ALA-PDT Amplies Intense Inammatory Response in the Treatment of Acne Vulgaris via CXCL8

Background: Acne vulgaris is a chronic inammatory cutaneous disease. 5-Aminolaevulinic acid photodynamic therapy (ALA-PDT) is a novel and effective therapy for severe acne vulgaris. However, the specic mechanism of ALA-PDT for acne still remain unclear. Here, we investigate the possible mechanism of intense inammatory response of ALA-PDT for acne vulgaris. Results: It appeared that ALA-PDT suppresses proliferation and lipid secretion of primary human sebocytes. And ALA-PDT could up-regulate the expression of CXCL8 in vivo and in vitro, amplifying inammatory response by recruiting T cells, B cells, neutrophils and macrophages. We also found that ALA-PDT elevated the expression of CXCL8 via p38 pathway. SB203580, a p38 pathway inhibitor, decreased the expression of CXCL8 after ALA-PDT in sebocytes. Conclusion: ALA-PDT amplies intense inammatory response in the treatment of acne vulgaris via CXCL8. Our data deciphers the mechanism of intense inammatory response after ALA-PDT for acne vulgaris.


Introduction
Acne vulgaris is a chronic in ammatory cutaneous disease involving pilosebaceous follicles, and it affects almost 85% of the population.(1) Excessive sebum production and complex in ammatory reaction are two of the key factors of the pathogenesis of acne vulgaris. The dysregulation of innate immune signaling is a new sight for the physiopathology of acne.(2, 3) Topical or oral antibiotics and isotretinoin are routinely used to treat acne. (1,4,5) However, antibiotic resistance is increasing, more than 50% of Propionibacterium acnes strains are resistant to topical macrolides, affecting its effectiveness, as well as the system adverse effects of isotretinoin involving mucocutaneous, musculoskeletal, and ophthalmic systems. (1,6) And long course of the traditional treatments also limits their application. (7) 5-Aminolaevulinic acid photodynamic therapy (ALA-PDT) is a relatively novel therapy for acne vulgaris.
After applied to the lesion for a period of time, ALA preferentially accumulated in the pilosebaceous units, then light of an appropriate wavelength activates the process of reactive oxygen species (ROS) generating, promoting the photodynamic reaction. (8) ALA-PDT has been involved in many kinds of guidelines of acne vulgaris for the advantages of positive curative effects, tissue targeting, minimal invasion, repeatability, and no system side effects like oral agents. (1,9) (10) However, the speci c mechanism of the ALA-PDT for acne still remain undetermined.
Clinically, "intense in ammatory response", such as erythema and pustule, occurs in the early phase after ALA-PDT. (5) We also observed that there was a positive correlation between the degree of in ammatory reaction and the e cacy. (11,12) However, the related research is rarely reported on the ampli cation of in ammation. The correlative mechanism is being unknown. In our study, we found that interleukin-6 (IL-6), tumor necrosis factor α (TNF-α), interleukin-1β (IL-1β) and C-X-C motif chemokine (CXCL8) were increased signi cantly after ALA-PDT for acne vulgaris. We focused the role of CXCL8, which is a pro-in ammatory chemokine, secreted by different cell types including blood monocytes, alveolar macrophages, broblasts, endothelial cells, and epithelial cells. (13) Various cytokines (IL-1, IL-6 and TNFα) and ROS could stimulate the up-regulation of CXCL8. (13,14) However, the mechanism of CXCL8 in amplifying acute in ammatory response after ALA-PDT on acne vulgaris has not been fully elucidated.
Here, we found the optimum parameters of ALA-PDT on primary sebocytes and show that ALA-PDT could inhibit the proliferation and lipid secretion of sebocytes. Furthermore, we found that ALA-PDT regulates the expression of CXCL8 via p38 pathway, recruiting a variety of immune cells to amplify the in ammatory response. Our data deciphers the mechanism of intense in ammatory response after ALA-PDT for acne vulgaris.
We observed that ALA-PDT had a direct effect on sebaceous gland in acne vulgaris in clinical. To verify this, we detected if ALA accumulated in targeting cells and tissues. As shown in Fig. 1A, red uorescence of PpIX in sebocytes induced by ALA was observed by uorescence microscope. As the uorescence intensity of PpIX was related to the effect of ALA-PDT, we investigated the optimum parameters of ALA.
Our results showed that uorescence intensity of PpIX was strongest in sebocytes when incubated with 0.5 mM ALA for 6 h (Fig. 1B, 1C). But if incubated for 6 h, the proliferation of sebocytes would be decreased signi cantly. No signi cant differences were detected when incubated for 4 h (Fig. 1D). So the optimum parameter of ALA was 0.5 mM for 4 h on sebocytes.
To further con rm that ALA was accumulated in sebaceous gland directly, 8% ALA was applied on the right sebaceous gland on back of golden hamster for 3 h. As shown in Fig. 1E, red uorescence of PpIX was observed around both right and left sebaceous gland. Freezing section of the tissue also showed that PpIX was accumulated in follicle sebaceous gland units directly (Fig. 1F). Taken-together, ALA could speci cally accumulate in sebaceous gland.

