Control of Human Hair Growth by Neurotrophins: Brain-Derived Neurotrophic Factor Inhibits Hair Shaft Elongation, Induces Catagen, and Stimulates Follicular Transforming Growth Factor β2 Expression

Control of Human Hair Growth by Neurotrophins: Brain-Derived Neurotrophic Factor Inhibits Hair Shaft Elongation, Induces Catagen, and Stimulates Follicular Transforming Growth Factor b2 Expression Eva M. J. Peters, Marit G. Hansen,w Rupert W. Overall,w Motonobu Nakamura,wz Paolo Pertile,y Burghard F. Klapp, Petra C. Arck, and Ralf Pausw

Department of Internal Medicine, Biomedical Research Center, University Medicine Charite´, Campus Virchow Hospital, Berlin, Germany; wDepartment of Dermatology, University Hospital Hamburg-Eppendorf, Hamburg, Germany; zDepartment of Dermatology, Kyoto University, Kyoto, Japan; yCutech Srl, Padova, Italy

Neurotrophins are important modulators of epithelial–mesenchymal interactions. Previously, we had shown that brain-derived neurotrophic factor (BDNF) and its high-affinity receptor tyrosine kinase B (TrkB) are prominently involved in the control of murine hair follicle cycling. We now show that BDNF and TrkB are also expressed in the human hair follicle in a manner that is both hair cycle dependent and suggestive of epithelial–mesenchymal crosstalk between BDNF-secreting dermal papilla fibroblasts of anagen hair follicles and subpopulations of TrkB þ hair follicle keratinocytes. As functional evidence for an involvement of BDNF/TrkB in human hair growth control, we show in organ-cultured human anagen hair follicles that 50 ng per mL BDNF significantly inhibit hair shaft elongation, induce premature catagen development, and inhibit keratinocyte proliferation. Quantitative real-time rtPCR analysis demonstrates upregulation of the potent catagen inducer, transforming growth factor b2 (TGFb2) by BDNF, whereas catagen induction by BDNF was partially reversible through co-administration of TGFb-neutralizing antibody. This suggests that TrkB-mediated signaling promotes the switch between anagen and catagen at least in part via upregulation of TGFb2. Thus, human scalp hair follicles are both a source and target of bioregulation by BDNF, which invites to target TrkB-mediated signaling for therapeutic hair growth modulation.

Key words: brain-derived neurotrophic factor/hair cycle/human/neurotrophin/tumor growth factor b2/tyrosinekinase B J Invest Dermatol 124:675 – 685, 2005

Since its discovery more than 20 y ago (Barde et al, 1982), the neurotrophin brain-derived neurotrophic factor (BDNF), the second member of the nerve growth factor family of growth factors, has become known to guide and sustain sensory and autonomic neuron development and their differentiation within peripheral neural networks (Barde et al, 1982; Botchkarev et al, 1998b; Ernfors, 2001; Metsis, 2001; Farinas et al, 2002). BDNF is produced by peripheral innervation targets such as visceral epithelia (Lommatzsch et al, 1999) and skin including the murine and rat hair follicle (Botchkarev et al, 1999a, b; Bergman et al, 2000; Peters et al, 2003) and cutaneous nerve fibers express the highaffinity receptor for BDNF tyrosine kinase B (TrkB) in mice (Botchkarev et al, 1999a) and humans (Peters et al, 2003). As a result of more recent research efforts, neurotrophins such as BDNF have also been shown to hold growth- and immunmodulatory capacities far beyond their sole engage-

ment as neurotrophins that was previously assigned to them. They evolve as potent neuro-immunmodulators, nonneuronal growth factors, and even as stress mediators in various tissues, transgressing their role in the regulation of nerve damage repair and peripheral neuronal plasticity (Alleva and Santucci, 2001; Renz, 2001; Marshall and Born, 2002; Bayas et al, 2003). TrkB for example is expressed by dendritic cells and macrophages in lymphoid organs (Hannestad et al, 1998) and BDNF is expressed by T cells (Barouch et al, 2000; Moalem et al, 2000). BDNF is also required for the sustenance of corneal epithelial cells (You et al, 2000) and the survival and proliferation of pigment cells during neural crest development and in the retina (Sieber-Blum, 1998; dos Santos et al, 2003), and may also play a role in bone formation (Asaumi et al, 2000). As a very densely innervated mini-organ, the hair follicle is subject to life-long recurring tissue-remodeling events (hair cycle) with alternating phases of rapid growth and pigmentation (anagen), apoptosis-driven regression (catagen), and relative quiescence (telogen). These tissue remodeling events during hair follicle cycling involve both the organ itself (Stenn and Paus, 2001), its innervation (Botchkarev et al, 1997, 1998b, 1999c; Peters et al, 2001), its

Abbreviations: BDNF, brain-derived neurotrophic factor; IR, immunoreactivity; rhBDNF, recombinant human BDNF; TGFb2, transforming growth factor b2; TrkB, tyrosine kinase B; TUNEL, terminal deoxynucleotidyl transferase (TdT)-mediated dUTP-biotin Nick end labeling

Copyright r 2005 by The Society for Investigative Dermatology, Inc.

675

676 PETERS ET AL

vasculature (Mecklenburg et al, 2000), and profound changes of the skin architecture and immune function (Paus et al, 1999a). BDNF expression in the murine hair follicle fluctuates in a strikingly hair cycle-dependent manner and BDNF overexpression causes premature catagen development (Botchkarev et al, 1999a) and alters cutaneous innervation density (Botchkarev et al, 1998b). Since nervous system and hair follicle epithelium share a common ectodermal origin, it is not surprising that TrkB is also expressed by cutaneous epithelial cells in mice and humans (Shibayama and Koizumi, 1996; Botchkarev et al, 1999a, b). In regressing hair follicles of C57BL/6 mice, TrkB is expressed on apoptotic keratinocytes during catagen and BDNF skin content is highest during anagen–catagen transformation (Botchkarev et al, 1999a). Simultaneous expression of BDNF by these cells suggests para- and autocrine growth-regulatory effects of BDNF on epithelial tissues (Shibayama and Koizumi, 1996; Botchkarev et al, 1999a). Although epidermal keratinocyte proliferation is stimulated by BDNF (Botchkarev et al, 1999b), a growth-inhibitory effect of BDNF on selected keratinocyte subpopulations was demonstrated with the report that BDNF induces premature catagen development in anagen hair follicles of C57BL/6 mice (Botchkarev et al, 1999a). Given the very substantial inter-species differences that a given bioregulatory compound can exert, e.g. on murine versus human hair follicles (cf. Paus and Cotsarelis, 1999; Stenn and Paus, 2001) and the contrasting effects of BDNF on epidermal (stimulation of proliferation) (Botchkarev et al, 1999b) and hair follicle keratinocytes (promotion of apoptosis) (Botchkarev et al, 1999a) in mice, it remains totally unknown whether BDNF/TrkB signaling plays any role in human hair biology. With these limitations in mind, we have addressed the following specific questions: 1. Do human hair follicle keratinocytes and fibroblasts in normal scalp skin express BDNF and/or its high-affinity receptor TrkB mRNA and immunoreactivity (IR)? 2. Does BDNF promote catagen development in organcultured human anagen hair follicles, as it does in mice? 3. If so, does this include the recruitment of recently recognized common mechanisms of catagen induction in human hair follicles, such as downregulation of hair follicle keratinocyte proliferation, upregulation of keratinocyte apoptosis, and transforming growth factor b (TGFb)-mediated catagen induction? 4. Can negative BDNF effects on human hair growth in vitro be antagonized? These questions were addressed by analyzing the expression pattern of BDNF and TrkB in human scalp skin biopsies as well as in isolated anagen VI hair follicles by rtPCR analysis, in situ hybridization, and immunofluorescence. Analysis of human hair follicle cycling is difficult. Most available hair follicles for rtPCR, in situ analyses, or immunohistochemistry derive from scalp skin excised during elective plastic surgery. In such biopsies, 80%–90% of all hair follicles are in anagen VI, and the remainder are in telogen, with catagen or early anagen hair follicles being detectable only extremely rarely (Paus and Cotsarelis, 1999). Therefore, in situ expression analyses of growth fac-

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

tors and receptors largely are limited to studying anagen VI and telogen hair follicles. In addition, human anagen VI scalp skin hair follicles were organ cultured (Philpott et al, 1994; Philpott, 1999) to investigate the impact of recombinant human BDNF (rhBDNF) or BDNF plus TGFbneutralizing antibodies on active hair shaft elongation and maintenance of the hair shaft-producing unit in the anagen hair bulb.

