Expression, Purification, Crystallization, and Biochemical ...

3 downloads 0 Views 4MB Size Report
Shaoqiu ZhuoS, James C. Clemens, David J. Hakes, David Barford§, and Jack E. Dixonq. From the Department of ... Millar et al., 1991). The dual specificity ...

THEJOGRNAL OF BIOLOGICAL CHEMISTRY 0 1993 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 268, No. 24, Issue ofAugust 25, pp. 17754-17761, 1993 Printed in U.S.A.

Expression, Purification, Crystallization, and Biochemical Characterization of a Recombinant Protein Phosphatase* (Received for publication, March 22, 1993)

Shaoqiu ZhuoS, James C. Clemens, David J. Hakes, David Barford§, and Jack E. Dixonq From the Department of Biological Chemistry, University of Michigan Medical School, Ann Arbor, Michigan 48109-0606 and the PWalther Cancer Institute, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York 11724-2220

A proteinphosphatase (PPase) from the bacterioA number of protein phosphoseryUphosphothreony1 phosphage A was overexpressed in Eschericha coli. The re- phatases (PPasesI2have been isolated and characterized (Balcombinant enzyme was purified to homogeneity yield- lou and Fischer, 1986; Cohen, 1989). These enzymes are clasing approximately17 mg ofenzyme from a single liter of sified as Types 1(PPase l),2A (PPase 2A), 2B (PPase 2B), and bacterial culture. Biochemical characterization of the 2C (PPase 2C). This nomenclature is based upon their sensienzyme showed thatit required Mn2+ or Ni2+ as an acti- tivity to natural occurring inhibitor 1or 2 and on their prefervator. The recombinant enzyme was active toward ser- ential specificity for dephosphorylation of the a or p subunit of ine, threonine, and tyrosine phosphoproteins and phos-phosphorylase kinase (Cohen,1978;Ingebritsen and Cohen, phopeptides. Surprisingly, the bacterial histidyl 1983). Primary sequence analysis shows that PPase 1, PPase , also dephosphorylated by the phosphoprotein, N R ~ I was 2A, and PPase 2B are structurally relatedto one another (Fig. A-PPase. The A-PPase shares a number of kinetic and 1).Interestingly, several bacteria phages also have open readstructural properties withthe eukaryotic Ser/Thr phosing frames,which encode proteins with structural similarity to phatases, suggesting that the A-PPase will serve as a good model for structure-function studies. Crystalliza- PPase 1, 2A, and 2B (Cohen et al., 1988). The type 1 and 2 tion of the recombinant purifiedA-PPase yie!ded mono- serinelthreonine protein phosphatase family is not structurally clinic crystals. The crystals diffract to 4.0 A when ex- related to the tyrosine phosphatase family, which utilizes an active site cysteine in catalyses (Fischer et al., 1991). posed to synchrotron x-ray radiation. The type 1and 2 phosphatases areinvolved in many important biological processes such as the regulation of glycogen metabolism (Ingebritsen et al., 1983; Alemany et al., 1984) and Kinases and phosphatases modulate the protein phospho- muscle contraction (Chisholmand Cohen, 1988a, 198813). It was rylation “status” of a cell, which in turn governs numerous recently reported that PPase1and PPase2A play a role in the fundamental biological phenomenon such as cell division and cell cycle in fission yeast division and deletion of these two development (Hunter, 1987; Walton and Dixon, 1993). Kinases enzymes causes mitotic defects (Kinoshita et al., 1990). PPase have been classified according to their substrate specificity, 2A has also been shown to dephosphorylate and inactivate the being either tyrosine kinase, serinehhreonine kinases,or dual p34cdc2-cyclincomplex (Lee et al., 1991). Mutation of a PPase 1 specificity kinases (which have both tyrosine as well as serine/ gene blocks mitotic progression and terminatesDrosophila dethreonine kinase activities),Although the specificity of the ki- velopment at an early stage (Axton et al.,1990; Dombradi et al., nases differ, they are all structurally related (Hunter,1987). 1990; Gatti andGoldberg, 1991). The Drosophila retinal degenPhosphatases have also been classified according to their eration C gene, which is required to prevent light-induced retsubstrate specificity. Members of the protein tyrosine phospha- inal degeneration, has also been identified t o be a PPase (Steele tase family have anactive site cysteine residue and mechanis- et al., 1992). PPase 2B (calcineurin) was recently shown to be tically proceed through a thiol-phosphate enzymeintermediate the target of the immunosuppressant drugs used t o prevent (Guan and Dixon, 1991). Dual specificity phosphatases with host rejection in organ transplantation(Liu et al., 1991). Phosactivitiestoward serinehhreonine as well as tyrosine phos- phatases are also important in ion channel function. A reconphate were initially described in vaccinia virus, but other cell stituted epithelial chloride channel is shown to be closed upon cycle proteins such as p8OCdcz5 also show dual specificity (Guan the addition of PPase 2A and reopened by addition of Mg-ATP et al., 1991; Dunphy and Kumagai, 1991; Gautier et al., 1991; and thecatalytic subunit of protein kinase A (Finn et al., 1992). Millar et al.,1991). The dual specificity phosphatases are struc- Interestingly, PPase 2Aisfound to be a target of the middle and turally related to the protein tyrosine phosphatases and also small antigens ofDNA tumorviruses(Pallas et al., 1990; use an active site cysteine residue in catalyses.’ Walter et al., 1990). Although the type 1 and 2 PPases have been studied for a number of years, thereis a limitedamount of information avail* This work was supported in part by Grant NJDDKD 18849 from the able on their structure and catalytic mechanism. In order to National Institutes of Health (to J. E. D.). The costs of publication of better understand themechanism and structure of this family this articlewere defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement”in accordance of enzymes, we have focused our attention on a PPase from the bacteriophage A. An open reading frame in the A genome (orf with 18 U.S.C. Section 1734 solely to indicate this fact. $ Recipient of a fellowship from the Scottish Rite SchizophreniaRe221) was reported to encode a bacteriophage protein phosphasearch Foundation. tase (A-PPase) (Cohen et al., 1988; Cohen and Cohen, 1989). 7 To whom correspondence should be addressed: Dept. of Biological Chemistry, University of Michigan Medical School, 5416 Medical Sci* The abbreviations used are: PPase, protein phosphatase; pNPP, p Fax: 313-763ence I, Ann Arbor, MI 48109-0606.Tel.: 313-764-8192; nitrophenyl phosphate; PCR, polymerase chain reaction; PAGE, poly4851. acrylamide gel electrophoresis; DTT, dithiothreitol. K.-L. Guan and J. E. Dixon, unpublished data.






