The presence of dark melanin (eumelanin) within human epidermis represents one of the strongest predictors of low skin cancer risk. Topical rescue of eumelanin synthesis, previously achieved in “redhaired” Mc1r- deficient mice, demonstrated significant protection against UV damage. However, application of a topical strategy for human skin pigmentation has not been achieved, largely due to the greater barrier function of human epidermis. Salt-inducible kinase (SIK) has been demonstrated to regulate MITF, the master regulator of pigment gene expression, through its effects on CRTC and CREB activity. Here, we describe the development of small-molecule SIK inhibitors that were optimized for human skin penetration, resulting in MITF upregulation and induction of melanogenesis. When topically applied, pigment production was induced in Mc1r-deficient mice and normal human skin. These findings demonstrate a realistic pathway toward UV-independent topical modulation of human skin pigmentation, potentially impacting UV protection and skin cancer risk.
The incidence of nonmelanoma and melanoma skin cancers has been increasing in the United States over recent decades. Epidemiological evidence suggests that there is a causal relationship between sun/UV exposure and the three major histologic forms of skin cancer: squamous cell carcinoma, basal cell carcinoma, and cutaneous melanoma. Individuals with fair skin and/or poor tanning ability are at higher risk for developing these malignancies, which are uncommon in darkly pigmented individuals. During UV-induced tanning, DNA damage in keratinocytes triggers p53-mediated transcription of the pro-opiomelanocortin (POMC) gene. Proteolytic cleavage of POMC produces alpha-MSH (melanocyte-stimulating hormone), which binds to the melanocortin-receptor-1 (MC1R) on melanocytes, activating adenylate cyclase. Elevated cyclic AMP (cAMP) activates protein kinase A (PKA), which phosphorylates the cAMP-responsive-element-binding protein (CREB), which, in turn, stimulates the transcription of the microphthalmia-associated transcription factor (MITF) gene. MC1R non-signaling variants are associated with lighter skin tones and red hair and are linked to poor tanning responses. Previously, topical application of the cAMP agonist forskolin was shown to rescue the cAMP-MITF-eumelanin pathway in Mc1r-deficient mice. Subsequent studies identified the phosphodiesterase PDE4D3 as a key regulator of melanocytic cAMP homeostasis, and its suppression produced hyperpigmentation similar to forskolin treatment in red-haired mice. However, attempts to apply both of these small-molecule approaches to human skin have been unsuccessful, likely related to poor skin penetration of the active species.
Genetic data in mice have suggested the presence of a pathway in which CREB-regulated transcription co-activator (CRTC) positively regulates and salt-inducible kinase 2 (SIK2) negatively regulates MITF and pigment synthesis independently of CREB phosphorylation by PKA. In macrophages, the small-molecule SIK inhibitor HG 9-91-01 has been shown to regulate CREB-dependent gene transcription by suppressing phosphorylation of CRTC, thereby inhibiting cytoplasmic sequestration and permitting its nuclear translocation. We hypothesized that small-molecule SIK inhibitors could be generated and optimized as topical agents capable of inducing cutaneous pigmentation independently of UV irradiation in human skin.
To test regulation of the pigmentation pathway by the previously published SIK inhibitor HG 9-91-01 (HG) in vitro, we treated normal human melanocytes, UACC62 human melanoma cells, and UACC257 human melanoma cells.Dose-dependent increases in expression of MITF were observed in these cells in response to SIK inhibitor application (Figures 1A, S1A, and S1D). RNA levels of the MITF target gene TRPM1 also increased and followed the anticipated delayed kinetics relative to MITF induction in normal human melanocytes (Figures 1B and 1C) and UACC257 human melanoma cells (Figures S1G and S1H). Gross pigmentation was observed in cell pellets of UACC257 human melanoma cells after 3 days of HG 9-91-01 treatment (Figure 1D). Since SIK kinase activity is known to be dependent on LKB1 we next evaluated whether SIK-inhibitor treatment of LKB1-null G361 melanoma cells would induce MITF. In LKB1-null G361 melanoma cells, there is no MITF induction with SIK-inhibitor treatment (Figure S1J). In contrast, when LKB1 is introduced in G361 melanoma cells (Figure S1I), we observed a 6-fold induction of MITF expression with SIK-inhibitor treatment (Figure S1K), demonstrating the dependence of SIK-inhibitor effect on active SIK. These data suggest that small-molecule SIK inhibition can stimulate the pigmentation pathway in vitro.
