Biochimica et Bioptzysica Acta, 1094 (1991) 130-133 © 1991 Elsevier Science Publishers B.V, 0167-4889/91/$03.50 ADONIS 0167488991002222
130
BBAMCR 12982
The use of phosphopeptides to distinguish between protein phosphatase and acid/alkaline phosphatase activities: opposite specificity toward phosphoseryl/phosphothreonyl substrates Arianna Donella-Deana ~, Helmut E. Meyer 2 and L.A. Pinna I Dipartimento di Chimica Biologica, Universita' di PadoL'a and Centre per Io Studio della Fisiologia Mitocondriale, Padoca (Italy) and 2 lnstitut fur Physiologische Chemie, Ruhr-Unit'ersitat Bochum, Bochum (F.R. G.)
(Received 4 April 1991)
Key words: Acid phosphatase;Alkaline phosphatase;Synthetic peptide; Substrate specificity; Enzymology; Peptide dephosphorylation
The four main classes of protein phosphatases (PP-I, 2A, 2B and 2C), although differing in their ability to dephosphorylate phosphopeptide substrates, invariably display a marked preference toward phosphothreonyi peptides over their phosphoseryi counterparts. Conversely, all the acidic and alkaline phosphatases tested so far dephosphorylate phosphoseryl derivatives far more readily than phosphothreonyi ones. This opposite behaviour provides a criterion for discriminating between protein dephosphorylating activity due to authentic protein phosphatases as compared to nonspecifie acid a n d / o r alkaline phosphatases. In particular the phosphothreonyl peptides RRATt,VA and RRREEETpEEEAA appear to be especially suited for detecting the activity of PP-2C and PP-2A, since they are hardly dephosphorylated by acid and alkaline phosphatases. Conversely, the phosphoseryl peptides SpEEEEE and RRASpVA can provide a sensitive evaluation of the majority of acid and alkaline phosphatases, while being refractory to protein phosphatases. Protein pbosphatases involved in the dephosphorylation of phosphoseryl and phosphothreonyl residues of proteins can be grouped into four major classes, termed PP-1, ZA, 2B and 2C according to their sensitivity to heat-stable protein inhibitors and different response to Ca 2+ and Mg 2+ [1]. Protein pbosphatases 2A are also termed PCS-protein pbosphatases after their stimulation by polycationic compounds, like polylysine [2]. These four classes of protein phospha. tases also differ in their specificity toward small phospborylated peptides: while PP-1 and 2B display negligible activity towards small peptides (Ref. 3 and unpub. lished data in collaboration with C.B. Klee) suggesting they recognize higher order structural features, both PP-2A (PC$) [3,4] and PP-2C [5] readily dephospho-
Abbreviations. CK-2, casein kinase-2; PK-A, cAMP dependent protein kinas¢ PP-1, -2A, -2B, -2C. protein phosphatase-l, 2A, -2B, 2C. Corrvslmndence: A. Doneila-Deana, Dipartimento di Chimica BioIogica, Via Trieste 75, 35121 Padova, Italy.
rylate a variety of small phosphopeptides at rates comparable with those of their protein substrates. The specificity of PP-2C, however, appears to be narrower than that of PP-2A, since some peptides which are excellent substrates for the latter [4], like RRREEETpEEEAA and RRRRAASpVA, are nearly unaffected by the former [5]. Irrespective of these differences a curious feature shared by all protein phospharases is their remarkable preference for phosphothreonyl peptides over their phosphoseryl homologues. Such a preference, firstly described for a protein phosphatase termed protein phosphatase-T [6] and subsequently identified as a subspecies of the PP-2A (PCS) family [7] has been shown to be a constant feature of all the members of this class of protein phosphatases [3,4] as well as of PP-1 [3], PP-2C [5] and PP-2B (unpublished data in collaboration with C.B. Klee). A synopsis of the data supporting this concept is presented in Table I showing that the phosphopeptide RRATpVA, although dephosphorylated at quite different rates by the four classes of protein phosphatases, is nevertheless preferred by all of them over its phosphoseryl counterpart. Such a preference is especially remarkable with PP-2A and PP-2C, while less dramatic
131 TABLE I
Preferential dephosphorylation of phosphothreonyl over phosphoseryl peptides by different protein phosphatases Enzymes
Relative PPase activity toward a
Ratio
RRATpVA
RRASpVA
Tp/Sp
2.0 883.0 0.7 253.0
0.1 17.0 0.2 9.0
Ref.
