c-Jun Amino-terminal Kinase Is Regulated by Gα12/Gα13 and Obligate for Differentiation of P19 Embryonal Carcinoma Cells by Retinoic Acid*

Retinoic acid induces P19 mouse embryonal carcinoma cells to differentiate to endoderm and increases expression of the heterotrimeric G-protein subunits Gα12 and Gα13. Retinoic acid was found to induce differentiation and sustained activation of c-Jun amino-terminal kinase, but not of ERK1,2 or of p38 mitogen-activated protein kinases. Much like retinoic acid, expression of constitutively active forms of Gα12and Gα13 induced differentiation and constitutive activation of c-Jun amino-terminal kinase. Expression of the dominant negative form of c-Jun amino-terminal kinase 1 blocked both the activation of c-Jun amino-terminal kinase and the induction of endodermal differentiation in the presence of retinoic acid. These data implicate c-Jun amino-terminal kinase as a downstream element of activation of Gα12 or Gα13 obligate for retinoic acid-induced differentiation.

The role of heterotrimeric guanine nucleotide binding proteins (G-proteins) 1 in cell differentiation and development has been shown in several systems (1)(2)(3)(4). Adipogenesis of NIH 3T3-L1 cells in response to inducers such as insulin or dexamethasone plus methylisobutylxanthine is accompanied by a sharp decline in G␣ s subunit (5). Both suppression of G␣ s by antisense oligodeoxynucleotides and overexpression of the constitutively active mutant form of G␣ i2 promote adipogenesis in the absence of classical inducers (5). Differentiation of F9 embryonic stem cells to endodermal cells provokes a sharp reduction of G␣ i2 (2), a de-repression of phospholipase C, and activation of protein kinase C and mitogen-activated protein (MAP) kinase, especially via ERK1,2 (6). The expression of constitutively active forms of G␣ q or G␣ 16 can induce neuronal differentiation of PC 12 cells (7), via activation of c-Jun aminoterminal kinase but not via ERK1,2.
The extracellular signal-regulated kinase (ERK), stress-acti-vated protein kinase (SAPK/JNK), and mammalian homolog of the yeast-osmosensing ERK HOG1 (p38 MAPK) are conserved members of a MAP kinase cascade for regulation of targets such as transcription factors in response to growth factors or environmental stresses, such as ultraviolet light, and protein synthesis inhibitors (8 -16). ERK appears to play a major role in provoking cell proliferation and differentiation (17)(18)(19), whereas JNK mediates stress responses and some forms of apoptosis (20 -23). JNK has been implicated also in the induction of differentiation (7,24) and oncogenesis (25). The ability of overexpression of c-Jun, which is a target for JNK, to stimulate differentiation of P19 embryonal carcinoma cells to a mixed endoderm/mesoderm population supports a role of JNK in differentiation (26). P19 embryonal carcinoma (P19) cells have been used as a model system for murine pre-implantation development (27,28). These pluripotent cells have the ability to differentiate into derivatives of three germ layers, endoderm, mesoderm, and ectoderm, upon different inducer stimulation (29,30). Monolayer cultures of P19 cells challenged with 50 nM retinoic acid (RA) differentiate to endodermal-like cells (31).
In the current study we show that JNK activity is increased during RA-induced differentiation of P19 cells and the progression of P19 cell differentiation by G␣ 12 and G␣ 13 is mediated by JNK, but not ERK or p38 MAPK. Expression of the dominant negative form of JNK1 is shown to block RA-induced differentiation of embryonal cells to endoderm, establishing an obligate role of JNK in a model of early mouse embryogenesis.

EXPERIMENTAL PROCEDURES
Cell Culture and Differentiation-The P19 cells were obtained from the American Type Culture Collection (Rockville, MD). Cells were cultured in Dulbecco's modified Eagle's medium (Life Technologies, Inc.) supplemented with 10% fetal bovine serum (HyClone, Logan, UT) in 6% CO 2 humidified chamber. To induce endodermal differentiation, P19 cells were cultured as monolayers on tissue culture plates in growth media supplemented with 10 nM RA (all-trans; Sigma) for 2-5 days.
