Feeling the Heat. Mapping the Epigenetic Modifications of Histone during Burn Wound Healing

Abstract

Burn injuries are observed throughout a wide range of ages, with over 1.1 million Americans suffering burns yearly, and half of these require hospitalization. Epigenetic modifications are fast-acting mechanisms that allow the human body to respond and adapt to environmental changes, including burn injuries. There is a lack of understanding of the epigenetic role during burn-induced tissue repair. Here, we characterize the histone modifications that follow burn injury, aiming at future pharmacological intervention using drugs capable of targeting epigenetic events. A clinically relevant porcine burn model was used to recapitulate the skin healing process after the burn. Isolated skin tissues at different time points were used to detect the acetylation levels of histones H3K27, H4K5, H4K8, and H4K12 as significant players of gene transcription using MetaXpress High-Content Imaging Analysis. We observed that the acetylation of histones is dynamically adjusted throughout healing, and its modifications are uniquely expressed according to the anatomical location and time of healing. We also observed that histone H4K5 is the most widely expressed during healing, followed by histone H3K27. We observed that histones expressed in intact skin tissue adjacent to the burn site could sense the burn injury by changing its histone acetylation pattern compared to control skin from uninjured and distant skin. Using a clinically relevant animal model, we have generated a comprehensive landscape of epigenetic modifications during burn healing. Our data will help us identify novel epi-drugs capable of manipulating histone modifications during healing to accelerate the healing process.

INTRODUCTION

Histone modifications constitute a series of epigenetic events that actively modify gene transcription, presenting a fast-paced control over gene expression triggered by environmental exposure.^1,2 Similar to the dynamic epigenetic modifications observed in cancer, the maintenance of tissue homeostasis also depends on histones’ epigenetic modifications.^3,4 From a translational perspective, a better understanding of histone modifications that occur during tissue integrity loss or disease development is needed. Epi-drugs constitute a new class of agents capable of interfering with different epigenetic mechanisms, including histone modifications, thereby influencing gene transcription.⁵ In cancer, we have shown that different epi-drugs can effectively sensitize tumors to chemotherapeutic agents while interfering with the maintenance of cancer stem cells.^6–8 The disruption of cancer stem cells directly impacts the overall tumor resistance to chemotherapy. Interestingly, although we are learning about the impact of cancer epigenetics on tumor development and the development of drug resistance to therapy, we still lack a basic understanding of how histone modifications react to injury, followed by tissue repair.

Here, we decided to characterize the epigenetic events that follow tissue burn and the epithelial repair process. Toward this goal, we used a well-defined and human-relevant animal model for burn research. We characterized the epigenetic modifications of histones, focusing on acetylation changes of different lysines. Histones H3 and H4 play a major role in gene transcription when acetylated at lysine 27 from histone H3 and lysines 5, 8, and 12 from histone H4.

We found that histone acetylation dynamically responds to burn injury. Surprisingly, we observed that the first signs of histone modifications from a burn injury are found on intact skin cells adjacent to the burn area but not in contact with the injured tissue. We also observed that some histones are well expressed throughout tissue healing. In contrast, others are acetylated only at a specific time or anatomical location during tissue healing.

By identifying and understanding the histone acetylation changes during burn healing, we can develop therapeutic strategies capable of interfering with burn wounds to accelerate healing.

METHODS

Experimental animal handling

All animal work was previously approved by the University of Michigan Institutional Animal Care & Use Committee under protocol number PRO00007135. All procedures followed the National Institutes of Health guidelines for laboratory animal care and use.

Briefly, we used 12- to 14-week-old female Yorkshire pigs weighing between 22 and 30 kg (Michigan State University Swine Resource Center; Lansing, MI). All pigs were kept in 12 h of light/dark cycles and, upon receiving, were acquainted for one week at the University of Michigan facility. Pigs were fed a porcine diet (Lab Diet 5801; PMI Nutrition, IN) and water ad libitum. Before surgical interventions, pigs fasted overnight in preparation for anesthesia. Anesthesia was induced using intramuscular injection of 2.5-3 mg/kg of Telazol (Tiletamine/Zolazepam, Zoetis, Parsippany, NJ) and 2.2 mg/mL of xylazine and maintained by isoflurane inhalation (VetOne, Boise, Idaho) under an oxygen flow of 2 L/min. Animals were constantly monitored during anesthesia for distress signs and hypothermia using a rectal thermometer while kept warm using a water heater blanket. After the procedure, pigs were monitored for discomfort while receiving intramuscular buprenorphine at a concentration of 0.01 mg/kg, complemented with transdermal delivery of buprenorphine (Butrans patch, Purdue Pharma, Stamford, CT) at a 5 µg/h rate for 7 days.

