Impaired Wound Repair in Adult Endoglin Heterozygous Mice Associated with Lower NO Bioavailability

ORIGINAL ARTICLE

Impaired Wound Repair in Adult Endoglin Heterozygous Mice Associated with Lower NO Bioavailability Eduardo Pe´rez-Go´mez1,5, Mirjana Jerkic2,3, Marta Prieto2, Gaelle del Castillo1, Ester Martı´ n-Villar1, Michelle Letarte3, Carmelo Bernabeu4, Fernando Pe´rez-Barriocanal2, Miguel Quintanilla1 and Jose´ M. Lo´pez-Novoa2 Endoglin (Eng) is a transmembrane glycoprotein that is mainly expressed in endothelial cells, but it is also present in the epidermis and skin appendages. To address the role of Eng in cutaneous wound healing, we compared the kinetics of reepithelialization in Eng heterozygous null (Eng þ /  ) mice and their normal littermates (Eng þ / þ ) following skin wounds. The wound area was significantly larger in Eng þ /  than in Eng þ / þ mice from 2 to 8 days after injury; overall wound closure was delayed by 1 to 2 days. In Eng þ /  mice, keratinocytes at the wound edges exhibited impaired proliferation but were more migratory, as shown by their elongated morphology and increased keratin 17 expression. Inhibition of nitric oxide (NO) synthesis delayed healing in Eng þ / þ but not in Eng þ /  mice. Administration of the NO donor LA-803 accelerated wound closure in Eng þ /  mice, with no effect on normal littermates. The acute stimulation with 12-O-tetradecanoylphorbol-13-acetate (TPA) enhanced Eng expression in mouse epidermal keratinocytes in vivo and in vitro associated with hyperproliferation. Similarly, the skin of Eng þ /  mice failed to mount a hyperplastic response to acute stimulation with TPA. These results demonstrate an important involvement of Eng in wound healing that is associated with NO bioavailability. Journal of Investigative Dermatology (2014) 134, 247–255; doi:10.1038/jid.2013.263; published online 18 July 2013

INTRODUCTION Physiological wound healing is a highly organized process requiring tight temporal and spatial coordination of various events and signaling networks (Martin, 1997; Grose and Werner, 2004; Gurtner et al., 2008). Cutaneous wound healing begins immediately after a skin injury in an attempt to provide a protective barrier against further external stimuli or infections. Proper skin wound repair requires the coordinated activation of several cell and tissue responses, including reepithelialization, connective tissue contraction, 1 Instituto de Investigaciones Biome´dicas Alberto Sols, Spanish National Research Council (CSIC), Autonomous University of Madrid (UAM), Madrid, Spain; 2Renal and Cardiovascular Physiopathology Unit, Instituto Reina Sofia de Investigacio´n, Department of Physiology and Pharmacology, University of Salamanca, Salamanca, Spain; 3Molecular Structure and Function Program, Hospital for Sick Children, Department of Immunology, University of Toronto, Toronto, Ontario, Canada and 4Centro de Investigaciones Biolo´gicas, CSIC, and Centro de Investigacio´n Biome´dica en Red de Enfermedades Raras (CIBERER), Madrid, Spain 5

Current address: Department of Biochemistry and Molecular Biology I, School of Biology, Complutense University of Madrid, Madrid, Spain

Correspondence: Jose M. Lopez-Novoa, Edificio Departamental, Campus Miguel de Unamuno, Universidad de Salamanca, Salamanca, 37007, Spain. E-mail: [email protected] Abbreviations: ALK, activin-like kinase receptor; Eng, endoglin; K17, keratin 17; L-NAME, Nw-Nitro-L-arginine methyl ester; NO, nitric oxide; NOS, nitric oxide synthase; TGF-b, transforming growth factor-b; TPA, 12-O-tetradecanoylphorbol-13-acetate Received 13 September 2012; revised 18 April 2013; accepted 19 May 2013; accepted article preview online 13 June 2013; published online 18 July 2013

& 2014 The Society for Investigative Dermatology

and angiogenesis, the latter process leading to the formation of a dense network of blood vessels in the wound granulation tissue. A robust inflammatory response commencing soon after tissue damage, followed by tissue remodeling, is also an important step in tissue repair (Martin, 1997). The wound healing process is interactive and dynamic, involving the extracellular matrix, multiple soluble mediators, blood cells, keratinocytes, and parenchymal cells (Coulombe, 2003). Delayed skin wound healing is a serious complication in diabetes and is caused primarily by microangiopathy and impaired cutaneous blood flow, hypoxia, accelerated inflammation, edema, and endothelial dysfunction (Reiber et al., 1999; Martin et al., 2003). Moreover, expression of vascular endothelial growth factor-A is markedly reduced in wounds of diabetic patients (Frank et al., 1995; Kampfer et al., 2001). The genetic control of wound healing involves multiple families of growth factors, of which transforming growth factor-b (TGF-b) is a classic example (Singer and Clark, 1999). TGF-b expression occurs throughout wound healing and regulates many processes involved in tissue repair (Valluru et al., 2011a). Endoglin (CD105) is a 180-kDa homodimeric transmembrane glycoprotein that modulates cellular responses to ligands of the TGF-b superfamily (Gougos and Letarte, 1990; Barbara et al., 1999; Li et al., 2000; Lo´pez-Novoa and Bernabe´u, 2010). Endoglin (Eng) is upregulated in tissues undergoing angiogenesis (Fonsatti et al., 2003; Jerkic et al., 2006a; Lo´pez-Novoa and Bernabe´u, 2010), www.jidonline.org

