High mobility group box1 (HMGB1) in relation to cutaneous inflammation in systemic lupus erythematosus (SLE)

Transcription

1 (2013) 0, PAPER High mobility group box1 (HMGB1) in relation to cutaneous inflammation in systemic lupus erythematosus (SLE) DA Abdulahad 1, J Westra 1, E Reefman 2, E Zuidersma 1, J Bijzet 3, PC Limburg 3, CGM Kallenberg 1 and M Bijl 4 1 Department of Rheumatology and Clinical Immunology, University Medical Center, University of Groningen, The Netherlands 2 Department of Metabolic Health Research, TNO, The Netherlands; 3 Pathology and Laboratory Medicine, University Medical Center Groningen, University of Groningen, The Netherlands; and 4 Department of Internal Medicine and Rheumatology, Martini Hospital, The Netherlands Summary: Photosensitivity is characteristic of systemic lupus erythematosus (SLE). Upon ultraviolet B (UVB) exposure, patients develop inflammatory skin lesions in the vicinity of sunburn cells (SBCs). High mobility group box 1 (HMGB1) is released from apoptotic and activated cells and exerts inflammatory actions through ligation to its receptors. Methods: Eleven SLE patients and 10 healthy controls (HCs) were exposed to UVB. Skin biopsies were taken before and at one, three and 10 days after irradiation. Sections were stained for SBC, HMGB1, CD3, CD68, interferon-induced protein MxA and cleaved caspase 3. In vitro experiments with UVB-irradiated keratinocytes were also performed. Higher numbers of cells that had released HMGB1 were seen in the skin of SLE patients compared to HCs before and after irradiation. HMGB1-negative nuclei correlated with the presence of SBCs, and with the number of cleaved caspase 3 positive cells in lupus skin. Results: HMGB1 release is increased in the skin of SLE patients compared to HCs. Upon UVB exposure, HMGB1 release further increases in SLE patients and is related to the number of apoptotic cells. Our data suggest that HMGB1, probably released from apoptotic keratinocytes, contributes to the development of inflammatory lesions in the skin of SLE patients upon UVB exposure. (2013) 0, Key words: HMGB1; SLE; skin Introduction Systemic lupus erythematosus (SLE) is a systemic autoimmune disorder with various clinical manifestations of which cutaneous involvement is most common. The pathophysiological mechanisms leading to cutaneous lesions are not fully elucidated, but ultraviolet B (UVB) irradiation has been associated with the onset of lesions. 1 3 UVB irradiation initiates inflammatory reactions in the skin via several mechanisms. It can induce production and release of chemokines and cytokines such as interleukin (IL)-8, IL-6, and tumor necrosis factor (TNF) by keratinocytes. 1,2 It induces apoptosis in keratinocytes, which turn into so-called sunburn cells (SBCs). 3,4 Correspondence to: Johanna Westra, Department of Rheumatology and Clinical Immunology, AA21, University Medical Center Groningen, University of Groningen, PO Box , 9700 RB Groningen, The Netherlands. johanna.westra@umcg.nl Received 29 November 2012; accepted 25 February 2013 Interestingly, inflammatory lesions in the skin of SLE patients after UVB exposure develop in the vicinity of these SBCs. 4 Moreover, it has been reported that apoptotic cells accumulate in tissues and in the circulation of SLE patients. 5 These cells present autoantigens on their cell surface but also release high mobility group box 1 (HMGB1). 6,7 HMGB1 is a nuclear DNA-binding protein that resides inside cells. 8 It can be released from activated cells but release occurs also during the late phase of apoptosis as well as during necrosis. 6,9,10 Extracellular HMGB1 acts as an alarmin and exerts its activity through binding to receptors such as the receptor for advanced glycation end products (RAGE) and Toll-like receptors (TLR)-2, -4 and An important role for HMGB1 in the pathogenesis of SLE has been described by Voll et al. 15 They demonstrated that this protein is tightly attached to chromatin released from late apoptotic cells. These complexes are able to induce inflammatory and immune responses and as such form an important factor in the pathogenesis of SLE. 16 Also the fact that HMGB1 undergoes reversible! The Author(s), Reprints and permissions: /

