Letter to the Editor
(PD98059) or Akt inhibitor (LY294002). Both PD98059 and LY294002 abolished the inhibition of melanin synthesis by clotrimazole (Figure S3b). Moreover, PD98059 and LY294002 attenuated the inhibition of tyrosinase expression by clotrimazole (Figure S3c). Taken together, these results indicate that clotrimazole suppressed melanogenesis through the proteasomal degradation of tyrosinase and that this may be associated with the activation of the ERK and Akt signalling pathways.
safe depigmenting agent that is suitable for the topical treatment of hyperpigmentary disorders related to fungal infection and of other sources of inflammation.
Conclusions
BYC wrote the manuscript; SYK performed the research; JMJ assisted in writing the manuscript; CHW, JHC and MWL analysed the data; and SEC supervised the study and the manuscript.
Herein, we have demonstrated a novel role for the antimycotic agent clotrimazole in the inhibition of melanogenesis by accelerating ERK and PI3K-/Akt-mediated tyrosinase degradation, which is a novel mechanism. These findings indicate that clotrimazole is a
Acknowledgements This research was supported by a Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology(2011-0009121).
Author contributions
Conflict of interests The authors have declared no conflicting interests.
References 1 Sur R, Babad J M, Garay M et al. J Invest Dermatol 2008: 128: 336–344. 2 Tsuji G, Takahara M, Uchi H et al. J Invest Dermatol 2012: 132: 59–68. 3 Hirata N, Naruto S, Ohguchi K et al. Bioorg Med Chem 2007: 15: 4897– 4902. 4 Gledhill K, Rhodes L E, Brownrigg M et al. Pigment Cell Melanoma Res 2010: 23: 394– 403.
5 Shin J M, Kim M Y, Sohn K C et al. PLoS ONE 2014: 9: e96035. 6 Bellei B, Pitisci A, Izzo E et al. PLoS ONE 2012: 7: e33021. 7 Bellei B, Maresca V, Flori E et al. J Biol Chem 2010: 285: 7288–7299. 8 Lee M S, Yoon H D, Kim J I et al. Exp Dermatol 2012: 21: 471–473. 9 Madi L, Rosenberg-Haggen B, Nyska A et al. Exp Dermatol 2013: 22: 74–77.
Supporting Information Additional supporting data may be found in the supplementary information of this article. Data S1. Materials and methods. Figure S1. Cell viability assay of clotrimazole. Figure S2. Clotrimazole inhibits melanogenesis in Mel-Ab cells. Figure S3. The effects of clotrimazole on melanogenesis are related to ERK and Akt signalling pathways.
DOI: 10.1111/exd.12670
Letter to the Editor
www.wileyonlinelibrary.com/journal/EXD
Paracrine crosstalk between human hair follicle dermal papilla cells and microvascular endothelial cells Eleonora Bassino1, Franco Gasparri2, Valentina Giannini2 and Luca Munaron1 1
Department of Life Sciences and Systems Biology, University of Turin, Turin, Italy; 2Rottapharm-Madaus, Monza, Italy Correspondence: Luca Munaron, PhD, Dept. Life Sciences & Systems Biology, University of Torino, Via Accademia Albertina 13, 10123 Torino, Italy, Tel.: 00390116704667; Fax: 00390116704508; e-mail: [email protected] Abstract: Human follicle dermal papilla cells (FDPC) are a specialized population of mesenchymal cells located in the skin. They regulate hair follicle (HF) development and growth, and represent a reservoir of multipotent stem cells. Growing evidence supports the hypothesis that HF cycling is associated with vascular remodeling. Follicular keratinocytes release vascular endothelial growth factor (VEGF) that sustains perifollicular angiogenesis leading to an increase of follicle and hair size. Furthermore, several human diseases characterized by hair loss, including Androgenetic Alopecia, exhibit alterations of skin vasculature. However, the molecular mechanisms underlying HF vascularization remain largely unknown. In vitro coculture approaches can be successfully
employed to greatly improve our knowledge and shed more light on this issue. Here we used Transwell-based co-cultures to show that FDPC promote survival, proliferation and tubulogenesis of human microvascular endothelial cells (HMVEC) more efficiently than fibroblasts. Accordingly, FDPC enhance the endothelial release of VEGF and IGF-1, two well-known proangiogenic growth factors. Collectively, our data suggest a key role of papilla cells in vascular remodeling of the hair follicle.
