Targeting β2-adrenergic receptor reduces UV-induced cutaneous damage and inflammation in a murine model

Abstract

Preclinical studies demonstrated that the β-adrenergic receptor antagonists (β-blockers) inhibit skin damage and cancer development induced by ultraviolet (UV) radiation, but the mechanism remains unknown. β2-adrenergic receptor (β2-AR) is the predominant adrenergic receptor expressed on skin keratinocytes and immune cells that bind to catecholamines, but its function in UV-induced skin lesions is unknown. Here, the role of β2-AR in UV-induced acute skin damage was investigated using a β2-AR knockout (KO) mouse model. The β2-AR KO mice exhibited attenuated UV-induced skin edema, sunburn, erythema, barrier disruption, apoptosis, and overexpression of IL-6, accompanied by a transient elevation in expression of β1- and β3-ARs. Cytokine array and immunohistochemical analysis of the KO skin revealed reduced UV-induced overexpression of multiple cytokines and chemokines involved in leukocyte infiltration and inflammation. RNA-sequencing analysis confirms that UV triggers a differential transcriptional response between the WT and KO skin. Furthermore, RNA-sequencing identified multiple gene regulatory pathways involved in the KO skin, including reduced activity of the pro-inflammatory transcription factor NF-κB, increased activity of Interferon Response Factors (IRFs) and the glucocorticoid receptor (GR), and reductions in myeloid immune cell/macrophage-related signaling pathways such as CEBP-β and GATA transcription factors. Collectively, these gene regulatory alterations were associated with a substantial reduction in innate immune, inflammatory, and mesenchymal tissue differentiation responses to the UV radiation in the KO skin. These data identify β2-AR as a critical neurobiological pathway involved in UV-induced skin damage and inflammation and support that β2-AR blockade might be useful for preventing UV-related skin lesions and sequelae (e.g., cancers).

Introduction

Skin is not only a barrier organ defending against environmental stressors, but also a receiver of internal stressors. The intricate relationship between the brain and skin has been documented since ancient times [1]. Psychosocial stress exerts negative effects on the skin and exacerbates many skin diseases [2]. Interestingly, skin is also a local source of stress hormones [1]. A peripheral catecholaminergic system exists in the skin, where epinephrine can be synthesized in keratinocytes in response to environmental stressors such as UV light [3]. Many cell types in skin express β-adrenergic receptors (β-ARs) [4,5], which are the receptors for catecholamines, mediating local skin reactions to physical and psychosocial stress.

The β2-adrenergic receptor (β2-AR) is a member of the adrenergic family of G-protein-coupled receptors that bind to catecholamines, which function as both local neurotransmitters (from sympathetic nerve fibers) and blood stress hormones (resulting from adrenal epinephrine release and neurotransmitter spillover into blood). The predominant adrenergic receptor expressed in epidermal keratinocytes is the β2-AR subtype [6]. β2-ARs are also expressed on melanocytes and fibroblasts [4,5,7], and are the most prevalent subtype on immune cells and therefore, play a role in modulating immunity, allowing for neural regulation of immune responses [8,9]. Chronic stress, partially mediated by the catecholaminergic system, promotes skin carcinogenesis in SKH-1 hairless mice [10]. Pharmacologic β-AR antagonists (“β-blockers”) inhibit catecholamine binding to the β-ARs and previous studies have indicated skin cancer-protective effects of several β-blockers, including the non-selective β-blocker carvedilol and the β2-AR selective blockers ICI-118,551 and butoxamine [11,12]. However, the neurobiological mechanism of this stress effect has not been identified, and the role of β2-ARs in UV-induced skin damage (i.e., in the absence of psychosocial stress) is also unknown.

The role of epinephrine in skin has been described in several studies using preclinical models of skin wounds and repair. Epinephrine at the concentration range of 50 nM–1 μM, which is detected in the circulation of stressed individuals for days or weeks after initial trauma, activates β2-AR and negatively affects keratinocyte motility and wound healing in vitro, ex vivo, and in vivo [13–17]. In a murine model of the skin burn wound, stress-induced increases in systemic and local epinephrine delayed wound repair [15]. Blocking the binding of the ligand to the β-ARs by the administration of β-blockers improved the healing of the burn wound [15]. Studies using excisional skin wound models demonstrated that loss of the β2-AR gene or β2-AR-specific blocker treatment promoted wound repair [18]. In another study using tape stripping to disrupt the skin barrier, the β2-AR agonists delayed skin barrier recovery while the β2-AR antagonists blocked the delay, but the agonists and antagonists of β1-AR, α1- and α2-AR did not affect barrier function [19]. Furthermore, topical β-blocker timolol has shown clinical efficacy in treating chronic skin wounds with various etiologies [20–23]. Therefore, β2-AR may play an important role to enhance the detrimental consequences in the skin induced by many environmental factors. Although how β2-AR activation promotes skin damage is not completely understood, the downstream signaling pathways may involve cAMP-PKA and MAPK, which lead to gene regulation changes associated with oxidative stress, apoptosis, inflammation and barrier dysfunction [24]. Therefore, it is crucial to investigate the role of β2-AR in mediating skin damage following exposure to UV radiation, a ubiquitous environmental stressor. This basic scientific information is essential in the development of effective treatments to prevent UV-induced skin cancer.

