Clinical and molecular change induced by repeated low–dose visible light exposure in both light-skinned and dark-skinned individuals

Abstract

Background:

Visible light (VL) is known to induce pigmentation in dark-skinned individuals and immediate erythema in light-skinned individuals. However, the effects of accumulated low-dose VL exposure across skin types are not well established.

Methods:

31 healthy subjects with light (Fitzpatrick skin types [FST] I-II, n=13) and dark (FST V-VI, n=18) skin types were enrolled. Subjects’ buttocks were exposed daily to VL, wavelength 400–700 nm, with a dose of 120 J/cm² at 50 mW/cm², for four consecutive days. Microarray using Affymetrix GeneChip (49,395 genes) was performed followed by qRT-PCR on skin samples.

Results:

Repeated low-dose VL irradiation-induced immediate pigment darkening and delayed tanning in dark-skinned individuals while no discernable pigmentation and erythema were observed in light-skinned individuals. Top ten upregulated genes by repeated VL exposure in microarray included melanogenic genes such as tyrosinase (TYR), tyrosinase-related protein-1 (TYRP1), dopachrome tautomerase (DCT), premelanosome protein (PMEL), melan-A (MLANA), and solute carrier family 24, member 5 (SLC24A5) and genes involved in inflammation/matrix remodeling/cell signaling including chemokine (C-C motif) ligand 18 (CCL18 ), BCL2-related protein A1 (BCL2A1), and cartilage oligomeric matrix protein (COMP). In qRT-PCR CCL18 was upregulated in light skin with a greater extent (mean fold change ± SD; 4.03 ± 3.28, p = 0.04) than in dark-skinned individuals (1.91 ± 1.32, p = 0.07) while TYR was not significantly upregulated in both skin types.

Conclusion:

This study highlights the genes upregulated by cumulative VL exposure involved in pigmentation, immune response, oxidation/reduction, and matrix remodeling across skin types providing relevant information on daily solar exposure.

1. INTRODUCTION

Solar irradiation beyond the ultraviolet (UV) domain can cause biological effects on the skin^1–4. Visible light (VL) has been shown to cause pigmentation and erythema as well as induction of inflammatory cytokines, collagenolytic enzymes, and free radicals⁵. Solar radiation consists of 44% VL (400–700 nm), 3–7% UV (290–400 nm), and 53% infrared (700–1440 nm)⁵. Although VL has less energy than UV, it penetrates deeper into the dermis, potentially causing widespread alterations in the skin⁶. Additionally, VL is not entirely filtered by glass or clothing; thus, the skin is exposed to more VL daily and cumulatively over a lifetime⁷. However, our knowledge of the clinical and molecular effects of VL on human skin remains limited.

The effects of UV irradiation are largely dependent on skin types^8,9. Similarly, VL-induced clinical responses differ across skin types^1,3,10. In Mahmoud’s study, overt pigmentation and erythema were induced in dark skin types in a dose-dependent manner, whereas no pigmentation was observed in light skin types after single VL irradiation³. Later, Randhawa et al. showed that multiple VL exposure was able to induce pigmentation and melanogenesis also in Caucasian skin ex vivo while that was more prominent in dark skin types¹⁰. Along with dose, irradiance is another factor in pigment formation, as more intense pigmentation is observed with increased irradiance at equivalent doses of VL exposure¹¹. Previous studies have used single VL at higher irradiance (200 mW/cm²) for 40 min to deliver a maximal dose of 480 J/cm² to induce a biological response in human skin^3,11. However, few studies have examined the cumulative effects of VL at low irradiance and physiologic doses. This poses a challenge in estimating the real-life effects of VL on the skin. Thus, we evaluated pigmentary changes and gene expression induced by repetitive low-dose exposure of VL in light- and dark-skinned individuals in vivo. The approximate irradiance level of VL in sunlight is 50 mW/cm² (American Society for Testing and Materials standard, 2008)¹¹. Using a VL dose of 120 J/cm² (at 50 mW/cm² for 40 minutes), comparable to daily exposure to natural sunlight, we aimed to investigate the clinical and molecular responses induced by cumulative VL exposure in both dark and light skin types.