247

E Pe´rez-Go´mez et al. Wound Repair and Endoglin

RESULTS Temporal and spatial expression of Eng during repair

We first analyzed Eng expression during wound healing in normal mice (Supplementary Figure S1 online). Eng expression was low in control skin (day 0), but gradually increased after wounding, reaching maximal values 4 days after the injury and decreasing by 8 days (Supplementary Figure S1a online). The localization of Eng expression in the wounded area was also determined by immunohistochemical analysis. Basal-like keratinocytes at the wound edges and blood vessels of the subepithelial area were the main cellular sources of Eng protein (Supplementary Figure S1b online). Note the faint expression of Eng in keratinocytes in the interfollicular, nonwounded skin (arrowheads) and the high expression of Eng in the keratinocytes of the reepithelializing skin (black arrows, insert). Delayed wound closure in Eng-deficient mice

We introduced full-thickness wounds into the dorsal skin of Eng þ / þ and Eng þ /  mice and analyzed wound closure 2–14 days after the injury, as indicated in the Materials and Methods. Eng þ /  mice exhibited delayed wound closure compared with Eng þ / þ mice, showing a significantly larger wound from days 2 to 8 after injury (Figure 1a). At 4 days after wounding, B60% wound closure was observed in Eng þ / þ mice as compared with 40% in Eng þ /  mice. In Eng þ / þ mice, wound closure was complete by day 12, whereas in Eng þ /  animals this occurred 1 or 2 days later. The rate of healing, expressed as the half closing time, was 3.2±0.2 days in Eng þ / þ and 4.2±0.2 days in Eng þ /  mice (Figure 1c). 248

Journal of Investigative Dermatology (2014), Volume 134

Wound area (%)

a

100 80

Eng+/+ Eng+/– Eng+/+, L-NAME Eng+/–, L-NAME

** **

60

**

40

**

20 0 2

Wound area (%)

b

4

6 8 10 Days after injury

100 80

*

&

60

&

14

Eng+/+ Eng+/– Eng+/+, LA-803 Eng+/–, LA-803

*

40

12

* &

* &

20 0 2

4

10 6 8 Days after injury

12

14

c

HCT (days)

and inhibition of its expression in endothelial cells impairs this process (Li et al., 2000). Endoglin-null mice die at mid-gestation from defective angiogenesis and severe cardiovascular abnormalities, whereas Endoglin heterozygous (Eng þ /  ) mice have a normal lifespan (Bourdeau et al., 1999; Li et al., 1999; Arthur et al., 2000). Eng has been shown to regulate vascular function through modulation of endothelial nitric oxide (NO) synthase (NOS) expression and activity, with Eng haploinsufficiency being associated with reduced NO production and increased endothelial NOS–derived superoxide production (Jerkic et al., 2004; Toporsian et al., 2005; Jerkic et al., 2006a). Reduced NO production has now been demonstrated in lungs, heart, and liver of Eng þ /  mice (Jerkic et al 2012). Eng has also been shown to have a major role in the regulation of angiogenesis (Jerkic et al., 2006b; Lebrin and Mummery, 2008). Eng is expressed in keratinocytes of the interfollicular epidermis and hair bulbs where it acts to attenuate TGF-b signaling (Quintanilla et al., 2003; Pe´rez-Go´mez et al., 2007). Eng expression increases after dermal wound in association with inflammation (Torsney et al., 2002). In addition, Eng is overexpressed in some skin diseases such as scleroderma or psoriasis (Rulo et al., 1995; Maring et al., 2012). The purpose of this study was to assess the role of Eng in epidermal reepithelialization by studying the effect of Eng insufficiency on cutaneous healing after an excisional wound.

5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0

#

# &

Eng+/+

Without treatment L-NAME LA-803

Eng+/–

Figure 1. Wound healing is delayed in endoglin (Eng)-deficient mice. (a) Wound closure kinetics in normal and mutant mice before and after oral administration of the inhibitor of nitric oxide (NO) synthesis Nw-Nitro-L-arginine methyl ester (L-NAME). (b) Wound closure kinetics in normal and mutant mice before and after oral administration of the NO donor LA-803. Data are shown as mean±SEM (n ¼ 11 animals per group). *Po0.01 versus Eng þ / þ ; &Po0.01 versus Eng þ /  . (c) The half closing time (HCT) of the wounds for the different treatments was determined. #Po0.01 versus Eng þ / þ without treatment; &Po0.01 versus Eng þ /  without treatment.

Oral administration of the NOS inhibitor Nw-Nitro-L-arginine methyl ester (L-NAME) significantly delayed wound healing in Eng þ / þ mice, but the effect was minimal in Eng þ /  mice (Figure 1a and c). In contrast, administration of the NO donor LA-803 accelerated the rate of healing in Eng þ /  mice, with almost no effect in Eng þ / þ mice (Figure 1b and c). These results suggest that reduced NO levels, previously reported in several organs of Eng þ /  mice (Jerkic et al., 2012), have a crucial role in the delayed wound healing observed in Eng þ /  mice. Histological examination of the wounds revealed that 4 days after the injury the hyperproliferative epithelium at the wound edges was thinner in Eng þ /  mice as compared with Eng þ / þ mice (Figure 2a and b). However, keratinocytes forming the new epithelium in Eng þ /  wounds exhibited a more elongated morphology than those of normal mice (see below). The difference in thickness of the regenerated

E Pe´rez-Go´mez et al. Wound Repair and Endoglin

P

S

S P G G

Figure 2. Reepithelialization is defective in endoglin (Eng)-deficient mice. Representative histological appearances of wounds in (a, c) wild-type and (b, d) Eng þ /  mice (a, b) 4 days and (c, d) 12 days after wounding. Black arrows in a and b indicate the estimated limit between unwounded and reepithelialized skin, whereas red arrow marks the tip of the regenerating tongue. G, granulation tissue; P, panniculus carnosus; S, scab. Scale bar ¼ 210 mm.

epidermis between Eng þ / þ and Eng þ /  mice was also evident at 12 days after the injury when wounds were completely re-epithelialized (Figure 2c and d). Thus, the regenerated epithelium in Eng þ /  mice was formed by only 1–2 layers (Figure 2d), which was in clear contrast with the 3–5 layers that were formed in the new epidermis of Eng þ / þ mice (Figure 2c). To substantiate this difference further, epidermal thickness was measured in five sections of each genotype, each section corresponding to a mouse (four fields per section). Epidermal thickness was 153±11 mm in Eng þ / þ mice and 72±5 mm in Eng þ /  mice (Po0.01).

