The Ras Suppressor, RSU-1, Enhances Nerve Growth Factor-Induced Differentiation of PC12 Cells and Induces P21cip Expression1

Vol. 10, 555–564, August 1999 Cell Growth & Differentiation 555

The Ras Suppressor, RSU-1, Enhances Nerve Growth Factor-induced Differentiation of PC12 Cells and Induces p21CIP Expression1

L. Masuelli, S. Ettenberg, F. Vasaturo, Introduction 2 3 K. Vestergaard-Sykes, and M. L. Cutler Rsu-1, which was isolated based on its ability to suppress

Department of Pathology, Uniformed Services University of the Health Ras transformation, encodes a Mr 33,000 protein that con- Sciences, Bethesda, Maryland 20814 [L. M., S. E., F. V., M. L. C.]; tains a series of leucine-rich amphipathic repeats homolo- Department of Experimental Medicine, First University of Rome, Rome 00161, Italy [L. M.]; and United States Department of Agriculture gous to the leucine repeats found in yeast adenylyl cyclase Laboratories, Beltsville, Maryland 20705 [K. V-S.] (1–3). These repeats are required for the activation of adenylyl cyclase by Ras in Saccharomyces cerevisiae (4, 5), and similar repeats are required for a Ras-induced differentiation Abstract pathway in Caenorhabditis elegans (3). Rsu-1 binds to Raf-1 The Rsu-1 Ras suppressor gene was isolated based on in in vitro binding assays, suggesting that Rsu-1 may stabi- its ability to inhibit v-Ras transformation. Using Rsu-1 lize Ras-Raf association and/or inhibit the association of Ras transfectants of the pheochromocytoma cell line PC12, with other effectors. Rsu-1 expression inhibited RasGAP we demonstrated previously that Rsu-1 expression activity, resulting in an increase in Ras-GTP, and inhibited the inhibited Jun kinase activation but enhanced Erk2 activation of Jun kinase by EGF4 (6). Activation of the Erk activation in response to epidermal growth factor. In kinase pathway was not inhibited by Rsu-1 expression; in the present study, the Rsu-1 PC12 transfectants were contrast, Rsu-1 expression resulted in increased stimulation used to investigate the role of Rsu-1 in nerve growth of Erk in response to EGF (6). factor (NGF)- and v-Ki-ras-mediated neuronal The infection of PC12 pheochromocytoma cells with Ki- differentiation. NGF-induced neurite extension was MSV or Ha-MSV and microinjection of activated Ras p21 enhanced, not inhibited, by the expression of Rsu-1 in results in the induction of a program of differentiation phe- PC12 cells. The activation of Erk kinase activity in notypically demonstrated by neurite outgrowth (7, 8). This response to NGF was sustained longer in the Rsu-1 program of differentiation is biochemically characterized by transfectants compared with the vector control cells. induction of a specific gene expression program (9, 10). The During NGF-mediated differentiation, an increase in the stimulation of PC12 cells with NGF results in the activation of expression of specific mRNAs for the early response a program of neuronal differentiation (11). NGF activation of genes Fos, cJun, and NGF1a was detected in both the the tyrosine kinase activity of trk, as part of a high affinity vector control and Rsu-1 transfectants. The expression NGF receptor, results in tyrosine phosphorylation of trk and of the differentiation-specific genes VGF8 and SCG10 activation of downstream effectors (12–14). This activation was similar in Rsu-1 transfectants compared with the requires Ras (15, 16) and depends upon sustained activation vector control cells. The induction of Rsu-1 expression of the Erk pathway (17, 18). In the present study, we tested in these cell lines did not inhibit v-Ki-ras-induced the ability of Rsu-1 expression, which can inhibit Ras-in- differentiation, as measured by neurite extension. duced transformation and Jun kinase activation, to inhibit These data suggest that although Rsu-1 blocked some v-Ki-Ras-induced and NGF-induced differentiation of PC12 Ras-dependent response(s), these responses were not cells. Previous studies have demonstrated that increased required for differentiation. Moreover, the induction of expression of RasGAP, resulting in a decrease in Ras-GTP Rsu-1 expression in the PC12 clones resulted in levels, inhibited NGF induced-differentiation (19) and that WAF/CIP growth inhibition and p21 expression. Hence, Jun kinase activation was not necessary for differentiation in Rsu-1 expression enhances NGF-induced response to NGF (20). Hence, overexpression of Rsu-1 dur- differentiation while inhibiting the growth of cells. ing NGF stimulation of PC12 cells will assess the importance of elevated and sustained activation of Erk kinase as well as the requirement for activation of other pathways downstream Received 2/24/99; revised 4/14/99; accepted 7/1/99. of Ras for the initiation of the differentiation program. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked Previous studies of NGF-induced differentiation of PC12 advertisement in accordance with 18 U.S.C. Section 1734 solely to indi- cells indicated there are numerous changes brought about cate this fact. by cell cycle regulatory proteins that can promote growth 1 This work was supported in part by a grant from the Childhood Brain Tumor Foundation (to M. L. C.) and by DOD Grant DAMD 17-97-1-7089 arrest in G1. NGF-stimulated PC12 cells exhibit an increase (to M. L. C.). 2 Present address: Department of Surgical Oncology, Medical College of Virginia, Richmond, VA. 3 To whom requests for reprints should be addressed, at B3122, Uni- 4 The abbreviations used are: EGF, epidermal growth factor; NGF, nerve formed Services University of Health Sciences, 4301 Jones Bridge Road, growth factor; cdk, cyclin-dependent kinase; GPDH, glyceraldehyde-3- Bethesda, MD 20814. Phone: (301) 295-3453; Fax: (301) 295-1640; E- phosphate dehydrogenase; MMTV, mouse mammary tumor virus; MuLV, mail: [email protected]. murine leukemia virus. 556 Rsu-1 Enhances PC12 Differentiation

