Advertisement

Involvement of Oxidative Stress in Apoptosis Induced by a Mixture of Isothiazolinones in Normal Human Keratinocytes

      A 3:1 combination of 5-chloro-2-methyl-4-isothiazolin-3-one (CMI) and 2-methyl-4-isothiazolin-3-one (MI) is widely used to preserve cosmetic products. We show here that CMI/MI induced apoptosis in normal human keratinocytes (NHK) as at low concentrations (0.001–0.05% documented by subdiploid DNA content and phosphatidylserine exposure, while at the highest concentration (0.1% as supplied, 15 p.p.m.) the response was necrosis. Various molecular events accompanied the cytotoxic effects of CMI/MI. Generation of ROS and hyperpolarization of mitochondrial transmembrane potential (Δψm) were early events, followed by increased Fas expression and activation of caspase-8, and then activation of caspase-3 and -9. The drop in Δψm occurred only later in the cell death pathway, when NHK showed signs of apoptosis. Pretreatment of cells for 2 h with the redox-active agent N-acetyl-L-cysteine conferred complete protection against the CMI/MI-induced cytotoxic effects, Δψm loss, and apoptosis. The pan-caspase inhibitor Z-Val-Ala-Asp(OMe)-CH2F blocked the CMI/MI-induced apoptosis without preventing ROS generation and the drop in Δψm. These results indicate that the generation of ROS plays an important part in mediating apoptosis and necrosis associated with CMI/MI treatment. This new aspect of the in vitro toxicity of CMI/MI may provide important information about the relationship between the preservative's in vitro apoptotic activity and its in vivo toxicity.

      Keywords

      Abbreviations

      Ac-DEVD-pNA
      Ac-Asp-Glu-Val-Asp-p-Nitroaniline
      Ac-IETD-pNA
      Ac-Ile-Glu-Thr-Asp-pNA
      Ac-LEHD-pNA
      Ac-Leu-Glu-His-Asp-pNA
      AnnxV
      Annexin-V
      CHAPS
      3-[(3-cholamidopropyl)dimethyl-ammonio]-1-propane-sulfonate
      CMI/MI
      mixture of 5-chloro-2-methyl-4-isothiazolin-3-one and 2-methyl-4-isothiazolin-3-one in an approximate ratio of 3:1
      DCFH-DA
      2′,7′-dichlorofluorescein diacetate
      FITC
      fluorescein isothiocyanate
      HEPES buffer
      N-[2-Hydroxyethyl]piperazine-N′-[2-ethanesulfonic acid]
      Δψm
      mitochondrial transmembrane potential
      MTT
      3-[4,5 dimethylthiazol-2-yl]-2,5-diphenyl-tetrazolium bromide
      NAC
      N-Acetyl-L-cysteine
      NHK
      normal human keratinocytes
      R-PE
      R-phycoerythrin
      PI
      propidium iodide
      Rhod123
      Rhodamine123
      ROS
      reactive oxygen species
      Z-VAD-FMK
      Z-Val-Ala-Asp(OMe)-CH2F
      Preservatives are an important class of chemicals used to inhibit the growth of pathogenic and nonpathogenic microorganisms in a variety of final products such as cosmetic creams and lotions, pharmaceuticals, household detergents, and hygiene products. The combination of 5-chloro-2-methyl-4-isothiazolin-3-one (CMI) and 2-methyl-4-isothiazolin-3-one (MI), in an approximate ratio of 3:1, is commonly used to preserve cosmetic and body care products such as shampoos, skin creams and lotions, as well as toilet paper and household detergents. CMI/MI is also used as a biocide in swimming pool water and in various industrial applications. It is active at very low concentrations against bacteria, fungi, and yeast. The chlorinated molecule is the more active of the two but both ingredients are considered biologically functional, although the mechanism of action is not well understood (
      • Law A.B.
      • Moss N.J.
      • Lashen E.S.
      • Kathon C.G.
      A new single-component, broad spectrum preservative system for cosmetics and toiletries.
      ). It has been reported that the biocidal activity of isothiazolinone compounds is due to reactions with the sulphydryl groups of enzymes and other proteins (
      • Morris S.L.
      • Walsh R.C.
      • Hansen J.N.
      Identification and characterization of some bacterial membrane sulphydryl groups which are targets of bacteriostatic and antibiotic action.
      ;
      • Collier P.J.
      • Ramsey A.
      • Waigh R.D.
      • Douglas K.T.
      • Austin P.
      • Gilbert P.
      Chemical reactivity of some isothiazolone biocides.
      ). This mixture has also been shown to be a bacterial mutagen (
      • Monte W.C.
      • Ashoor S.H.
      • Lewis B.J.
      Mutagenicity of two non-formaldehyde-forming antimicrobial agents.
      ).
      The irritation potential of CMI/MI has been demonstrated by the hen's egg chorioallantoic membrane test (
      • Luepke N.P.
      Hen's egg chorioallantoic membrane test for irritation potential.
      ), whereas its allergenic potential has been shown by the lymphocyte transformation test (
      • Stejskal V.D.
      • Forsbeck M.
      • Nilsson R.
      Lymphocytes transformation test for diagnosis of isothiazolinone allergy in man.
      ). Studies on guinea pigs and humans have demonstrated that CMI/MI can induce contact hypersensitivity; in particular, CMI and 4,5-dichloromethylisothiazolinone (a contaminant present in Kathon CG) are strong sensitizers, whereas MI is a weak one (
      • Bruze M.
      • Gruvberger B.
      • Persson K.
      Contact allergy to the active ingredients of Kathon CG in the guinea pig.
      ,
      • Bruze M.
      • Freget S.
      • Gruvberger B.
      • Persson K.
      Contact allergy to the active ingredients of Kathon CG in the guinea pig.
      ,
      • Bruze M.
      • Dahlquist I.
      • Gruvberger B.
      Contact allergy to dichlorinated methylisothiazolinone.
      ). Allergic contact dermatitis reactions to the preservative and chemical burns have often been described (
      • Alexander B.R.
      An assessment of the comparative sensitization potential of some common isothiazolinones.
      ). Epidemiologic studies have indicated that cosmetics are the most important cause of sensitization to CMI/MI, even though the CMI/MI concentrations are low: the maximum concentration allowed in “leave on” cosmetic products is 7.5 p.p.m.
      Many studies have investigated the effect of CMI/MI as an allergen and irritant in vivo, but there has been little work on the cytotoxic potential of CMI/MI in cell cultures.
      • Rivalland P.
      • Vie K.
      • Coiffard L.
      • De Roeck-Holtzhauer Y.
      Cytotoxicity test of antibacterial agents on human fibroblasts cultures.
      reported CMI/MI cytotoxicity in single layers of human fibroblast culture, as shown by the MTT (3-(4,5 dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) test. We previously reported that CMI/MI was the most toxic of four preservatives in HL60 cells and that apoptosis was one mechanism of the CMI/MI-mediated cytotoxicity (
      • Anselmi C.
      • Ettorre A.
      • Andreassi M.
      • Centini M.
      • Neri P.
      • Di Stefano A.
      In vitro induction of apoptosis vs necrosis by widely used preservatives: 2-phenoxyethanol, a mixture of isothiazolinones, imidazolidinyl urea and 1,2-pentanediol.
      ).
      As CMI/MI is an extensively used preservative in cosmetics, we investigated its potential effects in vitro using normal human keratinocyte (NHK) cell cultures and focusing on its apoptotic potential. Keratinocytes are the predominant cells in the epidermis and cultured keratinocytes may be used experimentally to predict the potential toxicity of chemicals in contact with the skin.
      Andreassi L, Di Stefano A, Ettorre A, Andreassi M, Sbrana S, Pianigiani E, Neri P: Keratinocytes apoptosis as a parameter of chemical toxicity: Flow cytometric analysis in different conditions of cultivation. J Invest Dermatol 117(2):405, 2001 (Abstr.)
      1Andreassi L, Di Stefano A, Ettorre A, Andreassi M, Sbrana S, Pianigiani E, Neri P: Keratinocytes apoptosis as a parameter of chemical toxicity: Flow cytometric analysis in different conditions of cultivation. J Invest Dermatol 117(2):405, 2001 (Abstr.)
      Apoptosis is known to occur in the skin after physical injury caused by UV radiation (resulting in sunburn cells) or by primary skin irritation, in graft-versus-host disease and in cutaneous and systemic lupus erythematosus (
      • Kanerva L.
      Electron microscopic observations of dyskeratosis, apoptosis, colloid bodies and fibrillar degeneration after skin irritation with dithranol.
      ;
      • Leverkus M.
      • Yaar M.
      • Gilchrest B.A.
      Fas/Fas ligand interaction contributes to UV-induced apoptosis in human keratinocytes.
      ). Knowledge of the molecular mechanism of CMI/MI may be clinically important for an understanding of its in vivo effects. Using the MTT test and independent biochemical indices of apoptosis, we first showed that CMI/MI induces caspase-dependent apoptosis and necrosis in a concentration-dependent manner in normal human keratinocytes. We also monitored the production of reactive oxygen species (ROS), mitochondrial transmembrane potential (Δψm) by rhodamine 123 (Rhod123) uptake over time, and Fas expression. We found that CMI/MI increased the production of ROS and induced a drop in Δψm. These events were prevented by preincubation with the antioxidant N-acetyl-L-cysteine (NAC) but not by pretreatment with the pan-caspase inhibitor Z-Val-Ala-Asp(OMe)-CH2F (Z-VAD-FMK), which only prevented caspase activation and the appearance of apoptotic features.

