N-Acetylcysteine Conjugate of Phenethyl Isothiocyanate Enhances Apoptosis in Growth-Stimulated Human Lung Cells Free

Isothiocyanates have been shown to have a broad chemopreventive activity against carcinogenesis in laboratory animals (1–7). Evidence obtained from epidemiologic studies also supports the protective role of dietary isothiocyanates against lung and colon cancers (8, 9). Therefore, certain isothiocyanates are potentially promising chemopreventive agents for human cancers (8, 10, 11). After administration of isothiocyanates, thiol conjugates of isothiocyanates are formed by nonenzymatic or enzymatic reaction with reduced glutathione (12). Stepwise enzymatic removal of glutamate and glycine from reduced glutathione yields l-cysteine-isothiocyanate conjugates, which are subsequently acetylated to yield N-acetyl-l-cysteine (NAC) conjugates of isothiocyanates (mercapturic acids) that are excreted in the urine (10, 12–14). NAC conjugates of some isothiocyanates also display chemopreventive activity against 4-(methylnitrosamino)-1-(pyridyl)-1-butanone (NNK)- and benzo(a)pyrene-induced lung tumors in A/J mice (15, 16), and they inhibit azoxymethane-induced colonic aberrant crypt foci formation in Fischer rats (17). Thiol conjugates of isothiocyanates have been reported to inhibit growth of human cancer cells in culture (18–20). Although the conjugates as chemopreventive agents seem to be less potent than corresponding parent isothiocyanates (16, 21), compared with the parent isothiocyanates, their slow release of isothiocyanates, and consequently, the lower toxicity, decreased pungency, and increased water solubility suggest that the thiol conjugates of isothiocyanates would be potentially more efficacious and a more desirable form for clinical trials.

Two major mechanisms have been proposed for the chemopreventive effects of isothiocyanates. One mechanism involves isothiocyanate binding and inactivation of phase I enzyme cytochrome p450s (7, 22) and induction of phase II enzymes, e.g., glutathione S-transferase, quinone reductase, and glucuronosyltransferases, possibly through activation of activator protein (AP-1; refs. 23–25). The other mechanism emerging in recent years involves the induction of apoptosis (16, 18, 20), which results in the deletion of genetically damaged cells, so that clonal expansion is aborted. Isothiocyanates and their conjugates have been shown to induce apoptosis or block proliferation in cultured cells (26–33). A/J mice treated with the N-acetylcysteine conjugate of phenethyl isothiocyanate (PEITC-NAC) had an increased apoptotic rate in the lung, which seemed to underlie the inhibition of benzo(a)pyrene-induced lung tumorigenesis by PEITC-NAC (16). Our recent study showed that PEITC, sulforaphane, and their NAC conjugates inhibit the progression of lung adenomas to adenocarcinomas in A/J mice, possibly through increasing apoptosis (34). Therefore, induction of apoptosis seems to play an important role in chemoprevention by isothiocyanates.

In various cell types, PEITC induced a sustained c-Jun NH₂-terminal kinase (JNK) activation in a dose-dependent manner, and the JNK activation caused by isothiocyanates was associated with the induction of apoptosis (27, 35). PEITC-induced apoptosis was suppressed by interfering with the JNK pathway using dominant-negative mutants of JNK1 or MEKK1, implying that the JNK pathway is required for apoptotic signaling. That isothiocyanate-induced JNK activation was blocked by the antioxidants, 2-mercaptoethanol and NAC, suggests that cell death signaling was triggered by oxidative stress. Furthermore, reactive oxygen- or nitrogen-induced AP-1 binding was associated with apoptosis in rat lung epithelial cells (36, 37). Our study showed that the induction of JNK activation and AP-1–binding activity by PEITC-NAC were accompanied by apoptosis in lung of A/J mice (16). However, AP-1 activation caused by PEITC-NAC in human lung cells has not been investigated and its relationship with apoptosis is not presently well-understood. In this study, we have shown that PEITC-NAC induces apoptosis in human lung cancer cells and established the role of AP-1 in apoptosis induced by PEITC-NAC by using c-jun-transfected and 12-O-tetradecanoyl phorbol-13-acetate (TPA)-treated human lung cells. A proapoptotic condition is predisposed by AP-1 activity in A549 cells, i.e., preexisting AP-1 activation sensitizes cells to apoptosis induced by PEITC-NAC.

Cell lines. The human lung adenocarcinoma cell line A549 was obtained from the American Type Culture Collection (Bethesda, MD), and maintained in DMEM supplemented with 10% fetal bovine serum. Normal human lung cells (small airway epithelial cells) were purchased commercially (Bio Whittaker, Inc., Walkersville, MD), and maintained in small airway epithelial growth medium provided by the same company. All cells were grown at 37°C under a 5% CO₂ atmosphere.

