ABSTRACT
In an investigation of cellular differentiation processes, it is of great value to be able to study the appearance of specific cell products. Systems permitting this include pancreatic exocrine cells which make zymogen granules and amylase (Grobstein, 1964), the myoblast which produces myosin (Konigsberg, 1965), the chondrocyte which makes chondroitin and chondroitin synthesizing en-zymes (Lash, Glick & Madden, 1964), and the melanocyte which makes melanin (Wilde, 1961). In other cases, such well-defined products are not known or are difficult to obtain and it is therefore necessary to turn to other methods for investigations of specific cell changes. Histochemical methods have been widely used but frequently give unspecific results. Advantages are obtained by using immuno-chemical methods which can demonstrate the appearance of organ- or tissue-specific antigen (Dumonde, 1966). The differentiation of the lens has been extensively studied by immuno-histological methods (for reviews, see Halbert & Manski, 1963; Dumonde, 1966). The appearance of kidney-specific antigens during metanephros differentiation was studied by Lahti & Saxén (1966).
The differentiation of the Müllerian epithelium in female mice has been studied by morphological, experimental, and histochemical methods (Forsberg, 1963, 1965; Forsberg & Olivecrona, 1964, 1965). In this tissue, however, no specific cell products are as yet known. Therefore we decided to study the antigen content of the Müllerian epithelium and its possible changes during differentiation.
Preparation of antigen
Several preliminary attempts to produce an immune serum by immunizing rabbits with homogenates of Mullerian epithelium from newborn mice were fruitless. We therefore tried older animals and used uterine epithelium as antigen.
Young, fertile female albino mice approximately equal in size were used. The uterus was dissected free and opened by a longitudinal incision. The epithelium was scraped free from the rest of the uterine wall with a scalpel, care being taken to avoid contaminating it with underlying tissue. It was stored at —35°C.
The epithelium was homogenized in an equal weight of cold phosphatebuffered saline, pH 71. The homogenate w’as then thoroughly emulsified in Freund’s complete adjuvant. When the two components were inseparable on water, the emulsion was considered adequate. A fresh emulsion was made for every injection.
PREPARATION OF ANTISERA
Rabbits weighing about 2 kg were used for immunization. The emulsion was injected into the foot-pads of the rabbits. Each rabbit received 0 05 ml/foot-pad, 0-2 ml in all at each injection time.
The following time-table was used:
If the test was negative, the immunization was continued with one injection a month until the serum showed activity. When antibodies were detected the rabbit was bled to death through the common carotid artery.
IMMUNOFLUORESCENCE METHOD
The tissue to be examined was immediately frozen and sectioned in a cryostat at a thickness of 10 μ. One section of the tissue to be examined was placed on a slide together with one control section from the liver and one from the musculature. The sections were thawed on to a slide and processed within 24 h. For every experiment, 8-10 slides were used.
The following procedure was followed:
1. The sections were dried for 15 min at 56°C on a sand bath.
2. Fixed in acetone, 15 min at room temperature.
3. Dried for 10 min at 56°C on the sand bath.
4. Sections dipped into buffered saline.
5. Rabbit antibody applied on the sections, 30 min at 37°C.
6. Rinsed with buffered saline.
7. Sections fixed in buffered 10% formol for 20 min at room temperature.
(Formol buffered with phosphate-buffered saline, pH 71.) When buffered formol was used and the time for fixing was less than 20 min, the non-specific fluorescence was reduced, and the histology of the sections was superior to that in sections not fixed in formol. No reduction of the specific fluorescence was seen under these circumstances (cf. Mayersbach & Schubert, 1960).
8. Sections thoroughly rinsed with buffered saline.
9. FITC-conjugated goat-antiglobulin versus rabbit globulin applied for 15 min at room temperature.
10. Sections rinsed in buffered saline.
11. Sections mounted in glycerine.
Sections prepared in this way kept their fluorescence for 3-5 days. The various rinsing procedures were made with phosphate-buffered saline, pH 71. FITCconjugated goat-antiglobulin versus rabbit globulin (Microbiological Associates, Bethesda, Maryland) was diluted 8 times. To the diluted solution of conjugated goat-antiglobulin, lissamine-rhodamine B 200 conjugated bovine albumin (Microbiological Ass.) was added (1:5 vol. of diluted solution) to provide sufficient counterstaining. The sections were studied with a Zeiss fluorescence microscope equipped with an Osram high-pressure Mercury arc Lamp HB0200, BG 3 and BG 12 exciter filters, and barrier filters 44 and 50 (‘cut off’ at 500 m/t). Photographs could be taken on 35 mm High-Speed Ektachrome film EH 135-20 (for daylight).
