mouse oviduct and uterus associated with pregnancy block by autologous antiprogesterone monoclonal antibody. A. Whyte, C. Yang*, F. Rutter and R.B. Heap.
Journal of Reproductive Immunology, 11 (1987) 209-219 Elsevier Scientific Publishers Ireland Ltd.
209
JRI 00481
Lectin-binding characteristics of mouse oviduct and uterus associated with pregnancy block by autologous antiprogesterone monoclonal antibody A. Whyte, C. Yang*, F.
Rutter and R.B. Heap
AFRC Institute of Animal Physiology and Genetics Research, Babraham, Cambridge CB2 4AT (U.K.) (Accepted for publication 13 April 1987)
Oviducts and uteri were removed from BALB/cJ and F1 (CBA/Cax BALB/cJ) mice at known stages post coitum and following treatment with an anfiprogesterone monoclonal antibody (DB3) or a non-specific immunoglobulin (DNP). Thin sections of tissue were prepared and reacted with fluorescent conjugates of a wide range of lectins to determine if saccharide alterations were associated with the pregnancy-inhibiting effect of the DB3 antibody in BALB/c, but not FI, individuals. The ampullary region of the DB3-treated BALB/c mice showed the most marked changes, with an almost total inhibition of lectin binding, particularly for N-acetylglucosamine residues. There was also a reduced affinity for a lectin reactive with N-acetylgalactosamine in the uteri of DB3treated BALB/c mice, associated with an extended expression during gestation of this saccharide in the proximal region of the oviduct in such mice. These are the first.biochemical alterations in reproductive tract epithelia to be associated with the efficacy of the DB3 antibody in preventing pregnancy.
Correspondence to: Dr. A. Whyte. *Present address: Tianjin Institute for Family Planning, I0 Ying Shui Road, Nan Kai District, Tianjin, People's Republic of China. 0165-0378/87/$03.50 © 1987 Elsevier Scientific Publishers Ireland Ltd. Published and Printed in Ireland
210
Key words: antiprogesterone monoclonal antibody; pregnancy block; lectins; mouse reproductive tract.
Introduction
The establishment of pregnancy in BALB/c mice can be blocked by passive immunization with an antiprogesterone monoclonal antibody at 32 h post coitum (Wang et al., 1984). Following such immunization, normal cleavage and embryonic development is arrested, and endometrial sensitivity and the initiation of implantation are reduced or prevented (Rider et al., 1985). The antiprogesterone monoclonal antibody is also effective in blocking pregnancy in another inbred strain (CBA), but is less effective in FI hybrid animals (females from mating CBA males with BALB/c females) (Wang et al., 1986). The results suggest that antibody binding of circulating progesterone prevents the normal activation of target cell receptors which normally initiates the processes leading to implantation. There was no evidence that the pregnancy blocking effect was the result of altered tubal transport (Rider et al., 1987). Evidence suggests that the glycosylation of uterine and oviducal epithelia revealed by lectin binding changes qualitatively during pregnancy (Lee et al., 1983; Thie et al., 1986), although the significance of these changes remains unclear. Equine endometrium has been shown to have specialized regions with greater affinity for lectins around the time of implantation when compared to endometrium from non-pregnant animals (Whyte and Allen, 1985). In view of the pregnancy inhibiting effect of the antiprogesterone monoclonal antibody (Wright et al., 1982), we decided to study oviducal and endometrial glycosylation in antiprogesterone antibody treated animals (syngeneic and F~ hybrid females) to determine if altered glycosylation was associated with the difference in antibody efficacy with respect to the pregnancy block. Materials and Methods
Animals Two stocks of mature virgin mice (from Olac 1976 Ltd) were used, one inbred (BALB/cJ) and the other an FI hybrid (CBA/Ca c~x BALB/cJ 9). After overnight housing with males, the appearance of a vaginal plug was denoted day 1 of pregnancy (day 1 p.c.). Antibodies A murine IgG1 (DB3) raised against progesterone (Wright et al., 1982) was used and immunoglobulin was purified from ascitic fluid of pristane-
211
primed animals and characterized using the methods detailed by Rider et al. (1987). Animals received 250~1 i.p. which contained 9.5nmol (1.425mg) IgG. Control animals received murine myeloma ascites (MOPC 1748, anti-DNP specificity) at the same immunoglobulin cone -ntration. Injections were given 32 h p.c.
