Functional Involvement of PTP-U2L in Apoptosis Subsequent to Terminal Differentiation of Monoblastoid Leukemia Cells*

A large family of protein tyrosine phosphatases (PTPs) bidirectionally regulate intracellular signaling pathways by reversing agonistic or antagonistic phosphorylation events derived from the action of protein tyrosine kinases. Receptor-like PTPPTP-U2 is expressed during phorbol ester-induced differentiation of monoblastoid leukemia U937 cells. We found that the shorter isoform, PTP-U2S, was expressed at an earlier phase in the course of differentiation and the longer isoform, PTP-U2L, was induced at a later phase. In the presence of 12-O-tetradecanoylphorbol-13-acetate, ectopic expression of PTP-U2L in U937 cells enhanced several characteristics of terminally differentiated cells. Most striking was that PTP-U2L enhanced apoptosis of the differentiated cells, which was only partially inhibited by caspase inhibitor Z-Asp-CH2-DCB. The catalytically inactive mutant PTP-U2L(C → S) still retained the ability to enhance the differentiation but retained the ability to enhance the following apoptosis of the cells to a lesser extent. These data indicate a functional involvement of PTP-U2L in apoptosis subsequent to terminal differentiation of U937 cells. Since terminally differentiated blood cells often undergo apoptosis, the data also suggest that PTP-U2L might be involved in physiological turnover of hematopoietic cells in vivo.

in their catalytic domains, these PTP isozymes exhibit diverse structures outside the catalytic domains, suggesting involvement of the PTPs in each isozyme-specific function.
In fact, recent studies have demonstrated specific functions of PTP isozymes in physiological events (see for review, Ref. 7). For example, even among structurally similar PTP isozymes, entirely distinct functions have been documented: SHP1 and SHP2 are both non-receptor PTPs with two tandem Src homology 2 (SH2) domains in their amino-terminal regions (8). SHP1 negatively regulates hematopoietic signaling pathways (9 -11). At the pathological level, genetic failure in the shp1 locus leads to autoimmune disease with hyperproliferation of hematopoietic cells (12,13). On the other hand, SHP2 is required for mitogenic signal transduction from receptor tyrosine kinases to the Ras/mitogen-activated protein kinase (MAPK) pathway (14 -17). These observations suggest that other PTP isozymes, even if they are structurally related through entire sequences, also have distinct functions although some redundancy might exist. Meanwhile, a novel PTP isozyme PTEN/MMAC1/TEP1 has been identified as a candidate tumor suppressor gene (18 -20), still suggesting a functional involvement of certain PTPs in anti-oncogenesis.
Previous studies, including ours, suggest a functional involvement of PTPs in terminal differentiation of myeloid leukemia cells. Pharmacological differentiation of leukemia cells is often accompanied by an increase in cellular PTP activity (21)(22)(23)(24). Meanwhile, such differentiation is often inhibited by sodium orthovanadate, a PTP inhibitor (25). Previously, we identified 13 PTP gene fragments from the differentiated monoblastoid leukemia U937 cells by employing the reverse transcription-polymerase chain reaction strategy (24). Among them, gene expression for four isozymes, including PTP-U2 (also known as GLEPP1/PTP) (26 -28), PTP-MEG2 (29), P19-PTP (30), and PTP-U1/DEP-1/HPTP (24,31,32), is induced during monocytic and granulocytic differentiation of myeloid leukemia cells (33). 2 PTP-U2 is a receptor-like PTP with a single transmembrane domain and a single intracellular catalytic domain (26). Physiologically, the PTP-U2 gene is expressed as tissue-specific isoforms, which are probably due to alternative splicing of the transcript. The extracellular domain of the longest isoform contains 14 putative N-glycosylation sites and eight repeats of a fibronectin-type III-like motif. Especially, in U937 cells, PTP-U2 gene expression is greatly induced by phorbol ester, 12-O-tetradecanoylphorbol-13-acetate (TPA), which arrests growth and induces monocytic differentiation of U937 cells (24). Furthermore, PTP-U2 gene expression is also induced or enhanced during various types of differentiation of myeloid and erythroid leukemia cells, including HL-60, HEL, and K562.