2 2 oxidative modifications at cysteine residues during cell death contributes to its biological properties. 17 Recent studies have shown an association between HMGB1 and skin inflammation in SLE patients with photosensitivity. 18,19 However, the role of UVB in the development of cutaneous lesions has not been fully clarified. It has been shown that patients without a history of photosensitivity can develop UV-induced cutaneous lesions as well. 4,20 In accordance, in a previous study we demonstrated that in only one-third of SLE patients, exposure to UVB resulted in inflammation of the skin after three days, irrespective of photosensitivity as defined by patients history and photoprovocation. 4,21 These signs of inflammation are called infiltrates, and are not comparable to the skin lesions formed by induction through standardized photoprovocation as described by Kuhn et al. 22 This study was conducted to further substantiate the involvement of HMGB1 in cutaneous lesion development upon UVB exposure in SLE patients. We determined the percentage of cells showing HMGB1-negative nuclei as a result of its release in the epidermis, along with its mrna expression following UVB irradiation, both in patients with and without UVB-induced inflammation, and in healthy controls (HCs). We show that the number of cells with HMGB1-negative nuclei is significantly increased in the skin of SLE patients even before UVB irradiation. After UVB exposure, the presence of extracellular HMGB1 in SLE patients, particularly in patients who developed infiltrates, further increased. Extracellular HMGB1 coincided with the development of apoptotic keratinocytes. Material and methods Patients In the current study skin biopsies were used from 11 SLE patients and 10 HCs who participated in studies on UVB irradiation as described previously. 4,21,23 All patients fulfilled at least four of the criteria of the American College of Rheumatology for SLE and had inactive disease as defined by the SLE Disease Activity Index (SLEDAI) 4 at the time of skin biopsy in combination with the absence of active cutaneous disease. None of the patients was using topical or high doses of oral corticosteroids. Patients and HCs whose buttock skin had been exposed to sunlight or other sources of UVB in the past six months preceding the biopsy were excluded. Table 1 shows patient characteristics. There was no difference in use of prednisone or hydroxychloroquine between patients who did or did not develop infiltrates. In the patients who developed inflammatory infiltrates, none (0 out of five) was on azathioprine, which differed (not significantly) from use of azathioprine in patients who did not develop infiltrates (four out of six, p ¼ 0.06 Fisher s exact test). As HCs, three males and seven females were included with a mean age of years. The local ethics committee of the University Medical Center Groningen (UMCG) (The Netherlands) had approved the study and all study subjects had given written informed consent. Table 1 Characteristics of SLE patients included in the study. ACR criteria Sex Age (years) Disease duration (years) Infiltrates after UVB irradiation Prednisone (mg/day) Hydroxychloroquine (mg/day) Azathioprine (mg/day) F Yes þ þ þ þ M 46 7 Yes þ þ þ þ þ 5 F Yes þ þ þ 200 F 41 6 Yes þ þ þ þ þ 10 F Yes þ þ þ þ 400 M No þ þ þ þ F 47 4 No þ þ þ þ F No þ þ þ F 31 8 No þ þ þ þ þ þ F No þ þ þ þ þ þ þ 5 75 F 51 8 No þ þ þ þ þ þ 400 SLE: systemic lupus erythematosus; ACR: American College of Rheumatology; F: female, M: male, UVB: ultraviolet B; ACR criteria, numbered according to Hochberg et al. 33 ; þ and indicate cumulative presence or absence of a particular criterion: 1) malar (or butterfly) rash, 2) discoid rash, 3) sensitivity to light, or photosensitivity, 4) oral ulcers, 5) arthritis, 6) serositis, 7) kidney disorder, 8) neurological disorder, 9) blood abnormalities, 10) immunologic disorder, including positive anti-double-stranded DNA, anti-sm, anti-phospholipid antibodies, or 11) positive antinuclear antibodies.

