- Email: [email protected]
Collagenase inhibitor stimulates cleft formation during early morphogenesis of mouse salivary gland
DEVELOPMENTAL
BIOLOGY
113,201-206
(1986)
Collagenase Inhibitor Stimulates Cleft Formation during Early Morphogenesis of Mouse Salivary Gland
A collagenase inhibitor obtained from the culture medium of bovine dental pulp markedly enhanced the cleft formation of mouse embryonic salivary gland epithelium when the inhibitor was included in the culture medium for 12-day and 13-day salivary glands. Determination of collagenase activity using [3H]collagen as substrate indicated that there was a latent collagenase activity in 12-day glands. In addition, a highly purified Clmstritlicrl collagenase freed from protease and hyaluronidase activities, strongly inhibited initiation of the cleft formation of the 12-day epithelium. Scanning electron microscopic observation showed that abundant collagen-like fibrils were seen on the epithelium in the collagenaseinhibitor-treated glands compared to those in the control. These findings suggest that during early morphogenesis d 19x6 Academic Press. Inc. tissue collagenase may regulate the cleft formation in the epithelium. INTRODUCTION
The branching morphogenesis of mouse embryonic salivary gland has been examined extensively in analysis of epithelial-mesenchymal interaction. Specifically, these studies have assessed the mechanism of interlobular cleft formation necessary for the proper morphogenesis of the epithelium (Grobstein, 1967). It is well established that such interactions are achieved through the extracellular space, in which glycosaminoglycans and collagens are the major components (Bernfield ef al., 1984). Among those, basal lamina glycosaminoglycan plays an essential role in maintaining epithelial morphology, since enzymatic removal of the glycosaminoglycan disrupts the basal lamina of the 13-day mouse salivary epithelium and causes loss of its multilobular morphology (Bernfield et al., 1973). Furthermore; the turnover rates of the glycosaminoglycan fraction at the distal ends of the lobules, interlobular clefts, and stalk are different from each other (Bernfield and Banerjee, 1982). A novel neutral hyaluronidase in the mesenchymal tissue (Banerjee and Bernfield, 1979) is believed to be a key enzyme for such precise metabolism of the glycosaminoglycan, leading to the characteristic branching of mouse salivary gland. Another important substance in the extracellular space responsible for the branching morphogenesis of salivary gland is collagen, which is present in the mesenchyme and on the epithelial surface, especially on the cleft and stalk regions (Spooner and Faubion, 1980). In the experiment using L-azetidine 2-carboxylic acid (Spooner and Faubion, 1980), collagen synthesis and secretion were suggested to be involved in the branching 201
phenomenon. If collagenolytic activity in the salivary gland is concerned with clefting, introduction of a collagenase inhibitor into the culture medium might modulate the branching pattern. In addition, since the inhibition experiments of branching with some bacterial collagenase preparations (Grosbtein and Cohen, 1965; Wessells and Cohen, 1968) were obscure due to possible contamination of mucopolysaccharidases (Bernfield et al., 1972), it seemed to be important to re-evaluate the effect of a bacterial collagenase on the epithelial morphology. In this report, we will describe a stimulative effect on the cleft formation of an animal collagenasespecific inhibitor (CI), which was isolated from the culture medium of bovine dental pulp and purified to a single protein (Kishi and Hayakawa, 1984), and an inhibitory effect on the cleft formation of a highly purified collagenase from Clostridiunz hisfolyticum. We will also present evidence demonstrating the existence of collagenase activity in 12-day salivary gland and some scanning electron micrographs of the glands treated with CI. MATERIALS
AND
METHODS
Mnteriuls. [3H]NaBH4 (8.7 Ci/mmole) and [3H]thymidine (23.8 CVmmole) were obtained from New England Nuclear Corporation, Boston, Massachusetts. A collagenase inhibitor (CI) was isolated from the culture medium of bovine dental pulp and purified to homogeneity as described previously (Kishi and Hayakawa, 1984). A highly purified collagenase from Clostridium hysfolyficum, which was freed from caseinolytic activity, was kindly donated by Dr. T. Ohya and Dr. N. Yokoi of 0012-1606/86 Copyright All rights
$3.00
Cc 1986 hy Academx Press, Inc. of reproduction in any form rrserved
202
DEVELOPMENTAL
BIOLOGY
Amano Pharmaceutical Company, Japan. No hyaluronidase activity of the collagenase preparation was detected by using [3H]hyaluronic acid as substrate. 