Dual roles for c-Jun N-terminal kinase in developmental and stress responses in cerebellar granule neurons

Abstract

c-Jun N-terminal kinases (JNKs) typically respond strongly to stress, are implicated in brain development, and are believed to mediate neuronal apoptosis. Surprisingly, however, JNK does not respond characteristically to stress in cultured cerebellar granule (CBG) neurons, a widely exploited CNS model for studies of death and development, despite the regulation of its substrate c-Jun. To understand this anomaly, we characterized JNK regulation in CBG neurons. We find that the specific activity of CBG JNK is elevated considerably above that from neuron-like cell lines (SH-SY5Y, PC12); however, similar elevated activities are found in brain extracts. This activity does not result from cellular stress because the stress-activated protein kinase p38 is not activated. We identify a minor stress-sensitive pool of JNK that translocates with mitogen-activated protein kinase kinase-4 (MKK4) into the nucleus. However, the major pool of total activity is cytoplasmic, residing largely in the neurites, suggesting a non-nuclear role for JNK in neurons. A third JNK pool is colocalized with MKK7 in the nucleus, and specific activities of both increase during neuritogenesis, nuclear JNK activity increasing 10-fold, whereas c-Jun expression and activity decrease. A role for JNK during differentiation is supported by modulation of neuritic architecture after expression of dominant inhibitory regulators of the JNK pathway. Channeling of JNK signaling away from c-Jun during differentiation is consistent with the presence in the nucleus of the JNK/MKK7 scaffold protein JNK-interacting protein, which inhibits JNK-c-Jun interaction. We propose a model in which distinct pools of JNK serve different functions, providing a basis for understanding multifunctional JNK signaling in differentiating neurons.

Fig. 1.

Cerebellar granule neuron-derived JNK activity is elevated above that of stressed U937 cells, whereas p38 activity from neurons is low and responds to stress. A, Equal amounts of protein (25 μg) from 1 DIV cerebellar granule neuron (CBG) or U937 cell extracts were electrophoresed and immunoblotted with anti-SAPK. JNK expression was on average 1.7-fold higher in CBG extracts. B, Kinase assay lysates were normalized for JNK expression, and specific JNK activity was measured by immune-complex kinase assay. The specific activity of JNK from 1 DIV CBG extracts was higher than that in anisomycin-treated U937 cells. Treatments were with calyculin A (1 μm) for 30 min or anisomycin (50 μg/ml) for 45 min. After stimulation, JNK was isolated using polyclonal anti-SAPK, and immune-complex kinase assays were performed using GST-c-Jun(5–89) as substrate. C, Treatment of cerebellar granule neuron-derived JNK with alkaline phosphatase indicated that the basal activity measured was phosphorylation dependent. Representative images are shown, and quantitated data from four experiments are in Figure 2.D, p38 expression was compared in 1 DIV cerebellar granule neurons and U937 cells. Forty micrograms of CBG and U937 cell lysates were resolved by SDS-PAGE and immunoblotted using polyclonal anti-p38 (n = 4). U937 cells expressed on average sevenfold more p38 than cerebellar granule neurons. E, As in B, lysates were normalized for kinase expression, and specific activities of p38 from cerebellar granule neurons and U937 cells after treatment for 45 min with anisomycin (50 μg/ml) or sorbitol (300 mm) are shown. Treatment with anisomycin lead to a 399 ± 59% increase in p38 activity in neurons, comparable with that in U937 cells of 302 ± 80%. Representative images from four separate experiments are shown (A–E).

Fig. 2.