Figure 3. Reduced keratinocyte proliferation in the wounded epidermis of Endoglin (Eng)-deficient mice. Representative Ki67 immunostaining (brownstained nuclei in the epidermis) in (a, b) unwounded and (c, d) wounded skin at 4 days and (e, f) 12 days after the injury of (a, c, e) normal and (b, d, f) Eng þ /  mice. In c and d, the tip of the regeneration tongue is marked with a red arrow. In e and f, approximate limits of regenerating skin in the wound are marked with red arrows. Scale bar ¼ 210 mm.

Disturbed epidermal proliferation in Eng-deficient mice

To determine the proportion of proliferating keratinocytes in the regenerating epithelium of normal and mutant mice, we performed Ki67 staining in the epidermis adjacent to the wounds in unwounded areas, and no evident differences in terms of the number of Ki67-positive keratinocytes in the basal layer were observed between Eng þ / þ and Eng þ /  (Figure 3a and b). Nevertheless, in wounded areas, a clear reduction in the number of Ki67-positive keratinocytes was observed in mutant mice as compared with normal mice (as shown in Figure 3c and d for wounds 4 days after the injury). Even at 12 days after wounding, proliferation was still observed in the basal layer of the regenerated epidermis of Eng þ / þ mice (Figure 3e), whereas only faint Ki67 staining was seen in the new epithelium of Eng þ /  mice (Figure 3f). These results, together with the histological analysis presented in Figure 2, suggest that the activation of keratinocyte proliferation after cutaneous injury is defective in Eng-deficient animals. Keratin 17 (K17) is expressed by activated suprabasal keratinocytes migrating onto the wound bed (Patel et al., 2006). Immunofluorescence staining of K17 in wounds at 6 days after the injury revealed that both the number of

K17-positive cells and the intensity of staining were increased in Eng þ /  mice (Supplementary Figure S2c and d online) as compared with Eng þ / þ mice (Supplementary Figure 2a and b online), suggesting that the epidermis of Eng þ /  mice adjacent to the wound, although unable to accumulate proliferating cells, contained a higher number of migratory keratinocytes compared with the epidermis of normal mice. To further analyze the relationship between epidermal proliferation and Eng, the skin of normal and mutant mice was subjected to an acute proliferative stimulus with the tumor promoter 12-O-tetradecanoylphorbol-13-acetate (TPA). Skin sections were taken 24 and 48 hours after the application of either TPA or vehicle (acetone), and Eng expression was analyzed by real-time quantitative reverse transcription–PCR and western blotting. TPA significantly increased Eng levels in both Eng þ / þ and Eng þ /  mice, although the increase was lower in Eng-deficient animals (Figure 4a and b). Eng expression in the epidermis of mice treated with vehicle alone was barely detected. However, in TPA-stimulated mice, the expression of Eng in the interfollicular epidermis was highly increased, was restricted to the basal-like layer, and was www.jidonline.org

249

E Pe´rez-Go´mez et al. Wound Repair and Endoglin

a Eng relative transcript levels

2

–TPA +TPA

1.5 1

**

0.5

**

0 Eng+/+

b

Eng+/–

Eng+/+ –TPA

+TPA

Eng+/– –TPA

+TPA

Endoglin

β-Actin

c TPA

Endoglin

Ki67

Merge

DAPI

–

+

Figure 4. TPA stimulates endoglin (Eng) expression in the epidermis associated with hyperproliferation. Normal and mutant mice were subjected to a single topical application of 12-O-tetradecanoylphorbol-13-acetate (TPA) or vehicle (acetone), and the expression of Eng was determined 48 hours after the treatment. (a) Transcriptional expression of Eng determined by real-time quantitative PCR. Normalized Eng transcripts were measured relative to glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Data are expressed as mean±SEM from triplicate determinations.**Po0.01 versus Eng þ / þ . (b) Western blot analysis of Eng protein expression. b-Actin was used as a control for protein loading. (c) Representative immunofluorescence analysis of Eng and Ki67 expression in the interfollicular epidermis of wild-type mice treated or not treated with TPA. Nuclei were stained with 4’,6-diamidino-2-phenylindole (DAPI). Note the strong colocalization of Eng and Ki67 in basal-like keratinocytes. Scale bar ¼ 50 mm.

associated with hyperproliferation as found by colocalization of Eng and Ki67 immunofluorescence staining (Figure 4c). TPA was also able to stimulate Eng expression in cultured premalignant keratinocytes (Supplementary Figure S3 online). It is well known that marked hyperplasia is an early event seen after topical application of TPA (Argyris, 1982). 250

Journal of Investigative Dermatology (2014), Volume 134

The comparison of epidermal thickness in vehicle-treated and TPA-treated animals revealed that the hyperplastic response was strongly inhibited in Eng þ /  as compared with Eng þ / þ mice (Figure 5a and b), as expected because of the restrained proliferative capacity of keratinocytes in Eng-deficient animals (Figure 5c).

