Ultraviolet B (UVB) is a major cause of skin inflammation, leading to skin damage. Our previous in vivo study revealed that a natural flavonoid silibinin had marked anti-inflammatory effect on UVB-exposed murine skin. UVB exposure caused reduced autophagy in epidermis while it promoted autophagy in dermis. Nevertheless, silibinin inhibited the inflammatory flux in the skin epidermis as well as dermis through the modulation of autophagy. In order to elucidate the underlying protective mechanisms of silibinin for UVB damage on skin, separate studies on epidermis and dermis are helpful. Derived from the normal tissue of the mouse, L929 cells are capable of representing some characteristics of dermal cells. UVB irradiation caused L929 cell apoptosis in a time- and dose-dependent manner. Ataxia-telangiectasia-mutated (ATM) protein and p53 were activated to cause cell apoptosis, accompanying upregulation of the autophagic flux. The pharmacological inhibition of ATM, p53 and autophagy or the transfection with autophagy-associated protein-targeted small interfering RNAs showed that the UVB-activated ATM-p53 axis and autophagy formed a positive feedback loop, which synergistically promoted cell apoptosis. Silibinin treatment simultaneously repressed the activation of ATM-p53 and autophagy and thereby protected UVB-irradiated L929 cells from apoptotic death.
terminal deoxynucleotidyl transferase dUTP nick end labelling
Most skin diseases, such as cancer, photo-aging, sunburn and pigmentation, are closely linked to solar irradiation, typically ultraviolet B (UVB, wavelength 280–315 nm) . In addition to its effect on the epidermal layer, UVB can also reach the dermis and influence physiological functions of fibroblasts [2, 3]. Previous studies on UVB-irradiated dermal fibroblasts were mainly focused on inflammation [4, 5] and photo-aging [2, 6], with little attention to apoptosis. An in vivo study indicated that murine epidermis and dermis responded differently towards UVB irradiation in terms of apoptosis, autophagy and inflammation  tempting us to examine the responses separately. Studies on human epidermoid carcinoma A431 cells [7, 8] showed that autophagy induced by silibinin was protective in UVB-irradiated A431 cells consistently with the in vivo findings on epidermis . Here, the effect of UVB irradiation on L929 cells is examined from the viewpoint that the cells represent some characteristics of dermal fibroblasts.
The ataxia-telangiectasia-mutated (ATM) protein kinase, which senses DNA aberration in eukaryotic cells, is a member of the phosphatidylinositol 3-kinase-related serine/threonine protein kinases (PIKKs) family . DNA damage can induce the rapid auto-phosphorylation of ATM and thereby activate the kinase . p53, which mediates multiple responses including cell cycle arrest , senescence  and apoptosis  towards cellular stress, is the typical substrate of ATM. p53 is rapidly ubiquitinated by the binding of MDM2 , but can be stabilized through the phosphorylation mediated by ATM .
Autophagy is a cellular digestive process which might either recycle the damaged organelles and macromolecules for energy and cell constituents  or induce programmed cell death through cell lysis . It occurs under various cellular stresses, such as starvation, hypoxia and DNA damage, acting synergistically with  or contrarily to apoptosis . The role of autophagy under UVB stress has not been well established.
Silibinin (Fig. 1A), a flavonoid, is suggested as an anti-UVB agent . Here, we attempt to elucidate the mechanisms of silibinin effects involved. UVB irradiation activated the ATM-p53 axis, which caused apoptosis in murine fibroblast L929 cells. The autophagy induced by UVB formed a positive feedback loop with ATM-p53 and promoted apoptosis. Silibinin inhibited the activation of both ATM-p53 and autophagy, leading to the rescue of L929 cells from UVB-induced apoptotic death.
Treatment with silibinin rescues L929 cells from UVB-induced apoptosis
UVB irradiation has been reported to cause apoptosis in several cell lines [7, 12]. In this research, the dose- and time-dependent cytotoxic effect of UVB was observed in L929 cells (Fig. 1B). Silibinin treatment dose-dependently improved the viability of UVB-irradiated cells (Fig. 1C). However, due to the obvious growth inhibitory effect of silibinin at concentrations higher than 200 μm (Fig. 1C), silibinin at a concentration of 150 μm was selected for further study on UVB-irradiated L929 cells. The flow cytometric analysis of cellular DNA content showed that silibinin inhibited time-dependent cell death (Fig. 1D,E).
