Expression and the role of E- and P-cadherin in the histogenesis of the surface epidermis and hair follicles were examined using the upper lip skin of the mouse. P-cadherin is expressed exclusively in the proliferating region of these tissues, that is in the germinative layer of the surface epidermis, the outer root sheath and the hair matrix. E-cadherin is coexpressed in these layers but this molecule was also detected in non-proliferating regions such as the intermediate layer of the surface epidermis and the immature regions of the inner root sheath. Neither P- nor E-cadherin was detected in fully keratinized layers such as the horny layer of the surface epidermis, the outermost layer of the outer root sheath and the mature hair fibres. These two cadherins were not detected in dermal cells. We cultured pieces of the upper lip skin in vitro in the absence or presence of a monoclonal antibody to E-cadherin (ECCD-1) or to P-cadherin (PCD-1). In control cultures, skin morphogenesis normally occurred in a pattern whereby the hair follicles grew and dermal cells were condensed to form the dermal sheath. A mixture of ECCD-1 and PCD-1, however, induced abnormal morphogenesis in the skin in several respects. (1) The cuboidal or columnar arrangement of basal epithelial cells was distorted. (2) Hair follicles were deformed. (3) Condensation of dermal cells was suppressed, causing a homogeneous distribution of these cells. These results suggest that cadherins present in epidermal cells are involved not only in maintaining the arrangement of these cells but also in inducing dermal condensation.

Differentiation of the skin is governed by reciprocal interactions between the epithelium and the mesenchyme (Sengel, 1976). The epithelial placode induces the condensation of mesenchymal cells, and then the condensed mesenchyme instructs the epithelium to differentiate into the region-specific cutaneous structures. The molecular mechanism of such interactions, however, remains to be solved.

It is known that various classes of cell-cell adhesion molecules are differentially expressed in the skin during development. The spatiotemporal pattern of expression of L-CAM and N-CAM was extensively studied in chicken feather development, and this pattern was implicated in embryonic induction (Chuong & Edelman, 1985a,b). Interestingly, the inhibition of L-CAM, which is expressed exclusively in the epidermal components of the skin, induced an abnormal pattern in mesenchymal condensation (Gallin et al. 1986), suggesting that epidermal cell-cell adhesion is involved in the induction process directly or indirectly.

The mouse epidermis expresses E- and P-cadherin differentially (Nose & Takeichi, 1986). P-cadherin is expressed only in the basal layer, while E-cadherin is expressed both in the basal and the intermediate layers. Neither of these molecules is expressed in the terminally differentiated keratinized layers. The expression pattern of E-cadherin in the mouse skin is similar to that of L-CAM observed in chicken feather morphogenesis (Chuong & Edelman, 1985a).

In the present study, we examined the expression pattern of E- and P-cadherin in the surface epidermis and vibrissa follicles of the mouse upper lip skin. Then, in order to elucidate a role for these molecules in skin morphogenesis, we cultured the skin fragments in the presence or absence of antibodies that inhibit the function of these adhesion molecules. The results show that the expression of cadherins in the epidermis and hair follicles plays a role not only in epithelial morphogenesis but also in the epithelial-mesenchymal interaction.

Antibodies and immunohistochemistry

The same reagents and methods as described in our accompanying paper (Hirai et al. 1989) were used. Briefly, rat monoclonal antibodies ECCD-1 (Yoshida-Noro et al. 1984) and PCD-1 (Nose & Takeichi, 1986) were used to block the function of E- and P-cadherin, respectively. A rat monoclonal antibody ECCD-2 to E-cadherin which has no activity in blocking the antigen function was used as a control. For immunohistochemical staining of E- and P-cadherin, a rabbit anti-E-cadherin serum (Nagafuchi et al. 1987) and PCD-1 were used, respectively. Unless otherwise stated, sections were double-stained with these antibodies using FITC- and rhodamine-conjugated second antibodies. See the accompanying paper for details on the immunohistochemical procedures (Hirai et al. 1989).

Organ cultures

Upper lip skins were isolated from 13-day embryos of ICR mice. Pieces of the skin were placed on Nuclepore filters, coated with collagen as described by Hirai et al. (1989), with the epidermal side up. The explants were incubated in the presence or absence of ECCD-1 or PCD-1, as described (Hirai et al. 1989).

