Background
Wound healing is a complex phenomenon that requires an integrated network of repair mechanisms, including cell migration, proliferation and cellular production of structural molecules essential for tissue regeneration. The tissue response to injury is characterized by three overlapping phases: (a) Inflammation, (b) granulation and (c) matrix remodeling [
1]. These phases involve a dynamic cascade of events including clotting and hemostasis, leukocyte recruitment, tissue formation, epithelialization, angiogenesis, collagen synthesis and wound contraction [
1]. The major cell types involved in cutaneous wound healing span from hematopoietic cells such as blood platelets, immune regulators like neutrophils, T cells, dendritic cells and macrophages, to structural cells like fibroblasts, endothelial cells and keratinocytes [
1‐
6]. Specifically, wound repair is regulated by a highly orchestrated interplay of cytokines, growth factors, extracellular matrix and more recently described chemokines [
2,
3,
6‐
15]. Chemokines and their receptors play a crucial role in development, homeostasis, tumor development and most notably wound healing and angiogenesis [
16‐
18].
Chemokines are small, secreted proteins which mediate directional migration and regulate leukocyte trafficking [
19‐
21]. Notably, the chemokine family is likely to be one of the first complete protein superfamilies that has been identified and characterized at molecular level [
22]. This presents the opportunity to identify all relevant members of the chemokine superfamily involved in complex biological processes such as wound healing.
The aim of this work was to investigate the complex interactions of chemokines and their receptors in cutaneous wound healing. We systematically measured cytokines, growth factors, adhesion molecules, matrixmetalloproteinases and chemokines in a murine model for acute cutaneous wound healing. Strikingly, a subset of chemokines proved to be among the most highly regulated genes. The chemokine expression is highly orchestrated and coincides with the appearance of matching receptors. Accordingly, we demonstrate that structural cells, such as human keratinocytes, dermal fibroblasts and dermal microvascular endothelial cells, express a distinct and functionally active repertoire of chemokine receptors. Furthermore, migration assays suggest that chemokines and their cognate receptors accelerate wound healing via chemotaxis of structural cells.
Methods
Cell culture
Human primary epidermal keratinocytes, dermal fibroblasts, melanocytes and dermal microvascular endothelial cells were purchased from Clonetics (San Diego, CA) and cultured in keratinocyte (KGM-2), fibroblast (FGM-2), or endothelial cell (EGM-2) growth medium (all Clonetics, San Diego, CA) [
23]. We have used 2 × 10
4 cells for the culture and let it grow up to 80% confluence. Cells were treated with TNF-α (10 ng/ml)/IL-1β (5 ng/ml) (R&D Systems Inc., Minneapolis, MN) for 10 to 12 h or left untreated. RNA was extracted from cells immediately, after 8 and 18 h as described later.
Quantitative real time PCR analysis
Quantitative real-time PCR analyses were performed as previously described [
21,
24,
25]. Skin biopsies were homogenized in liquid nitrogen and RNA was extracted with RNAzol according to the manufacturer’s protocol (Tel-Test, Friedensburg, TX). 4 μg of RNA was reverse transcribed using standard protocols. 25 ng of cDNA was amplified in the presence of 12.5 μl of TaqMan
® universal master mix (Applied Biosystems, Foster City, CA), 0.625 μl of gene-specific TaqMan
® probe, 0.5 μl of gene-specific forward and reverse primers, and 0.5 μl of water. As an internal positive control, 0.125 μl of 18S RNA-specific TaqMan
® probe and 0.125 μl of 18S RNA-specific forward and reverse primers were added to each reaction. Gene-specific probes used FAM™ as reporter dye, whereas probes for the internal positive control (18S RNA) were associated with the VIC™ reporter dye. Alternatively, 25 ng of cDNA was amplified using target-specific primer combinations and SYBR green master mix (Applied Biosystems, Foster City, CA). Chemokine ligand- and receptor-specific primers and target-specific probes were obtained from Applied Biosystems (Foster City, CA). Gene-specific PCR products were measured by means of an ABI PRISM
® 7700 or 5700 Sequence Detection Systems (Applied Biosystems, Foster City, CA) continuously during 40 cycles. Target gene expression was normalized between different samples based on the values of the expression of the internal positive control (18S RNA) or ubiquitin.
