Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Many studies have shown that retinoids have an important regulatory function in lung morphogenesis during development, and in the control of proliferation and differentiation of tracheobronchial epithelial cells in vivo as well as in in vitro cell systems (1). In correlative studies, vitamin A deficiency has been shown to result in hyperplasia and squamous metaplasia of the tracheobronchial epithelium (1). The latter can be reversed by the addition of vitamin A to the diet (4, 6). These studies suggest that retinoids play a critical role in controlling homeostasis in the tracheobronchial epithelium. Regulation of the rate of cellular proliferation, differentiation, and apoptosis by retinoids is likely to be an intricate part of this control mechanism (14).

Many of the biologic and molecular effects of retinoids in tracheobronchial epithelial cells involve nuclear retinoid receptors (17). The nuclear retinoid receptors comprise two subfamilies, RARs and RXRs, each consisting of three different genes (, , and  [22, 23]). The RARs and RXRs alter gene expression by binding to specific response elements, RAREs or RXREs, respectively, in the regulatory regions of target genes. However, certain actions of retinoids on gene expression mediated by the nuclear retinoid receptors do not involve RARE- or RXRE-dependent transactivation, but are due to an anti-activator protein-1 (AP-1) activity (24, 25). Although the precise mechanism of this action has yet to be elucidated, it may involve interaction of the retinoid receptors with cyclic adenosine monophosphate-responsive element binding (CREB) protein CBP or p300 (26).

Normal tracheobronchial epithelial cells, as well as lung carcinoma cells, have been reported to express both RAR- and RXR-type receptors (17, 27). Treatment of normal tracheobronchial epithelial cells with retinoic acid (RA) induces expression of the RAR receptor (17). In many lung carcinoma cell lines, the inability of RA to induce RAR appears to correlate with defects in the retinoid signaling pathway and may contribute to the malignant phenotype of these cells (17, 21).

Although the growth of many non-lung-carcinoma cell lines is inhibited by retinoids, lung carcinoma cells have been reported to be rather resistant to them (28). This may be due in part to defects in the retinoid signaling pathways detected in many lung carcinoma cells (31). However, the retinoid 4-hydroxyphenyl retinamide (4-HPR) is an exception, and has been shown to inhibit the proliferation of small-cell carcinoma cell lines but not non-small-cell carcinoma cell lines (32). This growth-inhibitory effect is not mediated by an RAR- or RXR-dependent signaling pathway, but by an unknown mechanism.

Recently, the synthetic retinoid 6-[3-(1-adamantyl)-4-hydroxyphenyl]-2-naphthalene carboxylic acid (AHPN/ CD437), which binds weakly to and selectively activates the RAR receptor, has been shown to inhibit proliferation and induce apoptosis in several mammary carcinoma cell lines (33, 34). In contrast to the growth-inhibitory effect of RA, this action was independent of the expression of the estrogen receptor, and appeared to occur through a retinoid receptor-independent mechanism (33). In this study we characterize the effect of AHPN on the proliferation of several human lung carcinoma cell lines. We demonstrate that in contrast to 4-HPR, AHPN inhibits cell proliferation of both small-cell and non-small-cell carcinoma cell lines, whereas RA, retinoids selective for RAR and RXR receptors, and a retinoid with selective anti-AP-1 activity had little effect. In addition, we show that in some cell lines the inhibition of growth by AHPN was associated with an induction of apoptosis. To obtain greater insight into the mechanism by which AHPN inhibits cell proliferation, we analyzed its effect on cell-cycle progression with flow cytometry, and related this to changes in the expression and activity of several growth-regulatory genes (retinoblastoma [Rb], cyclin-dependent kinases [cdks], p21^WAF1/Cip1 [p21], p27^Kip1 [p27], and p53). The specificity with which AHPN and several other retinoids induce growth arrest and p21 expression in lung carcinoma cells indicates that the action of AHPN is not mediated by RAR or RXR receptors, but involves a novel signaling pathway.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Cell Culture and Materials

