The absence of H-2 antigens from mouse pancreatic beta-cells demonstrated by immunoferritin labeling.

Islets of Langerhans were isolated from mouse pancreases and fixed in periodatelysine-paraformaldehyde. The fixed islets were then dissociated with trypsin and EDTA to yield cell suspensions that contained mainly four cell types; beta-cells, capillary endothelial cells, acinar cells, and pancreatic duct epithelial cells. The nonislet cells were probably associated wtih the surface of the isolated islets. The H-2 antigens of the dissociated pancreatic cells were labeled with an immunoferritin technique. Pancreatic duct epithelial cells showed specific ferritin labeling on their lateral cell membranes but not on apical microvillus membranes. Acinar cells were also labeled on lateral membranes, and the capillary endothelial cells were labeled on both the luminal and albuminal aspects of their surface membranes. In contrast, pancreatic beta-cells were unlabeled. The number of ferritin molecules per unit length of beta-cell membrane was essentially the same on cells from the antigenic strain and the congeneic control strain, and was about 200-fold less than on the labeled pancreatic duct epithelial cell lateral membranes. Pancreatic beta-cells are therefore one of six known epithelial cell types on which H-2 antigens can not be detected by immunoferritin labeling. The apparent absence of H-2 antigens from these cells suggests a study of the viability of beta-cells in allografts of dissociated islet cells, in which the beta-cell would not be in contact with antigenic cells. Such studies might lead to a new approach to the control of diabetes mellitus by transplantation.

Epithelium of mouse yolk sac placenta lacks H-2 complex alloantigens.

The only fetal cell membrane exposed to the mother in the mouse yolk sac placenta is the apical membrane of the endodermal epithelial cells. In yolk sac preparations in vitro, this apical membrane was exposed to reagents or cells in the incubation medium. By using several techniques we were not able to detect fetal major histocompatibility complex (MHC) antigens in this membrane. Immunoferritin labeling with and without prefixation and after neurominidase and trypsin digestion indicated that the apical membrane could contain no more than approximately 1% of the H-2 complex antigens that were present on peritoneal macrophages. Incubation of yolk sac preparations in anti-H-2 complex antiserum and complement had no cytotoxic effect on the endodermal epithelium, nor did incubation in an excess of alloreactive lymphocytes. Dissociated preparations of prefixed yolk sac contained endodermal epithelial cells and vascular endothelial cells whose entire surface membranes were exposed to the medium. H-2-complex antigens were not detected by immunoferritin labeling in either the apical or the laterobasal membrane of the yolk sac endoderm, but they were present in low density on the vascular endothelium. Also, incubation of unfixed, dissociated cells in anti-H-2-complex serum and complement had no detectable cytotoxic effect on endodermal epithelial cells. These observations indicate that H-2 antigens are sparse or absent in both the apical and laterobasal membranes of endodermal epithelial cells. The deficiency of MHC antigens in the apical membrane may account for the failure of sensitized females to reject the yolk sac, whereas the composition of the laterobasal membrane is probably less important to maternal-fetal relations. The present observations are consistent with labeling studies of adult-lining epithelial cells, which indicate that self-marker MHC molecules are absent from the apical membranes oriented toward the outside world and variably expressed in the laterobasal self-side membranes. It is suggested that the corresponding exclusion of fetal self-marker molecules from the apical membranes of some kinds of placental epithelia would deprive the mother of target sites for an alloimmune reaction at the maternal-fetal interface.

The relationship of polymorphonuclear leukocytes to infertility in uteri containing foreign bodies.

