W1 AN77G
`
`1995
`v.11
`0.91 —————— *"SEO: SROOBSSBQ
`
`TI: ANNUAL REVIEW OF CELL AND
`DEVELOPMENTAL BIOL 08/08/96
`
`
`
`Page 1 0f 35
`
`Kaken Exhibit 2062
`
`Acrux V. Kaken
`
`IPR2017-00190
`
`

`

`TX 4—159—264
`
`lllllllllllllll[Illflljjlflllfllljlljlllllilllllllllll
`M
`"
`FE -
`ANNUAL REVIEW OF
`CELL AND
`DEVELOPMENTAL
`BIOLOGY
`
`VOLUME 11, 1995
`
`JAMES A. SPUDICH, Editor
`
`Stanford University School of Medicine
`
`JOHN GERHART, Associate Editor
`
`University of California, Berkeley
`
`STEVEN L. MCKNIGHT, Associate Editor
`
`TULARIK, South San Francisco
`
`RANDY SCHEKMAN, Associate Editor
`
`University of California, Berkeley
`
`ANNUAL REVIEWS INC.
`Page 2 0f
`
`4139 EL GAMING \myma'tfifiltmflflag
`:1. the NLM and may be
`Sub-fiat USC'ooooright Laws
`
`PALO ALTO. CALIFORNIA 94303-0139
`
`

`

`l
`
`I
`
`ANNUAL REVIEWS INC.
`
`Palo Alto, California, USA
`
`COPYRIGHT © 1995 BY ANNUAL REVIEWS th., PALO ALTO, CALIFORNIA, USA. ALL
`
`RIGHTS RESERVED. The appearance of the code at the bottom of the first page of an
`
`article in this serial indicates the copyright owner's consent that copies of the
`
`article may be made for personal or internal use, or for the personal or intemai use
`
`of specific clients. This consent is given on the condition, however,
`
`that the copier
`
`pay the stated per-copy fee of $5.00 per article through the Copyright Clearance
`
`Center, Inc. (222 Rosewood Drive, Danvers, MA 01923) for copying beyond that
`
`permitted by Sections 107 or 108 of the US Copyright Law. The per~copy fee of
`
`$5.00 per article also applies to the copying, under the stated conditions, of articles
`
`published in any Annual Review serial before January 1, 1978. Individual readers,
`
`and nonprofit libraries acting for them, are permitted to make a single copy of an
`
`article without charge for use in research or teaching. This consent does not extend
`
`to other kinds of copying, such as copying for general distribution, for advertising
`
`or promotional purposes, for creating new collective works, or for resale. For such
`
`uses, written permission is required. Write to Permissions Dept., Annual Reviews
`
`Inc., 4139 El Camino Way, PO. Box 10139, Palo Alto, CA 94303-0139 USA.
`
`International Standard Serial Number: 1081—0706
`
`International Standard Book Number: 0—8243—311 1-7
`
`Annual Review and publication titles are registered trademarks of Annual Reviews Inc.
`
`® The paper used in this publication meets the minimum requirements of
`
`American National Standard for Information Sciences—uPermanence of Paper
`
`for Printed Library Materials, ANSI 239.48-1984.
`
`Annual Reviews Inc. and the Editors of its publications assume no responsibility for the
`
`statements expressed by the contributors to this Review.
`
`Typesetting by Kachina Typesetting Inc., Tempe, Arizona; John Olson. President;
`
`Toni Starr, Typesetting Coordinator; and by the Annual Reviews Inc. Editorial Staff
`
`PRINTED AND BOUND IN THE UNITED STATES OF AMERICA
`
`Page 3 of 35
`
`