E Pe´rez-Go´mez et al. Wound Repair and Endoglin

Eng +/+

TPA

Eng +/–

–

35 30 25 20 15 10 5 0

Eng+/+ Eng+/–

**

–

+ TPA

Ki67-positive cells per field

Epidermal thickness (AU)

+

80 70 60 50 40 30 20 10 0

Eng+/+ Eng+/–

**

–

+ TPA

Figure 5. Mutant mice exhibit reduced hyperplasia in response to 12-O-tetradecanoylphorbol-13-acetate (TPA). (a) Representative histological appearances of the skin of a normal and mutant mouse 24 hours after a single topical application of TPA (or vehicle). Sections were counterstained with hematoxylin and eosin (H&E). Note the clear hyperplasia induced in normal skin. Scale bar ¼ 50 mm. (b, c) Quantification of epidermal thickness and the proportion of Ki67-positive keratinocytes 24 hours after TPA (or vehicle) stimulation. Data are shown as mean±SEM (n ¼ 6 animals per group; 5 different fields in each animal). AU, arbitrary units; Eng, endoglin. **Po0.01 versus TPA-treated Eng þ / þ mice.

DISCUSSION Eng has crucial functions in angiogenesis and inflammation (Bernabeu et al., 2007; ten Dijke et al., 2008; Lo´pez-Novoa and Bernabe´u, 2010), two processes intimately linked to wound healing (Valluru et al., 2011a). In this work, we show that skin wound healing is delayed in Eng-deficient mice, in particular at 2–8 days after the injury. In normal mice, Eng expression is enhanced promptly in keratinocytes and blood vessels of wounded skin (Supplementary Figure S1 online). A rapid increase in Eng expression was previously reported in proliferating endothelial cells and infiltrating inflammatory cells in skin punch biopsies from human volunteers (Torsney et al., 2002). Elevated Eng expression was also observed in blood vessels in surgical wounds at 3 days after surgery, suggesting a role for Eng in human cutaneous repair and scar formation (Valluru et al., 2011b). We show in this report that the lower rate of healing observed in Eng heterozygous mice is associated with impaired keratinocyte proliferation at the wound edges, and decreased NO bioavailability. We have previously reported the expression of Eng in epidermal keratinocytes from the proliferative basal layer, hair bulbs, and sebaceous glands (Quintanilla et al., 2003). Carcinogenesis experiments carried out in Eng þ /  mice suggested a crucial function of Eng in epidermal homeostasis by attenuating TGF-b signaling in order to allow keratinocyte proliferation (Quintanilla et al., 2003; Pe´rez-Go´mez et al.,

2005). Therefore, the burst of Eng expression in keratinocytes immediately after wounding appears to be related to the wave of epidermal hyperproliferation occurring in the epidermis adjacent to the wound to replenish the wounded area with new cells. In Eng-deficient mice, the lack of a hyperproliferative epithelium at the wound edges is evident, as shown by histological and Ki67 immunostaining analyses (Figures 2 and 3). A close relationship between Eng and epidermal proliferation was confirmed by analyzing the skin of Eng þ /  and Eng þ / þ mice after a single topical application of TPA. As occurred in wounded skin, TPA stimulated the expression of Eng in the basal layer of normal epidermis, and this enhancement was associated with hyperproliferation, as seen by Ki67 staining (Figure 3). Both Eng expression and the strong epidermal hyperplasia induced by TPA in normal mice were attenuated in Eng heterozygous mice (Figure 3). Interestingly, the epidermis of mutant mice contained a larger subpopulation of keratinocytes migrating into the wounded area than that of wild-type mice, as suggested by its elongated morphology and K17 staining (Figure 3). Although we cannot rule out the possibility of the sectioning angle influencing the shape of the cells, the repeated appearance of more elongated cells in Eng þ /  mice suggests this migratory phenotype. These results are compatible with in vitro studies showing that the knockdown of Eng in transformed keratinocytes constitutively activates TGF-b signaling, simultaneously leading to growth arrest and increased migration (Pe´rez-Go´mez et al., 2007). In fact, TGF-b1, which is released in large amounts from platelets immediately after wounding (Assoian et al., 1983), promotes keratinocyte migration and reepithelialization during wound healing (reviewed in Werner and Grose, 2003). In addition, TGF-b3 has been found to regulate epidermal cell migration (Bandyopadhyay et al., 2006). Several reports point to an important role of NO in wound healing (reviewed in Schwentker et al., 2002; Witte and Barbul, 2002). Thus, clinical studies have revealed that oral administration of the NO donor L-arginine has a beneficial effect on wound healing (Barbul et al., 1990; Kirk et al., 1993). The constitutive, endothelial, neuronal, and inducible NO synthases are expressed during wound healing, and several reports have shown that either reduced activity or expression of NOS leading to decreased NO production has a detrimental effect on wound healing (Schaffer et al., 1996; Yamasaki et al., 1998; Lee et al., 1999). We have reported decreased NO production in Eng þ /  mice caused by reduced endothelial NOS expression (Jerkic et al., 2004) or uncoupling of the enzyme with the chaperone Hsp90, leading to superoxide production rather than NO production (Toporsian et al., 2005; Jerkic et al 2012). In this work, we show that delayed wound closure in Eng-deficient mice is also associated with diminished NO bioavailability, as increasing NO production in Eng þ /  mice by oral administration of a NO donor accelerated the rate of healing, whereas inhibition of NO synthesis in normal mice significantly delayed wound healing (Figure 1). Deficient NO synthesis in the wounds of Eng þ /  mice may result in a marked loss of the keratinocyte proliferative capacity (Frank et al., 1999; Stallmeyer et al., 1999), www.jidonline.org

