Stress-induced JNK Activation Is Independent of Gadd45 Induction*

DNA damage and environmental stress activate signaling and induce genes involved in cell cycle and cell death. Expression of the Gadd45 protein is induced following DNA damage and other stress. Gadd45 is believed to play a role in growth arrest and possibly in cell death. The JNK signaling pathway is also activated by some DNA-damaging agents. This activation leads to phosphorylation and activation of transcription factors, such as c-Jun/AP-1 and ATF2, which mediate immediate early gene induction. Recently Gadd45 was suggested to be involved in JNK activation. However, as this suggestion relied onin vitro experiments and ectopic overexpression of Gadd45 protein, we examined whether physiological levels of Gadd45 that are induced following exposure to DNA damaging agents and stress can lead to JNK induction. We found that JNK activation by UV irradiation and anisomycin treatment precedes the induction of gadd45mRNA by these agents. Gadd45 protein induction by methyl methanesulfonate also lagged behind JNK activation. The use of protein synthesis inhibitors suggested that newly synthesized proteins, including the stress-induced Gadd45, make only a marginal contribution to JNK activation. We also found that stresses such as γ irradiation induce Gadd45 and do not activate JNK in mouse fibroblasts. Therefore, stress-induced JNK does not depend on Gadd45 induction.

Exposure of cells to environmental stress results in activation of several signal transduction pathways. A major factor in the cellular response to DNA damage and other types of stress, including nucleotide pool depletion and hypoxia, is the p53 tumor suppresser gene (1,2). Upon detection of DNA damage the p53 protein, which functions as a transcription factor, is stabilized, and its elevated level results in induction of p53 target genes (3). GADD45 is one of several known p53 target genes. However, while GADD45 induction by ␥ radiation is p53-dependent (4), other factors such as short wavelength ultraviolet (UV) light, methyl methanesulfonate (MMS), 1 ⌬12-prostaglandin J2, and camptothecin can induce Gadd45 in the absence of active p53 (5)(6)(7). Its rapid induction following exposure to DNA-damaging agents suggests that Gadd45 has a role in the response to DNA damage and other cellular insults. Several biological activities have so far been suggested for Gadd45, including involvement in DNA repair. Reduced Gadd45 expression correlates with decreased DNA repair (8). The ability of Gadd45 to modify DNA accessibility on damaged chromatin (9) and to interact with PCNA (10), a protein involved in DNA replication and repair, might support a role in this activity. Gadd45 was also shown to be an essential component of the G 2 /M cell cycle checkpoint induced by UV light or MMS (11). This activity is at least partially dependent on its ability to inhibit Cdc2/cyclin B kinase activity (11,12). Gadd45 has also been implicated in apoptosis, but this biological activity is controversial, as studies suggesting that Gadd45 can induce apoptosis (13) are opposed by others, which suggest that Gadd45 induction has a protective function (8,14).
Other pathways that are activated by various genotoxic stresses lead to activation of various MAP kinases (MAPKs), most notably JNKs and p38 (15). JNK is rapidly activated by exposure of cells to certain DNA damaging agents, such as UV light and MMS, as well as by treatment with proinflammatory cytokines and growth factors (15). Activated JNK phosphorylates and thereby enhances the transcriptional activity of several transcription factors, including c-JUN, ATF2, and elk1. This results in induction of immediate early genes, whose end result is cell type-specific, in some cells leading to increased cell proliferation and in others to apoptosis (16 -20). The duration of JNK activation was suggested to be the regulating factor that determines the cell fate (21). In some cases dominant negative mutants of c-Jun that can not be phosphorylated by JNK failed to mediate apoptosis activation (22).
A recent study presented data linking Gadd45 to the JNK pathway. Gadd45 and two other homologs, designated Gadd45␤ (also known as MyD118; Ref. 23) and Gadd45␥ (also known as CR6), were identified through a two-hybrid screen as proteins that interact with MTK1/MEKK4, a MAPK kinase kinase (MAPKKK) that can activate JNK and p38 subgroups of MAP kinases (13). Interestingly, Gadd45 proteins not only interact with MEKK4 but can also stimulate its protein kinase activity in vitro. Transient overexpression of Gadd45 was reported to activate JNK and p38 in intact cells, presumably in an MEKK4-dependent manner (13).