Glutamate receptor activity does not mediate elevated basal JNK activity. To determine whether glutamate receptor activity might contribute to basal JNK activity (∼10-fold elevated compared with cell lines), cerebellar granule neurons at 1 DIV were treated for 1 hr with 50 μm CNQX, 25 μm(RS)-MCPG, 2 μm (+)-MK801, or 1 μm nifedipine. JNK activity was assessed by phospho-JNK immunoblotting and normalized to control samples. These treatments did not reduce neuronal JNK activity; indeed, inhibition of mGluR-type receptors caused a small but significant increase in JNK activity. To investigate whether culture medium supplements might contribute to the basal JNK activity, the effects of serum depletion (−FCS) and low KCl were assessed by phospho-JNK immunoblotting. Total JNK activity was measured from 7 DIV CBG neurons grown for 24 hr in the absence of FCS but with supplementary KCl (25 mm) to prevent neuronal death. Both activity and expression of CBG JNK decreased; thus specific activities (s.a.) for JNK are shown for these treatments. Serum depletion did not significantly alter specific JNK activity. To assess whether KCl might contribute to the high basal activity, we measured JNK activity from 1 DIV CBG neurons grown for 24 hr in low KCl (5 mm). The activity was not significantly reduced. Quantitated data for JNK activation after treatment with classic JNK stimuli as described in Figure 1 are shown. Cerebellar granule neurons at 1 DIV were treated with anisomycin (50 μg/ml) or sorbitol (300 mm) for 45 min, or with alkaline phosphatase (see Materials and Methods); JNK activity was measured by immune-complex kinase assay. Normalized, mean data ± SEM are shown. Significance levels as assessed by Student's ttest are indicated (*p < 0.05 and **p < 0.001).

Fig. 3.

Anisomycin induces moderate activation of MKK4/JNK signaling and phosphorylation of c-Jun in cerebellar granule neurons.A, Cerebellar granule neurons at 1 DIV were stimulated with or without anisomycin (50 μg/ml) for 40 min. JNK and MKK4 were immunoprecipitated from lysates, and immune-complex kinase assays were performed. Representative images from multiple experiments (n ≥ 4) are shown. Lysates of anisomycin-treated 1 DIV neurons were also blotted for c-Jun protein and Ser-63 phosphorylation. Although anisomycin only moderately activates JNK, there is a mobility shift with a dramatic loss of the most mobile form of c-Jun. B, A time course for anisomycin-induced (10 μg/ml) p-46 JNK activation is shown (mean ± SEM,n = 3). C, Qualitative RT-PCR analysis of c-jun mRNA from 5 DIV cerebellar granule neurons after treatment for 3 hr with anisomycin (10 μg/ml) ± SB203580 (1 μm). The anisomycin-induced increase inc-jun mRNA is not eliminated by treatment with the p38 inhibitor, suggesting that JNK activation might contribute to the induction of c-jun. The bottom panelshows actin levels from parallel samples, indicating that the changes seen do not result from a general regulation of mRNA levels.Lanes 1 and 2 on this panel show reactions with equal (1×) or twice (2×) the amount of input cDNA as in other samples, indicating that the reactions were not saturating. D, SB203580 (1 μm)blocks p38α activity in these cells as detected by MEF2A-GAL4 induction of a GAL4-driven luciferase reporter. Quantitated data are shown (mean ± SEM).

Fig. 4.

Specific activities of JNK and JNK kinases from brain tissues and primary neurons are elevated compared with those from SH-SY5Y and PC12 cells. A, Rat brain tissues and cerebellar granule neurons (CBG neurons) were compared with non-neural tissue and neuron-like cell lines, SH-SY5Y and PC12 cells, for both JNK expression and specific activity. Top panel, Equal protein (25 μg/lane) from tissues and cell lysates was loaded and immunoblotted using anti-SAPK. Elevated levels of JNK protein are seen in brain tissues and primary neurons. The neuroblastoma cells also show elevated JNK expression compared with PC12 cells. Bottom panel, Lysates for kinase assays were normalized for JNK expression, and immune-complex kinase assays were performed. Although JNK levels were normalized between samples, JNK activities from brain tissues and primary neurons are still considerably higher than from liver, SH-SY5Y, and PC12 cells. Data shown for tissue samples and cells are representative of two and four repeats, respectively. B, Specific JNK expression and activity from forebrain was compared with other tissues as described inA except that activities were this time measured with phosphospecific JNK antibodies. Again, neural-derived JNK activity is elevated above that of JNK from non-neuronal tissues. Data shown are representative of tissues from three separate animals.C, Specific expression and activity of JNK from rat cerebellum of postnatal days 9 and 25 are shown. Normalization of samples was performed exactly as described in A. The specific activity of cerebellar JNK increases during this time. Data shown represent results from three experiments. D, The same procedure as described in A was used to compare MKK4 and MKK7 expression in brain tissues and CBG neurons to cell lines. Expression of both MKK4 and MKK7 is high in brain. In primary cultured neurons and cell lines, levels of MKK4 and MKK7 are similar, with the exception of MKK4 in neuroblastoma cells. Data from tissues and cells are representative of two and four separate experiments, respectively. E, Representative images from two-step immune-complex kinase assays showing the developmental regulation of MKK4 and MKK7 activities in cerebellar granule neurons are shown. Representative images of four and three repeats are shown, respectively. Quantitated data are shown in Figure 5.

