Analysis of Circadian Clock Gene Expression in Epidermal Human Skin Indicates that Body Location Impacts Rhythm Amplitude

Circadian rhythms allow cells, tissues, and whole organisms to adapt to and anticipate changes to environmental conditions and maximize biological fitness. Most mammalian tissues exhibit circadian rhythmicity, including the skin, where a variety of processes ranging from wound healing to immune defense and responses to external stress vary over the course of 24 hr (Duan et al., 2021Duan J. Greenberg E.N. Karri S.S. Andersen B. The circadian clock and diseases of the skin.FEBS Lett. 2021; 595: 2413-2436Crossref PubMed Scopus (26) Google Scholar; Lubov et al., 2021Lubov J.E. Cvammen W. Kemp M.G. The Impact of the Circadian Clock on Skin Physiology and Cancer Development.Int J Mol Sci. 2021; 22Crossref PubMed Scopus (21) Google Scholar). At the molecular level, the circadian clock is governed by a transcription-translation feedback system in which a transcription factor complex (CLOCK-BMAL1) binds to the promoters of many clock-controlled genes (CCGs) to regulate transcription, including that of cryptochrome (CRY) and period (PER), which feedback to inhibit CLOCK-BMAL1 function (Takahashi, 2017Takahashi J.S. Transcriptional architecture of the mammalian circadian clock.Nature reviews.Genetics. 2017; 18: 164-179Crossref PubMed Scopus (0) Google Scholar). Most analyses of circadian gene expression are derived from nocturnal rodent models. Given that humans are diurnal and are exposed to different environmental conditions, it is important to extend such work to human skin. Indeed, two studies with human subjects were conducted in recent years (Del Olmo et al., 2022Del Olmo M. Spörl F. Korge S. Jürchott K. Felten M. Grudziecki A. et al.Inter-layer and inter-subject variability of diurnal gene expression in human skin.NAR Genom Bioinform. 2022; 4lqac097PubMed Google Scholar; Wu et al., 2018Wu G. Ruben M.D. Schmidt R.E. Francey L.J. Smith D.F. Anafi R.C. et al.Population-level rhythms in human skin with implications for circadian medicine.Proc Natl Acad Sci U S A. 2018; 115: 12313-12318Crossref PubMed Scopus (78) Google Scholar), each with unique characteristics in terms of subject recruitment, biopsy location and frequency, and microarray platform (Fig. 1a). Using a previously outlined bioinformatic approach (Wu et al., 2018Wu G. Ruben M.D. Schmidt R.E. Francey L.J. Smith D.F. Anafi R.C. et al.Population-level rhythms in human skin with implications for circadian medicine.Proc Natl Acad Sci U S A. 2018; 115: 12313-12318Crossref PubMed Scopus (78) Google Scholar), we obtained the two transcriptomic datasets from the Gene Expression Omnibus (GEO) and then plotted the relative expression of core circadian clock genes over the course of the day. As shown in Fig. 1b, clear and similar oscillations were observed for most of the genes, including the times of peak and trough gene expression. However, calculation of the amplitudes of the gene expression oscillations revealed modestly higher amplitudes for most of the genes in the Wu et al. study, including for the CRY and PER transcriptional repressors. Because there were differences in the sex of the subjects in the two studies, which could potentially impact circadian rhythms (Cvammen et al., 2024Cvammen W. Rider S.D. Travers J.B. Kemp M.G. Effects of Age and Sex on the Expression of Core Circadian Clock Genes in Human Skin Epidermis.J Invest Dermatol. 2024; 144: 1172-1175Abstract Full Text Full Text PDF PubMed Scopus (0) Google Scholar), we then compared clock gene expression in males and females from the Del Olmo et al. study. However, though CLOCK and NR1D1 expression exhibited statistically significantly lower amplitudes in females, the other clock genes showed very similar rhythms (Figs 1d-e). Thus, we conclude that sex does not significantly affect the expression of core circadian clock genes in epidermal skin. Using meta3d (Wu et al., 2016Wu G. Anafi R.C. Hughes M.E. Kornacker K. Hogenesch J.B. MetaCycle: an integrated R package to evaluate periodicity in large scale data.Bioinformatics. 