Vojtiskova and Pokorna (1964) found that thymectomy of adult mice prevented induction of EAO by testicular antigens+CFA-immunization. Later in the rat, neonatal thymectomy at 3 days of age led to the development of EAO without any active immunization in about 65% of Lewis rats (Hattori and Brandon 1979; Lipscomb et al. 1979) (Fig. 7.1). This is in contrast to classically induced experimental autoimmune model systems, where previous investigators have reported that thymectomy lessens or prevents induction of autoimmune disease. This difference should be related to the timing of thymectomy. Thymectomy after puberty when autoimmunogenic haploid TGC are already present in the testis may be effective in reduction of pathogenic effector T cells rather than Treg; however, thymectomy before appearance of haploid TGC in the testis during neonatal period failed to develop Treg ontogenically. In about 20–30% of susceptible strains of mice that were thymectomized neonatally on day 3, a mild EAO occurred along with other organ-localized autoimmune diseases such as thyroiditis and gastritis, without any artificial immunization (Taguchi and Nishizuka 1981). It is noteworthy that in this post-thymectomy EAO, epididymitis and vasitis consistently occurred prior to the development of EAO in mice. However, neonatal thymectomy on day 7 after birth could not induce any organ-localized autoimmunity involving EAO and survived well until a natural death. This indicates that thymectomy around day 3 is critical for disturbance of immune regulation. The EAO lesion resulted in testicular atrophy and was characterized by disappearance of mature sperms, formation of multinuclear giant cells in the seminiferous tubules, and infiltration of lymphocytes in the testicular interstitium. Deposition of IgG, IgA, IgM, and C3 were also identified on the basal lamina of both seminiferous tubules and epididymal ducts. Serum autoantibodies in the neonatally thymectomized mice exclusively bound to acrosomes of mature spermatozoa, but not to round and oval spermatids. It seems, therefore, that the testicular autoimmunity following neonatal thymectomy on day 3 may be directed predominantly to acrosomal proteins of mature spermatozoa within epididymal ducts, rather than to immature TGC within seminiferous tubules.

The effect of vasectomy on EAO was also investigated in the post-thymectomy model. The incidence of EAO was found to increase when day 3-thymectomized mice received vasectomy on day 60 after birth (Taguchi and Nishizuka 1981). Kojima and Spencer (1983) also reported that vasectomy increased the incidence of testicular atrophy in day 3-thymectomized mice. Interestingly, in day 3-thymectomized males and females, the incidence of autoimmune inflammation is apparently higher in the ovary than in the testis. While approximately 95% female thymectomized animals develop autoimmune oophoritis, only 20–30% male thymectomized animals develop EAO in (C57BL6 × A/J)F1 mice. However, the prevalence of EAO increased to over 90% when day 3-thymectomized mice were vasectomized later. This indicates that testis is a relatively privileged organ against autoimmune attack compared with the ovary, but the vasectomy disrupts the immune privileged circumstance.

