Intrinsic regulation of differentiation markers in human epidermis, hard palate and buccal mucosa
-
Upload
susan-gibbs -
Category
Documents
-
view
212 -
download
0
Transcript of Intrinsic regulation of differentiation markers in human epidermis, hard palate and buccal mucosa
Intrinsic regulation of di�erentiation markers in humanepidermis, hard palate and buccal mucosa
Susan Gibbs*, Maria Ponec
Department of Dermatology, Leiden University Medical Centre, Wassenaarseweg 72, Bld 3, Sylvius Laboratory, 2333 AL Leiden,
The Netherlands
Received 8 December 1998; received in revised form 20 July 1999; accepted 13 August 1999
Abstract
Di�erent epithelia show extensive variation in di�erentiation. Epidermis and epithelium from the hard palate are
both typical examples of orthokeratinized epithelia whereas buccal mucosa is an example of a non-keratinizedepithelium. Each of these tissues can be distinguished morphologically and also by the expression of a number ofstructural proteins. Tissue explants derived from epidermis, hard palate or buccal mucosa were cultured at the air±liquid interface on collagen gels containing human dermal ®broblasts. Reconstructed epithelia that retained many of
the morphological and immunohistochemical characteristics of the original tissue were formed. Cultures derivedfrom epidermis and the hard palate both had a well-de®ned stratum basale, stratum spinosum, stratum granulosumand stratum corneum whereas cultures derived from buccal mucosa had no stratum granulosum or corneum and the
cells retained their nuclei. Signi®cantly more living cell layers were observed in both types of epithelia obtained fromthe mouth than in epidermis. The speci®c localization of proliferation and di�erentiation markers (Ki67, loricrin,involucrin, SPRR2, SPRR3 and keratin 10) closely resembled that of the tissue from which the cultures were
derived. As identical three-dimensional culture models were used here, it is concluded that the di�erences observedbetween these epithelia were due to intrinsic properties of the keratinocytes. # 2000 Elsevier Science Ltd. All rightsreserved.
Keywords: Reconstructed epithelium; Tissue explant
1. Introduction
Epithelia from di�erent regions of the body show
extensive variation in their di�erentiation. The end
product of epithelial di�erentiation is the formation of
a protective barrier against the environment. An epi-
thelium must form a barrier that will limit the loss of
body ¯uids and at the same time it must prevent the
entry of toxic substances, bacteria and viruses.
Di�erent types of epithelia have therefore evolved to
provide the optimal form of protection for their
speci®c location. For example, epidermis is dry, con-
stantly exposed to changing humidities and tempera-
tures lower than 378C whereas oral epithelium is
exposed to heavy abrasion, 100% humidity and mainly
a temperature of 378C. Epidermis and epithelium from
hard palate are both typical examples of orthokerati-
Archives of Oral Biology 45 (2000) 149±158
0003-9969/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0003-9969(99 )00116-8
www.elsevier.com/locate/archoralbio
* Corresponding author. Tel.: +31-71-5271917; fax: +31-
71-5271910.
E-mail address: [email protected] (S.
Gibbs).
Abbreviations: DMEM, Dulbecco's modi®ed Eagle med-
ium; PBS, phosphate-bu�ered saline; SPRR, small proline-
rich protein.
nized epithelia whereas buccal mucosa is an exampleof non-keratinized epithelium. Each of these tissues
can be distinguished morphologically and also by thedi�erences in expression of a number of structural pro-
teins (Brysk et al., 1995; Hohl et al., 1993, 1995;Mackenzie et al., 1991; Smack et al., 1994).
