| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Tumor Biology |
Departments of Radiation Oncology [T. M. G., E. V. R., H. Z., C. I. S., A. D. C.] and Pharmacology [A. D. C.], Lineberger Comprehensive Cancer Center [T. M. G., E. V. R., H. Z., C. I. S., A. D. C.] and Curriculum in Genetics and Molecular Biology [T. M. G., A. D. C.], University of North Carolina, Chapel Hill, North Carolina 27599
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
B (NF-B), and Raf>mitogen-activated protein kinase/extracellular signal-regulated kinase kinase (MEK)>extracellular signal-regulated kinase] as necessary or sufficient for Ras-mediated radioresistance in matched pairs of RIE-1 rat intestinal epithelial cells expressing oncogenic Ras or empty vector (RIE-Ras and RIE-vector). Inhibiting PI3-K with LY294002 sensitized RIE-1 cells to IR in a dose-dependent manner, indicating that PI3-K is necessary for radioresistance, whereas inhibition of NF-B with the super-repressor IB
had little effect on survival. Expression of either the constitutively active catalytic subunit of PI3-K, p110-CAAX, or the Ras effector domain mutant 12V/40C, which retains binding to PI3-K but is impaired in binding to other Ras effectors, was sufficient to confer partial radioresistance. Expression of either a constitutively active form of the serine/threonine kinase Raf-1 or the Ras effector domain mutant 12V/35S, which retains binding to Raf but is impaired in binding to other Ras effectors, was also sufficient to confer partial radioresistance. Surprisingly, however, even complete inhibition of MEK activity by using U0126 resulted in no change in post-IR survival whatsoever. Thus, whereas Raf contributes to Ras-mediated radioresistance, this is accomplished through a MEK-independent pathway. Taken together, these results indicate that multiple pathways, including both PI3-K-dependent and Raf-dependent but MEK-independent signaling, are required for Ras-mediated radioresistance in epithelial cells. Finally, we demonstrate that Ras-mediated radioresistance can be uncoupled from Ras-mediated transformation, in that PI3-K is required for radioresistance but not transformation, whereas MEK and NF-B are required for transformation but not radioresistance in RIE-1 epithelial cells. |
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Evidence for Ras-mediated radioresistance has been demonstrated both in rodent model systems, including NIH 3T3 fibroblasts, REF fibroblasts, rat rhabdomyosarcomas, and other cell types (4, 5, 6, 7 , 9) , and in human tumor cell lines including EJ Ras-transformed bladder carcinoma, DLD-1 colon carcinoma, and HT1080 fibrosarcoma (8 , 10) . However, overexpression of Ras or the presence of activated Ras does not increase radioresistance in all contexts (11, 12, 13) . Nevertheless, blocking Ras activity by using pharmacological inhibitors or anti-Ras neutralizing antibodies or by genetic loss of the activated ras allele from a tumor cell line has been shown to increase radiosensitivity in certain contexts (9 , 10 , 14 , 15) . Thus, the preponderance of evidence suggests that the activity of oncogenic Ras induces radioresistance and that interfering with signaling from Ras increases radiosensitivity.
Whereas oncogenic Ras can clearly tip the balance toward survival and radioresistance, the signaling pathways by which this is accomplished are not fully understood and are likely to be cell type dependent. The fact that oncogenic transformation and radioresistance can be uncoupled is shown by the inability of the transforming oncogenes Myc and Abl to confer radioresistance (7)
. Therefore, although several signaling pathways downstream of Ras have been fairly well delineated, and their contribution to Ras-mediated transformation has been well studied, it remains unclear which of these pathways also contribute to Ras-mediated radioresistance, particularly in epithelial cells, from which the majority of human cancers arise. Likely candidates include the proliferative pathway Raf>MEK3
>ERK, the survival pathways PI3-K>Akt and NF-B, and stress response pathways involving JNK and p38 MAPKs.
Several of these Ras-mediated signaling pathways are up-regulated in response to IR. For example, the proliferative Raf>ERK kinase cascade is activated by IR (2–5 Gy) in laryngeal squamous carcinoma cells, in U937 human myeloid leukemia cells, and in MCF-7 and MDA-MB-231 mammary carcinoma cells (16, 17, 18, 19)
. The stressresponse MAPK, JNK, has been reported to be activated in NIH 3T3, Jurkat, and U937 cells in response to very high doses of IR (20–100 Gy; Refs. 20, 21, 22
) and also after a more clinically relevant 2-Gy dose in MDA-MB-231 mammary carcinoma cells and A431 squamous carcinoma cells (19
, 23) . Another stress response MAPK, p38, has also been shown to be activated in a variety of cell types, including 293 human embryonic kidney cells, SW480 colon carcinoma cells, HeLa cervical carcinoma cells, and U2OS osteosarcoma cells (24, 25, 26)
, and such activation has been proposed to increase radiation-induced apoptosis. Finally, IR has been shown to activate NF-B gene expression in several cell types as well (27
, 28)
.