Immunohistological localization of E- and P-cadherin in the upper lip skin

The surface epidermis of the upper lip skin expresses E- and P-cadherin differentially, as found in the back skin (Nose & Takeichi, 1986); the basal layer has both E- and P-cadherin, the intermediate layer has only E-cadherin and the top layer has neither of them (Figs 1, 2B,C). The differential expression of the two cadherins was also observed in the hair follicles (Figs 1, 2D-G). The proliferating cell layers in the outer root sheath and the hair matrix, which are continuous with the basal layer of the surface epidermis, express both P- and E-cadherin. The non-proliferating zone of the hair matrix and the immature region of the inner root sheath express only E-cadherin. The fully keratinized regions such as hair fibres had neither P- nor E-cadherin. Thus, in both the surface epidermis and hair follicles, the expression of P-cadherin occurs specifically in the proliferating cells, while that of E-cadherin occurs not only in proliferating cells but also in non-proliferating cells. Among the cells expressing P-cadherin, the hair matrix cells were most strongly stained with the antibodies to this molecule. Generally, E-cadherin is expressed at a lower level in the proliferating cell layers than in the non-proliferating layers. P- and E-cadherins were not detected in mesenchymal cells in the dermis.

Fig. 1.

Immunofluorescent localization of E- and P-cadherin in sections of the upper lip skin of a 13-day embryo. (A) Stained with hematoxylin. (B) Stained for E-cadherin. (C) Stained for P-cadherin. Arrows indicate a developing hair follicle. Bar, 50 μm.

Fig. 1.

Immunofluorescent localization of E- and P-cadherin in sections of the upper lip skin of a 13-day embryo. (A) Stained with hematoxylin. (B) Stained for E-cadherin. (C) Stained for P-cadherin. Arrows indicate a developing hair follicle. Bar, 50 μm.

Fig. 2.

Immunofluorescent localization of E- and P-cadherin in the upper lip skin of a 16-day embryo. (A) Longitudinal section of a hair follicle stained with hematoxylin. (B,C) Transverse section of the surface epidermis. (D,E) Transverse section of hair follicles through the sheath area. (F,G) Longitudinal section of a hair follicle. (B,D,F) Stained for E-cadherin. (C,E,G) Stained for P-cadherin. bl, basal cell layer of epidermis; der, dermis; hm, hair matrix; is, inner root sheath; os, outer root sheath. Bar, 50 μm.

Fig. 2.

Immunofluorescent localization of E- and P-cadherin in the upper lip skin of a 16-day embryo. (A) Longitudinal section of a hair follicle stained with hematoxylin. (B,C) Transverse section of the surface epidermis. (D,E) Transverse section of hair follicles through the sheath area. (F,G) Longitudinal section of a hair follicle. (B,D,F) Stained for E-cadherin. (C,E,G) Stained for P-cadherin. bl, basal cell layer of epidermis; der, dermis; hm, hair matrix; is, inner root sheath; os, outer root sheath. Bar, 50 μm.

The pattern of distribution of P- and E-cadherin in the upper lip skin of adult mice was similar to that observed in embryos, although P-cadherin tended to be distributed over a broader area in the hair matrix (Fig. 3A,B). When the regenerating hair follicles in adult mice were stained, the pattern of distribution of these cadherins was essentially the same as that in embryos (Fig. 3C,D).

Fig. 3.

Immunofluorescent localization of E-cadherin (A,C) and P-cadherin (B,D) in the upper lip skin of adult mice. (A,B) Normal hair follicle. (C,D) Regenerating hair follicle. Whiskers were removed from the skin and allowed to regenerate for three days, hm, hair matrix. Bar, 50 μm.

Fig. 3.

Immunofluorescent localization of E-cadherin (A,C) and P-cadherin (B,D) in the upper lip skin of adult mice. (A,B) Normal hair follicle. (C,D) Regenerating hair follicle. Whiskers were removed from the skin and allowed to regenerate for three days, hm, hair matrix. Bar, 50 μm.

