DOI: 10.1111/exd.12782
Commentary
www.wileyonlinelibrary.com/journal/EXD
Pericytes in wound healing: friend or foe? David M Ansell1 and Ander Izeta2 1
The Centre for Dermatology Research, Institute of Inflammation and Repair, The University of Manchester, Manchester, UK; 2Instituto Biodonostia, Hospital Universitario Donostia, San Sebastian, Spain Correspondence: David M Ansell, PhD The Centre for Dermatology Research, Institute of Inflammation and Repair, The University of Manchester, UK, Tel.: +44-161-306-0515, Fax: +44 (0)161 306 0693, e-mail: [email protected]
Key words: mesenchymal stromal cell – pericyte – repair – skin – stem cell
Accepted for publication 2 June 2015
Pericytes are cells that reside on the outer surface below the basal lamina of smaller blood vessels, where they provide structural integrity, and regulate blood flow (s1). These cells which are generally [although not always (s2)] of mesodermal origin and display multipotency have been differentiated into chondrocytes, adipocytes, phagocytes, osteoblasts and granulocytes (s3). Pericytes have sparked a great deal of interest in the field of regenerative medicine due to their ability to promote repair of various organs, which is primarily thought to be through enhancing angiogenesis. In addition, there is evidence for pericytes playing a role in coagulation, lymphocyte activation and even phagocytosis [see (s4) for a recent review]. However, some of their purported functions might be wrongly attributed due to confusion with surrounding cells (e.g. perivascular macrophages) (s5). On the other hand, pericytes (or subsets thereof) may also have detrimental effects for tissue regeneration, as a fibrosis-promoting role is now evident in diverse tissues [reviewed by (s6, s7, s8)]. Several cell surface receptors are used to detect pericytes, including PDGFRb, EGFR, aSMA and adenosine A2 receptors. However, none of these are entirely specific for pericytes (s9), while additional markers might be found on pericytes depending on the body site, developmental stage and activation in response to injury (s3). To further complicate matters, pericytes are classified into distinct subtypes based on morphology, cell markers and functional assessment (s4). Perivascular-localised cells expressing PDGFRb/ADAM12, PDGFRb/Glast or PDGFRb/Gli1 seem responsible for injury-induced fibrosis (s10, s11, s12), although the relationship among these populations has not been properly addressed. It is therefore clear that pericytes are a highly heterogeneous population and a better definition of functional subclasses is urgently needed. Although surprisingly little data exist on the regenerative effect of pericytes within the skin, it has previously been shown using an in vitro skin equivalent model that pericytes signal to keratinocytes to promote re-epithelialisation (1), raising the possibility that pericytes could be used therapeutically to stimulate wound repair. In fact, mesenchymal stromal cell (MSC) cultures are known to include (and possibly arise from) perivascular-localised cells (2); their wound-healing-promoting effect now being well established (recently reviewed by (3, s13, s14). For clarity, we chose to include here only studies using cells defined by the expression of perivascular cell markers. In their study presented in this issue of Experimental Dermatology, Mills et al. have inoculated human pericytes into murine
wounds in vivo and report a delay to dermal repair (4). This intriguing finding contradicts much of the in vivo skin pericyte literature, which indicates a strong role for the promotion of wound angiogenesis, at least by some of the pericyte subtypes (Table 1). In this study unlike in the authors’ previous in vitro work (1), Mills et al. did not find any effect on wound re-epithelialisation (4). One plausible explanation given for this discrepancy would be that mouse keratinocytes are not able to respond efficiently to paracrine signalling molecules released from the human pericytes in this model. Recently developed humanised mouse models of wound healing may provide a tool to address this specific question (s15). Another limitation is the need to use immune-compromised animals, which entails defective signalling mediated by immune cell infiltration and inflammation in the initial phases of healing (s16, s17, s18, s19). The delay to inflammatory resolution that Mills et al. observed was coupled with increased TGFb1 and Smad3 signalling that may be of importance with respect to later wound remodelling. It has been shown