The features of cellular apoptosis such as cell shrinkage and formation of apoptotic bodies were observed under the microscope after UVB irradiation (Fig. 2A). These changes were rescued by silibinin treatment (Fig. 2A). Terminal deoxynucleotidyl transferase dUTP nick end labelling (TUNEL) is a common method for detecting DNA fragmentation that results from apoptotic signalling cascades. UVB irradiation markedly elevated the ratio of TUNEL-positive cells (Fig. 2B). The ratio was also reversed by silibinin treatment (Fig. 2B). Analysis of protein levels revealed that UVB exposure resulted in the activation of caspase-3 (Fig. 2C). However, the activation was attenuated with silibinin (Fig. 2C). Cleavage of caspase-3 substrates, poly-ADP-ribose polymerase (PARP) and inhibitor of caspase-activated DNase (ICAD) was reversed with silibinin as well (Fig. 2D). Taken together, it was concluded that treatment with silibinin protected L929 cells from UVB-initiated apoptosis.
Silibinin protects UVB-irradiated L929 cells by repressing the ATM-p53 pathway
UVB irradiation was reported to cause DNA damage and then to initiate cell arrest, DNA repair or apoptosis through activation of ATM [9, 14, 18]. The expression of ATM, p-ATM and its downstream effector p53 was upregulated by UVB irradiation, and these changes were attenuated by silibinin treatment (Fig. 3A). KU55933, an ATP-competitive inhibitor , reduced the phosphorylation of ATM, as well as the expression of p53 and p-p53 (Fig. 3B). Pifithrin-α inhibits p53-dependent transactivation of p53-responsive genes in many cell lines owing to the decrease of p53 stability . In this study, pifithrin-α reduced the expression of p53 and p-p53 (Fig. 3B). Treatment with the ATM inhibitor KU55933 or p53 inhibitor pifithrin-α improved the viability of UVB-irradiated L929 cells (Fig. 3C). The ratio of hypo-diploid cells (Fig. 3D) was also reduced by treatment with KU55933 or pifithrin-α. p53 has been reported to transactivate many pro-apoptotic proteins such as Bax, PUMA and Noxa . In this study, the expression of Bax, PUMA and Noxa was markedly upregulated by UVB irradiation, while it was downregulated by silibinin treatment (Fig. 3E). The transcription of Bax, PUMA and Noxa was stimulated by UVB irradiation, as reflected by reverse transcription PCR of the targeted mRNAs (Fig. 3F), in accordance with the activation of p53 (Fig. 3A). But silibinin and pifithrin-α reduced the transactivation of these targets (Fig. 3F), indicating that silibinin inhibited the transactivation of pro-apoptotic Bax, PUMA and Noxa by suppressing the activity of p53. Consistent with these findings, inhibition of ATM and p53, with KU55933 and pifithrin-α respectively, both reduced the protein expression of Bax, PUMA and Noxa (Fig. 3G) as well as the activation of caspase-3 and the cleavage of PARP and ICAD (Fig. 3H). These findings suggested that UVB caused apoptosis in L929 cells through the ATM-p53 axis, and this process was blocked by treatment with silibinin.
Autophagy is induced by UVB irradiation and reversed by silibinin treatment in L929 cells
Autophagy is always involved in the responses of cells towards various stresses, such as nutrient deprivation, oxidative stress, hypoxia, DNA damage and intracellular pathogens [10, 17, 22]. UVB exposure activates the autophagic process, as shown by the marked increase of GFP-LC3 puncta in L929 cells (Fig. 4A), the elevated monodansylcadaverine (MDC) positive cell ratio (Fig. 4B) and time-dependently accelerated conversion of LC3 (Fig. 4C). However, this induction of autophagy was attenuated by treatment with silibinin (Fig. 4A–C). Pre-culture with chloroquine (CQ), a lysosomal degradation inhibitor , increased the amount of LC3-II, indicating that silibinin inhibited the autophagic flux induction caused by UVB irradiation but with no effect on LC3-II degradation in autolysosomes (Fig. 4D).
