preocedure for Pi determination reported so far (6,11), we were able to establish and characterize a simple method for following enzyme-catalyzed Pi release from phosphoproteins and peptides. As a result, the amount of released Pi needed to be present in the final assay volume (0.25 ml) in order to yield an absorbance of 0.1 unit is about 0.3 nmol. Although it is given that the radioactive assay is the most sensitive one for the determination of released Pi, the method we described offers several advantages over the radioactive protein phosphatase procedure: (a) It is inexpensive because nonradioactive ATP is used for labeling of the protein substrate. (b) Pi released from a soluble, low M, phosphopeptide can be easily determined in a one-step procedure without needing previous removal or stabilization of the peptide. (c) It offers the possibility to assay the enzymatic dephosphorylation of biologically important protein-bound endogenous phosphate when it is not possible to find suitable conditions or the particular protein kinase necessary for the radiolabeling of the specific site(s). Nevertheless, we must outline some disadvantages of our method: (a) Multisite dephosphorylation of polyphosphorylated protein substrates would complicate the interpretation of kinetic data. In this respect, the use of monophosphorylated peptides or protein substrates is advantageous. (b) The use of complex biological samples (especially at high protein concentrations) as a source of protein phosphatase, may produce large blank values because of enzymatic Pi release from endogenous phosphoproteins or from free and protein-associated (noncovalently) phosphate-containing compounds. In this case, one must select an appropriate method for the efficient removal of such phosphocompounds. (c) One cannot exclude the possibility that certain phosphoprotein substrates or protein containing bi-
AND
EVANGELOPOULOS
ological samples interfere with the color assay (even after deproteinization). Thus, it is necessary always to examine the effect of protein substrates and phosphatase containing mixtures on standard Pi calibration plots. In general, our modification of the malachite green procedure offers the advantage of avoiding overexposure of phosphoprotein substrates to acid and molybdate by enhancing the rate of color formation (increasing molybdate concentration (Fig. 1)) when acidified conditions are used for deproteinization. Moreover, protein stabilization after SDS treatment at neutral pH and subsequent removal of excess SDS with KCl, permits application of the highly sensitive malachite green procedure for assaying protein phosphatases as well as ATPases and other phosphohydrolases without deproteinization. Previous Pi assays (20,21) which have also used SDS for protein solubilization, were relatively insensitive. REFERENCES 1. Edelman, A. M., Blumenthal, Annu. Rev. Biochem. 56,567-613.
2. Cohen,
D. K.,
and
Krebs,
P. (1989) Annu. Rev. Biochem. 68.453-508. M. (1979) Protein Phosphorylation. Pion
3. Weller, don.
4. Buss, J. E., and Stull, J. T. (1983) in Methods (Corbin, J. D., and Hardman, Eds.), Vol. 99, Press, San Diego, CA. 5. Itaya,
E. G. (1987)
K., and Ui, M. (1966)
6. Baykov, Biochem.
A. A., Evtushenko,
Clin.
Chim.
Acta
Limited,
Lon-
in Enzymology 7-14, Academic
14,361-366.
0. A., and Avaeva,
S. M. (1988)
Anal.
171,266-270.
7. Nikolaropoulos, S., and Sotiroudis, T. G. (1985) Eur. J. Biochem. 151,467-473. 8. Shenolikar, S., and Ingebritsen, T. S. (1984) in Methods in Enzymology (Wold, F., and Moldave, K., Eds.), Academic Press, San Diego, CA. 9. Hartree,
E. F. (1972)
Anal.
Biochem.
Vol.
107,
pp. 102-129,
48,422-427.
10. Ames, B. N. (1966) in Methods in Enzymology (Neufeld, E., and Ginsburg, V., Eds.), Vol. 8, pp. 115-118, Academic Press, San Diego, CA. 11. Van Veldhoven, P. P., and Mannaerts, G. P. (1987) them. 161,45-48. 12. Penney, C. L. (1976) Anal. Biochem. 75,201-210. 13. Hohenwallner, W., and Wimmer, E. (1973) Clin. Chim. 169-175. 14. Terasaki, W. L., and Brooker, G. (1976) Anal. Biochem.
