The Csk Homologous Kinase Associates with TrkA Receptors and Is Involved in Neurite Outgrowth of PC12 Cells*

Csk homologous kinase (CHK), a member of the Csk regulatory tyrosine kinase family, is expressed primarily in brain and hematopoietic cells. The role of CHK in the nervous system is as yet unknown. Using PC12 cells as a model system of neuronal cells, we show that CHK participates in signaling mediated by TrkA receptors. CHK was found to be associated with tyrosine-phosphorylated TrkA receptors in PC12 cells upon stimulation with NGF. Binding assays and far Western blotting analysis, using glutathione S-transferase fusion proteins containing the Src homology 2 (SH2) and SH3 domains of CHK, demonstrate that the SH2 domain of CHK binds directly to the tyrosine-phosphorylated TrkA receptors. Site-directed mutagenesis of TrkA cDNA, as well as phosphopeptide inhibition of the in vitro interaction of the CHK-SH2 domain or native CHK with TrkA receptors, indicated that the residue Tyr-785 on TrkA is required for its binding to the CHK-SH2 domain upon NGF stimulation. In addition, overexpression of CHK resulted in enhanced activation of the mitogen-activated protein kinase pathway upon NGF stimulation, and microinjection of anti-CHK antibodies, but not anti-Csk antibodies, inhibited neurite outgrowth of PC12 cells in response to NGF. Thus, CHK is a novel signaling molecule that participates in TrkA signaling, associates directly with TrkA receptors upon NGF stimulation, and is involved in neurite outgrowth of PC12 cells in response to NGF.

Nerve growth factor (NGF) 1 regulates the survival, development, and differentiation of the sympathetic and sensory neurons in the peripheral nervous system and the differentiation of certain cholinergic neurons in the central nervous system (1,2). In addition, NGF promotes neuronal differentiation of the rat pheochromocytoma cell line, PC12, which has been used extensively to investigate ligand-receptor interactions and cellular differentiation in response to NGF (3,4). Upon stimulation with NGF, these cells acquire the phenotype of sympathetic neurons as characterized by neurite outgrowth and persistent activation of the ERK family of mitogen-activated protein (MAP) kinase (5). Two cell surface receptors for NGF have been identified: the receptor tyrosine kinase, TrkA, and the lowaffinity neurotrophin receptor, p75 NTR (6 -9). NGF exerts its growth-and survival-promoting effects on neurons through activation of TrkA and subsequent biochemical events that ultimately influence the expression of various genes, including those encoding ion channels, neurotransmitter-synthesizing enzymes, and cytoskeletal components (10). The binding of NGF to the p140 TrkA receptor induces dimerization of TrkA receptors and stimulates rapid tyrosine autophosphorylation of the receptor (11). The phosphotyrosines on activated TrkA serve as docking sites for signaling substrates such as phospholipase C-␥1 (PLC-␥1), phosphatidylinositol 3-kinase (PI-3 kinase), and the Shc adaptor protein (6,12). These molecules trigger kinase cascades resulting in the phosphorylation and activation of transcription factors that direct gene expression. Shc triggers the activation of Ras and the subsequent sequential phosphorylation and activation of the kinases Raf, mitogen-and extracellular-regulated kinase, and ERK (13,8). The Ras-ERK pathway plays a major role in the activation of transcriptional events by NGF and in NGF-induced neuronal differentiation. Mutation analysis of TrkA has defined critical tyrosines that specifically regulate the activities of PLC-␥1, PI-3 kinase and Shc (14 -18). PLC-␥1 and Shc regulation appear to play a major role in NGF-mediated neurite outgrowth (17,18). In addition, sustained PI-3 kinase activity is necessary for the neurite outgrowth of PC12 cells induced by NGF (19).