Burn wound model

Each pig was projected to have 10 dorsal burn sites measuring 5 × 5 cm each. Each wound was consistent with superficial partial thickness burn specifically created using copper blocks weighing 530 g and heated to 80°C in water.^9,10 Each copper block was immersed in hot water for 30 min to achieve the desired temperature. Copper blocks were maintained in place using an aluminum rod attached to the block and held in place for 40 s without pressure.¹¹

Histological evaluation

After each experiment, pigs were euthanized, and skin samples were collected (day 21 after burn). Eight samples at all time points (0, 6, 12, 18, and 21 days) were collected along with a healthy tissue margin of 3 cm around the original burn wound size, and all tissues were fixed using 10% paraformaldehyde for 48 h or frozen. Fixed tissues were further embedded in paraffin and processed for histological sections averaging 3-4 μm of thickness.

All tissue samples were deparaffinized in xylene substitute solution and rehydrated in a descending ethanol concentration. Tissue sections were either stained for hematoxylin and eosin (H&E) or processed for immunofluorescence. Histologically, the wound margin is identified by morphometry and characterized by an abrupt loss of continuity of the epidermis at the burn area. Re-epithelialization areas were defined as the continuity of epithelial cells migrating over the wound bed, typically consisting of 1-2 layers of cells positioned over the granulation tissue at the wound bed. Furthermore, the re-epithelialization area does not contain hair follicles, only seen at the wound margin. The wound edge is populated by highly proliferative cells.¹² Two experienced pathologists evaluated all samples.

Immunofluorescence

All samples were processed for immunofluorescence after antigen retrieval using citric acid buffer and blocked for unspecific binding using 3% (w/v) bovine serum albumin (BSA) in 0.5% (v/v) Triton X-100 phosphate-buffered saline (PBS). Immunofluorescence staining was further processed using the primary antibodies anti-histone specific for histone H3 lysine 27 (H3K27), histone H4 lysine 9 (H4K9), and 12 (H4K12) (Cell Signaling Technology, Danvers, MA), and histone H4 lysine 5 (H4K5) (Thermal Scientific, Waltham, MA). All antibodies are reactive to porcine tissues. After washing slides in PBS, a secondary antibody (Alexa Fluor 568) was applied, followed by Hoechst 3342 staining for DNA content (Sigma Aldrich, Burlington, MA).

Data quantification, representation, and statistical analysis

Immunofluorescence images were captured using a Nikon Eclipse 80i microscope (Nikon, Melville, NY, USA) equipped with a QImaging ExiAqua monochrome digital camera (Teledyne Photometrics, Tucson, AZ) and quantified using MetaXpress High-Content Imaging Analysis software (Molecular Devices, San Jose, CA). All immunofluorescence images were taken using a 20x Nikon FL objective. Individual pictures of the same field size were collected for each anatomical location and imported and further analyzed using MetaXpress High-Content Imaging Analysis software. Radar graphics displaying changes in histone acetylation levels were generated using Flourish studio software (Kiln Enterprises Ltd, London, UK). Statistical analyses were performed using GraphPad Prism 9.5.1 (GraphPad Software, San Diego, CA). Statistical differences between histone expression levels from normal porcine skin and skin adjacent to the burn wound and day 0 were performed using an unpaired Student’s t-test. Histone levels for days 6-21 were analyzed using one-way ANOVA followed by Tukey post-test with 95% confidence intervals. Data are expressed as mean ± standard deviation (SD). Asterisks denote statistical significance following GP values style (* P ≤ .05; ** P ≤ .01; *** P ≤ .001; **** P ≤ .0001, and ns P > .05).

