Proc. Natl. Acad. Sci. USAVol. 90, pp. 10270-10274, November 1993Biochemistry

Tissue- and stratum-specific expression of the human involucrinpromoter in transgenic mice

(epidermis/keratinocyte)

JOSEPH M. CARROLL*, KATHRYN M. ALBERSt, JONATHAN A. GARLICK, ROBIN HARRINGTON,AND LORNE B. TAICHMANSDepartment of Oral Biology and Pathology, School of Dental Medicine, State University of New York, Stony Brook, NY 11794

Communicated by William J. Lennarz, August 2, 1993 (received for review May 20, 1993)

ABSTRACT Involucrin is a marker of keratinocyte termi-nal differentiation and is expressed only in the suprabasallayers of stratified squamous epithelium. In a previous studywith various cell types in culture, we noted that expression ofthe putative human involucrin promoter was keratinocytespecific. To determine if this promoter is sufficient to directexpression to the suprabasal cells of stratified squamous epi-thelia in vivo, we have now generated transgenic mouse linesharboring the involucrin promoter sequences linked to a (3ga-lactosidase reporter gene. In the resulting lines, -galacto-sidase was expressed in the suprabasal compartment of strat-ified squamous epithelia and in hair follicles in a tissue-specificmanner. In the palate, distinct vertical stacks of 3galacto-sidase-expressing cells were present, suggesting movement ofclonally derived ceUs through the epithelium. The involucringene has a single intron upstream of the translational start site,and removal of this intron did not affect tissue- or stratum-specific expression. These results show that the 3.7-kb involu-crin upstream sequences contain all the information necessaryfor a high level of tissue- and stratum-specific expression.

Keratinocytes in stratified squamous epithelia exhibit spe-cific patterns of gene expression as they undergo a progres-sive terminal differentiation during movement from the basallayer to the surface (1, 2). One gene expressed in suprabasalcells of this tissue encodes involucrin (3). Involucrin is acomponent of the cornified envelope of stratified squamousepithelia (3-5) and is a substrate, as are other envelopeprecursors, for transglutaminase-mediated enzymatic cross-linking (for reviews see refs. 6 and 7). In humans, theinvolucrin gene has been shown to consist of a short non-coding exon, a single intron, and a single exon containing theentire coding region (8). Although much is known about thefunction and evolution of involucrin protein (9), little isknown about how the involucrin gene is regulated, other thana recent report of a phorbol 12-myristate 13-acetate-responsive element in the promoter (10).The onset of involucrin protein expression in vitro is

related to the appearance ofinvolucrin RNA in differentiatingkeratinocytes (11). Although involucrin protein is detected atvarious levels of stratification depending on the site of tissueorigin, it is invariably expressed at some point in the spinouslayer (12). Unlike other markers of terminal differentiation,such as the K1/K10 keratins (13) or loricrin (14), involucrinappears to be resistant to the effects of agents which alterkeratinocyte gene expression. For example, involucrin issynthesized in submerged cultures of keratinocytes (3) eventhough in these cultures there is no K1/K10 (15), no loricrin(16), few cornified envelopes, and no stratum corneum. Evenwhen stratification is inhibited by low concentrations of

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

calcium in the medium, involucrin continues to be expressedin a small fraction of cells in the basal layer (17). In raftcultures, where tissue differentiation is improved, treatmentwith retinoic acid suppresses both loricrin and transgluta-minase expression (14, 18), whereas in the intact skin appli-cations of retinoic acid result in an apparent increase ininvolucrin staining (19). In healing epidermis (20), as well aspsoriasis (21, 22), there is gross disruption of keratinocytedifferentiation, yet involucrin expression appears prema-turely but within the confines of the spinous layer. Becauseinvolucrin expression in normal epidermis is likely to beregulated at the transcriptional level (11), analysis of itspromoter is likely to provide insights into why involucrin isnot regulated in the same way as other markers of differen-tiation.

In an in vitro study of the involucrin upstream region wenoted that a 3.7-kb fragment conferred keratinocyte-specificexpression in transient assays (23). To determine ifthis 3.7-kbfragment encoded tissue and stratum specificity, transgenicmouse lines were generated with this construct linked to a,B3galactosidase ((3gal) reporter gene. In the transgenic miceexamined, reporter gene expression was confined to supra-basal cells of stratified squamous epithelia, and interestingpatterns of expression were noted in internal epithelia.