The Csk homologous kinase (CHK) (previously referred to as megakaryocyte-associated tyrosine kinase (MATK)) is a recently identified protein tyrosine kinase that shares high homology with Csk (COOH-terminal SRC kinase). CHK was independently identified as MATK (20,21), Lsk (22), Hyl (23), Ctk (24), Ntk (25), and Batk (26). Like Csk, CHK contains Src homology 3 (SH3), SH2, and tyrosine kinase domains and lacks the Src family NH 2 -terminal myristylation and autophosphorylation sites (20,21). Csk is ubiquitously expressed, whereas expression of CHK is restricted to hematopoietic cells and the nervous system. The expression of CHK in brain increases postnatally, whereas the expression of Csk decreases with age (27). Although the function of CHK is still unclear, recent studies indicated that, unlike Csk, CHK interacts with receptor tyrosine kinases, such as c-Kit in megakaryocytes (28,29) or the ErbB-2/neu receptor in breast cancer cells (30,31) via its SH2 domain. Because CHK is abundantly expressed in the nervous system, we sought to identify the signaling pathways that involve CHK and to characterize its function in neuronal cells. In this report, we identified and characterized the association of CHK in TrkA signaling and investigated CHK involvement in neuronal differentiation of PC12 cells.

EXPERIMENTAL PROCEDURES
Antibodies-Anti-GST monoclonal antibodies were produced as described previously (32). Affinity-purified polyclonal anti-TrkA antibodies against the COOH-terminal peptide of the TrkA receptor were purchased from Oncogene Science (Cambridge, MA). Anti-CHK, anti-Csk, anti-p85/PI-3 kinase, anti-ERK1, and anti-ERK2 antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Monoclonal anti-phosphotyrosine (PY-20) was obtained from Transduction Laboratories (Lexington, KY). The PhosphoPlus MAP kinase antibody kit was obtained from New England BioLabs (Beverly, MA). Mouse 7S NGF was purchased from Upstate Biotechnology (Lake Placid, NY). Glutathione-Sepharose 4B was purchased from Amersham Pharmacia Biotech. Reagents for electrophoresis were obtained from Bio-Rad. Horseradish peroxidase-conjugated anti-mouse Ig antibodies, horseradish peroxidase-conjugated anti-rabbit Ig antibodies, and enhanced chemiluminescence (ECL) reagents were purchased from Amersham Pharmacia Biotech.
Cell Culture and Stimulation-PC12 cells were grown as described (3). PC12-TrkA cells were stably transfected with TrkA cDNA (33). These cells expressed 1.3 ϫ 10 5 TrkA receptors/cell. Both PC12 and PC12-TrkA cells were grown in RPMI 1640 medium containing 10% heat-inactivated horse serum and 5% fetal bovine serum. The cells were grown on 10-cm dishes coated with poly-L-lysine. For stimulation, cells were stimulated with 100 ng/ml NGF for the indicated periods at 37°C.
To establish PC12 cells that stably overexpress CHK (PC12-CHK), the full-length CHK cDNA (1.6 kilobases) was subcloned into the pcDNA3 vector (Invitrogen, San Diego, CA). Transfection of the CHK cDNA into PC12 cells was performed using LipofectAMINE. The transfected cells were selected in 750 g/ml Geneticin. Positive transfectants were chosen based on the expression of CHK protein upon immunoblotting with anti-CHK antibodies. PC12-CHK cells were grown in RPMI 1640 medium containing 10% horse serum and 5% fetal bovine serum.
Precipitation with GST Fusion Proteins and Far Western Blotting-GST fusion proteins containing the CHK-SH3 domain, the CHK-NH 2terminal domain plus SH3 domain (NH 2 -SH3), the CHK-SH2 plus SH3 domains (CHK-SH2-SH3) or GST alone were prepared as described previously (30). The starved PC12-TrkA cells were stimulated with 100 ng/ml NGF and then lysed and precipitated with various GST fusion proteins. 10 g of GST fusion protein bound to glutathione-Sepharose 4B were used for each binding assay. After lysate protein content was normalized and precipitated with GST fusion proteins, bound proteins were immunoblotted with PY-20, affinity-purified polyclonal anti-TrkA antibody, or anti-GST antibody. The blots were developed using the ECL system. For far Western analysis, starved PC12-TrkA cells were stimulated with 100 ng/ml NGF for the indicated periods, lysed, and immunoprecipitated with anti-TrkA antibody. The immunoprecipitates were then separated on 7% SDS-PAGE gels, transferred onto membranes, and subjected to far Western blotting. Ten g of GST fusion proteins were used for each binding assay.