251

E Pe´rez-Go´mez et al. Wound Repair and Endoglin

as well as impaired keratinocyte production of vascular endothelial growth factor and chemotactic cytokines (Frank et al., 1999, 2000; Wetzler et al., 2000). In fact, Stallmeyer et al. (1999) have described that the amorphous mass of keratinocytes that spread over the wounded tissue present in untreated animals was reduced to a flat cell tongue covering the wound. They also reported in mice treated with L-N6(1-iminoethyl)-lysine (L-NIL), a selective inhibitor of inducible NOS enzymatic activity, a situation similar to that observed by us in Eng-deficient mice. In addition, angiogenesis may also be affected, as endothelial cells isolated from Eng þ /  mice show reduced proliferation, migration, and capillary tube formation associated with decreased endothelial NOS activity and vascular endothelial growth factor secretion (Jerkic et al., 2006a). Eng regulates angiogenesis by modulating TGF-b signaling through the type I activin-like kinase receptors ALK1 and ALK5 (reviewed in Lebrin and Mummery, 2008). Whereas there are conflicting results about the precise roles of these receptors in angiogenesis, it is clear that Eng þ /  mice exhibit a reduced angiogenic response in vivo (Du¨wel et al., 2007; Jerkic et al., 2006b), suggesting that Eng haploinsufficiency during wound healing leads to the inhibition of the angiogenic switch likely by disturbed TGF-b signaling. A schematic model depicting the effects of Eng haploinsufficiency on TGF-b signaling and NO production is presented in Figure 6. A normal level of Eng expression during wound healing favors angiogenesis and keratinocyte proliferation by modulating NO synthesis and TGF-b signaling in the endothelial and epidermal compartments (Figure 6a). Eng haploinsufficiency in mutant mice leads to defective angiogenesis in the endothelial compartment owing to disturbed TGF-b signaling, whereas in the epidermis, reduced Eng expression allows increased TGF-b signaling that, in turn, results in decreased proliferation and increased keratinocyte migration (Figure 6b). Furthermore, alterations in myofibroblast function have recently been reported after wounding in Eng þ /  mice (Pericacho et al., 2013). Thus, Eng seems to control many of the integrated processes involved in proper healing, including adequate function of endothelial cells, myofibroblasts, and keratinocytes. It should be noted that, in addition to being involved in wound healing, Eng is overexpressed in several skin pathologies, including scleroderma (Leask et al., 2002; Maring et al., 2012; Morris et al., 2011), psoriasis (Rulo et al., 1995; van de Kerkhof et al., 1998), and irradiated skin (Wang et al., 1995). All these data support the relevance of Eng in skin function and disease. MATERIALS AND METHODS Animals Generation and genotyping of Eng þ /  mice on a C57Bl/6 background has been previously described (Bourdeau et al., 1999; Rodrı´guez-Pen˜a et al., 2002). Mice were kept in ventilated rooms. Lighting and temperature were controlled by a timer that permitted light between 0800 and 2000 hours and a temperature of 20±1 1C. Mice were fed with standard mouse chow (Panlab, Madrid, Spain) and water ad libitum. All studies were conducted in parallel in Eng þ /  252

Journal of Investigative Dermatology (2014), Volume 134

Endothelial cell Endoglin NO

TGF-β signaling

Keratinocyte Endoglin NO

TGF-β signaling

Angiogenesis

Proliferation

Endothelial cell

Keratinocyte

Endoglin NO

TGF-β signaling

Angiogenesis

Migration

Endoglin NO

TGF-β signaling

Proliferation

Migration

Endoglin high expression Endoglin low expression

Figure 6. Schematic model representing the effects of endoglin (Eng) haploinsufficiency on the endothelial and epidermal cell compartments during wound healing. (a) Normal Eng. (b) Low Eng. In endothelial cells, Eng deficiency results in reduced nitric oxide (NO) synthesis and disturbed transforming growth factor-b (TGF-b) signaling that impairs the angiogenic response. In the epidermal compartment, reduction of NO synthesis and constitutive activation of TGF-b signaling as a consequence of Eng haploinsufficiency result in the inhibition of keratinocyte proliferation and stimulation of migration.

and Eng þ / þ littermate female mice aged 4–6 months (20–25 g). All animal procedures were in accordance with the guidelines for animal use published by Conseil de ´lEurope (No. L358/1–358/6, 1986), Spanish Government (BOE No. 67, pp. 8509–8512, 1988, and BOE No. 256, pp. 31349–31362, 1990), and were approved by the University of Salamanca Animal Care and Use Committee.

Wounding procedure and wound closure analysis Mice were anesthetized with isoflurane (starting with 2–3% and maintained with 1%; Forane, Abbott, Abbott Park, IL). Two full-thickness wounds (5 mm in diameter, 5 mm apart) were made on the back of the mice by excising the skin and the underlying panniculus carnosus. Every 2 days after wounding, mice were anesthetized with isoflurane, and wound borders were drawn on a transparent film. Wound areas (in mm2) were calculated from wound perimeter tracings by photographic analysis using the SCION IMAGE software (Scion, Houston, TX). The original area of each wound (day 0) was given a value of 100, and the wounded area on subsequent days was expressed as a percentage of the original area. To assess the rate of healing, we calculated the time at which wounds were closed by 50% (half closing time).

E Pe´rez-Go´mez et al. Wound Repair and Endoglin

Skin samples were obtained at different days after injury, and the mouse was then killed. At each time point, an area that included the scab and the complete epithelial margin was excised from each individual wound. As a control, a similar area of skin was taken from the back of unwounded mice.

Local NO donor administration and NO synthesis blockade To inhibit NO synthesis in vivo, L-NAME (Sigma, Madrid, Spain) was applied at a dose of 10 mg kg  1 per day in the drinking water of groups of 11 Eng þ /  and Eng þ / þ mice until full wound closure. The same procedure was used to dispense the nitrosothiol NO donor LA-803 (30 mg kg  1 per day), kindly provided by Lacer, Barcelona, Spain (Garcı´a-Criado et al., 2009).

