Malignant changes in New Zealand black mice

Ageing, Coombs positive, NZB mice, may spontaneously develop a neoplasia of the reticulum cell type which can be transferred by serial passage of their lymphoid tissues in young intact or neonatally thymectomized syngeneic recipients. The recipients (103/117) of cell suspensions prepared from the enlarged spleens and/or lymph nodes of four such donors developed an extensive and lethal reticulum cell neoplasia affecting the spleen, lymph nodes, lungs and liver but the bone marrow, thymus and kidneys were seldom involved. The recipients (16/17) of spleen cells from a fifth donor showed massive proliferations of eosinophils in all the organs examined.

Prematurely positive antiglobulin (Coombs) reactions were detected in only two recipients. Although there was an indication that the IgM content of the sera decreased as one of the passages progressed, the levels of IgG and IgA were not seriously distorted.

Particles resembling murine leukaemia virus were identified by electron microscopy in the spleen and in plasma or serum pellets of passage recipients. However, similar particles were also seen in the thymus and/or spleen or bone marrow of untreated NZB mice including an 18-day embryo and animals aged 1–56 weeks, although no particles were found in plasma and serum pellets of mice aged 6–70 weeks.

The theory that autoimmunity, malignancy and virus infection are directly related is discussed.

     

Autoimmune reactions and malignant changes in New Zealand black mice

Although NZB mice were bred and maintained in a germ-free environment their spleens enlarged and showed a sequence of histological events concomitant with the advent of positive antiglobulin (Coombs) reactions at 8–10 months which were similar to, but less intense than, those of their conventional NZB counterparts. The numerous large follicles with prominent germinal centres which developed in the white pulp and the proliferations of large pyroninophilic cells in the red pulp thus represented a humoral autoimmune reaction uncomplicated by external microbial antigenic stimuli. This burst of immunological activity in the spleen was followed by a reticulum cell neoplasia (apparently originating within the follicles and from the perifollicular mantles) which was transferable by intraperitoneal injection of spleen cell suspensions to syngeneic and allogeneic (BALB/c) recipients. By comparison, the inguinal lymph nodes of these same germ-free NZB mice were both immunologically inactive and exempt from the malignant process. Lesions in the thymus, and kidney lesions resembling human membranous glomerulonephritis or lupus nephritis, were found in both germ-free and conventional mice of this strain. Possible relationships between the autoimmunity, malignancy and the virus-like particles known to be present in germ-free NZB mice are discussed.     

The appearance of immunological competence at an early age in New Zealand black mice

Adult and 5-day-old New Zealand Black (NZB) and Ju control strain mice were injected with sheep red blood cells. At various times after injection they were killed and the spleens, lymph nodes and thymus tested for haemolysin-producing cells. When the response was expressed as plaque-forming cells (PFC) per million viable cells, the response curve of the spleens of baby NZB mice was very similar to the response in the spleens of the adults, and the response in the lymph nodes of the babies was as high and more sustained than that of the adults. In Ju control strain baby mice of this age, the response in the spleen and the lymph nodes was both reduced and delayed compared with the adults. Neither strain gave a significant response in the thymus. The spleen and lymph nodes of adult NZB mice showed a response which was delayed but not reduced as compared with the adults of the Ju control strain, whereas in baby NZB mice the spleen and lymph nodes showed a response which was advanced and increased (particularly in the lymph nodes) compared with control strain babies (Ju, Swiss, C57B1, CBA). The NZB mice did not reach this level of responsiveness until they were 4–5 days old.     

Age-decrease of cells sensitive to an autoantibody-specific for thymocytes and thymus-dependent lymphocytes in NZB mice

NZB mice naturally produce an autoantibody which in the presence of complement is specifically cytotoxic for thymocytes and thymus-dependent lymphocytes (T-cells) in the peripheral lymphoid tissues (lymph nodes and spleen) and the circulation of mice. Using a direct cytotoxicity test with a NZB mouse serum pool which contained the high titred autoantibody, a progressive decrease was observed with age in the proportion of the autoantibody-sensitive cells in mesenteric lymph node, spleen, and blood of NZB mice in comparison with mice of other strains (C57BL/6J and NZW). The numerical decrease in the population of autoantibody-sensitive cells was evident at younger age and greater degree in the peripheral blood than in the lymph node and spleen. The age-decrease in the number of autoantibody-sensitive cells in lymph node and spleen contrasted with the numerical increase in nucleated cells in these organs. The age-decrease in the proportion and number of the autoantibody-sensitive cells in the blood exceeded the decrease in the blood lymphocyte count. This finding indicated that T-cells in the blood are selectively depleted with the ageing of NZB mice. A similar observation was made on the blood lymphocytes of (NZB × NZW)F₁ hybrid mice. The depletion of T-cells in the blood in association with the production of natural thymocytotoxic autoantibody is termed autoimmune thymus-dependent lymphocytopenia.     