Construction of Wild-type and Mutant TrkA cDNA Constructs and Transfection Experiments-Site-directed mutagenesis of full-length TrkA cDNA in the pUC19 vector was performed using a Transformer site-directed mutagenesis kit (CLONTECH, Palo Alto, CA) according to the manufacturer's protocol. In order to change selected tyrosines (Tyr) to phenylalanine (Phe), we used the following synthetic oligonucleotides: 5Ј-CAC ATC ATC GAG AAC CCA CAA TTC TTC AGT GAT GCC TGT GTT CAC-3Ј (Y490F), 5Ј-TGC CCA CCA GAG GTC TTC GCC ATC ATG CGG GGC-3Ј (Y751F), and 5Ј-CAG GCA CCT CCT GTC TTC CTG GAT GTC CTG GGC-3Ј (Y785F). The mutations were confirmed by DNA sequencing. The mutant and wild-type TrkA inserts were excised from pUC19 vectors and subcloned into the pcDNA3 vector. Transfection into COS-7 cells was performed using LipofectAMINE (Life Technologies, Inc.) according to the manufacturer's protocol.
Inhibition of the CHK-SH2 Domain/TrkA Interaction by Phosphopeptides-Tyrosine-phosphorylated synthetic peptides were obtained from the Molecular Biology Core Facility at Dana Farber Cancer Institute. Peptides were analyzed for purity by high pressure liquid chromatography, mass spectroscopy, and amino acid analysis. Two phosphotyrosine peptides corresponding to autophosphorylation sites of TrkA, including the Shc association site and the PLC-␥1 association site, were synthesized. In addition, nonphosphorylated peptide corresponding to the PLC-␥1 association site was also synthesized. Unstimulated and NGF-stimulated PC12-TrkA cell lysates were added to the GST-CHK-SH2 fusion protein (10 g/incubation), which was preincubated with each phosphopeptide. The final concentration of phosphopeptides was 100 M for each incubation. Lysates were incubated for 1 h at 4°C and then precipitated by the addition of glutathione-Sepharose 4B for 30 min at 4°C. Precipitates were washed and separated by 7% SDS-PAGE.
MAP Kinase Assay-MAP kinase assay was performed as described (34). Briefly, starved PC12 and PC12-CHK cells (three independent clones that overexpress CHK) were stimulated with NGF (100 ng/ml). Cells were then lysed and immunoprecipitated with anti-ERK1 and anti-ERK2 antibodies (Santa Cruz Biotechnology). MAP kinase activity was assayed with myelin basic protein as a substrate in the presence of [␥-32 P]ATP. All reactions were allowed to proceed for 10 min at 30°C. The p44/42 MAP kinase activity was also measured in total lysates using the p44/42 MAP kinase assay kit (New England BioLabs). Phospho-p44/42 MAP kinase monoclonal antibody was used to immunoprecipitate the active p44/42 MAP kinase from the cell extracts, and then an in vitro kinase assay was performed using Elk-1 protein as a substrate.
Microinjection-Synchronized PC12 cells were microinjected with purified CHK-specific antibodies (at a concentration of 100 g/ml) in microinjection buffer with GFP vector plasmid (pEGFP-C2) (CLON-TECH, Palo Alto, CA) as a marker in a final concentration of 100 g/ml. IgGs were purified from control antiserum, Csk and CHK antisera using protein G columns (Mab Trap, Amersham Pharmacia Biotech). Pooled fractions containing electrophoretically pure IgGs were dialyzed against phosphate-buffered saline and concentrated using Centricon concentrators (Amicon Corp., Danvers, MA). Following microinjection, cells were incubated for 2 h at 37°C. Cells were refed with medium alone or medium containing 0.1% horse serum overnight. The medium was then changed to RPMI medium containing 10% horse serum and 5% fetal bovine serum, with or without NGF (100 ng/ml) as indicated. Cells were monitored and analyzed after 24, 36, and 48 h. Differentiated cells were defined as cells with refractile cell bodies extending at least two processes, one of which had to be longer than the diameter of the cell body. Subsequent phase-contrast micrographs were recorded on Technical Pan film (2415) (Eastman Kodak Co.).