Flow cytometry
To analyze chemokine receptor expression of non-hematopoietic cells, cultured primary epidermal keratinocytes, dermal fibroblasts and dermal microvascular endothelial cells from different donors were analyzed using flow cytometry and the following monoclonal antibodies (mAb): PE-conjugated anti-CCR1 (53,504.111, IgG2β), CCR2 (48,607.211, IgG2β), CCR3 (61,828.111, IgG2α), CCR7 (150,503, IgG2α), CCR9 (112,509, IgG2α), CXCR4 (12G5, IgG2α), CXCR5 (51,505, IgG2α), CXCR6/STRL33 (56,811, IgG2β) all R&D Systems, Mineapolis, MN. CCR4 (1G1.1, IgG1), CCR5 (2D7/CCR5, IgG2α), CCR6 (11A9, IgG1) CXCR1 (5A12, IgG2β), CXCR2 (6C6, IgG1), CXCR3 (1C6/CXCR3, IgG1) all Pharmingen, San Diego, CA, CCR8 (Goat, IgG) Alexis Biochemicals, Grünberg, Germany, CCR10 (1908, IgG1) DNAX Research Inc., Palo Alto CA (or appropriate isotype controls (Pharmingen, San Diego, CA and R&D Systems, Mineapolis, MN, Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA). Briefly, 106 cells were stained with anti-chemokine receptor mAb or isotype and analyzed using a FACScan and CELLQuest software (Becton Dickinson, San Jose, CA).
In vitro wound repair assay
2 × 104 human primary dermal fibroblasts, 2 × 104 dermal microvascular endothelial cells and 2 × 104 epidermal keratinocytes were cultured in 24-well plates in FGM-2, EBM-2 and KGM-2 (Clonetics, San Diego, CA) medium and grown to confluency analogously to the previously described cultivation. The medium was removed and a single path wound was made with a sterile pipette tip through the intact cell layer. The width of the wound was 200 µm. Medium supplemented with or without CCL1/I-309, CCL11/eotaxin, CCL17/TARC, CCL20/MIP-3α, CXCL8/IL-8, CXCL9/MIG, CXCL10/IP-10 for keratinocytes, CCL1/I-309, CCL11/eotaxin, CCL17/TARC, CCL27/CTACK for dermal fibroblasts and CCL1/I-309, CCL11/eotaxin, CCL17/TARC, CCL20/MIP-3α, CCL27/CTACK, CXCL8/IL-8, CXCL9/MIG, CXCL10/IP-10 for dermal microvascular endothelial cells was added to each well. To confirm the involvement of G(i) alpha protein-coupled receptor signaling in chemokine-induced wound repair responses, structural cells were pre-incubated with or without pertussis toxin (100 ng/ml PTX, Calibochem, Germany) 10–12 h before injury and chemokine stimulation. The status of the single path wound was determined immediately as well as 8 and 12 h after the wounding using a digital camera system (Olympus, Hamburg, Germany), or was monitored for 12–18 h by time lapse video microscopy system (inverse Leitz Microscope with cell culture equipment, Zeiss AxioCam HRc, Zeiss AxioVision Software, Carl Zeiss, Oberkochen, Germany).
Cutaneous wound healing model
A total of 27 (3 mice per defined time point) female BALB/c mice were anesthetized with an intraperitoneal injection of ketamine (120 mg/kg bw) and xylazine (16 mg/kg bw). The lumbar region was shaved and the surgical area was disinfected with betadine and isopropanol. Full-thickness 2 cm dorsal incisions were made with a scalpel through the epidermis and dermis leaving the subcutaneous muscle layer undisturbed. Wounds were closed using a single 5 W surgical staple. Mice were euthanized by CO2 inhalation 6, 12, 18 h, 1, 2, 3, 5, 7, and 10 days after incisions were done. At the time of harvest, dorsal pelts surrounding the incisions were removed and normal skin was trimmed to 1–2 mm of incision. One half of each sample was placed in 10% buffered formalin and processed for paraffin embedding. The other half of each sample was snap frozen in liquid nitrogen and processed for RNA extraction. Each cohort consisted of three animals and the subsequent experiment was repeated twice.
All experiments with mice were carried out after examination by the local ethics committee of the University of Düsseldorf and after an internal identification number was given. The tests were carried out by trained personnel, who respect the internal and international guidelines for animal research (e.g. ARRIVE Guidelines).
Patients
This study was approved by the local ethics committee and informed consent was secured from all patients participating in this study. A total of 4 patients were included in the study. Skin biopsies were taken from either healthy individuals or patients undergoing surgery and secondary wound healing. Prior, 24 h, 2, 3, 5, 7, 9 and 12 days after radical surgery for acne inversa, biopsies were taken from the non-lesional wound edges. Skin samples were immediately frozen in liquid nitrogen and stored at – 80 °C.