Results Human anagen hair follicles express BDNF and TrkB mRNA rtPCR revealed that micro-dissected human scalp skin anagen hair bulbs, which include the hair follicle epithelium proximal to the isthmus and bulge region as well as the dermal papilla and the proximal connective tissue sheath of the hair follicle, contained transcripts of the BDNF and TrkB genes (Fig 1). This offered an indication, that, just like their murine counterparts, human anagen hair follicles are both sources and targets of BDNF/TrkB-mediated signalling. BDNF mRNA is expressed both in the mesenchyme and epithelium of human anagen VI hair follicles, TrkB only in the epithelium (inner and outer root sheath) BDNF and TrkB transcription in human scalp anagen VI hair follicles was confirmed and localized by in situ hybridization using fluorescein-labeled oligonucleotide probe (Fig 2). All negative and positive controls (see Materials and Methods) confirmed the specificity and sensitivity of the used in situ stainings. Staining intensity for BDNF transcripts was strongest in the proximal inner root sheath and distal outer root sheath (Fig 2), and BDNF transcript-related signals were also detected in the dermal papilla. TrkB transcripts were present in the proximal inner and distal outer root sheath, but not in the dermal papilla (Fig 2). Immunohistochemistry confirms translation of BDNF and TrkB mRNA in human anagen VI hair follicles BDNF and TrkB translation in human scalp anagen VI hair follicles was confirmed by immunohistochemistry (Fig 2). All

Figure 1 Brain-derived neurotrophic factor (BDNF) and tyrosine kinase B (TrkB) are expressed in the human anagen hair follicle by rtPCR. rtPCR analysis was performed on isolated human hair follicles using primers for BDNF and TrkB. As positive control, RNA from full-thickness temporal human scalp skin was used and water was used as negative control.

124 : 4 APRIL 2005

negative and positive controls confirmed the specificity and sensitivity of the used immunohistochemical stainings: specific staining patterns were absent in negative controls without primary antibody, or were strongly diminished after preincubation with specific blocking peptide. Positive controls showed the expected IR patterns in human pancreatic epithelia (Miknyoczki et al, 1999). All extrafollicular IR patterns were not analyzed in detail, since they are the subject of a separate, ongoing study, and the schemes that summarize BDNF and TrkB IR have only been sketched in preliminary form (Fig 2).

BDNF/TRK SIGNALING IN HUMAN HAIR GROWTH

677

Semiquantitative analysis of BDNF IR demonstrated strongest protein expression in more differentiated keratinocyte cell populations such as the proximal inner root sheath and in the mesenchymal compartment, e.g. the dermal papilla. TrkB IR, in contrast, was strongest in more undifferentiated keratinocyte cell populations such as the basal layer of the hair follicle ostium. TrkB IR was also detected in the hair matrix and proximal outer root sheath but not in the dermal papilla (Table I). Low BDNF and TrkB is present in the telogen hair follicle Because of their low incidence in human scalp skin (10%–20% of all hair follicles), telogen hair follicles were detected only rarely in normal human scalp skin sections. In 13 (BDNF) and nine (TrkB) different donor samples (which contained more than 40 anagen hair follicles), only four telogen hair follicles, each of a different donor, were found per investigated antigen. Hair follicles in catagen or early anagen were not detected at all. Weakly positive BDNF staining (IR and in situ) was detected in the outer root sheath and secondary hair germ epithelium of all telogen hair follicles (Fig 2). BDNF staining was pronounced in the outer root sheath basal layer. TrkB staining was weak in the hair follicle epithelium (Fig 2). Interestingly and in contrast to our findings in anagen skin, in the telogen stage of the hair cycle the dermal papilla-stained negative for both antigens (Fig 2). BDNF concentration dependently inhibits hair shaft elongation in cultured human anagen hair follicles In order to assess the functional relevance of BDNF/TrkB signaling in the anagen hair follicle, microdisected lower anagen hair follicles from normal human scalp skin were cultured in the presence of 1, 5, 25, 50, and 150 ng per mL rhBDNF. These experiments revealed a significant reduction of hair growth in hair follicles cultured with 50 and 150 ng per mL BDNF, whereby hair follicles treated with 150 ng per mL exhibited dystrophic features such as pigment incontinence as a sign of toxification. Therefore 50 ng per mL was chosen for the treatment of a larger number of samples (Fig 3). Figure 2 Brain-derived neurotrophic factor (BDNF)- and tyrosine kinase B (TrkB)-expression patterns in human hair follicles. The schematic summaries display the most highly reproducible BDNF and TrkB IR patterns revealed in human anagen scalp skin hair follicles. Cryosections (immunohistochemistry [IHC]) or paraffin sections (in situ hybridization) from human scalp skin samples containing 80%–90% anagen hair follicles were stained with antisera against BDNF and TrkB (Santa Cruz Biotechnology) or were labeled with oligonucleotide probes to BDNF or TrkB. All images demonstrate representative IR patterns in anagen hair follicles and telogen hair follicles. If not otherwise labeled, arrows point at strongly stained structures. Boxes indicate anatomical location of the presented micrographs. The bottom panel gives examples of positive and negative controls. For the IHC-negative control, the primary antibody was omitted and the secondary antibody was applied (identical for BDNF and TrkB). The arrow points at the negative inner root sheath, which would stain strongly with primary antibody. Pancreatic ductal adeno carcinoma served as positive controls for IHC. Arrows point at strong staining. Pan-mRNA served as positive control and omittance of probe as negative control for in situ. Again, arrows point at the inner root sheath normally staining positive. ch, club hair; d, dermis; dp, dermal papilla; e, epidermis; hfe, hair follicle epithelium; hm, hair matrix; hs, hair shaft; irs, inner root sheath; ors, outer root sheath; sc, subcutis.

678 PETERS ET AL

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Table I. Summary of BDNF and TrkB staining intensities in various compartments of anagen VI scalp skin hair follicles BDNF

TrkB

mRNA

Protein

mRNA

Hair follicle ostium

þ

þ

þ

þ þ þa

Protein

Inner root sheath

þþþ

þþþ

þ

þþ

Outer root sheath þþ (including the isthmus and bulge region)

þ

þþ

þ
þ þ

Proximal hair follicle bulb þ

þ

þ

þ

Hair matrix

(þ)

(þ)

(þ)

0

Cuticle of the hair shaft

0

0

þ

þ

Dermal papilla

þþ

þþ

0

0

The table lists the mean staining intensities from nine (TrkB) or 13 (BDNF) different human scalp skin samples. From each sample ten microscopic fields or six hair follicles were evaluated and assigned arbitrary scores ( þ , weak; þ þ , obvious staining; þ þ þ , strong staining). a Basal layer, rest is þ like epidermis. BDNF, brain-derived neurotrophic factor; TrkB, tyrosine kinase B.

BDNF induces premature catagen development in vitro Hair cycle staging of hair follicles treated with 50 ng per mL rhBDNF after 48 h in culture did not show catage-like development. First catagen-like hair follicles could be detected after 96 h, whereas none were seen in control hair follicles (not shown). This difference became highly significant after 10 d in culture with 50 ng per mL rhBDNF (Fig 4).

Figure 3 Brain-derived neurotrophic factor (BDNF) reduces hair shaft elongation in organ culture. On the left, photodocumentation panels over a culture period of 10 d of one characteristic hair follicle per group are given. On the right, data pooled from six highly comparable, independent experiments (nine hair follicles per experiment, only pigmented hair follicles included) are shown. Growth rates are expressed as percent difference over day 0. Significances always refer to the growth rate of control hair follicles on the same day (p-values by Mann–Whitney U: þþþ o0.001, þ o0.05).