Anal.
Bio-
Acta 45, 75,447-
453. 15. Ingebritsen, 16. Goldsmith, Fletterick,
T. S., and Cohen,
P. (1983)
Science
221, 331-338.
E. J., Sprang, S. R., Hamlin, R., Xuong, R. J. (1989) Science 246, 528-532.
17. Hurd, S. S., Teller, Biophys. Res. Commun.
D., and Fischer, 24, 79-84.
E. H.
(1966)
N.-H.,
and
Biochem.
18. Glomset, J. A. (1959) Biochim. Biophys. Acta 32, 349-357. 19. Rose, S. P. R., and Heald, P. J. (1961) Biochem. J. 81, 339-347. 20. Haschke, R. H., and Heilmeyer, L. M. G., Jr. (1972) Anal. Biothem. 47,451-456. 21. Hegyvary, C., Kang, K., and Bandi, Z. (1979) And. Biochem. 94,
397-401.
192, 112-116
(1991)
A Malachite Green Calorimetric Assay for Protein Phosphatase Activity Taxiarchis P. Geladopoulos, Theodore G. Sotiroudis,l and Athanasios Institute of Biological Research,The National Hellenic ResearchFoundation, 48 Vassi.!eos Constantinou Avenue, Athens 116 35, Greece
Received
May
E. Evangelopoulos
31,199O
A simple and sensitive calorimetric assay for protein phosphatase activity based on the determination of released P, by an improved malachite green procedure (A. A. Baykov, 0. A. Evtushenko, and S. M. Avaeva, 1988, Anal. Biochem. 171,266-270) is described. Proteins must be removed or stabilized prior to Pi determination with 0.25 N sulfuric acid or 3% (w/w) perchloric acid. Alternatively, to avoid possible acid hydrolysis of phosphate groups from organic compounds during deproteinization, the protein present in the phosphatase assay mixture can be stabilized with sodium dodecyl sulfate. In this case, the excess detergent is subsequently removed by precipitation with KC1 because it colors with the malachite green reagent. The above procedure was applied to the determination of phosphorylase phosphatase activity in bovine brain extracts and the results are comparable to those obtained with the radioisotopic phosphatase assay. 0 1991 Academic Press, Inc.
by the reduction of phosphomolybdate, although simple, lacks sensitivity (3,4). Highly sensitive assays for Pi have relied upon formation of a colored complex between the acidified molybdate and the dye malachite green (5). The sensitivity of this procedure is 30 times greater than the widely used Fiske-SubbaRow method (4,5). Recently a modified malachite green procedure has been reported by Baykov et al. (6). In this method all necessary reagents are combined in one concentrated solution making the assay more sensitive and convenient (6). In the experiments presented in this paper the malachite green method for Pi determination of Baykov et al. (6) was adapted to measure protein phosphatase activity after precipitation or stabilization of the protein present in the assay mixture. MATERIALS
AND
METHODS
Chemicals The reversible phosphorylation of proteins catalyzed by protein kinases and protein phosphatases is now recognized to be a major process for regulating cellular functions (1,2). A number of methods are available for measuring Pi release in protein phosphatase assays (3). The most sensitive technique involves acid precipitation of a protein substrate containing bound [32P]phosphate and determination of the released radioactivity in the supernatant. This procedure is handicapped by restrictions for working with radioactivity and it cannot distinguish between the release of acid-soluble 32P-labeled peptides through proteolysis and the release of free [3”P]Pi by the phosphatase. On the other hand in nonradioisotopic phosphatase assays, the measurement of released Pi, using the formation of molybdenum blue
’ To whom
correspondence
should
be addressed.
