Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Lung cancer is the leading cause of cancer deaths in western countries (1). Worldwide, the number of cases of lung cancer continues to increase, despite the fact that we know the cause of the disease in the overwhelming majority of cases. The percentage of the United States adult population that smokes peaked at almost 50% in the 1960s and is still ~ 25%. Most ominously, tobacco use among high school students in 1997 was 43% (2, 3).

The mitogen-activated protein kinases (MAPKs) are a family of protein kinases that transmit signals from a variety of stimuli from the cell membrane to the nucleus. In the nucleus, these enzymes activate various transcription factors, which promote proliferative and/or inflammatory responses (4, 5). Therefore, these pathways could play a role in malignant transformation or malignant cell growth. The three relatively well-described pathways, which are interconnected, are the JNK, extracellular regulated kinase (ERK), and p38 pathways. Each cascade is composed of at least three enzymes activated in series. The ERK1/2 (p44 and p42) pathways are the best characterized of the group. ERK 1 and 2 are similar and likely functionally redundant, although they may have somewhat different substrate specificities. The ERK1/2 pathway is thought to be stimulated predominantly by growth factors, and plays an important role in cell growth and differentiation. The c-Jun N-terminal kinases (JNK1/2) were first identified by their ability to phosphorylate the c-Jun transcription factor following exposure to UV irradiation, growth factors, or expression of transforming oncogenes. A third group of mammalian MAPKs collectively known as p38 (including , , 2, /SAPK3/ ERK6, /SAPK4) are less well studied, but appear to be regulated in a fashion similar to that of JNK, except that they possess different substrate specificities (6). JNK and p38 are also called the stress-activated protein kinases (SAPK), and play an important role in stress responses. However, there is significant crosstalk between these pathways, and a single stimulus may activate more than one pathway. Other less well-studied mammalian MAPKs include ERK3, ERK5, and several other p38-related kinases (6). Some researchers have postulated that there may be more than 50 MAPKs (9). Regulation of MAPK activity occurs through phosphatases, which inactivate the enzymes by dephosphorylating them.

Activation of these signaling cascades has been noted in the malignant transformation of in vitro cell lines and chemical-induced colon cancers in vivo (10); however, the role of these signaling pathways in human cancers, although assumed, has not been demonstrated. A few studies have examined the possible association of ERK/MAPK cascade disorders with the development of human cancers (17, 18). Another study showed that a constitutively active and nuclear form of ERK2 was sufficient to induce activation of Elk-1 and AP-1, and to transform NIH3T3 cells (19). This mechanism of malignant transformation may be important in many lung cancers. A large number of non- small cell lung cancer cells (NSCLCs) overexpress epidermal growth factor receptor (EGFR) or EGFR-like receptor and/or have mutations in k-ras (20, 21). Activation of the EGFR leads to MAPK activation. Members of the ras gene family encode small guanine nucleotide-binding proteins on the inner surface of the plasma membrane that play central roles in the signal transduction pathways. Mutated Ras proteins have lost their inherent ability to become inactivated, thus stimulating growth of a number of cells in an uncontrolled fashion (22). The phosphorylation and activation of the MAPK signaling pathways, which ultimately transduce the signals to the nucleus, may mediate this effect (23).

We hypothesize that in some cancers, derangement of one or more of the MAPK pathways leads to uncontrolled proliferation. Because of their propensity for Ras mutations and EGFR overexpression, lung cancer may be particularly prone to this means of malignant transformation. In this study we measured MAPK protein levels and activity in resected non-small cell lung tumors compared with normal lung tissue from the same patients.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Tissue and Cell Lysate Preparation

Nineteen primary non-small cell lung tumors and paired normal lung tissue from the same patients were obtained from the New York University Medical Center tumor bank. Tissue samples were frozen immediately in liquid nitrogen and stored at 70°C. Tissues were homogenized on ice in a lysis buffer (0.5% NP-40, 10% glycerol, 50 mM Tris-HCL [pH 7.5], 0.3 mM sodium orthovanadate, 100 mM NaCl, 1 mM dithiothreitol (DTT), with phenylmethylsulfonyl fluoride, leupeptin, aprotinin, benzamidine, and soybean tryptin inhibitor). Whole cell lysates were prepared by scraping and pelleting cells, then resuspending the pellet in lysis buffer. Both tissue and cell lysates were then sonicated and centrifuged (20 min at 10,000 rpm). Protein concentrations in the supernatants were measured using the Bio-Rad protein determination assay (Bio-Rad, Hercules, CA).