derivative PP-1 b PP'2A c PP-2B PP-2C
20.0 51.9 3.5 28.1
3 3, 4 d 5
a Expressed as % of the reference substrates: phosphorylase-a (10 ttM) for PP-I and PP-2A as detailed in Ref. 3, synthetic peptide corresponding to residues 81-99 of RII subunit of cAMP dependent protein kinase (8 t~M) for PP-2B as detailed in Ref. 17 and 3"P-casein (2 ~M) for PP-2C as detailed in Ref. 5. Amino acid residues are denoted by the one letter symbols, Tp and Sp indicating phosphothreonine and phosphoserine, respectively. b Catalytic subunit (AMDc) [3]. c PCSM [4]. a Unpublished data in collaboration with C.B. Klee.
with PP-1 and PP-2B, whose activity toward any kind of short peptide substrates however is intrinsically very low. Similar results were obtained with protein phosphatase 2A by replacing RRATpVA/RRASpVA with
other pairs of phosphopeptides whose members differed for just S e r P / T h r P substitutions, like SpEEEEE/TpEEEEE and RRREEESpEEEAA/ RRREEETpEEEAA [4]. Although their dephosphorylation rate was deeply influenced by the overall structure of the peptide the phosphothreonyl homologues were invariably dephosphorylated at a much higher rate than the phosphoseryl ones, thus providing the demonstration that phosphothreonine is intrinsically preferred over phosphoserine by protein phosphatases. In particular the phosphoseryl peptide SpEEEEE proved totally refractory to all the protein phosphatases tested so far [4,5] It is well known on the other hand that many acidic and alkaline phosphatases do also display protein phosphatase activity at pH close to neutrality, a property which is routinely exploited for the nonspecific dephosphorylation of phosphorylated proteins in vitro (e.g., Refs. 8-14), Consequently such classes of phosphatases may also contribute substantially to the overall protein phosphatase activity when assayed in intact cells and crude extracts. Interestingly, however, the phosphopeptide RRRRPSpPA was reported to be a better substrate than its phosphothreonyl counterpart for an acid and alkaline phosphatase [15]. A similar
TABLE II
Rates of dephosphorylation of phosphopeptides by acid and alkaline phosphatases Phosphopeptide phosphatase activities are expressed as pmol of Pi" min-t released by 0.05 U of phosphatase, one unit being defined as the amount of enzyme that hydrolyzes 1 /.Lmolof p-nitrophenylphosphate, min-i under standard assay conditions [18], at the optimal pH for each enzyme. Acid phosphatase from potato, wheat germ, bovine prostate and alkaline phosphatase from Escherichia coli and bovine intestinal mucosa were purchased from Sigma. Repressible acid and alkaline phosphatases, purified from Saccharomyces cerevisiae cells, according to Ref. 19 and 20, respectively, were kindly supplied by Dr. S. Barbaric' (University of Zagreb, Yugoslavia). RRASVA, RRATVA, SEEEEE and TEEEEE were a generous gift from Dr. F. Marchiori (Universita' di Padova, Italy). RRASVA and RRATVA were phosphorylated by incubation with catalytic subunit of cyclic AMP-dependent protein kinase from bovine cardiac muscle [21,4], other peptides were labelled by casein kinase 2 [22,4], Incubations (4 h) were stopped by addition of 30% (v/v) acetic acid and [~/32P]ATP removed by ion exchange chromatography on a 2 x 0.5 cm Dowex AGIX8 column (Bio-Rad) equilibrated with 30% (v:v) acetic acid [23]. The eluate containing the phosphorylated substrates was lyophilised and dissolved in H20. The concentration of 32p-labelled substrates in the assays (calculated from specific radioactivity)was 2/~M. No significant influence of the concentration of nonophosphorylated substrate on the dephosphorylation rate could be observed over the range 3-300 /~M. Phosphatase activities were assayed in 50/zl of incubation mixture containing 10 mU of enzyme and 2/~ M 32P-labelled peptides for 10 rain at 30°C at optimal pH values for phosphopeptide substrates, namely: 7.5 (Saccharomyces cerevisiae alkaline phosphatase); 7.5 (intestinal mucosa); 7 (E. colD; 6 (wheat germ); 6 (potato); 6 (Saccharomices cerevisiae acid phosphatase); and 7 (prostatic acid phosphatase). 5 mM MgCl: was added to the incubation mixture, in the case of S. cerevisiae alkaline phosphatase. Reactions were stopped by the addition of 10% (w/v) trichloroacetic acid and the [32p]phosphate liberated was converted into its phosphomolybdic complex, e~aracted with isobutyl alcohol-toluene (1 : 1, v/v) and quantitated as in Ref. 24. Assays with each substrate were linear up to 10-20% phosphate released and dephosphorylation was kept within this limit to ensure that initial rate conditions were met. Amino acid residues are denoted by the one letter symbols Sp and Tp denbting phosphoserine and phosphothreonine, respectively. Average values calculated from 5-10 determinations are shown and the standard error was less than 14%. Substrates
Dhncnh~t~p~
alkaline RRASpVA RRATpVA SpEEEEE TpEEEEE RRREEES vEEEAA RRREEETp EEEAA
acid
S. cerevisiae
intestinal
E. coil
wheat germ
potato
S. cerevisiae
prostatic
