- Email: [email protected]
Murine Vibrissae Cultured in Serum-Free Medium Reinitiate Anagen
J Lee et al. Anagen Reinitiation In Vitro
screen use by people who spent their holidays on the beach and who, by definition, were not sun avoiders. We found that more detailed sunscreen labeling resulted in subjects’ more effective use of sunscreens and fewer sunburns, indicating that the labeling affected the habits of sunscreen users toward their use of sunscreens; their sun exposure and use of protective clothing were unchanged. In support of their concern, Boniol et al. refer to their study (Autier et al., 2001) in which students who use highsun protection factor sunscreens tend to increase their sun exposure. We did not find such an effect in the adult population we studied. In fact, a randomized study (Dupuy et al., 2005) suggested that young adults could be considered a separate population due to their propensity toward sunbathing and other risky behaviors. We consider sun behavior to be guided by strong societal and psycho-
logical factors (Grob and Bonerandi, 1997) that are unlikely to be influenced by sunscreen labels. We believe that people who want to protect their skin by using sunscreens can benefit from more informative labeling, although Boniol et al. may regret that these people do not choose to wear protective clothing or avoid the sun entirely. A long-term campaign will be necessary to change predominant sun behaviors. The benefit of changing sunscreen labeling, although modest, could have immediate effects. CONFLICT OF INTEREST The authors state no conflict of interest.
Jean-Jacques Grob1 and Caroline Gaudy-Marqueste1 1
Department of Dermatology Hoˆpital Ste Marguerite, Assistance Publique des Hoˆpitaux de Marseille and Research Unit LIMP EA 3291, Universite´ de la Me´diterrane´e, Marseille, France. E-mail: [email protected]
REFERENCES Autier P, Boniol M, Severi G, Dore JF, European Organization for Research and Treatment of Cancer Melanoma Co-operative Group (2001) Quantity of sunscreen used by European students. Br J Dermatol 144: 288–91 Boniol M, Dore JF, Autier P (2008) Changing the labeling of sunscreen, will we transform sun avoiders into sunscreen users? J Invest Dermatol 128:482 Dupuy A, Dunant A, Grob JJ (2005) Reseau d’epidemiologie en dermatology. Randomized controlled trial testing the impact of high-protection sunscreens on sun-exposure behavior. Arch Dermatol 141:950–6 Grob JJ, Bonerandi JJ (1997) Attitudes and behaviour toward sun exposure: implications for melanoma prevention. In: Epidemiology, Causes and Prevention of Skin Diseases (Grob JJ, Stern R Mackie R, Weinstock M, eds), Oxford: Blackwell Science, 144–51 Nicol I, Gaudy C, Gouvernet J, Richard MA, Grob JJ (2007) Skin protection by sunscreens is improved by explicit labeling and providing free sunscreen. J Invest Dermatol 127: 41–8
Murine Vibrissae Cultured in Serum-Free Medium Reinitiate Anagen Journal of Investigative Dermatology (2008) 128, 482–485; doi:10.1038/sj.jid.5701024; published online 9 August 2007
TO THE EDITOR Aberrant regulation of the hair cycle has been implicated in human baldness (reviewed in Paus and Cotsarelis, 1999; Nakamura et al., 2001). Although several modulators of mammalian hair follicle cycle have been recently described, discovery of effective treatment for baldness has suffered from the absence of a reliable in vitro culture system in which the hair cycle can be assessed easily and inexpensively (Stenn and Paus, 2001). Several laboratories have established serum-free culture systems where vibrissa can grow and differentiate in vitro (Table S1; Jindo et al., 1993; Robinson et al., 1997; Yano et al., 2001). Although rat vibrissae cultured for 23 days were reported to share histologic similarities with catagen or pro-anagen (telogen) stage follicles (Philpott and Kealey, 482
2000), these follicles did not show any progress beyond the pro-anagen phase nor did they produce a new hair shaft (Philpott and Kealey, 2000). In this study, we demonstrate that a simple modification permits murine vibrissae in the current in vitro culture system to reinitiate anagen. This will accelerate the development of screening systems aimed at modifying the hair cycle. To establish a modified in vitro vibrissa culture system, anagen-stage vibrissae were carefully isolated from 14-day-old mice, and the tips of the vibrissa shafts were anchored in a stripe of sterilized silicone grease placed on the culture dish through a 3-ml syringe. The plate was then filled with serum-free medium (Figure 1a and Supplementary Materials and Methods). Out of the 86 cultured vibrissae collected from three
Journal of Investigative Dermatology (2008), Volume 128