3H-Labeled co lla g en was prepared by the modified method essentially the same as those described by Rice and Means (1971) and Tack et al. (1980). Eight milligrams of acid-soluble native type I collagen (calf skin), which was dissolved in 2 ml of 0.01 M acetic acid, was mixed with 6 ml of 0.2 M borate buffer, pH 9.0, and placed on ice for 1 hr. To the vessel, 400 ~1 of 0.04 M formaldehyde was added immediately followed by the addition of 0.725 ml of [3H]NaBH4 solution (25.0 mCi of [3H]NaBH4 was diluted with 0.01 mg of unlabeled NaBH4 in 1.55 ml of 0.01 M NaOH). After 2 min, the same volume of [3H]NaBH4 solution was again added and the vessel was kept on ice for 30 min. The [3H]collagen thus formed was precipitated by the addition of acetic acid and NaCl so as to make final concentrations of 0.5 M and lo%, respectively, and was collected by centrifugation at 10,OOOg for 20 min. The precipitate was washed with 10% NaCl in 0.5 M acetic acid and dissolved in 6 ml of 0.5 M acetic acid. The solution was dialyzed against 0.01 M acetic acid and lyophilized (6 mg). The specific activity was determined to be 6.1 X lo7 dpm/mg. Culture and labeling. Salivary glands were obtained from DDY strain mice and the day of discovery of the vaginal plug was designated as Day 0. The glands were routinely cultured on ultrathin Millipore filter (THWP, 0.45-wrn pore size) assemblies as previously described (Banerjee et al, 1977) in BME medium supplemented with 50 pg/ml ascorbic acid and 10% fetal calf serum. Labeling medium contained 10.0 pCi/ml [3H]thymidine. The rate of DNA synthesis was determined by counting the radioactivity of acid-insoluble materials after 2 hr’s pulse with [3H]thymidine. Living cultures of salivary glands were monitored by taking photographs at appropriate intervals on an Olympus light microscope BH-2 equipped with an automated exposure unit. Determination of number of clefts into epithelium. Quantitative determination of epithelial morphogenesis was done by counting the number of clefts in the photographs of the living cultures taken under standardized conditions of lighting and magnification. Clefts to be counted were vertical ones at 24 hr of culture in the plane of the photograph, which had length of more than half of the radius of the lobule. Determination of collagenase activity. Twelve-day glands used for the assay were removed from embryos in Tyrode’s solution and transferred together with a small amount of the solution into an Eppendorf tube and frozen until use. The whole tissues (886 glands) were thawed and centrifuged to remove the supernatant. The glands were suspended in 2 ml of 30 mM Tris-HCl, pH
VOL~JME
113, 1986
7.8, containing 0.5 M NaCl, 5 mM CaClz and 0.1% Triton X-100 and homogenized in a Potter-Elvehjem homogenizer. The homogenate was used as enzyme source. The complete reaction mixture consisted of 350 ~1 of 0.1 M Tris-HCl, pH 7.8, containing 0.2 M NaCl, 5 mM CaC& and 1 Mglucose, 100 ~1 of [3H]collagen solution (6.1 X lo6 dpm) and 50 ~1 of tissue homogenate (-22 glands) which was preincubated at 35°C for 3.5 hr in the presence of 1 mM 4-aminophenylmercuric acetate. The reaction mixture was incubated at 35°C for 19 hr and 50 ~1 of 0.2% unlabeled collagen solution were added. The degraded collagen was then extracted into 50% dioxane solution as previously described (Terato et al., 1976) and the radioactivity was counted. Under these conditions, [3H]collagen-degrading activity of the homogenate activated by 4-aminophenylmercuric acetate was linear up to 100 ~1 of the homogenate. Processing for scanning electron fmicroscopy. Salivary glands for scanning electron microscopy were fixed with 1.6% glutaraldehyde, 1.6% paraformaldehyde in 0.1 M phosphate buffer, pH 7.2. Glands were postfixed for 1 hr with 2% osmium tetroxide in 0.1 M phosphate buffer, dehydrated with ethanol, and transferred in isoamyl acetate. Glands thus treated were critical point dried from liquid carbon dioxide, mounted onto aluminum stubs, and cracked with a tiny tip of the fine needle under microscope. The fragments of glands were sputter-coated with gold and viewed with an Akashi ALPHA-30 scanning electron microscope at 30 kV. RESULTS
AND
DISCUSSION
We first assessed the effect of CI on the branching morphogenesis of mouse salivary gland. Salivary glands at two different stages were cultured on Millipore filter assemblies in the presence and absence of 5 pg/ml CI (Fig. 1). In the control culture of 12-day glands, an average of 2 clefts was seen at 24 hr of culture (Figs. lAC). On the contrary, when CI was added, more clefts were consistently formed (Figs. lD-F), strongly suggesting a possible involvement of collagenase activity during early branching morphogenesis. It is of interest that lobules varying in size were often formed in the CItreated glands (Figs. lE, F) as compared with relatively uniform lobules in the control cultures (Figs. lB, C), possibly indicating that no pre-determined sites for clefts exist on the epithelial surface. The stimulative effect was further confirmed when CI was added in the culture medium of 13-day glands. However, since the branching pattern of the glands at this stage became more complex in culture, further studies using CI and the bacterial collagenase were performed mostly on 12 day glands. To find out an optimal concentration, different amounts of CI were added to the culture medium and
FIG. culture.
1. CI effects A CI-treated
on salivary (5 &ml)
gland gland
morphogenesis. A control 12-day gland is shown successively at (A) 1 hr, (B) 18 hr, is shown successively at (II) 1 hr, (E) 18 hr, and (F) 25 hr of culture. Bar = 0.2 mm.
clefts were counted, As seen in Fig. 2, a saturation was attained at 2-5 pg/ml and approximately 5 clefts on the average were observed as compared with 2 in the absence of CI. The concentration of 5 pg/ml was effective for 13day gland and used throughout this study. To confirm the collagen involvement, we re-examined the effect of Clostridial collagenase on salivary gland morphogenesis. On a dose-response experiment, 5 pg/
yjil..i Collagenase
Inhibitor
(uglml)
FIG. 2. Effects of various concentrations of CI on cleft formation of the glands cultured for 24 hr. Quantitative determination of epithelial morphogenesis was done as described under Materials and Methods.