RESULTS

Impact of burn wounds on the adjacent epidermis

Histone acetylation is a fast and effective system capable of changing gene transcription in response to environmental cues. Loss of the epithelial barrier triggers a repair process focused on the expedited re-establishment of functional and protective skin or mucosa. Initially, we sought to understand the early epigenetic modifications from a burn trauma compared to intact and distant normal skin during a 21-day healing process of a human-relevant burn animal model (Fig. 1A). Our model is characterized by a superficial partial thickness burn that results in complete loss of the epidermal barrier and compromised dermal layer (Fig. 1B, left). During healing, we can identify the thickening of the adjacent skin next to the burn wound compared to normal skin and the formation of an epithelial tongue at an early time point (day 6) (Fig. 1B, right). Histone modifications were identified using antibodies that recognize the acetylation sites of histone H3 at lysine 27 and histone H4 at lysine 5, 8, and 12 (Fig. 1C). All four acetylation sites from histones H3 and H4 are known major players in gene transcription. Although this study focused on characterizing the histone modifications during burn healing, we found unforeseen histone acetylation changes on intact skin adjacent to the burn areas. The epidermis adjacent to the burn wound became acetylated at histone H3K27 **** P ≤ .0001) and at histone H4K5 (** P ≤ .01) compared to the skin found far from the burn injury. Nonetheless, adjacent skin also showed reduced acetylation on histone H4K12 compared to normal skin (** P ≤ .01). Histone H4K8 was not modified (ns P > .05) (Fig. 1D–F). Next, we decided to evaluate changes in the expression of histones H3K27, H4K5, H4K8, and H4K12 at different time points of the burn healing process using the expression of histones from an intact normal skin (Fig. 1E and F, and Supplementary Fig. S1A) as a common reference to all time points.

Figure 1.

Burn Assay Design. (A) Twelve- to Fourteen-Week-Old Female Yorkshire Pigs Were Used in This Experiment. Tissue Samples From Adjacent Skin, Wound Margin, and Epithelial Tongue Were Collected at Time Points 0, 6, 12, 18, and 21 Days. (B) H&E Images Represent Superficial Partial Thickness Burn Wounds Measuring 5 × 5 cm Created Using Copper Blocks (Objective 4×, Scale Bar 600 µm—Left H&E Image). Note That Skin Burn Healing Is Characterized by the Thickening of the Adjacent Skin of the Wound and the Formation of an Epithelial Tongue Migrating Over the Wound When Compared With Normal Skin From Control (Upper H&E Taken Using 4× Magnification, Scale Bar: 600 µm; Lower H&E Images Taken Using 20× Objective, Scale Bar 200 µm—Right H&E Images). (C) Schematic Representation of a Nucleus Containing the Chromatin Represented by Histones and Its Acetylation Sites for Histone H3 Lys 27, Histone H4 Lysines 5, 8, 12, and 16. (D) Normal Histone Acetylation Levels on the Skin of Pigs Isolated Distant From the Burn Wound (Normal Skin) and the Skin Adjacent to the Burn Wound (Adjacent Skin) (** P ≤ .01, **** P ≤ .0001). (E) Bar Graphic Depicting the Acetylation Levels of Histones H3K27, H4K5, H4K8, and H4K12 in Tissues From Normal Skin and Adjacent Skin (** P ≤ .01, **** P ≤ .0001). (F) Representative Immunofluorescence Images of Histones H3K27, H4K5, H4K8, and H4K12 Expression Levels of Normal Skin and Skin Samples Adjacent to a Burn Wound (Histones Stained With Secondary Antibody Alexa 488, DNA Staining Using Hoechst 33342, Dashed Lines Identify Epithelial-Stromal Interface).

Histone modifications during burn-healing

The process of epithelial healing after trauma is characterized by the activation of epithelial proliferation, migration, differentiation, and, finally, the re-establishment of the epithelial barrier. During the day of the burn (day 0), we failed to identify any meaningful epithelial response to the injury (e.g., cellular proliferation, migration). This finding is consistent with many reports, including from our group. However, for the first time, we were able to observe that fast epigenetic modifications take place right after burning trauma (Fig. 2A, Supplementary Fig. S1B). By day 0, we focused on 2 anatomical locations: the intact adjacent skin (located next to the burn site), and the wound margin which is located right at the burn wound and compared with the histone expression levels of normal skin. As previously shown, histones H3K27 and H4K5 are upregulated on adjacent skin, while H4K12 is downregulated. Interestingly, the wound margin does not undergo similar epigenetic modifications as seen on adjacent skin; rather, it responds similarly to the distant normal skin, denoting its inability to respond to injury. This finding is interesting as it suggests that thermal injury interferes with the ability of tissues in contact with the burn to respond.

Figure 2.