Plasmids and Transfection-The antisense sequences, 5Ј-AGCT-TCAGGCCGCGGCCGCGCCCCGCTGGGGCCCGCGCG CCAT-3Ј and 5Ј-AGCTTTGCCGCCGCCGCCGCCTCGGCGGGCCCCTGCGGCTC-CTAT-3Ј derived from the cDNA sequences of mouse G␣ 12 and G␣ 13 (32), respectively, were engineered into the HindIII/ClaI sites of the pLNCX retroviral vector using standard recombinant DNA techniques (33). Plasmids (34) harboring the constitutively activated form of G␣ 12 (pcDNA3-G␣ 12 Q229L) or of G␣ 13 (pcDNA3-G␣ 13 Q226L) were obtained from Dr. Gary L. Johnson (National Jewish Center for Immunology, Denver, CO). The plasmid containing the cDNA for the Flag epitopetagged, dominant negative form of JNK1 replaces the dual phosphorylation sites Thr-Pro-Tyr with Ala-Pro-Phe (pCMV5-JNK1APF) and was employed to probe the role of JNK in P19 cell differentiation. P19 cells were transfected with plasmids using Lipofectin TM (Life Technologies, Inc.), according to the manufacturer's protocol. For pCMV5-JNK1-(APF), the pCW1-neo plasmid was co-transfected to provide a selectable marker. Positive transfectants were selected using G418 (400 mg/ml, * This work was supported by the American Cancer Society and by Grant DK-30111 from the NIDDKD, National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Life Technologies, Inc.) and stable transfectant clones maintained in media supplemented with 100 mg/ml G418. The ability of the antisense RNAs to suppress the expression of G␣ 12 and G␣ 13 in the transfectant clones was determined by immunoblotting analysis. Expression of pcDNA3-G␣ 12 (Q229L) or pcDNA3-G␣ 13 (Q226L) was established by reverse transcriptase-polymerase chain reaction.
Immunoblotting Analysis-Aliquots of protein were subjected to 10% SDS-polyacrylamide gel electrophoresis, and the separated proteins were transferred electrophoretically from the gel to nitrocellulose membrane. Antibodies to the following antigens employed in these studies were obtained from the indicated sources: JNK 1, Santa Cruz Biotechnology, Santa Cruz, CA; ERK1 and 2, Zymed Laboratories Inc., San Francisco; p38 MAPK, Upstate Biotechnology, Lake Placid, NY, G␣ 12 , Dr. Gunther Schultz, Freie University, Berlin, Germany and Santa Cruz Biotechnology, Santa Cruz, CA; and G␣ 13 , Dr. Gunther Schultz, Freie University, Berlin, Germany and Santa Cruz Biotechnology, Santa Cruz, CA, or a rabbit polyclonal antiserum obtained by immunization with a synthetic C-terminal dodecapeptide of G␣ 13 . The immune complexes formed were made visible by using alkaline phosphataselinked goat anti-rabbit IgG (Life Technologies, Inc.), rabbit anti-mouse IgG (Life Technologies, Inc.), or rabbit anti-goat IgG (Sigma).
Immunoprecipitation of JNK and JNK Assay-An aliquot (250 g to 1 mg of protein) of the clarified lysate was mixed with the JNK1-specific antibody and resuspended to 1.0 ml in the same cell lysis buffer. After 2 h at 4°C with constant rotation, 20 -40 l of Protein A/G Plus-agarose (Santa Cruz Biotechnology, Santa Cruz, CA) was added, and the incubation was continued for an additional 1.5 h. The immunoprecipitates were pelleted by centrifugation at 14,000 ϫ g for 5 min. The pellets were washed twice with same lysis buffer and then washed twice with JNK assay buffer (20 mM HEPES (pH 7.5), 20 mM MgCl 2 , 0.1 mM Na 3 VO 4 , 2 mM dithiothreitol, 20 mM ␤-glycerophosphate). Immunoprecipitates were resuspended in a final volume of 40 l of the JNK assay buffer. A 20-l aliquot was used for Western blot analysis and the other 20 l was used for JNK assay. Aliquots of immunoprecipitates were mixed with 2 g of GST-fusion proteins encoding for the N-terminal region of Jun-(1-79) (36) and placed on ice. Reaction was started by addition of [␥-32 P]ATP (10 Ci/tube, 100 M final concentration) and incubated for 15 min at 30°C. Reaction was stopped by addition of 6 l of 4 ϫ Laemmli buffer (200 mM Tris-HCl (pH 6.8), 0.4 mM dithiothreitol, 8% SDS, 0.08% bromphenol blue, and 40% glycerol). After separation of proteins by 10% SDS-PAGE and staining with Coomassie Brilliant Blue R250 and autoradiography, the bands for GST-Jun were excised and the incorporated 32 P was quantified by liquid scintillation counting.