Reagents. PEITC was obtained from Aldrich (Milwaukee, WI) and its NAC conjugate was synthesized using the method described previously (38). Unless specified otherwise, A549 cells were treated with 10 or 25 μmol/L of PEITC-NAC. TPA was purchased from Sigma (St. Louis, MO).

SuperArray analysis. GEArray Q Series Mouse Cancer Pathway Finder Gene Arrays (SuperArray Inc., Bethesda, MD) were employed for analysis of gene expression in lung tissue of mice treated with dietary PEITC-NAC during the progression stage of tumor development. Total RNA isolated from individual mouse lung lobes was pooled according to various treatment groups. There were two pooled RNA samples used for this experiment: (a) control RNA was pooled from mice treated with benzo(a)pyrene and NNK alone, whereas (b) PEITC-NAC treated RNA was pooled from mice treated with benzo(a)pyrene, NNK plus PEITC-NAC (34). cDNA probe synthesis was done using a ³²P-RT-Labeling Kit (SuperArray) that converts pooled RNA samples into labeled probes. The hybridization and the signal detection steps were done according to the instructions provided commercially. The quantification of gene expression was obtained using a digital imaging system (ChemImager 5500, Alpha Innotech, San Leandro, CA). Software provided by SuperArray was used for data analyses.

Electrophoretic mobility shift assay. Double-stranded oligonucleotides containing the consensus-binding site for AP-1 were purchased (Santa Cruz Biotechnology Inc., Santa Cruz, CA). The oligos were labeled with γ-³²P-ATP (6,000 Ci/mmol, Amersham Pharmacia Biotech Inc., Piscataway, NJ) using polynucleotide kinase (Promega, Madison, WI) according to standard procedures. The labeled DNA was incubated with 5 μg of total proteins as specified in Results for 10 minutes at room temperature in the presence of 1 μg of poly-d(I)-d(C) oligomer (Roche Molecular Biochemicals, Indianapolis, IN) and DNA-binding buffer as described previously (39). The complexes were then separated on a 7.5% polyacrylamide gel and autoradiographed.

In supershift reactions, 5 μg of total proteins from cells treated with 25 μmol/L PEITC-NAC for 24 hours was incubated with radiolabeled probe. After 10 minutes, specific antibodies were added to the DNA-binding reaction mixtures and incubated for another 2 hours. The reaction mixtures were then separated on a 4% polyacrylamide gel containing 0.7% glycerol.

Cell cycle analysis. A549 cells were harvested 24 hours after treatment with various concentrations of PEITC-NAC by fixation in 70% ethanol. Cellular DNA was determined following staining with 1 μg/mL of 4′,6-diamidino-2′-phenylindole (DAPI, Molecular Probes, Eugene, OR) dissolved in PBS. Cellular blue fluorescence was measured using the Elite ESP flow cytometer/cell sorter (Coulter, Miami, FL) following excitation with a Ni/Cad UV light–emitting laser. The data were collected and DNA histograms were deconvoluted using Multicycle software (Phoenix Flow Systems, San Diego, CA).

Transfection. The mammalian cell transfections were done using a standard method previously described (40). Twenty micrograms of pMEX MTH-jun plasmid or the parent pMEX MTH plasmid (41) were transfected into 5 × 10⁶ A549 cells using electroporation with 230 V, 960 mF (BTX electroporation system) in 2× HEPES buffer. After transfection, cells were maintained in normal growth medium for 24 hours, followed by the addition of G418 (800 μg/mL). For selection of stable neomycin-resistant transfectants, the cells were cultured in G418 selection medium for 10 days, then maintained in medium with 400 μg/mL G418. Two transfected cell lines were generated: (a) A549/c-jun was transfected with wild-type c-jun and (b) A549/control vector was transfected with the empty vector, pMEX MTH.

Annexin V apoptotic cell staining. The cells were harvested and cyto-centrifuged on slides; slides were then stained using a Annexin V-Cy3 Apoptosis Detection Kit (Sigma), which provides Annexin V (red) and 6-carboxyfluorescein diacetate (6-CFDA, green) dual staining. Two transfected A549 cell lines, A549/control vector and A549/c-jun, were treated with 25 μmol/L PEITC-NAC for 16 hours, followed by affixation to slides and staining. The staining was done by following the manufacturer's recommendations.