Other myeloid leukemia THP-1 cells, which often differentiate spontaneously even without inducers, express a relatively high amount of PTP-U2 transcript. On the contrary, TPA does not induce or enhance PTP-U2 gene expression in solid tumor cells that do not differentiate upon treatment with TPA (26). Thus, we supposed that PTP-U2 would be functionally involved in drug-induced differentiation of U937 cells. However, the precise role of PTP-U2 in terminal differentiation has yet to be determined.
In this study, we investigated the function of the longest PTP-U2 isoform (PTP-U2L) during TPA-induced differentiation and subsequent cell death of U937 cells. We found that PTP-U2L enhances the apoptotic event that follows TPA-induced differentiation of U937 cells. The physiological importance of PTP-U2L function is also discussed.

MATERIALS AND METHODS
Cell Culture and Differentiation Induction-Human monoblastoid leukemia U937 cells were grown as described previously (24). A phorbol ester-resistant U937 variant, UT16, was established and characterized as described (33). Human embryonal kidney 293T cells were kindly provided by Dr. T. Suzuki (University of Tokyo, Tokyo, Japan). Cellular differentiation was induced by treatment with 5 ng/ml TPA (Sigma) and characteristic changes in cellular morphology, development of adherence to plastic substratum, and cell surface expression of CD14 or CD11b glycoproteins were monitored. Expression of CD14 or CD11b was determined by flow cytometry, as described previously (33).
Preparation of Cellular Lysates-Cells were washed with ice-cold phosphate-buffered saline (PBS) and resuspended in 0.5 ml of buffer E consisting of 50 mM HEPES, pH 7.0, 2 mM EDTA, 2 mM EGTA, 10 mM 2-mercaptoethanol, 1 mM phenylmethylsulfonyl fluoride, 10 g/ml aprotinin, and 2% Triton X-100 (34). Differentiated cells were harvested by scraping in buffer E after PBS washing. The lysate was incubated with occasional mixing at 4°C for 1 h, and centrifuged (10,000 ϫ g for 10 min at 4°C) to obtain the solubilized fractions. The protein concentration in the extracts was determined using a Protein Assay System reagent (Bio-Rad).
Western Blot Analysis-The samples (30 g of protein) were separated by SDS-polyacrylamide gel electrophoresis and electrophoretically transferred to nitrocellulose membranes (Schreicher & Schuell, Dassel, Germany). The membranes were soaked in 10% skim milk, PBS at room temperature for 1 h and further incubated for 1 h with an affinity-purified rabbit antibody raised against PTP-U2 carboxyl-terminal 15-amino acid residues (26) or mouse anti-T cell PTP antibody (Oncogene Science, Uniondale, NY). The membranes were then washed with 0.1% Tween 20, 0.5% skim milk, PBS and incubated for 1 h with horseradish peroxidase-conjugated goat anti-rabbit or anti-mouse immunoglobulin (Amersham International, Buckinghamshire, United Kingdom). After washing the membranes extensively with 0.1% Tween 20, PBS, the specific signals were detected on X-Omat AR films (Kodak, Rochester, NY) using an ECL detection system (Amersham Int.).
Vector Constructions-The full-length cDNA for PTP-U2L (26) was cloned into a mammalian expression vector pCR3 (Invitrogen, Carlsbad, CA), and the resulting plasmid was designated as pCR3/PTP-U2L. pCR3/PTP-U2L(C 3 S), the expression vector for the catalytically inactive mutant of PTP-U2L (point mutated at Cys 1136 to Ser 1136 ) was constructed with pCR3/PTP-U2L as a template and a Chameleon mutagenesis kit (Stratagene, La Jolla, CA), according to the instruction manual. DNA sequences of the constructs were confirmed by an ABI PRISM Dye Primer Cycle Sequencing Kit (Applied Biosystems, Chiba, Japan) with a Perkin-Elmer DNA thermal cycler and an ABI PRISM 310 Genetic Analyzer.