Whether these Ras-regulated pathways that are altered by IR also contribute to Ras-mediated increases in radiation survival remains unclear. Antisense Raf increased radiosensitivity and caused tumor regression in irradiated SQ-20B laryngeal carcinoma-bearing athymic nude mice (29 , 30) , but whether this pathway specifically mediates radioresistance downstream of oncogenic Ras was not examined. Some studies have shown that inhibition of MEK causes radiosensitization, whereas others have shown that MEK is dispensable for radioresistance in several carcinoma cell lines (31 , 32) .
We examined the activation of and requirement for several Ras-mediated signaling pathways in response to IR, using the RIE-1 rat intestinal epithelial model system. To isolate specific signaling pathways in the radioresponse phenotype, we used pharmacological inhibitors as well as ectopic expression of effector domain mutants of Ras and of downstream Ras effectors. Our results suggest that PI3-K is a major contributor, but not the sole contributor, to Ras-mediated radioresistance and that a novel MEK-independent pathway downstream of Raf also plays an important role. Furthermore, our results also demonstrate specific ways in which signaling to Ras-mediated radioresistance can be uncoupled from Ras-mediated transformation.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
B, SR-IB, which is insensitive to degradation by phosphorylation (36)
, was expressed from an adenoviral vector (UNC-CH Viral Vector Core Facility) at a multiplicity of infection of 50 to optimize cellular transduction. The negative control virus expressed only GFP, which was used to monitor transduction efficiency. Constitutively active forms of PI3-K (p110-CAAX) and Raf-1(22W) have been described in Ref. 37
and Ref. 38
, respectively, and were transferred to pBABE (35)
.
Cells.
RIE-1 rat intestinal epithelial cells (Ref. 39
; a gift of Robert J. Coffey, Vanderbilt University, Nashville, TN) were maintained in monolayer culture in DMEM-H (Life Technologies, Inc., Gaithersburg, MD) supplemented with 5% fetal bovine serum, 50 units/ml penicillin, and 50 mg/ml streptomycin at 37°C in a humidified atmosphere of 90% air and 10% CO2. Cells were transfected with the pBABE plasmids using Fugene (Boehringer Mannheim, Indianapolis, IN) according to the manufacturer’s instructions, and stable cell lines were established by selection in 2 ug/ml puromycin (Life Technologies, Inc.).
Clonogenic Survival Curve Assays.
RIE-1 cells stably expressing Ras or empty vector were plated at low density and irradiated with graded, single doses from a Co60 Theratron irradiator (Atomic Energy of Canada, Limited) at a dose rate of 100 cGy/min. The culture medium was changed immediately after IR. After 2 weeks of incubation, samples were fixed in methanol:acetic acid (3:1, v:v) and stained in 1% crystal violet, and the number of viable colonies (defined as those colonies containing 
50 cells) per dish was counted. Surviving fraction was calculated from the number of colonies formed in the irradiated dishes compared with the number formed in the unirradiated control, where plating efficiency is defined as the percentage of cells plated that form colonies, and surviving fraction = number of colonies formed/(number of cells plated x plating efficiency). To assure that plating efficiency was not affected by pharmacological inhibitors, we counted the number of colonies formed on control, inhibitor-treated but mock-irradiated dishes. The lack of effect of inhibitors on plating efficiency was expected because the cells are plated the day before irradiation and are not exposed to inhibitors until immediately before and during irradiation. Each point on the survival curve represents the mean surviving fraction from at least two dishes. The curve was fitted to the linear quadratic model (S = e-(D + BD2)) using a macro for KaleidaGraph, kindly supplied by Dr. Eric Bernhard (Department of Radiation Oncology, University of Pennsylvania) and originally based on a similar program developed by Dr. Michael Joiner, who was then at the Gray Laboratory, Northwood, United Kingdom.
Clonogenic Survival in the Presence of Pharmacological Inhibitors.