pups, 81 vibrissae (94%) showed measurable shaft growth (Table 1 and Figure 1b). Two of the 81 vibrissae were lost, and 17 developed abnormal (kinked) fiber and were omitted from growth measurement. Of the remaining 62 vibrissae, straight shafts were produced by all at a rate of 0.3–0.5 mm/day for the first 3 days. While some follicles maintained this growth rate through the fifth day of culture (Figure 1b and c and data not shown), others began a gradual decline in growth rate (Figure 1b and c and 2a and data not shown). As reported previously (Jindo et al., 1993; Robinson et al., 1997), growth rates of all follicles slowed down considerably after 5 days in culture, indicating independence from culture conditions (Table S1). Hair shaft elongation rate decreases and eventually stops when follicles
J Lee et al. Anagen Reinitiation In Vitro
1 day
1 day
15 days
21 days
21 days
2 days 3 days 5 days
Length of hair shaft (mm)
3 2.5
8 days
2 1.5
13 days
1 0.5 0
15 days
5 10 15 Days after culture
Figure 1. Vibrissae produce hair shafts and reinitiate anagen in serum-free culture medium. (a) Schematic diagram of the in vitro vibrissa culture procedure. The culture media; Williams E medium (Invitrogen, Carlsbad, CA) supplemented with 2 mM L-glutamine, insulin (10 g/ml, Invitrogen), hydrocortisone (10 ng/ml, Sigma-Aldrich, St Louis, MO), penicillin (100 U/ml), and streptomycin (100 ug/ml, Invitrogen). In addition, tips of the harvested vibrissae were immobilized in silicone grease on culture dish. (b) Growth curve of the vibrissal shafts over time. Error bars represent SD. (c) Representative pictures of a vibrissal shaft growth cultured for 15 days. Pictures were taken on 1, 2, 3, 5, 8, 13, and 15 days after culture. Note that some vibrissae show growth retardation 3 days after culture, as shown in Figure 2a. Bar ¼ 1 mm. (d) Representative vibrissa grown in culture exhibit an emergence of new shaft (arrows) and bulb (open arrowhead). New shafts were clearly distinguished from the original shafts which become club hairs (arrowheads). Bar ¼ 300 mm.
Table 1. Summary of the in vitro vibrissa culture Intact collagen capsule
Without collagen capsule
3
5
Total no. of mice used Total no. of vibrissae cultured No. of vibrissae with shaft growth No. of vibrissae lost during culture No. of vibrissae with shaft kinks No. of vibrissae with a bud emergence No. of vibrissae reinitiating anagen1
86
10
81/86 (94.19%)
0/10 (0%)
2/81 (2.47%)
0/10 (0%)
17/79 (21.52%)
0/10 (0%)
0/79 (0%)
7/10 (70%)
79/79 (100%)
1/10 (10%)
1
Formation of new hair shaft and bulb was considered as anagen reinitiation.
enter the catagen/telogen phase in vivo (Alonso and Fuchs, 2006). Several days after cessation of hair shaft growth, cultured vibrissae follicles continuously changed their morphology, at times visible through their collagen capsules (Figure 1d). Strikingly, dissection and removal of collagen capsules from 21day-cultured vibrissae revealed that all vibrissae produced a second shaft (Figures 1d and 2a–h, arrows; Figure S1), readily distinguishable from the original shaft whose proximal end formed club-hair, indicative of a completed catagen (Figures 1 and 2 closed arrowheads). Histological and immunohistochemical analyses using AE13, a marker for cortex/cuticle-specific kera-
tins, confirmed the formation of new hair shafts (Figure 2c–f and Figure S1). The new shafts were produced in proximal bulbs that appeared smaller in size than the original ones before culture. Since the original hair growth stopped within the first 5 days of culture, and since a new shaft emerged from a secondary bulb, we concluded that vibrissa follicles from CD1 mice reinitiated the anagen phase by regenerating hair bulbs (Figure 2). Staining for alkaline phosphatase activity identified an intact (but small) dermal papilla within the new bulb (Figure 1d, open arrowhead; Figure 2g, alkaline phosphatase), as seen in the early anagen phase of vibrissae in vivo (Young and
Oliver, 1976; Oshima et al., 2001). Moreover, the hair matrix cells were positive for Ki-67, a proliferation marker, even after 13 days in culture (Figure 2h), indicating that the formation of a new shaft was the result of active proliferation and differentiation of matrix cells in cultured vibrissae. Similar cycling patterns were also observed with vibrissae from 14-day-old C57BL/6 mice (Figure S2), indicating that hair cycle in vitro is independent of mouse strains. To visualize better the hair cycle process, we next performed vibrissae culture after careful removal of their collagen capsules (Figure 2i–n, Supplementary Materials and Methods). Within 2 days, the proximal end of the vibrissa keratinized, forming a club hair, at that time when dermal papilla structures were no longer discernible (Figure 2j and data not shown). A day later, however, a bud-like structure emerged from the proximal part of the bulb in eight of the 10 cultured vibrissae. Of these, one vibrissa continued growing outwards, eventually forming a new bulb producing a vibrissa thinner than the club hair (Figure 2m and n). Strikingly, we observed that this vibrissa appeared to enter a third anagen, producing a new bud from www.jidonline.org
