Differential Activation of c-Jun NH₂-terminal Kinase and p38 Mitogen-activated Protein Kinases by Methyl Methanesulfonate in the Liver and Brain of Rats: Implication for Organ-specific Carcinogenesis¹

DNA-damaging agents such as UV light, ionizing radiation,1-β-d-arabinofuranosylcytosine, cis-platinum,mitomycin C, chemotherapeutic drugs, hydrogen peroxide, and alkylating agents are known to activate members of the MAPK³family in cultured mammalian cell lines (1, 2, 3, 4, 5, 6, 7). The MAPK family consists of three subgroups: (a) the extracellular signal-regulated protein kinases; (b) the JNKs; and(c) the p38 MAPKs (8, 9). The JNKs and p38 MAPKs are collectively termed SAPKs because they are activated by similar stress-related stimuli (10). Despite the apparent coordinate regulation of JNK and p38, these protein kinases have distinct substrate specificity: JNK phosphorylates c-Jun, ATF-2, Elk-1,and p53 (11, 12, 13); whereas p38 phosphorylates MAPK-activated protein kinases 2 and 3 and CHOP as well as ATF-2 and Elk-1 (14, 15, 16). Moreover, the kinase cascade leading to the activation of JNK is distinct from the kinase cascade leading to the activation of p38. Whereas SEK1/MKK4 phosphorylates and activates both JNK and p38 (17, 18, 19), MKK3/MKK6 and MKK7 only phosphorylate and activate p38 and JNK, respectively (17).

Recently, in vivo studies have demonstrated that the coordinate activation of JNK/p38 is not always observed and may reflect an in vitro phenomenon: JNK activation by oxidative stress in the liver correlated with decreased p38 MAPK activity(10); and the signal-dependent activation of the SAPK cascades is distinct in different cell types (20). Because the signal transduction cascades leading to the activation of JNK/p38 and to subsequent gene induction are associated with stress responses that promote either cell recovery and survival after cellular damage or apoptotic death (21, 22, 23, 24), differential activation of the signal transduction pathways may result in distinct phenotypic consequences of stressful stimuli in target cells.

MMS is a direct-acting alkylating agent known to cause cell death,mutation, chromosome damage, and neoplastic transformation(25). Intriguingly, MMS is a strong inducer of brain tumors, whereas it is a weak hepatocarcinogen, even in the regenerating liver of rats (26, 27, 28, 29, 30, 31, 32). The tissue specificity of the carcinogenic effect of MMS is of particular interest because MMS does not require metabolic activation and is therefore not tissue specific. Moreover, the initial amount of DNA damage caused by MMS has been shown to be similar in the liver and brain of rats after i.p. injection of MMS (26, 33, 34, 35, 36). Therefore, the cellular response(s)induced by damaged DNA, such as apoptosis or activation of signaling systems, rather than the DNA damage itself may determine the tissue specificity of tumor induction by MMS.

In the present study, we have addressed the potential involvement of two major stress signaling pathways, namely the JNK and p38 pathways,in tissue-specific responses induced by MMS in rats. We demonstrate the differential activation of JNK and p38 and a strong correlation between the patterns of stress kinase activation and apoptotic cell death in the liver and brain of rats exposed to MMS. To our knowledge, our study is the first to show that a tissue-specific signaling pathway(s)in response to MMS exists, implying that this distinct kinase signaling pathway may be an important molecular component responsible for differential responses of tissues after exposure of the whole animal to alkylation damage.

Male Sprague Dawley rats weighing 110–120 g were given an i.p. injection of 500 μl of MMS (Sigma) diluted in water to the concentrations indicated in the figures and then decapitated at the given time points. Livers and brains were homogenized immediately at 4°C with a glass-Teflon homogenizer in 10 volumes of prechilled homogenation buffer [25 mm HEPES (pH 7.5), 300 mm NaCl, 1.5 mm MgCl₂,0.2 mm EDTA, 0.05% Triton X-100, 20 mmβ-glycerophosphate, 1 mm orthovanadate, 0.5 mm DTT, 0.4 mm phenylmethylsulfonyl fluoride, 2μg/ml leupeptin, and 1 μg/ml pepstatin]. The extracted tissue homogenates were centrifuged at 12,000 × gfor 30 min at 4°C, and the supernatants were boiled for 5 min in Laemmli’s sampling buffer [50 mm Tris (pH 6.8),2% SDS, 10% glycerol, 0.001% bromphenol blue, and 50 mm β-mercaptoethanol]. After determining the protein concentration, the prepared samples were stored at −80°C. A minimum of six animals per data point was used in three independent experiment, i.e., a minimum of two animals per point, and representative data are shown in the figures.