Source 15Q FPLC-Undifferentiated and differentiated P19 cells were harvested and lysed. An aliquot of sample (2.5 mg of protein) was resuspended in 1 ml of lysis buffer and fractionated by chromatography on a Mono Q HR 5/5 FPLC column (Pharmacia Biotech Inc., Uppsala, Sweden). The column was washed with 8 column volumes of buffer A (70 mM ␤-glycerophosphate (pH 7.2), 0.1 mM Na 3 VO 4 , 2 mM MgCl 2 , 1 mM EGTA, and 1 mM dithiothreitol), and the bound proteins were eluted with a 12.5-ml linear gradient of NaCl (0 -400 mM) in buffer A. Twenty-five fractions (0.5 ml) were collected at a flow rate of 1 ml/min. ERK and p38 Protein Kinase Assays-A 15-l aliquot of FPLC fraction was mixed with 5 l of 4 ϫ kinase assay buffer (100 mM ␤-glycerophosphate, 160 mM HEPES (pH 7.2), 0.2 mM Na 3 VO 4 , 40 mM MgCl 2 , 100 g/ml protein kinase A inhibitor peptide, 2 mM EGTA) and 2 g of either MBP (Sigma) or GST-ATF2-(1-109) (36) used as substrates. Each reaction was started by addition of [␥-32 P]ATP (10 Ci/tube, 100 M final concentration) and incubated for 15 min at 30°C. The reactions were terminated by addition of 6 l of 4 ϫ Laemmli buffer. After separation of proteins by SDS-PAGE on 12% acrylamide gels, staining of the resolved proteins with Coomassie Brilliant Blue R250, and autoradiography, the bands for either MBP or GST-ATF2-(1-109) were excised, and the incorporated 32 P into these substrates was quantified by liquid scintillation counting.
Indirect Immunofluorescent Methods and Antibodies-Cells were grown in Lab-Tek chamber slides (Nunc, Rochester, NY) and were fixed for 10 min with 3% paraformaldehyde. After fixation, slides were rinsed three times with modified Shield's media/Pipes (modified Shield's media; 18 mM MgSO 4 , 5 mM CaCl 2 , 40 mM KCl, 24 mM NaCl, 5 mM Pipes (pH 6.8), 0.5% Triton X-100, 0.5% Nonidet P-40). A monoclonal antibody MC-480 (SSEA-1) that reacts with a stem-cell surface antigen marker (37) was used to identify the embryonal-stage cells. MC-480 was obtained from the NIH Developmental Studies Hybridoma Bank (Balti-FIG. 1. P19 embryonal carcinoma cells differentiate to endodermal phenotype in response to retinoic acid. P19 cells were induced to endoderm by treatment with 100 nM RA for 3 days. Cells were fixed using 3% paraformaldehyde. Phase contrast (PC) (panels A, C, E, and G) and indirect immunofluorescence (IF) were performed upon cells following staining with either the embryonal marker SSEA-1 (panels B and D) or the endodermal marker TROMA (panels F and H). Note that positive staining for the embryonic-specific antigen SSEA-1 was confined to the untreated P19 cells and lost following RA-induced differentiation to endoderm. Likewise, positive staining for the endoderm-specific marker TROMA-1 was confined to the RA-treated cells and absent in the untreated P19 embryonic cells. The images shown are representative of at least five separate experiments, performed on different occasions for each treatment. Bar, 50 microns. more, MD). The fixed cells were incubated with primary antibody for 30 min at 37°C and then washed three times with modified Shield's media/Pipes buffer. Fluorescein-conjugated goat anti-mouse (Kirkegaard & Perry Laboratories, Gaithersburg, MD) was added and incubated for 30 min at 37°C. The slides were washed then three times with blotting buffer (560 mM NaCl, 10 mM KPO 4 (pH 7.5), 0.1% Triton X-100, 0.02% SDS). The cells were examined by phase contrast and epifluorescence microscopy on a Zeiss Axiphot microscope and photographed on Kodak-64T or Kodak-TMax 400 film.

