Archs oral Viol.Vol. 35, No. IO, pp. 813422, 1990 Printed in Great Britain. All rights reserved

0003-9969/90 $3.00 + 0.00

Copyright 0

1990 Pergamon Press plc

IMMUNOLOCALIZATION OF TRANSFORMING GROWTH FACTOR p1 AND EPIDERMAL GROWTH FACTOR RECEPTOR EPITOPES IN MOUSE INCISORS AND MOL,ARS WITH A DEMONSTRATION OF IN ?‘ITRO PRODUCTION OF TRANSFORMING ACTIVITY Y. CAM, M. R. NEUMANNand J. V. RUCH INSERM 13F No. 88-08, Institut de Biologie Medicale, Fact&C de Medecine, II, rue Humann, 67085 Strasbourg Cedex, France (Accepted 1 May 1990) Summary-Day-14 lower incisors and day-18 first lower molars of mouse embryos produced in vitro transforming activities for non-confluent NRK cells co-cultured in agar, and mitogenic activities for exponentially growing NRK and BHK cells. The patterns of distribution of TGF/?, and EGF receptor, both known to regulate cell proliferation, differentiation and transformation in vitroand suspected to play important roles in developmental processes, were studied during mouse odontogenesis by means of indirect immunofluorescence on fixed or frozen fixed sections. TGFj, epitopes were detected in the stellate reticulum of day-13 to day-16 incisors and of molars from day-17 onwards. Dental mesenchyme of day-14 incisors and postnatal molars, and peridental mesenchyme of bud and cap stage molars and incisors were also stained by TGFj, antibodies. EGF receptor was localized in the enamel organs of incisors and molars: the inner dental epithelium and later the outer dental epithelium rapidly became negative while the stellate reticulum remained stained. Incisor apical mesenchyme showed an intense reaction with EGF receptor antibodies after birth. Key words: immunohistochemistry, transforming growth factor /?, , epidermal growth factor receptor, teeth, transforming activity, growth factors, odontogenesis.

INTRODUCTION Odontogenesis is dependent on both cell-matrix interactions and specific cell kinetics (reviewed by Ruth, 1987). Cytodfferentiation of post-mitotic cells (odontoblasts and ameloblasts) gives rise to dentine and enamel (Ahmad and Ruth, 1987). Circulating and paracrine or autocrine growth factors and their specific receptors ml.ght provide epigenetic control of odontogenesis in vitro and in vivo (reviewed by Partanen and Thesleff, 1989) by modulating cell growth and ceil interactions. Numerous macromolecules and mRNAs of the growth factor families have been located in embryonic tissues. Among these factors, expression of TGFa, but not EGF, (both potential ligands for the same receptor, the EGF receptor), and of TGF& have been described (Mercola and Stiles, 1988; Wilcox and Derynck, 1988). Partanen and Thesleff (1987) and. Abbott and Pratt (1988) have suggested that EGF receptor plays a crucial role during tooth morphogenesis. Expression of TGFP, has also been reported in incisors of mouse embryos (Heine et al., 1987).

Abbreviations: BHK, baby hamster kidney; BSA, bovine serum albumin; DMEM, Dulbecco’s modified Eagle’s medium; EGF, epidermal growth factor; FCS, fetal calf serum; HBSS, Hank’s balanced salt solution, NRK, normal rat kidney; PBS, phosphate-buffered saline; TBS, tris-buffered AOB WI_

saline; TGF,

transforming

growth

factor.

It appears that TGFcl and 8, act synergistically on proliferation and elicit reversible changes of phenotype (transformation in vitro) in NRK cells grown in agar (De Larco and Todaro, 1978; Anzano et al., 1983). Bifunctional effects (stimulation or inhibition) of TGFP, on proliferation and/or differentiation of various mesenchymal and epithelial cell lines have been described (reviewed by Roberts et al., 1988). Transcriptional and post-transcriptional regulation of different genes (in particular those encoding TGFP, itself, EGF receptor, fibronectin, tubulin, type I collagen, proteoglycans and integrins) occur in vitro in the presence of TGF& (Van Obberghen-Schilling et al., 1988; Bascom et al., 1989; Thompson, Assoian and Rosner, 1988; Ignotz and Massague, 1986; Bassols and Massague, 1988; Heino et al., 1989). We have now first sought to characterize mitogenic and transforming activities produced by mouse tooth germs (incisors and molars) in co-culture with fibroblastic cell lines NRK and BHK. We also describe the patterns of distribution of TGFfl, and EGF receptor during odontogenesis obtained by means of indirect immunofluorescence using polyclonal monospecific antibodies. MATERIALSAND METHODS Collection, preparation and dissociation of tissues

