Nuclear Localization of Non-receptor Protein Tyrosine Phosphatase ϵ Is Regulated by Its Unique N-Terminal Domain

Nuclear Localization of Non-receptor Protein Tyrosine Phosphatase ϵ Is Regulated by Its Unique N-Terminal Domain

Experimental Cell Research 281, 182–189 (2002) doi:10.1006/excr.2002.5661 Nuclear Localization of Non-receptor Protein Tyrosine Phosphatase ⑀ Is Regu...

262KB Sizes 0 Downloads 28 Views

Experimental Cell Research 281, 182–189 (2002) doi:10.1006/excr.2002.5661

Nuclear Localization of Non-receptor Protein Tyrosine Phosphatase ⑀ Is Regulated by Its Unique N-Terminal Domain Judith Kraut, Gloria Volohonsky, Hila Toledano-Katchalski, and Ari Elson 1 Department of Molecular Genetics, Weizmann Institute of Science, Rehovot 76100, Israel

Precise subcellular localization is an important factor in regulation of the functions of protein tyrosine phosphatases. The non-receptor form of protein tyrosine phosphatase ⑀ (cyt-PTP⑀) can be found in cell nuclei, among other cellular locations, while p67 PTP⑀, a naturally occurring isoform which lacks the 27 N terminal residues of cyt-PTP⑀, is exclusively cytosolic. Using deletion and scanning mutagenesis we report that the first 10 amino acid residues of cytPTP⑀, in particular residues R4, K5, and R9, are critical components for its nuclear localization. We also establish that increased oxidative stress enhances accumulation of cyt-PTP⑀ in cell nuclei. Of the four known protein forms of PTP⑀, cyt-PTP⑀ is the only one which includes the extreme N-terminal sequence containing R4, K5, and R9. The role of the unique N terminus of cyt-PTP⑀ is therefore to regulate its subcellular localization. The existence of naturally occurring forms of PTP⑀ which lack this sequence and which are generated by translational and posttranslational mechanisms, suggests that nuclear localization of cyt-PTP⑀ can be actively regulated by cells. © 2002 Elsevier Science (USA) Key Words: tyrosine phosphatase; nuclear localization signal; oxidative stress; hydrogen peroxide.

INTRODUCTION

Accurate phosphorylation of tyrosine residues in proteins is well established as a vital mechanism by which protein structure and function are regulated in a reversible manner. Tyrosine dephosphorylation is carried out by members of the protein tyrosine phosphatase (PTP) superfamily, which are genetically, structurally, and functionally distinct from the tyrosine kinases whose activities they counter [1– 4]. In recent years it has become evident that subcellular localization of nonmembranal PTPs is crucial in regulating their physiological roles [5, 6]. Examples of this phenomenon include PTPs such as the non-receptor-type PTP1B, PTP-PEST, and STEP, whose proteolytic cleavage in platelets and T-cells alters 1 To whom correspondence and reprint requests should be addressed. Fax: ⫹972-8-934-4108. E-mail: [email protected].

0014-4827/02 $35.00 © 2002 Elsevier Science (USA) All rights reserved.

their subcellular localization patterns and can result in their activation [7–12]. Additional examples are SHP1 and SHP2, whose SH2 domains mediate their recruitment to activated growth factor receptors in whose vicinity they act [13]. The protein tyrosine phosphatase ⑀ (PTP⑀) subfamily contains four distinct protein species, all products of a single gene [14, 15]. The receptor-type form of PTP⑀ (RPTP⑀) [16, 17] is an integral membrane protein, which has been linked to assisting with mouse mammary tumorigenesis [17, 18; Gil-Henn and Elson, submitted] and to down-regulation of insulin receptor signaling in cultured cells [19, 20]. The non-receptor form of PTP⑀ (cyt-PTP⑀), which is expressed from the PTPR⑀ gene by use of an alternative promoter [15, 21, 22], is predominantly cytoplasmic but can also be detected at the cell membrane and within the cell nucleus [23]. The absence of this form from Schwann cells of young PTP⑀-deficient mice increases phosphorylation and activity of the delayed-rectifier, voltage-gated potassium channels Kv1.5 and Kv2.1 and correlates with myelination defects in sciatic nerves in these mice [24]. p67 PTP⑀, which is produced by internal initiation of translation from PTP⑀ mRNAs, and p65 PTP⑀, which is produced by calpain-mediated proteolytic processing of the larger PTP⑀ forms, are N-terminally truncated forms of PTP⑀ and are exclusively cytoplasmic [23, 25]. The distinct N-termini of the four PTP⑀ proteins dictate their unique subcellular localization patterns and affect their physiological roles; this is exemplified in the significantly reduced ability of the cytosolic p67 and p65 PTP⑀ proteins to dephosphorylate the integral membrane potassium channel Kv2.1 [23, 25]. Other roles for PTP⑀ include suppression of endothelial cell proliferation [26], ensuring proper function of mouse macrophages [27], and inhibition of JAK-STAT signaling in M1 leukemia cells in response to various cytokines [28, 29]. p67 PTP⑀ can be produced by initiation of translation from cyt-PTP⑀ mRNA at an alternative ATG initiation codon located 81 bp downstream of the initiator ATG for full-length cyt-PTP⑀ [23]. As a result, the only structural difference between the cyt-PTP⑀ and p67