Our aim here was to determine whether, in variousepithelia, the tissue architecture and the expression of a
number of keratinocyte di�erentiation markers (keratin10, loricrin, involucrin, SPRR2, SPRR3) and prolifer-
ation markers (Ki67) can be reproduced in vitro andwhether these proteins are regulated by intrinsic or
extrinsic factors. Therefore, tissue explants derived fromepidermis, hard palate and buccal mucosa were used to
reconstruct epithelia using the skin equivalent modelsystem as described by Saiag et al. (1985). In this model
small, full-thickness punch biopsies are placed epithelialside upwards on to a ®broblast-populated collagen gel
and cultured at the air±liquid interface until the epi-thelial outgrowth covers the gel. With this approach,
extensive mechanical and chemical manipulations withenzymes that result in the death of an important pro-
portion of the cells can be avoided. Furthermore, thereconstructed tissue can be grown from a very small tis-
sue biopsy (2 mm2) whereas traditional culture methodsrequire a larger amount of the original tissue. The
explants are small, easily detectable and can easily beremoved to avoid any ambiguity in the results.
2. Materials and methods
2.1. Morphology and immunohistochemistry
Epidermis was obtained from healthy patientsundergoing surgical corrections; hard palate and buc-cal mucosa were obtained as 4-mm punch biopsiesfrom two healthy volunteers. The biopsies were divided
into two parts: one piece was used for culture (2 mm2)and the other piece was washed in PBS, ®xed in 4%formaldehyde, dehydrated and embedded in para�n.
Sections (5 mm) were cut, depara�nized and rehy-drated in preparation for morphological or immuno-histochemical analysis.
For morphological observations, sections werestained with haematoxylin and eosin. For immunohis-tochemical analysis, the avidin±biotin±peroxidase com-plex method was used essentially as described by the
suppliers (strepABComplex/HRP; DAKO code no. K377). Antigen retrieval for Ki67 was as described byGibbs et al. (1996). Antigen for keratin 10 was
retrieved by incubating the sections for 25 min atroom temperature in 0.01 M sodium citrate, pH 6.0,that had immediately before been preheated to boiling.
After washing in PBS, the sections were incubated for15 min at room temperature in 0.04 mg/ml pepsin in0.2 M HCl.
Fig. 1. Morphology of native tissue. Di�erences in tissue morphology between human epidermis (A), hard palate (B) and buccal
mucosa (C).
S. Gibbs, M. Ponec / Archives of Oral Biology 45 (2000) 149±158150
2.2. Antibodies
Antibodies to human SPRR2 and SPRR3 were agift from Dr. D. Hohl, Lausanne (Hohl et al., 1995);antibodies to human involucrin (SY5) were a gift from
Dr. F. Watt, London (Hudson et al., 1992); antibodiesto human loricrin were a gift from Dr. D. Roop(Mehrel et al., 1990). Anti-Ki67 was purchased from
DAKO (Glostrup, Denmark) and antikeratin 10 fromICN Pharmaceuticals (California, USA).
2.3. Cell culture
Dermal ®broblasts were obtained from healthy
patients undergoing surgical corrections and culturedas described by Ponec et al. (1977). Fibroblast-popu-lated collagen gels (containing 1 � 105 cells/ml) wereprepared as described by Smola et al. (1993).
Reconstructed epithelium was generated by placing atissue biopsy (approx. 2 mm2), with the epithelial sideupwards, on to a ®broblast-populated collagen gel
(approx. 2 cm2). The explants were cultured immedi-ately at the air±liquid interface at 378C and 100%humidity until the epithelium had expanded over the
gel (2±3 weeks). The medium used was a mixture ofDMEM and Ham's F12 (3:1) supplemented with 5%HyClone calf serum, 0.4 mg/ml hydrocortisone, 5 mg/
ml insulin, and 1 mM isoproterenol. Great care wastaken to cut an entire cross-section of the culture for
histological and immunohistochemical analysis. Theexplant was easy to identify due to the presence of adermal matrix.