`

`Alum. Rev. Cell Dev. Biol. I995. 11323—53
`
`Copyright © 1995 by Animal Reviews Inc. All rights reserved
`
`KERATINS AND THE SKIN
`
`EZaine Fuchs
`
`Howard Hughes Medical Institute, Department of Molecular Genetics and Cell
`Biology, The University of Chicago. 5841 South Maryland Avenue, Chicago.
`Illinois 6063’?
`
`KEY WORDS:
`
`keratin filaments. genetic disease, epidermis, hair follicles, multigene family
`
`
`
`CONTENTS
`
`SKIN AND ITS PROGRAMS OF EPITHELIAL DIFFERENTIATION . . . . . . . . . . . ..
`
`124
`
`PATTERNS OF KERATIN EXPRESSION IN EMBRYONIC AND ADULT
`MAMMALIAN SKIN . . . .
`.
`.
`. . . . . . . .
`.
`.
`.
`. . . . .
`.
`. . . . .
`.
`.
`. . . . .
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`. ..
`
`REGULATION OF KERATIN GENE EXPRESSION IN THE SKIN . .
`
`.
`
`. .. .. .. .. ..
`
`ASSEMBLY AND STRUCTURE OF KERATIN FILAMENTS .
`
`.
`
`.
`
`. . .
`
`. . .
`
`.
`
`.
`
`.
`
`. . .
`
`.
`
`.
`
`.
`
`.
`
`ASSOCIATIONS BETWEEN KERATIN IFS AND OTHER PROTEINS AND
`STRUCTURES: FORMING A CYTOPLASMIC IF NETWORK IN
`EPIDERM IS AND HAIR . .
`.
`.
`.
`.
`.
`.
`.
`.
`. . . . . . .
`.
`. . . . .
`. . . . .
`.
`.
`.
`.
`.
`.
`. . . . . . .
`
`.
`. . . . .
`KERATIN MUTATIONS AND BLISTERING HUMAN SKIN DISORDERS .
`.
`.
`.
`.
`.
`.
`Epidermolysr's Buliosa Simplex .
`.
`.
`. .
`.
`.
`.
`.
`.
`.
`. .
`.
`.
`.
`.
`. .
`. .
`.
`.
`. .
`. . . .
`. .
`.
`.
`.
`.
`. .
`.
`.
`.
`.
`. .
`Epidermolyric Hyperkerarosi: . .
`. . .
`.
`. . . .
`.
`.
`. . . . .
`.
`. . .
`. . .
`.
`.
`.
`.
`.
`. . . . .
`. . .
`.
`. .
`.
`.
`. .
`Epidermal Nevr' of the EH Type .
`.
`.
`.
`.
`.
`. .
`. .
`.
`.
`.
`.
`. .
`.
`.
`.
`.
`. .
`. .
`.
`.
`. .
`.
`.
`.
`.
`.
`.
`. .
`I’ahnopianrar Keraroderma and Paehyonyehia Congem'm . . .
`.
`.
`. . . . .
`.
`.
`.
`. . . . .
`.
`. .
`KERATIN IFS FUNCTION TO IMPART MECHANICAL INTEGRITY TO CELLS:
`POSSIBLE INSIGHTS INTO ADDITIONAL GENETIC DISORDERS OF
`KERATIN .
`.
`.
`. .
`.
`.
`. .
`. .
`. . . . .
`. . .
`.
`.
`. . . . .
`.
`. . . . .
`.
`.
`.
`. . . . . . . . . . .
`.
`. . . . .
`.
`
`CONCLUSIONS AND PERSPECTIVES . . . . . .
`
`.
`
`. . . . .
`
`.
`
`. . . . . .
`
`.
`
`.
`
`. . . . .
`
`.
`
`.
`
`.
`
`. . .
`
`.
`
`.
`
`.
`
`.
`
`.
`
`128
`
`130
`
`135
`
`133
`
`I40
`I40
`144
`144
`14S
`
`146
`
`I46
`
`ABSTRACT
`
`Keratins are the major structural proteins of the vertebrate epidermis and its
`appendages, constituting up to 85% of a fully differentiated keratinocyte.
`Together with aetin microfilaments and microtubules, keratin filaments make
`up the cytoskeletons of vertebrate epithelial cells. Traced as far back in the
`evolutionary kingdom as mollusks, keratins belong to the superfamily of in-
`termediate filament (IF) proteins that form oc-helical coiled-coil dimers which
`associate laterally and end—to-end to form 10-nm diameter filaments. The
`evolutionary transition betWecn organisms bearing an exoskeleton and those
`with an endoskeleton seemed to cause considerable change in keratin. Keratins
`expanded from a single gene to a multigene family. Of the ~60 IF genes in
`
`123
`
`Page 4 Of35
`
`1081—0706/951’1115—0123$05.00
`
`