Background
reservoir of multipotent stem cells (1,2). Growing evidence supports the hypothesis that HF cycling is associated with vascular remodelling (3). Follicular keratinocytes release vascular endothelial growth factor (VEGF) that sustains perifollicular angiogenesis
Human follicle dermal papilla cells (FDPCs) are a specialized population of mesenchymal cells located in the skin. They regulate hair follicle (HF) development and growth and represent a
388
Key words: Coculture – endothelial cells – follicle dermal papilla cells – hair growth – hair follicle
Accepted for publication 11 February 2015
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 381–400
Letter to the Editor
(a)
(a)
(b)
(b)
(c)
(c)
(d)
(e)
(d)
(e)
(f)
Figure 1. FDPCs promote HMVEC survival, proliferation and tubulogenesis and inhibit IL-1a production. (a) Scheme of the experimental set-up. Arabic numbers in graphs (1–3) are referred to the monoculture and coculture configurations. 1) HMVEC grown alone, 2) HMVEC grown in presence of the insert filled with cell culture medium only (in the absence of FDPCs) and 3) HMVEC cocultured with FDPCs. (b) HMVEC viability is maximal in DMEM 10% FCS and is reduced in DMEM with lower percentage of FCS (0–2%) (24 h). HMVEC survival is significantly increased when cocultured with FDPCs grown in DMEM 2 or 10% FCS (24 h). (c) HMVEC proliferation is significantly increased when cocultured with FDPCs grown in DMEM 2 or 10% FCS (48 h). Light grey bars (in both survival and proliferative experiments) represent the negative control condition in coculture: the insert is filled with cell culture medium only (DMEM 0% FCS or DMEM 10% FCS) in the absence of FDPCs. (d) Tubule formation at 18 h is maximal when HMVECs are grown in Endogro medium and is significantly reduced when grown in DMEM (0, 10%). Tubulogenesis is significantly increased when HMVECs are cocultured with FDPCs maintained in DMEM 10% FCS. (e) VEGF is maximal when HMVECs are grown in Endogro medium and it is significantly reduced when grown in DMEM (0, 10% FCS). VEGF levels are significantly increased in the culture medium of HMVECs cocultured with FDPCs for 18 h. (f) H2O2 (400 lM, 2 h) treatment significantly increases IL-1a levels in HMVECs (24 h). IL-1a production is significantly decreased in HMVECs cocultured with FDPCs (24 h). Only a slight, but not significant, reduction is observed for a prolonged period of coculture (48 h). All values are expressed in A.U. as mean SEM.
leading to an increase of follicle and hair size (3). Furthermore, several human diseases characterized by hair loss, including androgenetic alopecia, exhibit alterations of skin vasculature (4–6). However, the molecular mechanisms underlying HF vascularization remain largely unknown (7). In vitro coculture approaches can be successfully employed to greatly improve our knowledge and shed more light on this issue. Here, we used transwell-based cocultures to show that FDPCs promote survival, proliferation and tubulogenesis of human microvascular endothelial cells (HMVECs) more efficiently than fibroblasts. Accordingly, FDPCs enhance the endothelial release of VEGF and IGF-1, two
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 381–400
Figure 2. HMVECs support short-term FDPC survival and enhance their b-catenin production. (a) Scheme of the experimental set-up. Arabic numbers in graphs (1–3) are referred to the monoculture and coculture configurations. 1) FDPC grown alone, 2) FDPC grown in presence of the insert filled with cell culture medium only (in the absence of HMVECs) and 3) FDPC cocultured with HMVECs. (b) FDPC viability is maximal in DMEM 10% FCS and is reduced when grown in DMEM with lower percentage of FCS (0–2%) (24 h). FDPC survival is significantly increased when cocultured with HMVECs grown in DMEM 10% FCS (24 h). (c) FDPC proliferation is not modified in coculture with HMVECs grown in DMEM with low percentage of FCS (0–2%) (48 h). In both viability and proliferation assays, light grey bars represent the negative control condition in coculture: the insert is filled with cell culture medium only (DMEM 0% FCS or DMEM 10% FCS) in the absence of HMVECs. Arabic numbers (1–3) are referred to the coculture configurations. (d) HMVECs increase b-catenin expression by quiescent FDPCs only upon shortterm treatment (viability culture protocol as in Fig. 1a, 24 h), while no significant changes are detected for prolonged periods (48 h). (e) No significant changes in b-catenin expression by proliferating FDPCs are observed for both 48 h and 72 h of coculture (proliferation culture protocol as in Fig. 1c). All values are expressed in A.U. as mean SEM.
