- Email: [email protected]
Immunocytochemical localization of vitamin D-dependent calcium-binding protein in the yolk sac of the rat
Placenta (1983), 4, 263-270
Immunocytochemical Localization of Vitamin D-dependent Calcium-binding Protein in the Yolk Sac of the Rat A. C . D E L O R M E
a, P. C A S S I E R b, B. G E N Y c
& H. MATHIEU
a
"Unite de Recherches sur le MEtabolisme Hydro-MinEral, I N S E R M U. 12o, 44 Chemin de Ronde, 78IIO Le VEsinet, France bCytophysiologie des Arthropodes, E.R.A. 62o, Universit~ P. et M. Curie, Paris, France cCin~tique cellulaire en HEmatologie et Canc&ologie, F.R.A. 4, H6pital Saint-Louis, Paris, France Correspondence should be addressed to A. C. Delorme
INTRODUCTION In the pregnant rat, the early 'inversion of the germ layers' results in the development of the yolk sac which persists until the end of pregnancy (Panigel et al, I963; Steven and Morriss, 1975). The yolk sac is a true placental site involved in the physiological transfer of nutrients between the mother and the embryo (Morriss, I975). It is composed of an external columnar endodermal epithelium resting on a layer of mesenchyme which contains the vitelline vessels. The opposite face exposed to the exocoelom is covered by a thin layer of mesothelium (Padykula, Deren and Wilson, i966 ). In addition, the endodermal cells of the yolk sac resemble structurally and functionally the intestinal absorptive cells of the newborn, which are also of endodermal origin (Morriss, I975; Gardner and Papaioannou, i975; Rodewald, I980 ). The rat yolk sac is now considered as a target organ for vitamin D 3. Receptors for 1,25dihydroxyvitamin D 3 (I,25(OH)zD3) , the hormonal form of vitamin D3, have recently been reported in this organ (Danan, Delorme and Cuisinier-Gleizes, i98-i ). Furthermore, as in the intestine, which is one of the major target organs for vitamin D3, a vitamin D-dependent calcium-binding protein (CaBP) similar to the intestinal one is present in the rat yolk sac (Delorme, Marche and Garel, 1979; Garel et al, i981 ). CaBP is described as one of the molecular expressions of the steroid-like action of the 1,25(OH)zD 3 (Lawson, 1978). In the present work, in order to specify which of the cell types in the yolk sac respond to 1,25(OH)2D3, we have used immunocytochemical methods to determine the cellular localization of CaBP in the rat yolk sac.
MATERIALS AND M E T H O D S Tissue s a m p l i n g and cell isolation procedures Tissue sampling. Yolk sacs were obtained from normal pregnant rats of the Wistar strain between days t5 and 19 of gestation. The visceral yolk sacs were carefully dissected free from fetus,
263
z64
A. C. Delorme, P. Cassier, B. Geny, H. Mathieu
amnion and placenta, rinsed in phosphate-buffered saline (PBS) and fixed immediately. Two fixatives were used: calcium-neutral formol (Baker's fluid) for six hours or Bouin's fluid for 24 hours. The samples were then dehydrated in graded ethanol solutions and embedded in paraffin. Paraffin sections (7 ~tm thick) were deparaffinized in xylene. Endodermal cellpreparation. Pregnant rats at I6 to 19 days of gestation were anaesthetized with ether, the uterine wall was slit and each conceptus was exposed with its placenta attached. A thread ligature was tied around the umbilical cord at its insertion into the placenta. The fetal membranes and the cord were cut above the ligature to remove the placenta and to obtain a closed yolk sac encapsulating the fetus. This preparation was suspended by the thread and washed in isotonic NaCI solution. Closed yolk sacs were preincubated at 37~ in a shaking water-bath in phosphate buffer (8mMKHzPO4, 5.6mMNa2HPO4, pH 7-3) containing I. 5 mM KCI, 96mM NaCI and 27 mM Na citrate for 15 minutes. They were then incubated in PBS added with 1.5 mM E D T A for two hours, being transferred at 3o-minute intervals to fresh solution. After 1.5 to 2 hours of incubation, the sacs were removed and the resulting cell suspension was centrifuged for io minutes at ioo g. The cell pellet was resuspended in PBS containing i mM CaCI 2 and I mM MgCIz, pH 7.4, and 5 to io #1 of the cell suspension were spread on microscope slides and fixed by air-drying. The cell smear was treated with cold acetone for four minutes, then quickly washed with PBS and processed for immunological reactions. The quality of the cell preparation was examined under the light microscope following May-Grunwald-Giemsa staining. In one set of experiments, after removal of endodermal cells, the yolk sacs were fixed in Bouin's fluid, dehydrated in ethanol, embedded in paraffin and stained for histological observation using Heidenhan's azan method.
Anti-CaBP antiserum The preparation of the rabbit antiserum against pure intestinal rat CaBP has been described elsewhere (Marche et al, i977). Ouchterlony double immunodiffusion plates contained 1.5 per cent agarose in PBS pH 7-4 and were incubated overnight. The antigenic preparations were ioo ooog supernatants prepared from rat duodenal mucosa or yolk sacs (Delorme, Marche and Garel, I979).