Fig. 5.

JNK and MKK7 specific activities increase, whereas expression and activity of c-Jun decrease during maturation of cerebellar granule neurons. A, Lysates from differentiating cerebellar granule neurons in culture were loaded according to cell number and immunoblotted for JNK, p38, MKK4, MKK7, and c-Jun expression (♦) or total phospho-JNK, phospho-p38, or p63-c-Jun immunoreactivity, or MKK4 and MKK7 immune-complex kinase activities were measured (▪). Quantitated data from multiple experiments (mean ± SEM, n ≥ 3) are shown. Error bars are shown for each data point; SEMs smaller than the symbols are not visible. Although JNK and MKK7 expression are not significantly upregulated during differentiation, their activities rise sharply. Levels of c-Jun, p38, and c-Jun serine 63 phosphorylation, and activation of p38, all decrease as cells mature invitro.B, Bright-field images of cerebellar granule neurons differentiating in culture. By 6 DIV, a dense network of processes has formed.

Fig. 6.

Developmental upregulation of CBG JNK activity occurs in cells grown with or without elevated KCl. A, Specific JNK activities from differentiating cerebellar granule neurons grown in high (25 mm) or low (5 mm) KCl were measured as described in Figure 4B. Lysates were normalized for expression of JNK, and kinase activity was assessed using phosphospecific JNK. B, Quantitated data, means ± SEM from four separate experiments, are shown.

Fig. 7.

JNK activity is predominantly extranuclear in cerebellar granule neurons in culture. A, Cytosolic and nuclear fractions of cerebellar granule neurons at 7 DIV were analyzed for expression of JNK and JNK kinases MKK4 and MKK7. Cultures were lysed, cytosolic and nuclear extracts were separated, and amounts corresponding to ∼75,000 cytosols and nuclei were loaded on gels, and expression levels were analyzed. Nuclear and cytosolic fractions were then normalized for JNK expression, and specific phospho-JNK immunoreactivity was measured (P-JNK). Nuclear and cytosolic fractions were also blotted for c-Jun and Iκ-B to validate the fractionation procedure. B, Quantitated data (mean ± SEM, n = 3) are shown.C, Lysates from cerebellar granule neurons at 6 DIV were immunoblotted with isoform-specific antibodies detecting JNK1, JNK2/1, and JNK3/1, respectively. All three JNK isoforms are expressed in cerebellar granule neurons in culture. D, Western blots of CBG lysates show the relative specificity of the MAPK and MAPKK antibodies used in the immunofluorescent analysis (Fig. 8).

Fig. 8.

JNK/MKK4 are cytosolic, whereas MKK7 is exclusively nuclear in cerebellar granule neurons. Immunofluorescent microscopy was used to investigate the subcellular localization of endogenous JNK and JNK kinases in cerebellar granule neuronsinvitro. Shown are confocal scans of immunofluorescent staining for JNK, P-JNK, MKK4, MKK7, and p38 in cerebellar granule neurons at 6–7 DIV(green). Nuclei were counterstained with propidium iodide (red). Overlapping staining appearsyellow. Sections through the nuclei are shown withinsets of scans closer to the coverslip that highlights staining in neuritic processes. Neuronal staining in the absence of primary antibodies, using biotin-conjugated anti-rabbit (−1°αr) and anti-goat (−1°αgt) is shown. Allpanels are scaled according to the scale bar depicted in the p38 panel except for the HeLa staining, which has its own scale. For comparison, endogenous MKK7 immunoreactivity in HeLa cells is shown. Propidium iodide gives relatively higher background cytoplasmic staining in HeLa cells, resulting in some yellow showing in HeLa cytoplasm. Staining in HeLa cells in the absence of 1° antibody is also shown (HeLa−1°αgt).