2016; 32: 3351-3353Crossref PubMed Scopus (296) Google Scholar) to identify genes from these datasets with 24 hr periodicity in expression, the Del Olmo et al. dataset was found to contain approximately 10 times more genes with rhythmic expression than Wu et al. (Fig. 2a), which is likely due to the additional biopsy frequency providing better resolution of expression. Interestingly, among the 71 genes that were rhythmic in both studies, the amplitudes of the rhythms were on average 4.5 times higher in the Del Olmo et al. dataset than in Wu et al. (Fig. 2b), which was the opposite pattern than what was observed for the core clock genes (Fig. 1c). Consistent with the much higher number of rhythmic genes in Del Olmo et al., many more biological processes were found to exhibit rhythmic expression in skin epidermis (Fig. 2c), including for genes that comprise the cellular DNA damage response (DDR) that is critical to skin biology. Analyses of a subset of these gene pathway amplitudes similarly showed higher average amplitudes in expression in the Del Olmo et al. dataset (Fig. 2d), which indicates that many biological processes exhibit less robust rhythmicity in skin epidermis in Wu et al. than in Del Olmo et al. To explore this phenomenon further, we manually curated a list of meta3d-derived DDR genes into different DNA repair systems and pathways (Fig. 2e) and then plotted the expression of a subset of these genes over the course of the day from the two datasets. As shown in Fig. 2f, the magnitude of change within the gene expression patterns was found to be generally lower in Wu et al. than in Del Olmo et al. Quantitation of gene expression amplitudes confirmed that the expression of these genes was significantly higher for most of these genes in Del Olmo et al. (Fig. 2g), including for the essential nucleotide excision repair factor XPA, which exhibits rhythmic expression in mouse tissues (Gaddameedhi et al., 2015Gaddameedhi S. Selby C.P. Kemp M.G. Ye R. Sancar A. The circadian clock controls sunburn apoptosis and erythema in mouse skin.J Invest Dermatol. 2015; 135: 1119-1127Abstract Full Text Full Text PDF PubMed Scopus (0) Google Scholar; Kang et al., 2009Kang T.H. Reardon J.T. Kemp M. Sancar A. Circadian oscillation of nucleotide excision repair in mammalian brain.Proc Natl Acad Sci U S A. 2009; 106: 2864-2867Crossref PubMed Scopus (170) Google Scholar). However, some genes, such as the transcription factor KLF9 (Spörl et al., 2012Spörl F. Korge S. Jürchott K. Wunderskirchner M. Schellenberg K. Heins S. et al.Krüppel-like factor 9 is a circadian transcription factor in human epidermis that controls proliferation of keratinocytes.Proc Natl Acad Sci U S A. 2012; 109: 10903-10908Crossref PubMed Scopus (115) Google Scholar) and dNTP synthesis protein RRM2, oscillated with similar magnitudes in both studies. Thus, though some genes show similar profiles in both datasets, the level of oscillation of most CCGs differs significantly between the two studies. As described in Fig. 1a, there were several differences in the design of these two studies that could be responsible for the more robust rhythms of core clock genes in Wu et al. (Fig. 1 b, c) but more robust CCG rhythms in Del Olmo et al. (Fig. 2c-g). However, sex does not appear to be responsible (Fig. 1d, e), and the age range is only marginally broader and not provided in the de-identified Wu et al. dataset. It also seems unlikely that the method of epidermal isolation and microarray platform could have such a dramatic and opposing effect on the gene expression profiles of core clock versus CCGs, though these factors cannot be ruled out. The major remaining variable is biopsy location, which was forearm in Wu et al. and lower back in Del Olmo et al. As these body locations are typically exposed to different amounts of incidental visible and ultraviolet radiation, it is possible that cell signaling pathways that respond to light signals could further modulate clock function in the epidermis. However, other factors may contribute to the differences reported here, such as body temperature or other unknown human subject demographics. Nonetheless, it will be important in future analyses of circadian gene expression and function in the skin to consider biopsy location when designing experiments and to determine whether the loss of rhythmicity in DDR and other CCGs is associated with more stable and elevated expression throughout the day, which could provide better and/or more steady protection against external stressors that are relevant to certain regions of the body. Conceptualization: MGK; Data curation: WC, SDR; Formal Analysis: WC, SDR; Funding Acquisition: MGK; Investigation: WC, SDR; Methodology; WC, SDR, MGK; Project Administration: MGK; Supervision: MGK; Writing-Original Draft: WC, MGK; Writing-Review & Editing: WC, SDR, MGK. Gene expression data sets GSE205155 and GSE112660 were analyzed in this study. Not applicable Not applicable The authors state no conflicts of interest. This work was supported by grants from the National Institutes of Health (R01GM130583) and the Veterans Administration (I01CX00224) to MGK.