(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



Initiation of wound healing is regulated by the convergence of mechanical and epigenetic cues [1]

['Tanay Bhatt', 'Ifom-Instem Joint Research Laboratory', 'Center For Inflammation', 'Tissue Homeostasis', 'Institute For Stem Cell Science', 'Regenerative Medicine', 'Bangalore', 'National Centre For Biological Sciences', 'Rakesh Dey', 'Akshay Hegde']

Date: 2022-09

Wound healing in the skin is a complex physiological process that is a product of a cell state transition from homeostasis to repair. Mechanical cues are increasingly being recognized as important regulators of cellular reprogramming, but the mechanism by which it is translated to changes in gene expression and ultimately cellular behavior remains largely a mystery. To probe the molecular underpinnings of this phenomenon further, we used the down-regulation of caspase-8 as a biomarker of a cell entering the wound healing program. We found that the wound-induced release of tension within the epidermis leads to the alteration of gene expression via the nuclear translocation of the DNA methyltransferase 3A (DNMT3a). This enzyme then methylates promoters of genes that are known to be down-regulated in response to wound stimuli as well as potentially novel players in the repair program. Overall, these findings illuminate the convergence of mechanical and epigenetic signaling modules that are important regulators of the transcriptome landscape required to initiate the tissue repair process in the differentiated layers of the epidermis.

Funding: This work was supported by grants to C.J. from the Department of Biotechnology of the Government of India (BT/PR8738/AGR/36/770/2013 and DBT/PR32539/BRB/ 10/1814/2019); and inStem Core funds. TB was supported by PhD scholarship from Council of Scientific & Industrial Research (CSIR). Animal work in the NCBS/inStem Animal Care and Resource Center was partially supported by the National Mouse Research Resource (NaMoR) grant # DBT/PR5981 /MED/31/181/2012;2013-2016 & 102/IFD/SAN/5003/2017-2018 from the Department of Biotechnology of the Government of India. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Introduction

The wound healing program in an epithelial tissue is fundamentally a product of cell state transitions from homeostasis to a repair program. In particular, cutaneous wound healing in the adult is an intricately regulated system wherein keratinocytes and many other cell lineages exhibit their plasticity as they undergo reprogramming, to carry out otherwise dormant functions, to rebuild the damaged skin. Many of the phenomena that occur in the repair process in adult skin are, in fact, reminiscent of cellular events that operate during fetal development [1]. At the other extreme, inappropriate activation of these repair processes can manifest as tissue pathology, which forms the foundation of the perception of diseases with a “wound signature” [2]. The question that arises is how the whole scale changes in gene expression are accomplished in order to facilitate this cellular reprogramming.

Recently, epigenetic regulators have emerged as a vital component capable of transiently rewiring the cell’s transcriptional program to mediate the continual regeneration of the mouse epidermis [3,4]. This mode of gene regulation operates at multiple levels ranging from histone and DNA modifications, chromatin remodeling, and activity of various subtypes of RNA species such as non-coding RNAs and micro-RNAs (miRNAs) [5,6]. These epigenetic mechanisms can thus have a profound impact on the transcriptional landscape of the cell and can easily be envisioned to participate in the transient activation or repression of approximately 1,000 genes that are required for wound closure [7]. Circumstantial evidence in support of a role for epigenetics in tissue repair comes from reports of the dynamic expression of several epigenetic regulators following injury to the skin. For instance, Ezh2, Suz12, and Eed, which are components of the polycomb repressive complex 2 (PRC2), are down-regulated, whereas the histone methylases JMJD3 and Utx are up-regulated upon tissue damage and all return to homeostatic levels upon the completion of wound closure [8]. While the description of various epigenetic players in epidermal homeostasis and wound healing are reported, the identity and function of their upstream regulators are, to a large extent, absent in the literature.

