Protein-tyrosine phosphatase MEG2 is expressed by human neutrophils. Localization to the phagosome and activation by polyphosphoinositides.

Signaling pathways involving reversible tyrosine phosphorylation are essential for neutrophil antimicrobial responses. Using reverse transcriptase PCR, expression of the protein-tyrosine phosphatase MEG2 by peripheral neutrophilic polymorphonuclear leukocytes (PMN) was identified. Polyclonal antibodies against MEG2 were developed that confirmed expression of MEG2 protein by PMN. Through a combination of immunofluorescence and cell fractionation followed by immunoblotting, we determined that MEG2 is predominantly cytosolic with components present in secondary and tertiary granules and secretory vesicles. MEG2 activity, as determined by immunoprecipitation and in vitro phosphatase assays, is inhibited after exposure of cells to the particulate stimulant opsonized zymosan or to phorbol 12-myristate 13-acetate but largely unaffected by the chemoattractant N-formyl-methionyl-leucyl-phenyalanine. Studies using bacterially expressed glutathione S-transferase MEG2 fusion protein indicate that cysteine 515 is essential for catalytic activity, whereas the noncatalytic (N-terminal) domain of MEG2 negatively regulates the enzymatic activity of the C-terminal phosphatase domain. The activity of MEG2 is enhanced by specific polyphosphoinositides with the order of potency being phosphatidylinositol (PI) 4,5-diphosphate > PI 3,4,5-triphosphate > PI 4-phosphate. MEG2 associates at an early stage with nascent phagosomes. Taken together, our results indicate that MEG2 is a polyphosphoinositide-activated tyrosine phosphatase that may be involved in signaling events regulating phagocytosis, an essential antimicrobial function in the innate immune response.

Neutrophils are an essential component of the innate immune system. To function in host defense, they have evolved a variety of mechanisms, including chemotaxis, phagocytosis, and intracellular killing of invading microbial pathogens. Paradoxically, unregulated activation of these responses can result in tissue injury, as is believed to occur in inflammatory disorders such as rheumatoid arthritis (1), inflammatory bowel disease (2,3), ischemia-reperfusion (4,5), and acute lung injury (6,7). Leukocyte microbicidal responses must be precisely regulated by selective activation and rapid cessation of signaling cascades once the initial stimulus has been removed.
Signaling pathways involving tyrosine phosphorylation are pivotal to neutrophil antimicrobial responses, including adherence (8), chemotaxis (9), phagocytosis (10), and oxidant production (11)(12)(13). The level of tyrosine phosphorylation is regulated by the reciprocal activities of protein-tyrosine kinases (PTK) 1 and protein-tyrosine phosphatases (PTP). Although PTK have been extensively studied, current knowledge of the identity and functional importance of PTP expressed in neutrophils is comparatively less. To date, CD45 (14 -16), SHP-1 (17), SHP-2 (18), and CD148 (19) have been identified in neutrophils, but, by inference from studies in other cell types, additional PTP are likely to be expressed and participate in cell regulation (20).
The objectives of the current study were to characterize the subcellular localization of MEG2 and to begin to elucidate its function in myeloid cells. Our results indicate that MEG2 is present in both the cytosol and granules and becomes incorporated into nascent phagosomes. The activity of MEG2 is enhanced by the polyphosphoinositides PI 4,5-P 2 , PI 3,4,5-P 3 , and PI 4-P. These observations suggest a role for MEG2 in the regulation of signaling pathways involved in phagocytosis.

EXPERIMENTAL PROCEDURES
Materials-MEG2-specific and degenerate primers were synthesized by General Synthesis and Diagnostics (Toronto, Ontario, Canada). Restriction enzymes and all other primers were purchased from Invitrogen (Burlington, Ontario, Canada). The TA Cloning kit and One-Shot kit were purchased from Invitrogen (San Diego, CA). MuLV reverse transcriptase and Ampli Taq DNA polymerase were purchased from PerkinElmer Life Sciences (Applied Biosystems, Mississauga, Ontario, Canada). The T7 Sequencing kit was purchased from Amersham Biosciences, Inc. (Bai D'Urfey, Quebec). Pfu polymerase was purchased from Stratagene (La Jolla, CA). Diisopropylfluorophosphate (DFP), latrunculin B, and wortmannin were purchased from Calbiochem (San Diego, CA). Glutathione-agarose and p-nitrophenyl phosphate were obtained from Sigma Chemical Co. (St. Louis, MO). Protein A/G Plusagarose was purchased from Santa Cruz Biotechnologies (Santa Cruz, CA). Malachite Green reagent and phosphopeptide were purchased from Upstate Biotechnology Inc. (Lake Placid, NY). A glutathioneagarose column was purchased from Pierce (Rockford, IL). Cholesterol, ceramide, cardiolipin, phosphatidylcholine, dioleoyl-phosphatidylcholine, phosphatidylethanolamine, phosphatidic acid, and phosphatidylinositol were purchased from Avanti (Alabaster, AL). All phosphatidylinositol phosphatases were purchased from BIOMOL (Plymouth Meeting, PA) except PI 3,5-P 2 and PI 5-P, which were purchased from Echelon (Salt Lake City, UT).
Antibodies-Anti-HA mouse monoclonal antibody was from the Berkeley Antibody Co. (Berkley, CA) and used at a 1:5000 dilution for immunoblotting. Anti-mouse and anti-rabbit horseradish peroxidaseconjugated secondary antibodies (Amersham Biosciences, Inc.) were used at a 1:5000 dilution. Goat anti-rabbit and anti-mouse Texas Redconjugated secondary antibodies were purchased from Molecular Probes (Eugene, OR). Rabbit immune and preimmune serum were used in a 1:1000 dilution for immunoblotting, and affinity-purified immune serum was used at a 1:2000 dilution for immunoblotting. Anti-hemagglutinin (HA) rabbit polyclonal antibody and nonimmune rabbit IgG were purchased from Zymed Laboratories Inc. (San Francisco, CA). For immunofluorescence, all antibodies were used at a 1:100 dilution. Antiphosphotyrosine (4G10) mouse monoclonal antibody was purchased from Upstate Biotechnology Inc. and used at a 1:2000 dilution. Murine monoclonal anti-human LAMP2 antibody was purchased from the University of Iowa Repository (Iowa City, IA) and used at a 1:40 dilution. Anti-gp91phox antibody was a gift from Drs. Al Jesaitis and Mark Quinn (University of Montana, Bozeman, MT) and used at a 1:1000 dilution.