182

NUCLEAR LOCALIZATION OF PTP⑀

PTP⑀ proteins is the absence of a stretch of 27 amino acid residues from the N-terminus of the latter. The absence of this region from p67 correlates with the inability of p67 to associate with cell membranes and with its absence from cell nuclei [23], suggesting that the N-terminal domain of cyt-PTP⑀ regulates the precise subcellular localization of the protein. The present study dissects the 27 N-terminal residues of cyt-PTP⑀ and identifies residues within it that are required for nuclear localization of cyt-PTP⑀. Importantly, residues required for this effect reside in the extreme N-terminal region of this domain, which are absolutely unique to cyt-PTP⑀. The study also links nuclear localization mediated by the above domain of cyt-PTP⑀ with increases in oxidative stress. The existence of p67 and p65 PTP⑀, which are naturally occurring isoforms of PTP⑀ that lack these sequences, suggests that the ability of cyt-PTP⑀ to enter the nucleus is dynamically regulated within cells. MATERIALS AND METHODS Reagents. cDNAs for full-length mouse cyt-PTP⑀ [21] and p67 PTP⑀ [23] cloned in pCDNA3 (Invitrogen) were used in this study. The N-terminal deletion mutants of cyt-PTP⑀ were constructed by PCR, using the following series of recessed 5⬘ oligos:R4 (CTGGATCCACCATGAGAAAGAACTTTTCCCGG), S8 (CTGGATCCACCATGAGCCGGCTCACCTGGTTC), T11 (CTGGATCCACCATGACCTGGTTCCGGAAGCAG), and S22 (CTGGATCCACCATGAGCAGCAACGACAAGAAA) (initiating ATG underlined) in conjunction with cyt-PTP⑀ cDNA. R4A, K5A, N6A, F7A, S8A, and R9A point mutants of cyt-PTP⑀ were constructed by PCR using 5⬘ oligomers which contained the relevant point mutations. ⌬11–27 (cyt-PTP⑀ missing amino acid residues 11–27) was constructed by PCR using the following 5⬘ oligomer, which contained the required deletion (indicated by a slash; initiator ATG codon underlined): TAAGCTTGCCATGAGCAGCAGAAAGAACTTTTCCCGGCTC/ATGCCTAACGGGATCTTAGAAGAG. Following sequence verification cDNAs were cloned either into pcDNA3 or into pEGFP-N1 (Clontech); the latter vector attaches a GFP molecule at the C terminus of PTP⑀. Antibodies used for blotting in this study included polyclonal anti-PTP⑀ [17] and polyclonal anti-SOS (Santa Cruz). Cell culture. 293T, COS, and NIH3T3 cells were grown in Dulbecco’s modified Eagle medium (Gibco-BRL), supplemented with 10% (vol/vol) fetal calf serum (Gibco-BRL), 2 mM glutamine, 50 units/ml penicillin, and 50 ␮g/ml streptomycin. Cells were transfected using the Lipofectamine reagent (Gibco-BRL) or by the calcium-phosphate method [30]. In some experiments NIH3T3 cells were treated with 1 mM H 2O 2 for 10 min. Staining of nuclei was performed by adding Hoechst 33342 (Molecular Probes, Eugene, OR) to growth medium at a final concentration of 0.05%. When required, cells were examined by confocal microscopy. Protein blotting experiments indicated that expression levels of the various cyt-PTP⑀ molecules examined in this study were comparable. Also, as individual transfected cells may express GFP-fusion proteins at different levels, care was taken in fluorescence microscopy studies to compare cells which displayed similar intensities of GFP fluorescence. Cell fractionation and protein blotting. Cells were fractionated into nuclear and nonnuclear (i.e., combined cytosol and membrane) fractions by lysis in an NP-40-based buffer (10 mM Tris–Cl, pH 7.5, 10 mM NaCl, 3 mM MgCl 2, 0.5% Nonidet P-40, pH 7.5, supplemented with protease inhibitors (1 mM N-(␣-aminoethyl) benzenesulfonyl fluoride, 40 ␮M bestatin, 15 ␮M E64, 20 ␮M leupeptin, 15