17755 89 89 83 79 124 46 120 120


113 110 173






261 261 254 251 321 221

FIG.1.Sequence similarity between A-PPase andtype 1 and 2 PPase. Identical residues are blackened, and conserved replacements are shaded. The full A-PPase sequence is shown. All other proteinsare truncated. HPPl is human protein phosphatase1 (Barker etal., 1990).YFISPPZ is a putative type 1 protein phosphatase encoded by the fission yeast dis2(+) gene (Ohkura et al., 1989). HPP2AA is human lung PPase 2A a catalytic subunit (Stone et al., 1988). YSIT4 is a suppressor of a HIS4 transcriptional defect in yeast and isclosely related to PPase 2A(Arndt et al., 1989).HCALAl is human calcineurin Afrom human brain stem basal ganglia (Guerini and Nee, LAMBDAPP 1989). is the sequence of A-PPase (Cohen and Cohen, 1989).

The first 115 residues of this 25-kDa protein have 35% sequence identity to the N-terminalregion of PPase 1 and PPase 2A (Fig. 1).If conserved replacements are included in this comparison, the similarity between sequences increases to 49%. This suggests that the A-PPase may have some catalytic properties in common with the eukaryotic SerPThr phosphatases. Since the A-PPase has only 221 residuesand is smaller than the eukaryotic Ser/Thr phosphatases, we felt it would serve as a good model for understanding the structure andmechanism of the PPasefamily. We describe the expression, purification,biochemical characterization, and crystallization of the recombinant A-PPase.

and the mixture was heatedto 100 "C for 5 min. Five pl of the heated sample were then used in the PCR reaction. The PCR cycle conditions were 94 "C (melt) for 1 min, 55 "C (anneal)for 1 min, and72 "C (extend) for 1 min. The reaction was repeatedfor 30 cycles. The resulting PCR fragment, containing the entire coding sequence of the A-PPase, was digested with NdeI and EcoRI (sites present in the oligonucleotides) and ligated into NdeI- and EcoRI-digestedpT7-7 producing the vector pT77/LPP. The complete sequence of the A-PPase PCR product was determined. Asingle nucleotide mutation resulting inA119T was discovered. The mutation probably arose due to PCR error. This mutation was corrected by site-directed mutagenesis, and both the wild-type enzyme and the A119T mutants were expressed, purified, and examined kinetically. There were no differences in the apparent KT,,or V,,,, values for the two enzymes from pH 6.5 to 8.5, suggesting that this mutation has no effect on catalysis. The mutation falls within a non-conserved region EXPERIMENTALPROCEDURES of the protein, and the corresponding residuein thephage 680-PPase is Materials-Bio-Gel A-0.5m was purchasedfrom Bio-Rad and Phenyl- a serine (Cohen and Cohen, 1989). Expression of A-PPase-The various pT7-7LPP clones were transSepharose fkom Sigma. Phosphorylated peptides were products of the Protein and Carbohydrate Facility, University of Michigan. ~ ~ 4 3 " ~ "formed " ~ into BLZl(DE3) cells for overexpression. Single transformed colonies were grown 16 h at 37 "C in 5 ml of 2 x YT media containing was from Oncogene Science. Protein kinase A was the gift of Dr. M. ampicillin (100 pg/ml). These cultures were then used to inoculate 1 Uhler, University of Michigan. Casein and p-nitrophenyl phosphate (pNPP)were from Fluka, andall other chemicals werefrom Sigma and liter of the same media, and the cultures incubated atroom tempera1.0and (600 nm). Aldrich. NRTIand Che A were gifts from Dr. A. Ninfa and Dr. L. Ninfa, ture until the absorbance reached a value between 0.6 At this point, isopropyl-1-thio-P-o-galactopyranoside was added to the Wayne State University, Detroit,MI. Construction ofpT7-7iLPP"The A-PPase coding sequence is present cultures to a final concentrationof 0.4 mM and thecells were incubated in the "arms" of gtl0 phage libraries. Two synthetic oligonucleotides 16 h at room temperature with shaking. (5"TTTCATATGCGCTATTACG and 5"TTTGAATTCTCATGCGCCT) Purification of the Recombinant A-PPase-Isopropyl-1-thi0-p-nwere used as primers in the polymerase chain reaction (PCR, GeneAmp galactopyranoside-induced cells from a 1-liter overnight culture were PCR Reagent Kit, Perkin-Elmer Cetus) utilizing DNA from a Agtl0 harvested by centrifugation at 4200 x g for 15 min and resuspended in library as a template. One pl of the library was addedt o 19 p1 of water, 25 ml of 50 mM Tris-HC1 buffer, pH 7.5, containing 2 mM EGTA, 0.5 M


