Failure of CD25+ T Cells from Lupus-Prone Mice to Suppress Lupus Glomerulonephritis and Sialoadenitis1
http://www.100md.com
免疫学杂志 2005年第14期
Abstract
The development of organ-specific autoimmune diseases in mice thymectomized on day 3 of life (d3tx mice) can be prevented by transferring CD4+CD25+ T cells from syngeneic, normal adult mice. Using a d3tx model, we asked whether CD4+CD25+ T cell deficiency contributes to glomerulonephritis (GN) in lupus-prone mice. New Zealand Mixed 2328 (NZM2328) mice spontaneously develop autoantibodies to dsDNA and female-dominant, fatal GN. After d3tx, both male and female NZM2328 mice developed 1) accelerated dsDNA autoantibody response, 2) early onset and severe proliferative GN with massive mesangial immune complexes, and 3) autoimmune disease of the thyroid, lacrimal gland, and salivary gland. The d3tx male mice also developed autoimmune prostatitis. The transfer of CD25+ cells from 6-wk-old asymptomatic NZM2328 donors effectively suppressed dsDNA autoantibody and the development of autoimmune diseases, with the exception of proliferative lupus GN and sialoadenitis. This finding indicates that NZM2328 lupus mice have a selective deficiency in T cells that regulates the development of lupus GN and sialoadenitis. After d3tx, the proliferative GN of female mice progressed to fatal GN, but largely regressed in the male, thereby revealing a checkpoint in lupus GN progression that depends on gender.
Introduction
The CD4+CD25+ regulatory T cell (Treg)3 subset has been established as a powerful regulator of the host immune response to foreign and self Ags (1, 2). Adoptive transfer of Treg cells from a normal host silences the autoreactive T cell response and prevents or abrogates autoimmune disease development. Study of the spontaneous fatal human immune dysregulation, polyendocrinopathy, enteropathy, X-linked syndrome, and the fatal lymphoproliferative disorder in scurfy mice has provided compelling evidence that autoimmune disease can result from functional mutations of the forkhead/winged helix transcription factor, which is essential for Treg development and function (3, 4, 5). This exciting finding begs the question of whether deficiency in Treg cells also contributes to the pathogenesis of the more common human autoimmune diseases, such as systemic lupus erythematosus (SLE), multiple sclerosis, and rheumatoid arthritis. Indeed, recent studies have described a deficiency of in vitro CD4+CD25+ T cell function in patients with multiple sclerosis and rheumatoid arthritis (6, 7). Reduced numbers of CD25+ T cells in the peripheral blood of SLE patients have also been reported (8). However, the clinical studies are not able to evaluate the disease-suppressing capacity of Treg cells in vivo. We therefore turned to murine models of SLE to determine whether Treg cells in these mice are deficient in their capacity to inhibit spontaneous systemic autoimmune disease.
SLE is a complex multisystemic autoimmune disease, characterized by circulating Ab to nuclear and cytoplasmic Ags (9). Lupus glomerulonephritis (GN) affects 50% of lupus patients and is associated with the deposition of immune complexes and complement components in the glomeruli, with infiltration of inflammatory cells and, in severe disease, glomerular and interstitial fibrosis, tubular atrophy, and renal failure (10). The lupus-prone New Zealand Mixed 2328 (NZM2328) mouse, a recombinant inbred strain that originated from the crosses among New Zealand Black and New Zealand White mice and their progenies also develops GN (11). NZM2328 mice have autoantibodies to nuclear Ags including dsDNA. At 12 mo of age, 72% of females and 25% of males develop severe proteinuria with lupus GN (12).
Previous studies have explored the presence of Treg cells for the dsDNA autoantibody response. Seo et al. (13) showed that the transfer of Treg cells into mice expressing anti-dsDNA BCRs abrogated the maturation process of transgenic B cells into Ab-forming cells. La Cava et al. (14) showed that Treg cells from (New Zealand Black x New Zealand White) F1 mice injected with a tolerogenic peptide from the CDR3 region of an anti-dsDNA Ab prevented B cell synthesis of dsDNA Ab. In these studies the question of whether CD25+ regulatory T cells can suppress lupus GN was not investigated.