Figure 4 Fifty nanograms per milliliter recombinant human brain-derived neurotrophic factor (rhBDNF)-induced catagen after 10 d in culture. This figure shows characteristic examples of photodocumented hair follicles cultured over 10 d and corresponding H&E-stained cryosections. Note the wide anagen hair bulb of the control hair follicle (left). In comparison, the hair bulb of a hair follicle treated with 50 ng per mL rhBDNF appears narrow, the dermal papilla has rounded up, and the hair follicle epithelium is not enclosing it. The graph represents data pooled from four highly comparable independent experiments (nine hair follicles per experiment). Data are represented as percentage hair follicles in anagen VI, early or late catagen, respectively, or as hair cycle score. Each anagen VI hair follicle was ascribed an arbitrary value of 100, early catagen like of 200, mid-catagen like of 300, and late catagen like of 400 to calculate the hair cycle score. Values were added per group and divided by the number of staged follicles. The score thus represents the mean hair cycle stage of all hair follicles per group. Significances refer to the hair cycle score of control hair follicles on the same day (p-values by Mann–Whitney U: þ þ þ o0001, þ þ o0.01, þ o0.05). c, club; dp, dermal papilla; hs, hair shaft; irs, inner root sheath, ors, outer root sheath; pu, pigmentary unit.

BDNF upregulates apoptosis and inhibits proliferation of human hair follicle keratinocytes in situ Quantitative histomorphometry of terminal deoxynucleotidyl transferase (TdT)-mediated dUTP-biotin Nick end labeling (TUNEL) or Ki67 IR nuclei in organ-cultured human hair bulbs revealed significantly less Ki67 þ , proliferating keratinocytes in hair follicles treated for 10 d with 50 ng per mL rhBDNF when compared with controls (Fig 5). Complementary hair follicles showed significantly more TUNEL þ cells after 10 d of culture with 50 ng per mL rhBDNF (Fig 5). Quantitative TUNEL/ Ki67 histomorphometry of hair follicles harvested after only 2 or 4 d of culture did not show significant differences (not shown). BDNF upregulates TGFb2 gene transcription in human anagen hair follicles To investigate possible signaling loops recruited by BDNF in the control of catagen induc-

124 : 4 APRIL 2005

BDNF/TRK SIGNALING IN HUMAN HAIR GROWTH

679

TGFβ2 normalised against GADPH 100

arbitrary units

80

60

40

20

0 BDNF 50 ng/ml

control

Figure 6 Brain-derived neurotrophic factor (BDNF) increases tumor growth factor b2 (TGFb2) transcription. Nine human hair follicles were cultured with or without 50 ng per mL BDNF over 48 h. Proximal (lower) portions were collected with surgical scalpels after the culture. The expression of TGFb2 was measured by quantitative real-time PCR (TaqMan PCR) and normalized with GAPDH expression. Each PCR reaction was performed in triplicate. Graphs represent pooled date from these triplicates. Significance refers to control levels (p-values by Mann–Whitney U: þ þ þ o0.001, þ þ o0.01, þ o0.05).

Figure 5 Terminal deoxynucleotidyl transferase (TdT)-mediated dUTP-biotin Nick end labeling (TUNEL)/Ki67 double staining analysis on hair follicles cultured over 10 d. Per experiment harvested on day 10 after onset of culture, one to four of nine hair follicles allowed quantification of TUNEL (green) or Ki67 IR (red) nuclei. Only analysis of apoptosis and proliferation in keratinocytes yielded significant differences. Significances always refer to control levels (p-values by Mann–Whitney U: þ o0.05). Please note the numerous TUNEL IR terminally differentiating cell nuclei in the hair shaft. cs, connective tissue sheath; dp, dermal papilla; hs, hair shaft. Arrowheads point at TUNEL þ nuclei.

tion, we performed quantitative real-time rtPCR analysis of TGFb2 steady-state mRNA levels in hair bulbs, since TGFb2 is now appreciated as a key catagen-promoting growth factor in the human system (Soma et al, 2002; Tsuji et al, 2003; Hibino and Nishiyama, 2004). This showed substantial upregulation of TGFb2 mRNA in hair bulbs that were cultured in the presence of 50 ng per mL rhBDNF over 48 h (Fig 6). TGFb-neutralizing antibodies abrogate BDNF-induced inhibition of hair shaft elongation Simultaneous treatment of cultured human anagen hair follicles with 50 ng per mL BDNF and 10 (not shown) or 25 (Fig 7) mg per mL TGFbneutralizing antibodies completely abrogated the catageninductive and anti-proliferative effects of 50 ng per mL rhBDNF, e.g. hair shaft elongation rates did not differ from controls after 4 d in culture, when rhBDNF inhibition of hair shaft elongation became significantly different from control (Fig 7). Likewise, hair cycle staging, apoptosis (TUNEL), and

proliferation analysis (Ki67) on hair follicles cultured over 4 or 10 d revealed anagen to early catagen hair bulbs with high numbers of Ki67 þ proliferating and low numbers of TUNEL þ apoptotic keratinocytes in the control as well as the groups treated with BDNF together with TGFb-neutralizing antibody (not shown). Hair follicles that were treated with 50 ng per mL rhBDNF showed striking upregulation of TGFb2 expression in the hair bulb by immunohistochemistry especially in the hair follicle matrix, the inner root sheath, and the outer root sheath (Fig 7), whereas TGFb2 expression was restricted to the most differentiated layer of the inner root sheath in control hair follicles and hair follicles treated with 50 ng per mL BDNF and 25 mg per mL TGFbneutralizing antibodies.

Discussion Since it was first shown in mice that mammalian skin and especially keratinocytes may be a source for neurotrophins (Tron et al, 1990), a body of data has accumulated, in support of the concept that the skin epithelium is also an important target of neurotrophin signaling (Paus et al, 1994b; Pincelli and Yaar, 1997; Botchkarev et al, 1999a, b; Lambiase et al, 2000). More recently, our own studies have highlighted that neurotrophins are prominently involved in the control of murine hair follicle development and cycling and that the murine hair follicle is actually a prominent source of neurotrophins (Botchkarev et al, 1998a, c, 1999a, b, 2000). In this investigation on BDNF signaling in human hair growth control, we now show that selected neurotrophins act as hair growth regulators in the human system as well, e.g. via high-affinity tyrosinekinase receptors such as TrkB, and that the human hair follicle is both a source and target of neurotrophins. This underscores their importance in

680 PETERS ET AL

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Figure 7 Tumor growth factor b (TGFb)-neutralizing antibody abrogates Brain-derived neurotrophic factor (BDNF)-induced inhibition of hair shaft elongation. On the left, data pooled from three independent experiments (nine hair follicles per experiment; only pigmented hair follicles were included) are shown. Growth rates are expressed as percent difference over day 1. Significances always refer to the growth rate of control and recombinant human BDNF (rhBDNF)/TGFbneutralizing antibody treated hair follicles on the same day (p-values by Mann–Whitney U: þ 0.05). On the right, representative photomicrographs are shown of TGFb2 immunoreactivity in control hair follicles (top row), hair follicles treated with 50 ng per mL rhBDNF, and 25 ng per mL TGFb-neutralizing antibody, and hair follicles treated with 50 ng per mL rhBDNF (bottom row). The left column shows immunofluorescence micrographs; the right column shows inverted micrographs for better anatomical orientation. dp, dermal papilla; hs, hair shaft; hm, hair matrix; irs, inner root sheath; ors, outer root sheath.