MATERIALS AND METHODSPlasmid Construction for the Transgenes. The construct

used in this study is derived from pNAss,B pL (23) and, asshown in Fig. 1A, includes the 3.7-kb involucrin sequences,a simian virus 40 (SV40) intron, the f(3gal gene, and an SV40poly(A) signal sequence. The HindIII site on the 5' endrepresents a site 2.5 kb upstream of the involucrin transcrip-tional start site. A second construct lacking the involucrinintron was used and is shown in Fig. 1B.

Preparation of Transgenic Mice. The involucrin expressioncassette was isolated on a 0.8% agarose (Boehringer Mann-heim) gel, extracted from the gel by using Genecleanglassmilk (Bio 101) purification, run through an NACS*52Prepac column (Bethesda Research Laboratories), precipi-tated with ethanol, resuspended in Ca- and Mg-free phos-phate-buffered saline (PBS) at a concentration of 5 pg/ml,and microinjected into mouse embryos (strain C3H-B6 Fl;Harlan-Sprague-Dawley). Injections and implantations were

Abbreviations: 3-gal, 3-galactosidase; SV40, simian virus 40; RT,reverse transcriptase; X-Gal, 5-bromo-4-chloro-3-indolyl P3D-galactopyranoside.*Present address: Imperial Cancer Research Fund, KeratinocyteLaboratory, P.O. Box 123, Lincoln's Inn Fields, London WC2A3PX, England.

tPresent address: Department of Pathology, Lucille P. MarkeyCancer Center, University of Kentucky, Lexington, KY 40536.iTo whom reprint requests should be addressed at: Department ofOral Biology and Pathology, Westchester Hall, State University ofNew York, Stony Brook, NY 11794-8702.

10270

Dow

nloa

ded

by g

uest

on

Oct

ober

24,

202

0

Proc. Natl. Acad. Sci. USA 90 (1993) 10271

-- vluvhcrinA , l!2olri!-Promote -- istronm

B clvolveri Prooter =-o 0TAGSol I

ATG Baa

kNSV40 Intron

B gal

SV4O lntro.

FiG. 1. Depictions of transgene constructions. Sal I fragments of involucrin upstream sequences in plasmid pNAssp-pL2 are shown. Theconstruct in A encompasses 3.7 kb of the upstream sequences and contains the involucrin intron, whereas the construct in B lacks the 1188-bpinvolucrin intron. Both transgenes contain the SV40 16S/19S intron and a SV40 poly(A)+ signal.

carried out by using standard protocols (24). For mousescreenings, genomic DNA was isolated from ears of 4-week-old mice and polymerase chain reaction (PCR) analysis wasperformed with (3-gal primers. Primers to the (3-gal (lacZ)gene were chosen to yield a fragment of 680 bp (upstreamprimer, 5'-TGCGTGACTACCTACGGGTAACAGT; down-stream primer, 5'-GATCGACAGATTTGATCCAG-CGATA). DNA-positive lines were then histochemicallyscreened for 5-bromo-4-chloro-3-indolyl (-D-galactoside (X-Gal) expression in the tail epidermis (see X-Gal Histochem-istry below). One founder (368-2) with intense (3-gal stainingwas outbred to generate an F1 line which was used for allsubsequent analysis.Southern Analysis. Genomic DNA isolated from tails was

digested with BamHI, which cuts twice within the transgeneto yield a 4-kb band, and the resultant DNA was electropho-resed on a 1% agarose gel. After blotting to nitrocellulose,filters were incubated with a 635-bp 32P-labeled (-gal probe,washed, and exposed to x-ray film. Copy number wasestimated by comparing band intensities, as measured with alaser densitometer, to known standards of the 4.0-kb (-galfragment, and variations in loading were assessed by reprob-ing the blots for the single-copy (3-actin gene band.RNA Analysis. Total RNA was isolated from the organs of