Later, Tung et al. (1987a, b) reexamined the findings of post-neonatal thymectomy model by Taguchi and Nishizuka (1981). They found epididymo-vasitis in 70–90% of (SWR/J × A/J)F₁, 50% of (C57BL6 × A/J)F₁, and 64% of balb/cBy mice after day 3-thymectomy, whereas orchitis occurred in approximately 20% in these strains. In some mice, epididymo-vasitis also developed after not only day 3 but also day 7 thymectomy. It may be that the time window of neonatal thymectomy for induction of autoimmune epididymo-vasitis is wider than that for EAO. In day 3-thymectomized mice, the incidences of inflammation in the vas deferens, the cauda, the corpus, and the caput of the epididymis were almost equal. At 5–7 weeks, polymorphonuclear leukocytes dominated and were replaced by lymphocytes and macrophages between 8 and 18 weeks. The inflammatory cells were distributed in perivascular and peritubular spaces, and they rarely invaded epithelial cell linings. Maximal incidence of epididymitis preceded immune complex deposition, which is probably a consequence of tissue injury. Immune complexes were found in less than 25% of day 3-thymectomized mice during the first 11 weeks, when the epididymo-vasitis reached its peak incidence. Later, at 12–14 weeks, the frequency of immune complexes rose to 70%. Typically, linear deposits of mouse IgG were detected along the basal lamina of epididymal ducts and seminiferous tubules. However, differing from findings by Taguchi and Nishizuka (1981), deposits of IgA, IgM, and C3 were absent. In the testis, inflammatory cells infiltrate focally, and invasion of lymphocytes and macrophages into the seminiferous tubules with extensive disturbance of spermatogenesis was rare. Granular deposits of IgG and C3 were detected only in 30% of day 3-thymectomized mice. In regard to serum autoantibodies, a positive and significant correlation was found between the levels of autoantibodies against testicular autoantigens and EAO occurrence (Tung et al. 1987a, b). In contrast, there was no correlation between the autoantibodies levels and occurrence of autoimmune epididymo-vasitis. Furthermore, there was no correlation between autoantibodies against acrosome of epididymal spermatozoa and occurrence of EAO or autoimmune epididymitis. Although Taguchi and Nishizuka (1981) specifically found autoantibodies against acrosome of elongated spermatids and epididymal spermatozoa, Tung et al. (1987a, b) demonstrated autoantibodies against (1) large granular speckles in nuclei of epithelial cells of the caput and corpus epididymal ducts, (2) fine spikes at the luminal surface of epithelial cells in corpus epididymal ducts, (3) ring-shaped antigens surrounding basal cells in corpus epididymal ducts, and (4) linear antigens surrounding Sertoli cell nuclei. Additionally, unusual autoantibodies reactive with vascular smooth muscle were also found.

EAO with autoimmune epididymitis in day3-thymectomized mice could be prevented by injection of adult normal spleen cells on day 4 (Taguchi and Nishizuka 1981). The most effective donor source was of normal male. Spleen cells from normal females and day 0-orchidectomized donors were less effective for EAO prevention, and spleen cells of day 3-thymectomized male and female donors failed to prevent EAO, indicating that tolerance for autoantigens relevant to this EAO is ontogenically regulated. It also became apparent that spleen cells from normal males had adequate Treg for prevention of EAO but that spleen cells from day 3-thymectomized males have little Treg. Normally, the neonatal repertoire is enriched in peripheral autoreactive T cells in immature system of immunoregulation. Since neonatal thymectomy should affect the T cell repertoire in the neonate, followed by decline of Treg and prevalence of autoreactive T cells, this could explain why autoimmune diseases occur spontaneously in day 3-thymectomized mice. Indeed, V beta 11-positive autoreactive T cells have been found to be enriched in adult I-E⁺ day 3-thymectomized mice (Smith et al. 1989, 1991, 1992; Jones et al. 1990). Therefore, if the regulatory balance between the expansion of the autoreactive neonatal T cell repertoire and the relatively late ontogeny of Treg is tipped in favor of autoreactive T cell activity, EAO could occur. Testicular autoantigens in developing testis may leak latently for expansion of autoreactive T cells after puberty. There is another possibility that immune responses against some antigens with molecular mimicry can cause EAO through pathogenic T cell activation. In other words, the activation of testis-specific T cells by non-testicular peptides that mimic the target autoantigens cause enrichment of the EAO-inducing autoreactive T cells.

A previous study suggested that H-2-linked genes play little role in determining autoimmune disease outcome in neonatally thymectomized mice (Kojima and Prehn 1981). Studies using recombinant inbred strains showed that EAO susceptibility was not associated with the H-2 haplotype but appeared to be influenced by a minor histocompatibility locus and inherited as a recessive trait. In addition, the disease pathology was strictly controlled by the genetic background of the mice that carried the T cell receptor transgene. On the other hand, Del Rio et al. (2011) reported that H-2 control of natural Treg frequency in the lymph node correlates with susceptibility to day 3-thymectomy-induced autoimmunity. Quantitative trait loci on chromosome 17 (H-2) and also chromosomes 1, 2, 3, 7, and 16 control day 3-thymectomy-induced autoimmunity, and quantitative differences in the frequency of natural Treg in the lymph nodes, but not spleen or thymus, are present between the disease-resistant and disease-susceptible strains of mice. Using H-2-congenic mice, the observed difference in frequency of lymph node natural Treg is chiefly controlled by H-2 on chromosome 17. This indicates the existence of a lymph node-specific, H-2-controlled mechanism regulating the prevalence of natural Treg is critical for EAO in day 3-thymectomy model.

