Zn2+-dependent Activation of the Trk Signaling Pathway Induces Phosphorylation of the Brain-enriched Tyrosine Phosphatase STEP

Excessive release of Zn2+ in the brain is implicated in the progression of acute brain injuries. Although several signaling cascades have been reported to be involved in Zn2+-induced neurotoxicity, a potential contribution of tyrosine phosphatases in this process has not been well explored. Here we show that exposure to high concentrations of Zn2+ led to a progressive increase in phosphorylation of the striatal-enriched phosphatase (STEP), a component of the excitotoxic-signaling pathway that plays a role in neuroprotection. Zn2+-mediated phosphorylation of STEP61 at multiple sites (hyperphosphorylation) was induced by the up-regulation of brain-derived neurotropic factor (BDNF), tropomyosin receptor kinase (Trk) signaling, and activation of cAMP-dependent PKA (protein kinase A). Mutational studies further show that differential phosphorylation of STEP61 at the PKA sites, Ser-160 and Ser-221 regulates the affinity of STEP61 toward its substrates. Consistent with these findings we also show that BDNF/Trk/PKA mediated signaling is required for Zn2+-induced phosphorylation of extracellular regulated kinase 2 (ERK2), a substrate of STEP that is involved in Zn2+-dependent neurotoxicity. The strong correlation between the temporal profile of STEP61 hyperphosphorylation and ERK2 phosphorylation indicates that loss of function of STEP61 through phosphorylation is necessary for maintaining sustained ERK2 phosphorylation. This interpretation is further supported by the findings that deletion of the STEP gene led to a rapid and sustained increase in ERK2 phosphorylation within minutes of exposure to Zn2+. The study provides further insight into the mechanisms of regulation of STEP61 and also offers a molecular basis for the Zn2+-induced sustained activation of ERK2.

The intracellular tyrosine phosphatase, STEP (striatalenriched tyrosine phosphatase, also known as PTPN5) is expressed exclusively in the central nervous system (16,17) and is emerging as a key regulator of neuronal survival and death. The STEP-family of PTPs includes both membrane-associated (STEP 61 ) and cytosolic (STEP 46 ) variants that are formed by alternative splicing of a single gene (18). STEP is expressed in neurons of the cortex, hippocampus, and striatum and its substrate affinity is regulated through phosphorylation/dephosphorylation of a serine residue in a conserved regulatory domain termed the KIM (kinase interacting motif) domain (18 -22). Dopamine/D1 receptor/cAMP/PKA pathway-mediated phosphorylation of this residue hinders the ability of STEP to bind to its substrates (22). In contrast, dephosphorylation of this residue following glutamate-NMDA receptor mediated activation of calcineurin allows STEP to bind to its substrates and inhibit their activities (21). This dephosphorylated form of STEP also referred to as the active form has been shown to contribute to neuroprotection in cell culture models of excitotoxicity and oxygen glucose deprivation (23,24). Evidence for a neuroprotective role of STEP also comes from in vivo studies demonstrating that degradation of active STEP following an ischemic insult allows activation of detrimental cascades involved in neuronal injury and brain damage. In contrast, restoration of STEP function, using a brain-permeable STEP-derived peptide, is effective in limiting ischemic brain injury (24).
These findings indicate that loss of function of endogenous STEP increases the vulnerability of neurons to excitotoxic insult. Since Zn 2ϩ has been associated with excitotoxic brain injury, the present study sought to examine the role of excessive Zn 2ϩ exposure in regulating the function of STEP 61 , the predominant isoform expressed in cultured neurons, cortex, and hippocampus. The results show that Zn 2ϩ -mediated Trk receptor activation leads to phosphorylation of STEP 61 at multiple PKA sites with a concomitant increase in the phosphorylation of ERK MAPK. The findings suggest that loss of affinity of phosphorylated STEP 61 toward its substrates facilitates the sustained phosphorylation of ERK MAPK that is known to be involved in Zn 2ϩ -induced neurotoxicity (13).

Experimental Procedures
Materials-Pregnant female Sprague-Dawley rats (16-day gestation) were obtained from Harlan Laboratories. STEP knock-out mice (STEP KO or STEP Ϫ/Ϫ ) were developed on a C57BL6 background (25) and were bred at the University of New Mexico Animal Care Facility. ZnCl 2 , glutamate, kainic acid, APV, CNQX, pyrithione sodium salt, Ca-EDTA, and BDNF were from Sigma-Aldrich. MK801, phorbol 12-myristate-13-acetate (phorbol ester or PMA), nifedipine, K252a, PP2, bisindolylmaleimide I (Bis), thapsigargin, and H89 were from EMD Biosciences. NGF was from R&D Systems. GM6001 was from Millipore. Antibodies used were as follows: polyclonal anti-ERK2 and anti-TrkB antibodies from Santa Cruz Biotechnology, monoclonal anti-phosphorylated-ERK1/2 (T P EY P ), anti-phospho-PLC␥1 and PLC␥1 antibody from Cell Signaling Technology; anti-phosphotyrosine (anti-pTyr, 4G10) from Millipore; polyclonal anti-␤-tubulin antibody from Sigma-Aldrich; anti-BDNF antibody from Promega and monoclonal anti-STEP (recognizes all STEP isoforms) from Novus Biologicals. All secondary antibodies were from Cell Signaling. All tissue culture reagents were obtained from Invitrogen. All other reagents were from Sigma-Aldrich. Approval for animal experiments was given by the University of New Mexico, Health Sciences Center, Institutional Animal Care and Use Committee.