Blocking autophagy represses UVB-induced apoptosis through inhibiting the ATM-p53 axis in L929 cells
3-Methyladenine (3-MA), which targets PI3K, is used to repress the induction of autophagic flux . Blocking autophagy with 3-MA (Fig. 5A) improved the viability of UVB-irradiated L929 cells (Fig. 5B). The percentage of hypo-diploid cells (Fig. 5C) and the activation of caspase-3 (Fig. 5D) were both reduced by 3-MA treatment. The expression of p-ATM and p53 was also reduced with 3-MA (Fig. 5D), suggesting that blocking autophagy induction rescued the cells from UVB-induced apoptosis through inhibition of the ATM-p53 axis. Treating UVB-irradiated cells with CQ  also reduced the apoptosis (Fig. 5E,F) and the expression of p-ATM and p53 (Fig. 5F), indicating that blocking the lysosomal degradation in autophagy attenuated ATM-p53 pathway mediated apoptosis in UVB-irradiated L929 cells. To confirm the effects of autophagy, the autophagy-associated proteins Beclin 1 and LC3 were specifically silenced with the corresponding small interfering RNAs (siRNAs). After transfection with siRNA targeting Beclin 1 (si-Beclin 1) the formation of LC3-II was suppressed and the transfection of siRNA targeting LC3 (si-LC3) reduced the expression of LC3 (Fig. 5H). Silencing of Beclin 1 or LC3 improved the cell viability (Fig. 5G) and inhibited the expression of p-ATM, p53 and caspase-3 (Fig. 5H).
Suppression of the ATM-p53 axis reduces autophagy in UVB-treated L929 cells
Some reports have shown that autophagy is regulated by ATM and p53 [24, 25]. In this study, suppressing the activity of ATM and p53 respectively with KU55933 and pifithrin-α both reduced the autophagic level, as reflected by the decrease of the MDC positive cell ratio (Fig. 6A), the reduction in numbers of GFP-LC3 puncta (Fig. 6B) and the blocking of LC3 conversion in UVB-irradiated L929 cells (Fig. 6C). These results suggested that there was a positive feedback interaction between the ATM-p53 axis and autophagy in UVB-irradiated L929 cells. Their activation synergistically promoted the induction of apoptosis.
This study elucidated that apoptosis of L929 cells induced by UVB was caused and strengthened by a positive feedback loop of ATM-p53 pathway activation and autophagy induction. Silibinin simultaneously inhibited the activation of ATM-p53 and autophagy in UVB-irradiated L929 cells and thereby attenuated the cell apoptosis. These findings provided new insight into the process of UVB injury in the cells. The protective potential of silibinin on UVB-irradiated L929 cells can be accounted for on this basis. The relationship between apoptosis and autophagy illustrates the crosstalk involving the activation of ATM and p53 proteins between these typical programmed cell death (PCD) pathways.
In our study, we found that silibinin markedly protected cells from UVB-induced apoptosis, as indicated by the reduction in the TUNEL-positive cell ratio and caspase-3 activation. However, the improvement in cell viability was only 20%–40%. This result suggested that there might be some other growth inhibitory mechanisms (but not death) involved in UVB-irradiated L929 cells besides apoptosis. Actually, in our study, we observed the growth arrest of UVB irradiation, which was not significantly improved by silibinin treatment. For example, according to the fluorescence activated cell sorting (FACS) analysis of cell cycle in Fig. 1D, it seemed that there might be S-phase arrest in UVB-irradiated L929 cells, which was not reversed by silibinin treatment.
Activation of ATM and p53 under DNA damage is essential for the initiation of the cellular self-repair process [9, 10, 14], but sustained hyper-activation directs the cells into apoptosis [12, 25]. Inhibiting ATM by KU55933 in a low amount (2 μm) in this study improved cell viability but a higher concentration (20 μm) of KU55933 caused obvious cytotoxic effects (unpublished data) in UVB-irradiated L929 cells, suggesting that it would be beneficial to maintain the ATM-p53 activation in a certain range where the cells initiate a self-repair process without activating the caspase pathway. UVB-stimulated hyper-activation of ATM caused cell apoptosis, while silibinin modestly repressed the ATM-p53, thereby improving cell viability.