RESULTS AND DISCUSSION
The P19 cells differentiate to neuroectoderm-like phenotypes (29,30) when allowed to aggregate in the presence of 100 nM retinoic acid. In the absence of aggregation, P19 cells are induced to differentiate by RA but only to the endoderm-like phenotype (29, Fig. 1), providing a useful model for study of this process. P19 embryonic cells stain positive in indirect immunofluorescence study of cells probed with antibody to the embryonic marker SSEA-1 (Fig. 1, panel B). The cells induced morphologically to endoderm by RA no longer stain positive with the same antibody, displaying a loss in the embryonic marker antigen SSEA-1 (Fig. 1, panel D). Endodermal cells stain positive with TROMA-1, a monoclonal antibody to the endoderm-specific marker cytokeratin-1 (37,38). Untreated P19 cells stain negative for TROMA-1 (Fig. 1, panel F), reflecting their embryonal character, whereas RA-treated P19 cells stain prominently for TROMA-1 (Fig. 1, panel H), a hallmark for endodermal development (37,38). The absence of epifluorescence signals from the RA-treated cells stained with SSEA-1, from the untreated P19 cells stained with TROMA-1, and from the controls performed without primary antibody (not shown) provide ample evidence for the specificity of the staining in defining endoderm versus embryonal character (Fig. 1). Since heterotrimeric G-proteins have been implicated in control of P19 cell differentiation (39) and G␣ 12 and G␣ 13 are known to mediate the activation of mitogen-activated protein kinases, especially c-Jun amino-terminal kinase (16,40), we explored the role of MAP kinases in P19 cell differentiation, with an emphasis on G␣ 12 and G␣ 13 .
To explore the role of MAP kinases in P19 cell differentiation, cells were treated with retinoic acid (100 nM) for 3 days to achieve full differentiation to endodermal phenotype and a cell extract prepared and fractionated by FPLC on a Source Q column ( Fig. 2A). Immunoblotting of samples from the fractions revealed ERK1,2 in fractions 9 -11 and p38 in fractions 14 -19 of extracts from untreated P19 cells. MAP kinase activity was measured using myelin basic protein (MBP), which is a substrate for both p38 and ERK1,2. The profile of activities coincides with the appearance of ERK1,2 and p38 kinases in the immunoblots, shown as insets (Fig. 2, panel A). When MAP kinase activity was measured in fractions from an FPLC of RA-treated, endodermal cells, a modest but highly reproducible (n ϭ 3) decline in ERK1,2 activity was observed in each profile (Fig. 2, panel A). MAP kinase activity in the region of the profile in which p38 was found increased slightly in the extracts from the RA-treated, endodermal P19 cells in each of three separate analyses (Fig. 2, panel A). The status of p38 kinase activity in the FPLC fractions from untreated and RAtreated P19 cells was measured further, using GST-ATF2 fusion protein, as ATF2 is a specific substrate for p38 and not ERK1,2 Samples of each fraction used in the activity assays were analyzed by immunoblotting using antibodies specific for either ERK1,2 or p38 (panels A and B). Peak of MAP kinase activities correlated well with profile of MAP kinase established by immunoblotting antibodies specific for ERK1,2 or p38 (insets to panels A and B). Panel C, undifferentiated (NO) cells and cells differentiated with 100 nM RA for 1, 2, and 3 days (RA) were cultured and then lysed. JNK was immunoprecipitated (IP) with JNK1 antibody (Santa Cruz Biotechnology, Santa Cruz, CA) from 1.0 mg of lysate protein. Kinase assays were performed using the substrate GST-Jun (upper panel). The product of phosphorylation was resolved by SDS-PAGE on 10% gels and detected by autoradiography and scintillation counting. An aliquot of the immunoprecipitated JNK was subjected to SDS-PAGE, the resolved proteins transferred to nitrocellulose membranes, and the blots stained with an antibody specific for JNK1 (bottom). The data shown are representative of at least three independent experiments, replicated on separate occasions. (Fig. 2B). The activity was confined to the same FPLC fractions as judged by immunoblotting and phosphorylation of the ATF2fusion protein. In extracts from RA-treated P19 cells, again, the p38 activity displayed a small but reproducible increase compared with the activity in profiles from untreated P19 cells.