483
J Lee et al. Anagen Reinitiation In Vitro
c
0 day
d
i
l
j
m
k
n
1 day 2 days 3 days 5 days 7 days 9 days 11 days 13 days
b
e
f
g
h
Figure 2. Histological and immunohistochemical characteristics of new anagen follicles. (a) Representative pictures of a vibrissa cultured for 13 days. Pictures were taken on 0, 1, 2, 3, 5, 7, 9, 11, and 13 days after culture. (b) Magnified view of the 13-day cultured vibrissa shown in (a). Note the shape of bulb and dermal papilla. (c) Hematoxylin and eosin staining of the sections from (b). Note that both new hair shaft (arrow) and the club hair (arrowhead) are shown. (d) AE13 staining (green) on the adjacent section of (c). (e and f) Magnified views of the area marked by green and red box shown in (d). Note both club hair (e) and new hair shaft (f) are AE13-positive. (g) The presence of dermal papilla was confirmed by alkaline phosphatase substrate staining (purple) in adjacent section of the same vibrissa. (h) Ki-67 staining (green) on the adjacent section revealed the presence of proliferating matrix cells in the vibrissal bulb. (i–n) Pictures were taken on 1 (i), 2 (j), 3 (k), 7 (l), 14 (m), and 18 (n) days after culture of a vibrissa without collagen capsule. Inset in (n) is a magnified view. Note that a bud grew from the proximal part of the follicle to form a hair bulb that produces a new hair shaft (arrow in (m)) distinguished from a club hair (arrowheads in (l) and (m)). This follicle also shows a possible second bud formation (open arrowhead in (n)). Bars ¼ 1 mm (a); 300 mm (b and i–n); 50 mm (e–h). See serial sections of follicle (c) in Figure S1.
the second bulb (Figure 2n, open arrowhead). Collectively, the data presented here indicate that in our in vitro culture system, vibrissa follicles enter a second anagen phase and it may be possible for a few to enter a third cycle in culture. We are not certain which modification in our culture system permits vibrissae to cycle. We do not think that this is caused by differences in media composition, mouse strain, or frequency of media change. Our attempt to grow vibrissae at the air–fluid interface, as in Jindo et al. (1993), showed that one of the five vibrissae formed a new hair shaft (Figure S3, arrows), indicating that hair cycle is not strictly dependent on culture media or on the frequency of media change. However, we observed that on a solid substrate, most vibrissae gradually lost their hair matrix, possibly due to pressure caused by friction generated while extending a hair shaft on a solid surface or in the absence of sebum (Sundberg et al., 484
2000; Figure S3, arrowheads). Although not proven, we speculate that the critical modification was our method of immobilizing the hair shaft in silicon grease that provided a friction-free environment; anchoring the tip of shafts may have allowed gravity to facilitate the separation of club hairs from the hair matrix, and that supports new hair shaft formation. In addition, the smaller size of mouse vibrissae, in comparison with human or rat vibrissae, could also be a contributing factor in the anagen reinitiation in vitro. As this culture system is the first to support reinitiation of anagen in a culture dish, we hope this improvement will expedite the testing and development of hair cycle modifying pharmaceuticals.
Jonghyeob Lee1,2,4, Wei Wu3 and Raphael Kopan1,2 1 Department of Molecular Biology and Pharmacology, Washington University School of Medicine, Saint Louis, Missouri, USA; 2 Division of Dermatology, Department of Medicine, Washington University School of Medicine, Saint Louis, Missouri, USA and 3Synaptic Transmission Unit, National Institute of Neurological Disorders and Stroke, NINDS, National Institutes of Health, Bethesda, Maryland, USA. E-mail: [email protected] 4 Current Address: Department of Developmental Biology, Stanford University School of Medicine, 279 Campus Drive, Stanford, California 94305-5329, USA.
SUPPLEMENTARY MATERIAL Supplementary Text. Materials and Methods. Table S1. Comparison of rodent vibrissa hair cycle in different in vitro culture systems.
CONFLICT OF INTEREST The authors state no conflict of interest.
Figure S1. Histological analysis of serial sections from cultured vibrissa.
ACKNOWLEDGMENTS
Figure S2. Cultured vibrissae from C57BL/6 also exhibit hair cycling.
This work was supported by National Institutes of Health Grant no. RO1 GM55479 to RK.
Figure S3. Air–fluid interface culture of vibrissae in RPMI1640 culture medium.