in p21WAF/CIP when propagated in serum-containing media (21). However, stimulation of cells maintained in low levels of serum did not require induction of p21WAF/CIP for differentiation (22). Recent studies on the role of p21CIP in cell cycle regulation indicated that activation of Raf can result in induction of p21CIP expression (23, 24). Therefore, in addition to the activation of the Ras-Raf pathway in response to NGF, differentiation relies on induction of p21WAF/CIP or serum deprivation to inhibit cell cycle progression. In this study, Rsu-1 transfectants of PC12 cells, which contain Rsu-1 under the control of a dexamethasone-inducible MMTV promoter, were induced to differentiate by the addition of NGF or by infection with Ki-MSV. The activation of early-response genes as well as differentiation-specific markers was determined by Northern blotting after both infection and NGF treatment. The activation of Erk and Jun kinase as well as the extent of neurite outgrowth was measured in response to NGF exposure and Ki-MSV infection. The results demonstrate that the effect of increased Rsu-1 expression is not an inhibition of Ras-induced differentiation but rather an enhancement of the process. Moreover, the result of NGF-induced stimulation of the Rsu-1 transfectants is the induction of a rapid differentiation process.

Results The Ras suppressor, Rsu-1, was introduced into PC12 cells to test its effect on Ras-dependent processes of growth factor-induced mitogenesis and differentiation. Our previous work described the construction of a vector carrying the Rsu-1 cDNA under the control of a dexamethasone-induc- Fig. 1. Thymidine incorporation after stimulation of Rsu-1 transfectant ible promotor and its introduction into PC12 cells (6). Two cell line with mitogens. PC12-a, a vector control transfectant cell line, and clones expressing Rsu-1 in response to dexamethasone, PC12-20, an Rsu-1 transfectant cell line, were treated with dexamethasone for 24 h in reduced serum medium. EGF (100 ng/ml) or insulin (400 PC12-20 and PC12-26, were used in the studies described ng/ml) was added to the cells, and [3H]thymidine pulse labeling was here, along with a pMAM vector control clone, PC12-a (6). performed at 0, 24, 48, and 72 h after the addition of mitogen. Triplicate samples were used to calculate the mean for each time point; bars, SD. The cells were stimulated with NGF or infected with Ki-MSV SDs are shown and were Ͻ10% of the mean at all time points. to initiate differentiation; the response of the cells to mitogenic compounds was tested by stimulation with insulin and EGF. Growth of PC12 Rsu-1 Transfectants in Response to regulating signal transduction leading to neuronal differenti- Mitogens. To investigate the effects of mitogenic factors ation, two Rsu-1 transfectant PC12 cell clones and a vector signaling through Ras on the growth of PC12 cells overex- control PC12 cell line were exposed to NGF. The phenotypic pressing Rsu-1, the effect of two growth factors, epidermal appearance of the cells in the form of neurite outgrowth and growth factor and insulin, was tested. The PC12 vector con- the expression of differentiation-specific RNA were investi- trol and Rsu-1 transfectant cells were plated in serum con- gated over a period of 4 days. taining medium. Rsu-1 expression was induced for 24 h with The extent of neurite outgrowth after NGF addition to the 0.5 ␮M dexamethasone before the addition of EGF or insulin. cells was analyzed in cultures of PC12-a vector control cell The cells were harvested at different time points after [3H]thy- line and in the PC12-20 Rsu-1 transfectant clones. PC12 cell midine pulse labeling, as indicated in Fig. 1. [3H]Thymidine clones growing on dishes were stimulated with two different incorporation studies demonstrated that the vector control concentrations of NGF. The degree of neurite extension was cell line PC12-a proliferated rapidly in response to EGF or determined by microscopically counting cells at 24 and 72 h insulin, whereas proliferation of the Rsu-1 transfectant after NGF addition and quantitating the number of cells ex- PC12-20 was nearly completely blocked. These data sug- tending short or long neurites. As shown in Table 1, both cell gest that overexpression of Rsu-1 can inhibit proliferation of lines differentiated when exposed to NGF. The Rsu-1-over- these cells in response to mitogens. expressing cell lines, PC12-20, differentiated earlier and Analysis of NGF-mediated Differentiation in PC12- faster than the vector control cell line PC12-a. The overall Rsu-1 Transfectant Cell Lines. Stimulation of PC12 cells level of differentiation was approximately 3-fold greater for with NGF can induce neuronal differentiation, characterized the Rsu-1 transfectants than the vector control cells. More by neurite outgrowth and expression of specific differentia- striking were the appearance of long neurites in the Rsu-1 tion markers (25–27). To analyze a possible role for Rsu-1 in transfectants (Table 1). The photograph in Fig. 2 illustrates Cell Growth & Differentiation 557

