(Received for publication, October 16, 1996, and in revised form, February 3, 1997)
From the Department of Medicine, University of Colorado Health Sciences Center, Denver, Colorado 80262
Exposure of cultured small cell lung cancer (SCLC) cells to UV radiation induces apoptosis. We observed that the UV sensitivity of a panel of SCLC lines and the activation of c-Jun NH2-terminal kinases (JNKs) by UV in the individual SCLC lines, assessed by binding and phosphorylation of glutathione S-transferase (GST)-c-Jun fusion proteins, ranged widely. In fact, increased JNK activity in this assay was closely correlated with decreased sensitivity to apoptosis following UV irradiation. Increased JNK activity was also detected in anti-JNK1 immune complexes collected from UV-irradiated SCLC cells, although the level of activity was similar among the various SCLC lines and correlated poorly with UV sensitivity. Immunoblot analysis of JNK polypeptides that bound to GST-c-Jun revealed at least two JNK polypeptides, one of which appeared only in extracts from UV-irradiated SCLC. To test the role of JNKs in UV-induced apoptosis, nonphosphorylatable mutants of JNK1 or JNK2 in which the phosphorylation site Thr-Pro-Tyr is changed to Ala-Pro-Phe (JNK-APF) and are predicted to behave as competitive inhibitors were stably expressed in SCLC. Expression of JNK1-APF or JNK2-APF significantly reduced UV-stimulated JNK activity. However, JNK1-APF markedly increased the resistance of the cells to UV-induced apoptosis, while JNK2-APF did not influence SCLC sensitivity to UV. The findings suggest that UV-stimulated JNK1 activation promotes UV-induced SCLC apoptosis, while a JNK isoform that is variably activated among the SCLC lines may signal a UV-protective response. We hypothesize that integration of distinct JNK activities dictates the relative responsiveness of SCLC to UV and ionizing radiation.
Small cell lung cancer (SCLC)1 is a rapidly growing human cancer that accounts for approximately 25% of primary lung cancers. Compared with the majority of non-small cell lung cancers, SCLC exhibits neuroendocrine features (1) and is characteristically more sensitive to ionizing radiation and cytotoxic drugs (2-4). Emerging evidence suggests that the sensitivity of tumor cells such as SCLC to ionizing radiation and cytotoxic drugs is not related to their rapid rate of cell division, but to the ability of these cell stresses to induce programmed cell death, or apoptosis, relative to growth arrest of untransformed cells (5). However, as with many cancers, selection for SCLC tumor cells that are relatively resistant to the cytotoxic insult frequently occurs and is followed by relapse, translating to a cure rate of 10% or less. Cell lines derived from relapsed SCLC tumors often exhibit the variant phenotype (1, 6-8) associated with loss of neuroendocrine differentiation, gene amplification of the Myc family members, and increased growth rates in vitro. This general problem emphasizes the importance of defining both the mechanism of radiation and cytotoxic drug-induced tumor cell death as well as the molecular mechanisms accounting for cellular resistance.
Recent advances in the understanding of cellular responses to cytotoxic
stresses such as ionizing and UV radiation, tumor necrosis factor-,
heat shock, and toxic drugs have defined the regulation of a protein
kinase signaling cascade that targets specific transcription factors
including c-Jun, c-Fos, ATF-2, and NF
B (9-15). UV irradiation is
the prototypical activator of this highly conserved cellular pathway
referred to as the UV response (16). Transactivation of genes by c-Jun
and ATF-2 is achieved by phosphorylation by specific mitogen-activated
protein (MAP) kinase family members, the stress-activated protein
kinases (SAPKs), or c-Jun NH2-terminal kinases (JNKs)
(12-14, 17-19). Molecular cloning has defined three genes encoding
rodent SAPK or human JNK enzymes, and alternative splicing of the
mRNAs yields as many as 12 distinct enzymes (19-22). In a manner
parallel to that previously defined for the related extracellular
signal-regulated kinases (ERKs) or p42/44 MAP kinases (23), the JNKs
are activated following their phosphorylation on threonine and tyrosine
by dual specificity MAP kinase kinases (24). While DNA damage is often
a consequence of exposure to agents that stimulate the UV response, the
regulation of JNKs and NF
B proceeds normally in enucleated cells
(11), indicating that DNA damage is not the initiating signal. In fact, the Ras and Rho families of low molecular weight GTP-binding proteins, which integrate signaling via a host of extracellular signals, play a
role in mediating the UV response in mammalian cells and yeast (11,
16). This finding further supports the notion that the UV-induced
transcriptional activation through JNK protein kinase cascades is
independent of DNA damage and is probably initiated at the plasma
membrane.
