UVA1 radiation attenuates pro- inflammatory functions in human monocytes

UVA1 therapy is effective in the treatment of inflammatory and autoimmune skin diseases. The mode of action of UVA1 therapy is not completely understood and especially data on cells of the innate immune system like monocytes, which are critically involved in many inflammatory processes, are sparse. We wanted to answer the question whether UVA1 irradiation alters functional properties of human monocytes. We treated human peripheral blood monocytes in vitro with 2 J/cm 2 UVA1 light, incubated the cells for 48 h and examined both functional properties and alterations in the gene and protein expression profile. While UVA1 did not alter cell viability or susceptibility to apoptosis inducing agents, it decreased the capacity of monocytes for phagocytosis and to eliminate infectious agents like Leishmania major . Moreover, we measured a significantly reduced production of interleukin (IL)- 1 β mRNA in lipopolysaccharide activated monocytes after UVA1 treatment. Importantly, UVA1- treated monocytes not only produce less IL- 1 β , but also upregulate expression of the anti- inflammatory IL- 1 β decoy receptor. Our data provide evidence that UVA1 radiation not only interferes with fundamental monocyte properties like phagocytosis, pathogen killing and activation, but could also specifically attenuate pro- inflammatory IL- 1 effects. This might constitute a hitherto unknown antiinflammatory mechanism of UVA1 in human monocytes.

Narrowband UVB (311 nm) and UVA1 (340-400 nm) are commonly used by dermatologists due to their effectiveness and their safety profile with few adverse effects. 1,2 While the clinical benefit of UV therapy has been known for many years, the mode of action of UV therapy is only partially understood. It is well established that UV light has effects both on resident skin cells like keratinocytes but also on leukocytes. 3,4 With respect to direct effects of UV therapy on immune cells, T-and B-cells and considering innate immunity dendritic cells (DCs) have been most extensively studied. UV irradiation is able to induce apoptosis in T-cells and to inhibit antibody production in B-cells. 3,5-7 UV exposure induced epidermal depletion of CD1apositive Langerhans cells which was compensated by the appearance of inflammatory DC subsets as well as an increased production of interleukin (IL)-12p40. 8,9 In dermal dendritic cells noncanonical autophagy was recently reported to mediate suppressive effects of UV exposure on these cells. 10 There is, however, only a little data on the impact of UV therapy on other cells of the innate immune system like monocytes and macrophages.
Our group has a long standing interest in macrophages and monocytes, especially in anti-inflammatory mechanisms of these cells. [11][12][13][14][15][16][17] Monocytes and macrophages are critically involved in the pathophysiology of inflammatory and autoimmune skin diseases. [18][19][20][21] Monocytes are a central part of innate immunity and are the progenitor cells of both dendritic cells as well as of macrophage subsets. 22,23 They play an essential role in the generation of inflammatory mediators and the initiation of innate and adaptive immune responses. 24 Equally important, they also contribute to resolution of inflammation, e.g. by secreting anti-inflammatory cytokines and by removal of cellular debris and apoptotic cells. 25,26 We demonstrated previously that glucocorticoids, the clinically most frequently used inhibitors of inflammation, induced a monocyte subtype with an anti-inflammatory phenotype which actively induces resolution of inflammation and could contribute to the clinical efficacy of these drugs. 11,12,14 It has only been sparsely investigated if UV therapy modifies monocyte functions and whether UV-mediated mechanisms in monocytes might also contribute to the immunomodulatory effects of UV therapy.
We decided to focus on the effects of UVA1 radiation.
Approximately 40% of UVA1 radiation penetrates through the epidermis reaching the dermis, while UVB radiation affects mainly the epidermis. 6 In the dermis, blood capillaries are present and hence the residing or infiltrating immune cells can be targeted by UVA1 radiation therapy.
Regarding potential effects of UVA1 radiation on monocytes, there is, to our best knowledge, no study addressing the question whether UVA1 light affects gene and protein expression as well as functional properties of human monocytes. We are only aware of one report which analyzed monocyte viability. 6 Higher doses of UVA1 radiation were associated with decreased monocyte survival.
However, in this study monocyte survival was not influenced by lowdose UVA1 (<2 J/cm 2 ) in vitro. 6 Since, in human skin, approximately 40% of UVA1 radiation reaches the dermis, this dose is comparable to low-dose UVA1 used in clinical practice and which is known to have profound anti-inflammatory effects, as shown e.g. in patients suffering from the autoimmune disease systemic lupus erythematosus (SLE). [27][28][29] Therefore, we wanted to answer the question whether UVA1 therapy alters monocyte properties and whether such effects on monocytes could contribute to the anti-inflammatory effects of this therapy. To this end, we treated monocytes with increasing doses of UVA1 in vitro and studied cell death, gene expression and surface protein expression. We also systematically studied the impact of UVA1 radiation on important cellular properties of monocytes like phagocytosis, parasite killing and migration in functional assays.
We demonstrate that UVA1 radiation attenuated pro-inflammatory properties of human monocytes in vitro.

