Guanine Deaminase in Human Epidermal Keratinocytes Contributes to Skin Pigmentation

Epidermal keratinocytes are considered as the most important neighboring cells that modify melanogenesis. Our previous study used microarray to show that guanine deaminase (GDA) gene expression is highly increased in melasma lesions. Hence, we investigated the role of GDA in skin pigmentation. We examined GDA expression in post-inflammatory hyperpigmentation (PIH) lesions, diagnosed as Riehl’s melanosis. We further investigated the possible role of keratinocyte-derived GDA in melanogenesis by quantitative PCR, immunofluorescence staining, small interfering RNA-based GDA knockdown, and adenovirus-mediated GDA overexpression. We found higher GDA positivity in the hyperpigmentary lesional epidermis than in the perilesional epidermis. Both UVB irradiation and stem cell factor (SCF) plus endothelin-1 (ET-1) were used, which are well-known melanogenic stimuli upregulating GDA expression in both keratinocyte culture alone and keratinocyte and melanocyte coculture. GDA knockdown downregulated melanin content, while GDA overexpression promoted melanogenesis in the coculture. When melanocytes were treated with UVB-exposed keratinocyte-conditioned media, the melanin content was increased. Also, GDA knockdown lowered SCF and ET-1 expression levels in keratinocytes. GDA in epidermal keratinocytes may promote melanogenesis by upregulating SCF and ET-1, suggesting its role in skin hyperpigmentary disorders.