The lung carcinoma cell lines Calu-1, Calu-6, H441, A549, H209, and H82 were obtained from the American Type Culture Collection (Rockville, MD). The cell lines H460 and H1355 were obtained from Dr. A. Gazdar (Hamon Cancer Center, University of Texas, Dallas, TX). All cell lines were grown in RPMI1640 supplemented with 10% fetal bovine serum, penicillin, and streptomycin. AHPN/ CD437, described previously (35, 36), was obtained from Dr. U. Reichert (CIRD Galderma, Valbonne, France) or from Dr. M. Dawson (SRI, Menlo Park, CA). RAR-(SRI-6751-84/TTAB, 4-[5,6,7,8-tetrahydro-5,5,8,8-tetramethyl-2-anthracenyl]benzoic acid), RXR-(SR11217, 4-[1-(5,6,7,8-tetrahydro-5,5,8,8,-tetramethyl-2-naphthalenyl)-2-methylpropenyl]benzoic acid), and anti-AP1-selective (SR11302, [E]-3-methyl-9-[2,6,6-trimethylcyclohexenyl]-7-[4-methyl-phenyl]-2,4,6,8-nonatetraenoic acid) retinoids were provided by Dr. M. Dawson (37). All-trans retinoic acid was obtained from Hoffmann-La Roche, Nutley, NJ.

Northern Blot Analysis

Total RNA was isolated from cultured cells with Tri Reagent (Sigma Chemical, Inc., St. Louis, MO) according to the manufacturer's protocol. Total RNA (30 µg) was electrophoresed through a formaldehyde 1.2% agarose gel as described (38), blotted to a Nytran Plus membrane (Schleicher & Schuell, Keene, NH), and UV-crosslinked. Hybridizations were performed for 1 to 2 h at 68°C, using QuikHyb^TM reagent (Stratagene, La Jolla, CA) and blots were washed twice with 2× standard saline citrate (SSC), 0.05% sodium dodecyl sulfate (SDS) at room temperature for 15 min. The final wash was with 0.5× SSC, 0.1% SDS at 65°C for 30 min. Autoradiography was done with Hyperfilm-MP (Amersham, Arlington Heights, IL) at 70°C, using double intensifying screens. The clonal DNA (cDNA) probes for p21 and p53 were obtained from Dr. J. Fontana (University of Maryland, Baltimore, MD) and Dr. M. Oren (Weizman Institute of Science, Rehovot, Israel), respectively. A 1.26-kb fragment of the chicken glyceraldehyde-3-phosphate dehydrogenase gene (GPDH) (39) was used as a control probe.

Western Blot Analysis

Cells were washed in phosphate-buffered saline (PBS) and then collected in sample buffer (60 mM Tris, pH 6.8; 2% SDS; 10% glycerol; 10 mM dithiothreitol [DTT]; 1 mM phenylmethylsulfonyl fluoride [PMSF]; aprotinin; and leupeptin). Immunoblot analyses were performed as described previously (40). The primary antibodies against p53 (1:500), Bax (1:1,000), and Rb (1:2,000) were purchased from SantaCruz Biotechnology (Santa Cruz, CA), p21 (1:500) from Calbiochem (La Jolla, CA) and p27 (1:250) from Transduction Laboratories (Lexington, KY). Antibody against human Bcl-2 was purchased from DAKO Corp. (Carpinteria, CA). Peroxidase-conjugated antimouse or antirabbit IgG (Chemicon, Temecula, CA) were used as secondary antibodies. Antibodies were diluted in PBS containing 1% milk powder and 0.05% Tween 20. Detection was done with horseradish peroxidase-based Supersignal Substrate reagents (Pierce, Rockford, IL) and exposure to Hyperfilm (Amersham).

Flow-cytometric Analysis

A total of 1 × 10⁵ cells/100-mm dish were plated in complete RPMI 1640 medium and treated after 2 d with 2.5 µM AHPN for 24 or 48 h. Fluorescence activated cell sorter (FACScan) analysis of DNA fragmentation was done with a variation of the nick end-labeling technique of Gavrieli and colleagues as previously described (41, 42), substituting deoxyuridine triphosphate (dUTP)-fluorescein isothiocyanate for biotinylated dUTP in the labeling reaction. Briefly, the cell pellets were washed with cold PBS and fixed with 1% paraformaldehyde in PBS for 15 min. The cell pellets were then washed again in PBS, permeabilized with 70% ethanol, washed a third time in PBS containing 1% serum albumin, and labeled. Cell samples were divided into two halves. One half was used to determine background fluorescence by omitting the terminal deoxynucleotidyltransferase enzyme (TdT) from the reaction mixture. TdT was added to the second half of the sample, and increases in fluorescence (FL1) were evaluated by FACScan analysis. The analysis was done with the Flow Cytometric Analysis Reagent Set (Boehringer Mannheim, Indianapolis, IN) containing 0.5 mg/ml propidium iodide. Clumps and doublets were excluded from the analysis by gating on forward scatter versus propidium iodide (PI) fluorescence (FL2-width).