Studies suggest that the UV response serves a protective role in lower organisms such as yeasts (16). Yet, the functional significance of the UV response in mammalian cells is unclear. Ras-transformed mammalian cells are often more resistant to UV and ionizing radiation (25, 26). However, the profound and rapid activation of JNKs by many cytotoxic stimuli suggests that the JNKs may, in fact, initiate cell death in mammalian cells rather than a protective response. Support for this possibility is provided by findings that MAP kinase/ERK kinase kinase, which has recently been defined as a proximal activator of the JNK pathway (27), can induce apoptosis when expressed in fibroblast cell lines (28), and cellular expression of molecular inhibitors of the JNK pathway reduces the induction of apoptosis (29, 30). To investigate the role of the JNK pathway in SCLC, we examined the influence of UV irradiation on induction of cell death and signal transduction through JNKs in a panel of cultured SCLC lines.
Materials
Recombinant GST-c-Jun(1-79) and ATF2-NT(1-254) were expressed in bacteria and purified using glutathione agarose (Sigma) and Ni+-nitrilotriacetic acid-agarose (Qiagen, Studio City, CA), respectively, as described previously (17, 31). Sera and powdered growth media were from Life Technologies, Inc. The rabbit polyclonal antibodies to JNK-1 (C-17), p38 MAP kinase (C-20), ERK1 (C-16), and ERK2 (C-14) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The mouse monoclonal antibody directed against the influenza hemagglutinin (HA) epitope (12CA5), dUTP-fluorescein, and terminal deoxytransferase were from Boehringer Mannheim. Protein A- and protein G-Sepharose were purchased from Pharmacia Biotech Inc.
Cell Culture and Retrovirus-mediated Gene Transfection
NCI-H345 and H187 cells were cultured in HITES medium (RPMI 1640 containing 10 nM hydrocortisone, 5 µg/ml insulin, 10 µg/ml transferrin, 10 nM 17-estradiol, 30 nM sodium selenite, and 0.1% bovine serum albumin). SCLC
lines N417, SHP-77, and H69 were routinely cultured in RPMI 1640 containing 10% fetal bovine serum. Cell lines H69, H187, and H345
exhibit features of classic SCLC (1), whereas N417 is a variant SCLC
line (1, 6), and SHP-77 exhibits features of both classic and variant
SCLC. Swiss 3T3 fibroblasts were cultured in Dulbecco's modified
Eagle's medium containing 10% fetal bovine serum.
For experiments examining the influence of UV irradiation on cell growth and signaling, suspensions (2 ml) of SCLC cells in HITES medium were placed in 6-cm plastic tissue culture plates and irradiated uncovered with a Mineralight R-52G (UVP, Inc., San Gabriel, CA) short wave UV lamp. The radiance of the lamp at 254 ± 20 nm at a distance of 2 inches was 3.2 J/m2/s. Following irradiation, the cells were incubated at 37 °C for the indicated times in a CO2 incubator. For determination of cell counts, SCLC cell suspensions were mixed with an equal volume of 0.4% trypan blue in 0.85% NaCl, and cells excluding dye were counted with a hemocytometer. Swiss 3T3 cells were similarly counted after trypsinization.
The cDNAs (18, 32, 33) encoding HA-JNK1-APF and HA-JNK2-APF were
ligated between the HindIII and HpaI sites of
pLNCX (34) and packaged into replication-defective retrovirus using 293T cells and the retrovirus component-expression plasmids
SV--A-MLV and
SV-
-env
-MLV as described previously (35,
36). Conditioned growth medium containing secreted retrovirus was
collected, supplemented with 8 µg/ml polybrene, filtered through a
0.45-µm filter, and incubated with the indicated SCLC lines for
24 h. SCLC cells expressing the retrovirus-encoded cDNAs were
selected for growth in medium containing G418, and pooled cultures of
G418-resistant cells were used for the described studies.