Unlike the modest changes observed in the activity of p38 or ERK1,2 in the RA-treated cells, c-Jun amino-terminal kinase activity (JNK) displays marked activation upon treatment of the cells with RA (Fig. 2C). JNK activity in the solid-state assay using GST-Jun fusion protein increases from 2.7-fold on day 2 to 7.8-fold on day 3 following induction of differentiation in response to RA. Based upon multiple experiments, the increase in JNK activity (fold over basal) was found to be 4.3 Ϯ 0.6 (n ϭ 3) on day 2 and 7.0 Ϯ 0.2 (n ϭ 4) on day 3 following treatment with RA. The amount of the JNK itself, in contrast, was found to be relatively constant over the time frame of the induction of differentiation, whereas the activity of JNK was markedly enhanced. These data demonstrate among the various MAP kinases that JNK is markedly activated in the P19 cells induced to endoderm by RA, p38 activities increase little, and ERK1,2 activities show a modest decline.
Anisomycin is a potent activator of JNK activity (15) and was employed to explore the nature of the activation of JNK that accompanies RA treatment (Fig. 3). P19 cells were incubated with or without anisomycin (100 ng/ml) for 2 h, and the JNK was immunoprecipitated from the cell extracts and the kinase activity measured using the solid-state assay (Fig. 3A). Anisomycin proved to be a potent stimulator of JNK, increasing JNK activity by more than 3-fold within 2 h of stimulation. The stimulation by anisomycin was reversible, simply removing the anisomycin precipitated a return to base line of the JNK activ-ity in cells treated with anisomycin, washed free of the agent, and measured 1 day later (Fig. 3A). Anisomycin treatment alone was not able to induce the loss of SSEA-1 antigen nor positive staining by TROMA. The same situation was not observed with regard to the ability of RA to induce differentiation and activation of JNK in the P19 cells (Fig. 3B). Treating cells with 100 nM RA activates JNK by at least 6-fold by day 5, whether the RA is included throughout the 5-day period or added only for the first 2 of the 5-day treatment. The RAinduced increase in JNK activity is sustained following disruption of P19 cell aggregates, harvest, and re-plating of the cells (data not shown). These data suggest a fundamental difference in the activation of JNK by RA as compared with anisomycin. Commitment to the differentiated state or activation of JNK observed in response to RA in the first 2 days continues well beyond the presence of the RA, much like the expression of G␣ 12 and G␣ 13 induced by RA in P19 cells (39).
RA treatment increases the expression of G␣ 12 and G␣ 13 in P19 cells (39). Both G␣ 12 and G␣ 13 have been shown to mediate signals that act on downstream MAP kinases, particularly JNK (16,40,41). Taken together these observations stimulated us to examine the relationship between G␣ 12 and G␣ 13 levels in P19 cells and the activity of JNK (Fig. 4). P19 cells stably expressing mutant forms of either G␣ 12 (Q226L) or G␣ 13 (Q229L) that are constitutively active were probed with respect to JNK activation, in the absence of RA treatment. Expression of either Q226LG␣ 12 or Q229LG␣ 13 in P19 cells resulted in the activation of JNK activity in the absence of RA treatment. Cells expressing Q226LG␣ 12 displayed activation of JNK activity ranging from 1.9-to 2.7-fold over basal levels. P19 cells expressing Q229LG␣ 13 displayed activation of JNK activity ranging from 2.4-to 3.8-fold over basal (Fig. 4). In the absence of RA treatment, P19 cells expressing Q229LG␣ 13 have been shown to differentiate to the endodermal phenotype (39). Expression of Q226LG␣ 12 induces the differentiation from embryonal phenotype and loss of positive staining for the SSEA-1 embryonicspecific marker but not to endodermal-like phenotype (39). These data are consistent with a role for G␣ 12 and G␣ 13 controlling aspects of differentiation induced by morphogens (38) but go further to demonstrate a potential role for activation of the MAP kinase regulatory network, and specifically JNK, by RA.