For the analysis of BrdU incorporation and PI labeling, cells were pulse-labeled for 30 min with 10 mM BrdU, washed in ice-cold PBS, and pelleted. Cell pellets were resuspended in PBS and cells fixed in ice-cold ethanol as described previously (43). Incorporation of BrdU was measured with a fluorescein isothiocyanate (FITC)-conjugated anti-BrdU antibody (Becton Dickinson, Bedford, MA) and the Flow Cytometric Reagent Set. Flow-cytometric analysis was done with a Becton Dickinson FACSort and accompanying CellQuest software.

Electron Microscopy

Cells were collected by trypsinization, fixed in Fowler's fixative for 24 h, dehydrated, and embedded in Epon, as described previously (13). Thin sections (60 to 90 nm) were stained with 5% uranyl acetate and 2.7% lead citrate, and were examined in a Phillips 400 transmission electron microscope.

Kinase Assays

Cells were lysed in Tween 20 lysis buffer (50 mM 4-(2-hydroxyethyl)-1-piperazine-N'-2-ethanesulfonyl acid [Hepes], pH 7.5; 150 mM NaCl; 0.1% Tween 20; 1 mM ethylene diamine tetraacetic acid [EDTA]; 2.5 mM ethylene glycol-bis- (-aminoethyl ether)-N,N,N',N'-tetraacetic acid [EGTA]; 10% glycerol) supplemented with protease and phosphatase inhibitors as described (40). The kinase reactions were performed as reported previously, using the fusion protein GST-Rb as a substrate (40).

Transfection Assay

Cells (5 × 10⁵) were plated in six-well dishes and transfected the next day. The cells were transfected with 2 µg WWP-luciferase (LUC) (44) and 0.5 µg -actin-chloramphenicol acetyltransferase (CAT) using 6 µl DOSPER liposomal transfection reagent (Boehringer Mannheim) in 1 ml of serum- and antibiotic-free Opti-Minimal Essential Medium (MEM) (Life Technologies, Grand Island, NY) according to the manufacturer's protocol. Sixteen hours after transfection, cells were treated with 2.5 µM AHPN. After 24 h, cells were collected and assayed for LUC and CAT activity, using assay systems from Promega and Boehringer, respectively. The WWP-LUC plasmid, in which the LUC-reporter is under the control of a 2 kb 5'-promoter flanking region of the p21 gene, was obtained from Dr. K. Kinzler (Johns Hopkins University, Baltimore, MD).

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Inhibition of Cell Proliferation by AHPN/CD437

Treatment of the lung adenocarcinoma cell lines A549, H1355, and Calu-6; the large-cell carcinoma cell line H460; and the small-cell carcinoma cell lines H82 and H209 with 1 µM of AHPN significantly inhibited the proliferation of all cell lines (Figure 1). In addition, the proliferation of the squamous carcinoma cell line Calu-1 was also inhibited by AHPN (not shown). These observations suggest that growth inhibition by AHPN occurs independently of the type of lung carcinoma. This is in contrast to the action of the retinoid 4-HPR, which has been shown to inhibit only the growth of small-cell carcinoma cell lines (32) (Figure 1). These differences in specificity suggest that the actions of 4-HPR and AHPN are mediated by different signaling pathways. Several studies have demonstrated that many lung carcinoma cell lines are resistant to the growth-inhibitory effects of retinoic acid (28, 29). In accord with these reports, treatment of A549, H1355, Calu-6, H460, and H209 with 1 µM RA had little effect on cell proliferation, whereas H82 cells were moderately inhibited by RA (Figure 1). All cell lines were resistant to the growth-inhibitory effect of the RAR-selective retinoid TTAB, which binds to and selectively activates RAR receptors at nanomolar concentrations (19, 24) (Figure 1). In addition, the RXR-selective retinoid SR11217 (1 µM), which selectively binds to and activates RXR receptors, and the retinoid SR11302, which exhibits a selective anti-AP-1 activity (24), had little effect on the growth of these cell lines (not shown). These results suggest that the growth-inhibitory effect of AHPN is not related to either RAR- or RXR-dependent transactivation or to anti-AP-1 activity, and are in agreement with findings reported recently with mammary carcinoma cells (33).

View larger version (72K): [in this window] [in a new window]   Figure 1.   Effect of various retinoids on the growth of several human lung carcinoma cell lines. Cells were plated in six-well dishes and treated the next day with 1 µM of the indicated retinoid. After 4 d of treatment, cell numbers were determined and plotted. Control cells were treated with vehicle (0.1% dimethylsulfoxide [DMSO]). 4-HPR = 4-hydroxyphenyl retinamide; RAR = RAR-selective retinoid TTAB; RA = all trans-retinoic acid; CD437 = the retinoid 6-[3-(1-adamantyl)-4-hydroxyphenyl]-2-naphthalene carboxylic acid (AHPN). The SD was less than 5%.