Between 100 and 120 μg of protein extracts were electrophoresed on an 8% SDS-polyacrylamide gel and then transferred to nitrocellulose membranes (Amersham). The blots were blocked with TTBS containing 5%nonfat dry milk for 1 h at room temperature. The blots were incubated with a primary antibody [phospho-SAPK/JNK(Thr¹⁸³/Tyr¹⁸⁵),phospho-SEK1/MKK4 (Thr²²³), phospho-p38(Thr¹⁸⁰/Tyr¹⁸²),phospho-c-Jun (Ser⁷³), or phospho-ATF-2(Thr⁷¹; all from New England Biolabs)] in TTBS with 5% nonfat dry milk for 2 h at room temperature, washed three times with TTBS, and incubated with horseradish peroxidase-conjugated donkey antirabbit IgG (Amersham). The signals were visualized by an enhanced chemiluminescence system (Pierce). The blots were deprobed in a deprobing solution [62.5 mm Tris (pH 6.5), 10%SDS, and 100 mm β-mercaptoethanol] for 10 min at 50°C. The deprobed blots were blocked, reprobed with a primary antibody[SAPK/JNK, SEK1/MKK4, p38, c-Jun, or ATF-2 (all from New England Biolabs)], and the signals were visualized as described above.

Immunohistochemical detection of apoptosis was carried out using an In Situ Cell Death Detection Kit (Boehringer Mannheim)following the procedures provided by the manufacturer. Briefly,paraffin-embedded tissue sections were dewaxed and rehydrated according to standard protocols (37) and then treated with 20μg/ml proteinase K in 10 mm Tris (pH 7.4) for 15 min at 30°C. After rinsing the slides with PBS, the sections were incubated with a blocking solution (0.3%H₂O₂ in methanol) for 1 h at room temperature. After rinsing with PBS, the sections were incubated in a permeabilization solution (0.1% Triton X-100 in 0.1%sodium citrate) for 2 min at 4°C. After rinsing twice with PBS, the sections were incubated with 50 μl of terminal deoxynucleotidyl transferase-mediated nick end labeling reaction mixture for 1 h at 37°C, 50 μl of Converter-POD for 30 min at 37°C, and 100μl of 3,3′-diaminobenzidine substrate solution for 2 min at room temperature, and the sections were rinsed three times with PBS between each step. The slide was analyzed under a light microscope. The number of apoptotic cells was scored from pictures of six different slides made from two independent experiments with three animals per time point, and the average numbers of apoptotic cells/mm² were plotted against time (in hours)after MMS treatment.

To examine whether JNK is activated by MMS in vivo, we treated the rats with increasing concentrations of MMS for 1 h and monitored the degree of JNK activation in the liver and brain. Because dual phosphorylation of JNK at Thr¹⁸³ and Tyr¹⁸⁵ is essential for kinase activity, we used phosphorylation at these sites as a marker for JNK activity. We confirmed that the phosphorylation reflected the actual activity of JNK with an in vitro kinase assay using glutathione S-transferase-c-Jun fusion protein as a substrate(38).

Liver and brain extracts from control rats or rats treated with varying concentrations of MMS were analyzed on an immunoblot with an anti-phospho-specific JNK(Thr¹⁸³/Tyr¹⁸⁵) antibody that reacts with the dually phosphorylated isoforms of both JNK p46 and JNK p54. Neither JNK p46 nor JNK p54 phosphorylation was detected in the control liver (Fig. 1,A). However, JNK p54 phosphorylation was detected in the liver 1 h after an injection of 1.2 mmol/kg MMS and increased as the concentration increased (Fig. 1 A). A dose of 1.5 mmol/kg MMS corresponds to the dose levels used to induce tumors in the brain after a single injection into an adult rat (27). At a concentration of 1.8 mmol/kg MMS, both JNK isoforms were strongly phosphorylated in the liver.

The lack of JNK phosphorylation in the control liver was not due to the absence of JNK expression because JNK p46 and JNK p54 were readily detectable by Western analysis with an antibody recognizing both isoforms (Fig. 1,A). Furthermore, the Western analysis confirmed that similar levels of JNKs were present in all samples. In the liver, JNK p54 was more abundant than JNK p46, but both isoforms were equally phosphorylated on MMS treatment (Fig. 1, A and B).