DNA Constructs-Full-length STEP 61 cDNA was constructed in mammalian expression vector pcDNA3.1 encoding C-terminal V5 and His tags. Mutations of serine residues in STEP 61 were obtained by polymerase chain reaction (PCR)based site-directed mutagenesis using Pfu Turbo DNA polymerase (Stratagene, La Jolla, CA) according to the manufacturer's protocol. All mutations were verified by nucleotide sequencing.
Cell Culture and Stimulation-Primary neuronal cultures were obtained from 16 -17-day-old rat or wild type (WT) and STEP knock-out (KO) mice embryos as described previously (21). Briefly, the cortex was dissected under a microscope, the tissue dissociated mechanically and resuspended in DMEM/ F-12 (1:1) containing 5% fetal calf serum. Cells were plated on 60 mm poly-D-lysine-coated tissue culture dishes and grown for 12-14 days at 37°C in a humidified atmosphere (95:5% air:CO 2 mixture). To inhibit proliferation of non-neuronal cells, 10 M of cytosine D-arabinofuranoside was added to the cultures 72 h after plating. For neuronal stimulation, cells were washed twice with minimum essential medium (MEM) followed by treat-ment with ZnCl 2 , BDNF, or NGF for the indicated times at 37°C. For some experiments, cells were returned back to its original medium following treatment with Zn 2ϩ . APV, CNQX, MK801, nifedipine, K252a, thapsigargin, Ca-EDTA, PP2, bisindolylmaleimide, H89, or GM6001 were added 15 min before stimulation with Zn 2ϩ or BDNF. The selective Zn 2ϩ chelator, CaEDTA or the Zn 2ϩ ionophore, pyrithione were added along with Zn 2ϩ without any pre-incubation. For some experiments, function blocking anti-BDNF antibody was added 15 min prior to treatment with BDNF. Some cultures were treated with thapsigargin for the specified times without stimulation with Zn 2ϩ , BDNF, or NGF. Cells were harvested at the specified times after stimulation and processed for immunoblot analysis.
HeLa cell lines were routinely maintained in DMEM supplemented with 10% fetal bovine serum at 37°C in a humidified atmosphere (95:5% air:CO 2 mixture). Cells were transiently transfected with 1 g of each DNA using Lipofectamine 2000 (Invitrogen). After 24 -48 h, the cells were treated with forskolin (40 M) or phorbol ester (10 ng/ml) at 37°C. Both treated and untreated cells were harvested at the specified times after stimulation and either processed for immunoprecipitation with anti-ERK2 antibody or immunoblot analysis.
Alkaline Phosphatase Assay-Cells were lysed in a buffer containing 50 mM Tris/HCl (pH 7.4), 150 mM NaCl, 10 mM MgCl 2 , 0.1 mM DTT, 0.5% Nonidet P40, and a mixture of protease inhibitors for 30 min on ice. Lysates were centrifuged for 10 min at 14,000 rpm and the supernatant was used for the phosphatase assay. Equal amount of protein from each sample was incubated with 0.5 l of alkaline phosphatase from calf intestine (257 units/l) at 37°C for 30 min. The reactions were terminated by the addition of Laemmli sample buffer (1ϫ, final concentration) and the samples were then boiled at 100°C for 10 min followed by immunoblot analysis.
Immunoprecipitation-Cells were lysed in a buffer containing 50 mM Tris/HCl (pH 7.4), 150 mM NaCl, 50 mM NaF, 10 mM Na 4 P 2 O 7 , 1 mM Na 3 VO 4 , 0.1% Nonidet P40, and a mixture of protease inhibitors. Lysates were centrifuged for 10 min at 14,000 rpm to remove insoluble material, and then pre-cleared with protein-G Sepharose for 1 h. For immunoprecipitation of tyrosine-phosphorylated proteins, STEP or ERK, samples were incubated overnight with anti-phosphotyrosine antibody (4G10), anti-STEP antibody, or anti-ERK2 antibody. respectively. The immune complexes were incubated with 30 l of protein-G-Sepharose for 2 h at 4°C and washed five times with lysis buffer. Proteins were eluted using SDS-sample buffer and processed for SDS/PAGE and immunoblotting analysis with antibodies, as described in the individual experiments.
Immunoblotting-Equal amount of protein lysates, as estimated using bicinchoninic acid protein assay kit (Pierce Biotechnology), were resolved by 8% SDS-PAGE and transferred to polyvinylidene difluoride membrane. The membranes were blocked for 1 h at room temperature with 5% nonfat dry milk or 5% bovine serum albumin and then incubated for 1 h at room temperature or at 4°C overnight with the appropriate primary antibody. Horseradish peroxidase coupled to anti-rabbit or anti-mouse IgG that were raised in goat were used as secondary antibodies. Immune complexes were detected on x-ray film after treatment with West Pico super signal chemiluminescence reagent (Pierce Biotechnology).