RIE-1 cells were incubated with pharmacological inhibitors of PI3-K (LY294002; Biomol, Plymouth Meeting, PA), MEK (U0126; Alexis, San Diego, CA), or p38 MAPK (SB203580; Calbiochem, San Diego, CA) for 25 min and then exposed to a single dose of 7 Gy of IR. The concentrations of inhibitors used are shown in the figure legends. Inhibitors were removed 10 min after irradiation through replacement of the culture medium. Cells were cultured for 14 days before fixation, staining, and counting. The number of colonies formed in inhibitor-treated dishes was compared with the number of colonies formed in the vehicle (DMSO)-treated dishes. The absolute surviving fraction in each case was normalized to that of a "no drug-no irradiation" control.
Western Blotting.
Protein expression levels and phosphorylation status were determined by Western blot analysis. RIE-1 cells stably expressing H-Ras(12V) or empty vector were starved for 24 h in 0.5% serum, exposed to a single dose of 7 Gy of IR, and then lysed at various times as shown. For inhibitor treatment, cells were starved for 24 h in 0.5% fetal bovine serum, treated with an inhibitor or an equivalent amount of vehicle (DMSO) for 30 min, and then exposed to 7 Gy of IR. Cells were lysed in Triton X-100 lysis buffer [50 mM Tris-HCl (pH 7.5), 100 mM NaCl, 1% (v/v) Triton X-100, 10% (v/v) glycerol, 1 mM sodium orthovanadate, 10 mM p-nitrophenyl phosphate, 20 mM ß-glycerophosphate, 0.5 mM Pefabloc, 5 ug/µl leupeptin, and 10 ug/mg aprotinin] at 55 min after drug treatment (25 min after IR). Protein concentration in the lysates was determined using a Bio-Rad protein assay kit (Bio-Rad Laboratories, Hercules, CA). Twenty µg of protein were loaded for each sample. Proteins were separated by SDS-PAGE on 10% SDS, 15% acrylamide low cross-linker gels for anti-ERK blots (40)
and on Bio-Rad Criterion precast gels 12.5% Tris-HCl for all other probes. Separated proteins were electroblotted onto polyvinylidene difluoride membranes and blocked in TBS/T containing 5% nonfat powdered milk. Primary and secondary antibodies were added sequentially. After washing with TBS/T, the blots were developed using SuperSignal chemiluminescent substrate (Pierce, Rockford, IL).
Antibodies.
Primary antibodies were all rabbit polyclonal antibodies diluted 1:1,000 in Tris-buffered saline/Tween/5% BSA and were as follows: anti-phospho-Akt (specific for Akt phosphorylated at serine 473), anti-Akt, anti-phospho-p38 MAPK, and anti-p38 MAPK (Cell Signaling, Beverly, MA); anti-ERK (sc-94; Santa Cruz Biotechnology, Santa Cruz, CA); and anti-phospho-JNK and anti-JNK (Promega, Madison, WI). For all primary antibodies, the secondary antibody used was horseradish peroxidase-conjugated antirabbit IgG (Amersham, Arlington Heights, IL) diluted 1:30,000 in TBS/T + 5% nonfat milk.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, for each dose given, RIE-Ras cells consistently survived better than RIE-vector cells, demonstrating that Ras does confer radioresistance in RIE-1 cells. For our additional studies with inhibitors, we chose a dose of 7 Gy because, as shown in Fig. 1
, this dose produces a modest amount of cell death (surviving fractions of between approximately 0.2 and 0.5) and also falls on the exponential portion of the survival curve.
|
The PI3-K>Akt pathway has been shown to enhance survival after a variety of cellular stresses, but the nature of its response to IR has not been established. Because the level of phosphorylation of the PI3-K downstream target Akt can be used as an indirect measure of PI3-K activity, we determined the levels of phospho-Akt after IR. We observed that these are responsive to IR in a biphasic manner (Fig. 2A)
: immediately after IR, phospho-Akt levels drop below the basal level in both RIE-Ras and RIE-vector cells. They begin to rise again at 30 min and 10 min in RIE-vector and RIE-Ras cells, respectively. Phospho-Akt levels rise to near-basal levels at 60 min and then decrease again by the 2 h time point, suggesting that phosphorylation of Akt is dynamically regulated after IR. In response to IR, RIE-Ras cells down-regulate phospho-Akt to a lesser extent than RIE-vector cells. RIE-Ras cells display higher levels of phospho-Akt than RIE-vector cells both basally and in response to IR, indicating that the PI3-K>Akt survival pathway may contribute to Ras-mediated radioresistance.
|
. Phospho-ERK levels began to drop by 4 h after IR in RIE-vector cells but remained higher than the basal level in RIE-Ras cells for at least 4 h. IR activated ERK1 and ERK2 in both RIE-Ras and RIE-vector cells, with higher levels of activation in RIE-Ras cells, indicating that the Raf>MEK>ERK pathway may also be important for radioresistance.