Some reports have shown that ATM upregulated the level of autophagy. DNA-damage-activated ATM phosphorylated the ΔNp63α transcription factor to initiate the expression of the autophagy-related proteins ATG1/ULK1, ATG3, ATG4A, ATG5, ATG6/BECN1, ATG7, ATG9A and ATG10 . Glucose-deprivation-induced ATM/AMPK/p53 activation led to the induction of autophagy in human fibroblasts . However, reports concerning the direct effect of autophagy on the activation of ATM are limited. One research paper showed that genetic or pharmacological disruption of autophagy attenuated capsaicin-induced ATM activation . In addition to the accelerating effect of UVB-activated ATM on autophagic flux, this study also demonstrated that ATM phosphorylation was stimulated by autophagy. The positive feedback loop of the ATM-p53 pathway and autophagy synergistically promoted apoptosis in UVB-irradiated L929 cells. Therefore, in L929 cells the ATM pathway and autophagy should be blocked at the same time to achieve the best prevention against UVB damage. Silibinin could be such an ideal agent.
The tumour suppressor protein p53 (also known as TP53) influences a large range of cellular processes in the determination of cellular fate. p53 targets DRAM (damage-regulated autophagy modulator) to induce cell autophagy . p53 can also inhibit the anti-autophagic protein Bcl-2 to initiate autophagy [27, 28]. These findings suggest that, besides the effect of p53 on apoptosis and senescence, p53 also behaves as an autophagy-inducer. Results in this study, together with the related published information [25, 27, 28], enriched the recognition of this protein on the regulation of autophagy.
Recently, the interaction between apoptosis and autophagy is attracting more and more attention. Autophagy, as a cellular digestive process, is required for complete cell degradation , and in the early stage of studies on autophagy researchers observed enhanced autophagy in cells undergoing death . Therefore autophagy is defined as the second type of PCD [16, 28, 29]. With further studies, more and more research has shown that, as a cellular recycling pathway, autophagy exerts an antagonizing effect towards apoptosis  and the role of autophagy in the determination of cell fate has become complicated . Autophagy was reported to regulate the stability of p53 through inhibiting the de-ubiquitination by Beclin 1 . It might augment the apoptosis. Autophagy is essential for the clearance of apoptotic cellular components as well . A previous study on UVB-induced skin inflammation in vivo showed that autophagy was reduced in epidermal cells, while it was promoted in dermal cells after UVB irradiation . The upregulation of autophagy in dermis retarded the apoptosis of inflammatory cells and fuelled the inflammation induced by UVB irradiation . However, the dermis contains different clusters of cells, such as inflammatory cells and resident fibroblasts. This study focused on fibroblast L929 cells in vitro. UVB-induced autophagy in L929 cells augmented the apoptosis. Based on all these findings, although defined as a type of PCD, autophagy might be better considered as a process occurring along with apoptosis, playing a promoting or inhibitory role on cell apoptosis depending on the context [3, 10, 29-31].
Dual effects of autophagic level were noted in UVB-irradiated L929 cells. UVB irradiation induced hyper-activation of autophagy, attenuation of which contributed to cell survival. However, on the other hand, excessive block of autophagy accelerated cell death. A higher concentration of CQ (20 μm) reduced the viability of UVB-irradiated L929 cells (unpublished data). Silencing Beclin 1 and LC3 at a higher degree with 50 nm siRNA duplex exhibited a slight repression of cell viability in UVB-irradiated L929 cells (unpublished data). This means that autophagy in irradiated cells should be kept in a moderate range for cell survival. Hyper-activation of autophagy might stimulate the pro-apoptotic signals, while hypo-activation might accumulate cellular damage as well leading to cell death.
Silibinin is a multi-functional reagent in modulating cell responses towards UV irradiation, which might involve the reduction of DNA damage , the modulation of inflammation  and the regulation of pro-apoptotic and autophagic signalling [7, 8]. Therefore, it can be a useful drug for the improvement of cell bio-functions [3, 8]. Considering the dual effects of silibinin on ATM-p53 activation [10, 12, 25] and autophagy induction [3, 8], it was suggested that silibinin might have a balancing potential towards stress-induced alterations in mammalian cells. Therefore, it exhibits efficacy in multiple diseases including hepatitis , cancer [35, 36] and inflammation [3, 37], possibly by adjusting the adaptation of cells to stresses. From the cell protective effect found in this study, silibinin might be an ideal agent to prevent UVB damage of mammalian skin.
