Canine Follicle Stem Cell Candidates Reside in the Bulge and Share Characteristic Features with Human Bulge Cells  Tetsuro Kobayashi, Toshiroh Iwasaki, Masayuki Amagai, Manabu Ohyama  Journal of Investigative Dermatology  Volume 130, Issue 8, Pages 1988-1995 (August 2010) DOI: 10.1038/jid.2010.68 Copyright © 2010 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 1 The anatomical location of LRCs in canine hair follicles resembled that in human hair follicles. (a) A canine skin xenografting model allowed the detection of BrdU label-retaining cells (LRCs) in canine hair follicles. (b, c) In anagen hair follicles, LRCs (arrowheads) were predominantly detected in the outermost layer of the outer root sheath around the insertion point of the arrector pili muscle. This anatomical location of LRCs resembles that in human anagen hair follicles. (d, e) In telogen hair follicles, LRCs were detected in the bulge (d) and in the secondary hair germ (e). Scale bars=100μm for b, 40μm for c–e. Journal of Investigative Dermatology 2010 130, 1988-1995DOI: (10.1038/jid.2010.68) Copyright © 2010 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 2 Upregulation of human bulge cell biomarkers in canine keratinocytes enriched for bulge cells. (a) Keratin 15 immunoreactivity (red) was observed in the bulge area (arrowheads) of canine hair follicles. Arrector pili muscle was highlighted by alpha smooth muscle actin staining (brown) (b) Canine anagen hair follicles were microdissected into four fragments: P1 (bulb), P2 (suprabulbar), P3 (isthmus including bulge), and P4 (infundibulum). (c) Keratin 15 and the human specific bulge markers CD200 and DIO2 were up-regulated in bulge cell-enriched P3 fragments (*P<0.05) (P3). (d) Follistatin, a human bulge specific marker, was also expressed in the canine bulge area (arrowheads). SG, sebaceous gland; APM, arrector pili muscle. Scale bars=100μm for a, 80μm for d. Mean±SEM; n=3. Journal of Investigative Dermatology 2010 130, 1988-1995DOI: (10.1038/jid.2010.68) Copyright © 2010 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 3 Bulge cell key regulator genes were similarly expressed in canine and human hair follicles. SOX9, LHX2, TCF3, and NFATC1 were up-regulated in canine and human hair follicles (*P<0.05). Note that SOX9, the master regulator of bulge cell development, was distinctively overrepresented in both hair follicles (P=0.002). Mean±SEM; n=3. Journal of Investigative Dermatology 2010 130, 1988-1995DOI: (10.1038/jid.2010.68) Copyright © 2010 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 4 Bulge cell-enriched canine hair follicle keratinocytes demonstrated high proliferative capacity in vitro. (a) Single canine anagen hair follicles were cut into four fragments: P1 (bulb), P2 (suprabulbar), P3 (isthmus enriched for the bulge cells), P4 (infundibulum). (b) P3 and P1 fragments predominantly formed colonies. Each row represents a respective single hair follicle. (c, d) P3-derived colonies demonstrated holoclone colony phenotype, while P1-derived colonies did not. (e) The percentage of holoclone colonies was significantly higher with P3 than with P1 keratinocytes. (f, g) P3 keratinocytes could be cultured for multiple passages. The panels in f demonstrate consecutive passages. HF: hair follicle, KC: keratinocyte. Scale bar=500μm for d. Mean±SEM; *P<0.05; (c, e) n=9, (g) n=3. Journal of Investigative Dermatology 2010 130, 1988-1995DOI: (10.1038/jid.2010.68) Copyright © 2010 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 5 Patch grafting of canine bulge cell–enriched keratinocytes and trichogenic canine/mouse dermal cells regenerated complete pilosebaceous structures and the epidermis. (a, b) Co-transplantation of canine isthmus keratinocytes and canine dermal papilla cells formed hair shaft-like structures (arrowheads). (c) Combination of canine isthmus keratinocytes and mouse neonatal dermal cells successfully regenerated hair follicles. (d) Neonatal mouse dermal cells alone could not regenerate hair follicles. (e) The reconstituted tissue contained complete anagen follicle structures with sebaceous glands and contiguous epidermis. (f, g) Both mouse-size (arrow) and larger hair follicles (arrowhead) were observed in the patch structure. KC, keratinocyte; DP, dermal papilla; SG, sebaceous gland. Scale bars=40μm for (b), 80μm for (e), 250μm for (f), 300μm for (g). (b), (e), (f): Hematoxylin and eosin staining. Journal of Investigative Dermatology 2010 130, 1988-1995DOI: (10.1038/jid.2010.68) Copyright © 2010 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 6 Canine bulge cell enriched keratinocytes participated in the neogenesis of pilosebaceous structures, including the reorganization of the bulge area and the sebaceous gland. (a) Canine Y chromosome specific fragments (650bp) were amplified exclusively from hair follicle-like structures (dog KC+dog DP) and reconstituted pilosebaceous structures (dog KC+mouse dermal cells: each lane represents samples from different patches), suggesting that canine cells contributed to tissue regeneration. (b) Canine mRNA of bulge markers, K15, CD200, follistatin and a sebocyte lineage gene, PPARG, were successfully detected from a dog hair follicle (HF) and reconstituted HFs (each lane represents respective HFs), but not from a mouse HF (negative control), indicating that canine keratinocytes repopulated the bulge area and the sebaceous gland. Journal of Investigative Dermatology 2010 130, 1988-1995DOI: (10.1038/jid.2010.68) Copyright © 2010 The Society for Investigative Dermatology, Inc Terms and Conditions