Ras also regulates the JNK and p38 MAPKs. Because JNK and p38 are up-regulated in response to many cellular stresses, including UV radiation, they might also be up-regulated in response to IR. Therefore, we examined the phosphorylation status of JNK and p38 MAPK after IR (Fig. 2, C and D
, respectively). We detected no phosphorylation of either JNK or p38 MAPK in response to IR, although we could demonstrate that these SAPKs can be phosphorylated in RIE-1 cells in response to the cellular stressor anisomycin (Fig. 2, C and D)
. Therefore, JNK and p38 MAPK can be phosphorylated in RIE-1 cells but are not phosphorylated after IR under these conditions.
Signaling through PI3-K Is Necessary for Ras-mediated Radioresistance in RIE-1 Epithelial Cells.
To determine which signaling pathways are necessary for Ras-mediated radioresistance in RIE-1 epithelial cells, we used pharmacological inhibitors of specific signaling pathways downstream of Ras. Western blot analysis was used to determine the effectiveness of the inhibitors at the concentrations used, and clonogenic survival assays of colony formation after 7 Gy of IR were performed to determine whether the inhibitors altered cell survival.
The PI3-K>Akt pathway is now commonly regarded as a cellular survival pathway and was therefore a logical candidate for analysis. The inhibitor LY294002 blocks the activity of PI3-K, so it also blocks the phosphorylation and activation of downstream targets of PI3-K, such as the serine/threonine kinase Akt. As expected, LY294002 completely blocked phosphorylation of Akt at both 1 and 10 µM (Fig. 3A)
in RIE-vector cells but unexpectedly failed to completely block phosphorylation of Akt, and presumably PI3-K function, in RIE-Ras cells. Thus, RIE-Ras cells may display greater up-regulation of LY294002-resistant isoforms of PI3-K.
|
. Interestingly, higher levels of phospho-Akt were seen in samples that corresponded to those treatments with the greatest surviving fraction. In RIE-Ras cells, where LY294002 was unable to completely inhibit phosphorylation of Akt, cell survival was less affected by inhibitor treatment. We observed similar radiosensitization of RIE-vector cells with the PI3-K inhibitor wortmannin (data not shown). From these data, we conclude that PI3-K is a necessary mediator of radiation resistance in both RIE-Ras and RIE-vector cells.
Ras Activation of Either PI3-K or Raf Is Sufficient for Partial Radioresistance in RIE-1 Epithelial Cells.
H-Ras effector domain mutants have become a widely used means of determining which signaling pathways are sufficient or necessary for a particular Ras activity (41, 42, 43, 44)
. The Ras effector domain mutant H-Ras(12V,35S) retains signaling through Raf but is impaired in signaling through the RalGDS and PI3-K signaling pathways. The effector domain mutant H-Ras(12V,37G) retains signaling through RalGDS but is impaired in signaling to Raf and PI3-K. The effector domain mutant H-Ras(12V,40C) retains signaling through PI3-K but is impaired in signaling to Raf and RalGDS. To determine which signaling pathways were sufficient for Ras-mediated radiation resistance, RIE-1 cell lines stably expressing each of these effector domain mutants, RIE-H-Ras(12V,35S) (RIE-35S), RIE-H-Ras(12V,37G) (RIE-37G), and RIE-H-Ras(12V,40C) (RIE-40C), were established and analyzed for radioresponse in clonogenic survival assays. Both RIE-35S and RIE-40C conferred radioresistance intermediate between that of RIE-vector and RIE-Ras cells at lower doses of radiation (1–6 Gy), whereas at higher doses (10–15 Gy), cells expressing the effector domain mutants were not more radioresistant than RIE-vector cells (data not shown). These results indicate that Ras activation of either Raf (i.e., RIE-35S) or PI3-K (RIE-40C) may be partially sufficient to confer radioresistance but that additional activities are required. Survival of RIE-37G cells was similar to that of RIE-vector cells, implying that RalGDS is not involved in radioresistance in this context. We conclude that Ras uses multiple signaling pathways to confer radioresistance and that no one pathway can be singled out as solely sufficient for Ras-mediated radioresistance by this method.