4- to 10-week-old F1 mice by using RNAzolB reagent (Cinna/Biotecx Laboratories, Houston), followed by digestion withDNase I and precipitation with ethanol. A reverse transcrip-tase (RT)-PCR kit (Perkin-Elmer/Cetus) and random prim-ers were used to amplify 0.5 ,g of total RNA to cDNA. Halfof the reaction mixture was subjected to PCR using 3galprimers and half to PCR using actin primers to determine ifcomparable amounts ofRNA were present in all tissues. The(-gal primers yielded a 680-bp band, and the actin primersyielded a 217-bp band. RNA was mock-PCR amplified (noreverse transcription) to ensure that all traces ofDNA wereremoved.X-Gal Histochemistry. Tissue was harvested from mice and

placed overnight in a solution of 30o sucrose in PBS. Thenext day tissue samples were snap frozen at -70°C in O.C.T.embedding compound (Tissue-Tek, Miles) and 8-,um-thicksections were cut by using a Reichert Histostat Cryostat(model 855) and placed on gelatin-coated slides. Sectionswere postfixed for 10 min in 2% paraformaldehyde, washed,and stained en face with X-Gal (25). After staining in X-Galsolution for 2 hr at 37°C, slides were rinsed and counter-stained by means of the Feulgen reaction (using the Schiffreagent). Nontransgenic mouse tissues were also stained toassess endogenous (3-gal activity.

RESULTS

Generation of Transgenic Mice. Transgenic mice weregenerated with an expression vector containing the involu-crin upstream sequence, an SV40 intron, the (-gal codingregion, and an SV40 poly(A)+ sequence (Fig. 1A and ref. 23).

Three to five weeks after birth, 35 founder mice werescreened for (3-gal sequences and 4 positive mice weredetected. These were further screened for (-gal expressionby using X-Gal histochemical staining on tail skin. Transgenecopy number in the DNA-positive founders was determinedby Southern analysis and was 2-50 copies per cell (Fig. 2).The intensity of X-Gal staining in tail skin did not correlatewith transgene copy number. For example, line 359-2 withapproximately 42 copies per cell had very light X-Gal stain-ing, whereas line 353-6 with approximately 2 copies per cellhad quite intense staining. Line 368-1, which was negative for(3gal DNA, was also negative for X-Gal staining. Mouse line368-2 was chosen for further study as preliminary examina-tion of tail skin showed that staining was more intense andmore uniformly distributed than in the other founder mice.Line 368-2 had approximately 41 copies of the transgene percell. To avoid the possibility ofmosaic expression sometimesseen in founder mice (26), all further analysis ofline 368-2 wasperformed on F1 offspring derived by mating with outbredstrain C3H.

Tissue- and Stratum-Specific Transgene Expression. Todetermine the tissue distribution of transgene expression,total RNA was isolated from various tissues of mouse line368-2 and subjected to RT-PCR amplification. Results areshown in Fig. 3. Mouse (-actin RNA served as a positivecontrol and yielded a PCR band of 217 bp of similar intensityin all tissues. An intense 680-bp (-gal band was observed in

cN

cD

Co

IC)

Co

Cuc)CY)

0

Co

cb pL2Coco

Co) C\J CUIC\J I

Copy Number

FIG. 2. Southern analysis of transgenic DNA from founder mice.Samples (2 ,g) ofgenomicDNA isolated from the ears offive founderlines were digested with BamHI, electrophoresed, and blotted tonitroceliulose. The blot was probed with a 635-bp 32P-labeled ,B-galprobe, washed, and exposed to x-ray film. To calibrate for copynumber, DNA derived from plasmid pNAssf-pL2 was electropho-resed at two concentrations (far right). Blots were stripped andreprobed with a mouse /-actin gene probe (not shown) representinga single-copy gene in each lane. The intensity of the bands wasscanned with a laser densitometer and the intensity was compared toknown copy number amounts of plasmid pNAss,B-pL2 (after adjust-ing for intensity of mouse ,-actin signal). The approximate copynumber of the transgene was then noted in the bottom of the figure.One founder line, 368-1, was negative for (3-gal DNA and is includedhere as a negative control.

SV40pAO \

Sal I

SV4O PA.aiSol I

. I r- T- 11 IM40*0- 1.

a i

Biochemistry: Carroll et al.

Dow

nloa

ded

by g

uest

on

Oct

ober

24,

202

0

Proc. Natl. Acad. Sci. USA 90 (1993)

a 0 'Z. 0i-a

e _

.0

h- .2O0L 0-

c

0

P.

