• unlimited access with print and download
    $ 37 00
  • read full document, no print or download, expires after 72 hours
    $ 4 99
More info
Unlimited access including download and printing, plus availability for reading and annotating in your in your Udini library.
  • Access to this article in your Udini library for 72 hours from purchase.
  • The article will not be available for download or print.
  • Upgrade to the full version of this document at a reduced price.
  • Your trial access payment is credited when purchasing the full version.
Buy
Continue searching

Functional analysis of Trps1 in skin and hair follicle development

Dissertation
Author: Katherine A. Fantauzzo
Abstract:
Hair follicle morphogenesis involves a series of reciprocal epithelial-mesenchymal interactions that drive the growth and differentiation of the developing follicle. The goal of this thesis was to address the question of how a simple epithelium and underlying mesenchyme interact to generate the diverse cell types of the hair follicle. My thesis focused on identifying novel transcriptional regulators of hair follicle morphogenesis and characterizing the expression and function of such molecules in both mice and humans. I initially used microarray hybridization analysis to generate a global transcriptional profile of murine skin and hair follicle morphogenesis. Through separate examination of the epithelial and mesenchymal compartments, we identified novel secreted and membrane-bound factors regulating either epidermal placode formation, including members of the stratified epithelium secreted peptides complex, the Wnt inhibitor Dickkopf 4 and the potential Hoxa13 target Derp7 , or the reciprocal condensation of the underlying dermis, such as Syndecan-1 and members of the epidermal and keratinocyte growth factor families. To further refine the signaling molecules that coordinately control the regulation of both compartments, I focused on the zinc-finger transcription factor Trps1. I demonstrated that Trps1 localizes to the nuclei of dermal papillae cells and the highly proliferative epithelial cells of anagen hair follicles, indicating a role for Trps1 in promoting growth of the follicle. To identify downstream target genes of Trps1, I performed microarray hybridization analysis comparing expression patterns in wild-type versus Trps1Δgt/Δgt embryonic whisker pads. I established that Trps1 directly activates a number of secreted Wnt antagonists and transcription factors in the vibrissa follicle, including the hair follicle stem cell regulators Lhx2 and Sox9 . Upon analysis of the vibrissa follicle defects in Trps1Δgt/Δgt embryos, I uncovered a transcriptional hierarchy including Trps1, Shh and Sox9 that controls the specification of early hair follicle progenitors and their subsequent proliferation. Finally, I used comparative genomics and morphological analyses to characterize a position effect on TRPS1 associated with hypertrichosis in both humans and mice. I demonstrated that TRPS1 is subject to complex tissue- and temporal-specific regulation, as underexpression of the gene can result in both hypotrichosis and hypertrichosis, underscoring its critical role in hair follicle development.

Table of Contents Chapter I: General introduction 1 1. Ectodermal appendage formation 2 2. Regional specificity of hair types in the mouse 3 3. Hair follicle morphogenesis 4 4. The hair follicle cycle 8 5. Hypertrichosis 12 6. Hypotrichosis 14 7. Trichorhinophalangeal syndrome 15 8. TRPS1 structure and function 17 9. Transcriptional targets of TRPS1 21 10.TRPS1 protein interactions and posttranslational modifications 22 11. Expression of TRPS1 24 12. Mouse models: Trp^9t/Agt and TrpsrA null 25 13. Position effects on genes regulating hair follicle development 28 14. Work described in this thesis 31 Chapter II: Transcriptional profiling of murine skin and hair follicle morphogenesis: Epidermal placode formation and differentiation 34 1. Overview 35 2. Transcriptional profiling of developing mouse epidermis reveals novel patterns of coordinated gene expression (Manuscript #1) 39 3. The Wnt inhibitor, Dickkopf 4, is induced by canonical Wnt signaling during ectodermal appendage morphogenesis (Manuscript #2) 50 i