      Materials and methods

      A commercial mixture of 5-chloro-2-methyl-4-isothiazolin-3-one and 2-methyl-4-isothiazolin-3-one, in an approximate ratio of 3:1 (Kathon CG), was obtained from Rohm and Haas Company (Philadelphia, Pennsylvania).

      Keratinocyte monolayer culture

      —primary basal keratinocytes were obtained from samples of skin taken from patients during dermatologic operations at the Policlinico Le Scotte, Siena (Italy). All patients gave their informed consent to the tissue sampling. The investigations have been carried out in agreement with the Helsinki Principles after approval of The Institutional Review Board. Cells were harvested by trypsinization, washed in serum-free medium, counted, and stored frozen until their use. The cells were then resuspended in serum-free basal keratinocyte medium supplemented with keratinocyte medium supplement and antibiotics (100 U per mL penicillin and 100 μg streptomycin per mL) (Sigma-Aldrich, Milan, Italy).

      Cell treatment

      Keratinocytes at second to fourth passage, plated in 25 cm2 flasks and grown until 75% to 80% or complete confluence, were exposed to increasing concentrations of CMI/MI (0.001–0.1%) for 10 min. They were then washed with phosphate-buffered saline (PBS) and fresh medium was added. Cells treated with PBS served as controls.
      We chose the concentrations 0.001%, 0.01%, 0.05%, and 0.1% of CMI/MI solution as supplied by Rohm and Haas Company (respectively 0.15, 1.5, 7.5, and 15 p.p.m. of active ingredients). In fact the maximum allowed concentration in “rinse off” products is 0.1% of Kathon CG (sold as a dilute aqueous solution of 1.5% of the two isothiazolinones, corresponding to 15 p.p.m. of active ingredients), whereas in “leave on” products it is 0.05% (corresponding to ≈ 7.5 p.p.m.), as reported in The Cosmetic Ingredient Review, Inc. 2001.
      Final report on the safety assessment of methylisothiazolinone and methylchloroisothiazolinone (eds). Mary Ann Liebert, Inc. Publishers. J Am Coll Toxicol 11:75–128, 1992
      Spontaneously floating cells were collected with the supernatant, pelleted by centrifugation, and washed with PBS. They were then combined with adherent cells that were trypsinized (trypsin-ethylendiamine tetraacetic acid 0.5%, Sigma-Aldrich) until complete cell detachment. The trypsin was neutralized with medium plus fetal bovine serum (Gibco, Milan, Italy) and the cells were harvested for analysis at different times after treatment.

      Cytotoxicity and cell viability

      In the preliminary study, the cytotoxic effect of increasing concentrations of CMI/MI (0.001–0.1%) was estimated at different times (0–24 h) by measurement of the rate of mitochondrial metabolism of MTT (Sigma-Aldrich) (
      • Mosmann T.
      Rapid colorimetric assay for cellular growth and survival: Application to proliferation and cytotoxicity assays.
      ). Briefly, the control and treated cells were seeded at 1 × 105 cells per well in 100 μL of medium in 96-well plates and 10 μL of MTT solution (5 mg per mL in PBS) were added to each well. After 4 h of incubation at 37°C, 100 μL of a lysing buffer (10% sodium dodecyl sulfate, 45% dimethylformamide, adjusted to pH 4.5 with glacial acetic acid) were added to each well and the blue formazan crystals were dissolved by pipetting. The plates were read with a microplate reader (Bio-Rad, Milan, Italy) using a test wavelength of 595 nm and a reference wavelength of 655 nm. All cell viability assays were performed in triplicate.

      Measurement of cellular DNA content

      To quantify the percentage of apoptotic cells by means of DNA content analysis, we used the simple flow cytometric method described by
      • Nicoletti I.
      • Migliorati G.
      • Pagliacci M.C.
      • Grignani F.
      • Riccardi C.
      A rapid and simple method for measuring thymocyte apoptosis by propidium iodide staining and flow cytometry.
      . At different times after CMI/MI exposure, 1 × 106 cells for each sample were washed in PBS and the pellet was fixed overnight in ice-cold ethanol 70% at –20°C. The cell suspension was centrifuged, washed twice with 1 mL of PBS and resuspended in 1 mL of a PBS solution containing RNAse (type I-A, Sigma-Aldrich; 1 mg per mL final concentration) and propidium iodide (PI) (Sigma-Aldrich; 50 μg per mL final concentration). The tubes were placed on ice in the dark until the cellular red fluorescence of PI was measured on a linear scale using a FACSCalibur flow cytometer (Becton Dickinson, Mountain View, California) equipped with an excitation laser line at 488 nm and a 575±15 nm bandpass filter. At least 20,000 events were recorded for each sample using Cell Quest software (Becton Dickinson) and the pulse processing module for doublet discrimination; debris was excluded from the analysis by an appropriate morphologic gate of forward scatter versus side scatter.

