ORIGINAL ARTICLE
Profiling the Response of Human Hair Follicles to Ultraviolet Radiation Zhongfa Lu1,2,5, Tobias W. Fischer1,5, Sybille Hasse1, Koji Sugawara1, York Kamenisch3, Sven Krengel1, Wolfgang Funk4, Mark Berneburg3 and Ralf Paus1 Excessive UVR ranks among the most harmful environmental influences on human skin. However, the direct impact of UVR on human skin appendages remains to be systematically investigated. Organ-cultured human anagen hair follicles in vitro were irradiated, and reduction of hair shaft elongation, premature catagen entry, and reduced hair matrix keratinocyte proliferation were observed upon irradiation with UVB (20/50 mJ cm2). At 20 mJ cm2, apoptotic cell death prevailed (casp-3/p53 activation), whereas at 50 mJ cm2, necrotic cell death was predominant (lactate dehydrogenase increase). Mitochondrial common deletion and oxidatively damaged genomic DNA (8-OH-dG) was mainly observed at 20 mJ cm2. Follicular melanogenesis and ACTH immunoreactivity drastically declined, but a-melanocyte-stimulating hormone remained unchanged, whereas transforming growth factor-b2 expression shifted from the outer toward the inner root sheath. Both the number of Giemsa þ mast cells and the degree of mast-cell degranulation increased in the connective tissue sheath (CTS), and CD117 immunoreactivity of CTS cells and matrix keratinocytes was upregulated. Thus, UVR differentially modifies hair growth and cycle, promotes cell death, and induces complex regulatory events in human hair follicles in vitro. The leads from this human organ model, which is a living and human tissue interaction system under physiologically relevant in situ conditions, may encourage its use for general investigation of UV-induced effects as well as for testing possible agents for their UV-protective potency. Journal of Investigative Dermatology (2009) 129, 1790–1804; doi:10.1038/jid.2008.418; published online 22 January 2009
INTRODUCTION UVR exerts beneficial effects in human skin (for example, it is essential for dihydroxycholecalciferol synthesis; Lehmann et al., 2004) and stimulates the intracutaneous synthesis of important bioregulators such as melanocortins (for example, melanocyte-stimulating hormone (MSH) and ACTH (Luger et al., 1999; Slominski, 2005)), corticotropin-releasing hormone (Slominski et al., 1996a, 2006), and melatonin (Slominski et al., 2005a; Fischer et al., 2006a), but it is also considered one of the most important damaging environmental factors to the skin (Cleaver and Crowley, 2002; 1
Department of Dermatology, University Hospital Schleswig-Holstein, University of Lu¨beck, Lu¨beck, Germany; 2Department of Dermatology, 2nd Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, China; 3Molecular Oncology and Aging, Department of Dermatology, Eberhard Karls University, Tuebingen, Germany and 4Klinik Dr. Koslowski, Munich, Germany
5
These authors contributed equally to the published work
Correspondence: Dr Tobias W. Fischer, Department of Dermatology, Allergology and Venerology, University Hospital of Schleswig-Holstein, University of Lu¨beck, Ratzeburger Allee 160, 23538 Lu¨beck, Germany. E-mail:
[email protected] Abbreviations: 8-OH-dG, 8-hydroxy-deguanosin; casp-3, caspase-3; CTS, connective tissue sheath; LDH, lactate dehydrogenase; MED, minimal erythema dose; MSH, melanocyte-stimulating hormone; ORS, outer root sheath; TGFb2, transforming growth factor-b2 Received 21 December 2006; revised 4 September 2008; accepted 30 September 2008; published online 22 January 2009
1790 Journal of Investigative Dermatology (2009), Volume 129
