Vitamin D protects keratinocytes from deleterious effects of ionizing radiation

Summary

Background Radiotherapy can induce severe skin responses that may limit the clinically acceptable radiation dose. The responses include erythema, dry and moist desquamation, erosions and dermal–epidermal blister formation. These effects reflect injury to, and reproductive failure of, epidermal cells and may also be due to dysregulation of the tissue remodelling process caused by excessive proteolytic activity. Calcitriol, the hormonally active vitamin D metabolite, protects keratinocytes from programmed cell death induced by various noxious stimuli.

Objective To examine whether calcitriol protects proliferating keratinocytes from the damage inflicted by ionizing radiation under conditions similar to those employed during radiotherapy.

Methods Autonomously proliferating HaCaT keratinocytes, used as a model for basal layer keratinocytes, were irradiated using a linear accelerator. Cell death was monitored by vital staining, executioner caspase activation, lactic dehydrogenase release and colony formation assay. Induction of matrix metalloproteinase‐9 was assessed by gelatinase activity assay and mRNA determination. Levels of specific proteins were determined by immunoblotting.

Results Treatment with calcitriol inhibited both caspase‐dependent and ‐independent programmed cell death occurring within 48 h of irradiation and increased the colony formation capacity of irradiated cells. These effects may be attributable to inhibition of the c‐Jun NH₂‐terminal kinase cascade and to upregulation of the truncated antiapoptotic isoform of p63. Treatment with the hormone also attenuated radiation‐induced increase in matrix metalloproteinase‐9 protein and mRNA levels.

Conclusions The results of this study suggest that active vitamin D derivatives may attenuate cell death and excessive proteolytic activity in the epidermis due to exposure to ionizing radiation in the course of radiotherapy.

apoptosis, c‐Jun NH₂‐terminal kinase, ionizing radiation, keratinocytes, metalloproteinase, p63, vitamin D

External beam radiation during cancer therapy inevitably involves skin irradiation and can induce severe skin responses. These side‐effects exacerbate the patient’s suffering and can limit the clinically acceptable radiation dose. After exposure of human skin to ionizing radiation, an acute reaction develops characterized by erythema, dry and moist desquamation, erosions and epilation. Dermal–epidermal blisters may form which lead to epidermal necrosis.¹ These acute changes depend on the dose of radiation and reflect injury to, and reproductive failure of, germinative epidermal cells. It is the ionizing effect of irradiation that accounts for the skin responses. The ionizing products are short‐lived free radicals that in the presence of oxygen form additional reactive species, notably H₂O₂ and the highly toxic ˙OH radical. Irradiated cells may die within hours of exposure to radiation by necrosis or apoptosis and thus lead to tissue damage. Surviving cells may lose their reproductive integrity even after exposure to a moderate radiation dose, and cell death may be preceded by a time span encompassing several division cycles. Radiosensitivity is directly correlated with the rate of cell division and inversely with the degree of cell differentiation.¹

The response of the skin to injury and stress often involves tissue remodelling, which requires coordinated degradation and synthesis of extracellular matrix (ECM) constituents. Matrix metalloproteinases (MMPs) play an important role in ECM degradation.² Dysregulation of the remodelling process due to excessive MMP proteolytic activity may result in tissue damage. Pertinent to radiation‐induced damage is excessive activity of MMP‐9, which is capable of disrupting the basement membrane and may contribute to the impairment of the epidermal–dermal barrier observed in irradiated skin.³

Calcitriol (1,25‐dihydroxyvitamin D₃), the hormonally active metabolite of vitamin D, is produced by a cascade of reactions starting with the photochemical synthesis of vitamin D₃ in the skin and followed by two consecutive hydroxylations that take place in the liver and kidney, but may take place also in several extrarenal sites. The epidermis is unique among calcitriol‐producing organs in containing the full enzymatic complement needed for calcitriol generation as well as its nuclear receptor, VDR (vitamin D receptor).⁴, ⁵ Calcitriol is known to modulate keratinocyte proliferation and enhance their differentiation.^{5  6–7} Also, noncalcaemic vitamin D analogues are widely used in the treatment of the skin disorder psoriasis, characterized by inflammation and keratinocyte hyperproliferation.⁸ Cumulative evidence strongly suggests that hormonally active vitamin D derivatives have an additional role in the epidermis: the protection of keratinocytes from the consequences of severe stress. Vitamin D derivatives reduced the number of apoptotic cells in ultraviolet (UV)‐irradiated epidermis⁹ and reduced chemotherapy‐induced apoptosis in hair follicles.¹⁰ We have recently shown that calcitriol protects keratinocytes from programmed cell death (PCD) induced by a variety of environmental and pathophysiological noxious stimuli, such as exposure to the inflammatory cytokine tumour necrosis factor‐α, oxidative stress, hyperosmotic shock and heat shock.^{11  12–13} This protective effect was at least partially due to attenuation of the stress‐induced activation of p38 mitogen‐activated protein kinase (MAPK) and c‐Jun N‐terminal protein kinase (JNK), known modulators of death signalling cascades.¹³

In this study we used the immortalized nontumorigenic HaCaT keratinocytes as an experimental model to examine the effect of calcitriol on the response of keratinocytes to ionizing radiation. These cells are a suitable model for proliferating basal keratinocytes, the epidermal keratinocyte population most vulnerable to ionizing radiation damage. Like basal layer keratinocytes these cells proliferate in low calcium media and their proliferation is driven by autocrine epidermal growth factor receptor ligands.⁷

This in vitro study provides evidence in support of the notion that hormonally active vitamin D derivatives may protect proliferating keratinocytes from the damage induced by ionizing radiation under conditions similar to those employed during radiotherapy.

