Journal of Dermatological Science
Volume 65, Issue 1 , Pages 12-18, January 2012

Hypoxia regulates the expression of extracellular matrix associated proteins in equine dermal fibroblasts via HIF1☆

Université de Montréal, Département de biomédecine vétérinaire, Canada

Received 6 July 2011; received in revised form 5 September 2011; accepted 15 September 2011. published online 14 October 2011.

Article Outline

Abstract 

Background

Exuberant granulation tissue (EGT), a fibrotic healing disorder resembling the human keloid, occurs almost exclusively in limb wounds of horses and may be caused in part by a relative state of hypoxia within the wound.

Objective

The objectives of this study were therefore to (1) assess the effects of hypoxia on equine dermal fibroblast (EDF) proliferation and apoptosis, (2) study the effects of hypoxia on the expression of key extracellular matrix (ECM) associated proteins and determine if such effects are dependent on hypoxia-inducible factor (HIF), and (3) determine if EDFs from the body or limb respond differently to hypoxia.

Methods

EDFs were isolated and cultured from skin from body or limb under normoxic or hypoxic conditions for up to 7days.

Results

Hypoxia significantly stimulated EDF proliferation, but had no effect on cell survival. The hypoxia-mimetic agent CoCl2 up-regulated COL1A1 expression and down-regulated MMP2 expression, suggesting an increase in ECM synthesis and a decrease in turnover. Both regulatory effects were inhibited by the addition of echinomycin, indicating that they are mediated by the transcriptional regulatory activity of HIF. No differences were observed between EDFs originating from body or limb for any effect of hypoxia or CoCl2, suggesting that EGT development does not depend on intrinsic properties of limb fibroblasts.

Conclusions

We conclude that hypoxia regulates ECM remodeling via HIF1 in EDFs, and that this may be an important determinant in the pathogenesis of equine EGT.

Keywords: Wound healing, Hypoxia, HIF1A, Fibroblast, Extracellular matrix, Horse

 

Back to Article Outline

1. Introduction 

Wound repair is remarkably uniform across mammalian species [1] and includes four distinctive phases: hemostasis, inflammation, proliferation and remodeling [2]. The proliferative phase aims to replace damaged tissues and relies on three major mechanisms: epithelialisation, angiogenesis and fibroplasia. Dermal fibroblasts predominate during fibroplasia and perform multiple functions important to wound repair, in particular the synthesis of extracellular matrix (ECM) proteins, ECM reorganization and wound contraction resulting in mature scar formation [3]. Excessive ECM accumulation leads to fibrosis and can be the result of increased production of proteins such as collagen, proteoglycan and elastin [4], and/or decreased turnover [5]. The latter occurs when an imbalance arises in the secretion of matrix metalloproteinases (MMPs) vs. tissue inhibitors of metalloproteinases (TIMPs) by fibroblasts or other wound cells [1].

Dysregulation of the wound healing process can lead to fibrotic disorders, which in humans manifest clinically as hypertrophic scars or keloids [2]. The equine counterpart is exuberant granulation tissue (EGT), commonly known as “proud flesh”, wherein the wound is trapped in the proliferative phase of repair. Fibroblasts persist and synthesize ECM rather than differentiating into myofibroblasts or being eliminated [6]. This condition occurs primarily on the horse's lower limb and is detrimental to performance in both show and race horses. 7% of injuries leading to retirement are the result of a wound [7], exerting a significant financial impact of the horse industry. Studies have shown both equine EGT and human keloid to display deficient fibroblast apoptosis and microvascular dysfunction caused by luminal occlusion [8], [9], [10]. The latter was incriminated in the state of tissue hypoxia measured in both keloids [11] and, more recently, equine EGT [9]. Hypoxia is proposed as a major mechanism underlying fibrosis and overscarring. Proliferation of human dermal fibroblasts is increased 71% following 72h of culture at 1% oxygen [12], fibroblast synthesis of profibrotic TGF-β1 is amplified 9-fold after 72h of culture at 2% oxygen [13] and hypoxia is a powerful stimulus to the synthesis and transcription of the procollagen gene (COL1A1) [14]. In murine fibroblasts, hypoxia decreases MMP9 activity by 64% and mRNA by 80% [15].

Hypoxia's foremost influence is to generate adaptive responses, such as angiogenesis and cell proliferation, aimed at cellular survival, which depend or not on the hypoxia inducible factor (HIF)1 pathway. HIF1 is a transcription factor composed of a HIF1A subunit with an oxygen-dependent degradation domain (ODD) having a half-life of less than 5min in normoxia [16]. In normoxia, prolyl hydroxylase (PHD) can efficiently hydroxylate the HIF1A subunit, leading to rapid degradation by the proteasome, whereas hypoxia prevents this reaction [17]. HIF1A translocates to the nucleus where it forms a heterodimer with aryl hydrocarbon receptor nuclear translocator (ARNT) and binds to DNA on hypoxia response elements (HREs) in the promoters of target genes [18]. The latter include >60 genes, such as vascular endothelial growth factor (VEGF) and the facilitated glucose transporter solute carrier family 2 member 1 (SLC2A1) [19]. Up-regulation of HIF1A occurs in adult wound healing but not in fetal wound regeneration, implying the contribution of the HIF1A pathway in the pathogenesis of fibrosis and scarring [20].