1-

0

0C.)

gal

actin

FIG. 3. RT-PCR analysis of tissue-specific expression patterns intransgenic mice. Samples (0.5 pg) of total RNA isolated from variousmouse tissues of line 368-2 were reverse transcribed randomly andamplified with either p-gal primers or mouse 3-actin primers by 30cycles of PCR. The products were then electrophoresed on a 2%agarose gel and stained with ethidium bromide. The positive controlis RNA isolated from keratinocytes transfected in vitro with thetransgene (see ref. 23).

tissues or organs which have stratified squamous epithelia.However, in brain, heart, and liver, organs which have nostratified squamous epithelia, the 680-bp band was not de-tected. These results indicate that transgene expression wastissue specific.

Histochemical staining of the skin of tail, dorsum, maxilla,ear, penis, and footpad indicated that the transgene wasalways expressed in the suprabasal layer, beginning at somepoint in the spinous layer. This is illustrated for the skin ofthetail and dorsum, where a uniform and intense blue stainingwas seen throughout the suprabasal layers (Fig. 4 A and B).Hair follicles were also noted to stain intensely, and thelocation of staining (data not shown) was similar to that notedfor involucrin protein in human hair follicles (27). Internalorgans with simple epithelia-liver, kidney, and lungs-werenegative for (3-gal expression, as was muscle. The stratifiedsquamous epithelia of internal organs (palate, buccal mucosa,tongue, esophagus, forestomach, and cervix) were positive,as was the stratified urothelium of bladder. Three of these-palate, buccal mucosa, and cervix-are shown in Fig. 4 C, D,and E. X-Gal staining was generally present in the suprabasallayers, although the staining pattern of these internal epitheliawas somewhat less uniform than in the epidermis. In fact, inthe hard palate, stacks of stained cells appeared as a clusterof one or several cells, beginning in the immediate suprabasallayer and extending upwards to the surface as an expandingvertical column of cells (see arrows in Fig. 4C). Theseordered stacks are reminiscent of the vertical columns ofcornified cells seen in normal mouse epidermis (28). Closeexamination of the base of these columns suggests that basalcells exhibit (3-gal staining, but it is not certain if the basallocation of these cells is a result of the plane of tissue section.These histochemical results indicate that the 3.7-kb involu-crin promoter sequence can direct stratum-specific expres-sion in vivo and, when considered along with the RNAexpression pattern (Fig. 3), confirm tissue-specific expres-sion.The involucrin gene has a single intron located between the

transcriptional and translational start sites (8). In transientassays in cultured keratinocytes, removal of this intronresults in a dramatic reduction of expression (23). As intronsin this upstream location site are known to have regulatoryactivity (29), we queried if the involucrin intron was requiredfor tissue- and stratum-specific expression. Several lines ofmice were constructed with 2.5 kb of involucrin upstream

sequences without the involucrin intron (Fig. 1B). Of theselines, line 367-2 (containing approximately 30 copies of thetransgene) exhibited the most intense staining and was bredfor further analysis. It should be noted that an SV40 intronwas present in the upstream region of this construct to ensurethat a splicing event did take place. RT-PCR analysis ofRNAextracted from an F1 derivative ofthe 367-2 founder indicatedthat, as in line 368-2, (-gal RNA was present only in organshaving stratified squamous epithelia (data not shown). His-tochemical analysis of skin and oral mucosa of line 367-2revealed X-Gal staining confined to suprabasal cells of theepithelia (staining of buccal mucosa shown in Fig. 4F).Staining in the epidermis of line 367-2 was patchy andnoticeably less intense compared with staining in line 368-2.It is evident that the involucrin intron is not required forstratum-specific expression. The reduced levels of transgeneexpression in mouse line 367-2 made an analysis of tissuespecificity less meaningful, but when expression was de-tected, either by histochemistry or by RNA analysis (data notshown), it was always in stratified squamous epithelia. Theinvolucrin intron is therefore not essential for tissue-specificexpression, although its presence in the promoter appears toensure high expression in all tissues.

DISCUSSIONTo determine if a 3.7-kb segment of DNA upstream of theinvolucrin coding region was sufficient for tissue- and stra-tum-specific expression in vivo, this segment was linked to a,(3gal reporter gene and used in the construction oftransgenicmice. Expression in the resulting mice was confined to thesuprabasal cells of stratified squamous epithelium, therebydemonstrating its promoter regulatory role in vivo. Theinvolucrin promoter thereforejoins with the human K5, K14,and Ki promoters, as well as the bovine K10 promoter, inbeing correctly expressed with both tissue and stratum spec-ificity in stratified epithelia of transgenic mice (26, 30-32).