We therefore also used a complementary approach by stably expressing constitutively activated forms of PI3-K and Raf-1 in RIE-1 cells and analyzing the consequences to radioresistance via clonogenic assay. Both a constitutively active form of PI3-K, p110-CAAX, and a constitutively active form of Raf-1, 
Raf22W, conferred radioresistance intermediate to that of RIE-Ras versus RIE-vector (Fig. 4)
. These results are in agreement with those of the Ras effector domain mutants and demonstrate that both the PI3-K and Raf pathways are partially sufficient for Ras-mediated radioresistance in this context.
|
at both 1 and 10 µM in both RIE-Ras and RIE-vector cells. However, when cells were treated with 1 or 10 µM U0126 and irradiated with 7 Gy, the surviving fraction was not altered (Fig. 5B)
. Similar results were seen with the MEK inhibitor PD98059 (data not shown). Therefore, although Ras signaling through Raf is partially sufficient, and Ras signaling to ERK is up-regulated upon IR, MEK activation is not necessary for Ras-induced radioresistance in RIE-1 cells. Thus, a novel Raf-dependent, MEK-independent pathway mediates radioresistance in this epithelial cell type.
|
, p38 MAPK inhibition did not significantly alter clonogenic survival after IR. Although the levels of phospho-p38 MAPK in response to IR were too low to examine an effect of SB203580, the inhibitor could block p38 MAPK activity in response to anisomycin treatment (data not shown). Therefore, Ras does not mediate radioresistance by up-regulating p38 MAPK.
|
B Survival Pathway Only Weakly Impairs Postradiation Survival.B. To determine the relative importance of NF-B in Ras-mediated radioresistance, we blocked NF-B activation by use of an adenovirus expressing the super-repressor SR-IB, a nondegradable mutant form of inhibitory B (S32A/S36A) that locks NF-B in the cytoplasm. Electrophoretic mobility shift assays for the presence of NF-B in nuclear lysates of RIE-Ras and RIE-vector cells in the presence or absence of Ad-SR-IB confirmed that the super-repressor blocked NF-B translocation to the nucleus under these conditions (data not shown). We then performed clonogenic survival assays in the presence or absence of Ad-SR-IB (Fig. 7)
. In contrast to inhibition of the PI3-K survival pathway by LY294002, which resulted in decreases in survival of >40% (Fig. 3B)
, inhibition of NF-B resulted in only a small decrease in clonogenic survival. Indeed, the presence of the negative control adenovirus created a greater difference in survival than did the additional expression of the inhibitory SR-IB. This was true regardless of whether the negative control adenovirus expressed only GFP (as in Fig. 7
) or only lacZ (data not shown). Thus, we conclude that NF-B plays only a minor role in Ras-mediated radioresistance in these cells.
|
. H-ras-transformed RIE cells were the most radioresistant at the clinically relevant dose of 2 Gy, with the remaining Ras isoforms showing intermediate resistance. Thus, all three major isoforms of Ras can confer radioresistance to RIE-1 epithelial cells. This finding is consistent with the ability of all three Ras isoforms to activate both PI3-K and Raf.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Activation of the PI3-K Survival Pathway Is a Critical Component of Ras Signaling That Is Both Necessary and Partially Sufficient for Radioresistance.
The PI3-K survival-associated signaling pathway allows cells to better tolerate a variety of cellular insults, including serum deprivation, extracellular matrix detachment, and DNA damage (46, 47, 48, 49, 50)
. Ras is known to activate the PI3-K>Akt pathway, but it was not clear whether PI3-K would play a role in Ras-mediated radioresistance because PI3-K is thought to promote survival by suppressing apoptosis (51)
. For most cells exposed to IR, death while attempting to divide, i.e., clonogenic death, rather than apoptotic death, is the predominant mechanism of cell killing (52)
. Now we demonstrate that PI3-K also is a critical component that is both necessary and partially sufficient for Ras-mediated radioresistance in RIE-1 epithelial cells.
Using clonogenicity (colony formation) as an end point, we show here that blocking PI3-K activity with the inhibitor LY294002 sensitized RIE-Ras cells to IR, i.e., that PI3-K is necessary for radioresistance. The observation that PI3-K is partially sufficient for radioresistance is supported here by two different lines of evidence. First, the Ras effector domain mutant 40C that retains binding to PI3-K but is impaired in binding to other Ras effectors confers partial radioresistance. Second, expression of a constitutively activated form of PI3-K, p110-CAAX, confers partial radioresistance. These results are consistent with the ability of Ras to use PI3-K as an immediate downstream effector target and to up-regulate PI3-K activity to increase cellular survival.