3 Photoprovocation (in vivo irradiation protocol) UVB irradiation was performed as previously described. 4,21 The minimal erythemal dose (MED) was determined in each patient and HCs after 24 hours and subjects were irradiated with two MEDs on four small areas on the buttock. After one, three and 10 days, skin biopsies were taken from the area of irradiated skin. As a control, a biopsy was taken after one day from nonirradiated skin at a distance of at least 2 cm from the site of irradiation to avoid influence of wound healing on the reaction to UVB (non-irradiated, that is day 0 in further reading). Biopsies were split: One-half was fixed in formaldehyde and the other half was snap frozen. The formaldehyde samples were used in immunohistochemical staining for determining numbers of SBCs and HMGB1-, RAGE-, CD68-, CD3-, cleaved caspase 3 and MxA expression. Frozen biopsies were used for determination of mrna levels. Staining of skin biopsies and morphometry Next, sections were deparaffinized and antigen retrieval and endogenous peroxidase blocking was performed. Slides were incubated with monoclonal antibodies against CD68 (clone PG-M1) and CD3 (clone F7.2.38) (DakoCytomation, Denmark), rabbit anti-hmgb1 antibody (Abcam, UK), goat anti-rage antibody (AbD Serotec, Germany), rabbit anti-cleaved caspase 3 antibody (Cell Signaling Technology, The Netherlands) and mouse antibody clone 2C12 directed against MxA (gift of Prof. Otto Haller, Germany). Subsequently, slides were incubated with horseradish peroxidase (HRP)-labeled secondary antibodies (DakoCytomation). Next, slides were incubated in diaminobenzidine solution and counterstained with hematoxylin and eosin (H&E). Numbers of SBC per mm 2 were determined by counting their numbers in three sequential stained sections and dividing their numbers by the epidermal surface area as described previously. 4 H&E stained sections were scored semi-quantitatively for the presence of perivascular infiltrating cells in the dermis using a score from 0 to 5: no infiltrating cells (0), one to two infiltrating cells (one), one perivascular layer of infiltrating cells (two), two to three layers of infiltrating cells (three), >three layers of infiltrating cells (four), >3 layers of infiltrating cells outside perivascular region (five). The final score was determined by averaging the mean vessel score of three consecutive sections. Morphometry was performed on CD3-, CD68-, cleaved caspase 3- and MxAstained sections. 4 Evaluation of HMGB1 staining The stained slides were coded and analyzed independently by two observers (DA and JW) in a blinded fashion. Cellular distribution of HMGB1 was determined in the epidermal layer by counting 100 cells in three brightfield pictures for HMGB1-positive (brown) and HMGB1-negative (blue) nuclei. Results are expressed as the percentage of cells being negative. Intraclass correlation (ICC agreement) between the two observers was RNA isolation from the skin and real-time polymerase chain reaction (RT-PCR) Frozen skin samples were used for determination of mrna levels of HMGB1, RAGE, TLRs (2, 4 and 9) and IL-8 before and after UVB irradiation. mrna was isolated using RNeasy kit (Qiagen Benelux, The Netherlands) and cdna was synthesized from total RNA using M-MLV Reverse Transcriptase and oligo(dt) 24 (Life Technologies, USA). mrna expression of HMGB1, RAGE, TLRs (2, 4 and 9), IL-8 and glyceraldehyde-3- phosphate dehydrogenase (GAPDH) was measured by the real-time quantitative PCR system (ABI Prism 7900HT Sequence Detection System, Applied Biosystems, USA) with specific Taqman probes. The amount of target was normalized to an endogenous reference (GAPDH) and expressed as relative expression (2 CT ). Culture of normal human keratinocytes and irradiation Human primary keratinocytes (obtained from healthy persons undergoing breast correction operations) were provided by the Department of Dermatology, UMCG, and cultured in serum-free medium (Gibco, CA) supplemented with 25 mg/ml bovine pituitary extract and 0.2 mg/ml recombinant epidermal growth factor (regf) at 37 C in a 95% humidified atmosphere containing 5% CO 2. Cells were used for experiments after reaching 80% confluence. Primary keratinocytes were irradiated with 0.29 J/cm 2 UVB at a distance of 15 cm using a TL12 lamp (Philips, The Netherlands). Supernatants were collected after 0, four, eight, 16, 24, 48 and 72 hours, and the remaining cells were lysed with sodium dodecyl sulphate (SDS) buffer. Supernatants and cells from non-irradiated keratinocytes at similar time points served as control. The experiments were performed using keratinocytes from three different donors. 3