3. Results
3.1. Morphological characteristics of reconstructedepithelia
Epidermis (Fig. 1A) and hard palate epithelium(Fig. 1B) are both typical examples of orthokeratinized
epithelium. They both displayed a stratum basale, spi-nosum, granulosum and corneum. Nuclei were lostfrom the terminally di�erentiating cells when the cellsreached the stratum corneum. In contrast, epithelium
from buccal mucosa (an example of non-keratinizedepithelium) (Fig. 1C) had no stratum corneum or gran-ulosum. It consisted of three strata (basal, ®lamento-
sum and distendum) and the terminally di�erentiatedcells of the stratum distendum retained their nuclei.Signi®cantly more living cell layers were observed in
both types of epithelia obtained from the mouth thanin the epidermis. This observation correlated with theproliferation status of the di�erent epithelia, where it
Fig. 2. Tissue-speci®c expression of proliferation and di�erentiation markers. Sections derived from epidermis (A), hard palate (B)
and buccal mucosa (C) stained immunohistochemically using antibodies directed against human Ki67, loricrin, involucrin, SPRR2,
SPRR3 and keratin 10.
S. Gibbs, M. Ponec / Archives of Oral Biology 45 (2000) 149±158 151
can be seen that Ki67 was more frequently expressed
in epithelia derived from the mouth than in epidermis
(Fig. 2; Table 1).
When epithelia from epidermis, hard palate and buc-
cal mucosa (Fig. 3) were reconstructed on a ®broblast-
populated collagen gel, the cultured epithelia retained
many of the morphological characteristics of the orig-
inal tissue. Cultures derived from epidermis and hard
palate both had a well-de®ned stratum basale, spino-
sum, granulosum and corneum whereas cultures de-
rived from buccal mucosa had no stratum granulosum
or corneum and the cells retained their nuclei. A thick
upper layer of nuclei containing cells was formed in
cultured buccal mucosa, which presumably accumu-
lates due to the lack of constant abrasion that occurs
in the mouth. Whether or not a stratum ®lamentosum
Fig. 2 (continued)
S. Gibbs, M. Ponec / Archives of Oral Biology 45 (2000) 149±158152
and stratum distendum were formed cannot be deter-
mined without the aid of electron microscopy and isnot within the scope of this communication. As in thecorresponding native epithelia, a signi®cantly larger
number of living cell layers were observed in bothtypes of reconstructed epithelia obtained from themouth than in the epidermis, but the e�ect was slightlyless pronounced than in the native epithelia (Table 1).
The number of living cell layers correlated with theproliferation status of the cultures; Ki67 was more fre-quently expressed in reconstructed epithelia derived
from the mouth than from the epidermis (Fig. 4; Table1).
3.2. Tissue-speci®c expression of di�erentiation markers
3.2.1. EpidermisThe expression of proliferation and di�erentiation
markers was examined in epidermis and epithelia from
the mouth. In the epidermis (as shown in Fig. 2A),
positive immunohistochemical staining for the corni®edenvelope precursors loricrin, involucrin and SPRR2was observed in the stratum granulosum. Keratin 10
was present in all suprabasal cells and SPRR3 wasabsent. As in native epidermis, in reconstructed epider-mis (Fig. 4A) loricrin and SPRR2 were expressed inthe stratum granulosum and keratin 10 in all supraba-
sal layers. In contrast to native epidermis, involucrinwas expressed in all suprabasal layers and SPRR3 wasintermittently and weakly expressed in a very few cells
of the stratum spinosum.
3.2.2. Hard palateIn the hard palate (another example of orthokerati-
nized epithelium) (Fig. 2B), loricrin was weakly
expressed in the stratum granulosum whereas involu-crin and SPRR2 expression extended into the spino-sum. In striking contrast to epidermis, SPRR3 was
Fig. 2 (continued)
S. Gibbs, M. Ponec / Archives of Oral Biology 45 (2000) 149±158 153
strongly expressed in the stratum granulosum andupper spinosum and keratin 10 was expressed only
intermittently in these cell layers. Similarly to thenative epithelium, in reconstructed hard palate (Fig.4B) loricrin was expressed in the stratum granulosum;
involucrin was expressed in all suprabasal layers;SPRR2 and SPRR3 were expressed in the upper spino-sum and the granulosum, and keratin 10 was expressed
intermittently in these two layers.