Immune suppression to nucleosides: differences between NZB and NZW mice

Previous studies (Y. Borel and M. C. Young, Proc. Natl. Acad. Sci. USA 1980. 77: 1593) have shown that one can raise nucleic acid-specific suppressor T cells which diminish either the T-dependent immune response in vivo or the T-independent immune response in vitro. The results presented here confirm and extend these observations in several different strains of mice. Administration of nucleoside-modified spleen cells diminishes antibody-forming cells to nucleoside in mice immunized with nucleoside linked to keyhole limpet hemocyanin (KLH). Immune suppression was obtained in all strains except SJL and NZW, which are known to be high responders to denatured DNA. Both the primary and secondary immune responses were suppressed in C57BL/6 mice. Autologous cells exhibit a different ability to function as carriers. Spleen cells are the most effective, and to a certain extent, thymus cells. In contrast, bone marrow cells and red cells fail to induce immune suppression. A strain difference was found between NZB and NZW mice in their susceptibility to immunosuppression by nucleoside-modified spleen cells. Whereas NZB mice are high responders to nucleoside-KLH, they were easily suppressed by nucleoside coupled to spleen cells. In contrast, NZW mice, although relatively low responders to nucleoside-KLH, were not suppressed by administration of nucleoside coupled to spleen cells. Both male and female (NZB × NZW)F₁ mice appeared to behave like the NZW parental strain and were resistant to immunosuppression by nucleoside-modified spleen cells. The significance of this observation for the pathogenesis of murine systemic lupus erythematosus is discussed.     

Hyperactive T-cell function in young NZB mice; increased proliferative responses to allogenic cells.

The one-way mixed lymphocyte reaction was employed to study proliferative responses to antigens by mature, immunocompetent T cells from NZB mice 3 weeks to 4 months old. Compared to cells from control mice of the same H-2 type, thymus, spleen and lymph node cells from NZB mice were hyperactive in this response. The results are discussed in relation to possible effects of chronic stimulation by endogenous type C leukaemia virus upon differentiation of functional T cells or upon regulation by T cells of other T-cell functions, including augmentation of antibody responses.     

Characterization of CD4(+) T cell autoreactivity to self-MHC in New Zealand hybrid mice

The New Zealand white (NZW) H2(z) locus is strongly associated with the development of autoimmune disease in (NZB x NZW)F(1) mice, a model of systemic lupus erythematosus. To better understand the role of H2(z) in autoimmunity, we generated CD4(+) T cell hybridomas from the spleen and lymph nodes of unimmunized (NZB x NZW)F(1) mice and characterized their specificity. We found that over 50% of the hybridomas responded to syngeneic (H2(d/z)) spleen cells in the absence of exogenous antigen. Many of these autoreactive hybridomas responded to spleen cells expressing H2(z) and used H2(z) class II (I-A(z) or I-E(z)) molecules for presentation. Interestingly, nearly one third of the H2(z)-reactive hybridomas could not respond to spleen cells expressing only H2(z) class II molecules. These studies characterize a frequent population of autoreactive CD4(+) T cells in lupus mice and indicate that major histocompatibility complex molecules in addition to class II may be important for this self-recognition.     

Genesis of antinuclear antibody in NZB/W mice: role of genetic factors and of viral infections

Antinuclear antibodies could be induced in young NZB/W mice long before the natural occurrence of such antibodies by immunization with heat denatured DNA coupled to methylated bovine serum albumin (DNA m BSA). While induction of antinuclear antibodies was possible in several strains of mice (NZB/W, A/J, DBA/2, CBA and AKR), NZB/W mice had by far the highest titre of antibody. A genetic determination of this immune hyperreactivity to DNA was suggested by study of the parental strains. The NZW mice which have a low incidence of spontaneously occurring antinuclear antibody made as much antinuclear antibody upon immunization with DNA m BSA as did the NZB/W mice, while NZB mice which develop naturally a moderate incidence of antinuclear antibody responded relatively poorly to immunization.