183

␮M pepstatin; Sigma). Following incubation on ice for 10 min, lysates were centrifuged at 15,000g for 10 min. The supernatant (nonnuclear fraction) was collected and stored; the pellet (nuclei) was washed by resuspension in a fresh aliquot of lysis buffer and centrifugation as above. The pellet from this second centrifugation was then suspended in RIPA buffer, sonicated, and centrifuged to remove insoluble debris, and the supernatant was stored. The quality of fractionation was verified by probing blots prepared from cellular fractions with anti-SOS antibodies; SOS was detected exclusively in nonnuclear fractions, attesting to the absence of nonnuclear contamination of nuclear fractions; this was further confirmed by the absence of p67 PTP⑀ from nuclear fractions. Equal amounts (10 –20 ␮g total protein, measured by the Bradford technique using bovine serum albumin as a standard) were analyzed on 7% SDS–polyacrylamide gels, followed by transfer to nitrocellulose membranes (Protran, Schleicher & Schuell), and hybridization to antibodies. Complete protein transfer following blotting was verified routinely by noting the transfer of prestained molecular size marker proteins of the proper size range and by occasionally staining gels following transfer with Coomassie brilliant blue. The absence of lane-to-lane variations in blotting was verified by staining the blotted membranes with Ponceau S (Sigma).

RESULTS AND DISCUSSION

The 10 Extreme N-Terminal Residues of cyt-PTP⑀ Are Required for Nuclear Localization As indicated, p67 is identical to cyt-PTP⑀ except that p67 PTP⑀ lacks 27 amino acid residues located at the extreme N terminus of cyt-PTP⑀ (Fig. 1A). This structural difference strongly affects the subcellular localization pattern of p67—this molecule is exclusively cytosolic, while cyt-PTP⑀ is predominantly cytosolic but can be found also at the cell membrane and within the cell nucleus [23]. In order to demonstrate this, full-length and p67 PTP⑀ were linked at their C termini to green fluorescent protein (GFP) to enable visualization of their localization in live cells (Fig. 1B). Images presented in Fig. 1B were acquired by confocal microscopy, indicating that GFP-tagged cyt-PTP⑀ is located inside the cell nucleus and occupies the same focal plane as Hoechst-stained DNA and that the large unstained regions in cells expressing p67-PTP⑀–GFP are in fact unstained nuclei. Protein blotting of lysates prepared from cells expressing cyt-PTP⑀–GFP or p67– GFP fusion proteins did not reveal the presence of GFP not linked to cyt-PTP⑀ (not shown), indicating that the nuclear fluorescence in Fig. 1B did not arise from free GFP, which can enter the nucleus on its own. The diffuse pattern of cyt-PTP⑀–GFP staining vs the punctate pattern of Hoechst-stained DNA within the nucleus suggests that cyt-PTP⑀ is most likely not associated with chromatin. These results indicate that the 27 N-terminal residues of cyt-PTP⑀ contain signals which regulate the subcellular localization of the molecule. Of note, the amino acid composition of this region is atypical. Over half of the residues in this region are serines or threonines (6 residues of 27) or basic amino acids (9/27) (Fig. 1A), suggesting that phosphorylation or the

184

KRAUT ET AL.

FIG. 1. p67-PTP⑀ is excluded from cell nuclei. A. Amino acid sequences at the N-termini of full-length cyt-PTP⑀ (top) and p67-PTP⑀ (bottom). Translation of p67 starts at an ATG codon located at a position which corresponds to M28 of the full-length cyt-PTP⑀ protein [23]. B. p67-PTP⑀ is excluded from cell nuclei. 293T cells were transfected with GFP-tagged cyt-PTP⑀ or p67 PTP⑀ (green); Nuclei were stained in parallel with Hoechst 33342 (blue). Cells were analyzed by confocal microscopy.

presence of a basic-type nuclear localization signal [31] may regulate cyt-PTP⑀ localization. In order to identify regions within the N-terminal domain which regulate nuclear localization of cyt-PTP⑀ we created a series of N-terminal deletion mutants, in which an initiator methionine was inserted upstream of R4, S8, T11, R14, or S22 in cyt-PTP⑀. These Nterminally truncated proteins were fused to GFP as described. Expression of this series of mutants revealed that R4 cyt-PTP⑀ was expressed throughout the entire cell in a pattern indistinguishable from that of full-length cyt-PTP⑀. In contrast, T11, R14, and S22 cyt-PTP⑀ mutants were excluded from the nucleus in a manner similar to p67 PTP⑀, while S8 cyt-PTP⑀ exhibited intermediate levels of nuclear localization (Fig. 2A and data not shown). Lack of nuclear staining of p67 PTP⑀ and of the T11, R14, and S22 deletion mutants was widespread and was detected in virtually all cells expressing these molecules. In order to rule out the possibility that the GFP tag affected nuclear localization we expressed in 293T cells nontagged versions of the same series of deletion mutants and examined

