UVB-generated Microvesicle Particles: A Novel Pathway by Which a Skin-specific Stimulus Could Exert Systemic Effects

Abstract

Ultraviolet B radiation (UVB) exerts profound effects on human skin. Much is known regarding the ability of UVB to generate a plethora of bioactive agents ranging from cytokines and other bioactive proteins, lipid mediators and micro-RNAs. It is presumed that these agents are in large part responsible for the effects of UVB, which only is absorbed appreciably in the epidermis. However, the exact mechanism by which these bioactive agents can leave the epidermis are as yet unclear. This review addresses the potential role of microvesicle particles (MVP) as UVB signaling agents through transmitting biologic mediators. New data is provided that UVB treatment of human skin explants also generates MVP production. We hypothesize that UVB production of MVPs (UVB-MVP) could serve this important function of transmitting keratinocyte-derived bioactive agents. Moreover, we propose that UVB-MVP formation involves the lipid mediator Platelet-activating factor. This novel pathway has the potential to be exploited pharmacologically to modulate UVB effects.

Graphical Abstract

UVB generates microvesicle particles through Involvement of the lipid mediator Platelet-activating Factor (PAF). The cartoon depicts theoretical pathway by which UVB-induced reactive oxygen species (ROS) which generate PAF agonists that then trigger microvesicle particle release. This cartoon also suggests potential targets by which this pathway can be modulated.

INTRODUCTION

Ultraviolet (UV) radiation is divided into three types by wavelength range: UVA (320–400 nm), UVB (290–320 nm), UVC (100–290 nm). Of the three types, UVB is the most biologically active. UVB radiation is absorbed by the epidermis and cannot penetrate appreciably into the dermal layer of the skin (1). When the epidermis is exposed to UVB radiation, the major cell type found in the epidermis, the keratinocyte, produces many cytokines including: IL-1α, IL-1β, IL-6, IL-8, granulocyte colony stimulating factor (G-CSF), macrophage-CSF, interferon gamma (INF-γ), platelet-derived growth factor (PDGF), transforming growth factor alpha (TGF-α), TGF-β, and tumor necrosis factor alpha (TNF-α) (1,2). UVB also generates antimicrobial peptides including human β-defensin-2, -3 and ribonuclease-7 and psoriascin (S100A7) (3). UVB-mediated upregulation of antimicrobial peptides has been suggested to provide an explanation as to why skin treated with this immunosuppressant, UVB, is resistant to secondary infection (4). UVB also upregulates the alarmin high mobility group box 1 (HMGB1), which acts via toll-like receptors (TLRs 2,4 & 9) and the receptor of advanced glycation end products (RAGE) to generate sterile inflammation (5).

UVB exposure also stimulates the production of many bioactive lipids such as prostaglandin E2 and platelet-activating factor (1-alkyl-2-acetyl-glycerophosphocholine; PAF) (2,6). Regarding PAF, UVB has been shown to induce the enzymatic synthesis of PAF in keratinocytes as well as the production of 1-alkyl glycerophosphocholines with sn-2 oxidatively modified fatty acids (ox-GPC) PAF-receptor agonists in a non-enzymatic process via reactive oxygen species (6,7). Indeed, UVB-mediated PAF-R agonists are diminished by antioxidants, indicating the important role of ROS in this process. UVB irradiation of keratinocytes also results in the upregulation of multiple microRNAs, which are small (~22 nucleotides in length) “non-coding” RNAs that negatively regulate gene expression post-transcriptionally (8).

The production of these bioactive proteins and lipids and nucleic acids by UVB can lead to acute inflammation, erythema, degenerative aging, and cancer (9,10). UVB radiation, especially at high doses, can result in systemic effects including fever as well as immunosuppression (11,12). Given that UVB only absorbs appreciably into the epidermis, it is presumed that bioactive agents released by the epidermis then serve as effectors for UVB responses. Yet, the exact mechanism(s) by which keratinocyte-derived bioactive products can leave the epidermis to exert effects is unknown. In keeping with the theme of honoring Dr. Hasan Mukhtar for his important contributions in photobiology and carcinogenesis including providing an enhanced understanding of keratinocyte-generated bioactive agents in these processes, this review presents a novel hypothesis and provides additional evidence for microvesicles as a potential mechanism for the UVB-mediated release of bioactive products.