4. Dermal papilla derived protein 7 61 5. Derp7 materials and methods 73 Chapter III: Transcriptional profiling of murine skin and hair follicle morphogenesis: Condensation of the dermis 76 1. Overview 77 2. KGF and EGF signaling block hair follicle induction and promote interfollicular epidermal fate in developing mouse skin (Manuscript #3) 79 3. Dynamic expression of Syndecan-1 during hair follicle morphogenesis (Manuscript #4) 92 Chapter IV: Dynamic expression of Trpsl during hair follicle morphogenesis and cycling 100 1. Overview 101 2. Dynamic expression of the zinc-finger transcription factor Trpsl during hair follicle morphogenesis and cycling (Manuscript #5) 103 3. Discussion 111 Chapter V: Identification of downstream molecular targets of Trpsl during murine hair follicle morphogenesis 115 1. Overview 116 2. Identification of downstream transcriptional targets of Trpsl 118 3. Trpsl colocalizes with hair follicle stem cell markers 124 4. Target gene expression levels are altered in Trps1A9t/Agt mutant mice 126 ii

5. TRPS1 directly binds GATA sites in the promoters of its target genes 128 6. Discussion 131 7. Materials and methods 147 Chapter VI: Regulation of the Shh pathway in the developing murine hair follicle by Trpsl 154 1. Overview 155 2. Abnormalities in hair follicle development in Trps1Agt/Agt mutant mice 157 3. Trps1Agt/Agt mutant vibrissae follicles exhibit increased levels of proliferation and decreased expression of Shh 160 4. Trpsl and Shh colocalize in the matrix and inner root sheath of vibrissae follicles 163 5. Shh pathway members are transcriptionally downregulated in Trps1Agt/Agt mutant vibrissae follicles 165 6. Trpsl forms stable complexes with Gli2 166 7. SOX9, a Shh target gene, directly binds to the TRPS1 promoter 168 8. Generation of shh+/tm1(EGFP/cre);Trps1+/Agt compound heterozygous mice 172 9. Discussion 174 10. Materials and methods 182 Chapter VII: A position effect on TRPS1 is associated with Ambras syndrome in humans and the Koala phenotype in mice 190 iii

1. Overview 191 2. A position effect on TRPS1 is associated with Ambras syndrome in humans and the Koala phenotype in mice (Manuscript #6) 194 3. Discussion 208 Chapter VIII: General discussion 213 References 236 IV

List of Abbreviations All abbreviations are listed as they first appear in the text. In general, human genes and RNA are written in italicized capital letters (e.g. TRPS1 gene or RNA), while human proteins are written in non-italicized capital letters (e.g. TRPS1 protein). Mouse genes, RNA and proteins follow the same format, with the exception that only the first letter is capitalized (e.g. Trpsl gene or RNA; Trpsl protein). ABCA - ATP-binding cassette, sub-family A Alk Phos - alkaline phosphatase Ape - adenomatosis polyposis coli APCDD1 - adenomatosis polyposis coli down-regulated 1 AS - Ambras syndrome P-Gal - p-galactosidase B2m - beta-2-microglobulin BAC - bacterial artificial chromosome BCC - basal cell carcinoma BGLAP - osteocalcin BMP - bone morphogenetic protein bp - base pair(s) BrdU - 5-bromo-2-deoxyuridine Calbl - calbindin-28K CCD - cleidocranial dysplasia CD - campomelic dysplasia CDSN - comeodesmosin CGH - comparative genomic hybridization CGHT - congenital generalized hypertrichosis terminalis with gingival hyperplasia ChIP - chromatin immunoprecipitation cKO - conditional knockout CNV - copy number variation CoM - collagen type I Col8a1 - procollagen, type VIM, alpha 1 Cre - Cre recombinase CtBP - C-terminal binding protein CtnnM - p-catenin DAPI - 4',6-diamidino-2-phenylindole DC - dermal condensation Den - decorin Derp7- dermal papilla derived protein 7 (also known as Tmem45a, p19.5, M32486) Dkk - dickkopf homolog DLX3 - distal-less homeobox 3 Dmkn - dermokine DP - dermal papilla v