Materials and methods

Materials

Minimum essential medium (MEM), fetal calf serum (FCS), l‐glutamine, antibiotic mixture (penstrepnystatin) and trypsin–ethylenediamine tetraacetic acid (EDTA) solution B were purchased from Biological Industries (Beit Haemek, Israel). Tissue culture dishes were purchased from Corning Glass Works (Corning, NY, U.S.A.). 1,25(OH)₂D₃ was a generous gift from Teva Pharmaceutical Company, Israel. Bovine serum albumin (BSA) fraction V was purchased from MP Biomedicals Inc. (Irvine, CA, U.S.A.). A BCA Protein Assay Kit was obtained from Pierce Biotechnology Inc. (Rockford, IL, U.S.A.). Ac‐Asp‐Glu‐Val‐Asp‐7‐amido‐4‐methylcoumarin (ac‐DEVD‐AMC), z‐Asp‐2,6‐dichlorobenzoyloxymethylketone (zD‐DCB) and SB203580 were purchased from Alexis Biochemicals (Lausen, Switzerland). Hoechst 33342, propidium iodide, pyruvate and β‐nicotinamide adenine dinucleotide reduced form (NADH) were from Sigma Chemical Co. (St Louis, MO, U.S.A.). Mouse monoclonal antibodies directed against the dually phosphorylated JNK and against p63 (sc‐8431) were purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, CA, U.S.A.); mouse monoclonal antibody against p73 was from Oncogene Research Products (Cambridge, MA, U.S.A.). Rabbit anti‐JNK polyclonal antibody and the monoclonal antibodies against dually phosphorylated p38 MAPK and ERK were purchased from Sigma Chemical Co.; rabbit anti‐p38 polyclonal antibody was purchased from Cell Signaling Technology Inc. (Beverly, CA, U.S.A.). Peroxidase‐conjugated goat antimouse IgG was purchased from Jackson ImmunoResearch Laboratories Inc. (West Grove, PA, U.S.A.) and peroxidase‐conjugated goat antirabbit IgG was purchased from Sigma Chemical Co. All other reagents were of analytical grade.

Cell culture

The human keratinocyte cell line HaCaT was kindly provided by Professor N. Fusenig, German Cancer Research Center, Heidelberg, Germany. Cells were grown in MEM containing 0·075 mmol L⁻¹ Ca⁺² (MEM‐75) supplemented with 10% FCS and antibiotics. Cells were cultured in 6‐cm Petri dishes and subcultured every 4 days.

Ionizing irradiation

HaCaT cells were plated in MEM‐75 containing 10% FCS in 3·5‐cm Petri dishes (250 000 cells per dish). Twenty‐four hours later the medium was replaced with serum‐free MEM‐75 containing 0·5 mg mL⁻¹ BSA. Cultures were irradiated 24 h later using a linear accelerator 6 MV (Varian, Palo Alto, CA, U.S.A.) at a rate of 2 Gy min⁻¹. Culture medium was replaced immediately after irradiation with fresh serum‐free MEM‐75 containing 0·5 mg mL⁻¹ BSA.

Staining of cultures with Hoechst 33342 and propidium iodide

After removal of culture medium, cells were exposed to staining solution containing 0·01 mg mL⁻¹ propidium iodide and 0·005 mg mL⁻¹ Hoechst 33342 in phosphate‐buffered saline (PBS) containing 0·125 mg mL⁻¹ BSA. Cultures were photographed using a digital fluorescence microscope 15 min after the addition of the staining solution.

Caspase activity assay

Caspase‐3‐like activity in cell extracts was measured as previously described¹⁴ with slight modifications. Briefly, adherent and detached cells were harvested and washed with ice‐cold PBS, centrifuged for 15 s at 16 000 g and resuspended in 50 μL of lysis buffer (130 mmol L⁻¹ NaCl, 10 mmol L⁻¹ Tris–HCl, 10 mmol L⁻¹ phosphate buffer, 10 mmol L⁻¹ Na‐pyrophosphate and 1% Triton‐X 100, pH 7·5). After 30–60 min on ice, cell extracts were centrifuged for 2 min at 16 000 g. The supernatants were collected, frozen immediately and kept at −20 °C and enzyme activity was determined in thawed extracts. The reaction buffer contained 2 mmol L⁻¹ dithiothreitol and 10% glycerol in 20 mmol L⁻¹ HEPES buffer (pH 7·4). The fluorogenic substrate ac‐DEVD‐AMC was added just before the reaction to a final concentration of 100 μmol L⁻¹. The reaction was started by addition of the extract (10 μL containing approximately 10 μg protein) to the reaction mixture (100 μL) and monitored using a fluorescence plate reader (FLUOstar, BMG Labtechnologies, Offenburg, Germany) at excitation wavelength 390 nm and emission wavelength 460 nm. Specific caspase activity was expressed as the ratio between the reaction rate (increase in fluorescence over time) and protein content as determined by the BCA Protein Assay Kit.