Table 1 Differentiation of Rsu-1 transfectant in response to NGF Infection of PC12 Rsu-1 Transfectants with Ki-Ras. The introduction of viral Ras into PC12 cells by infection with Cells were plated on laminin-coated dishes in the presence of 5% horse serum induced with 0.5 ␮M dexamethasone and stimulated with NGF. Ki-MSV, microinjection of v-Ras p21, or transfection of a Twenty-four or 72 h after the addition of growth factor, the number of vector carrying v-Ras under the control of an inducible pro- differentiated cells was determined microscopically by counting Ͼ1000 cells. moter resulted in initiation of differentiation in PC12 cells (7, 8, 10). Because Rsu-1 expression was able to inhibit Ras- % cells extending neurites induced transformation of fibroblasts, the ability of Rsu-1 to 24 h, NGF 72 h, NGF alter v-Ras-induced differentiation of PC12 cells was tested. Clone (50 ng/ml)a (1 ng/ml)c Ki-MSV was used to infect PC12 vector control cells and Long Short Long Short Rsu-1 transfectants. The cells were examined for phenotypic neurites neuritesb neurites neuritesb differentiation, and RNA was isolated to check for the induc- PC12, clone a 1.7 6.7 0.8 2.7 tion of Ras-induced transcriptional response. As a control for PC12, clone 20 8.2 26.5 3.5 10.2 changes occuring in response to v-Ras, a PC12-derived cell a Cells were plated on laminin-coated dishes in low-serum medium in- line containing activated v-Ras under the control of a dexa- duced with 0.5 ␮M dexamethasone for 48 h, and then NGF was added to the medium at 50 ng/ml. Twenty-four h after the addition of growth factor, methasone-inducible MMTV promoter (GSRas-1) was used the number of differentiated cells was determined microscopically by (10). In addition, MuLV infection of the PC12 clones served Ͼ counting 1000 cells. as a negative control. b Neurite projections were classified as “short” if they were 0.5–1 cell body length and “long” if they were Ͼ1 cell body length. The effect of v-Ras infection on phenotypic differentiation c Cells were plated on laminin-coated dishes in low-serum medium in- of PC12 vector control and Rsu-1 transfectants was exam- duced with 0.5 ␮M dexamethasone for 24 h, and then NGF was added to the medium at 1 ng/ml. Twenty-four h after the addition of growth factor, ined. The percentage of differentiated cells in the cultures fresh medium was added to the cells. Seventy-two h after the initial was determined microscopically at 48 and 96 h after infec- addition of growth factor, the number of differentiated cells was deter- tion. As seen in Table 2, the percentage of differentiated cells mined microscopically by counting Ͼ1000 cells. was similar in control and Rsu-1 transfectants. Again, these results indicate that the expression of Rsu-1 does not inhibit v-Ras-induced differentiation. the appearance of the vector control and the Rsu-1 trans- Alteration of expression of early-response genes in the fectant clones maintained on plastic dishes in complete me- GSRas-1 cells was checked 24 h after dexamethasone in- dium in the presence and absence of NGF. duction of v-Ras expression. The results shown in Fig. 5a The expression of RNA for early response genes (Fos, indicate that Fos but not Jun expression increased in c-Jun, and NGF1A/EGR1) and for differentiation-specific GSRas-1 cells. RNAs isolated from Ki-MSV-infected control genes (VGF8 and SCG10/strathmin) in response to NGF was and Rsu-1 transfectant PC12 clones at 72 h after infection analyzed. Initially, RNA was extracted from dexamethasone- were analyzed for expression of early-response genes as induced PC12-a control and PC12-20 Rsu-1 transfectant well as for expression of v-ras- and Rsu-1-specific messages by Northern blotting (Fig. 5B). An increase in Fos expression cells after a time course of NGF treatment (0–60 min); the was present in the Rsu-1 transfectant cell line PC12-20 after level of specific RNA was determined by Northern blotting v-Ras infection compared with the vector control cell line and hybridization to specific probes. As shown in Fig. 3A, PC12-a. The expression of cJun after Ki-MSV infection was there was an increase in the expression of c-fos, c-Jun, and less in the Rsu-1 transfectants relative to the control cell line. NGF1A by 30 min after NGF addition in both the Rsu-1 and Equivalent expression of Ki-MSV-specific RNA was present vector control cell lines. Quantitation revealed that the mag- in both cell lines (Ras, Fig. 5). These data indicate that the nitude of the Fos response was 2-fold greater in the Rsu-1- introduction of v-Ras into the Rsu-1 transfectants resulted in overexpressing cell line, PC12-20, compared with the vector v-Ras expression and induction of specific Ras-induced control cell line, PC12-a (Fig. 3B). Fos induction by NGF has genes. been reproducibly elevated 2–2.5-fold over the control cell Erk and RafB Kinase Activation in Rsu-1 Transfectants. line at 45 and 60 min. Expression of late-response genes The activation of Erk kinase and Jun kinase pathways after specific for neuronal differentiation, VGF8 and SCG10, were exposure to NGF was analyzed in PC12 Rsu-1 transfected also analyzed over a longer time course (0–72 h) in both and vector control cell lines. Two Rsu-1 transfectant cell Rsu-1 transfectants and the vector control clone. There were clones, PC12-20 and PC12-26, were used for these exper- no significant differences in SCG10 expression between the iments. The cells were plated, induced for 24 h with 0.5 ␮M Rsu-1 transfectant cell lines and the control. However, VGF8 dexamethasone, and treated with NGF (100 ng/ml) for 0, 1, 4, expression was sustained longer in response to NGF in the 6, or 24 h before lysis. Cell lysates were used to analyze Erk Rsu-1 transfectants (Fig. 4). Again, the greater intensity of kinase by immunoprecipitation and phosphorylation of my- Fos induction was evident in the Rsu-1 transfectants com- elin basic protein substrate. As shown on Fig. 6A, the Erk pared to the control clone. The somewhat elevated neurite kinase pathway was activated in response to NGF in both the extension and induction of VGF8 RNA, which is specific to PC12-20 and PC12-26 cell lines as well as the vector control the induction of the differentiation program by NGF, in the cell line PC12-a. The results of three experiments were used Rsu-1 transfectants demonstrates that this Ras-dependent to quantitate the level of Erk activity and revealed a sustained process is not inhibited by Rsu-1 expression. In fact, it is increase in Erk activity in the Rsu-1 transfectants exposed to somewhat enhanced compared with the control cell line. NGF compared with identically treated control cells. These 558 Rsu-1 Enhances PC12 Differentiation