Immunocytochemical localization of CaBP Immunocytological reactions were investigated by indirect methods. Tissue sections or endodermal cell smears were first exposed to the whole anti-rat-CaBP rabbit serum, diluted i : io to i : 2o, for three hours at 2o~ and washed 3 x 1o minutes with PBS. This was followed by a second incubation for x hour with the globulin fraction of a sheep antiserum to rabbit immunoglobulins linked either to fluorescein isothiocyanate ( F I T C ) or to horseradish peroxidase (Institut Pasteur, Paris) at a dilution of i : 2o0. Slides were rinsed 3 x IO minutes in PBS. For the peroxidase technique, enzyme activity was revealed by staining at pH 7-4 with diaminobenzidine and hydrogen peroxide (o.o2 per cent final concentration) (Graham and Karnovsky, 1966 ). Five types of control were used: (a) first incubation with non-immune rabbit serum; (b) preincubation for one hour at room temperature with non-immune rabbit serum; (c) experiments without anti-CaBP antiserum for the non-specific reaction; (d) experiments with only PBS (autofluorescence) or with diaminobenzidine solution (endogenous peroxidase activity).
z65
Localization o["Ca-binding Protein in Rat Yolk Sac
RESULTS Figure I shows the specificity of the rabbit anti-CaBP antiserum. When the crude i0o ooo g supernatant from duodenal mucosa or yolk sac reacted with the antiserum, a single precipitin line was obtained. T h e antigen in duodenal mucosa and yolk sac is identical, as indicated by the complete coalescence of the single precipitin lines. No immunoreactive material was detected in rat renal cortex, liver, spleen, pancreas, muscle, skin, cerebellum, mammary gland. No precipitate occurred when antigens reacted against non-immune rabbit serum.
8
C
[I
Figure I. Ouchterlony double immunodiffusion.The centre well contained 15I~1of rabbit antiserum against rat pure CaBP. The wells labelledA and D contained 15 ,ul of non-immune rabbit serum. The wells B and E contained 2.6/lg CaBP in 15 ktl aliquots of rat yolk sac ioo ooog supernatant. The well C contained 2.5/~g CaBP in t 5 kd aliquots of rat duodenal mucosa supernatant.
Although similar results were obtained with both methods of fixation (see Materials and Methods), the intensity of specific staining was better in the Baker's fluid-fixed tissues and therefore reference will be made to this material. Figure 2 shows the results obtained with yolk sac sections. Positive reactions were found in the villous zone facing the chorioallantoic placenta and in the smooth zone. T h e endodermal cell layer was specifically stained with the anti-CaBP antiserum. T h e majority of the endodermal cells contained CaBP, but the staining intensity seemed to vary from one cell to another. CaBP was Shown to be an intracellular protein, located in the cytoplasm of endodermal cells. There was no CaBP in the nucleus or on the external surface of the cells. Staining ofendodermal cells was not seen when non-immune rabbit serum was used in place of the anti-CaBP immune serum. These appearances did not vary between the 15th and the i9th days of gestation. The same results were obtained by both immunofluorescence and immunoperoxidase methods. Figure 3 shows that E D T A in isotonic buffer was an effective agent for dissociating endodermal cells from the closed yolk sac. It allowed a release of pure endodermal cells without damaging the underlying layers. This prevented contamination by mesothelial elements and by blood cells. The isolated endodermal cells attached to a microscope slide seemed to be morphologically intact. Although they had lost their columnar shape, they remained asymmetric with their prominent nucleus at the basal pole and their extensive vacuolation at the apex (Figure 3b). When a smear of pure endodermal cells was exposed to the specific anti-CaBP antiserum (Figure 4) a positive reaction was observed, indicating the presence of CaBP within
~,66
A. C. Delorme, P. Cassier, B. Geny, H. Mathieu
Figure z. Cytochemica| localization of CaBP in rat yolk sac by (A) immunofluorescence ( x 25o ) or by (B) immunoperoxidase staining ( x 250 ). e = endodermal cells; m = mesothelium; v = vitelline vessel.
Localization of Ca-binding Protein in Rat Yolk Sac
.
9
9 .~
9
267
~',i
."
9 r+
9
+"
.
.~+..
Figure 3. Preparation of endodermal cells isolated from rat yolk sac. (A) Aspect of the yolk sac after EDTA treatment ( • 250 ) (staining with Heidenhain's azan method), e = endodermal cells; m = mesothetium+ (B) An isolated endodermal cell ( x xooo) (May-Grunwald-Giemsa staining).
268
A. C. Delorme, P. Cassier, B. Geny, H. Mathieu
Figure 4. Localization of CaBP in isolated endodermal cells by immunofluorescence. Incubation with anti-CaBP antiserum ( • 400).
these isolated endodermal cells. The staining was distributed throughout the cytoplasm without a preferential location of CaBP within the cytoplasm. The nucleus appeared unstained. The results obtained with immunoperoxidase and immunofluorescence were the same. The controls were negative.