PREPARATION AND ABSORPTION OF THE IMMUNE GLOBULIN
The crude globulin fraction was precipitated from serum with ammonium sulphate, 40% saturation. The precipitate was dissolved in phosphate-buffered saline pH 7·1 and was used after dialysis against the same buffer. When nonpretreated immune globulin was applied to the sections, it caused an extensive non-specific staining of the whole section which had to be removed by various treatments of the globulin solution. To test the effect of these treatments a standard set-up of tissue sections was used. These consisted of adult liver, adult muscle, adult uterus, and a section through the rectum, the Müllerian vagina, and the urinary bladder of newborn female mice.
A homogenate was prepared from the upper body half of newborn mice. The same part of newborn mice bodies was used to prepare an acetone powder. This preparation was made in a similar way to the preparation of acetone powder from mouse liver (Nairn, 1964).
The globulin solution was incubated for 60 min at 37°C with an equal volume of the homogenate. After centrifugation the supernatant was incubated for 60 min at 37°C with the acetone powder (100 mg/ml, amount not critical). After centrifugation, the supernatant was ready for use. The results are summarized in Table 1. As fluorescence remained in urinary bladder epithelium, in rectum connective tissue, and in muscular connective tissue, the incubation homogenate was altered thus: I part newborn mice-homogenate excluding uterus and vaginal anlage+ I part rectum homogenate from newborn mice+ 1 part urinary bladder homogenate from newborn mice+ 4 parts buffered saline pH 7·1 (w/w for every part). Otherwise the incubation scheme was that followed before: incubation with homogenate, centrifugation, incubation with acetone powder, centrifugation. The sections showed no unspecific fluorescence, but good fluorescence in the uterine epithelium and the Müllerian epithelium (Table 1). Immune globulin solutions prepared in this latter way were used as standard starting material for every part of the investigation. A newly prepared solution (prepared at the latest on the preceding day) was used.
TEST FOR SPECIFICITY OF THE SYSTEM
The following conditions should be fulfilled to ensure specificity of the system:
(1) Specificity of the immune globulin solution, obtained from rabbits immunized against uterine epithelium and pretreated in the way described in the preceding paragraph—(a) Non-immune globulin pretreated in exactly the same way as immune globulin should give no fluorescence whatsoever, (b) Immune globulin treated with homogenate of uterine epithelium (60 min at 37°C) prior to the application on the tissue sections should show no fluorescence.
(2) Specificity of goat-antiglobulin versus rabbit globulin. Tissue sections treated with buffered saline instead of immune globulin should show no fluorescence.
All these tests were negative; thus, the fluorescence seen after treating sections with immune globulin is significant.
TEST FOR PRECIPITATING ANTIBODIES
Immune globulin solution was tested for presence of precipitating antibodies by agar gel diffusion according to Ouchterlony’s (1962) micro-method with standard equipment (LKB, Stockholm).
Uterine epithelium obtained in the same way as for immunization was extracted in 0·9% NaCl and used as antigen solution. Extracts from adult mouse liver and skeletal muscle were also tested.
Immune globulin and non-immune globulin solutions, both non-pretreated and pretreated as above, were used.
A dilution of antigen solutions was tested against the globulin solutions as well as a dilution of the globulin solutions against the antigen solutions.
The untreated immune globulin solution contained precipitating antibodies reacting with adult liver and adult skeletal muscle as well as with uterine epithelium. Pretreated immune globulin contained no precipitating antibodies.
REACTIONS RELATED TO THE SPECIFIC ANTIBODY TESTED IN DIFFERENT ORGANS
Immune globulin solution pretreated with homogenate and acetone powder, as described in an earlier paragraph, was used. The object was to test whether antibodies reacting with antigens in the uterine epithelium and Müllerian epithelium also reacted with antigens in other tissues or organs.
Various organs from adult and newborn mice were tested. The results are summarized in Table 2. In the kidney tubules and jejunal epithelium the whole The antigen with which the specific antibody reacts thus occurs in different organs. A certain difference in distribution, as recorded in our system, seems to exist between newborn and adult mice.
When the immune globulin was incubated prior to application on the tissue sections with adult jejunal epithelium or newborn kidney, all reactivity of the immune globulin solution disappeared. This was true for fluorescence in all tissues mentioned in Table 2, including uterine epithelium and the Müllerian epithelium of newborn mice.
No fluorescence was seen when the anti-mouse uterine immune globulin was tested on sections from the rat uterus.
STUDIES ON THE ANTIGEN CONTENT OF THE MULLERIAN EPITHELIUM DURING DEVELOPMENT USING THE SPECIFIC ANTIBODY
Neither the Mullerian nor the Wolffian ducts exhibited any fluorescence with immune globulin in 14- and 16-day foetuses. The Mullerian epithelium was still negative for fluorescence in 17-day female foetuses, but at 18 days there was an indication of the thin superficial fluorescent zone in the Mullerian epithelium, both in its uterine and vaginal region. The sinus epithelium showed no fluorescence. In male foetuses of the latter stages, both the Wolffian and the degenerative Müllerian epithelia were negative.