and
(C) 25 hr of
ml collagenase was shown to be sufficient for the inhibition of initiation of the cleft formation for 24 hr of cultivation (Fig. 3). Of particular interest is that although each 12-day gland had a distorted lobule sometimes having several indentations, addition of the enzyme gave rise to formation of round or oval lobule with a smooth contour (Fig. 3E). This implies a possible role of collagen as a source of constraint probably necessary for the cleft formation (Nogawa, 1983). It should be added here that, in the enzyme-treated glands, neither distinct flattening of the glands nor spreading of the mesenchyme on the filter was observed except rounding up of the epithelium compared with the control cultures. The inhibitory effect of the bacterial collagenase was also demonstrated in 13-day gland. We, therefore, assumed that collagen could be a crucial factor for the cleft formation. We next carried out an experiment for estimating collagenase activity in 12-day glands to further support our hypothesis. By using [3H]collagen with a high specific activity as substrate, a latent collagenase activity could be detected in the tissue homogenate, but not in the supernatant obtained after thawing the frozen tissue. As shown in Table 1, [3H]collagen-degrading activity of the homogenate activated by 4-aminophenylmercuric acetate was inhibited by each dithiothreitol, o-phenanthroline and EDTA, all of which have been known as inhib-
204
DEVELOPMENTAL
BIOLOGY
VONIME
FIG. 3. Effects of C2ostridial collagenase on salivary gland morphogenesis. A control hr, and (C) 25 hr of culture. A collagenase-treated (5 pg/ml) gland is shown successively glands were cultured as in Fig. 1 except that younger glands were used. Bar = 0.2 mm.
itors for various animal collagenases (Sellers et al., 1977; Gross, 1982). The sustained activity might come from the telopeptide or the denatured portion of the substrate, which could be susceptible to proteases in the homogenate (Table 1). The most important finding was that the enhanced activity by the activator was markedly inhibited by CI, which did not have any effect on serine proteases such as trypsin and elastases, and metallopro-
REQUIREMENTS
TABLE 1 FOR TISSUE COLLACENASE
OFU-DAY
Requirements Complete ~ APMA* + Dithiothreitol” + o-phenanthroline” + EDTA’ + CId
SALIVARY
ACTIVITY
GLANDS
[3H]Collagen-degrading (dpm)
activity”
325,560 135,860 114,780 158,020 115,480 87,480
’ The values were corrected for the activity without the tissue mogenate (368,910 dpm). * APMA, 4-aminophenylmercuric acetate, 1 mM. ’ 10 mA4. ’ CI, bovine dental pulp collagenase inhibitor, 4.4 pg/ml.
ho-
113, 1986
12-day gland is shown successively at (A) 1 hr, (B) 18 at (D) 1 hr, (E) 18 hr, and (F) 25 hr of culture. Salivary
teases such as thermolysin, leucocyte gelatinase and Clostridial collagenase (Kishi and Hayakawa, 1984). This fact seems consistent with the observation that CI stimulates the cleft formation (Fig. l), probably because of inhibiting the collagen degradation due to tissue collagenase activity. We, therefore, tried to demonstrate the possibility that collagen accumulates in the glands treated with CI by using scanning electron microscope. The glands were cultured for 17 hr in the presence and absence of CI, fixed, and dehydrated as described under Materials and Methods. After cracking the fixed tissues with a fine needle, we were often able to observe the epithelial and mesenchymal surfaces, which had faced each other in the glands. Collagen-like fibrils, 50-100 nm in diameter, were found scattered over the epithelial surface, probably on the basal lamina, of the lobule of the control gland (Fig. 4A). In the CI-treated gland, there were much abundant fibrils, thicker in diameter, on the epithelium as shown in Fig. 4B. The fibrils were localized rather at the bases of the clefts and also were either attached to the epithelial surfaces or present in bundles free on the surfaces. Similar changes of the fibrils were observed on the mesenchymal surfaces. The study on the glands cultured with Clostridial collagenase was hampered by the difficulty to obtain the tissue surfaces fractured at
F~ti. 4. Scanning and the presence
electron micrographs of epithelial surfaces of CI, 5 n&ml (B). (A) X1040; (B) X990
facing
the epithelial-mesenchymal interface by unknown reason. Extensive observations of their transversely fractured surfaces, however, indicated that the fibrils in the collagenase-treated glands were much less abundant than those in the untreated ones. These results support the notion that CI affected the tissue collagenase acting on the interstitial collagens and changed the fibrous architecture at the epithelial-mesenchymal interface, leading to the alteration of the branching pattern, although it is uncertain whether CI did not inhibit the type IV collagenase or other enzyme(s) in the salivary gland. To pursue the mechanism of collagen involvement, it was essential to ask whether the reagents employed in
EFFECTS (‘OKPOKATION
OF
TABLE COLLACENASE
Chtritlitrl
INTO ACID-INSOLUBLE
2 AND CI ON [3H]T~~~~n~~~
MATERIELS
OF E-DAY
IN-
GLANDS
AF-
TER 14 hr OF CULTIVATIOK Number
Treatment None Collagenase (5 &ml)
glands
of
used
Radioactivity (dpm/one gland)
25
4240
25
3960
25
4290
Relative rate (57 ) 100 93.4
Cl (5 &ml)
101
Kotcx Glands were cultured for 14 hr on Millipore filter assemblies as described under Materials and Methods and then pulse-labeled for 2 hr with 10.0 pCi/ml of [3H]thymidine in the same medium. Acidinsoluble materials were counted after washing the glands once with 107 and twice with 5% trichloroacetic acid.