An intriguing candidate for an upstream regulator in a highly tensile tissue such as the epidermis are mechanical cues. The epidermis is a stratified epithelium comprised of a basal layer of proliferation competent keratinocytes and suprabasal layers of differentiated cells glued together via intercellular adhesion complexes that partly endows the tissue with its barrier function. In different cell types, changes in mechanical tension have been documented to induce the nuclear translocation of important transcription factors—a notable example of which is YAP/TAZ that has proliferation stimulating gene targets [9]. Many studies, including those on epidermal homeostasis and wound healing, have primarily focused on the changes in gene expression in proliferating cells [10,11]. On the other hand, differentiated cells, such as the suprabasal keratinocytes near the surface of the epidermis, have largely been relegated to bystander status. In spite of this, a few reports suggest that these neglected pools of differentiated cells are not inert in the cellular crosstalk that mediates the early responses of the tissue to injury. In particular, the uppermost layer of differentiated keratinocytes in the epidermis expresses caspase-8 that has a non-canonical role in regulating the wound healing program. We previously demonstrated that the down-regulation of caspase-8 is a natural phenomenon upon application of an excisional wound to the mouse skin [12]. This down-regulation is particularly relevant as genetically ablating caspase-8 in the epidermis is sufficient to induce a wound healing response even in the absence of any damage to the organ. In addition, the down-regulation of caspase-8 in the upper, differentiated layer of the epidermis mediates signaling networks to incite epithelial stem cell proliferation in the epidermis [12] and the hair follicle [13,14] to fuel wound closure. We have thus used the down-regulation of caspase-8 as a cellular biomarker to identify the higher order regulatory machinery that reprograms the cell to enter the wound healing process in differentiated keratinocytes, which are emerging as an important participant in the tissue repair program.

Wound induce down-regulation of caspase-8 RNA correlates with the degree of promoter methylation Previously, we have established the importance of the down-regulation of caspase-8 RNA in both physiological (wound healing [12]) as well as pathological (atopic dermatitis [15] and psoriasis [16]) scenarios. The mechanisms responsible for this down-regulation, however, remain unknown. Uncovering the regulatory machinery of caspase-8 RNA also holds the promise of understanding the process by which cells transition from a state of homeostasis to repair. Moreover, it can provide potential new therapeutic targets for common inflammatory skin diseases where this regulation is perturbed. RNA down-regulation can be achieved either via blocking the synthesis and/or active degradation. In order to distinguish between these 2 possibilities, we determined the half-life of caspase-8 in homeostasis compared to wound conditions. In differentiated primary epidermal keratinocytes, we observed that the half-life of caspase-8 mRNA under homeostatic conditions in vitro is approximately 2 hours (S1A Fig). In an in vitro scratch wound assay with multiple scratches, the level of caspase-8 RNA is significantly reduced by 8 hours (Fig 1A). Since the reduction of caspase-8 is faster under homeostatic conditions compared to the wound healing context, merely blocking RNA synthesis can achieve the reduction of caspase-8 mRNA and initiate the downstream wound healing response. Interestingly, the reduction caspase-8 RNA is localized in cells near the front of the scratch wound in vitro (Figs 1B and S1B). In situ hybridization of caspase-8 RNA demonstrates that the down-regulation can clearly be visualized in the cells immediately adjacent to the leading edge of a single scratch wound as early as 4 hours post wounding. By 8 hours post wounding, the caspase-8 RNA is down-regulated in about 3 to 4 cell layers from the wound front. These findings are consistent with our observation in excisional wounds on the back skin of mice where the decrease of caspase-8 RNA is visible as early as 4 hours in the wound proximal region (Figs 1C and S1C). Together, these results suggest that simply blocking transcription post injury is sufficient to down-regulate caspase-8. We hypothesized that the block in caspase-8 RNA synthesis is achieved through promoter methylation, which is consistent with previous reports documenting the same phenomenon in a variety of cancer cells through the hypermethylation of regulatory DNA sequence [17,18]. To understand whether this process in cancer cells is an aberration of the physiological healing program, we have assessed the methylation status of important regulatory sequences in the caspase-8 promoter, namely the CpG loci and SP1 binding sites (S1D Fig) [19]. Analysis of methylation of SP1 sites and other CpG loci reveals a time-dependent increase of promoter methylation in a sheet of differentiated epidermal keratinocytes subjected to multiple scratch wounds (Fig 1D). This progressive increase in the methylation of the caspase-8 promoter correlates well with the kinetics of the decrease in caspase-8 RNA (Fig 1A–1C). This suggests DNA methylation may play a critical role in regulating the wound healing response. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 1. Kinetics of caspase-8 promoter methylation and expression. (A) Levels of caspase-8 mRNA at different time points post-scratch wound (fold change) (n = 4). (B) In vitro ISH of caspase-8 mRNA showing its levels at scratch margins over time [scale = 10 μm]. (C) In vivo ISH of caspase-8 mRNA showing its levels at wound proximal and distal regions over time (dotted line represents basement membrane, Epi = Epidermis, Der = Dermis) [scale = 20 μm]. (D) Bisulphite sequencing of caspase-8 promoter proximal region (265 bp) shows methylation status of 10 individual CpG sites (columns) from 10 cloned PCR products (rows) at various time points post-scratch wound. Percentage value denotes the percent methylation for each group of CpG sites over time (refer S1D Fig for the sequenced region and primer sites, n = 5 with 2 technical replicates). (Data are shown as mean ± SEM, P-values were calculated using 1-way ANOVA with Dunnett’s test and 2-tailed t test (A), *** P ≤ 0.001, ns = P > 0.05). Data underlying the graphs can be found in Fig 1A of S1 Raw Data. https://doi.org/10.1371/journal.pbio.3001777.g001