      Measurement of phosphatidylserine expression using fluorescein isothiocyanate (FITC)-labeled Annexin V (AnnxV)

      AnnxV binding and PI uptake were assessed by flow cytometry using a commercial kit (Boehringer Ingelheim Bioproducts, Vienna, Austria) according to the manufacturer's instructions. Briefly, at different times of incubation after CMI/MI exposure, approximately 2.5 × 105 cells for each sample were washed twice in PBS and the pellet was resuspended in 200 μL of the binding buffer provided in the kit. 5 μL of the AnnxV-FITC kit stock solution were added to the cell suspension (1 μg per mL final concentration) and incubated for 10 min at room temperature in the dark. The cells were then washed in PBS and resuspended in 190 μL of binding buffer plus 10 μL of the PI stock solution (1 μg per mL final concentration). The cells were immediately analyzed with a FACSCalibur flow cytometer (Becton Dickinson) equipped with Cell Quest software (Becton Dickinson). The AnnxV-FITC (green fluorescence) and the PI (red fluorescence) were both measured on a log scale through a 530±20 and 575±15 nm bandpass filter, respectively (
      • Di Stefano A.
      • Ettorre A.
      • Sbrana S.
      • Giovani C.
      • Neri P.
      Purpurin-18 in combination with light leads to apoptosis or necrosis in HL60 leukemia cells.
      ).

      Evaluation of transmembrane potential using double staining with Rhod123 and PI

      Mitochondrial transmembrane potential was assessed by flow cytometry uptake of the cationic lipophilic dye Rhod123 and PI using a commercial product (Sigma-Aldrich) according to the method described by
      • Gorczyca W.
      • Melamed M.R.
      • Darzynkiewicz Z.
      Analysis of apoptosis by flow cytometry.
      . Briefly, at different times of incubation after CMI/MI exposure, approximately 5 × 105 cells for each sample were washed twice in PBS and the pellet was resuspended in 500 μL of PBS. Four microliters of the Rhod123 stock solution were added to the cell suspension (1 μg per mL final concentration) and incubated for 30 min at 37°C in the dark. The cells were then washed in PBS and resuspended in 200 μL of binding buffer plus 10 μL of the PI stock solution (10 μg per mL final concentration). The cells, kept in ice, were analyzed with a FACSCalibur flow cytometer (Becton Dickinson) equipped with an excitation laser line at 488 nm and Cell Quest software (Becton Dickinson). The Rhod123 (green fluorescence) and the PI (red fluorescence) were both measured on a log scale through a 530±20 and 575±15 nm bandpass filter, respectively.

      Evaluation of ROS production

      The formation of intracellular ROS was measured using 2′,7′-dichlorofluorescein diacetate (DCFH-DA, Sigma-Aldrich). By this method, it is possible to measure the amount of H2O2 generated by increased oxidative metabolism. Viable cells can deacetylate DCFH-DA to 2′,7′-dichlorofluorescein; the latter is not fluorescent but can react quantitatively with oxygen species within the cell to produce 2′,7′-dichlorofluorescein (DCF), which is fluorescent and is trapped inside the cell. The cytofluorimetric measurement of the DCF produced can provide an index of intracellular oxidation (
      • Sureda F.X.
      • Gabriel C.
      • Comas J.
      • Palls M.
      • Escubedo E.
      • Camarasa J.
      • Camins A.
      Evaluation of free radical production, mitochondrial membrane potential and cytoplasmic calcium in mammalian neurons by flow cytometry.
      ). The cells, plated in 25 cm2 flasks and grown until 75% to 80% or complete confluence, were incubated for 30 min with DCFH-DA (100 μM final concentration), then washed, resuspended in fresh medium and treated with different concentrations of CMI/MI (0.001–0.1%). The fluorescence intensity was measured 1 h after treatment. The cells, kept in ice, were analyzed with a FACSCalibur flow cytometer (Becton Dickinson) equipped with an excitation laser line at 488 nm and Cell Quest software (Becton Dickinson). The DCF (green fluorescence) was measured on a log scale through a 530±20 bandpass filter.

      Assay of caspase-3, -8, and -9 activity

      At different times of incubation after CMI/MI exposure, the cells were washed in PBS and resuspended in ice-cold lysis buffer (50 mM HEPES, 1 mM dithiothreitol, 0.1 mM ethylendiamine tetraacetic acid, 10% glycerol, 0.1% 3-[(3-cholamidopropyl)dimethyl-ammonio]-1-propane-sulfonate (CHAPS), pH 7.4, supplemented with 5 mg leupeptin per mL). After centrifugation at 10,000 ×g at 4°C, the supernatant was used for the assay of caspase-3 activity. The protein concentration in the lysate was determined by the Bradford assay (
      • Bradford M.M.
      A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding.
      ). For the assay of caspase-3, -8, and -9, 25 μg of cell lysate were incubated in 100 μL of assay buffer (50 mM HEPES, 10 mM dithiothreitol, 0.1 mM ethylendiamine tetraacetic acid, 10% glycerol, 0.1% CHAPS, pH 7.4, 100 mM NaCl) containing, respectively, 200 μM Ac-DEVD-pNA, Ac-IETD-pNA, Ac-LEHD-pNA (BioSource International distributed in Italy by Prodotti Gianni, Milan, Italy). Samples were incubated at 37°C in a microtiter plate reader for 16 h. The enzyme-catalyzed release of p-nitroaniline was monitored at 405 nm. The conversion of the substrate was linear in time and in amount of protein (
      • Di Stefano A.
      • Ettorre A.
      • Sbrana S.
      • Giovani C.
      • Neri P.
      Purpurin-18 in combination with light leads to apoptosis or necrosis in HL60 leukemia cells.
      ). The caspase-3, -8, and -9 activity assays were also performed in the presence of specific inhibitor substrates, respectively, Ac-DEVD-CHO, Ac-IETD-CHO, and Ac-LEHD-CHO, final concentration 0.1 μM.

      Flow cytometric evaluation of Fas expression

      After different times of treatment with CMI/MI, approximately 5 × 105 cells for each sample were washed in PBS and incubated for 30 min at 4°C with 1 μg per mL (final concentration) of an R-phycoerythrin (R-PE)-conjugated anti-human Fas monoclonal antibody (anti-CD95; Ancell, Milan, Italy) and IgG1-R-PE (IgG1; Ancell). After washing in PBS, cells were resuspended and analyzed on a FACSCalibur flow cytometer (Becton Dickinson). At least 20,000 events were recorded after exclusion of dead cells and debris on a double morphologic scatter cytogram. A sample with an unrelated IgG1-PE was also used to evaluate nonspecific labeling. Monoparametric histograms of the fluorescence distribution were plotted for the estimation of Fas surface expression.

      Cell pretreatment with NAC and Z-VAD-FMK

      In some experiments, cells were pretreated with NAC (10 mM) or Z-VAD-FMK (50 μM) 2 h before exposure to the different concentrations of CMI/MI. In another series of experiments, NAC was added to the culture medium immediately after treatment with CMI/MI. Using the methods and times after treatment reported above, we analyzed the cells for ROS production, changes in Δψm, AnnxV-binding, DNA content, caspase activity, and Fas expression.