to the
mesenchyme
of the glands
cultured
for
17 hr in the absense
(A)
this study might influence the cleft formation through the changes of cell proliferation rates. Two hours of pulse-labeling experiments with [3H]thymidine after 14 hr of cultivation showed that 93 and 101% of the radioactivity were incorporated into DNAs of collagenaseand CI-treated glands, respectively, compared with that of the control (Table 2), indicating that the differences in DNA synthetic rates might not be involved in these phenomena. Regarding the distribution of collagen in the extracellular matrix, immunofluorescent staining by using antibodies for type I and III collagens revealed that they are specifically located at the cleft and stalk (Bernfield et al., 1984). In the present study, introduction of CI into the culture system of 12-day salivary gland gave rise to accumulation of collagen-like fibrils at the epithelialmesenchymal interface, resulting in the enhanced cleft formation. This observation seems in accord with the finding of a tissue collagenase activity in 12-day glands. Furthermore, the fact that removal of collagen by the bacterial collagenase prevented the cleft formation confirmed the evidence reported by Spooner and Faubion (1980) that the use of L-azetidine 2-carboxylic acid inhibited the cleft formation. These findings, together with the data by immunofluorescent staining (Bernfield et al., 1984), further support the hypothesis that accumulation of collagen is one of the motive forces which ensure the cleft formation in the epithelium. We thank Dr. M. R. Bernfield and Dr. S. D. Banerjee for valuable discussions and the critical reading of the manuscript; Dr. K. Yamazaki-Yamamoto for the help in electron microscopic observation and Dr. K. Takata for the discussion and the use of scanning electron mi-
206
DEVELOPMENTAL
BIOLOGY
croscope of Radioisotope Center, Nagoya University; Mr. M. Yamagata and Dr. K. Kimata for the assay of hyaluronidase activity in the bacterial collagenase preparation. Y.N. and F.S. wish to express their thanks to Professor S. Suzuki for the continuous support. Supported by the grants from the Ministry of Education, Science and Culture of Japan, and the Ishida Foundation. REFERENCES BANERJEE, S. D., and BERNFIELD, M. R. (1979). Developmentally regulated neutral hyaluronidase activity during epithelial-mesenchymal interaction, J. Cell Biol. 83,469a. BANERJEE, S. D., COHN, R. H., and BERNFIELD, M. R. (1977). Basal lamina of embryonic salivary epithelia. Production by the epithelium and role in maintaining lobular morphology. J. CeZl Biol. 73, 445463. BERNFIELD, M. R., and BANERJEE, S. D. (1982). The turnover of basal lamina glycosaminoglycan correlates with epithelial morphogenesis. Dev. Biol. 90, 291-305. BERNFIELD, M., BANERJEE, S. D., KODA, J. E., and RAPRAEGER, A. C. (1984). Remodeling of the basement membrane as a mechanism of morphogenetic tissue interaction. In “The Role of Extracellular Matrix in Development” (R. L. Trelstad, ed.), pp. 545-572. Liss, New York. BERNFIELD, M. R., COHN, R. H., and BANERJEE, S. D. (1973). Glycosaminoglycans and epithelial organ formation. Am. Zool. 13,1067-1083. GROBSTEIN, C. (1967). Mechanisms of organogenetic tissue interaction. Natl. Cancer Inst. Monogr. 26,279-299.
VOL~JME GROBSTEIN, phogenesis 626-628.
113, 1986 C., and COHEN, of embryonic
J. (1965). salivary
Collagenase: epithelium
GROSS, J. (1982). An essay on biological “Cell Biology of Extracellular Matrix” Plenum, New York. KISHI, J., and HAYAKAWA, T. (1984). of bovine dental pulp collagenase 395-404.
Effect in vitro.
degradation (E. D. Hay,
Purification inhibitor.
on the Science
mor150,
of collagen. Zn ed.), pp. 217-253.
and characterization J. Biochem. (Tokyo)
96,
NOGAWA, H. (1983). Determination of the curvature of epithelial cell mass by mesenchyme in branching morphogenesis of mouse salivary gland. J. Embryol. Exp. Mwrphol. 73, 221-232. RICE, R. H., and MEANS, G. E. (1971). Radioactive labeling of proteins in vitro. J Biol. Chem. 246, 831-832. SELLERS, A., CARTWRITE, E., MURPHY, G., and REYNOLDS, J. J. (1977). Evidence that latent collagenases are enzyme-inhibitor complexes. Biochem. J. 163, 303-307. TACK, B. F., DEAN, J., EILAT, D., LORENZ, P. E., and SCHECHTER, A. N. (1980). Tritium labeling of proteins to high specific radioactivity by reductive methylation. J. Biol. Chem. 255, 8842-8847. TERATO, K., NAGAI, Y., KAWANISHI, K., and YAMAMOTO, S. (1976). A rapid assay method of collagenase activity using ‘*C-labeled soluble collagen as substrate. Biochim. Biophys. Acta 445,753-762. WESSELLS, N. K., and COHEN, J. (1968). Effects of collagenase on developing epithelia in vitro: Lung, ureteric bud, and pancreas. Deu. Biol. 17, 294-309.