well-known proangiogenic growth factors. Collectively, our data suggest a key role of papilla cells in vascular remodelling of the hair follicle (8).
Questions addressed This study provides an experimental validation of quantitative, robust and reliable protocols to investigate paracrine interactions occurring between different cell types associated to hair follicle vascularization by the use of in vitro coculture approach.
Experimental design See Data S1.
Results FDPCs promote HMVEC survival, proliferation and tubulogenesis and inhibit IL-1a production The maximal HMVEC viability was measured when the cells were grown in DMEM 10% FCS (see Methods and Fig. 1a for set-up configurations) and resulted drastically reduced upon serum deprivation (DMEM 0% or 2% FCS, 24 h) (Fig. 1b, Table S1). In coculture with FDPCs (24/48 h), the viability of starved HMVECs was significantly increased (Fig. 1b, Table S1). FDPCs were also
389
Letter to the Editor
able to enhance HMVEC proliferation upon 48 h or 72 h of coculture: this evidence is confirmed by propidium iodide assay (Fig. 1c, Table S1, Figure S1). In addition, FDPCs strongly promoted in vitro tubulogenesis of HMVECs cultured in 3D matrigel (Fig. 1d, Figure S2). These functional effects are in nice agreement with the ability of FDPCs to release VEGF and IGF-1, two wellknown proangiogenic growth factors (Fig. 1e, Table S2). As IL-1a is a cytokine prominent in skin wounding and during inflammatory responses, as well as a negative hair growth regulator, we decided to evaluate the effects of FDPCs on its production by the endothelium. HMVECs were treated with H2O2 (400 lM, 2 h) to induce a strong oxidative stress. In this condition, IL-1a levels in endothelial medium drastically increased, as expected (24 h) (Fig. 1f). Upon 24 h of coculture, FDPCs significantly reduced IL-1a production by HMVECs (Fig. 1f), while no significant change was observed for a prolonged period of coculture (48 h). To determine the specificity of the effects exerted by FDPCs on HMVECs, we cocultured endothelium with normal human dermal fibroblasts (NHDFs). NHDFs actually exhibited a trophic and protubulogenic activity on HMVECs (Figure S3a,b), but at much lower extent when compared to FDPCs (cell viability: 45% and 19% of increment from negative control promoted by FDPCs and NHDFs, respectively; tubulogenesis index: 70% and 28% of increment induced by FDPCs and NHDFs, respectively). Accordingly, VEGF production was less pronounced in fibroblasts compared to FDPCs (Figure S3c). Moreover, unlike FDPCs, human fibroblasts completely failed to affect IL-1a release triggered by oxidative stress (Figure S3d).
HMVECs support short-term survival and enhance b-catenin production by FDPCs In the complementary configuration (see schemes 1–3 in Fig. 2a), HMVECs sustained FDPC viability only for a short period of coculture (24 h), resulting ineffective at 48 h (Fig. 2b). In addition, endothelial cells were not mitogenic for FDPCs (48/72 h; Fig. 2c and Table S1). Finally, we evaluated the production of b-catenin, a key player in the regulation of follicle cell adhesion and signalling (9), by quiescent and proliferating FDPCs (see detailed Methods in Data S1). HMVECs enhanced nuclear b-catenin expression by FDPCs
only upon short-term treatment (24 h; Fig. 2d), while no changes were detected for prolonged periods of coculture (48/72 h) (Fig. 2e).
