Apoptosis in Skin Wound Healing

Introduction Apoptosis is a highly conserved physiological cell death process that removes unwanted cells.[1,2] Since 1972 when Kerr, et al., introduced the term apoptosis to distinguish from necrosis, there has been tremendous interest in programmed cell death as a critical component in maintaining homeostasis and growth in tissues.[3] Failure of apoptosis can lead to a variety of cancers, viral infections, and autoimmune diseases.[4–6] Since the past three decades when the mystery of this cellular phenomenon began to unravel, scientists have discovered its valuable applications in clinical research and biotechnology. One area of particular importance is the wound healing process in which apoptosis is responsible for the removal of inflammatory cells and granulation tissue.[7] While extensive literature in apoptosis has been published in many fields, very few have focused on the role of apoptosis in the skin wound healing process.[8,9] This review describes the basic cellular and biochemical aspects of apoptosis and the occurrence of apoptosis during the skin wound healing and regeneration process. Morphological Features of Apoptosis Apoptosis, derived from the Greek word for “falling off” of leaves from a tree, is a physiological process in which cells are strategically programmed to die. The hallmarks of this phenomenon include chromatin condensation and membrane blebbing, cell shrinkage, and cell disassembly into apoptotic bodies. Morphological and biochemical changes occur throughout the cell. In the nucleus, chromatin condenses and nuclear endonucleases cleave the DNA into 180 base pair fragments. The cytoplasm shrinks in size due to electrolytic loss and fragmentation into blebs of cytoplasm with condensed nuclear material.[10] Mitochondrial alterations occur as the voltage and pH gradient across the inner membrane is lost, leading to matrix swelling, rupture of the outer membrane, and release of pro-apoptotic proteins from the intermembrane space.[11,12] The normal asymmetric distribution of phospholipids in the inner and outer leaflets of the plasma membrane is lost as phosphatidyl serine attaches to binding proteins on the outer cell surface,[13,14] signaling neighboring macrophages bearing phosphatidyl serine receptors to recognize and engulf the apoptotic bodies.[15] Since the cell membrane retains its integrity throughout most of the apoptotic progress, no intracellular substances leak out of the cell to cause inflammation. Apoptosis is distinct from another form of cell death known as necrosis, which can be triggered by cell injury from such conditions as hyperthermia, hypoxia, trauma, and the accumulation of toxic reagents.[16–19] Whereas apoptosis affects single cells, necrosis often affects sheets of cells within a tissue. This is due to the deleterious effects that necrotic cells have on surrounding cells. Necrotic cells are characterized by the loss of membrane integrity, organelle swelling, and lysosomal leakage.[5,19] The cellular DNA also non-specifically degrades into a variety of molecular base sizes. Another distinct feature of necrosis is a significant inflammatory response. The release of cellular material into the extracellular fluid mobilizes inflammatory cells to release chemotactic agents and remove the necrotic cells.[20] Molecular and Biochemical Pathways of Apoptosis The nematode Caenorhabditis elegans (C. elegans) has been a useful model for studying apoptosis because of similarities to the apoptosis process in mammalian organisms. After extensive research, it was found that this nematode has 1090 somatic cells as an adult, of which precisely 131 undergo programmed cell death.[21] Three gene products that were found to be crucial for apoptosis in C. elegans are ced-3, ced-4, and ced-9.22 As apoptosis promoters, ced-3 is homologous to the family of caspases and ced-4 is homologous to apoptotic protease-activating factor-1 (APAF-1).23 The apoptosis suppressor, ced-9, is homologous to the Bcl-2 family.[8,24] Due to the shared homology between C. elegans and mammalian species, the evolutionary relationship has helped elucidate the apoptosis pathway for higher forms of organisms. Apoptosis is a complex process that can be divided into the following four basic steps: induction, detection, effector, and removal.9 In the induction step, the cell receives an apoptotic signal. Various external stimuli can trigger apoptosis, including nutrient deprivation, cytokine depletion, ionizing radiation, and oxidative stress.