Skin Graft Take Rates, Granulation, and Epithelialization: Dependence on Myeloid Cell Hypoxia-Inducible Factor 1α | Dermatology | JAMA Otolaryngology–Head & Neck Surgery | JAMA Network
[Skip to Navigation]
Figure 1. 

Representative photographs showing skin grafts in situ before harvesting the specimens. Wild-type (A) and knockout (B) mice. Note the difference between healthy areas of the skin grafts with hair follicles (solid lines) and areas healed by secondary healing with eventual scarring and lack of hair follicles (dotted lines).

Representative photographs showing skin grafts in situ before harvesting the specimens. Wild-type (A) and knockout (B) mice. Note the difference between healthy areas of the skin grafts with hair follicles (solid lines) and areas healed by secondary healing with eventual scarring and lack of hair follicles (dotted lines).
Figure 2. 

Comparison of percentage “take” rates of skin grafts 14 days after surgery in wild-type (WT) and knockout (KO) mice.

Comparison of percentage “take” rates of skin grafts 14 days after surgery in wild-type (WT) and knockout (KO) mice.
Comparison of Histopathologic Parameters Between Wild-type (WT) and Knockout (KO) Mice
Comparison of Histopathologic Parameters Between Wild-type (WT) and Knockout (KO) Mice
Original Article
July 19, 2010

Skin Graft Take Rates, Granulation, and Epithelialization: Dependence on Myeloid Cell Hypoxia-Inducible Factor 1α

Arch Otolaryngol Head Neck Surg. 2010;136(7):720-723. doi:10.1001/archoto.2010.103

Objective  To compare the “take” rates of skin grafts between myeloid-selective hypoxia-inducible factor (HIF) 1α knockout (KO) and wild-type (WT) mice. Production of the α subunit of HIF-1α is increased in healing wounds, which stimulates expression of vascular endothelial growth factor (VEGF) to promote angiogenesis. Therefore, the take rate of skin grafts may be closely associated with the presence or absence of HIF-1α production in the recipient bed.

Design  The percentage of healthy graft areas obtained by planimetry and scores for epithelialization and granulation tissue formation obtained by histopathologic analysis were compared in 12 KO and 12 WT mice following skin grafting.

Results  The graft take rate was significantly impaired in the KO group ( = .009), whereas epithelialization ( = .46) or granulation ( = .41) tissue formation scores did not reveal any significant differences.

Conclusion  Hypoxia-inducible factor 1α in myeloid cells may be an important molecule for revascularization of avascular tissues such as skin grafts, probably owing to its stimulating effect on angiogenesis.

One of the most important factors involved in wound healing is granulation tissue formation and angiogenesis, which provide an infrastructure for epithelialization. Angiogenesis in healing wounds is closely associated with the production of vascular endothelial growth factor (VEGF), which promotes neovascularization by selectively targeting endothelial cells for proliferation.1 There is strong evidence in the literature that the angiogenic effect of VEGF in wound healing is primarily regulated by the α subunit of hypoxia-inducible factor 1 (HIF-1).2 Hypoxia-inducible factor 1 is a heterodimeric protein composed of α and β subunits, and the HIF-1α subunit is primarily regulated by tissue oxygen.3 In the presence of hypoxia, HIF-1α accumulates within the tissue instead of being degraded, which occurs constantly in normoxic state, and activates the expression of its target genes, including VEGF.4,5 It is well known that HIF-1α production is increased in healing wounds, which are hypoxic, at least at the initial stages of healing.4,6 This increased production of HIF-1α stimulates expression of VEGF and its receptors in healing wounds to promote angiogenesis.2

We have previously shown that exogenously administered VEGF can significantly increase “take” rates of transplanted avascular skin grafts either in healthy conditions or in irradiated recipient beds in a rat model.7,8 However, the role of endogenous VEGF production without external administration of VEGF in skin graft survival is unknown. We speculated that take rates of skin grafts would be closely associated with the presence or absence of HIF-1α production in the recipient bed as an initial step of angiogenesis. We performed this study to find out if the take rate of full-thickness skin grafts would be different between wild-type (WT) and myeloid-selective HIF-1α knockout (KO) mice.


