The sunburn response in human skin is characterized by sequential eicosanoid profiles that may mediate its early and late phases

Abstract

Sunburn is a commonly occurring acute inflammatory process, with dermal vasodilatation and leukocyte infiltration as central features. Ultraviolet (UV) B-induced hydrolysis of membrane phospholipids releases polyunsaturated fatty acids, and their subsequent metabolism by cyclooxygenases (COXs) and lipoxygenases (LOXs) may produce potent eicosanoid mediators modulating different stages of the inflammation. Our objective was to identify candidate eicosanoids formed during the sunburn reaction in relation to its clinical and histological course. We exposed skin of healthy humans (n=32) to UVB and, for 72 h, examined expression of proinflammatory and anti-inflammatory eicosanoids using LC/ESI-MS/MS, and examined immunohistochemical expression of COX-2, 12-LOX, 15-LOX, and leukocyte markers, while quantifying clinical erythema. We show that vasodilatory prostaglandins (PGs) PGE(2), PGF(2alpha), and PGE(3) accompany the erythema in the first 24-48 h, associated with increased COX-2 expression at 24 h. Novel, potent leukocyte chemoattractants 11-, 12-, and 8-monohydroxy-eicosatetraenoic acid (HETE) are elevated from 4 to 72 h, in association with peak dermal neutrophil influx at 24 h, and increased dermal CD3(+) lymphocytes and 12- and 15-LOX expression from 24 to 72 h. Anti-inflammatory metabolite 15-HETE shows later expression, peaking at 72 h. Sunburn is characterized by overlapping sequential profiles of increases in COX products followed by LOX products that may regulate subsequent events and ultimately its resolution.

Figure 1.

Time course of sunburn erythema in human skin. Buttock skin of healthy human volunteers (n=22) was exposed to a UVB dose equivalent to 4× each volunteer’s sunburn threshold (4×MED). Erythemal response was quantified using a reflectance technique, giving an erythema index (EI) proportional to the blood content of the superficial dermis. Results are expressed as mean ± se EI in unexposed skin and in skin at 4, 18, 24, 48, and 72 h after UVB exposure. *P < 0.05, **P<0.001 vs. unexposed skin.

Figure 2.

Levels of ω-6 and ω-3 PUFA-derived prostanoids and their metabolites in cutaneous blister fluid during sunburn response. Panels show PGE₂ (A), 13,14-dihydro-15-keto PGE₂ (B), PGF_2α (C), 13,14-dihydro-15-keto PGF_2α (D), PGE₁ (E), 13,14-dihydro-15-keto PGE₁ (F), and PGE₃ (G). Buttock skin of healthy volunteers (n=22) was exposed to 4 × MED of UVB. Blisters were formed on UVB-exposed and unexposed skin, using suction cups and a vacuum pressure of 250 mmHg. Blister fluid samples were semipurified by solid phase extraction, followed by analysis by ESI/LC-MS/MS. Eicosanoids were identified by their m/z value and mass spectra, and quantitation was performed using multiple reaction monitoring-based assays. Results are expressed as mean ± se mediator (pg/mg protein) in unexposed skin and at 4, 18, 24, 48, and 72 h after UVB exposure. *P < 0.05, **P < 0.01, ***P < 0.001 vs. unexposed levels.

Figure 3.

Levels of 8-, 11-, 12-, and 15-HETE in cutaneous blister fluid during sunburn response. Panels show 8-HETE (A), 11-HETE (B), 12-HETE (C), and 15-HETE (D). Blister fluid samples from unexposed skin and 4 × MED of UVB-exposed skin were collected and analyzed (n=22 subjects; see Fig. 2). Results are expressed as mean ± se mediator (pg/mg protein) in unexposed skin and at 4, 18, 24, 48, and 72 h after UV exposure. *P < 0.05, **P < 0.01, ***P < 0.001 vs. unexposed levels.

Figure 4.