Given that equine EGT develops in an hypoxic environment which may influence the equine dermal fibroblast (EDF) during the proliferative phase of wound healing, the present study sought to measure ECM gene expression and the mitotic activity of these cells when subjected to hypoxia. In addition, given the clinical propensity of limb wounds for fibrosis, fibroblasts were collected from two different anatomic sites (body and limb) to determine if intrinsic cellular factors contribute to the differences in healing. We hypothesized that hypoxia would increase fibroblast proliferation, that protein synthesis would favor ECM accumulation (increased collagen/decreased MMP expression) and that these changes would differ according to the cell's anatomic origin. An additional aim of the study was to determine whether hypoxia-dependent changes in ECM protein expression are HIF1-dependent. The ultimate objective of the research program is to elucidate the pathogenesis of dermal fibrosis and scarring in an effort to provide new pharmacological targets to prevent and/or resolve EGT.

Back to Article Outline

2. Materials and methods 

2.1. Animals 

Two horses euthanized at the Centre Hospitalier Universitaire Vétérinaire de l’Université de Montréal for conditions unrelated to skin were used for sampling, which was sanctioned by the Canadian Council on Animal Care. Tissue samples were collected immediately upon death.

2.2. Tissue collection, fibroblast extraction and culture 

Following aseptic preparation, a full-thickness 8cm2 sample of skin was taken with a scalpel from both the dorsolateral surface of one metacarpal area and from the lateral thoracic wall. Tissues were rinsed in PBS and transported to the laboratory in DMEM supplemented with gentamicin, amphotericin-B, HEPES, l-glutamine and FBS 10% at 4°C. Cell culture products were from Invitrogen (Burlington, ON, Canada). Tissue were cut into ∼1mm2 pieces after removal of subcutaneous fat, then subjected to enzymatic digestion in DMEM containing 4mg/ml of collagenase at 37°C for 30min. Thereafter, collagenase activity was inactivated with fresh, unsupplemented DMEM and supernatant was collected and centrifuged at 750RPM for 10min. Collagenase digestion was repeated twice. Cell pellets were re-suspended in DMEM then plated on Petri dishes for 48h to allow adherence. Cells were incubated at 37°C in a 5% CO2 humidified environment (i.e. 21% O2) corresponding to normoxia; medium was changed every 3days. Equine dermal fibroblasts (EDF) were identified based on morphology and were passages 4–6. Hypoxia was produced by culturing EDF in an incubator set at 1% O2 in a 5% CO2 humidified environment at 37°C. Alternatively, hypoxia was mimicked by adding 200μM CoCl2 [21] to fibroblasts cultured in normoxia. The HIF1 pathway was inhibited by adding 10nM echinomycin (VWR International, Mississauga, ON, Canada) [22].

2.3. Cell mortality and proliferation 

EDF were seeded in 35mm Petri dishes at a concentration of 6×104cells/dish and left to attach for 24h prior to the experiment. Cell mortality and proliferation were measured using Trypan blue exclusion and counting with a hemacytometer at various times (0h, days 1, 3, 5 and 7) after culturing in normoxia or hypoxia. Experiments were repeated in triplicate for cells from each anatomic site of each horse. Cell divisions (log base 2 of the cell count) as a function of time were measured. Proliferation rate was obtained by linear regression of cell divisions vs. time. Cell mortality was estimated as a percentage ((dead cell/total cell)×100). Four groups were thus compared: limb EDF in normoxia (LN), body EDF in normoxia (BN), limb EDF in hypoxia (LH) and body EDF in hypoxia (BH).

2.4. RNA extraction, reverse transcription and qRT-PCR 

Total RNA was extracted from cells using Qiagen RNeasy® Mini Kit (Qiagen, Mississauga, ON, Canada). Dosage, reverse transcription and qRT-PCR, were done like in a precedent study [23]. Gene-specific PCR primers were designed from equine cDNA sequences available on GenBank using Primer Express Software (Applied Biosystems); sequences for VEGFA were (sense: 5′-CAA CGA CGA GGG CCT AGA GT-3′; anti-sense: 5′-CAT CTC TCC TAT GTG TGG CTT TG-3′) while sequences for SLC2A1 (sense: 5′-CCA TCC TCA TCG CTG TGA TG-3′; anti-sense: 5′-TGC ACC CCC GCT TTC TC-3′). ATCB (sense: 5′-CCG ACG GCC AGG TGA TC-3′; anti-sense: 5′-TCG TGG ATA CCA CAA GAC TCC AT-3′) was used as a housekeeping gene. VEGFA amplicons were 100 base pairs (bp) in length, SLC2A1 amplicons were 100bp whereas those for ATCB contained 100bp. PCR were performed on 1.5μl cDNA in 25μl reaction volume in duplicate. qRT-PCR annealing temperature was 58°C. PCR amplification efficiencies were similar for all genes. Negative controls (water) were included in each run. The relative expression of target gene was calculated using the ΔΔCt method with efficiency correction [24]; the control was a cDNA sample derived from equine scrotal skin.