| Isolation and stimulation of human monocytes
Human monocytes were isolated from fresh human blood leukocyte reduction chambers of platelet apheresis sets from healthy, volun- was controlled by flow cytometry using an anti-human CD14 antibody (BioLegend). 2 × 10 6 cells per mL were cultivated in RPMI 1640 medium (Biochrom), which was supplemented with 15% fetal calf serum (FCS), 2% L-glutamine and 1% penicillin/streptomycin (Biochrom), and allowed to rest overnight in Teflon bags preceding stimulation. Monocytes were irradiated with increasing doses of UVA1 (or with 2 J/cm 2 unless otherwise noted) by Sellamed 3000 (Sellas Medizinische Gerate GmbH) and were incubated at 7% CO 2 for 48 h.

| Real-time quantitative polymerase chain reaction
Monocytes were stimulated with 100 ng/mL LPS at 7% CO 2 for 3 h as earlier described. 13 RNA from control monocytes and UVA1irradiated human monocytes was extracted using the Quiagen RNeasy micro kit according to the manufacturer's instructions.
Semi-quantitative real-time polymerase chain reaction (RT-PCR) was performed as described previously on a Bio-Rad CFX384 Touch Real-Time PCR detection system. 13 The PCR data was normalized to the mean of the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH).

| Flow cytometry analysis
Flow cytometry analysis was performed as previously described. 12 Monocytes were harvested and washed with phosphate buffered saline (PBS). Surface cell proteins were stained with 1 μg/mL human monoclonal antibodies against IL-1RII and CD163 and the

| Phagocytosis and killing assay
To determine the phagocytosis rate of Leishmania (L.) major, complement opsonized parasites were labeled with the fluorescent dye carboxyfluoresceindiacetate succinimidylester (CFDA-SE; Thermo Fisher Scientific). Phagocytosis assay was performed as previously described. 12 After the harvest of the monocytes, 0.5 × 10 6 cells were seeded into a sterile FACS tube and co-cultivated with either labeled L. major at a 1:3 ratio or 1.0 μm yellow-green fluorescent polystyrene microspheres called latex beads (Thermo Fisher Scientific) at a 1:10 ratio for 4 h at 37°C. Monocytes were washed three times and the supernatant was discarded to remove extracellular parasites.

Phagocytosis was measured by FACS and assessed as MFI shift (tak-
ing into account that monocytes take up multiple parasites) and as the rate of monocytes which phagocytosed at least a single parasite/ latex bead.
For assessment of parasite killing, monocytes were incubated with CFDA-SE labeled L. major for 4 h, washed three times and further incubated for another 16 h. Elimination was assessed by comparing the amount of intracellular parasites determined by FACS after phagocytosis for 4 h and after additional 16 h culture for parasite killing. Both the percentages of monocytes, which contain at least a single parasite and the fluorescence intensity, measured as MFI shift, was determined after 4 h and additionally after 16 h cell culture, subtracted and depicted as difference (Δ). was added right before FACS measurement to assess late stages of apoptosis and necrosis.

| Transmigration assay
A total of 1 × 10 6 monocytes in FCS-free medium were placed in the upper chamber of a 5 μm pore-size Transwell filter (Thermo Fisher Scientific) and allowed to transmigrate for 4 h at 37°C. Leukotriene B4 (LTB4) (100 nM) was added to the FCS-containing medium in the lower chamber. The number of migrated monocytes was determined using a cell counter. Experiments were done in duplicate.

| Statistical analysis
Each experiment was conducted with ≥3 samples from different donors. Values represent mean ± SEM. Paired T-Test or Wilcoxon-Mann-Whitney test was applied to determine significance.