Northern blot analysis. RNA extraction, electrophoresis, gel transfer to nylon membranes and blot hybridization were done as described previously (42). The blots were washed twice at 65°C in 2× SSC and 0.1% SDS for 15 minutes. The c-jun cDNA insert (1.2 kb) of the pMEX MTH-jun plasmid was excised as a probe for hybridization and labeled by incorporation of α-³²P-dCTP (6,000 Ci/mmol, Amersham) into the c-jun sequence using a Random Primed DNA Labeling Kit (Roche Molecular Biochemicals). The amount of RNA loaded was monitored using 28 S and 18 S rRNA stained by ethidium bromide.

Analysis of DNA fragmentation. To confirm the appearance of apoptotic cells, nuclear DNA fragmentation was analyzed by agarose gel electrophoresis (43). The ethanol-fixed cells were centrifuged at 800 × g for 5 minutes; the cell pellets were resuspended in 40 μL of phosphate-citrate buffer, consisting of 192 parts of 0.2 mol/L Na₂HPO₄ and 8 parts of 0.1 mol/L citric acid (pH 7.8), at room temperature for at least 30 minutes. After centrifugation at 1,000 × g for 5 minutes, the supernatant was concentrated in a Speed Vac concentrator. A 3 μL aliquot of 0.25% NP40 (v/v in H₂O) and 3 μL of RNase A solution (1 mg/mL) were added. After 30 minutes of incubation at 37°C, 3 μL of proteinase K solution (1 mg/mL) was added and the extract was incubated for an additional 30 minutes at 37°C. After adding the loading buffer, the extract was subjected to electrophoresis on 0.8% agarose gel. The presence of the characteristic “ladder” pattern of discontinuous DNA fragments was visualized by ethidium bromide staining.

Assay for single-cell DNA strand breaks. A549 cells treated with or without TPA, PEITC-NAC, or TPA plus PEITC-NAC were collected at the times specified in the text, cytocentrifuged, and stained with DAPI. Photographs were taken using an Olympus AX 70 microscope (Melville, NY) with fluorescence and SPOT RT Slider digital image system (Diagnostic Instruments Inc., Sterling Heights, MI).

Detection of cell cycle phase-specific DNA strand breaks (terminal nucleotidyl transferase–mediated nick end labeling). A549 cells were fixed with 1% formaldehyde for 15 minutes and then permeabilized by post-fixation in 70% ethanol. The presence of in situ DNA strand breaks, a characteristic feature of apoptosis, was detected by labeling with a fluorochrome-conjugated nucleotide in the reaction catalyzed by the terminal deoxynucleotidyl transferase; cellular DNA was counterstained with propidium iodide. The kit provided by Phoenix Flow Systems was used in this assay. This method is described in detail elsewhere (44). Green and red fluorescence of cells probed for DNA strand breaks and DNA content was measured using a FACScan flow cytometer (Becton Dickinson, San Jose, CA) with standard settings of green (strand breaks) and red (DNA) fluorescence detection. To calculate the percentage of apoptotic cells in respective phases of the cell cycle, DNA content frequency histograms were deconvoluted using CELLQuest software.

Protein isolation. Cells were incubated at 37°C with 5% CO₂, followed by mock treatment or by treatments specified in Results. Total proteins were then isolated using radioimmunoprecipitation assay buffer. The protease inhibitors aprotinin (1 μg/mL), leupeptin (1 μg/mL), pepstatin (1 μg/mL), phenylmethylsulfonyl fluoride (0.1 mmol/L), the phosphatase inhibitor Na₃VO₄ (1 mmol/L), and NaF (1 mmol/L) were added to all buffers. Protein concentrations were determined by means of the Coomassie Plus protein assay reagent (Pierce, Rockford, IL); aliquots of the proteins were stored at −80°C.

Western blot analysis. Western blot analysis was carried out as described previously (16). Briefly, total proteins were prepared from the A549 cells with sham treatment or PEITC-NAC stimulation. Fifty micrograms of the total proteins were boiled for 5 minutes in the presence of Laemmli sample buffer, and then separated by 10% SDS-PAGE. The proteins on the gel were transferred to nitrocellulose membranes. The anti-phospho-JNK 1/2, anti-Bax (Santa Cruz Biotechnology), anti-p53 (Oncogene Research Products, Cambridge, MA), and anti-Bcl-2 (Cell Signaling Technology, Inc., Beverly, MA) antibodies were diluted to the concentration according to commercial recommendations. Immunoreactive bands were detected with an enhanced chemiluminescence kit (Amersham). The loading amount and transfer quality were carefully monitored by Ponceau S staining of each blot (Sigma). The quantification of the gene expressions was done using a gel documentation system (ChemImager 5500, Alpha Innotech). The apoptotic index was determined by the following method: gene expression level results were normalized by dividing the densitomer reading for each band by the reading for the corresponding band from untreated A549/control vector. The apoptotic index at each treatment of the cell line is the normalized expression level of Bax divided by the normalized expression level of Bcl-2.