that overexpression of TGFb1 in normal mouse skin activates pericytes, leading to fibrosis (s20). However, the authors did not find significantly altered collagen deposition in the transplanted mice. TGFb - mediated - signalling leading to fibrosis is common to several other tissues (s21). The complex nature of wound repair makes it excruciatingly difficult to tease apart the specific effect that an intervention has on individual healing processes, which occur concurrently during wound repair in vivo. It might be that pericytes act to stimulate some aspects of healing (e.g. angiogenesis and re-epithelialisation), while simultaneously repressing others (e.g. inflammatory resolution), with the overall healing outcome observed being the cumulative effect. By employing a panel of in vitro assays, one could examine several wound-healing processes in isolation, in an attempt to unpick this conundrum (5). The apparent plasticity of pericytes, coupled with the lack of an agreed list of cell surface markers, makes it virtually impossible to assess the degree of similarity of published studies across different research groups. In addition, some studies utilise freshly isolated pericytes, while others culture cells prior to experimentation, which may alter their characteristics during subsequent analysis. Thus, much of the conflicting data within the field likely arises from selecting fundamentally different cell populations for study. A tantalising possibility is that pericyte subtypes, rather than being constitutively distinct cells, may in turn reflect an activated/dedifferentiated phenotype upon injury (6,7). If correct, this would mean that the local
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 833–834
833
Commentary
Table 1. Reported effect of pericytes in cutaneous wounds (studies shown in reverse chronological order) Animal
Wound model
Pericyte phenotype
Brief description of method
Reported activity of pericytes
Ref
Human foreskin-derived CD45-VLA1bri cells
Topical application of pericytes in collagen type I gels
(4)
Human ASCs (aSMA+, CD140b+, NG2+, PDGFRb+, ANGPT+) from discarded burn skin samples PKH67-labeled human umbilical cord perivascular cells (HUCPVCs)
Topical application of pericytes in PEG-fibrin gels
Delay in dermal wound resolution Prolonged inflammatory response (↑ Neutrophils and macrophages) No effect in re-epithelialisation and wound closure Earlier collagen deposition and wound remodelling ↑ Vascularisation
Topical application of pericytes in fibrin gels
Accelerated wound closure ↑ Vascularisation and granulation tissue Improved skin architecture and tensile strength
(9)
5-mm excisional fullthickness wounds
PDGFRb+, aSMA+, NG2+
4-mm excisional fullthickness wounds
aSMA+, NG2+
F344/DuCrj male rats
5-mm excisional fullthickness wounds
PDGFRb+, aSMA+, Desmin+, Thy-1+, A3+ cells surrounding vWF+ vessels
Immunohistochemical evaluation up to 26 days postwounding
Male Balb/c mice
2-mm excisional fullthickness wounds
PDGFRb+, Desmin+, NG2+, aSMA+, VEGFA+ surrounding Cd31+ vessels
Intravital blood vessel labelling with lectin, immunohistochemistry and TEM
Col1a2 promoter-driven LacZ mice.
4-mm excisional fullthickness wounds
NG2+, surrounding Cd31+ vessels
Imatinib used to block PDGFRb signalling.
MFG-E8 expressed by PDGFRb+, aSMA+, NG2+ pericytes stimulates angiogenesis CCN2 is required to recruit pericytes to wounds. No effect on wound healing Promote angiogenesis Related to early (less differentiated) myofibroblast formation Inhibition (vessel stabilization) and promotion (in the tip) of endothelial cell proliferation during angiogenesis, depending on the pericyte subtype ↑ Vascularisation Stimulate myofibroblast differentiation
(S26)
Col1Cre; CCN2-/- mice
Wounding of MFG-E8-/- mice. MFG-E8 is a glycoprotein expressed by PDGFRb+, aSMA+, NG2+ pericytes. Wounding of Col1 Cre CCN2-/- mice.
Epidermal regeneration of bilayered skin constructs
mAb HD-1bri cells (↑ for aSMA, MYH9, TPM1, TPM2, RGS5, NG2, PDGFRb, ENEP, CNN1, CALD1) Human foreskin-derived HMW-MAA+ vascular cells
Incorporated pericytes into dermal gels, then overlayed with keratinocytes to form skin constructs. Pericytes and fibroblasts loaded into collagen gels
In vivo pericyte transplantation studies Balb/c SCID male mice 4-mm excisional fullthickness wounds
Male Rowett nude rats (athymic)
1.5-cm excisional full-thickness wounds
Female Balb/c nude mice
4-mm and 8-mm excisional full-thickness wounds
In vivo pericyte manipulation studies MFG-E8-/-, C57Bl6 background
In vitro studies N/A
N/A
Gel contraction assay.