`

`124
`
`FUCHS
`
`the human genome, half encode keratins, and at least 18 of these are expressed
`in skin. Vertebrate keratins are subdivided into two sequence types (I and II)
`that are typically coexpressed as specific pairs with complex expression pat-
`terns. The filament-forming capacity of a pair is dependent upon its intrinsic
`ability to self-assemble into coiled-coil heterodimers, a feature not required of
`the invertebrate keratins (Weber et a1 1988). Approximately 20,000 heterodi-
`mers of type I and type II keratins assemble into an IF. Mutations that perturb
`keratin filament assembly in vitro can cause blistering human skin disorders
`in vivo. From studies of these diseases, an important function of keratins has
`been unraveled. These filaments impart mechanical strength to a keratinocyte,
`without which the cell becomes fragile and prone to rupturing upon physical
`stress. In this review, studies on the pattern of expression, structure, and
`function of skin keratins are summarized, and new insights into the functions
`of these proteins and their involvement in human disease are postulated.
`
`SKIN AND ITS PROGRAMS OF EPITHELIAL
`DIFFERENTIATION
`
`The single-layered embryonic ectoderm of mammals receives mesenchymal
`cues that specify its programs of differentiation. Early cues influence epidermal
`cell fate, and later signals influence the ectodermal cell to become epidermis
`vs hair follicle. These later cues happen shortly after stratification, where
`condensates of Specialized mesenchyme, referred to as dermal papilla (anlage)
`cells. form in a dotted pattern beneath the embryonic basal layer (Figure I).
`Where contact is made, basal cells of the epidermis differentiate downward to
`craft what will ultimately be the hair follicles of the mammalian skin. In the
`absence of these mesenchymaI-epithelial interactions. basal cells conunit to
`an epidermal cell fate.
`At the single-layer stage, eetodermal proliferation occurs lateral 1y, with the
`mitotic plane perpendicular to the embryo surface (Figure 2). Upon stratifica—
`tion, mitotic activity occurs in all layers, with the mitotic plane often parallel
`
`
`
`Figure I The choica between epidermis and hair follicles during embryonic development in the
`skin. Shortly after stratification, condensates of specialized mescnchyme, or dermal papilla anlage.
`assemble beneath the embryonic basal
`layer of the skin. These condensates provide as—yet-
`undetermined external cues that stimulate basal cells to migrate downward and form a primary hair
`germ. This morphegenic process eventually gives rise to an adult hair follicle.
`
`Page 5 of 35
`
`

`

`KERATINS AND THE SKIN
`
`125
`
`mitosis basally
`single layered epithelium
`
`terminal differentiation suprahasally
`
`mitosis suprabasally
`stratification but no differentiation
`
`mitosis basally
`
`Figure 2 Cell division during embyronic development in the skin. At the single-layer epithelial
`stage, ectodermal proliferation occurs laterally, with the mitotic plane perpendicular to the embryo
`surface. The mitotic plane soon shifts 90°, causing stratification. Suprabasal mitoses are necessary
`during the time that the skin surface is rapidly expanding. Later, the mitotic plane reverts back,
`allowing dividing cells to move only laterally. At this stage, signs of differentiation appear in the
`suprabasal layers.
`
`to the embryo surface (Figure 2). Suprabasal mitoses occur during the time
`that the skin surface undergoes rapid expansion. Later, as suprabasal cells
`display morphological signs of differentiation, mitotic activity becomes re—
`stricted to a single layer of basal cells, with the mitotic plane reverting to a
`perpendicular orientation. This pattern persists in postnatal epidermis.
`In the adult, a basal epidermal cell responds to an as-yet-unidentified trigger
`of terminal differentiation. When it ceases to divide and begins its journey to
`the skin surface, it alters its adhesive properties, evoking changes that are likely
`to be central to the control of the differentiative program (for a recent review,
`see Watt et al 1993). In transit, the cell undergoes a series of morphological
`and biochemical changes culminating in the production of dead, flattened,
`enucleated squames, which are sloughed from the surface, continually replaced
`by inner cells differentiating outward (Figure 3). The trek takes 2 to 4 weeks
`in humans and continues throughout the life of the individual.
`In the adult hair follicle, differentiation is far more complex than in the
`epidermis (Figure 4; Hardy I992). The hair is surrounded by two sheaths, an
`outer root sheath (ORS), continuous with the epidermis but thought to have
`its own compartment of stem cells (Rochat ct al 1994), and an inner root sheath
`
`Page 6 of 35
`
`