Lactic dehydrogenase release

Lactic dehydrogenase (LDH) released from irradiated cultures was determined in aliquots from culture medium following centrifugation (16 000 g, 2 min). The assay was performed in 96‐well plates monitoring the decrease in NADH absorbance at 340 nm (PowerWave plate reader, BioTek Instruments Inc.). Reaction buffer contained Tris (0·1 mol L⁻¹, pH 8), sodium pyruvate (1 mmol L⁻¹) and NADH (0·2 mmol L⁻¹). Initial reaction rates (5–15 min) were used as a measure of LDH activity. LDH activity in culture medium was normalized to the total LDH activity in cell extracts obtained by sonication.

Colony formation assay

HaCaT cells were irradiated as described above and immediately thereafter cultures were rinsed, trypsinized to obtain single cell suspensions and plated in 96‐well plates (50 cells per well, 60 wells for each 3·5‐cm culture dish) in MEM‐75 containing 10% FCS. This protocol was adopted as we observed that irradiated HaCaT cells plated in Petri dishes tend to migrate and form aggregates, a phenomenon that was markedly reduced in 96‐well cultures. Seven days after seeding, wells were stained with crystal violet and the number of colonies with more than 50 cells was scored visually under a microscope. The cut‐off point for colonies above 50 cells was checked visually and found to be a diameter of 180 μm. In order to obtain the size distribution of colonies, microscope images were processed using the IMAGE‐PRO PLUS 5.1 software. For this purpose all microscopic colonies (sized from five cells upward) were counted; 250 colonies were measured for each treatment and sorted according to their diameter.

Western blot analysis

Cell extracts were prepared as follows: cultures were washed with ice‐cold PBS, lysed with sodium dodecyl sulphate (SDS) sample buffer without mercaptoethanol and bromophenol blue and boiled for 15 min; mercaptoethanol and bromophenol blue were added after determination of the protein content of the cell extracts. Samples were centrifuged before electrophoresis and subjected to SDS–polyacrylamide gel electrophoresis under reducing conditions using 10% polyacrylamide gels (20–30 μg protein per lane). Proteins were transferred to nitrocellulose membranes and probed with the appropriate antibodies. Detection was carried out by horseradish peroxidase‐conjugated secondary antibodies and enhanced chemiluminescence. The protein content of the cell extracts in sample buffer without mercaptoethanol and bromophenol blue was determined with the BCA Protein Assay Kit, which allows the quantification of proteins in the presence of detergents.

Gelatin zymography

Immediately following irradiation, culture medium in 3·5‐cm Petri dishes was replaced with 1 mL of fresh serum‐free MEM‐75 containing 0·5 mg mL⁻¹ BSA. Pro‐MMP‐9 in culture media was quantified by gelatin zymography 24 h later. Nondenaturating sample buffer X4 [0·24 mol L⁻¹ Tris–HCl (pH 6·8), 7·6% (w/v) SDS, 38% (w/v) glycerol in distilled H₂O] was added to aliquots of conditioned medium. Samples were resolved on an 8% SDS–polyacrylamide gel containing 0·1% gelatin. After electrophoresis gels were incubated in 2·5% (v/v) Triton X‐100 for 30 min to remove SDS and then washed with developer solution [200 mmol L⁻¹ NaCl, 5 mmol L⁻¹ CaCl₂, 50 mmol L⁻¹ Tris–HCl (pH 7·6), 0·02% Brij 35] for 30 min. Cells were further developed overnight at 37 °C in fresh developer solution and subsequently stained with 0·5% Coomassie Blue dissolved in 30% methanol and 10% acetic acid solution. After destaining in 30% methanol and 10% acetic acid solution proteolytic activity was visualized as colourless bands on a blue background. Intensity of bands was quantified by densitometry and normalized against protein content of the cells in cultures at the time of medium harvesting.

Total RNA isolation and mRNA determination by real‐time polymerase chain reaction

Total RNA was isolated using the EZ‐RNA total RNA isolation kit (Biological Industries) according to the manufacturer’s instructions. Total RNA (1 μg) was then reverse transcribed by EZ‐First Strand cDNA Synthesis Kit for reverse transcription‐polymerase chain reaction (RT‐PCR) (Biological Industries) using random hexamer primers according to the manufacturer’s instructions. Transcribed cDNA was then amplified using TaqMan gene expression assay (Hs00234579‐m1 for MMP‐9 and Hs999999901‐S1 for the endogenous control 18S ribosomal RNA) supplied by Applied Biosystems (Foster City, CA, U.S.A.) according to the manufacturer’s instructions by means of the Applied Biosystems Prism 7000 Sequence Detector.