Collins et al., 2004; de Vries et al., 2004). UVR inhibits basal epidermal keratinocyte proliferation (Yaar and Gilchrest, 2001), induces inflammatory changes (Averbeck et al., 2006), oxidative stress (Sander et al., 2004), and keratinocyte apoptosis (Fischer et al., 2006b), and promotes both skin aging and carcinogenesis (Hadshiew et al., 2000; Krutmann, 2003; Sander et al., 2006; Berneburg et al., 2006a). Most research on the effects of UVR on mammalian skin has focused on epidermis and dermis as well as on the immune system (Pisarchik et al., 2004; Schwarz, 2005; Averbeck et al., 2006; Claerhout et al., 2006; McCullough and Kelly, 2006; Nishigori, 2006; Rabe et al., 2006; Sander et al., 2006; Berneburg et al., 2006a), whereas skin appendages have not been investigated as targets of UVR. Although the damaging effect of UVR on the extracutaneously located hair shaft is well recognized (Nogueira et al., 2004), the impact of UVR on the hair shaft production unit (that is, proliferating hair matrix keratinocytes, anagen dermal papilla, melanocytes of the hair follicle pigmentary unit) is not yet known. So far, it has only been shown that UV-induced chronic skin inflammation may exert indirect damaging effects on hair growth (for example, on the balding, sun-burned scalp; Camacho et al., 1996). In mice, UVB irradiation can induce apoptosis in the upper hair follicle (Muller-Rover et al., 2000), and reportedly can exert either stimulatory or inhibitory effects in vivo, depending on the wavelength & 2009 The Society for Investigative Dermatology
Z Lu et al. UV Response of Hair Follicles
administered (Johnson, 1960; Johnson and Daniels, 1969). However, essentially nothing is known about the direct effects of UVR on the human hair follicle production unit, that is, the anagen hair bulb. Apart from their general biological significance, these effects may also be clinically relevant in the context of UV-induced aging of skin appendages and phototherapy of alopecia areata (Claudy and Gagnaire, 1983; Mohamed et al., 2005). As a first step toward closing this gap of knowledge, we have exploited the well-established organ-culture model using microdissected human scalp hair follicles (pioneered by Philpott et al., 1990) to study the effects of direct exposure of human anagen hair follicles to UVB irradiation (280 nm) at the UV doses of 20 and 50 mJ cm2 which correspond to the UV-dose range of the minimal erythema dose (MED) of a white skin phototype II/III (Fitzpatrick classification; Fitzpatrick, 1988). It also represents a widely accepted UVB-dose range used in in vitro models (Schwarz et al., 2002; Sitailo et al., 2002; Takasawa and Tanuma, 2003; Lee et al., 2004; Fischer et al., 2006b). In this initial, exploratory study we aimed to investigate whether repetitive, moderately dosed, monochromatic UVB irradiation would exert any effect on (1) hair shaft elongation, (2) hair follicle cycling, (3) proliferation/apoptosis of hair matrix keratinocytes, (4) mitochondrial and genomic DNA, and (5) melanogenesis of the follicular pigmentary unit. Given the central importance of transforming growth factorb2 (TGFb2) as a catagen inducer in human hair follicles (Soma et al., 2002; Hibino and Nishiyama, 2004; Foitzik et al., 2005), we also evaluated whether follicular TGFb2 immunoreactivity would be modulated. In view of the stimulatory effects of UVB irradiation on melanocortin expression in human skin and on human cutaneous cell populations (Chakraborty et al., 1999; Paus et al., 1999; Bohm and Luger, 2004), we also aimed to assess the expression of ACTH and a-MSH protein in hair follicles under UVB exposure. Finally, UV-induced urticaria (Tuchinda et al., 2005) highlights that human skin mast cells may also be important targets of UVR (Guhl et al., 2005; Maurer and Metz, 2005; Grimbaldeston et al., 2006). If UVR affects mast-cell function, this would be another mechanism of UV-induced impairment of hair follicle integrity. Mast cells are abundantly found in the connective tissue sheath (CTS) of the hair follicle, represent potent hair cycle modulators in mice and induce catagen keratinocyte proliferation in murine skin organ culture via their main secretory products, histamine and serotonin (Paus et al., 1994; Maurer et al., 1997). We therefore carefully analyzed the expression of masts cells under the influence of UV exposure and evaluated the possible impact of UVR on mast-cell degranulation.