36.2 0.0 35.5 9.4 0.7 0.0
163.2 20.1 783.0 124.2 -
282.0 0.0 89.2 28.3 2.0
4.3 0.7 166.2 19.8 35.7
2.5 0.2 809.0 43.3 40.0 9.9
25.7 1.2 0.0 0.0 1.0 0.0
102.5 16.2 104.2 20.5 34.2 2.5
-
0.0
5.9
132 preference for a phosphoseryl vs. phosphothreonyl peptide was exhibited by a yeast acid phosphatase [16]. These observations prompted us to undertake a systematic study on the phosphopeptide specificity of a variety of acidic and alkaline phosphatases, in order to assess whether they can be distinguished from authentic protein phosphatases by a different selectivity toward phosphoserine and phosphothreonine residues. in Table II the results relative to three alkaline and four acidic phosphatases are reported. They were assayed for their ability to dephosphorylate three pairs of phosphopeptides each consisting of the phosphoseryi and phosphothreonyl derivatives of otherwise identical compounds, namely RRASr(Tp)VA, Sp(Tp)EEEEE and RRREEESe(Tp)EEEAA. Invariably, phosphoser?l peptides are by far preferred by these phospharases over their phosphothreonyl counterparts, the latter in some instances being totally unaffected. Such behaviour is opposite to that of protein phosphatases whose predilection for phosphothreonyl peptides is well established (see Table 1). It should be noted however that acidic and alkaline phosphatases, apart from this common feature, differ widely in their peptide substrate specificity. Thus E. coli alkaline phosphatase is especially active on RRASpVA, while the phosphopeptide SpEEEEE is by far preferred by potato acid phosphatase and by intestinal mucosa alkaline phosphatase. The narrowest peptide substrate specificity is observed with yeast repressible acid phosphatase displaying significant activity only toward RRASpVA. This phosphopeptide, moreover, is more or less readily dephosphorylated by all acid and alkaline phosphatases tested so far and it could represent a suitable substrate for the detection and evaluation of these enzymes by virtue of its resistance to protein phosphatases. Likewise, the phosphopeptide SpEF~EE is also totally unaffected by protein phosphatases, while being dephosphorylated efficiently by all the acid and alkaline phosphatases apart from Saceharomyc~ cerevisiae acid phosphatase. Consequently, the peptide $pEEEEE is especially suited for evaluating wheat germ and potato acid phosphatases, whose activity toward RRASpVA is rather low. On the other hand the phosphothreonyl peptide RRATpVA is an excellent substrate for both PP-2A and PP-2C, but a very poor one for all the acid and alkaline phospharases except intestinal alkaline phosphatase. As a general rule therefore efficient dephosphorylation of RRATpVA can be taken as an indication of high protein phosphatase activity mainly attributable to PP2A and/or PP-2C, whereas the contribution of acid and alkaline phosphatases at nearly neutral pH can be disclosed by determining the dephosphorylation rate of the phosphoseryl peptides RRASpVA a n d / o r SpEEEEE. Either one or the other, or both, are readily dephesphorylated by all the acid and alkaline phos-
phatases tested so far, while they are nearly unaffected by protein phosphatases. All the phosphoueptides described here can be easily obtained as 32p-derivatives by incubation of the corresponding peptides with [3,3zp]ATP in the presence of either cyclic AMP-dependent protein kinase or casein kinase-2 [4,5]. Apart from these methodological applications it is also possible that phosphopeptide substrates will prove useful for mechanistic studies on the specificity and catalytic mechanism of acid and alkaline phosphatases considering how strongly and variably their dephosphorylation efficiency is influenced by the nature of the phosphoaminoacid and the surrounding residues. It should be noted in this connection that the different dephosphorylation efficiencies outlined in Table II using a 2 /zM concentration of phosphopeptide substrates may reflect differences not only in Vm~,x but also in K m values, these latter being quite variable for the same phosphopeptide RRASpVA depending on the phosphatase. In particular, while the g m with yeast acid phosphatase (0.3/~M) is lower than the substrate concentration generally used, the K m values with the other phosphatases tested are higher (5.7, 8.3 and 33 /zM with E. coil alkaline phosphatase, yeast alkaline phosphatase and prostatic acid phosphatase, respectively). As expected therefore, at saturating concentration, RRASpVA would become an even better substrate especially for prostatic acid phosphatase, which displays the highest K m value. The superiority of RRASrVA over its phosphothrconyl derivative however is not expected to cbap.ge by varying the substrate concentration, since the phosphoseryl and phosphothreonyl peptides exhibit the same K m, albeit quite different Vm~x values. Likewise, the opposite selectivity of protein phosphatase-2A for the phosphothreonyl over phosphoseryl derivative is also due to sharply different Vm~, values [3].