Neutrophil and CD4 ϩ T-lymphocyte Isolation-Human neutrophils were isolated from whole blood obtained by venipuncture and anticoagulated with citrate. Dextran sedimentation and discontinuous plasma-Percoll gradients were used as previously described (31). The remaining monocyte-lymphocyte band was collected and passed through a human CD4 immunocolumn (Cedarlane Laboratories, Ontario, Canada) to isolate CD4 ϩ T-lymphocytes. The purity of the neutrophils isolated exceeded 98% as determined by modified Wright-Giemsa staining of cytocentrifuge preparations. The purity of the CD4 ϩ T-lymphocytes exceeded 85% as determined by fluorescence-activated cell sorting analysis. After isolation, cells were resuspended in Krebs-Ringer phosphate dextrose buffer with glucose at a concentration of 8 ϫ 10 6 cells/ml and gently rotated at room temperature until used (generally within 1 h).
Cell Culture-HL-60 and U937 cells were obtained from the American Type Culture Collection. Cells were grown in RPMI 1640 media containing 10% heat-inactivated fetal bovine serum, penicillin (100 units/ml), streptomycin (100 g/ml), and 0.5 mM L-glutamine.
RNA Isolation-Total RNA was isolated from human peripheral blood neutrophils and HL-60 cells by the guanidinium isothiocyanatecesium chloride protocol (32,33).
Reverse Transcriptase (RT) PCR-cDNA was reverse-transcribed from total RNA isolated from human peripheral blood neutrophils and HL-60 cells. Murine leukemia virus (MuLV) reverse transcriptase, and random hexamers from the Gene Amp RNA PCR kit were used for the reaction, which was carried out in a DNA Thermal Cycler 480 (PerkinElmer Life Sciences, Emeryville, CA) according to the manufacturer's instructions. Transcribed cDNA was then amplified using the degenerate primers for an initial five cycles with the following parameters: denaturation at 94°C for 30 s, annealing at 42°C for 30 s, and extension at 72°C for 2 min. Another 30 cycles followed with denaturation at 94°C for 30 s, annealing at 55°C for 30 s, and extension at 72°C for 2 min. Transcribed cDNA was amplified using MEG2 primer sets 1 and 2 for 35 cycles with denaturation at 94°C for 30 s, annealing at 50°C for 30 s, and extension at 72°C for 2 min. All components of the amplification mixtures were tested for contamination by running 35 PCR cycles in absence of the template RNA. Genomic DNA contamination was tested for by performing the RT-PCR in absence of the MuLV reverse transcriptase. All reactions were analyzed by agarose gel electrophoresis using ethidium bromide staining. Amplified cDNA sequences from the RT-PCR were cloned into the Invitrogen TA vector. Sequencing of inserts from 150 individual clones was performed using the Pharmacia T7 sequencing kit. Automated fluorescence sequencing was additionally performed on selected sequences at the Biotechnology Center of the University of Toronto.
Construction of MEG2-hemagglutinin Fusion Protein Vector-The cloned MEG2 cDNA in pBluescript II SK(ϩ) (Stratagene) was received as a gift from P. Majerus (Washington University School of Medicine). The MEG2 insert was then transferred to the pCDNA3 vector (Invitrogen). An XhoI restriction site was added to 5Ј-end of the MEG2 insert. A 3Ј-primer (CCCCCTCGAGTTACGCATAGTCAGGAACATCGTATG-GGTACTGACTCTCCACGGCCAGCAGG) was constructed containing the final 24 bp of the coding MEG2 sequence, the hemagglutinin (HA) sequence, a stop codon, and an XhoI restriction site. A 5Ј-primer (GG-CAGAAACGCCAGGTGACCC) was synthesized that targeted an upstream region in the MEG2 cDNA containing a BstEII restriction site. Twenty-five cycles of the polymerase chain reaction (PCR) were performed with denaturation for 1 min at 94°C, annealing at 55°C for 1 min, and extension at 72°C for 1 min using Pfu polymerase. The PCR product was double-digested (BstEII, XhoI) and ligated into the MEG2-pcDNA3 construct in place of the existing BstEII-XhoI sequence in the native MEG2 sequence.
GST-MEG2 Fusion Protein Constructs-A 5Ј-primer (TTTTCG-GAATTCATGGAGCCCGCCACCGCG) was synthesized containing an EcoRI site immediately followed by the first 18 bp of the coding MEG2 sequence. A 3Ј-primer (ACTTGCGAATTCTTACTGACTCTCCACGGC) was synthesized containing the final 15 bp of the coding MEG2 sequence immediately followed by a stop codon and an EcoRI site. Twenty-five cycles of PCR were performed with denaturation for 1 min at 94°C, annealing at 53°C for 1 min, and extension at 72°C for 1 min using Pfu polymerase and the MEG2-Bluescript construct as a template. The PCR product was digested with EcoRI and ligated into the EcoRI site of the pGEX-4T1 vector (Amersham Biosciences, Inc., Bai D'Urfey, Quebec).
A catalytically inactive pGEX-MEG2 construct was constructed using a 3Ј-primer (CAGAAGGTACCTGTCCTGCCAATGCCTGCACTG-CTATGGA), which coded for a T to A substitution at position 1543 of the MEG2-coding cDNA sequence thus converting the cysteine 515 to serine. PCR was performed (as above) using a 5Ј-primer (TTTTCGGAAT-TCATGTATGAAGACATTCGTCGT) targeting bp 919 -936 and the pGEX-wild type MEG2 construct as template DNA. The BstEII-KpnI region of the PCR product was used to replace the corresponding region in the pGEX-wild type MEG2 construct.