Fig. 2. Differentiation of Rsu-1 transfectant cell lines in response to NGF. PC12 Rsu-1 transfectant and vector control transfectant cells were seeded in plastic tissue cultures dishes in complete medium, induced with 0.5 ␮M dexamethasone for 48 h, followed by maintenance in the presence or absence of NGF (50 ng/ml) for 24 h. Cells were photographed by confocal microscopy using a ϫ40 objective (A) and by phase contrast microscopy at ϫ400 (B). Panels a and b, clone a; panels c and d, clone 20; panels e and f, clone 26. Panel a, c, and e, cells exposed to only dexamethasone; panels b, d, and f, cells exposed to dexamethasone and NGF. Cell Growth & Differentiation 559

Fig. 2. Continued.

differences in Erk activation are similar to those obtained NGF treatment in the control cells in agreement with the after EGF stimulation of the Rsu-1 transfectants maintained published results of others (data not shown; Ref. 20). These in the same way (6). Analysis of Jun kinase activation re- data support the conclusions of others that although the Erk vealed little or no increase of c-Jun phosphorylation after mitogen-activated protein kinase pathway is important for 560 Rsu-1 Enhances PC12 Differentiation

ated kinase activity for Rb, indicating that p21CIP induction

may play an important role in G1 growth arrest necessary for the initiation of the differentiation program (21, 22). To assess the contribution of Rsu-1 to p21CIP induction and because Rsu-1 expression inhibits growth in these cells (Fig. 1), the level of p21CIP was determined at time points after the induction of Rsu-1 expression. The results of Western blotting of Rsu-1 transfectants and control lysates are shown in Fig. 7A. Increased p21CIP expression is evident 48 h after the induction of Rsu-1 expression in the Rsu-1 transfectant cell lines, but identical treatment of the vector control cell line did not induce p21CIP. As a control, the level of another cell cycle inhibitor, p27KIP, was checked on the same cell lysates. The results in Fig. 7A show that p27KIP, was induced equally in the control and Rsu-1 transfectants. Hence, the expression of Rsu-1 in these cells results in specific p21CIP induction. We also examined NGF-stimulated transfectant and control cells for p21CIP expression. We noted that after dexamethasone and NGF stimulation of PC12 Rsu-1 transfectants, p21CIP expression was maximal by 48 h; however, a large increase in p21CIP was not detected in the vector control cell line until 72 h of NGF treatment. Again, the level of p27KIP did not increase in response to NGF in either the Rsu-1 transfectants or the control cells. Although the earlier increase in p21CIP expression in the Rsu-1 transfectants stimulated with NGF supports a role for p21CIP in NGF-induced differentiation, our results indicate that increases in p21CIP alone are clearly not sufficient to initiate differentiation.

Fig. 3. Expression of early-response genes and differentiation-specific Discussion genes after NGF stimulation of Rsu-1 transfectant cell lines. A, Northern Activation of a program of differentiation in the PC12 cell line blot of RNA from NGF-stimulated PC12-a and PC12-20 transfectant cell lines. Northern blots were prepared from 10 ␮g of total RNA extracted by NGF requires sustained activation of the Erk kinase path- from cells exposed to NGF (200 ng/ml) for 0, 30, 45, or 60 min. The blots way (17). This is achieved by a Ras-dependent mechanism were hybridized sequentially to the indicated specific probes. B, the hybridization signals from the blot in Fig. 2A were quantitated and nor- (28). Induction of differentiation by v-Ras also results in an malized to the GPDH hybridization signal. The results are shown graphi- increase in Erk kinase activation via the Raf-Mek-Erk path- cally. way (10, 28). In this study we demonstrate that PC12 transfectant cell lines expressing the suppressor of Ras transformation, Rsu-1, are capable of differentiating in response to NGF-induced differentiation, the Jun kinase pathway is not NGF. The induction of the differentiation process is accom- significantly activated by NGF and is not required for the panied in both control and Rsu-1 transfectant cell lines by induction of the differentiation process. The ability of Rsu-1 activation of the Erk kinases and early-response gene tran- to enhance differentiation in response to NGF is most likely scription. The activation of Jun kinase, a step blocked by due to the sustained activation of Erk kinase in these cells. Rsu-1 in PC12 cells (6), occurs at a low level in the control The likely mechanism for the activation of Erk in response cells as a result of NGF stimulation. However, the inhibition of to NGF is via the activation of RafB and Mek. Therefore, the Jun kinase activation may prevent proliferation of Rsu-1 level of RafB activation was determined using an immune transfectants in response to insulin and EGF. The Rsu-1 kinase assay with purified Mek1 as a substrate. The results in transfectants differentiate more rapidly than the control cells Fig. 6B demonstrate that the level of RafB activation was in response to NGF, indicating that, as shown by others, this enhanced in the Rsu-1 transfectants relative to the vector pathway does not require a high level of Jun kinase activa- control cells. Analysis of Raf-1 activation using the same cell tion. However, the Rsu-1 transfectants exhibit increased lysates revealed a much lower level of Mek phosphorylation RafB and Erk activation as well as an increase in the level of in all clones (data not shown). These results support the Fos transcriptional activation compared with the control conclusion that there is an enhanced RafB-Mek-Erk stimu- cells. lation by NGF in the Rsu-1 transfectants. After infection of the PC12 transfectant cell lines with Induction of p21CIP Expression after Induction of Rsu-1 Ki-MSV and expression of v-Ki-Ras, little activation of Jun Expression in PC12 Cells. The addition of NGF to PC12 kinase was detected in control cell lines or the Rsu-1 trans- cells maintained in serum containing medium results in ex- fectants. The Rsu-1 transfectants exhibited the same or pression of p21CIP. Several studies have demonstrated this higher levels of Erk kinase activation and phenotypic differ- increase, as well as a decrease in cdk-4- and cdk-6-associ- entiation as the control transfectants. These data indicate Cell Growth & Differentiation 561