In the present study, we sought to determine the role of β2-ARs in the basic effects of UV radiation on skin damage, sunburn, and inflammation, which are the early-stage drivers of later skin carcinogenesis [25]. We focused on β2-ARs in particular because they are the predominant adrenergic receptor expressed in keratinocytes and immune cells, and developed a whole-body β2-AR gene (Adrb2) knockout mouse model in the SKH-1 genetic background, which is a widely used mouse strain for skin research [26]. We examined the effects of single-dose or multiple-dose UV radiations on the skin in β2-AR KO mice and wild-type (WT) littermate controls, analyzed histological effects on the skin, and used RNA sequencing (RNA-seq) to explore the dermal and immune cell dynamics, and gene regulatory pathways involved in mediating these effects.

Materials and methods

Animal models

All animal studies were carried out under the recommendations and guidelines established by the Western University of Health Sciences’ Institutional Animal Care and Use Committees which has approved the study protocol (protocol #: R23/IACUC/015). Mice had access to water and food ad libitum and were housed on a 12-hour light/dark cycle with 35 % humidity.

Supplementary Fig. 1A depicts the procedure to establish the β2-AR gene KO model in the SKH-1 background. Mice that were homozygous null for both β1-AR and β2-AR, encoded by Adrb1 and Adrb2 genes, respectively, were purchased from Jackson Laboratory (Bar Harbor, ME) (Adrb1^tm1Bkk Adrb2^tm1Bkk/J, Stock Number 003810) and were described previously (Rohrer et al., 1999). The strain of these double knockout mice was in a mixed background of 129S1/Sv*129 × 1/SvJ*C57BL/6J*DBA/2*FVB/N. The double KO mice were bred with the WT C57BL/6J mice which were purchased from Jackson Laboratory to generate a heterozygous F1 generation. The F1 generation was bred together to generate the F2 generation of which 1/16th of the offspring was a β2-AR single gene KO mouse. To obtain the hairless β2-AR KO mice, SKH-1 mice (Crl:SKH1-Hr^hr) which were purchased from Charles River (Wilmington, MA) were bred with the β2-AR KO mice generating a second heterozygous F1 generation. The F1 generation was bred with the WT SKH-1 mice for more than ten rounds to generate hairless mice that are in SKH-1 background. The heterozygous mice (Adrb2 ±) were bred together to generate hairless WT and β2-AR KO mice for experiments. To confirm deletion of the Adrb2 gene, we examined the mRNA expression of three β-AR subtypes using RT-PCR (Supplementary Fig. 1): in skin tissues isolated from the β2-AR KO mice, only Adrb1 and Adrb3 mRNA were detectable, while the WT skin robustly expressed all three Adrb subtypes.

UV light source and animal exposure device

The UV light source and animal exposure equipment have been described in our previous report [27]. In brief, the UV light source came from UV lamps that emit UVB (280 – 320 nm; 54 % total energy), UVA (320–400 nm; 37 % of total energy), UVC (100–280 nm; 2.0 % of total energy) and visible light (400–450 nm; 7.0 % of total energy) (#95-0042-08 and #95-0043-13: UVP, Upland, CA) used in all mouse experiments. The UV lamp power output (mW/cm²) was measured using a UVX radiometer (#97-0015-02, UVP) coupled with a calibration point of 280 – 340 nm (#97-0016-04, UVP), and exposure time was calculated using the following formula: dose (mJ/cm²) = exposure time (s) × output intensity (mW/cm²). The UV output and exposure time were measured and calculated before each use of the lamps to account for changes in power output over time. During UV exposure, mice roamed freely in acrylic cages on a rotating platform, ensuring consistent and equal dorsal distribution of UV irradiation. The distance from the lamps to the mice is approximately 30 inches.

UV exposure studies (seven-dose protocol)

The KO and WT mice (females, 8~10-week-old) were randomly divided into 2 groups. The mice in group 1 were set up as a negative control (no UV) (n = 3 per genotype), and in group 2 the mice were irradiated with 180 mJ/cm² UV every other day for a total of seven UV doses (n = 5 per genotype) which encompassed two weeks (days 1, 3, 5, 7, 9, 11, and 13). Bi-folded epidermal thickness was measured using digital calipers on the dorsal skin immediately before irradiation. Three repeated measurements of epidermal thickness were taken along the middle region of the dorsal skin to obtain the mean epidermal thickness for each mouse. Skin injury was non-invasively evaluated by the measurement of skin biophysical parameters using the Multi-Probe Adapter (MPA) System (Courage & Khazaka Electronic GmbH, Germany) with the following probes: Mexameter MX18 and Tewameter TM 300. The Mexameter MX18 probe measures the coloration of the skin as a measure of erythema (reddening), and the Tewameter TM 300 probe measures transepidermal water loss (TEWL) as a measure of the barrier function of the skin. The skin parameters were taken on days 0, 4, 8, and 14. The sunburn lesion index, according to reported method (Lu et al., 2007), was calculated as the lesion area multiplied by the arbitrary intensity: 0, no lesion; 1, barely detectable red lesion; 2, moderate red lesion; and 3, bright red lesion. The results were expressed as dynamic changes in sunburn index on days 6, 8, 10 and 13. This experiment was repeated using a separate cohort of WT and KO mice (females), with group 1 as a negative control (no UV) (n = 5 per genotype), and in group 2 the mice were irradiated with 180 mJ/cm² UV every other day for a total of seven UV doses (n = 8 per genotype).

UV exposure studies (single-dose protocol)

Male KO or WT mice (8–10 weeks old) were divided randomly into two groups: Group 1 was a “no UV” control group, and the other was exposed to a single dose of 224 mJ/cm² UV. The mice were euthanized at different time points (6, 24 or 48 h) after the UV exposure, and their dorsal skin samples and plasma were collected for molecular and histological analysis. The experiments were repeated twice, and the data were combined in qRT-PCR analysis (n = 10 per time point per genotype per group).