DNA Transfection-Twenty micrograms of pCR3, pCR3/PTP-U2L, or pCR3/PTP-U2L(C 3 S) were transfected into 10 7 U937 cells by electroporation using a Gene-Pulser (Bio-Rad). After the recovery culture of 48 h, transfectants were selected by treatment with 0.8 mg/ml geneticin (Sigma), and the resistant cells were cloned by limiting dilution. Expression of PTP-U2L or PTP-U2L(C 3 S) protein was monitored by Western blot analysis, as described above, and the positive clones were used for subsequent experiments.
Measurement of Apoptotic Cells-Apoptosis was monitored through characteristic changes in cellular morphology, externalization of phosphatidylserine (37,38), and the appearance of chromosomal DNA fragmentation (39). Externalization of phosphatidylserine was detected by annexin V-fluorescein isothiocyanate (Kamiya Biomedical, Seattle, WA) staining, as described previously (40). DNA fragmentation was measured, as described previously (41). In brief, PBS-washed cells (1 ϫ 10 6 cells) were resuspended in 20 l of solution A (50 mM Tris-Cl, pH 8.0, 10 mM EDTA, 0.5 mg/ml proteinase K) and incubated at 50°C for 90 min. Then, 10 l of solution B (10 mM Tris-Cl, pH 7.6, 15 mM NaCl, 1 g/ml ribonuclease A) was added and the mixture was further incubated at 50°C for 90 min. Finally, DNAs in the reaction were resolved by agarose gel electrophoresis (2% agarose, TBE buffer). Propidium iodide staining of cellular DNA was carried out as follows: cells were washed with ice-cold PBS, resuspended in 70% ethanol for 30 min, and treated with 2 mg/ml ribonuclease A/PBS at 37°C for 30 min. After washing in PBS, cells were resuspended in 50 g/ml propidium iodide (Sigma), PBS, passed through a 70-m nylon mesh filter, and analyzed by using a FACScan system (Beckton-Dickinson, Franklin Lakes, NJ).

PTP-U2 Enzyme Activation during TPA-induced Differentiation of U937
Cells-When U937 cells were continuously treated with 5 ng/ml TPA, they differentiate into monocytes and macrophages (42,43). As shown in Fig. 1A, the cell surface expression of CD11b glycoprotein, one marker for monocytic differentiation, gradually increased during the course of differentiation. Similarly, CD14 glycoprotein, another marker, was also induced (data not shown). Meanwhile, adhesion of cells to a plastic substratum was one of the most drastic changes observed in the differentiated U937 cells. While the undifferentiated U937 cells were grown in suspension, the TPA-treated cells adhered to a culture dish and the percentage of the adherent cells increased up to 70% in a time-dependent manner (Fig. 1B).
To investigate the kinetics of PTP-U2 activation, we first monitored the PTP-U2 enzyme activity in U937 cells with an anti-PTP-U2 immunocomplex PTP assay (see "Materials and Methods"). As shown in Fig. 2A, PTP-U2 enzyme activity against phosphotyrosyl substrate 1 (PS1; END(pY)INASL; derived from a highly conserved region of T cell PTP) (35) and PS2 (DADE(pY)LIPQQG; corresponding to the autophosphorylation site of epidermal growth factor receptor) (36) was substantially induced during differentiation, and this enzyme activation was parallel to the appearance of differentiation markers (Fig. 1). Maximal activation of the enzyme was observed at 72 h (10.5-and 4.5-fold activation against PS1 and PS2, respective- ly). A nonspecific PTP inhibitor, sodium orthovanadate (1 mM), efficiently inhibited the activity, indicating that the release of free phosphate from the phosphotyrosyl substrates was due to the catalytic activity of the PTP. The control immunocomplex precipitated without anti-PTP-U2 antibody did not exhibit any trace of PTP activity (data not shown). On the other hand, UT16 cells, TPA-resistant variant of U937 (33), did not respond to TPA with the PTP-U2 enzymatic activation.