Materials and methods
Cells and culture
Murine fibroblast L929 cells (derived from normal subcutaneous areolar and adipose tissue) were purchased from American Type Culture Collection (ATCC) (Manassas, VA, USA) and cultured in RPMI-1640 medium (Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (Beijing Yuanheng Shenyang Research Institution of Biotechnology, Beijing, China), 100 μg·mL−1 streptomycin and 100 U·mL−1 penicillin. Cells were incubated at 37 °C with 5% CO2 in a humidified atmosphere. All experiments were performed on logarithmically growing cells.
Silibinin with a purity of 99% was obtained from Beijing Institute of Biological Products (Beijing, China). The reagent was dissolved in dimethylsulfoxide to make a stock solution. It was diluted with RPMI-1640 complete medium and the dimethylsulfoxide concentration was kept below 0.1% in cell culture, which had no detectable effects on cells.
Methylthiazolyldiphenyl-tetrazoliumbromide (MTT), propidium iodide (PI), RNase A, MDC, CQ, 3-MA, KU55933, pifithrin-α and primary antibody against LC3 were purchased from Sigma Chemical (St Louis, MO, USA). Primary antibodies against caspase-3, ATM, phosphorylated ATM, p53, p-p53, PARP, ICAD and β-actin, as well as horseradish-peroxidase-conjugated secondary antibodies, were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). The SuperSignal® West Pico Chemiluminescent Substrate for horseradish peroxidase enzyme was obtained from Thermo Scientific (Rockford, IL, USA).
UVB lamps (220 V, 40 W) (Beijing Lighting Research Institute, Beijing, China), which were equipped with a UVB spectra radiometer (Photoelectric Instrument Factory of Beijing Normal University, Beijing, China), emitted UVB radiation ranging from 280 to 340 nm (~ 80% of total energy), with a peak at 314 nm. L929 cells were irradiated with UVB radiation at an intensity of 30 μW·cm−2, and the total irradiative dose was controlled by timing: cells were irradiated for 5, 10 and 15 min to get a radiation of 90, 180 and 270 J·m−2, respectively.
To avoid possible UVB absorption by proteins or chemical compounds, the culture medium was replaced with a thin layer of fresh NaCl/Pi prior to the irradiation. Silibinin treatment was performed 1 h before UVB exposure. If protein inhibitors were administered, they were pre-cultured for 1 h. After UVB irradiation, the medium that had been removed before irradiation was added back to the cell culture plates. The cells were harvested or analysed at indicated time points after UVB irradiation (the time point of UVB irradiation was taken as the baseline).
L929 cells were seeded into 96-well cell culture clusters (Corning, NY, USA) at a density of 6 × 103 cells·well−1 and cultured for 24 h. Then the cells were subjected to different treatments for another 24 h or the indicated time course. Thereafter, the cells were rinsed twice with ice-cold NaCl/Pi and incubated with 100 μL of 0.5 mg·mL−1 MTT solution at 37 °C for 3 h. After removing the supernatant the residual cell layer was dissolved in 150 μL dimethylsulfoxide, and the optical density (A value) was measured at 490 nm wavelength using a microplate reader (Thermo Scientific Multiskan MK3, Shanghai, China). Cell viability was calculated using the equation.
Flow cytometric analysis of hypo-diploid cells
PI is a fluorescent dye that is commonly used to quantify DNA content. Cells subjected to the indicated treatments were collected (5 × 105 cells per sample), fixed with 70% (v/v) ethanol at 4 °C overnight, rinsed with ice-cold NaCl/Pi twice and incubated with 1 mL of PI solution (PI 50 mg·L−1 and RNase A 1 g·L−1) at 4 °C in the dark for 30 min. The cellular DNA content was next analysed using a FACScan flow cytometer. Cells with DNA content less than normal diploid (at subG0/G1 phase) are dying cells.