The inhibition of proliferation by AHPN was concentration- and time-dependent. In most carcinoma cell lines, 1 µM AHPN was able to reduce proliferation by more than 80% (Figure 2), and proliferation was blocked within 24 h of treatment (Figure 3). Interestingly, a dramatic decrease (80 to 95%) in cell number was observed for H460 cells, whereas the cell number decreased only modestly in other cell lines over a 48-h period of treatment. The growth arrest of Calu-6 and A549 cells but not that of H460 cells was at least partly reversible, since removal of AHPN from the medium resulted in a steady increase in cell number for these cell lines (not shown). As subsequently discussed, the decrease in cell number observed for H460 cells was due to induction of apoptosis in these cells.

View larger version (10K): [in this window] [in a new window]   Figure 2.   Concentration-dependent inhibition of cell proliferation in several lung carcinoma cells by AHPN (CD437). Cells were plated as described in MATERIALS AND METHODS and treated with AHPN at the concentrations indicated. (A) Solid circles, Calu-6; open circles, H460; x, A549. (B) Solid circles, H209; open circles, H82. (C) Solid circles, H441; open circles, H1355. The SD was less than 7%.

View larger version (12K): [in this window] [in a new window]   Figure 3.   Inhibition of cell proliferation of the human lung carcinoma cell lines H460, A549, and Calu-6 by AHPN as a function of time. Cells were plated as described in MATERIALS AND METHODS and were treated with 2.5 µM AHPN (CD437). At different time intervals, cell number was determined and plotted. Solid circles, cells treated with AHPN; open circles, cells treated with vehicle (DMSO).

Effect on Cell-cycle Progression

To determine whether AHPN affects cells in a particular phase of the cell cycle, we performed flow cytometry on H460 cells pulse-labeled with BrdU at different times after the addition of AHPN. Cells were analyzed on the basis of their DNA content (PI labeling) as well as according to the extent of BrdU labeling, which served as a measure of the number of cells in the S phase of the cell cycle. As shown in Figures 4A through 4C, treatment of H460 cells with AHPN caused a rapid reduction in the number of cells in the G₂/M phase of the cell cycle. Moreover, AHPN completely blocked the incorporation of BrdU into H460 cells within 8 h of treatment (Figure 4D), indicating that AHPN prevents entry into the S phase of the cell cycle. In a second experiment (not shown), H460 cells were pulse-labeled with BrdU immediately prior to the addition of AHPN, and exit of labeled cells from the S phase was monitored over time with flow cytometry. These results demonstrated that although cells progressed slowly from the S through the G₂/M and into the G₁ phase over a 24-h period, a substantial number of BrdU-labeled cells appeared to remain in the S phase, suggesting an inhibition of their exit from the S phase. A closer examination of this potential block is in progress.

View larger version (26K): [in this window] [in a new window]   Figure 4.   AHPN inhibits the entry of lung carcinoma H460 cells into the S phase of the cell cycle. Cells were treated with 2.5 µM AHPN (CD437) for the times indicated. A 30-min pulse of BrdU was given just prior to harvest, after which cells were fixed in ethanol and stained with PI and FITC-labeled anti-BrdU. Cells were analyzed with flow cytometry as described in MATERIALS AND METHODS. (A through C). Analysis of cells based on DNA content. (D) Effect of AHPN on the percent of cells labeled by BrdU.

Inhibition of Rb Hyperphosphorylation

Progression of the cell through the G₁ phase of the cell cycle is controlled by several cdks which regulate the activity of regulatory proteins of the Rb family (45, 46). Because growth arrest is generally associated with a reduction in the hyperphosphorylation of Rb, we examined the effect of AHPN on Rb phosphorylation. As shown in Figure 5A, treatment of H460 cells with 2.5 µM AHPN significantly reduced the hyperphosphorylated (inactive) form of Rb, concurrently with an increase in the hypophosphorylated (active) form. Reduction in Rb hyperphosphorylation was also observed in A549, H441, and H1355 cells, and to a lesser extent in Calu-6 cells.