| UVA irradiation of monocytes does not induce cell death
Since it has been published that higher doses of UVA1 radiation induce cell death in human monocytes, we irradiated human monocytes with increasing doses of UVA1 (1, 2, 5, 10 J/cm 2 ) and cultured the cells in Teflon bags for 48 h. AnV staining was used to detect early stages of apoptosis and PI staining was used as marker for necrotic and late apoptotic events. In agreement with previous data on cell survival, 6

| Phagocytosis and killing in irradiated monocytes is diminished
Phagocytosis is a pivotal property of monocytes. It is a prerequisite for presentation of antigens and autoantigens. Phagocytosis of cellular debris and apoptotic leukocytes by monocytes and macrophages is essential for resolution of inflammation. Probably most important, however, is the phagocytic uptake of and consecutive killing of pathogens by human monocytes during defense against infections.
Thus, we examined the phagocytic capacities of UVA1-irradiated and control monocytes. Cells were allowed to phagocytose either fluorescence labeled latex beads or L. major parasites for 4 h F I G U R E 1 Monocytes are not susceptible to spontaneous apoptosis and apoptotic stimuli after low-dose UVA1 radiation. Human monocytes from healthy blood donors were irradiated with increasing doses of J/cm 2 UVA1 (1, 2, 5, 10 J/cm 2 ) and cultured in Teflon bags for 48 h. Monocytes were left untreated (spontaneous apoptosis) or were challenged with 200 nM staurosporine (STS) for 6 h to determine susceptibility to an apoptotic stimulus. Subsequently, monocytes were stained for Annexin V (AnV) and propidium iodide (PI) to determine stages of cell death. Fluorescence was measured by fluorescence-activated cell sorting (FACS). (a) shows representative charts of FACS analysis and bars show the percentage of AnV − /PI + (b, e), AnV + /PI + cells (c, f) and AnV + /PI − (d, g) of human monocytes irradiated with 1, 2, 5 and 10 J/cm 2 UVA1 during spontaneous and STS-induced apoptosis, respectively. The bars show mean and SEM of six independent experiments in spontaneous and STS-induced apoptosis. at 37°C. We were able to demonstrate that phagocytosis of latex beads (Figure 2a,c,e) and L. major (Figure 2b,d,f) was significantly decreased. To determine parasite elimination, monocytes were incubated with L. major for 4 h followed by vigorously washing and the amount of phagocytosis was determined. Cells were then cul-

| Monocytes still migrate after UVA1 radiation
During an inflammatory response monocyte infiltration into tissues follows infiltration of granulocytes and supports defense against pathogens (in case of an infection), but also helps to initiate the resolution phase of inflammation (by uptake of apoptotic cells and cellular debris). 25 The migration rate of human monocytes in vitro was assessed using Transwell filters. Chemotaxis was assessed by analyzing migration towards the monocyte chemoattractant LTB4. We could show that both unstimulated cell migration and chemotaxis was not significantly affected by UVA1 light (Figure 4). statistical significance. This, however, has to be interpreted with caution bearing in mind that using this UVA1 dose significant apoptotic mechanisms are already triggered.

| UVA1-treated monocytes produce reduced levels of pro-inflammatory cytokines in response to classical activation by LPS
Overall we could detect significantly decreased expression and secretion of IL-1β in low-dose irradiated monocytes stimulated with LPS.

| IL-1RII is significantly increased by UVA radiation
Having shown that LPS-induced production of pro-inflammatory IL-1β and IL-6 mRNA as well as IL-1β secretion was diminished in UVA1-treated monocytes while anti-inflammatory IL-10 was not affected, we next investigated whether the cell surface expression of anti-inflammatory proteins could be modulated by UVA1 treatment. Therefore monocytes were stained with the wellestablished anti-inflammatory proteins IL-1RII, the decoy receptor of IL-1β, the IL-1 receptor antagonist IL-1RA and CD163, the scavenger receptor for the hemoglobin-haptoglobin complex. 12,33,34 IL-1RII and CD163 are significantly increased during induction of an anti-inflammatory monocyte phenotype by treatment with glucocorticoids. 12 While we did not observe an altered protein level of CD163 on the cell surface of monocytes (Figure 6a), we demonstrate that the amount of IL-1RII was significantly increased in UVA1-irradiated monocytes (Figure 6b). In agreement with the cell surface expression, CD163 mRNA levels were not altered by UVA1-treatment (Figure 6d), while gene expression of IL-1RII was also significantly increased ( Figure 6e). Interestingly, we also observed a trend towards increased levels of IL-1RA on protein level but these differences did not reach statistical significance (Figure 6c), while gene expression seemed not to be affected ( Figure 6f). Since even IL1-R1 has been reported to interact with secreted IL-1β in some cases, 35 we also analyzed IL1-R1 expression ( Figure 5c) but did not observe an effect of UVA1-treatment.
In summary, the capacity of low-dose UVA1-treated monocytes to produce IL-1β was diminished while on the other hand the expression of the IL-1β decoy receptor IL-1RII was increased indicating that IL-1β signaling could be specifically affected by UVA1 treatment.