4 4 Western blotting for HMGB1 and cleaved caspase 3 Cleaved caspase 3 in keratinocytes and HMGB1 released from the cells into supernatants were determined by Western blot. Cleaved caspase 3 was used as a marker of apoptosis of keratinocytes. Lysed cell samples and supernatants were resolved in SDS-polyacrylamide gel electrophoresis (PAGE) and blotted to a polyvinylidene fluoride membrane (Millipore, The Netherlands) followed by incubation with rabbit anti-cleaved caspase 3 (Cell Signaling Technology, The Netherlands), mouse anti-a-actin (Santa Cruz Biotechnology, Germany) and mouse anti- HMGB1 (R&D Systems, UK). Detection was performed with goat anti-rabbit IgG labeled with IRDye800 or goat anti-mouse IgG labeled with IRDye680. Blots were scanned with Odyssey infrared Imaging System (LI-COR Biotechnology, USA) and analyzed with Odyssey software (version 2.1). A preparation of 1 million lysed Jurkat cells was used as a positive control in the HMGB1 blot. For cleaved caspase 3 detection, anti-actin antibody was included as an internal control. IL-8 detection by enzyme-linked immunosorbent assay (ELISA) IL-8 levels were measured in cell supernatants by ELISA, using antibody pairs for ELISA and recombinant protein as the standard (R&D Systems, UK). Statistical analysis Data are presented as median with interquartile range unless stated otherwise. The nonparametric Mann-Whitney U test was used for comparison between groups. Wilcoxon signed-rank test was used to compare paired samples at different time points. Correlations were calculated by Spearman s rank correlation test. P values less than 0.05 were considered significant. Statistical analysis was performed using the statistical package Graph Pad Prism, version 3.02 (Graph Pad Software Inc, USA). Figure 1 Expression in the skin of SLE patients and HCs before and after UVB exposure. (a) Isotype and HMGB1 staining in the epidermis of an HC and an SLE patient before UVB irradiation. (b) Percentage of HMGB1-negative nuclei in the epidermis in HCs and patients before and after UVB irradiation at day 1, 3, and 10. Open dots are HCs, filled dots are SLE patients. Horizontal line represents median with interquartile range. SLE: systemic lupus erythematosus; HC: healthy control; UVB: ultraviolet B; HMGB1: high mobility group box 1.

5 5 Figure 2 Expression in the skin of patients who did and did not develop infiltrates before and after UVB exposure. (a) Percentage of HMGB1-negative nuclei in patients with infiltrates (open dots) and without infiltrates (filled dots) before and at different time points after UVB-irradiation. Horizontal lines represent median with interquartile range. (b) HMGB1 staining in the epidermis and dermis of a patient with infiltrates at day 3 after UVB at 10 and 40 magnifications. In the oval blue nuclei, without HMGB1, are shown. HMGB1: high mobility group box 1; UVB: ultraviolet B. Results HMGB1 protein and HMGB1 mrna expression in the skin before and after UVB irradiation A representative picture of the HMGB1 staining in HC and SLE skin is shown in Figure 1(a). In nonirradiated skin (day 0), a significantly higher number of HMGB1-negative nuclei was seen in SLE patients as compared to HCs (p ¼ 0.003) (Figure 1(b)). In general, the number of HMGB1- negative nuclei increased after UVB irradiation, reaching a maximum at day 3 in HCs and patients (Figure 1(b)). Levels in patients were significantly higher compared to HCs at day 1 (p ¼ 0.04) day 3 (p ¼ 0.02) and day 10 (p ¼ 0.05). Five out of the 11 patients included in the study developed infiltrates at day 3 after UVB irradiation that persisted until day 10 and occurred in the vicinity of apoptotic keratinocytes as reported previously. 4 No significant difference in epidermal HMGB1 expression before UVB irradiation was seen between patients who developed infiltrates and those who did not, although the number of HMGB1-negative nuclei was higher in patients who developed infiltrates (Figure 2(a)). Of note, cells with HMGB1-negative nuclei were observed not only in the epidermal layer but also in infiltrated cells at the basal layer of the epidermis in patients who developed infiltrates. A representative picture is shown in Figure 2(b) at magnification 10 and 40 (blue nuclei visible in the oval form). Comparison of HMGB1 mrna expression between patients and HCs showed that most of the patients had higher relative mrna expression before UVB exposure (median 2.0, range (0.2 16)) than HCs (0.8, (0.2 4)), although this did not reach significance (Figure 3(a)). In HCs and patients, HMGB1 mrna expression decreased after irradiation, with lowest levels at day 3 and a tendency to normalization at day 10. No difference in HMGB1 mrna synthesis was seen between patients who