DISCUSSION The immunocytochemical techniques employed proved suitable for CaBP localization studies in the rat yolk sac, although the level of immunoreactive CaBP is lower than in the intestine (Delorme, Marche and Garel, 1979; Garel et al, 198i ). Both indirect immunofluorescence and immunoperoxidase staining methods showed that CaBP is located only in the endodermal cells of the yolk sac, and this finding was confirmed by the use ofendodermal cells isolated from the yolk sac. It is well known from electron microscopical studies that the endodermal cells have the cytological characteristics of absorptive cells. They are highly polarized with numerous microvilli at the apex (Lambson, I966; Padykula, Deren and Wilson, i966 ). Numerous studies have demonstrated that they possess intense absorptive abilities (Padykula, Deren and Wilson, 1966; King and Enders, i97o; Seibel, i974; Rodewald, x98o). Therefore, in respect of transport between mother and fetus, the endodermal cells are the most important elements of the yolk sac (Morriss, 1975). Our demonstration that CaBP is localized in the functional absorptive cells of the yolk sac parallels findings in the mammalian intestine. In pigs (Arnold, Kovacs and Murray, 1976) and
Locah'zation of Ca-binding Protein in Rat Yolk Sac
269
rats (Marche, Le Guern and Cassier, 1979) , the CaBP is found only in the columnar absorptive intestinal cells. Furthermore, it has been shown that the rat yolk sac in vitro is able to transport calcium from the maternal to the fetal side (Kernis and Johnson, I969). The amount of CaBP is correlated with active calcium absorption in the intestine although its exact role in calcium intestinal transport remains unclear (Armbrecht et al, 1979; Feher and Wasserman, 1979). Thus, our data are consistent with an in vivo calcium transport through th e 3)olk sac endodermal .cells by a mechanism that is at least partially similar to the intestinal one. The presence of CaBP in the endodermal cells suggests that they are the vitamin D target cells containing the 1,25(OH)2D 3 receptors and are involved in vitamin D-mediated processes. At the subcellular level, our results on tissue sections and on isolated cells indicate that immunoreactive CaBP is rather uniformly distributed throughout the endodermal cytoplasm and absent from the nucleus. The use of the two different experimental procedures (tissue sections and isolated cells) with similar results makes an artefact in the localization unlikely. A cytoplasmic localization has been shown in the rat intestine (Marche, Le Guern and Cassier, 1979) and in other vitamin D target tissues (Taylor, 1981; Jande, Tolnai and Lawson, I 9 8 I ; Rhoten and Christakos, 198I ). But in the rat isolated intestinal cell, CaBP seems to be specifically located in the terminal web region (Marche, Cassier and Mathieu, I980 ). We have not observed such a preferential localization of CaBP in the yolk sac endodermal cells. This difference may be ascribed to the structural dissimilarities between the two cell types. The external appearance of these two types of epithelial cells, when they are isolated, is suggestive of structural differences. The isolated intestinal cells remain rectangular (Bretscher and Weber, I978; Marche, Cassier and Mathieu, 1980), but the isolated yolk sac endodermal cells appear less rigid and lose their columnar shape. The differences are particularly striking in the structural organization of the cell apex. Indeed, it is now clear that the terminal web is a well-defined region in the highly differentiated epithelial cell of the small intestine (Hull and Staehelin, 1979). By contrast, such organization in the apical cytoplasm of the yolk sac endodermal cells has not, to our knowledge, been described.
SUMMARY The yolk sac from normal pregnant rats contains a vitamin D-dependent calcium-binding protein (CaBP). To localize CaBP within the yolk sac, a rabbit antiserum specifically directed against purified rat CaBP, in iiadirect immunofluorescence and immunoperoxidase, has been used. Using yolk sac sections, positive reactions indicative of the presence of CaBP were found only in the endodermal cell layer. A method for isolation of pure endodermal cells, using E D T A treatment, was developed. On smears of isolated endodermal cells exposed to anti-CaBP antiserum, CaBP was distributed throughout the endodermal cytoplasm; no CaBP was detected in the nucleus. Our results suggest that the endodermal cells, which are the functional absorptive cells of the yolk sac, may play a role in vitamin D-mediated processes.
ACKNOWLEDGEMENTS We arc grateful to J. Bigler and A. Porteu de la Morandi+re for their technical assistance. We wish to thank P. Schuman, C. Chabot and M. Courat for the preparation of the manuscript. This work was supported by grant CRL 79.5.22o.4 from lnstitut National de la Sant~ et de la Recherche M6dicale and grant BDR 8o.7.o355 from the D+16gation G~n+rale fi [a Recherche Scientifique et Technique.