Quantification of phosphorylated STEP 61 and ERK MAPK levels was done by computer-assisted densitometry and Image J analysis. For quantification of the increased basal phosphorylation of STEP 61 (band 2) and the hyper-phosphorylation of STEP 61 (band 3) the intensities of all the three STEP bands (band 1, band 2, and band 3) were measured, and the intensity of band 2 or band 3 is presented as a percentage of the total intensity of band 1, band 2, and band 3. Band intensities of phosphorylated ERK were normalized to total ERK levels from the same blot and presented as the number of fold increase as compared with the untreated control.
Statistical Analysis-Data in the text and figures are expressed as means Ϯ S.E. Statistical differences between multiple groups were assessed using one-way ANOVA followed by Bonferroni's post-hoc comparisons test. Differences were considered statistically significant when p Ͻ 0.05.

Results
Zn 2ϩ -mediated Phosphorylation of STEP 61 -In initial studies, neuronal cultures (12-14 days in vitro) were treated with increasing concentrations of Zn 2ϩ (0, 25, 50, 100, 200, and 300 M) for 30 min and analyzed by immunoblotting with an anti-STEP antibody (Fig. 1A). As previously reported STEP 61 , the predominant isoform expressed in cultured neurons was present as a doublet (21), where the lower band (band 1) represents the non-phosphorylated form and the upper band (band 2) represents the phosphorylated form of STEP 61 (Fig. 1A, upper row, lane 1). Treatment with Zn 2ϩ progressively decreased the intensity of band 1 with a concomitant increase in the intensity of band 2 and the appearance of a third band at an even higher molecular weight (band 3) suggesting concentration-dependent hyperphosphorylation of STEP 61 . To examine the temporal profile of STEP 61 phosphorylation, we next treated neuronal cultures with Zn 2ϩ (300 M) for various time periods (0, 5, 15, 30, 45, and 60 min). A rapid increase in the phosphorylation of band 2 of STEP 61 was observed within 5 min of stimulation with Zn 2ϩ that remained unchanged over time (Fig. 1B). The level of the hyperphosphorylated band 3 of STEP 61 increased progressively, peaked at 30 min and then remained elevated throughout the rest of the time studied (Fig. 1B, upper row). This increase in phosphorylation and hyperphosphorylation of STEP 61 (bands 2 and 3) was completely blocked following addition of a cell-impermeable selective Zn 2ϩ chelator (CaEDTA, 1 mM) either at the onset of Zn 2ϩ exposure ( Fig. 2A) or 15 min after the onset of Zn 2ϩ exposure (Fig. 2B).
The increase in phosphorylation and hyperphosphorylation of STEP 61 returned to basal levels within 30 min of Zn 2ϩ removal from the medium, implying that extracellular Zn 2ϩmediated phosphorylation of STEP 61 is a reversible phenomenon (Fig. 2C). To confirm that the upward shift in the mobility of the STEP 61 band was due to phosphorylation, lysates obtained from neuronal cultures treated with Zn 2ϩ (300 M, 30 min) were incubated with alkaline phosphatase, an enzyme that catalyzes the hydrolysis of phosphates from phosphorylated substrates (26,27). The complete loss of band 2 and 3 under these conditions (Fig. 2D, lane 3) confirms that Zn 2ϩ exposure led to increase in phosphorylation of STEP 61 at multiple sites resulting in a partial hypershift in the mobility of the STEP 61 band. Our result also reveal an additional STEP 61 band below the non-phosphorylated band 1 that did not change with alkaline phosphatase treatment (Fig. 2D, lane 1 versus lane 3), indicating that this band also represents a non-phosphorylated form of STEP 61 . The difference in mobility between this band and band 1 could be attributed to other post-translational modifications and is an important topic for future study. To our knowledge this is the first study that demonstrates such a hyperphosphorylation of STEP 61 in response to any stimuli.   2E shows that, in contrast to extracellular Zn 2ϩ application, approaches to preferentially increase intracellular Zn 2ϩ concentrations did not increase STEP 61 phosphorylation. Neurons were exposed to varying concentrations of Zn 2ϩ (1, 10, and 100 M for 30 min) in the presence of pyrithione (20 M), an ionophore that facilitates Zn 2ϩ entry into the cytosol (28). 1 or 10 M of Zn 2ϩ had no effect on the basal phosphorylation of STEP 61 , in the presence or absence of pyrithione. Treatment with Zn 2ϩ (100 M) in the absence of pyrithione led to increase in basal phosphorylation as well as hyperphosphorylation of STEP 61 (Fig. 2E, lane 6), as shown earlier (Fig. 1). In contrast, exposure to Zn 2ϩ (100 M) in the presence of pyrithione led to a marked decrease in the basal phosphorylation of STEP 61 , as evident from the decrease in intensity of band 2 (phosphorylated form of STEP 61 ) and concomitant increase in that of band 1 (dephosphorylated form of STEP 61 ). Taken together these findings raise the possibility that hyperphosphorylation of STEP 61 by exogenous Zn 2ϩ is mediated through stimulation of cell surface receptors.