      Results

      Cytotoxic effects of CMI/MI in NHK

      In the first series of experiments, we evaluated the effect of CMI/MI on keratinocyte viability using the MTT test. As shown in Figure 1(a), the cell viability of keratinocytes from 0 to 24 h decreased in a time- and concentration-dependent manner. When human keratinocytes were treated with 0.001% CMI/MI there was no evident decrease of cell viability, with the exception of 24 h after exposure; however, cell viability strongly decreased as the CMI/MI concentration was increased from 0.01% to 0.05%. At the highest concentration (0.1% CMI/MI), there was an acute loss of cell viability even at 0 time. At 3 h after treatment, many cells exposed to the highest concentration (0.1%) appeared to become rounded and floated in the medium (data not shown). It is well accepted that an early parameter of acute cellular damage is the loss of ability to grow adhering to the substrate (
      • Chen Q.M.
      • Liu J.
      • Merrett J.B.
      Apoptosis or senescence-like growth arrest: Influence of cell cycle position, p53, 21 and bax in H2O2 response of normal human fibroblasts.
      ).
      Figure thumbnail gr1
      Figure 1Cytotoxicity and induction of apoptosis by increasing concentrations of CMI/MI in NHK. Cells were incubated with increasing concentrations of CMI/MI (0.001–0.1%) for 10 min, resuspended in fresh medium and incubated at 37°C. (A) Cell viability was evaluated at 0, 3, 6, and 24 h by the MTT test. Values are expressed as percentage ± SEM (average of five separate experiments). (B) Frequency distribution histograms of DNA content of PI-stained NHK at 24 h after CMI/MI exposure. The x-axis shows PI fluorescence intensity; the y-axis indicates the cell number. Similar data were obtained in three independent experiments (C). Bivariate flow cytometry dot plots of AnnxV-FITC (FL1)-stained versus PI (FL3)-stained NHK at 12 h after treatment with CMI/MI. The lower left quadrant (AnnxV/PI) represents viable cells, whereas the upper left (AnnxV+/PI) and upper right (AnnxV+/PI+) quadrants show apoptotic and necrotic or late apoptotic cells, respectively, and the lower right quadrant represents nonviable cells, positive only for PI. The x-axis shows log FL3-fluorescence intensity; the y-axis indicates the log FL1-fluorescence intensity. Similar data were obtained in five independent experiments.

      CMI/MI induces apoptosis and necrosis in NHK

      To determine whether the loss of cell viability was due to apoptosis, we stained NHK with PI and analyzed the DNA content by flow cytometry. Exposure of NHK to CMI/MI for 10 min caused the appearance of subdiploid DNA content, which has been reported to be apoptotic DNA (
      • Darzynkiewicz Z.
      • Bedner E.
      • Smolewski P.
      Flow cytometry in analysis of cell cycle and apoptosis.
      ). The sub-G0/G1 DNA peak was barely evident at 12 h after exposure to 0.001% to 0.05% CMI/MI (data not shown), but it was clearly visible at 24 h and the reduction of the diploid peak was evident (Figure 1b). In contrast, no sub-G0/G1 peak was noted at a final concentration of 0.1% CMI/MI at any time after treatment. In parallel experiments we measured phosphatidylserine exposure, considered an early marker of apoptosis, by AnnxV-FITC binding. Figure 1(c) compares representative flow cytometry plots obtained from control keratinocytes and CMI/MI-treated cells at 12 h after treatment when the phosphatidylserine exposure became evident. Cells were simultaneously stained with PI and analyzed by flow cytometry. Figure 1(c) indicates that 6.11% of untreated NHK were AnnxV positive, 12.74% were AnnxV and PI positive, and 6.35% were PI positive. The high number of apoptotic cells was likely due to membrane damage during their detachment from the culture flask by trypsinization (
      • van Engeland M.
      • Nieland L.J.W.
      • Ramaekers F.C.S.
      • Schutte B.
      • Reutelingsperger C.P.M.
      Annexin V-affinity assay. A review on an apoptosis detection system based on phosphatidylserine exposure.
      ). When cells were treated with 0.001% CMI/MI, however, there was a slight, nonsignificant increase of both AnnxV-positive and AnnxV/PI-positive cells. At higher concentrations (0.01–0.05%), most cells were positive for AnnxV and AnnxV/PI, confirming a late stage of apoptosis. These data are consistent with CMI/MI-induced apoptosis. At the highest CMI/MI concentration (0.1%), only the percentage of AnnxV-positive and PI-positive cells increased dramatically; this suggests that cell death was by necrosis, as confirmed by cell detachment and lack of a subdiploid DNA peak.

      CMI/MI induces changes in mitochondrial transmembrane potential and an early increase of ROS production

      It is widely accepted that the mitochondrion plays a crucial part in necrosis and in many types of apoptotic responses. Therefore, we wished to evaluate the mitochondrial involvement in cell death induced by CMI/MI. One of the biochemical events noticed in the apoptotic process is the variation of mitochondrial transmembrane potential. We used Rhod123 to monitor ΔΨm: Rhod123 is readily incorporated into mitochondria in a manner dependent on ΔΨm. Untreated cells and cells treated with CMI/MI were loaded with Rhod123; the fluorescence intensity was then analyzed by flow cytometry and plotted as a fluorescence histogram. At 1 h after the 10 min exposure to CMI/MI, an increase of Rhod123 fluorescence was detected in NHK at final concentrations of 0.05% and 0.1% CMI/MI (Figure 2a). At 6 h, the fluorescence intensity of treated cells had returned to control levels, whereas the 0.1% CMI/MI-treated cells were still hyperpolarized (data not shown). At 12 h, the Rhod123 fluorescence was significantly reduced at all concentrations; the highest decrease of ΔΨm was observed at 0.05% CMI/MI (Figure 2b).
      Figure thumbnail gr2
      Figure 2Flow cytofluorimetric analysis of mitochondrial transmembrane potential and ROS production in NHK exposed to increasing concentrations of CMI/MI. Untreated and CMI/MI-treated NHK were loaded for 30 min with Rhod123 at 1 h (A) and 12 h (B) after treatment and the fluorescence intensity was determined by flow cytometry. These experiments were performed three times with very similar results. Open curves correspond to control cells, whereas shaded curves represent treated cells. The x-axis shows log FL-1 fluorescence intensity; the y-axis indicates the cell number. Decreased fluorescence (highlighted by the marker) indicates reduction of mitochondrial transmembrane potential (Δψm). The percentage of cells with decreased Δψm is reported. (C) Cells were incubated for 30 min with DCFH-DA, then washed and resuspended in fresh medium and treated with CMI/MI. At 1 h from treatment the fluorescence intensity was determined by flow cytometry. Similar data were obtained in four independent experiments. Open curves correspond to control cells, whereas shaded curves represent treated cells. The x-axis shows log FL-1 fluorescence intensity; the y-axis indicates the cell number. The amount of ROS production was quantified as the percentage of cells with increased fluorescence relative to control.
      To detect the production of ROS, we loaded NHK with a chemical probe, i.e., DCFH-DA, which becomes fluorescent upon oxidation. Figure 2(c) shows a concentration-dependent increase in fluorescence, measured by flow cytometry, just 1 h after exposure of the cells to CMI/MI (0.01–0.1%). The ROS production in NHK treated with 0.001% CMI/MI was not significantly different from that in control cells.