Cyclosporine A, a fungal metabolite, may depress the synthesis of certain cytokines that support T cell growth, or may affect thymopoiesis or deplete the thymic stromal cells, especially in the medulla. Cyclosporine A abrogated CD4⁺CD8⁻ and CD4⁻CD8⁺ thymocytes and also affected thymic epithelial cells (Hiramine et al. 1989a, b). Therefore, it may destroy mature thymocytes or block the maturation of immature thymocytes. Depletion of thymic stromal cells in the medulla may interfere with the proliferation/differentiation of thymocytes (Hiramine et al. 1989b). Cyclosporine A has been routinely used in fields of transplantation immunology for establishment of nearly permanent immunologic tolerance to allografts; however, treatment with cyclosporine A paradoxically causes multiple organ-localized autoimmune diseases in balb/c mice later in life when the cyclosporine A (10 mg/kg body weight per day) was administered daily for 1 week to newborn (Fig. 7.1). Although the incidences of autoimmune gastritis and oophoritis were 20% and 10%, respectively, in mice that had received neonatal cyclosporine A treatment, EAO was not observed. However, daily treatment with cyclosporine A from day 1 day to day 6 after birth followed by thymectomy on day 7 increased the incidences of gastritis (95%) and oophoritis (58%) and also developed EAO (36%), thyroiditis (30%), insulitis (13%), adrenitis (9%), and sialadenitis (22%). Cyclosporine A does not appear to induce de novo production of forbidden clones by interfering with clonal deletion mechanism in the thymus. It is likely that cyclosporine A would eventually affect the education of Treg in the thymus, as the formation of autoimmune lesions can be prevented by injection of splenic T cells from normal mice (Sakaguchi and Sakaguchi 1989, 1992). By the fact that cyclosporine A caused organ-localized autoimmune diseases involving EAO in mice when the drug was administered to newborns for a limited period, it is suggested that cyclosporine A primarily affects the neonatal thymus and interferes with the thymic production of Treg that regulates the expansion of autoreactive T cells. On the other hand, administration of cyclosporine A to adult mice failed to induce autoimmune disease presumably by that a sufficient number of Treg had been already produced and prevailed before starting cyclosporine A treatment, and cyclosporine A could not effectively eliminate them already delivered in the periphery. To investigate whether neonatal but not adult thymus is critical for cyclosporine A-induced autoimmunity or not, engrafting of the thymus from cyclosporine A-treated euthymic (nu/+) mice into syngeneic, athymic nude (nu/nu) mice was performed (Sakaguchi and Sakaguchi 1988). Athymic nude mice were engrafted with either one thymus from 7-day-old nu/+ mice treated daily with cyclosporine A (10 mg/kg body weight per day) for 1 week from the day of birth, or one thymic lobe from the adult mice administered daily with cyclosporine A (20 mg/kg per day) for 2 weeks. When examined histologically 3 months later, multiple organ-localized autoimmune diseases were developed, accompanied by appearance of circulating autoantibodies in the recipient nu/nu mice that had been transplanted with cyclosporine A-treated donor thymic lobes of not only newborn but also adult mice (Fig. 7.2). At the time of sacrifice, the transplanted thymi from both neonatal and adult donors were histologically populated with massive lymphocytes, and the composition of thymocyte subsets was similar to that of a normal adult thymus. These results suggest the following: (1) normal thymus is continuously producing potential autoreactive T cells as well as Treg; (2) cyclosporine A can selectively abrogate the thymic production of Treg cells in mice at any age; (3) however, for autoreactive T cells to proliferate and differentiate into pathogenic effector T cells, active Treg must be absent in the periphery or have not yet migrated to the periphery. Thus, engrafting of the thymus from cyclosporine A-treated mice of any age into athymic nude mice or cyclosporine A administration to euthymic newborn mice can cause autoimmune disease, but its administration to euthymic adult mice does not.