BIOLOGY
113,201-206
(1986)
Collagenase Inhibitor Stimulates Cleft Formation during Early Morphogenesis of Mouse Salivary Gland
A collagenase inhibitor obtained from the culture medium of bovine dental pulp markedly enhanced the cleft formation of mouse embryonic salivary gland epithelium when the inhibitor was included in the culture medium for 12-day and 13-day salivary glands. Determination of collagenase activity using [3H]collagen as substrate indicated that there was a latent collagenase activity in 12-day glands. In addition, a highly purified Clmstritlicrl collagenase freed from protease and hyaluronidase activities, strongly inhibited initiation of the cleft formation of the 12-day epithelium. Scanning electron microscopic observation showed that abundant collagen-like fibrils were seen on the epithelium in the collagenaseinhibitor-treated glands compared to those in the control. These findings suggest that during early morphogenesis d 19x6 Academic Press. Inc. tissue collagenase may regulate the cleft formation in the epithelium. INTRODUCTION
The branching morphogenesis of mouse embryonic salivary gland has been examined extensively in analysis of epithelial-mesenchymal interaction. Specifically, these studies have assessed the mechanism of interlobular cleft formation necessary for the proper morphogenesis of the epithelium (Grobstein, 1967). It is well established that such interactions are achieved through the extracellular space, in which glycosaminoglycans and collagens are the major components (Bernfield ef al., 1984). Among those, basal lamina glycosaminoglycan plays an essential role in maintaining epithelial morphology, since enzymatic removal of the glycosaminoglycan disrupts the basal lamina of the 13-day mouse salivary epithelium and causes loss of its multilobular morphology (Bernfield et al., 1973). Furthermore; the turnover rates of the glycosaminoglycan fraction at the distal ends of the lobules, interlobular clefts, and stalk are different from each other (Bernfield and Banerjee, 1982). A novel neutral hyaluronidase in the mesenchymal tissue (Banerjee and Bernfield, 1979) is believed to be a key enzyme for such precise metabolism of the glycosaminoglycan, leading to the characteristic branching of mouse salivary gland. Another important substance in the extracellular space responsible for the branching morphogenesis of salivary gland is collagen, which is present in the mesenchyme and on the epithelial surface, especially on the cleft and stalk regions (Spooner and Faubion, 1980). In the experiment using L-azetidine 2-carboxylic acid (Spooner and Faubion, 1980), collagen synthesis and secretion were suggested to be involved in the branching 201
phenomenon. If collagenolytic activity in the salivary gland is concerned with clefting, introduction of a collagenase inhibitor into the culture medium might modulate the branching pattern. In addition, since the inhibition experiments of branching with some bacterial collagenase preparations (Grosbtein and Cohen, 1965; Wessells and Cohen, 1968) were obscure due to possible contamination of mucopolysaccharidases (Bernfield et al., 1972), it seemed to be important to re-evaluate the effect of a bacterial collagenase on the epithelial morphology. In this report, we will describe a stimulative effect on the cleft formation of an animal collagenasespecific inhibitor (CI), which was isolated from the culture medium of bovine dental pulp and purified to a single protein (Kishi and Hayakawa, 1984), and an inhibitory effect on the cleft formation of a highly purified collagenase from Clostridiunz hisfolyticum. We will also present evidence demonstrating the existence of collagenase activity in 12-day salivary gland and some scanning electron micrographs of the glands treated with CI. MATERIALS
AND
METHODS
Mnteriuls. [3H]NaBH4 (8.7 Ci/mmole) and [3H]thymidine (23.8 CVmmole) were obtained from New England Nuclear Corporation, Boston, Massachusetts. A collagenase inhibitor (CI) was isolated from the culture medium of bovine dental pulp and purified to homogeneity as described previously (Kishi and Hayakawa, 1984). A highly purified collagenase from Clostridium hysfolyficum, which was freed from caseinolytic activity, was kindly donated by Dr. T. Ohya and Dr. N. Yokoi of 0012-1606/86 Copyright All rights
$3.00
Cc 1986 hy Academx Press, Inc. of reproduction in any form rrserved
202
DEVELOPMENTAL
BIOLOGY
Amano Pharmaceutical Company, Japan. No hyaluronidase activity of the collagenase preparation was detected by using [3H]hyaluronic acid as substrate. 3H-Labeled co lla g en was prepared by the modified method essentially the same as those described by Rice and Means (1971) and Tack et al. (1980). Eight milligrams of acid-soluble native type I collagen (calf skin), which was dissolved in 2 ml of 0.01 M acetic acid, was mixed with 6 ml of 0.2 M borate buffer, pH 9.0, and placed on ice for 1 hr. To the vessel, 400 ~1 of 0.04 M formaldehyde was added immediately followed by the addition of 0.725 ml of [3H]NaBH4 solution (25.0 mCi of [3H]NaBH4 was diluted with 0.01 mg of unlabeled NaBH4 in 1.55 ml of 0.01 M NaOH). After 2 min, the same volume of [3H]NaBH4 solution was again added and the vessel was kept on ice for 30 min. The [3H]collagen thus formed was precipitated by the addition of acetic acid and NaCl so as to make final concentrations of 0.5 M and lo%, respectively, and was collected by centrifugation at 