Conclusions In this study, we employed and validated coculture technique to investigate the paracrine crosstalk between human FDPCs and endothelium in vitro. Remarkably, FDPCs provide a strong support for endothelial survival, proliferation and tubulogenesis (Fig. 1 and Table S1). Accordingly, FDPCs release VEGF and IGF-1, two potent proangiogenic mediators whose expression changes during hair growth (3,10) (Table S2). Other crucial players in skin development such as human dermal fibroblasts (NHDFs)(11) appear less powerful than FDPCs as trophic and protubulogenic sources for endothelium, releasing smaller amount of VEGF. Moreover, NHDFs are completely unable to mimic the protective activity of FDPCs on HMVECs exposed to oxidative injury, as shown by IL-1a assays. This observation further highlights the differential paracrine effects of follicle papilla cells and fibroblasts in sustaining skin vasculature (Figure S3). The relationship between FDPCs and HMVECs appears somehow asymmetrical with the inductive power of the first prevailing on the reciprocal one. Indeed, endothelium is not mitogenic and supports only short time FDPC survival (Figure 2). We acknowledge the limitations of our simplified in vitro set-up based on cell culture that takes into account only a part of the complex architecture of native organs. Nonetheless, the validation of this method, that allows to test the effects of pharmacological compounds and drugs by the use of highthroughput and standardized protocols, could successfully integrate and support the well-established ex vivo and in vivo approaches (12).
Acknowledgements E.B. performed the research, analysed the data and wrote the manuscript. F.G. contributed essential reagents or tools and designed the research study. V. G. contributed essential reagents or tools. L.M. designed the research study, analysed the data, wrote the manuscript and contributed essential reagents or tools. We would like to thank Dr. Daniele Avanzato for graphical assistance.
Conflicts of interest None declared.
References 1 Driskell R R, Clavel C, Rendl M et al. J Cell Sci 2011: 124: 1179–1182. 2 Jahoda C A, Christiano A M. Cell 2011: 146: 678–681. 3 Yano K, Brown L F, Detmar M. J Clin Invest 2001: 107: 409–417. 4 Montagna W, Ellis R A. J Natl Cancer Inst 1957: 19: 451–463. 5 Ellis R A, Moretti G. Ann N Y Acad Sci 1959: 83: 448–457. 6 Braverman I M, Fonferko E. J Invest Dermatol 1982: 78: 444–448. 7 Xiao Y, Woo W-M, Nagao K et al. J Invest Dermatol 2014: 133: 2324–2331.
390
8 Hill R P, Gardner A, Crawford H C, et al. Exp Dermatol 2013: 22: 236–238. 9 Fagotto F. EMBO Rep 2013: 14: 422– 433. 10 Li W, Man X-Y, Li C-M, et al. Exp Cell Res 2012: 318: 1633–1640. 11 Driskell R R, Lichtenberger B M, Hoste E, et al. Nature 2013: 504: 277–281. 12 Jahoda C A, Horne K A O R. Nature 1984: 311: 560–562.
Supporting Information Additional supporting data may be found in the supplementary information of this article.
Figure S1. Propidium iodide test confirms the trophic effect of FDPC on HMVEC shown in figure 1 of the paper. Figure S2. Representative images used to quantify the tubulogenic effect of FDPC on HMVEC described in Fig. 1c of the manuscript. Figure S3. Effects of normal human dermal fibroblasts (NHDF) on HMVEC. Table S1. Cell viability (48 h) and proliferation (72 h) assays performed on HMVEC or FDPC grown separately or in co-culture (see Methods for detailed protocols). Table S2. VEGF and IGF-1 production by HMVEC and FDPC grown separately. Data S1. Material.
