==== Front Cell Death Discov Cell Death Discov Cell Death Discovery 2058-7716 Nature Publishing Group UK London 37385992 1509 10.1038/s41420-023-01509-8 Article Down-regulation of peptidylarginine deiminase type 1 in reconstructed human epidermis disturbs nucleophagy in the granular layer and affects barrier function Alioli Adebayo Candide 13 Briot Julie 1 Pons Carole 1 Yang Hang 14 Gairin Marie 15 Goudounèche Dominique 2 Cau Laura 16 http://orcid.org/0000-0003-3655-6329 Simon Michel michel.simon@inserm.fr 1 http://orcid.org/0000-0001-8915-8223 Méchin Marie-Claire marie-claire.mechin@inserm.fr 1 1 grid.15781.3a 0000 0001 0723 035X Toulouse Institute for Infectious and Inflammatory Diseases (Infinity), University of Toulouse, CNRS, INSERM, University Paul Sabatier, Toulouse, France 2 grid.508721.9 Centre de Microscopie Électronique Appliquée à la Biologie (CMEAB), University of Toulouse, Medical Faculty of Toulouse, Toulouse, France 3 grid.25697.3f 0000 0001 2172 4233 Present Address: University of Lyon, INSERM UMR1033, Lyon, France 4 grid.424530.4 0000 0001 0725 2364 Present Address: Capgemini, Issy les Moulineaux, France 5 Present Address: Pierre Fabre Dermo-Cosmétique, Muret, France 6 grid.510047.7 Present Address: Silab, Saint-Viance, France 29 6 2023 29 6 2023 2023 9 19813 12 2022 25 5 2023 20 6 2023 © The Author(s) 2023 https://creativecommons.org/licenses/by/4.0/ Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Deimination is a post-translational modification catalyzed by a family of enzymes named peptidylarginine deiminases (PADs). PADs transform arginine residues of protein substrates into citrulline. Deimination has been associated with numerous physiological and pathological processes. In human skin, three PADs are expressed (PAD1-3). While PAD3 is important for hair shape formation, the role of PAD1 is less clear. To decipher the main role(s) of PAD1 in epidermal differentiation, its expression was down-regulated using lentivirus-mediated shRNA interference in primary keratinocytes and in three-dimensional reconstructed human epidermis (RHE). Compared to normal RHEs, down-regulation of PAD1 caused a drastic reduction in deiminated proteins. Whereas proliferation of keratinocytes was not affected, their differentiation was disturbed at molecular, cellular and functional levels. The number of corneocyte layers was significantly reduced, expression of filaggrin and cornified cell envelope components, such as loricrin and transglutaminases, was down-regulated, epidermal permeability increased and trans-epidermal-electric resistance diminished drastically. Keratohyalin granule density decreased and nucleophagy in the granular layer was disturbed. These results demonstrate that PAD1 is the main regulator of protein deimination in RHE. Its deficiency alters epidermal homeostasis, affecting the differentiation of keratinocytes, especially the cornification process, a special kind of programmed cell death. Subject terms Post-translational modifications Nuclear organization https://doi.org/10.13039/501100004718 Université de Toulouse (Toulouse University) - Simon Michel https://doi.org/10.13039/501100001677 Institut National de la Santé et de la Recherche Médicale (National Institute of Health and Medical Research) - Simon Michel https://doi.org/10.13039/501100004794 Centre National de la Recherche Scientifique (National Center for Scientific Research) - Simon Michel issue-copyright-statement© Cell Death Differentiation Association (ADMC) 2023 ==== Body pmcIntroduction Deimination, or citrullination, is a calcium-dependent post-translational modification of arginine to citrulline residues catalyzed by peptidylarginine deiminases (PADs). Deimination alters the charge of targeted proteins thereby causing changes in their conformation and in interactions with partners. In mammals, five conserved PAD isotypes exist [PAD1-4 and PAD6], encoded by distinct and clustered PADI genes [1, 2]. PADs display different tissue expression and are involved in many physiological and pathological processes [3–7]. Three PADs (PAD1-3) are expressed in the human skin [8]. Recently, mutations of human PAD3 have been described in hair and scalp diseases, affecting hair follicle shape [9] and causing a particular alopecia with perifollicular inflammation [10]. However, the specific role of each PAD isotype in the epidermis remains unknown. PAD1 is present in the cytoplasm of keratinocytes and in the matrix of corneocytes, from the basal to the most superficial cornified layer [8, 11, 12]. The ubiquitous PAD2 is detected in all suprabasal layers of living keratinocytes [8, 13]. PAD3 is mainly expressed in granular keratinocytes and in the deepest corneocyte layers [8, 14]. For years, the characterized epidermal targets of PADs were filaggrin, keratins K1 and K10, trichohyalin, filaggrin-2, and hornerin [12, 15–17]. Using