histomorphometry, and revealed that 91.7% (33/36) of the examined, organ-cultured hair follicles prematurely transformed into hair follicles with characteristic catagen-like morphology after irradiation by UVB at 50 mJ cm2, compared with 72.2% (26/36) in the group of 20 mJ cm2 UVB exposure and 16.7% in control hair follicles (6/36; Po0.01; Figure 1b). Hair shaft elongation of 72 human scalp hair follicles from three different individuals was measured over 9 days. The lower UVB-dose (20 mJ cm2) led only to a slight reduction after day 5, with no statistical significance (Figure 1c). At 50 mJ cm2, however, hair shaft elongation was significantly inhibited compared to sham-irradiated control follicles as early as day 5 (Po0.05), with increasing levels of significance reached at later time points (Po0.001; Figure 1d). DNA fragmentation in human hair follicles increases significantly in a linear dose-dependent manner, whereas proliferation rapidly declines after low UVB irradiation and cytotoxicity occurs only at higher UVB doses
Lactate dehydrogenase (LDH) release into the culture medium was assessed as an indicator of cytotoxicity (Vandana et al., 2006). LDH levels increased over time in UV-irradiated as well as in sham-irradiated hair follicles; however, at the lower UVB dose of 20 mJ cm2, no difference was observed between the UVB group and control (data not shown). In contrast, LDH-activity under irradiation with 50 mJ cm2 significantly increased in the supernatant of UVBirradiated follicles as early as day 5 (Po0.05) with increasing significance at later time points (day 7 and day 9: Po0.001; Figure 1e). Next, 72 hair follicle sections (24 hair follicles on days 1, 3, and 5, respectively) derived from three different individuals were analyzed for apoptosis (TUNEL) and proliferation (Ki67) in the hair matrix area after 9 days in culture and UVB irradiation on days 1, 3, and 5. In a defined reference area (displayed in Figure 2a–i) the number of TUNEL/Ki67 þ cells was counted and normalized to the number of DAPI-stained cells (equals 100%). In nonirradiated follicles (Figure 2a–c), Ki67 positivity was shown in 22% of hair matrix keratinocytes (Figure 2c and j), as opposed to only 1.36 and 0.76% of hair matrix keratinocytes in hair follicles irradiated with 20 mJ cm2 (Figure 2f and j) and 50 mJ cm2 (Figure 2i and j), respectively (Po0.01). Quantification of apoptotic cells by the TUNEL technique showed a strong UVB dose-dependent increase of 27 and 57% TUNEL þ cells per defined reference area after UVB irradiation with 20 (Figure 2e and j) and 50 mJ cm2 (Figure 2h and j), respectively, as opposed to only 2.5% TUNEL þ cells in nonirradiated follicles (Po0.01; Figure 2b and j).
RESULTS UVB-irradiated human anagen hair follicles show reduced hair shaft elongation and premature catagen entry
Specific pathways of apoptosis are induced only at lower UVB doses
Compared to nonirradiated hair follicles, human hair follicles that were irradiated with 50 mJ cm2 UVB at 280 nm prematurely entered catagen as early as day 4 in culture (Figure 1a). Hair follicle cycling was assessed by quantitative
Two independent apoptotic pathways in human hair follicles exposed to UVB have been studied. Immunoreactivity of caspase-3 (casp-3), which is a major effector caspase of both the casp-9-related mitochondrial (intrinsic) pathway www.jidonline.org 1791
Z Lu et al. UV Response of Hair Follicles
b 120 *
Percentages
100
1
4 7 Days cultured
9
c 140
*
80 Anagen Catagen
60 40 20 Control 20 mJ 50 mJ cm–2 cm–2
90 80
Control
#
UVB (50 mJ cm–2)
60 40
0
e 0.9
*
50 40 30 20
1
3 5 Days in culture
Control UVB (50 mJ cm–2)
0.8 0.7 Absorbance
Elongation (%)
80
0
#
70 60
100
20
0
d
Control UVB (20 mJ cm– 2)
120 Elongation (%)
a
7
9
# #
0.6
*
0.5 0.4 0.3 0.2 0.1
10 0
0 0
1
3 5 Days in culture
7
9
1
3
5 Day
7
9
Figure 1. Increased catagen induction and LDH release, and reduced hair shaft elongation with increasing doses of UVB irradiation. Hair follicles were irradiated with either 20 or 50 mJ cm2 over a period of 9 days. Catagen development (a and b) and hair shaft elongation reduction (c and d) of 72 hair follicles was measured and the results are summarized graphically. After day 5 in culture, there was a linear and significant increase of catagen induction (*Po0.05) and hair shafts grew significantly less compared with control (*Po0.05; #Po0.001) when irradiated with 50 mJ cm2 (d), whereas there was only a minor, nonsignificant reduction in hair shaft elongation after 20 mJ cm2 UVB (c). There was a strong and significant increase of LDH in the supernatant of follicles irradiated with 50 mJ cm2 as early as day 5 (*Po0.05) with increasing significance at later time points (#Po0.001) (e). Arrowheads mark the days of UVB exposure (days 1, 3, and 5).
