Transglutaminase Participates in UVB-Induced Cell Death Pathways in Human Corneal Epithelial Cells

Abstract

purpose. Ultraviolet light (UVB) is known to cause apoptosis in human corneal epithelial cells. This study evaluates the role of transglutaminase in regulating tumor necrosis factor (TNF) receptor clustering as well as caspase activation in UVB-induced apoptosis in human corneal epithelial cells.

methods. A human corneal epithelial cell line was used. A single dose of UVB (20 mJ/cm²) was used as a stimulus. Cell viability and cell death were investigated by MTT, terminal deoxynucleotidyl transferase-mediated dUTP-digoxigenin nick end labeling (TUNEL), and caspase-3 assays. Immunofluorescent staining was used to investigate TNF receptor-I clustering at various time intervals after UVB. Short interfering RNA was used to knock down transglutaminase-2 expression. Fluorescein-cadaverine uptake was used to assess transglutaminase activity. A noncovalent peptide delivery system was used to transfect guinea pig liver transglutaminase into corneal epithelial cells.

results. UVB increased transglutaminase activity, reduced cell viability, and increased TUNEL staining. UVB or TNF-α promoted TNF-receptor-I clustering, a process inhibited by the transglutaminase inhibitor, mono-dansyl cadaverine. UVB also increased activated caspase-3, in a manner suppressible by mono-dansyl cadaverine. Intracellular delivery of exogenous transglutaminase markedly increase caspase-3 activation compared with the vehicle control.

conclusions. Transglutaminase enzymatic activity is involved in corneal epithelial cell death after UVB and appears to participate in two steps regulating this process, clustering of TNF receptor-I and caspase-3 activation.

Ultraviolet radiation (UVB) causes apoptosis ^{1 2 3} in corneal epithelial cells. UVB radiation of the corneal epithelium also produces clinically significant consequences, such as inflammation and cell death, with features such as nuclear fragmentation and loss of tissue cohesion. ⁴ A previous study has shown that transglutaminase activity is detectable in the corneal epithelium. ⁵ On the ocular surface, transglutaminase has been proposed to play a role in wound healing ⁶ and diverse clinical diseases such as pterygium formation, ⁷ allergic conjunctivitis, ⁸ dry eye, ^{9 10} and Stevens-Johnson syndrome. ¹¹ Although UVB has been observed to activate transglutaminase, this response varies greatly, depending on the cell type, ¹² and its effect on transglutaminase activation in corneal epithelium has not been investigated. 

Mammalian transglutaminases belong to a group of at least eight enzymes that catalyze the polymerization of cellular proteins, a few of which are known to be involved in the regulation of apoptosis. ¹³ Type 2 transglutaminase or tissue transglutaminase (TGM-2), in particular, is known to be involved in the effector pathways ^{13 14} for cell death. It can be proapoptotic by polymerizing BAX, ¹⁵ or it can be antiapoptotic by reducing anoikis. ¹⁵ In a nonocular cell line, cytosolic transglutaminase has been found to be proapoptotic, whereas nuclear transglutaminase was noted to attenuate apoptosis. ¹⁶ The pro- or antiapoptotic effects of TGM-2 in corneal epithelial cells are unknown. 

In a nonocular cell line, UVB has been shown to induce clustering of tumor necrosis factor (TNF) receptors. ¹⁷ Murine corneal epithelial cells has been shown to express TNF receptors that mediate TNF-α-induced cell death. ¹⁸ Because TNF receptor-I is the major cell surface receptor subtype responding to TNF-α ligand, a known stimulus for apoptosis, clustering of these receptors to form signaling complexes is expected to be an important early event in apoptosis. ¹⁹ Similarly, caspase activation is a well-recognized phenomenon in UVB-induced cell death in corneal epithelial cells. ² The effects of TGM-2 on the key apoptotic signaling molecules TNF receptor-I and caspase-3 are unknown. 

The purpose of this study was to evaluate the role of TGM-2 in the UVB-induced death pathways in corneal epithelial cells—in particular, the role of TGM-2 on TNF receptor clustering, caspase-3 activation, and nuclear morphology. 