A. C. Delorme, P. Cassier, B. Geny, H. Mathieu
270 REFERENCES
Armbrecht, H. J., Zenser, T. V., Bruns, M. E. H. & Davis, B. B. (1979) Effect of age on intestinal calcium absorption and adaptation to dietary calcium. American Journal of Physiology, 236, E769-E779. Arnold, B. M., Kovacs, K. & Murray, T. M. (I976) Cellular localization of intestinal calcium-binding protein in pig duodenum. Digestion, i4, 77-84. Bretscher, A. & Weber, K. 0978) Localization of actin and microfilament-associated proteins in the microvilli and terminal web of the intestinal brush border by immunofluorescence microscopy. Journal of Cellular Biology, 79, 839-845. Danan, J. L., Delorme, A. C. & Cuisinier-Gleizes, P. 0 9 8 0 Biochemical evidence for a cytoplasmic lCC,25dihydroxyvitamin D 3 receptor-like protein in rat yolk sac. Journal of Biological Chemistry, 256 , 4847-485o. Delorme, A. C., Marche, P. & Garel, J. M. (i979) Vitamin D--dependent calcium-binding protein. Changes during gestation, prenatal and postnatal development in rats. Journal of Developmental Physiology, i, 181-194. Feher, J. J. & Wasserman, R. H. 0979) Calcium absorption and intestinal calcium-binding'protein: quantitative relationship. American Journal of Physiology, 236 , E556-E56I. Gardner, R. & Papaioannou, V. E. 0975) Differentiation in the trophectoderm and inner cell mass. In The Early Development of Mammals (Ed.) Balls, M. & Wild, A. E. pp. to7-132. Cambridge: Cambridge University Press. Garel, J. M., Delorme, A. C., Marche, P., N'Guyen, T. M. & Garabedian, M. 0981) Vitamin D 3 metabolite injections to thyroparathyroidectomized pregnant rats: effects on calcium-binding proteins of maternal duodenum and of feto-placental unit. Endocrinology, lO9, 284-289. Graham, A. C. & Karnovsky, M. J. (1966) The early stages of absorption of injected horseradish peroxidase in the proximal tubules of mouse kidney: ultrastructural cytochemistry by a new technique. Journal of Histochemistry and Cytochemistry, t4, 291-325. Hull, B. E. & Staehelin, L. A. (t979) The terminal web. A re-evaluation of the structure and function. Journal of" Cellular Biology, 81, 67-82. Jande, S. S., Toinai, S. & Lawson, D. E. M. (i981) lmmunohistochemical localization of vitamin D-dependent calcium-binding protein in duodenum, kidney, uterus and cerebellum of chickens. Histochemistry, 71, 99-1 I6. Kernis, M. M. & Johnson, E. M. (1969) Effects oftrypan blue and Niagara blue 2B on the in vitro absorption of ions by the rat visceral yolk sac. Journal of Embryology and Experimental Morphology, 22, 115-125. King, B. F. & Enders, A. C. ( 197o) Protein absorption and transport by guinea-pig visceral yolk sac placenta. American Journal of Anatomy, 129, 261-288. Lambson, R. O. (1966) An electron microscopic visualization of transport across rat visceral yolk sac. AmericanJournal of Anatomy, 118, 21-52. Lawson, D. E. i . (1978) Biochemical responses of the intestine to vitamin D. In Vitamin D (Ed.) Lawson, D. E. M. pp. 167-2oL London: Academic Press. Marcbe, P., Pradelles, P., Gros, C. & Thomasset, M. 0977) Radioimmunoassay for a vitamin D-dependent calcium-binding protein in rat duodenal mucosa. Biochemical and Biophysical Research Communications, 76, '
1020--I O26.
i a r c b e , P., Le Guern, C. & Cassier, P. 0979) Immunocytochemical localization of a calcium-binding protein in the rat duodenum. Cell Tissue Research, 197, 69-77. Marche, P., Cassier, P. & Mathieu, H. (198o) Intestinal calcium-binding protein. A protein indicator of enterocyte maturation associated with the terminal web. Cell Tissue Research, 212, 63-72. Morriss, G. (1975) Placental evolution and embryonic nutrition. In Comparative Placentation. Essays in Structure and Function (Ed.) Steven, D. H. pp. 87-1o 7. London: Academic Press. Padykula, H. A., Deren, J. J. & Wilson, T. H. (t966) Development of structure and function in the mammalian yolk sac. I. Developmental morphology and vitamin Bt2 uptake of the rat yolk sac. Developmental Biology, 13, 3 t 1-348. Panigel, M., Nguyen, H. Anh & Le Goascogne, C. 0963) Interpr6tation au microscope de la signification omphaloplacentaire des organites et inclusions cellulaires de l'endoderme vitellin chez les rongeurs. Annales des Sciences naturelles, Zoologie et Biologie animale, 12~me s~rie 5, n~ 451-46o. Rhoten, W. B. & Christakos, S. (I981) Immunocytochemical localization of vitamin D-dependent calcium-binding protein in mammalian nephron. Endocrinology, lO9, 981-983. Rodewald, R. (198o) lmmunoglobulin transmission in mammalian young and the involvement of coated vesicles. In Coated Vesicles (Ed.) Ockleford, C. D. & Whyte, A. pp. 69-1Ol. Cambridge: Cambridge University Press. Seibel, W. (1974) An ultrastructural comparison of the uptake and transport of horseradish peroxidase by the rat visceral yolk sac placenta during mid and late gestation. American Journal of Anatomy, 14o, 213-236. Steven, D. & Morriss, G. 0975) Development of the foetal membranes. In Comparative Placentation. Essays in Structure and Function (Ed.) Steven, D. H. pp. 58-86. London: Academic Press. Taylor, A. N. (I98Q Immunocytochemical localization of the vitamin D-induced calcium-binding protein: relocation of antigen during frozen section processing. Journal of Histochemistry and Cytochemistry, 29, 65-73.
