Mitochondrial Permeability Transition Is a Central Coordinating Event in N-(4-Hydroxyphenyl)retinamide-induced Apoptosis¹

4HPR³ is a synthetic analogue of vitamin A belonging to a growing family of compounds known as retinoids. 4HPR has shown efficacy as an antineoplastic agent in experimental models and clinical trials. In animal models, 4HPR can inhibit carcinogenesis in breast, bladder,lung, ovary, and prostate (reviewed in Refs. 1 and 2). Various clinical trials involving chemoprevention of cancers of the breast, prostate, cervix, skin, and lung have also been conducted (reviewed in Refs. 1 and 3). In skin, topical administration of 4HPR promoted regression of precancerous skin lesions (actinic keratosis) after short-term treatment. However, this was not sustainable after discontinuation of therapy (4). Nonetheless, this clinical trial demonstrated that 4HPR has a potential application for treating cutaneous neoplasms and points to the need for additional investigation of the biological activity of this compound. The skin is especially suitable for chemopreventive and chemotherapeutic interventions because cutaneous precancers and cancers can be directly targeted by topical administration of the chemical agent of choice, thereby avoiding the possibility of systemic toxicity.

The finding that 4HPR can promote apoptosis in tumor cell lines implies a possible common cellular event that may be important with respect to both the chemopreventive and therapeutic effects of this compound (5, 6, 7, 8). The mechanism through which 4HPR induces apoptosis is not well understood. Retinoids are believed to act via nuclear receptor-mediated transactivation of target genes by retinoic acid receptors α, β, and γ and retinoid X receptors α, β, and γ (9). In vitro studies have shown that 4HPR can induce the transcription of retinoic acid response elements by retinoic acid receptors, particularly retinoic acid receptor γ (10, 11). Yet, the ability of 4HPR to induce apoptosis in cell lines refractory to ATRA (12, 13, 14), the natural ligand for nuclear retinoic acid receptors (9), points to the possibility that 4HPR may trigger apoptosis in a receptor-independent manner (6). This is supported by recent findings that 4HPR acts as a prooxidant in leukemia cell lines and in carcinoma cell lines derived from the cervix and prostate (5, 7, 8). Work with these cell lines has shown that 4HPR can enhance the production of ROS,and its ability to induce apoptosis can be inhibited by exogenous antioxidants like ascorbic acid, pyrrolidine dithiocarbamate, and butylated hydroxyanisole. The ability of 4HPR to promote ROS does not appear to be receptor-mediated (5, 7).

This study was conducted to determine whether 4HPR could promote apoptosis in human SCC cells. Furthermore, ways to modulate the effects of 4HPR in SCC cells were investigated to gain a clearer understanding of the way in which 4HPR functions at the cellular level. Data are presented that further substantiate the possibility that 4HPR can operate in a novel pathway with respect to therapeutic effectiveness (15), supporting its usefulness as a treatment for skin neoplasms and perhaps for other hyperproliferative cutaneous disorders.

The SCC-13 cell line was derived from a biopsy of a primary cutaneous SCC (Ref. 16; a kind gift from Dr. Anton Jetten, National Institute of Environmental Health Sciences, Research Triangle Park,NC). The COLO 16 cell line was derived from a metastatic lesion in a female patient who succumbed to metastatic SCC (17). SRB-1 and SRB-12 were derived from biopsies of primary SCC from patients at the University of Texas M. D. Anderson Cancer Center. COLO 16, SRB-1,and SRB-12 cell lines were a kind gift from Dr. Janet Price (Department of Cell Biology, M. D. Anderson Cancer Center). COLO 16 and SRB-1 were subcloned in agarose and SRB-12 was subcloned in Matrigel as a way of selecting clones with more transformed (anchorage-independent)phenotypes.