In RIE-Ras cells, the PI3-K inhibitor LY294002 reduced PI3-K activity but failed to return it to the basal level seen in RIE-vector cells. This is consistent with the lesser effect of LY294002 on the clonogenic survival of RIE-Ras cells and also supports the idea that Ras up-regulates PI3-K to enhance radioresistance. PI3-K-related kinase family members such as ATM, ATR, and DNA-PK are not inhibited at a concentration of 1 µM LY294002 as used here, nor are these proteins known to be activated by Ras. The observation that 1 µM LY294002 significantly decreased the survival for RIE-vector cells indicates that PI3-K is a major target of radiosensitization by the kinase inhibitor LY294002. It is also possible that Ras may activate additional, LY294002-insensitive isoforms of PI3-K that also contribute to the PI3-K activity remaining after LY294002 treatment and thereby contribute to radioresistance.
While this manuscript was in preparation, Gupta et al. (53) showed that blockade of PI3-K with LY294002 radiosensitizes Ras-transformed REFs and two human tumor cell lines containing activated ras alleles. Furthermore, expression of a constitutively active PI3-K in REFs increased radiation resistance. Our studies were carried out with epithelial cells, confirming and extending the finding that Ras uses PI3-K to enhance radiation resistance to multiple cell types. Indeed, PI3-K may play an even greater role in epithelial cells than in fibroblasts because in our study PI3-K inhibition resulted in decreased survival in both RIE-Ras and RIE-vector epithelial cells, whereas Gupta et al. (53) observed that PI3-K inhibition decreased survival in Ras-transformed REFs, but not in normal REFs.
Akt Phosphorylation Is Regulated Dynamically after IR.
Unexpectedly, we observed that phospho-Akt levels were regulated in a dynamic manner after irradiation, including a dramatic drop in phospho-Akt in as short a time as 5 min after IR. The rapid kinetics of changes in Akt phosphorylation after IR have not been reported previously, and they suggest the existence of an IR-regulated Akt-phosphatase. To date, no phosphatases that regulate Akt activity have been reported. It is not clear whether Akt is a necessary downstream component of the PI3-K activity that mediates radioresistance, but Akt phosphorylation is restored more rapidly and to a greater extent in RIE-Ras cells than in RIE-vector cells. Additional studies will be required to determine both the mechanism of rapid down-regulation of phospho-Akt and its role in radioresistance.
The NF-B Survival Pathway Is Not Critical for Ras-mediated Radioresistance in RIE-1 Cells.
NF-B is a major survival pathway that is activated downstream of Ras. Although Ras requires NF-B for transformation and IB blocks Ras focus formation (36)
, we show here that inhibition of NF-B by use of the super-repressor IB had relatively little effect on post-IR survival in RIE-Ras or RIE-vector cells. A previous study showed that SR-IB moderately decreased clonogenic survival in HT1080 fibrosarcoma cells after a single 10-Gy treatment, although it is not known whether the presence of the oncogenic ras mutation (N-Ras61K) modulated this response (54)
. Furthermore, SR-IB only modestly decreased survival in several human tumor cell lines, whereas inhibition of NF-B by use of the 26S proteasome inhibitor PS-341, which blocks the activity of both NF-B and other proteins also regulated by proteasomal degradation, displayed somewhat better radiosensitizing activity (55)
. However, this effect was independent of ras mutation status. Taken together, these results suggest that NF-B alone is not a significant contributor to Ras-mediated radioresistance, although proteasome inhibition can impair survival in a non-Ras-dependent manner.
The SAPKs JNK and p38 MAPK Do Not Play a Role in Ras-mediated Radioresistance.
The SAPKs JNK and p38 MAPK are activated in response to multiple cellular stresses including tumor necrosis factor, osmotic stress, and heat shock (56
, 57)
. Therefore, we expected that both JNK and p38 MAPK would also be activated in response to IR. However, neither JNK nor p38 MAPK phosphorylation was up-regulated by IR in either RIE-Ras or RIE-vector cells, and the p38 MAPK inhibitor SB203580 did not alter clonogenic survival of RIE-1 cells. These results were surprising because the SAPKs are responsive to DNA-damaging agents and are up-regulated by UV radiation (24)
. However, it has been shown recently that SAPKs are not always up-regulated by IR (26
, 58)
, and evidence exists that there are cell type differences in SAPK induction (24)
. In addition, SAPKs are most often reported to be activated at doses of IR that are much higher (20–100 Gy) than the daily doses used for radiotherapy (2–3 Gy) and in our studies (1–7 Gy). Because neither JNK nor p38 MAPK was phosphorylated in response to IR in RIE-1 cells, and inhibition of p38 MAPK did not alter radiation survival, we conclude that Ras does not induce radioresistance through the SAPKs in RIE-1 cells.