Immunocytochemical Localization of Vitamin D-dependent Calcium-binding Protein in the Yolk Sac of the Rat A. C . D E L O R M E
a, P. C A S S I E R b, B. G E N Y c
& H. MATHIEU
a
"Unite de Recherches sur le MEtabolisme Hydro-MinEral, I N S E R M U. 12o, 44 Chemin de Ronde, 78IIO Le VEsinet, France bCytophysiologie des Arthropodes, E.R.A. 62o, Universit~ P. et M. Curie, Paris, France cCin~tique cellulaire en HEmatologie et Canc&ologie, F.R.A. 4, H6pital Saint-Louis, Paris, France Correspondence should be addressed to A. C. Delorme
INTRODUCTION In the pregnant rat, the early 'inversion of the germ layers' results in the development of the yolk sac which persists until the end of pregnancy (Panigel et al, I963; Steven and Morriss, 1975). The yolk sac is a true placental site involved in the physiological transfer of nutrients between the mother and the embryo (Morriss, I975). It is composed of an external columnar endodermal epithelium resting on a layer of mesenchyme which contains the vitelline vessels. The opposite face exposed to the exocoelom is covered by a thin layer of mesothelium (Padykula, Deren and Wilson, i966 ). In addition, the endodermal cells of the yolk sac resemble structurally and functionally the intestinal absorptive cells of the newborn, which are also of endodermal origin (Morriss, I975; Gardner and Papaioannou, i975; Rodewald, I980 ). The rat yolk sac is now considered as a target organ for vitamin D 3. Receptors for 1,25dihydroxyvitamin D 3 (I,25(OH)zD3) , the hormonal form of vitamin D3, have recently been reported in this organ (Danan, Delorme and Cuisinier-Gleizes, i98-i ). Furthermore, as in the intestine, which is one of the major target organs for vitamin D3, a vitamin D-dependent calcium-binding protein (CaBP) similar to the intestinal one is present in the rat yolk sac (Delorme, Marche and Garel, 1979; Garel et al, i981 ). CaBP is described as one of the molecular expressions of the steroid-like action of the 1,25(OH)zD 3 (Lawson, 1978). In the present work, in order to specify which of the cell types in the yolk sac respond to 1,25(OH)2D3, we have used immunocytochemical methods to determine the cellular localization of CaBP in the rat yolk sac.
MATERIALS AND M E T H O D S Tissue s a m p l i n g and cell isolation procedures Tissue sampling. Yolk sacs were obtained from normal pregnant rats of the Wistar strain between days t5 and 19 of gestation. The visceral yolk sacs were carefully dissected free from fetus,
263
z64
A. C. Delorme, P. Cassier, B. Geny, H. Mathieu
amnion and placenta, rinsed in phosphate-buffered saline (PBS) and fixed immediately. Two fixatives were used: calcium-neutral formol (Baker's fluid) for six hours or Bouin's fluid for 24 hours. The samples were then dehydrated in graded ethanol solutions and embedded in paraffin. Paraffin sections (7 ~tm thick) were deparaffinized in xylene. Endodermal cellpreparation. Pregnant rats at I6 to 19 days of gestation were anaesthetized with ether, the uterine wall was slit and each conceptus was exposed with its placenta attached. A thread ligature was tied around the umbilical cord at its insertion into the placenta. The fetal membranes and the cord were cut above the ligature to remove the placenta and to obtain a closed yolk sac encapsulating the fetus. This preparation was suspended by the thread and washed in isotonic NaCI solution. Closed yolk sacs were preincubated at 37~ in a shaking water-bath in phosphate buffer (8mMKHzPO4, 5.6mMNa2HPO4, pH 7-3) containing I. 5 mM KCI, 96mM NaCI and 27 mM Na citrate for 15 minutes. They were then incubated in PBS added with 1.5 mM E D T A for two hours, being transferred at 3o-minute intervals to fresh solution. After 1.5 to 2 hours of incubation, the sacs were removed and the resulting cell suspension was centrifuged for io minutes at ioo g. The cell pellet was resuspended in PBS containing i mM CaCI 2 and I mM MgCIz, pH 7.4, and 5 to io #1 of the cell suspension were spread on microscope slides and fixed by air-drying. The cell smear was treated with cold acetone for four minutes, then quickly washed with PBS and processed for immunological reactions. The quality of the cell preparation was examined under the light microscope following May-Grunwald-Giemsa staining. In one set of experiments, after removal of endodermal cells, the yolk sacs were fixed in Bouin's fluid, dehydrated in ethanol, embedded in paraffin and stained for histological observation using Heidenhan's azan method.
Anti-CaBP antiserum The preparation of the rabbit antiserum against pure intestinal rat CaBP has been described elsewhere (Marche et al, i977). Ouchterlony double immunodiffusion plates contained 1.5 per cent agarose in PBS pH 7-4 and were incubated overnight. The antigenic preparations were ioo ooog supernatants prepared from rat duodenal mucosa or yolk sacs (Delorme, Marche and Garel, I979).