`

`126
`
`FUCHS
`
`
`
`
`Splnous Layers:
`Keratln filaments
`made of K1 and K10
`
`Basal Layers:
`Kerntin filaments
`made 01 K5 and K14
`
`-- Basement Membrane
`
`Dermil
`
`F'
`
`c
`
`l
`
`I
`
`.
`
`n
`
`u
`
`s
`
`.
`
`te‘f;::;d'germlnaldifferentiation in adult epidermis. The diagram illustrates the four stages of
`b
`I
`l
`l erenttation and the pattern of keratin expression. All cells above the innermost. i.e.
`
`333 . ayer of the epidermis are considered to be suprabasal.
`
`(IRS), whose cells are derived from the same precursors as the hair shaft. The
`((31138 18 composed of multiple layers, with the outermost layer being the least
`1 ferentiated (Coulombe et a1 1989). In the lower part of the ORS, cells move
`upward and inward as they differentiate, whereas in the upper part of the
`lf)(])llicle, movement 'is largely inward, and differentiation more closely resem-
`es that of the epidermis. The IRS is composed of three layers: the outer
`(1612:: layer, the Huxley’s layer, and the cuticle of the IRS. These three layers
`theg“first;C1,113 5:1: Liliipesrhpptrtior;1 of the follicle..Another layer of cuticle forms
`surface (Figure 4).
`a
`an remains With it as it breaks through the skin
`
`Page 7 of 35
`
`

`

`
`
`
`KERATINS AND THE SKIN
`
`127
`
`I Epidermis
`spinous layers K1, K10
`basal layer -— WKS, K14
`
`fi—l r3
`,‘ '
`
`— Sebaeeous
`Gland
`
`Arrector
`
`_
`
`’3
`
`‘
`
`Muscle
`
`7-- :H ‘-
`
`Bulge -—fi——% K5, K14
`
`Isthmus
`
`
`
`—~ Outer Root Sheath
`
`Hgirr wee—k -—---
`ulicle
`Cortex
`Medulla
`
`inner rool sheath:
`Henle‘s layer
`Huxley's layer
`Cuticle of inner root shealh
`
`("1 K5' K” 'hmugmm
`K6, K16 inner layers
`- Ha. Hb Keratins
`HS. HGT lFAP Keratins
`
`-
`
`-~ 7
`
`K1. K10
`
`
`
`‘ Marix cells *‘
`
`—— no major keratin network
`
`Dermal papilla
`
`vimenlin
`
`Figure 4 Terminal differentiation in the hair follicle. In the growth phase of the hair cycle, cells
`from the dermal papilla interact with proliferating matrix cells. Under an unidentified trigger, matrix
`cells cease to divide, begin to migrate upward, and commit to at least five concentric rings of
`differentiated states: The inner root sheath (IRS), consisting of Henle's layer, Huxley's layer, and
`outer cuticle, guides the hair shaft as it emerges from the immature cortex cells. The shaft consists
`of the inner cuticle and medulla. The outer root sheath (ORS) is contiguous with the basal epidermal
`layer and has a proliferating compartment distinct from matrix cells. These cells move upward and
`inward as they grow and differentiate. The bulge (see text) is a possible silo for stem cells. The
`keratins expressed in different cells of the hair follicle are indicated according to the nomenclature
`of Moi] et al (1982).
`
`The cells of the IRS and the hair shaft are thought to arise from upward
`modes of differentiation that are controlled by the adult dermal papilla cells,
`which maintain a condensate at the base of the follicle. The precursors of the
`hair shaft and IRS are matrix cells, which are relatively undifferentiated epi-
`thelial cells that surround the dermal papilla to form the hair bulb. In their
`mitotically active state, matrix cells maintain close association with the dermal
`
`papilla. As they move upward and away from this compartment, they cease to
`divide and begin to differentiate. These postmitotic cortical cells then give rise
`to the hair shaft as transcriptional activity comes to a halt.
`
`Page 8 of 35
`
`