3.2.3. Buccal mucosa
In buccal mucosa, both in vivo (Fig. 2C) and invitro (Fig. 4C), loricrin and keratin 10 were absentwhereas involucrin, SPRR2 and SPRR3 were expressedin the upper suprabasal layers. The thick upper layer
of nuclei-containing cells formed in cultured buccalmucosa did not stain positively with any of the anti-bodies used.
4. Discussion
Epithelia from the epidermis, hard palate and buccalmucosa represent three distinct types of strati®ed squa-
mous epithelia which have di�erent histological fea-tures and a distinct expression of di�erentiationmarkers (Table 1). Epidermis and hard palate are com-
posed of a stratum basale, spinosum, granulosum andcorneum. Buccal mucosa has the same environment ashard palate (submerged in saliva), but its architecture
is very di�erent. It is composed of a stratum basale, astratum ®lamentosum and a stratum distendum. Thestratum corneum is absent and nuclei are retained in
the terminally di�erentiating cells. The rate of prolifer-ation in both types of oral epithelia is greater than inepidermis, as is illustrated by the larger number of liv-ing cell layers and the higher number of Ki67 positive-
staining cells. The observation that the corni®ed envel-ope precursors are expressed di�erently in these tissues(Table 1) supports the conclusions made by other
researchers (Hohl et al., 1993) that the composition ofthe corni®ed envelope is variable and is dependent onthe body site. It is possible that the composition of the
envelope and the interlinking intermediate ®laments(Steinert and Marekov, 1995) is related to the rigidityand/or permeability of the epithelia, the relative orderof rigidity being epidermis > hard palate > buccal
mucosa and the relative order of permeability beingbuccal mucosa > hard palate > epidermis (Squier etal., 1986). For example, we found that loricrin and
keratin 10 are strongly expressed in the epidermis,weakly expressed in the hard palate and not expressedin buccal mucosa, whereas SPRR3 is strongly
expressed in buccal mucosa and hard palate but isabsent in the epidermis.It is of interest that many of the characteristics that T
able
1
Summary
ofcharacteristics
ofnativeandreconstructed
epidermis,hard
palate
andbuccalmucosa
a
Epidermis
Hard
palate
Buccalmucosa
Native
Reconstructed
Native
Reconstructed
Native
Reconstructed
Tissuearchitecture
SB,SS,SG,SC
SB,SS,SG,SC
SB,SS,SG,SC
SB,SS,SG,SC
SB,SF,SD
SB,SF,SD
Number
oflivingcelllayers
5±8
6±8
20±30
10±20
20±30
10±20
Ki67
SB+
+;SS+
SB+
+SB+
++
;SS(1)+
++
SB+
+;SS(1)+
++
SB+++;SS(1)+
+++
SB+++;SS(1)+
++
Loricrin
SG
strong
SG
strong
SG
weak
SG
weak
Absent
Absent
Involucrin
SG
SS,SG
SS,SG
SS,SG
Suprabasal(u)
Suprabasal
SPRR2
SG
SG
SS(u),SG
SS(u),SG
Suprabasal(u)
Suprabasal(u)
SPRR3
Absent
Interm
ittent/weak
SS(u),SG
SS(u),SG
Suprabasal(u)
Suprabasal(u)
Keratin10
SS,SG
SS,SG
SS(u),SG
SS(u),SG
Absent
Absent
aTissuearchitecture
wasdetermined
withtheaid
oflightmicroscopy.In
nativeepithelia
thenumber
oflivingcelllayerswasestimatedfrom
therangebetweenthetopandbot-
tom
oftherete
ridges;in
reconstructed
epithelia
thenumber
oflivingcelllayerswasestimatedfrom
di�erentregionswithin
theculture.Im
munoreactivityusinganti-K
i67scored
onascale
ofincreasingfrequency
ofnuclearstainingis
shownas+
to+
++
+.Abbreviations:
SB,stratum
basale;SC,stratum
corneum;SD,stratum
distendum;SF,stratum
®lementosum;SG,stratum
granulosum;SS,stratum
spinosum;l,lower
celllayers;u,upper
celllayers.