CHK Is Associated with Activated TrkA Receptors upon Stimulation with NGF-Because
CHK is a second member of the Csk family and both kinases are expressed in brain, we investigated whether CHK and/or Csk might be involved in one of the main signaling pathways in neuronal cells that is mediated by the neurotrophin NGF. To identify proteins that associate with CHK and Csk in PC12 cells, we analyzed anti-CHK immunoprecipitates from untreated or NGF-stimulated cells. The PC12 cells were starved for 36 h and stimulated with 100 ng/ml NGF for 10 min. Cells were then lysed and immunoprecipitated with either anti-CHK antibodies, affinity-purified anti-TrkA antibodies, anti-Csk antibodies, or control antibodies. Upon NGF stimulation, anti-CHK antibodies, but not anti-Csk or control antibodies, co-immunoprecipitated the tyrosine-phosphorylated protein, p140, corresponding to the activated TrkA receptor (Fig. 1, A and B), whereas anti-TrkA antibodies coimmunoprecipitated CHK protein (Fig. 1C) but not Csk (Fig.  1D). Therefore, these results suggest that CHK, but not Csk, is associated with activated TrkA receptors upon NGF stimulation.
Characterization of the Association of the CHK-SH2 Domain with Activated TrkA Receptors-In order to determine which domain of CHK interacts with the activated TrkA receptors, we used GST fusion proteins containing the SH2 domain of CHK (CHK-SH2). Serum-starved PC12 cells were stimulated with 100 ng/ml NGF for the indicated periods, lysed, and then precipitated with the GST fusion protein containing the CHK-SH2 domain. The CHK-SH2 domain precipitated tyrosine-phosphorylated TrkA ( Fig. 2A) upon stimulation with NGF. The phosphorylation of TrkA upon NGF stimulation was rapid and became prominent at 5 min. The association of CHK with TrkA receptors was also observed after 3 days of NGF stimulation, indicating that this association is a long-term one, based upon the length of the stimulation (Fig. 2B). Similar results of the association of CHK with activated TrkA receptors were obtained using PC12-TrkA cells (PC12 cells stably transfected with TrkA receptors) (data not shown).
The potential involvement of other domains of CHK in the interaction with TrkA was examined using GST fusion proteins containing the CHK-SH3 domain, the CHK-SH3-SH2 domain, or GST protein alone. NGF-stimulated PC12 cell lysates were incubated with the different GST fusion proteins, analyzed by 7% SDS-PAGE, and immunoblotted with PY-20 or anti-TrkA antibodies. Anti-phosphotyrosine and anti-TrkA Western blotting revealed that the CHK-SH3-SH2 domain precipitated activated TrkA receptors upon NGF stimulation (Fig. 3). However, the CHK-SH3-GST or GST alone or a nonrelevant SH2 domain did not precipitate any proteins from the same lysates.
These results indicate that the CHK-SH2 domain specifically precipitated activated TrkA receptors upon NGF stimulation.