A chronic infiltration of polymorphonuclear leukocytes was invariably found in the infertile regions of uteri containing foreign bodies in conventional rats, germfree rats, mice, and rabbits. Polymorphonuclear leukocytes were never found in the fertile regions of these uteri. A foreign body in the uterus of the rat, and probably also the mouse, was associated with a bacterial infection which spread the inflammatory response throughout the horn containing the foreign body, and in the mouse occasionally into the control horn as well. No bacteria could be cultured from the rabbit uterine horn containing a foreign body. In the germfree rat, both the infiltration of polymorphonuclear leukocytes into the uterus and fertility were significantly different from that observed in the conventional rat. Whereas in the conventional rat the inflammation and infertility extended along the entire length of the uterine horn containing a small foreign body, in the germfree rat the inflammation and infertility were closely correlated to the position of the foreign body. As judged by measurements of lysozyme in the uterine lumens of rats and rabbits, polymorphonuclear leukocytes released their contents into solution in the uterine lumen. It is concluded that some substance derived from polymorphonuclear leukocytes may exert toxic effects on fertilized ova or on spermatozoa and thus be responsible for the infertility of uteri containing foreign bodies.

Perforin-expressing granulated metrial gland cells in murine deciduoma.

It has previously been shown that granulated metrial gland (GMG) cells of the pregnant uterus express abundant quantities of the lymphocyte pore-forming protein, perforin. No perforin was present before implantation of the embryo, but large numbers of perforin-producing GMG cells were observed after implantation, which coincides with decidualization of the uterus. The possible source of the activation factors responsible for perforin gene induction in GMG cells was studied here with the pseudopregnancy model, in which cervical stimulation of mice during estrus leads to a series of hormonal changes resembling those seen in pregnancy, but in the absence of an embryo. Subsequent stimulation of the uterus of pseudopregnant mice with oil causes the stimulated portion of the endometrium to differentiate into decidual tissue. Perforin-containing GMG cells were in fact present in the deciduomata, but not in adjacent nondecidualized tissues of the same mice. These results suggest that maternal factors associated with decidual tissue are responsible for the local expression of perforin in GMG cells.

The surface concentration of H-2 antigens on mouse pancreatic beta-cells and islet cell transplantation.

Measurement of the concentration of H-2 complex antigens in the surface membrane of mouse pancreatic beta-cells by a quantitative immunoferritin-labeling technique revealed that the antigens were present in very small amounts on the beta-cells of five strains. A comparison of the labeling densities in the five strains suggested that beta-cells from C57BL/10Sn congenic strains have about half the H-2 antigen density of BALB/c and C3H/HeJ cells. In C57BL/10Sn mice the antigen density on beta-cells was slightly greater than that on erythrocytes, about 20% of that on thymocytes, and about 2.5% of that present on peritoneal macrophages. Intraperitoneal allografts of c57BL/10Sn islets were uniformly rejected by B10.BR/SgSn diabetic recipients only when the islets were accompanied by 10(7) peritoneal lymphoid cells. When transplanted without peritoneal cells, C57BL/10Sn islets were only marginally rejectable. In a group of nine such allografts, three diabetic recipients were permanently cured and three others showed rejection times of about 30 days. Sensitization of the three mice showing permanent cures, using 10(7) allogeneic peritoneal cells at about 40 days after the transplant, did not cause rejection of the allografts. Isogeneic transplantation of cell suspensions from dissociated islets of Langerhans was markedly less effective in controlling diabetes than intact islets, and dissociation did not obviously improve the rate of allograft survival. However, 5/19 diabetic mice receiving allografts of dissociated islet cells showed long-term reversals of diabetes that were unaffected by administration of 10(7) peritoneal cells at about 100 days after the transplant. Recipient mice whose diabetes was reversed by islet allografts and unaffected by specific sensitization had pancreatic insulin concentrations characteristic of diabetic mice. Our reversals of diabetes with untreated islet allografts may be due to the cleanliness of islet preparations obtained with a modified isolation technique, and to the very low density of H-2 complex antigens on C57BL/10Sn beta-cells.

Purification and measurement of secretory IgA in mouse milk.