4HPR was obtained from Dr. Ronald Lubet (Division of Cancer Prevention and Control, National Cancer Institute, Bethesda, MD). ATRA and 9-cis RA were obtained from Dr. Werner Bollag (F. Hoffman-La Roche, Basel, Switzerland). The retinoic acid receptor β, γantagonist CD2665, and retinoic acid receptor γ agonist CD437 (18, 19) were obtained from Dr. Braham Shroot (Galderma Research and Development, Sophia Antipolis, France).

Vit-C, CsA, DMSO, and hydrogen peroxide 30% solution were purchased from Sigma Chemical Co. (St. Louis, MO). R123 and 2′,7′-dichlorofluorescin diacetate were purchased from Molecular Probes, Inc. (Eugene, OR). Cyp was purchased from Calbiochem (La Jolla,CA). Retinoids, CsA, R123, Cyp, and 2′,7′-dichlorofluorescin diacetate were dissolved in DMSO. Vit-C was dissolved in sterile deionized water,and hydrogen peroxide 30% solution was diluted in sterile deionized water. All solutions were stored at −20°C before use. BA, ∼30 mg/ml in 2 m NH₄OH, was purified as described (20) and kindly provided by Dr. J. A. Duine(Delft University of Technology, Delft, The Netherlands).

For experimental manipulation, SCC cells were cultured in keratinocyte growth medium consisting of keratinocyte basal medium supplemented with 100 ng/ml human recombinant epidermal growth factor and 0.4% bovine pituitary extract (BioWhittaker/Clonetics, San Diego, CA). The calcium concentration in the keratinocyte growth medium was 0.15 mm. Cell cultures were incubated at 37°C in humidified air containing 5% CO₂.

Proliferation rates for SCC cells were obtained by seeding 400,000 cells in 10-cm diameter dishes. After allowing the cells to attach and proliferate for 24 h, cells were detached after a brief incubation with 0.025% trypsin/0.01% EDTA at 6, 12, 24, 36, or 48 h later and counted with a hemacytometer. The doubling times obtained for the SCC cell lines were COLO 16, 23.9 h; SCC-13, 24.5 h; SRB-1,38.6 h; and SRB-12, 26.9 h. Treatment with retinoids and other agents was conducted on ∼50% confluent cultures.

For the proliferation inhibition study with 4HPR, SCC cells were seeded in 10-cm dishes (750,000 cells/dish). After seeding, the cells were allowed to attach and proliferate for 24 h (COLO 16, SCC-13, and SRB-12) or 36 h (SRB-1) before treatment with 10 μm4HPR. In addition, SCC-13 cells also received 1 and 5 μm4HPR for characterization of doseresponse relationships. All control dishes received the same amount of the vehicle, DMSO, as the retinoid-treated cultures. After 6, 12, 24, or 48 h, both the floating and the attached cells were harvested and counted with a hemacytometer. The trypan blue exclusion assay indicated that viability exceeded 90% in both control and treatment populations at 6 and 12 h.

Detection of intranucleosomal DNA fragmentation was performed using a flow cytometry apoptosis detection kit (Phoenix Flow Systems, Inc., San Diego, CA), which is based on the TUNEL technique and labels the 3′-hydroxyl termini of DNA fragmented during apoptosis (21). Cell samples were also stained with propidium iodide to indicate DNA content as a relative indicator of cell cycle progression. SCC cells were treated with 4HPR and harvested at the respective time points after the procedure described above for the proliferation inhibition study.

Cells were fixed and stained using the protocol provided in the apoptosis detection kit, with limited modifications. The first modification involved additional permeabilization of fixed cells in 1 ml of 1% Triton X-100 (Sigma Chemical Co.) in deionized water for 10 min on ice before rinsing and suspension in 70% (v/v)ethanol/deionized water. The second modification involved passing the cells slowly three times through a 1-ml syringe fitted with a 25-gauge needle to disperse cell clumps before incubation in the TUNEL reaction solution. The third modification required an overnight incubation at room temperature in the TUNEL reaction solution.