| DISCUSS ION
Although UVA1 radiation therapy is used for treatment of inflammatory and autoimmune skin disorders for nearly three decades, the molecular mechanisms of irradiation on particular cell types are still unclear. It is known that UVA1 radiation penetrates the epidermis and reaches the underlying cells of the dermis, including leukocytes. 6 UVA1 therapy is effective in the treatment of chronic skin disorders like atopic dermatitis and interestingly also in patients suffering from the important autoimmune disease SLE. [18][19][20][21] In these patients, usually prone to UV-induced disease exacerbations, low-dose UVA1 therapy was shown to have beneficial effects both on skin and systemic symptoms.
Most work on the cellular effects of UVA1 therapy focused on cells of the adaptive immune system. 5,6 However, it is becoming increasingly evident that monocytes and their successor cells are critically involved in the pathogenesis of chronic inflammatory and autoimmune skin disorders. [18][19][20][21]36 Thus, we pursued the concept that UVA1 therapy could also act on these cells. Data on UVA1 effects on monocytes are currently very sparse. 6 Based on the clinical experiences of anti-inflammatory effects of UVA1 radiation therapy and our expanding knowledge of anti-inflammatory mechanisms in monocytes both an inhibition of pro-inflammatory mechanisms and an induction of anti-inflammatory capacities could be possible. 16,37,38 Thus, we studied effects of UVA1 on basic monocyte functions and were able to show that low-dose UVA1 radiation specifically interferes with both pro-and anti-inflammatory capacities, while it does not affect viability and migratory properties of these cells. UVA1treated monocytes show decreased phagocytic and killing capacity and decreased production of IL-1β but increased production of the IL-1β decoy receptor.
Anti-inflammatory effects of UVA1 light have partly been attributed to its capacity to induce apoptosis of leukocytes, as shown for effector T-cells and B-cells. To analyze whether UVA1-treated monocytes display a reduced capacity to react to pro-inflammatory stimuli we next analyzed the response of monocytes to LPS, used as prototypic agent for proinflammatory activation of human monocyte by PAMPS.
In LPS-stimulated monocytes which have been previously irradiated with UVA1 we observed a trend towards diminished production of IL-6, TNFα mRNA, protein secretion and a significant reduction of IL-1β mRNA as well as the secreted protein.
Other cytokines and chemokines were not affected by UVA1 light. Thus, UVA1 radiation does not result in a general unresponsiveness towards LPS stimulation, but seems to act on a limited set of LPS-responsive genes. IL-1β, a member of the IL-1-family, is a potent pro-inflammatory cytokine in the acute and chronic phase of inflammation. It is the most important cytokine in autoinflammation but also known to act on adaptive immunity. [40][41][42] Therefore, reduced production of IL-1β is in agreement with a reduced capacity of UVA1-treated monocyte in the amplification phase of inflammation.
While we detected a reduced capacity of UVA1-treated monocytes to pro-inflammatory activation by LPS stimulation, we did not F I G U R E 5 UVA1 irradiation attenuates lipopolysaccharide (LPS) induced cytokine expression. Human monocytes were treated with increasing doses of UVA1 as described in Figure 1 and subsequently stimulated with 100 ng/mL LPS for 3 h. Afterwards, cells were lysed and analyzed for the expression of IL-1β, IL-6, IL-10, TNFα, CXCL9 and CXCL10 as well as IL-1R1 by qRT-PCR (a, c). Data were normalized to the mean of the housekeeping gene glyceraldehyde-3-phosphat dehydrogenase (GAPDH). The supernatant of these LPS-stimulated cells was also analyzed for IL-1β, IL-6, IL-10, TNFα protein secretion using the LEGENDplex Human Inflammation Panel 1 (b). The bars show mean and SEM of four independent experiments.
observe an induction of anti-inflammatory cytokine IL-10. However, cell surface proteins are also well-established as anti-inflammatory principle in human monocytes as shown for action of glucocorticoids on these cells. 12,16,34 Although we could not detect a general anti-inflammatory monocyte subtype when analyzing cell surface protein expression, UVA1 radiation -similarly to anti-inflammatory glucocorticoidsenhanced the expression of the anti-inflammatory IL-1β decoy receptor IL-1RII and there was a trend towards increased expression of the IL-1RA.