a proteomic approach, we identified new substrates such as the keratinocyte proline rich protein, a cornified envelop component that is important for the epidermal barrier [4, 18]. In three-dimensional reconstructed human epidermis (RHE), PAD inhibition by the pan inhibitor Cl-amidine slows down cornification, the last step of keratinocyte differentiation at the granular to cornified cell transition and disturbs the associated autophagy process [19]. PAD1 is suspected of regulating the full degradation of filaggrin, a key protein for the skin barrier function, and hence the production of the filaggrin degradation-related natural moisturizing factor. Indeed, we demonstrated that reducing external relative humidity increases expression of PAD1 (with no impact on the expression of other PADs) and drives human filaggrin breakdown by increasing its deimination rate [20]. Therefore, deimination may play an active role in protecting human skin and in Stratum corneum hydration in physiological conditions. PAD1 has also recently been shown to be less immunodetected in lesional skin of psoriatic patients and to be down-regulated by IL-22 [21]. The aim of the present study was to decipher the specific role of PAD1 in human epidermis. PAD1 expression was down-regulated using RNA interference in human primary keratinocytes used to produce RHEs. The effects of PAD1 down-regulation on RHE morphology, on epidermal barrier function, and on the expression of several differentiation markers were analyzed. An impact of PAD1 down-regulation on granular keratinocyte nucleophagy was highlighted (see graphical abstract). Results RHE produced at 50% relative humidity RHEs were produced in an incubator with relative humidity adjusted to 50% (Supplemental Material and Fig. S1 online). At day 10 (D10), compared to day 4 (D4), RHEs displayed all the characteristics of normal human epidermis, with the four constitutive layers (cornified, granular, spinous, and basal) (Fig. S1A). A marked increase in PAD1 expression and in the deimination rate was observed between D10 and D4 with ratios of about 7 and 14, respectively (Table S1 and Fig. S1B–D), confirming the already known induction of PAD1 at the mRNA and protein levels during keratinocyte differentiation in vitro [19, 21] and in vivo [11]. As expected, profilaggrin and filaggrin were largely detected at mRNA and protein levels (Fig. S1B–D). In this study, RHEs were produced at 50% relative humidity and harvested at D10. PADI1 down-regulation in RHEs almost turns off protein deimination Down-regulation of PAD1 expression was performed using lentivirus-mediated shRNA interference in normal human primary keratinocytes obtained from five different donors, and the transduced keratinocytes were used to produce RHEs. Two shRNAs targeting PADI1 were validated (shPADI1_1, and shPADI1_2) as well as a control shRNA (sh-ctrl), the sequences of which are reported in Table S2. PADI1 down-regulation were probed by RT-qPCR, Western blotting, and using an indirect immunofluorescence assay. Effective and specific down-regulation of PAD1 expression was obtained at the mRNA level with both shPADI1 compared to sh-ctrl (89.50 ± 11.36% for shPADI1_1 and 90.00 ± 1.00% for shPADI1_2) while the amount of PADI3 was maintained, and PADI2, 4 and 6 mRNA remained undetectable (Figs. 1A, 3D and data not shown). Therefore, no compensatory variations were observed in other PADIs. At the protein level, the detection of PAD1 in shPADI1 RHE extracts was reduced to 58.40 ± 6.50% of controls (Fig. 1B, top panel). Consistently, PAD1 was weakly immunodetected in sections of shPADI1_1 and shPADI1_2 compared to sh-ctrl RHEs (Fig. 1C). Importantly, deiminated proteins were barely detected in both shPADI1 RHEs, either by Western blotting (Fig. 1B, middle panel) or by indirect immunofluorescence (Fig. 1D). Only one band (>250 kDa) was weakly detected in the shPADI1 RHE extracts compared to strong detection of numerous deiminated proteins in controls. The global rate of deimination in shPADI1 decreased significantly (~85%) (Fig. 1B and Fig. S2A, B), and deiminated keratins in particular, almost disappeared (Fig. 1B and Fig. S2C, D).Fig. 1 Efficiency and specificity of PAD1 down-regulation. A RT-qPCR of RHEs transduced by sh-ctrl, shPADI1_1, and shPADI1_2. Relative mRNA steady states of PADI1 and PADI3 are reported, other PADI mRNA were undetectable (Cycle threshold, Ct > 35). B–D Expression of PAD1, deiminated proteins, and actin. B Western blot analysis of total extract RHEs transduced by sh-ctrl, shPADI1_1, and shPADI1_2. C PAD1 immunofluorescence of RHE sections after transduction by sh-ctrl, shPADI1_1, and shPADI1_2. Representation using a LUT scale reported to highlight the down-regulation of PAD1 in shPADI1_1 and shPADI1_2 compared to sh-ctrl RHEs. D sh-ctrl, shPADI1_1, and shPADI1_2 RHEs were analyzed by indirect