An important factor limiting better understanding of the protective role of sIgA at mucosal surfaces is the limited availability of the purified immunoglobulin. Among other things, purified sIgA is needed for use as a standard in measurements of the concentration of this immunoglobulin in mucosal secretions, particularly in mice, where several models of mucosal infections are available. We describe here a simple method by which one can obtain a mean of 3.5 ml of milk per mouse without a breast pump. Immunoblotting studies after native PAGE demonstrated that the milk contained mainly 420 kDa dimeric sIgA and higher polymeric forms of sIgA; only a trace of monomeric IgA was present. Similar immunoblotting studies after SDS-PAGE revealed that a portion of the sIgA was dissociated by this treatment. The 420 kDa sIgA was purified by salt fractionation, gel filtration, and affinity chromatography, and the purity of the final product was demonstrated by immunoblot analysis of biotinylated polypeptides after reduction of biotinylated protein. The concentration of 420 kDa sIgA in whey was measured by densitometry of immunoblot bands, using the purified 420 kDa sIgA as a standard, and found to be 1.0 +/- 0.3 mg/ml.

Diversity of expression of H-2 antigens on mouse liver cells demonstrated by immunoferritin labeling.

Antigens of the mouse H-2 locus were studied on isolated liver cells by immunoferritin labeling. Cell suspensions were obtained by gentle homogenization of mouse livers that had been perfused with hypertonic sucrose to loosen cell junctions. Hepatocytes showed sparse labeling of H-2 antigens on the order of 1% of that present on peritoneal monocytes. The slight hepatocyte labeling appeared to be specific for the H-2 locus, since there was essentially no label on congenic control hepatocytes. Other cells in the liver cell preparations were more heavily labeled. Bile duct epithelial cells showed heavy labeling on their lateral surface membranes, but none on their apical brush borders. The preparations contained another nonparenchymal cell type that was densely labeled, but it was not positively identified. Spleen lymphocytes added to liver homogenates and carried through the cell isolation procedure with liver cells showed typically heavy ferritin labeling. These observations suggest that the expression of histocompatibility antigens by tissue cells is likely to be quite variable.

Cellular changes in cultured mouse thyroid glands and islets of Langerhans.

Islets of Langerhans cultured 7 days in vitro no longer contained any capillary endothelial cells, but their endocrine cells remained ultrastructurally normal up to 14 days. Vascular endothelial cells were also lost from cultured thyroid lobes, but more slowly. Thyroid endothelium was readily identified after 7 days of culture, although many cells appeared to be degenerating, and a few degenerating endothelial cells were still present after 14 days in culture. Erythrocytes and leukocytes in the lumina of thyroid vessels were observed to degenerate at about the same rate as the endothelial cells, while those in islet capillary lumina were largely washed out during isolation of the islets. Thyroid lymphatic endothelium and the numerous adipose cells present in this tissue also degenerated during the culture period. Follicle epithelial cells remained viable throughout the culture period, but the number of colloid droplets and endocytic vesicles they contained was markedly, decreased. Thyroid fibroblasts remained viable and appeared to enlarge and accumulate dense granules during culture. These cells were a prominent feature of thyroid lobes after 14 days of culture. Parathyroid tissue associated with the thyroid lobes showed viable endocrine cells but a loss of vascular endothelium after 14 days in culture. The loss of blood leukocytes and vascular endothelial cells in probably the major factor in the altered behavior of thyroid and islet allografts after culture in vitro.

H-2 complex and Ia antigens on cells dissociated from mouse thyroid glands and islets of Langerhans.