In the control brain, JNK p46 was shown to be highly phosphorylated,whereas JNK p54 phosphorylation was barely detectable (Fig. 1),although a greater amount of JNK p54 isoform was present. This indicates that JNK p46 is constitutively activated in the rat brain. In contrast to the liver, however, JNKs were not further phosphorylated in the brain after MMS treatment with up to 2.4 mmol/kg MMS (Fig. 1 A). This is not consistent with previous reports, in which JNKs were strongly activated by MMS treatment in many different in vitro cultured cell lines (6, 7), suggesting a difference in response to MMS between in vitro cultured cells and in vivo tissues. It is interesting to note that the levels of JNK p46 appeared to be lower in the liver than in the brain, whereas JNK p46 in the liver and control brain was phosphorylated to a similar degree after treatment with MMS.

We next determined the time course of the JNK activation for various lengths of time ranging from 30 min to 3 h, using 1.5 mmol/kg of MMS as a dose. In the liver, the phosphorylation of both JNK isoforms reached a maximum at 1 h after MMS treatment, where the increase in phosphorylation was more than 20-fold (Fig. 1,B). The activation of JNK was transient and returned to basal level by 5 h(data not shown). In the brain, however, no significant changes in JNK phosphorylation were observed for up to 5 h (Fig. 1 B;data not shown). Our results on the activation of JNK in the liver and brain demonstrate very distinct tissue-specific patterns in rats exposed to MMS. Furthermore, the pattern of JNK isoform expression and basal activity differs in the liver and brain.

The results presented above indicated that MMS activates the JNK in the liver but not in the brain of rats. To determine whether MMS selectively activated specific MAPK cascades, p38 activity was measured by Western analysis with an anti-phospho-specific p38 antibody. In a parallel experiment, p38 protein levels were measured by Western blot analysis with an anti-p38 antibody to ensure that similar levels of p38 proteins were present in all samples.

The liver expressed about twice the amount of p38 seen in the brain,and the basal level of p38 phosphorylation was relatively high in the liver compared with that in the brain (Fig. 2). After MMS treatment, p38 phosphorylation increased in both the liver and brain in a dose- and time-dependent manner (Fig. 2, A and B). There was about a 3-fold (3.2 ± 0.4; n = 6) increase in the liver and a 5–6-fold(5.4 ± 0.8; n = 6) increase in the brain, with a maximum level reached at 2 h after treatment(Fig. 2,B). Unlike the activation of JNK p46 in the liver,the level of phosphorylated p38 in the brain at its maximum was far less than the basal level in the liver. Our results indicate that MMS activates p38 in both the liver and brain, whereas it activates JNKs only in the liver (Figs. 1,A and 2 A). As shown in vitro (39), MMS did not appear to activate the extracellular signal-regulated protein kinases in either the liver or the brain (data not shown).

Biochemical studies have demonstrated that SEK1/MKK4 can phosphorylate and activate both JNK and p38 MAPKs (17, 40). However, the homozygous SEK1/MKK4^−/−embryonic stem cells were found to be defective in activating JNK but not in activating p38, whereas the homozygous SEK1/MKK4^−/− fibroblast cells were defective in activating both JNK and p38. This implies that there are cell type-specific roles of SEK1/MKK4 in vivo(20).

We examined the activation of SEK1/MKK4 in the liver and brain after MMS injection by Western blot analysis with an anti-phospho-specific SEK1/MKK4 antibody. MMS-induced phosphorylation at Thr²²³ of SEK1/MKK4, a marker for SEK1/MKK4 activation, was monitored in the liver and brain of rats after treatment with 1.5 mmol/kg MMS. The amounts of SEK1/MKK4 were similar in the rat liver and brain (Fig. 3). In the liver, the phosphorylation of SEK1/MKK4 was observed at 30 min and reached a maximum at 1 h after MMS treatment (Fig. 3),preceding the activation of JNK and p38 (Figs. 1,B and 2,B). The basal phosphorylation of SEK1/MKK4 in the brain was relatively high compared with that in the liver, but MMS treatment did not induce further phosphorylation of SEK1/MKK4 in the brain (Fig. 3).