wound environment would direct activated pericytes into regeneration- or fibrosis-promoting roles, depending on the circumstances. For transplantation studies, this would mean that the different biomaterials and culture media used for cell delivery (Table 1) might have a key role to obtain the desired clinical outcome. In studies where the wound-healing-promoting role of endogenous pericytes is to be enhanced, the role of PDGF (s20, s22, s23, s24, s25, s26), CXCL12/CXCR4 (s27), TGFb (s20, s21) and WNT/b-catenin (s28, s29, s30) signalling pathways should be explored. In summary, pericytes display a number of mesenchymal stem cell properties, making them a potentially exciting area for developing cell-based therapies to tackle wound pathology, yet little work has been conducted in this area when compared to other skin resident stem cell populations, such as those found within
(8)
(S36) (S37)
(S38)
(S24)
Enhance epidermal regeneration
(1)
Stimulate myofibroblast differentiation
(S24)
the dermal adipose tissue, or the hair follicle (s31, s32, s33, s34, s35). The work of Mills et al. highlights that a detailed understanding of pericyte activity during skin injury repair remains critically lacking, which needs to be urgently addressed if we ever hope to exploit this intriguing cell population to tackle the evergrowing burden of chronic wounds.
Author contribution DA and AI wrote the manuscript.
Conflict of interest The authors declare no conflict of interest
Supporting Information Additional supporting data may be found in the supplementary information of this article. Appendix S1. References.
References 1 Paquet-Fifield S, Schluter H, Li A et al. J Clin Invest 2009: 119: 2795–2806. 2 Murray I R, West C C, Hardy W R et al. Cell Mol Life Sci 2014: 71: 1353–1374. 3 Khosrotehrani K. Exp Dermatol 2013: 22: 307– 310.
834
4 Mills S J, Zhuang L, Arandjelovic P et al. Exp Dermatol 2015: 24: 881–883. 5 Ansell D M, Holden K A, Hardman M J. Exp Dermatol 2012: 21: 581–585. 6 Etxaniz U, Perez-San Vicente A, Gago-Lopez N, et al. Stem Cell Reports 2014: 3: 774–788.
7 Johnston A P, Naska S, Jones K et al. Stem Cell Reports 2013: 2013: 38–45. 8 Zamora D O, Natesan S, Becerra S et al. Angiogenesis 2013: 16: 745–757. 9 Zebardast N, Lickorish D, Davies J E. Organogenesis 2010: 6: 197–203.
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 833–834
Commentary
www.wileyonlinelibrary.com/journal/EXD
Pericytes in wound healing: friend or foe? David M Ansell1 and Ander Izeta2 1
The Centre for Dermatology Research, Institute of Inflammation and Repair, The University of Manchester, Manchester, UK; 2Instituto Biodonostia, Hospital Universitario Donostia, San Sebastian, Spain Correspondence: David M Ansell, PhD The Centre for Dermatology Research, Institute of Inflammation and Repair, The University of Manchester, UK, Tel.: +44-161-306-0515, Fax: +44 (0)161 306 0693, e-mail: [email protected]
Key words: mesenchymal stromal cell – pericyte – repair – skin – stem cell
Accepted for publication 2 June 2015
Pericytes are cells that reside on the outer surface below the basal lamina of smaller blood vessels, where they provide structural integrity, and regulate blood flow (s1). These cells which are generally [although not always (s2)] of mesodermal origin and display multipotency have been differentiated into chondrocytes, adipocytes, phagocytes, osteoblasts and granulocytes (s3). Pericytes have sparked a great deal of interest in the field of regenerative medicine due to their ability to promote repair of various organs, which is primarily thought to be through enhancing angiogenesis. In addition, there is evidence for pericytes playing a role in coagulation, lymphocyte activation and even phagocytosis [see (s4) for a recent review]. However, some of their purported functions might be wrongly attributed due to confusion with surrounding cells (e.g. perivascular macrophages) (s5). On the other hand, pericytes (or subsets thereof) may also have detrimental effects for tissue regeneration, as a fibrosis-promoting role is now evident in diverse tissues [reviewed by (s6, s7, s8)]. Several cell surface receptors are used to detect pericytes, including PDGFRb, EGFR, aSMA and adenosine A2 receptors. However, none of these are entirely specific for pericytes (s9), while additional markers might be found on pericytes depending on the body site, developmental stage and activation in response to injury (s3). To further complicate matters, pericytes are classified into distinct subtypes based on morphology, cell markers and functional assessment (s4). Perivascular-localised cells expressing PDGFRb/ADAM12, PDGFRb/Glast or PDGFRb/Gli1 seem responsible for injury-induced fibrosis (s10, s11, s12), although the relationship among these populations has not been properly addressed. It is therefore clear that pericytes are a highly heterogeneous population and a better definition of functional subclasses is urgently needed. Although surprisingly little data exist on the regenerative effect of pericytes within the skin, it has previously been shown using an in vitro skin equivalent model that pericytes signal to keratinocytes to promote re-epithelialisation (1), raising the possibility that pericytes could be used therapeutically to stimulate wound repair. In fact, mesenchymal stromal cell (MSC) cultures are known to include (and possibly arise from) perivascular-localised cells (2); their wound-healing-promoting effect now being well established (recently reviewed by (3, s13, s14). For clarity, we chose to include here only studies using cells defined by the expression of perivascular cell markers. In their study presented in this issue of Experimental Dermatology, Mills et al. have inoculated human pericytes into murine