cutaneous biology and further supports that neurotrophins are important hair growth-modulatory agents in hair growth control across species barriers: 1. The distribution of BDNF and TrkB in human scalp skin hair follicles suggests auto- and/or paracrine BDNF activities within the human hair follicle: a. paracrine activities and mesenchymal–epithelial cross-talk between BDNF-expressing fibroblasts in the anagen dermal papilla on the one hand, and TrkB þ hair follicle keratinocytes on the other. b. para- and/or autocrine-signaling activities of BDNF within the hair follicle epithelium expressing BDNF and TrkB on gene and protein expression level. c. suggestive involvement of BDNF in the autonomous, intrafollicular control of follicle cycling because of the hair cycle-dependent changes in BDNF expression in the mesenchymal ‘‘command center’’ of the hair follicle, i.e. the dermal papilla, between anagen VI (present) and telogen (absent) (Paus et al, 1999b; Stenn and Paus, 2001). 2. This is supported by the observation that BDNF inhibits hair shaft elongation and promotes catagen development in organ-cultured human anagen VI hair follicles, and suggests that BDNF functions rather as a terminally

differentiation-promoting factor than as a growth factor in human scalp hair follicles. 3. In line with this concept, higher doses of BDNF inhibit hair follicle keratinocyte proliferation and stimulate expression of the catagen-promoting and keratinocyte proliferation-inhibiting factor TGFb2. 4. That BDNF effects on human hair growth in vitro can be antagonized by pan-TGFb-neutralizing antibodies suggests that TGFb-mediated pathways along BDNF/ TrkB signaling serve at least as one important cellular ‘‘second messenger’’ system in the regressing human hair follicle. The expression pattern of BDNF and TrkB in human scalp skin hair follicles in anagen VI and telogen shown here is similar to the expression patterns described in mice and rats with low expression during the resting phase (telogen) and high expression in the high growth phase (anagen VI) (Botchkarev et al, 1998b, 1999b, 2000; Bergman et al, 2000), suggesting a parallel function of BDNF in murine and human hair growth regulation. Here, we can confirm this hypothesis by showing that exogenous BDNF does indeed cause premature catagen development of organ-cultured human anagen hair follicles as does BDNF overexpression in mice (Botchkarev et al, 1999a).

124 : 4 APRIL 2005

Catagen hair follicles are only rarely detected in human scalp skin samples; we therefore cannot provide data on TrkB expression of apoptotic cells in the regressing hair follicle as are seen in the mouse. In the mouse, the highaffinity BDNF-receptor TrkB is upregulated during catagen and co-localizes with markers of apoptotic cell death in the epithelium of the regressing hair bulb (Botchkarev et al, 2000). The promoting effect of BDNF on apoptosis-driven catagen development in the human anagen hair follicle may thus occur via BDNF secretion of the dermal papilla and/or more differentiated keratinocytes in the hair follicle, acting on TrkB þ hair follicle keratinocytes initially in the inner and outer sheath and upon catagen onset in the upregulating hair bulb. In this context, it is crucial to understand that neurotrophins utilize more than one mode of interaction with their receptors (Lopez-Sanchez and Frade, 2002): acting as high-affinity growth factors activated by low concentrations of neurotrophins (Yaar et al, 1994; Barker, 1998; Roux and Barker, 2002), activation of Trk receptors by high levels of neurotrophins has been shown to induce apoptosis depending on receptor availability and receptor density (cf. Roux and Barker, 2002). Apoptosis through the common low-affinity neurotrophin receptor p75, a member of the TNF-receptor family leading to rapid apoptosis of the expressing cells, is another possible pathway involved in BDNF-mediated hair follicle regression (Botchkarev et al, 2000) and is the subject of an ongoing study on neurotrophin signaling in human hair growth control through this receptor (Peters et al, 2005).1 Moreover, careful examination for morphological indicators of toxicity (see Materials and Methods) did not reveal any evidence that the hair growthinhibitory effects of BDNF presented here were related to hair follicle toxicity. In our model, BDNF was also able to upregulate the keratinocyte differentiation factor TGFb2 (Min et al, 1999; Hashimoto, 2000; Amendt et al, 2002) in the hair bulb as early as 48 h after start of culture. The observed TGFb2 IR pattern in hair follicles treated with 50 ng per mL BDNF is in good accordance with the expression pattern observed by others during progression of the hair cycle and induction of catagen (Soma et al, 2002). Moreover, in our study hair growth could be maintained by co-administration of BDNF with anti-TGFb1, 2, and 3-neutralizing antibodies. These observations suggest that TGFb2, known to be a potent catagen inducer in murine (Foitzik et al, 2000) and human hair follicles (Philpott et al, 1994; Soma et al, 1998; Inui et al, 2002; Soma et al, 2002, Tsuji et al, 2003; Hibino and Nishiyama, 2004), is at least one major player in BDNF-induced premature catagen development. As evidenced by our functional experiment utilizing a pan-TGFb-neutralizing antibody, TGFb1 or 3 may also contribute to the BDNF-induced premature catagen development observed in our experiments. But TGFb2 was shown to be a more potent catagen inducer in human scalp hair follicles than TGFb1 (Soma et al, 1998, 2002; Tsuji et al, 2003; Hibino and Nishiyama, 2004). Furthermore, to the best of our knowledge, 1 Peters EMJ, Hansen M, Overall R, Pertile P, Arck PC, Klapp BF, Paus R: p75, a brake on life and death decisions in human hair growth control, 2005, in preparation

BDNF/TRK SIGNALING IN HUMAN HAIR GROWTH

681

no effect of BDNF has previously been reported on TGFb3 expression, and TGFb3 is widely considered to be of negligible importance in both human and murine hair growth control since there is no published evidence to support any such role (cf. Soma et al, 1998, 2002; Foitzik et al, 1999; Stenn and Paus, 2001). In glial astrocytes, BDNF can induce cell cycle arrest proteins such as p53 (Jordan-Sciutto et al, 2001) and in corneal keratinocytes, BDNF promotes differentiation as shown by colony formation but not proliferation (You et al, 2000). Together with our findings of TGFb upregulation in unison with catagen induction and reduced keratinocyte proliferation by BDNF presented here, these observations led us to the hypothesis that BDNF/TrkB signaling in the context of the growing human hair follicle promotes terminal differentiation rather than proliferation directly and indirectly, e.g. by modulating other growth factor-mediated pathways. Our findings support an important role for BDNF as a trophic factor outside neural networks in the human and suggest to further use the cultured human hair follicle as an easily accessible and all the while complex miniature organculture system for future analysis of neurotrophins in epithelial and mesenchymal growth control. Local BDNF has been applied, with few side effects to patients suffering from diabetic polyneuropathy (Wellmer et al, 2001). Therefore, the topical application of TrkB agonists or the upregulation of endogeneous BDNF production in terminal anagen hair follicles may be useful future tools in the management of unwanted hair growth (hirsutism), whereas the local application of specific TrkB antagonists not yet available holds promise for the future management of telogen effluvium in human. Materials and Methods Tissue and hair follicle sources Temporal scalp skin containing mainly anagen VI hair follicles was obtained from disposed excess skin samples derived from patients undergoing elective plastic surgery after obtaining informed consent, following accepted ethic guidelines of the University of Hamburg, Germany. The study was conducted according to the Declaration of Helsinki Principles. After excision, tissue was maintained in Williams E Medium (Biochrom KG seromed, Berlin, Germany) for transportation at 41C up to 24 h. Upon arrival, the samples were divided by three: one part was immediately snap-frozen in liquid nitrogen for immunohistochemical analysis, the second was processed for paraffin embedding to allow routine hair follicle staging and in situ hybridization, and the third was processed for rtPCR and quantitative real-time rtPCR analysis and hair follicle culture as described below. Conventional rtPCR analysis Single human anagen VI hair follicles were microdisected from human skin biopsies. Microdisected hair follicles consisted of the hair bulb and the distal hair follicle, up to the level where the arector pili muscle inserts into the bulge region (including the hair shaft, the inner and outer root sheath, the connective tissue sheath, and the dermal papilla). RNA was extracted using the RNeasy Mini Kit (Qiagen, Hilden, Germany). 0.8 mg of total RNA was reverse transcribed using the first Strand cDNA Synthesis Kit for rtPCR (Roche, Mannheim, Germany) in an oligo-d(T)15-primed 20 mL reaction. rtPCR was carried out using the Core PCR Kit (Qiagen) in 25 mL reactions containing 1 mL cDNA, 2.5 mM MgCl2, 200 mM dNTP, 0.5 U Taq polymerase, and 0.5 pmol of each primer. The following oligonucleotide primers were designed to amplify target cDNA. 50 -TCCACCAGGTGAGAA-