Stages of mouse embryos are referred to day 0 (= vaginal plug). Day-14 and -15 incisors, day-18 and -20 molars, day-16, -17, -18 and -19 mandibles 813

814

Y. CAMet al.

and day-13, -14 and -15 heads of laboratory-raised Swiss mouse embryos and postnatal animals were dissected in HBSS (Gibco, Paisley, Scotland). Day-18 molars were incubated at room temperature for 5 min in 0.01 M EDTA (Merck, Darmstadt, F.R.G.) in Ca*+-free, PBS (Gibco) at pH 7.3 and the epithelial and dental papilla components dissociated, as described by Osman and Ruth (1981a). Cultures

NRK-49F cells (2432th passages) obtained from the American Tissue Culture Collection (ATCC, Rockville, MD) and BHK-21 cells provided by Dr G. Rebel (Centre de Neurochimie du CNRS, Strasbourg, France) were co-cultured with tooth germs or tooth germ components. Mitogenic activities. To determine mitogenic activities, Costar plates (6 x 24mm 4 wells; Cambridge, MA equipped with Transwell filters (0.4 p pore) were used. NRK-49F and BHK cells were seeded in DMEM (Gibco) containing 10% FCS (G&co) at concentrations of 2000 cells/ml and 1000 cells/ml respectively. Two, four and six lCday-old incisors and 18-dayold molars were deposited gently on the surface of filters just after cell seeding and co-cultured for 3 days in a humidified atmosphere containing 5% CO2 and 95% air. Experiments were run in duplicate or triplicate. Control cultures omitted teeth or tooth components. Cultures were arrested by removing the filters, discarding the media and rinsing both cells and teeth 6 times with PBS, pH 7.3. Cells were then fixed in 10% (v/v) neutral buffered formalin (i.e. minimum 35% formaldehyde solution (Merck) diluted 10 times in PBS, pH 7.2) and stained with Papandreou’s haematoxylin solution (40% in distilled water) and kept in 70% ethanol till cell counting. Teeth were fixed in Bouin’s Holland solution, embedded in paraffin and cut at 5 pm; sections were stained with Gomori’s solution. A minimum of 1000 cells per well was counted in a Leitz microscope equipped with a special tray. Cells in late prophase, metaphase, anaphase and early telophase were also counted to establish the mitotic indices. Means of numbers of cells and mitotic indices were compared by using the parametric Student’s t-test. Transforming activities. For this purpose one volume (500 ~1) of gelled Noble Agar (Difco Labs, Detroit, MI) was melted and quickly mixed at 39°C in one Petri dish (30 mm 4) with the same volume of 2 x (DMEM + FCS) containing NRK-49F cells so that the final concentrations of agar, cells and serum were 0.3% (w/v) 2000 cells/ml and 11.2% (v/v), respectively. Freshly dissected day-14 (2, 4 and 8 specimens per Petri dish) and day-l 5 (3, 6 and 9 specimens) incisors and day-18 molars (2, 4, 6 and 12 specimens) were added to the above mixture before gelling occurred. Dishes were then incubated in a humidified 5% CO2 atmosphere without further feeding for 7 days. Parallel assays were performed in the absence of teeth by adding either 0.3 and 1.5 nM mouse EGF (receptor grade, Collaborative Research, Lexington, MA) or 5, 25 and 125 pM human TGF/I, (from