View larger version (21K): [in this window] [in a new window]   View larger version (30K): [in this window] [in a new window]   Figure 5.   Inhibition of Rb hyperphosphorylation and cdk2 and cdk4 activity by AHPN. Lung carcinoma cells were treated with 2.5 µM AHPN (CD437) or DMSO (0.1%). (A) After 48 h, cells were collected and proteins examined by Western blot analysis, using an antibody against the retinoblastoma protein Rb. (B) H460 cells were treated with AHPN, and at the times indicated were collected and lysed as described in MATERIALS AND METHODS. Cdk2 and cdk4 complexes were immunoprecipitated from the extracts and used in a kinase assay with a GST-Rb fusion protein as substrate. Phosphorylated GST-Rb was analyzed with polyacrylamide gel electrophoresis. Immunoprecipitated samples were also examined for p21 by immunoblot analysis (bottom row). P indicates preincubation of the antibody with its cognate peptide.

Cyclin-dependent kinases cdk2 and cdk4 normally phosphorylate Rb late in G₁. We next examined whether the inhibition of Rb phosphorylation could be attributed to a reduction in cdk activity. As shown in Figure 5B, treatment of H460 cells with AHPN resulted in a decrease in both cdk2 and cdk4 activity. The activity of cdk complexes has been reported to be controlled by specific cdk inhibitors such as p21 and p27, through interaction with these complexes (45, 47). As shown in Figure 6A, treatment of several lung carcinoma cell lines with 2.5 µM AHPN reduced rather than increased the level of p27 protein; however, AHPN dramatically increased the p21 protein content in H460 and A549 cells, whereas no p21 was detected in Calu-6, H1355, or H441 cells. The observed enhancement in total p21 protein in H460 cells resulted in an increase in the amount of cdk2-associated p21 (Figure 5B), and may in part have been responsible for the reduced cdk2 activity in these cells.

View larger version (34K): [in this window] [in a new window]   View larger version (43K): [in this window] [in a new window]   Figure 6.   (A) Effect of AHPN on the protein levels of p53 and the cdk inhibitors, p21 and p27, in several lung carcinoma cell lines. Cells were treated with 2.5 µM AHPN (CD437) (+) or DMSO (). After 48 h, cells were collected and proteins examined by Western blot analysis, using antibodies against p53, p21, and p27. (B) Effect of AHPN on the expression of p21 and p53 mRNA in several human lung carcinoma cell lines. Carcinoma cells were treated with 2.5 µM AHPN for 24 h. RNA was isolated and 30 µg of RNA was examined with Northern blot analysis, using radiolabeled probes for p21 and p53.

In H460 and A549 cells, the induction of p21 protein correlates well with an increase in respective messenger RNA (mRNA) (Figures 6A and 6B). In H460 cells, an increase in p21 protein was detectable between 4 h and 8 h, and was maximal at 24 h after the addition of AHPN (Figure 7A), whereas p21 mRNA increased rapidly after 2 h, reaching a maximum at 8 h (Figure 7B). That AHPN inhibited proliferation in all of the cell lines tested, even in the absence of increased p21, suggests that induction of p21 is not the only pathway by which AHPN induces growth inhibition. As was observed for the inhibition of cell proliferation, the effect on p21 mRNA expression in H460 cells was found to be specific for AHPN. Treatment with 2.5 µM RA or with the RAR-selective, RXR-selective, or anti-AP-1 selective retinoid did not cause an increase in p21 mRNA levels (Figure 8). These results are in agreement with the concept that AHPN mediates its action through a novel signaling pathway.

View larger version (40K): [in this window] [in a new window]   View larger version (35K): [in this window] [in a new window]   Figure 7.   (A) Time-dependent induction of p21 and p53 protein by AHPN in H460 cells. Cells were grown in the presence or absence of 2.5 µM AHPN (CD437), and at the indicated time intervals cells were collected and proteins (20 µg) examined with Western blot analysis for the presence of p21 and p53 protein. (B) Time-dependent induction of p21 mRNA by AHPN in H460 cells. Cells were treated with 2.5 µM AHPN (CD437) for the periods indicated. Total RNA was then isolated and examined with Northern blot analysis, using radiolabeled probes for p21 and p53 protein.

View larger version (36K): [in this window] [in a new window]   Figure 8.   The induction of p21 mRNA in H460 cells is specific for AHPN (CD437). H460 cells were treated with various retinoids (2.5 µM) for 8 h. RNA was isolated and examined with Northern blot analysis, using a radiolabeled probe for p21 or glyceraldehyde-3-phosphate dehydrogenase (GPDH). RARsel = RAR- selective retinoid TTAB; RXRsel = RXR-selective retinoid SR11217; Anti-AP1sel = anti-AP1-selective retinoid SR11302.