[25–27] The cell then integrates numerous signals derived from signal transduction pathways to decide whether to commit to apoptosis. Once the cell commits to apoptosis, the signal to activate death machinery is detected and transduced to downstream effectors. Various regulators then carry out the apoptotic response,[8,28,29] and finally the cell is removed by phagocytosis. Each step of the apoptosis process requires the concerted effort of many molecules, and among the most influential ones are the caspases, the Bcl-2 family of proteins, and p53. Caspases. Since the discovery of similar homology between ced-3 and cysteine-dependent aspartate-directed proteases (caspases), there has been great interest in caspases as important players in apoptosis. At least 14 caspases have been identified. All caspases have in common three domains: a large subunit (20 kD), a small subunit (10 kD), and a NH2-terminus. Initially zymogens, caspases become activated by selectively cleaving after aspartic acid and assembling into heterotetramers. In general, they can be categorized into three groups. The first group (caspase-1, caspase-4, and caspase-5) plays a role in inflammatory response.[30] Members of the second group are initial transducers (caspase-2, caspase-8, caspase-9, caspase-10), and members of the third group are effectors (caspase-3, caspase-6, and caspase-7).[22,31,32] Caspases -12 through -14 have not yet been categorized due to insufficient information. One family of cellular proteins that regulates caspase activity is the inhibitor of apoptosis (IAP).[33] IAPs are the first class of endogenous cellular inhibitors of caspases to be found in mammalian species.[34] These proteins have in common a baculovirus IAP repeat (BIR), a ~70-residue binding domain.35 At least eight members of the IAP family have been identified, including X-chromosome-linked IAP (XIAP), cellular IAP1 (cIAP1), and cellular IAP2 (cIAP2).[36] Originally discovered in baculoviruses, IAPs suppress host cell death response.[37] It has been shown by ectopic expression that some IAPs block apoptosis in mammalian cells.[38,39] Although it is unclear how IAPs suppress apoptosis, some studies suggest that XIAP, cIAP1, and cIAP2 bind and inhibit caspases -3, -7, and -9.[40,41] Bcl-2 family. The bcl-2 gene was originally identified to be translocated in human follicular lymphoma.[42] Since this discovery, at least 18 members of the Bcl-2 family of proteins have been discovered.[43] As effectors of the apoptosis pathway, the Bcl-2 family of proteins either promotes or prohibits apoptosis. Pro-apoptotic proteins include Bax and Bak, while Bcl-2 and Bcl-xL are anti-apoptotic proteins.[44–46] The members of this family share up to four conserved Bcl-2 homology (BH) domains, namely BH1, BH2, BH3, and BH4.[47–49] In general, anti-apoptotic members possess all four conserved domains, and pro-apoptotic members display less conservation.[50] Genome p53. Often regarded as the guardian of the genome, p53 controls the fate of damaged cells by detecting and arresting the cell cycle.[51–53] It consists of a N-terminal transcriptional domain, a C-terminal regulatory domain, and a central DNA binding domain[54] Normally, the supply of p53 is continually replenished because of its interaction with mouse of double minute 2 homologue (MDM2).55 MDM2 binds to p53 in a negative feedback loop in which p53 stimulates the production of MDM2, which then degrades p53.[56] However, when there is DNA damage, p53 becomes stabilized and targets genes like p21WAF1/CIP1 to inhibit cyclin-dependent kinases, which leads to cell cycle arrest at the G1 and G2 stages.[57,58] p53 also regulates other transcriptional targets that arrest the cell cycle, including GADD45, wildtype p53-induced phosphatase 1 (Wip1), 14-3-3 sigma, and BTG2.[59–62] At the same time, p53 is regulated by proteins like ataxia telangiectasia mutated (ATM) and DNA-dependent protein kinase (DNA-PK).55 ATM and DNA-PK stabilize p53 by phosphorylation.[63–65] When DNA damage cannot be repaired, p53 activates the apoptosis machinery. Apoptosis Pathways Multiple pathways are known to induce apoptosis, and their detailed mechanisms are described elsewhere.