ALA-PDT suppresses proliferation and lipid secretion of sebocytes
To substantiate the inhibitory effect we evaluated the proliferation, apoptosis and lipid secretion of sebocytes after ALA-PDT. After 0.5 mM ALA-PDT (10 J/cm 2 ), ROS accumulates in sebocytes ( Fig. 2A). According to the previous study, the accumulation of ROS was related to the effect of ALA-PDT.
Next we used CCK-8 assay to evaluate the cell viability. As the energy intensity of 0.5 mM ALA-PDT increased from 0 to 40 J/cm 2 , cell viability of sebocytes was reduced from 100 ± 8.99% to 20.18 ± 8.09% and IC50 was at 10 J/cm 2 (Fig. 2B). The apoptotic population in sebocytes was evaluated by ow cytometry. As expected, 0.5 mM ALA-PDT (10 J/cm 2 ) signi cantly decreased the relative number of live sebocytes and concurrently increased the relative number of apoptotic sebocytes (Fig. 2C, 2D).
Fluorescence microscope was used to detect the secretion of lipids in sebocytes treated by ALA, Red light and ALA-PDT respectively. (Fig. 2E). Lipid in sebocytes was found to be signi cantly reduced 24 h after ALA-PDT. The above results indicated that ALA-PDT inhibited the ability of lipid secretion and proliferation of sebocytes.
Clinically, intense in ammatory reaction occurs in the early phase after ALA-PDT. Therefore, we tested whether in ammatory cytokines changed after ALA-PDT. Sebocytes treated by ALA, Red light and ALA-PDT were collected 24 h later. The results showed that the expression of CXCL8, IL-6, TNF-α and IL-1β was signi cantly increased after ALA-PDT (Fig. 3A). In view of the effect of the recruitment of CXCL8 on immune cells, our research focused on the regulatory effect of ALA-PDT on CXCL8.
Further experiment substantiated that the expression of CXCL8 was elevated by ALA-PDT in sebaceous gland of golden hamster. After treated by ALA-PDT, tissues were taken from treated sites at 1 h, 3 h, 6 h, 12 h, 24 h after ALA-PDT and untreated tissues were used for comparisons. We found that positive staining for CXCL8 started to increase at 3 h after ALA-PDT around sebaceous gland. These ndings illustrated ALA-PDT upregulated the expression of CXCL8 on sebocytes and sebaceous gland.