Immunocytochemical localization of CaBP Immunocytological reactions were investigated by indirect methods. Tissue sections or endodermal cell smears were first exposed to the whole anti-rat-CaBP rabbit serum, diluted i : io to i : 2o, for three hours at 2o~ and washed 3 x 1o minutes with PBS. This was followed by a second incubation for x hour with the globulin fraction of a sheep antiserum to rabbit immunoglobulins linked either to fluorescein isothiocyanate ( F I T C ) or to horseradish peroxidase (Institut Pasteur, Paris) at a dilution of i : 2o0. Slides were rinsed 3 x IO minutes in PBS. For the peroxidase technique, enzyme activity was revealed by staining at pH 7-4 with diaminobenzidine and hydrogen peroxide (o.o2 per cent final concentration) (Graham and Karnovsky, 1966 ). Five types of control were used: (a) first incubation with non-immune rabbit serum; (b) preincubation for one hour at room temperature with non-immune rabbit serum; (c) experiments without anti-CaBP antiserum for the non-specific reaction; (d) experiments with only PBS (autofluorescence) or with diaminobenzidine solution (endogenous peroxidase activity).
z65
Localization o["Ca-binding Protein in Rat Yolk Sac
RESULTS Figure I shows the specificity of the rabbit anti-CaBP antiserum. When the crude i0o ooo g supernatant from duodenal mucosa or yolk sac reacted with the antiserum, a single precipitin line was obtained. T h e antigen in duodenal mucosa and yolk sac is identical, as indicated by the complete coalescence of the single precipitin lines. No immunoreactive material was detected in rat renal cortex, liver, spleen, pancreas, muscle, skin, cerebellum, mammary gland. No precipitate occurred when antigens reacted against non-immune rabbit serum.
8
C
[I
Figure I. Ouchterlony double immunodiffusion.The centre well contained 15I~1of rabbit antiserum against rat pure CaBP. The wells labelledA and D contained 15 ,ul of non-immune rabbit serum. The wells B and E contained 2.6/lg CaBP in 15 ktl aliquots of rat yolk sac ioo ooog supernatant. The well C contained 2.5/~g CaBP in t 5 kd aliquots of rat duodenal mucosa supernatant.
Although similar results were obtained with both methods of fixation (see Materials and Methods), the intensity of specific staining was better in the Baker's fluid-fixed tissues and therefore reference will be made to this material. Figure 2 shows the results obtained with yolk sac sections. Positive reactions were found in the villous zone facing the chorioallantoic placenta and in the smooth zone. T h e endodermal cell layer was specifically stained with the anti-CaBP antiserum. T h e majority of the endodermal cells contained CaBP, but the staining intensity seemed to vary from one cell to another. CaBP was Shown to be an intracellular protein, located in the cytoplasm of endodermal cells. There was no CaBP in the nucleus or on the external surface of the cells. Staining ofendodermal cells was not seen when non-immune rabbit serum was used in place of the anti-CaBP immune serum. These appearances did not vary between the 15th and the i9th days of gestation. The same results were obtained by both immunofluorescence and immunoperoxidase methods. Figure 3 shows that E D T A in isotonic buffer was an effective agent for dissociating endodermal cells from the closed yolk sac. It allowed a release of pure endodermal cells without damaging the underlying layers. This prevented contamination by mesothelial elements and by blood cells. The isolated endodermal cells attached to a microscope slide seemed to be morphologically intact. Although they had lost their columnar shape, they remained asymmetric with their prominent nucleus at the basal pole and their extensive vacuolation at the apex (Figure 3b). When a smear of pure endodermal cells was exposed to the specific anti-CaBP antiserum (Figure 4) a positive reaction was observed, indicating the presence of CaBP within
~,66
A. C. Delorme, P. Cassier, B. Geny, H. Mathieu
Figure z. Cytochemica| localization of CaBP in rat yolk sac by (A) immunofluorescence ( x 25o ) or by (B) immunoperoxidase staining ( x 250 ). e = endodermal cells; m = mesothelium; v = vitelline vessel.
Localization of Ca-binding Protein in Rat Yolk Sac
.
9
9 .~
9
267
~',i
."
9 r+
9
+"
.
.~+..
Figure 3. Preparation of endodermal cells isolated from rat yolk sac. (A) Aspect of the yolk sac after EDTA treatment ( • 250 ) (staining with Heidenhain's azan method), e = endodermal cells; m = mesothetium+ (B) An isolated endodermal cell ( x xooo) (May-Grunwald-Giemsa staining).
268
A. C. Delorme, P. Cassier, B. Geny, H. Mathieu
Figure 4. Localization of CaBP in isolated endodermal cells by immunofluorescence. Incubation with anti-CaBP antiserum ( • 400).
these isolated endodermal cells. The staining was distributed throughout the cytoplasm without a preferential location of CaBP within the cytoplasm. The nucleus appeared unstained. The results obtained with immunoperoxidase and immunofluorescence were the same. The controls were negative.