Lectin histochemistry Animals were killed at known stages post-coitum and the reproductive tract was removed. This was then separated into ampulla, isthmus I, isthmus II and uterine segments numbered consecutively from the ovary (see Nilsson and Reinius, 1969), and tissue fragments were either snapfrozen in isopentane in liquid nitrogen or fixed in Bouin's fluid or absolute methanol at 4"C overnight for histological processing. Following sectioning, the tissues were reacted with fluorescein isothiocyanate (FITC)-lectin conjugates as described elsewhere (Whyte and Robson, 1984). A list of the lectins studied and the saccharides used as specificity controls is given in Table 1. All lectins and saccharides were obtained from Sigma Ltd. Because variable amounts of autofluorescence occurred in oviducal and uterine tissue, sections were also reacted by indirect immunoperoxidase methods using antilectin antisera. Tissue was selected at random and reacted with 1, 10 and 100 I~g/ml solutions of wheatgerm lectin (WGL) and peanut lectin (PNL) in PBS containing 2% normal goat serum, then with rabbit polyclonal antisera to these lectins (BDH) at 1 : 20 dilution in PBS. This was followed by peroxidase-conjugated goat antiserum to rabbit IgG (Miles Ltd, 1:40) and the reaction developed by diaminobenzidine (BDH). The controls were sections of tissue incubated in the presence of the appropriate haptenic saccharide at 0.2 M, '~,nd the antilectin antiserum which had previously been absorbed with the lectin at a concentration of 1 mg/ml. Sections were examined on Nikon Diaphot or Leitz Diavert microscopes and photographed on Pan F (rated at 40 ASA) or XP1 (rated at 1600ASA) films. Fluorescence was assessed on a qualitative scale, - (negative) to + + + + (strongly positive), by three independent observers. Results
The appearance of the ampullary region of the oviduct is shown in Fig. la. The main features are the relatively thin wall and the long finger-like processes projecting into its lumen. Isthmus I (Fig. lc) and isthmus II (Fig. le) have fewer protuberances into the lumen and have thicker walls. The results obtained with frozen material were the same as those obtained with tissue fixed in absolute methanol. Bouin's fluid fixation
Fig. 1. (a) Cross-section of oviduct showing ampulla (A) and isthmus If*). Hazmatoxylin and eosin, x60. (b) Reverse-contrast print of amnul~ re~cted with fluorescent wheatgerm lectin (FITC-WGL). This was a + + + reaction. Note that the deposits occur predomimmfly in the intervHlom crypts of this oviduct from a day 3 p.c. DNP-treated BALBIc female, x200. The DB3-injected BALB/c equivalent animals showed no such binding of FITC-WGL. (e) H & E section of isthmus I material, x120. (d) Isthmus I material reacted with WGL (not conjugated) followed by a rabbit polyclonal antibody to WGL, then peroxidase-conjugated goat antiserum to rabbit IgG, and showing patchy localization of WGL binding to the luminalepithelium (arrow), x200. (e) H & E section of isthmus lI, x120. (f) Reaction of isthmus H tissue with FITC-WGL. Note that the reaction deposits are more apical than in (b) and are both continuous (arrow) and granular (arrowhead) in nature. Reverse contrast print, x200.