A-Protein Phosphatase

NaC1, and 20% glycerol. The cells were disrupted by passage through a Frenchpress at 1000 p. s. i.Thelysateswerethencentrifuged at 145,000 g for 60 minat 4 "C, and the supernatants were applied to a gel filtration column. Bio-Gel A-0.5m Chromatography-All procedures were carried out at 4 "C. Supernatant (25 ml) was loaded onto aBio-Gel A-0.5m column (2.5 x 100 cm), which was equilibrated with 50 mhl Tris-HCI, pH 7.5, containing 1 mM EGTA, 0.5 M NaC1, and 20% glycerol. The column was eluted with the same buffer a t a flow rate of 10-12 ml/h. Fractionsof 5.5 ml were collected and assayed for enzyme activity a s described below. Fractions having high enzyme activity and purity (assessed by SDSPAGE) were pooled, producing a total volume of approximately 60 ml. Phenyl-Sepharose Chromatography-The pooled fractions from the Bio-Gel A-0.5m column were applied directly to a Phenyl-Sepharose column (2.5 x 10 cm), which had been equilibrated in 50 mM Tris-HC1, pH 7.5, containing 1 mM EGTA and 0.5 M NaCI. The column was then washed with 200-300 ml of the samebuffer and subsequently200 ml of 20 mM Tris, pH 7.5. The enzyme was eluted with 250 ml of 50 mw Tris-HCI in 50% glycerol, pH 7.5, a t a flow rate of approximately 2 mllmin. All procedures were carried out a t 4 "C. Fractions of 6 ml were collected and assayed as noted. Active fractions were pooled after assessing their purity (examination of SDS-PAGE) and stored at -20 "C in 50 mM Tris-HCI and50% glycerol, pH 7.5. Theenzymemaintains greater than 90% of its original activity for at least 6 months under these storage conditions. Sample for x-ray crystallization was concentrated t o >10 mg/ml by using Centriprep-10 (Amicon) at 4 "C and dialyzed for 16 h against a buffer of 10 mM Tris-HCI, pH7.0,50 mM NaCI, 2 mM MnCI,, and 0.2% a-mercaptoethanol. Preparation of ~~2PICasein-[32PITyr-caseinwas prepared by phosphorylation ofcy-casein (Fluka)with ~ ~ 4 3 " "(Oncogene ~' Science). Three hundred pl of reaction mixture, pH 7.4, with 50 mM Tris-HCI, containing 2.8 mg/ml casein, 1.3 mM [y-"PlATP, 13 mM DTT, 20 mM magnesium acetate, and 15 units of p ~ 4 3 " - "were ~ ' incubated at 30 "C for 4 h. The reaction was stoppedby addition of trichloroacetic acid t o final concentrationof 20%. The phosphorylated casein was recovered a s precipitate and was washed five times with 10% trichloroacetic acid. The precipitated ["2Plcasein was then dissolved in 0.5 ml of 50 mM Tris-HC1, pH 7.8, and extensively dialyzed against the same buffer overnight. Incorporation of ["2Pl was 0.01mollmol of protein, and phosphoamino acid analysis showed that only tyrosine was labeled.[3ZP1Sercaseinwaspreparedunderthesameconditionsina 1-ml reaction volume containing 50 mM Tris-HC1, pH 7.4, 2.8 mg/ml casein, 0.4 mM [y-32P]ATP,20 mM DTT, 20 mM magnesium acetate, and 5 pg of the catalytic subunit of protein kinase A. Only serine was found t o be phosphorylated following phosphoamino acid analysis. Autophosphorylation of NR,, and Che A-Nine pg of NRll or 19.2 pg of Che Awere autophosphorylated in 10 pl of 50 mM Tris-HC1, pH 8.0, containing 0.4 m~ [y-32PlATP,60 mM KCl, and 5 mM MgC12. The reaction mixtures were incubated at 37 "C for 20 min and quenched by addition of 40 ml of 10 mM EDTA. Excess [ Y - ~ ~ P I Awas T P removed by passing the50-pl reaction mixture through a 1-ml centrifuge columnof Sephadex G-25 equilibrated with 50 mM Tris-HC1, pH 7.8. The phosphorylation yields were 12-30% for NR,,. The final concentrations of Che A and NR,, were 0.32 and 0.15 mg/ml,respectively. Assay of Phosphatase Activity-The phosphatase activity was assayed in1ml of 50 mM Tris-HCl buffer, pH 7.8, containing 20 mM pNPP and 2.0mM MnCl, with or without DTT. A-PPase (20-50 ng) were added t o start the reactions. Increase of p-nitrophenol was monitored at 410 nm on a Beckman DU-64 Spectrophotometer a t 30 "C. The extinction coefficient of p-nitrophenol were determined a t various pH levels in order t o determine the micromoles of product produced. One unit of enzyme activity was defined as 1 pmol of pNPP hydrolyzed/min. The assays for [32PlSer-casein were performed in 300 pl of a solution containing 50mM Tris-HC1, pH 7.8,0.08mM MnCI,, 5 mM DTT, 0.089 pg/ml A-PPase, and different amountsof [52PlSer-caseina t 26 "C. Forty pl of the reaction mixture was quenched with trichloroacetic acid added to a final concentration of 20% at thetimes specified. The casein wasrecovered by centrifugation and washed with 100 pl of 10% trichloroacetic acid three times. The supernatants were combined and countedfor The precipitated casein was dissolved in 200 p1 of 0.1 M NaOH and also counted for 32P. The same procedures were used t o assay for dephosphorylation of ["PlTyr-casein, except the enzyme concentration was 0.53 pg/ml. Dephosphorylation assays of phosphorylated peptides were carried out at 26 "C in 600 p1 of 50 mM Tris-HC1, pH 7.8, 2 mM MnCI, with an enzyme concentration of 1.5 pg/ml for Qr-phosphorylated peptide and 0.42 pg/ml for Thr-phosphorylated and Ser-phosphorylated peptides. The reaction was quenched by addition of 100 pl of reaction mixture t o 100 pl of 50% trichloroactic acid at the indicated times.