DISCUSSION The immunocytochemical techniques employed proved suitable for CaBP localization studies in the rat yolk sac, although the level of immunoreactive CaBP is lower than in the intestine (Delorme, Marche and Garel, 1979; Garel et al, 198i ). Both indirect immunofluorescence and immunoperoxidase staining methods showed that CaBP is located only in the endodermal cells of the yolk sac, and this finding was confirmed by the use ofendodermal cells isolated from the yolk sac. It is well known from electron microscopical studies that the endodermal cells have the cytological characteristics of absorptive cells. They are highly polarized with numerous microvilli at the apex (Lambson, I966; Padykula, Deren and Wilson, i966 ). Numerous studies have demonstrated that they possess intense absorptive abilities (Padykula, Deren and Wilson, 1966; King and Enders, i97o; Seibel, i974; Rodewald, x98o). Therefore, in respect of transport between mother and fetus, the endodermal cells are the most important elements of the yolk sac (Morriss, 1975). Our demonstration that CaBP is localized in the functional absorptive cells of the yolk sac parallels findings in the mammalian intestine. In pigs (Arnold, Kovacs and Murray, 1976) and
Locah'zation of Ca-binding Protein in Rat Yolk Sac
269
rats (Marche, Le Guern and Cassier, 1979) , the CaBP is found only in the columnar absorptive intestinal cells. Furthermore, it has been shown that the rat yolk sac in vitro is able to transport calcium from the maternal to the fetal side (Kernis and Johnson, I969). The amount of CaBP is correlated with active calcium absorption in the intestine although its exact role in calcium intestinal transport remains unclear (Armbrecht et al, 1979; Feher and Wasserman, 1979). Thus, our data are consistent with an in vivo calcium transport through th e 3)olk sac endodermal .cells by a mechanism that is at least partially similar to the intestinal one. The presence of CaBP in the endodermal cells suggests that they are the vitamin D target cells containing the 1,25(OH)2D 3 receptors and are involved in vitamin D-mediated processes. At the subcellular level, our results on tissue sections and on isolated cells indicate that immunoreactive CaBP is rather uniformly distributed throughout the endodermal cytoplasm and absent from the nucleus. The use of the two different experimental procedures (tissue sections and isolated cells) with similar results makes an artefact in the localization unlikely. A cytoplasmic localization has been shown in the rat intestine (Marche, Le Guern and Cassier, 1979) and in other vitamin D target tissues (Taylor, 1981; Jande, Tolnai and Lawson, I 9 8 I ; Rhoten and Christakos, 198I ). But in the rat isolated intestinal cell, CaBP seems to be specifically located in the terminal web region (Marche, Cassier and Mathieu, I980 ). We have not observed such a preferential localization of CaBP in the yolk sac endodermal cells. This difference may be ascribed to the structural dissimilarities between the two cell types. The external appearance of these two types of epithelial cells, when they are isolated, is suggestive of structural differences. The isolated intestinal cells remain rectangular (Bretscher and Weber, I978; Marche, Cassier and Mathieu, 1980), but the isolated yolk sac endodermal cells appear less rigid and lose their columnar shape. The differences are particularly striking in the structural organization of the cell apex. Indeed, it is now clear that the terminal web is a well-defined region in the highly differentiated epithelial cell of the small intestine (Hull and Staehelin, 1979). By contrast, such organization in the apical cytoplasm of the yolk sac endodermal cells has not, to our knowledge, been described.
SUMMARY The yolk sac from normal pregnant rats contains a vitamin D-dependent calcium-binding protein (CaBP). To localize CaBP within the yolk sac, a rabbit antiserum specifically directed against purified rat CaBP, in iiadirect immunofluorescence and immunoperoxidase, has been used. Using yolk sac sections, positive reactions indicative of the presence of CaBP were found only in the endodermal cell layer. A method for isolation of pure endodermal cells, using E D T A treatment, was developed. On smears of isolated endodermal cells exposed to anti-CaBP antiserum, CaBP was distributed throughout the endodermal cytoplasm; no CaBP was detected in the nucleus. Our results suggest that the endodermal cells, which are the functional absorptive cells of the yolk sac, may play a role in vitamin D-mediated processes.
ACKNOWLEDGEMENTS We arc grateful to J. Bigler and A. Porteu de la Morandi+re for their technical assistance. We wish to thank P. Schuman, C. Chabot and M. Courat for the preparation of the manuscript. This work was supported by grant CRL 79.5.22o.4 from lnstitut National de la Sant~ et de la Recherche M6dicale and grant BDR 8o.7.o355 from the D+16gation G~n+rale fi [a Recherche Scientifique et Technique.