Journal of Investigative Dermatology (2008), Volume 128
B Choi et al. Long-Term Evaluation of Microvascular Response
REFERENCES Alonso L, Fuchs E (2006) The hair cycle. J Cell Sci 119:391–3 Jindo T, Imai R, Takamori K, Ogawa H (1993) Organ culture of mouse vibrissal hair follicles in serum-free medium. J Dermatol 20:756–62 Nakamura M, Sundberg JP, Paus R (2001) Mutant laboratory mice with abnormalities in hair follicle morphogenesis, cycling, and/or structure: annotated tables. Exp Dermatol 10:369–90 Oshima H, Rochat A, Kedzia C, Kobayashi K, Barrandon Y (2001) Morphogenesis and
renewal of hair follicles from adult multipotent stem cells. Cell 104:233–45 Paus R, Cotsarelis G (1999) The biology of hair follicles. New Engl J Med 341:491–7 Philpott MP, Kealey T (2000) Cyclical changes in rat vibrissa follicles maintained in vitro. J Invest Dermatol 115:1152–5 Robinson M, Reynolds AJ, Jahoda CA (1997) Hair cycle stage of the mouse vibrissa follicle determines subsequent fiber growth and follicle behavior in vitro. J Invest Dermatol 108:495–500 Stenn KS, Paus R (2001) Controls of hair follicle cycling. Physiol Rev 81:449–94
Sundberg JP, Boggess D, Sundberg BA, Eilertsen K, Parimoo S, Filippi M et al. (2000) Asebia-2J (Scd1(ab2J)): a new allele and a model for scarring alopecia. Am J Pathol 156: 2067–75 Yano K, Brown LF, Detmar M (2001) Control of hair growth and follicle size by VEGFmediated angiogenesis. J Clin Invest 107: 409–17 Young RD, Oliver RF (1976) Morphological changes associated with the growth cycle of vibrissal follicles in the rat. J Embryol Exp Morphol 36:597–607
The Importance of Long-Term Monitoring to Evaluate the Microvascular Response to Light-Based Therapies Journal of Investigative Dermatology (2008) 128, 485–488; doi:10.1038/sj.jid.5700991; published online 2 August 2007
TO THE EDITOR Optimization of laser therapy for disfiguring vascular birthmarks is one specific clinical application (Kelly et al., 2005). Current treatment protocols involve the use of high-power pulsed laser irradiation with parameters chosen to induce selective photocoagulation of the targeted blood vessels, a method known as selective photothermolysis (Anderson and Parrish, 1983). Protocol design is based largely on results from numerical modeling studies (van Gemert et al., 1997), which are designed to predict the laser light distribution within the skin and subsequent photothermal response leading toward selective photocoagulation. However, current modeling methods do not incorporate adequately the complex dynamics associated with changes in light absorption due to conversion of hemoglobin to methemoglobin (Barton et al., 2001; Kimel et al., 2005) and convective mixing of blood during pulsed laser irradiation (Kimel et al., 2003), limiting their overall predictive capability. Furthermore, these models do not consider the chronic, biological response of the microvasculature to therapeutic laser intervention, which remains a poorly researched field. Knowledge of the biological response is critical to understand the repair processes initiated with photothermal injury and to assess the ultimate efficacy of the treatment.
Animal models used as a platform to study light-based, microvascular-targeted therapies include the chick chorioallantoic membrane (Kimel et al., 1994, 2003), hamster cheek pouch (Suthamjariya et al., 2004), and rodent dorsal window chamber (Barton et al., 1998, 1999, 2001; Choi et al., 2004; Babilas et al., 2005; Smith et al., 2006). Optical imaging modalities used to evaluate noninvasively therapeutic outcome include video imaging, fluorescence microscopy, Doppler optical coherence tomography, and laser speckle imaging. Typically, short-term (o24 hours after intervention) evaluation of the microvasculature is performed. Babilas et al. (2005) proposed that a (1) 24-hour monitoring period allows for evaluation of ‘‘delayed biological effects’’ in the microvascular response to pulsed laser irradiation, and (2) the short-term response correlates well with numerical modeling predictions of photocoagulation. Longer (424 hour after intervention) monitoring periods usually are not performed with nontumor-bearing window chambers, presumably due to the reduced clarity of the chamber imposed by poor maintenance of window integrity secondary to infection. However, we hypothesized that the short-term response of the microvasculature is a poor predictor of the long-term response. With emphasis on aseptic
methods, we have been able to maintain clear window chamber preparations for as long as 45 days after intervention. With this model, we have studied the long-term microvascular response to light-based, microvasculartargeted therapies. The data presented herein were acquired from adult male Golden Syrian hamsters. The surgery was performed as defined in a protocol approved by the University of California, Irvine, Animal Use Committee. The surgical protocol was a modified version of one described previously (Papenfuss et al., 1979). For all steps, aseptic conditions were maintained. We used wide-field color reflectance imaging and laser speckle imaging (Choi et al., 2004, 2006; Smith et al., 2006) to document and evaluate quantitatively and chronically ensuing blood flow dynamics. In one set of experiments, we irradiated select arteriolevenule pairs with laser pulse sequences to evaluate the efficacy of various therapeutic protocols. In the presented example (Figure 1a), we irradiated an arteriole-venule pair (upper circle in ‘‘Before’’ image) with five laser pulses containing both 532 and 1064 nm laser wavelengths and a second pair (lower circle) with a single 532/1064 nm laser pulse. Numerical modeling data suggested that both sets of laser parameters should induce photocoagulation in the www.jidonline.org
485
screen use by people who spent their holidays on the beach and who, by definition, were not sun avoiders. We found that more detailed sunscreen labeling resulted in subjects’ more effective use of sunscreens and fewer sunburns, indicating that the labeling affected the habits of sunscreen users toward their use of sunscreens; their sun exposure and use of protective clothing were unchanged. In support of their concern, Boniol et al. refer to their study (Autier et al., 2001) in which students who use highsun protection factor sunscreens tend to increase their sun exposure. We did not find such an effect in the adult population we studied. In fact, a randomized study (Dupuy et al., 2005) suggested that young adults could be considered a separate population due to their propensity toward sunbathing and other risky behaviors. We consider sun behavior to be guided by strong societal and psycho-
logical factors (Grob and Bonerandi, 1997) that are unlikely to be influenced by sunscreen labels. We believe that people who want to protect their skin by using sunscreens can benefit from more informative labeling, although Boniol et al. may regret that these people do not choose to wear protective clothing or avoid the sun entirely. A long-term campaign will be necessary to change predominant sun behaviors. The benefit of changing sunscreen labeling, although modest, could have immediate effects. CONFLICT OF INTEREST The authors state no conflict of interest.