their subcellular localization by biochemical fractionation. Results of this study were similar to those obtained with the GFP-tagged PTP⑀ mutants (Fig. 2B). Note that despite somewhat higher expression levels of S22 and p67 PTP⑀ in the experiment shown these molecules were not detected in the nuclear fraction. We conclude that the region between R4 and L10 in cytPTP⑀ contains sequences which participate in regulating the nuclear localization of cyt-PTP⑀. In order to determine whether residues located between R4 and L10 were both necessary and sufficient for nuclear localization of cyt-PTP⑀ we fused this region to the p67 PTP⑀ protein (construct ⌬11–27, Fig. 2A). Examination of cells expressing this protein revealed significantly reduced nuclear staining, similar in intensity and affecting virtually all transfected cells in a manner like that seen in the T11, R14, and S22 mutants and in the p67 PTP⑀ protein (Fig. 2A). Residues R4 –L10 of cyt-PTP⑀ are therefore necessary, although not sufficient, for nuclear localization of cytPTP⑀. Residues R4 –L10 are part of the 12 N-terminal residues of cyt-PTP⑀ which are unique to this form of

NUCLEAR LOCALIZATION OF PTP⑀

185

FIG. 2. Nuclear localization is driven by a sequence between R4 and T11 of cyt-PTP⑀. A. Top: Representation of the N-termini of deletion mutants beginning at R4, S8, T11, and S22 of cyt-PTP⑀; dashed line denotes region missing from ⌬11–27 cyt-PTP⑀. Bottom: Pattern of subcellular localization of these mutants when expressed in 293T cells as GFP fusion proteins. B. Biochemical fractionation of 293T cells transfected with untagged deletion mutants. Cells were fractionated into nuclear (N) and nonnuclear (C, cytosol plus cell membrane); equal amounts of total protein from each fraction were electrophoresed and blotted with anti-PTP⑀ antibodies.

PTP⑀. Residues further downstream, starting with F13, are included in at least one of the three other known isoforms of PTP⑀, none of which are found in the nucleus [23, 25]. This indicates that the role of residues R4 –L10, which are found only in cyt-PTP⑀, is to help regulate subcellular localization of the phosphatase. R4, K5, and R9 of cyt-PTP⑀ Are Part of a Nuclear Localization Signal In order to further dissect the region between R4 and L10 we prepared a series of point mutants of fulllength GFP-tagged cyt-PTP⑀, in which residues R4 through R9 were individually replaced with alanine. Expression of these molecules in 293T cells revealed

that nuclear localization of R4A-, K5A-, and R9A-cytPTP⑀ was reduced (Fig. 3A). This effect was detected in all transfected cells expressing the above three constructs, although its intensity was somewhat weaker than that observed in p67 PTP⑀–GFP. In contrast, F7A and S8A mutants were expressed throughout the cells in a manner indistinguishable from full-length cytPTP⑀. We conclude that R4, K5, and R9 are part of a nuclear localization signal, most likely a basic-type NLS, and are necessary for driving nuclear localization of cyt-PTP⑀. This result agrees with the intermediate nuclear localization phenotype of the S8 deletion mutant (Fig. 2A); this mutant includes R9 but not R4 and K5. In agreement with the above findings, all cells

186

KRAUT ET AL.

FIG. 3. Nuclear localization of cyt-PTP⑀ is determined by R4, K5, and R9. A. The N-terminal domain, with the region analyzed (R4-R9) underlined (top). GFP-tagged, full-length cyt-PTP⑀ molecules containing individual point mutations indicated were expressed in 293T cells; subcellular localization was analyzed by fluorescence microscopy and compared with that of nonmutated cyt-PTP⑀ (WT) and p67 (p67). B. Same analysis as in A, performed on the double mutant R4A/R9A cyt-PTP⑀ and on the triple mutant R4A/K5A/R9A-cyt-PTP⑀. C. Comparison of the first 28 amino acids of mouse, rat, and human cyt-PTP⑀. Arrows denote N-termini of full length cyt-PTP⑀ and of p67 PTP⑀. Nonconservative changes vs mouse cyt-PTP⑀ are underlined, residues R4, K5, and R9 of mouse are in bold type. Sequence data are from GenBank Accession Nos. U36758 (mouse), U36623 (human), and A78610 (rat).