Flow cytometric analysis was conducted using a Coulter EPICS Profile II flow cytometer (Coulter Corp., Miami, FL). Approximately 10,000 events(cells) were evaluated for each sample. Gating of control nonapoptotic populations (cells treated with DMSO) was used as a reference to compare with treatments with 4HPR. An internal control (HL-60 cells treated with camptothecin to induce apoptosis) provided in the apoptosis detection kit was also used to insure the TUNEL reaction was occurring during the staining procedure.

The flow cytometer detected fluorescence at 623 nm for propidium iodide and at 520 nm for fluorescein using an excitation wavelength of 488 nm provided by an argon laser. These were recorded as single-parameter histograms along the X-axis as an indication of DNA content (propidium iodide) and DNA strand breaks (fluorescein-labeled dUTP incorporation),relative to population frequency. The appearance of a sub-G₁ population in the propidium iodide histogram was used to provide a second indicator of DNA degradation (22). The fluorescence values for each probe were also recorded as a dual-parameter scatter diagram with DNA content (linear propidium iodide fluorescence) along the X axis and fluorescein-labeled dUTP incorporation (log fluorescein fluorescence) on the Y axis. This diagram provided the option of determining the phase of the cell cycle,if any, during which the cells were most sensitive to apoptosis induction after treatment (23).

Generation of intracellular ROS was measured using 2′,7′-dichlorofluorescin diacetate. This nonpolar compound is taken up by cells and converted to the nonfluorescent derivative DCF by cellular esterases. DCF is membrane impermeable, localizing in the cytosol where it can be oxidized to the fluorescent compound 2′,7′-dichlorofluorescein by reactions with ROS, primarily hydroperoxides, via cellular peroxidases (24). SCC cells were seeded in six-well tissue culture plates (90,000 cells/2 ml/well)and allowed to reach 50% confluence. The wells were washed twice with 2 ml of buffer A [Krebs-Ringer buffer containing 10 mmd-glucose, 120 mm NaCl, 4.5 mm KCl,0.15 mm CaCl₂, 0.7 mmNa₂HPO₄, 1.5 mmNaH₂PO₄, and 0.5 mm MgCl₂ (pH 7.4) at 37°C]. The wells were then covered with 2 ml of buffer A containing 10 μg/ml 2′,7′-dichlorofluorescin diacetate and the appropriate concentration of the agent of choice. The plates were rocked for 2 min to insure adequate mixing. Fluorescence emission at 530 nm (representing oxidation of DCF) after excitation at 485 nm was measured at time 0(immediately after mixing) and subsequently at 30-min intervals over a 150-min period using a CytoFluor 4000 spectrofluorimeter (Perseptive Biosystems, Inc. Framingham, MA). Cultures were incubated at 37°C between the 30-min intervals for fluorescence determination. Examination of the fluorescence characteristics of cells treated in buffer A containing 4HPR without 2′,7′-dichlorofluorescin diacetate present, and buffer A containing 4HPR and 2′,7′-dichlorofluorescin diacetate without cells present in a six-well tissue culture plate revealed a signal of ∼2 fluorescence units at time 0, which remained unchanged during the exposure period described above (not shown).

To determine whether the effects of 4HPR could be modulated in SCC cells, the treatment effects of other agents were examined. As an example of the treatment procedure used for the TUNEL analysis, COLO 16 cells were seeded in 10-cm tissue culture plates and cultured 24 h. Two h before treatment with 4HPR, CD2665 was added directly to the culture medium to give final concentrations of 10 μm. Vit-C was diluted to 500 μm in keratinocyte growth medium at the initial seeding for a 24-h exposure. 4HPR was added directly to the medium at a final concentration of 10 μm for a 12-h exposure. CsA and Cyp were added to the culture medium of SCC-13 cells at the same time as 4HPR. BA was added to the culture medium of SCC-13 cells 2 h before 4HPR. After the appropriate exposure period, the cells were harvested for TUNEL staining as described previously. The same treatment procedures were used for ROS andΔψ _m determinations.