It seemed possible that while the native hyperreactivity of NZB/W mice to DNA might be inherited from the NZW parent, the antigenic stimulus calling forth the anti-DNA response might be transmitted from the NZB parent. A possible source of this stimulus might be a viral infection such as that caused by the leukemia virus found in NZB mice which might liberate unusual amounts or kinds of host nuclear antigens or provide viral nucleic acid antigens. In support of this, induced neonatal infection of NZB/W mice with Polyoma virus markedly enhanced their antinuclear antibody response. Also, neonatal thymectomy of NZW mice which predisposes to reduced resistance to infection also enhanced antinuclear antibody formation. And finally, Statolon treatment designed to increase and maintain interferon levels in both normal and thymectomized NZW mice reduced antinuclear antibody formation.

     

Pathological changes of thymic epithelial cells and autoimmune disease in NZB, NZW and (NZB × NZW)F1 mice

An extensive histological study was carried out of NZB, NZW and (NZB × NZW)F₁, (BWF₁), mice of all ages between birth and 18 months. The thymuses of these mice were compared to those of three normal mouse strains. The study of the NZW mice showed that these mice, although they only occasionally have weakly positive Coombs' tests, may develop a renal disease probably of an autoimmune nature, similar to that of the NZB and the BWF₁ mice. Mice of all the three NZ strains developed lesions of the skin, liver, intestines, lymphatic tissues and kidneys much resembling those occurring in human systemic lupus erythematosus (SLE), neonatally thymectomized mice and, with the exception of the renal changes, the lesions of graft versus host disease.

The comparative study of the thymus in autoimmune and normal strains, revealed that important changes of the large medullary epithelial cells, involved in the formation of Hassall's corpuscles, occur very early in the three autoimmune strains. In the NZB mice the large epithelial cells are severely decreased in number in the first weeks following birth. The depletion of epithelial cells could be ascribed to a secondary degeneration of these cells soon after birth.

In contrast with the NZB mice, an extensive hyperplasia of the large epithelial cells and Hassall's corpuscles was observed in the NZW and BWF₁ mice, and was already apparent in the newborn animal. Many of the epithelial aggregates seemed to have been invaded by lymphoid cells. Both epithelial cells and the lymphoid cells engaged in this process showed a variety of degenerative changes. As in the NZB, a depletion of epithelial cells occurred in a later phase, at the age of 8 months in the BWF₁ and at 1 year in the NZW. In the majority of young mice of the normal strains invasion of islands of epithelial cells by lymphoid cells may also be observed, although this process is far less extensive than in the autoimmune strains and does not result in either epithelial hyperplasia or depletion of epithelial cells.

The described phenomenon of invasion of epithelial structures in the thymus by subsequently disintegrating lymphoid cells seems to support Burnet's concept, that so-called `forbidden clones' of lymphoid cells are eliminated in the thymus.

     

Effects of oestrogen and exercise on caspase-3 activity in primary and secondary lymphoid compartments in ovariectomized mice

This study investigated the effect of oestrogen exposure and exercise on caspase-3 activity, a measure of apoptosis, in lymphocytes from the thymus, spleen, and lymph nodes in ovariectomized mice. Fifty-nine female B6D2F1 mice were randomized to hormone and exercise conditions. Hormone treatment consisted of implantation with oestradiol pellets (0.72 mg oestradiol) or placebo pellets (0 mg) for 21 days following bilateral ovariectomy (OVX). Exercise consisted of a single treadmill exercise bout (26 m min(-1), 6 degrees slope, 90-min) or sedentary condition. Mice were killed and the thymus, spleen and lymph nodes were removed for the determination of caspase-3 expression by enzyme-linked immunosorbent assay (ELISA), serum oestrogen levels by RIA, and tissue weights. Body weights were monitored throughout the study. In the thymus, oestrogen exposure, exercise and both treatments together were associated with higher caspase-3 activity (P < 0.05) and lower thymus weights (P < 0.05). In contrast, oestrogen exposure and exercise treatment were not associated with greater caspase-3 activity or change in tissue weight in secondary lymphoid tissues (spleen, lymph nodes). Oestrogen-replaced OVX mice had a higher concentration of plasma oestradiol than placebo OVX mice (P < 0.05). CONCLUSION: The results suggest that oestrogen and treadmill exercise are associated with greater apoptosis, as measured by caspase-3 activity, in the thymus but not in the spleen or lymph nodes. Clinical studies will be necessary to determine if women who take oestrogen have higher rates of apoptosis in primary lymphoid tissues and the significance of thymocyte apoptosis for maintenance of cellular immune function during the post-menopausal years.     