Acknowledgments This work was supported by Italian M.U.R.S.T. and C.N.R. (Target Project on Biotechnology and Bioinstrumentation and grant 90.02460.CT04). The skilful technical assistance of Mr G. Tasinato is gratefully acknowledged.
References 1 Cohen, P. (1989) Annu. Rev. Biochem. 58, 453-508. 2 Waelkens, E., Goris, L and Merlevede, W. (1987) J. Biol. Chem. 262, 1049-1059. 3 Agostinis, P., Goris, J. Waelkens, E., Pinna, L.A., Marchiori, F. and Merlevede, W. (1987) J. Biol. Chem. 262, 1060-1064. 4 Agostinis, P., Goris, J., Pinna, LA., Marchiori, F., Perich, J. W., Meyer, H.E. and Merlevede, W. (1990) Eur. J. Biochem. 189, 235-241. 5 Donella-Deana, A., Mac Gowan, C.H., Cohen, P., Marchiori, F.,
133 Meyer, H.E. and Pinna, L.A. (1990) Biochim. Biophys. Acta 1051, 199-202. 6 Donella-Deana, A., Marchiori, F., Meggio, F. and Pinna, L.A. (1982) J. Biol. Chem. 257, 8565-8568. 7 Donella-Deana, A. and Pinna, L.A. (1988) Biochim. Biophys. Acta 968, 179-185. 8 Stinson, R.A. and Chan, J.R.A. (1987) Adv. Prot. Phosphatases 4, 127-151. 9 Lin, M.-F. and Clinton, G.M. (1987) Adv. Prot. Phosphatases 4, 199-228. 10 Lougovoi, C.P. and Kyriakidis, D.A. (1989) Biochim. Biophys. Acta 996, 70-75. 11 Ackerman, P. and Osheroff, N. (1989) J. BioI.Chem. 264, 1195811965. 12 Maelandsmo, G.M., Ostvold, A. and Laland, S. (1989) Eur. J. Biochem. 184, 529-534. 13 Yamashita, T., Hoshi, M., Kawakami, M. and Sakai, H. (1990) Eur. J. Biochem. 193, 661-669. 14 Litwin, C.M.E., Cheng, H. and Wang J.H. (1991) J. Biol. Chem. 266, 2557-2566.
15 Kemp, B.E. (1980) FEBS Lett. I10, 308-312. 16 Doneila-Deana, A., Lopandic, K., Barbaric, S. and Pinna, L.A. (1986) Biochem. Biophys. Res. Commun. 139, 1202-1209. 17 Biumenthal, D.K., Takio, K., Hansen, R.S. and Krebs, E.G. (1986) J. Biol. Chem. 261, 8140-8145. 18 Barbaric', S., Kozulic', B., Ries, B. and Mildner, P. (1980) Biochim. Biophys. Res. Commun. 95, 404-409. 19 Barbaric', S., Kozulic', B., Ries, B. and Mildner, P. (1984) J. Biol. Chem. 259, 878-883. 20 Onishi, H.R., Tkacz, J.S. and Lampen, J.O. (1979) J. Biol. Chem. 254, 11943-11952. 21 Chessa, G., Borin, G., Marchiori, F., Meggio. F., Brunati, A.M. and Pinna, LA. (1983) Eur. J. Biochem. 135, 609-614. 22 Meggio, F., Marchiori, F., Borin, G., Chessa, G. and Pinna, L.A. (1984) J. Biol. Chem. 259. 14576-14579. 23 Kemp, B.E., Bylund, D.B., Huang, T.S. and Krebs, E.G. (1975) Proc. Natl. Acad. Sci. USA 72, 3448-3452. 24 Pinna, L.A., Donella, A., Clari, G. and Moret, V. (1976) Biochem. Biophys. Rcs. Commun. 70, 1308-1315.