2.5. Protein extraction and Western blot analyses 

Total protein was extracted from confluent cell cultures using M-PER® Mammalian Protein Extraction Reagent (Pierce, Thermo Fisher Scientific, Rockford, IL, USA) supplemented with Complete, a protease inhibitor cocktail (Roche Applied Science, Laval, QC, Canada). Petri dishes were scraped, on ice, then centrifuged at 600×g for 10min at 4°C. The protein concentration of the supernatant containing whole cell extract was measured with a ND-1000 NanoDrop spectrophotometer. Protein extracts were stored at −80°C until analysis. For HIF1A Western blotting, 160μg protein samples were separated by 7.5% acrylamide SDS-PAGE; for precursor collagen α1 type I (COL1A1), 50μg protein samples and for MMP2 protein, 70μg protein samples were separated on a 10% acrylamide gel, while for VEGFA and cleaved caspase-3 (cleaved CASP3), 100μg protein samples were separated on a 15% acrylamide gel then transferred to PVDF membranes (Hybond-P) (GE Amersham, Pittsburgh, PA, USA). Membranes were blocked in 5% milk, except for HIF1A and VEGFA for which blocking was done in a 2% blocking solution provided in the ECL™ Advance Western Blotting Detection Kit, for 1h at room temperature. Membranes were then incubated with mouse anti-human HIF1A monoclonal antibody (sc-53546; dilution 1:1000; Santa Cruz Biotechnology, Santa Cruz, CA, USA), rabbit anti-bovine collagen type 1 polyclonal antibody (CL50121; dilution 1:1000; Cedarlane Labs, Burlington, ON, Canada), rabbit anti-MMP2 polyclonal antibody (NB200-193; dilution 1:1000; Novus Biologicals, Littleton, CO, USA), or rabbit anti-human cleaved CASP3 polyclonal antibody (cs9661; dilution 1:1000; Cell Signaling Technology, Danvers, MA, USA) for 2h at room temperature. Secondary antibody (sheep anti-rabbit antibody NA931VS, dilution 1:10,000 or donkey anti-rabbit antibody NA934, dilution 1:20,000; GE Amersham, Pittsburgh, PA, USA) was applied for 1h at room temperature. Detection of immunoreactive proteins was achieved with ECL™ Advance Western Blotting Detection Kit for HIF1A and cleaved CASP3 or ECL™ Plus Western Blotting Detection Kit for the other proteins (Amersham Pharmacia Biotech Inc.). α-Tubulin (12G10; dilution 1:100,000; Developmental Studies Hybridoma Bank, Iowa City, IA, USA) was used as a loading control. Autoradiographic images were digitized and densitometric measures were obtained using Kodak 1D version 3.6 software.

2.6. Statistical analysis 

An ANCOVA model, with treatment (LN, BN, LH, BH) and culture time as factors, was used to determine the effects of site and time on proliferation, mortality and protein/gene expression. A priori contrasts, with comparison-wise alpha levels adjusted with the Bonferroni sequential correction to maintain the family-wise error rate at the desired level, were then used to compare pre-selected individual means. A non-parametric Kruskal–Wallis test was used to compare treatments (normoxia, CoCl2 and CoCl2+echinomycin) in the experiment aimed at evaluating HIF1-dependency. All analyses were carried out with a family-wise error rate of 0.05, using SAS v.9.2. (Cary, NC, USA). To be significant, a result needed to be below the postcorrection threshold while a tendency was defined as a p-value between 0.05 and the postcorrection threshold.

Back to Article Outline

3. Results 

3.1. Hypoxia stimulates equine dermal fibroblast proliferation 

To determine the potential effects of hypoxia on proliferation and mortality, EDFs from body and limb were cultured for 7days in normoxic or hypoxic conditions. The morphology of EDFs did not vary according to anatomic origin or culture conditions (not shown). However, comparisons of proliferation rates made over 7days of culture showed significant differences between groups (Fig. 1). Proliferation rate of EDF in hypoxia was higher than in normoxia whether for body EDF (p=0.0002) or limb EDF (p<0.0001). Conversely, no significant differences were seen between EDF from different anatomic sites in either condition.

  • View full-size image.
  • Fig. 1. 