by the molybdate assay Release of inorganic phosphate was determined (Fiske and Subbarow, 1925). The A-PPase used for all kinetic analyses had a specific activity between 1500 and 2500 unitsimg. Typically a dephosphorylation assay of ["PINRII was carried out in 10-20 pl of 50 mM Tris-HC1, pH 7.8, containing 0.4mM MnC1, and 6 mM DTT. The reaction was started by adding A-PPase. The mixture was incubated at 37 "C for 10 min and stopped by addition of 10 ml of SDS-PAGE loading buffer (100 mhl Tris-HC1, pH 8.0, containing 8 M urea, 0.4% SDS, and 10 mM EDTA). Samples were incubated at room temperature for 30-60 min and loaded on a 15% gel (Laemmli system). The gel was run a t constant voltage of 90 V for about 2 h to avoid hydrolysis by heating. After furation in 50% methanol for 20 min and 20% methanol for 20 min, the gel was sealed in a plastic bag and x-ray film used t o detect the position of the radioactive proteins. Radioactive bands were cutfrom the gel and counted in scintillation fluid. Crystallization-A sparse matrix method (Jancarik and Kim, 1991) was used to screenfor initial crystallization conditions using the hanging drop vapor diffusion technique (McPherson, 1982). Protein at 10 mg/ml was mixed with a n equal volume of precipitant solution and applied to silanized microscope coverslips and inverted over 1 ml of precipitant in a24-well tissue culture tray (ICN). Trays were incubated at both 4 and 16 "C. X-ray Diffraction-Single crystals were mounted in thin-walled glass capillary tubes (Charles Supper Co.) between plugs of mother liquor. X-ray data were collected on an Enraf Nonius FAST area detector attached to the National Synchrotron Light Source beamlineX12C. The FAST area detector was controlled by the program, MADNES (MesserSchmidt and Pflugrath, 1987). The wavelength was 1.0 A . Protein Sequencing-Purified enzyme (0.1 mg) was precipitated by dialysis against distilled water and recovered by centrifugation. The material was then redissolved in 0.5 ml of 0.1% trifluoroacetic acid in 50% acetonitrile and subjected to N-terminal sequence analysis using a n Applied Biosystems model 4704 ProteidPeptide Sequencer (University of Michigan, Protein Sequencing Facility). Other Methods-Protein concentration were determined byCoomassie (Pierce ChemicalCo.) (Bradford, 1976) and BCA(Pierce) (Smith et al., 1985) assays using bovine serum albumin as a standard. DNA sequencesweredetermined by the dideoxy nucleotidetermination method of Sanger (Sanger et al., 19771, using T7 DNA polymerase (Sequenase, U. S. Biochemical Corp.). SDS-PAGE was performed on 6.5 x 8 x 0.15-cm vertical slabs in 15% acrylamide with Laemmli system (Laemmli, 1970). Gels were stained for protein with Coomassie Blue R-250. RESULTS

Expression and Purification of A-Protein Phosphatase (APPase)-Cohen and Cohen (1989) observed that infection of Escherichia coli with phage A g t l O resulted in the appearanceof a Mn2+-dependent proteinphosphatase activity in bacterialextracts. Using two synthetic oligonucleotides as primers and a A g t l O cDNA library as a template, the complete coding sequence of the A-PPase was obtained by PCR. The PCR fragment was inserted intoa pT7-7 vector using NdeI andEcoRI restriction sites. Transformation of the recombinant plasmid into E. coli, BL21(DE3) cells, led to overexpression of a soluble APPase. TheE. coli extracts preparedfrom the overnight culture contained a pronounced Mn2+-dependent pNPP phosphatase activity. Most of the high molecular weight impurities were removed by passing thecleared lysate througha size exclusion column. The protein phosphatase was further purified from the lower molecular weight contaminants by hydrophobic interaction chromatography using a phenyl-Sepharose column. The expression and purification procedures employed here are easily repeatable steps, which produce pure protein in approximately a 80% yield. Approximately 17 mg of the pure enzyme can be obtained from 1 liter of culture, with specific activities ranging from 2500 to 3800 unitdmg of protein. The phosphatase activity was not particularly stable unless 50% glycerol was added to the storagebuffer. The purification data from one of the preparations is shown in Table I. The purified enzyme was homogeneous, as shown by SDS-PAGE (Fig. 2). An estimated molecular weightof 25,000 was observed on SDS-PAGE. Thecalculated molecular weight of the A-PPase is 25,222.

A-Protein Phosphatase


TABLEI Purification of a protein phosphatase from bacteriophage A expressed in E. coli pNPP phosphatase activity step



Crude extract Bio-Gel A-0.5 m Phenyl-Sepharose

kD 200 97.4 68.0 43.0 29.0

437 82 17






units x I O 5




77.3 76.3 61.4

177 932 3870

1 5 22

100 95.2





Activation and inhibition of A-PPase by metal ions Activities were assayed in 1ml of 50 mM Tris-HC1, pH 7.8,containing 20 mMpNPP and metal ions as indicated a t 25 "C. Purified A-PPase (47 ng) was added in each assay to start the reaction. Activation ions

14.3 6.2 -


FIG.2. SDS-polyacrylamide gel electrophoresisof purified APPase. Samples were denatured and reduced in loading buffer containing 200 m~ ?tis-HC1, pH 6.8, 0.48 SDS, and 10 mM dithiothreitol. The proteins were separated by 15% SDS-PAGE and stained with Coomassie Blue. Lane l , crude cell extract; lane 2, purified A-PPase.

Amino acid sequencing of the first 21 residuesof the purified enzyme resultedinthe observed sequence, MRYYEKIDGSKYRNIWWGDL, corresponding to the expected sequence of the A-PPase. Activation and Inactivation of A-PPase by Metal Zons-The activity of A-PPase toward pNPP, phosphoproteins, and phosphopeptides was shown to be metal ion-dependent. The effects of metal ions on pNPP phosphataseactivity are shown in Table 11. Of all the metal ions examined, only Mn2+ and Ni2+ were capable of activating theenzyme. Alkali metal ions, Mg2' and Ca2+, were neither activators nor inhibitors of the Mn2+ activated enzyme activity. At a concentration of 0.2 mM, V3+, C f + , Fez+,Co2+,Pd2+,and Sn2+ salts inhibited Mn2+-pNPP the phosphatase activity to varying degrees (838%).The saltsof Sc3+, Y b 3 + , Cu2+, Zn2+, and Hg2' were potent inhibitors of the enzyme, generally resulting in >90% inhibition of activity. Inhibition by Zn2+was shown to competitive with Mn2+ (data not shown). Nonmetal Zon Inhibitors of X-PPase--Almost all common oxyanions were found to inhibit the Mn2+-pNPP phosphatase activity as shown in Table 111. Among them, vanadate was the most potent inhibitor having an IC50 of 0.72 PM. Tungstate inhibited the enzyme with a n IC50 of 58 p,while phosphate showed a value of 0.71 mM. Molybdate and arsenate were less potent, and their [email protected] were observed at concentration where precipitation with Mn2+ was observed. Sulfate and carbonate were also poor inhibitors. As was the case with type 1 and 2 protein phosphatases (Ballou and Fischer, 19861, fluoride inhibited the A-PPase. Inhibition of A-PPaseby phosphate monoesters was also tested (Table IV). Phosphoenolpyruvate and adenosine 5'monophosphate (AMP) wererelatively good inhibitors with IC50 values of 2.7 and 6.7 mM, respectively. Surprisingly, phos-





