(Received for publication, June 20, 1996, and in revised form, October 4, 1996)
From the Departments of Internal Medicine and
Molecular Biophysics, Division of Hematology, Washington University
School of Medicine, St. Louis, Missouri 63110 and
§ Department of Internal Medicine, University of
Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104
Pleckstrin is the major substrate phosphorylated on serine and threonine in response to stimulation of human platelets by thrombin (Abrams, C. S., Zhao, W., Belmonte, E., and Brass, L. F. (1995) J. Biol. Chem. 270, 23317-23321). We now show that pleckstrin in platelets is in a complex with inositol polyphosphate 5-phosphatase I (5-phosphatase I). This enzyme hydrolyzes the 5-phosphate from inositol 1,4,5-trisphosphate and inositol 1,3,4,5-tetrakisphosphate and thus serves as a calcium signal-terminating enzyme, since the substrates but not the products mobilize intracellular calcium. Pleckstrin co-immunoprecipitates with 5-phosphatase I in homogenates of platelets. Platelet homogenates fractionated by anion exchange chromatography show co-elution of pleckstrin and 5-phosphatase I. Fractions containing phosphorylated pleckstrin have 7-fold greater 5-phosphatase activity than those containing unphosphorylated pleckstrin. Mixing experiments with recombinant 5-phosphatase I and pleckstrin in vitro show that they form a stoichiometric complex. A mutant form of pleckstrin, in which the serine and threonine residues that are phosphorylated by protein kinase C are substituted with glutamic acid (pseudophosphorylated pleckstrin), activates recombinant 5-phosphatase I 2-3-fold while native unphosphorylated pleckstrin does not stimulate the enzyme. Thus pleckstrin functions to terminate calcium signaling in platelets when it is phosphorylated by binding to and activating 5-phosphatase I.
Inositol polyphosphate 5-phosphatases comprise a large family of enzymes that share two short amino acid motifs that define the family (2). They show varying substrate specificities toward the 5-phosphate containing inositol phosphates and phospholipids. One of these enzymes, inositol polyphosphate 5-phosphatase I (5-phosphatase I),1 that was originally identified in platelets (3) hydrolyzes only the inositol phosphate substrates, inositol 1,4,5-trisphosphate (Ins(1,4,5)P3) and inositol 1,3,4,5-tetrakisphosphate (Ins(1,3,4,5)P4) (4). This enzyme is most likely to function as a signal-terminating enzyme since its substrates function to mobilize calcium ions while its products do not. A previous study suggested that 5-phosphatase I was phosphorylated in response to thrombin and thereby activated. 5-Phosphatase I purified from platelets was phosphorylated in vitro by protein kinase C and activated 4-fold under this condition (5). This would provide one mechanism for terminating Ins(1,4,5)P3-stimulated calcium mobilization. Subsequent to this study, cDNA clones encoding 5-phosphatase I were isolated, and the recombinant protein was expressed in heterologous systems (6-8). The specific activity of the recombinant enzyme is about 50 times greater than that purified from platelets suggesting that the platelet enzyme was not homogeneous. We now report that 5-phosphatase I exists as a complex with another protein, pleckstrin, in platelets. When pleckstrin is phosphorylated in the complex, 5-phosphatase I is activated. In retrospect the previous results must have been obtained from a preparation that included both 5-phosphatase I and pleckstrin.
Pleckstrin is a 40-kDa protein of human platelets that was originally
identified as the major substrate for phosphorylation in platelets upon
stimulation by thrombin (9, 10). The protein was found to contain 100 amino acid repeats at either end that are approximately 30% identical.
Since this discovery, similar motifs have been found in over 70 proteins, including a number of proteins involved in intracellular
signaling reactions (11-17). These proteins include phospholipase C
enzymes, small guanine nucleotide binding proteins, -adrenergic
receptor kinases,
-spectrin, and dynamin. The function of pleckstrin
homology (PH) motifs is uncertain although elucidation of their
three-dimensional structures indicates that they share a fold and
therefore constitute a common motif despite limited amino acid
similarity. Several of these motifs have been shown to bind to acidic
phospholipids, especially phosphatidylinositol 4,5-bisphosphate (13,
15, 17). PH motifs have also been shown to bind to the
subunits
of trimeric G proteins and thereby presumably inhibit further signaling
from an agonist-activated receptor (12-14). Platelet pleckstrin has been transfected into COS-1 cells and shown to diminish agonist-induced phosphatidylinositol turnover (1, 18). This effect was observed in
response to both G protein-linked and tyrosine kinase-linked receptors
and thus cannot be explained solely by binding to
subunits. We
now report a function for platelet pleckstrin. The protein forms a 1:1
complex with 5-phosphatase I, and upon phosphorylation of the former,
the latter is activated.