Western Analysis

Tissue or cell lysates were mixed with 3× Laemmli sample buffer containing -mercaptoethanol and denatured by heating to 95°C for 5 min. Samples containing 50 µg of protein were then fractionated on a 10% sodium dodecyl sulfate (SDS)-polyacrylamide gel, and electroblotted to nitrocellulose membranes. Filters were blocked for 1 h at room temperature in blocking solution (phosphate-buffered saline with 0.1% Tween-20 and 5% milk), and then incubated overnight at 4°C with the various primary antibodies. The antibodies used and their isoforms specificities were ERK (specific for ERK1/2 isoforms), JNK (reacts with JNK 1, 2, and 3 isoforms), p38 (specific for p38 and p38), p-ERK (specific for phosphorylated Tyr 204 on ERK1 and 2 isoforms), p-JNK (reacts with phosphorylated Tyr 185 on JNK 1, 2, and 3 isoforms), and p-p38 (specific for phosphorylated Tyr 182 on p38, Mxi2, and p38), all from Santa Cruz Biotechnology (Santa Cruz, CA). After washing in PBST, the filters were incubated with horseradish peroxidase-conjugated anti-goat, anti-rabbit, or anti-mouse immunoglobulin G (Amersham-Pharmacia Biotech, Piscataway, NJ) for 2 h at room temperature. After washing again, detection of the immune complexes was performed using enhanced chemiluminescence (ECL; Amersham-Pharmacia Biotech) according to the manufacturer's directions. All Western blots were repeated at least three times.

Relative changes in activated MAPK expression were quantitated from the Western blot results using the NIH Image Program. To control for differences in protein loading, for each sample, we normalized the phosphorylated forms of the MAPK to total MAPK expression. Statistical analysis (matched pair t test) was performed using these values (which were obtained in triplicate). We then set the levels of the activated forms in each normal tissue to 1, and thus obtained the average relative change in expression (fold increase or decrease) of activated MAPK in each tumor compared with its matched normal tissue.

In vitro Kinase Assay for MAPK Activity

The MAPK in vitro kinase assays were performed according to the manufacturer's (New England BioLabs, Inc., Beverly, MA) instructions, except that chemiluminescent detection was performed using ECL as described above. Briefly, whole cell lysates containing 200 µg of protein were immunoprecipitated at 4°C with 4 µg of monoclonal antibody (to p38, ERK, or JNK). The immune complexes were washed three times with lysis buffer and once with kinase buffer (25 mM Tris pH 7.5, 5 mM -glycerolphosphate, 2 mM DTT, 0.1 mM Na₃VO₄, and 10 mM MgCl₂). Pellets were resuspended in 40 µl of kinase buffer containing substrate (activating transcription factor [ATF]-2-GST for p38 and JNK, and ELK-1-GST for ERK) and 1 mM adenosine triphosphate (ATP) at 30°C for 30 min. The kinase reaction was terminated by addition of SDS sample loading buffer. The samples were then heated to 95°C for 5 min and fractionated on 10% SDS-polyacrylamide gel electrophoresis gels. Following electrophoretic separation and transfer onto nitrocellulose, the blots were immunoblotted with phosphorylated ATF-2 or ELK-1 antibody. Phospho-ATF-2 or phospho-ELK-1 were detected using ECL, as described above. Again, these experiments were repeated in triplicate.

Cell Culture and Treatments

The non-small cell lung cancer cell line A549 was obtained from the American Type Culture Collection, and grown in Ham's F12K medium supplemented with 10% fetal bovine serum.

Cells were plated equally on 10-cm tissue culture dishes. After an overnight incubation to allow the cells to adhere, treated cells were placed in a hypoxic chamber (CO₂ water jacketed incubator, Model #3130; Forma Scientific, Inc., Marietta, OH) with 1% oxygen. Control cells were incubated at 20% oxygen in a standard incubator. Treatment and control cells were harvested at 30 min, 1 h, 1 d, 2 d, and 3 d.