Histology and immunohistochemistry
Frozen skin samples, cut to a thickness of 6 µm on a cryomicrotome, were used for immunohistochemical analyses of human secondary wound healing, skin sections were fixed with 3% acetone and preprocessed with H2O2 followed by an avidin and biotin blocking step (VECTOR Blocking Kit). Sections were stained with monoclonal antibodies against CCR2α, CCR2β, CCR3, (all polyclonal goat IgG), CCR4 (rabbit IgG), CCL1/I-309, CCL11/eotaxin, CCL24/eotaxin-2 (goat IgG) (all Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA), CD31 (JC/70A, mIgG1), CD68 (KP1, mIgG1) (DAKOCytomation, Hamburg, Germany), CCL4/MIP-1β, CCL17/TARC, CCL18/DC-CK1, CCL20/MIP-3α, CCL27/ALP (124,302.11; mIgG2α), CXCL9/MIG, CXCL10/IP-10 (all goat IgG), CXCL12/SDF-1α (K15C; mIgG1; Unite d’Immunologie Virale, Institute Pasteur, Paris, France), CCR1 (53,504.111; IgG2β), CCR5 (45,549.111; mIgG2β), CCR6 (53,103.111; mIgGβ), CXCR1 (42,705.111; mIgG2α), CXCR2 (48,311.111; mIgG2α), CXCR3 (49,801.111; mIgG1), CXCR4 (44,716.111; mIgG2α), CXCL10/IP-10, CXCL9/MIG, CCL17/TARC (all goat IgG) all R&D Systems, Minneapolis, MN), CXCL8/IL-8 (B-K8; mIgG2α) (Diaclone Res., France), CCR8 and CCR10 (German Cancer Research Center, Heidelberg, Germany) at 4oC. The staining was performed with an ABC-Kit (VECTOR Vectastain ABC-Kit) and an ACE-Kit (VECTOR). Sections were counterstained with hematoxylin. The immunohistochemical examinations were conducted by three independent pathologists. If there were differences in the evaluation, a joint result was finally established.
Discussion
In recent years, several mediators including cytokines and chemokines have been suggested to participate in the process of wound healing [
1,
2,
17,
18]. So far, most reports have mainly focused on the role of single molecules or single ligand–receptor combinations and comprehensive studies are missing. The rapid discovery of the complete chemokine superfamily offers the opportunity to take a “global view” in order to identify all family members of a particular molecular family involved in a biological process.
In the present study, we systematically investigated the potential role of several members of the chemokine superfamily in cutaneous wound healing, characterized their cellular origin as well as their regulation by proinflammatory mediators and their putative function. We have performed the analysis of the chemokine network on the mRNA- and also on the protein level, and secondly we investigated mouse and human wound healing models in vitro and in vivo. Our results suggest a complex interplay of structural cells via chemokine ligand–receptor interactions and we identified several members of the chemokine superfamily as being among the most highly regulated genes during cutaneous wound healing.
Cutaneous wound healing occurs via sequential overlapping phases starting with the injury and consecutive with the hemostasis followed by inflammation, re-epithelization, granulation tissue- formation and finally tissue remodeling. Within hours after tissue injury, neutrophils are recruited to cleanse the wound of foreign, potentially infectious particles [
1,
2]. Neutrophil infiltration into the wound is followed by an influx of monocytes which differentiate to activated macrophages and release an array of growth factors such as platelet-derived growth factor (PDGF) and vascular endothelial growth factor (VEGF) together with chemoattractants initiating the formation of granulation tissue [
1,
2]. Among the leukocyte subsets infiltrating cutaneous wounds, macrophages and macrophage-derived factors have been shown to play a pivotal role in the transition from inflammation to tissue repair since macrophage-depleted animals exhibit defective wound repair [
4,
5]. In this phase, a plethora of cytokines along with chemokines is released. Chemokines such as CCL2, CCL3, CCL4 and CCL5 are candidates to mediate monocyte/macrophage recruitment to wound sites via CCR1, 2, 3 and 5 [
2,
3,
26]. Our findings confirmed strong CCL2, CCL3 and CCL4 expression during the inflammatory phase of mouse wound healing and show matching temporal expression of their main receptors CCR1 and CCR2 on wound-associated CD68
+ macrophages [
2]. Macrophages and dendritic cells are also an abundant source of chemokine ligands during cutaneous wound healing. In humans, macrophages and dendritic cells at wound sites secrete high levels of CCL18, CXCL9, 12 and CXCL22. In fact, previous findings in arteriosclerosis and lung diseases support strong CCL18 expression by lesional macrophages and dendritic cells [
27‐
29].