      CMI/MI stimulates caspase-3, -8, and -9 activity

      To evaluate the involvement of caspases in CMI/MI-induced apoptosis, we assayed caspase-3, -8, and -9 in lysates made from keratinocytes harvested at 3, 6, and 12 h after the 10 min exposure to different CMI/MI concentrations. Figure 3 shows that CMI/MI treatment induced an increase of caspase-3 (Figure 3a) and caspase-9 (Figure 3c) activity 6 and 12 h after exposure. CMI/MI also induced an increase of caspase-8 activity (Figure 3b) before the activation of caspase-3 and -9. The activation of caspases was concentration dependent at final concentrations of 0.001% and 0.01%. There was a smaller increase of caspase activation with 0.05% CMI/MI than with 0.01% CMI/MI, except for caspase-3 at 12 h. No activation was observed with 0.1% CMI/MI, confirming that cell death was by necrosis. The activation of caspase-3, -8, and -9 was inhibited by addition of specific inhibitor substrates, respectively, Ac-DEVD-CHO, Ac-IETD-CHO, and Ac-LEHD-CHO, final concentration 0.1 μM (data not shown).
      Figure thumbnail gr3
      Figure 3Time-course of caspase-3, -8, and -9 activity in NHK treated with increasing concentrations of CMI/MI. Lysates (25 μg), made from NHK treated with different concentrations of CMI/MI at different times after treatment (3, 6, and 12 h), were incubated in 100 μL of assay buffer containing the specific tetrapeptide substrates, Ac-DEVD-pNA (200 μM), Ac-IETD-pNA (200 μM), and Ac-LEHD-pNA (200 μM). Caspase activity is expressed as the OD 405 nm value ± SEM. Similar data were obtained in three independent experiments. (A) Caspase-3 activity; (B) caspase-8 activity; (C) caspase-9 activity.

      CMI/MI increases the expression of Fas on the NHK cell surface

      To investigate the molecular mechanism of CMI/MI-induced apoptosis, we evaluated the regulation of cell surface expression of Fas by flow cytometric analysis. Controls and treated cells were stained with anti-Fas-R-PE monoclonal antibody and analyzed by flow cytometry. Figure 4 shows that cell surface expression of Fas increased 3 h after treatment with 0.01% and 0.05% CMI/MI (Figure 4a). At 6 h, only a slight shift of fluorescence was detected at a final concentration of 0.01% (Figure 4b). At 3 h and 6 h, no variation of Fas was detected at the lowest CMI/MI concentration (0.001%). At 12 h and 24 h, no variations of Fas expression were observed at any of the concentrations (data not shown).
      Figure thumbnail gr4
      Figure 4Fas expression in NHK treated with increasing concentrations of CMI/MI. Monoparametric fluorescence histogram (Fl-2) of Fas (Anti-CD95-R-PE) expression plotted versus unrelated control (IgG1-R-PE) in NHK at 3 h and 6 h after treatment with increasing concentrations of CMI/MI. The increase of Fas expression relative to IgG1 was evaluated as the mean fluorescence shift. Open curves correspond to control anti-IgG isotype, whereas shaded curves represent cells positive to anti-CD95. The x-axis shows log FL-1 fluorescence intensity, the y-axis indicates the cell number. (A) Fas expression at 3 h after treatment; (B) Fas expression at 6 h after treatment. Similar data were obtained in three independent experiments.

      Effects of NAC or Z-VAD-FMK on ROS production, ΔΨm changes, induction of apoptosis, caspase activation, and Fas expression

      To determine the roles of oxidative stress and caspases in CMI/MI-induced cytotoxicity, we tested the effect of a 2 h preincubation of NHK with NAC (10 mM) or Z-VAD-FMK (50 μM) before and after (NAC only) treatment with the different concentrations of CMI/MI. Figure 5 only reports the data for 0.05% CMI/MI, as the results were similar for the other concentrations. As shown in Figure 5(a), NAC pretreatment inhibited all features of apoptosis, such as phosphatidylserine exposure, appearance of subdiploid DNA content and the loss of Δψm 12 h after CMI/MI treatment. Moreover, NAC completely inhibited ROS generation and protected cells from the mitochondrial hyperpolarization observed only 1 h after CMI/MI treatment (Figure 5b). NAC also prevented the increase of Fas expression and caspase activation (data not shown). When NAC was added after CMI/MI treatment, no protective effect was recorded. This confirmed that ROS production was the early event generating cytotoxic effects of CMI/MI in NHK (data not shown).
      Figure thumbnail gr5
      Figure 5Effects of the antioxidant NAC and the pan-caspase inhibitor Z-VAD-FMK on AnnxV-binding, changes in mitochondrial transmembrane potential, sub-G0/G1 peak and ROS production in NHK treated with CMI/MI. Cells were preincubated in the absence or in the presence of NAC (10 mM) for 2 h or Z-VAD-FMK (50 μM) and then exposed for 10 min to 0.05% CMI/MI. (A) After 12 h the cells were stained with AnnxV-FITC to evaluate phosphatidylserine exposure or with Rhod123 to monitor the drop in Δψm or with PI to evaluate the percentage of cells in the sub-G0/G1 phase. (B) After 1 h the cells were stained with Rhod123 to evaluate the increase of mitochondrial transmembrane potential or with DCFH-DA to evaluate ROS production.
      Pretreatment of NHK with the pan-caspase inhibitor Z-VAD-FMK (final concentration 50 μM) did not prevent the CMI/MI-induced generation of ROS, the increase in ΔΨm observed after only 1 h and the drop of Δψm at 12 h (Figure 5a, b). The appearance of subdiploid DNA content, phosphatidylserine exposure, caspase activation, and Fas expression, however, were inhibited at apoptotic concentrations. Nevertheless, when caspase activation was inhibited during cell culture up to 24 h and the NHK were exposed to apoptotic concentrations of CMI/MI, they showed necrotic phenotypes. This suggests that in the absence of caspases the cells are already committed to death (data not shown).