10,OOOg for 20 min. The precipitate was washed with 10% NaCl in 0.5 M acetic acid and dissolved in 6 ml of 0.5 M acetic acid. The solution was dialyzed against 0.01 M acetic acid and lyophilized (6 mg). The specific activity was determined to be 6.1 X lo7 dpm/mg. Culture and labeling. Salivary glands were obtained from DDY strain mice and the day of discovery of the vaginal plug was designated as Day 0. The glands were routinely cultured on ultrathin Millipore filter (THWP, 0.45-wrn pore size) assemblies as previously described (Banerjee et al, 1977) in BME medium supplemented with 50 pg/ml ascorbic acid and 10% fetal calf serum. Labeling medium contained 10.0 pCi/ml [3H]thymidine. The rate of DNA synthesis was determined by counting the radioactivity of acid-insoluble materials after 2 hr’s pulse with [3H]thymidine. Living cultures of salivary glands were monitored by taking photographs at appropriate intervals on an Olympus light microscope BH-2 equipped with an automated exposure unit. Determination of number of clefts into epithelium. Quantitative determination of epithelial morphogenesis was done by counting the number of clefts in the photographs of the living cultures taken under standardized conditions of lighting and magnification. Clefts to be counted were vertical ones at 24 hr of culture in the plane of the photograph, which had length of more than half of the radius of the lobule. Determination of collagenase activity. Twelve-day glands used for the assay were removed from embryos in Tyrode’s solution and transferred together with a small amount of the solution into an Eppendorf tube and frozen until use. The whole tissues (886 glands) were thawed and centrifuged to remove the supernatant. The glands were suspended in 2 ml of 30 mM Tris-HCl, pH
VOL~JME
113, 1986
7.8, containing 0.5 M NaCl, 5 mM CaClz and 0.1% Triton X-100 and homogenized in a Potter-Elvehjem homogenizer. The homogenate was used as enzyme source. The complete reaction mixture consisted of 350 ~1 of 0.1 M Tris-HCl, pH 7.8, containing 0.2 M NaCl, 5 mM CaC& and 1 Mglucose, 100 ~1 of [3H]collagen solution (6.1 X lo6 dpm) and 50 ~1 of tissue homogenate (-22 glands) which was preincubated at 35°C for 3.5 hr in the presence of 1 mM 4-aminophenylmercuric acetate. The reaction mixture was incubated at 35°C for 19 hr and 50 ~1 of 0.2% unlabeled collagen solution were added. The degraded collagen was then extracted into 50% dioxane solution as previously described (Terato et al., 1976) and the radioactivity was counted. Under these conditions, [3H]collagen-degrading activity of the homogenate activated by 4-aminophenylmercuric acetate was linear up to 100 ~1 of the homogenate. Processing for scanning electron fmicroscopy. Salivary glands for scanning electron microscopy were fixed with 1.6% glutaraldehyde, 1.6% paraformaldehyde in 0.1 M phosphate buffer, pH 7.2. Glands were postfixed for 1 hr with 2% osmium tetroxide in 0.1 M phosphate buffer, dehydrated with ethanol, and transferred in isoamyl acetate. Glands thus treated were critical point dried from liquid carbon dioxide, mounted onto aluminum stubs, and cracked with a tiny tip of the fine needle under microscope. The fragments of glands were sputter-coated with gold and viewed with an Akashi ALPHA-30 scanning electron microscope at 30 kV. RESULTS
AND
DISCUSSION
We first assessed the effect of CI on the branching morphogenesis of mouse salivary gland. Salivary glands at two different stages were cultured on Millipore filter assemblies in the presence and absence of 5 pg/ml CI (Fig. 1). In the control culture of 12-day glands, an average of 2 clefts was seen at 24 hr of culture (Figs. lAC). On the contrary, when CI was added, more clefts were consistently formed (Figs. lD-F), strongly suggesting a possible involvement of collagenase activity during early branching morphogenesis. It is of interest that lobules varying in size were often formed in the CItreated glands (Figs. lE, F) as compared with relatively uniform lobules in the control cultures (Figs. lB, C), possibly indicating that no pre-determined sites for clefts exist on the epithelial surface. The stimulative effect was further confirmed when CI was added in the culture medium of 13-day glands. However, since the branching pattern of the glands at this stage became more complex in culture, further studies using CI and the bacterial collagenase were performed mostly on 12 day glands. To find out an optimal concentration, different amounts of CI were added to the culture medium and
FIG. culture.
1. CI effects A CI-treated
on salivary (5 &ml)
gland gland
morphogenesis. A control 12-day gland is shown successively at (A) 1 hr, (B) 18 hr, is shown successively at (II) 1 hr, (E) 18 hr, and (F) 25 hr of culture. Bar = 0.2 mm.
clefts were counted, As seen in Fig. 2, a saturation was attained at 2-5 pg/ml and approximately 5 clefts on the average were observed as compared with 2 in the absence of CI. The concentration of 5 pg/ml was effective for 13day gland and used throughout this study. To confirm the collagen involvement, we re-examined the effect of Clostridial collagenase on salivary gland morphogenesis. On a dose-response experiment, 5 pg/
yjil..i Collagenase
Inhibitor
(uglml)
FIG. 2. Effects of various concentrations of CI on cleft formation of the glands cultured for 24 hr. Quantitative determination of epithelial morphogenesis was done as described under Materials and Methods.