Quantitation of Δψ_m was determined by R123 retention. R123 is a cationic fluorescent dye, which localizes in the mitochondria of viable cells because of the relatively high negative electric potential across the mitochondrial inner membrane (25). SCC cells were seeded in six-well tissue culture plates and cultured as described for the ROS determination. Treatment with 4HPR and other agents followed the protocol describe above. The cultures were incubated at 37°C for the appropriate time, after which R123 was added to each well to a final concentration of 1 μg/ml. The plates were incubated an additional 15 min at 37°C. The media was removed from each well, which was then washed once with 2 ml of buffer A, and replaced with 2 ml of buffer A. Fluorescence emission at 530 nm(representing R123 retention as a function ofΔψ _m) after excitation at 508 nm was measured with a CytoFluor 4000 spectrofluorimeter. As an internal control, the mitochondrial uncoupling agent 2,4-dinitrophenol (26) at a concentration of 100 μm was used to diminishΔψ _m. In COLO 16 cells, this treatment resulted in ∼60% reduction of R123 fluorescence relative to control after a 2-h exposure (not shown).

4HPR at a concentration of 10 μm promoted a time-dependent inhibition of proliferation in SCC cells (Fig. 1 A). The treatment concentration of 4HPR was determined empirically based on achieving maximum reduction of cell number relative to controls over 48 h among all of the cell lines examined without causing necrosis,determined by trypan blue exclusion, at the 6- or 12-h time points. This concentration has been used by other investigators for in vitro treatment of various cell types. Decreased cell survival became apparent between 12 and 24 h, with the exception of SRB-1 cells. SCC-13 cells were the most sensitive to 4HPR treatment, with∼50% cell survival observed at 12 h. SRB-1 cells did not show a similar effect until ∼36 h later. This differential response to 4HPR is likely attributable to the extended doubling time observed for SRB-1 cells with respect to other SCC cell lines (see “Materials and Methods”).

SCC-13 cells adequately illustrated both a dose- and a time-dependent decrease in relative cell numbers after exposure to varying concentrations of 4HPR (Fig. 1 B). This is especially evident at the 24-h time point, where 10 μm 4HPR caused∼90% reduction in cell number relative to control, whereas 5 and 1μ m 4HPR promoted ∼50% and 20% reduction,respectively. COLO 16 cells treated with 1 and 5μ m 4HPR required ∼48 h to experience similar effects in proliferation inhibition to those observed in SCC-13 cells at 24 h (not shown). SRB-12 and SRB-1 cells were not examined using lower concentrations of 4HPR for this reason.

Within 6–12 h after exposure to 10 μm 4HPR, SCC cells exhibited early morphological changes (specifically, chromatin condensation and cell shrinkage) typical of apoptotic cells (27). Over time, SCC cells exposed to 4HPR would shrink to approximately one-half of their normal size and detach from the tissue culture dish. To determine whether this was actually apoptosis and to possibly quantify the event, the TUNEL assay was conducted to measure DNA fragmentation associated with apoptosis induction (21).

Fig. 2 depicts typical flow cytometry scatter diagrams obtained for SCC-13 cells. Control populations were gated, and apoptosis induction was assessed as 4HPR-treated cells migrated above the gate, indicating increasing degrees TUNEL staining[presented in the upper right corner of each panel as a percentage of the total population (SD)]. SCC-13 cells treated with increasing concentrations of 4HPR exhibited apoptosis induction in both a concentration- and time-dependent fashion. Secondary necrosis (27) was evident in SCC-13 cells after a 48-h exposure to 5 μm 4HPR and after a 24-h exposure to 10μ m 4HPR. This advanced degradation prevented additional TUNEL evaluation because the remaining intact cells were not sufficient in quantity for the labeling procedure (1–1.5 ×10⁶ cells/sample were required for labeling).