Fig. 4. Expression of early-response genes and differentiation- specific markers after 72 h of NGF stimulation of Rsu-1 transfectant cell lines. Northern blots were prepared from total RNA isolated from PC12-a, PC12-20, and PC12-26 transfectant cell lines exposed to NGF (200 ng/ml) for 0, 1, 6, 24, 48, or 72 h. The blots were hybridized to probes for the differentiation-specific markers SCG10, VGF8, as well as the early-response genes Fos and NGF1A. Rsu-1 and GPDH (GAPDH) hybridization are shown as controls. Arrows, position of specific induced bands.

Table 2 Differentiation of Rsu-1 transfectant in response to Ki-MSV function has been associated with activation of the small GTPase, Rho (30, 31). Rho regulates actin cytoskeletal % cells extending b changes (32, 33), and Rho can function as a negative regu- Cell linea Virus neurites lator of PC12 differentiation (34–38). Hence, the inhibition of 48 h 96 h RasGAP activity in response to Rsu-1 expression might be PC12, clone a MuLV 0.6 0.8 expected to inhibit Rho activity and to sustain neurite out- PC12, clone a Ki-MSV 11.8 19.1 growth. PC12, clone 20 MuLV 2.6 0.75 PC12, clone 20 Ki-MSV 14.4 12.2 Another potential role for Rho in PC12 differentiation is in WAF/CIP a regulating levels of p21 , the cdk inhibitor (39). In PC12 PC12 vector control (clone a) or Rsu-1 transfectant cell line (clone 20) CIP was infected and treated with dexamethasone (500 nM) for 48 h, then the cells, the activation of p21 apparently contributes to the dexamethasone was removed from the culture. growth arrest required or associated with the induction of the b Neurite extension was determined microscopically by counting at least 1000 cells and determining the number with and without neurite exten- differentiation program by NGF (21, 22). In PC12 cells that sion. Projections that were Ͼ0.75 cell body length were classified as are maintained in very low serum during the NGF-induced neurites. differentiation process, there is no induction of p21CIP, pre-

sumably because G1 growth arrest is accomplished by re- moval of mitogenic factors (21, 22). However, high level that the pathway(s) downstream of Ras that are inhibited by activation of the Raf-Mek-Erk kinase pathway induced ex- Rsu-1 in the process leading to transformation in fibroblasts, pression of p21CIP (23, 24). Olson et al. (39) have demon- i.e., the activation of Jun kinase, may be only selectively strated that this growth-inhibitory event, i.e., high levels of required for the induction of a stable differentiation-specific p21CIP, can be down-regulated by activated Rho. In these response to v-Ras in PC12 cells. instances, the control of the level of p21CIP by Rho allows

Activation of Ras results in the activation of Raf kinase as cells to avoid G1 arrest. Analysis of the Rsu-1 PC12 cell well as other Ras effectors. Previous studies suggest two transfectants found a higher level of p21CIP in these cells mechanisms by which Rsu-1 expression may result in an than the control cells, even without the addition of NGF. This increase in Erk activation. Rsu-1 binds Raf1 in in vitro binding result suggests that Rsu-1 inhibits the activation of a specific assays. Although a similar in vivo association has not been Ras and Raf-dependent pathway by inhibiting a Rho-depen- detected, there may be stabilization of a complex involving dent event. Raf and Rsu-1 that promotes signaling by RafB. Also, Rsu-1 In conclusion, expression of Rsu-1 can enhance Ras-reg- expression inhibits the activity of RasGAP (6). RasGAP, the ulated differentiation of PC12 cells. Because another leucine Ras GTPase-activating protein, has Ras effector function in repeat protein has been identified as important in a Ras- addition to its regulatory function. Expression of RasGAP dependent differentiation process (3), this may be a common inhibits NGF-induced differentiation of PC12 cells, presum- function for a small group of leucine repeat proteins. The dual ably due to inhibition of Erk activation (19, 29). In Rsu-1 PC12 Rsu-1 activities of enhancing Ras-dependent differentiation, transfectants, the increase in RafB and Erk activation in but inhibiting Ras-dependent transformation, may result response to NGF may be a direct result of decreased Ras- from the requirement for activation of specific Ras effectors GAP activity and an increase in GTP-Ras. RasGAP effector for each function. Elucidation of the mechanism by which 562 Rsu-1 Enhances PC12 Differentiation