Role of TrkB Receptor in the Phosphorylation of STEP 61 -Earlier studies have reported that exogenous Zn 2ϩ can modulate the activation of several receptors, including tropomyosin-related kinase (Trk), NMDA, AMPA, and L-type voltage gated channels (8, 9, 29 -33). To clarify the mechanism by which Zn 2ϩ mediates STEP 61 hyperphosphorylation we evaluated the effect of pharmacological inhibition of these receptors on STEP 61 . To determine the involvement of Trk receptors, neurons were incubated with Zn 2ϩ (300 M, 30 min) in the presence of K252a (100 nM), an inhibitor of Trk tyrosine kinases (34). Fig. 3A shows that K252a can effectively block the Zn 2ϩinduced hyperphosphorylation of STEP 61 (band 3) and the increase in basal phosphorylation of STEP 61 (band 2). Since cultured neurons predominantly express TrkB receptors (12) we next wanted to confirm whether Zn 2ϩ enhances tyrosine phosphorylation of TrkB, a surrogate marker of its activation (35,36). Lysates from neurons treated with Zn 2ϩ (300 M, 30 min) in the absence or presence of K252a (100 nM) were immunoprecipitated with anti-phosphotyrosine (p-Tyr) antibody followed by immunoblot analysis with an anti-TrkB antibody. The results show that Zn 2ϩ exposure markedly increased the phosphorylation of TrkB without altering its expression, while pre-incubation with K252a completely blocked Zn 2ϩ -induced TrkB phosphorylation (Fig. 3B, panels 1 and 2). To confirm that phosphorylation of TrkB led to its activation we further inves- We next evaluated whether NMDA, AMPA and L-type voltage gated channels could also contribute to Zn 2ϩ -mediated STEP 61 phosphorylation. Neurons were incubated with Zn 2ϩ (300 M) in the presence of antagonists for AMPA/kainate (CNQX, 20 M) or NMDA (APV, 200 M or MK801, 15 M) receptors, or the L-type voltage-gated channel (nifedipine, 15 M). Fig. 3, C and D show that none of these agents significantly altered Zn 2ϩ -induced STEP 61 phosphorylation.
Zn 2ϩ /TrkB Receptor Activation Leads to Sustained Phosphorylation of ERK MAPK-ERK MAPK, a downstream effector of TrkB, regulates a variety of cellular responses including cell proliferation, migration, and differentiation. An earlier study has also demonstrated a role of ERK MAPK in extracellular Zn 2ϩ -induced neurotoxicity (13). Since ERK2 is a substrate of STEP (21), we evaluated the profile of ERK2 phosphorylation following treatment with Zn 2ϩ . Our findings show a dose and time-dependent increase in the magnitude of ERK2 phosphorylation following exposure to Zn 2ϩ (Fig. 4, A and B), which correlates with the profile of increase in STEP 61 phosphorylation and hyperphosphorylation (Fig. 1, A and B). The complete loss of ERK2 phosphorylation in the presence of CaEDTA, even when administered 15 min after the onset of Zn 2ϩ (300 M, 30 min) exposure (Fig. 5, A and B), further confirms the role of extracellular Zn 2ϩ in regulating ERK2 phosphorylation. Zn 2ϩinduced increase in ERK2 phosphorylation also returned to basal levels within 30 min of Zn 2ϩ removal from the medium (Fig. 5C), which is in agreement with the reversible nature of STEP 61 phosphorylation (Fig. 2C). Blocking Trk receptors with K252a significantly reduced Zn 2ϩ (300 M, 30 min) induced phosphorylation of ERK2 (Fig. 5D). These findings strongly indicate a primary role of TrkB receptors in extracellular Zn 2ϩinduced regulation of both STEP 61 and ERK2.
As described above for STEP 61 , inhibition of AMPA/kainate and NMDA glutamate receptors or voltage-gated channels failed to block Zn 2ϩ (300 M, 30 min)-induced phosphorylation of ERK2 (Fig. 5, E and F). To ensure the efficacy of the pharmacological inhibitors in the above experiments, in control studies neurons were treated with glutamate (100 M, 5 min) in the presence of MK801 or APV; kainic acid (100 M, 5 min) in the presence of CNQX; or KCl (60 mM, 5 min) in the presence of nifedipine and analyzed for phosphorylation of ERK2. Consistent with earlier findings (21,38,39) each of these pharmacological inhibitors effectively blocked glutamate, kainic acid, or KCl-mediated phosphorylation of ERK2 (Fig. 5, E and F).
Role of Ca 2ϩ in the Phosphorylation of STEP 61 -Earlier studies have indicated that exogenous Zn 2ϩ could facilitate increases in intracellular Ca 2ϩ levels through multiple pathways (9,40,41). Since Ca 2ϩ has been shown to play a role in modulating STEP phosphorylation (21) we investigated whether Zn 2ϩ -induced influx of Ca 2ϩ or release of Ca 2ϩ from   intracellular pools could play a role in the hyperphosphorylation of STEP 61 . Neurons were exposed to Zn 2ϩ (300 M, 30 min) in the absence of Ca 2ϩ in the medium. Fig. 6A shows that lack of extracellular Ca 2ϩ had no effect on the ability of Zn 2ϩ to phosphorylate STEP 61 . To deplete Ca 2ϩ released from intracellular sources, in the next set of experiments neurons were preincubated with thapsigargin, prior to treatment with Zn 2ϩ . Fig.  6B demonstrates the effect of treatment with thapsigargin alone. Incubation with thapsigargin (1 M, 15 min) led to a transient increase in ERK2 phosphorylation within 5 min of stimulation, which returned to basal levels by 15 min. Such transient increase in ERK2 phosphorylation following exposure to thapsigargin has been previously attributed to depletion of Ca 2ϩ stores (21). Therefore, in subsequent experiments neurons were treated with thapsigargin (1 M) for 15 min prior to Zn 2ϩ exposure (300 M, 30 min). Fig. 6C shows that such pretreatment with thapsigargin failed to alter the effect of Zn 2ϩ on STEP 61 phosphorylation (compare lane 2 with lane 4). Fig. 6D further shows that the Zn 2ϩ -mediated increase in ERK2 phosphorylation correlates with the phosphorylation of STEP 61 both in the absence or presence of thapsigargin (compare lane 2 with lane 4). Thus depletion of intracellular Ca 2ϩ stores also has no effect on the ability of exogenous Zn 2ϩ to increase the phosphorylation of STEP 61 and ERK2.