682 PETERS ET AL

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

GAGTG-30 and 50 -ACTTGACTACTGAGCATCAC-30 to amplify a 319 bp fragment corresponding to nucleotides 277–595 of BDNF mRNA (Genbank: X60,201), and 50 -ATGCTGTGGTGGTGATTGCG30 and 50 -CCTTTCATGCCAAACTTGGAGTGTC-30 to amplify a 97 bp fragment corresponding to nucleotides 1652–1748 of the TrkB mRNA (Genbank: U12,140 common to all known isoforms of TrkB). The reactions were run in a thermal cycler (PE Biosystems, Foster City, California) for 5 min at 951C, then 30 s at 951C, 30 s at 601C, 45 s at 721C for 35 cycles, and then 5 min at 721C. Products were run on a 1% agarose gel (Sigma, St Louis, Missouri). The identity of the products was confirmed by sequencing. Quantitative real-time rtPCR analysis Hair follicles cultured over 48 h in the presence of 50 ng per mL rhBDNF (for details, see Table II) were dissected into proximal region consisting of the hair bulb and distal region consisting of the hair shaft and inner and outer root sheaths with the adjacent connective tissue sheath with a surgical scalpel under a microscope. Snap-frozen proximal hair follicle regions were homogenized with a glass homogenizer (neoLab, Heidelberg, Germany) and processed as described above. Amplification reactions were performed as 50 mL triplicates in a 96well microplate format containing 1  SYBR Green Master Mix (PE Biosystems), first strand cDNA, and each primer set. Primers were designed using the Perkin Express 1.0 Software (PE Biosystems) and each set of primers are situated in different exons as human TGFb2 50 -AAAGTGGACGTAGGCAGCAATTA-30 forward primer, human TGFb2 50 -GACCAACCGGCGGAAGA-30 reverse primer, human glyceraldehydes-3-phosphate dehydrogenase (GAPDH) 50 TGGGTGTGAACCATGAGAAG-30 forward primer, and human GAPDH 50 -GCTAAGCAGTTGGTGGTGC-30 reverse primer, respectively. The reaction mixture was subjected to the following thermal cycle conditions: 501C for 2 min, 951C for 10 min and 40 cycles of 15 s at 951C, and 1 min at 601C. Analysis of the reactions was carried out in an ABI Prism 7700 Sequence Detection System (PE Biosystems) monitoring after each cycle. The expression of TGFb2 was normalized with that of GAPDH. Each TaqMan experiment was performed in triplicate.

In situ hybridization Oligonucleotide probes to TrkB (Genebank: AF410,899 50 -GAT TGC CCA CCA GGA TCA GTT CAG ACA AGT30 ) and BDNF (Genebank: M61,176 50 -ATG ATG TCA AGC CTC TTG AAC CTG CCT TG-30 ), labeled with fluorescein at both the 30 and 50 ends, were obtained from Biognostik (Go¨ttingen, Germany; designed and used as recommended by vendor, Biognostik). Four micrometer paraffin sections of human skin were dewaxed and rehydrated before being treated with 50 mg per mL proteinase K (Roche) in proteinase K buffer (50 mM Tris, pH 7.5, 10 mM EDTA,

10 mM NaCl) for 15 min at 371C. Sections were then post-fixed for 10 min with 1% paraformaldehyde in PBS (0.68 M NaCl, 26.82 mM KCl, 17.64 mM KH2PO4, 0.10 M Na2HPO4, pH 7.4) followed by washes in PBS and water for 10 min each. Sections were then prehybridized in hybridization buffer (2  SSC (0.15 M NaCl, 0.015 M sodium citrate, pH 7.0), 5 mg yeast tRNA (Sigma, Taufkirchen, Germany), 150 mM NaCl, 50% formamide) for 2 h at 301C. Hybridization was carried out for 16 h at room temperature (251C) with 1.5 pmol probe in the hybridization buffer. Following hybridization, excess unbound probe was washed off by a brief rinse followed by two washes for 15 min each with 0.1  SSC at 371C. Sections were then equilibrated in TN buffer (pH 7.5, 100 mM Tris, pH 7.5, 150 mM NaCl) for 5 min before being incubated with blocking solution (TN (pH 7.5) containing 0.1% Triton-X 100% and 1% normal goat serum) for 30 min at 371C. The blocking solution was then replaced with antibody (anti-fluorescein AP FAB fragments (Roche)) diluted 1:1000 in blocking solution, and the sections were incubated for another 1 h at 371C. The slides were then washed twice in TNT (100 mM Tris, pH 7.5, 200 mM NaCl, 0.1% Tween-20) for 15 min each time, and then equilibrated in TN (pH 9.5, 100 mM Tris, pH 9.5, 150 mM NaCl) for 5 min. Sections were then incubated in the dark with color solution (TNM (100 mM Tris, pH 9.5, 100 mM NaCl, 50 mM MgCl2) to which 20 mL NBT/BCIP mix (Roche) had been added directly before use) for 3  1 h, with sections rinsed in TN (pH 9.5) and a fresh color solution added each hour. Immunohistochemistry Cryostat sections (8 mm) of human scalp skin, fixed in acetone (at
201C, 10 min), were preincubated with 10% normal goat serum (20 min, room temperature) and then incubated overnight at room temperature with the primary antisera to BDNF, TrkB, or TGFb2 (BDNF; rabbit, polyclonal, raised against AA 128–147 mapping at C-terminus of h-BDNF; Santa Cruz Biotechnology, Santa Cruz, California; cf. Lommatzsch et al, 1999; Botchkarev et al, 1999a, b; diluted 1:50; n ¼ 14. BDNF; mouse, monoclonal, raised against rhBDNF; Sigma, St Louis, MI, USA; diluted 1:200; n ¼ 5. TrkB; rabbit, polyclonal, raised against AA 794–808 mapping at the C-terminus of TrkB gp 145; Santa Cruz Biotechnology; cf. Lommatzsch et al, 1999; Botchkarev et al, 1999a, b; diluted 1:75; n ¼ 9. TGFb2; rabbit, polyclonal, raised against peptide mapping at the C-terminus of TGFb2 of human origin; Santa Cruz Biotechnology; cf. Lyall et al, 2001; diluted 1:100; n ¼ 4). For BDNF we found two working antibodies yielding comparable staining results, which mapped different epitopes of BDNF. For TrkB only antibodies mapping the same epitope were commercially available. Working BDNF and TrkB antibodies were available from Santa Cruz Biotechnology; we therefore chose these for immunoflourescence screening of all available skin sam-

Table II. Hair follicle organ-culture treatment schedules Treatment

Duration of culture

Number of donors

Processed for

A

5 or 50 ng per mL rhBDNF

10 d

11 different donors

Seven for hair shaft elongation; four for proliferation, staging analysis, TGFb2 immunohistochemistry; eight for apoptosis analysis

Aa

50 ng per mL rhBDNF

96 h

One donor

Apoptosis/proliferation and staging analysis

Ab

50 ng per mL rhBDNF

48 h

Two different donors

One for apoptosis/proliferation and staging analysis, one for TGFb2 rtPCR

B

5, 10, 25, 50, 150 ng per mL rhBDNF

10 d

Five different donors

Hair shaft elongation

C

5 and 50 ng per mL anti-BDNF

10 d

Eight different donors

Hair shaft elongation, two for apoptosis/proliferation and staging analysis

D

5, 10 or 25 mg per mL anti-TGFb 250 ng per mL rhBDNF þ

96 h

Three different donors

Apoptosis/proliferation and staging analysis, TGFb2 immunohistochemistry

Each donor also contributed to controls. rhBDNF, recombinant human brain-derived neurotrophic factor; TGFb2, transforming growth factor b2.