Fig. 6. Induction of Erk 1 and RafB after NGF stimulation. A, dexamethasone-induced PC12 clone a, clone 20, and clone 26 were stimulated with NGF (100 ng/ml). At 0, 1, 4, 6, and 24 h after NGF addition, the cells were lysed, and the amount of ERK-1 and ERK-2 activity in the lysates was determined using an immune kinase reaction. The results are the means of three experiments and are shown as fold-increase in phosphorylation of myelin basic protein substrate. Bars, SD. B, dexamethasone-induced PC12 clone a, clone 20, and clone 26 were stimulated with Fig. 5. Induction of Fos and cJun RNA expression by Ras infection of NGF (100 ng/ml). At 0, 1, and 4 h after NGF addition, the cells were lysed, Rsu-1 transfectant cell lines. A, GS-Ras1 cells, which contain v-Ras under and the amount of RafB activity in the lysates was determined using an the control of an MMTV promoter, and PC12 clone a vector control cells immune kinase reaction. The autoradiogram of labeled substrate, Mek-1, were treated with dexamethasone for 24 h. RNA was extracted from the is shown. cells, and 10 ␮g of total RNA were analyzed by Northern blotting for levels of Fos and c-Jun RNA. The amounts of specific hybridization were quantitated and are represented graphically. B, PC12-clone a and PC12-clone 20 cells were treated with dexamethasone and infected with Ki-MSV. ␮ Forty-eight h after infection, the dexamethasone was removed from the g/ml, washed in TBS/0.1% Tween 20 (T-TBS), and incubated with the culture, and 24 h later (72 h after infection), the cells were lysed for specific horseradish peroxidase-conjugated secondary antibody (Amer- isolation of RNA. RNA was assayed for Rsu-1, Ras, Fos, and c-Jun sham Life Science, Inc., Arlington Heights, IL). The filters were developed transcripts as in Fig. 4A. by chemiluminescence using the ECL chemiluminescence system (Am- ersham Life Science, Inc.) and Kodak X-OMAT films. Northern Blotting. Total RNA was isolated from cell cultures induced with 0.5 ␮M dexamethasone for 48 h and treated with 100 ng/ml of NGF Rsu-1 inhibits RasGAP function should provide additional 2.5S (Boehringer Mannheim) for different periods of time, by lysis with understanding of the control of these pathways. guanidine thiocyanate and purification on cesium chloride gradient or using Tripure reagent (Boehringer Mannheim) according to the manufac- turer’s protocol. Ten ␮g of total RNA were separated on a 1% agarose/ Materials and Methods 4-morpholinepropanesulfonic acid/formaldehyde gel and transferred to Cell Culture. PC12 transfectants were grown in DMEM with 10% char- nylon membrane. The filters were cross-linked and hybridized with the coal-stripped horse serum and 4% fetal bovine serum “complete me- following probes: a 237-bp fragment corresponding to amino acids 101– dium” and induced with dexamethasone as described previously (6). 179 from the cDNA sequence of SCG10 (27); a 299-bp fragment corre- Low-serum medium contained 3% charcoal-stripped horse serum and sponding to bases 481–780 of cDNA sequence of VGF8 (25); and a 0.5% fetal bovine serum. 307-bp fragment corresponding to bases 1051–1358 of the cDNA se- Antibodies. Monoclonal antibodies against ERK-2 were purchased quence of NGF1A (26). These probes were obtained by reverse transcrip- from Upstate Biotechnology, Inc. (Lake Placid, NY) or Santa Cruz Bio- tion and PCR amplification. Other probes included the entire open reading technology, Inc. (Santa Cruz, CA). Polyclonal antibodies against ERK-1 frame of human c-jun, a 2-kb fragment of human c-fos, and the probe for and ERK-2 and RafB were obtained from Santa Cruz Biotechnology, Inc. GDPH were obtained from American Tissue Culture Collection. Hybrid- Monoclonal Jun kinase antibody was from PharMingen, and monoclonal izations were performed as described previously (6). The filters were antibody recognizing phosphorylated Jun kinase was obtained from quantitated using a Packard Beta Scan system. The filters were stripped Santa Cruz Biotechnology. Antibodies to p21WAF/CIP and cyclin D1 were by boiling for 5 min in washing buffer and were reused for hybridization. obtained from Santa Cruz Biotechnology. Kinase Assays. PC12 transfectants were maintained in regular me- Western Blotting. Seventy-five ␮g of lysate were separated by SDS- dium before stimulation with 100 ng/ml NGF 2.5S (Boehringer Mannheim). PAGE, transferred to nitrocellulose (Schleicher & Schuell, Keene, NH), and For Erk-2 phosphorylation gel shift assays, 50 ␮g of NP40/DOC lysates blocked overnight in 5% nonfat milk in Tris-buffered saline (TBS). The from stimulated or unstimulated cells were resolved by SDS-PAGE on a filters were reacted with primary antibodies at the concentration of 1 10% Tris-glycine gel, transferred on nitrocellulose, reacted by Western Cell Growth & Differentiation 563

Acknowledgments We are grateful to Dr. Simon Halegoua for the GS-Ras1 cell line.