TUNEL staining of apoptotic cells
After 24 h treatment of UVB and/or silibinin, the cells were collected, planted into the wells of a blue slide and air dried. Then the cell smears were treated with the TUNEL kit (in situ cell death detection kit, POD, Roche Applied Science, Mannheim, Germany) according to the manufacturer's illustration. Briefly as follows: cell smears after fixation, blocking and permeabilization were incubated with TUNEL reaction mixture for 1 h in the dark at 37 °C and then visualized using Converter-POD and DAB substrate.
After the indicated treatments, both adherent and floating cells were collected at 24 h or the predetermined time points and lysed with corresponding lysis. For the whole cell lysate, RIPA lysis buffer (Beyotime, Haimen, Jiangsu, China) supplemented with phenylmethylsulfonyl fluoride (1 mm) was used. The protein concentration was determined using the Bio-Rad protein assay reagent (Bio-Rad, Hercules, CA, USA). After denaturation, lysates with equal protein were separated by 10%–13% SDS/PAGE and transferred onto Millipore Immobilon®-P Transfer Membrane (Millipore Corporation, Billerica, MA, USA). After incubation with primary antibodies and corresponding horseradish-peroxidase-conjugated secondary antibodies, the blots were visualized using the SuperSignal® West Pico Chemiluminescent Substrate.
Reverse transcription PCR
Total RNA was isolated from L929 cells after different treatments using Total RNA Extraction Kit (Tiangen Biotech, Bejing, China) according to the manufacturer's instructions. Then the target mRNAs were reverse-transcripted into the first strand cDNA using RevertAid First Strand cDNA Synthesis Kit (Thermo Scientific, Waltham, MA, USA). The cDNAs were amplified with PCR MasterMix (Tiangen Biotech). Primers used in this study were as follows (forward and reverse, respectively): Bax, 5′-CTAGCAAACTGGTGCTCAAGG-3′ and 5′-CGAAGTAGGAGAGGAGGCCT-3′; PUMA, 5′-CTGTATCCTGCAGCCTTTGC-3′ and 5′-ACGGGCGACTCTAAGTGCT-3′; Noxa, 5′-ACTGTGGTTCTGGCGCAGAT-3′ and 5′-TGAGCACACTCGTCCTTCAAGT-3′; GAPDH, 5′-TCCCACTCTTCCACCTTC-3′ and 5′-CTGTAGCCGTATTCATTGTC-3′. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as the loading control to determine the relative changes in the target samples.
Quantification of the GFP-LC3 puncta in cells
GFP-LC3 puncta in cells represent the formation of autophagosomes and can be used to assess the autophagic flux . Cells cultured in 24-well cell culture clusters were transfected with 2 μg of GFP-LC3 plasmid (kindly provided by Y. Chen, Peking University Centre for Human Disease Genomics) using 1 μL of TranSmarter (Abmart, Shanghai, China) per well according to instructions of the manufacturer. The dots of GFP-LC3 were observed under a fluorescence microscope. In each treatment group, 30 cells were randomly selected for the quantification of GFP-LC3 puncta per cell.
Fluorescent microscopy of MDC staining
Fluorescent compound MDC stains lysosomes and it is nowadays widely used in studies of autophagy together with the measurement of other parameters. In this study, cells in 24-well cell culture clusters (Corning) were stained with 0.05 mm MDC solution in the dark at 37 °C for 30 min, and then observed under a fluorescence microscope (Olympus, Tokyo, Japan).
Flow cytometric analysis of autophagic cells
Cells were collected (5 × 105 cells per sample), incubated with 0.05 mm MDC solution in the dark at 37 °C for 30 min and then analysed using a FACScan flow cytometer (Becton Dickinson, Franklin Lakes, NJ, USA).
Transfection of siRNA
Murine Beclin 1-targeted (si-Beclin 1 or si-Beclin), LC3-targeted (si-LC3) and negative control (si-con) siRNAs were purchased from GenePharma (Suzhou, China). Cells were transfected with 5 nm si-Beclin 1, si-LC3 or si-con using siRNA-Mate™ (GenePharma) according to the manufacturer's protocols. The transfected cells were used for subsequent experiments 24 h later.
Comparisons between groups were determined using Student's t test. All P values were one-tailed and were considered significant for P <0.05.