and
(C) 25 hr of
ml collagenase was shown to be sufficient for the inhibition of initiation of the cleft formation for 24 hr of cultivation (Fig. 3). Of particular interest is that although each 12-day gland had a distorted lobule sometimes having several indentations, addition of the enzyme gave rise to formation of round or oval lobule with a smooth contour (Fig. 3E). This implies a possible role of collagen as a source of constraint probably necessary for the cleft formation (Nogawa, 1983). It should be added here that, in the enzyme-treated glands, neither distinct flattening of the glands nor spreading of the mesenchyme on the filter was observed except rounding up of the epithelium compared with the control cultures. The inhibitory effect of the bacterial collagenase was also demonstrated in 13-day gland. We, therefore, assumed that collagen could be a crucial factor for the cleft formation. We next carried out an experiment for estimating collagenase activity in 12-day glands to further support our hypothesis. By using [3H]collagen with a high specific activity as substrate, a latent collagenase activity could be detected in the tissue homogenate, but not in the supernatant obtained after thawing the frozen tissue. As shown in Table 1, [3H]collagen-degrading activity of the homogenate activated by 4-aminophenylmercuric acetate was inhibited by each dithiothreitol, o-phenanthroline and EDTA, all of which have been known as inhib-
204
DEVELOPMENTAL
BIOLOGY
VONIME
FIG. 3. Effects of C2ostridial collagenase on salivary gland morphogenesis. A control hr, and (C) 25 hr of culture. A collagenase-treated (5 pg/ml) gland is shown successively glands were cultured as in Fig. 1 except that younger glands were used. Bar = 0.2 mm.
itors for various animal collagenases (Sellers et al., 1977; Gross, 1982). The sustained activity might come from the telopeptide or the denatured portion of the substrate, which could be susceptible to proteases in the homogenate (Table 1). The most important finding was that the enhanced activity by the activator was markedly inhibited by CI, which did not have any effect on serine proteases such as trypsin and elastases, and metallopro-
REQUIREMENTS
TABLE 1 FOR TISSUE COLLACENASE
OFU-DAY
Requirements Complete ~ APMA* + Dithiothreitol” + o-phenanthroline” + EDTA’ + CId
SALIVARY
ACTIVITY
GLANDS
[3H]Collagen-degrading (dpm)
activity”
325,560 135,860 114,780 158,020 115,480 87,480
’ The values were corrected for the activity without the tissue mogenate (368,910 dpm). * APMA, 4-aminophenylmercuric acetate, 1 mM. ’ 10 mA4. ’ CI, bovine dental pulp collagenase inhibitor, 4.4 pg/ml.
ho-
113, 1986
12-day gland is shown successively at (A) 1 hr, (B) 18 at (D) 1 hr, (E) 18 hr, and (F) 25 hr of culture. Salivary
teases such as thermolysin, leucocyte gelatinase and Clostridial collagenase (Kishi and Hayakawa, 1984). This fact seems consistent with the observation that CI stimulates the cleft formation (Fig. l), probably because of inhibiting the collagen degradation due to tissue collagenase activity. We, therefore, tried to demonstrate the possibility that collagen accumulates in the glands treated with CI by using scanning electron microscope. The glands were cultured for 17 hr in the presence and absence of CI, fixed, and dehydrated as described under Materials and Methods. After cracking the fixed tissues with a fine needle, we were often able to observe the epithelial and mesenchymal surfaces, which had faced each other in the glands. Collagen-like fibrils, 50-100 nm in diameter, were found scattered over the epithelial surface, probably on the basal lamina, of the lobule of the control gland (Fig. 4A). In the CI-treated gland, there were much abundant fibrils, thicker in diameter, on the epithelium as shown in Fig. 4B. The fibrils were localized rather at the bases of the clefts and also were either attached to the epithelial surfaces or present in bundles free on the surfaces. Similar changes of the fibrils were observed on the mesenchymal surfaces. The study on the glands cultured with Clostridial collagenase was hampered by the difficulty to obtain the tissue surfaces fractured at
F~ti. 4. Scanning and the presence
electron micrographs of epithelial surfaces of CI, 5 n&ml (B). (A) X1040; (B) X990
facing
the epithelial-mesenchymal interface by unknown reason. Extensive observations of their transversely fractured surfaces, however, indicated that the fibrils in the collagenase-treated glands were much less abundant than those in the untreated ones. These results support the notion that CI affected the tissue collagenase acting on the interstitial collagens and changed the fibrous architecture at the epithelial-mesenchymal interface, leading to the alteration of the branching pattern, although it is uncertain whether CI did not inhibit the type IV collagenase or other enzyme(s) in the salivary gland. To pursue the mechanism of collagen involvement, it was essential to ask whether the reagents employed in
EFFECTS (‘OKPOKATION
OF
TABLE COLLACENASE
Chtritlitrl
INTO ACID-INSOLUBLE
2 AND CI ON [3H]T~~~~n~~~
MATERIELS
OF E-DAY
IN-
GLANDS
AF-
TER 14 hr OF CULTIVATIOK Number
Treatment None Collagenase (5 &ml)
glands
of
used
Radioactivity (dpm/one gland)
25
4240
25
3960
25
4290
Relative rate (57 ) 100 93.4
Cl (5 &ml)
101
Kotcx Glands were cultured for 14 hr on Millipore filter assemblies as described under Materials and Methods and then pulse-labeled for 2 hr with 10.0 pCi/ml of [3H]thymidine in the same medium. Acidinsoluble materials were counted after washing the glands once with 107 and twice with 5% trichloroacetic acid.