RNA isolation and qPCR analysis

Total RNA was isolated from whole skin tissue using the RNeasy Mini kit (Qiagen). cDNA was obtained with the High-Capacity cDNA Reverse Transcriptase Kit (Thermo Fisher Scientific). The cDNA was amplified with Power SYBR^™ Green PCR Master Mix (Applied Biosystems); primer sequences are available upon request. Real-time PCR was performed on a CFX96 Real-time Thermal Cycler Detection System (Bio-Rad) and analyzed with the 2^−ΔΔCt method with beta-actin as the normalization control.

RNA sequencing analysis

Skin tissue from the mice were collected from the back and stored at −80 °C. The total RNA was extracted from the skin tissues using the RNeasy Mini Kit (Qiagen, Germantown, MD). RNA samples (n = 4) were delivered to Novogene Corporation Inc. (Sacramento, CA) for library preparation and RNA sequencing. Briefly, the quality of RNA samples was determined with a Nanodrop and Agilent Bioanalyzer, and then mRNA was purified from total RNA using poly-T oligo-attached magnetic beads. After fragmentation, the first strand cDNA was synthesized using random hexamer primers, followed by the second strand cDNA synthesis using dTTP for non-directional library. The library was checked with Qubit and real-time PCR for quantification and a bioanalyzer for size distribution detection. Quantified libraries were pooled and sequenced on Illumina platforms, according to effective library concentration and data amount. The raw data were processed and annotated based on bioinformatics databases by Novogene. For bioinformatics analyses, transcript abundance was quantified as gene-specific reads per million total mapped reads (TPM), with values floored at 1 TPM and log2-transformed for linear statistical model analysis as described below (Statistical analysis).

ELISA

Plasma levels of IL-6, IL-10 and TNF-α were measured using ELISA kit (Biolegend; IL-6: Cat. No. 431307, IL-10: 431417, and TNF-α: 430907). Briefly, 50 μl of plasma was added to wells pre-coated with specific primary antibodies. Then, detection antibody coupled to a peroxidase was added with a substrate solution, and the color developed proportionally to the levels of IL-6, IL-10 and TNF-α. Optical density (OD) was measured at 450 nm using a microplate reader. The standard concentrations for IL-6 range from 500 pg/mL to 7.8 pg/mL, while for IL-10 and TNF-α, they range from 1000 to 15.6 pg/mL used in the calculations.

Cytokine array analysis

Expression of cytokines and chemokines in the skin tissue lysates was analyzed using the Proteome Profiler Mouse Cytokine Array Kit, Panel A (ARY006, R&D Systems, Minneapolis, MN) according to the manufacturer’s instructions. Briefly, skin tissues were excised into small pieces, homogenized in PBS containing protease inhibitors and Triton X-100 (1 %) and frozen at −80 °C. After thawing, the samples were centrifuged at 10,000 × g for 5 min to remove cellular debris. After protein quantification, 300 μg of each sample was mixed with a cocktail of biotinylated detection antibodies and then incubated with each membrane, which was spotted in duplicate with 40 different cytokine antibodies overnight at 4 °C. The membranes were washed and incubated with streptavidin-horseradish peroxidase, followed by chemiluminescent detection.

Histology and immunohistochemistry (IHC) analysis

The formalin-fixed skin tissues were embedded in paraffin, which were de-paraffinized using xylene and ethanol. The de-paraffinized sections were stained with hematoxylin and eosin (H&E) to analyze histology. We captured two images from each mouse to determine the thickness of the skin’s epidermis on H&E slides and conducted three measurements on each image. We then averaged the measurements to obtain the final data for each mouse. Before conducting immunohistochemistry, antigen retrieval was performed by boiling the samples in an antigen retrieval Buffer (Abcam Cat no. ab93678) for 20 min. Different expressions of proteins in skin tissue were determined using the Vectastain Elite ABC–HRP kit (PK-6104). In brief, sections were incubated with bloxall endogenous enzyme blocking solution for 10 min to quench endogenous peroxidase activity and then rinsed three times (5 min each) with TBST (0.05 % Tween-20). Blocking solution was applied for 20 min; then, sections were incubated with diluted antibodies, anti-CXCL2 (Thermo, 1:50) and anti-COX-2 (Cell Signaling, 1:200) overnight at 4 °C in a humid chamber. Further processing was performed according to the instructions of the Detection System. Scoring the images from IHC analysis was carried outperformed using a semi-quantitative scoring system. To quantitate CXCL2 and COX-2, staining, an expression index was calculated by the extent of staining multiplied by the intensity of staining. For the extent of staining, the following system was used: 0, no staining, 1 is < 25 % of cells positive, 2 is >25 % and ≤50 % of cells positive, and 3 is >50 % of cells positive. For staining intensity, the following system was used: 0 is no staining, 1 is faint staining, 2 is moderate staining, and 3 is strong staining. Skin sections were evaluated using a Leica DM750 LED Biological Microscope.

Statistical analysis

Data is expressed as a mean ± standard error of the mean (SEM) unless stated otherwise. All plots were made using GraphPad Prism version 10 (GraphPad Software, Inc, La Jolla, CA), and statistical analysis was conducted using GraphPad Prism or NCSS 2007 (NCSS, LLC, Kaysville, UT). For data with more than two groups a General Linear Models ANOVA was utilized, and when mice were measured repeatedly a Repeated Measures ANOVA was used. A Tukey-Kramer post-hoc test was used to examine two-factor interactions. The specific tests are detailed in the text and figure legends. For all statistical analysis, means were indicated to be statistically different when p < 0.05.