immunofluorescence to localize deiminated proteins (AMC staining) in situ. The line at the bottom represents the polycarbonate filter. SC Stratum corneum, LL living cell layers. Scale bars, 10 µm. PAD1 deficiency in RHEs alters cornification and keratohyalin granules RHE morphology was visualized by hematoxylin/eosin coloration and by transmission-electronic microscopy (TEM) (Fig. 2). At the histological level, when PAD1 was down-regulated, the thickness of the epidermis, in particular that of the cornified layer, seemed to decrease, and keratohyalin granule (KHG) staining was less visible (Fig. 2A). TEM analysis confirmed that the structure of shPADI1 RHEs was normal (Fig. 2B) compared to controls, but had a thinner Stratum corneum with a significant decrease in the number of corneocyte layers (mean number = 12.64 ± 3.81 for the sh-ctrl RHEs versus 6.27 ± 2.87 for shPADI1_1 and 9.18 ± 1.91 for shPADI1_2 RHEs; p < 0.0001) (Fig. 2C, D). Parakeratotic nuclei were sometimes observed in both shPADI1 but not in sh-ctrl RHEs (Fig. 2B, bottom right panel). In addition, the relative KHG surface area relative to the granular keratinocyte cytoplasmic area was significantly reduced [4.53 ± 3.28% for the sh-ctrl versus 3.21 ± 2.14% for shPADI1_1 (p = 0.0117) and 2.48 ± 1.15% for shPADI1_2 RHEs (p < 0.0001)] (Fig. 2C, E). These observations suggest that keratinocyte homeostasis, and consequently the proliferation/differentiation balance, may be impaired when PAD1 is down-regulated.Fig. 2 Tissue morphology and ultrastructural aspects of PAD1 down-regulated RHEs. A Hematoxylin and eosin staining of sh-ctrl, shPADI1_1, and shPADI1_2 RHEs. B Left to right: full vertical TEM sections of sh-ctrl, shPADI1_1, and shPADI1_2 RHEs. B Right bottom, white arrow: parakeratotic nucleus in shPADI1 RHEs. C Stratum corneum of sh-ctrl, shPADI1_1, and shPADI1_2 RHEs. D Number of corneocyte layers of RHEs (mean ± SD) for sh-ctrl and shPADI1_1, n = 5; shPADI1_2, n = 2. E KHG area of RHEs. (sh-ctrl and shPADI1_1, n = 5; shPADI1_2, n = 2). A, B Black lines: LL living cell layers, SC Stratum corneum, SG S. granulosum (bold line), SS S. spinosum, SB S. basale. Scale bars = 10 µm (A), 5 µm (B), 1 µm (C). PAD1 deficiency does not disturb keratinocyte proliferation To evaluate the potential defect in keratinocyte proliferation, we cultured post-transduced keratinocytes in a monolayer for two days, and calculated the number of cells at D1 and D2 (Fig. 3A). The ratio was similar whatever the condition (1.89 ± 0.26 for the sh-ctrl, 1.98 ± 0.27 for shPADI1_1, p = 0.4324 and 1.97 ± 0.43 for shPADI1_2, p = 0.6524). These data demonstrated that PAD1 deficiency did not alter keratinocyte proliferation.Fig. 3 Effect of PAD1 down-regulation on proliferation, differentiation, and barrier function. A Proliferation capacity of transduced NHEK by sh-ctrl (n = 11), shPADI1_1 (n = 11), and shPADI1_2 (n = 4). B LY permeability of sh-ctrl and shPADI1_1 RHEs (at 6 h, *p = 0.0317; at 24 h, *p = 0.0163; n = 7, RHEs; 2 banks). C Relative percentage of TEER for sh-ctrl, shPADI1_1, and shPADI1_2 RHEs (n = 6). D, E RT-qPCR analysis for D PADI1, PADI3, (pro)filaggrin (FLG), loricrin (LOR), involucrin (IVL) and for E transglutaminase (TGM) 1, 3 and 5 in sh-ctrl and shPADI1 RHEs. Data were pooled for statistical analysis: shPADI1_1 n = 7 to 13 and shPADI1_2 n = 2 from 5 banks. F–H Immunodetections of total protein extracts of sh-ctrl (C), shPADI1_1 (1), and shPADI1_2 (2) RHEs as indicated. Ladder sizes indicated on the right (kDa). PAD1 deficiency affects the epidermal barrier function of RHEs The effect of PAD1 down-regulation on Stratum corneum functionality was assessed using several complementary assays. First, inside-out permeability was investigated by measuring trans-epidermal water loss (TEWL). No significant differences were found between sh-ctrl and shPADI1_1 RHEs (Fig. S3A). In addition, the pH at the surface of RHEs was the same (equal to ~7.0) (Fig. S3B). Next, the outside-in permeability was investigated using the lucifer yellow (LY) assay (Fig. 3B). The amount of LY in the culture medium 6 and 24 h after its application on the top of RHEs was slightly but significantly increased by the down-regulation of PAD1 compared in the control condition; e.g., after 24 h, the mean concentrations were 29.11 ± 10.45 mM for sh-ctrl versus 35.74 ± 8.31 mM for shPADI1_1, p = 0.0163. Finally, mean trans-epidermal-electric resistance (TEER) values were drastically decreased in both shPADI1 (2205 ± 2090 ohm/cm² for shPADI1_1 and 1656 ± 992 ohm/cm² for shPADI1_2) compared to sh-ctrl RHEs (9657 ± 6977 ohm/cm², p < 0.0001, Fig. 3C). This result could be explained by modulations of mRNA and protein amounts of claudin-1 and occludin (Fig. S4A and S4D). Taken together, these results show that the barrier function is impaired in PAD1-deficient RHEs. PAD1 deficiency reduces the