Pulmonary adenocarcinoma (PAC) is the leading type of lung cancer and has a high mortality. The tobacco carcinogen nicotine-derived nitrosamine 4-(N-methyl-N-nitrosamino)-1-(3-pyridyl)-1-butanone (NNK) stimulates the proliferation of human PAC cells and small airway epithelial cells through β-1 adrenorecptor-mediated transactivation of the epidermal growth factor receptor (EGFR).
Using the NNK hamster PAC model and human PAC tissue arrays with matched and unmatched normal lung tissues, the authors tested the hypothesis that Raf-1, an effector of the EGFR, and P-CREB, an effector of the β-adrenoreceptor, are overexpressed in a significant subset of human PACs and are early markers of PAC development. Western blots from respiratory epithelial cells and microadenomas harvested by laser-capture microdissection from hamster lungs accompanied by immunostains were used to monitor the expression levels of Raf-1 and P-CREB after 5 weeks, 10 weeks, and 20 weeks of NNK treatment. Expression levels of these markers in human PAC tissue arrays were assessed by immunostains. Reverse-phase proteomics, Western blot analysis, and immunoprecipitation in immortalized human small-airway epithelial cells and in a human PAC cell line in the presence and absence of dominant-negative Raf were used to determine Raf dependence of extracellular signal-regulated kinase 1 and 2 (ERK1/2) activation in response to NNK or isoproterenol.
The data showed a time-dependent increase in the expression of Raf-1 and P-CREB after NNK treatment in small-airway epithelial cells and microadenomas of hamsters. The majority of human lung adenocarcinomas simultaneously overexpressed Raf-1 and P-CREB. Dominant-negative Raf completely abrogated ERK1/2 activation by NNK and isoproterenol.
Adenocarcinoma is the leading histologic type of lung cancer in both smokers and nonsmokers.1, 2 This malignancy has a very high mortality because of its resistance to conventional cancer therapy and a lack of diagnostic tools for its early detection. Recent therapeutic strategies for the treatment of adenocarcinoma have focused on the use of inhibitors of tyrosine kinases,3 farnesyl transferases,4 and Raf,5 all of which are members of the epidermal growth factor (EGF) receptor (EGFR) signaling pathway. However, the potential success of such selective inhibitors of signal-transduction proteins is hampered by the absence of methods for the monitoring of hyperactive signaling pathways in the individual patient prior to and during enrollment in clinical trials. The histopathology classification of lung cancer that is used currently as a basis for assigning treatment modalities does not provide information on the expression and activity status of signaling pathways, which may vary greatly within a given histologic category of lung cancer. A first step toward the design of such diagnostic tools is the identification of hyperactive signaling proteins that are overexpressed selectively during early stages of adenocarcinoma development and in the fully developed cancer.
It is believed that the majority of human lung adenocarcinomas are derived from the lining epithelium of small airways.6 The predominating cell type in this simple respiratory epithelium is the nonciliated Clara cell, which has the ability to proliferate,7 whereas it is believed that the less numerous ciliated cells are end-stage cells that do not proliferate. In humans, this type of epithelium is found in the bronchioles, whereas it coats bronchioles as well as the more centrally located subsegmental, segmental, and lobar bronchi in the hamster and the rat.7, 8 Lung adenocarcinoma induced by the nicotine-derived carcinogenic nitrosamine 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) in Syrian golden hamsters is an excellent animal model for this human malignancy.9 Similar to the human disease, the NNK-induced adenocarcinomas in hamsters harbor activating point mutations in K-ras,10 and sequential studies during tumor development have indicated that they are derived from Clara cells.9 The mutational and tumor-initiating activities of NNK