Introduction
Skin hyperpigmentation is caused by the interplay between melanocytes and neighboring cells of keratinocytes (KCs), fibroblasts, endothelial cells, and inflammatory cells [1][2][3][4][5][6]. Among them, KCs may have the most important role as they produce abundant melanogenic mediators. KCs, the major type of skin cells, act as the first barrier to UV or environmental stress. KC production and/or growth factor and cytokine secretion are essential stress response mechanisms in response to UV irradiation, inflammatory signals, or skin injury [1][2][3]. Indeed, many paracrine factors secreted from epidermal KCs upon stimulation regulate melanogenesis. Alpha-melanocyte-stimulating hormone (α-MSH) and adrenocorticotropic hormone (ACTH) are key melanogenic stimulators secreted by human epidermal KCs upon UV irradiation [7]. When bound to the melanocortin 1 receptor expressed We also performed quantitative PCR (qRT-PCR) in samples from four RM lesions and perilesional controls. The clinical information on the patients who participated in NGS and qRT-PCR analyses are shown in Supplementary Materials Table S1. GDA mRNA expression was higher in hyperpigmented lesions than in the controls ( Figure 1A). Hematoxylin and eosin (H&E) staining of the hyperpigmented lesions showed epidermal hyperpigmentation with dermal melanophages ( Figure 1B). Immunofluorescence staining showed that GDA was more strongly expressed in the epidermal KCs of RM lesions than in those of perilesional controls ( Figure 1C). These results suggested that GDA expression in KCs is associated with skin hyperpigmentation. However, serum ET-1, SCF, ACTH, and α-MSH levels from 15 patients with RM were not significantly different from those in healthy volunteers (n = 6, Figure 1D), although serum ET-1 level was found to be higher than that in volunteers Molecules 2020, 25, 2637 3 of 13 analyses are shown in Supplementary Materials Table S1. GDA mRNA expression was higher in hyperpigmented lesions than in the controls ( Figure 1A). Hematoxylin and eosin (H&E) staining of the hyperpigmented lesions showed epidermal hyperpigmentation with dermal melanophages ( Figure 1B). Immunofluorescence staining showed that GDA was more strongly expressed in the epidermal KCs of RM lesions than in those of perilesional controls ( Figure 1C). These results suggested that GDA expression in KCs is associated with skin hyperpigmentation. However, serum ET-1, SCF, ACTH, and α-MSH levels from 15 patients with RM were not significantly different from those in healthy volunteers (n = 6, Figure 1D), although serum ET-1 level was found to be higher than that in volunteers   (D) Box plots displaying the minimum, the maximum, the median, and the first and third quartiles for serum endothelin-1 (ET-1), stem cell factor (SCF), adrenocorticotropic hormone (ACTH), and alpha-melanocyte-stimulating hormone (α-MSH) levels in patients with RM (n = 15) and age-matched healthy volunteers (n = 6). * p < 0.05. Next, we attempted to compare GDA expression in primary human KCs, fibroblasts, and melanocytes. To investigate whether GDA is constitutively expressed in cultured skin cells, we performed qRT-PCR. GDA mRNA expression in normal human KCs (NHKs) was nearly 100-fold higher than that in NHMs and human dermal fibroblasts (HDFs, Figure 2A). Thus, we focused on the role of GDA expression in KCs interacting with melanocytes to develop pigmentation.
When GDA was visualized by immunofluorescence staining in KC and melanocyte coculture, it was predominantly localized in the cytosol of KCs. Transfecting the coculture with a small interfering RNA for GDA (siGDA) suppressed the GDA signal in KCs ( Figure 2B). In addition, melanosomes were far less abundant in siGDA-transfected cocultured cells than in cells transfected with negative control siRNA (siNC) ( Figure 2C).
To determine whether GDA expression can be associated with inflammatory stimuli, human KCs were treated with various inflammatory cytokines, and the GDA mRNA expression level was evaluated by qRT-PCR. The expression level of RELA, a member of the nuclear factor-kappa B (NF-κB) family known to be involved in the inflammatory response [33][34][35], was also investigated. The GDA mRNA expression level was elevated with interleukin 1 alpha (IL-1α) or tumor necrosis factor alpha (TNF-α), while TNF-α or interferon gamma (IFN-γ) increased RELA mRNA expression. Both GDA and RELA mRNA expression levels were increased the most after treatment with a mixture of IL-1α, TNF-α, and IFN-γ ( Figure 2D). KCs are the well-known neighboring cells to melanocytes, which secrete melanogenic mediators to melanocytes in the epidermis [1][2][3]. To investigate whether GDA can be induced by UVB irradiation in KCs, which is the most important physiologic triggering factor for hyperpigmentation, we measured GDA expression after exposure to low-dose UVB. Exposure of KCs to UVB led to an increase in GDA mRNA expression in a dose-dependent manner ( Figure 3A). UVB irradiation is known to increase melanogenesis by upregulating SCF and ET-1, melanogenic growth factors in KCs [3]. When we stimulated melanogenesis in a coculture of NHMs and human KCs by melanogenic growth factors from KCs, the combination of 10 ng/mL SCF and 0.1 nM ET-1 worked as the most consistent stimulator ( Figure 3B). GDA and tyrosinase mRNA levels were significantly increased in the SCF/ET-1-costimulated coculture ( Figure 3C).  we measured GDA expression after exposure to low-dose UVB. Exposure of KCs to UVB led to an increase in GDA mRNA expression in a dose-dependent manner ( Figure 3A). UVB irradiation is known to increase melanogenesis by upregulating SCF and ET-1, melanogenic growth factors in KCs [3]. When we stimulated melanogenesis in a coculture of NHMs and human KCs by melanogenic growth factors from KCs, the combination of 10 ng/mL SCF and 0.1 nM ET-1 worked as the most consistent stimulator ( Figure 3B). GDA and tyrosinase mRNA levels were significantly increased in the SCF/ET-1-costimulated coculture ( Figure 3C).

siGDA in KCs Downregulates Melanogenesis while GDA Overexpression promotes Melanogenesis in the Coculture
To examine the association between GDA and melanogenesis, we silenced GDA using siGDA in the coculture, which significantly reduced the melanin content and tyrosinase mRNA expression on day 5 ( Figure 4A,B).
Since we found that GDA mRNA level is increased under the UVB melanogenesis-stimulatory conditions in KCs, we sought to elucidate whether suppression of GDA expression also decreases the melanin content in the UVB-stimulated coculture. Melanin content was even lower in the UVBstimulated coculture of KCs and melanocytes after GDA suppression in KCs by siGDA than in the

siGDA in KCs Downregulates Melanogenesis While GDA Overexpression Promotes Melanogenesis in the Coculture
To examine the association between GDA and melanogenesis, we silenced GDA using siGDA in the coculture, which significantly reduced the melanin content and tyrosinase mRNA expression on day 5 ( Figure 4A,B).
Molecules 2020, 25, 2637 7 of 13 coculture without UVB stimulation, suggesting the regulating role of GDA in UVB-mediated melanogenesis ( Figure 4C). Next, we attempted to induce GDA overexpression using an adenovirus vector. While NHMs and NHKs could not be infected with GDA-adenovirus, HaCaT cells were successfully infected. Increased GDA expression in HaCaT cells promoted melanogenesis in a 1:1 coculture of NHMs and GDA-overexpressed HaCaT cells ( Figure 4D).