References 1. Cutler, M., Bassin, R., Zanoni, L., and Talbot, N. Isolation of rsp-1, a novel cDNA capable of suppressing v-ras. Mol. Cell. Biol., 12: 3750–3756, 1992. 2. Tsuda, T., and Cutler, M. L. Human RSU-1 is highly homologous to mouse Rsu-1 and localizes to human chromosome 10. Genomics, 18: 461–462, 1993. 3. Sieburth, D., Sun, Q., and Han, M. Sur-8, a conserved Ras-binding protein with leucine repeats, positively regulates Ras-mediated signaling in C. elegans. Cell, 94: 119–130, 1998. 4. Colicelli, J., Field, J., Ballester, R., Chester, N., Young, D., and Wigler, M. Mutational mapping of ras-responsive domains of the Saccharomyces cerevisiae adenylyl cyclase. Mol. Cell. Biol., 10: 2539–2543, 1990. 5. Field, J., Xu, H-P., Michaelli, T., Ballester, R., Sass, P., Wigler, M., and Fig. 7. Induction of p21CIP and p27KIP in Rsu-1 PC12 transfectants. A,at Colicelli, J. Mutations of the adenylyl cyclase gene that block ras function various times after dexamethasone addition, the vector control clone a in Saccharomyces cerevisiae. Science (Washington DC), 247: 464–467, and the Rsu-1 transfectant clones 20 and 26 were lysed, and the levels of 1990. CIP KIP ␮ p21 and p27 were determined by Western blotting of 50 g of cell 6. Masuelli, L., and Cutler, M. Increased expression of the Ras suppres- lysate. The results for vector control and two Rsu-1 transfectants are sor, Rsu-1, enhances Erk-2 activation and inhibits Jun kinase activation. shown. B, at various times after dexamethasone and NGF (100 ng/ml) addition, the vector control clone a and the Rsu-1 transfectant clones 20 Mol. Cell. Biol., 16: 5466–5476, 1996. and 26 were lysed, and the levels of p21CIP and p27KIP were determined 7. Bar-Sagi, D., and Feramisco, J. Microinjection of the Ras oncogene by Western blotting of 50 ␮g of cell lysate. protein into PC12 cells induces morphological differentiation. Cell, 42: 841–848, 1985. 8. Noda, M., Ko, M., Ogura, A., Liu, D., Amano, T., Takano, T., and Ikawa, Y. Sarcoma viruses carrying Ras oncogenes induce differentiation-asso- blotting analysis with a specific anti-ERK-2 monoclonal antibody, and ciated properties in a neuronal cell line. Nature (Lond.), 318: 73–75, 1985. detected by chemiluminescence. 9. Halegoua, S., Armstrong, R. C., and Kremer, N. E. Dissecting the mode For myelin basic protein phosphorylation assay, 100 ␮g of lysates of action of a neuronal growth factor. Curr. Top. Microbiol. Immunol., 165: (buffer: 0.5% Triton X-100 10% glycerol, 100 mM NaCl, and 20 mM HEPES) 120–170, 1991. from stimulated or unstimulated cells were immunoprecipitated using 10. D’arcangelo, G., and Halegoua, S. A branched signaling pathway for polyclonal antibody that recognize both ERK-1 and ERK-2. The immuno- nerve growth factor is revealed by Src-, Ras-, and Raf-mediated gene precipitates were washed twice with lysis buffer and twice with a kinase inductions. Mol. Cell. Biol., 13: 3146–3152, 1993. buffer containing 30 mM HEPES (pH 7.4), 10 mM MgCl2,and1mM DTT. The immunoprecipitates were resuspended in 30 ␮l of kinase buffer 11. Greene, L., and Tischer, A. Establishment of a non-adrenergic clonal containing 5 ␮g of myelin basic protein and 10 ␮Ci of ␥-ATP (32P; DuPont- line of rat pheochromocytoma cells which respond to nerve growth factor. NEN, Boston, MA) and incubated 30 min at 30°C. The reactions were Proc. Natl. Acad. Sci. USA, 73: 2424–2428, 1976. stopped by the addition of 2ϫ sample buffer and resolved by SDS-PAGE. 12. Klein, R., Jing, S., Nandouri, V., O’Rourke, E., and Barbacid, B. The trk After transfer on polyvinylidene difluoride (Millipore Corp., Bedford, MA) proto-oncogene encodes a receptor for nerve growth factor. Cell, 65: membrane, the filters were subjected to autoradiography and quantitated 189–197, 1991. using a Beta Scan system. For Mek-1 phosphorylation assay, 400 ␮gof 13. Kaplan, D., Martin-Zanca, D., and Parada, L. Tyrosine phosphoryla- lysates were precipitated with anti-RafB as described above. The reaction and tyrosine kinase activity of the trk proto-oncogene product by tions were carried out and quantitated as described above using kinase NGF. Nature (Lond.), 350: 158–160, 1991. buffer [25 mM Tris (pH 7.5), 2 mM MnCl ,1mM EDTA, 10 mM MgCl , and 2 2 14. Hempstead, B., Martin-Zanca, D., Kaplan, D., Parada, L., and Chao, 1mM DTT] and 2 ␮g of Mek-1 (Santa Cruz Biotechnology, Inc.). 6 M. High affinity NGF binding requires co-expression of the trk proto- Viral Infection. Cells (10 ) were plated on poly-D-lysine-coated, oncogene and the low affinity NGF receptor. Nature (Lond.), 350: 678– 100-mm dishes in complete medium containing 0.5 ␮M dexamethasone 683, 1991. for 48 h. Prior to the infection, the cells were incubated with 2.5 ␮g/ml of Polybrene in complete medium for4hintheincubator. The infection was 15. Thomas, S. M., DeMarco, M., D’Arcangelo, G., Halegoua, S., and carried out by incubation of cells with the Ki-MSV (MuLV) at a multiplicity Brugge, J. S. Ras is essential for nerve growth factor- and phorbol of infection of 1–5 in 2 ml of medium containing 2.5 ␮g/ml of Polybrene for ester-induced tyrosine phosphorylation of MAP kinases. Cell, 68: 1031– 4 h. The inoculum was then diluted to 10 ml and left on the cells overnight. 1040, 1992. After replacing the virus with dexamethasone containing medium, the 16. Wood, K. W., Sarnecki, C., Roberts, T. M., and Blenis, J. Ras medi- cells were incubated for up to 5 days, and differentiation was analyzed by ates nerve growth factor receptor modulation of three signal-transducting counting the number of neurite-bearing cells microscopically. For the RNA protein kinases: MAP kinase, Raf-1, and RSK. Cell, 68: 1041–1050, 1992. isolation experiments, the cells were incubated for 2 days after infection 17. Cowly, S., Paterson, H., Kemp, P., and Marshall, C. Activation of Map in dexamethasone containing medium and then for 2 additional days in kinase is necessary and sufficient for PC12 differentiation and for trans- regular medium lacking dexamethasone. RNA and proteins were ex- formation of NIH3T3 cells. Cell, 77: 841–852, 1994. tracted from these cultures and analyzed by Northern blotting and West- 18. Marshall, C. Specificity of receptor tyrosine kinase signaling: transient ern blotting as described. versus sustained extracellular signal-regulated kinase activation. Cell, 80: [3H]Thymidine Incorporation. Cells (4 ϫ 103) were plated on 96-well 179–185, 1995. plates in 5% horse serum containing medium; after 24 h, the cells were induced by addition of 0.5 ␮M dexamethasone, and triplicate wells were 19. Yao, R., and Cooper, G. Regulation of the Ras signaling pathway by treated with the appropriate growth factor for 24, 48, and 72 h prior to GTPase activating protein in PC12 cells. Oncogene, 11: 1607–1614, 1995. labeling with 0.5 ␮Ci/well [3H]thymidine (Amersham) for 6 h. The cells were 20. Xia, Z., Dickens, M., Raingeaud, J., Davis, R., and Greenberg, M. then harvested using a cell harvester, transferred onto filters, and quan- Opposing effects of ERK and JNK-p38 MAP kinases on apoptosis. Sci- tified by scintillation counting. ence (Washington DC), 270: 1326–1331, 1995. 564 Rsu-1 Enhances PC12 Differentiation