We were interested to test whether stress induction of endogenous Gadd45␣, the most abundant member of the family, can lead to JNK activation. In this work we report that JNK activation by UV light, MMS, and the protein synthesis inhibitor anisomycin is independent of Gadd45␣ induction. Furthermore, induction of Gadd45␣ expression after exposure of cells to ␥ irradiation is not linked to JNK activation. Others have found that Gadd45␣-deficient cells do not exhibit any defect in JNK or p38 activation (36). Thus, at least Gadd45␣ is unlikely to function as a physiologically relevant activator of the JNK pathway.

MATERIALS AND METHODS
Cell Culture-3T3 mouse fibroblasts used were grown in Dulbecco's modified Eagle's medium containing 10% fetal calf serum at 37°C and in the presence of 5% CO 2.
Immunoblotting-The levels of immunoprecipitated JNK were determined by gel separating the proteins used in the kinase assay and transferring them to Immobilon-P membranes (Millipore) that were probed with polyclonal antibody for JNK1 (Santa Cruz Biotechnology). Phospho-JNK levels were determined by probing membranes containing equivalent levels of whole cell extract with anti phospho-JNK antibodies (Promega). Gadd45␣ levels were determined using anti-Gadd45␣ polyclonal antibodies (Santa Cruz Biotechnology).

Stress-induced Activation of JNK Precedes gadd45␣
Induction-To test whether Gadd45 induction correlates with JNK activation following exposure of cells to environmental stress, we exposed p53 ϩ/ϩ mouse fibroblasts to short wavelength UV irradiation (UVC; 50 J/m 2 ), the alkylating agent MMS (100 g/ml), or to the protein synthesis inhibitor anisomycin (1 g/ml) and compared the kinetics of gadd45␣ mRNA induction to that of JNK activation. It is worth mentioning that UVC is an inducer of gadd45␣ and ␤, while MMS induces all 3 gadd45 homologs. Anisomycin is a more potent inducer of gadd45␣ and ␤ mRNA than UVC (13). All agents are potent JNK activators (25). Weak induction of gadd45␣ mRNA was observed only 1 h after UV treatment. The levels of gadd45␣ mRNA reached the peak after 4 h and remained high at least until 7 h (Fig. 1A). Anisomycin treatment led to gadd45␣ mRNA induction only after 4 h ( Fig. 2A), whereas MMS treatment induced gadd45␣ mRNA expression within 1 h, reaching a peak within 4 h and declining to lower levels 7 h after treatment, possibly due to cell death (Fig. 3A). Examination of JNK activation kinetics by these inducers revealed a faster activation pattern. In agreement with previous publications (26 -28), JNK activation, measured by the N-terminal phosphorylation of c-Jun, was maximal at 30 min after UVC exposure, remained high after 1 h, and declined to basal levels by 4-h postirradiation (Fig. 1,  ϪCycloheximide). The kinetics of JNK activation after anisomycin treatment was similar, with high levels of activated JNK observed up to an hour after induction and then declined to the basal levels (Fig. 2). By comparison to UVC and anisomycin, MMS activates JNK with slower kinetics: peak activity was reached at 4-h posttreatment, although considerable activation was observed by 1 h (Fig. 3, ϪCycloheximide). These experiments indicate that JNK activation at least by UV irradiation and anisomycin treatment actually precedes the induction of gadd45␣ mRNA expression.
JNK Activation Can Take Place in the Absence of Gadd45␣ Protein Induction-To further test whether Gadd45 induction is required for JNK activation by genotoxic stress p53 ϩ/ϩ and p53 Ϫ/Ϫ mouse fibroblasts were exposed to UVC irradiation and MMS in the presence or absence of protein synthesis inhibitor cycloheximide, and the effect on Gadd45␣ protein expression was compared with the effect on JNK activation. Gadd45␣ is considered to be a p53 target gene, and in one case a model suggesting that Gadd45␣ mediates p53 driven apoptosis by the FIG. 2. Activation of JNK by anisomycin is Gadd45␣-independent. p53 Ϫ/Ϫ and p53 ϩ/ϩ mouse fibroblasts were treated with anisomycin (1 g/ml) and harvested at the indicated times. A, gadd45␣ mRNA induction was determined by Northern blotting. GAPDH mRNA served as a loading control. B, Gadd45␣ expression was determined as described above, while JNK activation was determined by immunoblotting with an anti-phospho-JNK antibody (Promega).