In human epidermis, as well as in the epidermis of trans-genic mice, K14 is expressed in the basal layer, while Kl andK10 are expressed in the suprabasal cells (33). The K14promoter has been particularly useful in exploring the patho-logical consequences of expression of a mutant of K14keratin and in identifying human disease correlates of thiskeratin (34). Although a 12-kb fragment of the Kl promotergives correct tissue- and stratum-specific expression in trans-genic mice, it fails to respond as the endogenous mouse Klpromoter to modulators of keratinocyte differentiation, sug-gesting that additional sequences are needed to mediate thecorrect responses (31). As discussed in the introduction,involucrin is constitutively expressed during differentiationand is not subject to modulation in the same way as are othermarkers of differentiation. Analysis of the involucrin trans-gene promoter is therefore likely to lead to clues as to howthese different forms of regulation are achieved.

Involucrin (35) and several other precursors of the corni-fled envelope that are expressed specifically in differentiatingkeratinocytes map to chromosome location 1q21. Theseproteins include loricrin [a precursor of the cornified enve-lope (36)], trichohyalin [an intermediate filament-associatedprotein of hard keratin (37)], and profilaggrin [a filamentaggregating protein (38)]. The clustering of these genes atchromosomal location 1q21 is reminiscent of the globin genecluster on human chromosome 11 (39). The coding regions ofeach of these keratinocyte genes contain a multiple repeatsubunit structure and have no introns. In all cases there is anintron located upstream of the translational start site, and inthe case of profilaggrin, a second intron lies further upstreamin the noncoding region. The similar intron organization andthe genetic linkage of these genes suggest that these intronshave provided some essential function in that location to be

10272 Biochemistry: Caffoll et al.

Dow

nloa

ded

by g

uest

on

Oct

ober

24,

202

0

Biochemistry: Carroll et al.

A

Proc. Natl. Acad. Sci. USA 90 (1993) 10273

B

. . . .~~AN. A;

Fi-

-1. :: *

i..-

d-I

C

6.

.va

ii

I

- I~ ~ ~~~1IA

D_

,¼4' j%; > .p S ..........a....... .....4.. ...

4w*- ,e . A aa,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.6....f. V..C- W . ~~~~~~~~~~~~~~~~~~~... ... . ... .'*3W q4i;,. * Jiz,; ;.M)'. ~~~~~~~~~~. .::.....

* ~ 3*.-tAl

_s *~~~~~~~~~~~~~~~~9_~~~~~~tIt

-.. 4

FE

..i. r. ,ti-

4 ;.;' . >.

*'- 4.tA

-s ._ -/ * t v

-A~~~~~~ ~~~~~s\t

FIG. 4. X-Gal staining (blue) of various tissues of transgenic mice. Staining was confined to suprabasal cells of all stratified squamousepithelia. Hair follicles in skin also stained positively. Tissues from transgenic mouse line 368-2 are shown in A-E. (A) Orthokeratinizingepithelium of the tail cut in cross section. (B) Dorsum of skin. (C) Palate. (D) Buccal mucosa. (E) Cervix. Buccal mucosa from transgenic mouseline 367-2 is shown in F. Line 367-2 was constructed with a transgene lacking the involucrin intron. Tissues were removed from adult micebetween the ages of 12 and 40 weeks. (The bars represent 100 gim.)so preserved during evolution. The results of this studyindicate that the involucrin intron is not essential for tissue-or stratum-specific expression and may serve some otherfunction, possibly related to facilitating a high level of ex-pression in certain tissues. The initial intron in the humankeratin 18 gene is known to possess transcriptional regulatoryactivity through an AP-1 binding site (40). The involucrinintron contains an AP-2-like site, and such sites are requiredfor epidermal-specific expression of keratin 14 (41). The factthat involucrin intron is required for the high level of expres-sion in vitro (23) lends some support to the hypothesis thatthis intron is needed in vivo for high expression.

Stratified squamous epithelium is a tissue that undergoescontinual loss by desquamation at the surface and renewal byreplication of stem cells in the basal layer (42). The progenyofstem cell division, including amplifying cells and terminallydifferentiated cells, are organized into a spatially conservedunit known as the "epidermal proliferation unit" or EPU. Insome regions of the epidermis, where keratinization is com-plete and turnover is slow, vertical columns of cornified andgranular cells are seen and thought to be the progeny of singlestem cells (28). The vertical stacks of labeled cells seen in theoral mucosa beginning in the basal layer and spanning up-wards to the surface provide direct visualization of the

%s,. -4W

:, W.,t. V.