`

`128
`
`FUCHS
`
`Hairs follow rhythmic periods of growth and quiescence. In anagen, matrix
`cells are highly mitotically active, and the differentiating hair shaft moves
`Upward at about 0.3-0.4 mm/day in humans (Kobori & Montagna 1976).
`Matrix cells are inactive in catagen, a period where the hair follicle degenerates
`and regresses. This is followed by telogen, the period of rest in the hair cycle.
`In humans, the resting period is variable: relatively short for scalp hairs, and
`longer for body hairs. The hair cycle is controlled in part by FGFS, a member
`of the fibroblast growth factor (FGF) family. Mice homozygous for a null
`FGFS mutation have an extended cycle, resulting in the production of unusually
`long hairs (Hebert et a] 1994).
`Throughout an individual’s life, the epidermis and hair follicles must main-
`tain a balance of dividing and differentiating cells. Given the continuous
`renewal programs and ability to respond to injury, it is not surprising that these
`structures have reservoirs of cells capable of generating tremendous prolifera-
`tion. The pepulation of stem cells in the epidermis is likely to reside within
`the basal layer itself, and as judged by analysis of newborn human foreskin
`keratinocyte cultures, there is one clonal subtype, holoclones, whose cells
`possess extraordinary proliferative potential (>100 doublings) (Barrandon &
`Green 1987, Jones & Watt 1993, Jones et a1 1995). Holoclones thus have the
`capacity to generate enough cells from a single clone to completely cover an
`adult human (Rochat et al 1994 and references therein). In the follicle, matrix
`cells are able to amplify as long as they maintain contact with dermal papilla
`condensates. These cells could be the source of stem cells for the IRS, cuticle,
`and hair shaft. However, follicle stem cells have also been hypothesized to
`reside in the bulge, a region at the midpoint of the follicle where the arrector
`pili muscle attaches (Cotsarelis et al 1990). This region contains 95% of the
`keratinocyte colony-forming cells isolated from rat vibrissae, and these cells
`are slow cycling, features generally ascribed to stem cells (reviewed in Rocha;
`et al 1994). The location is attractive because the bulge effectively resides
`outside the hair follicle, which periodically degenerates. Cyclic stimulation of
`new stem cells in the bulge could then take place as the hair regresses upward,
`perhaps bringing dermal papilla cells with it (Cotsarelis et a1 1990). The nature
`of the epithelial-mesenchymal interactions that control the complex differen~
`tiation programs in the follicle are not yet known, although keratinocyte growth
`factor (KGF), another member of the FGF family, has been implicated (Finch
`et a1 1989, Guo et al 1993).
`
`PATTERNS OF KERATIN EXPRESSION IN EMBRYONIC
`AND ADULT MAMMALIAN SKIN
`
`The epidermis and its appendages devote the majority of their protein synthe-
`srzmg machinery to making keratins. Figure 5 provides a schematic of a
`
`Page 9 of 35
`
`T}! i: mzf'nrizll mm: rrxmine‘l
`
`

`

`KERATINS AND THE SKIN
`
`129
`
`Mr(x10'3)
`
`7.4
`
`6.4
`
`5.4
`
`pKi
`
`Figure 5 Two-dimensional gel profile of keratins expressed in the skin. The schematic illustrates
`the electrophoretic mobilities and isoelectric points of the various keratins found in the skin at
`different stages of development and differentiation. Keratins are expressed as specific pairs of type
`I and type 11 proteins. which form obligatory heteropolymers. Each pair is designated by like
`symbols, and the patterns of keratin expression are described in the text. The smaller sizes of like
`symbols indicate minor isoelectric variants of the same keratin (Expression patterns are according
`to Moll et al 1982.)
`
`two-dimensional gel, showing the type, size, and approximate isoelectric points
`of human skin keratins, which are divided into types I and II based on isoelec-
`tric point and sequence. Figures 3 and 4 indicate the pattern of expression of
`the major keratins of adult mammalian epidermis and hair follicles. K8 and
`K18 are the first skin keratins expressed, coincident with the emergence of a
`single-layered ectoderm. These keratins are typically associated with adult
`simple epithelial tissues and are not characteristic of adult stratified squamous
`epithelia. K5 and K14 are then induced (E95 in mice) in a defined pattern
`that is influenced by mesenchymc (Byrne et al 1994). As judged by expression
`of 21 K5 promoter driven B-galactosidase transgene in mice, it is not until E14.5
`that the entire surface ectoderm expresses these genes. In embryo and adult
`mice, K5 and K14 mRNAs seem to be restricted to cells that maintain their
`proliferative capacity (Byrne ct al 1994 and references therein). A minor type
`I keratin, K15, is also expressed in basal keratinocytes (Lloyd et a1 1995). As
`basal cells differentiate in adult skin, they downregulate expression of K5, K14,
`
`Page 10 of 35
`
`This material was {opted
`
`