For bioinformatic analysis of RNA-seq data, log2-TPM values were analyzed using a general linear model with factorial parameterization (Genotype: WT vs Adrb2KO × UV exposure: no vs yes) to identify all genes showing >2-fold differential expression as a function of Genotype, UV exposure, and their product-term interaction (to identify statistically significant differences in UV response by Genotype). Differentially expressed genes served as inputs into 1) TELiS bioinformatics analysis of core promoter transcription factor-binding motif ratios in up- vs. down-regulated genes (focusing in particular on pre-specified pathways involving inflammation and anti-inflammatory response, and immune and skin cell differentiation and activation), 2) Transcript Origin Analyses of epithelial vs mesenchymal differentiation profiles and major leukocyte subsets, as derived from previous transcriptome profiling reference studies of isolated cell populations, and 3) NetworkAnalyst mapping of gene regulatory relationships using TRRUST and SIGNOR 2.0 pathway annotation databases.

Results

Role of β2-AR in skin damage and sunburn induced by multiple UV exposures

We exposed mice to a standard sunburn model involving seven repeated UV radiations (180 mJ/cm², every other day) over the course of two weeks [28] (Fig. 1A). Compared to non-UV-exposed animals, skin edema evaluated by bi-fold thickness increased over time in both UV-exposed WT and β2-AR KO mice. Skin edema reached peak levels at day 7, but the magnitude of this increase was less pronounced for β2-AR KO mice; KO mice showed significantly less skin thickening on days 7 to 11 (Fig. 1B). Similarly, UV-exposed KO animals showed significantly less Sunburn Index (Fig. 1C) compared to UV-exposed WT mice at days 6 and 8, with this difference resolving by day 10 as all animals returned to the baseline. Erythema Index (Fig. 1D) demonstrated a similar pattern, with UV-exposed KO animals showing significantly less erythema index at the peak day 8. β2-AR KOs did not differ significantly from WT animals in trans-epidermal water loss (TEWL), a measure of skin barrier function (Fig. 1E), although the KOs showed a similar trend of reduced TWEL compared to WTs. Unlike other parameters, TEWL of both WTs and KOs did not return to baseline at day 14, suggesting that although sunburn and erythema was decreasing, skin damage persisted, and that the chronic damage to the barrier function was not dependent on the β2-AR. Representative photos taken on day 10 are shown in Fig. 1F, with a visible increase of reddening in the WT but substantially less sunburn in the KO mouse. These data indicate that repeated UV radiation damages the skin of both WT and KO mice, but the β2-AR KO mice show substantially less sunburn, cutaneous inflammation, and barrier function disruption.

Fig. 1.

Effects of seven-dose UV exposure on bi-fold skin thickness, sunburn index and skin parameters in β2-AR KO and WT mice. (A) Outline of experimental design: lightning bolts indicate administration of 180 mJ/cm² of UV to subgroups of the mice. (B) The bi-fold thickness of the skin was measured before and after each UV exposure. (C) Sunburn index. (D) Erythema index. (E) Transepidermal water loss (TEWL). In (b) to (e), days of UV radiation are indicated by the red arrows on the x-axis. (F) Photograph of representative mice taken at day 10. The data are expressed as mean ± SEM (n = 3 for controls; n = 5 for UV). A two-factor ANOVA with Tukey-Kramer Multiple-Comparison Test was run to determine individual differences. The asterisk (*) indicates individual differences with p < 0.05 between UV-exposed WT and KO mice at the given time points. A Repeated Measures two-way ANOVA with the Geisser-Greenhouse correction was used to determine the overall significance and the results are shown in each plot.

Role of β2-AR in cutaneous inflammatory response to single-dose UV radiation

Six hours after a single UV exposure (224 mJ/cm²) (Fig. 2A), quantitative RT-PCR (qRT-PCR) documented significant up-regulation of multiple inflammatory gene transcripts including Il6 (IL-6), Il1b (IL-1β), Ptgs2 (COX-2), Il10 (IL-10), and Tnf (TNF-α) in both WT and β2-AR KO animals. Compared to the WTs, β2-AR KOs showed significantly less induction of Il6 at this time point (6-hr) (Fig. 2B), but did not differ significantly in the level of other inflammatory gene transcripts. The pro-inflammatory cytokine Il6 can be induced by catecholaminergic signaling through the β2-AR based on previous research on inflammation, tissue injury and tumor growth [29–33]. Previous studies have also shown that the β₁/β₂-blocker carvedilol reduced UV-induced Il6 mRNA response [27]. The only other inflammatory gene transcript that showed a significant difference in β2-AR KO vs WT was Il12b (IL-12b), which showed significantly higher level in the KO skin at the baseline level compared to the WT skin (Fig. 2B). The Il12b gene encodes the p40 subunit of IL-12 and IL-23. IL-12 is an immunomodulatory cytokine that can reduce UV-induced DNA damage by inducing DNA repair [34,35], and the higher basal level of Il12b expression in the skin of the KO animals may contribute to their attenuated inflammatory response to UV exposure. Previous research has found that β2-adrenergic signaling can suppress Il12b expression [36–38], consistent with our observation that the β2-AR KO skin shows higher Il12b expression than WT. However, the basis for the UV-induced decline in IL12b expression in β2-AR KO animals is unclear. Consistent with the qRT-PCR data, the KO mice showed attenuated UV-induced IL-6 levels in the plasma six hours after the UV exposure (Fig. 2C). ELISA analysis was also conducted for TNF-α and IL-10; however, these cytokines were undetectable or very low in the mouse plasma.