are attributed generally to interactions of reactive NNK metabolites with the DNA molecule.11 In addition, it has been demonstrated that the unmetabolized NNK binds as a high-affinity agonist to β-adrenergic receptors.12 In human lung adenocarcinoma cells with Clara cell features and in their putative cells of origin, human small-airway epithelial cells, this interaction of NNK with β-adrenoreceptors caused the activation of the transcription factor, cyclic adenosine 3′–5′-monophosphatate (cAMP) response element-binding protein (CREB), and transactivation of the EGFR pathway in a protein kinase A (PKA)-dependent manner, resulting in increased cell proliferation.13 Thus, cells under dual growth control by β-adrenergic and EGFR signaling are provided with a growth advantage over other types of lung cells in smokers and in laboratory animals exposed to NNK. This may contribute to the extreme potency of NNK as an inducer of lung adenocarcinoma in laboratory animals and to the prevalence of this cancer type in smokers. The possibility of a tumor-promoting effect of NNK/β-adrenoreceptor interaction on the development of lung adenocarcinoma also is supported by the observation that pretreatment of hamsters with an antagonist of β-adrenoreceptors inhibited NNK-induced lung carcinogenesis in this animal model.14 We recently demonstrated that the NNK-induced adenocarcinomas in hamsters simultaneously overexpressed signaling proteins associated with β-adrenergic and EGFR pathways.15 The EGFR frequently is overexpressed in human lung adenocarcinomas, and it is believed that signaling downstream of this receptor is critical for the regulation of these cancers.16 Many of the proteins that were overexpressed in the NNK-induced lung tumors in hamsters also were overexpressed in endothelia and smooth muscle cells, the functions of which are regulated by β-adrenoreceptors. By contrast, the serine-threonine kinase Raf-1 and the phosphorylated transcription factor (P-CREB) were overexpressed selectively in the induced tumors and their epithelium of origin.17 Although it is well established that Raf-1 can be activated by RAS downstream of the EGFR,5 the overexpression of this kinase in human pulmonary adenocarcinoma (PAC) has not been reported to date. In addition, to our knowledge, the potential role of the transcription factor CREB in the regulation of lung cancer has not been addressed before, and no data are available on expression levels of P-CREB in any type of human lung cancer. Therefore, using the NNK-induced PAC hamster model in parallel with human tissue arrays from PACs with matched and unmatched control lung tissues, in the current study, we tested the hypotheses that overexpressed Raf-1 and P-CREB are overexpressed simultaneously in a subset of human PACs and that the simultaneous overexpression of these proteins is an early marker of PAC development. The important role of Raf as a mediator of mitogenic signaling in human PAC and small-airway epithelial cells was confirmed by reverse phase proteomics and Western blot analyses in the immortalized human small-airway epithelial cell line HPL1D and the human PAC cell line NCI-H322 in the presence and absence of transfection with vdominant-negative RAF.
MATERIALS AND METHODS
Unless otherwise specified, all reagents were purchased from Sigma Chemical Company (St. Louis, MO). NNK was purchased from Midwest Research Institute (Kansas City, MO). Reagents for avidin/biotin immunoperoxidase staining were purchased from BioGenex (San Ramon, CA) and Vector Laboratories (Burlingame, CA).
β-Actin, a mouse monoclonal antibody, was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Raf-1 (C-12), a rabbit polyclonal antihuman antibody that detects the carboxy terminus of Raf-1 p74 protein, was purchased from Santa Cruz Biotechnology. P-CREB (serine 133), a rabbit polyclonal antihuman antibody that detects CREB when it is phosphorylated at serine 133, was obtained from Upstate Biotechnology (Lake Placid, NY). Rabbit polyclonal antihuman antibody that recognizes total CREB was purchased from Upstate Biotechnology.