Results

Two modes of radiation‐induced cell death and their inhibition by calcitriol

Autonomously proliferating HaCaT keratinocyte cultures were exposed to ionizing radiation using a linear accelerator at a rate of 2 Gy min⁻¹. Cells were treated with calcitriol 24 h before and for an additional 24–48 h after irradiation. Cultures were then stained with Hoechst 33324 and propidium iodide and observed under a fluorescence microscope. As can be seen in Figure 1c, cells displaying typical features of the late phases of apoptosis appeared 24 h after irradiation. These features included nuclear condensation and fragmentation and nuclear staining with propidium iodide.¹⁵ Such apoptotic cells were virtually absent in cultures treated with calcitriol before irradiation (cf. Fig. 1c and d). Forty‐eight hours after irradiation the pattern was markedly different (Fig. 1e). The propidium iodide‐stained nuclei with condensed and/or fragmented chromatin were no longer seen, and propidium iodide‐stained nuclei of normal size with chromatin margination appeared. The change in morphology is indicative of a shift in the mode of cell death. Treatment with calcitriol brought about a reduction in the number of propidium iodide‐stained cells 48 h after irradiation without an apparent change in their morphology (Fig. 1f).

1

Fluorescent micrographs of HaCaT cell cultures irradiated following treatment with calcitriol. HaCaT cells were plated in minimum essential medium (MEM)‐75 containing 10% fetal calf serum. Twenty‐four hours later the medium was replaced with serum‐free medium containing 0·5 mg mL⁻¹ bovine serum albumin (BSA). Cultures were then treated with vehicle (ethanol 0·06%; a, c, e) or with calcitriol (100 nmol L⁻¹; b, d, f) for 24 h, irradiated (4 Gy; c–f) and further incubated in fresh BSA‐containing MEM‐75 in the presence of calcitriol for 24 (b, d) or 48 (f) h. Cultures were stained with Hoechst 33324 and propidium iodide.

Radiation‐induced caspase‐dependent cell death and its inhibition by calcitriol

Cell death programmes, activated in response to noxious stimuli, are executed by enzymes that degrade essential intracellular components. Such executioner enzymes are the caspases, which are considered to be the hallmark of the classical apoptotic process. Caspase‐3 is the main executioner caspase in most cells, including keratinocytes. Radiation‐induced caspase activity in HaCaT cells was quantified by cleavage of the fluorogenic caspase‐3 substrate ac‐DEVD‐AMC. As can be seen in Figure 2a, the rate of DEVDase reaction in cell extracts was constant over the time period of the measurement (50 min), and the very modest DEVDase activity in control cultures increased dramatically 24 h following irradiation. Both basal and radiation‐induced DEVDase activity was markedly inhibited in calcitriol‐treated cultures. Caspase‐3‐like activity was induced in HaCaT cells by irradiation with 4 Gy, increased moderately with increasing doses of radiation and was practically abolished following treatment with calcitriol (Fig. 2b). The effect of the hormone was dose dependent and already apparent at a concentration of 1 nmol L⁻¹ (Fig. 2c). The effect of timing of exposure to calcitriol is illustrated in Figure 2d, which shows that the hormone confers its full protective effect if added either only for the 24 h preceding irradiation, or only for the 24 h after irradiation.

2

Effect of calcitriol on radiation‐induced caspase activation. HaCaT cells were irradiated and treated with calcitriol 24 h before and/or 24 h after irradiation as described in Figure 1. (a) Cells were irradiated, treated with calcitriol (100 nmol L⁻¹) before and after irradiation and harvested for caspase activity assay 24 h after irradiation. Kinetics of accumulation of the fluorescent cleavage product of ac‐DEVD‐AMC (ac‐Asp‐Glu‐Val‐Asp‐7‐amido‐4‐methylcoumarin) representing caspase‐3‐like activity was monitored for 50 min. (b) Cells were irradiated with increasing doses and treated with calcitriol as in (a). (c) Cells were irradiated (12 Gy) and treated with increasing concentrations of calcitriol. (d) Cells were irradiated (12 Gy) and treated with calcitriol (100 nmol L⁻¹) before (pre), after (post) or both (pre & post). Caspase‐3‐like activity is expressed as enzyme activity rate during the initial 50 min normalized to culture protein content. The results are presented as the mean ± SD of three independent cultures. Caspase activity in calcitriol‐treated cultures was significantly lower than in untreated cultures [P <0·0001, two‐way anova in (b); P <0·05 in all data items shown in (c) or (d), unpaired t‐test].