Jean-Jacques Grob1 and Caroline Gaudy-Marqueste1 1
Department of Dermatology Hoˆpital Ste Marguerite, Assistance Publique des Hoˆpitaux de Marseille and Research Unit LIMP EA 3291, Universite´ de la Me´diterrane´e, Marseille, France. E-mail: [email protected]
REFERENCES Autier P, Boniol M, Severi G, Dore JF, European Organization for Research and Treatment of Cancer Melanoma Co-operative Group (2001) Quantity of sunscreen used by European students. Br J Dermatol 144: 288–91 Boniol M, Dore JF, Autier P (2008) Changing the labeling of sunscreen, will we transform sun avoiders into sunscreen users? J Invest Dermatol 128:482 Dupuy A, Dunant A, Grob JJ (2005) Reseau d’epidemiologie en dermatology. Randomized controlled trial testing the impact of high-protection sunscreens on sun-exposure behavior. Arch Dermatol 141:950–6 Grob JJ, Bonerandi JJ (1997) Attitudes and behaviour toward sun exposure: implications for melanoma prevention. In: Epidemiology, Causes and Prevention of Skin Diseases (Grob JJ, Stern R Mackie R, Weinstock M, eds), Oxford: Blackwell Science, 144–51 Nicol I, Gaudy C, Gouvernet J, Richard MA, Grob JJ (2007) Skin protection by sunscreens is improved by explicit labeling and providing free sunscreen. J Invest Dermatol 127: 41–8
Murine Vibrissae Cultured in Serum-Free Medium Reinitiate Anagen Journal of Investigative Dermatology (2008) 128, 482–485; doi:10.1038/sj.jid.5701024; published online 9 August 2007
TO THE EDITOR Aberrant regulation of the hair cycle has been implicated in human baldness (reviewed in Paus and Cotsarelis, 1999; Nakamura et al., 2001). Although several modulators of mammalian hair follicle cycle have been recently described, discovery of effective treatment for baldness has suffered from the absence of a reliable in vitro culture system in which the hair cycle can be assessed easily and inexpensively (Stenn and Paus, 2001). Several laboratories have established serum-free culture systems where vibrissa can grow and differentiate in vitro (Table S1; Jindo et al., 1993; Robinson et al., 1997; Yano et al., 2001). Although rat vibrissae cultured for 23 days were reported to share histologic similarities with catagen or pro-anagen (telogen) stage follicles (Philpott and Kealey, 482
2000), these follicles did not show any progress beyond the pro-anagen phase nor did they produce a new hair shaft (Philpott and Kealey, 2000). In this study, we demonstrate that a simple modification permits murine vibrissae in the current in vitro culture system to reinitiate anagen. This will accelerate the development of screening systems aimed at modifying the hair cycle. To establish a modified in vitro vibrissa culture system, anagen-stage vibrissae were carefully isolated from 14-day-old mice, and the tips of the vibrissa shafts were anchored in a stripe of sterilized silicone grease placed on the culture dish through a 3-ml syringe. The plate was then filled with serum-free medium (Figure 1a and Supplementary Materials and Methods). Out of the 86 cultured vibrissae collected from three
Journal of Investigative Dermatology (2008), Volume 128
pups, 81 vibrissae (94%) showed measurable shaft growth (Table 1 and Figure 1b). Two of the 81 vibrissae were lost, and 17 developed abnormal (kinked) fiber and were omitted from growth measurement. Of the remaining 62 vibrissae, straight shafts were produced by all at a rate of 0.3–0.5 mm/day for the first 3 days. While some follicles maintained this growth rate through the fifth day of culture (Figure 1b and c and data not shown), others began a gradual decline in growth rate (Figure 1b and c and 2a and data not shown). As reported previously (Jindo et al., 1993; Robinson et al., 1997), growth rates of all follicles slowed down considerably after 5 days in culture, indicating independence from culture conditions (Table S1). Hair shaft elongation rate decreases and eventually stops when follicles
J Lee et al. Anagen Reinitiation In Vitro
1 day
1 day
15 days
21 days
21 days
2 days 3 days 5 days
Length of hair shaft (mm)
3 2.5
8 days
2 1.5
13 days
1 0.5 0
15 days
5 10 15 Days after culture