A pGEX construct containing the MEG2 noncatalytic domain and a histidine tag was developed using a 5Ј-primer (TTTTCGGAATTCATG-GAGCCCGCGACCGCG) containing a EcoRI site immediately followed by the first 18 bp of the coding MEG2 sequence. The 3Ј-primer (CCC-CCTCGAGTTAGTGATGGTGATGGTGATGAGATCCTCTGGTCATA-GCATGGGGACCTGG) contained bp 841-861 of the MEG2 coding sequence, a 6xHis tag, a stop codon, and an XhoI site. The pGEX-WT-MEG2 construct was used as a template. The PCR product was doubledigested with EcoRI and XhoI and ligated with the corresponding sites of the pGEX-4T1 vector. A pGEX construct containing the MEG2 phosphatase domain and a hemagglutinin tag was developed using a 5Ј-primer (TTTTCGGAATTCATGACCATCCAAGAGTTG) containing a EcoRI site immediately followed by the bp 859 -873 of the coding MEG2 sequence. The 3Ј-primer (CCCCCTCGAGTTACGCATAGTCAGGAAC-ATCGTATGGGTACTGACTCTCCACGGCCAGCAGG) contained the final 24 bp of the coding MEG2 sequence, the hemagglutinin (HA) sequence, a stop codon, and an XhoI restriction site. The pcDNA3-wild type MEG2-HA tag construct was used as a template. The PCR product was double-digested with EcoRI and XhoI and ligated with the corresponding sites of the pGEX-4T1 vector.
GST Fusion Protein Expression-Bacteria were cultured in LB media (containing 50 g/ml ampicillin) and grown to an OD of 0.6 -0.8. Samples were induced with 0.1 mM isopropyl-1-thio-␤-D-galactopyranoside for 4 h, shaking at room temperature. Bacteria were centrifuged (3500 rpm in a Beckman GPR centrifuge), washed once in STE buffer (10 mM Tris, pH 8.0, 150 mM NaCl, 1 mM EDTA), and stored overnight as pellets at Ϫ80°C. The following day the pellets were resuspended in a lysozyme buffer (100 g/ml hen egg-white lysozyme, 1 mM PMSF, 1 g/ml aprotinin, 1 g/ml leupeptin in STE buffer) for 15 min. DTT (5 mM) and Triton X-100 (1%) were added, then samples were incubated for 15 min. MgCl 2 (10 mM) and DNase I (10 g/ml) were added, and samples were allowed to incubate for a final 15 min. After centrifugation at 23,000 ϫ g for 15 min, the supernatant was recovered and combined with glutathione-Sepharose (1 ml of packed beads per liter of bacteria) for 30 min. The Sepharose beads were sedimented (500 ϫ g, 5 min) and washed three times in STE buffer. Fusion protein was eluted by resuspension of the beads in a buffer containing 75 mM Hepes, pH 7.4, 150 mM NaCl, 10 mM reduced glutathione, 5 mM DTT, 0.1% Triton X-100. All steps were done at 4°C.
Anti-MEG2 Antibodies-A whole-molecule MEG2 antibody was produced as follows: The GST-wild type MEG2 protein was used to immunize three rabbits. Primary immunizations consisted of 1 mg of fusion protein combined with Freund's complete adjuvant. Subsequent boosts were performed at 3-week intervals using 0.5 mg of protein mixed with Freund's incomplete adjuvant. The rabbits were bled at 4-to 8-week intervals, and the resulting serum was tested by probing immunoblots of MEG2-transfected cell lysates.
A peptide antibody targeting the MEG2 N terminus was produced as follows: The Kyte-Doolittle, Hopp-Woods, and Surface algorithms were used to select a 20-amino acid region of the MEG2 sequence for immunization. The sequence consisting of amino acids 63-82 (ELFHSYR-ETRRKEGIVKLKP) was identified. This peptide was synthesized and used for immunization of two rabbits. Keyhole limpet hemocyanin was coupled as a carrier protein through a cysteine added at the peptide N terminus. Three bleeds (including a preimmune bleed) were obtained from the rabbits, from which the serum was prepared. Immunoreactivity against the immunizing peptide was confirmed by enzyme-linked immunosorbent assay by the manufacturer (Synpep Corp., Dublin, CA).
Specific tyrosine phosphatase activities were determined for MEG2 present in the immunoprecipitates as follows: optical densities were converted to units of phosphate released per 1 ϫ 10 6 cells by using a standard curve based on known phosphate concentrations. Background values for nonimmune IgG immunoprecipitates were subtracted from the values.
GST-MEG2 Affinity Purification Column-GST-wild type MEG2 fusion protein was covalently linked onto a glutathione-agarose column according to the manufacturer's instructions (Pierce). For each antibody purification, 1 ml of immune serum was filtered through a 0.45-m filter and then diluted with an equal volume of binding buffer (25 mM Tris, pH 7.6, 150 mM NaCl). The sample was loaded onto the column and allowed to incubate in the gel matrix for 1 h at room temperature. The column was then washed with 100 ml of wash buffer (25 mM Tris, pH 5.0, 500 mM NaCl, 0.1% Tween 20) after which the antibody was eluted into 1-ml fractions using Gentle Elution Buffer (Pierce). After analyzing the samples at 280 nm, fractions with high protein concen-trations were pooled and dialyzed extensively against binding buffer.
SDS-PAGE and Immunoblotting-SDS-PAGE and immunoblotting were carried out as described previously (34). The Western blots were then developed using the enhanced chemiluminescence (ECL) detection system following the manufacturer's instructions (Amersham Biosciences, Inc.).
Immunofluorescence-Prior to fixation, neutrophils were pretreated in 2.5 mM DFP for 30 min at room temperature. Cells were fixed in 1.6% paraformaldehyde for 20 min and then washed with PBS. Cells (75,000) were cytospun for 5 min at 800 rpm onto glass slides and then incubated in a permeabilization/blocking buffer (0.5% Triton X-100 and 10% skim milk in PBS) for 20 min at 37°C. Samples were incubated with primary antibody (diluted in 10% skim milk, 0.2% Tween 20, in PBS) for 1 h at 37°C. Samples were then washed 3 ϫ 5 min in wash buffer (10% skim milk, 0.2% Tween 20, in PBS) and then incubated with secondary antibody diluted in wash buffer for 1 h at 37°C. Samples were washed three times with 0.2% Tween 20 in PBS and then mounted (Dako mounting medium). Fluorescent images were captured using a Hamamatsu Orca cooled charge-coupled device camera mounted on a Leica DMIRB inverted fluorescence microscope.