Earlier studies found that treatment of mammalian cells with alkylating agents resulted in the induction of the immediate early gene c-jun. To examine whether c-jun is induced in response to MMS in vivo, c-Jun expression in the liver and brain was measured by Western blot with an anti-c-Jun antibody after MMS treatment. In confirmation of the earlier findings, the control liver expressed a nondetectable level of the c-Jun protein (Fig. 4,B). The c-Jun protein was detected at 30 min after MMS treatment in the liver, and the level increased in a time-dependent manner, reaching a maximum at 3 h (Fig. 4,B). However,in the brain, an increase in the c-Jun level was barely observed at 3 h and thereafter (Fig. 4 B), although the basal level of expression was higher than that in the liver.

The induction of c-jun in response to DNA-damaging agents,as well as other stress-related stimuli, is mediated by c-Jun-ATF-2 heterodimers, which are stimulated by phosphorylation (11, 41, 42, 43, 44, 45, 46). Because JNK phosphorylates c-Jun and ATF-2 and p38 phosphorylates ATF-2 on activation in vitro, we examined whether the MMS-induced activation of JNK/p38 in the liver or the MMS-induced activation of p38 in the brain leads to phosphorylation of these transcription factors using Western blot analysis with anti-phospho-specific ATF-2 or anti-phospho-specific-c-Jun antibodies. One h after treatment with 1.5 mmol/kg MMS, ATF-2 was found to be heavily phosphorylated in the liver of rats (Fig. 4,A),producing multiple shifted bands (Fig. 4,A). In the brain,however, no phosphorylation of ATF-2 was detected. Furthermore, the phosphorylation of the c-Jun protein was found to be increased in response to MMS in the liver, but the phosphorylation of c-Jun protein was barely detected in the brain up to 5 h after MMS treatment(Fig. 4 B; data not shown).

A number of studies have demonstrated that stress kinase activation is implicated in the initiation of apoptotic cell death (21, 22, 23, 24, 47). To examine whether apoptotic cell death is induced in the liver and brain after a MMS injection, we determined the apoptosis-induced DNA fragmentation by in situ labeling. In the control liver and brain, very few apoptotic cells were observed,indicating that a very low level of apoptotic cell death occurs in normal tissues (Fig. 5, A and E). However, cells bearing fragmented DNA were detected in the liver within 1 h after MMS injection, and the number of cells bearing fragmented DNA continued to increase for up to 5 h after MMS injection, at which point the number of apoptotic cells exceeded 50 cells/mm² (54 ± 4.2 cells/mm²; n = 6;Fig. 5, C and I).

This result indicates that a rapid increase in apoptotic cell death occurs in the liver after MMS treatment. However, in the brain, no detectable changes in the number of cells undergoing apoptosis were observed up to 5 h after MMS treatment (Fig. 5, G and I). Differential induction of apoptosis in the liver and brain after MMS treatment was also evident from nuclear morphology, as determined by the microscopic examination of H&E staining (Fig. 5, B, D, F, and H). Combined with the pattern of stress kinase activation described above, these results demonstrate that there is a strong correlation between the time course in the initiation of apoptotic cell death and the activation of JNKs in the liver after MMS injection.

Alkylating agents are the most potent and abundant chemical DNA-damaging agents found in our environment (25). They generate many forms of DNA damage and, as a result, are toxic,mutagenic, and carcinogenic. Intriguingly, alkylating agents induce the development of tumors in an organ-specific manner. In the past, many attempts have been made to find a correlation between the extent of the primary reaction of alkylating agents with nucleic acids in different tissues and the location of tumors induced by the respective compounds. However, no consistent differences were found in the initial extent of alkylation between the target and nontarget tissues (34, 48, 49, 50). It is therefore of substantial interest to define the intracellular signaling pathways that mediate distinct cellular responses in tissues of a whole animal exposed to alkylating agents.

Previous studies have demonstrated that alkylating agents, including MMS, coordinately activate the SAPKs, JNK and p38 in many cultured cell lines of varying origins (6, 7, 41), but the in vivo responses of cells to these agents have yet to be defined. In the present study, we have addressed the potential involvement of the JNK and p38 stress signaling pathways in eliciting tissue-specific responses induced by MMS in vivo. We reasoned that the differences in the capacity of different tissues to activate signal transduction pathways in response to alkylating agents may correlate with tissue-specific carcinogenicity.