214
resulted in samples which gave variable results. Because of its reproducibility and reasonable histological preservation, methanol was routinely adopted as fixative. Of all the lectins studied (Table 1), only PNL and WGL showed reactivity that was different between antiprogesterone antibody (DB3) treated and control animals. Some lectins bound strongly to the uterine or oviducal luminal epithelium, e.g. CON A, whereas others bound very weakly or not at all (LCA and UEL, respectively). Where these other lectins did bind, no qualitative difference in fluorescence intensity was observable between tissues taken from treated and control animals. The fact that UEL failed to bind to tissue sections whereas TPL did bind to similar sections underlines the need for care in any study with lectins purporting to have similar specificities. The results with PNL and WGL showed the greatest reproducibility with total inhibition of binding to the epithelium by D-galactose for PNL and N-acetylglucosamine or chitobiose for WGL. N-Acetylneuraminic acid did not alter WGL binding, indicating that sialic acid was not involved in the affinity of this lectin. Fluorescence in tissue underlying the luminal epithelium was seen with several lectins, particularly in association with mononuclear cells, but no consistent pattern was seen. All the results reported (Table 2) relate solely to binding of the lectins to the luminal aspect of the epithelial cells. The most marked changes observed can be summarized as follows.
TABLE 1 Summary of the lectins used in this study together with their sugar specificities and the saccharides used (at 0.2 M) for competitive inhibition. The abbreviations used are: o-glcNAc, N-acetylglucosamine; /3-a-gal, fl-D-galactose; D-galNAc, N-acetylgalactosamine, a-D-man, a-D-mannose; a-D-glc, a-D-glucopyranosid¢; NANA, N-acetylneuraminic acid. Lectin source
Abbreviation
Specificity
Inhibitor
Arachis hypogaea Canavalia ensi~ormis Lens culinaris Ricinus communis (60) Ricinus communis (120) Tetragonolobus purpureas Triticum vulgaris
PNL CON A LCA RCA (II) RCA (I) TPL WGL
Ulex europaeus Glycine max Limulus polyphemus
UEL (I) GML LPL
/~-D-gal (1-3)-D-galNAc D-Galactose a-D-man; a-D-glc MethyI-D-mannoside or-D-man; a-D-glc D-Mannose o-gaINAc; ,8-D-gal o-Galactose D-Galactose /3-o-gal t~-L-fucose L-Fucose (o-glcNAc); NANA t Acetylglucosamine Chitobiose Acetylneuraminic acid L-Fucose a-L-fucose D-Galactose D-gaINAc Sialic acid NANA (acctylneuraminic acid)
2
F1
DNP
DB3
3 4 6 8 10
9.0+
4
1.4
9.5 + 1.7 9.8±0.5
10.3 + 3.2 5.8 ± 3 . 6 7.8+3.5
-
-
9.6±1.8 6.3±3.6 9.8±2.1
0 0 0
N o . of implantation sites ( m e a n + S . D . )
3 3 4 4
3 6 4
6
8 10
3
4
3
4
5
4 4
8 10
5 3
3 4
DNP
6
4 4 3 5 4
3 4 6 8 10
DB3
BALBIc
No. of mice
D a y of autopsy
Antibody treatment
Strain
-
+++ ++++ ++ -
+++ ++++ ++ + +
+++ +++ ++ + --
---± --
++++ +++ + + --
++ ++ ++ ++ ++
++ + ++ ---
++ + +++ ++ --
++++ +++ ++ ++ +
+++ +++ ++++ +++ ++
+++ +++ +++ ++ ±
+++ +++ +++ +++ +++
++ ++ + ---
++ + +++ ++++
+ ++
--
---
---+ ++
Uterus
++ ++ + ---
++ ++ + + +++
++ +++ ++ ---
++ ++ ++ ++ ++
+++ ++ + ++ +
+ ++ + ++ +++
++ ++ + + +
+++ ++ ++ ++ ++
Isthmus I
Ampulla
Isthmus II
Ampulla
Isthmus I
P e a n u t lectin
Wheatgerml e c t i n
B A L B I c a r e i n b r e d s t o c k a n d t h e F1 h y b r i d s a r e f e m a l e s f r o m m a t i n g C B A m u l e s w i t h B A L B / c f e m a l e s .