      Discussion

      The results of this study show that short exposure (10 min) of NHK to CMI/MI at final concentrations of 0.001% to 0.1% (0.15–15 p.p.m.) induces cytotoxic effects in a concentration-dependent manner, as demonstrated by the MTT assay, which is dependent on mitochondrial redox reactions. Flow cytometric analysis using the ROS-sensitive probe DCFH-DA showed that CMI/MI treatment is associated with early ROS generation. In fact the primary event in NHK exposed to CMI/MI appears to be a concentration-dependent increase of ROS production. Recent evidence indicates that, whereas high concentrations of ROS are toxic to cells, at low concentrations they may function as intracellular messengers to modulate signaling pathways, including apoptosis (
      • Sarafian T.A.
      • Bredesen D.E.
      Is apoptosis mediated by reactive oxygen species?.
      ). In line with these observations, we show here that exposure of NHK to low concentrations of CMI/MI (0.001–0.05%) activates the apoptotic machinery in normal cells. The appearance of subdiploid DNA content occurred at 12 h and 24 h after treatment, whereas phosphatidylserine exposure became evident after 12 h. Both nuclear and cytoplasmic features of apoptosis in NHK appeared later than in tumor cells treated with CMI/MI in the same conditions (
      • Anselmi C.
      • Ettorre A.
      • Andreassi M.
      • Centini M.
      • Neri P.
      • Di Stefano A.
      In vitro induction of apoptosis vs necrosis by widely used preservatives: 2-phenoxyethanol, a mixture of isothiazolinones, imidazolidinyl urea and 1,2-pentanediol.
      ). CMI/MI (0.001–0.05%) seems to have induced a delayed apoptotic response in NHK, which is a peculiarity of keratinocytes; in fact delayed apoptosis is also triggered by UV radiation and thermal injury (
      • Matylevitch N.P.
      • Schuschereba S.T.
      • Mata J.R.
      • Gilligan G.R.
      • Lawlor D.F.
      • Goodwin C.W.
      • Bowman P.D.
      Apoptosis and accidental cell death in cultured human keratinocytes after thermal injury.
      ;
      • Godar D.E.
      Light and death: Photons and apoptosis.
      ). At the highest concentration (0.1%), most cells appeared AnnxV and PI positive only 12 h after exposure, suggesting that the predominant form of cell death is necrosis. Thus CMI/MI-induced apoptosis is the response of NHK to a less severe injury than the one causing necrosis (
      • Kanerva L.
      Electron microscopic observations of dyskeratosis, apoptosis, colloid bodies and fibrillar degeneration after skin irritation with dithranol.
      ). Our results confirm the data for benzalkonium chloride (
      • De Saint Jean M.
      • Brignole F.
      • Bringuier A.F.
      • Bauchet A.
      • Feldmann G.
      • Baudouin C.
      Effects of benzalkonium chloride on growth and survival of Chang conjunctival cells.
      ) and ammonium quaternary compounds (
      • Debbasch C.
      • Brignole F.
      • Pisella P.J.
      • Warnet J.M.
      • Rat P.
      • Baudouin C.
      Quaternary ammonium and other preservatives' contribution in oxidative stress and apoptosis on Chang conjunctival cells.
      ), preservatives used in many ophthalmic solutions, which showed that cells undergo either necrosis or apoptosis depending on the intensity of the death stimulus.
      Mitochondria are the main organelles producing ROS and one of the main cytotoxic targets of ROS. Hydrogen peroxides, pro-oxidants, and direct cross-linkers of SH groups can induce disruption of the ΔΨm, leading to either necrosis or apoptosis depending on their strength and duration (
      • Zoratti M.
      • Szabo I.
      The mitochondrial permeability transition.
      ). The loss of ΔΨm leads to the mitochondrial release of apoptogenic proteins; this transactivates caspase-9, which then cleaves caspase-3 (
      • Zou H.
      • Henzel W.J.
      • Liu X.
      • Lutsehg A.
      • Wang X.
      Apaf-1, a human protein homologous to C. elegans CED-4, participates in cytochrome c-dependent activation of caspase-3.
      . Thus the decrease of ΔΨm is an early biochemical event in many types of apoptosis, occurring upstream of caspase activation. The data obtained with Rhod123, a mitochondrial probe, clearly show that mitochondrial function is affected by CMI/MI, although the kinetics is peculiar. An early event in the molecular mechanism induced by CMI/MI is an increase of ΔΨm, which precedes the phosphatidylserine externalization and the appearance of subdiploid DNA. The mitochondrial transmembrane potential of NHK then returns to normal and at longer times of incubation there is an evident decline, suggesting that the drop of ΔΨm occurs late in CMI/MI-induced apoptosis, concomitant with the appearance of specific apoptotic markers.
      Our results are consistent with a recent study showing an early increase of ΔΨm followed by a drop in ΔΨm as a late event in Fas-mediated apoptosis (
      • Banki K.
      • Hutter E.
      • Gonchoroff N.J.
      • Perl A.
      Elevation of mitochondrial transmembrane potential and reactive oxygen intermediate levels are early events and occur independently from activation of caspases in Fas signaling.
      ); they also agree with other studies reporting an increase of ΔΨm at an early stage of necrosis (
      • Gorczyca W.
      • Melamed M.R.
      • Darzynkiewicz Z.
      Analysis of apoptosis by flow cytometry.
      ).
      • Debbasch C.
      • Pisella P.J.
      • De Saint Jean M.
      • Rat P.
      • Warnet J.M.
      • Baudouin C.
      Mitochondrial activity and glutathione injury in apoptosis induced by unpreserved and preserved Beta-blockers on Chang conjunctival cells.
      reported that mitochondrial injury is the point of no return in the mechanism involved in apoptosis induced in Chang conjunctival cells by preservative-containing eye-drops.
      These data prompted us to evaluate the role of caspases in CMI/MI-induced apoptosis and the relationship between changes of ΔΨm and caspase activation. Indeed one of the most widely recognized events at various stages of apoptosis is the activation of caspases, a family of cysteine proteases. Caspase-3 has been well characterized and seems to act in the central pathway of the apoptotic process.