KC GDA Expression is Involved in the Melanogenic Property of UV Treated KC-Conditioned Media
To investigate how KC GDA regulates melanocyte function, we treated melanocytes with UVBtreated KC-conditioned media, which increased the melanin content ( Figure 5A). ET-1 and SCF are known to be secreted from KCs after UVB irradiation and subsequently act on melanocytes via c-KIT or endothelin receptor type B to accentuate melanogenesis [3]. ET-1 and SCF mRNA expression levels were significantly lower in siGDA-transfected NHKs than in siNC-transfected NHKs, suggesting GDA involvement in the regulation of ET-1 and SCF production from KCs ( Figure 5B). In addition, Since we found that GDA mRNA level is increased under the UVB melanogenesis-stimulatory conditions in KCs, we sought to elucidate whether suppression of GDA expression also decreases the melanin content in the UVB-stimulated coculture. Melanin content was even lower in the UVB-stimulated coculture of KCs and melanocytes after GDA suppression in KCs by siGDA than in the coculture without UVB stimulation, suggesting the regulating role of GDA in UVB-mediated melanogenesis ( Figure 4C).
Next, we attempted to induce GDA overexpression using an adenovirus vector. While NHMs and NHKs could not be infected with GDA-adenovirus, HaCaT cells were successfully infected. Increased GDA expression in HaCaT cells promoted melanogenesis in a 1:1 coculture of NHMs and GDA-overexpressed HaCaT cells ( Figure 4D).

KC GDA Expression is Involved in the Melanogenic Property of UV Treated KC-Conditioned Media
To investigate how KC GDA regulates melanocyte function, we treated melanocytes with UVB-treated KC-conditioned media, which increased the melanin content ( Figure 5A). ET-1 and SCF are known to be secreted from KCs after UVB irradiation and subsequently act on melanocytes via c-KIT or endothelin receptor type B to accentuate melanogenesis [3]. ET-1 and SCF mRNA expression levels were significantly lower in siGDA-transfected NHKs than in siNC-transfected NHKs, suggesting GDA involvement in the regulation of ET-1 and SCF production from KCs ( Figure 5B). In addition, ET-1 concentration was significantly lower in the media of siGDA-transfected NHKs than those of siNC-transfected NHKs on day 2 ( Figure 5C).

KC GDA Expression is Involved in the Melanogenic Property of UV Treated KC-Conditioned Media
To investigate how KC GDA regulates melanocyte function, we treated melanocytes with UVBtreated KC-conditioned media, which increased the melanin content ( Figure 5A). ET-1 and SCF are known to be secreted from KCs after UVB irradiation and subsequently act on melanocytes via c-KIT or endothelin receptor type B to accentuate melanogenesis [3]. ET-1 and SCF mRNA expression levels were significantly lower in siGDA-transfected NHKs than in siNC-transfected NHKs, suggesting GDA involvement in the regulation of ET-1 and SCF production from KCs ( Figure 5B). In addition, ET-1 concentration was significantly lower in the media of siGDA-transfected NHKs than those of siNC-transfected NHKs on day 2 ( Figure 5C).
In our previous study using DNA microarray from melasma lesions, GDA was shown to be one of the most upregulated genes, which was validated by qRT-PCR analysis [17]. KCs were the most abundant type of cells obtained from a skin biopsy using a 2 mm punch, and high GDA expression in cultured KCs suggested that KCs are the source of GDA in hyperpigmented lesions. In this study, we have shown that GDA expression level is also higher in RM lesions than in non-lesional controls. RNA microarray analysis showed that GDA expression level was increased in RM lesions, but its expression in sporadically-occurring congenital hyperpigmentary lesions of café au lait macules was lower than that in perilesions in our preliminary study (data not shown). RM is a peculiar severe form of post-inflammatory hyperpigmentation (PIH) on the face and neck of dark-skinned adults, which is triggered by skin irritation or contact dermatitis and aggravated by UV exposure [44].
Based on the above results, we hypothesize that epidermal KC GDA expression promotes melanogenesis. Our present study showed that GDA in epidermal KCs may promote pigmentation and play a role in UV-induced melanogenesis by interacting with KC-derived melanogenic growth factors ET-1 and SCF, suggesting its role in UV-aggravated skin hyperpigmented disorders. These results suggest that inhibiting keratinocyte GDA expression can be a novel approach for treating skin hyperpigmentary disorders.