Expression Levels of Histone Modifications From Day 0 to 12. (A) Data From Day 0 After the Burn Demonstrate the Acetylation Levels of All Histones Expressed at Normal Skin, Adjacent Skin, and at the Wound Margin After the Burn (* P ≤ .05, **** P ≤ .0001, ns P > .05). Note Representative Immunofluorescence Images of Histones H3K27, H4K5, H4K8, and H4K12 for Day 0. The Dashed Line Represents the Epidermal/Dermal Interface. (B) Data From Day 6 After Burning, Including All Anatomical Locations of the Porcine Wound, Including the Epithelial Tongue That Is Now Present (* P ≤ .05, ** P ≤ .01, **** P ≤ .0001, and ns P > .05). Immunofluorescence Images Are Representative of All Histone Expression Levels. (C) Day 12 After Burn Depicting All Changes in Histone Acetylation Levels Distributed Throughout the Four Anatomical Locations (**** P ≤ .0001, and ns P > .05). Immunofluorescence Images Are Representative of All Histone Expression Levels.

On day 6 after the burn, we observed the formation of an epithelial tongue defined by the collective migration of epithelial cells toward the wound center. Epigenetically, we observed a drastic change in the histone landscape with the abrupt downregulation of histone H3K27 on all 3 anatomical sites (Fig. 2B; ** P ≤ .01, Supplementary Fig. S1C) compared to normal skin. Unlike H3K27, we observed a hyperacetylation of H4K5 at the adjacent skin (**** P ≤ .0001), wound margin (** P ≤ .01), and at the epithelial tongue (**** P ≤ .0001) when compared to histone levels in normal skin. Histone H4K8 and H4K12 presented focal hyperacetylation compared to normal skin, where the adjacent skin and the epithelial tongue (* P ≤ .05) become hyperacetylated for H4K8, and the epithelial tongue being hyperacetylated for H4K12 (* P ≤ .05). As the epithelial tongue continued to migrate throughout days 12 and 18 and wounds achieved partial to complete healing by day 21 (Figs. 2C and 3A and B, Supplementary Fig. S2A–C), the acetylation of histones differed significantly according to the anatomical locations (Figs. 2 and 3 and Supplementary Figs. 1 and 2).

Figure 3.

Expression Levels of Histone Modifications From Day 18 to 21. (A) Expression of Histones at Day 18 After Burn Depicts the Expression Levels of All Histone Acetylation Sites (* P ≤ .05, ** P ≤ .01, **** P ≤ .0001, and ns P > .05). Immunofluorescence Images Are Representative of All Histone Expression Levels. (B) Final Time Point After Burn (Day 21) Depicting Burn Wound Closure and Tissue Maturation (* P ≤ .05, ** P ≤ .01, *** P ≤ .001, **** P ≤ .0001, and ns P >.05). Representative Immunofluorescence Staining for Histones at Day 21 Is Shown.

Hyperacetylation pattern of histones

Interestingly, histones H3K27 and H4K5 are hyperacetylated at the adjacent skin and wound margin (Fig. 4A and B). Typically, we see high proliferation rates of epithelial cells on day 6 after trauma at the skin wound margin. On day 6, we observed that H4K5 and H4K12 play a role in gene expression compared with all other histone acetylation levels (Fig. 4B). H4K5 and H4K12 are the main hyperacetylated histones at the epithelial tongue, followed by H4K8 (Fig. 4C).

Figure 4.

Overall Expression of All Histones. Radar Diagram Depicting Each Histone Acetylation Site at Different Healing Time Points Found Positive in Porcine Skin Tissues. The Expression of Each Histone Acetylation Site Is Divided by the Anatomical Location as the (A) Adjacent Skin, (B) Wound Margin, and (C) Epithelial Tongue.

A global overview of histone acetylation throughout the healing process (Fig. 5A) brings our attention to the overwhelming hyperacetylation status of histone H4K5 in all stages of tissue regeneration. Some uniqueness is also evident in the consistent hypoacetylation of histone H3K27 on day 6 of cellular proliferation and early migration.

Figure 5.

Different Modifications of Histones During Healing. (A) Radar Graphical Representation of Each Histone Expression Distributed by the 21 Days of the Assay at the Adjacent Skin, Wound Margin, and Epithelial Tongue. (B) Representation of Log2 Histone Acetylation Levels Distributed in a Timeline. Note That the Wound Margin Does Not Have Data on Day 21 Due to Complete Tissue Healing, and the Epithelial Tongue Graphic Does Not Present Data on Day 0 Due to the Lack of Epithelial Migration Observed on the Day of Injury.