S. Gibbs, M. Ponec / Archives of Oral Biology 45 (2000) 149±158154
we observed in buccal epithelia are also characteristicof psoriatic epidermis (Bowden et al., 1983; Hohl etal., 1995; Ishida-Yamamoto et al., 1996). For example,
as in buccal mucosa, lesional psoriatic tissue does notlose its nuclei during di�erentiation and the tissue hassigni®cantly more living cell layers than in healthy epi-
dermis. Involucrin and SPRR2 expression extends dee-ply into the stratum spinosum and granulosum,loricrin is not expressed and K10 expression is
decreased. However, unlike buccal mucosa whereSPRR3 is strongly expressed, SPRR3 is not expressed
in psoriatic epidermis. It is therefore possible that inpsoriasis and healthy buccal mucosa a number of com-mon regulatory pathways are shared which are distinct
from those in healthy epidermis.Most of the morphological and immunohistochem-
ical characteristics observed in the epidermis, hard
palate and buccal mucosa were also observed in recon-structed epithelia derived from these tissues (Table 1).As identical culture systems were used to generate the
reconstructed epithelia the di�erences observed can beattributed to di�erences in intrinsic properties of theepithelial cells rather than to environmental factors. It
is highly unlikely that ®broblasts present in the smalltissue specimen (2 mm2) migrated out of the explant
into the collagen gel (2 cm2) and thus contributed tothe speci®c properties observed in the reconstructedepithelia, and furthermore the number of ®broblasts
present in the biopsy is very small when compared tothe number incorporated within the collagen gel(2 � 105 cells). These reasonings are further supported
by the data reported by Saiag et al. (1985), who, usinga similar culture system, were able to disprove that®broblasts are able to migrate out of the explant. They
showed that reconstructed epidermis from psoriaticnon-involved skin was not hyperproliferative whengenerated on a dermal equivalent populated with nor-
mal human ®broblasts whereas hyperproliferation wasobserved when a normal skin biopsy was implanted on
to a gel populated with ®broblasts derived from psor-iatic non-involved skin.Involucrin is the only corni®ed envelope precursor
which showed a major di�erence between its in vivoand in vitro expression and even then this was onlyobserved in the epidermis and not in the two other
types of epithelia. In native epidermis, involucrin isexpressed in the stratum granulosum whereas in vitroit was expressed in all suprabasal layers. At present it
is not clear whether the enhanced expression of involu-crin can be ascribed to conditions used to reconstruct
the epidermis. In both types of reconstructed oralepithelia, involucrin was also expressed in most supra-basal layers but this does correspond to involucrin ex-
pression in vivo in these tissues.It is clear from our observations that the rate of pro-
liferation in oral epithelia is higher than in the epider-
mis and that this is re¯ected in the correspondingreconstructed epithelia (Table 1). However, slightdi�erences between the number of living cell layers in
native and reconstructed tissues were observed forboth types of oral epithelia. The di�erences can beascribed to the presence of deep rete ridges in vivo,
which are absent in vitro. The number of living celllayers at the top of the rete ridges corresponds to thenumber of living cell layers observed in vitro when the
epithelia are reconstructed on ¯at, ®broblast-popu-lated, collagen gels.
Whether the di�erences in the di�erentiation patternof di�erent epithelia are due to intrinsic properties ofkeratinocytes or to in¯uences from subepithelial con-
nective tissue is an important question that has boththeoretical and clinical implications. There is muchvariation in the data reported so far, but in general it
can be concluded that both intrinsic and extrinsic fac-tors are involved in the formation of a speci®c epi-thelium (Bohnert et al., 1986; Brysk et al., 1995;
Doran et al., 1980; Chung et al., 1997; Karring et al.,1975; Kautsky et al., 1995; Lillie et al., 1988;
Fig. 3. Morphology of reconstructed tissue. Reconstructed epithelia derived from human epidermis (A), hard palate (B) and buccal
mucosa (C) retain the morphological characteristics of their native tissue.