0.2 m~ Mg2' 0.2 mM Ca2* 0.2 mM Sc"+ 0.2 mM V"+ mMYb3* 0.2 0.2 m~ Cr2+ 0.2 mM Mn2* 0.2 mM Fez+ 0.2 m M COS' 0.2 mM Ni2+ 0.2 mM cu2* 0.2 mM Zn2* 0.2 m~ Pd2* 0.2 mM Hg2+ 0.2 mM Sn2*

0 80 8 17 0 0 1384 0 37 1148 0.2 0 0 20 0 0




0.2 m~ Mn2++ 0.2 mM Mg2' 0.2 mM Mn2* + 0.2 m~ Ca2+ 0.2 mM Mn2* + 0.2 mM Scz+ 0.2 mM Mn2* + 0.2 mM 0.2 mM Mn2* + 0.2 mM Yb3* 0.2 mM Mn2* + 0.2 Cm 1.2M ' 0.4 mM Mn2+ 0.2 m~ Mn2* + 0.2 m M Fe2* 0.2 mM Mn2++ 0.2 mM Co2+ 62 m~ Mn2++ 0.2 mM NiZ* 0.2 mM Mn2* + 0.2 m~ Cu2* 0.2 mM Mn2* + 0.2 mM Zn2* 0.2 m~ MnZ*+ 0.2 mM Pd2+ 0.2 m . Mn2* + 0.2 mM Hg2+ 0.2 m~ Mn2* + 0.2 mM Sn2* \




103 101 6 77 11 81 100 92 105 2 0 92 0 59

TABLE111 Inhibition of the Mn2+-stimulated pNPPphosphatase activity by oxyanions The enzyme activity was assayed in 50 m~ Tris-HC1, pH 7.8, containing 20 mMpNF'P and 2 m~ Mn2*at 30 "C. Inhibitors were added to the assay buffer a t various concentrations. ICs0



Vanadate Tungstate Phosphate Molybdate Arsenate Fluoride Sulfate Carbonate


X 10-4 0.058 0.71 2.0 >1.9 3.9 18 >16

phoserine, phosphothreonine, and phosphotyrosine were not good inhibitors. Many inhibitors and stimulators of phosphoseryllphosphothreonyl phosphatases, such as histone H1, protamine, spermine, polylysines, and trifluperazine (Ballou and Fischer, 1986), did not have any effects on A-PPase activity. Specific and potent inhibitors of type 1 and type 2 protein phosphatases, okadaic acid (124 nM) and microcystin-LR (5 p), were also found to have no effects on the enzyme. The enzyme was not inhibited by the acid phosphatase inhibitor, L-(+)-tartrate, at concentration of 5 mM. of A-PPase-The physiological subCatalyticProperties strates for the A-PPase are unknown. The enzyme showed a broad range of activities toward phosphoproteins and phosphopeptides. The kinetic constants of some substrates tested are presented in Table V. As noted in Table V, the enzyme can dephosphorylate phosphoseryllphosphothreonyl- as well as phosphotyrosyl-containing substrates. The phosphotyrosyl peptide sequence corresponds to the substrate of the insulin receptor kinase (Pike andKrebs, 1986). The phosphoseryl and phosphothreonylcontain the recognition sites for kinase C


A-Protein Phosphatase TABLEIV

Inhibition of the Mn"+-stimulated pNPP phosphatase activity of A-PPase by phosphate monoesters The enzyme activity was assayed in 50 mM Tris-HC1, pH 7.8, containing 20 mM pNPP and2 mM Mn2+at 30 "C. Inhibitors were added to the assay buffer a t various concentrations. Compounds



Phosphoenolpyruvate Adenosine 5'-monophosphate ~(-)-3-Phosphoglyceric acid 5-phosphate Pyridoxamine o-Phosphoethanolamine Phosphoserine Phosphorylcholine Phosphotyrosine Phosphothreonine

2.7 6.7 26 52 63 90

110 No inhibition a t 20 mM No inhibition a t 25 mM

ing the ability of the A-PPase to dephosphorylate the bacterial phosphohistidine-containing protein, Che A. Although Che A could be dephosphated, the apparent rate of dephosphorylation was significantly slower than the dephosphorylation of NRII (Fig. 4). We also examined the metal dependence of the histidine dephosphorylation. Mn2+and Ni2+were the only two metal ions that were capable of activating the enzyme. No phosphatase enzymeactivitywas observed using [32P]NRII as substrate when Cu2+, Zn2+,Co2+, and Mg2+were added to the reaction mixture. These results parallel those seen in Table 11 when pNPP was used as a substrate, suggesting that the hydrolysis of both substrates, pNPP and 132P1NR~~, has a similar metal dependence. The pH rate profile for dephosphorylation of N R I ~ was also examined (Fig. 5 ) . The optimum rate of hydrolysis is seen between pH 7.0 and 7.8, which is approximately the same pH optimum noted for pNPP. The dephosphorylation of Che A was also much slower than the dephosphorylation of NRII by A-PPase at all pH values examined (Fig. 5 ) . Crystallization of A-PPase-The sparse matrixscreening condition containing 0.1 M sodium citrate, 0.2 M ammonium acetate, 30% polyethylene glycol 4000, pH 6.3, at 16 "C yielded microcrystals within 2 days. Refinement of pH, salt, and precipitant concentrations indicated optimal conditions of 0.1 M sodium citrate, pH 6.5, 24% polyethylene glycol 4000, 2 mM MnC12, and 0.3% p-mercaptoethanol. Under these conditions, crystals appear within 2 days and grow to a maximum size of 0.6 mm by 0.2 mm by 0.07 mm within a month. Crystals grow as either clusters or single crystals exhibiting a monoclinic morphology (Fig. 6). X - r u ~Diffraction of A-PPase Crystals-The crystals diffract to 4.0 A when exposed to synchrotron x-ray radiation. The unit cells dimensions were determined using theAUTOINDEX routine of MADNES and confirmed by analysis of the diffraction images. The cell dimensions are a = 80.8 A, b = 193.2 A, c = 53.9 A, a = 90.00, p = 93.30, y = 90.0". Systematic absences indicatethatthe crystal belongs to space group P21, with each asymmetric unit containing between 4and 8 molecules assuming a solvent contentof between 71 and42%, respectively, using themethod of Matthews (1968).