Radiation‐induced caspase‐independent cell death and its inhibition by calcitriol

Microscopic observation (Fig. 1) indicated that the mode of cell death shifted from apoptosis 24 h following irradiation to a different mode of PCD at later times. The apoptotic features at the first phase were accompanied by the induction of caspase‐3‐like activity. Cell death due to impairment of membrane integrity is often accompanied by release of LDH into the extracellular milieu.¹⁶Figure 3 shows that significant amounts of LDH are released from HaCaT cells 48 h after irradiation. The addition of a pan‐caspase inhibitor at a concentration previously shown to block caspase activity in HaCaT cells¹³ did not inhibit LDH release, indicating that this mode of cell death is caspase independent. The extent of caspase‐independent cell death increased with the radiation dose (Fig. 3b). At all doses, irradiation‐induced LDH release was attenuated in calcitriol‐treated cultures. The sensitivity of caspase‐independent cell death to inhibition by calcitriol is lower than that of irradiation‐induced caspase activation as evidenced by the different dose–response curves for the hormone (cf. Figs 3c and 2c). This difference may indicate that a different mechanism underlies the protective action of the hormone under the two scenarios. Figure 3d shows the results of an experiment in which calcitriol was added only for 24 h before irradiation, only for 48 h after irradiation and according to the usual protocol, both before and after irradiation. Similar to the results in Figure 2d, the protective effect of calcitriol is comparable in the three protocols.

3

Effect of calcitriol on radiation‐induced lactic dehydrogenase (LDH) release. HaCaT cells were irradiated and treated with calcitriol 24 h before and/or 48 h after irradiation as described in Figure 1. LDH released from irradiated cultures was determined in aliquots from culture medium by assaying LDH activity and normalized to the total LDH activity in cell extracts. The extent of LDH release is expressed as percentage of the overall LDH level in culture after subtraction of LDH released in control cultures. (a) Cells were irradiated (12 Gy), treated with the pan‐caspase inhibitor zD‐DCB (z‐Asp‐2,6‐dichlorobenzoyloxymethylketone; 50 μmol L⁻¹) 30 min before and 48 h after irradiation. (b) Cells were irradiated with increasing doses and treated with calcitriol (100 nmol L⁻¹) before and after irradiation. The effect of calcitriol was significant (P <0·001) as judged by two‐way anova. (c) Cells were irradiated (12 Gy) and treated with increasing concentrations of calcitriol. (d) Cells were irradiated (12 Gy) and treated with calcitriol (100 nmol L⁻¹) before (pre), after (post) or both (pre & post). The results are presented as the mean ± SD of five (a–c) or three (d) independent cultures. *P <0·05, **P <0·01 as judged by unpaired t‐test.

Radiation‐induced delayed reproductive cell death and its inhibition by calcitriol

Irradiated cells may die within a short period of time by necrosis or apoptosis and thus lead to tissue damage. Surviving cells may lose their reproductive integrity even after exposure to a moderate radiation dose, and cell death may be preceded by a time span encompassing several division cycles. This ‘reproductive cell death’ is assessed by means of the colony formation assay. Müller et al.¹⁷ recently showed that pretreatment with calcitriol increased the colony forming capacity of keratinocytes exposed to ionizing radiation. As seen in Table 1, we confirm this finding. However, we wondered if the increased number of colonies formed by irradiated hormone‐treated cells reflects a genuine protection against radiation‐induced delayed reproductive cell death or whether the effect is secondary to the protection against caspase‐dependent and ‐independent cell death induced during the first 48 h following irradiation. The size of colonies is a direct measure of the number of cell divisions occurring since plating and it is anticipated that a change in clonogenic capacity will be reflected in the size distribution of colonies. Indeed, cell irradiation resulted in a shift of colony size distribution to smaller sized colonies. Thus, if calcitriol indeed reduced ‘reproductive cell death’ we would expect that the size distribution of surviving colonies following irradiation will be shifted toward larger sized colonies in irradiated hormone‐treated cells. As can be seen from the results presented in Figure 4, pretreatment with calcitriol did not affect the size distribution of colonies derived from irradiated cells, although treatment with calcitriol induced a slight shift toward larger sized colonies in nonirradiated cells. This pattern is consistent with the notion that the increase in colony formation is fully accounted for by the increase in the number of cells that survived the first phases of caspase‐dependent and ‐independent cell death (Figs 1–3) occurring within 48 h of irradiation.

1

Effect of calcitriol on clonogenicity of irradiated HaCaT cells

	Number of colonies		Survival (%)
	Control	Irradiated
No calcitriol	813	70	8·6
With calcitriol	1098	145	13·2*

*P <0·005 as judged by χ² test.

4

Effect of irradiation and calcitriol on colony size distribution. HaCaT cells were irradiated (4 Gy) and treated with calcitriol (100 nmol L⁻¹) 24 h before irradiation as described in Figure 1. Single cell suspensions were prepared and seeded in 96‐well plates (50 cells per well). After 7 days 250 colonies for each treatment were photographed and their diameters measured (using IMAGE‐PRO PLUS 5.1 software).