DNMT3a directly regulates caspase-8 expression We further explored whether the de novo DNA methylation of caspase-8 promoter is the result of DNMT3a’s direct binding to this region (S1B Fig). This was accomplished with the use of chromatin immunoprecipitation (ChIP) to assess the level of DNMT3a occupancy on the caspase-8 promoter pre- and post-scratch wound. We found that scratch wounds lead to the higher occupancy of DNMT3a on caspase-8 promoter, which is not seen in the case of DNMT3b (Fig 3A). To understand the functional relevance of DNMT activity in maintaining caspase-8 levels, we pre-treated the differentiated keratinocytes with a generic DNMT inhibitor (5-Aza-2′-deoxycytidine). We observed that the inhibitor treated cells were unable to down-regulate caspase-8 mRNA in a scratch wound assay (S3A Fig). To specifically assess the role of DNMT3a, we performed shRNA-mediated knockdown of DNMT3a (S3B Fig). Compared to the scrambled RNA controls, keratinocytes with reduced DNMT3a expression were unable to down-regulate caspase-8 in response to scratch wound (Fig 3B). We further analyzed whether failure of caspase-8 mRNA down-regulation was due to the absence of promoter methylation. Indeed, scratch wounded keratinocytes, transduced with DNMT3a shRNA, showed significantly reduced DNA methylation pattern on the caspase-8 promoter compared to scrambled RNA control (Fig 3C). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 3. Involvement of DNMT3a and histone modification in regulating caspase-8 expression. (A) ChIP-qPCR analysis to check DNMT3a and DNMT3b occupancy at caspase-8 promoter in control and scratch wounded keratinocytes (n = 3). (B) qPCR analysis of caspase-8 mRNA in scratch wounded keratinocytes, transduced with either scrambled RNA or DNMT3a shRNA (n = 3). (C) DNA methylation status of caspase-8 promoter in scratch wounded keratinocytes, transduced with either scrambled RNA or DNMT3a shRNA. (D) ChIP-qPCR analysis of H3K9ac, H3K4me3, H3K9me3, and H3K27me3 at caspase-8 promoter in control and scratch wounded keratinocytes (n = 3). (E) Effect of DNMT3a down-regulation on in vitro wound healing assay (n = 3). (Data are shown as mean ± SEM, P-values were calculated using 2-tailed t test (A, B, D), * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001, ns = P > 0.05.) Data underlying the graphs can be found in Fig 3A, 3B, 3D, and 3E of S1 Raw Data. ChIP, chromatin immunop recipitation; DNMT3a, DNA methyltransferase 3A. https://doi.org/10.1371/journal.pbio.3001777.g003 Promoter activities are often dependent on the associated histone modifications. These histone marks generally guide the DNA methylation at a particular genic region and vice-a-versa [21–23]. DNMT3a occupancy and activity has also been shown to be influenced by the methylation status of certain lysine (K) residues on the histone 3 (H3) tail [22,24]. To investigate the core machinery required for DNMT3a-mediated methylation on the caspase-8 promoter, we assessed several activation and repression histone marks in scratch wounded keratinocytes (Fig 3D). We observed that 2 transcriptional activation marks, H3K9ac and H3K4me3, are decreased at the caspase-8 promoter. On the other hand, the H3K9me3 mark, which is associated with transcriptional repression, was significantly increased at the caspase-8 promoter following wounding. Interestingly, another classical repressive mark, H3K27me3, did not show a significant change. It is possible that the caspase-8 proximal promoter is another example of a bivalent promoter [25] having both activation (H3K9ac and H3K4me3) and repression (H3K27me3) marks. In this scenario, then, wound-mediated repression of caspase-8 is achieved via reduction of both H3K9ac and H3K4me3 along with an increase in the H3K9me3 mark and DNMT3a occupancy. These results establish the mechanism by which DNMT3a localizes to the caspase-8 promoter. An outstanding question is whether DNMT3a is required for a proper wound healing response. To address this issue, we tested the effect of the knockdown of DNMT3a in a scratch wound assay (Fig 3E). We found that keratinocytes with decreased DNMT3a exhibited an impaired wound closure response, thereby illustrating the necessity of this methyltransferase in the proper repithelialization of an in vitro wound.