We reported previously that PEITC-NAC, when administrated in the diet, induced AP-1 activity in lung tissue of A/J mice at a dose that inhibited tumorigenesis in the tumor promotion stage (16). More recently, we investigated the chemopreventive activity of PEITC and sulforaphane and their NAC conjugates during the tumor progression stage in lungs of A/J mice exposed to NNK and benzo(a)pyrene (34). In an effort to determine the underlying mechanism(s) for the tumor inhibition by PEITC-NAC, mouse lung tissues from the tumor progression study were used to detect possible gene expression alterations by employing a SuperArray technique. Pooled total RNA from the carcinogen-treated group was labeled as the hybridization probe for the control blot (Fig. 1A), whereas corresponding RNA from the carcinogens plus PEITC-NAC–treated group was used as the hybridization probe for the treatment blots (Fig. 1B). When control and treatment blots were compared (for details, see Materials and Methods), gene expression analysis showed that among 96 detected genes, which are related to carcinogenesis pathways, cyclin E, c-jun, Akt, Akt2, and Bak were up-regulated, whereas CD44, Muc1, FGFR2, Myc, Plaur, CD3, cyclin D1, Bad, Atm, and Bcl-2 were down-regulated in lung tissues of PEITC-NAC-treated mice. The spectrum of alterations of gene expression in the treatment group suggests possible mechanisms for inhibition of tumor development by PEITC-NAC. Oncogenes (Myc, cyclin D1, Atm, and Bcl-2), invasion- and metastasis-related genes (Muc1 and Plaur), and the angiogenesis-related gene, FGFR2, exhibited decreased expression levels in the PEITC-NAC treatment group. Unfortunately, three genes that are very important in the apoptotic pathway, p53, JNK, and Bax were expressed at levels too low to be reliably compared by our method. Because Akt was up-regulated, it is quite probable that apoptosis induced by PEITC-NAC is not mediated through the phosphatidylinositol 3-kinase/AKT pathway. These data provide suggestive evidence for the molecular mechanism of the chemopreventive activity of isothiocyanates, even though most of the alterations seen in the SuperArray blots were not confirmed by other methods. However, a transcriptional factor, c-Jun, drew our attention, not only because its expression was increased 2.5-fold (after being normalized by expressions of housekeeping genes) by treatment of PEITC-NAC, but also because it confirmed our previous observation that the AP-1 activity was up-regulated in lung tissue of mice treated with PEITC-NAC during the tumor promotion stage (16). We observed consistently that PEITC-NAC induces c-Jun expression and AP-1 activity in vivo, which was accompanied by an increased apoptotic ratio in mouse lung tissue. Therefore, in this study, we investigated the role of c-Jun and AP-1 in PEITC-NAC-induced apoptosis in human lung cell lines because these evidences were observed in both of our animal bioassays.

PEITC-NAC induces AP-1 activity in human lung cell line A549 in a dose- and time-dependent manner. To determine whether PEITC-NAC induces AP-1 activity in human lung cells, total proteins were isolated from A549 human lung cells treated with PEITC-NAC. The binding activity of these proteins to AP-1 target sequences was assayed with electrophoretic mobility shift assay (EMSA). Figure 2, I-A shows that 24 hours after treatment, AP-1 activation was stimulated by PEITC-NAC at a concentration as low as 1 μmol/L. The peak binding activity was observed with 10 μmol/L showing a clear double band. However, further increasing concentrations of PEITC-NAC (25-100 μmol/L) reduced the AP-1 activity, which is probably attributable to cell death. AP-1 activation appeared as early as 30 minutes after the treatment with 10 μmol/L PEITC-NAC; a double band appeared after 6 hours, and the activity remained elevated up to 24 hours (Fig. 2, I-B). At the 24-hour point, the double band appeared again as shown in Fig. 2, I-A, lane “10 μmol/L,” which was under identical treatment conditions. As shown in Fig. 2, I-C, the binding activity in Fig. 1A and B was specific for the AP-1 element, as it was competed by the unlabeled AP-1 probe, but not by the nonspecific DNA fragment or the mutant AP-1 probe.