Fig. 2.

Effect of β2-AR gene knockout and single-dose UV on gene expression time course analyzed by qRT-PCR. (A) Experimental design: mice were exposed to a single dose of 224 mJ/cm² UV radiation and were euthanized at 6, 24, and 48 h after UV exposure. (B) Total RNAs were isolated from skin tissues of KO and WT mice irradiated with UV, and qPCR was conducted for indicated genes (n = 10). (C) IL-6 levels were determined by ELISA in mouse plasma (6-hr post-UV). Data are presented as the mean ± SEM. An one-way ANOVA followed by a Tukey-Kramer multiple-comparison was used to assess statistical differences. **: p < 0.01; ***: p < 0.001; ****: p < 0.0001. (D) The mRNA expression of β1-, β2-, and β3-ARs from control and irradiated WT and KO mouse skin were assessed via qRT-PCR (n = 5). Data were normalized by the expression of β-actin and then normalized by average level of the WT controls. The data are expressed as mean ± SEM. A two-factor ANOVA with Tukey-Kramer Multiple-Comparison Test was run to determine individual differences. An asterisk (*) indicates individual differences between WT and KO (p < 0.05) at the given time points. An * above the bar indicates significant differences comparing UV-treated samples and baseline control samples.

Effect of single-dose UV on β1-, β2-, and β3-AR expression

Because β-AR expression is transcriptionally dynamic and can respond to stimuli, we examined how the skin expression of β1-, β2-, and β3-ARs might potentially change in response to a single dose of UV radiation. At 6 h or later after a single UV exposure, WT mice showed a modest increase in mRNA for both β1- (Adrb1) and β3-AR (Adrb3), but a more pronounced increase in β2-AR (Adrb2) (Fig. 2D). However, in the β2-AR KO mice lacking Adrb2 mRNA, UV exposure induced significantly greater up-regulation of Adrb1 and Adrb3 mRNA at 6 h compared to the WTs (Fig. 2D). The greater Adrb1 and Adrb3 responses to UV in β2-AR KO animals may reflect genetic compensation for the deleted gene by other genes in the same network [39]. In the KO mice, both β1- and β3-AR mRNA returned to the basal levels within 24 h. Collectively, these data suggest that the UV can induce dynamic alterations in β-AR homeostasis across all 3 receptor subtypes, implying that non-selective β-blockers might potentially be more effective than β2-selective antagonists in treating UV-induced skin reactions. However, despite evidence of genetic compensation by β1- and β3-ARs, the data shown in Fig. 1 indicate that β2-AR plays a dominant role in the cutaneous reaction to UV.

Role of β2-AR in cytokine and chemokine response to single-dose UV radiation

To quantify the effects of UV on skin inflammation at the protein level, forty cytokines and chemokines were assayed using the Proteome Profiler Mouse Cytokine Array, 24 hrs after a single UV exposure, using pooled skin tissue lysates from three randomly selected mice in each of the four treatment groups. Fig. 3A shows cytokine protein levels for WT and β2-AR KO mice under the control condition (Ctr) and 24 h after UV exposure (WT UV and KO UV). Fig. 3A highlights the assay spots showing >2-fold difference between groups and intensity >100,000 in at least one group, and Fig. 3B displays results as a heatmap. β2-AR KO mice did not much differ from WT at baseline, but they showed much weaker UV induction of multiple cytokines/chemokines, which have been reported involved in UV responses, including IL-1ra [40], Cxcl1 [41], Mcp-1/Ccl2 [42], Tnf-ɑ [43], Ccl3/Mip-1a, and Cxcl2/Mip-2 [44]. Two chemokines were detectable only in the skin of UV-exposed WT animals: Ccl3/Mip-1a and Cxcl2/Mip-2, both of which have been previously shown to be induced in mouse skin by UV exposure, and mediate leukocyte infiltration in inflamed tissues [44]. The absence of these molecules’ induction by UV exposure in β2-AR KO mice confirms the key role of β2-AR signaling in mediating inflammatory responses to UV exposure.

Fig. 3.

Effects of β2-AR gene deletion and single-dose UV radiation on expression of cytokines and chemokines in WT and KO mouse skin. (A) Skin tissue lysates pooled from three randomly selected skin samples in each of the four groups (WT/Ctr, KO/Ctr, WT/UV and KO/UV) were run on a Mouse Cytokine Array. Based on quantified expression levels, proteins that show differential expression between WT/Ctr vs. KO/Ctr, WT/UV vs. WT/Ctr, KO/UV vs. KO/Ctr, and WT/UV vs. KO/UV skin samples (at least two-fold difference), and with intensity level > 100,000, are highlighted with rectangles. (B) Heat map analysis of chemokines/cytokines that showed differential expression between WT/Ctr vs. KO/Ctr, WT/UV vs. WT/Ctr, KO/UV vs. KO/Ctr, and WT/UV vs. KO/UV skin samples 24-h after UV exposure. Values for each row were normalized by z-scores.

Role of β₂-ARs in skin histological responses to UV radiation

Histological analysis of the skin showed modest increases in epidermal thickness in WT mice and a non-significant trend toward less epidermal expansion in β2-AR KO mice at 6- and 24-hr time points (Fig. 4A), while at 48 h post-exposure, β2-AR KO mice showed significantly less epidermal thickness compared to WT (3.6 ± 0.3 μm, vs. 5.6 ± 2.1 μm, p < 0.01) (Fig. 4A). At 48 h after UV radiation, WT skin demonstrated visible apoptotic cells, whereas KO skin showed significantly less apoptosis (Fig. 4B).