[6,8,12] However, it is instructive to briefly describe three well-studied pathways: the Fas-mediated pathway, mitochondria-mediated pathway, and the p53-dependent pathway. Fas-mediated pathway. Fas (APO-1/CD95), the death receptor, is a 48 kD membrane protein from the tumor necrosis factor receptor (TNFR) family.[66,67] It contains an intracellular domain of 80 amino acids known as the death domain (DD). The Fas ligand (FasL), a member of the tumor necrosis family (TNF), is a 40 kD membrane protein that produces a soluble homotrimer.[68] The binding of three Fas molecules to a FasL homotrimer leads to subsequent binding of Fas-associated DD (FADD) and procaspase-8. The formation of this complex, known as the death-inducing signaling complex (DISC), triggers a cascade of caspase activation, including caspase-3, leading to cell death.[69] Besides the recruitment of FADD, Fas-induced apoptosis pathway can also be mediated by a receptor-interacting protein (RIP), RIP associated ICH/CED-3-homologous protein with a DD (RAIDD), and procaspase-2.[70] Mitochondria-dependent pathway. In contrast to death receptor-mediated apoptosis, mitochondria-dependent apoptosis is characterized by the mitochondrial permeability transition (MPT), in which the permeability of the inner mitochondrial membrane increases for small solutes (p53-dependent pathway. Responding primarily to DNA damage, activated p53 serves as a transcription factor that modulates the transcription of several apoptosis-related genes. For example, p53 upregulates the transcription of Bax while downregulating that of Bcl-2, thus favoring mitochondria-dependent apoptosis.[73,74] In addition, it upregulates transcription of Fas to support Fas-mediated apoptosis. Some studies suggest that transcriptional activity is not necessary for p53-dependent apoptosis. For example, in the presence of actinomycin D or cycloheximide, which block RNA and protein synthesis, respectively, p53 still induced apoptosis.[75] Further research on p53 will unravel more clues to its role in the regulation of apoptosis. Detection Methods Numerous methods to detect apoptotic cells have been established based on the morphological and biochemical events of apoptosis. Morphological features are the definitive standard for detecting apoptotic cells, but such changes may be difficult to observe in large samples. Biochemical techniques may be easier methods for detection, but they often do not distinguish between apoptotic and necrotic pathways. Together, however, morphological and biochemical methods can provide meaningful information about apoptosis. On the cellular level, microscopy of dyes is an indicator of cell viability. Using light microscopy, viable cells can be detected by their rejection of Trypan blue, while apoptotic cells whose plasma membranes are permeable will take up the dye. Ethidium bromide and propidium iodide both stain apoptotic cells red under fluorescent microscopy. For cell surface changes, annexin V, a phospholipid-binding protein, preferentially binds to phosphatidylserine. Since the externalization of phosphatidylserine occurs in the early stages of apoptosis, this assay is capable of detecting early apoptosis using flow cytometry. On the nuclear level, fluorescent dyes, such as ethidium bromide and propidium iodide, can detect DNA strand breaks associated with apoptosis by flow cytometry or laser scanning microscopy. Another nuclear detection method called terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) takes advantage of an apoptosis hallmark in which DNA cleaves into 180 to 200 bp fragments, generating 3’-OH groups. The terminal deoxynucleotidyl transferase (Tdt) enzyme recognizes and attaches to these hydroxy groups, thus labeling apoptotic cells. Finally, immunohistochemical staining can be performed to analyze the expression of apoptosis-related markers, such as caspase-3, Bcl-2, and APAF-1. A summary of some conventional methods are highlighted in Table 1, and further literature about these techniques can be found from the listed references. The Wound Healing Process Skin, the largest organ of the human body, consists of the epidermal and dermal layers. The stratified, keratinized epidermis acts as a physical barrier for the skin, and the