The involvement of stress‐activated MAPK cascades in irradiation‐induced apoptosis and its inhibition by calcitriol

We have previously shown that the stress‐activated MAPK cascades, p38‐MAPK and JNK, are activated in keratinocytes by several environmental and pathophysiological stresses, that calcitriol attenuates this activation¹¹, ¹³ and that modulation of these cascades plays a major role in stress‐induced PCD on the one hand, and in the protective effect of calcitriol on the other hand.¹³ We proceeded to examine the role of these cascades in the protective effect of the hormone against PCD induced in keratinocytes by ionizing irradiation. Figure 5a and b provides evidence that, similarly to other stresses, p38‐MAPK is activated following exposure to ionizing irradiation and this activation is attenuated by pretreatment with calcitriol. However, inhibition of p38‐MAPK did not affect radiation‐induced keratinocyte apoptosis (Fig. 5c) supporting the notion that the inhibitory effect of calcitriol on p38‐MAPK activation does not underlie its protective effect on irradiated HaCaT cells. On the other hand, the experiments depicted in Figure 6a–c strongly suggest a mediatory role of the JNK cascade: the pronounced activation of this cascade 3 h following irradiation is practically blocked by calcitriol and mimicking this effect of the hormone by the addition of the JNK inhibitor, SP600125, markedly inhibited radiation‐induced apoptosis. However, as seen in Figure 6c, although inhibition of JNK activation can partially account for the protective effect of calcitriol, the hormone also inhibits the residual apoptosis that persists in the presence of the JNK inhibitor, suggesting that inhibition of JNK cannot fully explain the protective effect.

5

Involvement of p38‐mitogen‐activated protein kinase (MAPK) in radiation‐induced caspase activation in the presence or absence of calcitriol. HaCaT cells were irradiated (12 Gy) and treated with calcitriol (100 nmol L⁻¹) 24 h before irradiation and until harvesting as described in Figure 1. (a) Cultures were incubated for 3 h after irradiation and cell extracts were subjected to Western blot analysis using antibodies recognizing dually phosphorylated p38 (p‐p38) and total p38. (b) Densitometric quantification of the results of two experiments like the one presented in (a). (c) Cultures were treated with the p38 inhibitor SB203580 (5 μmol L⁻¹) 30 min before and immediately after irradiation. Cells were incubated for 24 h after irradiation and harvested for caspase activity thereafter. Each data point represents the mean ± SD of three independent cultures.

6

Involvement of c‐Jun N‐terminal protein kinase (JNK) in radiation‐induced caspase activation in the presence or absence of calcitriol. Cells were treated and experiments performed as described in Figure 5 except that antibodies recognizing dually phosphorylated JNK (p‐JNK) and total JNK were used for Western blotting and SP600125 (30 μmol L⁻¹) was used to inhibit JNK. (b) Densitometric quantification of the results of two experiments like the one presented in (a). (c) The results are presented as the mean ± SE of three independent experiments.

Involvement of p53 family proteins in the protective effect of calcitriol

p53, p63 and p73 are a family of transcription factors involved in cell response to stress. Members of the p53 protein family are known to play a crucial role in the response of mammalian cells to ionizing irradiation and in the regulation of PCD induced by damage to the DNA.^18–21 p53 is mutated in HaCaT keratinocytes in ‘UV hot spots’, mutations that may interfere with its known proapoptotic activities.²² p63 and p73 have a high sequence similarity to p53 particularly in the DNA‐binding domain, which allows them to transactivate p53‐responsive genes causing cell cycle arrest and apoptosis. Both proteins are expressed as full‐length transactivating (TA) forms and truncated (ΔN) forms that act as antagonists to the proapoptotic family members.²³ As can be seen in Figure 7, four isoforms of p73 were detected; two of them were identified according to their molecular weight as the full‐length isoforms TAp73β and TAp73γ. Similarly, four isoforms of p63 were detected in HaCaT cells among which two were identified as the full‐length p63 isoform TAp63α that occurs in relatively low amounts, and the truncated isoform ΔNp63α which is the major isoform in HaCaT cells.²⁴ The cellular levels of p73 isoforms were not affected by either pretreatment with calcitriol, exposure to ionizing irradiation or the combination of both. On the other hand, using the same cell extracts we showed that the cellular levels of ΔNp63α were elevated following treatment with calcitriol or irradiation and the combined effect of both treatments is at least additive. Upregulation of the antiapoptotic ΔNp63α could contribute to the protective effect of the hormone against irradiation‐induced PCD.

7

Effect of irradiation and calcitriol on p63 and p73 expression. HaCaT cells were irradiated (8 Gy) and treated with calcitriol (100 nmol L⁻¹) 24 h before and 3 h after irradiation as described in Figure 1. Cell extracts were then prepared and subjected to Western blotting using antibodies against p63 or p73. The same cell extracts were used for the analysis of p63 and p73. Data of one out of two experiments with similar results.