Effect of cadherin antibodies on skin morphogenesis

We cultured pieces of the upper lip skin in vitro and found that the morphogenesis of the upper lip skin took place normally in the organ culture system; for example, the extensive outgrowth of vibrissa follicles occurred as reported by Hardy (1951) (Fig. 4A,C). Before testing the effect of antibodies to cadherins on the development of skin explants, we examined whether they can penetrate into them and obtained a positive result (Fig. 4B). When monoclonal antibodies ECCD-1 to E-cadherin or/and PCD-1 to P-cadherin were added to the cultures, abnormal morphogenesis was induced in the explants. The strongest effect of the antibodies was found when ECCD-1 and PCD-1 were added together. Under these conditions, the growth of hair follicles was partially suppressed (Fig. 4F). In histological analyses, we found abnormal morphology in these explants in the following respects. (1) In normal epidermis, the basal layers are packed with cells with a cuboidal or columnar shape (Fig. 5A,B), whereas in the presence of the antibodies, the basal layers became thinner and the cell-cell association in these layers appeared loose and irregular (Fig. 5G,H). (2) The overall morphology of hair follicles was distorted, as seen in Fig. 5G. (3) The distribution of the dermal cells was severely affected in the presence of antibodies. In normal skin, dermal cells were condensed around the outer root sheath to form the dermal sheath (Fig. 5A). Such condensation of dermal cells was virtually abolished by the antibodies (Fig. 5G). As a result, mesenchymal cells were distributed almost homogeneously in the dermis of treated tissues (Fig. 6). Before the skins were dissected for culture, they had already had some dermal condensation (Fig. 1A). Therefore, the preformed condensation seems to have also been abolished.

Fig. 4.

Organ culture of upper lip skin fragments. (A) A skin fragment removed from a 13-day embryo at the initiation of culture. (B) A test of penetration of antibodies. An explant was incubated with 300 μg mF1 ECCD-2 for one day, sectioned and stained with a FITC-labelled anti-rat IgG. Note positive staining on the surface epidermis and hair follicles. (C-F) Cultured for 4 days. (C) Without antibody; (D) 300 μg mF1 ECCD-1; (E) 300 μg mF1 PCD-1; (F) 300 pg mF1 ECCD-1 +300 μg ml-1 ECCD-1. Bar, 500 μm.

Fig. 4.

Organ culture of upper lip skin fragments. (A) A skin fragment removed from a 13-day embryo at the initiation of culture. (B) A test of penetration of antibodies. An explant was incubated with 300 μg mF1 ECCD-2 for one day, sectioned and stained with a FITC-labelled anti-rat IgG. Note positive staining on the surface epidermis and hair follicles. (C-F) Cultured for 4 days. (C) Without antibody; (D) 300 μg mF1 ECCD-1; (E) 300 μg mF1 PCD-1; (F) 300 pg mF1 ECCD-1 +300 μg ml-1 ECCD-1. Bar, 500 μm.

Fig. 5.

Sections of upper lip skin fragments cultured in vitro. (A,B) Control medium. (C,D) 300 μg ml 1 ECCD-1. (E,F) 300 μg ml-1 PCD-1. (G,H) 300 μg ml-1 ECCD-1 + 300 μg ml-1 PCD-1. The left column gives an overall view of a hair follicle and the surrounding mesenchyme. The right column represents a higher magnification of the basal layer of the surface epidermis. Sections were stained with hematoxylin, bl, the basal cell layer of epidermis; de, dermal condensation. Bar, 50 μm for the left column and 20 μm for the right column.

Fig. 5.

Sections of upper lip skin fragments cultured in vitro. (A,B) Control medium. (C,D) 300 μg ml 1 ECCD-1. (E,F) 300 μg ml-1 PCD-1. (G,H) 300 μg ml-1 ECCD-1 + 300 μg ml-1 PCD-1. The left column gives an overall view of a hair follicle and the surrounding mesenchyme. The right column represents a higher magnification of the basal layer of the surface epidermis. Sections were stained with hematoxylin, bl, the basal cell layer of epidermis; de, dermal condensation. Bar, 50 μm for the left column and 20 μm for the right column.

Fig. 6.

Effect of cadherin antibodies on dermal condensation. Upper lip skin fragments were cultured under the same conditions as in Fig. 5. Explants were sectioned in 8 urn thickness through the axis of hair follicles, stained with hematoxylin and photographed. On each photograph, the mesenchymal area adjacent to the outer root sheath was divided into sections with 16·4×32·8 μm2 at both sides of the hair follicle and these sections were numbered from proximal to distal to a hair follicle, as illustrated. Number of nuclei per section was then counted to estimate the cell density. ○, control; ▫, ECCD-1; ▵, PCD-1; ▴, ECCD-1 and PCD-1. Values represent the average of 16 samples.