124 : 4 APRIL 2005 ples. All anti-sera were supplemented with 2% normal goat serum. This was followed by an incubation of 45 min at room temperature with tetramethylrhodamine-isothiocyanate-conjugated F(ab)2 fragments of goat anti-rabbit or goat anti-mouse IgG (Dianova, Hamburg, Germany) diluted 1:200 in Tris-buffered saline (TBS). Incubation steps were interspersed by three washes with TBS (5 min each). Then sections were stained with 40 ,6-diamidin-20 -phenylindol-dihydrochlorid (DAPI, Boehringer Mannheim, Mannheim, Germany) with a concentration of 1 mg per mL (1 min, room temperature) for identification of cell nuclei. Two types of negative controls were run: (a) slides were incubated with the secondary antibody alone; (b) slides were incubated with a mixture (incubated overnight at 41C in tubes covered with 1% milk powder in TBS) of the primary antibody, and the control peptide for the specific anti-sera (Santa Cruz Biotechnology, 5 and 10 and 20 times higher concentrated than primary antibody, overnight, room temperature) after preincubation with 10% normal goat serum. As positive controls, cryopreserved samples from human pancreatic ductal adenocarcinoma (Miknyoczki et al, 1999) were kindly supplied by Prof. Kalthoff (Molecular Oncology, Department of Surgery, University of Kiel, Germany) and Dr Hosch (Department of Surgery, University Hospital Hamburg-Eppendorf, Germany). Histomorphometry A minimum of six sections of nine (TrkB) or 14 (BDNF) scalp skin samples, each derived from a different donor, were examined at  250 magnification under a Zeiss Axioscope 2 fluorescence microscope using appropriate filters (Zeiss, Go¨ttingen, Germany). IR was assessed by histomorphometry in defined compartments such as the hair follicle ostium (hair follicle epithelium between the epidermis and entry of the sebaceous gland into the hair canal), inner root sheath, outer root sheath, cuticle, dermal papilla, matrix, and proximal hair bulb epithelium (below Auber’s line; for details on these follicle compartments, see Mu¨ller-Ro¨ver et al, 2001). Intensity was recorded in a semi-quantitative manner by ascribing arbitrary units to different staining intensities (0, negative; þ , barely visible staining; þ þ , apparent staining; þ þ þ , strong staining). For each IR pattern, follicular compartments were evaluated in at least six different anagen hair follicles per sample and the extrafollicular compartments in a minimum of ten microscopic fields. The photodocumentation was carried out with the help of a modular imaging program (Openlab, improvision, Heidelberg, Germany). The results of the observed, representative IR patterns were also recorded qualitatively in schematized recording protocols. Hair follicle organ culture Normal human scalp skin hair follicles in the anagen VI stage of the hair cycle (Mu¨ller-Ro¨ver et al, 2001) were isolated following the protocol published by Philpott et al (1994) and Philpott (1999) with slight modifications. Briefly, after separation of the epidermis and dermis from subcutaneous fat under a binocular dissecting microscope, the proximal two-thirds of anagen hair follicles located in the subcutaneous fat were isolated using watchmakers forceps and subsequently collected in Petri dishes containing complete hair follicle culture medium (Williams E, Biochrom KG seromed); 1% penicillin–streptomycin (Life Technologies, Eggenstein, Germany); 1% L-glutamine 200 mM (Life Technologies); 0.02% hydrocortisone (Sigma, Taufkirchen, Germany); and 0.1% insulin (Sigma). Three hair follicles per well were then randomly distributed and cultured in 24-well plates (Costar, Bremen, Germany) containing 500 mL of complete hair follicle culture medium per well. Per experiment, a minimum of three wells (containing three hair follicles each) was assigned to each test group, and was supplemented with different concentrations of rhBDNF (Promega, Woods, Wisconsin) or rhBDNF with TGFb-neutralizing antibody (monoclonal anti-TGFb1–3 antibody (R&D Systems, Weisbaden, Germany); binds to all three TGFb isoform (Roberts et al, 1985; Cheifetz et al, 1990). An equal number of control follicles were cultured in complete hair follicle culture medium alone. Each ex-

BDNF/TRK SIGNALING IN HUMAN HAIR GROWTH

683

periment was repeated with hair follicles from different donors (for details see Table II). Every second day, each well was photodocumented, the total length of each hair follicle was measured, medium was replaced, and fresh supplements were added. After 2, 4, or 10 d hair follicles were snap-frozen in a drop of OCT Cryochrome (Shandon, Pittsburgh, Pennsylvania) and stored at –801C until cryosectioning. Hair cycle staging Cultured hair follicles were staged after 2, 4, and 10 d (for detail see Table II) in culture, either by morphometric evaluation of entire hair follicles photodocumented by inversion light microscopy or by routine H&E histomorphometry, following previously published hair cycle staging guidelines for the murine hair cycle (Mu¨ller-Ro¨ver et al, 2001) and morphological characteristics of the human hair cycle (Kligman, 1959). Macroscopically, hair follicles were classified as anagen VI if the hair bulb was wider than the rest of the hair follicle, the dermal papilla was partially covered by melanin, and if the pigmented hair shaft was emerging directly from the hair follicle pigmentary unit in the precortical hair matrix. Histologically, anagen VI hair follicles show a narrow, elongated, oval-shaped dermal papilla of loose consistency, which is fully enclosed by the hair bulb keratinocytes, residing in an onion-shaped, large hair bulb. Hair follicles were classified as ‘‘early catagen’’ if the hair bulb was more narrow, i.e. had essentially the same diameter as the rest of the hair follicle, if a slightly condensed, tear-drop-shaped dermal papilla was visible, if the pigmented zone of the hair shaft was beginning to retract in the distal direction from the pigmentary unit, and/or if the number of visible melanin granules in the distal hair bulb was substantially decreased, compared with high anagen. Hair follicles were classified as being in ‘‘mid-catagen’’ when melanin granules were almost absent from the hair bulb and if, on longitudinal sections, the bulb appeared opened at its proximal end, with a compact, ballshaped dermal papilla. Hair follicles were classified as being in ‘‘late catagen’’, if the hair bulb had become narrower than the distal rest of the cultured hair follicle, if the dermal papilla was clearly visible in an unpigmented hair bulb, yet was further condensed compared with mid-catagen, and if the proximal part of the pigmented hair shaft was fully separated from the dermal papilla. Histologically, these late catagen hair follicles display an epithelial strand, a column of epithelial cells between the compact dermal papilla and the germ capsule (secondary hair germ) (Mu¨ller-Ro¨ver et al, 2001), the likely seat of epithelial stem cells from which the hair matrix is reconstructed during the subsequent anagen stage (Panteleyev et al, 2001). Furthermore, late catagen hair follicles have already developed the so-called club hair, the hair shaft of follicles in the ‘‘resting’’ stage (telogen), whose proximal end is unpigmented and appears separated from the dermal papilla. In addition to hair cycle staging, signs of hair follicle dystrophy such as extraepithelial pigment (melanin incontinence), ectopic localization of melanin granules in the hair bulb (Paus et al, 1994a; Tobin et al, 1998), or hair shaft kinking were also documented. For statistical analysis (hair cycle score [HCS]), anagen VI hair follicles were arbitrarily attributed a score of 100, hair follicles in early catagen a score of 200, in mid-catagen of 300, and in late catagen of 400. The sum of scores per group was than divided by the number of investigated hair follicles (Maurer et al, 1997; Peters et al, 1999). The mean value of these scores therefore is a reliable quantitative indicator of the mean hair follicle stage that had been reached on average by a larger population of hair follicles after culture (Maurer et al, 1997; Peters et al, 1999). Apoptosis/proliferation assay Cryostat sections of cultured hair follicles were processed for TUNEL labeling (apoptotic cell nuclei) and Ki67 labeling (proliferating cell nuclei) after 2, 4, or 10 d (for detail, see Table II) according to our previously published protocol (Lindner et al, 1997) adapted for human antigens. The TUNEL assay detects DNA fragmentation by enzymatic labeling of free 30 -OH ends according to the manufacturer’s instructions (Apopdetect Fluorescein, QBiogene, Heidelberg, Germany). Subsequently, sec-

684 PETERS ET AL

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

tions were incubated with mouse anti-human Ki67 antiserum (anti-human Ki67 antigen, monoclonal, mouse, DAKO, Hamburg, Germany), detected by a rhodamine-conjugated goat anti-mouse antibody. Cell nuclei were counterstained with DAPI as described above. For negative controls, the TdT-enzyme step was omitted. Hair bulbs were photo-documented at  250 magnification with a Zeiss Axioscope 2 fluorescence microscope, and IR nuclei were counted in individual hair follicles in the following tissue compartments: epithelial bulb (hair growth effects of culture conditions), the dermal papilla, and the surrounding connective tissue sheath of all hair follicles (toxic effects of culture conditions), which allowed full appreciation of a longitudinal-sectioned dermal papilla (since this is very difficult to achieve, the total number of hair bulbs analyzed was restricted to 9–16 hair bulbs per group).