Figure 1. Vibrissae produce hair shafts and reinitiate anagen in serum-free culture medium. (a) Schematic diagram of the in vitro vibrissa culture procedure. The culture media; Williams E medium (Invitrogen, Carlsbad, CA) supplemented with 2 mM L-glutamine, insulin (10 g/ml, Invitrogen), hydrocortisone (10 ng/ml, Sigma-Aldrich, St Louis, MO), penicillin (100 U/ml), and streptomycin (100 ug/ml, Invitrogen). In addition, tips of the harvested vibrissae were immobilized in silicone grease on culture dish. (b) Growth curve of the vibrissal shafts over time. Error bars represent SD. (c) Representative pictures of a vibrissal shaft growth cultured for 15 days. Pictures were taken on 1, 2, 3, 5, 8, 13, and 15 days after culture. Note that some vibrissae show growth retardation 3 days after culture, as shown in Figure 2a. Bar ¼ 1 mm. (d) Representative vibrissa grown in culture exhibit an emergence of new shaft (arrows) and bulb (open arrowhead). New shafts were clearly distinguished from the original shafts which become club hairs (arrowheads). Bar ¼ 300 mm.
Table 1. Summary of the in vitro vibrissa culture Intact collagen capsule
Without collagen capsule
3
5
Total no. of mice used Total no. of vibrissae cultured No. of vibrissae with shaft growth No. of vibrissae lost during culture No. of vibrissae with shaft kinks No. of vibrissae with a bud emergence No. of vibrissae reinitiating anagen1
86
10
81/86 (94.19%)
0/10 (0%)
2/81 (2.47%)
0/10 (0%)
17/79 (21.52%)
0/10 (0%)
0/79 (0%)
7/10 (70%)
79/79 (100%)
1/10 (10%)
1
Formation of new hair shaft and bulb was considered as anagen reinitiation.
enter the catagen/telogen phase in vivo (Alonso and Fuchs, 2006). Several days after cessation of hair shaft growth, cultured vibrissae follicles continuously changed their morphology, at times visible through their collagen capsules (Figure 1d). Strikingly, dissection and removal of collagen capsules from 21day-cultured vibrissae revealed that all vibrissae produced a second shaft (Figures 1d and 2a–h, arrows; Figure S1), readily distinguishable from the original shaft whose proximal end formed club-hair, indicative of a completed catagen (Figures 1 and 2 closed arrowheads). Histological and immunohistochemical analyses using AE13, a marker for cortex/cuticle-specific kera-
tins, confirmed the formation of new hair shafts (Figure 2c–f and Figure S1). The new shafts were produced in proximal bulbs that appeared smaller in size than the original ones before culture. Since the original hair growth stopped within the first 5 days of culture, and since a new shaft emerged from a secondary bulb, we concluded that vibrissa follicles from CD1 mice reinitiated the anagen phase by regenerating hair bulbs (Figure 2). Staining for alkaline phosphatase activity identified an intact (but small) dermal papilla within the new bulb (Figure 1d, open arrowhead; Figure 2g, alkaline phosphatase), as seen in the early anagen phase of vibrissae in vivo (Young and
Oliver, 1976; Oshima et al., 2001). Moreover, the hair matrix cells were positive for Ki-67, a proliferation marker, even after 13 days in culture (Figure 2h), indicating that the formation of a new shaft was the result of active proliferation and differentiation of matrix cells in cultured vibrissae. Similar cycling patterns were also observed with vibrissae from 14-day-old C57BL/6 mice (Figure S2), indicating that hair cycle in vitro is independent of mouse strains. To visualize better the hair cycle process, we next performed vibrissae culture after careful removal of their collagen capsules (Figure 2i–n, Supplementary Materials and Methods). Within 2 days, the proximal end of the vibrissa keratinized, forming a club hair, at that time when dermal papilla structures were no longer discernible (Figure 2j and data not shown). A day later, however, a bud-like structure emerged from the proximal part of the bulb in eight of the 10 cultured vibrissae. Of these, one vibrissa continued growing outwards, eventually forming a new bulb producing a vibrissa thinner than the club hair (Figure 2m and n). Strikingly, we observed that this vibrissa appeared to enter a third anagen, producing a new bud from www.jidonline.org
483
J Lee et al. Anagen Reinitiation In Vitro
c
0 day
d
i
l
j
m
k
n
1 day 2 days 3 days 5 days 7 days 9 days 11 days 13 days
b
e
f
g
h