Mouse thyroid tissue was dissociated with collagenase, fixed in periodate-lysine-paraformaldehyde (PLP), and further dissociated with EDTA and trypsin to yield cell suspensions containing mainly follicle epithelial cells and vascular endothelial cells. H-2 complex antigens were detected on the vascular endothelial cells at about the same high density as on peritoneal macrophages, and at a lower concentration on the laterobasal membranes of follicle epithelial cells. Neither of these cell types expressed detectable Ia antigens, but a minor cell type was presented that showed dense expression of Ia antigens. This cell type was probably a passenger leukocyte. It showed ultrastructural characteristics closely resembling those of spleen dendritic cells, which are known to express Ia antigens and to be potent stimulator cells in mixed lymphocyte culture. Dissociation of thyroid glands that had been cultured in vitro for 14 days yielded only follicle epithelium, and these cells showed the same labeling density of H-2 complex antigens as on uncultured cells. Dissociation of islets of Langerhans yielded capillary endothelial cells and beta cells, neither of which expressed detectable Ia antigens. The labeling results are discussed in relation to the cellular changes that occur during culture in vitro and the altered behavior of cultured allografts.

Intracellular labeling with ferritin conjugates. A specificity problem due to the affinity of unconjugated ferritin for selected intracellular sites.

Nonspecific binding of ferritin to chromatin and the cytoplasmic aspect of the nuclear envelope was observed when nonantigenic, serum-washed hepatocyte nuclei were incubated in ferritin-antibody conjugates. This labeling was duplicated when nuclei from a wide range of species and cell types were exposed to unconjugated ferritin. Unconjugated ferritin binding to nuclei did not depend on a subpopulation of denatured molecules or on the ferritin purification procedure. Binding occurred equally on unfixed and formaldehyde-fixed nuclei, but no ferritin bound to glutaraldehyde-fixed nuclei. Inconjugated ferritin also bound to the cytoplasmic aspects of the rough endoplasmic reticulum and the plasma membrane. The tracer did not bind to lysosomes, mitochondria, Golgi vesicles, the extracellular surface of plasma membranes, or the intracisternal surfaces of ruptured nuclear envelopes. The addition of 0.4 M KCl or 0.7 M NaCl to ferritin solutions and washing media at neutral pH reduced the binding of conjugated and unconjugated ferritin to nuclei to about 3% of that seen in 0.10 M phosphate buffer alone. The added salts caused little extraction of nuclear contents from formaldehyde-fixed nuclei. The use of one of these salts in ferritin conjugates should considerably improve the specificity of intracellular labeling.

The occurrence and distribution of H-2 antigens on mouse intestinal epithelial cells.

This report describes an immunoferritin labeling study of mouse H-2 histocompatibility antigens on epithelial cells dissociated from the small intestine by EDTA and trypsin. Before cell dissociation, the intestine was prefixed in paraformaldehyde or periodate-lysine-paraformaldehyde in order to preserve the shape of the cells and to immobilize H-2 antigens in their native positions. The results demonstrated the presence of H-2 antigens on the lateral and basal cell membranes at about the same high density that was observed at the surface of mouse monocytes. No H-2 antigens could be detected at the apical surface of dissociated or undissociated epithelial cells. It is unlikely that the fuzzy coat masked H-2 antigens at the apical surface because it was essentially absent from the apical membranes of dissociated cells. These observations extend our knowledge of the cellular distribution of transplantation antigens, and provide further evidence of a discontinuity in the expression of membrane components at the junctional complex of epithelial cells.

An immunoferritin labeling study of H-2 antigens on dissociated epithelial cells.