Methods

Materials and Reagents

Rabbit anti-tissue transglutaminase polyclonal antibody (clone ab421) was from Novus Biologicals (Littleton, CO); anti-active caspase 3 antibody (no. 557035) from BD Biosciences (San Diego, CA); mouse monoclonal anti-β actin antibody (clone AC-15) from Sigma-Aldrich (St. Louis, MO); goat anti-rabbit AlexaFluor 594 antibodies and goat anti-mouse AlexaFluor 488 antibodies from Invitrogen-Molecular Probes (Eugene, OR); recombinant human TNF-α or TNFSF1A (no. 210-TA-010), MTT reagent (TA5355, TACS) and detergent reagent from R&D Systems, Inc. (Minneapolis, MN); a cell viability assay (ApopTag Fluorescein Kit; S7110) from Chemicon (Temecula, CA); mono-dansyl cadaverine (MDC) and fluorescent cadaverine (no. A10466) from Invitrogen-Molecular Probes; cresyl violet (no. 192) from Harleco (Gibbstown, NY); guinea pig liver transglutaminase (no. T5398) from Sigma-Aldrich; a protein delivery system (Chariot) from ActiveMotif (Carlsbad, CA); glass coverslips (no. 70462, Secureslip) from Electron Microscopy Sciences (Hatfield, PA); predesigned siRNA (Silencer) from Ambion (Austin, TX); RNA transfection reagent (no. 301705, HiPerfect) from Qiagen (Valencia, CA); first-strand beads (Ready-To-Go You-Prime) from GE Healthcare (Piscataway, NJ); and PCR master mix (AmpErase UNG) MGB probes (Taqman) for TGM-2 (assay Hs 00190278_m1) from Applied Biosystems (ABI; Foster City, CA). 

Cell Culture Procedure

A human corneal epithelial cell (T-HCEC) line at passages 60 to 70 was used in all experiments and was provided courtesy of Kaoru Araki-Sasaki (Kinki Central Hospital, Hyogo, Japan). ²⁰ Culture media and conditions were as described previously ²¹ except that 10% fetal bovine serum was used instead of 5% and epidermal growth factor, cholera toxin, and hydrocortisone were not used. In all cases in which immunofluorescent staining was desired, cells were cultured in wells with a removable glass coverslip at the base (Secureslip; Electron Microscopy Sciences), which could be transferred to a glass slide for purposes of imaging. 

Ultraviolet Procedure

Ultraviolet irradiation was performed as described previously. ²² For experiments with UVB exposure, cell cultures were observed daily under inverted light microscopy and experiments commenced when cells became just confluent. MDC, a chemical inhibitor used previously to inhibit transglutaminase in cell cultures, ²³ was used at a final concentration of 100 μM and was introduced 1 hour before any UVB exposure. 

RNA Interference and Real-Time Polymerase Chain Reaction

Short interfering (si)RNA against TGM-2 (no. 111472) and fluorescein-conjugated siRNA (no. 1022079; both from Ambion) were used. The sense and antisense sequences (5′–3′) of the siRNA against TGM-2 were GGCCCGUUUUCCACUAAGAtt and UCUUAGUGGAAAACGGGCCtt, respectively. The dsRNA sequence targeting mRNA for TGM2 (accession number in GenBank: NM_198951) corresponded to part of exon 3 of the TGM-2 gene. The sense and antisense sequences (5′–3′) of the fluorescein-conjugated siRNA were UUCUCCGAACGUGUCACGUdTdT and UUCUCCGAACGUGUCACGUdTdT, respectively. Details of siRNA transfection with the reagent (RNA HiPerfect; Qiagen) have been described (Chen LZ, et al. IOVS 2005;46:ARVO E-Abstract 2109). Briefly, the ratio of siRNA to transfection reagent used was 1:8. The silencing or nonsilencing (control fluorescein-conjugated siRNA) siRNA (1 μg per well in the case of a single well in a 12-well plate) was applied at a final concentration of 67 nM with the fast-throw technique according to the manufacturer’s protocol. UVB stimulus, when applicable, was applied 24 hours after transfection. 

Real-time reverse transcription–polymerase chain reaction was performed as previously described. ^{24 25} Briefly, the first-strand cDNA was synthesized from 1 μg of total RNA with random hexamer using M-MuLV reverse transcriptase (Ready-To-Go You-Prime First-Strand Beads; GE Healthcare). Real-time PCR was performed using specific MGB probes (Taqman), a universal PCR Master Mix (Taqman AmpErase UNG), and a thermocycler (Smart Cycler System; Cepheid, Sunnyvale, CA) according to the manufacturer’s recommendations. A nontemplate control was included to detect DNA contamination. The GAPDH gene was used as an endogenous reference for each reaction to correct differences in the amount of total RNA added. 