`

`130
`
`FUCHS
`
`and K15 and induce new sets of differentiation-specific keratins (Fuchs &
`Green 1980, M011 et a1 1982, Sun et a] 1984, Lloyd et al 1995). Most body
`regions express K1 and K10 suprabasally, along with a second type I keratin,
`K11. K2e is also quite broadly expressed suprabasally, but its production is
`delayed relative to K] and K10 (Collin et a1 1992). K9 is confined to suprabasal
`palmo and plantar skin (Fuchs & Green 1980, Langbein et al 1993). K6 and
`K16 are unusual in that they are induced suprabasally during wound healing,
`upon retinoic acid treatment, or in hyperproliferative diseases of the skin,
`including various skin cancers (Sun et a1 1984). K6, K16, and K17 are also
`induced when skin keratinocytes are cultured in vitro. The fianctional signifi-
`cance of. the multiplicity of keratins has not yet been resolved; however, the
`assembly properties of keratins differ as do their differential interactions with
`lF-associated proteins.
`Keratin expression in the hair follicle is as complex as its differentiation
`programs suggest (Figure 4). In the ORS, as cells differentiate and move
`inward, they maintain mitotic activity, upregulate expression of K5 and K14,
`and also initiate expression of K6 and K16. In the IRS, K1 and K10 are
`expressed. The IF network in the matrix cells of the hair follicle has been
`difficult to discern either ultrastructurally or biochemically. K19 and possibly
`very low levels of K14 have been the only keratins identified in the matrix,
`and whether they produce a bona fide IF network remains to be determined.
`As these cells differentiate, however, they initiate abundant expression of Ha
`and Hb keratins, which are exclusive to the cortex of the follicle (Lynch et a]
`1986, Heid et a1 1986, M011 et a1 1988, Kopan & Fuchs 1989, Rogers & Powell
`1993). In addition to Ha and Hb keratins, there are high sulfur (HS) and high
`glycine-tyrosine (HGT) proteins, which are also called keratins. These proteins
`share no sequence or structural homology to the IF keratins, but rather they
`are small IF-associated proteins that seal together the IFS into large macrofi-
`brillar structures. Expression of the Ha/Hb, HS, and HGT keratins in the cortex
`is complex, and not all of the genes encoding these proteins are induced
`simultaneously (Rogers & Powell 1993).
`
`REGULATION OF KERATIN GENE EXPRESSION IN THE
`SKIN
`
`All keratins seem to be encoded by separate genes, and as yet, no evidence of
`differential splicing has been reported. Human K14 was the first keratin whose
`cDNA was cloned (Hanukoglu & Fuchs 1982) and whose gene was sequenced
`(Marchuk et a1 1984). Following these studies, a flood of additional epidermal
`and hair follicle cDNA and gene sequences were reported (for review, see
`Fuchs & Weber 1994). As judged by nuclear run-off experiments, expression
`of skin keratin mRNAs is largely regulated at the transcriptional level (Stell-
`Page 11 of35
`
`

`

`KERATINS AND THE SKIN
`
`131
`
`mach et a1 1991). Transgenic mouse studies have revealed that for many skin
`keratin genes, the sequences involved in regulating their expression reside in
`the 5' upstream sequences of the genes (Vassar et a1 1989, Bailleu] et a1 1990,
`Powell et a1 1992, Takahashi et a1 1994). A knowledge of the major transcrip-
`tion factors controlling skin keratin gene expression is of central importance
`in the quest to elucidate the molecular mechanisms underlying keratinocyte
`specificity and epidermal and follicle differentiation.
`Among the candidates implicated in orchestrating epidermal gene expres-
`sion is the sequence 5’-GCCTGCAGGC—3’, first identified 5’ from the TATA
`box of vertebrate K14 genes (Leask et al 1990, Snape et al 1990). For the K14
`gene, the sequence acts in synergy with a distal element to regulate transcrip-
`tion in keratinocytes (Leask et al 1990). Epidermal nuclear extracts contain a
`protein(s) that binds to this sequence and cross-reacts with antibodies against
`the transcription factor AP2 (Leask et a1 1990, Snape et a] 1991). APZ-binding
`sites have now been found in the promoters of most epidermal- and some
`hair—specific genes, and where tested, they are functionally important for gene
`expression (Leask et a] 1990, Snape et a1 1990, Byrne & Fuchs 1993 and
`references within).
`During embryogenesis, multiple AP2 mRNAs are synthesized in the skin
`(Snape et a1 1991 and references within; Buettner et a1 1993), and an AP2
`CRNA recognizing these forms hybridizes to tissues of ectodermal and neural
`crest lineages (Mitchell et al 1991). As judged by in situ hybridizations of
`whole mouse embryos, the patterns of AP2 mRNA on the embryo surface are
`strikingly similar to and precede by about 1 day those patterns of K5 and K14
`mRNAs (Byrne et al 1994). In contrast, a dominant negative inhibitor of AP2,
`referred to as AP2B (Buettner et al 1993), is not in the ectoderm of developing
`skin (Byrne et al 1994). Thus APZ mRNAs are positioned temporally and
`spatially to play a role in controlling gene expression during skin development
`and differentiation.
`How important is AP2 in controlling epidermal—specific gene expression?
`In at least one case, AP2 appears to be central. Thus recombinant AP2A (the
`active form of AP2) can impart to cultured hepatocytes the ability to express
`K5 and K14 promoter-driven transgenes when cotransfected into these cells
`(Byrne et a1 1994). APZB counteracts these effects. However, the mere pres—
`ence of AP2 does not mandate the expression of endogenous K5 and K14
`genes in cultured cells or in animals (Leask et al 1990), and thus AP2 cannot
`be the key to unleashing the cascade of keratinocyte—specific genes. Neverthe—
`less,
`these findings do suggest that
`the AP2 forms present in embryonic
`ectoderm are likely to play a role in KS and K14 activation in vivo, and in
`some cells, the absence of these factors may be sufficient to maintain these
`genes in an inactive state.
`Members of the API family of transcription factors are also found in the
`
`Page 12 of 35
`
`Thai: m=+r=ri=l WIPES rnniafl
`
`