Radiation‐induced matrix metalloproteinase‐9 and its inhibition by calcitriol

The impairment of the epidermal–dermal barrier is one of the deleterious aftermaths of cutaneous irradiation. Excessive activity of MMPs could be one of the underlying mechanisms of this phenomenon. We screened MMP activities induced by irradiation by gelatin zymography of conditioned media of irradiated cultures. We found one detectable band of proteolytic activity migrating at a molecular weight of 92 kDa corresponding to that of pro‐MMP‐9 (Fig. 8a). In preliminary experiments (data not shown) we verified that the gelatinolytic activity of the band was due to a MMP (and not to the action of serine proteases) by incubating the gels with the MMP inhibitors EDTA and 1,10‐phenanthroline during the development stage of zymography. Figure 8a and b shows that irradiation of HaCaT cell cultures induced the secretion of pro‐MMP‐9 and that pretreatment with calcitriol markedly reduced this effect of ionizing radiation. Figure 8c establishes that irradiation increased MMP‐9 gene expression and that pretreatment with calcitriol reduced the levels of MMP‐9 mRNA in irradiated cells.

8

Effect of irradiation and calcitriol on pro‐matrix metalloproteinase (MMP)‐9 synthesis and activity. HaCaT cells were irradiated and treated with calcitriol (100 nmol L⁻¹) 24 h before and 24 h after irradiation as described in Figure 1. (a) Zymogram showing the effect of calcitriol and irradiation on pro‐MMP‐9 secretion into culture media. (b) Densitometric quantification of pro‐MMP‐9 zymograms of irradiated (12 Gy) cultures. The results are presented as the mean ± SD of three independent cultures and expressed as the percentage of gelatinolytic activity in the media of irradiated cultures. (c) mRNA levels of pro‐MMP‐9 were measured by quantitative real‐time polymerase chain reaction and normalized to the level of the endogenous control 18S ribosomal RNA. Data are presented as the mean ± SD of three independent cultures and expressed as percentage of mRNA level in untreated cultures.

Discussion

The findings of this study support the notion that hormonally active vitamin D derivatives have the capacity to attenuate the damage inflicted by ionizing radiation to epidermal keratinocytes. Treatment with calcitriol, the natural hormonal metabolite of vitamin D, reduced radiation‐induced PCD and the induction of MMP‐9 implicated in blister formation due to degradation of the basement membrane. This study was carried out using the spontaneously immortalized nontumorigenic HaCaT keratinocytes as an experimental model. HaCaT cells, when grown in low‐calcium medium retain their nondifferentiated phenotype and can proliferate autonomously in the absence of exogenous growth factors or active ingredients. This autonomous proliferation is driven by the same growth factors that support the proliferation of basal layer keratinocytes in vivo.²⁵ Autonomously proliferating HaCaT cells are therefore faithful representatives of the keratinocyte population that is most vulnerable to radiation damage, the proliferating basal layer cells. In this respect HaCaT cells are more suitable than primary keratinocytes for this study, as the latter require the presence of exogenous growth factors and other active ingredients to proliferate in culture, and are thus exposed to an extracellular milieu that is markedly different from that of basal keratinocytes in vivo.

The p53 protein is widely considered to play a pivotal role in determining the quality of the response to DNA damage inevitably incurred during exposure to ionizing radiation.¹⁸ p53 in HaCaT cells is mutated in ‘hot UV spots’,²² which results in impaired transcriptional activity. These mutations could interfere with the cellular response to ionizing radiation. However, recent studies performed in several cellular models including for instance osteosarcoma,²⁶ glioma²⁷ and colon cancer²⁸ cells challenge the notion that p53 is essential for the initiation of PCD following exposure to radiation. Patel et al.²⁹ compared the induction of apoptosis by γ‐irradiation of head‐and‐neck squamous carcinoma cells with wild‐type or transcriptionally impaired p53 mutants and concluded that p53‐ and Bcl‐2‐independent pathways predominate in regulation of apoptosis in keratinocytes suggesting that the efficacy of radiotherapy may not be solely determined by p53 status. p63 and p73 are two additional members of the p53 family. Overexpression of p73 can activate typical p53‐responsive genes and induce apoptosis like p53.¹⁹ p63 has also been shown to possess clear proapoptotic activity, mediated both by death receptors and mitochondrial pathways.²⁰, ²¹ The p53 family genes encode for multiple p63, p73 or p53 isoforms due to multiple splicing, alternative promoter and alternative initiation of translation, yielding full‐length TA isoforms, and also aminoterminally truncated (ΔN) isoforms that act as antagonists to proapoptotic p53 family members.²³ In this study we showed that under our experimental conditions HaCaT cells express the full‐length, proapoptotic p73, which may thus serve as a functional ‘back up’ to the mutated p53 in these cells.