Figure 2. Histological and immunohistochemical characteristics of new anagen follicles. (a) Representative pictures of a vibrissa cultured for 13 days. Pictures were taken on 0, 1, 2, 3, 5, 7, 9, 11, and 13 days after culture. (b) Magnified view of the 13-day cultured vibrissa shown in (a). Note the shape of bulb and dermal papilla. (c) Hematoxylin and eosin staining of the sections from (b). Note that both new hair shaft (arrow) and the club hair (arrowhead) are shown. (d) AE13 staining (green) on the adjacent section of (c). (e and f) Magnified views of the area marked by green and red box shown in (d). Note both club hair (e) and new hair shaft (f) are AE13-positive. (g) The presence of dermal papilla was confirmed by alkaline phosphatase substrate staining (purple) in adjacent section of the same vibrissa. (h) Ki-67 staining (green) on the adjacent section revealed the presence of proliferating matrix cells in the vibrissal bulb. (i–n) Pictures were taken on 1 (i), 2 (j), 3 (k), 7 (l), 14 (m), and 18 (n) days after culture of a vibrissa without collagen capsule. Inset in (n) is a magnified view. Note that a bud grew from the proximal part of the follicle to form a hair bulb that produces a new hair shaft (arrow in (m)) distinguished from a club hair (arrowheads in (l) and (m)). This follicle also shows a possible second bud formation (open arrowhead in (n)). Bars ¼ 1 mm (a); 300 mm (b and i–n); 50 mm (e–h). See serial sections of follicle (c) in Figure S1.
the second bulb (Figure 2n, open arrowhead). Collectively, the data presented here indicate that in our in vitro culture system, vibrissa follicles enter a second anagen phase and it may be possible for a few to enter a third cycle in culture. We are not certain which modification in our culture system permits vibrissae to cycle. We do not think that this is caused by differences in media composition, mouse strain, or frequency of media change. Our attempt to grow vibrissae at the air–fluid interface, as in Jindo et al. (1993), showed that one of the five vibrissae formed a new hair shaft (Figure S3, arrows), indicating that hair cycle is not strictly dependent on culture media or on the frequency of media change. However, we observed that on a solid substrate, most vibrissae gradually lost their hair matrix, possibly due to pressure caused by friction generated while extending a hair shaft on a solid surface or in the absence of sebum (Sundberg et al., 484
2000; Figure S3, arrowheads). Although not proven, we speculate that the critical modification was our method of immobilizing the hair shaft in silicon grease that provided a friction-free environment; anchoring the tip of shafts may have allowed gravity to facilitate the separation of club hairs from the hair matrix, and that supports new hair shaft formation. In addition, the smaller size of mouse vibrissae, in comparison with human or rat vibrissae, could also be a contributing factor in the anagen reinitiation in vitro. As this culture system is the first to support reinitiation of anagen in a culture dish, we hope this improvement will expedite the testing and development of hair cycle modifying pharmaceuticals.
Jonghyeob Lee1,2,4, Wei Wu3 and Raphael Kopan1,2 1 Department of Molecular Biology and Pharmacology, Washington University School of Medicine, Saint Louis, Missouri, USA; 2 Division of Dermatology, Department of Medicine, Washington University School of Medicine, Saint Louis, Missouri, USA and 3Synaptic Transmission Unit, National Institute of Neurological Disorders and Stroke, NINDS, National Institutes of Health, Bethesda, Maryland, USA. E-mail: [email protected] 4 Current Address: Department of Developmental Biology, Stanford University School of Medicine, 279 Campus Drive, Stanford, California 94305-5329, USA.
SUPPLEMENTARY MATERIAL Supplementary Text. Materials and Methods. Table S1. Comparison of rodent vibrissa hair cycle in different in vitro culture systems.
CONFLICT OF INTEREST The authors state no conflict of interest.
Figure S1. Histological analysis of serial sections from cultured vibrissa.
ACKNOWLEDGMENTS
Figure S2. Cultured vibrissae from C57BL/6 also exhibit hair cycling.
This work was supported by National Institutes of Health Grant no. RO1 GM55479 to RK.
Figure S3. Air–fluid interface culture of vibrissae in RPMI1640 culture medium.