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 381–400
(PD98059) or Akt inhibitor (LY294002). Both PD98059 and LY294002 abolished the inhibition of melanin synthesis by clotrimazole (Figure S3b). Moreover, PD98059 and LY294002 attenuated the inhibition of tyrosinase expression by clotrimazole (Figure S3c). Taken together, these results indicate that clotrimazole suppressed melanogenesis through the proteasomal degradation of tyrosinase and that this may be associated with the activation of the ERK and Akt signalling pathways.
safe depigmenting agent that is suitable for the topical treatment of hyperpigmentary disorders related to fungal infection and of other sources of inflammation.
Conclusions
BYC wrote the manuscript; SYK performed the research; JMJ assisted in writing the manuscript; CHW, JHC and MWL analysed the data; and SEC supervised the study and the manuscript.
Herein, we have demonstrated a novel role for the antimycotic agent clotrimazole in the inhibition of melanogenesis by accelerating ERK and PI3K-/Akt-mediated tyrosinase degradation, which is a novel mechanism. These findings indicate that clotrimazole is a
Acknowledgements This research was supported by a Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology(2011-0009121).
Author contributions
Conflict of interests The authors have declared no conflicting interests.
References 1 Sur R, Babad J M, Garay M et al. J Invest Dermatol 2008: 128: 336–344. 2 Tsuji G, Takahara M, Uchi H et al. J Invest Dermatol 2012: 132: 59–68. 3 Hirata N, Naruto S, Ohguchi K et al. Bioorg Med Chem 2007: 15: 4897– 4902. 4 Gledhill K, Rhodes L E, Brownrigg M et al. Pigment Cell Melanoma Res 2010: 23: 394– 403.
5 Shin J M, Kim M Y, Sohn K C et al. PLoS ONE 2014: 9: e96035. 6 Bellei B, Pitisci A, Izzo E et al. PLoS ONE 2012: 7: e33021. 7 Bellei B, Maresca V, Flori E et al. J Biol Chem 2010: 285: 7288–7299. 8 Lee M S, Yoon H D, Kim J I et al. Exp Dermatol 2012: 21: 471–473. 9 Madi L, Rosenberg-Haggen B, Nyska A et al. Exp Dermatol 2013: 22: 74–77.
Supporting Information Additional supporting data may be found in the supplementary information of this article. Data S1. Materials and methods. Figure S1. Cell viability assay of clotrimazole. Figure S2. Clotrimazole inhibits melanogenesis in Mel-Ab cells. Figure S3. The effects of clotrimazole on melanogenesis are related to ERK and Akt signalling pathways.
DOI: 10.1111/exd.12670
Letter to the Editor
www.wileyonlinelibrary.com/journal/EXD
Paracrine crosstalk between human hair follicle dermal papilla cells and microvascular endothelial cells Eleonora Bassino1, Franco Gasparri2, Valentina Giannini2 and Luca Munaron1 1
Department of Life Sciences and Systems Biology, University of Turin, Turin, Italy; 2Rottapharm-Madaus, Monza, Italy Correspondence: Luca Munaron, PhD, Dept. Life Sciences & Systems Biology, University of Torino, Via Accademia Albertina 13, 10123 Torino, Italy, Tel.: 00390116704667; Fax: 00390116704508; e-mail: [email protected] Abstract: Human follicle dermal papilla cells (FDPC) are a specialized population of mesenchymal cells located in the skin. They regulate hair follicle (HF) development and growth, and represent a reservoir of multipotent stem cells. Growing evidence supports the hypothesis that HF cycling is associated with vascular remodeling. Follicular keratinocytes release vascular endothelial growth factor (VEGF) that sustains perifollicular angiogenesis leading to an increase of follicle and hair size. Furthermore, several human diseases characterized by hair loss, including Androgenetic Alopecia, exhibit alterations of skin vasculature. However, the molecular mechanisms underlying HF vascularization remain largely unknown. In vitro coculture approaches can be successfully
employed to greatly improve our knowledge and shed more light on this issue. Here we used Transwell-based co-cultures to show that FDPC promote survival, proliferation and tubulogenesis of human microvascular endothelial cells (HMVEC) more efficiently than fibroblasts. Accordingly, FDPC enhance the endothelial release of VEGF and IGF-1, two well-known proangiogenic growth factors. Collectively, our data suggest a key role of papilla cells in vascular remodeling of the hair follicle.