Statistical analysis The HCS was assessed and calculated as described (Maurer et al, 1997; Peters et al, 1999). Scores, TUNEL IR, and Ki67 IR cell nuclei numbers were pooled per group for each experiment. The results between different experiments using samples from different donors were highly comparable so that the mean scores and numbers of TUNEL IR or Ki67 IR cell nuclei per group and experiment from all experiments were then pooled again and statistical differences between groups were determined by the Mann–Whitney U test for unpaired samples.

This study was supported in part by grants from the German Research Foundation (DFG) to Ralf Paus (Pa 345/6-3) and Eva Peters (Pe 890/11) and by a grant from Cutech Srl (Padova, Italy) to Burghard Klapp and Petra Arck. We are grateful to Dr Thomas Schmidt-Rose and Silke Altgilbers (Beiersdorf AG, Hamburg, Germany) for helping to perform quantitative real-time rtPCR analysis. DOI: 10.1111/j.0022-202X.2005.23648.x Manuscript received April 26, 2004; revised August 12, 2004; accepted for publication September 2, 2004 Address correspondenc to: Eva M. J. Peters, MD, Skin and Hair Research in Psychoneuroimmunology, Department of Internal Medicine, Biomedical Research Center, Room Nr. 2.0549, University Medicine Charite´, Campus Virchow Hospital, Augustenburger Platz 1, D-13353, Berlin, Germany. Email: [email protected]

References Alleva E, Santucci D: Psychosocial versus ‘‘physical’’ stress situations in rodents and humans: Role of neurotrophins. Physiol Behav 73:313–320, 2001 Amendt C, Mann A, Schirmacher P, Blessing M: Resistance of keratinocytes to TGFbeta-mediated growth restriction and apoptosis induction accelerates re-epithelialization in skin wounds. J Cell Sci 115:2189–2198, 2002 Asaumi K, Nakanishi T, Asahara H, Inoue H, Takigawa M: Expression of neurotrophins and their receptors (TRK) during fracture healing. Bone 26:625–633, 2000 Barde YA, Edgar D, Thoenen H: Purification of a new neurotrophic factor from mammalian brain. EMBO J 1:549–553, 1982 Barker PA: p75NTR: A study in contrasts. Cell Death Differ 5:346–356, 1998 Barouch R, Appel E, Kazimirsky G, Braun A, Renz H, Brodie C: Differential regulation of neurotrophin expression by mitogens and neurotransmitters in mouse lymphocytes. J Neuroimmunol 103:112–121, 2000 Bayas A, Kruse N, Moriabadi NF, et al: Modulation of cytokine mRNA expression by brain-derived neurotrophic factor and nerve growth factor in human immune cells. Neurosci Lett 335:155–158, 2003 Bergman E, Ulfhake B, Fundin BT: Regulation of NGF-family ligands and receptors in adulthood and senescence: Correlation to degenerative and regenerative changes in cutaneous innervation. Eur J Neurosci 12:2694–2706, 2000 Botchkareva NV, Botchkarev VA, Albers KM, Metz M, Paus R: Distinct roles for nerve growth factor and brain-derived neurotrophic factor in controlling the rate of hair follicle morphogenesis. J Invest Dermatol 114:314–320, 2000

Botchkareva NV, Botchkarev VA, Metz M, Silos-Santiago I, Paus R: Retardation of hair follicle development by the deletion of TrkC, high-affinity neurotrophin-3 receptor. J Invest Dermatol 113:425–427, 1999 Botchkarev V, Botchkareva N, Welker P, et al: A new role for neurotrophins: Involvement of brain-derived neurotrophic factor and neurotrophin-4 in hair cycle control. FASEB J 13:395–410, 1999a Botchkarev VA, Botchkareva NV, Albers KM, Chen LH, Welker P, Paus R: A role for p75 neurotrophin receptor in the control of apoptosis-driven hair follicle regression. FASEB J 14:1931–1942, 2000 Botchkarev VA, Botchkarev NV, Albers KM, van der Veen C, Lewin GR, Paus R: Neurotrophin-3 involvement in the regulation of hair follicle morphogenesis. J Invest Dermatol 111:279–285, 1998a Botchkarev VA, Botchkareva NV, Lommatzsch M, et al: BDNF overexpression induces differential increases among subsets of sympathetic innervation in murine back skin. Eur J Neurosci 10:3276–3283, 1998b Botchkarev VA, Eichmu¨ller S, Johansson O, Paus R: Hair cycle-dependent plasticity of skin and hair follicle innervation in normal murine skin. J Comp Neurol 386:379–395, 1997 Botchkarev VA, Metz M, Botchkareva NV, et al: Brain-derived neurotrophic factor, neurotrophin-3, and neurotrophin-4 act as ‘‘eliotrophins’’ in murine skin. Lab Invest 79:557–572, 1999b Botchkarev VA, Peters EM, Botchkareva NV, Maurer M, Paus R: Hair cycle-dependent changes in adrenergic skin innervation, and hair growth modulation by adrenergic drugs. J Invest Dermatol 113:878–887, 1999c Botchkarev VA, Welker P, Albers KM, et al: A new role for neurotrophin-3: Involvement in the regulation of hair follicle regression (catagen). Am J Pathol 153:785–799, 1998c Cheifetz S, Hernandez H, Laiho M, ten Dijke P, Iwata KK, Massague J: Distinct transforming growth factor-beta (TGF-beta) receptor subsets as determinants of cellular responsiveness to three TGF-beta isoforms. J Biol Chem 265:20533–20538, 1990 dos Santos AA, Medina SV, Sholl-Franco A, de Araujo EG: PMA decreases the proliferation of retinal cells in vitro: The involvement of acetylcholine and BDNF. Neurochem Int 42:73–80, 2003 Ernfors P: Local and target-derived actions of neurotrophins during peripheral nervous system development. Cell Mol Life Sci 58:1036–1044, 2001 Farinas I, Cano-Jaimez M, Bellmunt E, Soriano M: Regulation of neurogenesis by neurotrophins in developing spinal sensory ganglia. Brain Res Bull 57:809–816, 2002 Foitzik K, Lindner G, Mu¨ller-Ro¨ver S, et al: Control of murine hair follicle regression (catagen) by TGF-beta1 in vivo. FASEB J 14:752–760, 2000 Foitzik K, Paus R, Doetschman T, Dotto GP: The TGF-beta2 isoform is both a required and sufficient inducer of murine hair follicle morphogenesis. Dev Biol 212:278–289, 1999 Hannestad J, Germana A, Catania S, Laura R, Ciriaco E, Vega JA: Neurotrophins and their receptors in the pigeon caecal tonsil. An immunohistochemical study. Vet Immunol Immunopathol 61:359–367, 1998 Hashimoto K: Regulation of keratinocyte function by growth factors. J Dermatol Sci 24:46–S50, 2000 Hibino T, Nishiyama T: Role of TGF-beta2 in the human hair cycle. J Dermatol Sci 35:9–18, 2004 Inui S, Fukuzato Y, Nakajima T, Yoshikawa K, Itami S: Androgen-inducible TGFbeta1 from balding dermal papilla cells inhibits epithelial cell growth: A clue to understand paradoxical effects of androgen on human hair growth. FASEB J 16:1967–1969, 2002 Jordan-Sciutto KL, Murray Fenner BA, Wiley CA, Achim CL: Response of cell cycle proteins to neurotrophic factor and chemokine stimulation in human neuroglia. Exp Neurol 167:205–214, 2001 Kligman AM: The human hair cycle. J Invest Dermatol 33:307–316, 1959 Lambiase A, Manni L, Bonini S, Rama P, Micera A, Aloe L: Nerve growth factor promotes corneal healing: Structural, biochemical, and molecular analyses of rat and human corneas. Invest Ophthalmol Vis Sci 41:1063–1069, 2000 Lindner G, Botchkarev VA, Botchkareva NV, Ling G, van der Veen C, Paus R: Analysis of apoptosis during hair follicle regression (catagen). Am J Pathol 151:1601–1617, 1997 Lommatzsch M, Braun A, Mannsfeldt A, et al: Abundant production of brainderived neurotrophic factor by adult visceral epithelia. Implications for paracrine and target-derived neurotrophic functions. Am J Pathol 155:1183–1193, 1999 Lopez-Sanchez N, Frade JM: Control of the cell cycle by neurotrophins: Lessons from the p75 neurotrophin receptor. Histol Histopathol 17:1227–1237, 2002 Lyall F, Simpson H, Bulmer JN, Barber A, Robson SC: Transforming growth factor-beta expression in human placenta and placental bed in third trimester normal pregnancy, preeclampsia, and fetal growth restriction. Am J Pathol 159:1827–1838, 2001