130
BBAMCR 12982
The use of phosphopeptides to distinguish between protein phosphatase and acid/alkaline phosphatase activities: opposite specificity toward phosphoseryl/phosphothreonyl substrates Arianna Donella-Deana ~, Helmut E. Meyer 2 and L.A. Pinna I Dipartimento di Chimica Biologica, Universita' di PadoL'a and Centre per Io Studio della Fisiologia Mitocondriale, Padoca (Italy) and 2 lnstitut fur Physiologische Chemie, Ruhr-Unit'ersitat Bochum, Bochum (F.R. G.)
(Received 4 April 1991)
Key words: Acid phosphatase;Alkaline phosphatase;Synthetic peptide; Substrate specificity; Enzymology; Peptide dephosphorylation
The four main classes of protein phosphatases (PP-I, 2A, 2B and 2C), although differing in their ability to dephosphorylate phosphopeptide substrates, invariably display a marked preference toward phosphothreonyi peptides over their phosphoseryi counterparts. Conversely, all the acidic and alkaline phosphatases tested so far dephosphorylate phosphoseryl derivatives far more readily than phosphothreonyi ones. This opposite behaviour provides a criterion for discriminating between protein dephosphorylating activity due to authentic protein phosphatases as compared to nonspecifie acid a n d / o r alkaline phosphatases. In particular the phosphothreonyl peptides RRATt,VA and RRREEETpEEEAA appear to be especially suited for detecting the activity of PP-2C and PP-2A, since they are hardly dephosphorylated by acid and alkaline phosphatases. Conversely, the phosphoseryl peptides SpEEEEE and RRASpVA can provide a sensitive evaluation of the majority of acid and alkaline phosphatases, while being refractory to protein phosphatases. Protein pbosphatases involved in the dephosphorylation of phosphoseryl and phosphothreonyl residues of proteins can be grouped into four major classes, termed PP-1, ZA, 2B and 2C according to their sensitivity to heat-stable protein inhibitors and different response to Ca 2+ and Mg 2+ [1]. Protein pbosphatases 2A are also termed PCS-protein pbosphatases after their stimulation by polycationic compounds, like polylysine [2]. These four classes of protein phospha. tases also differ in their specificity toward small phospborylated peptides: while PP-1 and 2B display negligible activity towards small peptides (Ref. 3 and unpub. lished data in collaboration with C.B. Klee) suggesting they recognize higher order structural features, both PP-2A (PC$) [3,4] and PP-2C [5] readily dephospho-
Abbreviations. CK-2, casein kinase-2; PK-A, cAMP dependent protein kinas¢ PP-1, -2A, -2B, -2C. protein phosphatase-l, 2A, -2B, 2C. Corrvslmndence: A. Doneila-Deana, Dipartimento di Chimica BioIogica, Via Trieste 75, 35121 Padova, Italy.
rylate a variety of small phosphopeptides at rates comparable with those of their protein substrates. The specificity of PP-2C, however, appears to be narrower than that of PP-2A, since some peptides which are excellent substrates for the latter [4], like RRREEETpEEEAA and RRRRAASpVA, are nearly unaffected by the former [5]. Irrespective of these differences a curious feature shared by all protein phospharases is their remarkable preference for phosphothreonyl peptides over their phosphoseryl homologues. Such a preference, firstly described for a protein phosphatase termed protein phosphatase-T [6] and subsequently identified as a subspecies of the PP-2A (PCS) family [7] has been shown to be a constant feature of all the members of this class of protein phosphatases [3,4] as well as of PP-1 [3], PP-2C [5] and PP-2B (unpublished data in collaboration with C.B. Klee). A synopsis of the data supporting this concept is presented in Table I showing that the phosphopeptide RRATpVA, although dephosphorylated at quite different rates by the four classes of protein phosphatases, is nevertheless preferred by all of them over its phosphoseryl counterpart. Such a preference is especially remarkable with PP-2A and PP-2C, while less dramatic
131 TABLE I
Preferential dephosphorylation of phosphothreonyl over phosphoseryl peptides by different protein phosphatases Enzymes
Relative PPase activity toward a
Ratio
RRATpVA
RRASpVA
Tp/Sp
2.0 883.0 0.7 253.0
0.1 17.0 0.2 9.0
Ref.