Materials
[32P]Orthophosphate was from ICN and 32P-labeled Ins(1,4,5)P3 was prepared as described (19). Horseradish peroxidase-linked anti-rabbit IgG and Western blot detection reagents were purchased from Amersham Life Sciences. The PVL 1392 baculoviral transfer vector and BaculoGold transfection kit were from PharMingen. pBluescriptSK and competent Escherichia coli were obtained from Stratagene. Anti-pleckstrin antibody was made as described (18). Mono Q column and pGEX-5x-1 were from Pharmacia Biotech Inc. T4 DNA ligase, monoclonal HA antibody, and restriction enzymes were purchased from Boehringer Mannheim. Polymerase chain reaction reagents and Taq polymerase were from Perkin-Elmer. All other chemicals were obtained from Sigma.
Methods
Construction of Different Forms of Recombinant PleckstrincDNA molecules encoding pleckstrin and variant forms of pleckstrin were constructed as described previously (1, 18). These cDNA were then subcloned into a derivative of the plasmid vector PET-IIb that had been modified to contain sequences encoding six histidine residues at the amino terminus of the expressed protein. Proteins were expressed in E. coli as described previously (20) and purified using nickel-agarose (Qiagen). Recombinant proteins were eluted from nickel-agarose using 20% glycerol, 20 mM Tris, pH 7.9, 100 mM potassium chloride, 5 mM DTT, 0.5 mM PMSF, and 80 mM imidazole.
Expression of Recombinant 5-Phosphatase I in Sf9 CellscDNA encoding a truncated version of 5-phosphatase I in
pCMV2 plasmid was provided by C. A. Mitchell (6). A full-length cDNA was obtained by polymerase chain reaction using a human
umbilical vein endothelial cell cDNA library as template. The sense
primer included an EcoRI site followed by nucleotide 3 to
nucleotide 16 of human 5-phosphatase I cDNA
(5
-ttaagaattcaccatggcggggaaggcgg-3
); the antisense primer contained
the complement of nucleotides 107-128 (5
-gtgtgcacgacctggtaaaat-3
).
The polymerase chain reaction products were cleaved with
EcoRI and PstI. The digested products containing nucleotides
3 to 88 were subcloned into pBluescriptSK. Nucleotides 89-1784, obtained by PstI digestion of the truncated
5-phosphatase I in pCMV2 plasmid, were subcloned into pBluescriptSK.
The full-length cDNA was subsequently engineered into the
baculovirus transfer vector pVL 1392 between the EcoRI and
XbaI restriction sites. Sf9 cells grown at 27 °C in
TNM-FH insect medium (PharMingen) supplemented with 10%
heat-inactivated fetal bovine serum and 100 µg of gentamicin/ml were
infected with the baculovirus construct containing 5-phosphatase I. Sf9
cells were cultured either in monolayer or in suspension (2 × 106 cells/ml at the start of infection), and recombinant
protein was harvested at 72 h after infection.
Full-length 5-phosphatase I cDNA was isolated from
pVL 1392 by EcoRI and NaeI restriction and
transferred into pGEX-5x-1 plasmid between EcoRI and
SmaI restriction sites. The plasmid was used to transform
XL2 Blue E. coli (Stratagene). E. coli were grown in LB medium containing 20 µg of ampicillin/ml. After overnight culture, isopropyl-1-thio--D-galactopyranoside (1 mM) was added, and 4 h later E. coli were
harvested by centrifugation and the cells were resuspended in 20 mM Tris, pH 7.5, 5 mM EDTA, 1 mM sodium azide, 1 µM pepstatin A, and 50 µM
PMSF. The suspension was frozen at
70 °C and thawed slowly on ice;
lysozyme was added to the final concentration of 1 mg/ml, and the
mixture was incubated at 4 °C for 30 min. DNase I (0.25 µg/ml
original bacterial suspension) and MgCl2 (final
concentration, 10 mM) were added and incubated for 1 h
on ice. The final suspension was centrifuged at 12,000 × g, and the supernatant was passed through GST-agarose
columns equilibrated with 20 mM Tris, pH 7.5, 1 mM EDTA, and 1 mM DTT. After washing with this
buffer, the bound protein was eluted using 20 mM reduced
GSH in 20 mM Tris, pH 7.5.