This study was conducted in myeloid-selective HIF-1α KO mice and litter-mate WT mice after obtaining approval from the institutional animal care and use committee at the University of Arkansas for Medical Sciences. Animals were individually housed under standardized conditions with controlled temperature, humidity, and a 12-hour-day/12-hour-night light cycle. Animals had free access to water and standard mouse chow. Twelve adult male KO and 12 WT mice, weighing 25 to 30 g, were used. Myeloid-selective HIF-1α KO mice (HIF-1α-lysMcre) are double-mutant mice on a mixed Sv129/C57Bl/Cb.20 background. They carry as well a double-floxed HIF-1α gene as a Cre recombinase gene under control of the lysozyme M promoter (lysMcre).9 Wild-type mice lack the Cre recombinase gene. Breeders used to generate the mutant mice used in this study were on a mixed Sv129/C57Bl/6/CB.20 background and were kindly provided by R. S. Johnson, PhD, of the University of California at San Diego.9


On day 0, 12 male HIF-1α KO mice and 12 male WT mice were anesthetized using isoflurane inhalation anesthesia. The dorsum of the mice was shaved and prepared with alcohol. Subsequently, a circular area with a diameter of 20 mm was outlined on the dorsum at the midline using a surgical marking pen. The reason for choosing the dorsum at the midline for grafting was to minimize the mice's access to the surgical site with their mouths as well as taking the advantage of natural convexity of this area to minimize the dead space under the graft. Under aseptic conditions, an incision was made along the marking using a No. 15 scalpel blade. After making this incision just below the panniculus carnosus muscle, the outlined skin was harvested as a full-thickness graft by separating it from the deep dorsal muscular fascia layer. To simulate the removal of excess fat from undersurface of the harvested full-thickness skin grafts in clinical conditions, the panniculus carnosus layer was removed from the undersurface of the skin graft. Following this maneuver, the graft was placed back into its donor site by securing the edges with interrupted nonabsorbable sutures. To eliminate potential dead space underneath the graft, 1 tacking suture was placed through the center of the graft into its recipient bed. The mice were then caged individually as an additional measure to minimize the trauma to the surgical site.

Macroscopic assessment of the skin grafts

On day 14 following surgery, the skin grafted areas were macroscopically assessed by using planimetry, because this time frame has been shown to yield more objective assessment of the viability of the grafts, owing to epidermolysis that is almost invariably seen at earlier postoperative stages, which can be deceiving in the decision of graft viability.7,8 Areas with healthy graft tissue and areas that had healed by secondary intention after graft failure were identified. Regions with hair and/or follicles were considered to be healthy graft tissue, and areas with a smooth, whitish appearance without hair or follicles were considered to be areas that had healed by secondary intention owing to full-thickness loss (Figure 1). To calculate the size of the healthy regions and the regions healed by secondary intention, these areas were outlined on transparent paper that was placed on the skin-grafted dorsum. The transparency paper was digitally scanned, and the ratio of healthy area to the entire skin graft area was calculated by using computer software (Image Pro Plus, Silver Spring, Maryland) for each graft.

Tissue preparation

On day 14, after the macroscopic assessment of the skin grafts, the mice were humanely killed. The skin-grafted area was removed en bloc, including the recipient bed, and fixed in methanol–Carnoy solution (methanol, chloroform, and glacial acetic acid in a 6:3:1 ratio). Following this, representative parts composed of healthy graft areas and secondary intention healing were cut out of the main specimens, and 4-μm sections were obtained from each specimen for histopathologic evaluation.

Histopathologic evaluation

Standard hematoxylin-eosin staining was performed on representative sections for histopathologic evaluation of epithelialization and granulation tissue formation. Each of these parameters was semiquantitatively evaluated for each representative slide under low-power (original magnification ×100) light microscopy by the pathologist (C.-Y.F.), who was blinded to the source of specimens, in a 4-point scale scoring system (0, absent; 1, mild; 2, moderate; 3, abundant).