derivative PP-1 b PP'2A c PP-2B PP-2C
20.0 51.9 3.5 28.1
3 3, 4 d 5
a Expressed as % of the reference substrates: phosphorylase-a (10 ttM) for PP-I and PP-2A as detailed in Ref. 3, synthetic peptide corresponding to residues 81-99 of RII subunit of cAMP dependent protein kinase (8 t~M) for PP-2B as detailed in Ref. 17 and 3"P-casein (2 ~M) for PP-2C as detailed in Ref. 5. Amino acid residues are denoted by the one letter symbols, Tp and Sp indicating phosphothreonine and phosphoserine, respectively. b Catalytic subunit (AMDc) [3]. c PCSM [4]. a Unpublished data in collaboration with C.B. Klee.
with PP-1 and PP-2B, whose activity toward any kind of short peptide substrates however is intrinsically very low. Similar results were obtained with protein phosphatase 2A by replacing RRATpVA/RRASpVA with
other pairs of phosphopeptides whose members differed for just S e r P / T h r P substitutions, like SpEEEEE/TpEEEEE and RRREEESpEEEAA/ RRREEETpEEEAA [4]. Although their dephosphorylation rate was deeply influenced by the overall structure of the peptide the phosphothreonyl homologues were invariably dephosphorylated at a much higher rate than the phosphoseryl ones, thus providing the demonstration that phosphothreonine is intrinsically preferred over phosphoserine by protein phosphatases. In particular the phosphoseryl peptide SpEEEEE proved totally refractory to all the protein phosphatases tested so far [4,5] It is well known on the other hand that many acidic and alkaline phosphatases do also display protein phosphatase activity at pH close to neutrality, a property which is routinely exploited for the nonspecific dephosphorylation of phosphorylated proteins in vitro (e.g., Refs. 8-14), Consequently such classes of phosphatases may also contribute substantially to the overall protein phosphatase activity when assayed in intact cells and crude extracts. Interestingly, however, the phosphopeptide RRRRPSpPA was reported to be a better substrate than its phosphothreonyl counterpart for an acid and alkaline phosphatase [15]. A similar
TABLE II
Rates of dephosphorylation of phosphopeptides by acid and alkaline phosphatases Phosphopeptide phosphatase activities are expressed as pmol of Pi" min-t released by 0.05 U of phosphatase, one unit being defined as the amount of enzyme that hydrolyzes 1 /.Lmolof p-nitrophenylphosphate, min-i under standard assay conditions [18], at the optimal pH for each enzyme. Acid phosphatase from potato, wheat germ, bovine prostate and alkaline phosphatase from Escherichia coli and bovine intestinal mucosa were purchased from Sigma. Repressible acid and alkaline phosphatases, purified from Saccharomyces cerevisiae cells, according to Ref. 19 and 20, respectively, were kindly supplied by Dr. S. Barbaric' (University of Zagreb, Yugoslavia). RRASVA, RRATVA, SEEEEE and TEEEEE were a generous gift from Dr. F. Marchiori (Universita' di Padova, Italy). RRASVA and RRATVA were phosphorylated by incubation with catalytic subunit of cyclic AMP-dependent protein kinase from bovine cardiac muscle [21,4], other peptides were labelled by casein kinase 2 [22,4], Incubations (4 h) were stopped by addition of 30% (v/v) acetic acid and [~/32P]ATP removed by ion exchange chromatography on a 2 x 0.5 cm Dowex AGIX8 column (Bio-Rad) equilibrated with 30% (v:v) acetic acid [23]. The eluate containing the phosphorylated substrates was lyophilised and dissolved in H20. The concentration of 32p-labelled substrates in the assays (calculated from specific radioactivity)was 2/~M. No significant influence of the concentration of nonophosphorylated substrate on the dephosphorylation rate could be observed over the range 3-300 /~M. Phosphatase activities were assayed in 50/zl of incubation mixture containing 10 mU of enzyme and 2/~ M 32P-labelled peptides for 10 rain at 30°C at optimal pH values for phosphopeptide substrates, namely: 7.5 (Saccharomyces cerevisiae alkaline phosphatase); 7.5 (intestinal mucosa); 7 (E. colD; 6 (wheat germ); 6 (potato); 6 (Saccharomices cerevisiae acid phosphatase); and 7 (prostatic acid phosphatase). 5 mM MgCl: was added to the incubation mixture, in the case of S. cerevisiae alkaline phosphatase. Reactions were stopped by the addition of 10% (w/v) trichloroacetic acid and the [32p]phosphate liberated was converted into its phosphomolybdic complex, e~aracted with isobutyl alcohol-toluene (1 : 1, v/v) and quantitated as in Ref. 24. Assays with each substrate were linear up to 10-20% phosphate released and dephosphorylation was kept within this limit to ensure that initial rate conditions were met. Amino acid residues are denoted by the one letter symbols Sp and Tp denbting phosphoserine and phosphothreonine, respectively. Average values calculated from 5-10 determinations are shown and the standard error was less than 14%. Substrates
Dhncnh~t~p~
alkaline RRASpVA RRATpVA SpEEEEE TpEEEEE RRREEES vEEEAA RRREEETp EEEAA
acid
S. cerevisiae
intestinal
E. coil
wheat germ
potato
S. cerevisiae
prostatic