Determination of Apoptosis
Terminal Deoxytransferase LabelingFollowing UV irradiation (96 J/m2) as described above, the SCLC cultures were returned to the CO2 incubator for 24 h. Aliquots (1.5 ml) of control and irradiated cultures of SCLC were centrifuged (500 × g, 5 min) and washed once in phosphate-buffered saline (PBS). The cells were fixed for 15 min in PBS containing 1.5% paraformaldehyde and 1.5% sucrose and then washed with 1 ml of PBS and then with 250 µl of terminal deoxytransferase buffer (25 mM Tris-Cl (pH 6.6 at 25 °C), 200 mM potassium cacodylate, 0.25 mg/ml bovine serum albumin). The cell pellets were finally suspended in 50 µl of terminal deoxytransferase buffer containing 1.5 mM CoCl2, 3 µM fluorescein-dUTP, and 5 units of terminal deoxytransferase and incubated for 2 h at 37 °C. The reactions were terminated by the addition of 20 mM EDTA and the labeled cells were collected by centrifugation, washed once with PBS, and resuspended in 50 µl of PBS. The cells were immobilized onto coverslips precoated with Cell-Tak (Collaborative Biomedical Products, Bedford, MA) and viewed at × 60 magnification with a fluorescence microscope.
Propidium Iodide StainingTwenty-four and forty-eight hours later, aliquots of control and UV-irradiated cultures containing ~200,000 cells were centrifuged (5 min, 1000 × g), and the cell pellets were gently suspended in 100 µl of PBS. The cells were fixed by addition of 400 µl of methanol and stained for 15 min with propidium iodide (50 µg/ml). The cells were again collected by centrifugation, suspended in PBS containing 50% glycerol, and mounted under glass coverslips. Slides were viewed and photographed on a fluorescence microscope.
Assay of JNK Activity
GST-c-Jun(1-79) Binding/Protein Kinase AssaySCLC cells
were collected by centrifugation and lysed for 30 min at 4 °C in 0.5 ml of 25 mM HEPES (pH 7.7), 20 mM
-glycerophosphate, 0.1 mM sodium vanadate, 0.1% Triton
X-100, 0.3 M NaCl, 1.5 mM MgCl2,
0.2 mM EDTA, 0.5 mM dithiothreitol, 2 µg/ml
leupeptin, and 4 µg/ml aprotinin as described elsewhere (17).
Following a 5-min microcentrifugation (10,000 × g),
aliquots of the extracts containing 200 µg of protein were incubated
for 2 h at 4 °C with GST-c-Jun(1-79) immobilized to
glutathione-agarose (10 µl of packed beads per sample containing
~10 µg of protein). The GST-c-Jun(1-79)-agarose complexes were
washed three times by repetitive centrifugation in 20 mM
HEPES (pH 7.7), 50 mM NaCl, 2.5 mM
MgCl2, 0.1 mM EDTA, 0.05% Triton X-100, and
then incubated for 20 min at 30 °C in 40 µl of 50 mM
-glycerophosphate (pH 7.6), 0.1 mM sodium vanadate, 10 mM MgCl2, 20 µM
[
-32P]ATP (25,000 cpm/pmol). The reactions were
terminated with 10 µl of SDS-PAGE sample buffer, boiled, and
submitted to 12% SDS-PAGE. The GST-c-Jun(1-79) polypeptides were
identified in Coomassie-stained gels, excised, and counted in a
scintillation counter.
SCLC cultures
were collected by centrifugation and lysed in MAP kinase lysis buffer
(0.5% Triton X-100, 50 mM -glycerophosphate (pH 7.2),
0.1 mM sodium vanadate, 2 mM MgCl2,
1 mM EGTA, 1 mM dithiothreitol, 2 µg/ml
leupeptin, and 4 µg/ml aprotinin). Clarified extracts containing 200 µg of protein were incubated (4 °C, 2 h) with 1 µg of
anti-JNK1, 1 µg of anti-p38 MAP kinase, or 1 µg each of anti-ERK1
and anti-ERK2 and 10 µl of packed protein A-Sepharose in a total
volume of 0.5 ml. The immune complexes were washed three times in lysis
buffer and then suspended in 40 µl of 50 mM
-glycerophosphate (pH 7.2), 0.1 mM sodium vanadate, 10 mM MgCl2, 100 µM
[
-32P]ATP (5000 cpm/pmol), 25 µg/ml IP-20
(TTYADFIASGRTGRRNAIHD), 1 mM EGTA, and either 100 µg/ml
ATF2-NT, an excellent substrate for JNKs and p38 MAP kinase (12), or
200 µM EGFR662-681 peptide for analysis of
ERK activity (37). Following a 20-min incubation at 30 °C, the ATF2
kinase reactions were terminated with 10 µl of SDS sample buffer and
submitted to SDS-PAGE. The ATF2 polypeptides were excised from the
Coomassie-stained and dried gels, and incorporated radioactivity was
determined in a scintillation counter. The EGFR662-681
peptide kinase reactions were terminated with 10 µl of 25%
trichloroacetic acid, and phosphorylated peptide was quantified by P81
binding as described previously (37).