    Proliferation of EDF in normoxia and hypoxia. Cell proliferation measured using Trypan blue exclusion and counting. (A) Graph showing cell division (log base 2 of cell count) as a function of time (hypoxic chamber calibrated at 1% O2). N=3 for each group. Symbol: ♢, limb EDF in normoxia; □, body EDF in normoxia; ▵, limb EDF in hypoxia; and ×, body EDF in hypoxia. (B) Table compiling proliferation rate (linear regression of cell division as a function of time) and coefficient of determination (R2) for the four groups (mean±SEM; **statistically significant difference between normoxia and hypoxia, limb p<0.0001 and body p=0.0002).

EDF mortality data measured at days 0, 1, 3, 5 and 7 are shown in Table 1. No significant differences were found between groups at any time.

Table 1. Mortality of EDF in normoxia and hypoxia.
GroupDay 0Day 1Day 3Day 5Day 7
%±SEM%±SEM%±SEM%±SEM%±SEM
BN4.7±0.73.9±0.85±18±33.2±0.5
BH8±43.3±0.51.3±0.42.4±0.41.1±0.2
LN2.9±0.33.6±0.22.4±0.63.4±0.52.8±0.6
LH6±32.1±0.42.01±0.041.9±0.81.0±0.2

Cell mortality measured by Trypan blue exclusion staining. Table compiling cell mortality percentage at each day of culture (mean±SEM). Hypoxia obtained in hypoxic chamber calibrated at 1% O2. N=3 for each group (BN: body EDF in normoxia; BH: body EDF in hypoxia; LN: limb EDF in normoxia; LH: limb EDF in hypoxia).

EDF derived from skin of the limb showed a small, non-significant increase in expression of cleaved CASP3 in response to hypoxia at 3, 12 and 48h (Fig. 2). No changes in protein expression were measured in body EDF cultured for 48h in hypoxia. Furthermore, no differences between body and limb EDF were statistically significant at any time over 48h of hypoxia.

  • View full-size image.
  • Fig. 2. 

    Cleaved caspase-3 (CASP3) protein expression in EDF cultured in CoCl2 hypoxia. Western blot analysis of CASP3 in cell lysate total protein extraction. α-Tubulin was used as a loading control protein. (A) Histogram compiling results (mean±SEM; *tendency p<0.05; non significant after Bonferroni correction). Results are compared to normoxia values (time 0) and data are presented as fold-increase. Grey columns represent body EDF values and black columns represent limb EDF values. N=4 for each group at each time. (B) Western blot bands of CASP3 and α-tubulin for limb EDF.

3.2. Hypoxia mimetic (CoCl2) induces the expression of HIF1A 

A hypoxia mimetic was used for the protein expression studies in an effort to avoid fluctuations in HIF1A levels. The ability of CoCl2 to act as a hypoxia mimetic in the subsequent experiments was validated by measuring its effect on the expression of HIF1A. CoCl2 consistently and rapidly induced the expression of HIF1A protein (4-fold increase) at 3 and 6h; at 12h, HIF1A concentrations remained elevated (2-fold) but then declined to finally reach control levels by 24h (data not shown). Moreover, mRNA expression of both target genes VEGFA and SLC2A1 was increased up to 6-fold 6h following addition of CoCl2 (Fig. 3), confirming the reliability of the chosen hypoxia mimetic.

  • View full-size image.
  • Fig. 3. 

    Vascular endothelial growth factor A (VEGF A) and solute carrier family 2 (facilitated glucose transporter) member 1 (SLC2A1) mRNA expression in EDF exposed to 6h treatment of CoCl2 hypoxia, with or without echinomycin. qRT-PCR analysis of (A) VEGF A and (B) SLC2A1 in cell lysate RNA extraction. b-Actine (ACTB) used as a housekeeping gene. Histogram compiling results (mean±SEM). N=2 for each group.

3.3. Hypoxia regulates the expression of ECM-associated proteins (COL1A1 and MMP2) 

As hypoxia is thought to be involved in the pathogenesis of fibrosis, we next tested its effects on protein expression of COL1A1 and MMP2 by EDFs from body and limb. Hypoxia significantly upregulated the expression of COL1A1 in limb EDF by 3h (Fig. 4A). Concentrations then began to decrease between 12 and 24h of hypoxia. By 48h, COL1A1 concentrations in limb EDF had returned to normoxic values. Body EDF showed the same pattern of expression but the increase, in response to hypoxia, was never statistically significant. No differences in expression of COL1A1 protein between body and limb EDF were statistically significant at any time.

  • View full-size image.
  • Fig. 4. 

    Precursor collagen α1 type I (COL1A1) and matrix metalloproteinase 2 (MMP2) protein expression in EDF cultured in CoCl2 hypoxia. Western blot analysis of (A) COL1A1 or (B) MMP2 in cell lysate total protein extraction. α-Tubulin used as a loading control protein. Histogram compiling results (mean±SEM; **statistically significant difference between normoxia and hypoxia p<0.005, *tendency p<0.05; non significant after Bonferroni correction). Results are compared to normoxia values (time 0) and data are presented as fold-increase. Grey columns represent body EDF values and black columns represent limb EDF values. N=4 for each group at each time. Western blot bands of (C) COL1A1 or (D) MMP2 and α-tubulin for limb EDF.