The animal experiment was reviewed and approved by the Institutional Animal Care and Use Committee. Male outbred Syrian golden hamsters were purchased from Charles River (Wilmington, DE) and housed 1 animal per cage in plastic shoebox-type cages under standard laboratory conditions. The animals had free access to food (Purina Rodent Chow; Agri Feed, Knoxville, TN) and tap water. The animals were assigned randomly to 4 groups (10 hamsters per group). The hamsters in 3 groups received multiple subcutaneous injections of NNK (2.5 mg/100 g body weight 3 times per week) for 5 weeks, 10 weeks, or 20 weeks, respectively. This treatment regimen reproducibly induces the development of lung adenocarcinoma within 8 months, whereas microadenomas have developed after 20 weeks.9 Group 4 received identical injections with the vehicle of NNK (saline) for 20 weeks. At the end of each predetermined treatment period, the animals were killed by an overdose of anesthetic (Telazal). Lung tissues were fixed in 70% ethanol overnight at 4°C and embedded in paraffin. Serial 5-μm-thick sections were cut from each tissue block to be used for immunohistochemistry and laser-capture microdissection.
This technique was used to harvest cells from the simple respiratory epithelium of segmental and subsegmental bronchi, bronchioles, and microadenomas on the hamster lung slides. The lung sections were dewaxed by placing them in 2 baths of xylene for 5 minutes each and were passed through decreasing concentrations of ethanol to rehydrate (100%, 95%, 70%) and water for 30 seconds each. The tissue was stained according to the procedure outlined in the Histogene kit (Arcturus, Mountain View, CA). Briefly, the tissue was stained with Histogene staining solution, and excess stain was removed with a distilled water wash. The slides were dehydrated by successive immersion into 70% ethanol, 95% ethanol, 100% ethanol, and, finally, xylene. Then, the slides were stored in a desiccator for no longer than 30 minutes prior to cell harvesting. The PixCell IIe Laser-Capture Microdissection system (Arcturus) incorporates an Olympus IX-50 microscope that contains a microscope slide stage that is moved by a joystick. Laser power and duration settings were used to harvest cells as suggested by the instrument instruction manual. The capture spot size was 7.5 μm in diameter. To achieve this spot size, a power of 57 mV and duration of 1.3 msec were used. The number of cells captured onto CapSure caps was an approximation based on the number of laser pulses (2000 pulses per sample) that were used and the average capture of 5 cells per pulse (10,000 cells per sample). The proteins were extracted from the CapSure caps within 30 minutes after cell harvesting.
The cells were extracted from the CapSure caps by using protein extraction reagent (Pierce, Rockford, IL). Briefly, a 40-μL volume of extraction buffer was dispensed into a safe-lock 0.5-mL microcentrifuge tube (Eppendorf, Hamburg, Germany). The CapSure cap was inserted into the tube and inverted to allow the buffer to cover the captured cells on the cap. The microcentrifuge tube containing the cap was stored at −80°C until analysis of expressed protein levels by Western blotting.
Western blot analysis of cells harvested by laser-capture microdissection
After separation of the thawed cells that were extracted from the caps by centrifugation, protein was determined by using the BCA Protein Assay (Pierce). Equal amounts of protein samples were treated with loading buffer at 100°C for 5 minutes and were applied to each lane for electrophoresis in 12% polyacrylamide gel. The electrophoresed proteins were transferred onto nitrocellulose membranes in transfer buffer at 65 mV for 2 hours. After transfer, the membranes were treated with blocking buffer (5% nonfat dry milk in Tris-buffered saline/Triton-X 100 [TBST]) for 1 hour and incubated with the primary antibody overnight at 4°C. After a wash with TBST, the membranes were incubated with horseradish peroxidase-labeled secondary antibody (goat antimouse or goat antirabbit; Cell Signaling) for 1 hour. Immunoreactive bands were detected by using a chemiluminescent reaction (Enhanced chemiluminescence [ECL]; Amersham Biosciences, Piscataway, NJ) through autoradiography on Kodak Bio-Max film. Three separate Western blots were conducted for each antibody per sample and yielded similar data. Relative densities of the bands were determined by image analysis using National Institutes of Health (NIH) SCION image-analysis software. Mean values and standard errors from 5 densitometric readings per band were analyzed by 1-way analysis of variance (ANOVA) and a Tukey-Kramer multiple comparison test.