21. Yan, G-Z., and Ziff, E. NGF regulates the PC12 cell cycle machinery 31. McGlade, J., Brunkhorst, B., Anderson, D., Mbamalu, G., Settleman, through specific inhibition of the cdk kinase and induction of cyclin D1. J., Dedhar, S., Rozakis-Adcock, M., Chen, L. B., and Pawson, T. The J. Neurosci., 15: 6200–6206, 1995. N-terminal region of GAP regulates cytoskeletal structure and cell adhe- 22. van Grunsven, L., Billon, N., Savatier, P., Thomas, A., Urdiales, J., and sion. EMBO J., 12: 3073–3081, 1993. Rudkin, B. Effect of nerve growth factor on the expression of cell cycle 32. Ridley, A., and Hall, A. The small GTP binding protein Rho regulates regulatory proteins in PC12 cells: dissection of the neurotrophic response the assembly of focal adhesions and actin stress fibers in response to from the anti-mitogenic response. Oncogene, 12: 1347–1356, 1996. growth factors. Cell, 70: 389–399, 1992. 23. Woods, D., Parry, D., Cherwinski, H., Bosch, E., Lees, E., and McMahon, M. Raf-induced proliferation or cell cycle arrest is determined 33. LeBlanc, V., Tocque, B., and Delumeau, I. Ras-GAP controls Rho- by the level of Raf activity with arrest mediated by p21WAF/CIP. Mol. Cell. mediated cytoskeletal reorganization through its SH3 domain. Mol. Cell. Biol., 17: 5598–5611, 1997. Biol., 18: 5567–5578, 1998. 24. Sewing, A., Wiseman, B., Lloyd, A., and Land, H. High intensity Raf 34. Katoh, H., Aoki, J., Ichikawa, A., and Negishi, M. p160 RhoA-binding WAF/CIP signal causes cell cycle arrest mediated by p21 . Mol. Cell. Biol., 17: kinase ROKa induces neurite retraction. J. Biol. Chem., 273: 2489–2492, 5588–5597, 1997. 1998. 25. Levi, A., Eldridge, J., and Paterson, B. Molecular cloning of a gene 35. Gebbink, M., Kranenburg, O., Poland, M., vanHorck, F., Houssa, B., sequence regulated by nerve growth factor. Science (Washington DC), and Moolenaar, W. Identification of a novel, putative Rho-specific GDP/ 238: 393–397, 1985. GTP exchange factor and a RhoA binding protein. J. Cell Biol., 137: 26. Milbrandt, J. A nerve growth factor induced gene encodes a possible 1603–1613, 1997. transcriptional regulatory factor. Science (Washington DC), 238: 797–800, 1987. 36. Kozma, R., Sarner, S., Ahmed, S., and Lim, L. Rho family GTPases 27. Stein, R., Orit, S., and Anderson, D. The induction of a neural specific and neuronal growth cone remodeling: relationship between complexity gene, SCG10, by nerve growth factor in PC12 cells is transcriptional, induced by Cdc42Hs, Rac 1, and acetylcholine and collapse induced by protein synthesis dependent, and glucocorticoid inhibitable. Dev. Biol., RhoA and lysophosphatidic acid. Mol. Cell. Biol., 17: 1201–1211, 1997. 127: 316–325, 1988. 37. Tigyi, G., Fisher, D., Sebok, A., Yang, C., Dyer, D., and Miledi, R. 28. Thomas, S. M., DeMarco, M., D’Arcangelo, G., Halegoua, S., and Lysophosphatidic acid-induced neurite retraction in PC12 cells: control of Brugge, J. S. Ras is essential for nerve growth factor- and phorbol phosphoinositide signaling and rho. J. Neurochem., 66: 537–548, 1996. ester-induced tyrosine phosphorylation of MAP kinases. Cell, 68: 1031– 1040, 1992. 38. Tigyi, G., Fisher, D., Sebok, A., Marshall, F., Dyer, D., and Miledi, R. Lysophosphatidic acid-induced neurite retraction in PC12 cells: neurite 29. Nakata, H., and Watanabe, Y. Proliferation and differentiation of PC12 protective effects of cyclic AMP signaling. J. Neurochem., 66: 549–558, cells were affected by p21 ras GTPase activating proteins and its deletion mutant proteins. Biochem. Biophys. Res. Commun., 218: 538–543, 1996. 1996. 30. Clark, G., Westwick, J., and Der, C. p120 GAP modulates Ras acti- 39. Olson, M., Peterson, H., and Marshall, C. Signals from Ras and Rho vation of the Jun kinases and transformation. J. Biol. Chem., 272: 1677– GTPases interact to regulate expression of p21WAF/CIP. Nature (Lond.), 1681, 1996. 394: 295–299, 1998.