Effects of oestrogen and exercise on caspase‐3 activity in primary and secondary lymphoid compartments in ovariectomized mice

This study investigated the effect of oestrogen exposure and exercise on caspase‐3 activity, a measure of apoptosis, in lymphocytes from the thymus, spleen, and lymph nodes in ovariectomized mice. Fifty‐nine female B6D2F₁ mice were randomized to hormone and exercise conditions. Hormone treatment consisted of implantation with oestradiol pellets (0.72 mg oestradiol) or placebo pellets (0 mg) for 21 days following bilateral ovariectomy (OVX). Exercise consisted of a single treadmill exercise bout (26 m min^?1, 6° slope, 90‐min) or sedentary condition. Mice were killed and the thymus, spleen and lymph nodes were removed for the determination of caspase‐3 expression by enzyme‐linked immunosorbent assay (ELISA), serum oestrogen levels by RIA, and tissue weights. Body weights were monitored throughout the study. In the thymus, oestrogen exposure, exercise and both treatments together were associated with higher caspase‐3 activity (P < 0.05) and lower thymus weights (P < 0.05). In contrast, oestrogen exposure and exercise treatment were not associated with greater caspase‐3 activity or change in tissue weight in secondary lymphoid tissues (spleen, lymph nodes). Oestrogen‐replaced OVX mice had a higher concentration of plasma oestradiol than placebo OVX mice (P < 0.05). Conclusion: The results suggest that oestrogen and treadmill exercise are associated with greater apoptosis, as measured by caspase‐3 activity, in the thymus but not in the spleen or lymph nodes. Clinical studies will be necessary to determine if women who take oestrogen have higher rates of apoptosis in primary lymphoid tissues and the significance of thymocyte apoptosis for maintenance of cellular immune function during the post‐menopausal years.     

Consequences of neonatal thymectomy in New Zealand black mice

The possible role of the thymus in autoimmune disease was studied by comparing the effects of neonatal thymectomy on New Zealand Black (NZB) mice (which develop a Coombs positive haemolytic anaemia) and on C3H/Bi, F₁ (C57BL × C3H/Bi), C57BL and TO mice.

The neonatally thymectomized NZB mice, in common with those of other strains, showed lethal wasting, a lymphocyte deficit in their lymphoid organs and blood, their packed cell volume was reduced and some had liver lesions associated with the hepatotrophic virus MHV-1. As in C3H/Bi and hybrid mice, thymectomy had little effect on the levels of immunoglobulins (IgG, IgA, IgM) present in their sera.

Removing the thymus from NZB mice did not prevent and could precipitate Coombs positive reactions; thymectomized mice of the other strains were Coombs negative. Although germinal centres develop and plasma cells occur in the lymphoid organs of most thymectomized mice, the striking feature of the NZB mice was the large number, size and activity of the germinal centres that persisted after thymectomy.

     

Transmission of auto-immune haemolytic anaemia and murine leukaemia virus in F1 (BALB/c X NZB) hybrid mice derived by ovum transplantation.

F1(BALB/c X NZB)hybrid progeny derived by ovum transplantation were used to study the transmission of auto-immune haemolytic anaemia and murine leukaemia virus (MuLV) by male New Zealand black (NZB) mice. Fertilized ova, collected from the normal BALB/c partners 3 1/2 days after mating, were transferred to other, surrogate, BALB/c mothers, which then carried, delivered, and reared the hybrid young. This technical manoeuvre effectively closed the congenital transplacental route theoretically available to any infectious MuLV originating from the NZB father. Nevertheless, such progeny developed exactly the same profile of auto-immune haemolytic disease and the same range of diverse malignancies as their normally-derived F1(BALB/c X NZB) counterparts, and they carried type C MuLV particles readily detectable by electronmicroscopy. We concluded, therefore, that both the auto-immunity and virus were transmitted before placentation, presumably by the NZB male at fertilization, and probably as genetic information.     

In vitro studies on the phagocytosis of Staphylococcus aureus by peritoneal macrophages of New Zealand mice

The ability of peritoneal macrophages from NZB, NZW and BWF1 mice to phagocytose and subsequently kill Staphylococcus aureus has been studied and compared to peritoneal macrophages from CBA, BALB/c and C₃Hf mice. Phagocytic indices determined in young and old New Zealand mice did not differ from those of the control strains. The bactericidal activity of the peritoneal macrophages did not show any interstrain differences.