p21 can be regulated by p53-dependent and -independent mechanisms (44, 48, 49). The p53-dependent regulation of the p21 promoter is mediated by a p53-responsive element in the 5'-promoter flanking region of the p21 gene (44, 49). We therefore examined whether the increase in the expression of p21 induced by AHPN was related to alterations in the expression of p53 mRNA and/or p53 protein levels. As shown in Figure 6, AHPN caused an increase in the levels of p53 protein in only two cell lines, H460 and A549, in which it also induced p21 levels. In H460 cells this increase could be observed after 2 h of treatment, and reached a maximum at 16 h (Figure 7A). The increase in p53 protein in H460 and A549 cells was not associated with increased levels of p53 mRNA (Figures 6 and 7), suggesting that the increase in p53 protein is controlled by post-transcriptional mechanisms. Since the induction of p21 expression correlated well with the increase in p53 proein (Figures 6 and 7), we investigated whether the increase in p21 was controlled at a transcriptional level by a p53- dependent mechanism, by examining the effect of AHPN on the transactivating activity of a 2-kb 5'-promoter flanking region of the p21 gene that contained the p53-responsive element. The results suggested that p21 is regulated by both transcriptional and post-transcriptional mechanisms, since a 2- to 3-fold increase in transcriptional activation of a CAT reporter was observed in H460 cells, whereas A549 cells exhibited a slight decrease in this activation (Figure 9). The involvement of post-transcriptional mechanisms is in agreement with findings obtained in mammary carcinoma cells, which demonstrated that the increase in p21 mRNA induced by AHPN is related to increased mRNA stability (34). Although p53 could be involved in the inhibition of cell growth in H460 and A549 cells, our observations indicate that different mechanisms are operative in other cell lines. H441 cells contained high levels of p53 protein but low levels of p53 mRNA, neither of which was changed after cells were treated with AHPN, whereas Calu-6 and H1355 cells expressed p53 mRNA but did not contain any p53 protein. Only H460 and A549 cells express wild-type (wt) p53, whereas H441, Calu-6, and H1355 cells have been reported to contain mutated p53 (50). It is interesting to note that p21 is induced only in those lung carcinoma cells that contain wt-p53 protein. This correlation could support a role for p53 in the control of p21 in these cells. A summary of these results is shown in Table 1.

View larger version (21K): [in this window] [in a new window]   Figure 9.   Transcriptional and post-transcriptional control of p21 expression in H460 and A549 cells. Cells were transfected with the p21 promoter-luciferase reporter construct WWP-LUC and with -actin-CAT, and 16 h later were treated with 2.5 µM AHPN (CD437). After 24 h, cells were assayed for LUC and CAT activity. The relative LUC activity was calculated and plotted.

Induction of Apoptosis

In order to determine whether AHPN induced apoptosis, DNA from H460, Calu-6, and A549 carcinoma cells was isolated and analyzed electrophoretically for DNA fragmentation (data not shown). However, no DNA ladder could be detected. It has been reported in other cell systems that the formation of a DNA ladder does not always accompany apoptosis (53, 54). Therefore, we analyzed DNA fragmentation by the TdT/dUTP-fluorescein isothiocyanate nick end-labeling technique, using FACScan analysis (41, 42). Background fluorescence was determined by omitting TdT from one half of the cell sample. The results demonstrated that within 24 h after the addition of AHPN, a substantial number (89%) of H460 cells undergo apoptosis (Figure 10). The percentage of apoptotic cells observed in AHPN-treated Calu-6 and A549 cells was very small (< 5%) (not shown), suggesting that in these two cell lines the induction of apoptosis is not as efficient, and may not be a major part of the growth-inhibitory effects of AHPN. This latter possibility is in agreement with the results presented in Figure 3, which show that AHPN did not cause as great a reduction in cell number in Calu-6 and A549 cells as in H460 cells. The induction of apoptosis in H460 cells was confirmed by the appearance of cells with an apoptotic phenotype in the AHPN-treated samples as observed with transmission electron microscopy. Figure 11B shows the morphology of a typical apoptotic cell in the early stages of apoptosis. In contrast to the morphology of a control cell, most of the chromatin is aggregated in large compact granular masses that abut on the nuclear membrane.

View larger version (29K): [in this window] [in a new window]   Figure 10.   FACScan analysis of DNA fragmentation induced by AHPN in H460 cells. Cells were collected by trypsinization, fixed, and examined with the FACScan system for apoptosis, using the nick end-labeling technique as described in MATERIALS AND METHODS. Background fluorescence was quantitated with samples in which the terminal transferase enzyme was omitted (TdT) (top and bottom left panel). Increased fluorescence above background is indicative of TdT-uridine nucleotide end-labeling technique (TUNEL)-labeled apoptotic cells. Cells treated with 2.5 µM AHPN (CD437) for 24 h were determined to undergo apoptosis (lower right panel ).