A summary of the TUNEL data for SCC cells involved in apoptosis is presented in Fig. 3. All of the SCC cell lines exhibited apoptosis induction in ∼80% of the cell population between 12 and 24 h after exposure to 10 μm 4HPR(Fig. 3,A–D). Interestingly, COLO 16 cells (Fig. 3,A) exhibited latency from the initial time of exposure to 10 μm 4HPR until the 12-h time point before TUNEL staining was markedly different from control cells. In the other SCC cell lines exposed to 10 μm 4HPR (Fig. 3,B, C, and D), apoptosis induction was between 30 and 60% at the 6-h time point. Secondary necrosis was evident in SRB-12 cells after a 48-h exposure to 10μ m 4HPR, which prevented additional TUNEL evaluation. Although initially SCC-13 cells did not respond to 4HPR,eventually apoptosis induction was evident at the 24-h time point for exposure to 5 μm 4HPR (Fig. 3,E) or at the 48-h time point for exposure to 1 μm4HPR (Fig. 3 F).

Cell cycle distributions were examined for SCC-13 cells treated with 4HPR and presented in Fig. 4. The appearance of a sub-G₁ population, indicating cells or apoptotic bodies with less than G₁ DNA content via incorporation of propidium iodide, can be used as an indicator of apoptosis (22). The maximal sub-G₁ values for 4HPR-treated cells did not exceed 36% of the total sample population at any of the times for the concentrations of 4HPR examined. As a secondary indicator of apoptosis,this was typically <50% of the values obtained from the TUNEL procedure. It should be stressed that the apoptosis detection method(using propidium iodide as a counterstain) used in this work is optimized for TUNEL-mediated fluorescein-labeled dUTP incorporation. The cells were fixed with paraformaldehyde before the permeabilization procedures with Triton X-100 and 70% ethanol. Thus, any soluble intracellular histone-associated DNA fragments would be covalently cross-linked to intracellular matrix and unable to escape the cell after permeabilization. As such, a substantial sub-G₁ population would not be expected until perhaps the formation of apoptotic bodies that would ultimately shift into the sub-G₁ population as a result of their reduced DNA content. This shift may require more time to develop, or may occur to variable degrees in keratinocytes (28). SCC-13 cells exposed to 1 μm 4HPR have less cells in G₁ and more cells in G₂-M than do controls at the 12- and 24-h time points; and cells exposed to 5 μm 4HPR have less cells in G₁,more cells in S phase, and less cells in G₂-M than do controls. This shows that 4HPR is altering progression through the cell cycle before apoptosis induction.

The data presented indicate that 4HPR was effective in promoting apoptosis in SCC cells. Recent studies have shown that 4HPR can function as a prooxidant in various tumor cell lines (5, 7, 8). To investigate the possibility that ROS may be associated with 4HPR-induced apoptosis in SCC cells, measurements of cellular fluorescence resulting from the oxidation of DCF were conducted. Fig. 5,A demonstrates the ability of various concentrations of 4HPR to promote the oxidation of DCF in SCC-13 cells. After 60 min of exposure to 4HPR, a linear increase in fluorescence, as a measure of ROS production, was observed. Interestingly, SCC-13 cells responded similarly in ROS production to both the 5- and 10-μm concentrations of 4HPR,indicating possible saturation of the source of ROS generation. The other SCC cell lines were equally or more effective producers of ROS after a 10-μm 4HPR treatment, with COLO 16 cells exhibiting the greatest degree of ROS generation. ATRA,9-cis RA, and the retinoic acid receptor γ-specific ligand CD437 (all at 10 μm) were unable to enhance ROS production after short-term treatment of SRB 12 cells (Fig. 5 B). Furthermore, as observed in SCC-13 cells, there was a similar response to both the 5- and 10-μm 4HPR treatments with respect to ROS production in SRB-12 cells.