collagen-rich dermis provides support and strength. Since skin protects the body from external insults, any injuries must be quickly and efficiently treated. Wounds generally heal by primary intention (e.g., closed incisions) or secondary intention (e.g., wounds with tissue loss, excisions, chronic skin ulcers, etc.). Wound healing, which requires the concerted effort of numerous cell types, involves cell migration, proliferation, differentiation, and apoptosis. Cell types that play dominant roles include epidermal cells, macrophages, fibroblasts, and platelets. Besides cells, a variety of cytokines and extracellular matrix (ECM) proteins are also important. Wound healing is a well-organized process that can be characterized by the following continuous sequence of events: inflammation, proliferation, and maturation.[92] Inflammatory stage. Immediately upon injury, blood coagulation and platelet aggregation form a fibrin clot as a provisional matrix for the migration of inflammatory cells and fibroblasts to the wound area. Cytokines, such as platelet-derived growth factor (PDGF), transforming growth factor alpha (TGF-a), transforming growth factor beta (TGF-b), epidermal growth factor (EGF), basic fibroblast growth factor (bFGF), and vascular endothelial growth factor (VEGF), are released into the wound bed.[93] In response to the influx of cytokines, leukocytes and fibroblasts generate other cytokines, such as tumor necrosis factor alpha (TNF-a) and interleukin 1 beta (IL-1b).[94] In early inflammation, neutrophils arrive at the wound site shortly after injury, followed by macrophages and monocytes. The leukocytes clear contamination by phagocytosis to establish favorable conditions for angiogenesis, cell migration, and cell proliferation.[95] When there is no substantial wound contamination, neutrophils are usually phagocytosed within a few days by macrophages.[96] Macrophages, in particular, play an essential role in the transition between inflammation and repair by secreting growth factors that are necessary for granulation tissue formation, including TGF-alpha, TGF-beta, PDGF, and FGF.[97–100] Mast cells have also been found to participate during inflammation by releasing cytokines, such as TGF-b and interleukin four (IL-4).[101] Reepithelization begins hours after injury by the migration and proliferation of epithelial cells to the wound. The hemidesmosomes between epidermis and basement membrane dissolve, allowing activated epidermal cells to migrate over the wound bed and secrete cytokines, such as interleukin one (IL-1), TNF, TGF-a, and TGF-b.[102] To facilitate migration to the wound space, the epidermal cells produce collagenase to degrade the extracellular matrix.[103] While migrating inward from the margin of the wound, they lay down basement membrane proteins in a zipper-like fashion.[104] About one day later, the wound margin undergoes proliferation of epidermal cells.[105] Proliferative stage. About four days after injury, macrophages, fibroblasts, and blood vessels invade the stroma to form granulation tissue.[106] Macrophages stimulate fibroplasia and angiogenesis by releasing growth factors, fibroblasts provide structural support by synthesizing extracellular matrix, and blood vessels transport nutrients to the site.[105] Neovascularization is mediated by a number of chemical inducers, including fibroblast growth factor 1 (FGF-1), PDGF, TGF-a, TGF-b, and VEGF.[107–109] Besides growth factors, the influx of endothelial cells into the wound also requires the deposit of fibronectin by nearby microvascular endothelial cells.[110] Two weeks after injury, fibroblasts undergo a phenotypic alteration into myofibroblasts and migrate to the wound bed, where they begin to contract the wound.[105] Maturation stage. The final stage of wound healing is maturation, in which the granulation tissue and fibroplasia recede. It is a gradual process that can take many months to complete. During this stage, the epidermis regenerates by undergoing reduction of transient hypertrophy, while the provisional matrix is replaced by a dermal matrix of collagen and later by a low cellularity scar.[92,93] Degradation of the collagen matrix is mediated by matrix metalloproteinases, which are secreted by the epidermal cells, fibroblasts, endothelial cells, and macrophages.