++ ++ + +++ ++
+++ ++ ++
+
++
++ ++ ++ ++ ++
+++ +++ ++ + ++
Isthmus II
+ + + + --
-++ +
--
+
+ + ± ++ +++
---+ ++
Uterus
L e c t i n - b i n d i n g c h a r a c t e r i s t i c s o f o v i d u c t a n d u t e r i n e tissues f o l l o w i n g a n t i b o d y t r e a t m e n t in m i c e . T h e d e g r e e of f l u o r e s c e n c e w i t h W G L a n d P N L is s c o r e d f r o m 0 t o + + + + in i n c r e a s i n g i n t e n s i t y . D B 3 is t h e a n t i p r o g e s t c r o n e m o n o c l o n a l a n t i b o d y a n d D N P is t h e M O P C 1 7 4 8 h y b r i d o m a c o n t r o l I g G . T h e
T~BLE
216
Acetylglucosamine distribution ( WGL binding) in oviduct No reactivity was detected in the ampullae of BALB/c mice treated with DB3 whereas it was prominent in the F1 experimental group. Reactivity in the ampullae showed a progressive decrease with time in the F1 experimental group and in the control groups of both F1 and BALB/c mice (Table 2). In both BALB/c and F~ control (DNP) groups there was equivalent intense activity in ampulla and isthmus I and II which showed a progressive decrease during days 3 to 10 p.c. (Table 2). This decrease with time was not seen in isthmus I of F~ and isthmus II of BALB/c experimental groups.
Acetylgalactosamine/ galactose distribution (PNL binding) in oviduct PNL showed a similar binding pattern to WGL except that the BALB/c experimental ampullae bound strongly at all stages. The site of lectin binding in the oviduct varied. In ampullae it occurred predominantly in the intervillous crypts (Fig. lb) whereas it was apical in distribution in isthmus II (Fig. lf).
Saccharide distribution (WGL and PNL binding) in uterus Much less reactivity with lectins was seen in the uteri at all stages except in some cases at days 8 and 10p.c. Antiprogesterone antibody treatment did not appear to affect WGL lectin binding in BALB/c animals whereas there was an indication of increased lectin binding in F1 animals where the pregnancy blocking activity of DB3 antibody is less than in BALB/c individuals. With PNL, however, there was little apparent change in F1 animals whereas there was marked binding of this lectin with increasing gestation in the BALB/c stock. DB3 antibody treatment appeared to diminish uterine PNL lectin binding in BALB/c females, unlike its slight effect in F1 females (Table 2). The value of the differences recorded in lectin affinity in these experiments is, of course, limited by the qualitative nature of the fluorescence assay. To reduce variability in the observations, three observers looked at each slide independently and the means of their results are presentedAn Table 2. It would now be possible to verify our subjective observations on this tissue by quantitative assays using, for example, lectin-affinity probes and isolation and characterization of reactive determinants. Discussion
Induction of the decidual cell reaction necessary for implantation in mice involves stimulation via uterine glycoprotein receptors containing, particularly, N-acetyl-D-glucosamine and N-acetyl-D-galactosamine
217
(Buxton and Murdoch, 1982). The regulation of the assembly of such N-linked glycoproteins in the mouse uterus is controlled by steroids; oestradiol-17~8 stimulates the incorporation of [3H]mannose into uterine glycoproteins and its action is antagonized by progesterone (Dutt et al., 1986). Progesterone alone also induces uterine species-specific glycoproteins which may be present in up to 1000-fold greater concentrations in the early stages of pregnancy compared to later in gestation (Joshi, 1983; Joshi et al., 1980). Receptors for oestrogens and progesterone have been localized histochemically in the epithelium of the uterine endometrium (Bergqvist et al., 1985). It is generally accepted, therefore, that steroids play an important role in the regulation of glycoprotein expression in the reproductive tract, and that such glycoproteins are essential in the implantation of the embryo (Enders and Schlafke, 1974; Surani, 1979). Glycosyl alteration in the reproductive tract can be monitored by a number of methods, but the affinity of lectins for glycosylated molecules provides a convenient and powerful probe. Thus, Sutton et al. (1979) have used lectin affinity after electroblotting to identify an oviducal glycoprotein which is associated with the presence of an embryo in the sheep, but which disappears after the embryo has entered