CHK Is Directly Associated with Activated TrkA Receptors-In order to determine whether the binding between activated TrkA receptors and the CHK-SH2 domain is direct or indirect, we performed far Western blotting analysis. The phosphorylation of TrkA upon NGF stimulation was rapid, and reached maximum levels after 20 min (Fig. 4, A and B). In the absence of NGF, the SH2 domain of CHK did not bind to TrkA, as observed by far Western blotting (Fig. 4D). However, upon NGF stimulation, binding of the CHK-SH2 domain to TrkA was observed at 2 min. This association was maintained up to 60 min after stimulation with NGF (Fig. 4D). In contrast, the CHK-SH3 domain or GST fusion protein alone did not bind to activated TrkA (Fig. 4, C and E). These results indicate that the The transfected COS-7 cells were starved and then stimulated with 100 ng/ml NGF for 10 min. Unstimulated and NGF-stimulated cell lysates were divided into two aliquots, and either precipitated with the GST fusion protein containing the CHK-SH2 domain or immunoprecipitated with anti-TrkA antibodies. Tyrosine phosphorylation of wild-type TrkA upon stimulation with NGF was observed in transfected cells. Furthermore, although the CHK-SH2 domain did not precipitate the unstimulated wild-type TrkA receptor, it did precipitate the tyrosine-phosphorylated wild-type TrkA receptor upon NGF stimulation (Fig. 5A). These results demonstrate that the CHK-SH2 domain interacts with activated TrkA receptors upon NGF stimulation in TrkA-transfected COS-7 cells.
Experiments in COS-7 cells transfected with TrkA wild-type or mutant constructs revealed that the CHK-SH2 domain precipitated TrkA Y490F and Y751F in the same manner as the wild-type TrkA receptor (Fig. 5B). The same results were ob-served in the binding of the CHK-SH2 domain to TrkA Y490F/ Y751F. However, the CHK-SH2 domain could not precipitate mutant TrkA that had a Y785F mutation, such as TrkA Y785F, TrkA Y490F/Y785F, or TrkA Y751F/Y785F. These results indicate that the CHK-SH2 domain binds to phosphorylated Tyr-785 on the TrkA receptor. In order to confirm the expression of wild-type or mutant TrkA, the same amount of cell lysates was first immunoprecipitated and then immunoblotted with anti-TrkA antibodies. Expression levels of wild-type and mutant TrkA were similar (Fig. 5C).
Furthermore, we have studied the ability of synthetic phosphopeptides to compete for the binding of the CHK-SH2 domain to TrkA receptors. We synthesized two kinds of tyrosinephosphorylated peptides derived from the Shc binding site and the PLC-␥1 association site of TrkA receptors. A peptide containing the Shc binding site of TrkA receptors (Y*FSDTCV, where the asterisk indicates a phosphotyrosine peptide) could not compete the binding of the SH2 domain of CHK to the activated TrkA receptors (Fig. 5D). The non-tyrosine-phosphorylated control peptide (SYLDVLG) also failed to compete the binding of the SH2 domain of CHK to the activated TrkA receptors. On the other hand, a peptide containing the PLC-␥1 association site of TrkA receptors (SY*LDVLG) was able to compete the binding of the SH2 domain to the activated TrkA receptors (Fig. 5D). To further test the binding of CHK to the Tyr-785 site, we linked the tyrosine-phosphorylated SY*LDVLG peptide or the nonphosphorylated SYLDVLG peptide to Affi-Gel 15 beads, and the association of either the CHK-SH2 GST fusion protein or native CHK to the beads was analyzed. GST-CHK-SH2 was associated in a phosphotyrosine dependent manner to the phosphorylated SY*LDVLG peptide (data not shown). Similar specificity was observed when we tested the association of native CHK to the peptide beads. The phosphorylated SY*LDVLG was able to associate with native CHK from extracts of PC12 or PC12-CHK cells, whereas the nonphosphorylated peptide SYLDVLG did not associate with native CHK (data not shown). This specificity was in agreement with the peptide inhibition experiments. Taken together, these results, along with the site-directed mutagenesis of TrkA and the far Western blotting analysis, indicate that CHK binds to Tyr-785 of TrkA upon NGF stimulation.