36.2 0.0 35.5 9.4 0.7 0.0
163.2 20.1 783.0 124.2 -
282.0 0.0 89.2 28.3 2.0
4.3 0.7 166.2 19.8 35.7
2.5 0.2 809.0 43.3 40.0 9.9
25.7 1.2 0.0 0.0 1.0 0.0
102.5 16.2 104.2 20.5 34.2 2.5
-
0.0
5.9
132 preference for a phosphoseryl vs. phosphothreonyl peptide was exhibited by a yeast acid phosphatase [16]. These observations prompted us to undertake a systematic study on the phosphopeptide specificity of a variety of acidic and alkaline phosphatases, in order to assess whether they can be distinguished from authentic protein phosphatases by a different selectivity toward phosphoserine and phosphothreonine residues. in Table II the results relative to three alkaline and four acidic phosphatases are reported. They were assayed for their ability to dephosphorylate three pairs of phosphopeptides each consisting of the phosphoseryi and phosphothreonyl derivatives of otherwise identical compounds, namely RRASr(Tp)VA, Sp(Tp)EEEEE and RRREEESe(Tp)EEEAA. Invariably, phosphoser?l peptides are by far preferred by these phospharases over their phosphothreonyl counterparts, the latter in some instances being totally unaffected. Such behaviour is opposite to that of protein phosphatases whose predilection for phosphothreonyl peptides is well established (see Table 1). It should be noted however that acidic and alkaline phosphatases, apart from this common feature, differ widely in their peptide substrate specificity. Thus E. coli alkaline phosphatase is especially active on RRASpVA, while the phosphopeptide SpEEEEE is by far preferred by potato acid phosphatase and by intestinal mucosa alkaline phosphatase. The narrowest peptide substrate specificity is observed with yeast repressible acid phosphatase displaying significant activity only toward RRASpVA. This phosphopeptide, moreover, is more or less readily dephosphorylated by all acid and alkaline phosphatases tested so far and it could represent a suitable substrate for the detection and evaluation of these enzymes by virtue of its resistance to protein phosphatases. Likewise, the phosphopeptide SpEF~EE is also totally unaffected by protein phosphatases, while being dephosphorylated efficiently by all the acid and alkaline phosphatases apart from Saceharomyc~ cerevisiae acid phosphatase. Consequently, the peptide $pEEEEE is especially suited for evaluating wheat germ and potato acid phosphatases, whose activity toward RRASpVA is rather low. On the other hand the phosphothreonyl peptide RRATpVA is an excellent substrate for both PP-2A and PP-2C, but a very poor one for all the acid and alkaline phospharases except intestinal alkaline phosphatase. As a general rule therefore efficient dephosphorylation of RRATpVA can be taken as an indication of high protein phosphatase activity mainly attributable to PP2A and/or PP-2C, whereas the contribution of acid and alkaline phosphatases at nearly neutral pH can be disclosed by determining the dephosphorylation rate of the phosphoseryl peptides RRASpVA a n d / o r SpEEEEE. Either one or the other, or both, are readily dephesphorylated by all the acid and alkaline phos-
phatases tested so far, while they are nearly unaffected by protein phosphatases. All the phosphoueptides described here can be easily obtained as 32p-derivatives by incubation of the corresponding peptides with [3,3zp]ATP in the presence of either cyclic AMP-dependent protein kinase or casein kinase-2 [4,5]. Apart from these methodological applications it is also possible that phosphopeptide substrates will prove useful for mechanistic studies on the specificity and catalytic mechanism of acid and alkaline phosphatases considering how strongly and variably their dephosphorylation efficiency is influenced by the nature of the phosphoaminoacid and the surrounding residues. It should be noted in this connection that the different dephosphorylation efficiencies outlined in Table II using a 2 /zM concentration of phosphopeptide substrates may reflect differences not only in Vm~,x but also in K m values, these latter being quite variable for the same phosphopeptide RRASpVA depending on the phosphatase. In particular, while the g m with yeast acid phosphatase (0.3/~M) is lower than the substrate concentration generally used, the K m values with the other phosphatases tested are higher (5.7, 8.3 and 33 /zM with E. coil alkaline phosphatase, yeast alkaline phosphatase and prostatic acid phosphatase, respectively). As expected therefore, at saturating concentration, RRASpVA would become an even better substrate especially for prostatic acid phosphatase, which displays the highest K m value. The superiority of RRASrVA over its phosphothrconyl derivative however is not expected to cbap.ge by varying the substrate concentration, since the phosphoseryl and phosphothreonyl peptides exhibit the same K m, albeit quite different Vm~x values. Likewise, the opposite selectivity of protein phosphatase-2A for the phosphothreonyl over phosphoseryl derivative is also due to sharply different Vm~, values [3].