Fig. 4.

Histological analysis of skin tissues isolated from β2-AR KO and WT mice after exposed to a single dose UV radiation. (A) Representative images of H&E staining of the skin sections (20x magnification) at various time points. The mice were exposed to single dose UV radiation and samples were collected at 0-hr, 6-hr, 24-hr and 48-hr after the UV exposure. To determine the thickness of the skin’s epidermis on H&E slides, three measurements were conducted for each mouse, and the data were averaged the measurements to obtain the final data per mouse. Data are presented as the mean ± SEM (n = 5). (B) Representative images of H&E staining of skin sections at 40x magnification to visualize and quantify apoptotic cells due to sunburn 48 h after UV radiation (yellow arrows). The graph showed the number of apoptotic cells per field for each mouse. (C) Immunohistochemical (IHC) analysis of CXCL2 expression in skin tissues of WT and β2-AR KO mice 24-hour after exposed to a single dose UV radiation. Representative images of IHC staining of skin sections for CXCL2 at 40x magnification. The mice were exposed to single dose UV radiation and samples were collected at 0-h or 24-h after UV exposure. (D) IHC analysis of COX-2 expression in skin tissues of WT and β2-AR KO mice 24-hour after exposed to a single dose UV radiation. Representative images of IHC staining of skin sections for COX-2 at 40x magnification. The graph showed the expression index of CXCL2 and COX-2 in skin samples obtained from WT control, WT UV, KO control, and KO UV (n = 5); data are presented as the mean ± SEM. An one-way ANOVA followed by a Tukey-Kramer multiple-comparison was used to assess statistical differences. *: p < 0.05; **: p < 0.01; ***: p < 0.001; ****: p < 0.0001.

To confirm the role of β2-AR in skin inflammation, the expression level of Cxcl2/MIP-2 was assessed by immunohistochemistry (IHC) in mouse skin 24-hour after UV exposure (Fig. 5A). Both WT and KO mice showed significant increases in epidermal Cxcl2, but this effect was attenuated in the β2-AR KO animals relative to WT (Fig. 4C). Similarly, UV exposure increased Cox-2 (Ptgs2) expression in epidermal and dermal layers, but this effect was again significantly attenuated in β2-AR KO animals (Fig. 4D). Given the key role of Cox-2 in skin inflammation and carcinogenesis [45–47], these data further confirm reduced skin inflammatory response to UV in β2-AR KO mice.

Fig. 5.

RNA Seq gene expression profiles of skin tissues of WT and β2-AR KO mice 6 h after exposure to a single dose of UV radiation. (A) A clustering heat map was generated using the log2(FPKM+1) value. Red color indicates genes with high expression levels, and blue color indicates genes with low expression levels. (B) Principal component analysis of variation within the data sets. PC1 represents the most variation in the data, capturing the distinction between UV-exposed vs unexposed tissues. PC2 represents the second most variation in the data beyond the effects of UV exposure, and captures the distinction between KO and WT animals on exposure to UV. Note that KO animals show attenuation of UV-induced transcriptional alterations on PC1. (C) The expression values (FPKM) for select genes within the inflammatory cluster are plotted for four groups (WT control, WT UV, KO control, and KO UV) (n = 4); data are presented as the mean ± SEM. A student t-test was used to assess statistical differences. *: p < 0.05.

Distinct RNA-seq profiles between WT and KO skin after UV radiation

Given the multiple transcriptional and proteomic effects of β2-AR KO observed above, we sought to identify potential gene regulatory pathways and cellular mechanisms that might structure these responses to UV using genome-wide transcriptional profiling to drive bioinformatic analyses of transcription factor activity, immune cell activity, and tissue epithelial-mesenchymal differentiation dynamics. Genome-wide RNA-seq analysis was performed on the skin from β2-AR KO mice vs. WT mice at 6 h post-UV exposure (the time point of peak mRNA differences detected in Fig. 2). Initial global analyses showed a broad similarity in skin genome-wide transcriptional profiles for WT and β2-AR KO animals under basal conditions (Fig. 5A, lower left quadrant). Principal component analysis (PCA) of skin transcriptome profiles confirmed these observations and found that the skin transcriptome of β2-AR KO broadly resembles that of WT mice under basal conditions (i.e., the control samples fell within a single overlapping cluster at the low end of PC1) (Fig. 5B). The primary axis of skin transcriptome differentiation (PC Axis 1) distinguished UV-exposed from control skin (Fig. 5B) and thus confirms a substantial effect of UV exposure on gene regulation within the skin. The second major axis of skin transcriptome differentiation (PC Axis 2) distinguished the UV-exposed skin transcriptomes of β2-AR KO mice vs. WT mice (i.e., distinct genotype clusters at the high end of PC1) (Fig. 5B). β2-AR KO mice showed substantially lower values on PC1 compared to WT mice, indicating a broad attenuation of the overall transcriptional reaction to UV in KO mice. Hierarchical clustering of transcriptome profiles (Fig. 5C) also shows a broad similarity between β2-AR KO and WT skin transcriptomes under basal conditions (i.e., broadly similar heatmaps in the first 2 columns). However, one small gene cluster present at the bottom of the heatmap (column 2 vs 1) does indicate a limited set of genes that are systematically upregulated in β2 KO skin relative to WT skin at baseline (e.g., as shown for Il12b above; specific genes in this cluster are listed in Supplementary Table 1). In contrast to these broad transcriptomic similarities at baseline, β2-AR KO skin transcriptomes differed much more substantially from WT skin transcriptomes at 6 h post-UV (the last 2 columns of the heatmap). Heatmap columns 3 and 4 show large numbers of gene transcripts that are induced by UV exposure in WT skin (column 3) and β2-AR KO skin (column 4), but the specific genes induced by UV differ substantially by genotype (i.e., distinct heat patterns in the lower right quadrant of the heatmap). Particularly notable is a large cluster of genes that were strongly upregulated by UV in WT animals but to a much lesser degree in the β2-AR KO (marked by the purple bracket adjacent to the heatmap). This cluster contains many inflammatory genes that showed UV-induced upregulation in the qRT-PCR analysis (Fig. 2) and cytokine array analysis (Fig. 3), including Cxcl1, Cxcl2, Il6 and Ccl3 (Fig. 5C; see Supplementary Table 2 for full list of genes in this cluster).