the uterus. The technique has not, however, been applied extensively to a study of murine implantation. Lee et al. (1983) have described lectin binding to mouse oviduct material and showed increased WGL binding to the proximal region, but their report was only on one stage of gestation, at day 12 p.c. We have examined glycosylation, as exemplified by binding of fluorescent lectins, in various sites along the reproductive tract in BALB/c mice where DB3 antibody causes pregnancy block, and in F~ hybrid animals (CBA/CaxBALB/cJ) in which the antibody is less effective. This reduced efficacy appears to be related to differential rates of embryo development in inbred compared to crossbred stocks (Rider et al., 1987). The most striking effect we observed was the virtually total absence of N-acetyI-D-glucosamine residues from the oviducal ampullae of BALB]c mice in which the antibody is known to block implantation: This result, obtained using direct FITC-WGL binding, was also seen when WGL, anti-WGL and antirabbit peroxidase were used for indirect immunoperoxidase visualization. The ampulla is the region of the oviduct where fertilization occurs and where nutrients necessary for the embryo are produced (Nilsson and Reinius, 1969). Exogenously administered IgG localizes specifically and within 1 h to the pre-ampullary region of the mouse oviduct (Parr and Parr, 1986), indicating a role for this region of the reproductive tract in immune responses. After fertilization, the eggs remain in the ampulla for up to 24 h before travelling into the isthmus. Although the eggs will have passed beyond the ampulla at the time of our experiments, it is possible that a substance secreted by the ampulla could
218
be transferred "downstream" by virtue of peritoneal fluid flow towards the uterine lumen. The prevention of synthesis and/or secretion of such a substance may influence the subsequent development of the embryo. The inability to detect this material in DB3-injected BALB/c mice that show inhibition of implantation may represent a breakdown in the glycosylation reactions normally occurring in these cells, resulting in incomplete synthesis of a glycoprotein. In view of the importance of saccharides in many intercellular recognition processes, such defective glycosylation could have a profoundly limiting effect on the normal physiological role of such a glycoprotein, e.g. by preventing interaction with receptors on the blastocyst or uterine epithelium (Surani, 1979). It is perhaps not without significance that the ampulla is considered to be the most important region of the oviduct for secretion of nutrients for the blastocyst (Nilsson and Reinius, 1969). Use of the lectin PNL indicated a modest reduction in N-acetyl-Dgalactosamine in the uteri of DB3-treated BALB/c mice compared to controls, though this was not observed with WGL. There appeared to be prolonged expression during gestation of PNL-detected saccharides in the ampulla (and probably in isthmus I) of DB3 antibody-treated BALB/c. The isthmus is the region of the oviduct which affects the transport and nutrition of both sperm and eggs (Nilsson and Reinius, 1969). Such an increase in glycoprotein expression following antibody treatment is consistent with the results of Dutt et al. (1986) who showed progesterone regulation of uterine glycosylation. However, the regulation of a receptive uterine state by progesterone and oestrogen is complex and administration of monoclonal antibody to progesterone probably has a number of effects. For example, although antiprogesterone antibody arrests embryo development predominantly at the 4-cell stage, it also prevents the formation of an oil-induced decidual cell reaction (Rider et al., 1987). Steroids are known dramatically to influence N-linked glycoprotein assembly at multiple levels in the hen oviduct (DeRosa and Lucas, 1982), although we know of no reports in mammals. The modification of glycosyl expression detected in the female reproductive tract is the first indication of a biochemical change induced at the apical epithelial surface by passive immunization with this antiprogesterone antibody. The results are also dependent upon genotype and in this respect are comparable to antibody efficacy in blocking pregnancy.
Acknowledgements C.Y. acknowledges the receipt of a Training Fellowship from the World Health Organization. We thank Mr. Ian King for his help with the histology.