FIG. 3. Activation of JNK by MMS does not depend on Gadd45␣
induction. p53 ϩ/ϩ mouse fibroblasts were incubated with 100 g/ml MMS in the presence or absence of cycloheximide (10 g/ml) and harvested at the indicated time points. A, gadd45␣ mRNA levels were determined by Northen blot hybridization. GAPDH was used as a loading control. B, Gadd45␣ expression was determined by immunoblotting, and JNK activity was measured by an immune complex kinase assay. Recovery of JNK1 was determined by immunoblotting. activation of JNK was drawn (13). However, in agreement with a previous report (5), Gadd45␣ was induced by both agents regardless of the p53 status ( Fig. 1 and data not shown). Therefore, we examined the relationship between Gadd45␣ induction and JNK activation only in p53 ϩ/ϩ cells. As expected from the induction kinetics of gadd45␣ mRNA, expression of Gadd45␣ protein can be detected only 4 h after UVC exposure (Fig. 1B). At this time point JNK activation has already declined. Treatment with cycloheximide completely abolished Gadd45␣ expression, but had no significant effect on JNK activation (except for a slight elevation of its basal level). Therefore, UVC irradiation activates JNK independently of new protein synthesis, including that of Gadd45␣. Furthermore, there was no change in the JNK activation kinetics that would support a role for Gadd45␣ in a late phase of the JNK activation response. Gadd45␣ protein induction by MMS is detected only 4 h after treatment, while JNK activity can be detected at 1 h after treatment (Fig. 3B). Therefore, induction of Gadd45␣ synthesis by MMS also lags behind JNK activation. Inhibition of Gadd45␣ expression by cycloheximide did not prevent JNK activation (Fig. 3B, ϩCycloheximide). Interestingly, inhibition of protein synthesis rendered the cells more sensitive to MMS exposure and enhanced the extent of apoptosis (data not shown). This observation suggests that newly synthesized proteins induced by an MMS-activated stress response pathway promote cell survival.
Anisomycin is a potent inducer of gadd45␣ and ␤ mRNA ( Fig. 2A and Ref. 13). However, at the concentration used for RNA induction, anisomycin, a potent protein synthesis inhibitor, completely blocked Gadd45␣ protein synthesis (Fig. 2B). On the other hand, as described previously (29), treatment with the same concentration of anisomycin resulted in efficient JNK activation in both p53 ϩ/ϩ and p53 Ϫ/Ϫ cells (Fig. 2B). As mentioned before, JNK activation preceded even the induction of gadd45␣ mRNA. These experiments strongly suggest that induction of Gadd45␣ protein is neither required for JNK activation by UVC, MMS, or anisomycin nor is it involved in prolonging the JNK activation response.
Elevated Levels of Gadd45 Do Not Increase JNK Activation-Exposure of cells to ionizing radiation (IR) results in a p53-dependent induction of Gadd45␣ (4). We took advantage of this response to test a second point, can DNA damage-induced Gadd45 activate JNK in intact cells? We exposed p53 ϩ/ϩ and p53 Ϫ/Ϫ cells to 20 grays of IR and examined JNK activation and its relationship to Gadd45␣ induction. As expected, Gadd45␣ expression increased after exposure to IR only in p53 ϩ/ϩ cells ( Fig. 4 and data not shown). In p53 ϩ/ϩ cells the induction of Gadd45␣ by IR occurred with faster kinetics than its induction by UV or MMS (compare Fig. 4 to Figs. 1 and 3). Despite the rapid induction of Gadd45␣, exposure to IR did not result in considerable JNK activation (Fig. 4). At most, the extent of JNK activation in IR exposed cells was 100-fold lower than in UV-irradiated cells. These results are consistent with an earlier report that IR is a very poor JNK activator in comparison with UVC and MMS (25). DISCUSSION DNA damage and environmental stress elicit diverse signaling and gene induction responses. The MAPK cascade that leads to JNK activation is regulated by a complicated network of upstream kinases and small G proteins, as well as by phosphatases (15). In view of this complexity, multilevel regulation is expected, and in fact different stimuli probably activate JNK via different upstream regulators. As demonstrated previously (26 -28) UV irradiation results in a rapid JNK activation that is terminated within 1 h. These rapid kinetics are consistent with a process that does not require de novo protein synthesis. The rapid activation of JNK in UV-irradiated cells was shown to depend on two very early signaling events: clustering of cell surface receptors (30) and inhibition of membrane-associated tyrosine phosphatases (31).