Immunoprecipitation and in Vitro Phosphatase Assays-For each condition, 1 ϫ 10 7 neutrophils were pretreated in 2.5 mM DFP for 30 min at room temperature. Cells were disrupted in cold lysis buffer (1% Nonidet P-40, 20 g/ml aprotinin, 20 g/ml leupeptin, 0.5 mM Pefabloc, and 1 mM PMSF) and cleared of insoluble material by two successive centrifugations (22,000 ϫ g ϫ 15 min, 4°C). Complexes of anti-MEG2 antibody pre-bound to Protein A/G Plus-agarose beads were added to the cleared supernatant. The samples were rotated at 4°C for 2 h, after which the beads were separated and extensively washed. One-fifth of the bead volume was removed and used for immunoblot analysis by resuspension in Laemmli sample buffer containing 20 mM N-ethylmaleimide and then subjected to SDS-PAGE and immunoblotting. The blots were probed with a 1:2000 dilution of MEG2 63-82 antibody followed by washing and then incubation with a 1:3000 dilution of horseradish peroxidase-labeled Protein A (Bio-Rad). The remaining beads were sedimented and resuspended in assay buffer containing 0.1 mM phosphopeptide (RRLIEDAEpYAARG) and 60 mM ␤-mercaptoethanol. Samples were shaken for 3 h at 37°C and centrifuged briefly. For endogenous phosphatase assays, Malachite Green was used to detect free phosphate released from the phosphopeptide by measurement of absorbance at 650 according to the manufacturer's instructions. Malachite Green was used in these assays due to its sensitivity.
Subcellular Fractionation-Neutrophil subcellular fractionation was performed according to the method of Kjeldsen et al. (35) as previously described (17). Briefly, neutrophils were disrupted by nitrogen cavitation, and the postnuclear supernatant applied on top of a three-layer Percoll gradient (1.05/1.09/1.12 g/ml). Centrifugation at 37,000 ϫ g for 30 min yielded four separate bands corresponding to the primary, secondary, and tertiary granules and a mixture of secretory vesicles and plasma membranes. Each fraction was assayed for markers of the corresponding subcellular compartment as described previously (36). Percoll was removed from the samples by centrifugation, and the remaining biological material was mixed with boiling, 2ϫ Laemmli sample buffer. Equal amounts of protein from each fraction were subjected to SDS-PAGE and immunoblotting.
Fusion Protein Phosphatase Assays-Phosphatase activity was assessed by incubation of the fusion protein in an assay buffer containing: 25 mM Hepes, pH 7.2, 50 mM NaCl, 2.5 mM EDTA, 5 mg/ml p-nitrophenyl phosphate (pNPP) substrate, and 10 mM DTT. Samples were incubated for 30 min shaking at 37°C, and absorbance was measured at 405 nm.
To test the effect of lipids on phosphatase activity, lipids were dissolved in methanol (10 g/ml) and used to coat octyl-Sepharose beads (Amersham Biosciences, Inc.) overnight at 4°C. The beads were washed twice in water and six times with assay buffer (25 mM Hepes, 25 mM NaCl, 2.5 mM EDTA, pH 7.2) and incubated with 10 pmol of each fusion protein for 30 min at 37°C. Subsequently, pNPP (10 mg/ml) and DTT (10 mM) were added and incubated for an additional 30 min at 37°C. The supernatant was collected and the absorbance at 405 nm determined.
Mixed Liposome Assays-Phosphatidylcholine liposomes containing additional test lipids as a minor molar constituent (5% molar concentration) were prepared by evaporating lipid mixtures in a rotary evaporator followed by resuspension in an assay buffer composed of 30 mM Hepes, pH 7.2, 100 mM NaCl, 1 mM EDTA. After one freeze-thaw cycle, the samples were extruded through a 400-nm filter and stored at 4°C until use. The liposomes were labeled with [ 3 H]phosphatidylethanolamine to quantify final concentrations.
To test the effect of the mixed liposome on MEG2 fusion protein function, an 80-l reaction was prepared containing 25 g/ml GST-MEG2 fusion protein, liposomes (250 M lipid concentration), 40 mM Hepes, 100 mM NaCl, 1 mM EDTA, 10 mM DTT, 0.025% Triton X-100, and 0.02% Nonidet P-40, pH 7.2. Following a 40-min incubation at 37°C, 40 l of assay buffer containing10 mg/ml pNPP and 10 mM DTT was added. Samples were incubated again at 37°C and then collected, and the absorbance at 405 nm was measured as described above.
[ 32 P]Orthophosphate Labeling-Orthophosphate labeling was performed essentially as described previously (17). Neutrophil suspensions (1.0 ϫ 10 7 cells/ml) were incubated for 3 h at 37°C in sodium buffer (containing in mM, NaCl 140, potassium 4, calcium 2, magnesium 1, glucose 10) with 0.5% bovine serum albumin in the presence of 32 Plabeled orthophosphate (1.0 mCi/ml). The cells were washed with sodium buffer, treated with DFP for 30 min, and incubated with or without stimuli before lysis and immunoprecipitation with the MEG2whole molecule antibody. Following washing, the samples were then subjected to SDS-PAGE and immunoblot analysis. Quantitation of phosphorylation was performed with a Molecular Dynamics Phosphor-Imager, using the ImageQuant software.