Statistical analysis

Comparison of the data obtained by planimetry for the percentage of healthy graft areas, as well as the data obtained by semiquantitative scoring for epithelialization and granulation tissue among WT and KO mice, was performed by using 2-tailed t test. The statistical significance was determined at P ≤ .05. All statistics were conducted using Microsoft Excel for Mac 2008 (Microsoft Inc, Seattle, Washington).


The mean (SEM) percentage take rates of skin grafts in WT and KO mice were 82.9 (6.5) and 56.3 (6.6), respectively (P = .009) (Figure 2). Epithelialization and granulation tissue formation scores did not show any statistically significant difference between WT and KO mice (see the Table for P values).


Although grafts in the form of skin, dermis, cartilage, bone, fascia, and fat are indispensable tools of reconstructive surgery, their reliability may be questionable owing to their avascular nature, especially in the presence of less-than-optimal recipient bed conditions, such as diabetes mellitus, burns, or prior irradiation.10 Increasing the reliability of grafts, regardless of their nature, would provide an invaluable benefit to patients undergoing reconstructive surgery involving the use of graft materials. Although extensive research is under way to improve wound healing by revealing the relationship between molecules such as HIF-1α and VEGF, most of these studies are being performed in models of excisional wounds with secondary intention healing.

It has been shown that HIF-1α production is increased in wounds as a response to ischemia, which promotes expression of VEGF and VEGF receptors to increase angiogenesis.2,11 However, elimination of VEGF or its receptors from the wound does not cause delay in wound closure, despite impaired angiogenesis in such conditions.12,13 Therefore, impaired angiogenesis may not mean impaired wound healing, at least in secondary intention healing.13 Our study is somewhat in agreement with this because we were not able to demonstrate any difference in granulation tissue formation or epithelialization scores between KO and WT mice. This suggested that skin graft take rates were independent of granulation tissue formation and/or epithelialization potential of the wound. However, unlike secondary intention healing models, take rates of skin grafts showed noticeable impairment in HIF-1α KO mice, and this is most probably due to diminished angiogenesis causing poor revascularization of the grafts. Because HIF-1α deficiency in KO mice involved only myeloid cells and not all HIF-1α–producing cells, a decreased take rate of skin grafts in this group of animals should not be solely attributed to diminished VEGF production. It is known that myeloid-selective HIF-1α KO mice revealed limited attraction and differentiation of circulating angiogenic cells in the wound, which may be another reason for poor graft take rates in this group of animals.14 Furthermore, inhibition of proangiogenic factor HIF-2 in addition to HIF-1 has been shown to limit attraction and differentiation of circulating angiogenic stem cells even more, as a proof of involvement of more than 1 mechanism in wound angiogenesis.15 Regardless of the mechanism, it seems that relative HIF-1α deficiency in wounds repaired with skin grafts revealed poor take rates, probably owing to diminished angiogenesis within the grafted skin.