Role of PKA in the Phosphorylation of STEP 61 -The Src family of tyrosine kinases, protein kinase C (PKC) and cAMP-dependent protein kinase A (PKA) are known to play important roles in transducing TrkB receptor mediated intracellular signaling (14). To evaluate the role of these kinases in the phosphorylation of STEP 61 and ERK2, neurons were treated with Zn 2ϩ (300 M, 30 min) in the presence or absence of inhibitors of Src, PKC and PKA. Fig. 7, A and B show that inhibition of Src tyrosine kinase with PP2 (0.4, 1.0, or 2.5 M) and PKC with bisindolylmaleimide I (Bis; 1 or 3 M) failed to inhibit Zn 2ϩinduced phosphorylation of STEP 61 and ERK2. The efficacy of PP2 and Bis was confirmed in control studies, where neurons were treated with glutamate (100 M, 5 min) in the presence of PP2; or phorbol ester (PMA; 10 ng/ml, 5 min) in the presence of Bis followed by analysis of ERK2 phosphorylation (Fig. 7, A and   B). Fig. 7C shows that inhibition of PKA with H89 (40 M) attenuated the Zn 2ϩ -induced hyperphosphorylation (band 3) as well as the basal phosphorylation (band 2) of STEP 61 . The PKA inhibitor also attenuated the phosphorylation of ERK2 (Fig. 7C). Using a phospho-specific antibody that recognizes STEP when phosphorylated at the PKA site within the KIM domain (21), we found that STEP 61 was partially phosphorylated under basal condition (Fig. 7D, lane 1). The phosphorylation of this site increased following exposure to Zn 2ϩ (Fig. 7D, lane  2). The recognition of the second STEP 61 band at a higher molecular weight by the phospho-specific antibody (Fig. 7D, lane 2) confirmed the phosphorylation of STEP 61 at additional sites by Zn 2ϩ , resulting in the mobility shift. The lack of detection of any STEP 61 band with the phospho-specific STEP antibody in the presence of H89 (Fig. 7D, lane 3) further suggested that Zn 2ϩ induced phosphorylation of STEP 61 at multiple PKA sites. The strong correlation between the phosphorylation of STEP 61 at the PKA sites (bands 2 and 3) and the concomitant increase in ERK2 phosphorylation (Fig. 7C) suggests that the loss of affinity of phosphorylated STEP 61 toward its substrate leads to sustained phosphorylation of ERK2.
To establish more directly that phosphorylation of STEP 61 at multiple PKA sites regulates ERK2 activation we carried out a number of studies in HeLa cells. Based on an earlier study that identified Ser-160 and Ser-221 as the two putative PKA phosphorylation sites in STEP 61 (22) we generated three mutants of STEP 61 that could be either partially phosphorylated or nonphosphorylatable by PKA. In the first mutant Ser-160 was converted to alanine (V5-tagged STEP 61 S160A), in the second mutant Ser-221 was converted to alanine (V5-tagged STEP 61 S221A) and in the third mutant both Ser-160 and Ser-221 were converted to alanine (V5-tagged STEP 61 S160A/S221A). HeLa cells transfected with V5-STEP 61 wild type (WT) or one of these mutants were treated with or without the PKA agonist, forskolin (40 M), followed by immunoblot analysis with anti-V5 antibody. Fig. 8B shows that in the absence of forskolin the mobility of the protein bands of STEP 61 mutants remained unchanged. Fig. 8C shows that treatment with forskolin resulted in an upward shift in mobility of V5-STEP 61 WT, V5-STEP 61 S160A, and V5-STEP 61 S221A, whereas V5-STEP 61 S160A/S221A did not show any change in mobility, suggesting that phosphorylation of at least one of the sites is necessary for the upward shift in mobility of STEP 61 . The upward shift in mobility of V5-STEP 61 WT was also relatively higher than V5-STEP 61 S160A or V5-STEP 61 S221A, suggesting that concomitant phosphorylation of both the sites (Ser-160 and Ser-221) resulted in the hyperphosphorylation of STEP 61 .