Immunohistochemistry was conducted using lung slides from the hamster experiment as well as human tissue arrays from 48 lung adenocarcinomas with self-matched, adjacent, normal tissue (human tissue array LC1002; US Biomax, Inc., Rockville, MD). The sections were deparaffinized in xylene; washed with phosphate-buffered saline (PBS), pH 7.4; and incubated with 3% hydrogen peroxide in 50% methanol for 20 minutes at room temperature. After the elimination of endogenous peroxidase activity, the sections were washed with PBS and incubated in protein block solution (normal goat serum in PBS; BioGenex) in a humid chamber for 30 minutes at 37°C. Protein block solution was used to reduce background staining caused by nonspecific binding of the secondary antibodies to the tissue. The primary antibodies were applied immediately after draining excess liquid from slide without rinse. The sections were incubated with each primary antibody in a humid chamber at 4°C overnight. Antibodies were diluted (anti-Raf-1, 1:100; anti-P-CREB, 1:50) in common antibody diluents (BioGenex). Sections exposed to diluents alone without primary antibody served as negative controls. The slides were then washed 3 times in PBS for 5 minutes each and incubated with Super Sensitive Link (biotinylated antirabbit or antimouse immunoglobulin; BioGenex) for 30 minutes at room temperature. After 3 washes for 10 minutes each with PBS, the sections were incubated with Super Sensitive Label (horseradish peroxidase-conjugated streptavidin; BioGenex) for 30 minutes at room temperature. The reaction was developed using the peroxidase substrate diaminobenzidine (Vector Laboratories). The sections were counterstained lightly with hematoxylin (Richard Allan Scientific, Kalamazoo, MI). No significant immunoreaction occurred in the negative controls.
Tissue Culture and Transient Transfection
The human lung adenocarcinoma cell line NCI-H322 was purchased from the European Collection of Cell Cultures (Health Protection Agency, Porton Down, Salisbury, Wiltshire, United Kingdom) and was maintained in RPMI-1640 culture medium (Gibco, Frederick, MD) supplemented with fetal bovine serum (10% volume/volume) at 37°C in an atmosphere of 5% CO2. The simian virus 40-immortalized human peripheral airway cell line HPLD1 (kindly provided by Dr. Takahashi; Center for Neurological Diseases and Cancer, Nogoya University Graduate School of Medicine, Nagoya, Japan) was maintained in F-12 medium (HAM; Gibco) medium supplemented with 1% fetal calf serum; 15 mM N-2-hydroxyethyl piperazine-N′-2-ethane sulphonate, pH 7.3; 5 μg/mL insulin; 5 μg/mL transferrin; 10−7 M hydrocortisone; and 2 × 10−10 M triiodo thyronine. Both cell lines were maintained in antibiotic-free medium. Assays with NCI-H322 cells were conducted in low-serum RPMI medium (0.1%), and assays with HPLD1 cells were conducted in their basal media without supplements. The dominant-negative Raf plasmid, Rafc4, was kindly provided by Dr. Silvio Gutkind (NIH, Bethesda, MD). Twenty-four hours before transfection, cells were plated at 5 × 105 cells/100-mm dish and transfected with 2 μg Rafc4 pcDNA3 using lipofectamine reagents (Invitrogen, Carlsbad, CA). Exposure of cells to NNK (10 nM) or isoproterenol (1 μM) was for 10 minutes.
Reverse-phase protein microarray assay in cultured cells
Cells were lysed in 6 M urea (Sigma Chemical Company); 65 mM dithiothrietol; 2% pharmalyte, pH 8.0 to 10.0; and 1% 3-[(3-cholamidopropyl) dimethylammonio]-1-propane sulfonate; and lysates were separated by centrifugation at 4°C for 15 minutes. Samples were spotted onto nitrocellulose-coated glass slides (Schleicher and Schuell Bioscience, Keene, NC) using a pin-in-ring format Affymetrix GMS 417 Arrayer (Affymetrix, Santa Clara, CA). Each sample was spotted in triplicate of 2-fold dilutions 1, 2, 4, 8, 16, and 32 dilution factors (Fig. 1). The arrayed slides were incubated with primary antibody for 1 hour at room temperature, washed, reincubated with secondary antibody conjugated to Alexa Fluor 680 probe, and scanned with a GenePix 4000B microarray scanner (Axon Instruments, CA). Spot images were converted to raw pixel values and analyzed by using a modified version of GenePiX 5.1 software. Background was subtracted from the intensity values.