-t4

I4a. ..w. e ..

:s

NI t. I ;.I.

I

i ...`'P.. 0 .l

'01

ilk, *.04#I' ip.bL .. - , - %

Dow

nloa

ded

by g

uest

on

Oct

ober

24,

202

0

Proc. Natl. Acad. Sci. USA 90 (1993)

ordered movement of progeny through the tissue. This alsomarks the first time of which we are aware that an EPU-typeorganization has been seen in oral mucosa. The lack of astratum corneum in this noncornifying tissue has preventedvisualization of ordered columns of cornified cells, as isfound in the epidermis. A similar columnar pattern has beenseen in transgenic mice in which a K14-directed transgenewas expressed in a mosaic fashion in the basal layer of theepidermis and the resulting transgene product was carriedalong in the progeny cells in the overlying spinous layer (26).Because the involucrin promoter is faithfully expressed in

all stratified squamous epithelia and is insensitive to envi-ronmental influences, it might be particularly well suited asa vehicle for keratinocyte-mediated gene therapy (43). Thismight be valuable for potential medical applications, includ-ing drug delivery and correction of epidermal disease states.The fact that expression from the involucrin promoter isconfined to suprabasal spinous and granular cells may alsoallow targeting of the new gene product to specific stratawhile sparing the basal population of cells.

Note Added in Proof. Similar tissue- and differentiation-specificresults have been obtained with the human involucrin promoter intransgenic mice. See Crish et al. (44).

We thank Frankie Davis (University of Kentucky) for expertassistance in constructing transgenic mice and Marcia Simon (StateUniversity of New York at Stony Brook) for many helpful discus-sions. We also thank Peter Brink and Beth Grine (State UniversityofNew York at Stony Brook) for assistance in preparing histologicalmaterial. This research was supported by Grants DE04511 andDC00203 from the National Institutes of Health. J.A.G. is therecipient ofPhysician Scientist Award for Dentists DE00263 from theNational Institute of Dental Research.

1. Watt, F. M. (1989) Curr. Opin. Cell Biol. 1, 1107-1115.2. Fuchs, E. (1990) Curr. Opin. Cell Biol. 2, 1028-1035.3. Rice, R. H. & Green, H. (1979) Cell 18, 681-694.4. Haftek, M., Serre, G., Mils, V. & Thivolet, J. (1991) J.

Histochem. Cytochem. 399, 1531-1538.5. Yaffe, M. B., Murthy, S. & Eckert, R. L. (1993) J. Invest.

Dermatol. 100, 3-9.6. Hohl, D. (1990) Dermatologica 180, 201-211.7. Simon, M. (1993) in The Keratinocyte Handbook, eds. Leigh,

I. M., Watt, F. M. & Lane, E. B. (Cambridge Univ. Press,Cambridge, U.K.), in press.

8. Eckert, R. L. & Green, H. (1986) Cell 46, 583-589.9. Green, H. & Djian, P. (1992) Mol. Biol. Evol. 9, 977-1017.

10. Takahashi, H. & lizuka, H. (1993) J. Invest. Dermatol. 100,10-15.

11. Watt, F. M. & Green, H. (1981) J. Cell Biol. 90, 738-742.12. Banks-Schlegel, S. & Green, H. (1981)J. CellBiol. 90, 732-737.13. Kopan, R., Traska, G. & Fuchs, E. (1987) J. Cell Biol. 105,

427-440.14. Magnaldo, T., Bernerd, F., Asselineau, D. & Darmon, M.

(1992) Differentiation 49, 39-46.15. Fuchs, E. & Green, H. (1981) Cell 25, 617-625.16. Mehrel, T., Hohl, D., Rothnagel, J. A., Longley, M. A., Bund-

man, D., Cheng, C., Lichti, U., Bisher, M. E., Steven, A. C.,

Steinert, P. M., Yuspa, S. H. & Roop, D. R. (1990) Cell 61,1103-1112.

17. Watt, F. M. & Green, H. (1982) Nature (London) 295,434-436.18. Asselineau, D., Bernard, B. A., Bailly, C. & Darmon, M.