32P-Labeled Ins(1,4,5)P3 hydrolysis was measured as described previously (3). Antisera against a peptide that represents amino acids 49-64 of 5-phosphatase I (MALHCQEFGGKNYEAC) were produced in rabbits (Pocono Rabbit Farm, Cacadensis, PA). For immunoblot analysis, proteins separated by 10% SDS-PAGE were transferred to nitrocellulose membranes and incubated with the antipeptide antiserum at 1:1000 dilution overnight at 4 °C. Detection was performed by enhanced chemiluminescence (ECL, Amersham).
Preparation of 32P-Labeled Platelets, Thrombin-stimulated Platelets, and Platelet ExtractsHuman platelets were obtained from normal donors. Platelet isolation, labeling, and stimulation were performed as described previously (5). Platelets (1 × 109 in 1 ml) were labeled in 15 mM Tris, pH 7.4, 140 mM NaCl, and 5.6 mM glucose containing 1 mCi of 32PO4. 5-Phosphatase I was extracted by suspending 1 × 109 platelets in 200 µl of 10 mM imidazole, pH 7.2, 3 mM MgCl2, 1 mM EDTA, 0.3 M sucrose, 5 mM 2-mercaptoethanol, 50 µM PMSF, and 1% Triton X-100. The suspension was agitated at 4 °C overnight and centrifuged at 14,000 × g. The supernatant containing both cytosolic and membrane proteins was collected.
Purification of 5-Phosphatase I by Mono Q ChromatographySf9 cell extracts were prepared by suspending Sf9 cells in 10 mM imidazole, pH 7.2, 3 mM MgCl2, 1 mM EDTA, 0.3 M sucrose, 5 mM 2-mercaptoethanol, 50 µM PMSF, and 1% Triton X-100. The extraction was as described above for platelets. The extract (8 mg of protein) was applied to a 1-ml Mono Q column equilibrated in 10 mM imidazole, pH 7.2, 3 mM MgCl2, 1 mM EDTA, 0.2% Triton X-100. The column was eluted at 1 ml/min with linear NaCl gradients as follows: 0-10 min, equilibration buffer; 10-30 min, 0-0.15 M NaCl in equilibration buffer; 30-40 min, 0.15-0.3 M NaCl in equilibration buffer; 40-50 min, 0.3-1.0 M NaCl in equilibration buffer. For Mono Q chromatography of labeled platelet extracts, the platelets were extracted at 4 °C overnight in 10 mM imidazole, pH 7.2, 3 mM MgCl2, 1 mM EDTA, 1 µM pepstatin A, 4 mM 2-mercaptoethanol, 50 µM PMSF, 1% Triton X-100, 1 mM sodium molybdate, and 4 mM sodium pyrophosphate. The extract from 1 × 109 labeled platelets was applied to the column equilibrated with 10 mM imidazole, pH 7.2, 3 mM MgCl2, 1 mM EDTA, 0.2% Triton X-100, 1 mM sodium molybdate, and 4 mM sodium pyrophosphate. The column was eluted with the same gradient as described above.
Immunoprecipitation of 5-Phosphatase I from Platelet ExtractsThe extract from 2.5 × 107 labeled platelets was incubated with 10 µg of anti-5-phosphatase I IgG at 4 °C for 1 h in 20 µl containing 20 mM Tris, pH 7.5, 1 mM EDTA, 1 µM pepstatin A, and 1% Triton X-100. Protein A-Sepharose (40 µl of a 50% slurry in 50 mM Tris, pH 7.5, 150 mM NaCl, and 1% Triton X-100) was added, and the mixture was agitated at 4 °C for 1 h. The protein A-Sepharose pellets were collected by centrifugation and washed three times with 1 ml of the same buffer. The washed pellets were subjected to 10% SDS-PAGE and transferred to nitrocellulose membranes. The extract from 1 × 107 labeled platelets was also run on the same gel to serve as a marker for phosphorylated proteins. In some experiments 1 mM N-ethylmaleimide was substituted for DTT in the gel loading buffer in order to visualize proteins migrating in the region near IgG when membranes were immunoblotted.