PEITC-NAC induces apoptosis in A549 cells. To examine whether PEITC-NAC causes apoptosis in human lung cells as it did in lungs of A/J mice, we did flow cytometry on the cells treated with 10 μmol/L PEITC-NAC for 24 hours. When compared with control cells (Fig. 2, II-A), a distinct sub-G₁ peak (Fig. 2, II-B, arrow) appeared; it represents 4.5% of the cells that had undergone apoptosis. This result shows that the treatment with 10 μmol/L PEITC-NAC induces apoptosis in a small fraction of cells without affecting the survival of the majority of cells. A549 cells without PEITC-NAC treatment did not show any detectable apoptosis. Although 4.5% is a small number, however, considering the dynamics, that small proportion will be a significant effect in the human body.

The dose 10 μmol/L is important, as it shows the effects of PEITC-NAC at a dose that is close to physiologic concentrations in humans. At this concentration, as shown in Fig. 2, I and II, we observed a high AP-1 binding activity, and ∼4.5% cells underwent apoptosis after 24 hours of PEITC-NAC treatment. This mild effect of PEITC-NAC on apoptosis may well reflect its chemopreventive activity, because of the large amount of cell death caused by high-dose PEITC-NAC in the culture dish, such as that shown in Fig. 3 below, is less probable in vivo. The dose-dependent curve in Fig. 2, I-A shows that the AP-1 activities were down-regulated at 50 and 100 μmol/L; this is due to the fact that at 100 μmol/L of PEITC-NAC, no lung cells appeared to survive after 24 hours. Actually, 24 hours after 50 to 100 μmol/L PEITC-NAC treatment, all of the A549 cells were detached from culture dishes. Consequently, most of these cells did not respond to AP-1 induction.

Apoptosis induced by PEITC-NAC is enhanced in c-Jun–transfected cells. To identify the constituents of the PEITC-NAC-induced AP-1 complex, supershift assays were done. Antibody to c-Jun, but not Jun B and Jun D transcription factors supershifted on AP-1 binding complex, the antibody anti-Fos family (including c-Fos, Fos B, fra-1, and fra-2) also shifted the AP-1 complex (Fig. 3, I-A), indicating that c-Jun and Fos are involved in the AP-1 complex. Because c-Jun is a component of the AP-1 binding complex induced by PEITC-NAC, to determine the relationship of c-Jun and AP-1 activity with apoptosis induction by PEITC-NAC in A549 cells, a full-length c-jun cDNA construct, pMEX MTH-jun and its empty vector pMEX MTH were transfected into A549 cells via electroporation. Twenty-four hours after transfection, the cells were selected by G418 (800 μg/mL) for 10 days to generate two cell lines: A549/control vector and A549/c-jun. Figure 3, I-B shows that the A549/c-jun cell line overexpresses c-jun mRNA (1.35 kb). AP-1 binding activity was elevated in the A549/c-jun cell line (Fig. 3, I-C).

To evaluate the role of AP-1 in apoptosis, several techniques were employed. Figure 3, II shows the morphology (A and B) and DNA fragmentation (C) of the two transfected cell lines after incubation with 25 μmol/L PEITC-NAC. Compared with the control vector, c-jun transfected cells with high AP-1 activity showed enhanced apoptosis (Fig. 3, II-B and C).

In addition, dual staining with Annexin V and 6-CFDA was done on the control vector- and c-jun–transfected cell lines to show the membrane evidence of apoptosis. Annexin V binds to phosphatidylserine moieties that become exposed on the outer surface of the cell membrane during apoptosis, whereas 6-CFDA staining serves as the marker for viable cells. This combination allows the detection of apoptotic cells (Annexin V-positive, 6-CFDA-positive), necrotic cells (Annexin V-positive, 6-CFDA-negative), and viable cells (Annexin V-negative, 6-CFDA-positive). Cells with green stain only are viable; those with red stain only (Annexin V) are necrotic cells, and those showing both red and green stains are apoptotic cells. The results showed that c-jun–transfected A549 cells showed enhanced apoptosis (Fig. 3, III-C and D) compared with the control vector–transfected cells. In the control vector transfectants, even though apoptosis had already begun 16 hours after PEITC-NAC treatment, it was not predominant (Fig. 3, III-A and B). These results indicate that in the cells with high AP-1 activity, the sensitivity to apoptosis induced by PEITC-NAC was enhanced, which is agreement with the data presented in Fig. 3, II.