(Heasley and Johnson,1989; Pike andKrebs, 1986).All of these peptides were prepared by chemical synthesis, incorporating phosphateintothe peptide in stoichiometric amounts.The ~ ~the ' protein kinase A phosphorylation of casein by p ~ 4 3 " -or to approxicatalyticsubunit incorporated[32P]phosphate mately 1%.Phosphoamino acid analysis of 32P-labeled casein demonstrated that only seryl or tyrosyl phosphorylation occurred via protein kinase A or the v-abl kinase, respectively. Nevertheless, both of these proteins are good substrates for the enzyme. The A-PPase seems to havea preference for "protein" substrates and iscapable of hydrolyzing both phosphotyrosine as well as phosphoserine-containing protein substrates. Although pNPP was dephosphorylated with avery high turnover rate, no hydrolysis could be detected for phosphoserine, phosphothreonine, and phosphotyrosine under similar conditions. Because no naturally occurring substrates for the A-PPase have been described, we also examined the ability of the enzyme t o hydrolyze ADP and ATP. Neither compounds were substrates (data not shown). Characterization of Histidine PhosphataseActivity of the APPase-Several bacterial proteins important gene in regulation and chemotaxisundergoautophosphorylation (Stock etal., 1989). NRII and Che A are two representative proteins in this family, and they undergo autophosphorylation of a histidine DISCUSSION residue (Ninfa and Bennett, 1991). Fig. 3 demonstrates the dephosphorylation of bacterial [32PlNR11 protein by AAlthough a number of PPases have been isolated and charPPase. The dephosphorylation of [32PlNRIIby increasing con- acterized, there are limited amounts of data available regardcentrations of the recombinant enzymeis shown in Fig. 3A.The ing their catalyticproperties. Few studies havebeen directeda t enzyme catalyzed dephosphorylation of[32P1NRII was also understanding their structure and function. The protein phosmonitored as a function of the time of incubation (Fig. 3B). phatase from the bacteriophage A contains only 221 amino acid Under the assay conditions used, NRII also undergoes a non- residues, and for this reason it would appear tobe a good model enzyme-catalyzed dephosphorylation. The rateof this reaction for structural studies of the PPases. Large amounts of the is also shown in Fig. 3B. The activity of the A-PPase catalyzed enzyme were obtained by the overexpression of the A-PPase reaction was estimated to be 25 pmol/midmg at a substrate protein in E.coli and thedevelopment and implementation of a concentration of 180 nM. Due to the limiting amounts of L3'PI simple and direct method of purification. The large quantities NRII, the enzyme concentrationin thereaction mixture exceeds of pure enzyme have made crystallization efforts possible. the substrateconcentration making itdifficult to establish that We were particularly interested in addressing thequestion: the A-PPase is indeed catalytic. Several observations,however, "is the A-PPase likely to be a good biochemical model for the suggest that this is likely an enzyme-catalyzed reaction. The type 1and 2 PPases?" Biochemical studies reported here show dephosphorylation of [32PlNRlrwas dependent on the storage that theA-PPase shares a number of properties with type 1and conditions of the phosphohistidine-containingprotein. The 2 PPases. The A-PPase requires Mn2+or Ni2+ for phosphatase highesthistidinephosphatase activity was observed when activity, whereas Zn2+ and Cu2+ are inhibitory. These results fresh phosphorylated sample was used. Decreased dephospho- mimic the metal ion dependence of calcineurin (Pallen and rylation could be detected after the phosphorylated [32PlNR~~Wang, 1984). Indeed, all type 1 and 2 PPases can be activated was frozen at -20 "C.Autophosphorylated NRll is known to be by Mn2+and inhibitedby Zn2+(Ballou and Fischer, 1986). Mg2+ denatured when frozen a t -20 O C 3 This suggests that the has been only reported to activate type 2C PPase (Binstock and structure of the substrate is importantfor the dephosphoryla- Li, 1979; Hiraga et ai., 1981; Mieskes et al., 1984; Pato and tion. Support for this suggestion was also obtainedby examin- Adelstein, 1983). Unlike the other PPases, which can also use Co2+ as an activator, A-PPase is inhibited by Co2+. NaF is an non-competitive inhibitor of phosphoseryVphosphothreony1 A. Ninfa, personal communication.


A-Protein Phosphatase

TABLE V Summary of the kinetic constants for the dephosphorylation of various substrates by A-PPase K , and k,,, values were determinedby Eadie-Hofstee plots (Eadie,1942; Hofstee, 1959).Assay conditions were described under “Experimental Procedures.” Substrates

KR(+~)IRR-OH KRP(SD)QRHGSKY-amide 0.23 [“P1Sei-cLein [32P1~r-casein pNPP Phosphoserine Phosphothreonine Phosphotyrosine





k c d L


4.3 260 1200 410 1000 68 0.13 0.32 0.067 7620 930 No detectable hydrolysis No detectable hydrolysis No detectable hydrolysis

Acetvl-RRLIEDAE (Yu)AARG-amide





lambda PPase (pg)




0.017 0.36 0.066 1.8

0.21 0.12




(minutes) Time

FIG.3. Phosphatase activity of A-PPase toward [32P]NEtI,. A , ”P-labeled NRII (10pg/ml) was incubatedat 37 “C for 10 min with different amounts of A-PPase in 20 ~1 of 50 mM ”is-HC1 containing 6 mM DTT and 0.4 mM MnCI2.The reactions were quenched by addition of equal volume of SDS-PAGE loading buffer, and the mixtures were subjected to electrophoresis and quantitated as described under “Experimental Procedures.” B,110 pl of reaction mixture containing50 mM Tris-HC1, pH 7.8,0.4m~ MnCI,, 6 mM DTT, 180 nM [32P1NRII,and 0.05 mg/ml(1.8p ~A-PPase ) were incubated at 37 “C.At the time indicated, an aliquot of 10 pl was quenched with equal volume of gel loading buffer and subjectedt o electrophoresis as described under “Experimental Procedures.” Solid line indicates hydrolysis in the presence of A-PPase. Dashed line is autohydrolysis of I”ZP]NR1,.