Immunoprecipitation of Recombinant Proteins by Monoclonal Anti-HA AntibodyFor immunoprecipitation with monoclonal anti-HA antibody, 2.5 µg of purified GST-5-phosphatase I fusion protein and 0.16 µg of HA-tagged pleckstrin were incubated with 0.3 µg of monoclonal anti-HA antibody (IgG2b) or irrelevant IgG2b in 30 µl of 80 mM imidazole, pH 7.5, 100 mM KCl, 50 µM PMSF, 20% glycerol at 4 °C for 1 h. 30 µl of 50% protein A-Sepharose previously equilibrated with 50 mM Tris, pH 7.5, 150 mM NaCl, 1% Triton X-100 was added, and after centrifugation the pellet was washed three times with 1 ml of the above buffer and mixed with SDS-PAGE loading buffer containing 1 mM N-ethylmaleimide in place of DTT. The samples were separated by 10% SDS-PAGE, transferred to nitrocellulose membranes, and immunoblotted with a mixture of anti-pleckstrin antibodies and anti-5-phosphatase I antibodies. The amount of each protein was estimated by comparing the band intensity of the immunoblots to those of known amounts of the recombinant proteins.
Binding of Pleckstrin to GST-5-Phosphatase I1.6 nmol of GST or GST-5-phosphatase I fusion protein were linked to 1-ml GST-agarose columns, equilibrated with 20 mM Tris, pH 7.5, 1 mM EDTA, 1 mM DTT, 100 mM NaCl, and 0.5% Triton X-100. 200 µg of platelet extract was applied to each column, and they were then washed with 10 column volumes of equilibrating buffer. Bound proteins were eluted with 5 column volumes of 20 mM Tris, pH 7.5, 20 mM GSH, 1 mM EDTA, and 1 mM DTT. Fractions (1 ml) were collected and 20 µl of each was subjected to 10% SDS-PAGE followed by immunoblotting with anti-pleckstrin antibody. The fraction of pleckstrin in the eluates was estimated by comparison of the immunoblots of varying amounts of platelet extracts with those of the eluates.
Activation of 5-Phosphatase I by Various Forms of PleckstrinPleckstrin, pseudophosphorylated pleckstrin, or PH
pleckstrin, in which both pleckstrin domains were deleted, were added to 0.4 µg of purified 5-phosphatase I in 10 µl containing 50 mM Tris, pH 7.5, and 3 mM MgCl2.
The mixtures were immediately diluted with 50 mM Tris, pH
7.5, and 3 mM MgCl2 and assayed for
5-phosphatase activity. The same procedures were used with
GST-5-phosphatase I fusion protein as a source of 5-phosphatase.
The quantities of recombinant proteins used in various experiments were estimated by Coomassie Blue staining of SDS-PAGE and were determined to be of >95% purity.
For immunoblotting, pleckstrin antibody was diluted 1:10,000 and incubated with membranes overnight at 4 °C.
It was previously reported that a human platelet inositol
polyphosphate 5-phosphatase (designated 5-phosphatase I) was
phosphorylated and activated by protein kinase C (5). Since this
protein is approximately the same size as the major protein
phosphorylated in platelets in response to thrombin, pleckstrin was
assumed to be that protein. Recently cDNA molecules encoding
5-phosphatase I have been isolated (6, 7), and the predicted sequence bears no primary sequence similarity to pleckstrin (10). There are now
tools including recombinant proteins and antibodies that allow a
re-examination of the relationship between pleckstrin and 5-phosphatase
I. Initially we carried out SDS-PAGE of platelet proteins and performed
Western blots using antibodies against 5-phosphatase I and pleckstrin
to identify these proteins. While the proteins have approximately the
same mass, they are readily separated on SDS-PAGE with 5-phosphatase I
migrating at an apparent molecular weight slightly greater than that of
pleckstrin as show in Fig. 1A (lanes
1 and 2). The platelets in this experiment were labeled
with [32P]orthophosphate to identify the proteins
phosphorylated in response to thrombin as shown in Fig. 1B
(compare Western blot in lanes 3 and 4 to
autoradiography in B). It is clear that pleckstrin is the
major phosphoprotein in this experiment. Since the amount of pleckstrin
in platelets (0.1% of platelet protein2)
is approximately 25 times that of 5-phosphatase I, it is not possible
to determine whether 5-phosphatase I in platelets was also
phosphorylated in this experiment. The amount of 5-phosphatase I was
estimated by Western blot analysis of different amounts of platelets
compared to Western blots of purified recombinant 5-phosphatase I.