Apoptosis induced by PEITC-NAC is enhanced in TPA-treated cells. A549 cells transfected with c-jun cDNA have an increased potential for apoptosis; therefore, we investigated the responses to PEITC-NAC treatment in TPA-pretreated cells that had elevated AP-1 activity. The results presented in Fig. 3, IV-A (cell morphology) or Fig. 3, IV-B (DAPI staining) were obtained from A549 cells incubated with TPA (100 nmol/L) for 12 hours prior to the addition of PEITC-NAC (25 μmol/L) for another 24 hours. A549 cells treated with TPA and subsequently with PEITC-NAC (Fig. 3, IV-A-d) had a much higher incidence of apoptosis compared with the cells treated with PEITC-NAC alone (Fig. 3, IV-A-c), whereas cells treated with control-vehicle (Fig. 3, IV-A-a) or treated with TPA alone (Fig. 3, IV-A-b) showed active growth. The cells shown in Fig. 3, IV-B, were treated identically to those in Fig. 3, IV-A, but were stained with the DNA fluorochrome DAPI. The cells showed characteristics of apoptosis, i.e., DNA strand breakage and nuclear fragmentation after PEITC-NAC or TPA plus PEITC-NAC treatment. These results clearly indicate that apoptosis induced by PEITC-NAC was enhanced in the TPA-pretreated cells.

Apoptosis induced by PEITC-NAC occurs predominantly in dividing cells. To explore the mechanism that may be involved in the enhancement of apoptosis in growth-stimulated cells, the cell cycle phase specificity in PEITC-NAC-induced apoptosis was investigated. Toward this end, A549 cells were treated with 25 or 50 μmol/L PEITC-NAC for 24 hours. DNA strand breaks during apoptosis were detected by end-labeling, which was combined with the analysis of cellular DNA content. This method allows the correlation of apoptotic cells with the specific phase of the cell cycle. Figure 3, V shows that 24 hours after treatment with 25 μmol/L PEITC-NAC, a substantial portion of the cells accumulate in the G₂-M phase. In cells treated with 50 μmol/L PEITC-NAC for 24 hours, the G₂-M fraction was significantly reduced, accompanied by a distinct population of apoptotic cells. These cells were identified by incorporation of BrdUTP into DNA strand breaks at a position consistent with those cells having originated from the G₂-M population. The slight shift to the left of the apoptotic population is a result of DNA degradation; small molecular weight DNA removed during cell washing decreased the total stainable DNA. Thus, actively growing cells, compared with cells resting in G₀ to G₁ phase, showed an increased propensity for undergoing apoptosis following exposure to PEITC-NAC.

In an effort to determine the underlying mechanism for PEITC-NAC-induced apoptosis, A549 cells transfected with c-jun or with control vector were analyzed by Western blot. The expressions of the genes JNK, p53, Bcl-2, and Bax, which are important in the apoptotic pathway, were detected. The results show that phosphorylation of JNK 1 and 2 in both transfectants were induced by 25 μmol/L PEITC-NAC treatment at 3 hours, and reduced near baseline level after 24 hours (Fig. 4, I-A). The accumulations of p53 were induced by PEITC-NAC treatment at 3 and 24 hours in these two cell lines, although A549/c-jun cells showed higher induction compared with A549/control vector cells (Fig. 4, I-A). A549/c-jun cells exhibited a higher level of p53 than control cells at 3 hours after treatment, and a lower level at 24 hours. This decrease of p53 at 24 hours in A549/c-jun cells could be caused by extensive cell death as shown in Fig. 3, II-B; or the cells which survived at 24 hours had already entered the next cycle. The differences in the expression of Bcl-2 and Bax in these two cell lines were weak, but were very constant. We observed repeatedly that the expression of Bcl-2 was increased after PEITC-NAC treatment; and increased expression of Bcl-2 was less pronounced in A549/c-jun cells than in control cells (Fig. 4, I-A and B). In contrast, both cell lines showed elevated expression levels of Bax at 3 hours and Bax remained elevated at 24 hours (Fig. 4, I-A and B). It is interesting to note that the expression of Bax was higher in untreated A549/c-jun cells than in untreated A549/control vector cells; the expression of Bax in A549/c-jun remained at high levels after PEITC-NAC treatment when compared with A549/control vector cells. On the other hand, expression of Bcl-2 level was slightly lower in untreated A549/c-jun cells than in control cells (Fig. 4, I-B), however, this deficiency became remarkable 3 hours after PEITC-NAC treatment. Because these changes were not especially dramatic, we quantified the expression of Bax and Bcl-2 from a representative gel by densitometry (Fig. 4, I-B). If we set the ratio of the expression of Bax versus Bcl-2 as the apoptotic index and then assigned the apoptotic index from untreated A549/control vector cells as “1”, the apoptotic indices in the control cells 3 and 24 hours after treatment were 1.25 and 0.48, respectively; the apoptotic indices in A549/c-jun cells: untreated, 3 and 24 hours after treatment with PEITC-NAC were 1.35, 2.18, and 0.7, respectively. This observation indicates the following: (a) the cells with c-Jun overexpression alone have tendency for apoptosis because their apoptotic index was higher than control cells in our system, although this tendency was not detectable at the cellular level. (b) At 3 hours after PEITC-NAC treatment, the cells actively underwent apoptosis since the apoptotic index was increased at 3 hours in our time course study. (c) After 24 hours, cells had recovered from the treatment. The surviving cells were the cells that had responded by induction of Bcl-2. At this time, the apoptotic index of these cells were lower than that in untreated cells, which revealed the remaining viable cells had passed through the period of apoptosis and had undergone a new cycle of cellular proliferation.