MMS activated both JNK p46/p54 and p38 in the liver but activated only p38 in the brain. In the brain, MMS caused a 5–6-fold (5.4 ± 0.8; n = 6) increase in p38 activation (Fig. 2) but failed to activate JNK and induce the expression of c-Jun protein (Figs. 1 and 4). The failure of MMS to activate JNK in the brain at concentrations more than sufficient to activate p38 indicates that p38 and JNK may not function in the same signaling pathway in the brain. More importantly, this suggests that JNK activation may not be a universal response to MMS-induced damage signals. After a MMS injection, the activation of JNK and p38 in the liver and the activation of p38 in the brain occurred at a clinically relevant drug concentration that was used to induce tumors in the brain after a single injection into an adult rat (27). The differential activation of the SAPKs in the liver and brain may provide a new possible explanation for different carcinogenic effects of MMS in these tissues.

Improper JNK activation is expected to affect the activities of its substrates, ATF-2, Elk-1, c-Jun, and p53, which were shown to participate in the DNA damage response in various pathways, including changes in cell cycle distribution, growth arrest, and the rate of DNA synthesis and repair (51, 52, 53, 54). Because recent studies indicate that the activation of JNK is necessary for the induction of apoptosis in response to diverse agents (21, 22, 23, 24), we examined MMS-induced apoptosis in the liver and brain. Consistent with the activation of JNK, a significant increase in apoptotic cell death was detected in the liver but not in the brain after MMS injection(Fig. 5). Moreover, a strong correlation was observed among the ability of a tissue to activate the JNK pathway, the induction of apoptosis,and the susceptibility of a tissue to the carcinogenic effect induced by MMS. It is of particular interest to find that the impaired JNK activation observed in the brain is limited to MMS or other alkylating agents, especially brain tumor chemotherapeutic agents.

Previous studies have demonstrated that DNA-damaging agents, which damage DNA by diverse mechanisms including exposure to UV light,ionizing radiation, 1-β-d-arabinofuranosylcytosine, cis-platinum, mitomycin C, chemotherapeutic drugs, hydrogen peroxide, and alkylating agents, activate JNK and p38(1, 2, 3, 4, 5, 6, 7). However, the initiating signals responsible for such responses are unclear. The findings that structurally distinct DNA-damaging agents are capable of activating SAPKs suggest DNA damage as an initial signal in this cascade. Nonetheless, because the amount of initial DNA damage induced by MMS was shown to be similar in the liver and brain of rats (26, 33, 34, 35, 36), MMS-induced DNA damage per se might not act as a trigger for the differential responses observed in this study. Alternatively, the DNA damage signal that activates JNK is not properly transduced in the brain. The fact that the ultimate biological consequence induced by MMS differs markedly in these rat tissues suggests that cellular context may determine the distinct biological response. Recent findings that the redox state of the cell is critical for JNK and p38 activation supports this notion (39).

Although it has been established that MMS and other alkylating agents coordinately activate JNK and p38 and induce the expression of c-jun mRNA in many different cell lines (6, 7, 55), the upstream mechanisms responsible for these responses are unclear. In the liver, the activation of JNK is preceded by the activation of SEK1/MKK4, suggesting that MMS-induced activation of JNK is mediated, at least in part, through SEK1/MKK4 (Figs. 1,B and 3). Recent studies have shown that homozygous SEK1/MKK4^−/− cells were defective in JNK activation but not in activation of p38 MAPK,suggesting that SEK1/MKK4 may function as a specific activator of JNK in vivo (20). The present study identifies the SEK, JNK, and c-Jun pathways as members of the pathways activated by MMS in the liver, implying the possibility that the impairment of this pathway may result in abnormal responses to MMS, which may then contribute to a greater sensitivity to MMS-induced malignant transformation. However, SEK1/MKK4 is not activated in the brain,although it is known to phosphorylate and activate both JNK and p38 in vitro (Fig. 3). Therefore, activation of p38 in response to MMS in the brain may be mediated by other p38 activators such as MKK3 and MKK6. Additional studies are required to better understand the role of MKK3/MMK6 in the MMS-induced p38 MAPK signal transduction pathway in the brain.