A. C. Delorme, P. Cassier, B. Geny, H. Mathieu
270 REFERENCES
Armbrecht, H. J., Zenser, T. V., Bruns, M. E. H. & Davis, B. B. (1979) Effect of age on intestinal calcium absorption and adaptation to dietary calcium. American Journal of Physiology, 236, E769-E779. Arnold, B. M., Kovacs, K. & Murray, T. M. (I976) Cellular localization of intestinal calcium-binding protein in pig duodenum. Digestion, i4, 77-84. Bretscher, A. & Weber, K. 0978) Localization of actin and microfilament-associated proteins in the microvilli and terminal web of the intestinal brush border by immunofluorescence microscopy. Journal of Cellular Biology, 79, 839-845. Danan, J. L., Delorme, A. C. & Cuisinier-Gleizes, P. 0 9 8 0 Biochemical evidence for a cytoplasmic lCC,25dihydroxyvitamin D 3 receptor-like protein in rat yolk sac. Journal of Biological Chemistry, 256 , 4847-485o. Delorme, A. C., Marche, P. & Garel, J. M. (i979) Vitamin D--dependent calcium-binding protein. Changes during gestation, prenatal and postnatal development in rats. Journal of Developmental Physiology, i, 181-194. Feher, J. J. & Wasserman, R. H. 0979) Calcium absorption and intestinal calcium-binding'protein: quantitative relationship. American Journal of Physiology, 236 , E556-E56I. Gardner, R. & Papaioannou, V. E. 0975) Differentiation in the trophectoderm and inner cell mass. In The Early Development of Mammals (Ed.) Balls, M. & Wild, A. E. pp. to7-132. Cambridge: Cambridge University Press. Garel, J. M., Delorme, A. C., Marche, P., N'Guyen, T. M. & Garabedian, M. 0981) Vitamin D 3 metabolite injections to thyroparathyroidectomized pregnant rats: effects on calcium-binding proteins of maternal duodenum and of feto-placental unit. Endocrinology, lO9, 284-289. Graham, A. C. & Karnovsky, M. J. (1966) The early stages of absorption of injected horseradish peroxidase in the proximal tubules of mouse kidney: ultrastructural cytochemistry by a new technique. Journal of Histochemistry and Cytochemistry, t4, 291-325. Hull, B. E. & Staehelin, L. A. (t979) The terminal web. A re-evaluation of the structure and function. Journal of" Cellular Biology, 81, 67-82. Jande, S. S., Toinai, S. & Lawson, D. E. M. (i981) lmmunohistochemical localization of vitamin D-dependent calcium-binding protein in duodenum, kidney, uterus and cerebellum of chickens. Histochemistry, 71, 99-1 I6. Kernis, M. M. & Johnson, E. M. (1969) Effects oftrypan blue and Niagara blue 2B on the in vitro absorption of ions by the rat visceral yolk sac. Journal of Embryology and Experimental Morphology, 22, 115-125. King, B. F. & Enders, A. C. ( 197o) Protein absorption and transport by guinea-pig visceral yolk sac placenta. American Journal of Anatomy, 129, 261-288. Lambson, R. O. (1966) An electron microscopic visualization of transport across rat visceral yolk sac. AmericanJournal of Anatomy, 118, 21-52. Lawson, D. E. i . (1978) Biochemical responses of the intestine to vitamin D. In Vitamin D (Ed.) Lawson, D. E. M. pp. 167-2oL London: Academic Press. Marcbe, P., Pradelles, P., Gros, C. & Thomasset, M. 0977) Radioimmunoassay for a vitamin D-dependent calcium-binding protein in rat duodenal mucosa. Biochemical and Biophysical Research Communications, 76, '
1020--I O26.
i a r c b e , P., Le Guern, C. & Cassier, P. 0979) Immunocytochemical localization of a calcium-binding protein in the rat duodenum. Cell Tissue Research, 197, 69-77. Marche, P., Cassier, P. & Mathieu, H. (198o) Intestinal calcium-binding protein. A protein indicator of enterocyte maturation associated with the terminal web. Cell Tissue Research, 212, 63-72. Morriss, G. (1975) Placental evolution and embryonic nutrition. In Comparative Placentation. Essays in Structure and Function (Ed.) Steven, D. H. pp. 87-1o 7. London: Academic Press. Padykula, H. A., Deren, J. J. & Wilson, T. H. (t966) Development of structure and function in the mammalian yolk sac. I. Developmental morphology and vitamin Bt2 uptake of the rat yolk sac. Developmental Biology, 13, 3 t 1-348. Panigel, M., Nguyen, H. Anh & Le Goascogne, C. 0963) Interpr6tation au microscope de la signification omphaloplacentaire des organites et inclusions cellulaires de l'endoderme vitellin chez les rongeurs. Annales des Sciences naturelles, Zoologie et Biologie animale, 12~me s~rie 5, n~ 451-46o. Rhoten, W. B. & Christakos, S. (I981) Immunocytochemical localization of vitamin D-dependent calcium-binding protein in mammalian nephron. Endocrinology, lO9, 981-983. Rodewald, R. (198o) lmmunoglobulin transmission in mammalian young and the involvement of coated vesicles. In Coated Vesicles (Ed.) Ockleford, C. D. & Whyte, A. pp. 69-1Ol. Cambridge: Cambridge University Press. Seibel, W. (1974) An ultrastructural comparison of the uptake and transport of horseradish peroxidase by the rat visceral yolk sac placenta during mid and late gestation. American Journal of Anatomy, 14o, 213-236. Steven, D. & Morriss, G. 0975) Development of the foetal membranes. In Comparative Placentation. Essays in Structure and Function (Ed.) Steven, D. H. pp. 58-86. London: Academic Press. Taylor, A. N. (I98Q Immunocytochemical localization of the vitamin D-induced calcium-binding protein: relocation of antigen during frozen section processing. Journal of Histochemistry and Cytochemistry, 29, 65-73.