Overexpression of CHK Results in Enhanced Activation of the MAP Kinase Pathway upon NGF Stimulation-Differentiation of PC12 cells upon NGF treatment requires activation of the MAP kinase pathway (17,18,35). Because CHK is associated with TrkA receptors upon NGF stimulation, we assessed the effects of overexpression of CHK on activation of the p44 and p42 MAP kinases following stimulation with NGF. We stably transfected PC12 cells with CHK-pcDNA3neo, generating several PC12-CHK clones that overexpress CHK. PC12 cells and PC12-CHK clones (three independent clones of PC12-CHK) were cultured in the absence or the presence of NGF (100 ng/ml) for the indicated times, and MAP kinase activity using myelin basic protein as a substrate was assayed. Although no MAP kinase activity was detected in control PC12 or PC12-CHK cells, an increase in this activity was observed in PC12 cells upon NGF stimulation, with a peak activity at 10 min, followed by a subsequent decline in activity (Fig. 6A). In PC12 cells overexpressing CHK, MAP kinase activity was enhanced as compared with untreated PC12 cells, and this activity was maintained for longer periods of up to 60 min (Fig. 6A). In addition, p44/42 MAP kinase activity was measured using antibodies specific to the activated p44/42 MAP kinases. The phosphokinase antibodies detect p44 and p42 MAP kinases only when they are activated by phosphorylation at Tyr-204 (36). PC12 cells and three independent PC12-CHK clones were cultured in the absence or presence of NGF (100 ng/ml) for the indicated times, and MAP kinase activity was assayed. In PC12 cells overexpressing CHK, tyrosine phosphorylation of MAP kinases was enhanced compared with untreated PC12 cells, and this phosphorylation was maintained for longer periods of up to 60 min (Fig. 6B). These results indicate a role of CHK in enhancing MAP kinase activity in PC12 cells upon NGF stimulation and thus its involvement in PC12 cell differentiation.
Microinjection of CHK-specific Antibodies Inhibited Neuronal Differentiation of PC12 Cells-To determine the requirement for the binding of the CHK-SH2 domain to TrkA receptors during neurite outgrowth, we examined the effects of decreasing the intracellular levels of CHK by microinjecting purified anti-CHK antibodies, anti-Csk antibodies, control preimmune rabbit antiserum, or control monoclonal antibody into living PC12 cells. Differentiation of these cells upon NGF stimulation was inhibited, as indicated by microinjection of purified anti-CHK antibodies, by morphological changes in neurite outgrowth and by rapid and dramatic changes in cell shape (cell rounding) (Fig. 7). When control antibodies or anti-Csk kinase antibodies were microinjected into PC12 cells followed by NGF stimulation, these cells were differentiated, as indicated by the appearance of prominent neurite outgrowth similar to that observed for the PC12 cells treated with NGF alone (Fig. 7). These data suggest that CHK is involved in neurite outgrowth of PC12 cells. DISCUSSION CHK protein tyrosine kinase is found predominantly in neuronal and hematopoietic cells. In the present study, we demonstrated direct binding of CHK to the TrkA receptors, upon NGF stimulation. Furthermore, we identified the binding site of the CHK-SH2 domain to the activated TrkA receptors as residue Tyr-785. In addition, overexpression of CHK resulted in an enhanced activation of the MAP kinase pathway upon NGF stimulation of PC12 cells, whereas microinjection of CHK-specific antibodies inhibited the neurite outgrowth of PC12 cells upon NGF stimulation. These results indicate for the first time that CHK, but not Csk, is involved in TrkA signaling and in neurite outgrowth of PC12 cells.
The association of the SH2 domain of CHK with activated FIG. 6. Effects of CHK overexpression on MAP kinase activation. A, quantification of MAP kinase activity in PC12 cells and PC12-CHK cells: PC12 and PC12-CHK cells were serum-starved and then stimulated with NGF (100 ng/ml) for the indicated periods. Total cell lysates were immunoprecipitated with anti-ERK1 and anti-ERK2 antibodies and assayed for MAP kinase activity. The phosphorylation level of the substrate, myelin basic protein, was determined. The plot shows fold increase Ϯ S.D. of MAP kinase activity over the basal level. B, activation of p44/42 MAP kinases: PC12 and PC12-CHK cells were serum-starved and then stimulated with NGF (100 ng/ml) for the indicated periods. Total cell lysates were prepared, and 10 g of protein were resuspended in SDS sample buffer. The samples were separated by 10% SDS-PAGE. The resolved proteins were analyzed by immunoblotting with anti-phospho-MAP kinase antibody using an antiserum specifically recognizing tyrosine-phosphorylated MAP kinases (New England BioLabs). Columns represent the means from densitometric scanning of three separate experiments. *, p Ͻ 0.05, representing the difference in MAP kinase activation of PC12-CHK clones versus PC12 cells using Student's paired or unpaired t test.