124 : 4 APRIL 2005 Marshall L, Born J: Brain–immune interactions in sleep. Int Rev Neurobiol 52: 93–131, 2002 Maurer M, Handjiski B, Paus R: Hair growth modulation by topical immunophilin ligands: Induction of anagen, inhibition of massive catagen development, and relative protection from chemotherapy-induced alopecia. Am J Pathol 150:1433–1441, 1997 Mecklenburg L, Tobin DJ, Mu¨ller-Ro¨ver S, et al: Active hair growth (anagen) is associated with angiogenesis. J Invest Dermatol 114:909–916, 2000 Metsis M: Genes for neurotrophic factors and their receptors: Structure and regulation. Cell Mol Life Sci 58:1014–1020, 2001 Miknyoczki SJ, Lang D, Huang L, Klein-Szanto AJ, Dionne CA, Ruggeri BA: Neurotrophins and Trk receptors in human pancreatic ductal adenocarcinoma: Expression patterns and effects on in vitro invasive behavior. Int J Cancer 81:417–427, 1999 Min BM, Woo KM, Lee G, Park NH: Terminal differentiation of normal human oral keratinocytes is associated with enhanced cellular TGF-beta and phospholipase C-gamma 1 levels and apoptotic cell death. Exp Cell Res 249:377–385, 1999 Moalem G, Gdalyahu A, Shani Y, et al: Production of neurotrophins by activated T cells: Implications for neuroprotective autoimmunity. J Autoimmun 15:331–345, 2000 Mu¨ller-Ro¨ver S, Handjiski B, van der Veen C, et al: A comprehensive guide for the accurate classification of murine hair follicles in distinct hair cycle stages. J Invest Dermatol 117:3–15, 2001 Panteleyev AA, Jahoda CA, Christiano AM: Hair follicle predetermination. J Cell Sci 114:3419–3431, 2001 Paus R, Christoph T, Mu¨ller-Ro¨ver S: Immunology of the hair follicle: A short journey into terra incognita. J Investig Dermatol Symp Proc 4:226–234, 1999a Paus R, Cotsarelis G: The biology of hair follicles. N Eng J Med 341:491–497, 1999 Paus R, Handjiski B, Eichmu¨ller S, Czarnetzki BM: Chemotherapy-induced alopecia in mice. Induction by cyclophosphamide, inhibition by cyclosporine A, and modulation by dexamethasone. Am J Pathol 144:719–734, 1994a Paus R, Luftl M, Czarnetzki BM: Nerve growth factor modulates keratinocyte proliferation in murine skin organ culture. Br J Dermatol 130:174–180, 1994b Paus R, Mu¨ller-Ro¨ver S, Botchkarev VA: Chronobiology of the hair follicle: Hunting the ‘‘hair cycle clock’’. J Invest Dermatol Symp Proc 4:338–345, 1999b Peters EM, Botchkarev VA, Botchkareva NV, Tobin DJ, Paus R: Hair-cycleassociated remodeling of the peptidergic innervation of murine skin, and hair growth modulation by neuropeptides. J Invest Dermatol 116:236–245, 2001 Peters EM, Maurer M, Botchkarev VA, Gordon DS, Paus R: Hair growth-modulation by adrenergic drugs. Exp Dermatol 8:274–281, 1999

BDNF/TRK SIGNALING IN HUMAN HAIR GROWTH

685

Peters EMJ, Botchkarev VA, Paus R: Neurotrophins act as regulators of hair follicle morphogenesis and cycling. In: Van ND (eds). Hair Science and Technology. Tournai, Belgium: Skinterface, 2003 Philpott M: In vitro maintenance of isolated hair follicles: Current status and future development. Exp Dermatol 8:317–319, 1999 Philpott MP, Sanders D, Westgate GE, Kealey T: Human hair growth in vitro: A model for the study of hair follicle biology. J Dermatol Sci 7:55–S72, 1994 Pincelli C, Yaar M: Nerve growth factor: Its significance in cutaneous biology. J Investig Dermatol Symp Proc 2:31–36, 1997 Renz H: The role of neurotrophins in bronchial asthma. Eur J Pharmacol 429: 231–237, 2001 Roberts AB, Anzano MA, Wakefield LM, Roche NS, Stern DF, Sporn MB: Type beta transforming growth factor: A bifunctional regulator of cellular growth. Proc Natl Acad Sci USA 82:119–123, 1985 Roux PP, Barker PA: Neurotrophin signaling through the p75 neurotrophin receptor. Prog Neurobiol 67:203–233, 2002 Shibayama E, Koizumi H: Cellular localization of the Trk neurotrophin receptor family in human non-neuronal tissues. Am J Pathol 148:1807–1818, 1996 Sieber-Blum M: Growth factor synergism and antagonism in early neural crest development. Biochem Cell Biol 76:1039–1050, 1998 Soma T, Ogo M, Suzuki J, Takahashi T, Hibino T: Analysis of apoptotic cell death in human hair follicles in vivo and in vitro. J Invest Dermatol 111:948–954, 1998 Soma T, Tsuji Y, Hibino T: Involvement of transforming growth factor-beta2 in catagen induction during the human hair cycle. J Invest Dermatol 118:993–997, 2002 Stenn KS, Paus R: Controls of hair follicle cycling. Physiol Rev 81:449–494, 2001 Tobin DJ, Hagen E, Botchkarev VA, Paus R: Do hair bulb melanocytes undergo apoptosis during hair follicle regression (catagen)? J Invest Dermatol 111:941–947, 1998 Tron VA, Coughlin MD, Jang DE, Stanisz J, Sauder DN: Expression and modulation of nerve growth factor in murine keratinocytes (PAM 212). J Clin Invest 85:1085–1089, 1990 Tsuji Y, Denda S, Soma T, Raftery L, Momoi T, Hibino T: A potential suppressor of TGF-beta delays catagen progression in hair follicles. J Investig Dermatol Symp Proc 8:65–68, 2003 Wellmer A, Misra VP, Sharief MK, Kopelman PG, Anand P: A double-blind placebo-controlled clinical trial of recombinant human brain-derived neurotrophic factor (rhBDNF) in diabetic polyneuropathy. J Peripher Nerv System 6:204–210, 2001 Yaar M, Eller MS, DiBenedetto P, et al: The trk family of receptors mediates nerve growth factor and neurotrophin-3 effects in melanocytes. J Clin Invest 94:1550–1562, 1994 You L, Kruse FE, Volcker HE: Neurotrophic factors in the human cornea. Invest Ophthalmol Vis Sci 41:692–702, 2000