Microvesicle particles (MVP), also named as microvesicles and microparticles, are small membrane-bound particles with a diameter between 100–1000 nm that can be shed from the surface of virtually all eukaryotic cells in an active energy-dependent process (13). Though originally MVP were felt to be unimportant cellular debris, increased interest in these agents has arisen as MVP contain a variety of bioactive substances such as proteins, bioactive lipids, cytokines, and nucleic acids. Moreover, the membranes of MVP can also contain membrane receptors and adhesion molecules. Though very little is known about keratinocyte MVP, cytokines that have been demonstrated to be found in MVP from other cell types includes TNFα derived from alveolar macrophages (14) and endothelial cells (15). The bioactive alarmin HMGB1 also has been demonstrated to be found in MVP released from macrophages stimulated with cigarette smoke extract (16). Thus, it is possible that numerous agents produced by UVB could travel in MVP.

Once they were released by the parental cells, MVP can transfer biological information from parental cells via direct fusion or internalization into the target cells and transfer content molecules directly or through activation of membrane receptors (13). Due to their important roles in cell to cell communication, the identification of MVP in multiple bodily fluids has raised interest in revealing their functions not only as biomarkers (17), but also as important mediators in pathogenesis mechanisms (18) and even potential targets for therapy (19,20). Despite a growing understanding of MVP functions, the formation and releasing mechanisms of MVP remain incompletely understood. Several studies have reported that MVP formation and release was associated with the ATP-dependent receptor P2X7R (21), small GTPases (22,23), and mitogen-activated protein kinase (MAPK) pathways (16,21). Indeed, MVP are released in response to cell activation by various stimuli such as inflammatory cytokines and shear stress, or even following serum withdrawal (24). Of interest, different pathways appear to be important in MVP release from various stressors. For example, in endothelial cells TNFα-induced MVP release is blocked by P38 MAPK and NFƙB inhibitors, whereas shear stress-mediated MVP release is attenuated by P42/P44 ERK MAPK and Rho kinase inhibitors (25). Thus, MVP can be generated in response to many different stressors via a variety of signaling pathways. Given that UVB irradiation of keratinocytes results in the activation of many of the same pathways important in MVP formation in other cell types, it is plausible that UVB could generate MVP in keratinocytes.

Recently our group reported in this journal in vitro preliminary studies indicating that keratinocytes irradiated with UVB can induce the production of MVP (26). Using the human keratinocyte-derived cell line HaCaT, we found that UVB generated MVP in a dose-dependent manner, seen as early as one hour post irradiation. Given that UVB generates PAF receptor (PAF-R) agonists in keratinocytes and in human skin (6,7), and that UVB-induced cytokine generation was augmented by the PAF-R (2) we asked if UVB-induced PAF agonists could be involved in MVP release. First, we found that treatment of HaCaT cells with the PAF-R agonist carbamoyl-PAF (CPAF) resulted in MVP release. Next, we demonstrated that pre-treatment of HaCaT cells with the antioxidants vitamin C and N-acetyl cysteine, at doses which block UVB-mediated generation of PAF-R agonists (6,7), diminished MVP production in response to UVB but not CPAF. Though competitive PAF-R antagonists are available, they are less potent than the native ligand (27). Thus, to more definitively assess the role of PAF-R agonists in the production of UVB-MVP, we used a PAF-R-negative human epithelial cell KB which was transduced with functional PAF-Rs. This model system consisting of KBP cells transduced with the leukocyte PAF-R, and KBM cells transduced with MSCV2.1 blank retroviral vector has been previously characterized (28). Using this model system of PAF-R-positive (KBP) and –negative (KBM) cells, we found that UVB only generated MVP in KBP cells.

RESULTS AND DISCUSSION

We report here additional experiments testing the ability of UVB and topical CPAF to generate MVP in human skin. To that end, we obtained de-identified human skin discarded from abdominoplasty surgeries (IRB exempted, Sycamore Hospital, Dayton, Ohio). The skin was rinsed, excess fat removed, and placed in PBS warmed at 37°C. Blisters were generated with a Vacuubrand^® vacuum pump, 100 mbar, attached to the barrel of a 20 mL syringe. Once blisters had formed (see Fig 1 A and B), the barrels were removed and the blisters were treated. As shown histologically in Fig 1C, the blisters were induced in such a fashion to separate epidermis from dermis as previously reported (29). The epidermal roofs of the blisters were treated with localized UVB using a Philips F20T12/UVB lamp as previously described (6), or topical CPAF (Avanti Polar Lipids) or DMSO (100 ul of 90% DMSO/10% ethanol) vehicle. Sham blisters received no treatment. To prevent migration, DMSO- and CPAF-treated blisters were separated from the other blisters by cutting through the dermis with a scalpel. Four hours after treatment, fluid was removed from inside the blisters using a 30ga needle on a 1 ml syringe and weighed in tared tubes. After adjusting the volumes of the blister fluid to 500 µl with HBSS, MVP were isolated by differential centrifugation and quantified using a NanoSight NS300 instrument (NanoSight Ltd) as previously described (26). As depicted in Fig 1D, both UVB irradiation as well as topical CPAF induced increased levels of MVP formation in the blister fluid. DMSO vehicle alone did not result in the increased generation of MVP. Given that UVB only appreciably reaches the epidermis, these findings fit with the notion that keratinocytes release MVP in response to UVB or the PAF-R agonist CPAF which then travel to the dermis.