Incorporation of serum from old and young NZB mice into the culture medium did not affect phagocytosis or the rate at which phagocytosed bacteria were killed. Studies performed in the absence of serum from the culture medium showed that probably both phagocytosis and bactericidal activity were unimpaired.

     

Loss of lymphocyte chalone activity in mice with autoimmune disease.

Lymphocyte chalone from the spleens of old BALB/c, young BALB/c and young NZB mice caused significant suppression of the proliferative response of BALB/c and NZB spleen cells to T and B mitogens, whereas lymphocyte chalone from old NZB spleen did not suppress. Lymphocyte chalone from young and old NZB mice was tested using different ages of NZB/NZW responding spleen cells; at all ages concanavalin A- and lipopolysaccharide-induced proliferation was suppressed less by the chalone from old NZB mice than from that of young NZB mice. The responding NZB/NZW cells were suppressed equivalently at all ages studied. The basis for the loss of lymphocyte chalone activity in old NZB mice remains unknown; however, it appears likely that this event has a role in the disturbance of the negative feedback control system which contributes to NZB autoimmune disease.     

Requirement for dendritic cells in the establishment of anti-phospholipid antibodies

Objective: The cross-presentation of cell-associated autoantigens contributes to systemic autoimmune diseases, including systemic lupus erythematosus (SLE). Little is known about the regulation of the immune response against soluble autoantigens targeted in these diseases.

Methods: We immunized the offspring of New Zealand Black and New Zealand White mice (NZB × NZW F₁) with syngeneic dendritic cells (DC) that had macropinocytosed β2-glycoprotein 1 (β₂GPI) during propagation in normal mouse serum or that had phagocytosed apoptotic thymocytes with syngeneic (murine) or xenogeneic (bovine) β₂GPI, which was associated to plasma membrane of the cells. Mice were in parallel immunized with apoptotic thymocytes that had associated the β₂GPI to their membranes in the absence of DC. The development of anti-β₂GPI antibodies and clinical features were monitored.

Results. Apoptotic cells alone, opsonized with β₂GPI, failed to induce anti-β₂GPI autoantibodies or clinical disease. In contrast, autoimmunity developed in the presence of DC. Furthermore, the syngeneic β₂GPI was a more effective antigen than the xenogeneic protein in re-boosted animals.

Conclusions. DC effectively initiate in NZB × NZW F₁ mice self-sustaining autoimmunity against the β₂GPI, either associated to apoptotic cells or macropinocytosed from the serum.     

The immunological responsiveness of germ-free mice thymectomized at birth. II. Lymphoid tissue and histopathology

The effect of neonatal thymectomy and antigenic stimulation on the lymphoid cell population has been studied in germ-free mice. Neither thymectomy nor injection of sheep erythrocytes induced any significant alteration in the blood lymphocyte levels. There was a clear-cut reduction in the cellularity of the periarteriolar lymphocyte sheaths of the spleen and of the paracortical regions of the lymph nodes in the thymectomized mice. Following stimulation with sheep erythrocytes, large pyroninophilic cells appeared in these areas in the intact germ-free controls but in only a few thymectomized mice and then in reduced numbers. Thymectomy did not influence the cellularity of the lymphoid follicles but less germinal centre and plasma cell activity occurred in response to an injection of sheep erythrocytes. Lesions suggestive of autoimmune reactivity were not found in lymphoid or nonlymphoid tissues of neonatally thymectomized germ-free mice. Lesions typical of viral infections were seen in some germ-free mice in both thymectomized and intact groups. It is concluded that the specific defect associated with the absence of the thymus is a reduction in a particular class of lymphocytes the development of which is under thymus control and the activities of which are to mediate certain defined immunological responses.     

Treatment of murine lupus with monoclonal antibody to L3T4. I. Effects on the distribution and function of lymphocyte subsets and on the histopathology of autoimmune disease