Transfection of U937 Cells-pcDNA3 ("empty vector") or pcDNA3 containing HA-tagged MEG2 was transfected into the monocytic cell line U937 using cationic liposomes (LipofectAMINE) according to manufacturer's instructions. Two micrograms of plasmid DNA per 5 ϫ 10 6 cells was used for transfections. Clones were selected in RPMI 1640 medium supplemented with 1 mg/ml G418 using limiting dilution in 96-well plates and expanded in tissue culture flasks. The expression of the recombinant MEG2 was confirmed using reverse transcriptase (RT)-PCR with primers bracketing the HA tag sequence and the 5Ј portion of MEG2. Expression of HA-tagged MEG2 protein was identified using Western blotting with a monoclonal anti-HA antibody.
Phagosome Isolation-Iron oxide beads (Sigma) were mixed with human serum for 30 min at 37°C with gentle agitation and then captured using a magnetic Capture-Tec stand (Invitrogen). After washing in PBS, the beads were resuspended in assay buffer (140 mM NaCl, 4 mM KCl, 10 mM glucose, 10 mM Hepes, 1 mM MgCl 2 , 1 mM CaCl 2 , pH 7.4), briefly sonicated, and cooled to 4°C. Neutrophils (1 ϫ 10 7 /sample) were cooled to 4°C and then combined with 60 l of beads (50:50 slurry) to allow bead binding without internalization. Cells were warmed to 37°C to allow phagocytosis to proceed and were collected at various time points as specified. Cells were washed two times with ice-cold PBS and then lysed in a buffer composed of 0.5% Triton X-100, 50 mM NaCl, 300 mM sucrose, 3 mM MgCl 2 , 20 g/ml aprotinin, 1 g/ml pepstatin, 1 mM PMSF, and 10 mM Pipes (pH 6.8). Samples were briefly sonicated then phagosome-containing beads were isolated by magnetic separation. The samples were resuspended in lysis buffer and homogenized by passage through a Dounce homogenizer (20 times). The beads were washed three times in lysis buffer, resuspended in Laemmli sample buffer, and analyzed by SDS-PAGE and Western blotting. This method has previously been shown to be nondisruptive of the phagosome membrane (37).
Data Analysis-Data were analyzed by analysis of variance with correction for multiple comparisons using Dunnett t tests. Statistical significance was considered for p values of Ͻ0.05 and indicated with an asterisk in the figures.

PTP-MEG2 Is Expressed by Human Neutrophils-During a search for PTP expressed by myeloid cells, expression of MEG2
in neutrophils was identified through an RT-PCR screen using degenerate primers targeted to the highly conserved catalytic domain of PTP. Of 150 clones that were sequenced, 98 shared complete sequence identity with four known PTP, including CD45 (n ϭ 30), SHP-1 (n ϭ 26), PTP1B (n ϭ 22), and MEG2 (n ϭ 20). The remainder of the clones had significant homology to ribosomal RNA (n ϭ 34) or less than 1% homology with known proteins (n ϭ 18) likely reflecting the degenerate nature of the primers and the relatively low stringency of the initial PCR cycles. Expression of PTP-MEG2 by cell lines of hematopoietic lineage (21,22), including the Jurkat T-cell line (23) has previously been reported. The expression of MEG2 by mature myeloid cells such as neutrophils has not been previously reported and accordingly, we focused our attention on this phosphatase.
Because our initial screening technique used degenerate primers, to confirm the expression of MEG2 in neutrophils an additional RT-PCR analysis was conducted with two additional sets of primers specific for MEG2. This additional analysis targeted two unique areas of MEG2: primer set 1 targeted base pairs 943-1124 of the MEG2 cDNA sequence and primer set 2 targeted a region in the C-terminal direction corresponding to base pairs 1402-1589. Bands of the predicted size (181 and 187 bp) resulted from RT-PCR amplification of total RNA from neutrophils ( Fig. 1). A similar strategy demonstrated the expression of MEG2 mRNA by the myeloid cell line HL-60 ( Fig.  1). DNA sequence analysis confirmed that these PCR products corresponded to MEG2 (not illustrated).
To determine if MEG2 protein was expressed in mature neutrophils, a polyclonal antibody was raised in rabbits against a GST-full-length MEG2 fusion protein. To test the specificity of this antibody (referred to as MEG2-whole molecule), Western analysis was performed on lysates of U937 cells transfected with HA-tagged MEG2. Fig. 2a illustrates that immune but not pre-immune serum recognized a protein of Ϸ70 kDa corresponding to the predicted molecular mass of the recombinant fusion protein. Independent blotting with a monoclonal anti-HA antibody confirmed that the position of the band recognized by the immune serum corresponded to that of the epitope-tagged MEG2. Fig. 2b demonstrates that the affinitypurified immune serum recognized a single 68-kDa immunoreactive band in neutrophils corresponding to the predicted molecular mass of endogenous MEG2. Immunodetection of this band could be competed by the immunizing protein, confirming the specificity of the antibodies. A second polyclonal antipeptide antibody was raised against the N-terminal noncatalytic domain of MEG2 (a 20-amino acid peptide corresponding to residues 63-82). The specificity of this antibody (referred to as MEG2 63-82) was confirmed using the same techniques described for the MEG2-whole molecule antibody (not shown). The specific tyrosine phosphatase activities observed in MEG2 immunoprecipitates were 6.85 Ϯ 0.98 pmol of phosphate released/10 6 cells/g of antibody for the MEG2 whole molecule antibody and 11.63 Ϯ 0.82 pmol of phosphate released/10 6 cells/g antibody for the MEG2 63-82 antibody (n ϭ 3).
MEG2 Is Predominantly Cytoplasmic in Neutrophils-To determine the subcellular localization of MEG2, immunofluorescence localization was initially conducted in U937 cells transfected with HA-MEG2 using an anti-HA antibody. Fig. 3a illustrates that the tagged recombinant MEG2 protein is predominantly cytoplasmic. The localization of the endogenous MEG2 protein was then examined in freshly isolated neutrophils using polyclonal anti-MEG2 antibodies. These studies demonstrated that, although MEG2 was predominantly cytoplasmic, the staining was distinctly punctate or granular, suggesting that MEG2 may be localized, at least in part, to the granules of mature neutrophils. To investigate this possibility, neutrophil subcellular fractions were isolated by Percoll gradient centrifugation and analyzed by immunoblotting (see "Ex-

FIG. 1. Detection of MEG2 RNA by RT-PCR in PMN and HL-60 cells using two distinct sets of primers.