2. MATERIALS AND METHODS

This study was approved by the Institutional Review Board of Johns Hopkin Hospital (IRB #00067712) and was conducted according to the Declaration of Helsinki principles. All participants provided written informed consent. Healthy volunteers age eighteen or older without a history of systemic diseases that can affect skin evaluation, without pigmentary disorders (i.e. vitiligo, melasma, and photosensitivity), and not taking any photosensitizing medications were included. Women who were lactating, pregnant, or planning to become pregnant were excluded. All participants were instructed to avoid sun exposure or tanning beds during the study period.

Subjects were exposed to a light source that predominantly emitted VL (400–700nm) as described in previous papers^3,5,10. We used a Fiber-Lite Model 180-D (Dolan-Jenner Industries, Boxborough, MA) with a 150 W quartz-halogen lamp. Inside of the lamp, three KG5/3mm Schott glass filters and one GG400/3mm (Schott North America, Inc., Duryea, PA) were installed in front of the light output to block IR radiation above 700nm and UVA radiation shorter than 400nm, respectively. A straight 8-mm Dolan Jenner glass optical fiber and the liquid light guide were attached to an acrylic holder fixed on the subjects’ buttocks during the irradiation. The irradiance was adjusted to 50 mW/cm² before each session by a power meter (model PM100D, Thorlab Inc., Newton, NJ). We irradiated at a dose of 120 J/cm² VL on two sites daily for four consecutive days (one exposure/day) and a nearby non-irradiated skin site served as the control.

Optical spectra from a Fiber-Lite, 180-D high-intensity fiber optic illuminator was measured using an Andor Optical Spectrum analyzer (AQ6317B). The measurement was carried out by coupling light from the illuminator into a multi-mode fiber (50 um diameter, 0.22 NA). Measured spectrum is shown in Figure 1. Spectrometer was configured to span 350–1400 nm range, with 1 nm optical resolution. The spectral output of our light source was measured as 95.3% VL (400–700nm), 1.5% UVA1 (350–400 nm), and 3.2% infrared (700–1400 nm). For the doses of 120 J/cm² of VL in this study, 1.8 J/cm² of UVA1 would be included.

FIGURE 1.

Spectral output of the light source.

Digital photographs and chromameter measurements were taken pre- and post-VL exposure during the four days of VL irradiation and at follow-up visits on Day 5 (24 hours following the last VL exposure) and Day 12 (8 days following the last VL exposure). Skin color change was measured using a CR-400 Chromameter (Konica Minolta Sensing Americas, Inc, Ramsey, NJ), which uses the L*a*b* color system determined by the Commission International de I’ Eclairage (CIE). The CIE L*a*b* color system represents the quantitative relationship of colors on three axes: L* indicates lightness with values from 0 (black) to 100 (white), a* indicates red/green intensity with positive (red) and negative (green), and b* indicates yellow/blue intensity, with positive (yellow) and negative (blue)¹². The L*, a*, and b* values can be transcribed to dermatological parameters, as the L* value correlates with the level of pigmentation, a* correlates with erythema, and b* correlates with pigmentation and tanning¹³. The temperature was measured every ten minutes during the visible light irradiation using a thermometer.