Patients
We included patients who were clinically diagnosed with RM in the Dermatology Department of the Asan Medical Center (Seoul, Korea). The patients were excluded if they had other possible causes of hyperpigmentation such as Addison's disease, hyperthyroidism, and hemochromatosis. We also excluded patients with hyperpigmentation in sites other than the face and neck. Clinical features such as age, sex, location of the lesion, Fitzpatrick skin type, morphology of the lesion (spotty, reticulated, diffuse, presence of erythema of hypopigmentation), and color of the lesion (bright/average, dark/very dark) were identified.

Next Generation Sequencing (NGS)
We performed punch biopsies of lesional skin and normal adjacent skin on 3 RM patients (one biopsy specimen for lesional and non-lesional control area, respectively, per person). Total RNA was isolated from tissue using the Trizol based method. One microgram of total RNA was processed for preparing the mRNA sequencing library using the TruSeq stranded mRNA sample preparation kit (Illumina, San Diego, CA, USA) according to the manufacturer's instructions. The first step involves purifying the poly-A containing mRNA molecules using poly-T oligo attached magnetic beads. Following purification, the mRNA is fragmented into small pieces using divalent cations under elevated temperature. The cleaved RNA fragments are copied into first strand cDNA using reverse transcriptase and random primers. Strand specificity is achieved by replacing dTTP with dUTP in the second strand marking mix (SMM), followed by second strand cDNA synthesis using DNA polymerase I and RNase H. These cDNA fragments then have the addition of a single 'A' base and subsequent ligation of the adapter. The products are then purified and enriched with PCR to create the final cDNA library. Finally, quality and band size of libraries were assessed using an Agilent 2100 bioanalyzer (Agilent, Santa Clara, CA, USA). Libraries were quantified by qPCR using the CFX96 Real Time System (Biorad, Hercules, CA, USA). After normalization, sequencing of the prepared library was conducted on the Nextseq system (Illumina) with 75 bp paired-end reads. The result was aligned using the reference human genome (hg19). Three independent analyses were performed per biopsy specimen.

Expression Analysis by qRT-PCR
Total RNA was isolated from skin biopsies (one biopsy specimen for lesional and non-lesional control area, respectively, per person) or cells using the RNeasy Mini kit (Qiagen; Valencia, CA, USA). Then, 1 µg RNA was reverse-transcribed using the RevertAid First Strand cDNA Synthesis Kit (Invitrogen). qRT-PCR was performed using the LightCycler ® 480II machine coupled with SYBR Green (Roche Applied Science; Indianapolis, IN, USA). For qRT-PCR, the initial denaturation was performed at 95 • C for 5 min, followed by amplification at 95 • C for 10 s, 60 • C for 10 s, and 72 • C for 10 s for 45 cycles. Relative gene expression levels were calculated after normalization to the RPLP0 gene using the ∆∆Ct method. Three independent analyses were performed per biopsy specimen. Three independent cellular experiments were performed in triplicate.
Primers for RPLP0 were used for loading control amplifications. Specific primer sets used for each gene are shown in Table 2.