Role of β2-ARs in gene regulatory response to UV

Following initial descriptive analyses (Fig. 5), we used genome-wide RNA-seq data as input into higher-order bioinformatic pathway analyses to determine which transcription control pathways and immune cell subsets might play a role in structuring the basic cutaneous response to UV exposure, which might structure the effects of β2-AR KO on skin under basal conditions (absent UV), and which might structure how the UV response differs in WT vs β2-AR KO mice (Fig. 6).

Fig. 6.

Gene regulation in skin from WT and β2-AR KO mice 6 h after a single dose of UV radiation. (A) Genes showing > 2-fold differential expression across experimental conditions served as input into TELiS promoter sequence-based bioinformatics analysis of transcription factor activity for pre-specified pathways involved in inflammation, cell differentiation and immune response. Data represent log2-fold difference in transcription factor-binding motif density in up- vs down-regulated genes for each condition (compared to control); red = up-regulated, blue = down-regulated. (B) Effect of experimental conditions on expression of pre-specified sets of mesenchymal and epithelial characteristic genes (and their difference, representing epithelial-mesenchymal transition). Data represent mean fold-difference from control condition; bars represent SEM (n = 4). (C) Immune cell Transcript Origin Analysis of genes showing >2-fold differential expression across experimental conditions. Data represent mean cell type diagnosticity score, red = up-regulated, blue = down-regulated. (D) NetworkAnalyst regulatory structure of gene transcripts down-regulated >2-fold in UV-exposed skin from Adrb2KO mice vs WT mice (edges represent regulatory associations as inferred from the TRRUST and SIGNOR 2.0 transcription control pathway databases).

In analysis of 2505 gene transcripts that showed > 2-fold difference in average expression in UV-exposed skin vs control skin among WT animals (1170 upregulated, 1335 downregulated; Supplemental Dataset 1, Column D), TELiS promoter-based bioinformatic analyses identified significant increases in activity of the proinflammatory NF-κB and AP-1 signaling pathways, as well as the CREB family of transcription factors (likely due to prostaglandin signaling through the cAMP-linked prostaglandin E receptor), the innate immune response Myeloid Zinc Finger (MZF) and CEBP-β pathways, and the glucocorticoid receptor (GR; likely as a physiological negative feedback response to systemic inflammation) (Fig. 6A). In analysis of pre-specified epithelial and mesenchymal cell differentiation profiles, UV also strongly activated the mesenchymal differentiation program (a standard part of injury response/healing) while repressing the epithelial differentiation program, leading to a net epithelial-mesenchymal transition in UV-exposed skin (Fig. 6B). Parallel Transcript Origin Analyses implicated classical and non-classical monocytes, and classical dendritic cells as primary immune cell mediators of UV-induced transcriptional differences (Fig. 6C). Genes downregulated by UV derived predominately from neutrophils and plasmacytoid DCs.

In analysis of 280 gene transcripts that showed > 2-fold difference in average expression in skin from β2-AR KO vs WT animals (UV-unexposed controls; 187 upregulated, 93 downregulated; Supplemental Dataset 1, Column C), TELiS promoter-based bioinformatic analyses identified significant reductions in NF-κB activity, increases in IRF activity and glucocorticoid signaling, and reductions in myeloid cell / macrophage-related signaling pathways such as CEBP-β and EGR family transcription factors (Fig. 6A). We also observed a paradoxical effect in increased CREB activity, which may be due to the well-known post-receptor sensitization that is also seen after long term pharmacologic β-AR blockade [48]. β2-AR KO also reduces mesenchymal bias (Fig. 6B) (consistent with previous research) [49]. Transcript Origin Analyses implicated monocytes in β2-AR KO-associated transcriptional up-regulation and neutrophils and dendritic cells in transcriptional down-regulation (Fig. 6C).