CXCL8 ampli es intense in ammatory response after ALA-PDT in sebaceous gland.
CXCL8 is a kind of chemokines, recruiting in ammatory cells. To further investigate the role of CXCL8, we detected the expression of CD3, CD11b, CD19 and CD68 in sebaceous gland after ALA-PDT by immunohistochemistry. After treated by ALA-PDT, tissues were taken from treated sites at 1 h, 3 h, 6 h, 12 h, 24 h after ALA-PDT and untreated tissues were used for comparisons. In HE staining, in ammatory cells started to collect towards sebaceous gland at 1 h after ALA-PDT. The results showed that positive staining for CD3(+) T cells and CD11b (+) neutrophils started to increase gradually at 1 h after ALA-PDT around sebaceous gland. Expression of CD19(+) B cells and CD68(+) macrophages slightly increased from 3 h after ALA-PDT (Fig. 4). Collectively, the results implied that CXCL8 could recruited several in ammatory cytokines, which were mainly T cells with a few neutrophils, B cells and macrophages.
The activation of CXCL8 gene promoter, trans-activation by JNK pathways and stabilization by p38 pathway are involved in the up-regulation of CXCL8 expression. To further study the mechanism of ALA-PDT's regulation of CXCL8 in sebocytes, we analyzed the expression of p38, Erk1/2, and JNK pathways by western blot. Sebocytes were collected for western blotting after ALA-PDT immediately. The results showed that ALA-PDT upregulated p-p38 and decreased expression of p-Erk1/2 and p-JNK signi cantly ( Fig. 5A, 5B).
Next, to verify whether suppression of p38 pathway could block the effect of ALA-PDT in sebocytes, p38 inhibitor SB203580 was employed. Cells were collected for further assays at 24 h post-treated with ALA-PDT and SB203580 (10 µM). The results showed that pretreated with SB203580 attenuated the upregulated CXCL8 mRNA expression by ALA-PDT in sebocytes (Fig. 5C). The results indicated that ALA-PDT induce CXCL8 production through p38 pathway.

Discussion
ALA-PDT is a common and effective treatment for severe acne vulgaris.(9) Key pathogenic factors involve hyperkeratinisation, disturbance of microbiota, sebum production and complex in ammatory mechanisms. (15) There are many research developments on the effect of ALA-PDT on keratinocytes and Propionibacterium acnes (P. acnes). But the effect of ALA-PDT on sebacytes received limited attention. In clinic, there is a speci c absorption of ALA on sebaceous gland, and after ALA-PDT, a temporary and serious in ammatory reaction happens around the target lesions to achieve the expected therapeutic e cacy. (9,16) A systematic review of randomized controlled trials to evaluate the e cacy and safety of PDT on acne showed that the ALA-PDT could signi cantly decrease the sebum secretion. Xiang et al have demonstrated that sebaceous glands atrophy and reduction of sebum secretion after ALA-PDT may be caused by the suppression of lipogenesis and cell growth in SZ95 cells by mTOR signaling pathway. (17) However, there are few studies about the effect of ALA-PDT on primary sebocytes in vivo. In this study, we used primary human sebocytes and animal models to verify the effect of ALA at different points in time and to further investigate the effect of ALA-PDT on sebaceous gland.
We provided the optimum parameters of ALA-PDT in vitro and in vivo. After incubated with 0.5 mM ALA for 4 h, red uorescence was observed in sebocytes, then after ALA-PDT (10J/cm 2 ), the lipid secretion was inhibited and sebocytes were induced to apoptosis (Fig. 1, 2). Sebaceous gland on the back of gold hamster could absorb ALA speci cally. When ALA was applied on one side, the other side of sebaceous gland could also absorb and transform to PpIX (Fig. 1).
Clinically, the obvious in ammatory response aggravated temporarily after ALA-PDT. And there is a positive relationship between in ammatory reaction and effectiveness of ALA-PDT.(9) However, the mechanism of complex in ammatory reaction after ALA-PDT is not clear. Multiple data have demonstrated ALA-PDT can enhance immunogenicity around target lesion, providing a promising strategy for inducing a systemic immune response. (18) The induction of CXCL8 by ALA-PDT was further supported by our ndings in vitro and in vivo. As shown in Fig. 3, expression levels of CXCL8 in sebocytes and sebaceous gland of golden hamster were increased after ALA-PDT. According to immunohistochemical results, we observed positive expression of T cells, neutrophils, B cells and macrophages at different points in time, nding a recruitment path of them (Fig. 4). This process changed the microenvironment in pilosebaceous unit, accelerating in ammatory process in order to eliminate the lesions. So up-regulation of CXCL8 after ALA-PDT to recruit more in ammatory factors is bene cial in treating acne vulgaris. Then, we found that p38 pathway might be involved in up-regulation of CXCL8 after ALA-PDT (Fig. 5). When treated with p38 inhibitor SB203580, the expression of CXCL8 showed decreased after ALA-PDT in sebocytes. Thus, CXCL8 was up-regulated after ALA-PDT via p38 MAPK pathway in sebocytes. However, there are still other in ammatory-related factors or pathways could regulate the microenvironment after ALA-PDT. Further studies are necessary to determine the mechanisms of complex in ammatory reactions induced by ALA-PDT for severe acne vulgaris.