wounds in vivo and report a delay to dermal repair (4). This intriguing finding contradicts much of the in vivo skin pericyte literature, which indicates a strong role for the promotion of wound angiogenesis, at least by some of the pericyte subtypes (Table 1). In this study unlike in the authors’ previous in vitro work (1), Mills et al. did not find any effect on wound re-epithelialisation (4). One plausible explanation given for this discrepancy would be that mouse keratinocytes are not able to respond efficiently to paracrine signalling molecules released from the human pericytes in this model. Recently developed humanised mouse models of wound healing may provide a tool to address this specific question (s15). Another limitation is the need to use immune-compromised animals, which entails defective signalling mediated by immune cell infiltration and inflammation in the initial phases of healing (s16, s17, s18, s19). The delay to inflammatory resolution that Mills et al. observed was coupled with increased TGFb1 and Smad3 signalling that may be of importance with respect to later wound remodelling. It has been shown that overexpression of TGFb1 in normal mouse skin activates pericytes, leading to fibrosis (s20). However, the authors did not find significantly altered collagen deposition in the transplanted mice. TGFb - mediated - signalling leading to fibrosis is common to several other tissues (s21). The complex nature of wound repair makes it excruciatingly difficult to tease apart the specific effect that an intervention has on individual healing processes, which occur concurrently during wound repair in vivo. It might be that pericytes act to stimulate some aspects of healing (e.g. angiogenesis and re-epithelialisation), while simultaneously repressing others (e.g. inflammatory resolution), with the overall healing outcome observed being the cumulative effect. By employing a panel of in vitro assays, one could examine several wound-healing processes in isolation, in an attempt to unpick this conundrum (5). The apparent plasticity of pericytes, coupled with the lack of an agreed list of cell surface markers, makes it virtually impossible to assess the degree of similarity of published studies across different research groups. In addition, some studies utilise freshly isolated pericytes, while others culture cells prior to experimentation, which may alter their characteristics during subsequent analysis. Thus, much of the conflicting data within the field likely arises from selecting fundamentally different cell populations for study. A tantalising possibility is that pericyte subtypes, rather than being constitutively distinct cells, may in turn reflect an activated/dedifferentiated phenotype upon injury (6,7). If correct, this would mean that the local
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 833–834
833
Commentary
Table 1. Reported effect of pericytes in cutaneous wounds (studies shown in reverse chronological order) Animal
Wound model
Pericyte phenotype
Brief description of method
Reported activity of pericytes
Ref
Human foreskin-derived CD45-VLA1bri cells
Topical application of pericytes in collagen type I gels
(4)
Human ASCs (aSMA+, CD140b+, NG2+, PDGFRb+, ANGPT+) from discarded burn skin samples PKH67-labeled human umbilical cord perivascular cells (HUCPVCs)
Topical application of pericytes in PEG-fibrin gels
Delay in dermal wound resolution Prolonged inflammatory response (↑ Neutrophils and macrophages) No effect in re-epithelialisation and wound closure Earlier collagen deposition and wound remodelling ↑ Vascularisation
Topical application of pericytes in fibrin gels
Accelerated wound closure ↑ Vascularisation and granulation tissue Improved skin architecture and tensile strength
(9)
5-mm excisional fullthickness wounds
PDGFRb+, aSMA+, NG2+
4-mm excisional fullthickness wounds
aSMA+, NG2+
F344/DuCrj male rats
5-mm excisional fullthickness wounds
PDGFRb+, aSMA+, Desmin+, Thy-1+, A3+ cells surrounding vWF+ vessels
Immunohistochemical evaluation up to 26 days postwounding
Male Balb/c mice
2-mm excisional fullthickness wounds
PDGFRb+, Desmin+, NG2+, aSMA+, VEGFA+ surrounding Cd31+ vessels
Intravital blood vessel labelling with lectin, immunohistochemistry and TEM
Col1a2 promoter-driven LacZ mice.