Skin samples (4mm punch biopsies) were taken on Day 5 from one VL-irradiated site and one control site and on Day 12 from the second VL irradiated site. All biopsy samples were collected and frozen immediately in liquid nitrogen and the specimens were stored at −80℃. Microarray analysis was performed in light skin type (n=4) and dark skin type (n=4) samples collected on Day 5 (post 24hr) and Day 12 (post 8days) using Affymetrix GeneChip DNA microarray (49,395 genes). Gene expression at messenger RNA(mRNA) level was ranked according to fold change (linearlized), post-24hrs compared to control site pair and post-8days compared to control site pair after the four consecutive daily VL irradiations, from the highest (top) to the lowest (bottom). To classify functional gene annotations, the top 200 upregulated genes after VL irradiation were uploaded and analyzed by Database for Annotation, Visualization, and Integrated Discovery (DAVID) bioinformatics resource (https://david.ncifcrf.gov)¹⁴. Quantitative real-time PCR (qRT-PCR) was performed to measure two highly upregulated genes, chemokine (C-C motif) ligand 18 (CCL18) and tyrosinase (TYR) with skin samples of both light (n=9) and dark skinned-individuals (n=7) from control site and VL-irradiated sites obtained on Day 5 and Day 12. Total RNA was isolated from the skin tissue by following the manual of Direct-zol™ RNA MiniPrep Plus kit (Zymo Research, R2073) according to the manufacturer’s instructions. The purity and concentration of RNAs were analyzed using a NanoDrop2000c (Thermo Fisher Scientific, ND-2000c). Following reverse transcriptase reactions using High-Capacity cDNA Reverse Transcription kit (ThermoFisher Scientific, 4374966), qRT-PCR was performed to measure target genes, chemokine (C-C motif) ligand 18 (CCL18) and tyrosinase (TYR), using TaqMan probes (Hs00268113_m1 and Hs00165976_m1, respectively, ThermoFisher Scientific) with iTaq Universal Probes Supermix (BioRad, 1725131). Relative expression of mRNAs was analyzed by the cycle of threshold (Ct) value of target genes and quantified by normalizing to ribosomal protein large P0 (RPLP0) as a housekeeping gene (Hs00420895 gH, ThermoFisher Scientific).

The colorimetric values were demonstrated as difference of L*, a*, and b* values between VL-irradiated sites and non-irradiated control sites. All values from each time point were adjusted to the difference at baseline (V1pre) set as to be zero. Difference of colorimetric values were analyzed using paired t-test followed by Bonferroni correction for multiple analyses, and presented as means (stacks) with ± standard error bars. Generalized linear regression analyses were performed to evaluate the incremental change of the immediate pigmentation and persistent pigmentation as VL was repeatedly exposed. The significance of gene expression from microarray was measured by Student’s t-test analysis. Fold changes of gene expression from qRT-PCR were analyzed using paired t-test and expressed as means ± SD. P <0.05 for 2-tailed hypothesis was considered as statistically significant for all analyses. Data were analyzed using SPSS v10.0 (SPSS Inc., Chicago, IL, USA).

3. RESULTS

Thirty-one healthy volunteers (light skin, FST I-II, n = 13, and dark skin, FST V-VI, n = 18) were enrolled. There were 14 males and 17 females and participants’ age ranges were 18–68 years (Table 1). The temperature measured during all VL sessions did not exceed 40°C and no participants complained of warmth during the treatment. There were no adverse events.

Table 1.

Demographics of participants

Subject	Age	Sex	Race	Skin types	Microarray	Colorimetry	qRT-PCR
1	27	M	W	II	O	O
2	33	F	W	II	O	O
3	31	M	W	II	O	O
4	32	F	W	II	O	O
5	47	M	B	V	O	O
6	20	M	B	VI	O	O
7	49	M	B	V	O	O
8	54	F	B	V	O	O
9	18	M	W	II		O	O
10	23	M	W	II		O	O
11	27	M	W	II		O	O
12	31	F	W	II		O	O
13	22	F	W	II		O	O
14	37	F	W	II		O	O
15	25	F	W	II		O	O
16	38	F	W	II		O	O
17	35	M	W	II		O	O
18	53	M	B	V		O
19	43	F	B	VI		O
20	35	F	B	VI		O
21	53	M	B	V		O
22	55	F	B	VI		O
23	51	M	B	VI		O
24	57	F	B	VI		O	O
25	30	M	B	VI		O	O
26	23	F	B	V		O	O
27	51	F	B	V		O	O
28	68	F	B	V		O	O
29	49	M	B	VI		O
30	38	F	B	V		O	O
31	50	F	B	V			O