DSG4 - desmoglein 4 E14.0 - 14.0 embryonic days post coitus Eda - ectodysplasin-A Edar - ectodysplasin-A receptor EDC - epidermal differentiation complex EGF - epidermal growth factor EGFP - enhanced green fluorescent protein Eif3s3 - eukaryotic translation initiation factor 3 ELISA - enzyme-linked immunosorbent assay EMSA - electrophoretic mobility shift assay Epyc - epiphycan EXT1 - exostosin 1 FACS - fluorescence-activated cell sorting FGF - fibroblast growth factor FISH - fluorescence in situ hybridization GFP - green fluorescent protein Gli - GLI family zinc finger GLI3A - activator form of the GLI3 transcription factor HA - hemagglutinin HaCaT - human adult low-calcium high-temperature keratinocytes Hd - hypodactyly HEK - human embryonic kidney cells HF- hair follicle HFG - hand-foot-genital syndrome HG - hair germ HLTS - hypotrichosis-lymphedema-telangiectasia syndrome Hox - homeobox HPE - holoprosencephaly HR - hairless HSPG - heparan-sulphate proteoglycan HTC - hypertrichosis universalis congenita IgG - immunoglobulin G Ihh - Indian hedgehog Ik - Ikaros IP - immunoprecipitation IRES - internal ribosome entry site IRS - inner root sheath K - keratin protein kb - kilo base pairs KCNJ16- potassium inwardly-rectifying channel J16 KFG - keratinocyte growth factor (also known as FGF7) Koa - koala Krt- keratin gene Krtdap - keratinocyte differentiation-associated protein Lc8a - dynein light chain 8, type a LEF1 - lymphoid enhancer binding factor 1 VI

Lhx2 - LIM homeobox 2 LIPH - lipase H Lum - lumican MAP2K6 - mitogen-activated protein kinase kinase 6 Mb - mega base pairs miRNA - microRNA NCBI - National Center for Biotechnology Information neo - neomycin resistance NFATd - nuclear factor of activated T-cells OMIM - Online Mendelian Inheritance in Man ORS - outer root sheath p - p (probability)-value P1 - postnatal day 1 P2RY5 - purinergic receptor P2Y, G-protein coupled, 5 PCAD - P-cadherin PCP - planar cell polarity PCR - polymerase chain reaction PGK - phosphoglycerine kinase PPD2 - preaxial Polydactyly II Prdml - PR domain containing 1, with ZNF domain PRS - Pierre Robin sequence PSA - prostate-specific antigen PSMAD - phospho-Smad Ptch - patched homolog Pthrp - parathyroid hormone-related protein qRT-PCR - quantitative real-time polymerase chain reaction Ra - ragged Rnf4 - ring finger protein 4 RUNX2 - runt-related transcription factor 2 Sbsn - suprabasin SD - standard deviation SG - sebaceous gland Shh - sonic hedgehog Smo - smoothened homolog Smugl - single-stranded selective monofunctional uracil SNP - single nucleotide polymorphism Sox - SRY (sex determining region Y)-box Sp1 - trans-acting transcription factor 1 SSC - stratified epithelium secreted peptides complex Stat3 - signal transducer and activator of transcription TA - transit amplifying Tcf - T cell factor Tg - transgenic Trie - tenascin C TRPS - trichorhinophalangeal syndrome TRPS1 -trichorhinophalangeal syndrome I vn

TUNEL - terminal deoxynucleotidyl transferase dUTP nick end labeling UBC9 - ubiquitin conjugating enzyme 9 UTR - untranslated region WB - Western blot WHN - winged-helix nude Wif1 - Wnt inhibitory factor 1 Wnt - wingless-related MMTV integration site Y10 - Wilms tumor 1 yeast artificial chromosome Vl l l