PTP-U2L Isoform Is Induced as the Late Event of TPAinduced Differentiation of U937 Cells-To determine whether enzyme activation of PTP-U2 during TPA-induced differentiation was due to enhanced expression of PTP-U2 protein, we analyzed the amounts of PTP-U2 protein using Western blot analysis. As shown in Fig. 2B, while expression of PTP-U2 protein was not detected after treatment with TPA for 0 -6 h, the 70-kDa protein that we reported to be the shorter isoform of PTP-U2 (26) was induced at 9 -72 h. Interestingly, further incubation with TPA led to reduction of the shorter isoform expression and induction of a larger protein at 220 kDa. The size of this protein corresponded to that of the PTP-U2 larger isoform ectopically expressed in 293T and U937 cells (Fig. 2B, right panel). In this report we refer to the 70-and 220-kDa isoforms as PTP-U2S and PTP-U2L, respectively. Employing the reverse transcriptase-polymerase chain reaction, we also detected gene expression of PTP-U2L in U937 cells treated with TPA for 96 h but not in the control cells (data not shown). On the other hand, the amount of T cell PTP protein, nonreceptor PTP isozyme, was down-regulated during the differentiation, indicating differential expression of PTPs in each isozyme-specific manner. In UT16 cells, neither PTP-U2S nor PTP-U2L were induced by treatment with TPA (data not shown).
Ectopic Expression of PTP-U2L Enhances the Growth Inhibitory Effect of TPA on U937 Cells-To elucidate the functional involvement of PTP-U2L in TPA-induced differentiation of U937 cells, we transfected the expression vector for the PTP-U2L gene (under control of a cytomegalovirus immediate-early promoter) into U937 cells. Resulting transfectants named as U21L2, U21L4, and U21L6 constitutively expressed PTP-U2L protein (the representative data are shown in Fig. 3A) and displayed normal growth rates, compared with parental U937 cells and mock transfectants (data not shown). Under normal growth conditions, we did not find significant morphological differences between the cell lines (Fig. 4, A-D). Upon treatment with 5 ng/ml TPA for 72 h, however, all the PTP-U2L transfectants exhibited higher induction of CD14 glycoprotein than control transfectants, such as mock4 (p Ͻ 0.01 or p Ͻ 0.025, evaluated by an unpaired Student's t test, Fig. 3B). Furthermore, the PTP-U2L transfectants exhibited higher sensitivity to TPA than the control cells (Fig. 3C). For example, sensitivity of the PTP-U2L transfectants to 1 ng/ml TPA was about 6-fold higher than that of mock4.
There were also morphological differences between the TPAtreated control and PTP-U2L transfectants (Fig. 4, E-H). In the case of U937 and the mock transfectants, the differentiated cells adhered to the culture dish, but still retained their round shape and looked phase-bright under phase-contrast microscopy. On the other hand, there were more drastic changes in the PTP-U2L transfectants. These cells adhered to the culture dish to such an extent that we could not distinguish them from the other kinds of adherent cells. Some of them lost the globular shape and showed a flat and polygonal morphology. The borders of the cell-cell contact were also observed. More interestingly, cellular fragments, probably due to the formation of apoptotic bodies, were often observed in the culture of TPA-treated PTP-U2L transfectants. This finding suggests that hypersensitivity to TPA of PTP-U2L transfectants (Fig. 3C), was due to enhanced susceptibility of the cells to TPA-induced apoptosis.