to the
mesenchyme
of the glands
cultured
for
17 hr in the absense
(A)
this study might influence the cleft formation through the changes of cell proliferation rates. Two hours of pulse-labeling experiments with [3H]thymidine after 14 hr of cultivation showed that 93 and 101% of the radioactivity were incorporated into DNAs of collagenaseand CI-treated glands, respectively, compared with that of the control (Table 2), indicating that the differences in DNA synthetic rates might not be involved in these phenomena. Regarding the distribution of collagen in the extracellular matrix, immunofluorescent staining by using antibodies for type I and III collagens revealed that they are specifically located at the cleft and stalk (Bernfield et al., 1984). In the present study, introduction of CI into the culture system of 12-day salivary gland gave rise to accumulation of collagen-like fibrils at the epithelialmesenchymal interface, resulting in the enhanced cleft formation. This observation seems in accord with the finding of a tissue collagenase activity in 12-day glands. Furthermore, the fact that removal of collagen by the bacterial collagenase prevented the cleft formation confirmed the evidence reported by Spooner and Faubion (1980) that the use of L-azetidine 2-carboxylic acid inhibited the cleft formation. These findings, together with the data by immunofluorescent staining (Bernfield et al., 1984), further support the hypothesis that accumulation of collagen is one of the motive forces which ensure the cleft formation in the epithelium. We thank Dr. M. R. Bernfield and Dr. S. D. Banerjee for valuable discussions and the critical reading of the manuscript; Dr. K. Yamazaki-Yamamoto for the help in electron microscopic observation and Dr. K. Takata for the discussion and the use of scanning electron mi-
206
DEVELOPMENTAL
BIOLOGY
croscope of Radioisotope Center, Nagoya University; Mr. M. Yamagata and Dr. K. Kimata for the assay of hyaluronidase activity in the bacterial collagenase preparation. Y.N. and F.S. wish to express their thanks to Professor S. Suzuki for the continuous support. Supported by the grants from the Ministry of Education, Science and Culture of Japan, and the Ishida Foundation. REFERENCES BANERJEE, S. D., and BERNFIELD, M. R. (1979). Developmentally regulated neutral hyaluronidase activity during epithelial-mesenchymal interaction, J. Cell Biol. 83,469a. BANERJEE, S. D., COHN, R. H., and BERNFIELD, M. R. (1977). Basal lamina of embryonic salivary epithelia. Production by the epithelium and role in maintaining lobular morphology. J. CeZl Biol. 73, 445463. BERNFIELD, M. R., and BANERJEE, S. D. (1982). The turnover of basal lamina glycosaminoglycan correlates with epithelial morphogenesis. Dev. Biol. 90, 291-305. BERNFIELD, M., BANERJEE, S. D., KODA, J. E., and RAPRAEGER, A. C. (1984). Remodeling of the basement membrane as a mechanism of morphogenetic tissue interaction. In “The Role of Extracellular Matrix in Development” (R. L. Trelstad, ed.), pp. 545-572. Liss, New York. BERNFIELD, M. R., COHN, R. H., and BANERJEE, S. D. (1973). Glycosaminoglycans and epithelial organ formation. Am. Zool. 13,1067-1083. GROBSTEIN, C. (1967). Mechanisms of organogenetic tissue interaction. Natl. Cancer Inst. Monogr. 26,279-299.
VOL~JME GROBSTEIN, phogenesis 626-628.
113, 1986 C., and COHEN, of embryonic
J. (1965). salivary
Collagenase: epithelium
GROSS, J. (1982). An essay on biological “Cell Biology of Extracellular Matrix” Plenum, New York. KISHI, J., and HAYAKAWA, T. (1984). of bovine dental pulp collagenase 395-404.
Effect in vitro.
degradation (E. D. Hay,
Purification inhibitor.
on the Science
mor150,
of collagen. Zn ed.), pp. 217-253.
and characterization J. Biochem. (Tokyo)
96,
NOGAWA, H. (1983). Determination of the curvature of epithelial cell mass by mesenchyme in branching morphogenesis of mouse salivary gland. J. Embryol. Exp. Mwrphol. 73, 221-232. RICE, R. H., and MEANS, G. E. (1971). Radioactive labeling of proteins in vitro. J Biol. Chem. 246, 831-832. SELLERS, A., CARTWRITE, E., MURPHY, G., and REYNOLDS, J. J. (1977). Evidence that latent collagenases are enzyme-inhibitor complexes. Biochem. J. 163, 303-307. TACK, B. F., DEAN, J., EILAT, D., LORENZ, P. E., and SCHECHTER, A. N. (1980). Tritium labeling of proteins to high specific radioactivity by reductive methylation. J. Biol. Chem. 255, 8842-8847. TERATO, K., NAGAI, Y., KAWANISHI, K., and YAMAMOTO, S. (1976). A rapid assay method of collagenase activity using ‘*C-labeled soluble collagen as substrate. Biochim. Biophys. Acta 445,753-762. WESSELLS, N. K., and COHEN, J. (1968). Effects of collagenase on developing epithelia in vitro: Lung, ureteric bud, and pancreas. Deu. Biol. 17, 294-309.