Treatment with the antioxidant Vit-C (29) was effective in inhibiting apoptosis induction in COLO 16 cells after a 12-h exposure to 4HPR (Fig. 6 A). This antioxidant has the ability to scavenge free radicals and hydrogen peroxide (30). Because Vit-C was unable to completely suppress 4HPR-induced apoptosis, we cannot exclude the possibility that some other mechanism, not associated with ROS production, could contribute to the overall effects of 4HPR in COLO 16 cells.

The most abundant retinoic acid receptor in normal skin is retinoic acid receptor γ (31), and this receptor is expressed to varying degrees in the SCC cell lines examined in this study (not shown). Treatment with the retinoic acid receptor β- and γ-specific antagonist CD2665 (32) was unable to influence 4HPR-mediated apoptosis at the 12-h time point (Fig. 6 A). Treatment with CD2665 at concentrations higher than 10μ m caused necrosis when combined with 10μ m 4HPR. Trypan blue exclusion was restricted to 60% of the total cell population after treatment with 20μ m CD2665 and treatment with 10μ m 4HPR for 12 h (not shown). This limited the use of this antagonist at higher concentrations with respect to possibly blocking apoptosis induction in COLO 16 cells. CD2665 at 10μ m and Vit-C at 500 μmwere unable to promote appreciable degrees of apoptosis as single agents in COLO 16 cells after 24 h of exposure (not shown).

Treatment with Vit-C could inhibit apoptosis in COLO 16 cells exposed to 4HPR. This prompted reexamination of ROS production to determine whether Vit-C was influencing the prooxidant activity of 4HPR as a means of modulating apoptosis induction. Fig. 6,Billustrates that treatment with Vit-C could diminish 4HPR-mediated ROS production in COLO 16 cells by ∼50%. ROS can affect various mitochondrial components, ultimately resulting in the loss ofΔψ _m and apoptosis induction (33). To investigate if 4HPR alone or in combination with Vit-C could modulate Δψ_m in SCC cells,retention of the cationic mitochondrial dye R123 (25) was examined. Fig. 6 C illustrates that COLO 16 cells treated with 10 μm 4HPR showed a gradual sustained decrease in R123 fluorescence (as a measure ofΔψ _m dissipation) over time. When 4HPR was combined with Vit-C, dissipation of Δψ_m was inhibited relative to the degree exhibited by 4HPR alone. Vit-C treatment promoted a slight reduction in Δψ_mrelative to DMSO-treated cells.

Having observed that R123 retention and the oxidation of DCF were directly associated with apoptosis or cell survival in SCC cells treated with 4HPR alone or in combination with Vit-C, the role of MPT in 4HPR-induced apoptosis was investigated. MPT is characterized as the opening of mitochondrial megachannels that can be triggered by ROS or other agents, resulting in the dissipation ofΔψ _m, ATP depletion, caspase/endonuclease activation, and ROS production (reviewed in Refs. 33 and 34). CsA has been used to inhibit MPT triggered by various agents, including those capable of promoting oxidative stress (35, 36). Therefore, the effects of CsA on SCC-13, the cell line most sensitive to the apoptotic effects of 4HPR, were examined to determine whether MPT was an essential characteristic of 4HPR-induced apoptosis.