Conclusions
In conclusion, our study demonstrated that ALA-PDT up-regulates the expression of CXCL8 in sebocytes and sebaceous gland via p38 pathway, then CXCL8 ampli es intense in ammatory reaction, suggesting a potential basis for the clinical application of ALA-PDT in the treatment of severe acne.

Cells and animals
Primary human sebocytes were isolated from pilosebaceous unit (face) and used at passages 4-8. The human sebocytes were maintained in Keratinocyte-SFM supplemented with 10% FBS, 10 ng/ml epidermal growth factor, 100 U/ml penicillin and 100 g/ml streptomycin at 37•C in 5% CO2 air atmosphere. Male golden hamster (42 ~ 48 days old) were obtained from Vital River Laboratories (Beijing, China).

Sebaceous gland
Sebaceous glands were obtained from the back of golden hamster after incubated with 8% ALA for 4 h in the dark. The samples were frozen at − 20 °C for about 30 min. The frozen tissue together with the vise were quickly observed under a stereomicroscope. The vertical plane of the skin sample was cut by a sterile scalpel to obtain slices that contained intact pilosebaceous units. The sebaceous glands in the hand-made slices were observed under white and UV light, and the spectrum of sebaceous glands was documented by the micro uorospectrometer. (19) Protoporphyrin IX (PpIX) absorption Sebocytes were treated with ALA (0.5 mM) for 4 h. After 24 h, the cells were observed by uorescence microscope (FM-YG100; Soochow FZM Optical Tech, China) with blue light (peak at 365 nm). After 10 min in dimethyl sulfoxide, the supernatant was collected, then detected by uorescence spectrophotometer with excitation at 405 nm wavelength.

ALA-PDT treatment
Sebocytes were incubated with ALA (0.5 mM) in serum-free medium for 4 h at 37 °C, then washed twice with phosphate buffered saline (PBS) and were irradiated by 10 J/cm 2 red LED light (635 nm) at 16.67 mW/cm 2 . In vivo study, 8% ALA cream was topically applied on the sebaceous gland on the back of golden hamster. After incubated for 3 h in the dark, excess ALA cream was removed, and the sebaceous gland was irradiated by a red LED light (635 nm) at a power density of 80 mW/cm 2 and energy density of 38.4 J/cm 2 .
Cell viability and apoptosis assay Cell viability was measured using a CCK-8 Kit (Beyotime Biotechnology, Shanghai, China). Sebocytes (5 × 10 3 per well) were seeded into 96-well plates, then treated with ALA (0.5 mM), red light and ALA-PDT respectively. 24 h after the treatments, 10 µL of CCK-8 was added to each well for 1 h. The absorbance of cell culture at 450 nm was detected with a Thermo Scienti c Microplate Reader. Cell apoptosis was monitored by ow cytometry. Sebocytes had been treated with ALA (0.5 mM), red light or ALA-PDT. 24 h later, these cells were harvested and quantitatively analyzed with a FACScan ow cytometer (BD, Franklin Lakes, NJ, USA).