Immunoblot Analyses
Samples were resolved by 10% SDS-PAGE and transferred to nitrocellulose. The filters were blocked in Tris-buffered saline (10 mM Tris-Cl (pH 7.4), 140 mM NaCl) containing 0.1% Tween-20 (TTBS) and 3% nonfat dry milk and then incubated with blocking solution containing the indicated antibodies at 1 µg/ml for 12-16 h. The filters were extensively washed in TTBS, and bound antibodies were visualized with horseradish peroxidase-coupled protein A or horseradish peroxidase-coupled sheep anti-mouse antibodies and ECL (Amersham Corp.) according to the manufacturer's directions.
The radiosensitivity of SCLC is well established, and
the survival of a variety of SCLC lines following ionizing radiation has been previously compared (2, 4). The growth and viability of
control SCLC line H345 or cultures irradiated with a UV-C (200 to 290 nM) source at intensities of 32 or 192 J/m2 is
shown in Fig. 1A. After 4 days in culture,
the density of control H345 cultures increased 3-fold, while cultures
of H345 cells UV-irradiated at 32 J/m2 failed to grow. In
fact, cell number decreased slightly, suggesting limited cell death. UV
irradiation of H345 cells at 192 J/m2 resulted, after a lag
of approximately 24 h, in a progressive and marked decrease in
cell number below the initial cell density, such that less than 5% of
the irradiated cells were viable as assessed by trypan blue exclusion
after 4 days of culture.
Examination of the dose dependence of UV-induced cell killing of five independent SCLC lines and Swiss 3T3 fibroblasts was performed (Fig. 1B). UV irradiation at intensities ranging from 6 to 192 J/m2 resulted in a dose-dependent decrease in viable cells measured 4 days later compared with untreated controls in all the cell lines tested. Notably, the SCLC lines were more sensitive to UV-induced cell killing than Swiss 3T3 fibroblasts at UV intensities of 16-64 J/m2, consistent with the known sensitivity of SCLC to ionizing radiation (2-4). Among the SCLC lines, H187 was the most sensitive to UV-induced killing such that less than 50% of the cells survived the lowest UV intensity tested, 6 J/m2 (Fig. 1B). H69 was somewhat more resistant and H345 cells were significantly more resistant to UV-C-induced killing than were H187 cells, but were still highly sensitive compared with cell lines SHP-77 and N417. H187, H69, and H345 cells exhibit features of classic SCLC, while SHP-77 and N417 cells express elevated c-Myc and/or lack of detectable neuropeptide autocrine loops, features indicative of the variant SCLC phenotype (1). Significantly, H69, H187, H345, and N417 have been previously examined for sensitivity to ionizing radiation, and a similar order of sensitivity to ionizing radiation was observed (2, 4).
The time-dependent decrease in viable cell number following
UV irradiation was suggestive of apoptosis, which is defined as cell
death that is physiologically regulated and is characterized morphologically by DNA breakage and nuclear condensation (38). Induction of DNA breaks following UV was assessed by terminal deoxytransferase-mediated labeling of nuclei with fluorescein-dUTP (Fig. 2). The terminal deoxytransferase assay resulted
in the labeling of ~1-5% of a control untreated suspension of H345
cells, while ~50% of H345 cells were labeled when assayed 24 h
after UV irradiation (96 J/m2). Pronounced condensation of
chromatin, another feature of apoptosis, was assessed by microscopic
inspection of UV-irradiated SCLC lines H187 and N417 that had been
fixed and then stained with the DNA dye, propidium iodide. The results
from these experiments revealed marked nuclear condensation that was
apparent 24 h after UV irradiation (Fig. 3).
Forty-eight hours after exposure to UV radiation (96 J/m2),
greater than 95% of the H187 cell nuclei were highly condensed relative to the nuclei observed in untreated control cells. Compared with H187 cells, fewer N417 cells (~50%) exhibited condensed nuclei 2 days after UV irradiation, consistent with the decreased sensitivity of N417 cells to UV-induced cell death shown in Fig. 1B.
Thus, the data indicate that UV irradiation induces cell death of five independent SCLC lines with variable efficacy and that the response occurs by apoptosis.