To compare the phenomenon that we observed in A549 cells with events in normal lung cells, small airway epithelial cells were used for EMSA. The results indicate that both AP-1 and p53 pathway were activated by 10 μmol/L PEITC-NAC (Fig. 4, II). AP-1 activity in small airway epithelial cells was induced within 0.5 hours after PEITC-NAC treatment, and at 6 hours reached the highest binding activity; AP-1 activity had returned to near baseline after 24 hours (Fig. 4, II-A). The time course of AP-1 activation in normal lung cells was very similar to that in A549 cells. p53 activity was strongly induced in small airway epithelial cells 0.5 hours after PEITC-NAC and further elevated at 6 hours, whereas within 24 hours this activity was decreased (Fig. 4, II-B). It was not surprising to observe that p53 activation was more dramatically induced in small airway epithelial cells than in A549 cells, because the p53 pathway is often attenuated in cancer cells. This accelerated and remarkable response to PEITC-NAC in normal lung cells may possibly be attributed to both accumulation and phosphorylation of p53.

Concordant with our previous observation in lung tissue of A/J mice treated with PEITC-NAC, in this study, we showed that PEITC-NAC induced apoptosis in human lung cancer cell line A549 and the induction of apoptosis was associated with induction of AP-1 binding activity. We observed that PEITC-NAC enhanced apoptosis in c-Jun-transfected cells; it also enhanced apoptosis in TPA-treated cells; moreover, it enhanced apoptosis in cells at G₂-M phase of the cell cycle. These observations indicated that PEITC-NAC enhanced apoptosis in cells that were growing more actively because the cells that are not actively growing would stay in G₁ phase most of the time, and actively growing cells have more opportunities to pass through the G₂-M phase, thus, PEITC-NAC-induced apoptosis has more chance to occur in actively growing cells.

c-Jun expression and AP-1 activation have been associated with apoptosis induced by various agents, such as ionizing radiation, SV40 T antigen, vitamin E succinate, and antitumor drugs, etc. (45–48). It has been reported that c-Jun expression and its NH₂-terminal phosphorylation can promote neuronal apoptosis (49–51). Experiments with sympathetic neurons cultured in vitro, as well as with cerebellar granule neurons and differentiated PC12 cells, have shown that JNK/c-Jun signaling can promote apoptosis following survival factor withdrawal. When c-Jun phosphorylation site mutants were expressed in cerebellar granule neurons, c-Jun phosphorylation has been shown to be necessary for apoptosis. c-Jun[asp], a constitutively active c-Jun mutant in which the known and potential serine and threonine phosphoacceptor sites in the transactivation domain have been mutated to aspartic acid, induces apoptosis under all conditions tested. In contrast, c-Jun[ala], which cannot be phosphorylated because the same sites have been mutated to alanine, blocks apoptosis caused by survival signal withdrawal. In nonneuronal tissues, however, the role of AP-1 activation in apoptosis was reportedly very different (52–54). In apoptosis induced by H₂O₂ and certain antitumor agents, cells transfected with the dominant-negative c-jun mutant (TAM67) exhibited delayed apoptosis and increased overall cell survival (55, 56). The studies in 3T3 mouse fibroblasts and human umbilical vein endothelial cells showed that increased c-Jun activity was sufficient to trigger apoptotic cell death (53, 56). On the other hand, anticancer drugs induced apoptosis in human acute leukemia cells without inducing of c-jun expression (57). Furthermore, it has been shown that c-Jun protects cells from UV-induced cell death and cooperates with nuclear factor κB in the prevention of apoptosis induced by tumor necrosis factor-α (54). Our observation with c-jun–transfected cells (Fig. 3, II and III), and TPA-stimulated cells (Fig. 3, IV), indicate that if the elevated AP-1 activity is a preexisting condition, either by overexpression of the c-Jun oncogene or by treatment with a tumor-promoting agent, PEITC-NAC enhances the apoptotic effect on these growth-stimulated cells.