expression of transglutaminases To further decipher the effect of PAD1 deficiency on cell differentiation and corneocyte formation at the molecular level, mRNA of other markers of keratinocyte differentiation were analyzed by RT-qPCR in RHEs produced with keratinocytes from up to five donors. In shPADI1 RHEs, the relative profilaggrin and loricrin mRNA levels were slightly decreased (0.8 and 0.5 fold, respectively) whereas no significant variations were observed in the levels of involucrin, corneodesmosin, desmocollin-1 and desmoglein-1 (Fig. 3D and Fig. S4B and S4D). A significant decrease in transcripts was measured for transglutaminases 1, 3, and 5 (0.5, 0.2, and 0.4 fold, respectively) (Fig. 3E). Some of the corresponding proteins were immunodetected by Western blotting (Fig. 3F–H). In PAD1 deficient extracts, a clear reduction in the detection of filaggrin monomers was observed (29.31 ± 24.28% relative to controls) while profilaggrin remained similar to controls (105.44 ± 40.64%) (Fig. 3F). The levels of loricrin and involucrin detected were also equivalent (Fig. 3G). In agreement with the RT-qPCR data, the amount of the native form (higher band, ~72 kDa) of tranglutaminases 3 and 5 detected decreased in the two shPADI1 (Fig. 3H). Furthermore and as expected, the in situ transglutaminase 1 activity showed a peripheral pattern in the late differentiated keratinocytes but with no significant modulations (Fig. S5). PAD1 deficiency affects the shape of nuclei in the granular keratinocytes TEM images highlighted changes in the nuclear shape in the granular layer of shPADI1 RHEs compared with controls (Fig. 4A–C). As illustrated in Fig. S6, these deformations were scored using images corresponding to seven independent RHE productions (7 RHEs for sh-ctrl, 6 for shPADI1_1, and 1 for sh_PADI1_2). The score for shPADI RHEs was almost doubled, with a mean ± SD equal to 1.42 ± 1.18 for sh-ctrl versus 2.66 ± 1.10 for shPADI1 RHEs (p < 0.0001) (Fig. 4D). The distribution of scores for shPADI1_1 and shPADI1_2 RHEs were similar (Fig. 4E, F). Furthermore, deep invaginations in the nuclei were observed, as illustrated in Fig. S6E, F (black arrow), for 29.17% of sh-ctrl versus 68.75% for shPADI1 RHEs (p = 4.6561e-7). These data demonstrated the major impact of PAD1 deficiency on nuclei at the granular to cornified cell transition. In addition, vesicles around nuclei of granular keratinocytes were notable and were scored (Fig. 4G, H). A significant increase in vesicles was observed in PAD1 deficient RHEs compared to controls [scores equal to 0.75 ± 0.69 for sh-ctrl versus 1.12 ± 0.79 for shPADI1 RHEs (p = 0.00219)]. These observations led us to analyze the expression of four autophagic key markers at protein and mRNA levels. PAD1 deficiency did not appear to have a major impact on the expression of these markers (Fig. S4C, D).Fig. 4 Effect of PAD1 down-regulation on nuclear shape. A–C Representative illustrations of nuclear shape for sh-ctrl, shPADI1_1, and shPADI1_2 RHEs, respectively. B Deep nuclear deformations (or invaginations) indicated by black arrows. D Nuclear deformation scoring of sh-ctrl (109 nucleus) and shPADI1 (96 nucleus) RHEs, NHEK banks, n = 5; 7 independent experiments, n = 6 for sh-ctrl, shPADI1_1 and 1 for shPADI1_2. E Distribution of nuclear deformation scores for sh-ctrl (gray) and shPADI1_1 (black) RHEs. F Distribution of nuclear deformation scores for sh-ctrl (gray) and shPADI1_2 (black) RHEs. G, H Representative perinuclear vesicles (small circular white area) indicated by white arrows. Also in (B) and (C). I Representative images of perinuclear vesicles (white arrows) and deep invaginations (black arrows) of the nuclear envelope in RHE treated with 400 µM Cl-amidine (ClA-400). Scale bars, 2 µm. N, nucleus. In a previous work, we observed cytosolic and perinuclear vesicles in the differentiated keratinocytes of RHEs treated with Cl-amidine, an irreversible inhibitor of PADs [19]. We therefore performed new TEM observations of the Cl-amidine-treated RHEs, focusing on nuclei. Nuclear deformations were significantly increased, with scores of 1.93 ± 1.30 for controls versus 3.09 ± 0.99 for RHEs treated with 400 µM, and 2.62 ± 1.23 for RHEs treated with 800 µM of Cl-amidine (Fig. S7A). The distribution of the scores for nuclear deformation and for perinuclear vesicles were also modulated by the PAD inhibitor (Fig. 4I and Fig. S7A-D). Furthermore, after 24 h of treatment with Cl-amidine, the proliferation capacity of keratinocytes was similar to control cells (Fig. S8). Taken together, these results support a major role for PAD1 and deimination in the nucleophagy process during the late differentiation of keratinocytes in human epidermis, at the transition between granular to cornified layers. Discussion In this study, tridimensional RHEs produced in 50% relative humidity were first