Detection of Multiple UV-regulated JNK Activities in SCLC
An understanding of the molecular regulation of the JNK/SAPKs initiated by cell stresses such as UV radiation has recently emerged. To examine the regulation of JNKs by UV in the panel of SCLC lines, an assay was employed in which the JNKs are adsorbed to a glutathione agarose-immobilized GST fusion protein encoding the NH2-terminal 79 amino acids of the transcription factor c-Jun and then assayed for phosphotransferase activity toward the resident GST-c-Jun(1-79) polypeptide (17). Preliminary experiments in which H345 cells were UV-irradiated at 96 J/m2 indicated that protein kinase activity collected with GST-c-Jun(1-79) increased rapidly following irradiation and was maximal by 10- 30 min of incubation following the UV exposure. Furthermore, this JNK activity was sustained for at least 2 h, the longest time point examined.
The UV dose response for activation of the GST-c-Jun-binding JNKs was
determined in the five SCLC lines (Fig. 4A).
SCLC lines H187 and H69, which are the most sensitive to UV-induced
cell death (Fig. 1B), exhibited a modest 2-4-fold
activation of protein kinase activity at a UV dose of 192 J/m2. In comparison, N417 and SHP-77, which are the most
UV-resistant of the SCLC lines, exhibited marked UV-stimulated
GST-c-Jun kinase activity. Note that a 5-10-fold activation was
observed with N417 at a UV intensity of only 16 J/m2.
Finally, H345 cells, which are of intermediate sensitivity to UV
irradiation (Fig. 1B), exhibited an intermediate
UV-stimulated JNK activation. When the UV-induced GST-c-Jun kinase
activation observed at 64 J/m2 (Fig. 4A) was
compared with the fraction of cells surviving the same UV dose (Fig.
1B), an excellent correlation was observed (r2 = 0.93, slope very significantly
(p < 0.01) different from zero). Thus, the data in
Figs. 1B and 4A reveal that the fold stimulation of a UV-regulated, GST-c-Jun-binding protein kinase activity correlates with the relative resistance of the SCLC lines to UV-induced cell death.
A polyclonal antiserum directed against the C terminus of the p46 JNK1
isoform was also used to measure the UV-regulated protein kinase
activity in an immune complex assay. The findings (Fig. 4B)
revealed a 3-8-fold stimulation of immune complex-associated protein
kinase activity by UV in extracts prepared from the panel of SCLC
lines. Similar analysis of ERK and p38 MAP kinases immunoprecipitated from extracts prepared from control and UV-irradiated H345 and SHP-77
cells revealed a 3.8- and 2-fold stimulation of ERKs, respectively, and
a 1.1- and 2.8-fold stimulation of p38 MAP kinase, respectively. Thus,
the ERK and p38 MAP kinases displayed little or no regulation by UV in
SCLC, despite reports of strong regulation of p38 MAP kinase by UV in
other cell types (39). Note that the fold stimulation by UV (96 J/m2) of anti-JNK1-precipitated protein kinase activity
from H345, N417, and SHP-77 cells was significantly less than that
observed in the GST-c-Jun complex assay (Fig. 4A). Also, the
fold stimulations observed with the immune complex assay did not
correlate with the relative resistance to UV-induced cell death
(r2 = 0.04, slope is not significantly different
from zero). The different fold stimulations observed with the GST-c-Jun
and immune complex assays in the various SCLC lines suggests that
distinct populations of cellular JNK enzymes were measured. The
sequence of the JNK1 peptide to which the commercial JNK1 antibody was generated is 100, 70, 59, 41, 82, and 53% identical to the sequences (20, 22) of rodent SAPK p46, p54
, p46
, p54
, p46
, and p54
, respectively. The human homologues of SAPK
,
, and
are JNK1, JNK3, and JNK2. Thus, the antibody would be predicted to exhibit a selectivity of p46 JNK1 > p46 JNK3 > p54
JNK1 > p46 JNK2 > p54 JNK3 > p54 JNK2. Immunoblot
analysis of SCLC cells expressing exogenous JNK1 and JNK2 molecules
(not shown) with the JNK1 antibody readily detected the JNK1
polypeptide, but not the JNK2 polypeptide, supporting the predicted
specificity of this antibody.
To further characterize the UV-stimulated JNK activity that bound to
GST-c-Jun and correlated with increased resistance to UV-induced
apoptosis, SCLC proteins were adsorbed to GST-c-Jun beads, resolved by
SDS-PAGE, and immunoblotted with the polyclonal JNK1 antibody. The
results (Fig. 5) revealed detectable JNK
immunoreactivity with an estimated mass of 46 kDa that was specifically
retained by GST-c-Jun, but not GST, from extracts prepared from
unirradiated H69, SHP-77, and N417 cells. Note, however, that the
majority of JNK immunoreactivity remained in the unbound fractions.