Fig. 6.

Effect of cadherin antibodies on dermal condensation. Upper lip skin fragments were cultured under the same conditions as in Fig. 5. Explants were sectioned in 8 urn thickness through the axis of hair follicles, stained with hematoxylin and photographed. On each photograph, the mesenchymal area adjacent to the outer root sheath was divided into sections with 16·4×32·8 μm2 at both sides of the hair follicle and these sections were numbered from proximal to distal to a hair follicle, as illustrated. Number of nuclei per section was then counted to estimate the cell density. ○, control; ▫, ECCD-1; ▵, PCD-1; ▴, ECCD-1 and PCD-1. Values represent the average of 16 samples.

A similar effect was observed when PCD-1 alone was added to the cultures, though to a lesser extent (Fig. 5E, F). The arrangement of the epidermal basal cells was somewhat perturbed, and the condensation of dermal cells was also partially inhibited (Fig. 6). On the other hand, ECCD-1 alone had a subtle effect on the morphogenesis of the skin (Figs 5C,D, 6). In a control experiment, the ECCD-2 antibody to E-cadherin, which is unable to block the activity of the antigen, displayed no effect on in vitro skin morphogenesis, although it bound to epidermal cells (ref. Fig. 4B). This indicates that the binding of antibodies to cadherins per se has no effect on skin morphogenesis, unless their function is inhibited.

The present observations showed that cadherins are involved at least partly in maintaining the epithelial structure in the skin, and that multiple cadherin subclasses cooperate in this process. These results are consistent with those obtained using the lung as described in the accompanying paper (Hira et al. 1989). In contrast to the situation in the lung, however, the inhibition of P-cadherin resulted in a stronger perturbation of skin morphogenesis than that of E-cadherin, suggesting that P-cadherin plays a substantial role in the morphogenesis of the skin while E-cadherin plays a supportive role in this tissue. The expression of P-cadherins seems to be associated with the proliferating activity of cells, since expression is seen only in the proliferating cell layers and is highest in the hair matrix cells which are known to have the highest mitotic activity in the epidermis (see Spearman, 1977, for review). Since the proliferating cell layers are regarded as a centre for epidermal histogenesis, the perturbation of the function of these cell layers must lead to the abnormal morphogenesis of this tissue, as observed in the present studies.

The inhibition of P- and E-cadherins with antibodies showed unexpected, interesting results. These antibodies suppressed the condensation of dermal cells around the outer root sheath, although these cells have no detectable E- or P-cadherin. A similar phenomenon was observed in the chick skin system. Gallin et al. (1986) tested the effect of antibodies to L-CAM, a chicken cadherin which is similar in tissue distribution to the mouse E-cadherin, on the histogenesis of chick embryonic back skin grown in organ culture. They found that the anti-L-CAM inhibited the condensation of dermal cells or perturbed the pattern of condensation depending on its concentration, although there was no gross distortion of the epithelial layer in the treated explants. Thus, the studies using different animal species reached a similar conclusion, namely that the inhibition of cadherin-mediated adhesion in the epithelial layer of the skin affects the distribution of dermal cells.

These results suggest that signals are transferred between the epidermis and the mesenchyme to cause condensation of dermal cells, and the mechanism of this signal transfer is associated with cadherin-mediated adhesion in the epidermis. A question arises of how the apparently independent cellular processes, the epidermal cell-cell adhesion and the signal transfer between the epidermis and the mesenchyme, can be connected to each other. The inhibition of dermal condensation observed in the present study was most effective when the two cadherins were blocked simultaneously and secondarily effective when P-cadherin was blocked. These results imply that this phenomenon is associated with the function of the basal layers of the epidermis. It is known that the condensation of dermal cells is induced by the epidermal placode at an early stage of skin morphogenesis (Sengel, 1976). Thus, the thick columnar morphology of epidermal basal cells may be coupled with their ability to induce dermal condensation. This notion is supported by the finding that the antibodies to cadherins destroyed the columnar arrangements of basal cells. It is therefore possible that the cadherin-mediated junctions are essential for the epidermal basal cells to exert their normal physiological functions. The cells whose cadherin-mediated junctions were blocked may therefore not be able to produce factors for inducing dermal condensation.