Intracellular lipids determination
Sebocytes were treated with ALA (0.5 mM), red light or ALA-PDT. After 24 h, the cells were washed with PBS and neutral lipids were labeled with the Nile red (10 µg/ml in DMSO). After 5 min in dimethyl sulfoxide, the cells were observed by uorescence microscope (FM-YG100; Soochow FZM Optical Tech, China) with excitation at 485 nm and emission at 565 nm to detect red uorescence. (20) Elisa analysis Sebocytes were cultured in six-well tissue culture plates and treated with ALA (0.5 mM), red light or ALA-PDT. After centrifugation, the supernatants were collected at 24 h post-treatment and analyzed using an ELISA-based IL-6, TNF-α, IL-1β and CXCL8 detection kit.

Quantitive real-time PCR analysis
Sebocytes were treated with ALA (0.5 mM), red light or ALA-PDT. Then, the cells were harvested using TRIzol according to the manufacturer's instruction and reverse transcribed to cDNA. Following reverse transcription, the samples were subjected to Taqman qPCR analysis on a 7900 H T Fast Real-Time PCR System (Life Technologies, ThermoFisher, Loughborough, UK). CXCL8 mRNA levels were detected by SYBR Green qPCR (Life Technologies) using the following primers: 5′-TTG CCA AGG AGT GCT AAA G-3′ (human CXCL8 forward primer), 5′-CACTCTCAATCACTCTCAGTTC-3′ (human CXCL8 reverse primer). GAPDH mRNA level was used as control.
Immunohistochemical studies Freshly isolated tissue from sebaceous gland on the back of golden hamster was obtained in 1 h, 3 h, 6 h, 12 h and 24 h after ALA-PDT. The tissue was stored in formalin and 5 µm sections were dewaxed (30 min 56 °C, 2 × 10 min xylene), followed by rehydration, antigen unmasking, and blocking. Then, the samples were stained with anti-CD3, anti-CD19, anti-CD68, and anti-CXCL8 primary antibodies at 1 µg/mL in blocking solution for 30 min at 37 °C. The slides were rinsed in PBS and incubated with a goat anti-rabbit IgG secondary antibody (Boster, China) diluted in blocking solution for 30 min. The slides were inculcated with streptavidin-biotin complex (Boster, China) for 30 min, rinsed in PBS, stained using DAB chromogen and hematoxylin counterstain, and observed under a light microscope. Exposure to PBS was used for negative control sections.

Western blot analysis
Sebocytes were cultured in six-well tissue culture plates and treated with ALA (0.5 mM), red light and ALA-PDT respectively. Total protein was extracted from the cells with RIPA lysis buffer. Afterwards, Western blotting was performed as described previously.

Declarations Ethics approval
The study protocol of golden hamsters was approved by the Ethics Committee of Shanghai Skin Disease Hospital (NO.11400700312027). The license of golden hamsters was SCXK2016-0011.   The secretion of lipids in sebocytes were observed by uorescence microscope after Nile red staining. All the results were shown as mean ± SD (n = 3), which were three separate experiments performed in triplicate. * = p < 0.05 signi cantly different from control. Tissues were taken at 1h, 3h, 6h, 12h, 24 h after ALA-PDT and untreated tissues were used for comparisons. It was observed markedly increased expression of CXCL8. All the results were shown as mean ± SD (n = 3), which were three separate experiments performed in triplicate. ***= p < 0.001 signi cantly different from control.

Figure 4
HE and immunohistochemical staining of T cell marker CD3, neutrophils marker CD11b, B cell marker CD19 and macrophages marker CD68 in sebaceous gland of golden hamster. Tissues were taken at 1h, 3h, 6h, 12h, 24 h after ALA-PDT and untreated tissues were used for comparisons. It was observed markedly increased expression of CD3 and CD11b, moderately increased expression of CD19 and CD68 from 3h after ALA-PDT. were determined by the relative intensities of the protein bands. (C) p38 inhibitor SB203580 inhibited the level of CXCL8 in sebocytes. Treated cells were collected for qRT-PCR 24h after ALA-PDT. All the results were shown as mean ± SD (n = 3), which were three separate experiments performed in triplicate. *** = p < 0.001 signi cantly different from control.