Nondegenerate primer sets 1 and 2 targeted two distinct regions of the MEG2 sequence. RT-PCR was performed on RNA isolated from neutrophils and HL-60 cells. Predicted PCR products for primer sets 1 and 2 were 181 and 187 bp, respectively. No product is detected in negative controls for RT-PCR, excluding template RNA. As a positive control, the plasmid pAW109 was used with specific primers that amplify a region of 308 bp.
perimental Procedures"). These studies (Fig. 4) demonstrated that, although MEG2 is predominantly located in the cytosol, significant amounts are present in the secondary granules and a fraction containing secretory vesicles and plasma membrane. At longer exposures, minor amounts of immunoreactive MEG2 were also observed in the tertiary granule fraction (not shown). Because the immunofluorescence studies (Fig. 3) did not demonstrate a staining pattern consistent with plasma membrane localization, we concluded that the signal from the combined plasma membrane/secretory vesicle fraction was predominantly from the secretory vesicle compartment. The 68-kDa immunoreactive bands in each of these fractions could be competed in the presence of the immunizing protein (not shown), confirming the specificity of detection.
MEG2 Phosphatase Activity Is Inhibited by Agonist Stimulation-Activation of leukocytes results in increased levels of cellular tyrosine phosphorylation that parallels inhibition of tyrosine phosphatases such as SHP-1 (17) and CD45 (20).
To determine if the catalytic activity of MEG2 was modulated by agonist exposure, phosphatase activity was quantified in MEG2 immunoprecipitates (using MEG2 62-83 antibody) from quiescent and activated neutrophils. Fig. 5 illustrates that the enzymatic activity of MEG2 was inhibited following treatment with either phorbol myristate acetate (PMA) or the phagocytic stimulus, opsonized zymosan. Treatment of neutrophils with diamide, a potent oxidizing agent and known phosphatase inhibitor (38,39), resulted in the greatest de-gree of inhibition of MEG2 activity. However, inhibition of MEG2 was not universally associated with agonist exposure, because stimulation of neutrophils with the chemoattractant fMLP, which binds to a receptor of the seven-transmembrane-spanning domain family, did not result in inhibition of FIG. 3. MEG2 localization in leukocytes by immunofluorescence. a, U937 cells were transfected with either MEG2-HA or the pcDNA3 vector alone. Cells were fixed, permeabilized, and incubated with monoclonal anti-HA antibody, followed by a Texas Red-conjugated secondary antibody. Samples were analyzed using confocal microscopy as described under "Experimental Procedures." Space bar ϭ 5 m. b, neutrophils were fixed, permeabilized, and stained with the MEG2whole molecule antibody or control (non-immune) rabbit IgG. Cells were then stained with a Texas Red-conjugated secondary antibody. Samples were analyzed using fluorescent microscope as described under "Experimental Procedures." Space bar ϭ 5 m.

FIG. 4. Subcellular distribution of MEG2 protein in neutro-
phils. Subcellular fractionation of human neutrophils was performed by nitrogen cavitation followed by Percoll gradient centrifugation (see "Experimental Procedures"). Fractions displaying the maximal activity of marker protein for cytosol (cyto), the primary (1°), secondary (2°), and tertiary (3°) granules, and the combined secretory vesicle and plasma membrane fraction (sv/pm) were selected, and an equal amount of protein (10 g) from each fraction was separated by SDS-PAGE and analyzed by immunoblotting with the MEG2-whole molecule antibody.
MEG2. 2 These observations suggest that the activity of MEG2 is selectively modulated during neutrophil stimulation and may regulate tyrosine phosphorylation-dependent signaling events involved in microbicidal processes such as phagocytosis.
The activity of other tyrosine phosphatases such as SHP-1 is regulated by phosphorylation on serine/threonine (17,40) or tyrosine (41)(42)(43) residues. To investigate the potential role of phosphorylation in regulation of the enzymatic activity of MEG2, its phosphorylation status was assessed in quiescent and stimulated cells. Two methods to assess phosphorylation in anti-MEG2 immunoprecipitates were used. First, cells were labeled with [ 32 P]orthophosphate followed by immunoprecipitation of MEG2, SDS-PAGE, and autoradiography. Second, MEG2 immunoprecipitates were immunoblotted with the phosphotyrosine-specific antibody 4G10. However, neither method detected phosphorylation of MEG2 in either resting or stimulated cells (not illustrated).
Mechanism of MEG2 Regulation: Studies with Recombinant MEG2-To study factors regulating MEG2 enzymatic activity, a series of GST fusion protein constructs, including full-length wild type MEG2, mutant C515S MEG2, the C-terminal catalytic domain, and the N-terminal CRALBP noncatalytic domain were engineered and expressed in bacteria. The catalytic activity of these fusion proteins was assayed using the phosphotyrosine analogue p-nitrophenyl phosphate. Fig. 6 indicates that the wild-type MEG2 protein (WT-MEG2) exhibits tyrosine phosphatase activity, approximately half that of SHP-1 when compared on a molar basis. Maximal enzymatic activity of the wild type MEG2 fusion protein occurred at a pH of 6.5 (not shown). Studies of other PTP have shown that loss of the signature motif cysteine will completely abolish the enzymatic activity. Consistent with these observations, mutation of cys-teine-515 to serine (Ser-515) eliminated the activity of wild type MEG2 (Fig. 6).
In addition to its catalytic domain, MEG2 contains an Nterminal domain with significant homology to cellular retinaldehyde binding protein and yeast SEC14p. The function of this domain is not known but, by analogy, may be involved in lipid metabolism, transport, or binding. We hypothesized that the noncatalytic domain regulates the catalytic activity of the phosphatase domain, as has been reported for SHP-1 (44,45) and SHP-2 (46,47). To investigate this possibility, we engineered separate GST constructs of the catalytic and the N-terminal domains. As illustrated in Fig. 6, the specific activity of the isolated catalytic domain (MEG2-Cat) was approximately double the activity of the full-length MEG2 protein on an equimolar basis. This observation suggests that the N-terminal cellular retinaldehyde binding protein (CRALBP) domain of MEG2 exerts an inhibitory effect on the holoenzyme. To explore this possibility further, we repeated the assay of activity of the catalytic domain in the presence of equimolar amounts of the N-terminal domain. However, contrary to our predictions, no inhibitory effect was observed (not shown).