View larger version (92K): [in this window] [in a new window]   Figure 11.   Electron micrograph of a typical H460 cell from untreated culture (A) and a H460 cell at an early stage of apoptosis from an AHPN-treated culture (B). Note the condensation of the nuclear chromatin into a compact mass that abuts the nuclear membrane.

Bcl-2 and Bax Expression

An increasing number of proteins are being reported that either inhibit or promote apoptosis. Bcl-2 is one of several proteins that have been shown to inhibit apoptosis, whereas Bax belongs to a class of proteins that can counteract this inhibitory effect (54, 55). Figure 12 shows that two of the cell lines used in the present study, H460 and A549, express Bcl-2, whereas Bcl-2 protein was essentially not detectable in H441, H1355, and Calu-6 cells. AHPN had no effect on the level of Bcl-2 in these cells. Bax was detected in all cell lines and its level was slightly decreased by AHPN treatment in several cell lines.

View larger version (27K): [in this window] [in a new window]   Figure 12.   Expression of Bcl-2 and Bax in several lung carcinoma cell lines. Cells were grown in the presence or absence of 2.5 µM AHPN (CD437). After 48 h, cells were collected and protein extracts (20 µg) examined with Western Blot analysis, using antibodies against Bcl-2 and Bax as described in MATERIALS AND METHODS.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

In this study we demonstrated that the novel retinoid AHPN/CD437 causes growth arrest and apoptosis in several lung carcinoma cell lines. The inhibition of cell proliferation was not restricted to a specific type of lung tumor, since small-cell carcinoma, adenocarcinoma as well as squamous carcinoma cell lines were inhibited. The effects of retinoids on the growth of human lung cancer cell lines have been examined by several investigators (28, 32). These studies have shown that most lung carcinoma cell lines are rather resistant to the growth-inhibitory effects of RA (28, 29). Recently, the retinoid 4-HPR has been reported to inhibit growth and induce apoptosis in several small-cell lung carcinoma cell lines, whereas non-small-cell lung carcinomas are refractory to it (32) (Figure 1). The differential effects of 4-HPR and AHPN on cell proliferation suggest that these two agents act through distinct signaling pathways.

To obtain insight into the mechanisms by which AHPN inhibits the growth of lung carcinoma cells, we compared the action of AHPN with that of several other retinoids. Previous studies have shown that AHPN selectively binds the RAR receptor and induces RARE-dependent transactivation, although its transactivating activity is significantly less than that of RA (33, 35). In addition, AHPN does not bind to RXR receptors, nor does it cause transactivation through RXRE. Examination of the growth-inhibitory effect of several of these retinoids in lung carcinoma cell lines indicates a high specificity for the action of AHPN. In contrast to AHPN, neither RA nor the RAR-selective retinoid TTAB or the RXR-selective retinoid SR12117 was able to inhibit the growth of most lung carcinoma cells or induce p21 expression. These retinoids bind RAR or RXR receptors with high affinity, and are more effective in inducing RARE- or RXRE-dependent transactivation than is AHPN (19, 33). TTAB, which acts as an RAR pan-agonist and can activate RARs at nanomolar concentrations, has no effect on the proliferation of lung carcinoma cells even at 1 µM. TTAB and SR12117 also exhibit a stronger anti-AP-1 activity than does AHPN (24, 33). These results suggest that the action of AHPN does not relate to its ability to induce RARE- or RXRE-dependent transactivation, or to its anti-AP-1 activity. This last concept is confirmed by our experiments showing that SR11302, which exhibits only anti-AP-1 activity, had no growth-inhibitory effect (not shown) and did not induce p21 in lung carcinoma cells. These results are in agreement with those in a recent report that the growth-inhibitory action of AHPN in mammary carcinoma cells also occurs independently of the retinoid receptors (33). That growth-inhibitory action of AHPN in mammary carcinoma cell lines is independent of the expression of the estrogen receptor, in contrast to the effect of RA, further supports the involvement of a novel signaling pathway in AHPN action (33). One possible mechanism by which AHPN may act is through binding and transactivation of either a new nuclear receptor or a known orphan receptor. Through differential display, we have been able to identify several genes that are rapidly induced after treatment with AHPN (M. Sakaue and A. M. Jetten, unpublished observations). These genes may be good candidates for direct regulation by AHPN, and should facilitate elucidation of the signaling pathway induced by AHPN.