3.1. Skin pigmentation and erythema induced by VL in light and dark skin types

The mean baseline colorimetric values in subjects with light skin type (average ± standard deviation, L* 69.68 ± 2.73, a* 8.59 ± 2.13, and b* 13.01 ± 2.69) and dark skin type (L* 38.68 ± 4.35, a* 10.66 ± 1.28 and b* 14.34 ± 3.35) were consistent with previous reports¹⁵. In individuals with light skin (FST I-II, n=13), clinically discernable erythema and pigmentation were not observed. In our colorimetric measurement, L* tended to be decreased and a* tended to be increased at immediately after each VL exposure in light-skinned individuals, but all were not significant. In subjects with dark skin (FST V-VI, n=17), pigmentation occurred immediately after each VL exposure and L* significantly decreased accordingly. Multiple exposures did not increase the immediate pigment darkening calculated by general linear regression model. Of note, while no distinguishable erythema was found immediately after each VL exposure but a significant decrease of a* was observed. This seemed that a* also reflected immediate pigmentary changes, toward more green color elements (which is opposite of the color red in CIE system) in dark-skinned individuals. The sallowness of skin, b* was also significantly decreased immediately after each VL exposure, in accordance with immediate pigment darkening. Persistent pigment darkening after 24hrs of each VL exposure occurred and it became more intense as the skin was repeatedly exposed to VL. In colorimetric value, this was detected by b* value, a significant decrease at 24hrs after the third VL treatment (Day 4 pre). Pigmentation persisted until Day 12 (8 days following the last VL treatment), showing a significant decrease of b* (Figures 2 and 3).

FIGURE 2.

Clinical appearance of pre-and post-VL exposures over four consecutive days as well as Day5 (24h post) and Day12 (8day post) in light (upper two panels) and dark (lower two panels) skin types.

FIGURE 3.

Difference in colorimetric luminescence (L*), erythema (a*), and sallowness (b*) measurements (mean values ± SEM) of pre-and post- VL exposure over four days as well as on Day5 (24h post) and Day12 (8day post) in light (n=13) and dark skin (n=17) types. Those values were adjusted to the difference at baseline (V1pre) set as to be zero, *p < 0.05.

3.2. Gene expression induced by repeated VL exposure analyzed by microarray

In microarray analysis with samples obtained at 24 hours and 8 days after the four consecutive daily VL exposures, gene expression (mRNA) fold changes of 49,395 genes were presented from highest (top) to the bottom according to skin types. The top ten upregulated genes with the highest fold changes compared to the control site among overall skin samples included CCL18, TYR, tyrosinase-related protein-1 (TYRP1), chloride channel accessory 4 (CLCA4), premelanosome protein (PMEL), BCL2-related protein A1 (BCL2A1), melan-A (MLANA), solute carrier family 24, member 5 (SLC24A5), dopachrome tautomerase (DCT), and cartilage oligomeric matrix protein (COMP) (p < 0.001 for all) (Table 2). Particularly, when stratified by skin types, CCL18 was the highest marker, as increased by 4.54-fold in light skin type while CCL18 was increased by 2.71-fold in dark skin type at post-24hrs. TYR was also ranked within the top fifth upregulated gene in both skin types, as increased by 1.83-fold in light skin type while 1.89-fold increase in dark skin type at post-24hrs (Suppl. Table 1).

Table 2.

Top ten upregulated genes in microarray at post-24hrs and post-8days repeated-VL exposure compared to control pair among all skin types.