6 6 Figure 3 mrna expression of HMGB1 (a), IL-8 (b), TLR2 (c) and TLR4 (d) in the skin of HCs and SLE patients before and after UVB irradiation. Data are presented as relative expression compared to GAPDH (median with interquartile range). Open dots are HCs, filled dots are SLE patients,*p < 0.05, **p < HMGB1: high mobility group box 1; IL-8: interleukin-8; TLR: Toll-like receptors; HC: healthy control; SLE: systemic lupus erythematosus; UVB: ultraviolet B; GAPDH: glyceraldehyde-3-phosphate dehydrogenase. developed inflammatory lesions and those who did not (data not shown). Furthermore, HMGB1 mrna expression did not differ significantly between patients and HCs at any time point after irradiation (Figure 3(a)). UVB irradiation is known to modulate intracellular signaling pathways that affect gene transcription and synthesis. To distinguish between general UVB-mediated suppression of mrna synthesis and a more specific effect, we determined mrna levels of IL-8, as it is known that the release of this cytokine is increased after UVB irradiation. IL-8 mrna expression increased 24 hours after irradiation in both groups (Figure 3(b)). There was no significant difference in IL-8 mrna expression at any time point before and after UVB irradiation between patients and HCs. Expression of HMGB1 receptors in the skin of patients and HCs Since we observed increased numbers of HMGB1- negative nuclei in the epidermal layer, we were interested in the expression of HMGB1 receptors, that is TLRs (2, 4 and 9) and RAGE. Staining biopsies for RAGE expression showed that RAGE was abundantly expressed on keratinocytes in the epidermal layer from HCs as well as SLE patients. No change was seen in RAGE expression after UVB irradiation (data not shown). TLR staining on paraffin sections could not be performed because of lack of biopsy specimens. mrna levels of TLRs (2, 4 and 9) were relatively low in non-irradiated skin (day 0) of HCs and patients. After UVB irradiation, mrna levels of TLR-2 and -4 decreased in HCs and in patients and were lowest at day 3 (Figure 3(c) and (d)). Similarly, mrna levels of RAGE decreased after UVB irradiation and were lowest at day 3 in HCs and patients (data not shown). No significant difference in TLR and RAGE mrna expression was observed between HC and patients. Correlation of HMGB1 with SBCs and apoptotic cells Apoptotic keratinocytes (SBCs) can be detected in the skin of lupus patients after UVB exposure. 3,4,21 A representative picture of SBC staining is shown in Figure 4(a) (left figure) and of cleaved caspase 3 staining in the right figure. As these cells are able to release HMGB1, we investigated whether the presence of SBCs was related to the percentage of HMGB1-negative nuclei. Indeed, only in patients

7 7 Figure 4 Correlation between HMGB1-negative nuclei in the epidermis and sunburn cells (SBCs) in the skin of HCs and SLE patients after UVB irradiation. (a) Representative picture of SBCs (left figure, SBCs indicated by arrows) and cleaved caspase 3 (right figure, positive cells indicated by arrows) in the skin of patients after UVB (magnification 40). (b) HMGB1-negative nuclei in irradiated skin of SLE patients correlate positively with the number of SBCs (left graph); no correlation is seen in HCs (right graph). HMGB1: high mobility group box 1; HC: healthy control; SLE: systemic lupus erythematosus; UVB: ultraviolet B. was a positive correlation found between the presence of SBCs and HMGB1-negative nuclei after UVB exposure combining data from time points 1, 3 and 10 days (r ¼ 0.60, p ¼ 0.005) (Figure 4(b), left graph). This correlation could not be found in HC skin slides (Figure 4(b), right graph). Also, the number of cleaved caspase 3-positive cells was correlated to the percentage