To evaluate the consequence of phosphorylation of STEP 61 at Ser-160, Ser-221, or both Ser-160 and Ser-221 in terms of their ability to bind to and dephosphorylate ERK MAPK we generated four mutant forms of STEP 61 . In the first mutant both Ser-160 and Ser-221 were converted to alanine to mimic the non-phosphorylated form of STEP 61 (V5-STEP 61 S160A/ S221A). In the second mutant Ser-160 was converted to alanine and Ser-221 was converted to glutamic acid (V5-STEP 61 S160A/S221E), whereas in the third mutant Ser-160 was converted to glutamic acid and Ser-221 was converted to alanine (V5-STEP 61 S160E/S221A) to mimic partially phosphorylated forms of STEP 61 . In the fourth mutant both Ser-160 and Ser-221 were converted to glutamic acid (V5-STEP 61 S160E/S221E) to mimic a completely phosphorylated form of STEP 61 . In initial studies HeLa cells were treated with phorbol ester (10 ng/ml) for varying time periods (0, 5, 10, or 15 min) to determine the temporal profile of ERK MAPK phosphorylation. Immunoblot analysis showed that phosphorylation of ERK MAPK was detectable within 5 min of phorbol ester application that peaked at 15 min (Fig. 8D). HeLa cells expressing V5-STEP 61 S160A/S221A, V5-STEP 61 S160A/S221E, V5-STEP 61 S160E/S221A, or V5-STEP 61 S160E/S221E were then treated with phorbol ester (15 min) followed by immunoprecipitation of ERK and immunoblot analysis. The data show almost complete dephosphorylation of ERK MAPK in cells transfected with V5-STEP 61 S160A/S221A (Fig. 8E, lane 1). In cells transfected with either V5-STEP 61 S160A/S221E or V5-STEP 61 S160E/ S221A a significant reduction in ERK MAPK phosphorylation was observed (Fig. 8E, lanes 2 and 3), whereas in cells transfected with V5-STEP 61 S160E/S221E did not show any reduction in ERK MAPK phosphorylation (Fig. 8E, lane 4). The incomplete dephosphorylation of ERK MAPK with the partially phosphorylated forms of STEP 61 (V5-STEP 61 S160A/S221E and V5-STEP 61 S160E/S221A) and the inability of the completely phosphorylated form of STEP 61 (V5-STEP 61 S160E/ S221E) to reduce ERK MAPK phosphorylation indicate that in intact cells both the Ser-160 and Ser-221 are involved in interacting with and dephosphorylating ERK2. The efficacy of V5-STEP 61 S160E/S221A in dephosphorylating ERK MAPK more effectively than V5-STEP 61 S160A/S221E further suggests a differential substrate binding ability of the two PKA sites with Ser-221 playing a greater role in substrate binding than Ser-160. These findings in conjunction with the observations in Figs. 1 and 4 indicate that Zn 2ϩ -induced progressive increase in the magnitude of ERK MAPK phosphorylation is regulated by sequential phosphorylation of STEP 61 at the two PKA sites.
STEP Is a Regulator of ERK2 Phosphorylation-To confirm that the loss of function of STEP 61 through hyperphosphoryla-  (30 min). a-c, samples were analyzed with anti-STEP (upper panels) or anti-phospho-ERK (middle panels, pERK1/2) antibody. Total ERK2 was analyzed by reprobing the blot with anti-ERK2 antibody (lower panels). D, STEP 61 was immunoprecipitated with anti-STEP antibody, analyzed by SDS-PAGE and probed with the anti-phospho-STEP antibody (upper panel). The blot was re-probed with anti-STEP antibody (lower panel). Each experiment was repeated at least four times.
tion is indeed responsible for Zn 2ϩ /Trk receptor mediated increase in ERK2 phosphorylation, neuronal cultures obtained from both WT and STEP KO mice were treated with Zn 2ϩ (300 M). Fig. 9A shows a time-dependent increase in ERK2 phosphorylation in the neuronal lysates obtained from the WT mice embryos that correlated with the temporal profile of STEP 61 phosphorylation. In neuronal cultures obtained from the STEP KO mice embryos, basal phosphorylation of ERK2 was significantly elevated as compared with the WT controls (Fig. 9B, lane  1). Moreover, within 5 min of exposure to Zn 2ϩ a large increase in ERK2 phosphorylation was also observed in the KO mice cultures that remained unchanged over time (Fig. 9B, lanes  2-4). This rapid increase in ERK2 phosphorylation in STEP KO mice cultures within minutes of exposure to Zn 2ϩ that remained unaltered thereafter establishes the role of STEP 61 as a regulator of Zn 2ϩ -mediated ERK2 phosphorylation. The lack of expression of STEP 61 in the STEP KO mice cultures was confirmed by immunoblotting with anti-STEP antibody (Fig.  9B).
Zn 2ϩ /TrkB-mediated STEP 61 Phosphorylation Is BDNF Dependent-Earlier studies have shown that Zn 2ϩ can activate TrkB receptor either by activation of matrix metalloproteinases (MMPs), which release pro-BDNF from cells and convert it to mature BDNF (12) or through intracellular activation of Src tyrosine kinase (33). The inability of the Src tyrosine kinase inhibitor, PP2, to attenuate Zn 2ϩ -mediated increase in STEP 61 phosphorylation (Fig. 7A) led us to hypothesize a role of extracellular BDNF in Zn 2ϩ /TrkB receptor mediated STEP 61 phosphorylation. Two different experimental approaches were taken to test this possibility. Fig. 10A shows that pharmacological inhibition of MMPs with GM6001 substantially reduced Zn 2ϩ -dependent STEP 61 and ERK2 phosphorylation. Fig. 10B shows that competitive inhibition of BDNF-TrkB receptor interaction using a function-blocking antibody against BDNF also attenuated Zn 2ϩ -dependent increase in STEP 61 and ERK2 phosphorylation. Consistent with these findings, a dose and time-dependent increase in the phosphorylation of STEP 61 and ERK2 was also observed following treatment with BDNF ( Fig.  10, C and D). The decrease in the level of STEP 61 observed at higher concentrations of BDNF (Fig. 10D, lanes 4 and 5) further suggested that BDNF regulates both STEP 61 phosphorylation and protein level. Pharmacological studies further demonstrated that inhibition of Trk receptors with K252a attenuated BDNF-induced STEP 61 and ERK2 phosphorylation (Fig. 10E).