4-mm excisional fullthickness wounds
NG2+, surrounding Cd31+ vessels
Imatinib used to block PDGFRb signalling.
MFG-E8 expressed by PDGFRb+, aSMA+, NG2+ pericytes stimulates angiogenesis CCN2 is required to recruit pericytes to wounds. No effect on wound healing Promote angiogenesis Related to early (less differentiated) myofibroblast formation Inhibition (vessel stabilization) and promotion (in the tip) of endothelial cell proliferation during angiogenesis, depending on the pericyte subtype ↑ Vascularisation Stimulate myofibroblast differentiation
(S26)
Col1Cre; CCN2-/- mice
Wounding of MFG-E8-/- mice. MFG-E8 is a glycoprotein expressed by PDGFRb+, aSMA+, NG2+ pericytes. Wounding of Col1 Cre CCN2-/- mice.
Epidermal regeneration of bilayered skin constructs
mAb HD-1bri cells (↑ for aSMA, MYH9, TPM1, TPM2, RGS5, NG2, PDGFRb, ENEP, CNN1, CALD1) Human foreskin-derived HMW-MAA+ vascular cells
Incorporated pericytes into dermal gels, then overlayed with keratinocytes to form skin constructs. Pericytes and fibroblasts loaded into collagen gels
In vivo pericyte transplantation studies Balb/c SCID male mice 4-mm excisional fullthickness wounds
Male Rowett nude rats (athymic)
1.5-cm excisional full-thickness wounds
Female Balb/c nude mice
4-mm and 8-mm excisional full-thickness wounds
In vivo pericyte manipulation studies MFG-E8-/-, C57Bl6 background
In vitro studies N/A
N/A
Gel contraction assay.
wound environment would direct activated pericytes into regeneration- or fibrosis-promoting roles, depending on the circumstances. For transplantation studies, this would mean that the different biomaterials and culture media used for cell delivery (Table 1) might have a key role to obtain the desired clinical outcome. In studies where the wound-healing-promoting role of endogenous pericytes is to be enhanced, the role of PDGF (s20, s22, s23, s24, s25, s26), CXCL12/CXCR4 (s27), TGFb (s20, s21) and WNT/b-catenin (s28, s29, s30) signalling pathways should be explored. In summary, pericytes display a number of mesenchymal stem cell properties, making them a potentially exciting area for developing cell-based therapies to tackle wound pathology, yet little work has been conducted in this area when compared to other skin resident stem cell populations, such as those found within
(8)
(S36) (S37)
(S38)
(S24)
Enhance epidermal regeneration
(1)
Stimulate myofibroblast differentiation
(S24)
the dermal adipose tissue, or the hair follicle (s31, s32, s33, s34, s35). The work of Mills et al. highlights that a detailed understanding of pericyte activity during skin injury repair remains critically lacking, which needs to be urgently addressed if we ever hope to exploit this intriguing cell population to tackle the evergrowing burden of chronic wounds.
Author contribution DA and AI wrote the manuscript.
Conflict of interest The authors declare no conflict of interest
Supporting Information Additional supporting data may be found in the supplementary information of this article. Appendix S1. References.
References 1 Paquet-Fifield S, Schluter H, Li A et al. J Clin Invest 2009: 119: 2795–2806. 2 Murray I R, West C C, Hardy W R et al. Cell Mol Life Sci 2014: 71: 1353–1374. 3 Khosrotehrani K. Exp Dermatol 2013: 22: 307– 310.
834
4 Mills S J, Zhuang L, Arandjelovic P et al. Exp Dermatol 2015: 24: 881–883. 5 Ansell D M, Holden K A, Hardman M J. Exp Dermatol 2012: 21: 581–585. 6 Etxaniz U, Perez-San Vicente A, Gago-Lopez N, et al. Stem Cell Reports 2014: 3: 774–788.
7 Johnston A P, Naska S, Jones K et al. Stem Cell Reports 2013: 2013: 38–45. 8 Zamora D O, Natesan S, Becerra S et al. Angiogenesis 2013: 16: 745–757. 9 Zebardast N, Lickorish D, Davies J E. Organogenesis 2010: 6: 197–203.
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 833–834
