A Potential Role of Keratinocyte-Derived Bilirubin in Human Skin Yellowness and Its Amelioration by Sucrose Laurate/Dilaurate

Abstract

Sallow and/or dull skin appearance is greatly attributable to the yellow components of skin tone. Bilirubin is a yellow chromophore known to be made in the liver and/or spleen and is transported throughout the body via the blood stream. Recent publications suggest bilirubin may be synthesized in other cells/organs, including the skin. We found human keratinocytes express the transcripts involved in bilirubin biosynthesis. In parallel, we also found human keratinocytes could indeed synthesize bilirubin in monolayer keratinocytes and in a 3D human skin-equivalent model. The synthesized amount was substantial enough to contribute to skin yellowness. In addition, oxidative stress enhanced bilirubin production. Using UnaG, a protein that forms a fluorescent species upon binding to bilirubin, we also visualized the intracellular expression of bilirubin in keratinocytes. Finally, we screened a compound library and discovered that the sucrose laurate/dilaurate (SDL) combination significantly reduced bilirubin levels, as well as bilirubin-mediated yellowness. In conclusion, bilirubin is indeed synthesized in epidermal keratinocytes and can be upregulated by oxidative stress, which could contribute to chronic or transient yellow skin tone appearance. Application of SDL diminishes bilirubin generation and may be a potential solution to mitigate yellowish and/or dull skin appearance.

Keywords: bilirubin; keratinocyte; oxidative stress; skin color; skin tone; skin yellowness; sucrose dilaurate; sucrose laurate.

Figure 1

Transcriptome expression of key genes involved in heme and bilirubin synthesis detected from tKC cell culture with and without hydrogen peroxide treatment. (A) Simplified diagram of heme biosynthesis and bilirubin production (B) List of key genes with transcriptome expression level, log2 fold changes, and p-values of hydrogen peroxide treatment group vs. no treatment control group. Genes encoding key enzymes (ALAS1, HMOX1, and BLVRB) were all significantly upregulated by hydrogen peroxide treatment (250 µM, 6 h). n = 6/group.

Figure 2

(A) Dose–response effect of hemin on de novo bilirubin production in keratinocytes for 48 h. Bar indicates mean ± SEM. *; p < 0.05 vs. no treatment group. (B) UnaG-mCherry bilirubin sensor. Observation of mCherry allows identification of transfected and expressing cells in absence of bilirubin-induced UnaG fluorescence and enables signal normalization to correct for cell-to-cell expression variation. (C) Human keratinocytes expressing sensors grown and maintained in bilirubin-free EpiLife media. mCherry (red) confirms sensor expression. UnaG fluorescence (green) indicates intracellular de novo synthesis of bilirubin in keratinocytes. Scale = 10 µm.

Figure 3

(A) De novo bilirubin synthesis in keratinocytes with or without succinyl acetone (SA, an inhibitor of ALA dehydratase) for 48 h. Bar indicates mean ± SEM. n = 4/group. *; p < 0.05 vs. untreated control group. (B) Representative images of untreated control keratinocytes (left, endogenous bilirubin) and keratinocytes treated with SA for 48 h (right). Scale = 10 µm. (C) Quantification of UnaG green fluorescence (bilirubin) by image analysis in nontreated control and 1 mM SA-treated keratinocytes for 48 h. SA treatment inhibited intracellular production of bilirubin. Bar indicates mean ± SEM. *; p = 0.001 versus control.

Figure 4

(A) Impact of hydrogen peroxide (H₂O₂) treatment (48 h) on de novo bilirubin synthesis in keratinocytes. Bar indicates mean ± SEM. n = 4/group. *; p < 0.05 vs. untreated control. (B) Time-course detection of bilirubin in EpiDerm™ 3D human epidermal equivalence cultures (sum of tissue and medium) with and without 30 µM hydrogen peroxide (H₂O₂) in medium for 72 h. Bilirubin was quantified using HPLC-MS method. Data are mean ± SEM. n = 4/group. *; p < 0.05 vs. baseline, ^; p < 0.05 vs. without H₂O₂ group.

Figure 5

Bright-field microscopy of human abdominal skin section (8 µm) incubated with either (A) 100 µL of DMSO or (B) 500 µg/mL bilirubin dissolved in DMSO for 1 h in dark. Arrowheads indicate visible bilirubin (yellow color) accumulation in epidermal layer of skin section. Scale bar = 100 µm.

Figure 6

Human epidermal skin explants were cultured with vehicle, 1.4 µg/mL bilirubin or 2.1 µg/mL of bilirubin for 44 h (treatment phase). At 44 h, the bilirubin treatments were replaced with DPBS vehicle without bilirubin and epidermal skin samples were equilibrated till 138 h (recovery phase). Data indicate mean ± SEM. n = 6/group. *; p < 0.05 of 1.4 µg/mL bilirubin leg vs. vehicle-treated control sample. #; p < 0.05 of 2.1 µg/mL bilirubin leg vs. vehicle-treated control sample. ^; p < 0.05 for 2.1 µg/mL group vs. 1.4 µg/mL group.

Figure 7

Key modulated biological pathways based on Gene Set Enrichment Analysis treated with different concentrations of bilirubin on human neonatal keratinocytes. Data are shown as log10 of GAGE p-values (negative represents inhibition, positive represents activation). n = 8/group. Color code indicates p-value range.

Figure 8

(A) Reduction of bilirubin concentration by incubation with 0.01% SDL for 20 h. Bar indicates mean ± SEM. n = 3/group. *; p < 0.05 vs. control. (B) Inhibitory effect of 0.001% SDL on H₂O₂-induced bilirubin production in keratinocytes at 48 h. Bilirubin concentrations were quantified using HPLC-MS method and were normalized to ATP level (an indicator of cell viability). Bar indicates mean ± SEM. n = 4/group. *; p < 0.05 vs. control. ^; p < 0.05 vs. 50 µM H₂O₂-treated group. (C) Visualization of bilirubin production by using UnaG-mCherry bilirubin sensor in keratinocytes in absence or presence of 0.001% SDL for 48 h. Scale = 10 µm. (D) Calculated green fluorescence intensity (bilirubin) normalized to red fluorescence intensity (mCherry). Bar indicates mean ± SEM. n = 13/group. *; p < 0.01.

Figure 9

Fluorescent images of frozen human skin sections. Green color in epidermis is bilirubin; blue color indicates DAPI staining. Green signals observed in dermis in nontreated control are due to autofluorescent property of elastin. White dotted line indicates dermal–epidermal junction: (A) nontreated control; (B) 50 µg/mL bilirubin; (C) 50 µg/mL bilirubin + 0.01% SDL. Scale bar = 100 µm.