To determine whether 5-phosphatase I was also phosphorylated, we
immunoprecipitated 5-phosphatase I from extracts of
32P-labeled platelets (Fig. 2).
Surprisingly, we found that essentially all of the major protein,
phosphorylated in response to thrombin, was immunoprecipitated with
5-phosphatase I antibody (compare lane 2 to lane
5 in Fig. 2). A lighter exposure of this autoradiograph confirmed
that most of the radiolabeled 40-kDa protein was immunoprecipitated (data not shown). The protein must be pleckstrin since this is the
major protein phosphorylated under these conditions as shown in Fig. 1.
This appears to be a specific association since other proteins
phosphorylated in platelets are not enriched in the immunoprecipitate, and there is very little radioactivity precipitated by preimmune serum.
We used a recombinant fusion protein consisting of 5-phosphatase I
linked to glutathione S-transferase (GST-5-phosphatase I) to
demonstrate directly that pleckstrin binds to 5-phosphatase I. Platelet
extracts were passed over columns of GST-agarose and GST-5-phosphatase I-agarose. The bound proteins were eluted with GSH
and then subjected to SDS-PAGE followed by Western blotting. Pleckstrin
was eluted from the GST-5-phosphatase column (lanes 3 and
4 in Fig. 3A) but not the GST
column (lanes 1 and 2 in Fig. 3A). By
comparing to Western blots of varying amounts of platelet extracts, we
estimated that approximately 22.5% of pleckstrin in the extracts were
recovered from the GST-5-phosphatase I-agarose column. We also
immunoprecipitated 5-phosphatase I from platelet extracts and then
separated the immunoprecipitate by SDS-PAGE followed by Western
blotting with anti-pleckstrin antibodies showing co-immunoprecipitation
(Fig. 3B).
The stoichiometry of the association was determined in an experiment
where HA-tagged pleckstrin was mixed with an excess of 5-phosphatase I
and the complex was precipitated with an anti-HA antibody followed by
SDS-PAGE and Western blotting with both anti-pleckstrin and
anti-5-phosphatase I as shown in Fig. 4A.
Pleckstrin was immunoprecipitated since no pleckstrin was detected in
the supernatant of the immunoprecipitate. The amount of 5-phosphatase I
in this immunoprecipitate was estimated by comparing this blot to
Western blots of serial dilutions of 5-phosphatase I (Fig.
4B). We found 0.7 mol of 5-phosphatase I/mol of pleckstrin
indicating a stoichiometric association between these proteins.
We next determined the effect of the association between pleckstrin and
5-phosphatase I on 5-phosphatase catalytic activity. In this experiment
we mixed pleckstrin, pseudophosphorylated pleckstrin, and PH
pleckstrin (in which both PH domains are deleted) with purified
recombinant 5-phosphatase I (specific activity, 120 µmol of
Ins(1,4,5)P3 hydrolyzed per min per mg of protein). Only
the pseudophosphorylated form of pleckstrin affected enzyme activity using Ins(1,4,5)P3 as substrate as shown in Fig.
5. There was a 2.1-fold increase in activity observed in
this experiment with the maximal effect observed at approximately 1 mol
of pleckstrin/mol of 5-phosphatase I. In other experiments the
activation ranged from 2- to 3-fold. This result implies that
phosphorylated pleckstrin binds and activates 5-phosphatase I. We
measured the effect of pleckstrin and pseudophosphorylated pleckstrin
on the hydrolysis of varying concentrations of Ins(1,4,5)P3
by recombinant 5-phosphatase I. We found a Km of 7.7 µM with both pleckstrins, while the
Vmax of pleckstrin-5-phosphatase I was 121 µmol of InsP2 formed per min per mg protein and that of
pseudophosphorylated pleckstrin-5-phosphatase I complex was 291 µmol
of InsP2 formed per min per mg of protein. Thus
phosphorylation of pleckstrin increases the Vmax
of 5-phosphatase I without changing Km as reported
previously, using partially purified 5-phosphatase I (5). The
Km value that we obtained is similar to that
reported by others (7, 21). We also carried out this experiment using
the fusion protein GST-5-phosphatase I with a similar result although
in this case native pleckstrin also stimulated enzyme activity when
added in large amounts (10 pleckstrin/GST-5-phosphatase). We also mixed
recombinant 5-phosphatase II (another form of inositol polyphosphate
5-phosphatase that is found in platelets) with pseudophosphorylated and
native pleckstrin and found no effect on enzyme activity (data not
shown).