PPases(Ingebritsenand Cohen, 1983; Shacter-Noiman and Chock, 1983),and italso inhibits theA-PPase. Compounds such as okadaic acid are effective inhibitors of the type 1 and type 2 PPases, but theydo not affect the activity of A-PPase. This may suggest that theinhibition by okadaic acid may require amino acid residues found outside of the region of sequence similarity noted between the A-PPase and other PPases shown in Fig. 1 (or require residues that are not conserved in the A-PPase). Collectively, there are a number of similarities, as well as several differences, between the metal ion requirements and inhibitors that affect the A-PPase and the mammalian protein phosphatases. One must therefore be cautious in concluding that theA-PPase will be a good model for all aspects of catalysis and inhibition properties of the mammalian A-PPases. The A-PPase has a number of catalytic properties that are particularly interesting. The enzyme can hydrolyze phosphoseryllphosphothreonyl as well as phosphotyrosyl substrates. A similar dual substrate specificity has been demonstrated for a phosphoseryUphosphothreony1 PPase from bovine cardiac muscle (Chernoff et al., 1983).The k,,,/K, values noted in Table V clearly show that the enzyme prefers protein as

opposed t o peptide substrates. This suggests that other residues in addition to the phosphorylated amino acid are important for binding of enzymes and substrates. The A-PPase was shown to dephosphorylate the phosphohistidyl-containing proteins, NRll and CheA, The ratesof dephosphorylation of Che A and NRI, by the A-PPase differed dramatically. Both NRII and Che A are protein kinases that catalyze transfer of y-phosphoryl group from ATP t o the N-3 position of one of their own histidines (Weiss and Magasanik, 1988; Ninfa and Bennett.1991; Stock et al., 1988; Hess et al., 1988).Among the family of related kinases,sequences surrounding the autophosphorylation sites are conserved (Stock et al., 1989). The major site of autophosphorylation of NRII is at His-139.Amino acids surrounding the siteof phosphorylation are: GLApHEIK (Ninfa and Bennett, 1991).The autophosphorylation site of Che A is at His-48 (near the N terminus of the protein; Hess et al. (1988)).The sequence surrounding thephosphorylation site in Che A is RAApHSIK. We do not know if the differences in dephosphorylation rates of Che Aand NRII are associated with differences in primaryor three-dimensional structure. The fact that denaturationof NRII leads to a loss in dephosphorylation

A-Protein Phosphatase



rate suggests that three dimensional structure is likely to be important. The dephosphorylation of phosphohistidyl N R I I by A-PPase shows the same metal ions and optimal pHas noted for pNPP. This suggests that the mechanismsof the catalyses of NRII the two substrates are likely to be similar. Prokaryotic histidine protein kinases play very important roles inbacterial signal transduction processes. These enzymes effect rapid transient change in motility as well as long term global reorganizations of gene expression and cell morphology Che A (Stock et al., 1989; Hess et al., 1988). More than 20 such signal transduction systems havebeen identified, and it is estimated that there mightbe as many as 50 regulatory systems present FIG.4. Dephosphorylation of [82PlNFt~r and P2P1Che A by A- in E. coli (Stock et al., 1989,1990; Bourretet al., 1991). We have PPase. 20 pl of 32P-labeledNRIl (25 pg/ml) and cheA (43 pglml) were not shown that N R I 1 is the substratefor A-PPase in vivo.Howincubated separately a t 37 "C for 10 min with 0.5 pg of A-PPase in 50 ever, the in vitro observation that demonstrates dephosphomM Ms-HCI, pH 7.8, containing 6 m w DTT and other additives as rylation of NRIl by A-PPase raises the interesting possibility indicated. The reactions were quenched and analyzed a s described in Fig. 4. Lune A, in 0.4 m~ MnClz and A-PPase was boiled for 10 min that thebacteriophage may use this enzyme to control aspects before addition; lane B , in 6 m~ EDTA, lane C, in 0.4 m~ MnC12;lane D, of host signal transductionmechanisms. in 0.3 mM NiCI2.



PH 7.4





+ - + - +





Acknowledgment-We are grateful to Elizabeth and Alex Ninfa for their gifts of Che A and NRII.We acknowledge Randy Stone for commenting on the manuscript. REFERENCES Alemany, S.,Tung, H. Y. L., Shenolikar, S.. Pilkis, S. J.. and Cohen. P. (1984) E u s J. Biochem. 145.51-56 Amdt. K. T., Styles, C. A,, and Fink, G. R. (1989) Cell 56,527-537 Axton. J. M.. Dombradi, V., Cohen. P. T..and Glover, D. M. (1990) Cell 63,33-46 Ballou, L. M., and Fischer. E. H. (1986) The Enzymes (Boyer, P. D., ed) Vol. XVII. pp. 311-361, Academic Press, Orlando, FL Barker, H. M.. Jones, T. A,, da Cruz e Silva, E. F.. Spurr, N. Sheer, D.. and Cohen, P. T.W. (1990) Genomics 7, 159-166 Binstock. J. F., and Li. H.-C. (1979) Biochem. Biophys. Res. Commun. 87, 1226-


Che A


Bourret, R. B.,Borkovich, K. A., and Simon, M. I. (1991)Annu. Reu. Biochem. 60, 401-441

FIG.5. Dephosphorylation of [82PlNRII by A-PPase at different pH. 20 pl of 32P-labeled NRII (25 pglml) and Che A (43 pg/ml) were incubated at 25 "C for 20 min with0.5 pg of A-PPase in 50 mM Ms-HCI a t different pH containing6 m~ D l T and 0.4 m MnCIz. At pH 5.8, 50 m~ sodium acetate-Ms was used instead of 50 mM Ms-HCI.