Malachite green (C.I. 42,000), ammonium heptamolybdate, perchloric acid 70%, and sulfuric acid suprapur, were obtained from Merck. Tween 20 and SDS’ were products of Serva and Sigma respectively. [T-~‘P]ATP (3 Ci/mmol) was purchased from Radiochemical Centre, Amersham. Protein Preparations Bovine albumin (pure), protamine sulfate, and casein (vitamin free) were obtained from Serva. Phosvitin, histone type II-AS from calf thymus, trypsin twice crystallized, and alkaline phosphatase from bovine intestinal mucosa (type VII-G) were products of Sigma. Bovine albumin, phosvitin, and human serum were purified by gel filtration on a Sephadex G-25 column. Crystalline rabbit skeletal muscle phosphorylase b * Abbreviation
used:
SDS,
sodium
dodecyl
sulfate.
112 All
Copyright 0 1991 rights of reproduction
0003-2697/91 $3.00 by Academic Press, Inc. in any form reserved.
COLORIMETRIC
ASSAY
FOR
PROTEIN
was prepared as described previously (7). [32P]Phosphorylase a was prepared as in (8). Prior to use phosphorylase b or a was chromatographed on a Sephadex G-25 column equilibrated with 50 mM Tris-HCI, pH 7.0, 1 mM EDTA, 10 mM 2-mercaptoethanol. Brain homogenate was prepared as follows: Bovine brain was homogenized in 3 vol of 50 mM Tris-HCl buffer, pH 7.4, containing 250 mM sucrose, 15 mM 2-mercaptoethanol, 0.1 mM EDTA, and 0.1 mM phenylmethylsulfonyl fluoride, with a Potter-Elvehjem homogenizer. The supernatant fluid obtained after centrifugation of the homogenate at 17,OOOg for 40 min was finally passed through a Sephadex G-25 column, equilibrated with the homogenization buffer. Casein peptides were prepared by incubating casein (20 mg/ml) with trypsin (2 mg/ml) for 2 h at 22°C in 5 mM Tris-HCl buffer pH 7.0 (1 ml). The reaction mixture was heated for 2 min at lOO”C, centrifuged at 14,000g for 5 min, and chromatographed on a Sephadex G-25 column (22 X 1.6 cm) equilibrated with the reaction buffer. The main phosphopeptide peak eluted from the column was used as a substrate of alkaline phosphatase. Color Reagent The malachite green reagent was prepared as described by Baykov et al. (6) except that the final reagent solution contained 3% (w/v) ammonium molybdate. Phosphate
Determination
The color reagent (50 ~1) was mixed with 200 ~1 of the solution to be analyzed, the mixture was allowed to stand for 10 min at 22°C and the absorbance at 630 nm was measured in lo-mm glass microcuvettes with a Hitachi U-2000 spectrophotometer. Blanks contained water (or the corresponding buffer) and the color reagent. Deproteinization
and Protein
Stabilization
Deproteinization with sulfuric or perchloric acid was performed by adding I vol of 2.25 N H,SO, or 70% (w/w) HClO, to 8 or 22 vol of protein or phosphatase reaction mixture, respectively. After standing for 10 min at O”C, the mixture was centrifuged for 3 min in an Eppendorf (type 54148) centrifuge and the supernatant was removed for phosphate determination. Deproteinization was also performed by heating the protein solution at 100°C for 2 min and removal of the precipitate was as described above. SDS-stabilization of the protein against acid precipitation was performed as follows: 50 ~1 of 10% (w/v) SDS was mixed with 150 ~1 of protein solution and, after 10 min incubation at 22”C, 133 ~1 of 2.5 M KC1 was added to the mixture. After standing for 1 h at 22°C the precipitate produced was removed by two successive centrifugations in the Eppendorf centrifuge
PHOSPHATASE
0
113
ACTIVITY
10 Time
(mlnj
30
FIG. 1. Effect of H,SO, and ammonium molybdate concentration on the time course of color development in presence of 9 gM Pi. Final concentrations in the assay: (A) 1.5% (w/v) and (B) 3% (w/v) ammonium molybdate; (0) 0.97 N and (0), 1.16 N H,SO,.