expressing the double (R4A/R9A) or triple (R4A/K5A/ R9A) mutants of cyt-PTP⑀ exhibited a decrease in nuclear staining that was similar in extent to p67 PTP⑀ (Fig. 3B). This result indicates that these combinations of mutations can recapitulate the effect of deleting the entire N-terminal domain of cyt-PTP⑀. Similar results were observed when the same constructs were expressed in NIH3T3 and in COS cells (not shown), indicating that this phenomenon was not unique to 293T cells. Of note, R4 and R9 of mouse PTP⑀ used here are

conserved in cyt-PTP⑀ from rat and human; K5 is conserved in rat cyt-PTP⑀ and is replaced in a nonconservative manner by a serine residue in human (Fig. 3C). Yet, despite its lack of conservation in human and the fact that the R4A/R9A mutant exhibited the same intensity of nuclear exclusion as the R4A/K5A/R9A mutant, K5 appears to play a role in nuclear localization of mouse cyt-PTP⑀ as the K5A mutant was excluded from the nucleus (Fig. 3A). Surprisingly, the N6A mutation resulted in a signif-

NUCLEAR LOCALIZATION OF PTP⑀

icant increase in nuclear localization of cyt-PTP⑀ (Fig. 3A). The strength of this phenotype varied among cell types; it was weaker in COS cells and was not detected in NIH3T3 cells (not shown). This suggests that nuclear localization of cyt-PTP⑀ might also be regulated by nuclear export regulated in part by N6; however, as this residue is not conserved in human cyt-PTP⑀ (Fig. 3C) and its effect appears to be variable in the cell types examined, this issue requires further study. Nuclear localization is regulated also by active export of proteins from the nucleus which, in many cases, relies on the presence of a nuclear export sequence (NES) of the type recognized by Crm1 [32]. Many NESs, such as those of STAT1, cyclin B, MEK, and the HIV accessory protein Rev, are of the form L(X) 2–3 L(X) 2L(X) 1L [33–36]. cyt-PTP⑀ contains no sequences which fit this consensus, although several sequences within this protein are somewhat similar ( 172IVMLTNL 175, 349LERHLQTL 356, 583IALSNIL 589). Note that the N6 residue noted above is not located in any of these sequences. Nonetheless, treatment of NIH3T3 and 293 cells expressing a full-length cyt-PTP⑀–GFP fusion protein with leptomycin B, an inhibitor of Crm1mediated nuclear export [37] (5–10 ng/ml, 1– 8 h), did not alter cellular localization of PTP⑀–GFP as assessed by fluorescence microscopy (not shown). This finding indicates that cyt-PTP⑀ does not contain an NES of the type recognized by Crm1, nor does it appear to associate with and be transported out of the nucleus by another protein which contains an NES of this type. cyt-PTP⑀ is ⬃70 kDa molecular mass, which is most likely too large to exit the nucleus by passive diffusion; we then conclude that if cyt-PTP⑀ exits the nucleus, this is probably regulated by other, non-Crm1-related mechanisms. Oxidative Stress Increases Nuclear Localization of cyt-PTP⑀ In order to better understand the physiological relevance of nuclear cyt-PTP⑀ we searched for stimuli which would result in increased amounts of cyt-PTP⑀ in cell nuclei. Subcellular localization was followed by direct visualization of cyt-PTP⑀–GFP fusion proteins and by biochemical fractionation in NIH3T3 cells. Oxidative stress in the form of H 2O 2 treatment was recently shown to promote inter- and intramolecular associations of RPTP␣ [38] and of cyt-PTP⑀ (ToledanoKatchalski et al., submitted). In order to examine the effects the N-terminal domain of cyt-PTP⑀ on subcellular localization of the protein in response to H 2O 2 we expressed cyt-PTP⑀–GFP fusion proteins in NIH3T3 cells and then treated the cells with 1 mM H 2O 2 as described under Materials and Methods. Upon treatment with H 2O 2, enhancement of nuclear and membranal localization of cyt-PTP⑀ was detected by GFP