Bradford, M.M. (1976)AnaL Biochem. 72,248-254 ChemoK, J., Li. H.-C., Cheng, Y.-S. E., and Chen, L.-B. (1983)J. Biol. Chem. 258, 7852-7857


Chisholm, A. A. and Cohen, P. (1988a) Biochim. Biophys. Acta 988,392400 Chisholm, A. A. K.. and Cohen, P. (1988b) Biochim. Biophys. Acta 971, 163-169 Cohen. P. (1978) Curs Top. Cell. Regul. 14, 117-196 Cohen. P. (1989)Annu. Reu. Biochem. 58,453-508 Cohen. P. T.W.. and Cohen. P. (1989) Biochem. J. 260,931-934 Cohen. P. T.W.. Collins, J. F., Coulson, A. F. W., Bemdt. N.. and daCruz e Silva,0. B. (1988) Gene (Amst.) 69, 131-134 Dombradi, V.. Axton. J. M., Barker, H. M.. and Cohen,P. T.(1990)FEES Lett. 275, 39-43

Dunphy, W. G., and Kumagai. A. (1991) Cell 67,189-196 Eadie. G. S. (1942) J. Biol. Chem. 146,85 Finn, A. L., Gaido, M. L.. Dillard, M.. and Brantigan, D. L. (1992)Am. J. Physiol. 263, C1724175

Fischer, E. H., Charbonneau, H.. and Tonks, N. K.(1991) Science 253,401-406 Fiske, C. H., and Subbarow, Y. (1925) J. Biol. Chem. 66.375-400 Gatti. M., and Goldberg. M. L. (1991) Methods Cell Biol. 35,543-586 Gautier, J.,Solomon, M. J., Booher, R. N. Bazan, J. E, andKirschner. M. W. (1991) Cell 67.197-211 Guan, K.-L.. and Dixon, J. E.(1991)J. Biol. Chem. 266,17026-17030 Guan, K.-L.. Broyles. S. S., and Dixon, J. E. (1991) Nature 350,359-362 Guerini, D., and Klee. C. B. (1989)Proc. Natl. Acad. Sci. U.S . A. 86,9183-9187 Heasley, L. E.,and Johnson, G . L. (1989) J. Biol. Chem. 264,86468652 Hess. J. F., Bourret. R. B., and Simon, M. I. (1988) Nature 336, 139-143 Hiraga, A,. Kikuchi. K., Tamura. S.. and Tsuiki, S. (1981) Eus J . Biochem. 119, 503-510

Hofstee, B. H. J. (1959) Nature 184, 1296 Hunter, T.(1987) Cell 50,823-829 Ingebritsen, T.S.,and Cohen, P.(1983) Eur. J. Biochem. 132,255-261 Ingebritsen, T.S..Stewart, A. A.,and Cohen. P. (1983) Eus J. Biochem. 132,297307

Jancarik, J., and Kim. S. H. (1991) J. Appl. Crystallogs 24.409-411 Kinoshita, N., Ohkura. H., and Yanagida, M. (1990) Cell 63,405-415 Laemmli, U. K. (1970)Nature 227,680-685 Lee. T.H., Solomon. M. J., Mumby. M.C.. and Kinchner. M. W. (1991) Cell 64, 415-123

FIG.6. Crystal of A-PPase. Conditions for crystal growth are described under "Results." Overall sizeis 0.6 x 0.2 x 0.07 mm.

Liu. J.. Farmer, J. D.. Jr.. Lane, W. S., Friedman, J.. Weissman, I., and Schreiber. S. L. (1991) Cell 66,807-815 Matthews, B. (1968) J . Mol. Biol. 33,491-497 Messerschmidt, A., and Pflugrath, J. W. (1987) J. Appl. Crystallogs 20,306-315 McPherson, A. (1982) Preparation and Analysis ofProtein Crystals. John Wiley & Sons, New York Mieskes, G.. Brand, I. A,, and Soling, H.-D. (1984) Eur. J. Biochem. 140.375-383 Millar, J. B.A.. McCowan. C. H., Lenaers, G., Jones, R., and Russell. P. (1991) EMBO J. 10,4302-4309 Ninfa. A. J.. and Bennett, R. L. (1991) J. Biol. Chem. 266,6888-6893

A-Protein Phosphatase Ohkura, H., Kmoshita, N., Miyatani, S., Toda, T., and Yanagida,M. (1989) Cell 57, 997-1007 Pallas, D. C., Shahrik, L. K., Martin, B. L., Jaspers, S., Miller, T. B., Brautigan, D. L.,and Roberts, T. M. (1990) Cell 60, 167-176 Pallen, C. J., and Wang, J. H. (19841 J. B i d . Chem. 259,6134-6141 Pato, M. D., and Adelstein, R. S.(1983) J . B i d . Chem. 258, 7055-7058 Pike, L. J., Eakes, A. T., and Krebs, E. G. (1986) J. Biol. Chern. 261,3782-3789 Sanger, F., Nicklen, S., and Coulson, A. R. (1977)Proc. Nutl. Acud. Sci. U. S. A. 74, 5463-5467 Shacter-Noiman, E., and Chock, P. B. (1983) J. Biol. Chem. 258, 4 2 1 4 4 2 1 9 Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H., Provenzano, M. D., Fujimoto,E. K., Goeke, N. M., Olson, B. M., and Klenk,D.C. (1985) Anal. Biochem. 150, 76-85


Steele, F. R., Washhurn, T., Rieger, R., and OTousa, J. E. (1992) Cell 69, 669-676 Stock, A. M., Chen, T., Welsh, D., and Stock, J . B. (1988) Proc. Nutl. Acud. Sci. U. S. A. 85, 1403-1407 Stock, J. B., Ninfa, A. J., and Stock, A. M. (1989) Microbiol. Reu. 53, 4 5 0 4 9 0 Stock, J. B., Stock, A. M., and Mottonen J. M. (1990) Nature 344, 3 9 5 4 0 0 Stone, S.R., Mayer, R., Wernet, W., Maurer, F., Hofsteenge, J., andHemmings, B. A. (1988) Nucleic Acids Res. 16, 11365 Walter, G., Ruediger, R., Slaughter, C., and Mumhy, M. (1990)Proc. Nutl. Acud. Sci. U. S. A. 87, 2521-2525 Walton, K., and Dixon, J. E. (1993) Annu. Reu. Biochem., in press Weiss, V., andMagasanik, B. (1988) Proc. Nutl. Acad. Sci. U. S. A. 85, 8919-8923

Suggest Documents