Acute hypoxia led to a steady drop of MMP2 protein expression in EDF cell lysates. Limb EDF expressed less MMP2 protein at 3h of hypoxia (p=0.03 non significant after Bonferroni correction). At 6h, expression continued its decrease to ultimately reach a level of 0.34±0.03-fold relative to cells in normoxia by 48h (Fig. 4B). MMP2 protein expression by body EDF followed the same pattern. While the decrease was non significant at 3h of hypoxia, from 6h forwards, the drop was significant until 48h inclusively. As for COL1A1 expression, no differences between body and limb EDF were statistically significant at any time.

3.4. Hypoxia regulates COL1A1 and MMP2 via HIF1A 

In order to determine whether the changes observed in response to CoCl2 treatment were HIF1A-dependent or not, additional EDFs were treated with echinomycin, which inhibits HIF transcriptional activity by suppressing its binding to the HRE site of target genes [25]. The addition of 10nM echinomycin to the CoCl2-induced hypoxic culture medium resulted in a drop of 30% in mRNA expression of VEGFA and of 65% in expression of SLC2A1 (Fig. 3).

Limb EDFs were subjected to one of three treatments (normoxia; CoCl2; CoCl2+echinomycin) for 6h, after which COL1A1 and MMP2 protein expression was measured. COL1A1 expression showed a statistically significant increase in hypoxia compared to normoxia and to hypoxia+echinomycin (Fig. 5A). In the case of MMP2, hypoxia decreased the expression while addition of echinomycin returned it to a level slightly higher than in normoxia (Fig. 5B). The differences between hypoxia and hypoxia+echinomycin were statistically significant.

  • View full-size image.
  • Fig. 5. 

    Precursor collagen α1 type I (COL1A1) and matrix metalloproteinase 2 (MMP2) protein expression in EDF exposed to 6h treatment of CoCl2 hypoxia, with or without echinomycin. Western blot analysis of (A) COL1A1 or (B) MMP2 in cell lysate total protein extractions. α-Tubulin used as a loading control protein. Histogram compiling results (mean±SEM; **statistically significant difference between treatments: COL1A1 p<0.01; MMP2 p<0.03). Results are normalized to normoxic values and data are presented as fold-increase. N=7 for normoxia and hypoxia; N=3 for hypoxia+echinomycin.

Back to Article Outline

4. Discussion 

Dysregulation of the wound healing process is a common pathology and can affect all species. Fibroproliferative disorders, such as hypertrophic scars or keloids in human [2] or exuberant granulation tissue (EGT) in horses, are examples. EGT is a major cause of retirement from show and race [7] and it exerts a significant financial impact on the horse industry. Recent studies conducted in an effort to elucidate the development of dermal fibroproliferative disorders in horses demonstrated a relative state of hypoxia in limb compared to body wounds, during early healing [26]. The present study attempts to clarify the influence of hypoxia on increased fibroblast proliferation, deficient apoptosis and aberrant collagen metabolism (abundant synthesis/impaired lysis) characteristic of equine EGT. In addition, contribution of the HIF1A pathway to ECM regulation was verified.

Dermal fibroblasts are the predominant cell type in both equine wound granulation tissue and EGT. The positive influence of acute hypoxia (1% O2) on their proliferation (Fig. 1) echoes what is reported for human dermal [27] and renal fibroblasts [5]. However, cell mortality and expression of cleaved CASP3, the ultimate effector of the apoptosis pathway, did not change in response to hypoxia (Table 1 and Fig. 2). Saed and Diamond [28] report that the influence of hypoxia on apoptosis of human peritoneal fibroblasts depends on the nature of the cell: normal fibroblasts become increasingly apoptotic when cultured at 2% O2 whereas hypoxia decreases the rate of apoptosis in adhesion-derived fibroblasts. It would be interesting to study the influence of hypoxia on EDF derived from EGT since previous studies show that both equine EGT [10] and human keloid [29] display deficient fibroblast apoptosis. While the effect of hypoxia on apoptosis merits further investigation, it must be emphasized that in general, severe and prolonged hypoxia initiates apoptosis whereas cells often adapt to acute, mild hypoxia and survive [30].

The anatomic origin of the fibroblast does not appear to exert a significant influence on its response to hypoxia. Microenvironmental factors must account for the distinct healing and scarring/fibrosis patterns observed in horses. Due to their closer proximity to the ground, limb wounds are more susceptible to colonization by microorganisms [31]. The relative state of hypoxia characterizing the healing process at limb level [9] may affect wound defense capabilities and heighten the risk of infection and the development of bacterial biofilm [32], [33]. Lower tissue perfusion also afflicts limb wounds [26]. Hypoperfusion may alter temperature but also moisture, pH and gaseous exchanges at the wound site [34]. These microenvironmental differences may exert a greater influence on fibroblast behavior than its specific anatomic origin.