On the other hand, we cannot rule out a more interesting possibility that cadherin molecules themselves are involved as a signal in inducing dermal condensation. It is known that mesenchymal cells make direct contact with epithelial cells when there is an inductive interaction (Slavkin & Bringas, 1976; Démarchez et al. 1981; Hardy et al. 1983). Therefore, there is a possibility that cadherin molecules on the epidermal surfaces could react directly with molecules present on dermal cells in certain conditions.

We thank Drs S. Tanaka (Mitsubishi Kasei Life Science Institute) and K. Toda (Kitano Hospital) for valuable advice on the literature. Part of this work was supported by research grants from the Ministry of Education, Science and Culture of Japan.

Chuong
,
C.
&
Edelman
,
G. M.
(
1985a
).
Expression of celladhesion molecules in embryonic induction. I. Morphogenesis of nestling feathers
.
J. Cell Biol
.
101
,
1009
1026
.
Chuong
,
C.
&
Edelman
,
G. M.
(
1985b
).
Expression of celladhesion molecules in embryonic induction. II. Morphogenesis of adult feathers
.
J. Cell Biol
.
101
,
1027
1043
.
Démarchez
,
M.
,
Mauger
,
A.
&
Sengel
,
D.
(
1981
).
The dermal-epidermal junctions in the development of skin and cutaneous appendage in the chick embryo
.
Archs Anal, microsc. Morphol. exp
.
70
,
205
218
.
Gallin
,
W. G.
,
Chuong
,
C.-M.
,
Finkel
,
L. H.
&
Edelman
,
G. M.
(
1986
).
Antibodies to liver cell adhesion molecule perturb inductive interactions and alter feather pattern and structure
.
Proc. natn. Acad. Sci. U.S.A
.
83
,
8235
8239
.
Hardy
,
M. H.
(
1951
).
The development of pelage hairs and vibrissae from skin in tissue culture
.
Ann. N. Y. Acad. Sci
.
53
,
546
561
.
Hardy
,
M. H.
,
Van Exan
,
R. J.
&
Songtegard
,
K. S.
(
1983
).
Basal lamina changes during tissue interactions in hair follicles. An in vitro study of normal dermal papillae and vitamin A induced gLondular morphogenesis
.
J. invest. Derm
.
80
,
27
34
.
Hirai
,
Y.
,
Nose
,
A.
,
Kobayashi
,
S.
&
Takeichi
,
M.
(
1989
).
Expression and role of E- and P-cadherin adhesion molecules in embryonic histogenesis. I. Lung epithelial morphogenesis
.
Development
105
,
263
270
.
Nagafuchi
,
A.
,
Shirayoshi
,
Y.
,
Okazaki
,
K.
,
Yasuda
,
K.
&
Takeichi
,
M.
(
1987
).
Transformation of cell adhesion properties by exogenously introduced E-cadherin
.
Nature, Lond
.
329
,
341
343
.
Nose
,
A.
&
Takeichi
,
M.
(
1986
).
A novel cadherin adhesion molecule: Its expression pattern associated with implantation and organogenesis of mouse embryos
.
J. Cell Biol
.
103
,
2649
2658
.
Sengel
,
P.
(
1976
).
Tissue interactions in skin morphogenesis
.
In Organ Culture in Biomedical Research
(ed.
M.
Balls
&
Monnickendam
), pp.
111
147
,
Cambridge
:
Cambridge University Press
.
Slavkin
,
H. C.
&
Brincas
,
P.
(
1976
).
Epithelial mesenchymal interactions during odontogenesis. IV. Morphological evidence for direct heterotypic cell-cell contacts
.
Devi Biol
.
50
,
428
432
.
Spearman
,
R. I. C.
(
1977
).
The structure and function of the fully developed follicle
.
In The Physiology and Pathophysiology of the Skin
(ed.
T.
Jarrett
), pp.
1294
1349
,
New York
:
Academic Press
.
Yoshida-Noro
,
C.
,
Suzuki
,
N.
&
Takeichi
,
M.
(
1984
).
Molecular nature of the calcium-dependent cell-cell adhesion system in mouse teratocarcinoma and embryonic cells with a monoclonal antibody
.
Devi Biol
.
101
,
19
27
.