Hypersensitivity to TPA of PTP-U2L Transfectants Is Associated with an Enhanced Rate of Apoptotic Cell Death-To determine whether TPA could induce apoptosis of PTP-U2L transfectants, we analyzed DNA content of each clone during TPA-induced differentiation. Upon treatment with 5 ng/ml TPA for 48 -96 h, mock4 cells arrested the cell cycle at the G 0 /G 1 phase (Fig. 5A). Subsequent to this growth arrest, a trace of a subdiploid (apoptotic) fraction was observed at 192 h (day 8). Under the same culture conditions, U21L4 and U21L6 cells exhibited the G 0 /G 1 arrest at 24 h (at least 24 h earlier than the mock4 cells). Furthermore, drastic accumulation of the subdiploid fraction was observed at 96 h. Like U21L4 and U21L6 cells, U21L2, but not mock1 and U937 cells, showed rapid accumulation of the G 0 /G 1 fraction (data not shown) and a drastic shift of the population to the subdiploid fraction (Fig.  5B). When a cell undergoes apoptosis, chromosomal DNA is degraded into nucleosome-size fragments, that is a most characteristic property of apoptotic cell death. We observed TPAinduced fragmentation of the chromosomal DNA in the PTP-  Fig. 1, and the cellular lysates were prepared. A, PTP-U2 enzyme activity. PTP-U2 was immunoprecipitated with anti-PTP-U2 antibody, and PTP activity in the immunocomplex was measured using the Tyrosine Phosphatase Assay System (Promega). PS1, phosphorylated substrate END-(pY)INASL (35); PS2, phosphorylated substrate DADE(pY)LIPQQG (36). Also shown are the activity in the presence of 1 mM sodium orthovanadate and the activity of TPA-treated UT16 cells. B, expression of PTP-U2 and T cell PTP proteins. The cellular lysates were electrophoresed and immunoblotted with either anti-PTP-U2 or anti-T cell PTP antibody. Also shown are 293T and U937 cells that are transiently or stably transfected pCR3/PTP-U2L as positive controls.
At the early phase of apoptosis, the plasma membrane phosphatidylserine is externalized to the cell surface (37,38). Consistent to this event, TPA-treated PTP-U2L transfectants significantly externalized phosphatidylserine but the mock4 cells did not (Fig. 5D). This also means that TPA-induced differentiation itself does not lead to externalization of phosphatidylserine. Taken together, these data indicate that PTP-U2L enhances TPA-induced cell death of U937 cells, and the mode of cell death seems to be apoptosis.

TPA-induced Apoptosis of PTP-U2L Transfectants Is Only Partially Inhibited by a Caspase
Inhibitor-Caspase family proteases (originally known as ICE-like proteases) are downstream regulators of apoptosis (44,45). We have previously reported that benzyloxycarbonyl-Asp-CH 2 OC(O)-2,6-dichlorobenzene (Z-Asp-CH 2 -DCB), an aspartate-based inhibitor of caspases, inhibited apoptosis of U937 cells induced by various stimuli (46). So, we examined the effect of Z-Asp-CH 2 -DCB on TPA-induced apoptosis of PTP-U2L transfectants. As shown in Fig. 6, a topoisomerase I inhibitor, camptothecin, drastically induced apoptosis of U937 cells, but it was completely blocked in the presence of 50 g/ml Z-Asp-CH 2 -DCB. On the other hand, the inhibitory effect of Z-Asp-CH 2 -DCB on TPA-induced apoptosis of PTP-U2L transfectants (U21L clones) was also detected, but it was more moderate than on the camptothecininduced apoptosis. In the presence of Z-Asp-CH 2 -DCB, TPAinduced apoptosis of U21L clones was only 35% inhibited.