characterized. A fully differentiated and stratified epidermis was obtained ten days after exposure to an air-liquid interface, with characteristics similar to the normal human interfollicular epidermis. In this model, the filaggrin monomer was increasingly detected during the course of the epidermis reconstruction process. The expression of PAD1 and 3 was confirmed, and as expected, deiminated proteins were detected in the cornified layer. These data are in line with previous results showing that relative humidity of 30–50% promotes epidermal differentiation [20, 22]. Using this model, we showed that down-regulation of PAD1 with two distinct shRNAs caused almost total suppression in the overall protein deimination. Only one deiminated protein with a high molecular mass (> 250 kDa) remained, which may correspond to the known PAD3 target trichohyalin [9, 14, 16]. This demonstrates that PAD1 is the main isotype involved in epidermis protein deimination, in full agreement with the recent hypothesis of its major role in the regulation of filaggrin breakdown [20]. Further, this work showed that PAD1 deficiency caused a dysregulation of keratinocyte differentiation, with a thinner cornified layer and a decrease of KHG. The functional properties of shPADI1 RHEs were disturbed with a small increase in outside-in permeability (LY assay) and, a huge collapse of the TEER, but no significant impact on inside out permeability (TEWL values) and surface pH. This deleterious phenotype could be partially explained by the lower number of corneocyte layers and by tight junction defects, as shown by the modulation of claudin-1 and occludin expression. In shPADI1 RHEs, the amount of filaggrin monomers was lower. Transglutaminases 3 and 5 were drastically down-regulated at both the mRNA and protein levels. Although no clear impacts were detected for involucrin and loricrin, two major components of the cornified envelope [23], we suspect that less transglutaminases could affect the protein or lipid composition and properties of the envelopes. This is in line with the impact on RHEs of treatments with the PAD inhibitor, Cl-amidine [19] (Table 1). Conversely, when PAD1 expression and the protein deimination rate were induced in RHEs produced under low relative humidity, the expression of transglutaminase 5 and the number of corneocyte layers increased [20] (Table 1).Table 1 Characteristics of RHEs with impaired PAD1 activity. Dry atmosphere Cl-A inhibitora shPADI1 Characteristics Cau et al., 2017 Cau et al., 2019 This work Deimination/citrullination Hypercitrullination Hypocitrullination Hypocitrullination Deiminated proteins/PAD activity Increase (~5-8x) Decrease (~50%) Decrease (~ 85%) mRNA/PADI1 Increase (~12x) n.d. Decrease (~ 90%) mRNA/other PADIs No compensatory effectb n.d. No compensatory effectb PAD1 protein Increase (~4x) n.d. Decrease (~ 40%) RHE morphology Thicker SC Similar Thiner SC Keratinocyte proliferation Unchanged Unchanged Unchanged Transitional cells Not observed Increased number Not observed Keratohyalin granules Increase (~1.6x) n.d. Decrease (~ 0.7-0.5x) Profilaggrin Tendency to increase Decrease (~0.8x) Decrease (~ 0.8x) Filaggrin (monomer) Decrease Decrease (~ 0.8x) Decrease (~ 0.3x) Deiminated filaggrin Increase Decrease Decrease Other differentiation markers IVL (~1.6x), LOR (~0.6x) IVL unchanged n.d. IVL unchanged, LOR (~ 0.5x) Claudin-1 (tight junction protein) n.d. n.d. Increase (~ 1.5x) Desmosomal proteins CDSN (1.6x) n.d. CDSN, DSG1, DSC1 unchanged Transglutaminases TGM5 increase (~2x) TGM5 decrease (~0.04x) TGM1,3,5 decrease (~0.2-0.5x) Number of corneocyte layers Increase (~2x) Decrease Decrease (~0.5–0.7x) TEWL (inside-out) Decrease n.d. Unchanged Surface pH Decrease n.d. Unchanged LY permeability (outside-in) Unchanged n.d. Increase TEER n.d. n.d. Decrease (~80%) Autophagic markers (LC3B, p62, etc.). n.d. Increase Unchanged Autophagic process n.d. Disturbed Unchanged Nucleophagic process n.d. Disturbed Disturbed Parakeratosis Not observed n.d. Observed Perinuclear vesicles Not observed Increase Increase Nuclear deformation Not observed Increase Increase Phenotypes Inverse Similar aChloro-amidine (Cl-A) is an inhibitor that irreversibly binds to the cysteine of the active site of all PADs. bPADI3 ~, others undetectable; n.d., not determined. Finally, down-regulation of PAD1 in RHEs caused major changes in nuclear shape in the Stratum granulosum, with deep nuclear invaginations, and the occurrence of perinuclear vesicles. Nuclear removal is a well-known characteristic of granular keratinocyte to corneocyte transition. Autophagy and particularly nucleophagy has been recently proposed as involved mechanisms [24, 25]. In our study, no impact on autophagic mRNA markers was observed, suggesting that autophagy was not induced by down-regulation of PAD1 expression. We observed similar morphological