platelets; R and D System Inc., Minneapolis, MN) and 0.3 or 1.5 nM EGF plus 5,25 and 125 pM TGFB, to the above culture mixtures. Control cultures without any teeth, EGF or TGFB, were also made. Two dishes per culture condition were used. All colonies of at least 8 cells were counted over each Petri dish and some of them photographed under a Leitz microscope equipped for microphotography. Immunocytochemical study Localization of TGFj?,. Freshly dissected mouse tissues were rinsed 3 times in PBS, pH 7.2, fixed at room temperature in 10% (v/v) neutral formalin for at least 4 h, and transferred to Bouin’s fixative (0.9% (w/v) picric acid (Merck), 9% (v/v) minimum 35% formaldehyde solution (Merck), 5% (v/v) acetic acid) for a further 4 h. Then, specimens were dehydrated in a graded series of ethanol and in toluene before embedding in paraffin (Prolabo, Paris, France) at 5456°C. Serial sections were cut at 5 pm in a microtome. Primary antibodies were affinity-purified rabbit polyclonal antibodies directed against two different synthetic preparations of a peptide corresponding to the N-terminal 30 amino-acids of TGF/i, (Ellingsworth et al., 1986; Flanders et al., 1988). These antibodies were generously provided by K. C. Flanders and N. L. Thompson (NCI, Bethesda, MD): the one called anti-CC (l-30) and the other anti-LC (l-30) antibody have been found to stain principally extracellular and intracellular TGF/I epitopes, respectively (Flanders et al., 1989). Secondary antibodies were fluoroisothiocyanateconjugated anti-rabbit IgGs (H + L) purchased from Cappel Labs (Cochranville, PA). The procedure described by Heine et al. (1987) for immunolocalization of TGF/?, with anti-CC (l-30) antibody was used, with some modifications. Sections were deparaffinized in toluene and rehydrated in decreasing concentrations of ethanol. Then they were rinsed 4 times quickly and 3 times for 3 min in TBS (0.01 M tris-HCl (Sigma, St Louis, MO), pH 7.4; 0.85% (w/v) NaCl (Merck) containing 0.1% (w/v) BSA and merthiolate (Sigma) at 50mg/l and pretreated with bovine testis hyaluronidase (Sigma) diluted at 1 mg/ml in 0.1 M sodium acetate buffer pH 5.5 (30 min at room temperature). Sections were rinsed as above and non-specific binding was prevented by blocking epitopes with TBS/O.S% BSA containing 1.5% (v/v) normal goat serum (normal goat serum, Gibco) for 15 min at room temperature. Primary antibodies, 20-50 p&depending on the size of the specimens-diluted at 20 pg/ml in TBS + 1% normal goat serum were applied to sections and incubated overnight at 4°C. Sections were then rinsed 6 times quickly and 4 times (3 min each) in TBS/O. 1% BSA and incubated with secondary antibodies diluted 1:40 in TBS/l% BSA for 1 h at room temperature. Sections were finally rinsed 4 times quickly and 3 times (3 min each) in TBS/O.l% BSA, mounted in glycerol/TBS (9: 1, v/v), viewed in a Leitz Orthoplan microscope equipped with epifluorescence and photographed with an Orthomat automatic camera.

815

Growth factor receptor in tooth germs

-5

. : : 8 x $ C

.3

-1 .s g 2 4 0 Nurnber of day - 18 molars

t

C

_

0

2 4 0 Number of day - 18 molars

6

(L3)

7

Iii

m

.7

7.

.5

5-

= s

6

(4)

.7 NRK

NRK 5

?;;

I

;/

?;;

; .j , 0 2 4 6 Number of day - 18 molar1 dental papillae

I.

a

.5

-

fi*j

.

z

,

(5)

7.

,-

,

0 2 4 Number of day - 18 enamel

z

6 organs

.7 NRK

5.

.5

I m

: :

:

3:

x

‘z 5 3 !,.

0

./L

j

t 0 2 4 Number of day - 14 incisors

,. 6

S

Mifogenic activities of tooth germs on NRK-49 F cells (2000cells/ml) (Figs 1, 3, 4 and 5) and BHK-21 cells (1000 cells/ml) (Fig. 2) expressed as numbers of cells and mitotic indices versus numbers of teeth molars (Figs 1 and 2) and incisors (Fig. 5) and molar components: papilla (Fig. 3) and epithelium (Fig. 4). Data are the average of 2-3 plates (bars are standard deviations). Figs l-5.

Control sections were obtained by replacing the primary specific antibodies with preimmune rabbit IgGs, affinity-purified in the same way as the specific IgGs (gift of K. C. Flanders and N. L. Thompson). Localization of EGF receptor. Freshly dissected

mouse tissues were rinsed in BPS, pH 7.2, and fixed at room temperature in freshly prepared 3.7% (w/v) paraformaldehyde (Serva, Heidelberg, F.R.G.) for 30 min-4 h depending on the size of the specimens, dehydrated in a graded series of ethanol and in toluene before embedding in paraffin and sectioned as above.

Freshly dissected day-14 heads and day-20 molars were quickly frozen in Tissue Tek (OCT compound, Miles, Elkhart, IN) over liquid nitrogen and cut at 8 pm ion a cryostat (Slee, Mainz, F.R.G.). Sections were fixed in 3.7% paraformaldehyde) for 30 min at 4°C just before immunolocalization. Primary antibodies were monospecific polyclonal rabbit antibodies to purified mouse EGF receptor (Weller, Meek and Adamson, 1987) generously provided by E. D. Adamson (CRF, La Jolla, CA). Secondary antibodies were fluoroisothiocyanate-conjugated goat anti-rabbit IgGs (H + L) from Cappel Labs.