Analysis of BrdU incorporation demonstrated that AHPN blocks the transition of H460 cells from G₁ to the S-phase of the cell cycle within 8 h of treatment. Members of the Rb family, including Rb, p107, and p130, play important roles in the control of progression through this stage of the cell cycle (45). In early G₁, hypophosphorylated Rb is in complex with the transcription factor E2F, thereby inactivating E2F. Phosphorylation of Rb in mid- to late G₁ by cdk:cyclin complexes results in dissociation of the Rb:E2F complex and allows E2F to activate transcription of several genes, such as those for cyclin A and thymidine kinase (tk), and c-myc, which are required for progression through late G₁ and into the S phase of the cell cycle (45, 46). Our results demonstrate that treatment with AHPN results in an inhibition of Rb hyperphosphorylation in several lung carcinoma cell lines. In addition, we show that AHPN treatment causes a reduction in cdk2 and cdk4 activity in H460 cells. This inhibition may be part of the mechanism by which AHPN causes G₁ growth arrest in these cells. The activity of cyclin(:)cdk complexes is controlled by several cdk inhibitors, including p21 and p27 (45, 46). AHPN treatment greatly increased the level of p21 protein in H460 and A549 cells but not in H441, H1355, or Calu-6 cells. Moreover, an increase in the amount of p21 protein associated with cdk2 was observed in H460 cells. It is likely that the reduction in cdk activity in H460 cells is at least in part due to the observed increase in p21, and is responsible for the inhibition of Rb phosphorylation. Different mechanisms (e.g., participation of other cdk inhibitors), not tested in this study, appear to be involved in the AHPN-induced growth arrest of other lung carcinoma cell lines.

Our results show that AHPN increased p53 protein levels only in H460 and A549 cells, which contain wt-p53, and that the increase in wt-p53 protein correlates well with the induction of p21 mRNA and protein (Table ). The enhancement in p53 was not associated with increased p53 mRNA levels, suggesting that the induction of p53 protein in these two cell lines is controlled at a post-transcriptional level. The expression of p21 can be regulated through p53-dependent and p53-independent mechanisms (44, 48, 49). The increase in p21 mRNA appears to be controlled largely by a both transcriptional and post-transcriptional mechanism, since AHPN had little or no effect on the transactivation of the CAT reporter placed under the control of the 2-kb 5'-promoter flanking region of the p21 gene. This conclusion is in agreement with recent observations on the growth-inhibitory effects of AHPN in mammary carcinoma cells (34), which demonstrated that AHPN increases the stability of p21 mRNA. It is interesting to note that the increase in p21 mRNA preceded that of the protein by several hours, which may indicate that the regulation of p21 protein synthesis is more complex and could involve control mechanisms at the translational level.

We provide evidence that AHPN induces apoptosis in H460 cells, as indicated by morphologic criteria and FACScan analysis of DNA fragmentation with the nick end-labeling technique. Apoptosis is a complex, multistage process involving many genes (20, 54, 55). p53 is a multifunctional protein that can modulate genes involved in the control of cell growth and apoptosis (56). However, the induction of growth arrest and apoptosis by AHPN does not require p53, since several lung carcinoma and mammary cell lines that do not express p53 are growth-arrested and undergo apoptosis after the addition of AHPN (33). To begin to understand the mechanism by which AHPN induces apoptosis, we examined its effect on two other proteins that can control apoptosis, Bcl-2 and Bax. The protooncogene Bcl-2, which encodes a mitochondrial protein, has been found to inhibit apoptosis (20, 54, 55). Bcl-2 forms heterodimers with a number of other proteins, including Bax, that have been shown to promote apoptosis. It has been suggested that it is not the level of Bcl-2 per se, but the ratio between members of this protein family, that is a critical factor in regulating apoptosis (54, 55). Our results show that AHPN has no affect on the level of Bcl-2 or Bax protein, suggesting that the induction of apoptosis by AHPN does not involve a downregulation of Bcl-2 or upregulation of Bax. In addition, the difference in sensitivity between the cell lines to CD437-induced apoptosis appears to be unrelated to the level of Bcl-2 protein. Studies are in progress to elucidate the signaling pathway involved in the induction of apoptosis by AHPN.

Our results demonstrate that AHPN induces growth arrest and apoptosis in a variety of lung carcinoma cells. The specificity with which AHPN and other retinoids induce growth arrest and p21 expression indicates that the action of AHPN is not mediated by RAR or RXR receptors, but involves a novel signaling pathway. Elucidation of this signaling pathway may lead to the development of new analogues of AHPN with a greater efficacy and which could be useful in the treatment of lung cancer.