	Post-24hrs vs. control All skin types			Post-8days vs. control All skin types
	Gene symbol	Fold change	p-value	Gene symbol	Fold change	p-value
1	CCL18	3.50	0.0001	CCL18	2.20	0.0114
2	TYRP1	2.05	2.50E-06	MLANA	1.98	0.0002
3	CLCA4	1.90	2.04E-05	TYRP1	1.96	0.0003
4	TYR	1.86	1.16E-05	TYR	1.87	0.0003
5	PMEL	1.85	9.86E-07	PMEL	1.81	0.0002
6	TYRP1	1.79	1.56E-05	MLANA	1.79	0.0007
7	TYR	1.73	1.96E-06	DCT	1.73	0.0016
8	BCL2A1	1.68	0.0007	COMP	1.73	0.0021
9	MLANA	1.66	2.41E-07	TYR	1.67	0.0003
10	SLC24A5	1.62	5.98E-07	DCT	1.62	0.0010

3.3. CCL-18 and TYR mRNA measurement by quantitative RT-PCR

Thus, we further evaluated the expression of these two genes of interest, CCL18 and TYR by qRT-PCR. We found that CCL18 was significantly elevated at post-24hrs (fold change ± SD, 4.03 ± 3.28, p = 0.04) and elevated at borderline significance at post-8days of VL exposure (2.67 ± 2.21, p = 0.09) in light-skinned individuals. Meanwhile, CCL18 elevation was to a lesser extent with borderline significance in dark-skinned individuals (post-24hrs 2.28 ± 1.78, p = 0.06; post-8days 1.91 ± 1.32, p = 0.07). TYR was upregulated in light-skinned individuals with borderline significance (post-24hrs 2.34 ± 1.56, p = 0.06; post-8days 1.95 ± 1.10, p = 0.06). In dark-skinned individuals, TYR was not significantly upregulated after VL exposure (post-24hrs 1.34 ± 0.74, p = 0.20; post-8days 1.23 ± 0.91, p = 0.47) (Figure 4).

FIGURE 4.

Upregulation of CCL18 and TYR mRNA at 24 hours and 8 days post VL exposure compared to unexposed skin (control) in light (n=8) and dark-skinned (n=10) individuals measured by qRT-PCR. Target gene normalized to RPLP0 (internal control), *p <0.05.

4. DISCUSSION

This exploratory study investigated the clinical and molecular changes induced by repeated low-dose VL exposure in light- and dark-skinned individuals. We irradiated VL (wavelength of 400–700 nm) for four consecutive days to unexposed human skin in vivo with a dose of 120 J/cm² (at the irradiance of 50 mW/cm²) that is equivalent to 40 minutes of sunlight exposure. Repeated VL exposure induced immediate/persistent pigment darkening and delayed tanning in dark-skinned individuals. Colorimetric values, L*, a*, and b*, decreased in accordance with the induction of immediate skin pigmentation in dark-skinned individuals. In contrast, neither erythema nor pigmentation was clinically observed after VL irradiation in light-skinned individuals, although a decrease in L* and an increase in a* immediately after each VL exposure was found. In microarray analyses, the top ten upregulated genes across all skin types after consecutive VL irradiation included inflammatory (CCL18), pigmentary (TYR, TYRP1, DCT, PMEL, MLANA, SLC24A5), anti-apoptotic (BCL2A1), and extracellular matrix (COMP) genes. qRT-PCR of CCL18 and TYR showed that CCL18 expression was significantly upregulated in light-skinned individuals four-fold, while a two-fold increase was found in dark-skinned individuals at post-24hr of repeated VL exposure. However, TYR gene expression was not significantly upregulated in either skin type.

Different skin types have distinct skin color composition^9,15. A study on constitutive skin colors of all skin types demonstrated that dark skin types have lower luminescence and higher redness than light skin types, which is consistent with our colorimetric measurements at baseline¹⁵. There are specific correlations between color parameters that are dependent on the skin type. L* is positively correlated with a* in dark skin types, whereas L* and a* are negatively correlated in light skin types¹⁵. When skin is tanned by UV exposure, the resulting pigmentation continues to follow these correlations in respective skin types and is independent of the proportion or wavelength of UVA and UVB. This could be extended to the VL range based on our results^16,17. After VL irradiation, L* decreased and a* increased in the light skin type, while L* and a* decreased in the dark skin type. This correlation was stronger when the skin was exposed to a high dose and high irradiance of VL (480 J/cm² at 200 mW/cm² for four days) (data not shown). This suggests that skin tanning occurs in its original shade, composed of the same constitution and nature of skin color, which are genetically programmed¹⁶.