modifications when PADs were inhibited with Cl-amidine (Table 1). However, in the latter case, expression of autophagic markers increased [19]. One possible explanation is that down-regulation by RNA interference acts during the full period of RHE production with a marked decrease in the deimination rate, whereas Cl-amidine treatment was performed only during the last 48 h leading to a reduction of only ~50% in PAD activity. Nevertheless in both cases, nucleophagy is clearly disturbed when PAD expression and/or activity decreases in RHEs. Nucleophagy during cornification is beginning to be deciphered at the cellular and molecular levels [24–27]. This requires Akt serine threonine kinase 1-dependent phosphorylation of lamins A/C inducing their subsequent dispersal and degradation, actin filament remodeling, and DNA degrading enzymes, namely DNase1L2 and DNase2. The present work is the first demonstration of the involvement of PAD1 and deimination, as a new post-translational modification, in this process. Lamin C carboxy terminus has been reported to be deiminated by PAD4 in nuclear fragmentation during apoptosis of several cell types [5, 28]. A 70 kDa perinuclear protein has also been proposed to be deiminated during in vitro apoptotic events in rat differentiated keratinocytes following calcium ionophore induction [29]. Deimination of similar nucleo-proteins and keratins (K1 and K10) by PAD1 could be involved in nuclear shape deformations in late keratinocyte differentiation steps. In conclusion, this work shows that knockdown of PADI1 in RHEs almost completely inhibits protein deimination. This leads to the collapse of TEER, a reduced number of corneocyte layers and KHG area, altered keratinocyte differentiation (especially transglutaminases, filaggrin and tight junction proteins) and impaired nucleophagic processes, highlighting the major role of PAD1 in epidermal homeostasis at the granular to cornified cell transition. Consequently, it seems to be important to take post-translational modifications such as deimination into account to better decipher keratinocyte physiology and skin diseases. Materials and methods shRNA lentiviral particles PAD1 down-regulation in keratinocytes was performed with two distinct shRNA targeting PADI1 (Table S2) and using the technical procedures previously reported [30]. Primary normal human keratinocyte culture, transduction, and production of RHEs Primary normal human epidermal keratinocytes (NHEK banks) were produced, with the approval of the French Minister of Research, as previously described [20]. For knockdown experiments, keratinocyte suspensions were infected by lentivirus particles containing either shPADI1_1, shPADI1_2, or the sh-ctrl at a multiplicity of infection of five, at 37 °C, 5% CO2. RHEs were produced at the air-liquid interface with controlled 50% relative humidity, and harvested at day 10 as previously described [19, 20, 31] and as detailed in Supplemental Material and at Fig. S1. Proliferation capacity During puromycin selection, after medium refreshment, ten images of adherent transfected keratinocytes (magnification ×4, EVOS digital inverted microscope, AMG) were randomly taken. Using the software imageJ (NIH, NY), the numbers of adherent keratinocytes were evaluated to calculate ratios between the cell numbers at days 2 and 1. Western blotting Total epidermal proteins were extracted in Laemmli sample buffer and immunodetected with antibodies listed in Table S3, as previously described [20]. Tissue staining and immunostaining RHEs were fixed in formaldehyde and embedded in paraffin. Five-micron thick sections were observed after a hematoxylin-eosin staining or immunodetected with the listed antibodies, as previously described [20]. Functional measurements TEWL and pH at the surface of RHEs were measured as previously described [20] and see Supplemental Material. LY (Sigma Aldrich, FR) permeability was quantified after 6 and 24 h of incubation as previously described [20] and Supplemental Material. Trans-epidermal-electric resistance (TEER) was evaluated using a Millicell ERS-2 apparatus, on 19 RHEs for shPADI1_1 (produced from 4 banks), on 6 RHEs for shPADI1_2 (1 bank), and compared to 23 RHEs for sh-ctrl (4 banks). Reverse-transcription and quantitative real-time PCR Total RNA extraction, reverse transcription, and real-time PCR were performed as previously reported [32]. RNA quality was confirmed for RIN > 8.5, on an Agilent 2100 Bioanalyzer, using a “RNAs 6000 Nano” kit according to the manufacturer’s instructions (Agilent, US), as previously described [19]. Quantitative PCR amplification was performed with the 7300 Real-Time PCR System (Applied Biosystems, Foster City, CA) using the “Sybr quantitative PCR SuperMix W/ROX” (Invitrogen Life Technologies, Fr.). The specific primer pairs are listed in Table S4. Transmission electron microscopy: Stratum corneum cell layers and nuclear scoring RHEs were processed on an HT7700 electron microscope (Hitachi, Tokyo, Japan), as previously described [19, 20]. The number of Stratum corneum cell (corneocyte) layers was quantified in at least three independent locations, and compared in sh-ctrl versus shPADI1_1 (n = 6 banks) or shPADI1_2 (n = 2 banks). Nuclear deformation was scored on a 0 to 4 scale, using images at ×2500 magnification (Fig. S6A–E). The absence or presence of at least one deep invagination of the nuclear envelope was scored 0 or 1, respectively (Fig. S6E, F). The perinuclear vesicles were scored 0, 1, or 2 according to their absence, abundance, and size. Statistical analysis A Shapiro-Wilk normality test was performed, and differences were evaluated using a parametric Student’s t-test, Welch’s t-test, or non-parametric Wilcoxon’s test using the GraphPad Prism (V 9.3.1). The statistical significance threshold was set at 0.05 (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001). Supplementary information Supplementary information final version Original Data File Supplementary information The online version contains supplementary material available at 10.1038/s41420-023-01509-8. Acknowledgements We are indebted to Dr. Pascal Descargues (Genoskin, Toulouse, France) for the human skin samples. We acknowledge all members of the electron microscopy (Toulouse University), histopathology (CREFRE Inserm US006, Toulouse), and cellular imaging (Toulouse Rio Imagerie, INFINITy Institute) facilities, and Corinne Rolland of the IRSD institute of Toulouse for her help with the Varioskan apparatus. Author contributions Conceptualization of the research approaches: M-CM and MS; biological data acquisitions and technical investigations: CA, CP, LC, DG, HY, MG, and M-CM; writing—preparation of the original draft: CA, LC, MS, and M-CM; writing—review and editing: M-CM, MS; literature research: CA, LC, JB, MS, and M-CM; funding acquisition: MS. All authors read and approved the final paper. Funding This work was supported by grants from Toulouse University, INSERM and CNRS, by the French Society for Dermatology (SFD), the French Society for Dermatological Research (SRD) and by the French Ministry of Research. Data availability The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request. Competing interests The authors declare no competing interests. The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. Ethics statement This manuscript is the authors’ own original work, which has not been published elsewhere. This study included human primary cells from informed consenting donors, and obtained according to procedures approved by the French Ministry of Research. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. ==== Refs References 1. Chavanas S Méchin M-C Takahara H Kawada A Nachat R Serre G Comparative analysis of the mouse and human peptidylarginine deiminase gene clusters reveals highly conserved non-coding segments and a new human gene, PADI6 Gene 2004 330 19 27 10.1016/j.gene.2003.12.038 15087120 2. Minato T Unno M Kitano T Evolution of S100A3 and PAD3, two important genes for mammalian hair Gene 2019 713 143975 10.1016/j.gene.2019.143975 31302167 3. Méchin M-C Sebbag M Arnaud J Nachat R Foulquier C Adoue V Update on peptidylarginine deiminases and deimination in skin physiology and severe human diseases Int J Cosmet Sci. 2007 29 147 68 10.1111/j.1467-2494.2007.00377.x 18489346 4. Méchin MC Takahara H Simon M Deimination and peptidylarginine deiminases in skin physiology and diseases Int J Mol. Sci 2020 21 566 10.3390/ijms21020566 31952341 5. Briot J Simon M Méchin MC Deimination, intermediate filaments and associated proteins Int J Mol Sci 2020 21 8746 10.3390/ijms21228746 33228136 6. Yu K Proost P Insights into peptidylarginine deiminase expression and citrullination pathways Trends Cell Biol. 2022 32 746 61 10.1016/j.tcb.2022.01.014 35197210 7. Yang L Yan Y Emerging roles of post-translational modifications in skin diseases: current knowledge, challenges and future perspectives J Inflam Res 2022 15 965 75 10.2147/JIR.S339296 8. Nachat R Méchin M-C Takahara H Chavanas S Charveron M Serre G Peptidylarginine deiminase isoforms 1-3 are expressed in the epidermis and involved in the deimination of K1 and filaggrin J Invest Dermatol. 2005 124 384 93 10.1111/j.0022-202X.2004.23568.x 15675958 9. Ü Basmanav FB Cau L Tafazzoli A Méchin MC Wolf S Romano MT Mutations in three genes encoding proteins involved in hair shaft formation cause uncombable hair syndrome Am J Hum Genet. 