The linkage between expression of G␣ 12 and G␣ 13 and JNK activity in differentiating P19 cells was approached from the  4. Expression of either constitutively active G␣ 12 (Q229L) or G␣ 13 (Q226L) activates c-Jun amino-terminal kinase in P19 embryonal carcinoma cells. JNK was immunoprecipitated with JNK1 antibody from 1.0 mg of lysate prepared from stable clones harboring empty vector alone (pCDNA3) or expressing constitutively active mutant forms of either Q229L G␣ 12 (G␣ 12 QL) or Q226L G␣ 13 (G␣ 13 QL). JNK kinase assays were performed using GST-Jun as the substrate (GST-Jun, top). The products of the kinase reactions were resolved by SDS-PAGE on 10% gels, detected by autoradiography, and quantified by scintillation counting. Portions of immunoprecipitates were subjected to SDS-PAGE, the resolved proteins transferred to nitrocellulose membranes, and the resultant blots probed with antibody specific for JNK1 (IP/blot:JNK1, bottom). The levels of JNK were equivalent among all of the experimental groups. These experiments were replicated twice, each with at least two separate clones, with identical results. opposite perspective, i.e. suppression of G␣ 12 and G␣ 13 with RNA antisense to each. Previously we demonstrated the suppression of either G␣ 12 or G␣ 13 in stable transfectants of P19 cells using the pLNCX vector (39). Suppression of G␣ 12 was 50 -70% in clones expressing RNA antisense to G␣ 12 (39), whereas induction of JNK activity upon RA treatment in those clones was attenuated significantly in comparison to clones harboring the empty vector (Fig. 5) or to wild-type cells (Fig.  2C). The same basic observations were true for the clones in which RNA antisense to G␣ 13 suppressed G␣ 13 levels 60 -80%, and induction of JNK activity upon RA treatment was found to be reduced by approximately half. Thus, expression of constitutively active forms of G␣ 12 and G␣ 13 results in constitutively elevated basal levels of JNK activity, whereas suppression of G␣ 12 and G␣ 13 results in constitutively attenuated induction of JNK activity upon RA treatment.
To test further the linkages among G␣ 12 and G␣ 13 , JNK activation and induction of differentiation by RA, we stably expressed a dominant negative mutant form of JNK1 (JNK1 (APF)) in P19 cells and examined JNK activity of the clones in the absence and presence of RA (Fig. 6A). Densitometric scanning of immunoblots reveals that the total level of JNK1 immunoreactivity increased about 50% in the cells stably expressing the immunoreactive, dominant negative form of JNK1 (not shown). The FLAG-tagged, dominant negative form of JNK1 was readily detected in the cells transfected with the vector harboring JNK1 (APF) with antibodies specific to the tag epitope (not shown). More importantly, expression of negative dominant JNK1 was observed to block completely the activation of JNK1 activity by RA, although slightly increasing the apparent level of JNK basal activity (Fig. 6A). Thus, expression of dominant negative JNK abolishes the ability of RA to activate JNK.
To test the ability of the dominant negative to block endogenous JNK1 activity, we probed activity of JNK following anisomycin activation in cells stably expressing either pCMV vector alone or pCMV vector harboring JNK1 (APF). Upon treatment with 200 ng/ml anisomycin, control cells displayed a 15-fold activation of JNK over basal levels, whereas cells expressing the dominant negative showed a 4-fold activation (Fig.  7B). These data strongly suggest that induction of JNK activity is a causal effect for differentiation rather than result of differentiation.
The capacity to abolish the activation of JNK in response to RA by expressing the dominant negative form of JNK1 pro-vided the opportunity to address directly if the activation of JNK is critical to the ability of RA to induce differentiation of the embryonal P19 cells. Clones harboring the empty vector (pCMV) as well as those expressing the dominant negative version of JNK1 were both treated with RA for 3 days and then FIG. 5. Suppression of either G␣ 12 or G␣ 13 by antisense RNA attenuates the activation of c-Jun amino-terminal kinase in P19 embryonal carcinoma cells. JNK was immunoprecipitated with JNK1 antibody from 1.0 mg of lysate prepared from stable clones harboring empty vector alone (pLNCX) or clones expressing RNA antisense to G␣ 12 (AS G␣ 12 ) or to G␣ 13 (AS G␣ 13 ). JNK kinase assays were performed using GST-Jun as the substrate (GST-Jun, top). The products of the kinase reactions were resolved by SDS-PAGE on 10% gels, detected by autoradiography, and quantified by scintillation counting. Portions of immunoprecipitates were subjected to SDS-PAGE, the resolved proteins transferred to nitrocellulose membranes, and the resultant blots probed with antibody specific for JNK1 (IP/blot:JNK1, bottom). The levels of JNK were equivalent among all of the experimental groups. These experiments were replicated twice, with two separate clones for each antisense construct, with identical results.