VL has been shown to induce dark blue-gray pigmentation that persists for months in the dark skin type. VL-induced skin tanning occurs in three steps. Immediate pigment darkening (IPD) within minutes, persistent pigment darkening (PPD) at 2–24 hours after VL exposure, and delayed tanning (DT) on days 5–7 which lasts for weeks to months¹¹. IPD and PPD are caused by photo-oxidation of pre-existing melanin and spatial redistribution of melanosomes within epidermal keratinocytes¹⁸. In contrast, DT results from the synthesis of new melanin¹⁹. In this study, repeated low-dose VL exposure induced IPD, PPD, and DT in the dark skin type. Considering that IPD intensity did not decrease and PPD gradually increased with repeated VL irradiation in dark skin types, IPD and PPD seemed to not offer protection against subsequent VL exposures. The threshold dose of VL for DT was greater than that for PPD and IPD². We found that DT was induced at an easily obtainable VL dose from sun exposure in dark-skinned individuals. Many genes involved in melanogenesis were upregulated by VL irradiation, including TYR, TYRP1, DCT, PMEL, MLANA, and SLC24A5. This implies that cumulative daily VL exposure might have a role in the development and/or aggravation of pigmentary disorders, which are more common in skin of color, and point to the importance of photoprotection outside of the UV range, especially in dark skin type.

Tyrosinase is the rate-limiting enzyme in melanogenesis²⁰. A previous study showed that TYR enzyme activity and TYR gene expression increased as melanin increased after multiple VL exposures using ex vivo Caucasian skin samples¹⁰. In this study, TYR was significantly upregulated approximately 2-fold at post-24 hrs and post-8 days in each skin type in the microarray, but this result could not be confirmed by qRT-PCR. This discrepancy may be due to the sample size, but one can also surmise that melanin synthesis could occur due to increased TYR activity rather than upregulation of the TYR gene at this low dose of VL exposure¹⁰.

VL has been suggested to have different pigmentary pathways from UV irradiation and the response from VL exposure presented differently in various skin types^3,18,21. The absence of p53 activation after blue light irradiation suggests other pathways are involved in VL-induced melanogenesis compared with UVB irradiation²¹. Opsin-3 is revealed to the key sensor of blue light in melanocytes responsible for hyperpigmentation induced by the shorter wavelengths of VL²². Opsin-3 mediates blue light-induced melanogenesis by upregulation of TYR and DCT and formation of a multimeric TYR/TYRP complex that leads to sustained TYR activity²². When comparing the response to blue light in different skin types, the opsin-3 level was not different between skin types but TYR/TYRP complexes were highly expressed in skin types Ⅲ-Ⅳ compared with skin types I and II. This was correlated with the increased TYR activity in dark-skinned melanocytes whereas TYR activity did not change in light skin type melanocytes providing an explanation for persistent hyperpigmentation in dark skin type by VL irradiation²². On the other hand, in response to UV exposure, the melanocortin-1 receptor (MC1R) on melanocytes promotes melanogenesis by binding to the α-melanocyte-stimulating hormone. However, opsin-3 activation by VL suppressed the MC1R, which supports that other pathways are involved in VL-induced hyperpigmentation ^6,23.