Total lymphoid irradiation is effective for treatment of lymphoid malignancies and autoimmune disease in humans and rodents. Furthermore, total lymphoid irradiation also establishes allograft tolerance when allogeneic bone marrow cells or solid organs are engrafted immediately after the irradiation. However, it was found that the irradiation can functionally alter the immune system and paradoxically break self-tolerance (Sakaguchi et al. 1994a). High dose (42.5 Gy), fractioned (2.5 Gy, 17 times) total lymphoid irradiation on mice caused various organ-localized autoimmune diseases involving EAO and autoimmune epididymitis (Fig. 7.1). The incidences of autoimmune gastritis, thyroiditis, sialadenitis, and EAO with epididymitis were approximately 75%, 4%, 7%, and 9%, respectively. Total lymphoid irradiation eliminated the majority of mature thymocytes and the peripheral T cells for 1 month, and inoculation of spleen cells, thymocytes, or bone marrow cells prepared from syngeneic non-irradiated mice within 2 weeks after total lymphoid irradiation effectively prevented the development of the autoimmune diseases. Depletion of CD4⁺ T cells from the inoculated donor lymphocytes abrogated the disease preventive activity. CD4⁺ T cells also appeared to mediate the autoimmune diseases because CD4⁺ T cells from disease-bearing irradiated mice adoptively transferred the autoimmune lesions to syngeneic naïve mice (Fig. 7.3), indicating that autoimmune disease is caused by affecting the T cell immune system, rather than the target autoantigens, presumably by altering CD4⁺ Treg-dependent control of autoreactive CD4⁺ T cells. Furthermore, balb/c athymic nude mice spontaneously developed EAO with other autoimmune diseases when transplanted adult thymuses were irradiated before transplantation (Sakaguchi and Sakaguchi 1990) (Fig. 7.2). In this transplantation model, thymi were removed from donor mice 2 days after 9.0 Gy whole-body irradiation.

It must be also noted that irradiation alters not only the immune system but also the BTB. The BTB plays an important role in the intact spermatogenesis, and uncontrolled permeability of the BTB results in leakage of TGC autoantigens with the resultant anti-TGC autoimmune responses. A single local application of microwave electromagnetic pulse irradiation of non-thermal intensity to the testes (400 kV/m) resulted in decreased levels of mRNA and protein expressions of tight-junction-associated proteins (zonula occludens-1 and occludin) of the BTB, followed by the development of autoimmune process in the testes of mice and rabbits (Grigorev et al. 1981; Wang et al. 2010; Hou et al. 2012). The pathology was produced by both humoral and cellular immunity against testicular autoantigens and characterized by structural disturbance of the seminiferous tubular walls and the spermatogenic disturbance. Transforming growth factor-beta 3 is also a key molecule involved in the BTB permeability via regulation of tight junctions. In mice that had received electromagnetic radiation, transforming growth factor-beta 3 significantly decreased with increase of serum anti-sperm autoantibodies levels (Wang et al. 2010). On the contrary, increase of both mRNA and protein expressions of transforming growth factor-beta 3 with increase of the apoptotic TGC was also reported in electromagnetic pulse-exposed mice (Luo et al. 2013).

Exposure to ionizing radiation also induced male infertility, accompanied by increasing permeability of the BTB in mice (Son et al. 2015). The diameter and epithelial depth of the seminiferous tubules were significantly decreased in 1.7 Gy-irradiated mice, which showed significantly decreased levels of tight junction-associated proteins such as zonula occludens-1 and occludin-1 and increased serum anti-sperm autoantibodies compared with those of the non-irradiated animals. In 6.0 Gy-irradiated mice, serum anti-TGC autoantibody levels were also significantly elevated; however, lymphocytic infiltration was hardly seen in the testes in spite of exhibiting severe spermatogenic disturbance (Takahashi et al. 2017). It seems apparent that damage to the BTB integrity results in leakage of TGC autoantigens, leading to the induction of anti-TGC autoimmunity. Therefore, the ionizing irradiation should induce the spermatogenic disturbance by direct killing of TGC and also the BTB damage-induced defect of TGC differentiation. The following production of serum anti-TGC autoantibodies might infiltrate the seminiferous epithelium through the damaged BTB region and further damage the spermatogenic state.