Y. CAMet al.

w - -w Day-i 8 molars

1

,

1

I

I

M

Day-l 5 incisors

+ -+

Day-l 4 incisors

I

I

5

I

I

I

I

I

8

I

10 Number of teeth

Fig. 7. Transforming activities of tooth germs on NRK cells (2000 cells/ml) expressed as numbers of colonies versus numbers of germs. Data are the average of duplicate plates.

Immunolocalization of EGF receptor was by the procedure described above for TGF/I localization except that:

1

25

: GF-O

concentration,

U 013 EGF concentration,

Fig. 6. Transforming

activities

125 pM (D-(7)

nM (OOOc?

of EGF,

TGFB and

EGF + TGFB on NRK cells (2000cells/ml) expressed as numbers of colonies versus concentrations of TGF/3 @M) and EGF (nM). Data are the average of duplicate plates.

(1) no hyaluronidase treatment was applied; (2) PBS, pH 7.2, was used instead of TBS; (3) primary antibodies were diluted in the range 1: 100-l : 250 (i.e. 20-50 pg/ml) in PBS containing 1% (v/v) normal goat serum and incubated with sections for 2 h at room temperature; and (4) monolayer of A431 cells, grown on coverslips, were washed in PBS and reacted unfixed with 1: 25 dilution of anti-EGF receptor antibodies (200 pg/ml) as described by Weller et al. (1987). Control Labs.

were normal

rabbit

IgGs from Cappel

817

Growth factor receptor in tooth germs

Table 1. Imrnunolocalization of TGFjf, with anti-CC( l-30) antibody in lower incisor (I) and first lower molar (M) of mouse embryo and new-born mouse Tissue dental mesenchyme Day ,after vaginal plug

I

13 14 I5 16 17 18 19 Z!O

M

I

M

I

M

I + +++ ++ +++

+

ND ND

Stellate reticulum

Inner dental Outer dental epithelium epithelium

+ +

ND ND

ND ND

ND ND

M

Peridental mesenchyme I

M

+ +

+ ++ ++

+ ++ ++ ND + + + ND +

Fainl. ( + ). main I + + ) and strona1.( + + + ) staining of incisor and molar tissues. ND := not’ddtermined. ’ RESULTS

Mitogenic

activities

Mitotoc indices of non-confluent NRK-cells increased steadily when co-cultured with increasing numbers of day-18 molar (Text Fig. l), molar dental papillae (Text Fig. 3) and enamel organs (Text Fig. 4). In the presence of 6 molars, 6 dental papillae and 6 enamel organs, the mitotic indices were respectively 5.6,2.8 and 3.2 times higher than those of NRK cells cultured in the absence of teeth. Effects of day-18 molars on the mitotic indices of co-cultured BHK cells were less intense (by approx. 35%) compared to those in NRK cells (Text Fig. 2). When co-cultured ,with day-14 incisors, NRK cells also showed an incrmeasein mitotic index, but this was less important than in the presence of day-18 molars (Text Fig. 5).

forming activities, expressed per co-cultured tooth germ, were: 5, 3 and 3 colonies for day-14, day-15 incisors and day-18 molars, repectively. Colonies of NRK cells produced by the addition of TGFfl (125 PM) plus EGF (1.5 nM) to the culture medium and after co-culture with day-14 lower incisors are shown in Plate Figs 9 and 10, respectively. Immunolocalization

Results of staining with anti-CC (l-30) antibody are summarized in Table 1. A preferential and transitory immunolocalization of TGFfl was found in stellate reticulum (Plate Figs 11, 13 and 15) and peridental mesenchyme (Plate Fig. 13) of incisors and molars. Dental mesenchyme appeared positively stained only in day-14 incisor (not shown). Anti-LC (l-30) antibody did not stain any cells of tooth germs. Immunolocalization

Transforming activities

Results are given in terms of numbers of colonies formed by NRK cells. Increasing concentrations of TGF/I (mainly those over 25 PM) in the presence of EGF resulted in a dramatic increase of the colony number (Text Fig. 6). In comparison, mouse teeth in co-culture with NRK cells (i.e. 8 day-14 and 9 day-l 5 incisors; 12 day-18 molars) produced as many colonies as with the addition of TGFfi (25 PM) plus EGF (0.3 and 1.5 nM) (Text Fig. 7). Maximal trans-

of TGFfi

of EGF receptor

Results are summarized in Table 2. Day-13 dental bud as well as oral epithelium (Plate Fig. 20) were specifically stained. During histomorphogenesis of the enamel organs of incisors and molars the inner dental epithelium (Plate Fig. 21) and later the outer dental epithelium rapidly became negative while the stellate reticulum remained stained (Plate Figs 22 and 24). Incisor apical mesenchyme showed an intense reaction after birth (day-19) (Plate Fig. 19).