IL-1RII is the decoy receptor of IL-1. When IL-1RII binds IL-1β, signal transduction cannot be initiated and the pro-inflammatory properties of IL-1β via the IL1-R1 are neutralized. 43 IL-1RII not only exists as a membrane bound protein but also as secreted protein and can be rapidly shed. Therefore, it can also act as a soluble protein binding and thereby neutralizing IL1β. 35 While increased expression of membrane bound IL-1RII indicates a reduced responsiveness of monocytes to IL1β stimulation, the soluble form of IL-1RII is able to inhibit effects of IL1β on other cells. 43 We also observed a trend towards increased protein production of IL-1RA in UVA1treated monocytes. IL-1RA is a member of the IL-1 family that exerts its effects in a similar way by binding to IL-1 receptors and thus blocking IL1-beta signaling. 44 The combination of UVA1-mediated reduction in IL-1β production and increased production of IL-1β neutralizing proteins could cooperatively diminish IL-1β-mediated proinflammatory effects of monocytes.
IL-1β is involved in the pathophysiology of many inflammatory disorders with skin symptoms and it is a major cytokine produced by keratinocytes. [45][46][47] An increased production of soluble IL-1RII receptor by monocytes following UVA1-radiation could limit the effect of IL-1β produced by both infiltrating leukocytes and resident skin keratinocytes and thereby contribute to the anti-inflammatory mechanisms of UVA1 radiation in the skin.
As mentioned above IL-1β is very important for innate immunity but also involved in adaptive immune responses. 40,42 As such a dysregulation of the IL-1 pathway was associated with disease activity in SLE, 48 a disease which is responsive to UVA1 treatment. [27][28][29] High levels of the soluble IL-1RII inversely correlated with autoantibody titers and with clinical disease activity in SLE patients. [49][50][51] In line with these reports, Anakinra, a recombinant human IL1RA has been demonstrated to ameliorate disease activity not only in disorders with prominent involvement of innate immunity, but also in clinical trials with SLE patients. 41,[52][53][54] Therefore, an UVA1 mediated reduction of pro-inflammatory IL-1β signaling could affect adaptive immunity and might contribute to the efficacy of UVA1 therapy in SLE.
Despite clinical efficacy, UVA1 therapy is currently not an approved treatment strategy. Our data indicate that effects of UVA1 on monocytes should be considered in the design of further trials on the effects of UVA1 in autoimmune and inflammatory skin disorders.
In conclusion, our data provide evidence for an effect of UVA1 irradiation on functional properties of human monocytes. UVA1 interferes with phagocytosis and elimination of infectious agents and modulates classical pro-inflammatory activation of monocytes.
The tendency towards an increased production of IL-1RA, the significant induction of the IL-1 decoy receptor and the concurrent F I G U R E 6 UVA1 radiation leads to an increased gene and protein expression of the cell surface protein IL-1RII in monocytes. Monocytes were treated as described in Figure 2 and subsequently stained for CD163, IL-1RII and IL-1RA and analyzed by fluorescence-activated cell sorting (FACS) (a-c). Bars show the quotient of specific stained/isotype control mean fluorescence intensity (MFI) shift. Gene expression (d-f) was assessed as described in Figure 5. and data were normalized to the mean of the housekeeping gene glyceraldehyde-3-phosphat dehydrogenase (GAPDH). The bars show mean and SEM of five independent experiments.
inhibition of IL-1β production by UVA1 treatment suggest that a mitigation of pro-inflammatory IL-1β effects could constitute a hitherto unknown anti-inflammatory mechanism of UVA1 in human monocytes.

ACK N OWLED G M ENTS
We thank Ulla Nordhues for excellent technical assistance and Dr.Theresa Ortkras as well as Dr. Nicole Freise for helpful discussions. Funding was provided by grant Eh2/019/11 by the intramural program "Interdisziplinäres Zentrum für Klinische Forschung" (IZKF).