Electron Microscopy

Corneal explant tissue from cadaveric human limbus was prepared and cut as previously described. ²¹ After UVB irradiation, as described earlier, the corneal epithelial explants were carefully removed and prepared for transmission electron microscopy, as previously described. ^{26 27} Briefly, the tissue was fixed in 3% glutaraldehyde, postfixed in 1% osmium tetroxide and dehydrated in a graded series of ethanol and acetone. This was followed by infiltration with acetone and Epon plastic resin and embedding in Epon plastic resin. Thick and thin tissue sections were obtained and examined. Imaging was then performed (model 902 electron microscope; Carl Zeiss MicroImaging, Inc., Thornwood, NY). 

Cell Viability Assays

MTT assay was performed per the manufacturer’s instructions. Briefly, a standard curve of known cell concentrations against absorbance showed that a portion of the curve was approximately linear. This was used to derive the optimal seeding concentration of 2 × 10⁵ cells/mL. For each experimental condition, cells were seeded at this density in a volume of 100 μL per well on a 96-well plate in triplicate. The tetrazolium compound MTT (3-[4,5-dimethylthiazol-2-yl]2,5-diphenyltetrazolium bromide) was added 24 hours after UVB. Detergent reagent was added 2 hours after addition of MTT and incubated at 37°C for a further 24 hours. The absorbance was then read at 570 nm using a reference wavelength of 650 nm. 

Cresyl violet staining was performed as a simple method of documenting attached cells. This stain has a known propensity for cell nuclei. ²⁸ Cells were washed 24 hours after UVB with PBS and stained with 1% cresyl violet for 30 minutes, followed by further washing with PBS. 

TUNEL Assay

The terminal deoxynucleotidyl transferase-mediated dUTP-digoxigenin nick end labeling (TUNEL) assay was performed (ApopTag Fluorescein kit; Chemicon), as previously described. ¹⁰ Briefly, cells were seeded into an eight-well chamber slide for this experiment. UVB irradiation was performed as described earlier, and the TUNEL protocol was performed 8 and 16 hours after UVB exposure. Sampling of cells was performed by counting propidium iodide stained nuclei for each condition. At least 400 cells were counted for each experimental group. Positive cells were identified by fluorescent imaging using the same exposure, brightness, contrast, and magnification in all experimental groups. This procedure was facilitated by overlapping fluorescent images with propidium iodide–stained images in different layers of a composite image (Photoshop, ver. 6.0; Adobe Systems Inc., San Jose, CA). 

Immunofluorescent Staining

Immunofluorescent staining ^{27 29} for TGM-2, TNF receptor-I, and active caspase-3 ¹⁰ was performed as previously described. Staining for TNF receptor-I was performed at various time intervals after UVB. Caspase-3 was evaluated at 6 hours after UVB. Anti-TNF receptor-I, anti-TGM-2, and anti-caspase-3 antibodies used were diluted 20-, 50-, and 100-fold respectively. The appropriate secondary antibody (1:300 dilution or 6.7 μg/mL) was applied in 5% goat serum for 30 minutes in a dark incubation chamber. Counterstaining of cell nuclei was performed using 1 μg/mL propidium iodide. Counting of the positively stained cells was performed as described for the TUNEL assay. 

Measurement of Nuclear Characteristics

Nuclear size and shape were evaluated in randomly obtained images of propidium iodide–stained cells at 400× magnification. The greatest linear dimension (L) was measured, as well as the longest width perpendicular to this measurement (w). The area of each cell was calculated as π(L/2)(w/2) and the roundness index as w/L, where an index of 1 indicates perfect roundness. A minimum of 400 randomly selected cells for each experimental condition were measured by a single observer. Nuclear images were digitally magnified by a standardized extent to facilitate measurements (Photoshop; Adobe Systems, Inc.). 

Fluorescein Cadaverine Assay

Transglutaminase activity was determined by a fluorescein cadaverine incorporation assay. ^{30 31} Cells for this assay was grown on glass coverslips until confluence. After UVB, FC was added to the culture media at a concentration of 0.5 mM, and the cells were incubated in the dark for 24 hours. After incubation, the coverslips with cells were transferred to a glass slide and washed four times in PBS for 2 minutes each. Immunofluorescent staining was imaged using the same exposure settings for all images. The relative intensity of the fluorescence as subjectively evaluated on digital images was a measure of the relative activity of transglutaminase. 