Western Blot and immunoprecipitation in cultured cells
Equal amounts of protein (≈20 μg) were resolved on sodium dodecyl sulfate (SDS)-polyacrylamide gels (12% acrylamide), transferred to nitrocellulose membranes, and probed with the appropriate antibodies. For immunoprecipitation, cell lysates were incubated with 2 μg of the appropriate antibody and precleared with protein AG-Sepharose for 2 hours. Immune complexes were washed 5 times with 1 × immunoprecipitation buffer (50 mM Tris-base, pH 7.5; 5 mM ethylenediamine tetraacetic acid; 20 mM α-glycerol-phosphate; 150 mM NaCl; and 1% NP-40). Proteins were eluted by boiling in 2 × sample buffer, separated by SDS-polyacrylamide gel electrophoresis (12%), then transferred to nitrocellulose membrane. ECL-plus was used for detection.
Extracellular signal-regulated kinase 1 and 2 assay in cultured cells
Equal amounts of proteins were incubated overnight with 15 μL of agarose hydrazide bead-immobilized phospho-p44/42 (p44/42 microtubule-associated protein kinase assay kit; Cell Signaling). Immune complexes were washed 3 times in 100 mM Tris, pH 7.5, 1% Nonidet P-40, and 2 mM sodium orthovanadate; washed once in 100 mM Tris, pH 7.5, and 0.5 M lithium chloride; and washed once in 12.5 mM 3-morpholinopropanesulfonic acid, pH 7.5, 12.5 mM β-glycerophosphate, and 7.5 mM MgCl2. Proteins were incubated for 20 minutes at 30°C in a 30-μL kinase reaction containing 2 μg Elk-1 fusion protein (GST-Elk-1 codons 307–428), and 10 μM adenosine triphosphate. After electrophoresis on a 12% SDS-polyacrylamide gel, proteins were transferred to nitrocellulose membranes and probed with antiphospho-Elk (serine 383) antibody. ECL was used for detection.
Protein Expression in Cells Harvested by Laser-capture Microdissection
Western blot analysis (Fig. 2) of epithelial cells that were harvested by laser-capture microdissection from segmental and subsegmental bronchi and from bronchioles of control and NNK-treated hamsters showed that the expression levels of Raf-1 protein increased with increasing duration of NNK treatment (2.3-fold at Week 5, 2.6-fold at Week 10). Raf-1 protein was increased further (3.4-fold) in cells that were harvested from microadenomas after 20 weeks of NNK treatment (Fig. 2). Expression levels of P-CREB (Fig. 2) had increased even more than Raf-1 in the respiratory epithelia from these airways in NNK-treated hamsters (3.2-fold at Week 5, 3.8-fold at Week 10). Similarly, expression of P-CREB in the microadenoma cells that were harvested after 20 weeks of NNK treatment showed a 4.3-fold increase over controls. The observed increases in protein expression of Raf-1 and P-CREB were significant (P < .001; ANOVA and Tukey-Kramer multiple comparison test) compared with the controls in all of the NNK-treated groups.