In order to evaluate whether phosphorylated pleckstrin also activates
5-phosphatase I in platelets, we fractionated extracts from platelets
on a Mono Q column and subjected each fraction to Western blotting with
both anti-pleckstrin and anti-5-phosphatase I antibodies and assayed
enzyme activity. Another sample of 32P-labeled platelets
was also fractionated on Mono Q and subjected to autoradiography to
locate phosphorylated pleckstrin as shown in Fig. 6.
Western blotting indicated that both proteins were eluted together in
fractions 21-27 with pleckstrin continuing to fraction 31 as shown in
Fig. 7. The phosphorylated pleckstrin eluted almost
entirely in fraction 21 as shown in Fig. 6. The 5-phosphatase activity
was also highest in this fraction indicating that pleckstrin also
activates the enzyme in platelets when it is phosphorylated. By
comparing the intensity of the 5-phosphatase I immunoblots, we estimate
that phosphorylation increases enzyme activity by at least 7-fold
(compare fractions 21 and 23 in Fig. 7, A and B).
The specific activity is also high in fraction 20, which contains some
phosphorylated pleckstrin (Fig. 6); however, the amount of
5-phosphatase I in this fraction is below the detection limit of
Western blotting (Fig. 7A) so that specific activity cannot
be calculated. It is possible that pleckstrin also binds to other
proteins in platelets since some phosphorylated pleckstrin is in
fraction 27. The 5-phosphatase activity in this fraction includes
5-phosphatase II, which is another major inositol
polyphosphate-5-phosphatase in platelets, that eluted in fraction 27, and 5-phosphatase I in this fraction cannot be estimated (21).
We have demonstrated that pleckstrin plays a role in phosphatidylinositol-mediated calcium signaling. When pleckstrin is phosphorylated on serine and threonine residues in response to thrombin, it binds to and activates 5-phosphatase I thereby accelerating the degradation of the calcium ion-mobilizing messenger molecule Ins(1,4,5)P3. Since the amount of pleckstrin in platelets is approximately 25 times that of 5-phosphatase I and the complex between the two is stoichiometric, it is clear that most of the platelet pleckstrin must have another function. Additionally, since most of the pleckstrin phosphorylated after thrombin treatment of platelets is immunoprecipitated by 5-phosphatase I, it is clear that most of pleckstrin is not phosphorylated in response to thrombin. It has been suggested that pleckstrin also leads to inhibition of phosphatidylinositol turnover in response to agonists (18). This would also serve to diminish calcium mobilization and thus the molecule may have a concerted action in terminating agonist responses. The effect on 5-phosphatase I is direct since pseudophosphorylated pleckstrin activates recombinant 5-phosphatase I in vitro. These results can explain our previous report that protein kinase C phosphorylates and activates platelet 5-phosphatase I. That preparation of enzyme had an activity of 4 µmol of Ins(1,4,5)P3 hydrolyzed per min per mg of protein compared with 120 µmol of Ins(1,4,5)P3 hydrolyzed per min per mg of protein for recombinant 5-phosphatase I purified from baculovirus-infected Sf9 cells. Since pleckstrin forms complexes with 5-phosphatase I and activates it, the earlier preparation undoubtedly contained pleckstrin as the major Coomassie Blue staining protein. Thus in the presence of protein kinase C, pleckstrin, which was the major protein in the partially purified enzyme preparation, was phosphorylated, and the phosphorylation of pleckstrin that bound to 5-phosphatase activated the enzyme. We cannot exclude the possibility that 5-phosphatase I is also phosphorylated. In two-dimensional gels in which one dimension is SDS-PAGE and the other is isoelectric focusing, 5-phosphatase I appears as a series of discrete spots (after Western blotting) along the pH dimension characteristic of phosphorylated proteins such as pleckstrin (data not shown). It is possible that this finding does not reflect phosphorylation but some other modification since at least in Sf9 cells labeled with [32P]orthophosphate, 5-phosphatase I appears not to be labeled and recombinant 5-phosphatase I purified from E. coli does not appear to be a substrate for protein kinase C (8). However, it has been reported that Sf9 cells are devoid of protein kinase C activity (22).
The function of pleckstrin domains remains uncertain in most cases. Several studies have suggested that proteins containing pleckstrin domains bind phospholipids in order to anchor proteins to membranes (13, 23, 24). In the studies reported here there is a clear action of pleckstrin that involves a direct protein-protein interaction without any role for lipids.