Taken together, high-dose ionizing irradiation induces Treg depletion, TGC death, and the BTB disruption. However, it remains unknown how the systemic Treg depletion, TGC death, the BTB damage, and the following leakage of TGC autoantigens from the disrupted BTB cooperate with each other for EAO induction in irradiated mice. However, effects of irradiation on the testicular tissues should not be noted in EAO in non-irradiated mice that received CD4⁺ T cells from disease-bearing irradiated mice and also in non-irradiated athymic nude mice engrafted with adult thymuses that had been irradiated before transplantation (Sakaguchi and Sakaguchi 1990) (Fig. 7.2).

Transfer of adult CD5^low spleen T cells of normal mice to athymic nude mice, to mice without T cells, or to scid mice induces multiple organ-localized autoimmune diseases (Sugihara et al. 1988; Smith et al. 1992). The incidence of EAO was 40% (Sakaguchi et al. 1985) (Fig. 7.3). Therefore, the existence of pathogenic T cells in normal individuals, and their regulation by Treg, has been demonstrated on this EAO model. The CD4⁺CD8⁻ subset represents mature thymocytes that have passed beyond the stage of T cell development in which the deletion of autoreactive T cells is expected. However, mature pathogenic autoreactive T cells are not deleted in normal thymus and can induce multiple organ-localized autoimmune diseases by their transfer to athymic nu/nu recipients. Transfer of unfractioned spleen T cells of normal adult donors did not evoke multiple organ-localized autoimmune diseases although transfer of fractioned CD4⁺ T cells that expresses a low level of cell surface CD5 molecules was able to do so. In this experimental system, splenic T cells of normal adult donors were treated with CD5 antibody and complement, and the residual T cells were then injected into recipients for induction of multiple organ-localized autoimmune diseases. Although 100% of T cells are known to be CD5⁺, this treatment eliminated 95% of the CD4⁺ T cells, as demonstrable by flow cytometry. It is noted that the residual 5%, which expressed a low level of CD5 (CD4⁺ CD5^low T cells), were responsible for disease transfer (Smith et al. 1992).

EAO also develops in athymic syngeneic recipients of neonatal but not adult splenic CD4⁺ T cells (Fig. 7.3). This finding is consistent with the results of previous studies of V beta 11-positive T cells that are undeleted in neonatal thymus and are enriched in the neonatal spleen for recognition of the endogenous retroviral peptides and superantigens (Smith et al. 1989; Woodland et al. 1991; Dyson et al. 1991).

balb/c athymic nu/nu mice spontaneously developed EAO with other autoimmune diseases such as oophoritis gastritis, thyroiditis, arteritis, glomerulonephritis, and polyarteritis when transplanted with “newborn” (0 or 1 day old) balb/c thymus (Sakaguchi and Sakaguchi 1990) (Fig. 7.2). The incidences of EAO, oophoritis gastritis, thyroiditis, vasculitis, and glomerulonephritis were approximately 8%, 25%, 61%, 4%, 23%, and 10%, respectively. Transplantation of thymus from adult balb/c mice was far less effective in inducing histologically evident autoimmune disease in athymic nu/nu mice. Furthermore, multiple organ-localized autoimmune diseases were not induced in athymic nu/nu mice grafted with syngeneic “embryonic” thymus (Nishigaki-Maki et al. 1999; Morimoto et al. 2000). To determine whether spontaneous autoimmune disease after thymus engrafting is unique to nu/nu mice, balb/c newborn thymi were engrafted into T cell-depleted balb/c mice that had received thymectomy at 6 weeks of age, irradiation at 2 weeks later at 9.0 Gy, and transplantation of 5 × 10⁶ syngeneic bone marrow cells treated with anti-Thy-1.2 plus rabbit complement. The results showed that similar autoimmune diseases were produced in “newborn” thymus-engrafted T cell-depleted recipient mice but not in the engrafted normal recipients. These results indicate that the balb/c mice have pathogenic autoreactive T cells in their thymi, and such autoreactive T cells spontaneously expand and cause autoimmune disease when released to the T cell-deficient/eliminated periphery. Various manipulations that deplete Treg (CD4⁺CD25⁺ T cells) potentially activate orchitogenic CD4⁺ T cells and promote EAO induction, but the presence of normal T cells including “natural” Treg suppresses the disease induction.