`

`132
`
`FUCHS
`
`epidermis ofskin. These includejunB (Wilkinson et a1 1989) and cfos (Smeyne
`et al 1992 and references therein) in postnatal skin, andfosB in effl‘bryotnfi‘:
`epidermis (Redemann-Fibi et a1 1991). JunB and cfos resrde prrmarr I): lnAPI
`differentiating layers of epidermis, implicating these members of 1 6 m
`family in regulating differentiation-Specific functions in skin. ConSIStent W.l
`this notion is the finding that promoters active in terminally qlffcfem‘at'lng
`keratinocytes contain APl sites (Lu et al 1994 and references Within, see a so
`Casatorres et a] 1994). Among the best evidence that these srtes are funCtlonal
`comes from studies on the AP] site 3' to the K1 gene, which .haS been
`implicated in mediating the calcium-inducible, differentiation-specrftc expres-
`sion of this gene (Lu et al 1994).
`.
`A number of other factors have been localized to the epidermis, although
`their roles in controlling keratinocyte-Specific gene expression. have not been
`fully elucidated. One of these proteins is a zinc finger proteln, basonuclln,
`which is expressed in the basal layer of epidermis (Tseng & Green 1994)-
`Basonuclin is interesting in that it persists in cells that have withdrawn from
`the cell cycle, but it is absent in terminally differentiating cells. BaSOHUCIin
`has been postulated to be a key to the switch controlling the balance between
`growth and differentiation in keratinocytes.
`Several POU-specific proteins have also been found in epidermis. A new
`class II POU sequence, called Skn-la but related or identical to Oct-11, was
`detected in epidermis and hair follicles (Anderson et al 1993). Skn-la can
`specifically upregulate expression of a human KlO-luciferase reporter gene,
`suggesting a role in terminal differentiation (Anderson et a] 1993). XLPOU],
`a Xenopus class III POU protein, is also expressed in adult skin (Agarwal &
`Sato 1991). XLPOUI shares ~90% sequence homology with human Oct-6,
`which has been cloned from epidermal keratinocytes (Faus et a1 1994). Oct-6
`is intriguing in that it can act in both positive and negative fashions, depending
`on the gene and the tissue. In basal keratinocytes, Oct—6 suppresses keratin
`gene expression, and it may possibly have a role in downregulation of these
`genes during terminal differentiation (Faus et al 1994).
`Researchers are just beginning to decipher the sequences and transcription
`factors involved in controlling the expression of the hair-specific keratin genes.
`A number of potential regulatory motifs have been identified on the basis of
`sequence comparisons (reviewed in Rogers & Powell 1993). One of these sites,
`5'-C'ITI‘GAAGA-3’, referred to as the HK—l motif, was detected in four
`published hair keratin promoters. When the HK—l sequence motif was seen to
`overlap with a sequence identical to the motif of a known pair of lymphoid
`enhancer factors, LEF—l and TCF-l, it became apparent that all thirteen pub-
`lished hair keratin promoters contained LEF—l/I‘CF—l binding sites (Zhou et
`al 1995). Further studies showed that LEF—l plays a functional role in hair-
`specific morphogenesis and keratin gene expression. LEF—l is expressed early
`
`Page 13 of35
`
`Tth.l'lH-V5"1r:—|I ....,—n— _._..:....t
`
`