(A) mRNA expression of MITF relative to RPL11 mRNA and vehicle control in normal human melanocytes 3 hr after HG 9-91-01 or vehicle control (70% ethanol, 30% propylene glycol) treatment, quantified by qRT-PCR (n= 3, mean ± SEM).
(B and C) mRNA expression of MITF (B) and MITF- dependent gene TRPM1 (C) relative to RPL11 mRNA and vehicle control at each time point, in normal human melanocytes over 24 hr after 4 μM HG 9-91-01 or vehicle control treatment, quantified by qRT-PCR (n= 3, mean ± SEM).
(D) Cell pellets of UACC257 melanoma cells after 3 days of treatment with vehicle control or 4 μM SIK inhibitor HG 9-91-01 (image is representative of n= 3 experiments).
For the graph in (A), statistical significance is reported as follows: ∗∗∗p< 0.001; ∗∗∗∗p< 0.0001, one-way ANOVA with Dunnett’s multiple comparisons test comparing treatment dose to vehicle control. For the graphs in (B) and (C), statistical significance is reported as follows: ∗p< 0.05; ∗∗p< 0.01; ∗∗∗p< 0.001; ∗∗∗∗p< 0.0001, repeated-measures one-way ANOVA with Dunnett’s multiple comparisons test comparing each time point to time point 0.
Since our in vitro results demonstrated that inhibition of SIK by HG 9-91-01 positively regulated MITF transcription, we next evaluated whether topical application of this compound could induce pigmentation independent of MC1R in vivo. To test this, we utilized a previously described mouse “red hair” model that carries the inactivating Mc1r e/e mutant allele and a transgene, K14-SCF, in which stem cell factor expression is driven by the keratin-14 promoter, allowing for epidermal homing of melanocytes. Albino mice harboring a mutation in the tyrosinase gene were combined with the K14-SCF transgene (Tyr c/c;K14-SCF mice) and served as controls to evaluate whether the pigmentation afforded by topical SIK inhibitor was dependent upon the canonical tyrosinase-melanin pathway. Daily application of the SIK inhibitor HG 9-91-01 for 7 days caused robust darkening in Mc1r e/e;K14-SCF mice (Figures 2A and S2A). No visible change in skin pigmentation was observed in Mc1r e/e;K14-SCF mice treated with vehicle or in Tyr c/ c;K14-SCF mice treated with vehicle or HG 9-91-01 (Figures 2A, S2A, and S2B). Reflective colorimetry analysis (Commission Internationale de l’Eclairage [CIE] L∗ white-black color axis) revealed significant darkening in Mc1r e/e;K14-SCF mice treated with SIK inhibitor, but not in vehicle-treated Mc1r e/e;K14-SCF mice or in Tyr c/c;K14-SCF mice treated with either SIK inhibitor or vehicle control (70% ethanol, 30% propylene glycol) (Figure 2B). Fontana-Masson staining, a specialized melanin stain, revealed strong induction of melanin production in Mc1r e/e;K14-SCF mice only in areas treated with HG 9-91-01 (Figures 2C and S2D) but no pigment induction in Mc1r e/e;K14-SCF mice treated with vehicle (Figure 2C) or in albino (Tyr c/c;K14-SCF) mice treated with either vehicle or SIK inhibitor (Figure S2C). Nuclear capping of melanin-laden melanosomes was observed within epidermal keratinocytes in Mc1r e/e;K14-SCF mice treated with HG 9-91-01 (indicated by white arrows) and represents a known subcellular localization typical of physiologic skin pigmentation (Figure 2C). This feature suggests that SIK-inhibitor treatment stimulates not only melanocytic pigment synthesis but also the export of melanin in a fashion that closely mimics the known pathway of UV melanogenesis. H&E staining revealed normal morphology of HG 9-91-01-treated Mc1r e/e;K14-SCF (Figure 2C) and Tyr c/c;K14-SCF epidermis (Figure S2C). NaOH lysis of skin samples revealed a visible increase in extractable eumelanin from Mc1r e/e;K14-SCF mice treated with HG 9-91-01, compared with all other treatment groups (Figure 2D).
(A–D) Shown here: (A) Mc1r e/e;K14-SCF mice and Tyr c/c;K14-SCF mice before treatment (day 0) and after 7 days of treatment (day 7) with 30 μL vehicle control (70% ethanol, 30% propylene glycol) or 37.5 mM HG 9-91-01 (image is representative of n= 4 experiments). (B) Reflective colorimetry measurements (L∗ white-black color axis; n= 4, mean ± SEM) and (D) melanin extraction (image is representative of n= 4 experiments) of the Mc1r e/e