Our observations also suggest that other monocyte/macrophage chemoattractants such as CCL1 and CXCL12 may be involved in the recruitment of CCR8/CXCR4-positive macrophages/dendritic cells during the inflammation and the transition to the tissue formation phase of cutaneous wound healing. However, this is in contrast to the more recent results of Zheng L et al. who described CXCL8 and CCL2 as important chemokines in the inflammation phase and hemostasis phase produced by macrophages [
30]. Our immunohistochemical analysis of acne inversa patients showed no significant expression of CXCL8 in macrophages but particularly in keratinocytes and endothelial cells, with a high expression of CXCL8 after 2 days.
The influx of macrophages is followed by the infiltration of lymphocytes into the wound site around days [
2,
3]. Lymphocytes remain in the demarcating zone for the entire wound healing process while the majority of neutrophils and macrophages disappear after the formation of granulation tissue. Strikingly, a large variety of lymphocyte chemoattractants such as CCL1, CCL2, CCL3, CCL4, CCL5, CCL7, CCL11, CCL18/DC-CK, CCL20, CCL24, CCL27, CXCL9, CXCL10, CXCL12, and XCL1 are present during the inflammation and tissue formation phase of wound healing. Correspondingly, a strong expression of CCR2, CCR3, CCR6, CCR8, CCR10, CXCR3 and CXCR4 can be detected on the surface of CD3
+ lymphocytes at wound sites. In summary, the immune system provides multiple and redundant mechanisms to attract lymphocytes to sites of injury and tissue repair. Yet, their precise role during tissue formation and remodeling, besides representing sentinels of the immune system, is currently not well understood. In this context, the predominant presence of CD3
+ lymphocytes in the perivascular pockets of sprouting blood vessels during cutaneous wound healing points toward a possible role in supporting endothelial cell proliferation, migration and invasion.
Tissue granulation and re-epithelialization of the wound bed are the next important processes during wound healing. Within days after injury, the phenotype of keratinocytes of the wound edge markedly changes. In fact, the cells reorganize their cytoskeleton, dissolve intercellular desmosomes, loose hemidesmosomal links between the epidermis and the dermis, and upregulate integrin receptors [
1]. Altogether, these processes enable keratinocytes to migrate along viable tissue into the wound bed. However, the stimuli determining the migration of keratinocytes during the formation of a neo-epidermis have not been completely elucidated yet. The in vitro findings presented in this study show that human keratinocytes express a distinct set of functionally active chemokine receptors including CCR3, CCR4, CCR6, CXCR1, CXCR2, CXCR3 on their cell surface. Ligands to particular receptors significantly enhance keratinocyte migration and promote wound repair in vitro. Chemokine receptor expression by keratinocytes was confirmed in vivo during human secondary wound healing pointing to several important chemokine ligand–receptor combinations in the process of re- epithelialization.
In normal skin, basal keratinocytes express high levels of CXCR3 on their cell surface. Yet, CXCR3 expression of keratinocytes appears to be differentiation-dependent since suprabasal keratinocytes of the stratum spinosum or granulosum loose CXCR3 expression. In contrast, differentiating keratinocytes gain CXCR4 expression in vivo. Previously, immunohistochemical analyses of Charbonnier et al. suggested CCR6 expression by keratinocytes but the biological function remained unclear [
31]. Here, we have shown that signaling through CCR6 can enhance keratinocyte wound repair in vitro.
In vitro, CCL27 markedly enhanced wound repair of endothelial cell and fibroblast indicating a role for CCL27–CCR10 interactions in tissue repair.
Approximately, 4 days after injury granulation tissue invades the wound bed. Fibroblasts start to produce a provisional matrix composed of fibrin, fibronectin, and hyaluronic acid, which is necessary to support cell ingrowth [
1]. Cell movement into this matrix may require proteolytic functions which are provided by matrix-metalloproteinases such as plasmin, plasminogen activator, collagenase, gelatinase A or stromelysin, which are markedly upregulated between days 2 and 5 and may cleave a path for cell migration. Previous studies showed CCL2-induced matrix-metalloproteinase induction in dermal fibroblasts [
32]. Although we could not detect significant CCR2 or CCR5 expression, both representing CCL2 ligands, on cultured fibroblasts, our in vitro findings show that primary dermal fibroblasts markedly express CCR3, CCR4, and CCR10 and respond with enhanced wound repair after stimulation with their ligands [
33]. Notably, a broad array of CCR3 ligands, including CCL5, CCL7, CCL11 and CCL24, are upregulated in cutaneous wound healing suggesting a role for these ligand–receptor interactions in fibroblast biology and the tissue formation phase of the wound healing process.