Effect of cellular tension on DNMT3a localization and caspase-8 expression We observed that caspase-8 down-regulation and DNMT3a nuclear localization initiate at the edge of wound site (Figs 1 and 2). Given that these are early responses to injury, understanding the mechanistic basis of this phenomenon can provide insights into the broader process of cellular wound sensing. The keratinocytes in the epithelial sheet are strongly connected to each other and an event of injury will result in the sudden relaxation in that tension, particularly in the cells at the boundary of the wound. Interestingly, the expanding number of cells exhibiting the down-regulation of caspase-8 RNA in the scratch wound assay over time (Fig 1A) closely parallels the changes in traction force previously reported for the collective cell migration of an epithelial sheet following a scratch wound [26]. We therefore investigated whether release of tension, caused by the severing of the epithelial sheet, can impact DNMT3a subcellular localization and subsequently caspase-8 expression. As shown in S4A Fig, modulation in cellular tension can be achieved via targeting the components of the adherens junction, which are known to play a role in generating and maintaining cellular tension [27,28]. We observed that tension release by disrupting calcium-dependent E-cadherin junctions via EGTA treatment resulted in the nuclear localization of DNMT3a (Figs 4A and S4B). Similarly, releasing cellular tension endowed by nonmuscle myosin II (NM-II) with the pharmacological inhibitor of NMII, blebbistatin, induced the DNMT3a’s nuclear translocation from the cytosol (Figs 4A and S4B). Furthermore, we examined the effect of blocking release of cellular tension in a scratch wounded sheet of epidermal keratinocytes. The release of tension was blocked by pre-treating keratinocytes with calyculin-A, which inhibits myosin light-chain phosphatase, thereby maintaining the active state of NMII [29]. The treatment of keratinocytes with calyculin-A prior to scratch wounding blocked the nuclear translocation of DNMT3a that was observed in cells treated with vehicle control (Figs 4B and S4C). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 4. Effect of cellular tension on DNMT3a localization and caspase-8 expression. (A) Effect of EGTA and blebbistatin on the localization of DNMT3a. (B) Effect of scratch wound DNMT3a localization in presence and absence of calyculin-A. (C) Effect of various matrix stiffness on the localization of DNMT3a. (D) Fold change in the levels of caspase-8 mRNA as a result of varios pharmacological and mechanical approaches of tension modulation (n = 4), [scale bar = 20 μm]. (Data are shown as mean ± SEM, P-values were calculated using 2-tailed t test (D), * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001, ns = P > 0.05). Data underlying the graphs can be found in Fig 4D of S1 Raw Data. DNMT3a, DNA methyltransferase 3A. https://doi.org/10.1371/journal.pbio.3001777.g004 In addition to a pharmacological approach, we also modulated cellular tension by altering the substrate stiffness on which the keratinocytes were growing. This was accomplished by utilizing polyacrylamide gels of various stiffness, which would alter cellular tension. We observed that differentiated keratinocytes seeded on “soft” matrices ranging from 10 kPa to 40 kPa mostly harbored DNMT3a in the nuclei (Figs 4C and S4D). However, cells grown on a “stiffer” matrix (100 kPa) predominantly showed a cytoplasmic localization of DNMT3a. We then evaluated whether DNMT3a’s dynamic localization in response to pharmacological and mechanical alterations in cellular tension has any transcriptional consequences. We observed that in all the scenarios where DNMT3a nuclear localization was favored (scratch wounds, EGTA/blebbistatin treatment, soft substrates), caspase-8 RNA was down-regulated compared to their respective controls (Fig 4D). On the other hand, inhibition of DNMT3a’s nuclear localization (via calyculin-A or a stiff substrate) resulted in the failure of caspase-8 down-regulation in spite of a scratch wound. These results suggest a correlation between DNMT3a localization and changes in tensile forces. It should be noted that these interventions may have other effects on the cell in addition to modulating cellular tension, and thus we cannot rule out additional pathways leading to cellular reprograming via epigenetic means.

[END]
---
[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3001777

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.