and the supernatant was further used for phosphate determination. Other Procedures Standard phosphorylase phosphatase activity measurements were performed by determining the release of trichloroacetic acid-soluble radioactivity from 32P-labeled phosphorylase a (8). Protein concentration was measured by a modified Lowry procedure (9). Total phosphate content of phosvitin or casein peptides was determined using the ashing procedure of Ames (10) and the malachite green assay of Pi (6). RESULTS
Kinetics of Color Development A prerequisite for the malachite green procedure for protein phosphatase activity determinations is to eliminate the interference due to precipitation of proteins in the acid environment of the color reagent. Preliminary experiments have shown that such interference may be effectively prevented by treatment with H,SO, or HClO, (final concentrations, 0.25 N and 3% (w/w), respectively). However, the increase of acidity in the final Pi assay mixture drastically reduces the rate of color development. As shown in Fig. 1, increase in the acidity of the standard assay mixture of Baykov et al. (6) (0.97 N H,SO, and 1.5% (w/v) ammonium molybdate) to 1.16 N H,SO, reduces the rate of color development (half-time of the kinetics 1.5 and 17 min at 0.97 N and 1.16 N H,SO,, respectively), in accordance with previous findings demonstrating a dependence of color formation on the final acidity (11). In our effort to overcome the retardation effect of overacidification we found that in-
114
GELADOPOULOS, TABLE
The Efficacy of Deproteinization Procedures in Removing Protein chite Green Assay
SOTIROUDIS,
1
or Protein Stabilization Interference with the MalaTreatment
Protein
No.”
100°C
Albumin Histone II-AS Protamine Phosphorylase b Phosvitin Brain homogenate Human serum
+ +b -
-
+
-
+
+
+e
-
+
-
+
+
+ +
+ +
fb
-
WO,
HClO,
+
SDS/KC1
+
+
Note. Proteins and protein mixtures (2 mg/ml) in 10 mM Tris-HCl buffer pH 7.0, were treated with acid or SDS/KC1 as described under Materials and Methods or heated for 2 min at 100°C. After centrifugation the supernatants (200 ~1) were mixed with the malachite green reagent (50 ~1) and examined for possible interference. (+) No interference;
(-)
formation
of precipitate.
a The proteins were added directly to the assay mixture. b The assay was not influenced by increasing protein concentration up to 6 mg/ml. ’ Protamine interferes at concentrations higher than 1 mg/ml.
creasing the final ammonium molybdate concentration up to 3% (w/v) highly accelerates color formation, so that the reaction is almost complete within 1 min, even in the presence of excess H,SO, (Fig. 1). Similar kinetic results were obtained when perchloric acid (final concentration, 0.28 N) was used for acidification (not shown). At this point it must be added that the doubling of the ammonium molybdate concentration over the formulation of Baykov et al. (6) does not affect the stability characteristics of the original. At 10 PM Pi, the color was stable for more than 30 h (not shown).