Monoclonal antibodies (MoAb) to L3T4 have been used successfully to suppress autoimmunity in murine models for several human autoimmune diseases. To clarify the immunologic and clinical consequences of treatment with anti-L3T4, we examined the effects of chronic administration of anti-L3T4 on the composition of lymphoid organs, the function of lymphocytes, and the histopathology of autoimmune disease in lupus-prone NZB/NZW F1 (B/W) mice. Weekly treatment with anti-L3T4 (2 mg/mouse) from age 5 to 8 months depleted L3T4+ cells from the spleen and lymph nodes, and prevented the development of splenomegaly and lymphadenopathy. The MoAb bound to target cells in the thymus and modulated their expression of the L3T4 antigen but, in contrast to its effect in extrathymic sites, anti-L3T4 did not deplete the target population from the thymus. In fact, after 3 months of therapy, mice that had been treated with anti-L3T4 had much larger thymuses than control mice that had been treated with saline, suggesting that treatment with anti-L3T4 prevented the thymic atrophy that occurs spontaneously in murine lupus. Despite depleting L3T4+ cells from the spleen, treatment with anti-L3T4 did not diminish the response of splenic lymphocytes to T and B cell mitogens, and it augmented splenic natural killer (NK) cell activity. Finally, treatment with anti-L3T4 decreased the diverse histopathologic manifestations of murine lupus. It dramatically reduced glomerular immunoglobulin and complement deposition and diminished lymphocytic infiltration and vasculitis in the kidneys. Treatment also reduced extrarenal immunopathology, including focal hepatitis and salivary gland infiltration. These observations have implications regarding the use of CD4 MoAb in people with autoimmune diseases.     

Natural cytotoxic autoantibody against thymocytes in NZB mice

NZB mice were found to produce natural thymocytotoxic autoantibody in high prevalence and antibody titre. This autoantibody in NZB mice was detectable by the cytotoxicity test at both 4°C and 37°C; the prevalence and antibody titre were generally higher at 4°C. Mice of other strains also produced natural thymocytotoxic autoantibody although in lower prevalence and antibody titre and in some instances the activity was greater at 37°C than at 4°C. Natural thymocytotoxic autoantibody in NZB mice reacted equally with the thymocytes of virtually all strains of mice tested but to a lesser degree with the thymocytes of SJL/J mice. A serum pool obtained from old NZB mice had an extremely high titre of natural thymocytotoxic autoantibody (1:1024 at 4°C). Nevertheless, the cells in lymph nodes, spleen and blood leucocytes were only partially sensitive to this serum pool, and bone marrow cells were for the most part negative. By absorption, the antigen reacting with natural thymocytotoxic autoantibody was found in thymus, lymph node, spleen and brain of adult mice, thymus of newborn mice and some leukaemias. Natural thymocytotoxic autoantibody in NZB mice was an IgM-globulin as determined by sensitivity to 2-mercaptoethanol treatment and by Sephadex G-200 column chromatography in contrast to other natural antibodies (antinuclear, antierythrocyte and G antibodies) of IgG-globulin class. NZB mice also produced natural antibodies against thymocytes of the rat and the hamster; these antibodies were species-specific and did not react with the thymocytes of any but the homologous species.     

Abnormal thymic expression of epithelial cell adhesion molecule (EP-CAM) in New Zealand Black (NZB) mice

New Zealand Black (NZB) mice have been well documented to have a variety of thymic epithelial cell microenvironmental abnormalities, including disruption of corticoepithelial cell networks and medullary cell clusters. These abnormalities of the thymic stromal network are particularly important because similar observations have been noted in other models of murine lupus. Thymic epithelial cells, a key component of the microenvironment, play an important role in selection of the mature T cell receptor repertoire. Recently, a homotypic calcium-independent human and murine epithelial cell adhesion molecule, Ep-CAM, has been described which is located at the thymocyto-cortical cell junction. The function of Ep-CAM is still unclear but its unique location within the thymus suggests that it is critical in the process of providing maturation signals. Consequently, we examined the thymic expression of Ep-CAM in a series of autoimmune prone mice by thymic distribution of Ep-CAM in NZB, NZW, NZB/W, BXSB-Yaa, MRL- lpr/lpr, C3H- gld/gld and the control strains BALB/c, C57BL6, C3H and MRL(+/+), by immunohistology and flow cytometry. Interestingly, NZB mice are similar to control mice from day 4 to 2 weeks of age, having a very low expression of Ep-CAM at the thymocyto-cortical junction. In control strains, there is a marked increased in expression of Ep-CAM beginning at 5 weeks of age. In contrast, NZB mice fail to show significant expression of Ep-CAM even well into adulthood. This abnormality of NZB mice was also noted in NZB/W F1 and BXSB mice, but not MRL- lpr/lpr or C3H- gld/gld mice. Given the potential importance of Ep-CAM in thymic selection, this study provides important evidence that a defective stromal microenvironment is likely to be of etiological significance in the susceptibility of NZB to autoimmune disease.