PTP-U2L Catalytic Activity Is Required for Maximal Induction of TPA-induced Apoptosis of U937 Cells-To determine whether apoptosis enhancement by PTP-U2L required catalytic activity of the enzyme, we constructed a catalytically inactive mutant PTP-U2L(C 3 S) gene, in which the codon for an essential cysteine residue (TGC) in the conserved active-site HCxxGxxRS(T) motif (for review, see Ref. 47) was point mutated to that for serine (AGC). The mutant PTP-U2L(C 3 S) gene was transfected into U937 cells, and the resulting U2L/ CS20 and U2L/CS21 were representative clones that stably expressed the mutant PTP-U2L(C 3 S) protein (Fig. 7A). Under normal growth conditions, growth rates of PTP-U2L(C 3 S) transfectants were comparable to those of wild-type PTP-U2L and mock transfectants (data not shown). Intracellular amounts of the ectopic PTP-U2L protein in the mutant transfectants were often higher than those in wild-type transfectants; the representative result is shown in Fig. 7A. Upon treatment with 5 ng/ml TPA for 72 h, U2L/CS20 and U2L/CS21 cells still differentiated into monocytic lineage and adhered to the culture dish. As shown in Fig. 8, the differentiated cells adhered extensively to the dish and showed flat and polygonal morphology, which was similar to the phenotypic changes observed in the TPA-treated wild-type PTP-U2L transfectants (Fig. 4). These observations indicate that PTP-U2L(C 3 S), despite the loss of its catalytic activity, retained its accelerative effect on TPA-induced differentiation of U937 cells. We further examined the susceptibility of U2L/CS20 and U2L/CS21 cells to TPA-induced apoptosis. As shown in Fig. 7B, PTP-U2L(C 3 S) still enhanced apoptosis of the transfectant to some extent, as compared with mock4 cells, which differentiated and was strictly arrested at the G 0 /G 1 phase of the cell cycle. However, the effect of PTP-U2L(C 3 S) on apoptosis enhancement was considerably reduced, compared with wildtype PTP-U2L (p Ͻ 0.01, Fig. 7, B and C; see also Fig. 8, C and  D). This observation suggests that PTP-U2L catalytic activity is required for maximal enhancement of TPA-induced apoptosis by PTP-U2L. These findings were repeated with other PTP-U2L(C 3 S) transfectants with similar results (Fig. 7C). DISCUSSION PTP-U2L transfectants exhibited progressive adhesion to the culture dish (Fig. 4). PTP-U2L(C 3 S) transfectants also retained this ability (Fig. 8), suggesting that the extracellular domain of PTP-U2L, consisting of eight fibronectin type III-like repeats, directly functioned as an adhesion molecule in a catalytic activity-independent manner (Fig. 9). Similar to these observations, previous reports have revealed that type II receptor-like PTPs, RPTP and RPTP, which have extracellular domains consisting of four fibronectin type III-like repeats, a single Ig-like domain, and an NH 2 -terminal MAM domain, mediate homophilic cell-cell adhesion even without the cytoplasmic domain (48 -50). Since the PTP-U2L transfectants were still in suspension in the absence of TPA, the extracellular domain of PTP-U2L would enhance cellular adhesion in cooperation with other molecules which expression was induced by TPA.
Among other characteristics of TPA-induced differentiation, CD14 expression was enhanced and the G 0 /G 1 arrest occurred earlier in the cell cycle of PTP-U2L transfectants than in that of mock transfectants (Figs. 3B and 5A). On the other hand, CD11b up-regulation by TPA was comparable with PTP-U2L and mock transfectants (data not shown). These observations indicate that there are parallel but distinct signaling pathways for TPA-induced differentiation and that PTP-U2L specifically participates in some of those pathways.
While PTP-U2L enhanced TPA-induced apoptosis of U937 cells (Figs. 4 and 5), it seems to be specific to terminal differentiation of leukemia cells or at least be limited to certain situations. First, there are normal tissues that constitutively express PTP-U2L (26). Among them are kidney and brain, in which podocyte and olfactory bulb exhibit dominant expression of PTP-U2L, respectively (27,51). Second, when 293T cells were transfected with the PTP-U2L gene, TPA sensitivity of the resulting transfectants was comparable to that of the mock cells. 2 Third, PTP-U2L was induced during TPA-induced differentiation ( Fig. 2B) but not during camptothecin-or etoposide-induced apoptosis of U937 cells (data not shown).
Apoptosis enhancement by PTP-U2L required exposure of the cells to TPA (Figs. 4 and 5). While TPA is a potent activator of protein kinase C (52) and causes activation of the further downstream ERK/MAPK cascade (53), 1-oleoyl-2-acetylglycerol, another protein kinase C activator, does not induce differentiation of myeloid leukemia cells (54). Consistently, 1-oleoyl-2-acetylglycerol did not induce apoptosis of the PTP-U2L transfectants at all despite ERK/MAPK activation (data not shown). These data suggest that apoptosis enhancement by PTP-U2L requires a differentiation-associated factor induced by TPA but not by 1-oleoyl-2-acetylglycerol (Fig. 9).