(1989) Dev. Biol. 133, 322-335.19. Rosenthal, D. S., Griffiths, C. E. M., Yuspa, S. H., Roop,

D. R. & Voorhees, J. J. (1992) J. Invest. Dermatol. 98, 343-350.

20. Mansbridge, J. N. & Knapp, A. M. (1987) J. Invest. Dermatol.89, 253-263.

21. Bernard, B. A., Reano, A., Darmon, Y. M. & Thivolet, J.(1986) Br. J. Dermatol. 114, 279-283.

22. Bernerd, F., Magnaldo, T. & Darmon, M. (1992) J. Invest.Dermatol. 98, 902-910.

23. Carroll, J. M. & Taichman, L. B. (1992) J. Cell Sci. 103,925-930.

24. Hogan, B., Constantini, F. & Lucy, E. (1986) Manipulating theMouse Embryo: A Laboratory Manual (Cold Spring HarborLab. Press, Plainview, NY).

25. Cepko, C. (1989) Neuromethods 16, 177-219.26. Vassar, R., Rosenberg, M., Ross, S., Tyner, A. & Fuchs, E.

(1989) Proc. Natl. Acad. Sci. USA 86, 1563-1567.27. Hashimoto, T., Inamoto, N., Nakamura, K. & Harada, R.

(1987) Br. J. Dermatol. 117, 325-332.28. Mackenzie, I. C. (1969) Nature (London) 222, 881-882.29. Kozak, M. (1991) J. Cell Biol. 115, 887-903.30. Byrne, C. & Fuchs, E. (1993) Mol. Cell. Biol. 13, 3176-3190.31. Rosenthal, D. S., Steinert, P. M., Chung, S., Huff, C. A.,

Johnson, J., Yuspa, S. H. & Roop, D. R. (1991) Cell GrowthDiffer. 2, 107-113.

32. Bailleul, B., Surani, M. A., White, S., Barton, S. C., Brown,K., Blessing, M., Jorcano, J. & Balmain, A. (1990) Cell 62,697-708.

33. Moll, R., Frnke, W. W., Schiller, D. L., Geiger, B. &Krepler, R. (1982) Cell 31, 11-24.

34. Coulombe, P. A., Hutton, M. E., Vassar, R. & Fuchs, E.(1991) J. Cell Biol. 115, 1661-1674.

35. Simon, M., Phillips, M., Green, H., Stroh, H., Glatt, K.,Bruns, G. & Latt, S. A. (1989) Am. J. Hum. Genet. 45,910-916.

36. Yoneda, K., Hold, D., McBride, 0. W., Wang, M., Cehrs,K. U., Idler, W. W. & Steinert, P. M. (1992) J. Biol. Chem.267, 18060-18066.

37. Lee, S.-C., Wang, M., McBride, 0. W., O'Keefe, E. J., Kim,I.-G. & Steinert, P. M. (1993) J. Invest. Dermatol. 110, 65-68.

38. McKinley-Grant, L. J., Idler, W. W., Bernstein, I. A., Parry,D. A. D., Cannizzaro, L., Croce, C. M., Huebner, K., Lessin,S. R. & Steinert, P. M. (1989) Proc. Natl. Acad. Sci. USA 86,4848-4852.

39. Maniatis, T., Fritsch, E. F., Lauer, J. & Lawn, R. M. (1981)Annu. Rev. Genet. 14, 145-178.

40. Oshima, R. G., Abrams, L. & Kulesh, D. (1990) Genes Dev. 4,835-848.

41. Leask, A., Byrne, C. & Fuchs, E. (1991) Proc. Natl. Acad. Sci.USA 88, 7948-7952.

42. Potten, C. S. (1981) Int. Rev. Cytol. 69, 271-318.43. Carroll, J. M., Fenjves, E. S., Garlick, J. A. & Taichman,

L. B. (1993) in Molecular Biology of the Skin: The Kerati-nocytes, eds. Darmon, M. & Blumenberg, M. (Academic, NewYork), pp. 269-284.

44. Crish, J. F., Howard, J. M., Zaim, T. M., Murthy, S. &Eckert, R. L. (1993) Differentiation 53, 191-200.

10274 Biochemistry: CaffoU et al.

Dow

nloa

ded

by g

uest

on

Oct

ober

24,

202

0