Finally, we examined the differential skin transcriptional responses to UV exposure vs control conditions in β2-AR KO vs WT mice (i.e., UV × Genotype interaction). In analysis of 1777 gene transcripts that showed > 2-fold difference in average UV response (i.e., UV - control) in skin from β2-AR KO vs WT animals (936 upregulated, 841 downregulated; Supplemental Dataset 1, Column E), TELiS promoter-based bioinformatic analyses found β2-AR KO skin to show a significant relative reduction in UV-induced myeloid cell activation pathways (MZF, CEBP-β) and glucocorticoid signaling (GR), as well as non-significant reductions in pro-inflammatory signaling pathways (NF-κB, EGR1–3) and immune regulatory pathways (e.g., ISRE, IRF2) (Fig. 6A). These results were accompanied by downregulation of genes expressed by classical and non-classical monocytes and classical dendritic cells, and up-regulated expression of genes derived from neutrophils and plasmacytoid dendritic cells (i.e., reversing the pattern of cell-mediated transcriptional responses to UV exposure in WT animals)(Fig. 6C). β2-AR KO animals also showed down-regulation of a broad network of genes involved in cellular bioenergetics (centered around the Irs insulin receptor substrate), catecholamine biosynthesis (centered around tyrosine hydroxylase, Th), angiogenesis (centered around the Vegf receptor, Flt1), MAP kinase and calmodulin signaling (centered around Mapk12, Camk2a and Camk2g), and other transcription control pathways involved in cell differentiation (Mef2c, Myog, Myod1) (Fig. 6D). As a consequence of these gene-regulatory and cellular alterations, β2-AR KO animals showed a marked reduction in UV-induced mesenchymal differentiation, up-regulated epithelial differentiation, and reversal of the epithelial-mesenchymal transition observed in UV-exposed skin in WT mice (Fig. 6B). Collectively, these results point to reduced myeloid cell infiltration, activation, and mesenchymal wound reaction as key cellular and molecular pathways through which β2-AR KO inhibits UV-induced skin damage and related tissue repair/wound healing processes.

Discussion

To study the role of β2-AR in UV-induced skin damage, we established a mouse model of whole-body β2-AR (Adrb2) gene deletion on the SKH-1 genetic background. Results of this study indicate that β2-ARs play a key role in mediating the effects of UV radiation on skin damage and inflammation, because the β2-AR KO mice exhibit significantly attenuated sunburn and reduced cutaneous inflammation in comparison with the WT mice. These effects are associated with significant reductions in skin expression of pro-inflammatory cytokines and chemokines (at both the mRNA and protein levels) as well as marked alterations in the activity of multiple gene-regulatory pathways involved in inflammation, innate immune regulation, and the infiltration and activation of myeloid lineage immune cells (i.e., neutrophils, monocytes/macrophages, and dendritic cells), as well as significant reductions in mesenchymal cellular differentiation in response to UV radiation. The widespread and marked effects of β2-AR deletion in this UV exposure paradigm are all the more remarkable because these studies utilized no exogenous activator of adrenergic signaling (i.e., there was no experimentally imposed behavioral stress or adrenergic pharmacology; UV radiation was the sole source of whatever β2-AR signaling impacted skin damage and inflammation in these studies). Based on these results, we speculate that UV radiation may potentially elicit local production of catecholamines that subsequently activate β2-ARs and contribute to inflammation as documented in these studies. In addition to highlighting an unanticipated catecholaminergic signaling dimension to UV exposure, the present results raise the possibility that other sources of catecholamine activity, such as behavioral stress, exercise, or adrenergic medications might potentially amplify the effects of UV on β2-AR-mediated inflammation, and that pharmacologic β-AR-antagonists might potentially inhibit the effects of UV on inflammation and skin injury (so long as the agents are either non-selective or selective inhibitors of β2-ARs).

The physiological source of UV-induced β2-AR signaling was not defined in these analyses and remains an important topic for future research. Catecholamines such as epinephrine may be synthesized locally in the skin after UV radiation, as previous evidence demonstrates that keratinocytes have the capacity to produce epinephrine in the context of a wound [50]. Since the animals were not under behavioral stress, locally synthesized epinephrine that binds to β2-ARs could promote chronic inflammation in the skin. In addition, the autocrine production of catecholamines by keratinocytes can increase the density of β2-AR on keratinocytes [51]. Moreover, previous research has shown that catecholamines can also be synthesized by other skin-resident immune cell types, such as macrophages and lymphocytes [52,53]. Future research using single-cell analyses of skin tissue should seek to determine the cellular sources of catecholamine signaling through β2-ARs in the skin following UV exposure.

The present results are subject to several limitations. In our global β2-AR KO mice, signaling through upregulated β1- and β3-ARs may potentially compensate for the loss of β2-AR signaling. Therefore, although this study is focused on β2-AR, it does not rule out an indirect contribution of altered β1- and β3-AR signaling. Furthermore, in a global knockout mouse model such as this one, whole-body β2-AR gene deletion may induce physiological changes, and genetic compensation is more likely to occur in such models. To address these challenges, future studies should evaluate the role of β2-AR in cell type-specific (e.g., skin-specific) KO mouse models. Another limitation of the present studies is that the protein assays, qRT-PCR assays, and RNA-seq analyses were all conducted on bulk tissue samples, and future analyses would benefit from single-cell analysis approaches (e.g., single-cell RNA-seq, spatial expression profiling, flow cytometry) to more clearly resolve the cellular mechanisms involved. The present study used UV lamps that emit a mixture of UVB, UVA, UVC and visible light, which is an artificial experimental system not exactly like the sunlight. Therefore, the current study cannot evaluate the effects of specific wavelengths. Future studies can investigate the effects of specific components of sunlight, including UVA, UVB, visible light and infrared on skin and their interaction with β2-ARs. Although these limitations are important and inform future studies, the present data nevertheless extend previous studies in identifying β2-ARs as a key molecular signaling pathway involved in UV-induced cutaneous damage and inflammation, and a potential target for therapeutic manipulation in the context of cutaneous diseases such as skin cancer.

Supplementary Material

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.jpap.2025.100265.

Abbreviations used:

β2-AR

β2-adrenergic receptor

KO

knockout

TEWL

trans-epidermal water loss

UV

ultraviolet

qRT-PCR

quantitative RT-PCR

WT

wild-type.

Data availability

Data will be made available on request.