In newborn female mice, the immune globulin gave a strong fluorescence in the superficial part of the pseudostratified Mullerian epithelium in the anterior part of the vaginal anlage (the Müllerian vagina: Plate 1, figs. A, B) and in the uterus. This fluorescence remained when the immune globulin was diluted to 1:40. The sinus epithelium in the urogenital sinus and in the short solid sinus vagina was completely negative for fluorescence. In 5and 6-day-old mice, a strong fluorescence was seen as a thin superficial zone in the anterior part of the vagina (Plate 1, fig. E), in the fornices, and in the uterus, but not in the now lumen-containing posterior part of the vagina (Plate 1, fig. F). In 20-day-old females, there was a strong fluorescence in the fornices (Plate J, figs. C, D) and in the uterus. Except in the fornices, the vaginal epithelium was negative for fluorescence. All sections of uterine epithelium from adult animals in different phases of the oestrous cycle showed a fluorescent zone in the superficial part of the epithelium.
Further studies were made to test whether the specific antibody reacted with any antigens in the male genital system. Sections from newborn and adult males were studied. The urethra, ejaculatory ducts, deferent ducts, prostrata, and seminal vesicles were negative for fluorescence, both in the newborn and the adult.
DISCUSSION
By immunizing rabbits with uterine epithelium from adult fertile mice, it was possible to prepare an immune globulin containing a specific antibody against uterine antigen. However, this antigen is not exclusive for uterine epithelium as it also occurs in other organs with a somewhat different pattern in newborn and adult animals (Table 2). Such cross-reactions between different organs and tissues are known also from other systems (cf. Dumonde, 1966; Weinberger & Boss, 1966; Clayton, Campbell & Truman, 1968). The occurrence of fluorescence in glomeruli from newborn and adult mice, but its absence in kidney tubules from adult animals, underlines the difference in antigen content between glomeruli and tubules (cf. Dumonde, 1966). In the case of the submaxillary gland epithelium, the antigen character also seems to change during development from the newborn to the adult stage.
A definite specific fluorescence was seen in the Müllerian epithelium in newborn female young. It occurred in a thin superficial zone in the whole Müllerian epithelium, not only in its uterine but also in its vaginal part. The latter, the Mullerian vagina, is easily delimited from the sinus epithelim in the sinus vagina in the posterior part of the vaginal anlage. The sinus epithelium did not exhibit any fluorescence.
A similar superficial occurrence of fluorescence as in the Müllerian and uterine epithelium was observed by Okada (1965) when studying the distribution of kidney-specific antigen in chick mesonephric tubules. It is interesting that electron microscope observations indicate a specialization of the apical part of kidney cells (cf. Okada, 1965). A similar specialization can also be seen in the apical part of mouse uterine epithelial cells, at least in oestrus and after oestrogen treatment of spayed animals (Nilsson, 1958a, b). The lumen cell membrane has many microvilli, and the surface is covered by a substance consisting of thin dense strands. There is a collection of small vacuoles in the apical part of the cell. Nothing is known about the submicroscopic structure of the Mullerian epithelium.
In the neonatal period, the epithelium in the Müllerian vagina undergoes a vigorous proliferation (Forsberg & Olivecrona, 1965) thereby forming two epithelial zones. The superficial zone can be considered as a matrix zone from which cells move basally and form the basal zone. The two zones are morphologically distinct. Later, the mitotic rate decreases in the superficial zone, the cells decrease in height, the zones become confluent, and the originally high pseudostratified columnar epithelium changes into a typical prepubertal vaginal epithelium. In the uterine cervix and in the fornices, the superficial zone retains much of its Müllerian character, with high cells in the prepubertal period (Forsberg, 1963).
These morphological changes are reflected in changes in the occurrence of uterine antigen in the vaginal Mullerian epithelium. The cells moving basally into the basal zone, forming the later basal layer of the vaginal epithelium, do not show any uterine antigen content in this investigation. These cells, from this point of view, thus become similar to the characteristically negative sinus epithelium, which gives rise to the epithelium in the posterior part of the vagina.
Parallel to the changes in the morphological character of the cells in the superficial zone, their specific antigen content changes, but in the fornices and in the cervix, where the superficial cells retain their morphological Mullerian character for a longer time, they also retain their specific uterine antigen content. In conclusion the Mullerian epithelium of female mice does not contain uterine antigen up to the newborn stage. Thereafter it appears in the whole length of the Miillerian epithelium only to disappear later from its vaginal part during the transformation process in this region.