`

`KERATINS AND THE SKIN
`
`133
`
`.
`
`a
`
`in the development of embryonic skin, and it is produced in cells that first
`Induce ha1r-_specrfic gene expression (Zhou et a1 1995). Interestingly, LEF—l
`knockout mice have no whiskers and few hair follicles, further strengthening
`thelrgiggpn that LEF—l is involved in follicle morphogenesis (van Genderen et
`LEF-l belongs to the high-mobility group (HMG) family of proteins that
`are not conventional transcription factors, but rather act by bending DNA and
`altering chromatin to create a structure conducive for conventional transcrip-
`tion factor binding (Giese et a1 1992). It has been postulated that in hair
`follicles, LEF-l is involved in creating a DNA structure conducive to initiation,
`but not necessarily maintenance, of gene expression, and that additional factors
`are likely to be necessary for hair-specific gene BXpl‘ession (Zhou et al 1995).
`one POSSiblc candidate iS a Zinc finger protein recently shown to be the product
`0f the hairless gene in mice (Cachon-Gonzalez et a1 1994). Like the LEF-I
`knOCROUt mice, hairless mice Show a marked reduction and malformation of
`hair follicles, suggesting that they may be involved in the same pathway.
`What governs transcriptional regulation early in development? Taking cues
`from other systems, researchers have begun to investigate skin-specific ex-
`pression of the Hex class of transcription factors, known to specify positional
`information in segmentally derived structures, as well as in appendages such
`as limbs. Although epidermis is not a segmental structure, it overlies and is
`influenced by a dermis, which is, in part, segmentally derived. During em—
`bryogenesis, patterning and regionalization of the epidermis is likely to be
`under the influence of both dermal and epidermal Hox genes. An example of
`the former may be the segmental expression of the Moxl homeoprotein in
`E85 dermamyotome in mouse embryos, in a pattern that parallels E9.5 ex~
`pression of basal epidermal keratin genes in the overlying ectoderm (Byrne et
`al 1994 and references therein). At later developmental stages, mesodermal
`gradients of Hox proteins have been found in epidermal appendages of other
`vertebrates (reviewed in Chuong 1993), which suggests that positional infor-
`mation directing skin appendage patterning is determined by dermal Hox
`genes, perhaps in a fashion analogous to their mode of action in body append-
`
`ages.
`Hox genes are also expressed temporally in the ectodermal component of
`skin. Early in development, Hox 2 gene family members are expressed in the
`branchial arches of the head (Hunt et al 1991), and later the pattern of transient
`expression of additional members of this family suggests a role in specifying
`the differentiation status of epidermis (Mathews et a1 1993). This notion is
`strengthened by the recent discovery of two additional differentiation-activated
`homeoproteins, Xdll-3/Dlx—3 (Morasso et al 1993) and HOXC4 (Rieger et al
`1994).
`It is well established that retinoids influence patterns of homeobox gene
`Page 14 of 35
`
`

`

`134
`
`FUCHS
`
`ave been shown
`
`ms of epithelial differen-
`expression, and thus it is not surprising that progra
`ids. Retinoic acid at high
`.
`tiation in the skin are intricately controlled by retino
`nhibitory effects on
`(10‘6 M) concentration has long been known to have .1
`mRNAs (Fuchs &
`epidermal differentiation and on the expression of keratin
`Green 1981). Recently, lower concentrations of retinoic acid h
`f
`'
`to have a beneficial effect on the differentiation process. The effects 0 reti-
`noids on keratin expression are transcriptionally regulated (StellmaCh 3t 31
`1991, Tomic-Canic et a1 1992, Lu et al 1994), and keratinocytes express a
`number of retinoid receptors. These include retinoic acid receptors on and “Y
`(RAROL and RARy) and RXRoc (Viallet & Dhouailly 1994 and references
`therein). RARY and RXROL are expressed in neonatal skin and 1n only a few
`other organs, whereas RAROL is more broadly expressed. Developmentally,
`RARy mRNAs appear in epidermis prior to RXROL mRNAS, and In neonatal
`skin, RAR'y mRNAs are most abundant in the keratinizing layers of epidermis.
`The control of gene transcription by RARs and RXRs is complex. involving
`a multitude of both indirect and direct mechanisms (see Kastner et a1 1993).
`RARs can heterodimerize with thyroid hormone receptors and With RXRS, and
`at least some of these interactions change the DNA affinity and activity of
`RARs. RARs and/or RXRs can bind to thyroid response elements, retinoic
`acid response elements (RAREs), and retinoid X response elements (RXREs),
`and the repertoire of complex interactions is further expanded by the capacity
`of RARs and RXRs to interact with AP] proteins. The direct binding of RARs
`to epidermal keratin genes has not yet been demonstrated, although RA-me—
`diated biochemical changes in differentiation do appear to involve DNA se-
`quences in keratin genes that are responsive to retinoids (Tomic-Canic