Angiogenesis represents a complex process depending on extracellular matrix in the wound bed as well as migration and mitogenic stimulation of endothelial cells. During recent years, a variety of angiogenic factors such as basic fibroblast growth factor (bFGF), vascular endothelial growth factor (VEGF), transforming growth factor-β (TGF-β), angiogenin, angiotrofin, angiopoietin-1, thrombospondin, low oxygen and elevated lactic acid, have been linked to the induction of angiogenesis [
1]. A current focus of chemokine research represents the investigation of the contribution of chemokines toward angiogenesis. Certain members of the CXC, e.g. CXCL8 and more recently also of the CC chemokine family induce endothelial cell migration in vitro and angiogenesis in vivo [
3,
15,
16,
34‐
39]. Furthermore, angiogenic growth factors such as VEGF and bFGF have recently been shown to regulate chemokine receptor expression on endothelial cell suggesting a complex relationship of growth factors and chemokines in the regulation of angiogenesis [
39]. In the present study, we show that human dermal microvascular endothelial cells express a distinct set of chemokine receptors, such as CCR3, CCR4, CCR6, CCR10, CXCR2, CXCR3 in vitro. Moreover, stimulation with the respective ligands (CCL1, CCL11, CCL27, CXCL8) enhanced wound injury repair of cultured endothelial cells. Another set of chemokines including CCL17, CCL20 and CXCL10 did not promote injury repair although their corresponding receptors (CCR4, CCR6 and CXCR3) are markedly expressed on endothelial cells. For other CXCR3 ligands, such as CXCL9, CXCL10 and CXCL11, there is accumulating evidence that they induce angiostasis through CXCR3 [
40].
Bernardini et al. and Haque et al. demonstrated that the CCR8 ligand CCL1 binds to endothelial cells, stimulates chemotaxis and invasion of these cells, enhances human umbilical vein endothelial cell differentiation into capillary-like structures in vitro and induces angiogenesis in vivo [
35‐
37]. Previously, it has been suggested that angiogenic ELR-motive-positive chemokines such as CXCL8 are expressed during the inflammation and tissue formation phase of wound healing and ELR-motive negative angiostatic chemokines such as CXCL10 appear later in the wound healing process dominating the tissue remodeling phase [
3]. Our findings indicate that both angiogenic and angiostatic factors are present at all phases of cutaneous wound healing and suggest a more complex balance and regulation.
Taken together, chemokines and chemokine receptors are expressed in a distinct temporal and spatial manner during wound healing and play many different functions during wound healing processes. Our findings elucidate an important role for chemokines in the complex wound healing cascade and underscore the central paradigm that cutaneous injury begins with proinflammatory cytokine and chemokine production which results in leukocyte recruitment and subsequent growth factor production. In turn, growth factor production by leukocytes and structural cells combined with chemokine signalling will orchestrate the dynamic process of migration, invasion, and proliferation of parenchymal cells for tissue repair. Not much has been done in the clinic regarding the use of chemokines and their receptors as a basis for treatment. Chronic inflammation and alteration in angiogenesis can potentially be reduced or eliminated by interfering with proinflammatory or angiogenic chemokines and their receptors using small molecules that interfere with receptor function. To our knowledge, this study is the most comprehensive analysis of chemokine and chemokine receptor expression both in vivo and in vitro. In addition to the extensive analysis of the central structural and soluble network partners in wound healing, our work was able to elucidate previously unknown interactions between chemokine ligands and chemokine receptors. We were able to gain information on a faster wound healing through the use of specific chemokines and we revealed patterns of expression that have not been described so far. Especially ligand–receptor signaling through CCR6 can enhance keratinocyte wound repair in vitro. CCL27 markedly enhanced wound repair of endothelial cell and fibroblast indicating an important role for CCL27–CCR10 interactions in tissue repair.
Nevertheless, there are some limitations to this study, which are based on the fact that a very special patient collective of the so-called Acne inversa was investigated. To support the results of our work, the number of samples should be increased and different patient groups should be examined in subsequent studies. The selection of the test mice can also contribute to the results. We have only examined female BALB/c mice in our work. However, recently published studies suggest that wound healing in female BALB/c mice may differ from male mice [
41]. To achieve a positive therapeutic effect, the exact interaction and influence of certain parts of this complex wound healing network must be examined in following work.
In conclusion, our findings enhance the understanding of the underlying mechanisms for cutaneous wound healing and may impact the development of novel therapeutic strategies for the treatment of wound healing disorders such as chronic ulcers.