Acute TPA treatment A single dose of TPA (12.5 mg) or the corresponding vehicle (200 ml of acetone) was applied onto the shaved dorsal skin of groups of Eng þ / þ and Eng þ /  mice. Animals were killed at 24 or 48 hours of treatment, and skin was collected. Samples were divided into portions that were either fixed in formaldehyde (for immunofluorescence, immunohistochemistry, and hematoxylin and eosin staining) or snap-frozen (for protein and RNA extraction). Frozen samples were stored at  80 1C until analysis. In addition, MCA3D premalignant keratinocytes (Pe´rez-Go´mez et al., 2007) were treated with TPA (100 ng ml  1 in DMSO) for 24 and 48 hours, or with vehicle alone for 48 hours, and Eng expression was determined by real-time quantitative reverse transcription–PCR and western blotting, as described later.

Western blot analysis Skin samples were homogenized in lysis buffer (1% Triton X-100, 20 mM Tris/HCl, pH 8.0, 137 mM NaCl, 10% glycerol, 5 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 1 mg ml  1 aprotinin, and 1 mg ml  1 leupeptin) and lysates were cleared by centrifugation at 14,000 g. Aliquots of 50 mg of total lysate protein were separated by SDS-gel electrophoresis and electrotransferred to PVDF membranes. For the detection of Eng, a-tubulin, and a-actin, the mAbs MJ7/18 (Ge and Butcher, 1994), sc-8035 (Santa Cruz Biotechnology, Dallas, TX), and AC-15 (Sigma-Aldrich, Madrid, Spain), respectively, were used, followed by appropriate secondary antibodies coupled to horseradish peroxidase. The peroxidase activity was developed using an enhanced chemiluminescence kit (Amersham Biosciences, Barcelona, Spain).

ethanol series, and they were then embedded in paraffin. Sections of 3 mm thickness were cut, mounted on glass slides, and processed. One set of slides were rinsed with water, counterstained with hematoxylin and eosin, washed in water, and dehydrated and mounted with a coverslip using Tissue-Tek (Miles, Fergus Falls, MN). Other sets of slides were subsequently incubated for 1 hour at room temperature using rabbit polyclonal antibodies recognizing mouse Eng (sc20632, dilution 1:10; Santa Cruz Biotechnology) or Ki67 (MAD-020310Q, dilution 1:25; Master Diagno´stico S.L., Granada, Spain). The slides were stained with the avidin–biotin peroxidase complex system from Santa Cruz Biotechnology using 3amino-9-ethylcarbazole as a chromogenic substrate and mounted with a coverslip using Tissue-Tek (Miles). For inmunohistochemical K17 detection, epitope retrieval was performed in a steamer for 3 minutes in 10 mM sodium citrate buffer (pH 6.0), followed by staining with a-K17 Ab-1 (Labvision, Kalamazoo, MI). Immunodetection was performed using the labeled streptavidin-biotin method (DAKO, Barcelona, Spain) with 3,30 -diaminobenzidine as the chromogen. Quantification of epidermal hyperplasia in skin treated with TPA or vehicle was performed in hematoxylin and eosin–stained sections of Eng þ / þ (n ¼ 6) and Eng þ /  (n ¼ 6) littermates in five different fields each. For immunofluorescence analysis in skin treated with TPA, paraffinembedded sections were fixed in 3.7% formaldehyde and subjected to heat-induced antigen retrieval for 9 minutes in Tris-EDTA buffer (10 mM Tris Base, 1 mM EDTA, pH 9.0) before exposure to primary antibodies. The mAb MJ7/18 recognizing mouse Eng (Ge and Butcher, 1994) and anti-Ki67 from Neomarkers/Lab Vision, (Fremont, CA) were used. Secondary anti-rat and anti-rabbit Alexa Fluor 594 and Alexa Fluor 488 antibodies were from Invitrogen (Carlsbad, CA). Cell nuclei were stained with 4’,6-diamidino-2phenylindole from Invitrogen. Fluorescence confocal images were acquired using the Leica TCS-SP2 software (Wetzlar, Germany). Quantification of Ki67-positive nuclei was performed in five different fields, with 3 mice for each condition.

Statistical analysis Between-group comparisons were performed by analysis of variance and corrected for repeated measures when appropriate. If analysis of variance revealed overall significant differences, individual means were evaluated post hoc using the Bonferroni procedure. Results are expressed as the mean±SEM.

Quantitative reverse transcription–PCR Quantitative reverse transcription–PCR analysis was carried out as previously described (Pe´rez-Go´mez et al., 2007). The gene encoding mouse glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal control of the RNA quality and amplification. Primers for mouse Eng were 50 -ACAAGGGGTGAGGTGACGTT-30 and 50 -GAAACGAGGACCAGGAACAC-30 ; and for GAPDH were 50 -TAATGAGCTGGTCATCCGTG-30 and 50 -CAGGCTTCCCTGAGT TCATC-30 . The annealing temperatures were 56 and 55 1C, respectively.

Histology, immunohistochemistry, and immunofluorescence analyses For histological analysis, dissected wounds were fixed in 4% buffered formaldehyde for 24 hours, followed by dehydration through a graded

CONFLICT OF INTEREST The authors state no conflict of interest.