of HMGB1-negative nuclei in patients (r ¼ 0.55, p ¼ 0.01, data not shown). Correlation of HMGB1 with inflammatory cells (macrophages and T-cells) and MxA The numbers of infiltrating CD3- and CD68-positive cells were not significantly different between patients and HCs, although patients who developed infiltrates had the highest numbers of CD3 and CD68 cells at day 3. There was no significant correlation between their numbers and the percentage of HMGB1-negative nuclei in patients and HCs (data not shown). The numbers of MxA-positive cells were higher in SLE patients compared to HCs at all time points, reaching significance in the dermis at day 10 (p < 0.05). A significant correlation with the percentage HMGB1-negative nuclei and MxA, however, could not be found. HMGB1 and IL-8 release by primary keratinocytes To further understand the dynamics and underlying mechanism of HMGB1 release from keratinocytes due to UVB irradiation, primary keratinocytes were irradiated with UVB. HMGB1 was not detectable in the supernatants of nonirradiated keratinocytes. However, after UVB irradiation, HMGB1 was detected in supernatants from 16 hours onwards up to 72 hours, but not at earlier time points (Figure 5(a)). To test whether released HMGB1 detected in supernatants of irradiated keratinocytes is a product of apoptosis or a result of cell activation, we measured the viability of keratinocytes by assessing cleaved caspase 3 and their activation state by measuring IL-8 released into the supernatants. Nonirradiated keratinocytes showed no cleaved caspase 3 expression. After UVB irradiation, apoptosis of keratinocytes occurred from 16 hours onwards (Figure 5(a)). IL-8 was released in low amounts before UVB irradiation and increased after

8 8 Figure 5 Effect of UVB irradiation on normal human keratinocytes. (a) Representative Western blot of HMGB1 released from primary keratinocytes into the supernatant before and after UVB irradiation. Representative Western blot of cleaved caspase 3 expression in normal human keratinocytes before and after UVB irradiation. Lanes represent samples collected at baseline and four, eight, 16, 24, 48 and 72 hours after UVB irradiation, respectively. PC is an HMGB1-positive control. (b) IL-8 release (mean SD) from normal human keratinocytes (HK) before and after UVB irradiation as measured by ELISA. UVB: ultraviolet B; HMGB1: high mobility group box 1; IL-8: interleukin-8; ELISA: enzyme-linked immunosorbent assay. 16 hours of irradiation and remained elevated up to 48 hours (Figure 5(b)). Discussion The current study shows increased numbers of HMGB1-negative nuclei in the epidermis of SLE patients before UVB irradiation with a further increase thereafter as compared to HCs. Also, the presence of HMGB1-negative epidermal nuclei, in all likelihood resulting from HMGB1 release, was associated with the presence of SBCs. Cutaneous lesions are common manifestations in patients with SLE. The processes that induce or exacerbate cutaneous lesions are still elusive. Recently, the role of HMGB1 in the pathogenesis of SLE has been discussed HMGB1 has been shown to be responsible for mediating inflammation in spontaneously occurring and UV-induced skin lesions of patients with SLE. 18,19 In these reports the included patients were all photosensitive by history. In the current study, we measured the percentage of nuclei negative for HMGB1 as an indicator for HMGB1 release. Formally, nuclei lacking HMGB1 could result from release, but also from decreased synthesis due to UVB-induced cell cycle arrest. There was, however, no significant difference in mrna expression of HMGB1 between patients and controls, while the number of HMGB1-negative nuclei was already increased in the skin of SLE patients compared to HC before UVB exposure. In normal healthy tissue all nuclei stain positive for HMGB1 as also recently was shown for kidney tissue in lupus patients. 