Despite the poor graft take rates in the HIF-1α–deficient mice in our study; granulation tissue formation and epithelialization scores were not significantly different in this group of animals when compared with WT animals. This may seem contradictory at first sight because one may expect poorer scores in these 2 major indicators of wound healing based on the finding of diminished graft take rates in HIF-1α deficiency. Although the KO mice used in this study were HIF-1α deficient selectively in myeloid cells, it is known that there are other sources of HIF-1α in wound healing in addition to myeloid inflammatory cells, such as basal keratinocytes at the wound edges.16 Therefore, HIF-1α deficiency in myeloid cell line alone may not cause a dramatic decrease in epithelialization because this process most probably continues from wound edges in both secondary intention healing and in skin grafting. Another explanation for the lack of difference in epithelialization scores between KO and WT mice would be the possibility of the absence of a direct relationship between HIF-1α and the epithelialization process. The lack of difference in granulation tissue formation scores between KO and WT mice in our study is in agreement with previous observations.17 It has been previously shown that the wounds left for secondary intention healing did not show any difference between myeloid-selective HIF-1α KO and WT mice after 11 days of wounding, although the wounds filled out with granulation tissue faster in KO mice at the earlier phase, most probably owing to diminished nitric oxide production and eventual improvement in wound inflammation.17 Because our evaluation of grafted recipient beds took place 2 weeks after wounding, we may not have observed the earlier accelerating effect of HIF-1α deficiency on granulation tissue formation in this study. Obviously, HIF-1α seems to be a molecule that can affect wound healing through more than 1 mechanism, which may cause different outcomes in secondary intention healing (granulation tissue formation followed by epithelialization) and skin grafting (neovascularization of avascular skin).

Although we have not specifically looked at microvascular density in our study as a measure of angiogenesis in the grafted areas, statistically significant impairment in graft take rates in HIF-1α KO mice may suggest diminished angiogenesis, probably owing to diminished HIF-1α production in this group of animals. In addition to our previously shown improvement of skin graft take rates with exogenously administered VEGF in healthy and irradiated recipient beds,7,8 our current study provides evidence that presumed impairment of endogenous VEGF production owing to impaired HIF-1α expression in the environment may also interfere with skin graft take rates.

It seems that HIF-1α may serve as another molecule to be investigated as VEGF, in revascularization of avascular tissues such as grafts, or in increasing the survival of tissue flaps with diminished blood supply.18,19 These molecules may have potentials for therapeutic manipulation in the future, especially in conditions with less-than-optimal neovascularization potential of tissue grafts or in ischemic conventional or microvascular flaps.

In conclusion, HIF-1α deficiency is related to poor skin graft takes, most probably owing to diminished angiogenesis.

Correspondence: Emre Vural, MD, Department of Otolaryngology–Head and Neck Surgery, University of Arkansas for Medical Sciences, 4301 W Markham, Slot 543, Little Rock, AR 72205 (

Submitted for Publication: October 14, 2009; final revision received January 2, 2010; accepted February 26, 2010.

Author Contributions: All authors had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Vural, Berbée, and Hauer-Jensen. Acquisition of data: Vural, Berbée, Acott, and Blagg. Analysis and interpretation of data: Vural, Berbée, Fan, and Hauer-Jensen. Drafting of the manuscript: Vural, Berbée, Acott, Blagg, and Hauer-Jensen. Critical revision of the manuscript for important intellectual content: Vural, Fan, and Hauer-Jensen. Obtained funding: Hauer-Jensen. Administrative, technical, and material support: Vural, Berbée, Blagg, Fan, and Hauer-Jensen. Study supervision: Vural and Hauer-Jensen.

Financial Disclosure: None reported.

Funding/Support: This study was supported by grants CA71382 and CA83719 from the National Institutes of Health.