Cellular Localization of Positive Immunoreactivity to RAF-1 and P-CREB
Immunohistochemical analysis of lung sections from hamsters revealed a distinct increase in positive immunoreactivity to Raf-1 of the single-layered respiratory epithelium in lobar, segmental, and subsegmental bronchi as well as bronchioles from animals that were treated for 5 weeks with NNK compared with tissues from control hamsters that had barely detectable immunoreactivity (Fig. 3). The staining reaction was cytoplasmic with nuclei that yielded negative reactions. Surrounding alveolar epithelium did not stain positive for Raf-1 (Fig. 3). The intensity of immunoreactions at these sites had increased further in hamsters that were treated for 10 (Fig. 3) or 20 weeks (not shown) with NNK. In addition, the hamsters that received NNK for 20 weeks had multiple microadenomas, all of which demonstrated intense positive immunoreactivity to Raf-1, whereas adjacent alveolar tissue remained unresponsive (Fig. 3). Positive immunoreactivity to P-CREB was undetectable in the lungs of control hamsters (Fig. 4). Distinct positive immunoreactivity with predominantly nuclear localization was observed after 5 weeks, 10 weeks (Fig. 4), and 20 weeks (data not shown) of treatment with NNK in epithelial cells of lobar, segmental, and subsegmental bronchi as well as bronchioles. Contrary to the immunoreaction to Raf-1, which was increased uniformly in all epithelial cells from these airways after NNK treatment (Fig. 3), reactivity to P-CREB was particularly prominent in foci of hyperplastic respiratory epithelium (Fig. 4). The NNK-induced microadenomas that were found after 20 weeks of NNK treatment also demonstrated strong positive immunoreactivity to P-CREB (Fig. 4). It is noteworthy that the cellular localization of P-CREB appeared to have shifted from predominantly nuclear at the earlier time points to predominantly cytoplasmic in the microadenomas (Fig. 4). Positive immunoreactivity to P-CREB was not observed in other lung cells at any of the time points investigated.
Of all 48 human lung adenocarcinoma tissues on the tissue array, 1 sample contained only necrotic tissue with no tumor cells and, thus, was excluded from the evaluation. Of the remaining 47 adenocarcinoma tissues, 43 samples demonstrated positive immunoreactivity in the epithelial tumor cells but not in connective tissue cells for Raf-1 (Fig. 5), whereas their matched normal lung tissue samples that contained alveolar and bronchial epithelia did not demonstrate such immunoreactivity (Fig. 5). The same samples that showed positive immunoreactivity to Raf-1 also selectively overexpressed P-CREB in the tumor cells (Fig. 5), whereas the 3 Raf-1-negative tumors or matched adjacent normal lung tissues yielded no positive immunoreactivity to P-CREB. Similar to the microadenomas in hamsters, immunoreactivity to P-CREB was predominantly cytoplasmic in the human lung adenocarcinomas. None of the 4 lung tissues from noncancer patients showed positive immunoreactivity to Raf-1 or P-CREB.
We showed previously that NNK activates extracellular signal-regulated kinase 1 and 2 (ERK1/2) in the CC10 Clara cell-specific, protein-positive human PAC cell line NCI-H322 and in small-airway epithelial HPL1D cells through β1-adrenergic, receptor-mediated transactivation of the EGFR. However, the potential role of Raf-1 in this signaling cascade has not been assessed to date. Using reverse-phase proteomics, Western blot analysis, and immunoprecipitation, our current data showed that NNK as well as the classic β-adrenergic agonist, isoproterenol, caused a pronounced increase in ERK1/2 activation in HPL1D cells (Fig. 6) and NCI-H322 cells (Fig. 7). These responses were abrogated completely in cells that were transfected transiently with dominant-negative Raf-1 (Figs. 1, 6). These findings support a crucial role for Raf-1 activity in the mitogenic signaling cascade initiated in both cell types through β-adrenergic and EGFR cross-talk.