29 Based on our finding that nuclei were found negative and that mrna levels for HMGB1 were not increased after UVB irradiation excluding new HMGB1 synthesis, we took the percentage of negative nuclei as a measure of HMGB1 release. The underlying mechanism by which UVB contributes to increase in HMGB1 release is not well understood. HMGB1 release from nuclei in the epidermis might be due to UVB-induced apoptosis of keratinocytes. Indeed, mrna levels of HMGB1 decreased after UVB irradiation and the percentage of HMGB1-negative nuclei correlated with the number of SBCs in vivo. Also, the percentage of HMGB1-negative nuclei correlated with the number of cleaved caspase 3-positive cells in SLE skin. This is further supported by our in vitro experiments showing that HMGB1 release from primary keratinocytes coincided with the occurrence of apoptosis after UVB irradiation. Thus it is likely that these uncleared apoptotic cells release HMGB1 into the extracellular space. Otherwise, UVB may induce activation of keratinocytes and release of HMGB1 and pro-inflammatory molecules (IL-6, IL-8, TNF), which attract inflammatory cells. 1,2 In turn, these inflammatory cells secrete HMGB1-sustaining inflammation. Indeed, a significant release of IL-8 from keratinocytes after UVB irradiation could be demonstrated in vitro. More strikingly, the number of cells expressing interferon (IFN)-a-induced protein MxA after UVB irradiation was higher in SLE skin than in HC skin. Expression of MxA in cutaneous lesions of SLE is reported to be increased and levels of IFN-a in the blood of SLE patients are elevated, both supporting the concept that IFNa is a central cytokine in the pathogenesis of SLE. 30,31 It has been shown that the presence of HMGB1 enhances CpG-mediated release of IFNa from plasmacytoid dendritic cells (pdcs). 12 Furthermore, HMGB1 is directly involved in the induction of IFN-a. 32 Yanai et al. showed in an

9 animal model that the absence of HMGB1 impairs the IFN-a- and cytokine-induced capacity of nucleic acids to activate TLRs. 32 Taken together, our data suggest that HMGB1 might play an important role in IFN-a production. UVB exposure induces translocation of HMGB1 from the nucleus to the cytoplasm and extracellular fluid in experimentally UV-induced skin lesions. 18 Similarly, our data showed a significantly higher HMGB1 release from the epidermal skin of SLE patients after UVB exposure, in particular in those patients who developed infiltrates. However, it seems that HMGB1 and UVB alone are not the only determinants for lesion induction, as some patients had a high level of released HMGB1 in the skin even before UVB irradiation but did not develop infiltrates after UVB. It is possible that there is a prerequisite threshold for full lesion development. Otherwise, the discrepancy between basal HMGB1 expression and UVB-induced infiltrate development might reflect inter-individual variation in skin types, suggesting involvement of other pathological parameters that predispose to cutaneous lesions. In conclusion, this study supports the role of HMGB1 in UVB-induced inflammatory lesions in SLE patients. Also, it shows a direct correlation between HMGB1 and apoptosis of keratinocytes in the skin of lupus patients after UVB irradiation. Acknowledgment Primary keratinocytes were kindly provided by Miranda Nijenhuis of the Department of Dermatology (UMCG). Funding This research received no specific grant from any funding agency in the public, commercial, or notfor-profit sectors. Conflict of interest None declared. References 1 Skov L, Hansen H, Allen M, et al. Contrasting effects of ultraviolet A1 and ultraviolet B exposure on the induction of tumour necrosis factor-alpha in human skin. Br J Dermatol 1998; 138: Yoshizumi M, Nakamura T, Kato M, et al. Release of cytokines/ chemokines and cell death in UVB-irradiated human keratinocytes, HaCaT. Cell Biol Int 2008; 32: Kuhn A, Herrmann M, Kleber S, et al. Accumulation of apoptotic cells in the epidermis of patients with cutaneous lupus erythematosus after ultraviolet irradiation. Arthritis Rheum 2006; 54: Reefman E, de Jong MC, Kuiper H, et al. Is disturbed clearance of apoptotic keratinocytes responsible for UVB-induced inflammatory skin lesions in systemic lupus erythematosus? Arthritis Res Ther 2006; 8: R Gaipl US, Kuhn A, Sheriff A, et al. Clearance of apoptotic cells in human SLE. Curr Dir Autoimmun 2006; 9: Bell CW, Jiang W, Reich III CF, Pisetsky DS. The extracellular release of HMGB1 during apoptotic cell death. Am J Physiol Cell Physiol 2006; 291: C1318 C Casciola-Rosen LA, Anhalt G, Rosen A. Autoantigens targeted in systemic lupus erythematosus are clustered in two populations of surface structures on apoptotic keratinocytes. J Exp Med 1994; 179: Harris HE, Andersson U, Pisetsky DS. HMGB1: A multifunctional alarmin driving autoimmune and inflammatory