FIG. 6. Expression of dominant negative c-Jun amino-terminal kinase blocks activation of c-Jun amino-terminal kinase in response to retinoic acid and attenuates activation by anisomycin. Panel A, P19 clones stably transfected either with empty vector (pCMV5) alone or with pCMV5-JNK1APF (JNK1 APF) vector harboring the dominant negative form of JNK1 were incubated in the absence or presence of 100 nM RA for 3 days. JNK was immunoprecipitated with JNK1 antibody from cell lysates (0.25 mg of protein) of vector alone and pCMV5-JNK1(APF) transfected cells. Kinase assays were performed using the substrate GST-Jun. The product of phosphorylation was resolved by SDS-PAGE on 10% gels, detected by autoradiography, and quantified by scintillation counting. JNK activity is expressed as the "fold induction" of JNK activity in response to the induction by RA. Aliquots of immunoprecipitated JNK were subjected to SDS-PAGE, transferred to nitrocellulose membranes, and the blots stained with antibody specific for JNK1 (inset). The immunoblotting demonstrates equivalence of JNK1 among the experimental groups. Panel B, P19 clones stably transfected either with empty vector (pCMV5) alone or with pCMV5-JNK1APF (JNK1 APF) vector harboring the dominant negative form of JNK1 were incubated in the absence or presence of 200 ng/ml of anisomycin for 2 h. JNK was immunoprecipitated with JNK1 antibody from cell lysates (0.25 mg of protein) of vector alone and pCMV5-JNK1(APF) transfected cells. Kinase assays were performed using the substrate GST-Jun. The product of phosphorylation was resolved by SDS-PAGE on 10% gels, detected by autoradiography, and quantified by scintillation counting. JNK activity is expressed as the fold induction of JNK activity in response to the induction by RA. Aliquots of immunoprecipitated JNK were subjected to SDS-PAGE, transferred to nitrocellulose membranes, and the blots stained with antibody specific for JNK1 (inset). The immunoblotting demonstrates equivalence of JNK1 among the experimental groups. These data are from experiments replicated at least twice with identical results. stained with antibody to the embryonic-specific marker SSEA-1 (Fig. 7). When treated with RA, clones harboring the empty vector differentiated to the endodermal phenotype just as wild-type P19 cells (Fig. 1), displaying negative staining for the embryonal marker antigen, SSEA-1 (Fig. 7). Clones expressing the dominant negative from of JNK1, in contrast, stain positive for the embryonal marker in the absence or presence of RA. These data demonstrate an obligate role for JNK1 activation in the differentiation of P19 cells to endoderm stimulated by RA.
P19 embryonal carcinoma cells are an attractive model for the study of differentiation, displaying a capacity to differentiate to endoderm, neuroectoderm, and beating cardiac myocytes under the appropriate culture conditions (29,30). The P19 cell line was adopted for the study of the role of heterotrimeric G-protein subunits G␣ 12 and G␣ 13 as well as MAP kinases in the differentiation of these embryonal cells to endoderm in response to RA. G␣ 12 and G␣ 13 are expressed in P19 cells, and their levels of expression are increased in response to RA (39). Moreover, expression of the constitutively active mutants of G␣ 12 and G␣ 13 provoke differentiation from the embryonal phenotype in the absence of RA (39). The role of heterotrimeric G-proteins in the regulation of the MAP kinase regulatory network has been appreciated recently, and the current study illuminates the role of G␣ 12 and G␣ 13 in differentiation mediated via specific activation of JNK but neither ERK1,2 nor p38. The JNK pathway is required for embryonic viability (42). Transgenic mice in which the gene for MKK4, the upstream activator of JNK, has been interrupted die prior to day 11 of embryonic development (42). This lack of viability is consistent with a role of JNK in mammalian development, perhaps differentiation (42). In Drosophila, the JNK pathway is required for embryonic development (43)(44)(45). Embryonic death occurs because of a failure of the migration of the dorsal epithelial cells required for dorsal closure. The early nature of this defect precludes analysis of the role of JNK in differentiation in this system (43)(44)(45). The most compelling evidence for the role of JNK1 in P19 embryonal cell differentiation was the ability of the dominant negative mutant of JNK1 to block endodermal differentiation in response to RA. Perhaps these heterotrimeric G-proteins activate JNK via known downstream elements (46) or perhaps through a tyrosine kinase that is regulated directly by G-proteins, as recently reported in an avian lymphocyte cell line (47).