Immunohistochemistry

Formalin-fixed, paraffin-embedded tissue sections of non-small cell lung cancer and normal lung tissues from the same patients were investigated by immunohistochemistry for activated p38 kinase. Tissue sections were deparaffinized in xylene, fixed in alcohol, and rehydrated before staining. Antigen unmasking was done in 0.01 M citric acid buffer (pH 6.0) for 10 min in a microwave. Staining was performed in a Ventanas Nexus automated immunohistochemical instrument using Ventanas' (Tucson, AZ) detection systems and TBST buffer (0.05 M Tris-HCL [pH 7.6], 0.3 M NaCl, 0.1% Tween 20). Endogenous peroxidase activity was blocked with 3% hydrogen peroxide. Slides were incubated with mouse monoclonal anti-phospho-p38 (Santa Cruz Biotechnology) overnight at room temperature. The primary antibody was detected using a standard combination of biotinylated anti-mouse and anti-rabbit secondary antibodies. This was followed by the addition of streptavidin-horseradish-peroxidase conjugate. The complex was visualized by the enzymatic reduction of 3,3' diaminobenzidine tetrahydrochloride (DAB) substrate and enhanced (darkened) with copper sulfate. Sections were counterstained with hematoxylin, then dehydrated and mounted.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

p38 Was Selectively Activated in Tumor Tissues Compared with Normal Lung Tissue

We analyzed 19 lung tumors compared with normal lung tissue from the same subjects. The samples consisted of eight adenocarcinomas, five bronchoalveolar cell, one large cell, four squamous cell carcinomas, and one metastatic colon adenocarcinoma (Table 1). Using Western analysis, we evaluated the levels of the active (phosphorylated) forms of ERK, JNK, and p38, as well as the total protein levels of these enzymes in the tissues. For both ERK and JNK there was no consistent pattern in the tumors compared with normal tissues (Figure 1). Some tumors had increased levels of active enzyme, whereas in others these levels decreased. However, when we evaluated p38, we found that in all 18 primary lung tumors the levels of active p38 were higher in the tumor tissues compared with the matched normal lung tissue. Only in the metastatic colon cancer was the level of activated p38 decreased (Figure 2). The average increase was a 2-fold increase (matched pair t test P value < 0.0001 for the difference between tumor and normal tissues). Total levels of p38 were essentially equal in all samples, confirming equal protein loading and also showing that the increase in p38 activity was not due to increased protein transcription. We found no correlation between pathologic type, tumor size, or lymph node metastases (tumor stage) and the degree of p38 activation (Table 1).

View larger version (43K): [in this window] [in a new window]   Figure 1.   Expression of p-ERK and p-JNK was variable in lung tumors. No clear pattern emerged of ERK or JNK activation in these lung tumors compared with their matched normal tissues. (A) Representative Western analysis showing the variable expression of activated (phosphorylated) ERK and JNK in lung tumors and paired normal lung tissues. Westerns were repeated three times for each sample. (B) Expression of p-ERK (black bars), p-JNK (shaded bars), ERK, and JNK was quantitated from the Western analysis using NIH-Image software. To control for differences in protein loading, we normalized the phosphorylated forms of the MAPK to total MAPK expression in each sample. We then set the levels of the activated forms in each normal tissue to 1, and thus obtained the relative change in expression of activated MAPK in each tumor compared with its matched normal tissue. This graph shows the average relative change in MAPK expression for each tumor/normal tissue pair in three experiments.

View larger version (32K): [in this window] [in a new window]   Figure 2.   Expression of activated p38 was increased in the lung tumor samples. p-p38 expression was increased in all 18 primary lung tumors compared with their matched normal tissues, and was decreased in one metastatic colon cancer. (A) Representative Western analysis showing that expression of activated (phosphorylated) p38 was increased in tumor tissues compared with normal tissues from the same patients. Total levels of p38 were relatively equal in all samples. Again, Westerns were repeated three times for each sample. (B) Quantitation of the change in p-p38 expression was done as described for p-ERK and p-JNK. The average change was a 2-fold increase in p-p38 in tumors compared with matched normal tissues. Sample number 13 was a metastatic colon adenocarcinoma. The P value reflects the matched pair t test of the differences between the tumor and normal tissues, based on the normalized quantitation.

In Vitro Kinase Reactions Confirmed that Levels of Active MAPK Correlated with Increased Enzyme Activity

For all 19 tumor-normal tissue pairs, p38, ERK, and JNK activity in the tissue lysates were measured by in vitro kinase reactions, using ATF2 as the substrate for p38 and JNK, and ELK1 as a substrate for ERK. MAPK immunoprecipitated from the tissue lysates was mixed with ATF-2-GST or ELK1-GST in the presence of ATP. We found that in every tumor-normal tissue pair, the relative levels of active MAPK correlated with the relative enzyme activitydemonstrated by levels of phosphorylated ATF-2 or ELK1 (Figure 3).