219
Relerences Bergqvist, A., Jeppsson, S. and Ljungberg, O. (1985) Histochemical demonstration of estrogen and progesterone binding in endometrial tissue and in uterine endometrium: a comparative study. J. Histochem. Cytochem. 33, 155-161. Buxton, E. and Murdoch, R.H. (1982) Lectins, calcium ionophore A23187 and peanut oil as deciduogenic agents in the uterus of pseudopregnant mice: effects of tranylcypromine, indomethacin, isoproniazid and propanolol. Aust. J. Biol. Sci. 35, 63-72. De Rosa, P.A. and Lucas, J.J. (1982) Estrogen-induced changes in chick oviduct membrane glycoproteins. J. Biol. Chem. 257, 1017-1024. Dutt, A., Tang, J.-P., Welply, J.K. and Carson, D.D. (1986) Regulation of N-linked glycoprotein assembly in uteri by steroid hormones. Endocrinology 118, 661-673. Enders, A.C. and Schlafl~e, S. (1974) Surface coats of the mouse blastocyst and uterus during the preimplantafion period. Anat. Rec. 180, 31-46. Joshi, S.G. (1983) A progestagen-associated protein of the human endometrium: basic studies and potential clinical applications. J. Steroid Biochem. 19, 751-757. Joshi, S.G., Ebert, K.M. and Smith, R.A. (1980) Properties of the progestagen-dependent protein of the human endometrium. J. Reprod. Fertil. 59, 287-296. Lee, M.-C., Wu, T.-C., Wan, Y.-J. and Damjanov, I. (1983) Pregnancy-related changes in the mouse oviduct and uterus revealed by differential binding of fluoresceinated lectins. Histochemistry 79, 365-375. Nilsson, O. and Reinius, S. (1969) Light and electron microscopic structure of the oviduct. In: The Mammalian Oviduct (Hafez, E.S.E. and Blandau, R.J., eds), pp. 57-84. University Press, Chicago. Parr, E.L. and Parr, M.B. (1986) Uptake of immunoglobulins and other proteins from serum into epithelial cells of the mouse uterus and oviduct. J. Reprod. Immunol. 9, 339-354. Rider, V. and Heap, R. B. (1986) Heterologous anti-progesterone monoclonal antibody arrests early embryonic development and implantation in the ferret (Mustela putorius). J. Reprod. Fertil. 76, 459-470. Rider, V., Heap, R.B., Wang, M.-Y. and Feinstein, A. (1987) Antiprogesterone monoclonal antibody affects early cleavage and implantation in the mouse by mechanisms that are influenced by genotype. J. Reprod. Fertii. 79, 33-43. Rider, V., McRae, A., Heap, R.B. and Feinstein, A. (1985) Passive immunization against progesterone inhibits endometrial sensitization in pseudopregnant mice and has antifertility effects in pregnant mice which are reversible by steroid treatment. J. Endocrinoi. 104, 153-158. Rider, V., Wang, M.-Y., Finn, C., Heap, R.B. and Feinstein, A. (1986) Anti-fertility effect of passive immunization against progesterone is influenced by genotype. J. Endocrinol. 108, 117-121. Surani, M.A.H. (1979) Glycoprotein synthesis and inhibition of protein glycosylation by tunicamycin in preimplantation embryos: influence on compaction and cell adhesion. Cell 18, 217-227. Sutton, R., Wallace, A. L. C. and Nancarrow, C.D. (1985) Characterisation of a glycoprotein in oviducal fluid by two-dimensional electrophoresis and lectin binding to protein gel blots. Electrophoresis 6, 516--520. Thie, M., Bochskani, R. and Kirchner, Ch. (1986) Glycoproteins in rabbit uterus during implantation. Differential localization visualized using 3H-N-acetyl-glucosamine labelling and F1TC-conjugated lectins. Histochemistry 84, 73-79. Wang, M.-Y., Rider, V., Heap, R.B. and Feinstein, A. (1984) Action of anti-progesterone monoclonal antibody in blocking pregnancy after postcoital administration in mice. J. Endocrinol. 101, 95-100. Whyte, A. and Allen, W.R. (1985) Equine endometrium at pre-implantation stages of pregnancy has specific glycosylated regions. Placenta 6, 537-542. Whyte, A. and Robson, T. (1984) Saccharides localized by fluorescent lectins on trophectoderm and endometrium prior to implantation in pigs, sheep and equids. Placenta 5, 533-540. Wright, L.J., Feinstein, A., Heap, R.B., Saunders, J.C., Bennett, R.C. and Wang, M.-Y. (1982) Progesterone monoclonal antibody blocks implantation in mice. Nature 295, 415-417.