187

fluorescence (Fig. 4A); increased nuclear localization was further verified by biochemical fractionation (Fig. 4B). No such changes were observed in cells expressing GFP alone, indicating that these changes were indeed mediated by cyt-PTP⑀. Interestingly, H 2O 2 treatment of cells expressing the R4A/K5A/R9A cyt-PTP⑀ triple mutant did not result in enhancement of nuclear and membranal localization of the mutant PTP⑀ (Fig. 4A). A similar result was obtained in cells expressing p67 PTP⑀, where nonfluorescent nuclei were clearly detected in the GFP staining patterns of H 2O 2-treated cells expressing this protein (not shown). We conclude that presence of the nuclear localization domain defined above significantly increased the ability of cytPTP⑀ to localize to the nucleus in response to H 2O 2. H 2O 2 treatment has previously been shown to cause redistribution of the 45-kDa nuclear variant of the T-cell tyrosine phosphatase into the cytosol [35], suggesting that cellular redistribution of PTPs may be a common outcome of enhanced oxidative stress. Fluorescence microscopy data presented here suggest that the same sequence region also affects membrane localization, although further studies are required in this respect. It should be noted that cyt-PTP⑀ nuclear localization was not affected by treating cells with inhibitors of serine-threonine or tyrosine phosphatases (okadaic acid or sodium pervanadate, respectively), UV (50 J/M 2) or gamma (1.5 Gy) irradiation, and transfection with various pro-apoptotic gene products, such as Bax, ZIP kinase, or DAP kinase [39 – 41], attesting to the specificity of the H 2O 2 effect. Nuclear localization of cyt-PTP⑀ may have two non mutually-exclusive consequences— gain of function, in which nuclear cyt-PTP⑀ gains access to previously inaccessible substrates, or loss of function, whereby cytPTP⑀ is removed from the cytosol and loses access to cytosolic substrates. As the amount of cyt-PTP⑀ found in the nucleus in response to stimuli examined in this study does not deplete cytosolic stores of cyt-PTP⑀, it is reasonable to believe that nuclear localization provides access to new substrates for PTP⑀. It is of particular interest to note that the two naturally-occurring, shorter forms of PTP⑀, p65 and p67, lack the region identified here as essential for nuclear localization of PTP⑀ [23]. As expression of these forms is regulated by translational and posttranslational mechanisms of gene expression, it is tempting to speculate that physiological stimuli can regulate nuclear localization of cyt-PTP⑀ by affecting these mechanisms. Coexistence of two mechanisms for removal of the N-terminal NLS of cyt-PTP⑀ may provide cells with flexibility in denying nuclear access to the enzyme. Along these lines, proteolytic processing of full-length cyt-PTP⑀ to p65 PTP⑀ may constitute a mechanism for preventing existing PTP⑀ molecules from entering the nucleus, while a shift in translation initiation toward production of

188

KRAUT ET AL.

FIG. 4. Nuclear localization of cyt-PTP⑀ is influenced by oxidative stress. A. NIH3T3 cells were transfected with GFP alone (GFP), with GFP-tagged full-length cyt-PTP⑀ (cyt-PTP⑀), or with GFP-tagged R4A/K5A/R9A cyt-PTP⑀ (R4A/K5A/R9A), and analyzed 16 –24 h later. Some cells were treated with 1 mM H 2O 2 for 10 min prior to photography. Shown are subcellular localizations of PTP⑀ in untreated cells (top) and in treated cells (bottom). Note the enhanced nuclear and membrane localization of full-length cyt-PTP⑀, which are missing from the R4A/K5A/R9A mutant. Cells were examined by fluorescence microscopy; original magnification ⫻60. B. NIH3T3 cells were fractionated into nuclear (N) and nonnuclear (C) fractions. Note the increase in amount of full-length cyt-PTP⑀ in nuclear fraction from H 2O 2-treated cells.

p67 PTP⑀ may be aimed at preventing synthesis of PTP⑀ molecules which can enter the nucleus. Further studies aimed at isolating nuclear substrates of cytPTP⑀ and linking them with stimuli which increase amounts of nuclear cyt-PTP⑀ are expected to help clarify these issues. This study was supported by The Israel Science Foundation, founded by The Israel Academy of Sciences and Humanities. A.E. is an incumbent of the Adolfo and Evelyn Blum Career Development Chair in Cancer Research at the Weizmann Institute of Science.

REFERENCES 1.

Tonks, N. K., and Neel, B. G. (2001). Combinatorial control of the specificity of protein tyrosine phosphatases. Curr. Opin. Cell Biol 13, 182–195. 2. den Hertog, J. (1999). Protein-tyrosine phosphatases in development. Mech. Dev. 85, 3–14. 3. Petrone, A., and Sap, J. (2000). Emerging issues in receptor protein tyrosine phosphatase function: Lifting fog or simply shifting? J. Cell Sci. 113, 2345–2354. 4. Andersen, J. N., Mortensen, O. H., Peters, G. H., Drake, P. G., Iversen, L. F., Olsen, O. H., Jansen, P. G., Andersen, H. S.,

Tonks, N. K., and Moller, N. P. (2001). Structural and evolutionary relationships among protein tyrosine phosphatase domains. Mol. Cell Biol. 21, 7117–7136. 5.