In the present study we used the classical approach to induce autoimmune disease by thymectomy (tx) on day 3 of life (d3tx) (15, 16). Although the mechanism of d3tx-induced autoimmune diseases is not fully understood, the current belief is that d3tx results in preferential depletion of CD4+CD25+ Treg cells relative to the autoreactive CD25– T effector population. The resultant state of regulatory and effector T cell imbalance is also exaggerated by homeostatic T cell expansion in the profoundly lymphopenic d3tx mice (17, 18). Regardless of the precise mechanism of disease induction, all d3tx-induced, organ-specific autoimmune diseases in nonlupus mice are readily prevented by early infusion of CD4+CD25+ T cells from normal syngeneic adults (19, 20).
In this study we have investigated the capacity of CD25+ regulatory T cells from lupus-prone NZM2328 mice to suppress autoimmune disease and autoantibody response in d3tx NZM2328 lupus mice. The study was possible because of the observation, described below, that d3tx NZM2328 mice developed severe proliferative GN of early onset, sialoadenitis, and accelerated autoantibody response to dsDNA. As an important control, we studied the capacity of the same population of CD25+ T cells from NZM2328 mice to suppress other organ-specific autoimmune diseases unrelated to the lupus syndrome.
Materials and Methods
Mice
A colony of NZM2328 has been maintained and housed under specific pathogen-free conditions at the University of Virginia Animal care facility. All mice used were generated in this facility. The animal housing and experimental protocols were performed in accordance with National Institutes of Health guidelines. Male and female NZM2328 mice were thymectomized (tx) on days 2–4 after birth using established protocols and aseptic technique (21). Briefly, 2- to 4-day-old mice were anesthetized using hypothermia, and the sternum was exposed through a midline incision. The sternum and ribs were cut, and the thymus was aspirated using a Pasteur pipette. The wound was closed using surgical glue. At the time of death, tissue from the superior mediastinum was collected and studied by histology for the presence of any residual thymus. Mice with residual thymus were excluded from the study. Sham-tx mice underwent surgery on days 2–4 after birth without removal of the thymus.
Monitoring and autopsy
The mice were monitored for proteinuria, and tail blood was collected at monthly intervals. At 20 or 30 wk of age, the mice were killed, and a complete autopsy was performed. Organs (kidneys, salivary glands, lacrimal glands, thyroid, ovary, testes, and prostate) were collected in Bouins’ fixative and processed for histopathology. One kidney was snap frozen in liquid nitrogen and used for immunofluorescence staining and Ab elution studies.
Histopathology, GN severity index and organ specific autoimmunity. Four-micron sections were cut from Bouins-fixed, paraffin-embedded tissues and stained with H&E. An observer blinded to the experimental design evaluated histopathology. The severity of acute GN was graded based on the percentage of glomeruli involved, the severity of inflammatory cell infiltration, and proliferative changes in mesangial and/or peripheral regions. Glomeruli were measured at their maximum diameter using an ocular micrometer, and numbers of nuclei in each glomerulus were counted. The data for glomerular size and cellularity presented are the mean of readings from 10 glomeruli for each mouse. Chronic GN was scored based on glomerular fibrosis, interstitial fibrosis, and tubular atrophy. Each of the above-mentioned changes was scored from 0 to 4, with 0 indicating no pathology and 4 being maximum pathology. The severity index represents a cumulative score. Severity of organ-specific disease was scored based on the presence of inflammatory infiltrates in different tissues, with 0 as normal, 1–3 indicating increasing severity of infiltration, and 4 associated with destruction of normal architecture.
Estimation of anti-dsDNA IgG and total IgG by ELISA
Ninety-six-well plates were coated with streptavidin (1 mg/ml), followed by biotinylated plasmid DNA as previously described (22). After blocking with 3% BSA in PBS, the plates were incubated with sera at a 1/100 dilution. Bound Abs were detected with HRP-conjugated goat anti-mouse IgG, followed by o-phenylene diamine and hydrogen peroxide in citrate phosphate buffer, pH 5.0. The reaction was stopped with 2.5 N sulfuric acid, and the OD was read at 490 nM in an ELISA reader. An mAb to dsDNA (R4A; gift from Dr. B. Diamond, Albert Einstein College of Medicine, Bronx, NY) was used as the standard. IgG levels in sera and kidney eluates were estimated using a sandwich ELISA as previously described (21). Purified mouse IgG at different concentrations (4–500 ng/ml) was used to construct a standard curve.