Besides genes involved in pigmentation, genes involved in immune response (CCL18, CCL13, CD1B, CD1C, CD1E, IGSF6 (immunoglobulin superfamily member 6), and TNFRSF14 (TNF receptor superfamily member14)) were also upregulated by VL (among the top 200 upregulated genes by VL). Among those, CCL18 is a cytokine which attracts T cells, B cells, and dendritic cells. Increased CCL18 expression has been associated with atopic dermatitis and cutaneous T cell lymphoma²⁴. The CCL18 gene was upregulated four-fold in light-skinned individuals and to a lesser extent in dark-skinned individuals after repeated VL exposure suggesting VL might induce greater immune response and inflammatory process in light-skinned individuals. Also, genes involved in oxidation-reduction (ALDH3A1 (aldehyde dehydrogenase 3 family member A1), OXR1 (oxidation resistance 1), SC5D (sterol-C5-desaturase), CP (ceruloplasmin), HSDL1 (hydroxysteroid dehydrogenase like 1)) were also upregulated. This is in line with well-known knowledge that VL induces ROS production in a dose dependent manner, especially in the blue light spectrum^25–27. In addition, EGF like domain (COMP, CRTAC1 (cartilage acidic protein 1), EGFL8 (EGF like domain multiple 8)) were also among the top 200 upregulated genes by VL. COMP is a part of the papillary dermis as a non-collagenous glycoprotein component of the extracellular matrix. UVA irradiation induce COMP expression via tissue growth factor-beta (TGF-β) in normal human dermal fibroblasts²⁸. In photo-aged skin, COMP was found to co-localize with amorphous elastotic materials and collagen fibrils spreading into the reticular dermis²⁸. Taken together, one can surmise that VL-induced oxidative stress and matrix remodeling would be sufficient to invoke photodamage in both skin types.

This study has several limitations. The patients were enrolled across various seasons; thus, daily sun exposure during the study period could not be controlled for our participants. This could affect the amount of melanin preconditioned for pigmentation, thus impacting the response to VL. Also, VL was irradiated to one’s buttock which is relatively unexposed skin, therefore it may have a lower tolerance to light-related effects than typical sun-exposed skin. This was an explorative study that focused on finding candidate genes upregulated in the microarray after consecutive VL exposure. Future studies can build off of this data and further explore gene and protein activity associated with VL. Tyrosinase activity was not measured, which may explain the pronounced pigmentation in dark skin types. The multiple exposure design makes it difficult to distinguish each phase of tanning as IPD can blur into PPD, and DT could overlap with IPD/PPD. There was a discrepancy in the proportions of UV, VL, and IR between this study and Mahmoud’s study although the same light source installation was used. The discrepancy could be due to the different methods of measurement of the spectral irradiance of the VL source. We used Andor Optical Spectrum analyzer (AQ6317B) coupling light from the illuminator with a light guide while Mahmoud et al. measured with OL754 spectroradiometer (Optronics, Orlando, FL) which uses an integration sphere and has a much larger dynamic range. Our spectrometer instrument yielded some noise below 400nm and above 750nm which might have contributed to the percentages of UV and IR that might not be real. Standardization of light sources and spectrometer measurements remain an important issue and expert consensus should be established²⁹. Lastly, our light source emitted a non-negligible amount of long UVA1 (350–400nm) at 1.5%, similar to prior studies.^3,11,29,30. Long-wavelength UVA1 has synergistic/additive effects with VL on pigmentation, erythema, and ROS production^11,18,30 Hence, the clinical and biological responses of VL demonstrated in this study may have been enhanced by the presence of long-wavelength UVA1.^11,18,29,30.

In conclusion, repeated low-dose VL induced IPD/PPD and DT in dark-skinned individuals, whereas no discernable pigmentation or erythema was found in light-skinned individuals. Most importantly, this study highlights the genes upregulated by cumulative VL exposure involved in pigmentation, immune response, oxidation/reduction, and matrix remodeling across skin types. Of note, the VL dose and irradiation used in this study, 120J/cm² and 50mW/cm² correspond to approximately 40 minutes of outdoor sun exposure and provide very relevant information on daily solar exposure³¹. Based on the upregulated gene expression, one could surmise that cumulative VL exposure would evoke photoaging changes that may be more pronounced in light-skinned individuals and exacerbate pigmentary diseases (i.e. melasma, and postinflammatory hyperpigmentation) especially in dark-skinned individuals. Further investigations are warranted to further clarify the effects of VL across skin types and to develop protection methods against VL effects.