Acknowledgments This work was supported by Italian M.U.R.S.T. and C.N.R. (Target Project on Biotechnology and Bioinstrumentation and grant 90.02460.CT04). The skilful technical assistance of Mr G. Tasinato is gratefully acknowledged.
References 1 Cohen, P. (1989) Annu. Rev. Biochem. 58, 453-508. 2 Waelkens, E., Goris, L and Merlevede, W. (1987) J. Biol. Chem. 262, 1049-1059. 3 Agostinis, P., Goris, J. Waelkens, E., Pinna, L.A., Marchiori, F. and Merlevede, W. (1987) J. Biol. Chem. 262, 1060-1064. 4 Agostinis, P., Goris, J., Pinna, LA., Marchiori, F., Perich, J. W., Meyer, H.E. and Merlevede, W. (1990) Eur. J. Biochem. 189, 235-241. 5 Donella-Deana, A., Mac Gowan, C.H., Cohen, P., Marchiori, F.,
133 Meyer, H.E. and Pinna, L.A. (1990) Biochim. Biophys. Acta 1051, 199-202. 6 Donella-Deana, A., Marchiori, F., Meggio, F. and Pinna, L.A. (1982) J. Biol. Chem. 257, 8565-8568. 7 Donella-Deana, A. and Pinna, L.A. (1988) Biochim. Biophys. Acta 968, 179-185. 8 Stinson, R.A. and Chan, J.R.A. (1987) Adv. Prot. Phosphatases 4, 127-151. 9 Lin, M.-F. and Clinton, G.M. (1987) Adv. Prot. Phosphatases 4, 199-228. 10 Lougovoi, C.P. and Kyriakidis, D.A. (1989) Biochim. Biophys. Acta 996, 70-75. 11 Ackerman, P. and Osheroff, N. (1989) J. BioI.Chem. 264, 1195811965. 12 Maelandsmo, G.M., Ostvold, A. and Laland, S. (1989) Eur. J. Biochem. 184, 529-534. 13 Yamashita, T., Hoshi, M., Kawakami, M. and Sakai, H. (1990) Eur. J. Biochem. 193, 661-669. 14 Litwin, C.M.E., Cheng, H. and Wang J.H. (1991) J. Biol. Chem. 266, 2557-2566.
15 Kemp, B.E. (1980) FEBS Lett. I10, 308-312. 16 Doneila-Deana, A., Lopandic, K., Barbaric, S. and Pinna, L.A. (1986) Biochem. Biophys. Res. Commun. 139, 1202-1209. 17 Biumenthal, D.K., Takio, K., Hansen, R.S. and Krebs, E.G. (1986) J. Biol. Chem. 261, 8140-8145. 18 Barbaric', S., Kozulic', B., Ries, B. and Mildner, P. (1980) Biochim. Biophys. Res. Commun. 95, 404-409. 19 Barbaric', S., Kozulic', B., Ries, B. and Mildner, P. (1984) J. Biol. Chem. 259, 878-883. 20 Onishi, H.R., Tkacz, J.S. and Lampen, J.O. (1979) J. Biol. Chem. 254, 11943-11952. 21 Chessa, G., Borin, G., Marchiori, F., Meggio. F., Brunati, A.M. and Pinna, LA. (1983) Eur. J. Biochem. 135, 609-614. 22 Meggio, F., Marchiori, F., Borin, G., Chessa, G. and Pinna, L.A. (1984) J. Biol. Chem. 259. 14576-14579. 23 Kemp, B.E., Bylund, D.B., Huang, T.S. and Krebs, E.G. (1975) Proc. Natl. Acad. Sci. USA 72, 3448-3452. 24 Pinna, L.A., Donella, A., Clari, G. and Moret, V. (1976) Biochem. Biophys. Rcs. Commun. 70, 1308-1315.