Neutrophil infiltrate in epidermis and dermis during sunburn response. Graphical presentation of time course (A), with representative photomicrographs of sections from unirradiated skin (B) and skin at 4 h (C), 24 h (D), and 72 h (E) after UVB exposure. Punch biopsy samples were collected from the buttock skin of healthy volunteers prior to and at intervals after exposure to 4 × MED of UVB. Frozen skin sections were immunohistochemically stained using antineutrophil elastase monoclonal antibody. Numbers of cells staining positive per hpf (×400 original view) were counted, using 3 fields in random epidermal/dermal interface locations for each time point. Results are expressed as mean ± se cell count; n = 10. *P < 0.05 vs. baseline.

Figure 5.

CD3⁺ lymphocytic infiltrate in epidermis and dermis during the sunburn response. Graphical presentation of time course (A), with representative photomicrographs of sections from unexposed skin (B) and skin at 4 h (C), 24 h (D), and 72 h (E) after UVB exposure. Punch biopsy samples were collected from buttock skin of healthy volunteers, and immunohistochemical analysis was performed to determine numbers of CD3⁺ cells per hpf (see Fig. 4). Results are expressed as mean ± se cell count; n = 10. *P < 0.05, ***P < 0.001 vs. baseline.

Figure 6.

Expression of COX-2 in the epidermis and dermis during sunburn response. Graphical presentation of time course (A), with representative photomicrographs of sections from unexposed skin (B) and skin at 4 h (C), 24 h (D), and 72 h (E) after UVB exposure. Biopsy samples were collected from buttock skin of healthy individuals, and immunohistochemical analysis was performed to determine levels of COX-2 expression in dermis and epidermis. Frozen sections were scored for relative staining intensity: 0 = no staining; 1 = weak staining; 2 = moderate staining; 3 = intense staining. Three hpf in random epidermal/dermal interface locations were examined for each time point. Arrows indicate COX-2 staining. Results are expressed as mean ± se staining intensity; n = 8. *P < 0.05 vs. baseline.

Figure 7.

Expression of 12-LOX in the epidermis and dermis during the sunburn response. Graphical presentation of time course of 12-LOX (A), with representative photomicrographs of sections from unirradiated skin (B) and skin at 4 h (C), 24 h (D), and 72 h (E) after UVB exposure. Punch biopsy samples were collected from buttock skin of healthy volunteers prior to and at intervals after exposure to 4 × MED of UVB. Immunohistochemical analysis was performed to determine relative levels of 12-LOX present in dermis and epidermis. Frozen sections were scored for staining intensity (×200 original view; see Fig. 6). Arrows indicate 12-LOX staining. Results are expressed as mean ± se staining intensity; n = 7. *P < 0.05; **P < 0.01 vs. baseline.

Figure 8.

Expression of 15-LOX in the epidermis and dermis following UVB irradiation. Graphical presentation of time course of 15-LOX expression (A), with representative photomicrographs of sections from unirradiated skin (B) and skin at 4 h (C), 24 h (D), and 72 h (E) after UVB exposure. Punch biopsy samples were collected from buttock skin of healthy volunteers prior to and at intervals after exposure to 4 × MED of UVB. Immunohistochemical analysis was performed on frozen skin sections to determine relative levels of 15-LOX in dermis and epidermis (see Fig. 6). Arrows indicate 15-LOX staining. Results are expressed as mean ± se staining intensity; n = 8. *P < 0.05, **P < 0.01 vs. baseline.

Scheme 1.

COX- and LOX-generated PUFA metabolites. Simplified illustration to show COX- and LOX-generated metabolites of the principal long-chain n-6 and n-3 PUFAs, AA, and EPA, respectively, focusing on eicosanoids relevant to the present study.

Scheme 2.

Composite illustration of the principal events of the sunburn response. Illustration shows the skin erythemal and leukocytic responses to UVB, indicating proposed key time points for contributions of COX- and LOX-generated eicosanoids.