Delivery of Exogenous Transglutaminase

Chariot (ActivMotif) is a transfection reagent that can transport biologically active proteins, peptides, and antibodies into mammalian cells. ³² The reagent contains a peptide that forms a noncovalent complex with the macromolecule to be delivered. On introduction to cells, the carrier–macromolecular complex is rapidly internalized and dissociated. The carrier peptide is then localized to the nucleus where it is degraded. Exogenous transglutaminase was delivered to cultured corneal epithelial cells using the transfection system according to the manufacturer’s instructions. Briefly, diluted reagent was incubated with 1 μg or 0.5 μg of guinea pig liver transglutaminase for 30 minutes at room temperature. Subsequently, the protein–reagent complex was added to wells containing cultured corneal epithelial cells in a 12-well plate for 1 hour and then replenished with complete serum containing medium, without removal of the existing medium. Immunofluorescence staining was performed 24 hours after incubation of cells with the introduced protein complex. 

Statistical Analysis

All experiments were performed at least three times. Statistical analysis was performed on computer (Excel; Microsoft, Redmond, WA; and Prism 4 for Windows, ver. 4.03; GraphPad Software Inc., San Diego, CA). Analysis of variance (ANOVA) with the Tukey post hoc test was used for statistical comparisons, and statistical significance was set at the level of α = 0.05. For proportions of cells staining positive with TUNEL or active caspase-3, the 95% confidence interval (CI) of the proportion was calculated, assuming a binomial distribution. 

Results

Effect of UVB on Apoptotic Cell Death and Transglutaminase Activity

Reduced cresyl violet staining of corneal epithelial cells was observed after UVB exposure (Fig. 1A) . Similarly, cell viability demonstrated by MTT assay decreased after UVB (Fig. 1B) . Ultraviolet irradiated cells showed ultrastructural changes in the form of blebbing (Fig. 1C , bottom left) and nuclear fragmentation and formation of apoptotic bodies (Fig. 1C , bottom right), features that were not seen in control cells (Fig. 1C , top panel). Increased fluorescent labeling of DNA nick ends with the TUNEL assay after UVB exposure, a feature indicative of apoptotic cell death, was detected (Fig. 1D , left). This labeling was shown to colocalize with fragmented nuclei (Fig. 1D , right). 

UVB stimulated transglutaminase functional activity (data not shown) in corneal epithelial cells compared with the control, according to the results of the fluorescent cadaverine uptake assay. MDC, a known inhibitor of transglutaminase, reduced the UVB-induced transglutaminase activity. This suggests that in the following experiments, MDC is an effective inhibitor for assessing the role of transglutaminase in key biological processes that are involved in cultured corneal epithelial cell death. 

Effect of Transglutaminase Activity on UVB-Induced Clustering of TNF Receptors

The effect of UVB exposure of cultured corneal epithelial cells on TNF receptor-I staining was investigated. The pattern of immunofluorescent staining for TNF receptors-I in corneal epithelial cells was altered as early as 5 minutes after stimulation (Fig. 2 , top row, second column from left). In control cells (Fig. 2 , top row, first column), weak TNF receptor-I staining was seen in the perinuclear area. In UVB-exposed cells (Fig. 2 , second and third columns), stronger staining could be seen in the vicinity of cellular membranes, corresponding to clustered receptors, as well as in the cytoplasm, corresponding to endocytosed receptors. Such an increase in intensity of staining for this cell surface receptor within minutes of stimulation has been considered to be evidence of clustering of receptors. ¹⁷ MDC had a suppressive effect on the UVB-induced clustering (Fig. 2 , bottom row). Short interfering RNA against TGM-2 (siTGM-2) was effective in lowering the level of TGM-2 transcripts in cells (Fig. 2 , bottom right). siTGM-2 was also found to have a suppressive effect on the UVB-induced clustering of TNF receptor-I. As a positive (ligand) control, the clustering effect was demonstrated by incubating cells with TNF-α (Fig. 2 , top right). 