TrkA was demonstrated using GST fusion proteins containing various domains of the CHK molecule. The association of CHK with TrkA occurred within the same time period in which tyrosine phosphorylation of TrkA was observed upon NGF stimulation. The phosphorylation of TrkA upon NGF stimulation was rapid and declined slowly, as previously reported (11). Our results indicated that the CHK-SH2 domain specifically interacts with the tyrosine-phosphorylated TrkA receptors upon NGF stimulation. The CHK-SH2 domain also precipitated other proteins, namely p112, p89, and p77. Although we were not yet able to identify these proteins, the possibility of their involvement in the association of CHK with TrkA signaling remains to be investigated in future studies.
Far Western blotting revealed that the binding between activated TrkA and the CHK-SH2 domain is direct (Fig. 4). The SH2 domain was found to bind to activated TrkA receptors, which were tyrosine-phosphorylated upon NGF stimulation, whereas the SH3 domain or GST fusion protein failed to bind to TrkA. Therefore, the CHK-SH2 domain can bind to the activated and phosphorylated TrkA.
CHK is undetectable in early embryos and begins to be expressed around E15 (27,40,41). It then increases progressively up to the time of birth and remains high in the adult central nervous system. CHK expression correlates with late stage development and neuronal differentiation. In contrast, Csk is expressed throughout embryonic development and remains high in the central nervous system until birth. Csk is then dramatically down-regulated in the adult brain, except in the olfactory bulb (27,41). Such diametrically opposite tempo- FIG. 7. Microinjection of anti-CHK antibody inhibited NGF-induced neuronal differentiation of PC12 cells. Subconfluent PC12 cells were subjected to single cell microinjection. A, PC12 cells were microinjected with purified anti-rabbit IgG (a-c) and with purified anti-CHK antibody (d-f). pEGFP-C2 vector was co-injected as a marker. Microinjected cells were grown in the absence of NGF (a and d) and in the presence of NGF (100 ng/ml) (b, c, e, and f) for 48 h and then analyzed for morphological changes under a UV microscope. Cells with green fluorescence (c and f) are microinjected cells shown in the same field as in b and e. B, microinjection of anti-mouse Csk antibody (as a negative control) and pEGFP-C2 into PC12 cells in the presence of NGF (100 ng/ ml) for 48 h. Arrows indicate microinjected cells. C, microinjection of control nonrelevant monoclonal antibodies (M4) and pEGFP-C2 into PC12 cells in the presence of NGF (100 ng/ml) for 48 h. a, microinjected cells were analyzed under a UV light microscope. b, cells in the same field as in a were analyzed using an immunofluorescent microscope. ral expression patterns of Csk and CHK suggest distinct roles for both these proteins, presumably mediated via different substrates. Csk is known to inhibit Src family kinases, and this inhibition results in a down-regulation of the MAP kinase pathway (42). However, our results indicate that overexpression of CHK resulted in enhanced activation of the MAP kinase pathway in PC12 cells upon NGF stimulation. In addition, we have observed that the neuronal differentiation of PC12 cells induced by NGF was inhibited by microinjection of anti-CHK antibodies, but not anti-Csk antibodies (Fig. 7), suggesting that CHK and Csk regulate different targets in the signaling pathways of neuronal cells. Thus, our findings strongly suggest that CHK plays a role in regulating the differentiation of PC12 cells. Future studies will further investigate the molecular mechanisms of the involvement of CHK in neuronal differentiation.