Raf but not MEK Plays a Critical Role in Ras-mediated Radioresistance in RIE-1 Cells.
Ras activates the Raf>MEK>ERK signal transduction pathway to promote cell proliferation (59)
and to play a role in cell survival in some contexts (60, 61, 62)
. However, RIE-1 epithelial cells with MEK activity transiently blocked using the MEK inhibitors U0126 or PD98059 did not show any changes in clonogenic survival after irradiation. There are conflicting reports in the literature concerning the consequences of MEK inhibitor treatment on radioresistance. Others using the same MEK inhibitors have likewise reported that MEK activity is not necessary for radioresistance of REFs, bladder carcinoma cells, or squamous cell carcinoma cells (31
, 32 , 53)
. However, long-term treatment with the MEK inhibitors PD98059 and U0126 has been shown to radiosensitize A431 squamous carcinoma cells and MDA-MB-231 mammary carcinoma cells (19
, 63) . These distinctions may be due to cell type differences or to differences in MEK inhibitor exposure. These inhibitors reduce proliferation and thus may block colony formation in the absence or presence of radiation, as opposed to specifically radiosensitizing cells. Long-term treatment may also inhibit autocrine loops that regulate PI3-K activity and/or the activity of other molecules involved in radioresistance.
The lack of effect of MEK inhibitors on radioresistance has generally been interpreted to mean that the Raf>MEK>ERK kinase cascade is not relevant to radioresistance, thereby leading to a potentially confusing situation wherein the role of Raf in mediating radioresistance is unclear. It is certainly possible that cell type and context are the simple and final determinants of a requirement for Raf and that these vary greatly. Another, more unifying option may be that although MEK itself and thus the Raf>MEK>ERK cascade do not mediate radioresistance, the Raf-1 kinase plays a significant role but with different downstream consequences than MEK activation. Such a model could resolve the apparent discordance between the ability of Raf antisense constructs to radiosensitize cells (29 , 30) and the lack of ability of MEK inhibitors to do the same thing.
We show here in two ways that Raf is partially sufficient for Ras-mediated radioresistance in RIE-1 epithelial cells. Both the Ras effector domain mutant 35S, which retains signaling through Raf but is impaired in signaling through other Ras effectors, and a constitutively active form of Raf, 22W, were able to induce radioresistance in RIE-1 epithelial cells. In addition, the basal levels of phosphorylated, active forms of ERK were increased in RIE-Ras cells compared with RIE-vector cells, and these were further increased in both cell lines by IR. Thus, Ras-mediated radioresistance proceeds at least in part through Raf. These results are in agreement with previous studies that have shown that Raf is up-regulated in response to IR in laryngeal squamous carcinoma cells (17)
and that antisense Raf reduces survival in the same cell type (29
, 30)
, although none of these reports concerned Raf regulation of radioresponsiveness in the context of signaling downstream of Ras activation.
MEK-independent Signaling Downstream of Raf.
Because we were able to block MEK activity without altering clonogenic survival, we concluded that MEK activity at the time of IR exposure does not play a role in Ras-mediated radioresistance of RIE-1 cells. Thus, although Raf is partially sufficient for Ras-mediated radioresistance, the signaling pathway downstream of Raf does not proceed via the conventional Raf>MEK>ERK kinase cascade, and instead, an unidentified non-MEK effector of Raf is critical in the context of RIE-Ras and radioresistance.
Several types of studies have suggested that Raf mediates both MEK-dependent and -independent signaling pathways. First, although both parental Ras and the 35S Ras effector domain mutant that retains Raf activation can inhibit myogenesis, blocking MEK with PD98059 fails to block this effect (64)
. Thus, Ras signaling can be Raf dependent but MEK independent. Second, a mutant of activated Raf engineered to disrupt MEK1/2 binding [RafBXB(T481A)] uncouples Raf from growth and morphological transformation but not from NF-B or p90 S6 kinase activation (65)
. Thus, Raf signaling can be both MEK dependent and independent. Third, a murine genetic knockout of Raf-1 that has been restored with a Raf mutant (340F/341F) that fails to activate MEK nevertheless causes ERK phosphorylation and activation (66)
. Thus, Raf can activate ERK in a MEK-independent manner. Our present results suggest that a MEK-independent pathway downstream of Raf is also important in mediating radioresistance. Additional investigations will be required to determine the components of such a pathway.