The choice to study the influence of acute, rather than chronic, hypoxia was based upon data from our laboratory, showing that the major differences in oxygenation between body and limb wounds occur 24 and 48h following wounding [9]. Hypoxia was thus considered a trigger to the development of EGT. Because a trend towards significantly greater hypoxia in limb wounds is still present at 4weeks of healing [9], future studies should measure the influence of chronic hypoxia (more than 6 passages in hypoxia) to characterize the condition of fibroblasts present in well-established EGT. Siddiqui et al. have shown that the stimulatory effects seen in acute hypoxia are suppressed or eliminated on chronic exposure of human dermal fibroblasts to hypoxic conditions [27]. This implies that EDF derived from EGT of limb wounds might suffer from slowed metabolism and/or be unable to progress to a contractile phenotype, explaining the deficient wound contraction characterizing limb wounds of horses [33]. In support of this, rat skin myofibroblast differentiation and contractility were significantly reduced by culturing for 5days at 2% O2 [35].

Despite several studies documenting changes in ECM proteins and fibrogenic growth factors in models of equine EGT and other dermal fibroproliferative disorders of various species, the mechanisms underlying accumulation of ECM remain obscure. Recently, hypoxia was incriminated in systemic sclerosis [36], scleroderma [37] and keloids [38]. Changes in ECM proteins were measure using a hypoxia mimetic, CoCl2, because it induces HIF1A more rapidly than physical hypoxia and generates a robust transcriptional response that is more repeatable. Given the largely synthetic function of dermal fibroblasts and their increased proliferative rate under hypoxic conditions, it is no surprise that collagen production was seen to increase in the present study. Perhaps more significant, in parallel with changes in collagen production, hypoxia also altered ECM turnover, suppressing expression of MMP2 (Fig. 4A and B). This collagenase degrades collagens I, III, IV, V and XI during tissue remodeling. Saed et al. showed that murine fibroblasts decrease MMP9 activity by 64% and mRNA expression by 80% in hypoxia (2% O2), but observed no changes in MMP2 [15]. Comparisons with the present study are impossible since protein expression rather than mRNA expression or protein activity was measured. ECM maturation is complex, involving a plethora of enzymes; ultimately, the fine balance between matrix-enhancing and matrix-degrading molecules will determine the clinical outcome.

Fibrosis is characterized by TGF-β1 release and fibroblast activation with a consequent increase in collagen synthesis and decrease in turnover [39]. Moreover, fibroblast synthesis of TGF-β1 is amplified 9-fold after 72h of culture at 2% oxygen [13] and neutralizing antibodies against TGF-β1 completely abrogate the induction of collagen in human fibroblasts cultured under hypoxic conditions [36]. Given the persistence of elevated concentrations of TGF-β1 in limb wounds of horses predisposed to the development of EGT [40], [41], it would be interesting to correlate the hypoxia-induced increase in COL1A1 and concomitant decrease in MMP2 synthesis (Fig. 4A and B) with TGF-β1 concentrations.

Changes in expression of COL1A1 and MMP2 under hypoxia were canceled by the addition of echinomycin (Fig. 5A and B). Echinomycin, a cyclic peptide antibiotic, specifically binds the 5′-CGTG-3′ sequence, which forms the core of the HRE found at the promoters of HIF-target genes [42]. COL1A1 and MMP2 regulation by hypoxia thus appear to depend on HIF1 signaling (Fig. 6). MMP2 is generally reported as stable or overexpressed in hypoxia, however, downregulation by HIF1 has been shown [43], [44], [45]. Results of the current study indicate that inhibition of HIF1 activity may prevent fibrosis and suggest the investigation of this pathway for the development of therapeutic strategies for equine EGT.

In conclusion, data suggest that hypoxia, via the HIF1 pathway, coordinately up-regulates matrix production and decreases matrix turnover in EDF. Results support a role for hypoxia in the pathogenesis of fibrosis (EGT), whose cardinal feature is ECM accumulation. Further studies should investigate the influence of chronic hypoxia and be conducted on other wound healing cells (in particular inflammatory cells and myofibroblasts), and should aim to study more proteins involved in ECM accumulation as well as their link with HIF1A. A better understanding of molecular pathways implicated in equine EGT will provide a basis for the development of appropriate therapies.

Back to Article Outline

Acknowledgments 

The 12G10 anti-a-tubulin developed by Frankel, J. and Nelsen, E.M. was obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the NICHD and maintained by The University of Iowa, Department of Biology, Iowa City, IA 52242. The authors thank Genevieve Langevin-Carpentier for technical assistance, as well as Dr. Guy Beauchamp for statistical analysis.