Manuscript Contributions Chapter II (manuscripts #1 and #2) The microarray hybridization analysis was a collaboration between our lab and the lab of Dr. Colin Jahoda. Dr. Jahoda microdissected the murine epidermal and dermal samples, and Hisham Bazzi and I prepared the samples for hybridization. Hisham performed clustering analyses of the samples and generated lists of differentially expressed transcripts. Together, Hisham and I proceeded to group the transcripts based on their biological function and/or known genetic interactions, and validate the dynamic expression patterns of selected transcripts. In addition to my role in the original microarray hybridization analysis discussed above, I validated the expression of numerous transcripts highlighted in the manuscripts by semi-quantitative reverse-transcriptase PCR and quantitative real-time PCR (qRT-PCR). Furthermore, I performed whole mount in situ hybridization analysis with a Dkk4 probe at multiple timepoints during murine embryogenesis and cloned the 2.5 kb Dkk4 promoter for use in reporter gene assays. Chapter III (manuscripts #3 and #4) For these studies, I validated the expression of multiple FGF pathway transcripts in the epidermis and dermis during hair follicle morphogenesis by qRT-PCR. Furthermore, I helped establish the skin organ culture system in our laboratory, and performed whole mount in situ hybridization analyses with Ctnnbl and GH1 probes on the organ cultured embryonic skin samples. Chapter IV (manuscript #5) In addition to my role in the original microarray hybridization analysis discussed above, I validated the expression of Trpsl in the epidermis and dermis by qRT-PCR. I also performed whole mount in situ hybridization analysis with a Trpsl probe at multiple timepoints during murine embryogenesis, and examined the expression of Trpsl protein by immunofluorescence analysis in mouse, rat and human follicles with a polyclonal Trpsl antibody. Chapter VII (manuscript #6) For this study, I compared the expression levels of TRPS1 and additional transcripts spanning human chromosome 8q23-8q24 between an Ambras syndrome patient and an unaffected parent by qRT-PCR. Similarly, I determined the levels of Trpsl expression in the whisker pads and dorsal skin of wild-type, Koa/+ and Koa/Koa mice at multiple timepoints from E14.0 to P3 by qRT-PCR, and examined the expression of Trpsl transcripts and protein in Koa mutant embryos by whole mount in situ hybridization and immunofluorescence analyses, respectively. I additionally performed histological analysis of the vibrissae and pelage follicles of Koa/+ and Koa/Koa mice during embryogenesis and late postnatal morphogenesis, and identified an altered configuration of transcription factor binding sites at the Koa proximal inversion breakpoint. IX

Acknowledgments I would like to acknowledge a number of individuals who have been instrumental in contributing to my thesis work and my development as a research scientist. First and foremost, I would like to thank my mentor, Dr. Angela Christiano. She has been extremely generous with advice and guidance, while allowing me to independently pursue novel lines of questioning. Above and beyond imparting her extensive scientific expertise, she has taught me how to better write about and discuss my ideas and results, which are valuable skills that I cannot thank her enough for. Her enthusiasm and dedication have created an ideal environment to study science. I also thank all members of the Christiano lab, both past and present, for their scientific insight, assistance and friendship. I would especially like to thank Dr. Yutaka Shimomura, who taught me a great deal about molecular biology and patiently answered so many of my questions. His work ethic and generosity made him a true role model for me in the lab. I also acknowledge Hisham Bazzi and Marija Tadin-Strapps, who contributed to the microarray and hypertrichosis projects, respectively. I would also like to thank Ming Zhang for his superb technical skills and tireless work to keep the lab well-stocked and running smoothly. I also thank Hyunmi Kim, Alex Casta, Lynn Petukhova, Carolina Cela, Claire Higgins and Courtney Luke for stimulating scientific discussions and friendship. Working alongside each of these individuals every day was a real pleasure. x

I would like to thank the members of my thesis committee, Drs. Colin Jahoda, Gerard Karsenty, David Owens and Virginia Papaioannou. I would especially like to thank Ginny, who graciously served as my second reader, and Gerard, both of whom served on my TRAC committee for a number of years and whose insight and advice have been invaluable in helping my thesis project to evolve. I extend my gratitude to my qualifier examiners, Drs. Frank Costantini, Eric Schon and Dorothy Warburton, as well as Drs. Schon and Debra Wolgemuth and the members of their labs for welcoming me as a rotation student. I am additionally grateful to the administrative staff in both the Genetics and Development and Dermatology BSRG offices. I thank Dr. Colin Jahoda and his entire lab at the University of Durham, especially Dr. Gavin Richardson, for continued scientific collaboration through which I learned a great deal about hair biology, and further, in their roles as hosts and friends. A number of investigators have generously provided reagents or protocols. Within Columbia University, Dr. Kathryn Calame and her student Erna Magnusdottir shared the Prdml antibody and immunohistochemistry protocol, and Dr. Thomas Jessell provided the Lhx2 antibody. Additionally, Dr. Ed Laufer provided the shh'oxp-ires-nLacZ-'oxP mice and Dr. Dorothy Warburton contributed to the characterization of the Ambras syndrome patient chromosomal rearrangements. Outside of Columbia University, Dr. Ramesh Shivdasani of the Dana- Farber Cancer Institute, Harvard Medical School generously shared the Trpsl xi