Mauro, L. J., and Dixon, J. E. (1994). “Zip codes” direct intracellular protein tyrosine phosphatases to the correct cellular “address”. Trends Biochem. Sci. 19, 151–155.

6.

Fischer, E. H. (1999). Cell signaling by protein tyrosine phosphorylation. Adv. Enzyme Regul. 39, 359 –369.

7.

Frangioni, J. V., Oda, A., Smith, M., Salzman, E. W., and Neel, B. G. (1993). Calpain-catalyzed cleavage and subcellular relocation of protein phosphotyrosine phosphatase 1B (PTP-1B) in human platelets. EMBO J. 12, 4843– 4856.

8.

Gu, M., and Majerus, P. W. (1996). The properties of the protein tyrosine phosphatase PTPMEG. J. Biol. Chem. 271, 27751–27759.

9.

Gurd, J. W., Bissoon, N., Nguyen, T. H., Lombroso, P. J., Rider, C. C., Beesley, P. W., and Vannucci, S. J. (1999). Hypoxia– ischemia in perinatal rat brain induces the formation of a low molecular weight isoform of striatal enriched tyrosine phosphatase (STEP). J. Neurochem. 73, 1990 –1994.

10.

Nguyen, T. H., Paul, S., Xu, Y., Gurd, J. W., and Lombroso, P. J. (1999). Calcium-dependent cleavage of striatal enriched tyrosine phosphatase (STEP). J. Neurochem. 73, 1995–2001.

11.

Haj, F. G., Verveer, P. J., Squire, A., Neel, B. G., and Bastiaens, P. I. (2002). Imaging sites of receptor dephosphorylation by

NUCLEAR LOCALIZATION OF PTP⑀

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

PTP1B on the surface of the endoplasmic reticulum. Science 295, 1708 –1711. Rock, M. T., Brooks, W. H., and Roszman, T. L. (1997). Calciumdependent signaling pathways in T cells. Potential role of calpain, protein tyrosine phosphatase 1b, and p130Cas in integrinmediated signaling events. J. Biol. Chem. 272, 33377–33383. Stein-Gerlach, M., Wallasch, C., and Ullrich, A. (1998). SHP-2, SH2-containing protein tyrosine phosphatase-2. Int. J. Biochem. Cell. Biol. 30, 559 –566. Elson, A., Kozak., C. A., Morton, C. C., Weremowicz, S., and Leder, P. (1996). The protein tyrosine phosphatase epsilon gene maps to mouse chromosome 7 and human chromosome 10q26. Genomics 31, 373–375. Tanuma, N., Nakamura, K., and Kikuchi, K. (1999). Distinct promoters control transmembrane and cytosolic protein tyrosine phosphatase epsilon expression during macrophage differentiation. Eur. J. Biochem. 259, 46 –54. Krueger, N. X., Streuli, M., and Saito, H. (1990). Structural diversity and evolution of human receptor-like protein tyrosine phosphatases. EMBO J. 9, 3241–3252. Elson, A. & Leder, P. (1995). Protein-tyrosine phosphatase epsilon. An isoform specifically expressed in mouse mammary tumors initiated by v-Ha-ras or neu. J. Biol. Chem. 270, 26116 –26122. Elson, A. (1999). Protein tyrosine phosphatase ⑀ increases the risk of mammary hyperplasia and mammary tumors in transgenic mice. Oncogene 18, 7535- 7542. Moller, N. P., Moller, K. B., Lammers, R., Kharitonenkov, A., Hoppe, E., Wiberg, F. C., Sures, I., and Ullrich, A. (1995). Selective down-regulation of the insulin receptor signal by protein-tyrosine phosphatases alpha and epsilon. J. Biol. Chem. 270, 23126 –23131. Andersen, J. N., Elson, A., Lammers, R., Romer, J., Clausen, J. T., Moller, K. B., and Moller, N. P. H. (2001). Comparative study of PTP⑀ isoforms: Membrane localization confers specificity in inhibiting insulin-induced cell rounding and cell detatchment. Biochem. J. 354, 581–590. Elson, A., and Leder, P. (1995). Identification of a cytoplasmic, phorbol ester-inducible isoform of protein tyrosine phosphatase epsilon. Proc. Natl. Acad. Sci. USA. 92, 12235–12239. Nakamura, K., Mizuno, Y., and Kikuchi, K. (1996). Molecular cloning of a novel cytoplasmic protein tyrosine phosphatase PTP epsilon. Biochem. Biophys. Res. Commun. 218, 726 –732. Gil-Henn, H., Volohonsky, G., Toledano-Katchalski, H., Gandre, S., and Elson, A. (2000). Generation of novel cytoplasmic forms of protein tyrosine phosphatase epsilon by proteolytic processing and translational control. Oncogene 19, 4375– 4384. Peretz, A., Gil-Henn, H., Sobko, A., Shinder, V., Attali, B., and Elson, A. (2000). Hypomyelination and increased activity of voltage-gated potassium channels in mice lacking protein tyrosine phosphatase ⑀. EMBO J. 19, 4036 – 4045. Gil-Henn, H., Volohonsky, G., and Elson, A. (2001). Regulation of RPTP alpha and PTP epsilon by calpain-mediated proteolytic cleavage. J. Biol. Chem. 276, 31772–31779. Thompson, L. J., Jiang, J., Madamanchi, N., Runge, M. S., and Patterson, C. (2001). PTP-epsilon, a tyrosine phosphatase ex-