Back to Article Outline

References 

  1. Gurtner GC, Werner S, Barrandon Y, Longaker MT. Wound repair and regeneration. Nature. 2008;453:314–321
  2. Li J, Chen J, Kirsner R. Pathophysiology of acute wound healing. Clin Dermatol. 2007;25:9–18
  3. Mollenhauer J, Bayreuther K. Donor-age-related changes in the morphology, growth potential, and collagen biosynthesis in rat fibroblast subpopulations in vitro. Differentiation. 1986;32:165–172
  4. Broughton G, Janis JE, Attinger CE. Wound healing: an overview. Plast Reconstr Surg. 2006;117:1e-S–32e-S
  5. Norman JT, Clark IM, Garcia PL. Hypoxia promotes fibrogenesis in human renal fibroblasts. Kidney Int. 2000;58:2351–2366
  6. Wilmink JM. Differences in wound healing between horses and ponies. In:  Stashak TS,  Theoret CL editor. Equine wound management. Iowa: Wiley-Blackwell; 2008;p. 29–46
  7. Perkins NR, Reid SW, Morris RS. Profiling the New Zealand thoroughbred racing industry. 2. Conditions interfering with training and racing. N Z Vet J. 2005;53:69–76
  8. Atiyeh BS, Costagliola M, Hayek SN. Keloid or hypertrophic scar: the controversy: review of the literature. Ann Plast Surg. 2005;54:676–680
  9. Celeste CJ, Deschene K, Riley CB, Theoret CL. Regional differences in wound oxygenation during normal healing in an equine model of cutaneous fibroproliferative disorder. Wound Repair Regen. 2011;19:89–97
  10. Lepault E, Celeste C, Dore M, Martineau D, Theoret CL. Comparative study on microvascular occlusion and apoptosis in body and limb wounds in the horse. Wound Repair Regen. 2005;13:520–529
  11. Zhang Q, Wu Y, Ann DK, Messadi DV, Tuan TL, Kelly AP, et al. Mechanisms of hypoxic regulation of plasminogen activator inhibitor-1 gene expression in keloid fibroblasts. J Invest Dermatol. 2003;121:1005–1012
  12. Tandara AA, Mustoe TA. Oxygen in wound healing – more than a nutrient. World J Surg. 2004;28:294–300
  13. Falanga V, Qian SW, Danielpour D, Katz MH, Roberts AB, Sporn MB. Hypoxia upregulates the synthesis of TGF-beta 1 by human dermal fibroblasts. J Invest Dermatol. 1991;97:634–637
  14. Falanga V, Martin TA, Takagi H, Kirsner RS, Helfman T, Pardes J, et al. Low oxygen tension increases mRNA levels of alpha 1 (I) procollagen in human dermal fibroblasts. J Cell Physiol. 1993;157:408–412
  15. Saed GM, Zhang W, Diamond MP. Effect of hypoxia on stimulatory effect of TGF-beta 1 on MMP-2 and MMP-9 activities in mouse fibroblasts. J Soc Gynecol Investig. 2000;7:348–354
  16. Pouyssegur J, Dayan F, Mazure NM. Hypoxia signalling in cancer and approaches to enforce tumour regression. Nature. 2006;441:437–443
  17. Distler JH, Wenger RH, Gassmann M, Kurowska M, Hirth A, Gay S, et al. Physiologic responses to hypoxia and implications for hypoxia-inducible factors in the pathogenesis of rheumatoid arthritis. Arthritis Rheum. 2004;50:10–23
  18. Weidemann A, Johnson RS. Biology of HIF-1alpha. Cell Death Differ. 2008;15:621–627
  19. Lee JW, Bae SH, Jeong JW, Kim SH, Kim KW. Hypoxia-inducible factor (HIF-1)alpha: its protein stability and biological functions. Exp Mol Med. 2004;36:1–12
  20. Scheid A, Wenger RH, Schaffer L, Camenisch I, Distler O, Ferenc A, et al. Physiologically low oxygen concentrations in fetal skin regulate hypoxia-inducible factor 1 and transforming growth factor-beta3. Faseb J. 2002;16:411–413
  21. Mace KA, Yu DH, Paydar KZ, Boudreau N, Young DM. Sustained expression of Hif-1alpha in the diabetic environment promotes angiogenesis and cutaneous wound repair. Wound Repair Regen. 2007;15:636–645
  22. Kim J, Bagchi IC, Bagchi MK. Signaling by hypoxia-inducible factors is critical for ovulation in mice. Endocrinology. 2009;150:3392–3400
  23. Deschene K, Celeste C, Boerboom D, Theoret CL. Constitutive expression of hypoxia-inducible factor-1 alpha in keratinocytes during the repair of skin wounds in horses. Wound Repair Regen. 2011;19:250–259
  24. Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res. 2001;29:e45
  25. Kong D, Park EJ, Stephen AG, Calvani M, Cardellina JH, Monks A, et al. Echinomycin, a small-molecule inhibitor of hypoxia-inducible factor-1 DNA-binding activity. Cancer Res. 2005;65:9047–9055