Immunofluorescence Staining and Serum Analysis
For immunofluorescence staining, the samples obtained from RM lesional skin and adjacent perilesional skin were fixed in 10% neutral buffered formalin and embedded in paraffin. The specimens were cut into 4 µm-thick sections, and serial sections were deparaffinized and rehydrated. For antigen retrieval, the sections were heated in antigen unmasking solution (Vector Laboratories; Burlingame, CA, USA) using a pressure cooker (Biocare Medical, Pacheco, CA, USA) at 120.5 • C for 30 s and 90 • C for 10 s. Then, they were immunofluorescently stained with the primary antibodies for GDA (1:200) and Melan-A (1:50) at 4 • C for 8 h. We used a FITC-conjugated anti-mouse (1:500) secondary antibody to detect Melan-A and anti-rabbit Alexa Fluor 546 (1:500) to detect GDA at 4 • C for 30 min. Images were acquired using a Zeiss LSM 780 laser scanning confocal microscope (Leica; Wetzlar, Germany).

Cell Culture and Melanin Content Assay
NHMs were maintained in Medium 254 (Invitrogen) containing human melanocyte growth supplement (Invitrogen). NHKs were cultured in KC growth medium (EpiLife; Invitrogen) supplemented with human KC growth supplement (Invitrogen). HDFs and HaCaTs were cultured in Dulbecco s modified Eagle s medium (DMEM) supplemented with 10% fetal bovine serum (FBS). All cells were maintained at 37 • C in a humidified atmosphere with 5% CO 2 . For coculture of NHMs and NHKs, 6 × 10 4 NHM cells were seeded in each well of a 6-well plate. The next day, 3 × 10 5 NHK cells were added to each well for the coculture. NHMs and NHKs were cocultured at a 5:1 seeding ratio in KC growth medium.
Cells were dissolved in 550 µL 1 N NaOH at 100 • C for 30 min and centrifuged at 13,000 rpm for 5 min. The optical density of the supernatant was measured at 405 nm using a microplate reader (Molecular Devices, Sunnyvale, CA, USA).

Knockdown and Ectopic Expression of GDA
siGDA and siNC were transfected into cells using the Lipofectamine ® RNAi MAX reagent (Invitrogen) according to the manufacturer's instructions. The siGDA sequence is AGUUGUCAGGAGAACACUA. The detailed knockdown protocol is shown in Table 3. For ectopic expression of GDA, human GDA cDNA was cloned into lentiviral vectors (pcDH-CMV-EF1-puro plasmids (System Biosciences, Mountain View, CA, USA)) and lentiviruses containing either empty vector (Empty) or human GDA (GDA) were produced and introduced into HaCaT cells. After puromycin selection and confirmation of GDA overexpression by qRT-PCR, HaCaT cells overexpressing GDA or infected with empty virus were cocultured with NHMs for 5 d and melanin content was measured as described above.

UVB Radiation and Preparation of KC-Conditioned Media
The cells were exposed twice to a narrow band-UVB lamp (Dermalight Psoracomb UV-B-311 nm-narrowband; National Biological Corp.; Beachwood, OH, USA). The cells were washed and resuspended in phosphate-buffered saline (PBS) prior to UVB radiation exposure. Non-exposed control samples were maintained in the dark under the same conditions. Following UVB radiation exposure, the cells were grown in fresh medium.
NHK cells (6 × 10 5 ) were seeded into a 60 mm dish, cultured for 24 h, washed with PBS, and treated UVB or siGDA. After treatment, the cells were incubated in KC growth medium for 48 h; the conditioned media was harvested, centrifuged, and stored at −20 • C.
ET-1 levels in the conditioned media were measured using an ELISA kit (R&D Systems; Minneapolis, MN, USA) according to the manufacturer's instructions.

Statistical Analysis
The experimental data were expressed as means and standard deviations for three independent experiments performed in triplicate. Differences between results were assessed using the Kruskal-Wallis test in Figure 1 and Student's t-test in Figures 2-5. p-Values < 0.05 were considered significant.

Supplementary Materials:
The following are available online at http://www.mdpi.com/1420-3049/25/11/2637/s1. Table S1: Clinical features of patients with Riehl's melanosis who were included in next generation sequencing and quantitative PCR, Figure S1: The standard curve with synthetic melanin, Figure S2: Phase contrast images of keratinocyte-melanocyte coculture for 5 days.

Conflicts of Interest:
The authors declare no conflict of interest.