This report describes an immunoferritin labeling study of mouse H-2 histocompatibility antigens on epithelial cells dissociated from stomach, duodenum-jejunum, ileum, trachea, diestrus uterus, gall bladder, and vas deferens. Before cell dissociation, most of the organs were prefixed in periodate-lysine-paraformaldehyde to preserve the shape of the cells and to immobilize H-2 antigens in their native positions. Five kinds of epithelial cells expressed H-2 antigens on lateral and basal membranes but not on apical membranes. These were the lining cells of the upper intestine, ileum, gall gladder, uterus, and the tracheal brush cell. The antigens were continuously distributed on the lateral and basal membranes of these cells and appeared to be absent from the apical membranes, rather than masked by the fuzzy coat. On four other epithelial cell types H-2 antigens could not be detected. These were the lining cells of the vas deferens, parietal and chief cells from the stomach, and ciliated tracheal cells. It does not seem to be uncommon for normal nucleated cells to lack H-2 antigens. On fixed and labeled epithelial cells from the upper intestine the zonula occludens membranes were unlabeled, while the zonula adherens and desmosome membranes were labeled as densely as the remainder of the lateral membranes. The zonula occludens membrane thus constituted the boundary betewen the unlabeled apical membrane and the labeled lateral membrane of these cells. Intestinal epithelial cells dissociated without prefixation showed a patchy distribution of H-2 antigens on their lateral membranes after indirect labeling, indicating antigen mobility in this membrane. On the same unfixed dissociated cells the antigens were able to migrate from lateral to apical membranes, a movement which appears to be prevented in the intact epithelial layer by the occluding junction. The absence of H-2 antigens from apical membranes and their inability to migrate through an intact zonula occludens suggest that these molecules must reach the lateral membranes of epithelial cells by a pathway which is distinct from that followed by apical membrane components.

Anatomic correction of transposition of the great arteries without significant ventricular septal defect or patent ductus arteriosus.

A case of transposition of the great arteries without a significant ventricular septal defect or patent ductus arteriosus was treated when the patient was 5 months of age by retransposition of the great arteries and coronary arterial supply to the appropriate ventricle. The patient has done well for 13 months, despite the late occurrence of aortic regurgitation.

Contraception with intrauterine devices.

PIP: A review of the history of contraception with intrauterine devices, characteristics of present devices, and directions of current research is presented. The serious need for population control is not yet being met by today's inconvenient, ineffective, or unsafe methods. Intrauterine devices have been best for international family planning programs because they are cheap, easily installed, and provide continuous protection. There are many different models that have been and are being used, with different effectiveness and complication rates. The most commonly used today is the Lippes Loop, with a pregnancy rate of 2.8/100 years of woman use and an expulsion rate of 10.4. Most of these failures occur in the first few months of use, after which these rates are greatly reduced. The removal rate because of bleeding or pain for the Lippes device is 14.0. Other devices commonly used have pregnancy rates ranging 1.3-4.7, expulsion rates of 2.6-25.8, and removal rates of 13.5-22.1. Expulsion is directly related to the size and design of the IUD and the age and parity of t,e recipient. It is important to match the size of the device used to the individual characteristics of the patient. Research is seeking a design that will implant itself in the endometrium to resist expulsion, but not too deeply so that it is covered. Removal for bleeding and pain remains the most frequent complication of the IUD, and it partly depends on the skill of the inserting physician and how well the patient is psychologically prepared for side effects in the first months of use. Pregnancy is the most significant IUD complication. The key to an effective IUD is an understanding of its antifertility mechanism, which has thus far eluded researchers. The IUD prevents implantation of the blastocyst in the uterine wall, which may be due to a foreign-body reaction in the endometrium. IUDs with copper cause a greater reaction than plastic devices and provide hope for a very effective device; particularly the T-shaped design, which resists expulsion. The most promising new IUD is the Dalkon Shield. It has small projections that imbed in the endometrium and a broad surface for contact with the uterine wall. In preliminary experiments the pregnancy rate with this device was 1.1, the expulsion rate 2.3, and the removal rate 2.0, much lower than that with any other device yet developed. It is concluded that IUDs such as the Dalkon Shield can provide safe contraception with high effectiveness.^ieng

Uptake of immunoglobulins and other proteins from serum into epithelial cells of the mouse uterus and oviduct.