The occurrence of this uterine antigen in the Müllerian epithelium does not run parallel to the distribution of previously studied enzymes or PAS-positive material in the vagina and vaginal anlage (Forsberg & Olivecrona, 1964, 1965; Forsberg, 1967), For instance, in young mice about 20 days old, there is no longer any fluorescence in the vaginal epithelium, except in the fornices. The superficial cell layer contains a material which can be stained by the periodic acid-Schiff reaction (Forsberg & Olivecrona, 1965). This could be taken as an indication that the antigen studied is not related to any epithelial mucin. However, it is possible that if the antigen is related to a mucin, its antigen character has been changed at some time between 6 and 20 days after birth. Kent (1961, 1963) has studied the antigenic nature of epithelial mucins and found them to have varying degrees of organ specificity. In uterine epithelium from adult mice, Fuxe & Nilsson (1963) found an apical rim of PAS-positive substance. The greatest amount occurred in oestrus, and in dioestrus granules were seen only occasionally.
With the Ouchterlony gel diffusion technique, it has been shown that rat uterine fluid contains at least one protein component not occurring in serum (Albers & Neves e Castro, 1961). Rabbit uterine fluid has three precipitin fines not shared by serum (Stevens, Hafs & Hunter, 1964). However, in this investigation no precipitating antibodies could be demonstrated by Ouchterlony’s micro-method.
SUMMARY
A specific antibody was prepared by immunizing rabbits with uterine epithelium from adult fertile female mice. Attention was concentrated on the distribution of the uterine antigen in different organs in foetal, newborn, immature, and adult animals, especially during the development of the mouse uterovaginal region. The antibody reacted with Miillerian epithelium in newborn and older female mice but not with sinus or Wolffian epithelium. It cross-reacts with antigen in some other organs, with a different pattern in newborn and adult animals.
RÉSUMÉ
Eludes par immunofluorescence de la distribution de l’antigène de l’épithélium de l’utérus de souris chez des souris foetales, néo-natales et adultes
Un anticorps spécifique a été obtenu en injectant à des lapins de l’épithélium utérin de souris adultes fertiles.
La distribution de l’antigène, dans différents organes des animaux injectés, a été étudiée, principalement pendant le développement de la région utérovaginale. Une réaction s’observe entre l’épithélium Mullérien des femelles néo-natales et adultes, et l’antigène; par contre, on n’observe pas de réaction avec l’épithélium du sinus ni avec l’épithélium Wolffien. L’antigène réagit également avec l’antigène d’autres organes selon une topographie différente chez des animaux adultes ou nouveaux-nés.
ACKNOWLEDGEMENTS
This investigation was supported by grants from the Swedish Medical Research Council (12X-579-03) and the Swedish Cancer Society (67:46).
REFERENCES
These micropholographs are black and white negative copies from multi-colour Ektachrome film. Bright fluorescence here reproduces as black.
Figs. A, B. Sections from the anterior part of the vaginal anlage, newborn mouse. Treatment of a section with immune globulin results in a bright fluorescence in the superficial part of the p scud os t ratified columnar Müllerian epithelium (fig. A). Non-immune globulin gives no fluorescence (fig. B).
Figs. C, O. Sections through the vaginal fornix, 20-day-old mouse. The section in fig. C is treated with immune globulin which gives a strong superficial fluorescence in the epithelium. No fluorescence is seen after treating a section with non-immune globulin (fig. D).
Figs. E, F. Fig. E is a section from the anterior part of the vagina, Fig. F from the posterior part of the vagina, 5-day-old mouse. Both sections are treated with immune globulin. In the anterior part of the vagina (Müllerian epithelium) there is fluorescence in the superficial part of the epithelium. This fluorescence is lacking in the posterior vaginal part, derived from sinus epithelium.
These micropholographs are black and white negative copies from multi-colour Ektachrome film. Bright fluorescence here reproduces as black.
Figs. A, B. Sections from the anterior part of the vaginal anlage, newborn mouse. Treatment of a section with immune globulin results in a bright fluorescence in the superficial part of the p scud os t ratified columnar Müllerian epithelium (fig. A). Non-immune globulin gives no fluorescence (fig. B).
Figs. C, O. Sections through the vaginal fornix, 20-day-old mouse. The section in fig. C is treated with immune globulin which gives a strong superficial fluorescence in the epithelium. No fluorescence is seen after treating a section with non-immune globulin (fig. D).
Figs. E, F. Fig. E is a section from the anterior part of the vagina, Fig. F from the posterior part of the vagina, 5-day-old mouse. Both sections are treated with immune globulin. In the anterior part of the vagina (Müllerian epithelium) there is fluorescence in the superficial part of the epithelium. This fluorescence is lacking in the posterior vaginal part, derived from sinus epithelium.