Table 2. Immunolocalization of EGF receptor in lower incisor (I) and first lower molar (M) of mouse embryo and new-born mouse Tissue dental mesenchyme Day after vaginal plug 13 14 15 16 1’7 18 l!J 20

I

M

+

v(+ +)v(+ +) v(+) v(+ +)

Inner dental Outer dental epithelium epithelium I

M

I

M

+ ++

++

+ ++

++ ++ +

Stellate reticulum I

M

Peridental mesenchyme 1

M

+++

+

+++ ++ + ++ + +++ ++++++

+ +

Faint ( + ), plain ( + + ) and strong ( + + + ) staining of incisor and molar tissues. Faint [VI:+ )] and plain [v( + + )] staining of blood vessels.

Y. CAMet al.

818

An intense pericellular staining of unfixed A431 cells was observed (not shown). DISCUSSION

Carrel and Baker provided direct evidence of the presence of growth-promoting activity in tissue extracts in 1926. Our demonstration here of the in vitro production of mitogenic activities by tooth germs is not proof of either restricted production of growth factors of production of a given growth factor by mouse incisors and molars. Mitogenic activities were also released in the same culture system by heart and toe buds of mouse embryos (not shown). This is true also for transforming activities. The production of mitogenic activities for NIH 3T3 cells and transforming activities for NRK cells by limb (wing and leg) buds have been described (Bell and McLachlan, 1985; Bell, 1986; McLachlan et al., 1988). Different combinations of growth factors can result in transformation of cultured cells; not only TGFB, associated with TGFc( or EGF but platelet derived growth factor and basic fibroblast growth factor, added together or apart, exert a transforming action on NRK fibroblastic cells (Rizzino and Ruff, 1986; Rizzino, Ruff and Rizzino, 1986) and other cell types such as chondrocytes (Kato, Iwamoto and Koike, 1987) and human fibroblasts (Palmer, Maher and McCormick, 1988). In our experiments, cultures were made in the presence of serum, which might explain why control values of colony number differ slightly from zero and that either EGF (nM concentrations) or TGF& (PM concentrations) added to the culture medium resulted in transformation of NRK cells. Nevertheless, as shown in Text Fig. 7, it is evident that significant transforming activity for NRK cells was secreted into the culture medium due to the presence of mouse tooth germs. To explain this, two suggestions may be offered. The first is that transforming activity actually resides within cultured teeth. The second is that autocrine transforming activity was expressed by NRK cells themselves under the influence of the metabolizing tooth germs. We carried out a further experiment consisting of adding portions of an acidic extract of 100 day-18 molars (i.e. approx. 25 mg wet wt) prepared according to Roberts et al. (1980) to the culture medium of non-confluent NRK cells in agar. A significant

(although less than above) transforming activity was found after 7 days in culture (not shown). As yet, neither of the two hypotheses can be ruled out. Both effects might amplify each other as, for instance, TGFB, is known for its ability to induce its own message in normal and transformed cells (Van Obberghen-Schilling et al., 1988). Our findings concerning immunolocalization of TGFB, with anti-CC(l-30) antibody corroborate the partial findings of Heine et al. (1987) on mouse incisors using the same specific antibody. The earlier expression of TGF/?, in stellate reticulum of incisors than molars does not reflect the known delay of 1.5 day between odontoblast polarization in incisors and first lower molars. No relation can be established between the differential ability of labial and lingual incisor preameloblasts to differentiate into functional ameloblasts as stellate reticulum on both sides (lingual and labial) was stained by anti-CC( l-30) TGFB, antibody (Plate Fig. 13). As suggested by Lehnert and Akhurst (1988), the indirect immunofluorescence technique might not be sensitive enough to detect intracellular TGF/?, . They described localization of TGFP, mRNA as “a diffuse signal associated with mouse dental and peridental mesenchyme at 12 days” (stage of dental lamina) and they noted at day 14 (i.e. early cap stage) a preferential localization of TGFPl mRNA within dental epithial cells. We suggest that cells of inner and/or outer dental epithelial cells synthesize and secrete TGF/?, ; receptors for TGF/?, would be located on cells of dental mesenchyme and stellate reticulum and participate in the regulation of cell proliferation at early stages of odontogensis (