Deproteinization or Protein Stabilization in Acid Solution Prior to the Malachite Green Phosphate Assay
AND EVANGELOPOULOS
lybdate largely eliminates the inhibition of color formation (Fig. 2). All proteins examined (Table 1) were either removed and/or stabilized after treatment with H,SO, (0.25 N), which is a component of the color reagent, or with HClO, (3% w/w), so that no interference with subsequent Pi determination was observed. In the case of glycogen phosphorylase, only HCiO, was effective. Taking into account that there is always the possibility of acid hydrolysis of labile phosphate during deproteinization with strong acids (13), we tried to establish a procedure for protein removal or stabilization at neutral pH. In this context, we examined the possibility of protein removal by heat treatment at neutral pH, but as shown in Table 1, only glycogen phosphorylase can be efficiently removed under these conditions. In contrast, when we first denature the proteins with SDS at neutral pH and then remove the excess detergent by precipitation with KC1 (SDS gives intense color reaction with malachite green reagent (14)), proteins do not interfere with Pi determination. In spite of that, phosvitin, a highly phosphorylated protein, could not be stabilized after treatment with SDS/KC1 (Table 1). Standard Phosphate Curves after Deproteinization Protein Stabilization
or
Although proteins may be removed or stabilized by treatment with strong acids or SDS/KC1 at neutral pH prior to Pi determination, it is possible that some Pi may be carried away with the precipitated proteins or that the solubilized proteins interfere with the malachite green Pi assay. Standard plots obtained with bovine albumin, phosvitin, and brain homogenate after treatment with H,SO, (Fig. 3) or with glycogen phosphorylase b and brain homogenate after treatment with SDS/KC1 (Fig. 4) were essentially identical with those obtained in absence of protein. Moreover, Pi standard curves obtained with glycogen phosphorylase b (1 mgl ml) and brain homogenate (0.5 mg/ml) after treatment with HClO, (3% w/w) were again linear up to 4 nmol Pi,
A number of proteins and proteins containing biological materials were used as standards in order to establish methods suitable for deproteinization or protein stabilization in acid solutions prior to the malachite green Pi assay. As shown in Table 1, bovine albumin, Histone II-AS, and human serum can be used in the final Pi assay mixture without interference; a mixture of casein peptides was also soluble in the final assay system, up to of 0.1 mg/ml, without affecting color formation (not shown). All the other proteins or protein mixtures examined interfere precipitation. The popular
because of acid-induced protein precipitants, perchlo-
ric acid and trichloroacetic acid, were observed to provide interference (12) but, although trichloroacetic acid always gives color by itself (not shown), in the case of HClO,
an increased
concentration
of ammonium
mo-
HC104
(%
w/w)
FIG. 2. Effect of HClO, concentration on the color formation by 7.8 pM Pi in presence of 1.5% (w/v) (0) or 3% (w/v) (0) ammonium molybdate. The mixtures were allowed to stand for 10 min before absorption measurement.
COLORIMETRIC
ASSAY
FOR
PROTEIN
PHOSPHATASE
0
nmol
curves in presence of protein solutions treated FIG. 3. Pi standard with H&SO,. Phosphate solutions in 10 mM Tris-HCl buffer, pH 7.0, containing proteins were treated with H,SO, and assayed as described under Materials and Methods. (0) Without protein; (0) albumin, 2 mg/ml; (Al brain homogenate, 0.5 mg/ml; (0) phosvitin, 0.2 mg/ml.
although in this case the assay values showed a difference of about 5% compared with those in absence of proteins (not shown). Application of the Method Determination
to Protein
10
20
Time
Pi
Phosphatase
Based on the previous results, the malachite green micromethod (with or without deproteinization) was applied to the determination of protein phosphatase activities. We used as substrates, phosvitin, a highly phosphorylated naturally occuring protein (3), and glycogen phosphorylase a, a physiological substrate for type 1 and type 2A protein phosphatases (15). A commercial preparation of alkaline phosphatase and a freshly prepared bovine brain extract were used as sources of protein phosphatase activity. As shown in Fig. 5, using the H,SO,-treatment procedure, we were able to follow the alkaline phosphatase-catalyzed release of Pi from phosvitin (present at PM concentration of protein bound phosphate). On the other hand, using a mixture of casein peptides as substrate we could determine Pi released by the phosphatase, directly, in one step, without
Pi
FIG. 4. P, standard curves in presence of protein solutions treated with SDS/KCl. Phosphate solutions in 10 mM Tris-HCl buffer, pH 7.0, containing proteins were treated with SDS/KC1 and assayed as described under Materials and Methods. (A) Without protein; (0) phosphorylase b, 1.5 mg/ml; (0) brain homogenate, 1.2 mg/ml.