Role of Transglutaminase in Caspase-3 Formation

The proportion of caspase-3–positive cells was increased by UVB or by incubation with recombinant TNF-α (Fig. 3A) . At 6 hours after UVB, caspase-3 activation was reduced in cells incubated with MDC compared with the control (Fig. 3A) . In control cells, the proportion of cells staining with active caspase-3 was similar to UVB-exposed cells incubated with MDC (Fig. 3A) . In cells treated with TNF-α, the UVB-induced caspase activation at 6 hours was also reduced by incubation with MDC (Fig. 3A) . 

The protein delivery system (Chariot; ActivMotif) was able to deliver exogenous transglutaminase into the cytoplasm and nuclei of cultured corneal epithelial cells, demonstrated by significantly increased transglutaminase immunofluorescence staining (Fig. 3B , top row), compared with control cells treated with vehicle alone. In Western blot assays, the anti-transglutaminase antibody used in this study (ab421) strongly detected the guinea pig transglutaminase in lysate-free protein solution, as well as endogenous transglutaminase in corneal epithelial cell lysates, in preliminary experiments (data not shown). Therefore, the immunofluorescent staining in cells after delivery of exogenous transglutaminase is likely to represent exogenous and endogenous transglutaminase. Delivery of exogenous transglutaminase to cells was associated with increased activated caspase-3 on immunofluorescent staining (Fig. 3B , bottom row). 

Suppression of Apoptotic Nuclear Changes by MDC, a Transglutaminase Inhibitor

UVB decreased nuclear size (Fig. 4A)and reduced nuclear roundness (Fig. 4B) . Cell nuclei exposed to UVB were significantly smaller and more ovoid than were the nonirradiated cells. The ranges of shape and size were also greater than the control, suggesting a more variable morphologic appearance of nuclei in irradiated cells. MDC suppressed the UVB-induced change in nuclear size (Fig. 4A)and roundness (Fig. 4B)and also suppressed the UVB-induced DNA fragmentation shown by TUNEL staining (Fig. 4C) . Only data for the TUNEL staining at 8 hours are shown. 

Discussion

In this study, UVB reduced corneal epithelial cell viability and increased the markers of apoptosis, including nuclear shrinkage, alteration of nuclear shape, nuclear fragmentation, and immunocytochemical labeling of DNA nick ends. UVB also upregulated transglutaminase activity, which is associated with the acute, time-dependent clustering of TNF receptor-I, as well as caspase-3 activation. 

The localization of TGM-2 in cell membranes ³³ and its incorporation during invagination ³⁴ support the existence of a membrane-related role. In addition, TGM-2 has been detected in cycling endosomes ³⁵ and it is implicated in receptor externalization. ³⁶ The role of MDC in the inhibition of receptor-mediated endocytosis is supported by the fact that it failed to inhibit endocytosis when multivalent antibodies cross-linked cell surface receptors in the place of transglutaminase. ³⁷ The specificity of MDC for transglutaminase is shown by the finding that it does not inhibit endocytosis in transglutaminase-deficient cells. ³⁸ Although the role of transglutaminase in endocytosis is not fully resolved, ³⁹ UVB and hyperosmolarity have been shown to induce the clustering of cell surface receptors in a fashion similar to that of their cytokine ligands. ¹⁷  

Although the involvement of cell surface receptors such as TNF receptor-I is a well-known cell signaling mechanism for activating apoptotic pathways, ⁴⁰ the role of transglutaminase in this process has not been evaluated. In the present study, recombinant TNF-α was used as a ligand control for immunofluorescent staining, and it is interesting to note that in liver cells this cytokine actually stimulated transcription driven by the TGM-2 gene promoter. ⁴¹ As UVB is a prominent stimulus of corneal disease ^{3 42} and apoptosis is a major event in corneal disease, ^{1 2 10} mechanisms regulating apoptosis (such as transglutaminase activation) in these cells have therapeutic implications. Transglutaminase inhibition using a small peptidomimetic that reduces transglutaminase-induced phospholipase function has proven clinical efficacy in a guinea pig model of allergic conjunctivitis. ⁸ Our study suggests that inhibition of clustering of cell surface receptors such as TNF receptor-I by transglutaminase inhibitors may be an alternative powerful strategy to minimize apoptosis in ocular surface diseases such as keratoconjunctivitis sicca. 