disease. Nat Rev Rheumatol 2012; 8: Scaffidi P, Misteli T, Bianchi ME. Release of chromatin protein HMGB1 by necrotic cells triggers inflammation. Nature 2002; 418: Chen G, Li J, Ochani M, et al. Bacterial endotoxin stimulates macrophages to release HMGB1 partly through CD14- and TNF-dependent mechanisms. J Leukoc Biol 2004; 76: Kokkola R, Andersson A, Mullins G, et al. RAGE is the major receptor for the proinflammatory activity of HMGB1 in rodent macrophages. Scand J Immunol 2005; 61: Tian J, Avalos AM, Mao SY, et al. Toll-like receptor 9-dependent activation by DNA-containing immune complexes is mediated by HMGB1 and RAGE. Nat Immunol 2007; 8: Yu M, Wang H, Ding A, et al. HMGB1 signals through toll-like receptor (TLR) 4 and TLR2. Shock 2006; 26: Park JS, Gamboni-Robertson F, He Q, et al. High mobility group box 1 protein interacts with multiple Toll-like receptors. Am J Physiol Cell Physiol 2006; 290: C917 C Voll RE, Urbonaviciute V, Herrmann M, Kalden JR. High mobility group box 1 in the pathogenesis of inflammatory and autoimmune diseases. Isr Med Assoc J 2008; 10: Urbonaviciute V, Furnrohr BG, Meister S, et al. Induction of inflammatory and immune responses by HMGB1-nucleosome complexes: Implications for the pathogenesis of SLE. J Exp Med 2008; 205: Urbonaviciute V, Meister S, Furnrohr BG, et al. Oxidation of the alarmin high-mobility group box 1 protein (HMGB1) during apoptosis. Autoimmunity 2009; 42: Barkauskaite V, Ek M, Popovic K, Harris HE, Wahren-Herlenius M, Nyberg F. Translocation of the novel cytokine HMGB1 to the cytoplasm and extracellular space coincides with the peak of clinical activity in experimentally UV-induced lesions of cutaneous lupus erythematosus. 2007; 16: Popovic K, Ek M, Espinosa A, et al. Increased expression of the novel proinflammatory cytokine high mobility group box chromosomal protein 1 in skin lesions of patients with lupus erythematosus. Arthritis Rheum 2005; 52: Lokitz ML, Billet S, Patel P, et al. Failure of physiologic doses of pure UVA or UVB to induce lesions in photosensitive cutaneous lupus erythematosus: Implications for phototesting. Photodermatol Photoimmunol Photomed 2006; 22: Reefman E, Kuiper H, Jonkman MF, Limburg PC, Kallenberg CG, Bijl M. Skin sensitivity to UVB irradiation in systemic lupus erythematosus is not related to the level of apoptosis induction in keratinocytes. Rheumatology (Oxford) 2006; 45: Kuhn A, Wozniacka A, Szepietowski JC, et al. Photoprovocation in cutaneous lupus erythematosus: A multicenter study evaluating a standardized protocol. J Invest Dermatol 2011; 131: Reefman E, Kuiper H, Limburg PC, Kallenberg CG, Bijl M. Type I interferons are involved in the development of ultraviolet 9

10 10 B-induced inflammatory skin lesions in systemic lupus erythaematosus patients. Ann Rheum Dis 2008; 67: Abdulahad DA, Westra J, Bijzet J, Limburg PC, Kallenberg CG, Bijl M. High mobility group box 1 (HMGB1) and anti-hmgb1 antibodies and their relation to disease characteristics in systemic lupus erythematosus. Arthritis Res Ther 2011; 13: R Abdulahad DA, Westra J, Limburg PC, Kallenberg CG, Bijl M. HMGB1 in systemic lupus erythematosus: Its role in cutaneous lesions development. Autoimmun Rev 2010; 9: Hayashi A, Nagafuchi H, Ito I, Hirota K, Yoshida M, Ozaki S. antibodies to the HMGB1 chromosomal protein: Epitope mapping and association with disease activity. Mod Rheumatol 2009; 19: Li J, Xie H, Wen T, Liu H, Zhu W, Chen X. Expression of high mobility group box chromosomal protein 1 and its modulating effects on downstream cytokines in systemic lupus erythematosus. J Rheumatol 2010; 37: Harris HE, Andersson U, Pisetsky DS. HMGB1: A multifunctional alarmin driving autoimmune and inflammatory disease. Nat Rev Rheumatol 2012; 8: Abdulahad DA, Westra J, Bijzet J, et al. Urine levels of HMGB1 in systemic lupus erythematosus patients with and without renal manifestations. Arthritis Res Ther 2012; 14: R Mikita N, Ikeda T, Ishiguro M, Furukawa F. Recent advances in cytokines in cutaneous and systemic lupus erythematosus. J Dermatol 2011; 38: Sun CL, Zhang FZ, Li P, Bi LQ. LL-37 expression in the skin in systemic lupus erythematosus. 2011; 20: Yanai H, Ban T, Wang Z, et al. HMGB proteins function as universal sentinels for nucleic-acid-mediated innate immune responses. Nature 2009; 462: Hochberg MC. Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum 1997; 40: 1725.