30
40
50
(min)
FIG. 5. Time course of the release of Pi from phosvitin by alkaline phosphatase. Alkaline phosphatase (0.1 pg/mll was incubated with phosvitin (0.1 mg/ml) in 40 mM Tris-HCl buffer, pH 8.2, containing 1 mM MgCl,, at 30°C (final volume, 1.5 ml). At various time intervals aliquots of the assay mixture were treated with HeSO,, centrifuged, and assayed for Pi as described under Materials and Methods. The phosphate released at 50 min represents 7% of total phosphate present in the final assay system.
any previous treatment of the enzyme assay mixture (not shown). Furthermore, using the calorimetric Pi microassay, it was possible to determine phosphorylase phosphatase activity in brain homogenates. The comparison of the time course of dephosphorylation determined colorimetrically with that monitored by following the release of [32P]Pi, clearly shows that the results obtained by both methods are almost the same (Fig. 6). To our knowledge this is the first nonradioactive procedure applied for monitoring the enzymatic release of Pi from phosphorylase a, a large homodimer with only one phosphorylated residue in each polypeptide chain (97,400 Da) (16). Concerning the phosphorylase phosphatase assay based on the determination of phosphorylase a catalytic activity, it must be emphasized that this
Time
nmol
115
ACTIVITY
(min)
FIG. 6. Kinetics of Pi released from phosphorylase a catalyzed by phosphorylase phosphatase present in brain homogenate. 32P-labeled phosphorylase a (2 mg/mll was incubated at 30°C with brain homogenate (0.33 mg/mll in 50 mM Tris-HCl buffer, pH 7.0, containing 0.7 mM ethylene glycol bis(@aminoethyl ether) N,N’-tetraacetic acid, 0.3 mM EDTA, 1.3 mM MnCl,, 2 mM caffeine, 80 mM NaCI, 40 mM sucrose, 0.1 mg/ml albumin, and 15 mM 2-mercaptoethanol (final volume, 2 ml). At various time intervals aliquots of the assay mixture were withdrawn and, after protein precipitation with trichloroacetic acid or heating at 100°C. P, released in the supernatant was determined by Cerenkov counting (a) or malachite green assay (0) as described under Materials and Methods.
116
GELADOPOULOS,
SOTIROUDIS,
coupled enzymatic procedure may give quite different courses. This is due to the fact that variations of the conditions for the phosphorylase a assay may result in formation of partially phosphorylated intermediates (17). DISCUSSION Dephosphorylation of phosphopeptides and proteins is most commonly monitored by following the release of [32P]Pi. To avoid interference of soluble phosphopeptides derived by proteolysis of an original protein substrate, or in order to use a peptide as substrate of a phosphatase, specific isolation of Pi or separation of Pi from the phosphopeptide is needed. On the other hand, calorimetric procedures used previously for measuring Pi release in protein phosphatase assays suffered from a lack of sensitivity, so that the presence of high concentrations of a multiphosphorylated protein substrate (casein or phosvitin) in the assay mixture was necessary (18,19). Using the most sensitive calorimetric preocedure for Pi determination reported so far (6,11), we were able to establish and characterize a simple method for following enzyme-catalyzed Pi release from phosphoproteins and peptides. As a result, the amount of released Pi needed to be present in the final assay volume (0.25 ml) in order to yield an absorbance of 0.1 unit is about 0.3 nmol. Although it is given that the radioactive assay is the most sensitive one for the determination of released Pi, the method we described offers several advantages over the radioactive protein phosphatase procedure: (a) It is inexpensive because nonradioactive ATP is used for labeling of the protein substrate. (b) Pi released from a soluble, low M, phosphopeptide can be easily determined in a one-step procedure without needing previous