antibody and Trps1Agt mice, and Dr. Anthony Oro of Stanford University provided the HA-tagged mGN2 and hGLI3 constructs. Drs. Fritz Baumeister at the Universitat Munchen, Stefano Cianfarani at Tor Vergata University and Lionel van Maldergem at the Universite de Liege provided the Ambras syndrome patient samples. Additionally, Drs. John Sundberg and Yun You at Jackson Laboratories contributed to the characterization of the Koa hypertrichosis phenotype. Finally, I thank my wonderful family and friends. In particular, my dear friend Maggie O'Meara, whose humor and support have seen me through the highs and lows of graduate school, and my good friends and running partners, Andrew Goldsmith and Sean West. I would especially like to thank my Grandmother, Dolores Tyler, my uncle and aunt, Tom and Mauve Tyler, my cousin, Margaret Hotaling, my sister and my parents for their unconditional love and encouragement. xn

Dedication To my parents, who have always encouraged me to pursue my dreams and whose faith in my abilities has given me the confidence to do so. Their love and support throughout this process and every endeavor I have undertaken in my life has meant the world to me. Xl l l

1 Chapter I General introduction

2 1. Ectodermal appendage formation The various ectodermal appendages found in nature have evolved over time to allow organisms to better adapt to their environment. These include hair, feathers, scales, nails, teeth, beaks, horns and a wide array of eccrine glands (Chuong et al., 2002). While these appendages vary greatly in their shape and function, they share several common developmental features. Ectodermal appendage formation begins during embryogenesis through a series of interactions between the epithelia and adjacent mesenchyme. Ectodermal organogenesis is typically initiated by signals from the mesenchyme that specify the regional identity of the epithelia. The epithelial tissues in turn proliferate and differentiate to form the specified appendage (Chuong, 1998; Chuong et al., 2002; Pispa and Thesleff, 2003). Classic tissue recombination experiments performed more than 30 years ago between different regions of the body or different species demonstrated that the mesenchyme regulates the pattern and shape of appendages in the epithelia (Dhouailly and Sengel, 1975), yet the molecular mechanisms underlying these phenomena have only recently begun to emerge. The hair follicle is an ectodermal appendage unique to mammals that serves a wide array of functions, including thermoregulation, sensation and communication. The hair follicle undergoes continuous cycling between periods of growth and regression by recapitulating developmental processes first employed during embryogenesis, making it an ideal model to study epithelial- mesenchymal interactions and stem cell biology. The mechanisms by which the

3 diverse cells types of the hair follicle arise through interactions of a simple epithelium and underlying mesenchyme and the contribution of specified progenitor cells to the processes of growth and differentiation have been the driving questions behind my doctoral research. 2. Regional specificity of hair types in the mouse The murine pelage coat (body fur) is composed of four different hair types that are patterned at various timepoints during embryogenesis. Primary guard hair follicles begin to develop at approximately embryonic day 14.0 (E14.0). These large, straight fibers will eventually comprise 2-10 percent of the pelage coat. Secondary hair follicles include awl, auchene and zig-zag types and commence development at E15.5-E17.0. Awl hair fibers are straight, roughly half the length of guard fibers, and are the second-most prevalent hair type in the pelage. Auchene hairs are similar to awl hairs, with the addition of a single bend in the fiber. Finally, zig-zag hairs have multiple bends in the fiber and comprise approximately 70 percent of the pelage coat (Sundberg, 1994). Additionally, there are six types of vibrissae follicles in the mouse: supra orbital, post-orbital, mystacial, post-oral, infraramal and ulnar (Danforth, 1925). Mystacial vibrissae (whisker) follicles are the most abundant type and begin to develop at approximately E12.0. The 42 mystacial vibrissae of mice grow in a stereotyped, bilaterally symmetric arrangement on both sides of the face. These follicles are wider in diameter (caliber) than pelage follicles and surrounded by heavily innervated, large blood sinuses that are associated with the papillar