ACKNOWLEDGMENTS We thank Annette Du¨well for the care and genotyping of the laboratory animals. This work has been supported by grants from Ministerio de Ciencia e Innovacion (SAF2010-15881 to JML-N, SAF2010-19222 to CB, and SAF2010-19152 to MQ), Junta de Castilla y Leo´n (Excellence Group GR100 to JML-N), Centro de Investigacio´n Biomedica en Red de Enfermedades Raras (CIBERER to CB), Red de Investigacio´n Cooperativa en Enfermedades Renales (REDinREN to JML-N), and Comunidad Auto´noma de Madrid (S2010/BMD2359, SkinModel, to MQ). REDinREN and CIBERER are initiatives of the Instituto de Salud Carlos III (ISCIII) of Spain supported by European Regional Development Funds (FEDER). EP-G and EM-V were the recipients of a postdoctoral research contract from Fundacio´n Cientı´fica Asociacio´n Espan˜ola Contra el Ca´ncer (AECC).

www.jidonline.org

253

E Pe´rez-Go´mez et al. Wound Repair and Endoglin

SUPPLEMENTARY MATERIAL Supplementary material is linked to the online version of the paper at http:// www.nature.com/jid

REFERENCES Argyris TS (1982) Tumor promotion by regenerative epidermal hyperplasia in mouse skin. J Cutaneous Pathol 9:1–18 Arthur HM, Ure J, Smith AJ et al. (2000) Endoglin, an ancillary TGFbeta receptor, is required for extraembryonic angiogenesis and plays a key role in heart development. Dev Biol 217:42–53 Assoian RK, Komoriya A, Meyers CA et al. (1983) Transforming growth factorbeta in human platelets. Identification of a major storage site, purification, and characterization. J Biol Chem 258:7155–60 Bandyopadhyay B, Fan J, Guan S et al. (2006) A ‘‘traffic control’’ role for TGFbeta3: orchestrating dermal and epidermal cell motility during wound healing. J Cell Biol 172:1093–105 Barbara NP, Wrana JL, Letarte M (1999) Endoglin is an accessory protein that interacts with the signalling receptor complex of multiple members of the transforming growth factor-beta superfamily. J Biol Chem 274:584–94 Barbul AS, Lazarou A, Efron DT et al. (1990) Arginine enhances wound healing and lymphocyte immune responses in humans. Surgery 108:331–6 Bernabeu C, Conley BA, Vary CP (2007) Novel biochemical pathways of endoglin in vascular cell physiology. J Cell Biochem 102:1375–88 Bourdeau A, Dumont DJ, Letarte M (1999) A murine model of hereditary hemorrhagic telangiectasia. J Clin Invest 104:1343–51 Coulombe PA (2003) Wound epithelialization: accelerating the pace of discovery. J Invest Dermatol 121:219–30 Du¨wel A, Eleno N, Jerkic M et al. (2007) Reduced tumor growth and angiogenesis in endoglin-haploinsufficient mice. Tumour Biol 28:1–8 Fonsatti E, Sigalotti L, Arslan P et al. (2003) Emerging role of endoglin (CD105) as a marker of angiogenesis with clinical potential in human malignancies. Curr Cancer Drug Targets 3:427–32 Frank S, Hu¨bner G, Breier G et al. (1995) Regulation of vascular enthotelial growth factor expression in cultured keratinocytes. Implications for normal and impaired wound healing. J Biol Chem 270:12607–13 Frank S, Kampfer H, Wetzler C et al. (2000) Large induction of the chemotactic cytokine RANTES during cutaneous wound repair: a regulatory role for nitric oxide in keratinocyte-derived RANTES expression. Biochem J 347:265–73 Frank S, Stallmeyer B, Kampfer H et al. (1999) Nitric oxide triggers enhanced induction of vascular endothelial growth factor expression in cultured keratinocytes (HaCaT) and during cutaneous wound repair. FASEB J 13:2002–14 Garcı´a-Criado FJ, Rodrı´guez-Barca P, Garcı´a-Cenador MB et al. (2009) Protective effect of new nitrosothiols on the early inflammatory response to kidney ischemia/reperfusion and transplantation in rats. J Interferon Cytokine Res 29:441–9 Ge AZ, Butcher EC (1994) Cloning and expression of a cDNA encoding mouse endoglin, an endothelial cell TGF-beta ligand. Gene 138:201–6

Kirk SJ, Hurson M, Regan MC et al. (1993) Arginine stimulates wound healing and immune function in elderly human beings. Surgery 114:155–9 Leask A, Abraham DJ, Finlay DR et al. (2002) Dysregulation of transforming growth factor beta signaling in scleroderma: overexpression of endoglin in cutaneous scleroderma fibroblasts. Arthritis Rheum 46:1857–65 Lebrin F, Mummery CL (2008) Endoglin-mediated vascular remodeling: mechanisms underlying hereditary hemorrhagic telangiectasia. Trends Cardiovasc Med 18:25–32 Lee PC, Salyapongse N, Bragdon GA et al. (1999) Impaired wound healing and angiogenesis in eNOS-deficient mice. Am J Physiol 277:H1600–8 Li C, Hampson IN, Hampson L et al. (2000) CD105 antagonizes the inhibitory signaling of transforming growth factor b1 on human vascular endothelial cells. FASEB J 14:55–64 Li DY, Sorensen LK, Brooke BS et al. (1999) Defective angiogenesis in mice lacking endoglin. Science 284:1534–7 Lo´pez-Novoa JM, Bernabe´u C. (2010) The physiological role of endoglin in the cardiovascular system. Am J Physiol Heart Circ Physiol 299: H959–74 Maring JA, Trojanowska M, ten Dijke P (2012) Role of endoglin in fibrosis and scleroderma. Int Rev Cell Mol Biol 297:295–308 Martin P (1997) Wound healing. Aiming for perfect skin regeneration. Science 276:75–81 Martin A, Komada MR, Sane DC (2003) Abnormal angiogenesis in diabetes mellitus. Med Res Rev 23:117–45 Morris E, Chrobak I, Bujor A et al. (2011) Endoglin promotes TGF-b/Smad1 signaling in scleroderma fibroblasts. J Cell Physiol 226:3340–8 Patel GK, Wilson CH, Harding KG et al. (2006) Numerous keratinocyte subtypes involved in wound re-epithelialization. J Invest Dermatol 126:497–502 Pe´rez-Go´mez E, Eleno N, Lo´pez-Novoa JM et al. (2005) Characterization of murine S-endoglin isoform and its effects on tumor development. Oncogene 24:4450–61 Pe´rez-Go´mez E, Villa-Morales M, Santos J et al. (2007) A role for endoglin as a suppressor of malignancy during mouse skin carcinogenesis. Cancer Res 67:10268–77 Pericacho M, Velasco S, Prieto M et al. (2013) Endoglin haploinsufficiency promotes fibroblast accumulation during wound healing through Akt activation. PLoS One 8:e54687 Quintanilla M, Ramı´rez JR, Pe´rez-Go´mez E et al. (2003) Expression of the TGF-beta coreceptor endoglin in epidermal keratinocytes and its dual role in multistage mouse skin carcinogenesis. Oncogene 22:5976–85 Reiber GE, Vileikyte L, Boyko EJ et al. (1999) Causal pathways for incident lower-extremity ulcers in patients with diabetes from two settings. Diabetes Care 22:157–62