removal or stabilization of the peptide. (c) It offers the possibility to assay the enzymatic dephosphorylation of biologically important protein-bound endogenous phosphate when it is not possible to find suitable conditions or the particular protein kinase necessary for the radiolabeling of the specific site(s). Nevertheless, we must outline some disadvantages of our method: (a) Multisite dephosphorylation of polyphosphorylated protein substrates would complicate the interpretation of kinetic data. In this respect, the use of monophosphorylated peptides or protein substrates is advantageous. (b) The use of complex biological samples (especially at high protein concentrations) as a source of protein phosphatase, may produce large blank values because of enzymatic Pi release from endogenous phosphoproteins or from free and protein-associated (noncovalently) phosphate-containing compounds. In this case, one must select an appropriate method for the efficient removal of such phosphocompounds. (c) One cannot exclude the possibility that certain phosphoprotein substrates or protein containing bi-
AND
EVANGELOPOULOS
ological samples interfere with the color assay (even after deproteinization). Thus, it is necessary always to examine the effect of protein substrates and phosphatase containing mixtures on standard Pi calibration plots. In general, our modification of the malachite green procedure offers the advantage of avoiding overexposure of phosphoprotein substrates to acid and molybdate by enhancing the rate of color formation (increasing molybdate concentration (Fig. 1)) when acidified conditions are used for deproteinization. Moreover, protein stabilization after SDS treatment at neutral pH and subsequent removal of excess SDS with KCl, permits application of the highly sensitive malachite green procedure for assaying protein phosphatases as well as ATPases and other phosphohydrolases without deproteinization. Previous Pi assays (20,21) which have also used SDS for protein solubilization, were relatively insensitive. REFERENCES 1. Edelman, A. M., Blumenthal, Annu. Rev. Biochem. 56,567-613.
2. Cohen,
D. K.,
and
Krebs,
P. (1989) Annu. Rev. Biochem. 68.453-508. M. (1979) Protein Phosphorylation. Pion
3. Weller, don.
4. Buss, J. E., and Stull, J. T. (1983) in Methods (Corbin, J. D., and Hardman, Eds.), Vol. 99, Press, San Diego, CA. 5. Itaya,
E. G. (1987)
K., and Ui, M. (1966)
6. Baykov, Biochem.
A. A., Evtushenko,
Clin.
Chim.
Acta
Limited,
Lon-
in Enzymology 7-14, Academic
14,361-366.
0. A., and Avaeva,
S. M. (1988)
Anal.
171,266-270.
7. Nikolaropoulos, S., and Sotiroudis, T. G. (1985) Eur. J. Biochem. 151,467-473. 8. Shenolikar, S., and Ingebritsen, T. S. (1984) in Methods in Enzymology (Wold, F., and Moldave, K., Eds.), Academic Press, San Diego, CA. 9. Hartree,
E. F. (1972)
Anal.
Biochem.
Vol.
107,
pp. 102-129,
48,422-427.
10. Ames, B. N. (1966) in Methods in Enzymology (Neufeld, E., and Ginsburg, V., Eds.), Vol. 8, pp. 115-118, Academic Press, San Diego, CA. 11. Van Veldhoven, P. P., and Mannaerts, G. P. (1987) them. 161,45-48. 12. Penney, C. L. (1976) Anal. Biochem. 75,201-210. 13. Hohenwallner, W., and Wimmer, E. (1973) Clin. Chim. 169-175. 14. Terasaki, W. L., and Brooker, G. (1976) Anal. Biochem.
Anal.
Bio-
Acta 45, 75,447-
453. 15. Ingebritsen, 16. Goldsmith, Fletterick,
T. S., and Cohen,
P. (1983)
Science
221, 331-338.
E. J., Sprang, S. R., Hamlin, R., Xuong, R. J. (1989) Science 246, 528-532.
17. Hurd, S. S., Teller, Biophys. Res. Commun.
D., and Fischer, 24, 79-84.
E. H.
(1966)
N.-H.,
and
Biochem.
18. Glomset, J. A. (1959) Biochim. Biophys. Acta 32, 349-357. 19. Rose, S. P. R., and Heald, P. J. (1961) Biochem. J. 81, 339-347. 20. Haschke, R. H., and Heilmeyer, L. M. G., Jr. (1972) Anal. Biothem. 47,451-456. 21. Hegyvary, C., Kang, K., and Bandi, Z. (1979) And. Biochem. 94,
397-401.