Fig. 7,A shows that the addition of CsA with 4HPR could preserve R123 fluorescence in SCC-13 cells as compared with exposure to 4HPR alone. In addition, CsA promoted a discernable increase in R123 retention in SCC-13 cells,indicating a possible hyperpolarization ofΔψ _m. CsA was also able to decrease the oxidation of DCF when combined with 4HPR (Fig. 7,B),indicating a reduction in ROS production. To determine whether this reduction could be directly attributed to possible reactions between ROS and CsA, hydrogen peroxide was combined with CsA. This resulted in no decrease in the oxidation of DCF. This would indicate that CsA was not directly competing with DCF for oxidation by ROS. Fig. 7,C demonstrates that the CsA/4HPR treatment combination could completely block the DNA fragmentation promoted by 4HPR alone after 6 or 12 h as measured by the TUNEL assay. This combination promoted even less apoptosis than the CsA treatment. CsA is also an inhibitor of the Ca²⁺-regulated protein phosphatase calcineurin (37, 38). However, a 10-μm concentration of the calcineurin inhibitor Cyp (39) was unable to protect SCC-13 cells from 4HPR-induced apoptosis after a 6-h exposure (Fig. 7,D). A specific inhibitor of MPT, BA (40), at a concentration of 50 μm, was also able to block apoptosis in SCC-13 cells exposed to 4HPR (Fig. 7 D). BA also inhibited dissipation of Δψ_m, and was slightly more effective in reducing 4HPR-induced ROS production compared with CsA in SCC-13 cells (not shown). The ability of CsA and BA to inhibit dissipation of Δψ_m, ROS production, and apoptosis points strongly to the mitochondria and MPT as the primary regulators of 4HPR-induced cell death in SCC cells.

The data presented provide evidence that, in addition to functioning as a prooxidant, apoptosis promoted by 4HPR in SCC cells is dependent on MPT. There are several lines of evidence supporting this conclusion. With respect to ROS production, cellular redox potentials are crucial for maintenance of biochemical activity required for cell viability. Oxidative stress is triggered by overproduction of ROS and the inability of the cell to counter this insult, resulting in disruption of redox homeostasis. This can promote deleterious biochemical modifications in proteins, carbohydrates, fatty acids, and nucleic acids, causing proliferation inhibition, apoptosis, or necrosis (41, 42, 43). Exogenous antioxidants and enzymatic defense mechanisms targeting ROS can buffer redox potentials during oxidative stress, thus allowing some degree of protection during such events (29, 30).

The ability of 4HPR to increase oxidation of DCF in all of the SCC cell lines implies that oxidative stress is associated with apoptosis induction. This is consistent with the observed inhibitory effects of Vit-C on 4HPR-induced ROS production and apoptosis in COLO 16 cells. Still, the ability of COLO 16 cells to generate more ROS than the other SCC cell lines after exposure to 10 μm 4HPR, and to display a latency to apoptosis induction, points to the possible involvement of intracellular defenses against ROS, possibly enzymes responsible for hydrogen peroxide production from superoxide anions,specifically superoxide dismutases (29). This tenet is reflected by the similar responses in DCF oxidation obtained from both the 5- and 10-μm concentrations of 4HPR in SCC-13 and SRB-12 cells, as well as by the ability of SCC-13 cells to tolerate 5μ m 4HPR for at least 12 h before the majority of the cell population became apoptotic.

We speculate that 4HPR promotes two distinct types of cell death depending on the duration of exposure and the treatment concentration. This is especially evident in SCC-13 cells exposed to 1, 5, or 10μ m 4HPR. The 10 μm exposure completely inhibited proliferation by promoting interphase cell death (44), where apoptosis occurred in the entire cell population without cell cycle progression or mitosis. Lower concentrations of 4HPR partially inhibited proliferation but ultimately resulted in mitotic or delayed reproductive cell death. This type of cell death is believed to be attributable to secondary changes in cellular metabolism or to genetic damage causing unbalanced cell cycle progression and proliferation (44). SCC-13 cells treated with 1 or 5 μm 4HPR exhibited decreased proliferation and alterations in cell cycle progression relative to controls before apoptosis induction. The 5-μm 4HPR treatment promoted more ROS production than the 1-μm 4HPR treatment. Thus,it would appear that the basis of the delayed apoptosis after exposure to decreasing concentrations of 4HPR was associated in some respect with varying degrees of oxidative stress.