[111] Eventually, the wound is replaced by a new functional tissue. Apoptosis in Normal Wound Healing While apoptosis has been previously described for many physiological processes, one area that has only recently been under investigation is skin wound healing. Apoptosis is vital to normal wound healing, especially in the removal of inflammatory cells and scar formation. As cell populations rapidly proliferate during tissue reconstruction, cell growth is balanced by apoptosis. Inflammatory cells, for example, must be removed in order to begin the next stage of wound healing. Otherwise, persistent inflammation can lead to nonhealing wounds. Similarly, the granulation tissue must decrease in cellularity to evolve into a scar. Recent research has elucidated some of the key roles of apoptosis in the wound healing process. Research by Brown, et al., supports that inflammatory cells undergo apoptosis.[112] In comparative studies of full-thickness wounds in diabetic and nondiabetic mice, they observed that apoptosis was detected in the inflammatory cells as early as 12 hours in nondiabetic mice. The apoptosis level peaked at the fifth day after wounding and slowly decreased afterwards. Infiltration of inflammatory cells was present at 12 hours, consisting primarily of neutrophils. Apoptosis was consistently observed in the inflammatory cells beneath the leading edge of the migrating epithelium. This may be an indication that apoptosis signals the end of the inflammatory phase of healing. In comparison to nondiabetic mice, diabetic mice were characterized by delayed apoptosis, but the trend was reversed when topical PDGF and insulin-like growth factor II (IGF-II) treatment was applied to the wound. Kane, et al., further explored the role of apoptosis-related markers during inflammation at the leading edge of the epithelium.[113] They compared the expression of p53, Bcl-2, and apoptosis between normal and diabetic mice for 42 days. Their results suggest that normal mice exhibit an inverse relationship between Bcl-2 and p53 over time. Upon injury, Bcl-2 expression increased while that of p53 decreased, in order to allow for cellular proliferation to occur. As the inflammation process declined, p53 levels increased while Bcl-2 levels decreased. The rate of apoptosis appeared to parallel the rate of p53. For diabetic mice, there was no inverse relationship, and p53 expression was consistently higher than that of Bcl-2. The levels of both p53 and Bcl-2 decreased over the days 21 to 42, and the peak of apoptosis activity did not occur until day 14. Although the relationship of these apoptosis markers is not well-understood in diabetic mice, this study illustrates the necessary balance between p53 and Bcl-2 in normal mice during the inflammatory phase of wound healing. Desmouliere, et al., investigated the role of apoptosis in the evolution of granulation tissue to scar tissue.[7] Using eight-week-old Wistar rats, granulation tissue was sampled from 2 to 60 days. They observed many apoptotic inflammatory cells during the first few days after wound healing. In particular, isolated apoptotic cells were visible by eight days, peaking between 12 to 25 days. Furthermore, the frequency of apoptotic myofibroblasts and vascular cells were highest during this time period. Their data also suggest that apoptosis of the granulation tissue begins after wound closure, affecting cells in a consecutive fashion. This concurs with the observation of gradual granulation tissue resorption after wound closure. Apoptosis has also been implicated in dermal reconstruction. In a study by Rossio-Pasquier, et al., athymic nude mice were grafted with split-thickness human skin biopsies.[114] Two months later, partial-thickness wounds were made, and the healing grafts were harvested from 1 to 4 days after injury. The same experiment was repeated using petrolatum-impregnated dressings. By Day 1 after the injury, TUNEL-positive cells were present in the uninjured human dermis beneath the scab for mice treated without petrolatum-impregnated dressings. Similar observations were subsequently noted until Day 4 when no apoptotic cells were seen. In contrast, for mice with petrolatum-impregnated dressings, very few TUNEL-positive cells were seen during the first three days after wounding, and no apoptotic cells were seen by Day 4. The results suggest that human fibroblasts disappeared from the uninjured dermis beneath the scab via apoptosis and that petrolatum-impregnated dressings reduced the level of apoptosis. Very few apoptosis studies have focused on burn wound healing. Nagata, et al., provide evidence that apoptosis and p53 work in a concerted effort to heal burn wounds.