Discussion
A key finding of the present study is that exposure of cultured neurons to high concentrations of exogenous Zn 2ϩ leads to hyperphosphorylation of STEP 61 involving multiple PKA sites. Mutational studies in cell lines further demonstrate that Ser-160 and Ser-221 are involved in the PKA-dependent hyperphosphorylation of STEP 61 . The differential abilities of the non-  -221). B and C, HeLa cells expressing V5-tagged STEP 61 -WT, -S160A, -S221A or -S160A/S221A were either (B) left untreated or (C) treated with 40 M forskolin for 5 min. D, HeLa cells were treated with phorbol ester (PMA, 10 ng/ml) for 0, 5, 10, or 15 min. E, HeLa cells expressing V5-tagged STEP 61 -S160A/S221A, S160E/S221A, S160A/S221E, or S160E/S221E were stimulated with phorbol ester (PMA, 10 ng/ml) for 15 min. B and C, expression and mobility shift of V5-tagged STEP 61 mutants in total lysate were analyzed with anti-V5 antibody. D and E, ERK was immunoprecipitated with anti-ERK2 antibody followed by immunoblot analysis of immune complexes with anti-phospho-ERK antibody (upper panels, pERK1/2). Total ERK2 was analyzed by re-probing the blots with anti-ERK2 antibody. E, expression of STEP 61 mutants was confirmed by immunoblot analysis of total lysates with anti-V5 antibody. Quantification of phosphorylated ERK2 is represented as mean Ϯ S.E. (n ϭ 4). *, p Ͻ 0.01 for phosphorylated ERK2 from STEP 61 S160E/221E. A, ⌬, p Ͻ 0.001 for phosphorylated ERK2 from STEP 61 S160A/221E; #, p Ͻ 0.01 for phosphorylated ERK2 from STEP 61 S160E/221A. phosphorylatable and the various phospho-mimicking mutants of STEP 61 to dephosphorylate ERK2 suggest that one of the functional consequences of such multi-site phosphorylation is to regulate the affinity of STEP 61 toward its substrate. The findings further suggest that the progressive loss of function of STEP 61 through phosphorylation of the two PKA sites regulates the magnitude of Zn 2ϩ -induced ERK MAPK phosphorylation.
Our findings further show that Zn 2ϩ -induced hyperphosphorylation of STEP 61 was attenuated by K252a, suggesting a role of TrkB receptor-dependent signaling in the modulation of STEP 61 phosphorylation by exogenous Zn 2ϩ . In vivo studies have further demonstrated that chronic Zn 2ϩ administration increases BDNF gene expression in the cerebral cortex resulting in increased BDNF level (42)(43)(44)(45), indicating a potential role of BDNF in mediating the actions of Zn 2ϩ . Consistent with this view, the present study shows that inhibiting BDNF-TrKB interaction, using a function-blocking antibody against BDNF, attenuates the Zn 2ϩ -induced phosphorylation of STEP 61 . Treatment with BDNF alone also leads to increase in STEP 61 phosphorylation. In contrast, inhibition of NMDAR, AMPAR and voltage-gated calcium channels that facilitate entry of extracellular Zn 2ϩ into the neurons (31,32,46,47) failed to block the Zn 2ϩ -mediated STEP 61 phosphorylation, suggesting that increases in cytosolic Zn 2ϩ level do not contribute significantly to phosphorylation of STEP 61 . The lack of increase in STEP 61 phosphorylation even when neurons were exposed to Zn 2ϩ in the presence of pyrithione further confirms this hypothesis. In fact, in the presence of Zn 2ϩ /pyrithione, a complete dephosphorylation of STEP 61 was observed suggesting that intracellular Zn 2ϩ modulates STEP 61 function by a dif-ferent mechanism than observed with extracellular Zn 2ϩ exposures.