S. Gibbs, M. Ponec / Archives of Oral Biology 45 (2000) 149±158 155
Fig. 4. Reconstructed epithelia retain the speci®c expression of proliferation and di�erentiation markers found in their corresponding
native tissue. Sections derived from reconstructed epidermis (A), reconstructed hard palate (B) and reconstructed buccal mucosa (C)
stained immunohistochemically using antibodies directed against human Ki67, loricrin, involucrin, SPRR2, SPRR3 and keratin 10.
Mackenzie and Fusenig, 1983; Mackenzie and Hill,1984; Mackenzie et al., 1993; Ouhayoun et al., 1988).We clearly show that not all of the di�erences
observed between epithelia are regulated by extrinsicfactors (such as neighbouring cell types Ð oral or epi-dermal ®broblasts, contact with saliva, air exposure,
temperature, relative humidity or contact with aspeci®c dermal matrix), but that at least the expressionof loricrin, involucrin, SPRR2, SPRR3, keratin 10 andKi67 is regulated by intrinsic properties of the kerati-
nocytes. The expression of these proteins is most prob-ably partially responsible for the di�erent phenotypesthat we observe in reconstructed epithelia which are
also characteristic of the di�erent epithelia.
Acknowledgements
We would like to thank Mary Verhoeven and
Johanna Kempenaar for their skilful technical assist-ance. This study was supported by a grant from theDutch Ministry of Health and Education PAD 92-16.
References
Bohnert, A., Hornung, J., Mackenzie, I., Fusenig, N., 1986.
Epithelial-mesenchymal interactions control basement
membrane production and di�erentiation in cultured and
transplanted mouse keratinocytes. Cell. Tissue Res. 244
(2), 413±429.
Bowden, P., Wood, E., Cunli�e, W., 1983. Comparison of
prekeratin and keratin polypeptides in normal and psoria-
tic human epidermis. Biochimica et Biophysica Acta 743,
172±179.
Brysk, M., Arany, I., Brysk, H., Chen, S., Calhoun, K.,
Tyring, S., 1995. Gene expression of markers associated
with proliferation and di�erentiation in human keratino-
cytes cultured from epidermis and from buccal mucosa.
Arch. Oral Biol. 40, 855±862.
Doran, T., Vidrich, A., Sun, T., 1980. Intrinsic and extrinsic
regulation of the di�erentiation of skin, corneal and eso-
phageal epithelial cells. Cell 22, 17±25.
Chung, J., Cho, K., Lee, D., Kwon, O., Sung, M., Kim, K.,
Eun, H., 1997. Human oral buccal mucosa reconstructed
on dermal substrates: a model for oral epithelial di�eren-
tiation. Arch. Dermatol. Res. 289, 677±685.
Gibbs, S., Backendorf, C., Ponec, M., 1996. Regulation of
keratinocyte proliferation and di�erentiation by all-trans-
Fig. 4c (continued)
S. Gibbs, M. Ponec / Archives of Oral Biology 45 (2000) 149±158 157
retinoic acid, 9-cis-retinoic acid and 1,25-dihydroxy vita-
min D3. Arch. Dermatol. Res. 288, 729±738.
Hohl, D., Olano, B., de Viragh, P., Huber, M., Detrisac, C.,
Schnyder, U., Roop, D., 1993. Expression patterns of lori-
crin in various species and tissues. Di�erentiation 54, 25±
34.
Hohl, D., Di Viragh, P., Amiguet-Barras, F., Gibbs, S.,
Backendorf, C., Huber, M., 1995. The small proline-rich
proteins constitute a multigene family of di�erentially
regulated corni®ed cell envelope precursor proteins. J.
Invest. Dermatol. 104, 902±909.