Exposure to ionizing radiation induced two distinct modes of PCD. The first mode, appearing 24 h following irradiation, was characterized by the classical features of apoptosis such as chromatin condensation, nuclear fragmentation and executioner caspase activation. The other mode was caspase‐independent ‘necrotic‐like’ PCD and was apparent 48 h after irradiation. Treatment with calcitriol attenuated both modes of PCD although the effect on the early apoptotic death was more pronounced. Another cellular effect of irradiation is reduction in clonogenic capacity.¹ To form a colony of at least 50 cells, irradiated cells must escape the PCD that occurs in our experimental system within 2 days after irradiation and also retain their reproductive capacity for at least five cell divisions. In agreement with a previous study¹⁷ we showed here that treatment with calcitriol before irradiation increases the clonogenicity of HaCaT cells. As delayed reproductive cell death is manifested during several cell divisions after exposure to the cytotoxic assault, we would expect that cells with impaired reproductive capacity will form smaller colonies. As we found no effect of calcitriol on the size distribution of the colonies formed by irradiated cells we conclude that the protective effect of the hormone is limited to PCD and that it does not affect delayed reproductive cell death.

We previously showed that the protective effect of calcitriol on keratinocytes exposed to various environmental and pathophysiological assaults is due to attenuation of the activation of the two stress‐activated MAPK cascades, JNK and p38‐MAPK, which is usually associated with the promotion of cell death.¹³ Irradiation activated both cascades in HaCaT keratinocytes and this activation was inhibited by pretreatment with calcitriol. However, only inhibition of JNK, but not of p38‐MAPK, by specific inhibitors, reduced radiation‐induced apoptosis thus establishing a cause–effect relationship between attenuation of JNK activation by vitamin D and its protective effect against radiation‐induced PCD. As the inhibition of apoptosis by calcitriol was more pronounced than the protective effect of a pharmacological JNK inhibitor, it seems that inhibition of JNK activation cannot fully account for the action of the hormone. We show here that whereas the level of p73, the major full‐length proapoptotic p53 family member in HaCaT cells, is not affected by either radiation or treatment with calcitriol, the level of the truncated, antiapoptotic form of p63 is upregulated by both. This effect of calcitriol may suggest an additional mechanism through which the hormone exerts its protective effect on irradiated keratinocytes.

The damage to the epidermis following exposure to ionizing radiation is not due only to cell death but may also be related to excessive proteolytic activity in the irradiated epidermis. During normal tissue remodelling MMPs carry out controlled degradation of ECM constituents, which enables their rebuild and promotes cell migration and activation of growth factors.³⁰ On the other hand, excessive activity of MMPs might cause degradation of growth factors and necessary ECM constituents and interfere with cell migration towards the injured tissue. MMP‐9 degrades collagen IV and other critical components of the basement membrane zone that separates the epidermis from the dermis and such excessive activity could lead to epidermal–dermal blister formation. As seen in this study exposure of keratinocytes to ionizing radiation increased pro‐MMP‐9 secretion into the extracellular milieu and this effect is due to upregulation of gene expression. Treatment with calcitriol attenuated the effect of radiation by decreasing MMP‐9 mRNA levels.

Taken together the findings of this study suggest that hormonally active vitamin D metabolites can ameliorate the deleterious aftermath of irradiation on the epidermis, by attenuating both radiation‐induced PCD and radiation‐induced excessive proteolytic activity. These findings are in line with the recent report that calcitriol prevented hair follicle loss induced by ionizing radiation.³¹ The possible clinical implications of these findings during radiotherapy are self‐evident, particularly as preparations of noncalcaemic, nontoxic active vitamin D derivatives are used for cutaneous applications and are thus readily available. We also show in this study that the timing of exposure to calcitriol is not critical and that the hormone can exert its protective effect whether present before or immediately after irradiation, a feature convenient for possible clinical use. A concern might be raised that if calcitriol and its analogues protect epidermal cells from radiation‐induced death they might also protect cancer cells at the same time and thus reduce the efficacy of radiotherapy. A few studies performed with breast cancer³² and prostate cancer³³ cells clearly demonstrate that this is not the case. In fact, treatment with active vitamin D derivatives sensitized these cells to radiation‐induced death. The opposite effects of calcitriol on PCD in keratinocytes and malignant cells are also observed in cells exposed to reactive oxygen species which are known to partake in ionizing radiation‐induced damage.³⁴ While the hormone sensitizes breast³⁵ and colon cancer³⁶ cells to PCD induced by exposure to hydrogen peroxide, it protects keratinocytes¹³ from the same insult. These findings are in favour of the notion that calcitriol and its analogues may add to the meagre arsenal of available measures to protect the vulnerable epidermis during radiotherapy The in vitro findings of this study are now being followed up in the Institute of Oncology at the Rabin Medical Center, by a small‐scale clinical trial examining the effect of topical treatment with the vitamin D analogue, calcipotriol, on the skin of breast cancer patients undergoing radiotherapy.

Acknowledgment

This work was supported by The Israel Science Foundation grant 703/04 and Israel Cancer Association grant B‐20052014.

Author notes

Conflicts of interest
None declared.

M.L. and C.R. contributed equally to this work.
R.K. contributed equally as senior author.