[115] Full-thickness burns were made on guinea pigs with a heated soldering iron for three seconds. Skin samples containing both the center and periphery of the wound were taken from 0.5 to 28 days after the burn. Their results showed that apoptosis and p53 protein expression increased in both the inflammatory (12 hours–2 days) and proliferation (2–14 days) stages of wound healing but decreased during the remodeling stage (14–28 days). Apoptotic cells were seen in the peripheral zone of the burn region from 12 hours to 10 days after the burn, and the level of apoptosis was highest by two days. The expression of p53 also increased from 12 hours to two days after wounding. For the regenerated epidermis, apoptosis occurred from Day 7 to 10, while for the granulation and scar tissue it was from Day 10 to 14. The p53 protein expression reached its peak by Day 7 in the regenerated epidermis and Day 4 for the granulation tissue. This study suggests that apoptosis plays a vital role to epidermal regeneration; p53 protein accumulation is associated with keratinocyte proliferation, and it down-regulates the proliferation process or it is associated with accelerated terminal differentiation of keratinocytes in the regenerated epidermis. Apoptosis in Abnormal Wound Healing Since the regulation of apoptosis is crucial to normal wound healing, altered apoptotic behavior results in a number of pathologic processes, of which the most notable types are hypertrophic scars and keloids. Hypertrophic scars and keloids, two forms of hyperproliferative healing, are characterized by hypervascularity and hypercellularity.[116] They are characterized by excessive scarring, inflammation, and an overproduction of extracellular matrix components, such as collagen I and proteoglycans.[117] Keloids are clinically similar to hypertrophic scars, with the exception that keloids expand beyond the original boundaries.[118] While the biochemical pathways leading to hypertrophic scar and keloid formation are incompletely elucidated, it is suggested that apoptosis is a significant factor. Wasserman, et al., studied the production of two apoptosis-modulating proteins, Bcl-2 and Fas, in peripheral blood mononuclear cell (PBMC) fractions of patients with hypertrophic scarring.[116] They found that the expression of Bcl-2 in PBMC fractions were significantly higher in patients with hypertrophic scarring than in those that healed normally. The expression of Fas did not vary significantly between the two cohorts. This study suggests that high Bcl-2 expression could be the cause of delayed fibroblast apoptosis, leading to abnormal healing patterns. Sayah, et al., examined the expression of 64 apoptosis-related genes in keloids and normal scars by cDNA arrayed hybridization and TUNEL.[119] They showed that eight apoptosis-related genes were significantly underexpressed in keloids, including glutathione S-transferase and defender against cell death 1 (DAD-1). TUNEL analysis also showed lower apoptosis indices for keloids in comparison to normal scars. In a combined study of keloids and hypertrophic scars, Bcl-2 intensely stained basal keratinocytes and scattered cells, but p53 expression was not detected.[117] As further research in proliferative healing continues, the specific role of apoptosis and apoptosis-related genes will be revealed. In conclusion, apoptosis is a physiological phenomenon that is necessary for maintaining homeostasis in an organism. In normal skin wound healing, apoptosis is responsible for events, such as the removal of inflammatory cells and the evolution of granulation tissue into scar. Dysregulation in apoptosis can lead to abnormal wound healing, such as hypertrophic scar and keloid formation. Further research to further the understanding of apoptosis will help develop novel methods to treat altered forms of wound healing. Acknowledgment The authors thank Dr. Melvin Silberklang of Ortec International, Inc. for critically reviewing this manuscript.