Background
reservoir of multipotent stem cells (1,2). Growing evidence supports the hypothesis that HF cycling is associated with vascular remodelling (3). Follicular keratinocytes release vascular endothelial growth factor (VEGF) that sustains perifollicular angiogenesis
Human follicle dermal papilla cells (FDPCs) are a specialized population of mesenchymal cells located in the skin. They regulate hair follicle (HF) development and growth and represent a
388
Key words: Coculture – endothelial cells – follicle dermal papilla cells – hair growth – hair follicle
Accepted for publication 11 February 2015
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 381–400
Letter to the Editor
(a)
(a)
(b)
(b)
(c)
(c)
(d)
(e)
(d)
(e)
(f)
Figure 1. FDPCs promote HMVEC survival, proliferation and tubulogenesis and inhibit IL-1a production. (a) Scheme of the experimental set-up. Arabic numbers in graphs (1–3) are referred to the monoculture and coculture configurations. 1) HMVEC grown alone, 2) HMVEC grown in presence of the insert filled with cell culture medium only (in the absence of FDPCs) and 3) HMVEC cocultured with FDPCs. (b) HMVEC viability is maximal in DMEM 10% FCS and is reduced in DMEM with lower percentage of FCS (0–2%) (24 h). HMVEC survival is significantly increased when cocultured with FDPCs grown in DMEM 2 or 10% FCS (24 h). (c) HMVEC proliferation is significantly increased when cocultured with FDPCs grown in DMEM 2 or 10% FCS (48 h). Light grey bars (in both survival and proliferative experiments) represent the negative control condition in coculture: the insert is filled with cell culture medium only (DMEM 0% FCS or DMEM 10% FCS) in the absence of FDPCs. (d) Tubule formation at 18 h is maximal when HMVECs are grown in Endogro medium and is significantly reduced when grown in DMEM (0, 10%). Tubulogenesis is significantly increased when HMVECs are cocultured with FDPCs maintained in DMEM 10% FCS. (e) VEGF is maximal when HMVECs are grown in Endogro medium and it is significantly reduced when grown in DMEM (0, 10% FCS). VEGF levels are significantly increased in the culture medium of HMVECs cocultured with FDPCs for 18 h. (f) H2O2 (400 lM, 2 h) treatment significantly increases IL-1a levels in HMVECs (24 h). IL-1a production is significantly decreased in HMVECs cocultured with FDPCs (24 h). Only a slight, but not significant, reduction is observed for a prolonged period of coculture (48 h). All values are expressed in A.U. as mean SEM.
leading to an increase of follicle and hair size (3). Furthermore, several human diseases characterized by hair loss, including androgenetic alopecia, exhibit alterations of skin vasculature (4–6). However, the molecular mechanisms underlying HF vascularization remain largely unknown (7). In vitro coculture approaches can be successfully employed to greatly improve our knowledge and shed more light on this issue. Here, we used transwell-based cocultures to show that FDPCs promote survival, proliferation and tubulogenesis of human microvascular endothelial cells (HMVECs) more efficiently than fibroblasts. Accordingly, FDPCs enhance the endothelial release of VEGF and IGF-1, two
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 381–400
Figure 2. HMVECs support short-term FDPC survival and enhance their b-catenin production. (a) Scheme of the experimental set-up. Arabic numbers in graphs (1–3) are referred to the monoculture and coculture configurations. 1) FDPC grown alone, 2) FDPC grown in presence of the insert filled with cell culture medium only (in the absence of HMVECs) and 3) FDPC cocultured with HMVECs. (b) FDPC viability is maximal in DMEM 10% FCS and is reduced when grown in DMEM with lower percentage of FCS (0–2%) (24 h). FDPC survival is significantly increased when cocultured with HMVECs grown in DMEM 10% FCS (24 h). (c) FDPC proliferation is not modified in coculture with HMVECs grown in DMEM with low percentage of FCS (0–2%) (48 h). In both viability and proliferation assays, light grey bars represent the negative control condition in coculture: the insert is filled with cell culture medium only (DMEM 0% FCS or DMEM 10% FCS) in the absence of HMVECs. Arabic numbers (1–3) are referred to the coculture configurations. (d) HMVECs increase b-catenin expression by quiescent FDPCs only upon shortterm treatment (viability culture protocol as in Fig. 1a, 24 h), while no significant changes are detected for prolonged periods (48 h). (e) No significant changes in b-catenin expression by proliferating FDPCs are observed for both 48 h and 72 h of coculture (proliferation culture protocol as in Fig. 1c). All values are expressed in A.U. as mean SEM.