These cell lines exhibit similar UV-induced JNK activity with the JNK1 immunoprecipitation reactions, but SHP-77 and N417 cells exhibit a much
greater UV-stimulated JNK response with the GST-c-Jun-binding assay
(Fig. 4). Similar analysis of extracts from UV-irradiated H69, SHP-77,
and N417 cells revealed the appearance in SHP-77 and N417 extracts, but
not H69 extracts, of a more slowly migrating species at ~48-50 kDa
that specifically bound to GST-c-Jun and was largely depleted from the
unbound fraction (compare GST- and GST-c-Jun-bound and -unbound
fractions from UV-treated cells in Fig. 5). The low level of p46-kDa
JNK binding to GST-c-Jun was not different between extracts from
control and UV-irradiated cells. Thus, these data are consistent with
the specific binding of at least two populations of JNK enzymes by
GST-c-Jun. These activities are represented by the p46-kDa
immunoreactivity that binds equally from control and activated extracts
and the p48-50-kDa species that appears in extracts from UV-irradiated
cells and is largely captured by GST-c-Jun. The observed tendency of
MAP kinases to migrate more slowly on SDS-PAGE following
phosphorylation and activation indicates that the more slowly migrating
JNK immunoreactivity is possibly a phosphorylated and activated JNK
isoform.
Influence of Expression of Inhibitory JNK Mutants on JNK Activation and Induction of SCLC Apoptosis
The findings in Figs. 4 and 5
support the existence of distinct JNK isoforms that are activated
following UV irradiation of SCLC. To test the role of JNKs in
UV-induced SCLC apoptosis, we stably expressed influenza hemagglutinin
epitope-tagged JNK1 and JNK2 mutants in which the phosphorylated
threonine and tyrosine in the TPY phosphorylation motif are changed to
alanine and phenylalanine, producing a nonphosphorylatable,
competitive-inhibitory JNK-APF. Transient expression of these JNK-APF
mutants has been previously shown to inhibit JNK-mediated
transcriptional regulation (32, 40). Fig. 6A
shows an immunoblot of cell extracts from SHP-77 cells in which the
HA-JNK1-APF and HA-JNK2-APF cDNAs were transduced by the indicated
retroviruses. A similar level of expression of the two HA-tagged
polypeptides was noted. JNK1-APF expression reduced UV-stimulated JNK
activation to 31 and 36% of the LNCX control response using the JNK
immune complex and GST-c-Jun binding assays, respectively (Fig.
6B). Likewise, JNK2-APF inhibited JNK activation to 51 and
39% of the LNCX response as defined with the JNK immune complex and
GST-c-Jun assays, respectively (Fig. 6B).
Having demonstrated that the JNK-APF constructs are expressed and that
they reduce UV-stimulated JNK activity, we investigated the influence
of these molecular reagents on induction of SCLC apoptosis by UV. Fig.
7 shows that SHP-77 cells expressing the JNK1-APF
construct were more resistant to UV-induced cell death compared with
the LNCX control cells over a range of UV intensities. By contrast, no
significant protection or sensitization to UV-irradiation resulted from
expression of HA-JNK2-APF. The inset (Fig. 7) shows the
average percent cell survival of the indicated SHP-77 transfectants following a 36 J/m2 UV exposure. At this UV intensity, the
JNK1-APF cells were ~3-fold more resistant to UV-induced apoptosis.
Thus, the findings indicate that expression of JNK1-APF inhibits the
activation of specific JNK isoforms that are involved in UV-induced
SCLC apoptosis.