View larger version (45K): [in this window] [in a new window]   Figure 3.   Activity of p38 was also increased in tumors. In vitro kinase reactions in all the tumor/normal tissue pairs confirmed that the relative levels of activated ERK, JNK, or p38 correlated with relative phosphorylating activity. A random selection of the Western analysis of the in vitro kinase reaction for p38 is shown here. These kinase reactions were performed by immunoprecipitating p38 from equal amounts of tumor and normal tissue lysates and then adding GST-ATF-2 as a substrate in the presence of ATP, and measuring levels of phosphorylated ATF-2 in equal volumes. Relative activity of p38 correlated with relative expression of p-p38.

Immunohistochemical Analysis Confirmed that the Increased Levels of Activated p38 Were Localized to Tumor Tissues

Because our Western analyses were done on tissue homogenates that may have contained adjacent or interspersed normal cells, we performed immunohistochemical analysis staining for phosphorylated p38 to confirm the cellular localization of the activated MAPK. Activated p38, unlike ERK and JNK, can be present in both the nucleus and cytoplasm. In our immunohistochemical analysis of matched lung tumor and normal tissues, we found that there was increased staining for activated p38 in both the nucleus and cytoplasm of malignant, and dysplastic cells, compared with normal cells in the same patients (Figures 4A, 4B, and 4C).

View larger version (95K): [in this window] [in a new window]   View larger version (103K): [in this window] [in a new window]   View larger version (106K): [in this window] [in a new window]   Figure 4.   Activated p38 was localized to malignant cells. Immunohistochemical staining of non-small cell lung cancer and matched normal tissues showed increased levels of activated p38 in both the nucleus and cytoplasm of malignant and dysplastic cells compared with normal cells in the same patients. Representative slides are shown here. (A) Normal bronchial epithelium (indicated by arrows), with minimal staining for activated p38. (B) Areas of squamous metaplasia (indicated by arrow) had increased expression of p-p38. (C) Adenocarcinoma strongly stained for p-p38, particularly in the cytoplasmic compartment.

Prolonged Exposure to Hypoxia Did Not Reproduce p38 Activation in a Lung Cancer Cell Line

p38 activation in the lung tumor samples could be a secondary response to stress. Cells in solid tumors are exposed to prolonged hypoxia, and hypoxia has been shown to stimulate p38. Some studies have shown that it may be the only stressor yet identified which produces selective p38 activationwhich we saw in our tumor samples. As noted above, in our samples tumor size (perhaps a rough measure of the degree of tissue hypoxia present) did not correlate with degree of p38 activation. Unfortunately it is difficult to test, in vivo, whether prolonged hypoxia will induce p38 activation in lung tumors. Exposure to systemic hypoxia will produce additional biochemical changes that would confound the analysis. Therefore, we elected to test in an in vitro system the possibility that the p38 activation in the lung tumor samples was a stress response. We incubated human lung cancer cells (A549 cells) in a hypoxic chamber with 1% oxygen to simulate the hypoxic environment in solid tumors. Treated and control A549 cells were harvested at various time points. We lysed the cells, and performed Western analysis to determine levels of activated MAPK in treated compared with untreated cells. Total MAPK levels were used to check for equal loading. We found that, contrary to our expectations, activated p38 levels were essentially unchanged, even after prolonged exposure to hypoxia (Figure 5).

View larger version (22K): [in this window] [in a new window]   Figure 5.   Hypoxia did not activate p38 in a lung cancer cell line. A549 cells were incubated in a hypoxic chamber with 1% oxygen for various time periods, and compared with cells incubated in 20% oxygen. Western analysis of the whole cell lysates showed that levels of phosphorylated p38 did not change, even after prolonged hypoxia.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Several studies suggest that derangement of one or more of the MAPK cascades can cause malignant transformation (24). Moreover, expression of activated Ras, which acts through the MAPK pathways, can confer metastatic potential upon some cells (25). Variable results from different investigators suggest that these effects may be cell type-specific. Constitutively activated forms of MAPK kinase (MEK) (an upstream member of the ERK MAPK pathway) are sufficient for fibroblast transformation, whereas MEK activation in PC-12 cells results in proliferative arrest and differentiation (10, 26). The NIH3T3 cells transfected with constitutively active MEK1 also caused macroscopic metastases when injected into athymic mice (27). Webb and colleagues showed that although tumorigenicity may be independent of ERK activity, metastasis required its activation (14). Another study showed that a constitutively active and nuclear form of ERK2 was sufficient to induce activation of Elk-1 and AP-1, and to transform NIH3T3 cells (19). Xiao and Lang showed that oncogenic Ras induced transformation of human lung cancer cells (NCIH82), perhaps by acting through the JNK pathway (16). Finally, a recent study in a cholangiocarcinoma cell line demonstrated that the p38 pathway plays an important role in anchorage-independent growth of transformed cells (28).