MEG2 Is Activated by Polyphosphoinositides-Because the N-terminal domain of MEG2 has homology to lipid-binding proteins, we postulated that the activity of the phosphatase might be regulated by presence of lipids. Accordingly, we studied the effects of various lipids on the activity of the MEG2 fusion proteins (Fig. 7). In our initial studies, lipids were bound to octyl-Sepharose beads and then incubated with recombinant MEG2. As illustrated by Fig. 7a, the activity of MEG2 was increased more than 3-fold by PI 4,5-P 2 compared with no significant changes with all other lipids tested. PI 4,5-P 2 did not affect the activity of GST-SHP-1 or GST-MEG2 phosphatase domain fusion proteins (not shown), indicating that its effect was specific for the wild type MEG2 protein.
To determine the effects of phospholipids on MEG2 activity in a more physiological system, we prepared mixed liposomes where phosphatidylcholine was the major constituent and other test lipids were present in minor quantities (5% molar). Fig. 7b demonstrates that, under these conditions, PI 4,5-P 2 robustly (ϳ6-fold) and specifically activates MEG2 phosphatase activity. In addition, in this mixed liposome system, the inclusion of PI 4-P also results in a smaller degree of MEG2 activation (ϳ2-fold). By contrast, another anionic lipid, cardio- lipin, did not affect MEG2 activity demonstrating that the affect of phosphatidylinositol phosphates on MEG2 is not simply based on charge interactions.
To test the positional specificity of MEG2 activation by PI 4,5-P 2 , phosphatidylcholine liposomes were prepared containing positional isomers of phosphatidylinositol phosphate (Fig.  7c). In addition to PI 4,5-P 2 and PI 4-P, significant activation of MEG2 phosphatase activity was observed with PI 3,4,5-P 3 . Other isomers tested had only minor and statistically insignificant effects on the activity of MEG2. Finally, PI 4-P, PI 4,5-P 2 , and PI 3,4,5-P 3 , when presented in mixed phosphatidylcholine liposomes, had no significant effect on the activity of the related PTP SHP-1 (not shown), indicating the specificity of the phospholipid activation on MEG2.
MEG2 Associates with Phagosomes-Neutrophil microbicidal activity is dependent on internalization (phagocytosis) of microorganisms followed by fusion of cytoplasmic granules with the phagosome and release of a variety of cytotoxic compounds that degrade the ingested organism. The observations that MEG2 activity is modulated during phagocytosis of opso-nized zymosan and that the phosphatase is associated with granules known to fuse with nascent phagosomes, suggested that MEG2 might be associated with phagosomes. To investigate this possibility, phagosomes were isolated from neutrophils using serum-opsonized magnetic beads. Fig. 8a illustrates a time-dependent increase in the amount of MEG2 associated with the maturing phagosome. Minimal binding of MEG2 to the opsonized beads was observed if phagocytosis was prevented by pretreatment of cells with either latrunculin B, a cytoskeletal disrupting agent, or the PI3K inhibitor wortmannin, both known inhibitors of phagocytosis (Fig. 8b). Other phagosomal markers copurified with MEG2 in the phagosome preparations including LAMP2, and gp91phox (Fig. 8, c and d). These observations also confirm the validity of the phagosomal preparations. DISCUSSION In the present study we have demonstrated that the tyrosine phosphatase MEG2 is expressed by human peripheral blood neutrophils and cultured cell lines of myeloid origin. The initial FIG. 7. Polyphosphatidylinositols stimulate MEG2 activity. a, lipids were coated on beads as described under "Experimental Procedures" and incubated with GST-MEG2 fusion protein. Phosphatase assays were performed using p-nitrophenyl phosphate as described. Values are reported as relative phosphatase activity compared with fusion protein incubated with beads alone. Values represent mean Ϯ S.D. of n ϭ 6 experiments. b and c, dioleoyl-phosphatidylcholine (DOPC) liposomes containing minor amounts of the test lipids (5% molar) were prepared as described under "Experimental Procedures." Liposomes were preincubated with GST-MEG2 fusion protein, following which phosphatase assays were performed using p-nitrophenyl phosphate as described. Values represent mean Ϯ S.D. of experiments (b, n ϭ 4; c, n ϭ 5). For control purposes, the reaction was also performed in the absence of liposomes (buffer alone, b) or with liposomes containing only DOPC (c). In c, values are reported as a percentage of PI 4,5-P 2 stimulation, which corresponds to maximal MEG2 activity. method of detection was degenerate PCR with a strategy that targeted highly conserved regions within the catalytic domain of PTP. This strategy yielded important information inasmuch as it identified expression of four PTPs, including CD45, SHP-1, PTP1B, and MEG2. The number of sequences identified as MEG2 (n ϭ 20) was comparable to the number of sequences identified as CD45 (n ϭ 30) or SHP-1 (n ϭ 26). To the extent that these values reflect the relative abundance of the respective transcripts in neutrophils, the level of expression of MEG2 is comparable to that of CD45 and SHP-1.
Using two distinct techniques, immunofluorescence and cell fractionation, MEG2 was observed to be present within the cytosol as well as in granular fractions, specifically secondary and tertiary granules, and secretory vesicles. This granular localization was unexpected, because the predicted amino acid sequence of MEG2 lacked a leader sequence or a strong hydrophobic stretch of amino acids, suggesting that the protein was cytosolic (21). We speculate that the association of MEG2 with granules is indirect and mediated by interactions with lipids such as PI 4,5-P 2 and PI 3-4,5-P 3 (discussed below) or via membrane-associated protein binding partners. The presence of MEG2 in the granular and vesicular fractions has important implications for the potential functions of the phosphatase. Specifically, in concert with the homology of the N-terminal domain of MEG2 with yeast Sec14P and CRALBP, this granular/vesicular association suggests the possibility that MEG2 functions in signaling or membrane fusion events involved in phagocytosis and/or maturation of the phagosome such as endosomal-phagosomal fusion (37,48). Our observation that the enzymatic activity of MEG2 is modulated during phagocytosis is also consistent with a role for the phosphatase in the signaling events associated with phagocytosis.