Certain compounds derived from fruits and vegetables, such as all-trans-retinoic acid and grape seed proanthocyanidin, have antiapoptotic action that occurs via JNK and AP-1 activation (58, 59). On the other hand, selenium compounds induced apoptosis through AP-1 activation, because TAM67-transfected cells treated with selenodiglutathione showed reduced apoptosis (60). Induction of apoptosis by various isothiocyanates, including PEITC, has been previously shown to be mediated by JNK (27, 61). We observed that PEITC-NAC induced a clear stress-response in A549 cells. Stress-induced genes such as JNK (Fig. 4, I, A549/control vector, 0 and 3 hours) was stimulated after treatment with PEITC-NAC. c-Jun is one of the direct substrates of JNK, so AP-1 activity induced by PEITC-NAC is attributed to the JNK pathway. In contrast to our previous results in mouse lung tissue (16), the accumulations of p53 were elevated 3 and 24 hours after PEITC-NAC treatment in A549 cell lines. Huang et al. (30) showed that p53 is essential for PEITC-induced apoptosis in mouse epidermal cells. The present study showed that the p53 pathway may be involved in isothiocyanate-induced apoptosis in A549 cells. Because p53 phosphorylated by JNK had been described in various p53 activation systems (62–65), our hypothesis is that PEITCNAC induced JNK activity, which enhanced apoptosis, perhaps through the p53 pathway. The difference in PEITC-NAC–mediated p53 accumulation between this study and the previous study in A/J mice might be attributed to the time-dependent accumulation in cultural cells versus a continuous stimulation for p53 in animals fed a diet containing the isothiocyanate conjugates. No remarkable differences between control and c-jun–transfected cell lines for PEITC-NAC–induced JNK1/2 phosphorylation and p53 accumulation were observed, indicating that JNK and p53 are upstream or independent from effects of AP-1 activity. On the other hand, basal expression levels of Bax were increased in c-jun-transfected cells. This finding suggests that AP-1 might be involved indirectly in the regulation of Bax expression in A549 cells, and that elevated intrinsic AP-1 activation confers the cells a proapoptotic status, that is related to higher Bax expression. At 3 hours after PEITC–NAC treatment, the expression of Bax was increased in both cell lines, and the apoptotic protein was dominant compared with antiapoptotic protein in both cell lines at the time. Afterwards, in control cells (A549/control vector), the expression ratio of Bax versus Bcl-2 showed a remarkable change: the Bcl-2 level was elevated and that of Bax was reduced 24 hours after PEITC-NAC treatment. This self-regulation prevented further apoptosis in certain control cells. In AP-1-activated cells (A549/c-jun) this self-rescue regulation did not appear; 24 hours after PEITC-NAC treatment, the level of Bax was still predominant. As a result, the majority of c-jun-transfected cells underwent apoptosis.

Phorbol esters have been used to mimic the action of the second lipid messenger diacylglycerol by activating PKC. Although phorbol esters induce apoptosis in certain cell lines (66, 67), they are well known as the promoters of mitogenesis in most cells. For example, it has been reported that TPA protects HL-60 cells from taxol-induced apoptosis and blocks Fas receptor–induced apoptosis in Jurkat and U937 cells (68–70). In the present study, TPA alone had a different effect on A549 cells, as it neither induces apoptosis in these cells nor protects the cells from PEITC-NAC–induced apoptosis, but instead enhanced PEITC-NAC–induced apoptosis. The observation that PEITC-NAC enhanced the apoptotic process in TPA-promoted A549 cells, combined with the results that PEITC-NAC enhance the apoptotic process in A549/c-jun, the oncogene-transfected cells, suggests an important role for PEITC-NAC as a chemopreventive agent by selectively enhancing apoptosis in promoted cells. We are currently investigating the apoptotic pathway induced by PEITC-NAC in human cell lines. These studies will eventually provide more detailed mechanism insights to help identify molecular targets in the chemopreventive activity of isothiocyanates in humans.

Grant support: NIH grant CA46535 (to F.L. Chung) and partial support from NIH grant 1RO3CA110057-01 (to Y-M. Yang).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Drs. Xiantao Wang (Department of Oncology, Georgetown University Medical Center) for reviewing this manuscript in its final form, and M.J. Birrer and Jerome E. Groopman for providing pMM and pMM c-jun plasmid constructs. Jan Kunicki assisted in the experiments with flow cytometry. We are grateful to Dr. Wei Dai, New York Medical College, for his support during this study.