Data from the current study indicate that the serine/threonine kinase RAF-1 and the phosphorylated transcription factor CREB are overexpressed simultaneously and selectively in the simple respiratory epithelium of bronchioles; in lobar, segmental, and subsegmental bronchi; and in microadenomas of hamsters prior to the development of NNK-induced PAC. These findings are in accord with reports that have identified epithelial cells in these airways as the origin of NNK-induced PACs in hamsters9 and rats.18 Moreover, these data are corroborated by the in vitro observation that NNK activated the transcription factor CREB as well as the EGFR and its downstream effectors in a PKA-dependent manner through binding to β-1-adrenoreceptors in human small-airway epithelial cells and PAC cells that expressed the Clara cell-specific CC10 protein.13 The simultaneous overexpression of Raf-1 and P-CREB in 43 of 48 investigated tissue microarrays from human PACs strongly suggests that the simultaneous activation of P-CREB and RAF-1 plays an important role in the regulation of a significant subset of human PACs. In addition, analysis by reverse-phase proteomics, Western blot, and immunoprecipitation in the presence and absence of dominant-negative Raf identified Raf as a critical mediator of the NNK or isoproterenol-induced ERK1/2 activity in HPL1D and NCI-H322 cells.
The EGFR, with its downstream effectors RAS, RAF, and the intracellular signal-regulated kinases ERK1/2, long have been recognized as important regulators of human lung adenocarcinoma.16 Surprisingly, overexpression of Raf-1 has not been reported to date in these human malignancies, even though RAF inhibitors currently are in clinical trials.5 In keeping with the documented activities of RAF in the transmission of signals from the EGFR through Ras to downstream effectors, positive immunoreactivity for this kinase was predominantly cytoplasmic. The transcription factor CREB and its upstream regulators, PKA and cAMP, have been implicated in the regulation of endocrine tumors and thyroid carcinomas.19 A potential role of CREB in lung cancer has been addressed only recently by reports from our laboratory, which indicated the phosphorylation of CREB by NNK in human small-airway epithelial cells and adenocarcinoma cells13 and the overexpression of P-CREB in NNK-induced adenocarcinomas.15 CREB is a nuclear transcription factor that is activated by PKA in response to increases in intracellular cAMP after the stimulation of G-protein-coupled receptors, such as the β-adrenoreceptors.20 The predominantly nuclear localization of P-CREB observed in the immunostained, hyperplastic foci of respiratory epithelium in bronchioles and in subsegmental, segmental, and lobar bronchi of hamsters after 5 weeks and 10 weeks of NNK treatment reflects this generally accepted function. By contrast, the predominantly cytoplasmic localization of positive immunoreactivity to P-CREB in the NNK-induced microadenomas and human adenocarcinomas is a novel finding suggestive of cytoplasmic activities of P-CREB. Studies using a heterokaryon assay to monitor cytoplasmic shuttling of nuclear transcription factors have indicated that CREB can shuttle from its classic nuclear localization to the cytoplasm.21 Even though the functional consequences of this phenomenon remain to be elucidated, it is intriguing to speculate that, in the context of the current studies, the cytoplasmic localization of P-CREB in human and hamster tumors (as opposed to nuclear localization in hyperplastic foci in the hamster model) is linked with signaling activities of this transcription factor with members of the EGFR pathway, such as RAF-1. Further studies are needed to address this important issue.
The selective overexpression of RAF-1 and P-CREB in the majority of human PACs investigated and at early stages of PAC development in NNK-treated hamsters suggests that these 2 proteins may be targeted for the development of molecular approaches for the early detection and assessment of hyperactive signaling pathways of this malignancy. Noninvasive molecular imaging with radiolabeled tracers that bind to such identified targets appears to be a promising approach. This technology currently utilizes 18F-fluorothymidine or 18-F-fluorodeoxyglucose for the detection of lung cancer by positron emission tomography.22 Although the use of these tracers allows for the detection of small neoplastic lesions, it does not provide information on the expression of regulatory signaling pathways. Noninvasive imaging of this type with tracers that bind to RAF-1 or P-CREB could offer an attractive alternative to this approach. In addition to the detection of small neoplastic and possibly even preneoplastic lesions, such imaging modalities would allow for the rational assignment of patients to therapeutics that inhibit RAF-1 and/or P-CREB. Further studies clearly are warranted to explore the feasibility of this approach.