Other interesting observations emerged from a timeline evolutionary analysis of acetylation for each histone throughout the healing process (Fig. 5B). At the adjacent skin, we observed that H4K5 presented a steady expression from day 0 to day 21. At the same time, H3K27 experienced a hypoacetylation on day 6 at the same point in which histones H4K8 and H4K12 became hyperacetylated (Fig. 5B—adjacent skin). These findings demonstrate the downregulation of genes associated with H3K27 and the upregulation of genes controlled by histones H4K8 and H4K12. It is also evident that histone H4K8 plays a focal effort in gene upregulation on day 6 but not so much on other days in the intact adjacent skin (Fig. 5B—blue H4K8). Wound margins undergo significant acetylation of histones on day 12. Histones H3K27 and H4K5 continue to support the wound margin until day 18, and histones H4K8 and H4K12 are downregulated (Fig. 5B—wound margin). The epithelial tongue demonstrates a unique histone acetylation pattern in which histone H3K27 is abruptly hyperacetylated by the end of the healing process, suggesting its role in cellular differentiation, reduced migration, and potential stratification of the epidermis as a result of complete healing (day 21).

DISCUSSION

Accidents resulting in burn injuries affect 11 million people worldwide, and in the US alone, 1.1 million Americans suffer from burn injuries, with a hospitalization rate of nearly half a million¹³ (American Burn Association). Along with the traumatic experience of coping with burns, many patients are scarred for life, leading to compromised limb movements or the impact of disfiguration and consequent social exclusion.

From a therapeutic perspective, new promising technologies have been generated and are currently available for clinical use,¹⁴ along with the continuous development of novel materials for burn wound dressing.^15,16 For the most part, current strategies aim to protect the wound while providing a favorable environment conducive to healing. More advanced techniques like the RECELL System use a spray-on technology that delivers autologous skin cells onto the burn wound, reestablishing an epithelial barrier in a shorter time frame. The technology’s limitations include managing deeper wounds requiring surgical skin grafting. Although we have seen a substantial advance in burn management using a diversity of new materials and technology, little has been achieved in harvesting the skin’s potential to heal to accelerate tissue healing. Our group has explored signaling pathways capable of accelerating skin regeneration through the modulation of the PI3K signaling pathway, as well as the delivery of Periostin and the induction of mTOR signaling through photobiomodulation of epithelial cells.^17–19 Our research observed that histone acetylation constitutes a common epigenetic modification during epithelial healing.²⁰ Histone modifications, particularly acetylation, are known to induce fast gene transcription in response to environmental changes, including trauma. Our current work aims to understand the dynamic epigenetic modifications after burn. Knowledge of which histones are modified upon injury will provide a solid basis for drug discovery of new compounds capable of recapitulating the normal epithelial healing process on a fast track.

Our current study shows strong data supporting an orchestrated succession of epigenetic events during burn healing. We found histones like H4K5 that are broadly hyperacetylated throughout 21 days of tissue healing and evenly distributeds throughout all segments of skin regeneration, including the adjacent skin, the wound margin, and the epithelial tongue. We also found that another histone acetylation site, like the histone H3 acetylated at lysine 27, is completely shut down early on tissue healing (day 6). Interestingly, some other acetylation sites of histone H4, like lysine 8 and 12, present modest elevation of the acetylation levels during the proliferative phase. In contrast, histone H4 lysine 5 was the only histone to show strong acetylation levels, while cellular proliferation was observed during early migration of the epithelial tongue on day 6. Nonetheless, these lysine residues become acetylated at very specific healing times and in specific anatomical areas during healing. For instance, histone H4K8 presents a pick of acetylation level on day 12, specifically at the wound margin, and on days 6 and 12 on the epithelial tongue. It is clear that the epigenetic machinery is actively involved in the healing process of burn injuries and that specific hyperacetylation of certain lysine from histones H3 and H4 are time- and tissue-dependent. Our data provide novel evidence on the potential use of epigenetic modifications to specifically target genes associated with cellular proliferation and migration. Although exciting, the precise pool of genes being upregulated through histone modifications is still unknown.

Author contributions: Larissa S. A. Rolim (Data curation [equal]; Formal analysis [equal]; Investigation [equal]; Methodology [equal]; Software [equal]; Validation [equal]), Patricia S. Nascente (Data curation [equal]; Investigation [equal]; Methodology [equal]), Rogerio M. Castilho (Conceptualization [lead]; Formal analysis [lead]; Funding acquisition [equal]; Investigation [equal]; Writing—original draft [lead]; Writing—review & editing [lead]), and Cristiane H. Squarize (Conceptualization [lead]; Funding acquisition [equal]; Project administration [lead]; Resources [lead]; Supervision [lead]; Writing—original draft [equal]; Writing—review & editing [equal])

Conflict of interest statement: The authors declare no conflict of interest.

Funding: This work was supported by the National Institute of General Medical Sciences Research Fund R01GM120056, R01120056-04S1, and R01GM143938.