This study used cultured corneal epithelial cells in serum-containing medium, thus the effect of TGM-2 in death signaling requires further study using intact corneas or whole animals, which would be of greater physiological relevance. Another limitation of our study is that we did not show a dose–response curve for activation of TGM-2, and we did not evaluate the effect of more than one dose of UVB radiation. A further limitation of the study is that we have not investigated the specificity of the various antibodies used in our experiments, although this has been previously evaluated. ^{43 44 45 46 47}  

In addition to the roles we have investigated, transglutaminase has been known to play additional roles in cell death. In dying cells, it was found to prevent inflammation by cross-linking and stopping the leakage of potentially antigenic molecules. ⁴⁸ It also facilitated phagocytic recognition and hence engulfment by exposing phosphatidylserine on surfaces of dying cells. ¹⁵ In certain nonocular systems, transglutaminase may protect against cell death. ⁴⁹ The desirability of inhibiting transglutaminase in clinical scenarios therefore require further evaluation. In some instances the role of TGM-2 in death signaling may be related to its GTPase activity instead of its transamidation activity. ¹⁵ Recently, TGM-2 has been found to possess kinase activity, ⁵⁰ which could affect apoptosis. 

We have shown that transglutaminase is involved in regulating two key stages of apoptosis, in TNF receptor-I clustering (extrinsic apoptotic pathway) and caspase-3 activation (common apoptotic pathway). In corneal epithelial cells, there may be multiple convergent pathways that are independent of cell surface receptors, such as those involving mitochondrial pathways of apoptosis. ^{13 14} In future studies, we hope to elucidate the transglutaminase dependence of other death signaling pathways. 

 Supported by National Eye Institute Grants, EY11915 (SCP) and EY014553 (D-QL); a fellowship from Agency for Science Technology and Research, Singapore; an unrestricted Grant from Research to Prevent Blindness; the Oshman Foundation; and The William Stamps Farish Fund.

 Submitted for publication April 11, 2006; revised May 21, 2006; accepted July 10, 2006.

 Disclosure: L. Tong, None; Z. Chen, None; C.S. De Paiva, None; R. Beuerman, None; D.-Q. Li, None; S. C. Pflugfelder, None

 Presented at the annual meeting of the Association for Research in Vision and Ophthalmology, Fort Lauderdale, Florida, May 2005.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

UVB stimulated apoptosis and increased transglutaminase activity. (A) Results of cresyl violet staining of cells 24 hours after UVB. (B) Results of MTT cell proliferation and viability assay 24 hours after UVB. Data are expressed as the mean ± SD. (C) Top: electron micrograph showing control corneal epithelial cells unexposed to UVB. Bottom left: electron micrograph showing blebbing and vacuolation (arrow) in a corneal epithelial cell exposed to UVB. Bottom right: electron micrograph showing intracellular apoptotic bodies (arrow, an example) in a corneal epithelial cell exposed to UVB. (D) TUNEL staining in control and UVB-exposed cells (left) and propidium iodide counterstaining in the UVB-exposed cells (right) show the colocalization of fluorescence from TUNEL and fragmented nuclei stained by propidium iodide in UVB-exposed cells.

Figure 2.

 

Transglutaminase inhibition and TNF receptor I clustering. Top row: immunofluorescent staining for TNF receptor-I performed various time intervals after UVB, and ligand control (incubation with 100 ng/mL TNF-α). Middle row: cells incubated with MDC before stimulation with UVB and then incubated for the indicated time intervals. Bottom row: cells incubated with siRNA against TGM-2 (siTGM-2) or fluorescein-conjugated nonspecific siRNA (siF) before exposure to UVB.

Figure 3.

 

Transglutaminase activity was related to caspase activation. (A) Proportion of cells with positive staining for active caspase-3. Error bars, 95% CI. (B) Top: immunofluorescent staining for TGM-2 after delivery of exogenous TGM-2 to nonirradiated corneal epithelial cells. Bottom: immunofluorescent staining for caspase-3 after delivery of exogenous TGM-2 to nonirradiated corneal epithelial cells. The secondary antibody was goat anti-rabbit AlexaFlor-594.

Figure 4.

 

Box plots showing apoptotic nuclear morphology measured on digital images: (A) Area of cell nuclei and (B) shape of cell nuclei (roundness index). Boxes: interquartile ranges; whiskers: 95% CI intervals. ANOVA probability (below the x-axes) and probabilities of significant multiple comparisons (above the boxes) are also shown. (C) Results of TUNEL staining performed at 6 hours after UVB. TUNEL-positive cells were counted. Data points are the mean and error bars, 95% CI.

The authors thank Ralph M. Nichols for his kind assistance with the electron microscopy.