4 muscle and contribute to nutritional vascularization (Danforth, 1925; Wright, 1965; Ibrahim and Wright, 1975; Fundin et al., 1997). Mystacial vibrissae follicles serve critical sensory functions through their association with cortical barrels in the brain; each vibrissa is correlated to a discrete barrel in the somatosensory cortex of the mouse (Van der Loos and Woolsey, 1973). While vibrissae follicle development has not been studied as extensively as that of the pelage, and the mechanisms underlying their regional specificity are not completely understood, these follicle types share a number of similarities in the morphogenetic processes and signaling events governing their formation. 3. Hair follicle morphogenesis Hair follicle morphogenesis involves a series of interactions between the mesenchyme and overlying surface epithelia. In the mouse, this process begins at approximately E14.0 in the case of primary pelage follicles of the trunk. A signal arising in the dermis leads to the formation of regularly-spaced epithelial thickenings, known as placodes (Fig 1). A reciprocal signal from the placode instructs the underlying mesenchymal cells to cluster, forming the dermal condensate. A second dermal signal arising from this condensate directs the follicular epithelium to proliferate and invade the dermis. The resulting epithelial hair peg will eventually surround the dermal condensate, which will develop into the mesenchymal signaling center of the hair follicle, the dermal papilla (Fig 1). The highly proliferative epithelial cells at the leading edge of the invaginating follicle lie in direct contact with the dermal condensate and give rise to the hair

matrix cells. Continued proliferation and differentiation of the hair matrix cells lead first to the formation of the various layers of the inner root sheath (IRS) (Henle's layer, Huxley's later and IRS cuticle) and subsequently to the production of the multiple layers of the hair shaft (hair cuticle, cortex and in some cases, medulla). Both the IRS and hair shaft are encased within the outer root sheath (ORS) (Fig 1), which is contiguous with the basal layer of the interfollicular epidermis (Hardy, 1992). Initiation of follicular cycling Sebaceous gland Gub S rArractor hai r M\ Pili muscle Dermal Bulge papilla New hair _ germ U Figure 1. Hair follicle morphogenesis and postnatal cycling. (Fuchs 2007)

6 Recent insights into the signaling molecules that mediate the cross-talk between mesenchyme and epithelia during hair follicle morphogenesis have revealed key pathways required for this process in mice (Millar, 2002; Schmidt- Ullrich and Paus, 2005). Several studies have demonstrated that Wnt signaling is active during early follicle formation and promotes the hair placode fate. Expression of a p-galactosidase reporter gene responsive to the Wnt effector complex LEF1/p-catenin (TOPGAL) was detected in the epithelial placodes and dermal condensates of transgenic mice, indicative of active Wnt signaling in both compartments of the developing hair follicle (DasGupta and Fuchs, 1999). Overexpression of stabilized p-catenin or targeted deletion of the Wnt pathway inhibitor Ape in the mouse epidermis resulted in the development of ectopic hair follicles in the interfollicular epidermis (Gat et al., 1998; Kuraguchi et al., 2006). Furthermore, mutation of endogenous epithelial p-catenin to a constitutively active form led to the premature induction of enlarged hair follicle placodes that were incapable of invaginating into the underlying dermis. The entire epidermis in these mutant mice eventually adopted a hair follicle fate at the expense of interfollicular epidermis, demonstrating that Wnt signaling in the epithelium is sufficient to induce hair follicle formation (Zhang et al., 2008). Conversely, targeted deletion of p-catenin or overexpression of the Wnt inhibitor Dkk1 in the epidermis completely inhibited hair placode formation, while Lef1'A mice have a drastic reduction in hair follicle number (Huelsken et al., 2001; Andl et al., 2002; van Genderen et al., 1994).

7 In contrast to the requirement for Wnt signaling in promoting placode fate, members of the bone morphogenetic protein (BMP) family inhibit placode development. Bmp2 is expressed in the epithelial placode, while Bmp4 and the secreted BMP antagonist Noggin are expressed in the dermal condensate (Botchkarev et al., 1999). Noggin''' mice developed fewer hair follicles than their wild-type counterparts and exhibited defects in the progression of morphogenesis, while ectopic expression of Noggin through implantation of Noggin-soaked beads into organ cultures of embryonic mouse skin resulted in an increase in follicle number and accelerated follicle morphogenesis (Botchkarev et al., 1999). Additionally, tumor necrosis factor-related ectodysplasin (Eda) and its receptor Edar have been demonstrated to play a role in follicle initiation (Ferguson et al., 1997; Srivastava et al., 1997; Headon and Overbeek, 1999) in primary, and not secondary, hair follicles (Laurikkala et al., 2002), underscoring the fact that while many factors involved in hair follicle initiation and morphogenesis are conserved throughout the body, there are some differences in the specification of the various hair types. Eda is expressed ubiquitously throughout the epithelium (Srivastava et al., 1997; Mikkola et al., 1999), while Edar localizes to the placode at the onset of hair follicle morphogenesis (Headon and Overbeek, 1999). Mice harboring mutations in Eda and Edar displayed a reduction in hair follicle number (Ferguson etal., 1997; Srivastava etal., 1997; Headon and Overbeek, 1999).