Romano Di Peppe  SMangoni  AZambruno  G  et al.  Adenovirus-mediated VEGF165 gene transfer enhances wound healing by promoting angiogenesis in CD1 diabetic mice.  Gene Ther 2002;9 (19) 1271- 1277PubMedGoogle ScholarCrossref
Tang  NWang  LEsko  J  et al.  Loss of HIF-1α in endothelial cells disrupts a hypoxia-driven VEGF autocrin loop necessary for tumorigenesis.  Cancer Cell 2004;6 (5) 485- 495PubMedGoogle ScholarCrossref
Wang  GLJiang  BHRue  EASemenza  GL Hypoxia-inducible factor I is a basic-helix-loop-helix-PAS heterodimer regulated by cellular O2 tension.  Proc Natl Acad Sci U S A 1995;92 (12) 5510- 5514PubMedGoogle ScholarCrossref
Liu  LMarti  GPWei  X  et al.  Age-dependent impairment of HIF-1α expression in diabetic mice: correction with electroporation-facilitated gene therapy increases wound healing, angiogenesis, and circulating angiogenic cells.  J Cell Physiol 2008;217 (2) 319- 327PubMedGoogle ScholarCrossref
Sen  CK Wound healing essentials: let there be oxygen.  Wound Repair Regen 2009;17 (1) 1- 18PubMedGoogle ScholarCrossref
Hopf  HWGibson  JJAngeles  AP  et al.  Hyperoxia and angiogenesis.  Wound Repair Regen 2005;13 (6) 558- 564PubMedGoogle ScholarCrossref
Richter  GTFan  CYOzgursoy  O McCoy  JVural  E Effect of vascular endothelial growth factor on skin graft survival in Sprague Dawley rats.  Arch Otolaryngol Head Neck Surg 2006;132 (6) 637- 641PubMedGoogle ScholarCrossref
Richter  GTBowen  T  IIIBoerma  MFan  CYHauer-Jensen  MVural  E Impact of vascular endothelial growth factor on skin graft survival in irradiated rats.  Arch Facial Plast Surg 2009;11 (2) 110- 113PubMedGoogle ScholarCrossref
Cramer  TYamanishi  YClausen  BE  et al.  HIF-1alpha is essential for myeloid cell mediated inflammation.  Cell 2003;112 (5) 645- 657PubMedGoogle ScholarCrossref
Vural  E Surgical reconstruction in patients with cancer of the head and neck.  Curr Oncol Rep 2004;6 (2) 133- 140PubMedGoogle ScholarCrossref
Zhang  QChang  QCox  RAGong  XGould  LJ Hyperbaric oxygen attenuates apoptosis and decreases inflammation in an ischemic wound model.  J Invest Dermatol 2008;128 (8) 2102- 2112PubMedGoogle ScholarCrossref
Frantz  SVincent  KAFeron  OKelly  RA Innate immunity and angiogenesis.  Circ Res 2005;96 (1) 15- 26PubMedGoogle ScholarCrossref
Tsou  RFathke  CWilson  LWallace  KGibran  NIsik  F Retroviral delivery of dominant-negative vascular endothelial growth factor receptor type 2 to murine wounds inhibits wound angiogenesis.  Wound Repair Regen 2002;10 (4) 222- 229PubMedGoogle ScholarCrossref
Milovanova  TNBhopale  VMSorokina  EM  et al.  Lactate stimulates vasculogenic stem cells via the thioredoxin system and engages an autocrin activation loop involving hypoxia-inducible factor 1.  Mol Cell Biol 2008;28 (20) 6248- 6261PubMedGoogle ScholarCrossref
Milovanova  TNBhopale  VMSorokina  EM  et al.  Hyperbaric oxygen stimulates vasculogenic stem cell growth and differentiation in vivo.  J Appl Physiol 2009;106 (2) 711- 728PubMedGoogle ScholarCrossref
Elson  DARyan  HESnow  JWJohnson  RArbeit  JM Coordinate up-regulation of hypoxia inducible factor (HIF)-1α and HIF-1 target genes during multi-stage epidermal carcinogenesis and wound healing.  Cancer Res 2000;60 (21) 6189- 6195PubMedGoogle Scholar
Owings  RABoerma  MWang  J  et al.  Selective deficiency of HIF-1α in myeloid cells influences secondary intention wound healing in mouse skin.  In Vivo 2009;23 (6) 879- 884PubMedGoogle Scholar
Taub  PJMarmur  JDZhang  WX  et al.  Locally administered vascular endothelial growth factor cDNA increases survival of ischemic experimental skin flaps.  Plast Reconstr Surg 1998;102 (6) 2033- 2039PubMedGoogle ScholarCrossref
Taub  PJMarmur  JDZhang  WX  et al.  Effect of time on the viability of ischemic skin flaps treated with vascular endothelial growth factor (VEGF) cDNA.  J Reconstr Microsurg 1998;14 (6) 387- 390PubMedGoogle ScholarCrossref