Hudson, D., Weiland, K., Dooley, T., Simon, M., Watt, F.,
1992. Characterization of eight monoclonal antibodies to
involucrin. Hybridoma 11, 367±379.
Ishida-Yamamoto, A., Eady, R., Watt, F., Roop, D., Hohl,
D., Iizuka, H., 1996. Immunoelectron microscopic analysis
of corni®ed cell envelope formation in normal and psoria-
tic epidermis. J. Histochem. Cytochem. 44, 167±175.
Kautsky, M., Fleckman, P., Dale, B., 1995. Retinoic acid
regulates oral epithelial di�erentiation by two mechanisms.
J. Invest. Dermatol. 104, 546±553.
Karring, T., Lang, N., LoÈ e, H., 1975. The role of gingival
connective tissue in determining epithelial di�erentiation.
J. Periodontal Res. 10, 1±11.
Lillie, J., MacCallum, D., Jepsen, A., 1988. Growth of strati-
®ed squamous epithelium on reconstructed extracellular
matrices: long term culture. J. Invest. Dermatol. 90, 100±
109.
Mackenzie, I., Fusenig, N., 1983. Regeneration of organized
epithelial structure. J. Invest. Dermatol. 81 (1), 189S±194S.
Mackenzie, I., Hill, M., 1984. Connective tissue in¯uences on
patterns of epithelial architecture and keratinization in
skin and oral mucosa of the adult mouse. Cell Tissue Res.
235, 551±559.
Mackenzie, I., Rittman, G., Bohnert, A., Breitkreutz, D.,
Fusenig, N., 1993. In¯uence of connective tissues on the in
vitro growth and di�erentiation of murine epidermis.
Epithelial Cell Biol. 2 (3), 107±119.
Mackenzie, I., Rittman, G., Goa, Z., Leigh, I., Lane, E.,
1991. Patterns of cytokeratin expression in human gingival
epithelia. J. Periodontal Res. 26 (6), 468±478.
Mehrel, T., Hohl, D., Rothnagel, J., Longley, M., Bundman,
D., Cheng, C., Lichti, U., Bisher, M., Steven, A., Steinert,
P., Yuspa, S., Roop, D., 1990. Identi®cation of a major
keratinocyte cell envelope protein, loricrin. Cell 61, 1103±
1112.
Ouhayoun, J., Sawaf, M., Go�aux, J., Etienne, D., Forest,
N., 1988. Re-epithelialization of a palatal connective tissue
graft transplanted in a non-keratinized alveolar mucosa: A
histological and biochemical study in humans. J.
Periodontal Res. 23, 127±133.
Ponec, M., de Haas, C., Bachra, B., Polano, M., 1977. E�ects
of glucocorticoids on primary human skin ®broblasts.
Arch. Dermatol. Res. 259, 117±123.
Saiag, P., Coulomb, B., Lebreton, C., 1985. Psoriatic ®bro-
blasts induce hyperproliferation of normal keratinocytes in
a skin equivalent model in vitro. Science 230, 669±672.
Smack, D., Korge, B., James, W., 1994. Keratin and keratini-
zation. J. Am. Acad. Dermatol. 30, 85±102.
Smola, H., ThiekoÈ tter, G., Fusenig, N., 1993. Mutual induc-
tion of growth factor gene expression by epidermal-dermal
cell interaction. J. Cell Biol. 122, 417±429.
Squier, C., Cox, P., Wertz, P., Downing, D., 1986. The lipid
composition of porcine epidermis and oral epithelium.
Arch. Oral Biol. 31, 741±747.
Steinert, P., Marekov, L., 1995. The proteins ela®n, ®laggrin,
keratin intermediate ®laments, loricrin, and small proline-
rich proteins 1 and 2 are isodipeptide cross-linked com-
ponents of the human epidermal corni®ed cell envelope. J.
Biol. Chem. 270, 17702±17711.
S. Gibbs, M. Ponec / Archives of Oral Biology 45 (2000) 149±158158