well-known proangiogenic growth factors. Collectively, our data suggest a key role of papilla cells in vascular remodelling of the hair follicle (8).
Questions addressed This study provides an experimental validation of quantitative, robust and reliable protocols to investigate paracrine interactions occurring between different cell types associated to hair follicle vascularization by the use of in vitro coculture approach.
Experimental design See Data S1.
Results FDPCs promote HMVEC survival, proliferation and tubulogenesis and inhibit IL-1a production The maximal HMVEC viability was measured when the cells were grown in DMEM 10% FCS (see Methods and Fig. 1a for set-up configurations) and resulted drastically reduced upon serum deprivation (DMEM 0% or 2% FCS, 24 h) (Fig. 1b, Table S1). In coculture with FDPCs (24/48 h), the viability of starved HMVECs was significantly increased (Fig. 1b, Table S1). FDPCs were also
389
Letter to the Editor
able to enhance HMVEC proliferation upon 48 h or 72 h of coculture: this evidence is confirmed by propidium iodide assay (Fig. 1c, Table S1, Figure S1). In addition, FDPCs strongly promoted in vitro tubulogenesis of HMVECs cultured in 3D matrigel (Fig. 1d, Figure S2). These functional effects are in nice agreement with the ability of FDPCs to release VEGF and IGF-1, two wellknown proangiogenic growth factors (Fig. 1e, Table S2). As IL-1a is a cytokine prominent in skin wounding and during inflammatory responses, as well as a negative hair growth regulator, we decided to evaluate the effects of FDPCs on its production by the endothelium. HMVECs were treated with H2O2 (400 lM, 2 h) to induce a strong oxidative stress. In this condition, IL-1a levels in endothelial medium drastically increased, as expected (24 h) (Fig. 1f). Upon 24 h of coculture, FDPCs significantly reduced IL-1a production by HMVECs (Fig. 1f), while no significant change was observed for a prolonged period of coculture (48 h). To determine the specificity of the effects exerted by FDPCs on HMVECs, we cocultured endothelium with normal human dermal fibroblasts (NHDFs). NHDFs actually exhibited a trophic and protubulogenic activity on HMVECs (Figure S3a,b), but at much lower extent when compared to FDPCs (cell viability: 45% and 19% of increment from negative control promoted by FDPCs and NHDFs, respectively; tubulogenesis index: 70% and 28% of increment induced by FDPCs and NHDFs, respectively). Accordingly, VEGF production was less pronounced in fibroblasts compared to FDPCs (Figure S3c). Moreover, unlike FDPCs, human fibroblasts completely failed to affect IL-1a release triggered by oxidative stress (Figure S3d).
HMVECs support short-term survival and enhance b-catenin production by FDPCs In the complementary configuration (see schemes 1–3 in Fig. 2a), HMVECs sustained FDPC viability only for a short period of coculture (24 h), resulting ineffective at 48 h (Fig. 2b). In addition, endothelial cells were not mitogenic for FDPCs (48/72 h; Fig. 2c and Table S1). Finally, we evaluated the production of b-catenin, a key player in the regulation of follicle cell adhesion and signalling (9), by quiescent and proliferating FDPCs (see detailed Methods in Data S1). HMVECs enhanced nuclear b-catenin expression by FDPCs
only upon short-term treatment (24 h; Fig. 2d), while no changes were detected for prolonged periods of coculture (48/72 h) (Fig. 2e).