The present study demonstrates that UV irradiation of human SCLC cell lines induces an apoptotic response where several classic SCLC lines were significantly more sensitive to UV-induced cell death than variant SCLC lines that are frequently derived from relapsed SCLC tumors. These findings are consistent with previous reports that classic SCLC lines are more sensitive to cell killing by ionizing radiation than are variant SCLC lines (2, 4). The varied sensitivities of the SCLC lines to both UV and ionizing radiation could be explained by enhanced activity of cytoprotective pathways or reduced functioning of pathways that signal cell death. Manipulation of cellular glutathione levels is known to influence the sensitivity of cells to UV and ionizing radiation (41) and thus represents a potential cytoprotective pathway. However, no consistent differences have been observed in the cellular content of glutathione or glutathione-metabolizing enzyme activities in classic and variant SCLC (42). A wealth of information has recently highlighted the p53 protein as an important modulator of apoptosis in many cancers (5) where loss of functional p53 may increase the threshold of tumor cells to entry into apoptosis. Significantly, a study (43) has demonstrated that the p53 gene was mutated in 100% of SCLC lines and 80% of primary SCLC tumors, indicating that p53 status is not related to the differential sensitivity of variant and classic SCLC to UV and ionizing radiation. Likewise, expression levels of the Myc family members can influence the apoptotic response of tumor cells (5). However, amplification of N-myc and c-myc has been noted in H69 and N417 cells, respectively. Yet, H69 is significantly more UV-sensitive than N417 cells (Fig. 1B). Our observations point toward distinct JNK pathways in SCLC as a mechanism for variable resistance to UV-induced apoptosis. We hypothesize that a GST-c-Jun-binding JNK activity that correlates closely with the resistance of the individual SCLC lines to UV-induced cell death regulates a UV-protective program, while the JNK activity or activities that are disrupted by expression of JNK1-APF signal an apoptotic program. Thus, the overall sensitivity of the various SCLC lines to UV may be dictated by the summation or integration of cell-death and cell-protective programs, which are regulated by distinct JNK pathways.
We observed that expression of an inactivatable JNK1 mutant (JNK1-APF) reduced UV-stimulated JNK activity (Fig. 6) and induction of apoptosis (Fig. 7), indicating the involvement of a JNK1-like enzyme in a UV-stimulated pathway that promotes SCLC apoptosis. The involvement of JNKs and the related p38 MAP kinases has been previously invoked in the induction of apoptosis (29, 30). Acute withdrawal of nerve growth factor from PC12 cells that had previously been differentiated with this factor induces apoptosis in a fraction of the cells. Expression of catalytically inactive forms of MAP kinase/ERK kinase kinase, MAP kinase kinase 3, or MAP kinase kinase 4 or dominant-negative c-Jun significantly reduced apoptosis induced by nerve growth factor withdrawal in this system (30). Likewise, expression of dominant-negative c-Jun or catalytically inactive MAP kinase kinase 4 inhibited apoptosis in U937 monoblastic leukemia cells in response to a host of cytotoxic stimuli, including x-rays and UV-C radiation (29). Together, these studies indicate that stress-induced apoptosis requires a functional JNK signaling cascade.
We also characterized a UV-stimulated, GST-c-Jun-associated JNK activity that correlated closely with the relative resistance of the SCLC lines to UV-induced apoptosis (Fig. 4A). Indeed, precedent exists to support a role for protein kinase pathways in cytoprotection. Pretreatment of a non-small cell lung cancer line with the protein kinase inhibitor, staurosporine, at concentrations that are not specific for protein kinase C, sensitized the cells to killing by ionizing radiation in a manner independent of the repair of double-stranded DNA breaks (44). Also, tyrosine kinase inhibitors significantly sensitized HeLa cells to UV irradiation (45). Clearly, the initiating signal in the UV response is not provided by damaged DNA (11) as the response proceeds in enucleated cells. Membrane-associated Ras proteins are involved in both yeast (16) and mammalian systems (11, 18), consistent with the notion that the UV response arises at the plasma membrane, not damaged DNA in the nucleus. Significantly, resistance of yeast to UV irradiation is correlated with Ras activity and functioning of Gcn4, a yeast homologue of c-Jun and ATF-2 (16). Finally, oncogene-transformed mammalian cells frequently exhibit increased resistance to UV and ionizing radiation compared with their untransformed counterparts (25, 26).