      We used two approaches to investigate the role of caspases. First we assessed the time-course of caspase activation. Caspase-3 is a likely candidate to mediate CMI/MI-induced apoptosis, as shown by the increase of DEVDase activity at all concentrations tested (except 0.1% CMI/MI), with the peak at 12 h. Moreover, caspase-9 and more interestingly also caspase-8 increased after CMI/MI treatment, with the maximal caspase-8 increase occurring before the increase of caspase-9 and -3.
      In a second approach we used the pan-caspase inhibitor Z-VAD-FMK. Whereas it inhibited the appearance of subdiploid DNA content and caspase activation, it did not prevent the CMI/MI-induced ROS generation and the changes in ΔΨm. Thus ROS production and the mitochondrial changes occurred independently of caspase activation.
      Although activation of caspase-3 and -9 is expected, the activation of caspase-8, an apical caspase, is surprising. Caspase-8 is involved in the apoptosis pathway mediated by the triggering of death receptor Fas (also known as CD95 or APO-1). Apoptotic signals triggered by the Fas system could be transmitted by caspase-8 followed by caspase-3 (
      • Wesselborg S.
      • Engels I.H.
      • Rossmann E.
      • Los M.
      • Schulze-Osthoff K.
      Anticancer drugs induce caspase-8/FLICE activation and apoptosis in the absence of CD95 receptor/ligand interaction.
      ) or by the mitochondrial pathway (
      • Luo X.
      • Budihardjo I.
      • Zou H.
      • Slaughter C.
      • Wang X.
      Bid, a Bcl2 interacting protein, mediates cytochrome c release from mitochondria in response to activation of cell surface death receptors.
      ). The increase in caspase-8 prompted us to test whether the Fas pathway is involved in CMI/MI-induced apoptosis.
      Fas expression was increased at 3 h after treatment with CMI/MI (0.01–0.05%), which is the time of maximal increase of caspase-8. The increase of Fas expression and caspase-8 activation appears to be important for apoptosis induced by CMI/MI, although it appears to be restricted to a few cells; however, the mitochondrial events could contribute to amplification of the cell death signal. The Fas system is involved in inflammatory and infectious skin diseases, such as contact dermatitis and herpesvirus infection (
      • Freiberg R.A.
      • Spencer D.M.
      • Choate K.A.
      • Peng P.D.
      • Schreiber S.L.
      • Crabtree G.R.
      • Khavari P.A.
      Specific triggering of the Fas signal transduction pathway in normal human keratinocytes.
      ;
      • Teraki Y.
      • Shiohara T.
      Apoptosis and the skin.
      ;
      • Wehrli P.
      • Viard I.
      • Bullani R.
      • Tschopp J.
      • French L.E.
      Death receptors in cutaneous biology and disease.
      ). Thus an intriguing hypothesis relating our in vitro data with in vivo effects is that Fas/FasL-mediated apoptosis may play a part in the widely described allergic contact dermatitis reactions to CMI/MI (
      • Alexander B.R.
      An assessment of the comparative sensitization potential of some common isothiazolinones.
      ); indeed most of the Fas increase occurred at a final concentration of 7.5 p.p.m., the maximum level allowed in “leave on” products. Further studies to test our hypothesis will be performed in a keratinocyte T cell coculture system, as an in vitro model.
      That the generation of ROS plays a crucial part in both necrosis and apoptosis induced by CMI/MI is supported by the finding that preincubation of NHK with the antioxidant NAC completely prevented the generation of ROS, the changes in ΔΨm and the appearance of all markers of apoptosis, including caspase activation and increased Fas expression. The timing of the NAC addition was important: NAC had to be present before CMI/MI treatment, as no protective effect was observed when it was added immediately after CMI/MI treatment. As an antioxidant, NAC can act as a scavenger of reactive intermediates either by itself or indirectly as a precursor of glutathione (
      • De Vries N.
      • De Flora S.
      N-acetyl-L-cysteine.
      ). How CMI/MI induces the generation of ROS remains to be determined. CMI/MI can interact with sulphydryl groups of enzymes and other proteins, however, and the specific chemical reaction of CMI/MI with glutathione has been reported in vitro (
      • Morris S.L.
      • Walsh R.C.
      • Hansen J.N.
      Identification and characterization of some bacterial membrane sulphydryl groups which are targets of bacteriostatic and antibiotic action.
      ;
      • Collier P.J.
      • Ramsey A.
      • Waigh R.D.
      • Douglas K.T.
      • Austin P.
      • Gilbert P.
      Chemical reactivity of some isothiazolone biocides.
      ;
      • Gruvberger B.
      • Bruze M.
      Can glutathione-containing emollients inactivate methylchloroisothiazolinone/methylisothiazolinone?.
      ). Therefore, we can speculate that one potential target for CMI/MI within NHK are cellular thiols such as glutathione (in addition to sulphydryl groups of critical proteins) and that the subsequent decrease of the endogenous antioxidant defense system provides an oxidative trigger for cell death. Our hypothesis is supported by recent evidence suggesting that the depletion of glutathione is a stimulus leading to apoptosis in different cell types (
      • Merad-Boudia M.
      • Nicole A.
      • Santiard-Baron D.
      • Saille C.
      • Ceballos-Picot I.
      Mitochondrial impairment as an early event in the process of apoptosis induced by glutathione depletion in neuronal cells: Relevance to Parkinson's disease.
      ;
      • Ghibelli L.
      • Coppola S.
      • Fanelli C.
      • Rotilio G.
      • Civitareale P.
      • Scovassi A.I.
      • Ciriolo M.R.
      Glutathione depletion causes cytochrome c release even in the absence of cell commitment to apoptosis.
      ). As NAC can counteract the cytotoxic effects of CMI/MI only if present before CMI/MI exposure, we believe that it exerts its protective effect as a glutathione precursor rather than as a radical scavenger. Nevertheless, we cannot rule out the possibility of direct reaction of CMI/MI with NAC during the short time of incubation.
      In summary we have demonstrated for the first time that ROS generation is an early and causal step in the apoptosis and necrosis associated with CMI/MI treatment. Moreover, the induction of caspase-dependent apoptosis and the involvement of the Fas pathway may play a part in allergic contact dermatitis reactions to the preservative. This new aspect of the in vitro toxicity of CMI/MI may provide important information about the relationship between the preservative's in vitro apoptotic activity and its in vivo toxicity.