The transport of immunoglobulins into the lumen of the female reproductive tract is not well understood, especially in the case of IgG. In mice, there are conflicting reports concerning the presence of immunoglobulins in uterine luminal and glandular epithelial cells, and immunoglobulins have not been detected in the luminal epithelial cells of the oviduct. In the present study we detected both IgA and IgG in uterine luminal and glandular epithelial cells on day 1 of pregnancy by immunolabeling. Also, we observed that fluorescein-conjugated mouse and bovine IgG and other proteins were taken up into vesicles in uterine luminal and glandular epithelial cells after intravenous administration. These observations indicate that both kinds of epithelial cells take up immunoglobulins from the interstitial fluid on day 1 of pregnancy, and that the cells may therefore be involved in the transport of immunoglobulins and other proteins to the uterine lumen at that time. In the oviduct, we detected IgA and IgG in vesicles in the luminal epithelial cells of the preampulla by immunolabeling, and we observed fluorescein-conjugated IgA and IgG in similar vesicles after intravenous administration. The presence of IgA and IgG in vesicles in the epithelial cells of the preampulla, together with the previous demonstration of plasma cells of both isotypes and large amounts of interstitial immunoglobulins in the lamina propria of this segment, suggests that the preampulla of the oviduct may be an important site for the local immune system in the mouse female genital tract.

The effect of sperm immunization in the gastrointestinal tract on anti-sperm antibody production and fertility in female mice.

The present investigation studied the effects of sperm immunization in the gastrointestinal tract on anti-sperm antibody production and fertility in female mice. For comparative purposes, mice were also immunized with sperm intraperitoneally. Intraperitoneal immunization with 5 X 10(6) washed epididymal and vas deferens sperm 3 times per week for 7 wk produced anti-sperm IgG in plasma at 1:20,000 and in vaginal washings at 1:100 as determined by ELISA. Such mice have been shown previously to have reduced fertility. In comparison, mice immunized intragastrically with 5 X 10(6) sperm once per week for 11-14 wk had anti-sperm IgA in vaginal washings at only about 1:8 as determined by ELISA. After mating at the 14th wk these mice delivered 6.5 +/- 1.4 pups, which was not significantly different from the 7.1 +/- 1.1 pups delivered by an untreated control group. Mice immunized twice intragastrically and once intravaginally during a 25-day period had no detectable anti-sperm IgA in vaginal washings by ELISA. These mice delivered 9.7 +/- 1.2 pups after mating beginning on day 32, as compared to 9.7 +/- 0.8 pups in a PBS-sham immunized group. Mice immunized once intraduodenally and then once intraperitoneally 14 days later delivered 10.4 +/- 0.9 pups after mating 10-14 days after the second immunization, while a similar group of mice whose primary sperm immunization was directly into Peyer's patches delivered 9.0 +/- 1.4 pups. We could not detect anti-sperm IgG or IgA bound to sperm in the uterine or oviduct lumen using immunohistochemical labeling after any of the groups of immunized mice were mated.(ABSTRACT TRUNCATED AT 250 WORDS)

Protective immunity against HSV-2 in the mouse vagina.

Herpes simplex virus type 2 (HSV-2) is a sexually transmitted pathogen that infects the genital tract. The high prevalence of HSV-2 in humans underscores the need to develop an effective vaccine. Efforts to develop vaccines to protect women against this and other sexually transmitted pathogens would be facilitated by a better understanding of the immune mechanisms that protect the female reproductive tract against infections in animal models. Such information would be invaluable in developing vaccine strategies to promote the type and magnitude of immune responses in the genital tract that would effectively protect against infection. This review focuses on recent studies using a progestin-treated adult mouse model to explore mucosal immunity to HSV-2 in the vagina. Evidence indicating a major role for both humoral and T cell immunity is presented.

Secretory immunoglobulin binding to bacteria in the mouse uterus after mating.