Although many researchers have assumed a role for deregulation of the MAPK pathways in the development of malignancy, few studies have looked at MAPK activity in human cancers. The results of the studies available have been variable. Colon cancer has been the most studied in this respect, and several investigators have failed to find MAPK activation in colon cancers (9, 29, 30). In one study ERK3 activity was increased in about half the cancers and in gel kinase results demonstrated proteins with increased autophosphorylating activity in the cancers compared with the normal mucosa, suggesting that other, potentially unidentified, protein kinases were activated in these samples (9). Another study demonstrated active NF-B, p38, and JNK in colon polyps (31).

In gastric cancer the results have been equivocal, with one group (32) reporting decreased and another (33) finding increased ERK activity. In breast and renal cancer, two studies suggest that ERK activity is increased. Sivarman and colleagues demonstrated that ERK activation and expression were dramatically elevated in all of the breast cancers analyzed as well as in metastatic lymph nodes, but not in normal breast tissue or benign conditions such as fibrocystic disease (18). Oka and colleagues found that nine of thirteen grade 2 renal cancers had increased ERK activity (17). Finally, Loda and colleagues assessed a panel of different cancers, including colon, prostate, bladder, breast, and liver cancers, along with corresponding normal tissues, and found an apparent elevation in ERK activity in the cancers compared with the paired normal tissue (34).

These studies suggest that the role of the MAPK cascades in human cancers may be tissue- or cell type-, as well as stage-specific, and in some cases, may involve as yet unidentified MAPK pathways. None of these studies looked at lung cancers, which, given the frequency of ras and EGF-R abnormalities, may be more prone to MAPK pathway derangement. Most of these studies looked only at ERK, and only one (9) included p38 in their analysis.

Our results suggest that p38 is selectively activated in human non-small cell lung cancer, and that this activation is not secondary to the hypoxic solid tumor environment. The p38 pathway is considered one of the stress response pathways, and traditionally is thought to play a role in differentiation, growth arrest, inflammation, immune activation and apoptosis. However, there are several mechanisms by which this pathway could cause cell proliferation and transformation. p38 has multiple cytoplasmic and nuclear substrates. One of the most well described nuclear targets of the MAPK cascades is the transcription factor activator protein-1 (AP-1). AP-1 is a dimer made up of members of the Fos and Jun families, and is important for cell proliferation in many different cell types. Phorbol esters, UV radiation, tumor necrosis factor-, and serum growth factors stimulate the different signal transduction pathways that ultimately converge onto and activate AP-1. The MAPKs can modulate AP-1 transcription, both by increasing the abundance of AP-1 components (Fos and Jun) and by directly stimulating their activity (35). Activated ERK translocates to the nucleus and activates the transcription factors Elk-1 and AP-1. JNK phosphorylates and thus activates the c-Jun component predominantly. P38 modulates AP-1 activity in several different ways. p38 activates Sap1a and ATF-2, which increase c-Fos and c-Jun transcription, respectively. Activated ATF-2 can also form a heterodimer with c-Jun, in place of c-Fos, and increase the expression of AP-1-controlled genes. Thus constitutive activation of any of the known MAPK pathways, including p38, could lead to sustained AP-1 activation and contribute to malignant transformation.

Our study is the first to describe the activity of the MAPK pathways in human lung cancers. Our results indicate that MAPK activation may play a role in human lung cancers. The p38 pathway was activated in all of the non- small cell lung tumors studied. Although this pathway is traditionally associated with stress response, there are several mechanisms by which it could play a role in malignant transformation and proliferation. Further study is required to determine the significance of this selective p38 activation in these lung cancers.