  26. Celeste CJ, Deschene K, Riley CB, Theoret CL. Kinetics of blood flow and anatomical-site differences in tissue perfusion during wound healing in an equine model of cutaneous fibroproliferative disorder. Vet Surg; under review.
  27. Siddiqui A, Galiano RD, Connors D, Gruskin E, Wu L, Mustoe TA. Differential effects of oxygen on human dermal fibroblasts: acute versus chronic hypoxia. Wound Repair Regen. 1996;4:211–218
  28. Saed GM, Diamond MP. Apoptosis and proliferation of human peritoneal fibroblasts in response to hypoxia. Fertil Steril. 2002;78:137–143
  29. Ladin DA, Hou Z, Patel D, McPhail M, Olson JC, Saed GM, et al. p53 and apoptosis alterations in keloids and keloid fibroblasts. Wound Repair Regen. 1998;6:28–37
  30. Greijer AE, van der Wall E. The role of hypoxia inducible factor 1 (HIF-1) in hypoxia induced apoptosis. J Clin Pathol. 2004;57:1009–1014
  31. Westgate SJ, Percival SL, Knottenbelt DC, Clegg PD, Cochrane CA. Microbiology of equine wounds and evidence of bacterial biofilms. Vet Microbiol. 2011;150:152–159
  32. Westgate SJ, Percival SL, Knottenbelt DC, Clegg PD, Cochrane CA. Chronic equine wounds: what is the role of infection and biofilms?. Wounds. 2010;22:138–145
  33. Wilmink JM, Stolk PW, van Weeren PR, Barneveld A. Differences in second-intention wound healing between horses and ponies: macroscopic aspects. Equine Vet J. 1999;31:53–60
  34. Knottenbelt DC. Basic wound management. In:  Knottenbelt DC editors. Handbook of equine wound management. London: Saunders; 2003;p. 54–73
  35. Modarressi A, Pietramaggiori G, Godbout C, Vigato E, Pittet B, Hinz B. Hypoxia impairs skin myofibroblast differentiation and function. J Invest Dermatol. 2010;130:2818–2827
  36. Distler JH, Jungel A, Pileckyte M, Zwerina J, Michel BA, Gay RE, et al. Hypoxia-induced increase in the production of extracellular matrix proteins in systemic sclerosis. Arthritis Rheum. 2007;56:4203–4215
  37. Hong KH, Yoo SA, Kang SS, Choi JJ, Kim WU, Cho CS. Hypoxia induces expression of connective tissue growth factor in scleroderma skin fibroblasts. Clin Exp Immunol. 2006;146:362–370
  38. Zhang Q, Wu Y, Chau CH, Ann DK, Bertolami CN, Le AD. Crosstalk of hypoxia-mediated signaling pathways in upregulating plasminogen activator inhibitor-1 expression in keloid fibroblasts. J Cell Physiol. 2004;199:89–97
  39. Kisseleva T, Brenner DA. Mechanisms of fibrogenesis. Exp Biol Med (Maywood). 2008;233:109–122
  40. Theoret CL, Barber SM, Moyana TN, Gordon JR. Expression of transforming growth factor beta(1), beta(3), and basic fibroblast growth factor in full-thickness skin wounds of equine limbs and thorax. Vet Surg. 2001;30:269–277
  41. van den Boom R, Wilmink JM, O’Kane S, Wood J, Ferguson MW. Transforming growth factor-beta levels during second-intention healing are related to the different course of wound contraction in horses and ponies. Wound Repair Regen. 2002;10:188–194
  42. Vlaminck B, Toffoli S, Ghislain B, Demazy C, Raes M, Michiels C. Dual effect of echinomycin on hypoxia-inducible factor-1 activity under normoxic and hypoxic conditions. Febs J. 2007;274:5533–5542
  43. Fan RH, Chen PS, Zhao D, Zhang WD. Hypoxia induced by CoCl2 influencing the expression and the activity of matrix metalloproteinase-2 in rat hepatic stellate cells. Zhonghua Gan Zang Bing Za Zhi. 2007;15:654–657
  44. Medina-Torres CE, Mason SL, Floyd RV, Harris PA, Mobasheri A. Hypoxia and a hypoxia mimetic up-regulate matrix metalloproteinase 2 and 9 in equine laminar keratinocytes. Vet J. 2011;
  45. Ye H, Zheng Y, Ma W, Ke D, Jin X, Liu S, et al. Hypoxia down-regulates secretion of MMP-2, MMP-9 in porcine pulmonary artery endothelial and smooth muscle cells and the role of HIF-1. J Huazhong Univ Sci Technolog Med Sci. 2005;25(407):382–384

☆ Research was funded by the Natural Sciences and Engineering Research Council (NSERC) of Canada.

PII: S0923-1811(11)00281-7

doi:10.1016/j.jdermsci.2011.09.006

Journal of Dermatological Science
Volume 65, Issue 1 , Pages 12-18, January 2012

Article Tools

* Image Usage & Resolution