Conclusions In this study, we employed and validated coculture technique to investigate the paracrine crosstalk between human FDPCs and endothelium in vitro. Remarkably, FDPCs provide a strong support for endothelial survival, proliferation and tubulogenesis (Fig. 1 and Table S1). Accordingly, FDPCs release VEGF and IGF-1, two potent proangiogenic mediators whose expression changes during hair growth (3,10) (Table S2). Other crucial players in skin development such as human dermal fibroblasts (NHDFs)(11) appear less powerful than FDPCs as trophic and protubulogenic sources for endothelium, releasing smaller amount of VEGF. Moreover, NHDFs are completely unable to mimic the protective activity of FDPCs on HMVECs exposed to oxidative injury, as shown by IL-1a assays. This observation further highlights the differential paracrine effects of follicle papilla cells and fibroblasts in sustaining skin vasculature (Figure S3). The relationship between FDPCs and HMVECs appears somehow asymmetrical with the inductive power of the first prevailing on the reciprocal one. Indeed, endothelium is not mitogenic and supports only short time FDPC survival (Figure 2). We acknowledge the limitations of our simplified in vitro set-up based on cell culture that takes into account only a part of the complex architecture of native organs. Nonetheless, the validation of this method, that allows to test the effects of pharmacological compounds and drugs by the use of highthroughput and standardized protocols, could successfully integrate and support the well-established ex vivo and in vivo approaches (12).
Acknowledgements E.B. performed the research, analysed the data and wrote the manuscript. F.G. contributed essential reagents or tools and designed the research study. V. G. contributed essential reagents or tools. L.M. designed the research study, analysed the data, wrote the manuscript and contributed essential reagents or tools. We would like to thank Dr. Daniele Avanzato for graphical assistance.
Conflicts of interest None declared.
References 1 Driskell R R, Clavel C, Rendl M et al. J Cell Sci 2011: 124: 1179–1182. 2 Jahoda C A, Christiano A M. Cell 2011: 146: 678–681. 3 Yano K, Brown L F, Detmar M. J Clin Invest 2001: 107: 409–417. 4 Montagna W, Ellis R A. J Natl Cancer Inst 1957: 19: 451–463. 5 Ellis R A, Moretti G. Ann N Y Acad Sci 1959: 83: 448–457. 6 Braverman I M, Fonferko E. J Invest Dermatol 1982: 78: 444–448. 7 Xiao Y, Woo W-M, Nagao K et al. J Invest Dermatol 2014: 133: 2324–2331.
390
8 Hill R P, Gardner A, Crawford H C, et al. Exp Dermatol 2013: 22: 236–238. 9 Fagotto F. EMBO Rep 2013: 14: 422– 433. 10 Li W, Man X-Y, Li C-M, et al. Exp Cell Res 2012: 318: 1633–1640. 11 Driskell R R, Lichtenberger B M, Hoste E, et al. Nature 2013: 504: 277–281. 12 Jahoda C A, Horne K A O R. Nature 1984: 311: 560–562.
Supporting Information Additional supporting data may be found in the supplementary information of this article.
Figure S1. Propidium iodide test confirms the trophic effect of FDPC on HMVEC shown in figure 1 of the paper. Figure S2. Representative images used to quantify the tubulogenic effect of FDPC on HMVEC described in Fig. 1c of the manuscript. Figure S3. Effects of normal human dermal fibroblasts (NHDF) on HMVEC. Table S1. Cell viability (48 h) and proliferation (72 h) assays performed on HMVEC or FDPC grown separately or in co-culture (see Methods for detailed protocols). Table S2. VEGF and IGF-1 production by HMVEC and FDPC grown separately. Data S1. Material.
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 381–400