It has recently been reported that a recombinant MEG2 fusion protein, heterologously expressed in Jurkat cells, localized to large, round cytoplasmic organelles of uncertain origin (23). We have observed a similar pattern of immunolocalization of endogenous MEG2 in Jurkat cells and primary CD4ϩ Tlymphocytes 3 using the anti-MEG2 antibodies described in this report. These observations highlight differences in the subcellular localization (and potentially function) of MEG2 in lymphoid and myeloid cells. The precise localization of MEG2 in lymphoid cells and the significance of these "cytoplasmic organelles" in lymphoid cells remain unclear and require additional investigation.
The noncatalytic (N-terminal) domain of MEG2 shares significant homology with the proteins CRALBP and yeast SEC14p. Because both of these proteins are known to bind hydrophobic ligands, it has been suggested that the function of MEG2 may also involve the binding of lipids (21). Indeed, in the current manuscript, we have provided definitive evidence that the activity of MEG2 is stimulated by polyphosphoinositides. Specifically, using two distinct assay systems, PI 4,5-P 2 consistently enhanced the activity of MEG2 by 5-to 8-fold. We also observed that the structurally related phospholipids, PI 3,4,5-P 3 and, to a lesser extent, PI 4-P, also stimulate MEG2 activity. However, this activation is not simply due to the anionic nature of PI 4,5-P 2 , because another highly anionic phospholipid, cardiolipin, failed to activate MEG2. We speculate that MEG2 associates with phospholipids present in the membranes of granular or vesicular compartments via its noncatalytic N-terminal domain leading to spatially restricted activation. Future studies will determine the significance of MEG2-lipid interactions in vivo and elucidate which polyphosphoinositides are of physiological relevance.
A functional role for the noncatalytic domain of MEG2 is further suggested by the studies that compared the activity of the full-length GST-MEG2 construct with that of the C-terminal phosphatase domain. These studies revealed that the catalytic activity of the single-domain construct was significantly enhanced when compared with full-length MEG2 suggesting that the noncatalytic domain inhibits the activity of the phosphatase domain. Based on these observations, we propose a model for the functional regulation of MEG2. In the resting state, MEG2 assumes a conformation in which interactions between the two domains restrain the catalytic potential of the phosphatase domain. Interaction of the N-terminal domain with membrane phospholipids such as PI 4,5-P 2 or PI 3-4,5-P 3 or other binding partners releases its autoinhibitory function resulting in enhanced catalytic activity. It should be recalled that, in experiments where the N-terminal domain fusion was mixed in equimolar amounts with the C-terminal catalytic domain fusion, no inhibition of the C-terminal domain catalytic activity was observed. This may reflect the inability to reconstruct complex intermolecular interactions with separate fusion proteins in vitro in dilute solution. We also investigated whether inter-domain interactions required the presence of lipids, but no effect of the N-terminal domain was observed on the catalytic activity of the C-terminal domain after the addi-3 J. M. Kruger and G. P. Downey, unpublished observations. FIG. 8. MEG2 associates with nascent phagosomes. Neutrophils were incubated with serum-opsonized iron oxide beads, which were allowed to phagocytose for the indicated times. Cells were lysed, and beads were recovered by magnetic isolation. Samples were separated by SDS-PAGE and probed for the indicated proteins using Western blotting. For control purposes and to inhibit phagocytosis, neutrophils were pretreated for 1 h (37°C) with either wortmannin (100 nM) or latrunculin B (1 M) and collected after a 20-min incubation with the coated beads. a, affinity-purified anti-MEG2 whole molecule antibody; b, anti-MEG2 antibodies were preincubated with the 2-fold molar excess of the full-length recombinant GST fusion of MEG2; c, monoclonal anti-LAMP2 antibody; d, monoclonal anti-gp91phox antibody.
tion of a variety of lipids. It is therefore possible that the conformations assumed by our fusion proteins do not sufficiently resemble the physical states of the domains within the native molecule that are required for inhibitory interactions to occur. Further studies are required to explore this possibility.
The role of specific protein-tyrosine phosphatases in myeloid leukocyte function is beginning to be defined. CD45 for example has been demonstrated to be critical for the maintenance of integrin-mediated adhesion in bone marrow-derived macrophages (49). Activation of the Src family members p56 Lck and p59 Fyn is dependent on a CD45-mediated dephosphorylation of these kinases (50). Overexpression of SHP-1 in established myeloid cell lines has been shown to suppress cell growth (51), and SHP-1 has been shown to associate with and regulate the function of a number of growth factor receptors, including the epidermal growth factor, platelet-derived growth factor, Kit, and interleukin-3 receptors (52). Recent studies from our laboratory have shown that SHP-1 negatively regulates adhesion and the oxidative burst in myeloid cells in response to PMA stimulation (53). In light of our evidence presented here that MEG2 associates with nascent phagosomes, is catalytically regulated during phagocytosis, is activated by PI 4,5-P 2 or PI 3-4,5-P 3 , we speculate that MEG2 participates in regulation of signaling or membrane fusion events involved in phagocytosis. These observations have potentially important implications for our understanding of the molecular mechanisms regulating phagocytosis, an essential antimicrobial function of neutrophils as part of the innate immune system. It is also possible that PTP such as MEG2 regulate other processes involving membrane fusion such as exocytosis or endocytosis. Therefore, further elucidation of the functions of MEG2 may have important ramifications for our understanding of disorders characterized by inflammatory tissue damage such as arthritis and ischemiareperfusion injury, (54 -57) the pathogenesis of which involves release of leukocyte-derived cytotoxic compounds. The possibility that PTPs such as MEG2 regulate leukocyte activation suggests potential therapeutic targets to restrain aberrant PMN responses in systemic inflammatory disease.