Journal of Investigative Dermatology (2008), Volume 128
B Choi et al. Long-Term Evaluation of Microvascular Response
REFERENCES Alonso L, Fuchs E (2006) The hair cycle. J Cell Sci 119:391–3 Jindo T, Imai R, Takamori K, Ogawa H (1993) Organ culture of mouse vibrissal hair follicles in serum-free medium. J Dermatol 20:756–62 Nakamura M, Sundberg JP, Paus R (2001) Mutant laboratory mice with abnormalities in hair follicle morphogenesis, cycling, and/or structure: annotated tables. Exp Dermatol 10:369–90 Oshima H, Rochat A, Kedzia C, Kobayashi K, Barrandon Y (2001) Morphogenesis and
renewal of hair follicles from adult multipotent stem cells. Cell 104:233–45 Paus R, Cotsarelis G (1999) The biology of hair follicles. New Engl J Med 341:491–7 Philpott MP, Kealey T (2000) Cyclical changes in rat vibrissa follicles maintained in vitro. J Invest Dermatol 115:1152–5 Robinson M, Reynolds AJ, Jahoda CA (1997) Hair cycle stage of the mouse vibrissa follicle determines subsequent fiber growth and follicle behavior in vitro. J Invest Dermatol 108:495–500 Stenn KS, Paus R (2001) Controls of hair follicle cycling. Physiol Rev 81:449–94
Sundberg JP, Boggess D, Sundberg BA, Eilertsen K, Parimoo S, Filippi M et al. (2000) Asebia-2J (Scd1(ab2J)): a new allele and a model for scarring alopecia. Am J Pathol 156: 2067–75 Yano K, Brown LF, Detmar M (2001) Control of hair growth and follicle size by VEGFmediated angiogenesis. J Clin Invest 107: 409–17 Young RD, Oliver RF (1976) Morphological changes associated with the growth cycle of vibrissal follicles in the rat. J Embryol Exp Morphol 36:597–607
The Importance of Long-Term Monitoring to Evaluate the Microvascular Response to Light-Based Therapies Journal of Investigative Dermatology (2008) 128, 485–488; doi:10.1038/sj.jid.5700991; published online 2 August 2007
TO THE EDITOR Optimization of laser therapy for disfiguring vascular birthmarks is one specific clinical application (Kelly et al., 2005). Current treatment protocols involve the use of high-power pulsed laser irradiation with parameters chosen to induce selective photocoagulation of the targeted blood vessels, a method known as selective photothermolysis (Anderson and Parrish, 1983). Protocol design is based largely on results from numerical modeling studies (van Gemert et al., 1997), which are designed to predict the laser light distribution within the skin and subsequent photothermal response leading toward selective photocoagulation. However, current modeling methods do not incorporate adequately the complex dynamics associated with changes in light absorption due to conversion of hemoglobin to methemoglobin (Barton et al., 2001; Kimel et al., 2005) and convective mixing of blood during pulsed laser irradiation (Kimel et al., 2003), limiting their overall predictive capability. Furthermore, these models do not consider the chronic, biological response of the microvasculature to therapeutic laser intervention, which remains a poorly researched field. Knowledge of the biological response is critical to understand the repair processes initiated with photothermal injury and to assess the ultimate efficacy of the treatment.
Animal models used as a platform to study light-based, microvascular-targeted therapies include the chick chorioallantoic membrane (Kimel et al., 1994, 2003), hamster cheek pouch (Suthamjariya et al., 2004), and rodent dorsal window chamber (Barton et al., 1998, 1999, 2001; Choi et al., 2004; Babilas et al., 2005; Smith et al., 2006). Optical imaging modalities used to evaluate noninvasively therapeutic outcome include video imaging, fluorescence microscopy, Doppler optical coherence tomography, and laser speckle imaging. Typically, short-term (o24 hours after intervention) evaluation of the microvasculature is performed. Babilas et al. (2005) proposed that a (1) 24-hour monitoring period allows for evaluation of ‘‘delayed biological effects’’ in the microvascular response to pulsed laser irradiation, and (2) the short-term response correlates well with numerical modeling predictions of photocoagulation. Longer (424 hour after intervention) monitoring periods usually are not performed with nontumor-bearing window chambers, presumably due to the reduced clarity of the chamber imposed by poor maintenance of window integrity secondary to infection. However, we hypothesized that the short-term response of the microvasculature is a poor predictor of the long-term response. With emphasis on aseptic
methods, we have been able to maintain clear window chamber preparations for as long as 45 days after intervention. With this model, we have studied the long-term microvascular response to light-based, microvasculartargeted therapies. The data presented herein were acquired from adult male Golden Syrian hamsters. The surgery was performed as defined in a protocol approved by the University of California, Irvine, Animal Use Committee. The surgical protocol was a modified version of one described previously (Papenfuss et al., 1979). For all steps, aseptic conditions were maintained. We used wide-field color reflectance imaging and laser speckle imaging (Choi et al., 2004, 2006; Smith et al., 2006) to document and evaluate quantitatively and chronically ensuing blood flow dynamics. In one set of experiments, we irradiated select arteriolevenule pairs with laser pulse sequences to evaluate the efficacy of various therapeutic protocols. In the presented example (Figure 1a), we irradiated an arteriole-venule pair (upper circle in ‘‘Before’’ image) with five laser pulses containing both 532 and 1064 nm laser wavelengths and a second pair (lower circle) with a single 532/1064 nm laser pulse. Numerical modeling data suggested that both sets of laser parameters should induce photocoagulation in the www.jidonline.org
485