Recently, however, it was suggested that induction of Gadd45␣ protein contributes to JNK activation in response to UV irradiation and other stress (13). This suggestion rested on two findings: a two-hybrid screen and overexpression experiments. Gadd45 proteins were found to bind to MEKK4, and recombinant Gadd45 proteins were found to activate MEKK4 (13). As MEKK4 was already known as one of many MAPKKKs that can lead to JNK and p38 activation, it seemed reasonable to assume that induction of Gadd45 proteins, a common event in cells exposed to a variety of stresses, including UVC radiation (13), can lead to JNK activation. Indeed, it was shown that overexpression of Gadd45 proteins can lead to JNK activation (13). However, it was not demonstrated that this process is mediated via MEKK4 or that it contributes to JNK activation in cells exposed to environmental stress. We examined the later possibility and found no correlation between the kinetics of Gadd45␣ induction, either at the protein or at the mRNA level, and JNK activation in response to three diverse stress stimuli: UVC, MMS, and anisomycin. In all of these cases, JNK activation clearly preceded Gadd45␣ expression. Furthermore, inhibition of Gadd45␣ expression and presumably expression of most other induced proteins, including Gadd45␤ and Gadd45␥, did not abolish JNK activation by these three stimuli or alter the kinetics of activation. Gadd45␣ was also induced by a fourth stimulus, ␥ irradiation, which did not lead to considerable JNK activation. These results strongly rule out the possibility that Gadd45␣ makes an essential contribution to JNK activation, either at an early phase or at a late phase of the response. Although we cannot rule out the involvement of Gadd45␤ and Gadd45␥ in JNK activation, by conjecture we suggest that it is unlikely that any of these proteins is a critical modulator of JNK activation, which is a rather rapid response to many forms of environmental stress. A more likely scenario is that JNK and/or p38 may contribute to Gadd45␣ induction. Indeed, it was shown recently that induction of Gadd45␣ by osmotic stress partially depends on p38 activity (32).
The ability of ␥ irradiation to induce Gadd45␣ without activating JNK shows clear dissociation between the events. Although the levels of p53 (33) and Gadd45␣ (this study) that are induced by ␥ radiation are lower than the ones induced by UVC radiation, the p53-dependent biological effects of the two forms of radiation on cell proliferation are similar; the doses that lead FIG. 4. Ionizing radiation induces Gadd45 expression but not JNK activity. p53 ϩ/ϩ mouse fibroblasts were exposed to 20 grays of ionizing radiation (␥ rays) and harvested at the indicated times. Gadd45␣ expression was detected by immunoprecipitation followed by immunoblotting with anti-Gadd45␣ antibodies. JNK activation was determined both by immunoblotting with an anti-phospho-JNK antibody and by an immune complex kinase assay. An identical amount of whole cell extract of p53 ϩ/ϩ cells exposed to UVC (50 J/m 2 ) was used as a positive control.
to Gadd45␣ induction totally inhibited colony formation by mouse fibroblasts (data not shown).
The huge difference in the magnitude of JNK activation by the two forms of radiation suggest that JNK is not responsible for the observed inhibition of cell proliferation.
However, Gadd45␣ expression in primary human fibroblasts was shown to lead to growth arrest (11). The induction of Gadd45␣ by osmotic stress was also shown to correlate with growth arrest (32). So far there is little evidence implicating JNK in radiation mediated growth arrest of mouse fibroblasts. Thus, while GADD45␣ induction is undoubtedly an important response to environmental stress, the downstream effects of this protein, which is mostly nuclear (34), are unlikely to be mediated by JNK, which in nonstimulated cells resides in the cytoplasm (35).