2016 99 1292 1304 10.1016/j.ajhg.2016.10.004 27866708 10. Malki L Sarig O Romano MT Méchin MC Peled A Pavlovsky M Variant PADI3 in central centrifugal cicatricial alopecia N. Engl J Med. 2019 380 833 41 10.1056/NEJMoa1816614 30763140 11. Guerrin M Ishigami A Méchin MC Nachat R Valmary S Sebbag M cDNA cloning, gene organization and expression analysis of human peptidylarginine deiminase type I Biochem J. 2003 370 167 74 10.1042/bj20020870 12416996 12. Méchin MC Enji M Nachat R Chavanas S Charveron M Ishida-Yamamoto A The peptidylarginine deiminases expressed in human epidermis differ in their substrate specificities and subcellular locations Cell Mol Life Sci 2005 62 1984 95 10.1007/s00018-005-5196-y 16091842 13. Coudane F Méchin MC Huchenq A Nachat R Ishigami A Adoue V Deimination and expression of peptidylarginine deiminases during cutaneous wound healing in mice Eur J Dermatol 2011 21 376 84 10.1684/ejd.2011.1394 21697043 14. Nachat R Méchin MC Charveron M Serre G Constans J Simon M Peptidylarginine deiminase isoforms are differentially expressed in the anagen hair follicles and other human skin appendages J Invest Dermatol. 2005 125 34 41 10.1111/j.0022-202X.2005.23763.x 15982300 15. Senshu T Kan S Ogawa H Manabe M Asaga H Preferential deimination of keratin K1 and filaggrin during the terminal differentiation of human epidermis Biochem Biophys Res Commun 1996 225 712 9 10.1006/bbrc.1996.1240 8780679 16. Tarsca E Marekov LN Andreoli J Idler WW Candi E Chung SI The fate of trichohyalin. Sequential post-translational modifications by peptidyl-arginine deiminase and transglutaminases J Biol Chem 1997 272 27893 901 9346937 17. Hsu C-Y Henry J Raymond A-A Méchin M-C Pendaries V Nassar D Deimination of human filaggrin-2 promotes its proteolysis by calpain 1 J Biol Chem. 2011 286 23222 33 10.1074/jbc.M110.197400 21531719 18. Suga H Oka T Sugaya M Sato Y Ishii T Nishida H Keratinocyte proline-rich protein deficiency in atopic dermatitis leads to barrier disruption J Invest Dermatol. 2019 139 1867 75.e7 10.1016/j.jid.2019.02.030 30905808 19. Cau L Takahara H Thompson PR Serre G Méchin MC Simon M Peptidylarginine deiminase inhibitor Cl-amidine attenuates cornification and interferes with the regulation of autophagy in reconstructed human epidermis J Invest Dermatol. 2019 139 1889 97 10.1016/j.jid.2019.02.026 30878672 20. Cau L Pendaries V Lhuillier E Thompson PR Serre G Takahara H Lowering relative humidity level increases epidermal protein deimination and drives human filaggrin breakdown J Dermatol Sci 2017 86 106 13 10.1016/j.jdermsci.2017.02.280 28242341 21. Padhi A Srivastava A Ramesh A Ehrström M Simon M Sonkoly E IL-22 downregulates peptidylarginine deiminase-1 in human keratinocytes: adding another piece to the IL-22 puzzle in epidermal barrier formation J Invest Dermatol. 2022 142 333 42.e6 10.1016/j.jid.2021.07.155 34352263 22. Sun R Celli A Crumrine D Hupe M Adame LC Pennypacker SD Lowered humidity produces human epidermal equivalents with enhanced barrier properties. Tissue Eng Part C Meth 2015 21 15 22 10.1089/ten.tec.2014.0065 23. Candi E Schmidt R Melino G The cornified envelope: a model of cell death in the skin Nat Rev Mol Cell Biol. 2005 6 328 40 10.1038/nrm1619 15803139 24. Rogerson C Bergamaschi D O’Shaughnessy RFL Uncovering mechanisms of nuclear degradation in keratinocytes: a paradigm for nuclear degradation in other tissues Nucleus 2018 9 56 64 10.1080/19491034.2017.1412027 29205081 25. Akinduro O Sully K Patel A Robinson DJ Chikh A McPhail G Constitutive autophagy and nucleophagy during epidermal differentiation J Invest Dermatol 2016 136 1460 70 10.1016/j.jid.2016.03.016 27021405 26. Park Y-E Hayashi YK Bonne G Arimura T Noguchi S Nonaka I Autophagic degradation of nuclear components in mammalian cells Autophagy 2009 6 785 804 27. Rogerson C Wotherspoon DJ Tommasi C Button RW O’Shaughnessy RFL Akt1-associated actomyosin remodelling is required for nuclear lamina dispersal and nuclear shrinkage in epidermal terminal differentiation Cell Death Differ 2021 28 1849 64 10.1038/s41418-020-00712-9 33462407 28. Tanikawa C Espinosa M Suzuki A Masuda K Yamamoto K Tsuchiya E Regulation of histone modification and chromatin structure by the p53-PADI4 pathway Nat Commun. 2012 14 676 10.1038/ncomms1676 29. Mizoguchi M Manabe M Kawamura Y Kondo Y Ishidoh K Kominami E Deimination of 70-kD nuclear protein during epidermal apoptotic events in vitro J Histochem Cytochem 1998 46 1303 9 10.1177/002215549804601110 9774629 30. Pendaries V Malaisse J Pellerin L Le Lamer M Nachat R Kezic S Knockdown of filaggrin in a three-dimentional reconstructed human epidermis impairs keratinocyte differentiation J Invest Dermatol 2014 134 2938 46 10.1038/jid.2014.259 24940654 31. Frankart A Malaisse J De Vuyst E Minner F de Rouvroit CL Poumay Y Epidermal morphogenesis during progressive in vitro 3D reconstruction at the air-liquid interface Exp Dermatol. 2012 21 871 5 10.1111/exd.12020 23163654 32. Méchin M-C Coudane F Adoue V Arnaud J Duplan H Charveron M Deimination is regulated at multiple levels including auto-deimination of peptidylarginine deiminases Cell Mol Life Sci 2010 67 1491 1503 10.1007/s00018-010-0262-5 20111885