and the receptor-dependent casp-8-related (extrinsic) apoptotic pathway, was activated in matrix keratinocytes on irradiation with 20 mJ cm2 (Figure 3b) compared to nonirradiated control follicles (Figure 3a). At 50 mJ cm2, there was no protein expression of casp-3 in matrix keratinocytes, but a weak expression in the inner root sheath (Figure 3c). The caspase-independent key regulator of apoptosis in cellular stress-related events, namely p53, was shown to be distinctively activated after irradiation with 20 mJ cm2 in matrix keratinocytes, but also in cells along the Henle’s layer of the inner root sheath compared to nonirridiated controls (Figure 4a and b). In hair follicles exposed to 50 mJ cm2 (Figure 4c), immunoreactivity of p53 in matrix keratinocytes was clearly less expressed compared to 20 mJ cm2, and p53 immunoreactivity expression in the Henle’s layer of the inner root sheath cells was slightly weaker. Mitochondrial common deletion is inducible by low doses of UVB
Irradiation of human hair follicles with a UV-dose of 20 mJ cm2 leads to the induction of the 4,977 bp common deletion of mitochondrial (mt) DNA. (Figure 5). 1792 Journal of Investigative Dermatology (2009), Volume 129
Oxidative damage of genomic DNA is induced by low UVBdoses
Oxidative DNA damage, as assessed by immunoreactivity of 8-hydroxy-deguanosin (8-OH-dG), was distinctively upregulated in hair follicles that had been irradiated with 20 mJ cm2 (Figure 6d and e) in the whole ‘‘living’’ bulb, that is, matrix keratinocytes around the dermal papilla (lower parts, lateral and above the convexity of the dermal papilla) and keratinocytes of the inner and outer root sheath compared to sham-irradiated hair follicles (Figure 6a and b). At 50 mJ cm2 (Figure 6g and h), there was a slight expression of 8-OH-dG, but considerably lower than in hair follicles that had been irradiated with 20 mJ cm2. UVB irradiation reduces melanin synthesis and disrupts the hair follicle pigmentary unit
As disruption of follicular melanogenesis is an early and highly sensitive indicator of hair follicle damage (Slominski et al., 1994, 1996b; Tobin et al., 1998; Hendrix et al., 2005), quantitative Masson–Fontana histochemistry was employed for melanin demarcation. UVR at 20 mJ cm2 (Figure 7b) and 50 mJ cm2 (Figure 7d) caused a significant decline of melanin content compared with controls (Po0.01; Figure 7a,
Z Lu et al. UV Response of Hair Follicles
DAPI
TUNEL
a
Ki67
b
c
ORS
Control
IRS CC
MK
MK
DP
d
e
20 mJ cm–2
f
ORS
ORS
ORS MK
MK
MK
g
h
50 mJ cm–2
i ORS
ORS
ORS MK
MK
j
Percentage of positive cells
CTS
80 70 60 50
CTS
P