Observations in cervical carcinoma cells suggest that 4HPR-induced ROS production is linked to components of the mitochondrial respiratory chain (45). In this study, COLO 16 cells treated with 4HPR showed the greatest increase in ROS production along with a gradual dissipation of Δψ_m. Interestingly, the uncoupling agent 2,4-dinitrophenol was able to reduce mitochondrial R123 retention in COLO 16 cells by ∼60% after a 2-h exposure. However, this exposure was unable to promote ROS generation in the same cell line (not shown). This would indicate that 4HPR-induced ROS production was associated with the loss ofΔψ _m. 4HPR-induced dissipation ofΔψ _m in COLO 16 cells could be prevented with Vit-C, and this was accompanied by a decrease in ROS production. In the case of the Vit-C/4HPR treatment combination, the maintenance ofΔψ _m can potentially be linked to the inhibition MPT via reduction of ROS, assuming MPT can be triggered by ROS (33, 46).

CsA, as an inhibitor of MPT, is believed to bind cyclophilin associated with the mitochondrial adenine nucleotide translocator, thereby inhibiting its function by promoting a closed-matrix conformation (46, 47). BA is believed to bind directly to the mitochondrial adenine nucleotide translocator, also inhibiting its function by promoting a closed-matrix conformation (47). CsA had the ability to maintain Δψ_m when combined with 4HPR in SCC-13 cells. This observation would suggest that 4HPR was not functioning simply as a protonophore, because CsA combined with aristolochic acid was unable to preserve R123 retention in fibroblasts treated with carbonyl cyanide m-chlorophenylhydrazone (38).

The ability of CsA to enhance R123 retention in SCC-13 cells demonstrates that CsA was functioning at the mitochondrial level. CsA and BA also had the ability to reduce ROS generation when combined with 4HPR in SCC-13 cells. As such, a substantial amount of the ROS production after exposure to 4HPR alone could be attributed to hyperproduction of ROS resulting from MPT (33),represented by the rapid dissipation of Δψ_m. The ROS production detected in the CsA/4HPR treatment combination was apparently independent of MPT, indicating ROS was responsible for MPT as observed in COLO 16 cells. Most notably, CsA and BA could inhibit 4HPR-induced DNA fragmentation, illustrating that MPT was required for apoptosis.

This study has highlighted the effects of 4HPR on ROS production and promotion of MPT, which appear to be the major mechanisms associated with apoptotic cell death in the SCC cell lines examined. It is noteworthy that in other cell types (e.g., neuroblastoma)other mechanisms may be important, such as increased production of ceramide (48). In neuroblastoma cells, 4HPR promoted necrosis in addition to apoptosis. A shift from predominately apoptotic to predominately necrotic cell death has been observed in SCC cells exposed to 20 μm 4HPR for 12 h (not shown). However, other potential treatment effects of this concentration were not examined because these conditions precluded accurate mechanistic evaluation of apoptotic cell death, which was the focus of this study.

The mechanism of action associated with the antitumor effects of 4HPR in vivo remains to be elucidated. This retinoid is appealing for chemopreventive and therapeutic application because it is effective without many of the undesirable side effects associated with the prolonged use of other synthetic and natural retinoids. In addition, it is conceivable that 4HPR could be developed for topical delivery in a cosmetic application. Such an application would be ideally suited as a prophylactic measure to prevent skin cancer development or to resolve premalignant actinic damage. Together, results obtained in clinical trials combined with in vitro findings lend support for the continued examination of 4HPR as a chemopreventive and therapeutic agent in skin.

We thank Karen Ramirez for her assistance in the acquisition of the flow cytometry data presented in this study; Dr. Janet Price for the SCC cell lines SRB-1, SRB-12, and COLO 16; Dr. J. A. Duine(Delft University of Technology, Delft, The Netherlands) for the gift of BA; Dafna Lotan for her invaluable assistance in retinoid preparation and subcloning the SCC cell lines examined in this study;and Dr. Shi-Yong Sun for his advice with the ROS assay.