Fig. 9.

Translocation of nuclear JNK occurs after stress and during maturation of cerebellar granule neurons. A, Nuclear levels of JNK, MKK4, MKK7, and P-JNK after treatment with anisomycin were assessed by immunoblotting of nuclear fractions from 6 DIV cerebellar granule neurons treated ± anisomycin (50 μg/ml) for 40 min. Levels of JNK, phospho-JNK, and MKK4 increase twofold in nuclear fractions after stress, consistent with activation and nuclear translocation of these kinases. B, The developmental increase in JNK activity occurs in both cytoplasmic and nuclear compartments. Nuclear and cytoplasmic fractions were prepared from CBG neurons at 1, 3, and 6 DIV and immunoblotted for phospho-JNK. To enable a direct comparison of nuclear and cytoplasmic fractions to be made, 10-fold more nuclei than cytoplasms were loaded (fraction of total).C, Quantitated data (mean ± SEM,n = 3) for multiple experiments as outlined inB.

Fig. 10.

JIP is expressed in nuclear fractions from cerebellar granule cells. A, JIP mRNA was detected in 9 DIV cell extracts by RT-PCR with oligos complementary to both JIP-1a and JIP-1b/1c/2/3, expected to produce fragments of size 324 (JIP⁻) and 465 (JIP⁺), respectively.B, Equal proportions of cytosolic and nuclear fractions from 6 DIV CBG neurons were loaded and immunoblotted with monoclonal anti-JIP-1. JIP expression was predominantly nuclear, although lower levels of JIP were detected in cytosolic fractions. C, Confocal sections through nuclei of 6 DIV CBG neurons stained for JIP-1 and the corresponding −1° sample are shown; JNK staining is also shown for comparison. The inset shows a section closer to the coverslip.

Fig. 11.

Inhibition of neuronal JNK activity results in increased projections from the cell body. A, Five daysin vitro cerebellar granule neurons were transfected with pEBG-JNK1α1, pEBG-SEK1KR, or pcDNA3-JIP-JBD as shown, and 24 hr after expression, JNK1α1 was isolated on GSH beads and its activity was measured. Recombinant JNK1α1 activity is elevated in unstimulated neurons, thus mimicking endogenous JNK activity. Coexpression of dominant inhibitory kinases SEK1KR or JIP-JBD effectively blocks JNK1α1 activity. B, Three days in vitrocerebellar granule neurons were transfected as above with the addition of a β-galactosidase transfection marker. Forty-eight hours after transfection, cells were stained for β-galactosidase expression, and the number of processes emerging from living cells with lengths more than or equal to the cell body radius were counted. The percentage of cells with a given number of processes (1–5 or >5) were calculated for randomly chosen fields from five to six coverslips per condition. The number of cells counted under each condition was 233 for pEBG (n = 5), 285 for SEK1KR (n = 5), and 305 for JIP-JBD (n = 6). Data are expressed as means ± SEM. C, A representative image from cells transfected as in A and B above is shown.

Fig. 12.

A model depicting the dual regulation of neuronal JNK signaling in cerebellar granule neurons in response to stress and during differentiation. In unstressed cerebellar granule neurons, JNK and MKK4 are localized predominantly in the cytoplasmic compartment.A, After treatment with the classic JNK activator anisomycin, MKK4 and JNK are activated, followed by their increased nuclear localization. Increased active JNK in the nucleus phosphorylates c-Jun, resulting in activation of stress-induced genes such as c-jun. B, During differentiation, cytoplasmic JNK (colocalized with MKK4) is activated threefold, and nuclear JNK (colocalized with MKK7) is activated 10-fold. Maintained localization of the major pool of MKK4/JNK activity to the cytoplasm suggests the presence of cytoplasmic JNK targets. Inhibition of JNK activity results in changes in neurite number, indicating a regulatory role for JNK signaling in morphological changes. JIP, a JNK-MKK7 scaffold that competes with c-Jun for binding to JNK, and MKK7 are colocalized in the nuclear compartment. This suggests a possible role for JIP in directing nuclear JNK activity away from c-Jun regulation and in upregulating maturation-specific genes.