Our findings thus support the existence of distinct JNK isoforms in
SCLC which can regulate respective protective and cell death-inducing
pathways. While the ability of JNK1-APF to inhibit JNK activation and
increase the resistance of the SCLC cells to apoptosis provides
evidence for a cell death-promoting pathway, JNK2-APF similarly
inhibited JNK activation, but did not influence UV-stimulated apoptosis
of SHP-77 cells. Therefore, we do not have direct evidence for a JNK
pathway involved in cytoprotection. We have observed, however, that
SHP-77 cells expressing HA-JNK2-APF, but not JNK1-APF, are
significantly more sensitive to induction of
apoptosis2 in response to a novel substance
P antagonist, [D-Arg1,
D-Phe5, D-Trp7,9,
Leu11]substance P, which has recently been shown to induce
apoptosis in SCLC (46). Clearly, published findings indicate that the defined JNK isoforms are not functionally redundant despite being activated by many of the same stimuli including UV irradiation and
TNF- (32, 33). In the yeast, Saccharomyces cerevisiae, where the HOG1 MAP kinase is required for growth on hypertonic medium,
expression of the JNK1 enzyme, but not JNK2 enzyme, restores the
ability of yeast lacking HOG1 to grow on hypertonic medium (33). It is
likely that divergent roles for highly homologous members of a MAP
kinase family such as JNK1 and JNK2 are dictated in part by
differential recognition and phosphorylation of cellular targets. In
fact, by virtue of a unique peptide sequence within its catalytic
domain, JNK2 exhibits a 25-fold higher affinity for the transcription
factor substrate, c-Jun, relative to JNK1 in vitro, which
translates to a selective ability of JNK2 to regulate a
c-Jun-controlled promoter (32). Thus, clear definition of the role of
the JNK signaling cascade in cell regulation demands analysis of the
individual JNK gene products and their respective cellular targets.
Immunoblotting of the cellular proteins that were specifically adsorbed
to GST-c-Jun revealed a 46-kDa JNK species that bound regardless of UV
irradiation as well as a JNK species with a slower mobility on SDS-PAGE
(~48 kDa) that appeared only in extracts from UV-irradiated cells and
was largely adsorbed from the extract by GST-c-Jun (Fig. 5). The
selective adsorption to GST-c-Jun of an activated JNK species with a
decreased mobility has been previously reported (47). It is unlikely
that this JNK activity represents JNK1 or JNK2, since these protein
kinases do not exhibit enhanced binding to GST-c-Jun when
phosphorylated and activated (32, 47). We are considering the
possibility that the GST-c-Jun-associated JNK activity that correlates
with increased UV resistance represents the JNK3 or SAPK isoform.
Human JNK3 as well as rodent SAPK
is reported to be most highly
expressed in neural tissue (21, 22) and were, in fact, cloned from
brain cDNA libraries (20-22). It is interesting to speculate that
a JNK3/SAPK
isoform may be more highly expressed in SCLC, as this
human tumor expresses neuroendocrine differentiation (1).
The JNKs and the related p38 MAP kinases are noted for their brisk
regulation in response to a host of cellular stresses including UV and
ionizing radiation, heat shock, hyperosmolarity, and cytotoxic drugs
(17-19, 39, 48). Furthermore, these protein kinases are activated by
tumor necrosis factor- and interleukin-1 receptors (19, 39),
receptor tyrosine kinases (19, 27), a growing number of G
protein-coupled receptors (49, 50), GTPase-deficient forms of
heterotrimeric G proteins (37, 51, 52), and the Ras and Rho families of
low molecular weight G proteins (53, 54). While JNK activation observed
in response to cell stresses is often accompanied by cell damage and
cell death, thereby complicating clear dissection of the role for the
JNK response, the regulation of the JNK pathway by growth factor
receptors and heterotrimeric G proteins is, in many instances,
associated with cellular transformation (52, 54) or differentiation
(37), not cell death. It is important to note that cell stresses often
selectively activate the JNK cascade with little or no activation of
the parallel ERK MAP kinase pathway. In contrast, stimulation of the
JNK pathway by receptor tyrosine kinases and G protein-coupled
receptors is accompanied by robust activation of the ERKs. Indeed,
coordinate activation of the ERKs with the JNKs or p38 MAP kinases is
associated with protection against apoptosis (30, 55). Thus, the
function of the JNKs in a particular cellular context will likely
depend not only on the cell type and its state of growth control, but also on additional signaling pathways that are coordinately activated.
In conclusion, the findings of this study are supportive of a role for distinct JNK pathways in signaling protective and apoptotic events in SCLC following UV radiation-induced cell damage. Enhanced activation of a JNK pathway that may promote protection against radiation-induced damage represents a potentially novel mechanism for conferring resistance to a variety of cytotoxic cell stresses. Molecular approaches leading to overexpression of dominant inhibitory forms of upstream regulatory protein kinases and downstream JNK targets coupled with pharmacologic inhibitors of the involved protein kinases as they become available will permit a definitive assignment for the role of the JNK pathway in the SCLC response to UV and ionizing radiation.
We appreciate the assistance of Nancy L. Johnson (National Jewish Center for Immunology and Respiratory Medicine, Denver, CO) for determination of SCLC apoptosis using the terminal deoxytransferase-based assay.