Gougos A, Letarte M (1990) Primary structure of endoglin, an RGD-containing glycoprotein of human endothelial cells. J Biol Chem 265:8361–4

Rodrı´guez-Pen˜a A, Eleno N, Du¨well A et al. (2002) Endoglin upregulation during experimental renal interstitial fibrosis in mice. Hypertension 40:713–20

Grose R, Werner S (2004) Wound-healing studies in transgenic and knockout mice. Mol Biotechnol 28:147–66

Rulo HF, Westphal JR, van de Kerkhof PC et al. (1995) Expression of endoglin in psoriatic involved and uninvolved skin. J Dermatol Sci 10:103–9

Gurtner GC, Werner S, Barrandon Y et al. (2008) Wound repair and regeneration. Nature 453:314–21

Schaffer MR, Tantry U, Gross SS et al. (1996) Nitric oxide regulates wound healing. J Surg Res 63:237–40

Jerkic M, Rivas-Elena JV, Prieto M et al. (2004) Endoglin regulates nitric oxidedependent vasodilatation. FASEB J 18:609–11

Schwentker A, Vodovotz Y, Weller R et al. (2002) Nitric oxide and wound repair: role of cytokines? Nitric Oxide 7:1–10

Jerkic M, Rodrı´guez-Barbero A, Prieto M et al. (2006a) Reduced angiogenic responses in adult endoglin heterozygous mice. Cardiovasc Res 69:845–54

Singer AJ, Clark RA (1999) Cutaneous wound healing. N Engl J Med 341: 738–46

Jerkic M, Rivas-Elena JV, Santibanez JF et al. (2006b) Endoglin regulates cyclooxygenase-2 expression and activity. Circ Res 99:248–56

Stallmeyer B, Kampfer H, Kol N et al. (1999) The function of nitric oxide in wound repair: inhibition of inducible nitric oxide synthase severely impairs wound reepithelialization. J Invest Dermatol 113:1090–8

Jerkic M, Sotov V, Letarte M (2012) Oxidative stress contributes to endothelial dysfunction in mouse models of hereditary hemorrhagic telangiectasia. Oxid Med Cell Longev 2012:686972

254

Kampfer H, Pfeilschifter J, Frank S (2001) Expressional regulation of angiopoietin-1 and -2 and the tie-1 and -2 receptor tyrosine kinase during cutaneous wound healing: a comparative study of normal and impaired repair. Lab Invest 81:361–73

Journal of Investigative Dermatology (2014), Volume 134

ten Dijke P, Goumans MJ, Pardali E (2008) Endoglin in angiogenesis and vascular diseases. Angiogenesis 11:79–89

E Pe´rez-Go´mez et al. Wound Repair and Endoglin

Toporsian M, Gros R, Kabir MG et al. (2005) Role for endoglin in coupling eNOS activity and regulating vascular tone revealed in hereditary hemorrhagic telangiectasia. Circ Res 96:684–92

Wang JM, Kumar S, van Agthoven A et al. (1995) Irradiation induces upregulation of E9 protein (CD105) in human vascular endothelial cells. Int J Cancer 62:791–6

Torsney E, Charlton R, Parums D et al. (2002) Inducible expression of human endoglin during inflammation and wound healing in vivo. Inflamm Res 51:464–70

Werner S, Grose R. (2003) Regulation of wound healing by growth factors and cytokines. Physiol Rev 83:835–70

Valluru M, Staton CA, Reed MWR et al. (2011a) Transforming growth factor-b and endoglin signaling orchestrate wound healing. Front Physiol 2:1–12

Wetzler C, Kampfer H, Pfeilschifter J et al. (2000) Keratinocyte-derived chemotactic cytokines: expressional modulation by nitric oxide in vitro and during cutaneous wound repair in vivo. Biochem Biophys Res Commun 274:689–96

Valluru M, Brown NJ, Cross SS et al. (2011b) Blood vessel characterization in human dermal wound repair and scarring. Br J Dermatol 165:221–4

Witte MB, Barbul A. (2002) Role of nitric oxide in wound repair. Am J Surg 183:406–12

van de Kerkhof PC, Rulo HF, van Pelt JP et al. (1998) Expression of endoglin in the transition between psoriatic uninvolved and involved skin. Acta Derm Venereol 78:19–21

Yamasaki K, Edington HDJ, McClosky C et al. (1998) Reversal of impaired wound repair in iNOS-deficient mice by topical adenoviral-mediated iNOS gene transfer. J Clin Invest 101:967–71

www.jidonline.org

255

Copyright of Journal of Investigative Dermatology is the property of Nature Publishing Group and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.