Bacteria of several species were present in the mouse uterus on the morning after mating, as demonstrated by bacterial cultures and microscopic examination of Gram-stained uterine luminal contents. The similarity between bacteria cultured from the vagina before mating and from the uterus after mating suggested that bacteria were introduced into the uterus from the vagina, possibly by coitus. The bacteria were cleared from the uterus about two days after mating. Immunohistochemical labeling of smears of the luminal contents on the morning after mating demonstrated IgA, IgG, and possibly IgM bound to many of the bacteria. The bacteria were often agglutinated, and there was a correlation between the intensity of immunoglobulin labeling on bacteria and the extent of agglutination. The amount of antibody bound to bacteria in multiparous mice was about the same as in mice that had not been mated previously. We observed both IgG and IgA on bacteria when organisms from vaginal cultures were incubated for 60 min in the uteri of estrogen-primed, virgin, female mice. This indicated that the uterus was the source of at least part of the immunoglobulins bound to bacteria. We did not demonstrate that the immunoglobulins bound to bacteria were specific anti-bacterial antibodies, but the binding persisted through three washing steps and there was no immunoglobulin binding to sperm in the same preparations. Neutrophils in the uterine lumen on the day after mating contained phagocytosed bacteria. These results suggest that the secretory immune system in the female mouse reproductive tract may play a role in returning the uterus to an aseptic state after mating by at least three mechanisms: direct blocking of attachment sites involved in bacterial binding to mucosal epithelium, agglutination of bacteria and thus reduction in the number of organisms available for binding to the epithelium, and opsonization of bacteria for phagocytosis by neutrophils.

Localization of immunoglobulins in the mouse uterus, embryo, and placenta during the second half of pregnancy.

Throughout the second half of pregnancy in mice there were many plasma cells containing immunoglobulins A (IgA) and G (IgG) in the uterine endometrium. There was intense staining of IgA in uterine glands at all stages, but little staining of IgG. The staining of both immunoglobulins (Igs) in the luminal epithelium was moderate to dark on day 11, slight on day 14, and increased from day 16 to term. From day 14 to term the endometrium exhibited folds or villi around each placenta. The cores of the villi contained many plasma cells of both isotypes, and the staining of extracellular Igs in the villous cores was darker than in nonvillous endometrium. Both Igs were detected in the uterine lumen, and in visceral and parietal yolk sac endoderm cells at all stages. Near term, the staining of Igs in the visceral yolk sac was darkest in the peripheral villous portion adjacent to the endometrial villi. From day 14 to term IgG was present in the visceral yolk sac mesenchyme and embryo, consistent with its transfer from the uterine lumen to the embryo via the vitelline circulation. In contrast, IgA was not detected in yolk sac mesenchyme until day 19, when only slight staining was observed, and IgA was never detected in the embryo. Most trophoblast giant cells contained both Igs on day 11. During the remainder of pregnancy, there was staining of both Igs in labyrinthine trophoblast and in a few giant cells adjacent to the parietal yolk sac on the placenta, but there was negligible staining in the spongiotrophoblast region. Our observations suggest that the local immune system in the mouse uterus may protect the embryo during the second half of pregnancy by secreting anti-microbial immunoglobulins A and G into the uterine lumen surrounding the visceral yolk sac, and may at the same time contribute to the transfer of maternal IgG to the embryo via the yolk sac and vitelline circulation.

Deposition of C3 on bacteria in the mouse uterus after mating.

A previous study demonstrated that several species of bacteria were present in the mouse uterine lumen on the day after mating, and that many of these bacteria had immunoglobulins bound to their surface. Neutrophilic leukocytes containing phagocytosed bacteria were also present in the lumen. Since bacteria are phagocytosed efficiently only when they are opsonized by the binding of specific antibody or C3b or both to their surface, we investigated whether the uterine bacteria were coated with C3. Immunolabeling demonstrated that an antigenic portion of C3, possibly C3b, was bound to many of the uterine bacteria. This observation suggests that bacteria in the mouse uterus after mating may be opsonized by both antibody and complement and that phagocytosis of these bacteria by neutrophils may play an important role in returning the uterus to an aseptic state before implantation.