The result in Fig. 6 suggests that Z-Asp-CH 2 -DCB-sensitive caspases are, in part, involved in the PTP-U2L-mediated apoptosis of differentiated U937 cells. Strictly speaking, the inhibitory effect of Z-Asp-CH 2 -DCB was only partial, and there could be other mechanisms of PTP-U2L-mediated apoptosis that are independent of Z-Asp-CH 2 -DCB-sensitive caspases. In fact, caspase-3 activation was significantly observed during camptothecin-induced apoptosis of U937 cells whereas that was only marginal during PTP-U2L-mediated apoptosis of the cells. 2 PTP-U2L(C 3 S) exhibited an intermediate effect on TPAinduced apoptosis, as compared with mock and wild-type PTP-U2L (Figs. 7, B and C, and 8). These data suggest that maximal enhancement of apoptosis by PTP-U2L requires its catalytic activity. In that sense, it is still possible that overexpression of PTP-U2L itself could elicit apoptosis in U937 cells. In fact, the wild-type PTP-U2L transfectants generally expressed lower amounts of PTP-U2L protein than the mutant types did (Fig.  7A). One explanation for this circumstance is that high producer clones for the wild-type PTP-U2L might die by apoptosis during the cloning procedures. Consistent with this idea, transfection experiments of wild-type PTP-U2L into TPA-resistant UT16 cells have not worked so far, whereas it has been easy to obtain the PTP-U2L(C 3 S)-transfected UT16 clones. 2 As UT16 cells have been established by prolonged exposure of U937 cells to TPA (33), differentiation-associated factor(s), required for the PTP-U2L-mediated and its catalytic activity dependent apoptosis (see above), seem to have irreversibly up-regulated in UT16 cells. If this were the case, introducing the wild-type PTP-U2L gene would kill the cells by apoptosis.
In the presence of a caspase inhibitor (Z-Asp-CH 2 -DCB), only 35% of the wild PTP-U2L-mediated apoptosis was inhibited (Fig. 6), whereas 53% of the C 3 S mutant-mediated apoptosis was inhibited (data not shown). These observations suggest that the PTP-U2L catalytic activity dependent apoptosis, which is abolished by the C 3 S mutation, represents the caspase-insensitive cell death. Meanwhile, it has been also suggested that PTP(C 3 S) mutants in some case might exhibit a dominant positive effect by trapping their substrates (7), and determination of specific substrate(s) for PTP-U2L is our future interest to understand the downstream pathway of PTP-U2Lmediated signaling.
PTP-U2 is expressed as at least three different isoforms in a tissue-specific manner (26). In addition, more shorter transcripts than PTP-U2S have been reported as PTP, the murine homolog of PTP-U2 (28). Quiescent macrophages express PTP, which is down-regulated when the cells are activated by colony-stimulating factor-1 (28). During the course of TPAinduced differentiation of U937 cells, PTP-U2S was induced as an early event whereas PTP-U2L was as a late event (Fig. 2B). Given the kinetics of PTP-U2S induction, one could speculate that the shorter isoforms, such as PTP-U2S and PTP, might play a role in triggering commitment to differentiation or in maintaining differentiated phenotypes of the cells (see the scheme shown in Fig. 9). Determination of PTP-U2S whole cDNA structure would afford transfection experiments to elucidate the precise role of PTP-U2S.
Terminally differentiated hematopoietic cells often die by apoptosis (55). Here we have demonstrated the first instance of PTP involvement in apoptosis subsequent to terminal differentiation of leukemia cells. The present data also suggest a functional involvement of PTP-U2L in the physiological events. For example, PTP-U2L might participate in turnover of peripheral myelocytes by eliminating them by apoptosis. If this is so and there are no functionally redundant PTP isozymes, failure in PTP-U2L function could have serious outcomes, such as hematopoietic disorders. The PTP-U2 gene is localized to chromosome 12p13.2-p13.3 (26), a region of interest because it is a common site of chromosomal abnormalities in such malignant proliferations as eosinophilia (56). Meanwhile, PTP-U2 is expressed as various isoforms in a tissue-specific manner. Thus, determination of the physiological role of PTP-U2 awaits the generation of PTP-U2 knock-out animals and cell lines.