8 While not required for the initiation of hair follicle development, the secreted morphogen Sonic hedgehog (Shh) plays a key role in regulating the proliferation and downgrowth of the follicular epithelium during morphogenesis and in the formation of the dermal papilla. Shh is expressed in the epithelium (Iseki et al., 1996), while its receptor and downstream transducers, encoded by Ptchl, GH1 and GH2, respectively, are expressed in the epithelial hair peg and dermal condensate, consistent with a role for Shh signaling in both compartments of the early hair follicle (St-Jacques et al., 1998). Pelage follicles of Shh''' and GH2'A mice arrest shortly after induction and form only rudimentary dermal condensates (St-Jacques et al., 1998; Chiang et al., 1999; Karlsson et al., 1999; Mill et al., 2003). Interestingly, expression of both Shh and Ptchl was lost in follicles of mice lacking epithelial p-catenin, placing Wnt signaling upstream of Shh during hair follicle morphogenesis (Huelsken et al., 2001). 4. The hair follicle cycle At the completion of hair follicle morphogenesis within approximately one week after birth, the mouse pelage follicle enters the postnatal hair follicle cycle. The follicle will undergo phases of regression and regeneration, producing a new hair fiber during each of these cycles. This postnatal hair cycle consists of three highly stereotyped stages: anagen (the growth phase), catagen (the regression phase), and telogen (the resting phase) (Dry, 1926) (Fig 1). While hair follicle cycling in the dorsal skin of mice proceeds in a synchronous fashion for the first

two hair cycles (Fig 2), more complex, mosaic patterns emerge over time (Plikus and Chuong, 2008). B — Telogen—f (- 2 weeks) -Anagen — 1-3 weeks) - Calagcn • (- 2 days) _U_- Telogen- t/3 CO s a> CO o f 3 .2 ™ 4» 60 .O c 60 a a < 3 * af •- - t - 10 11 12 13 trt H 1 16 19 28 35 42 49 Figure 2. A. Schematic representation of the morphology and length of the hair follicle during the postnatal hair follicle cycle. The approximate duration of each phase is indicated in brackets. B. Timeline of the hair follicle cycle during the first 14 weeks after birth. (Adapted from Muller-Rover et al., 2001) Once a mature hair follicle has been formed at the completion of morphogenesis, the follicle enters catagen, a stage characterized by extensive apoptosis of the matrix cells at the base of the follicle. The lower two-thirds of the follicle degenerate into an epithelial strand and regress (Fig 1). The dermal papilla, which remains associated with the follicular epithelium throughout the hair cycle, migrates to the base of the permanent, upper portion of the hair follicle (Straile et al., 1961; Chase et al., 1951; Chase, 1954). A reservoir of slow-

10 cycling, multipotent stem cells resides in a portion of the ORS at the base of this permanent part of the follicle in a compartment known as the "bulge" (Cotsarelis et al., 1990; Taylor et al., 2000; Oshima et al., 2001; Morris et al., 2004; Blanpain et al., 2004) (Fig 1). After the telogen resting phase, these stem cells become activated to generate a new follicle (Cotsarelis et al., 1990; Taylor et al., 2000; Blanpain et al., 2004; Tumbar et al., 2004). It has been proposed that stem cells from the bulge migrate along the ORS to the base of the follicle, giving rise to the transit-amplifying matrix cells (Oshima et al., 2001). The secondary hair germ, a specialized population located between the base of the bulge and the dermal papilla during telogen (Fig 1), has also been proposed to contribute to the regenerating follicle during anagen (Chase et al., 1951). While these cells are derived from the bulge at the end of catagen (Ito et al., 2004; Greco et al., 2009), they constitute a biochemically distinct population (Ito et al., 2002; Greco et al., 2009) that becomes activated prior to the bulge cells, albeit with less long-term proliferative potential (Greco et al., 2009). Similar to morphogenesis, signaling between the epidermal and dermal compartments of the hair follicle is once again required for anagen initiation and the subsequent proliferation and differentiation of the follicular epithelium (Oliver and Jahoda, 1988). Specifically, interactions between the mesenchyme-derived dermal papilla and the epithelial matrix cells result in growth of the hair shaft during anagen (Cotsarelis et al., 1990). A number of molecules are preferentially expressed in the bulge and have been shown to regulate bulge stem cell activity (Trempus et al., 2003; Blanpain