Figure 1. UVB irradiation of human skin explants results in MVP release.

A, B. Suction-induced blisters were generated on human skin tissue, which results in a sub-epidermal blister (C). D. Treatment of blister roofs with either 1 KJ/m² or 2.5KJ/m² UVB or topical CPAF (5.4 ng) resulted in increased MVP in the blister fluid at 4 h post-treatment in comparison to either no treatment (sham), or DMSO vehicle. The data depicted are mean +/− SE MVP per mg blister fluid from 6–8 separate experiments. * denotes statistically significant changes from Sham or DMSO vehicle treatment (p <0.05 using Student T test).

Combining the previously published in vitro findings that UVB-generates MVP (UVB-MVP) in a process which involves the keratinocyte PAF-R (26) with the present ex vivo studies allows us to hypothesize that the keratinocyte PAF-R could serve as an effector for UVB-MVP release from human skin (Fig 2). Several lines of evidence are supportive of the notion that activation of the PAF-R mediates UVB-MVP. First, UVB fluences that generate UVB-MVP also trigger the production of PAF and ox-GPCs (6,7). Of significance, UVB-mediated PAF-R agonist formation and UVB-MVP are both blocked by antioxidants (6,7). Second, the keratinocyte PAF-R is linked to the same signal transduction pathways that have been implicated in MVP formation in other cell types (30). Finally, our previously published in vitro studies (26) using PAF-R-positive KBP with PAF-R-negative KBM cells indicates that the PAF-R is necessary for the formation of UVB-MVP.

Figure 2. Hypothetical model of UVB-MVP formation.

UVB stimulates the production of PAF and ox-GPCs via reactive oxygen species, which then act upon the keratinocyte PAF-R resulting in MVP formation and release. This model denotes that this process of UVB-MVP production could be potentially inhibited by antioxidants, PAF-R antagonists, as well as by agents that could possibly interrupt the MVP formation in response to PAF-R activation (e.g., MAPK inhibitors).

UVB-mediated production of PAF and ox-GPCs is involved in the acute UVB response as well as in systemic immunosuppression (31,32). Our previous studies have demonstrated that intradermal injection of PAF results in immediate skin inflammation, and generates a similar cytokine profile as UVB (2,29,31). Moreover, we have discovered that cells lacking the xeroderma pigmentosum complementary group A (XPA) protein which exhibit augmented reactive oxygen species (ROS) in response to pro-oxidative stressors generate large amounts of PAF-R agonists following UVB (31). The photosensitivity associated with the XPA-deficient mouse is also PAF-R dependent (31). Regarding UVB-mediated immunosuppression, studies by the Ullrich laboratory have demonstrated that the PAF-R on the mast cell is necessary for this process (12,33). Of importance, PAF/ox-GPCs are metabolically labile, quickly degraded by PAF-acetylhydrolases found in serum and in cells (34). Thus, it is possible that UVB-generated keratinocyte-derived PAF-R agonists use MVP to transit to mast cells as a mechanism to avoid degradation.

The functions of various MVP in physiological and pathological processes depend on their carried contents (proteins, lipids and nucleic acids). As noted above, one advantage of this arrangement is that metabolically unstable compounds inside MVP can be somewhat protected from enzymatic degradation. A recent study described that circulating MVP are increased in inflammatory skin diseases such as psoriasis, in which many cytokines are also elevated (35). In translational studies, elevated levels of MVP have been documented in vascular diseases and correlated with the severity of disease (36–38). Though MVP have been examined in the context of various disease states ranging from psoriasis to cardiovascular stroke, the concept of MVP as a transmitter of systemic signals following a cutaneous insult such as UVB is as yet unexplored.

CONCLUSION

One long-standing question in photobiology concerns the ability of UVB irradiation, which only is absorbed appreciably in the epidermis, to generate systemic effects. This short review serves to provide the rationale and additional experimental evidence implicating MVP in this process. The significance of this new area of research is that if UVB-MVP serve as effectors for UVB, this provides a previously unappreciated target for the modulation of UVB effects. This novel pathway could be important in actinic neoplastic syndrome as well as in photosensitivity disorders. Strategies which would likely block UVB-MVP formation could include agents such as antioxidants to block PAF/ox-GPC formation, PAF-R antagonists, or blocking signal transduction pathways downstream of the PAF-R involved in UVB-MVP generation.