Of particular interest in this study is the high basal activity of the JNK p46 in the brain and of p38 in the liver, suggesting some roles for these protein kinases in the basic homeostatic mechanisms in rats. JNKs consist of three distinct genes, which, via differential splicing,yield at least 10 isoforms of JNK protein products of M_r 46,000 and M_r 54,000. It has been shown in vitro that JNK isoforms have different affinities for their substrates and are therefore likely to have different activities with regard to substrate phosphorylation (47). A recent study demonstrated stronger activation of the JNK p46 isoform than of JNK p54 after the cytokine stimulation of chondrocytes, and it was suggested that the JNK p46 isoform may play a more important role in chondrocytes (47). Differential basal activity of a JNK isoform in the brain has not been reported previously. Our data showing the selective activation of JNK isoforms in the control brain suggest a specific role for JNK p46 in brain function that requires further investigation. In the brain of the control group animals, the constitutively phosphorylated JNK p46 may preferentially phosphorylate substrate(s) other than c-Jun, a major substrate of JNK, because it is not phosphorylated (Fig. 4). Alternatively, phosphorylated JNK p46 may be localized in the cytoplasm and thus unavailable for the activation of nuclear targets, such as c-Jun and ATF-2. Consistent with the fact that the basal activity of JNK p46 is high in the brain, the basal activity of SEK1 is also high in the brain, suggesting that SEK1 may be a physiological activator of JNK p46 in the brain as well.

Multiple isoforms of p38 have been identified, but the studies performed to measure p38 MAPK activity do not distinguish between the multiple isoforms of the p38 MAPK identified to date. It is therefore possible that the individual isoforms of p38 MAPK are activated by specific conditions in the liver and brain, leading to distinct patterns of activation in response to MMS. Because p38 did not phosphorylate ATF-2, a known target of p38 in the brain, p38 activated in the liver may be a different isoform than the p38 activated in the brain (Fig. 4 A).

The carcinogenic potency of DNA-damaging agents depends on the type and persistence of DNA adducts, whereas the latter in turn depends on the cellular capacity for enzymatic DNA repair. MMS has a relatively high Swain-Scott factor (s = 0.86), indicating that it reacts with the strong nucleophilic nitrogen atoms in DNA via a S_N2 reaction mechanism (25). This results in the methylation of the nitrogen atoms at position 7 of guanine (7mG) and at position 3 of adenine (3mA), which account for approximately 95% of total alkylations produced by MMS, and therefore these lesions may play an important role in the observed cellular responses.

Alkylation at the much less nucleophilic oxygen atoms occurs at very low levels, and the mutagenic O⁶mG adducts account for less than 0.3% of the total methylation by MMS(55). O⁶mG is rapidly repaired by MGMT, whereas 7mG and 3mA are repaired by base excision repair, which is initiated by MAG (reviewed in Ref. 56). The repair activities of MGMT and MAG vary a great deal in different tissues and have been correlated with tumor induction in rats(57). MAG activities were shown not to be any higher in the liver than in the brain (58, 59). Therefore, the differential carcinogenicity of MMS in rat liver and brain tissues could not be explained solely by the differential cellular capacity of the enzymatic DNA repair responsible for repairing major types of DNA adducts produced by MMS in these rat tissue.

However, whereas MGMT activity was highest in the liver, it was undetectable in the brain (57, 60). Therefore, although MMS produces minute amounts of O⁶mG accounting for less than 0.3% of the total methylation product, a low level of MGMT activity in the brain may contribute to the development of tumors in the brain by MMS. Lack of MMS-induced apoptosis in the brain will augment this process by allowing the survival of DNA-damaged cells, ultimately leading to the accumulation of genetic defects associated with tumor progression. Moreover, the finding that expression of some of repair enzymes in the base excision repair pathway is induced by c-Jun and ATF-2 proteins (61, 62, 63),suggests that the impairment of the JNK signaling pathway to induce and activate these transcription factors in response to MMS could account for the sensitivity of the brain to MMS-induced carcinogenesis.

In summary, we demonstrated very distinct tissue-specific patterns in the activation of the SAPKs in rats treated with MMS. The upstream and the downstream pathways of the JNK signaling are activated only in the liver (the nontarget tissue) and not in the brain (the target tissue)in response to MMS. Consistently, apoptotic cell death is induced in the liver but not in the brain after MMS treatment. In contrast, p38 is activated in both the liver and brain after MMS treatment,demonstrating that p38 and JNK are not always coordinately regulated in vivo. Taken together, our data strongly suggest that the capacity of a tissue to activate the JNK pathway may be involved in the protective response of cells to prevent the development of tumor, presumably by inducing apoptosis to eliminate alkylation-damaged cells after exposure of an intact animal to alkylating agents. These findings may provide an explanation for the differential carcinogenic effects exerted by MMS in different rat tissues.