      ACKNOWLEDGMENTS

      The authors are grateful to Mrs Angela Petruzzelli, Mrs Giovanna Bernini, and Mr Giancarlo Mariotti for their excellent technical assistance with the keratinocyte cultures; we thank Dr Simona Tavarini and Dr Sandra Nuti for helpful discussion regarding the cytofluorimetric analysis and Dr Peter Christie for the careful English revision. This research was supported by grants from Piano di Ateneo per la Ricerca and Progetto Giovani Ricercatori (Dr Marco Andreassi).

      References

        • Alexander B.R.
        An assessment of the comparative sensitization potential of some common isothiazolinones.
        Contact Dermatitis. 2002; 46: 191-196
        • Anselmi C.
        • Ettorre A.
        • Andreassi M.
        • Centini M.
        • Neri P.
        • Di Stefano A.
        In vitro induction of apoptosis vs necrosis by widely used preservatives: 2-phenoxyethanol, a mixture of isothiazolinones, imidazolidinyl urea and 1,2-pentanediol.
        Biochem Pharmacol. 2002; 63: 437-453
        • Banki K.
        • Hutter E.
        • Gonchoroff N.J.
        • Perl A.
        Elevation of mitochondrial transmembrane potential and reactive oxygen intermediate levels are early events and occur independently from activation of caspases in Fas signaling.
        J Immunol. 1999; 162: 1466-1479
        • Bradford M.M.
        A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding.
        Anal Biochem. 1976; 72: 248-254
        • Bruze M.
        • Gruvberger B.
        • Persson K.
        Contact allergy to the active ingredients of Kathon CG in the guinea pig.
        Derm Beruf Umwelt. 1987; 35: 165-168
        • Bruze M.
        • Freget S.
        • Gruvberger B.
        • Persson K.
        Contact allergy to the active ingredients of Kathon CG in the guinea pig.
        Acta Derm Venereol. 1987; 67: 315-320
        • Bruze M.
        • Dahlquist I.
        • Gruvberger B.
        Contact allergy to dichlorinated methylisothiazolinone.
        Contact Dermatitis. 1989; 20: 219-239
        • Chen Q.M.
        • Liu J.
        • Merrett J.B.
        Apoptosis or senescence-like growth arrest: Influence of cell cycle position, p53, 21 and bax in H2O2 response of normal human fibroblasts.
        Biochem J. 2000; 347: 543-551
        • Collier P.J.
        • Ramsey A.
        • Waigh R.D.
        • Douglas K.T.
        • Austin P.
        • Gilbert P.
        Chemical reactivity of some isothiazolone biocides.
        J Appl Bacteriol. 1990; 69: 578-584
        • Darzynkiewicz Z.
        • Bedner E.
        • Smolewski P.
        Flow cytometry in analysis of cell cycle and apoptosis.
        Semin Hematol. 2001; 38: 179-193
        • De Saint Jean M.
        • Brignole F.
        • Bringuier A.F.
        • Bauchet A.
        • Feldmann G.
        • Baudouin C.
        Effects of benzalkonium chloride on growth and survival of Chang conjunctival cells.
        Invest Ophthalmol Visual Sci. 1999; 40: 619-630
        • De Vries N.
        • De Flora S.
        N-acetyl-L-cysteine.
        J Cell Biochem. 1993; 17F: 270-277
        • Debbasch C.
        • Brignole F.
        • Pisella P.J.
        • Warnet J.M.
        • Rat P.
        • Baudouin C.
        Quaternary ammonium and other preservatives' contribution in oxidative stress and apoptosis on Chang conjunctival cells.
        Invest Ophthalmol Visual Sci. 2001; 42: 642-652
        • Debbasch C.
        • Pisella P.J.
        • De Saint Jean M.
        • Rat P.
        • Warnet J.M.
        • Baudouin C.
        Mitochondrial activity and glutathione injury in apoptosis induced by unpreserved and preserved Beta-blockers on Chang conjunctival cells.
        Invest Ophthalmol Visual Sci. 2001; 42: 2525-2533
        • Di Stefano A.
        • Ettorre A.
        • Sbrana S.
        • Giovani C.
        • Neri P.
        Purpurin-18 in combination with light leads to apoptosis or necrosis in HL60 leukemia cells.
        Photochem Photobiol. 2001; 73: 290-296
        • van Engeland M.
        • Nieland L.J.W.
        • Ramaekers F.C.S.
        • Schutte B.
        • Reutelingsperger C.P.M.
        Annexin V-affinity assay. A review on an apoptosis detection system based on phosphatidylserine exposure.
        Cytometry. 1998; 31: 1-9
        • Freiberg R.A.
        • Spencer D.M.
        • Choate K.A.
        • Peng P.D.
        • Schreiber S.L.
        • Crabtree G.R.
        • Khavari P.A.
        Specific triggering of the Fas signal transduction pathway in normal human keratinocytes.
        J Biol Chem. 1996; 271: 31666-31669
        • Ghibelli L.
        • Coppola S.
        • Fanelli C.
        • Rotilio G.
        • Civitareale P.
        • Scovassi A.I.
        • Ciriolo M.R.
        Glutathione depletion causes cytochrome c release even in the absence of cell commitment to apoptosis.
        FASEB J. 1999; 13: 2031-2036
        • Godar D.E.
        Light and death: Photons and apoptosis.
        J Invest Dermatol Symp Proc. 1999; 4: 17-23
        • Gorczyca W.
        • Melamed M.R.
        • Darzynkiewicz Z.
        Analysis of apoptosis by flow cytometry.
        in: Jaroszeski M.J. Heller R. Methods in Molecular Biology, Vol. 91: Flow Cytometry Protocols. Humana Press Inc., Totowa, NJ1998: 217-238
        • Gruvberger B.
        • Bruze M.
        Can glutathione-containing emollients inactivate methylchloroisothiazolinone/methylisothiazolinone?.
        Contact Dermatitis. 1998; 38: 261-265
        • Kanerva L.
        Electron microscopic observations of dyskeratosis, apoptosis, colloid bodies and fibrillar degeneration after skin irritation with dithranol.
        J Cutan Pathol. 1990; 17: 37-44
        • Law A.B.
        • Moss N.J.
        • Lashen E.S.
        • Kathon C.G.
        A new single-component, broad spectrum preservative system for cosmetics and toiletries.
        in: Kabara J.J. Cosmetics and Drug Preservation, Principles and Practice. Marcel Dekker, New York1984: 533-557
        • Leverkus M.
        • Yaar M.
        • Gilchrest B.A.
        Fas/Fas ligand interaction contributes to UV-induced apoptosis in human keratinocytes.
        Exp Cell Res. 1997; 232: 255-262
        • Luepke N.P.
        Hen's egg chorioallantoic membrane test for irritation potential.
        Food Chem Toxicol. 1985; 23: 287-291
        • Luo X.
        • Budihardjo I.
        • Zou H.
        • Slaughter C.
        • Wang X.
        Bid, a Bcl2 interacting protein, mediates cytochrome c release from mitochondria in response to activation of cell surface death receptors.
        Cell. 1998; 94: 481-490
        • Matylevitch N.P.
        • Schuschereba S.T.
        • Mata J.R.
        • Gilligan G.R.
        • Lawlor D.F.
        • Goodwin C.W.
        • Bowman P.D.
        Apoptosis and accidental cell death in cultured human keratinocytes after thermal injury.
        Am J Pathol. 1998; 153: 567-577
        • Merad-Boudia M.
        • Nicole A.
        • Santiard-Baron D.
        • Saille C.
        • Ceballos-Picot I.
        Mitochondrial impairment as an early event in the process of apoptosis induced by glutathione depletion in neuronal cells: Relevance to Parkinson's disease.
        Biochem Pharmacol. 1998; 56: 645-655
        • Monte W.C.
        • Ashoor S.H.
        • Lewis B.J.
        Mutagenicity of two non-formaldehyde-forming antimicrobial agents.
        Food Chem Toxicol. 1983; 21: 695-696
        • Morris S.L.
        • Walsh R.C.
        • Hansen J.N.
        Identification and characterization of some bacterial membrane sulphydryl groups which are targets of bacteriostatic and antibiotic action.
        J Biol Chem. 1984; 259: 13590-13594
        • Mosmann T.
        Rapid colorimetric assay for cellular growth and survival: Application to proliferation and cytotoxicity assays.
        J Immunol Methods. 1983; 65: 55-63
        • Nicoletti I.
        • Migliorati G.
        • Pagliacci M.C.
        • Grignani F.
        • Riccardi C.
        A rapid and simple method for measuring thymocyte apoptosis by propidium iodide staining and flow cytometry.
        J Immunol Methods. 1991; 139: 271-279
        • Rivalland P.
        • Vie K.
        • Coiffard L.
        • De Roeck-Holtzhauer Y.
        Cytotoxicity test of antibacterial agents on human fibroblasts cultures.
        Pharm Acta Helv. 1994; 69: 159-162
        • Sarafian T.A.
        • Bredesen D.E.
        Is apoptosis mediated by reactive oxygen species?.
        Free Radic Res. 1994; 21: 1-8
        • Stejskal V.D.
        • Forsbeck M.
        • Nilsson R.
        Lymphocytes transformation test for diagnosis of isothiazolinone allergy in man.
        J Invest Dermatol. 1990; 94: 798-802
        • Sureda F.X.
        • Gabriel C.
        • Comas J.
        • Palls M.
        • Escubedo E.
        • Camarasa J.
        • Camins A.
        Evaluation of free radical production, mitochondrial membrane potential and cytoplasmic calcium in mammalian neurons by flow cytometry.
        Brain Res Protocols. 1999; 4: 280-287
        • Teraki Y.
        • Shiohara T.
        Apoptosis and the skin.
        Eur J Dermatol. 1999; 9: 413-426
        • Wehrli P.
        • Viard I.
        • Bullani R.
        • Tschopp J.
        • French L.E.
        Death receptors in cutaneous biology and disease.
        J Invest Dermatol. 2000; 115: 141
        • Wesselborg S.
        • Engels I.H.
        • Rossmann E.
        • Los M.
        • Schulze-Osthoff K.
        Anticancer drugs induce caspase-8/FLICE activation and apoptosis in the absence of CD95 receptor/ligand interaction.
        Blood. 1999; 93: 3053-3063
        • Zoratti M.
        • Szabo I.
        The mitochondrial permeability transition.
        Biochim Biophys Acta. 1995; 1241: 139
        • Zou H.
        • Henzel W.J.
        • Liu X.
        • Lutsehg A.
        • Wang X.
        Apaf-1, a human protein homologous to C. elegans CED-4, participates in cytochrome c-dependent activation of caspase-3.
        Cell. 1997; 90: 405-413