Uncoupling of Ras-mediated Radioresistance and Transformation.
Transformation by some oncogenes such as Abl and Myc is not sufficient to confer radioresistance (7
, 9)
, but the signaling pathways that permit transformation but fail to induce radioresistance have not yet been identified for these oncogenes. We demonstrate here for the first time and in three distinct ways that the Ras-mediated radioresistant phenotype can be uncoupled from that of Ras-mediated transformation. First, although we show here that PI3-K is necessary and partially sufficient for Ras-mediated radioresistance in RIE-1 cells, inhibition of PI3-K with LY294002 does not impair Ras-induced soft agar growth or protect against suspension-induced cell death (anoikis) in RIE-1 cells (35)
, nor does activated PI3-K confer resistance to anoikis in this cell type (35)
. These results demonstrate that PI3-K is important for Ras-mediated radioresistance but not for Ras-mediated transformation in RIE-1 epithelial cells. Conversely, MEK is not necessary for Ras-mediated radioresistance in RIE-1 cells, as shown here by the complete lack of effect of MEK inhibition on postradiation survival, whereas it is crucial for Ras-mediated transformation in this same cell type, as shown by the ability of a MEK inhibitor to reverse Ras-mediated growth and morphological transformation (67)
. Finally, like MEK, NF-B plays a more important role in Ras-mediated transformation than in Ras-mediated radioresistance in RIE-1 cells. Although we show here that inhibition of NF-B with the SR-IB does little to alter post-IR survival, the same procedure has been shown to block Ras-induced morphological transformation (36)
. Therefore, Ras uses distinct effector pathways to mediate distinct biological consequences.
In conclusion, we have provided new evidence that Ras mediates radioresistance in RIE-1 epithelial cells through both PI3-K and a novel Raf-dependent, MEK-independent pathway. The role of each of these Ras effector pathways in mediating radioresistance can be uncoupled from its role in mediating transformation. Our results support the use of inhibitors of both the PI3-K and Raf effectors as possible radiosensitizers and reinforce the importance of providing such inhibition at the level of the immediate downstream effectors of these Ras signaling pathways.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 This work was supported by NIH Grants CA67771, CA76092 (to A. D. C.), and CA83753 (to C. I. S.). T. M. G. was supported by a NIH Cancer Cell Biology Training Grant. 
2 To whom requests for reprints should be addressed, at Department of Radiation Oncology, University of North Carolina at Chapel Hill, Campus Box 7512, Chapel Hill, NC 27599-7512. E-mail: cox{at}radonc.unc.edu
3 The abbreviations in use are: MEK, mitogen-activated protein kinase/extracellular signal-regulated kinase kinase; ERK, extracellular signal-regulated kinase; MAPK, mitogen-activated protein kinase; SAPK, stress-activated protein kinase; PI3-K, phosphatidylinositol 3-kinase; JNK, Jun NH2-terminal kinase; phospho, phosphorylated; IR, ionizing radiation; NF-B, nuclear factor B; GFP, green fluorescent protein; TBS/T, 50 mM Tris, 150 mM NaCl, and 1% (v/v) Tween 20; REF, rat embryo fibroblast.
Received 12/26/01. Accepted 5/ 8/02.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
activates the epidermal growth factor receptor and mitogen-activated protein kinase pathway in carcinoma cells, leading to increased proliferation and protection from radiation-induced cell death. Mol. Biol. Cell, 10: 2493-2506, 1999.
radiation. Duration of JNK activation may determine cell death and proliferation. J. Biol. Chem., 271: 31929-31936, 1996. radiation in human cells. Mol. Cell. Biol., 10: 5279-5285, 1990.B. J. Clin. Investig., 88: 691-695, 1991.
B transcriptional activity, which is required for cellular transformation. J. Biol. Chem., 272: 24113-24116, 1997.B. Science (Wash. DC), 274: 784-787, 1996.B. Int. J. Radiat. Oncol. Biol. Phys., 50: 183-193, 2001.[Medline]
, RIE-vector; 
, RIE-Ras, where Ras = H-Ras(12V).
, RIE-Raf; 
, RIE-PI3-K.