11 et al., 2004; Morris et al., 2004; Tumbar et al., 2004), including the Wnt effector Tcf3 (Merrill et al., 2001; Nguyen et al., 2006), the LIM homeobox transcription factor Lhx2 (Rhee et al., 2006), the nuclear factor of activated T cells family transcription factor NFATd (Horsley et al., 2008), and the SRY box-containing transcription factor Sox9 (Nowak et al., 2008). Interestingly, two of these factors, Sox9 and Lhx2, are also expressed in the secondary hair germ during early anagen and have been shown to regulate early follicle progenitors (Nowak et al., 2008; Rhee et al., 2006). Expression of Sox9 and Lhx2 begins at the placode stage of hair follicle morphogenesis (Vidal et al., 2005; Rhee et al., 2006). Sox9 is required for early stem cell specification (Nowak et al., 2008), while Lhx2 is necessary for hair follicle stem cell maintenance (Rhee et al., 2006). Hair follicles of mice with conditional ablation of Sox9 in the epidermis during embryogenesis, or skin grafts from Lhx2 null embryos revealed decreased hair follicle stem cell quiescence (Nowak et al., 2008; Rhee et al., 2006). The mechanisms by which early hair follicle progenitors are specified and postnatal hair follicle stem cells are activated to generate all the epithelial cell layers of the mature hair follicle are not completely understood. However, many of the signaling pathways employed during hair follicle morphogenesis are also active during postnatal anagen initiation (Millar 2002; Schmidt-Ullrich and Paus, 2005). Dysregulation of these signaling events have been shown to lead to numerous hair defects in mice and humans, ranging from hypertrichosis to hypotrichosis.

Full document contains 277 pages
Abstract: Hair follicle morphogenesis involves a series of reciprocal epithelial-mesenchymal interactions that drive the growth and differentiation of the developing follicle. The goal of this thesis was to address the question of how a simple epithelium and underlying mesenchyme interact to generate the diverse cell types of the hair follicle. My thesis focused on identifying novel transcriptional regulators of hair follicle morphogenesis and characterizing the expression and function of such molecules in both mice and humans. I initially used microarray hybridization analysis to generate a global transcriptional profile of murine skin and hair follicle morphogenesis. Through separate examination of the epithelial and mesenchymal compartments, we identified novel secreted and membrane-bound factors regulating either epidermal placode formation, including members of the stratified epithelium secreted peptides complex, the Wnt inhibitor Dickkopf 4 and the potential Hoxa13 target Derp7 , or the reciprocal condensation of the underlying dermis, such as Syndecan-1 and members of the epidermal and keratinocyte growth factor families. To further refine the signaling molecules that coordinately control the regulation of both compartments, I focused on the zinc-finger transcription factor Trps1. I demonstrated that Trps1 localizes to the nuclei of dermal papillae cells and the highly proliferative epithelial cells of anagen hair follicles, indicating a role for Trps1 in promoting growth of the follicle. To identify downstream target genes of Trps1, I performed microarray hybridization analysis comparing expression patterns in wild-type versus Trps1Δgt/Δgt embryonic whisker pads. I established that Trps1 directly activates a number of secreted Wnt antagonists and transcription factors in the vibrissa follicle, including the hair follicle stem cell regulators Lhx2 and Sox9 . Upon analysis of the vibrissa follicle defects in Trps1Δgt/Δgt embryos, I uncovered a transcriptional hierarchy including Trps1, Shh and Sox9 that controls the specification of early hair follicle progenitors and their subsequent proliferation. Finally, I used comparative genomics and morphological analyses to characterize a position effect on TRPS1 associated with hypertrichosis in both humans and mice. I demonstrated that TRPS1 is subject to complex tissue- and temporal-specific regulation, as underexpression of the gene can result in both hypotrichosis and hypertrichosis, underscoring its critical role in hair follicle development.