Received May 13, 2002 Revised version received August 16, 2002 Published online November 11, 2002

189

pressed in endothelium, negatively regulates endothelial cell proliferation. Am. J. Physiol. Heart Circ. Physiol. 281, H396– 403. 27.

Sully, V., Pownall, S., Vincan, E., Bassal, S., Borowski, A. H., Hart, P. H., Rockman, S. P., and Phillips, W. A. (2001). Functional abnormalities in protein tyrosine phosphatase epsilon-deficient macrophages. Biochem. Biophys. Res. Commun. 286, 184–188.

28.

Tanuma, N., Nakamura, K., Shima, H., and Kikuchi, K. (2000). Protein-tyrosine phosphatase PTP epsilon C inhibits Jak-STAT signaling and differentiation induced by interleukin-6 and leukemia inhibitory factor in M1 leukemia cells. J. Biol. Chem. 275, 28216 –28221.

29.

Tanuma, N., Shima, H., Nakamura, K., and Kikuchi, K. (2001). Protein tyrosine phosphatase epsilon C selectively inhibits interleukin 6- and interleukin 10-induced JAK-STAT signaling. Blood 98, 3030 –3034.

30.

Chen, C., and Okayama, H. (1987). High-efficiency transformation of mammalian cells by plasmid DNA. Mol. Cell. Biol. 7, 2745–2752.

31.

Christophe, D., Christophe-Hobertus, C., and Pichon, B. (2000). Nuclear targeting of proteins: How many different signals? Cell Signal. 12, 337–341.

32.

Lei, E. P., and Silver, P. A. (2002). Protein and RNA export from the nucleus. Dev. Cell 2, 261–272.

33.

McBride, K. M., McDonald, C., and Reich, N. C. (2000). Nuclear export signal located within the DNA-binding domain of the STAT1 transcription factor. EMBO J. 19, 6196 – 6206.

34.

Perander, M., Bjorkoy, G., and Johansen, T. (2001). Nuclear import and export signals enable rapid nucleocytoplasmic shuttling of the atypical protein kinase C lambda. J. Biol. Chem. 276, 13015–130124.

35.

Lam, M. H., Michell, B. J., Fodero-Tavoletti, M. T., Kemp, B. E., Tonks, N. K., and Tiganis, T. (2001). Cellular stress regulates the nucleocytoplasmic distribution of the protein-tyrosine phosphatase TCPTP. J. Biol. Chem. 276, 37700 –37707.

36.

Henderson, B. R., and Eleftheriou, A. (2000). A comparison of the activity, sequence specificity, and CRM1-dependence of different nuclear export signals. Exp. Cell Res. 256, 213–224.

37.

Kudo, N., Wolff, B., Sekimoto, T., Schreiner, E. P., Yoneda, Y., Yanagida, M., Horinouchi, S., and Yoshida, M. (1998). Leptomycin B inhibition of signal-mediated nuclear export by direct binding to CRM1. Exp. Cell Res. 242, 540 –547.

38.

Blanchetot, C., Tertoolen, L. G., and den Hertog, J. (2002). Regulation of receptor protein-tyrosine phosphatase alpha by oxidative stress. EMBO J. 21, 493- 503.

39.

Hickman, J. A. (2002). Apoptosis and tumourigenesis. Curr. Opin. Genet. Dev. 12, 67–72.

40.

Kawai, T. Matsumoto, M., Takeda, K., Sanjo, H., and Akira, S. (1998). ZIP kinase, a novel serine/threonine kinase which mediates apoptosis. Mol. Cell. Biol. 18, 1642- 1651.

41.

Cohen, O., Feinstein, E., and Kimchi, A. (1997). DAP-kinase is a Ca2⫹/calmodulin- dependent, cytoskeletal-associated protein kinase, with cell death-inducing functions that depend on its catalytic activity. EMBO J. 16, 998 –1008.