In addition to our findings, two earlier studies have also addressed the regulation of ERK MAPK by extracellular Zn 2ϩ (12,13). The study by Hwang et al. (12) was performed in mixed cultures (neuron and astrocytes) and demonstrated TrkB-mediated increase in ERK MAPK phosphorylation using 10 M Zn 2ϩ . In contrast, our study was carried out in primary cortical neuron cultures and did not show increase in ERK MAPK phosphorylation with Zn 2ϩ concentrations below 50 M, suggesting that the increase in ERK MAPK phosphorylation observed by Hwang et al. (12) could be triggered, at least in part, by Zn 2ϩ signaling in astrocytes. Since astrocytes do not express STEP 61 , the observed increase in ERK MAPK phosphorylation by Hwang et al. (12) could not be attributed to the loss of function of STEP. On the other hand the study by He et al. (13) reported Zn 2ϩ -induced neuronal ERK MAPK phosphorylation in postzinc exposure time periods. Our findings now reveal an important additional phase of ERK MAPK phosphorylation during Zn 2ϩ exposure and provide a molecular basis for such sustained activation of ERK MAPK. We also observed that inhibition of TrkB receptor signaling attenuated Zn 2ϩ -induced phosphorylation of ERK2. Complementary studies in cells treated with BDNF showed increases in ERK MAPK phosphorylation, which was attenuated by TrkB receptor inhibition. In contrast, Zn 2ϩ -induced ERK MAPK phosphorylation was not blocked either by inhibition of channels/receptors that facilitates entry of Zn 2ϩ inside the neurons or by inhibition of Src kinase that transactivate TrkB in a BDNF-independent manner. Taken together, these findings suggest that by stimulating BDNF pro- FIGURE 9. Role of STEP in the Zn 2؉ -dependent phosphorylation of ERK. Neuronal cultures obtained from (A) wild-type and (B) STEP KO mice were treated with 300 M Zn 2ϩ for the specified times. Phosphorylation of ERK2 was analyzed by immunoblotting with anti-phospho-ERK antibody (upper panels, pERK1/2). Total ERK2 or STEP was analyzed by reprobing the blots with either anti-ERK2 (middle panels) or anti-STEP antibody (lower panels). Quantification of phosphorylated ERK2 is represented as mean Ϯ S.E. (n ϭ 4). A, *, p Ͻ 0.001 for phosphorylated ERK2 from untreated control in WT cultures; #, p Ͻ 0.001 for phosphorylated ERK2 from 5 min Zn 2ϩ treatment in WT cultures; B, ¶, p Ͻ 0.01 for phosphorylated ERK2 from untreated control in STEP KO cultures.
Although BDNF-TrkB receptor mediated intracellular signaling is mainly regarded as a contributor to synaptic development and plasticity (48), this pathway has also been shown to play a role in neurodegeneration (49) and promote cell death in vitro (50,51). In addition, several studies have indicated a role of BDNF-TrkB receptor signaling in the development of epilepsy (52)(53)(54)(55)(56). One explanation for this differential response to BDNF-induced TrkB receptor stimulation is that potentiation of synaptic efficacy requires a transient release of BDNF that elicits a signaling response, which is different from that observed following sustained BDNF release. Consistent with this hypothesis a recent study showed that acute and sustained BDNF application elicited different patterns of surface TrkB receptor expression and ERK MAPK phosphorylation (57). Acute administration of BDNF reduced surface TrkB receptor expression and induced transient increase in ERK MAPK phosphorylation. In contrast, progressive increase in BDNF application had little effect on surface expression of TrkB receptors, and induced a slow and sustained increase in ERK MAPK phosphorylation. Previous studies have also demonstrated that the small GTPases Ras and Rap1 differentially control the kinetics of ERK MAPK phosphorylation (57,58). Ras-dependent signaling to ERK MAPK is transient following acute BDNF treatment, whereas Rap1-dependent signaling to ERK MAPK is sustained following prolonged exposure to BDNF. Our findings now indicate that TrkB receptor mediated activation of the cAMP/PKA leads to progressive increase in phosphorylation of STEP at multiple sites resulting in loss of function of STEP. The temporal profile of Zn 2ϩ -mediated increase in ERK MAPK phosphorylation also strongly correlates with the pattern of STEP 61 phosphorylation. Thus it appears that simultaneous activation of the Rap1 signaling pathway and inhibition of the tyrosine phosphatase STEP may be essential for the sustained phosphorylation of ERK MAPK. While the Rap1 signaling pathway is involved in triggering ERK MAPK phosphorylation, the loss of function of STEP through phosphorylation facilitates maintenance of sustained ERK MAPK phosphorylation. This interpretation is further supported by our studies in neuronal cultures obtained from STEP KO mice demonstrating a rapid increase in the magnitude of ERK MAPK phosphorylation within minutes of exposure to Zn 2ϩ that remained unchanged over time. These findings provide a molecular basis for the sustained activation of ERK MAPK in Zn/BDNF/Trk receptor mediated signaling.
In conclusion, our findings present the first evidence that in addition to the neurotransmitters dopamine and glutamate, transition metal ions and neurotrophins can also modulate the function of STEP 61 . The study also establishes the signaling mechanism underlying Zn 2ϩ -induced regulation of STEP 61  (30 min) in the absence or presence of anti-BDNF antibody (C) treated with different concentrations of BDNF for 30 min, (D) treated with BDNF (1 ng/ml) for the specified times, or (E) pre-incubated with K252a (100 nM) for 15 min before treatment with BDNF (1 ng/ml) for 30 min. Blots were probed with anti-STEP (upper panels) or anti-phospho-ERK (middle panels, pERK1/2) antibody. Total ERK2 was analyzed by re-probing with anti-ERK2 antibody (lower panels). Each experiment was repeated at least four times.
function and provides a novel mechanism of regulating the magnitude of ERK MAPK activation by STEP 61 . Evaluation of the phosphorylation status of STEP 61 in animal models of Zn 2ϩ toxicity may therefore help determine whether targeting STEP may attenuate ERK-MAPK-dependent progression of brain damage in neurodegenerative disorders.
Author Contributions-R. P. and S. P. conceived and coordinated the study, analyzed the data, prepared the figures and wrote the paper. R. P. and S. R. performed the experiments. C. W. S. analyzed the data, provided important intellectual input and revised the manuscript. All authors approved the final version of the manuscript.