Detection of renal immune complexes by direct immunofluorescence
Immune complexes in renal glomeruli were detected by direct immunofluorescence (21). Five-micron sections of snap-frozen kidney were fixed in ice-cold acetone for 10 min. The sections were rinsed with PBS and blocked with normal goat serum in 3% BSA/PBS (1/10, v/v) for 20 min. The sections were then incubated with FITC-conjugated goat anti-mouse IgG (Southern Biotechnology Associates; 1/50 dilution) for 45 min. After two rinses in PBS, the tissue sections were mounted with Vectashield mounting fluid (Vector Laboratories), and fluorescence intensity was graded from 0 to 4 in a blinded fashion.
Acid elution of Igs from kidney
Igs were eluted from the kidney as previously described (23). Briefly, each individual frozen kidney was weighed and homogenized in cold PBS on ice. The homogenate was centrifuged at 1200 x g for 10 min at 4°C. The pellets were washed by resuspension in cold PBS, followed by vortexing and centrifugation to remove serum proteins. The washing was repeated until the OD280 of the supernatant was <0.05. The pellet was then suspended in 0.1 M glycine with 1% BSA, pH 2.8 (10 ml/g kidney), and mixed gently at 4°C for 20 min. The suspension was centrifuged, and supernatant was immediately neutralized with 1 M Tris base.
Purification of CD25+ and CD25– T cells
Lymph nodes were harvested from NZM2328 mice and enriched for T cells by negative selection on T cell enrichment columns (R&D Systems). The cells were incubated with biotinylated 7D4 Ab to CD25 (BD Pharmingen). After washing, the cells were incubated with streptavidin-PE (Rockland), followed by anti-PE magnetic beads (Miltenyi Biotec). The cells were washed and then passed over a magnetic column using the double-positive selection mode in an autoMACS cell separation system (Miltenyi Biotec). The unlabeled CD25– cells were collected in the flow-through fraction. The double-positive selected CD25+ fraction of >90% purity was used for passive transfer experiments as well as in vitro assays.
Transfer of CD25+ T cells into d3tx mice
Seven- to 10-day-old d3tx mice were given a single injection of purified CD25+ T cells (106cells/mouse i.p.) Tail blood was collected from the recipient mice at 13, 15, 17, and 20 wk. The mice were killed at 20 wk, and kidneys were studied. A cohort of mice was killed at 30 wk and also studied for organ-specific disease.
Statistical analysis
Methods for statistical analysis include Student’s t test, Mann-Whitney U test, 2 analysis, and ANOVA using PRISM version 3.02 (GraphPad).
Results
D3tx accelerated dsDNA autoantibody production and renal glomerular immune complex deposition in male and female NZM2328 mice
D3tx NZM2328 male and female mice were studied for serum dsDNA Ab and serum IgG at 5, 9, 13, and 20 wk. Compared with sham-tx mice, the dsDNA Ab response of d3tx mice was greatly accelerated (Fig. 1). A significant increase in total serum IgG was also seen in d3tx mice (females, p = 0.04; males, p < 0.0001; data not shown). To control for hypergammaglobulinemia, the data are expressed as the dsDNA Ab titer per U of serum IgG in Fig. 1.
The d3tx male and female mice had comparable incidence of proliferative GN (87 vs 90%, respectively) of comparable severity (Fig. 4). However, in 43% of d3tx females, in addition to the proliferative GN described above, there was evidence of chronic GN, including glomerulosclerosis, epithelial cell crescents, interstitial inflammation, and atrophy of renal tubules (Figs. 3D, 4D, and 5C).
Therefore, d3tx greatly accelerated the development of lupus GN in NZM2328 mice. In male mice, the changes included exclusively an early-onset proliferative GN characterized by profound glomerular cell proliferation and accumulation of mesangial immune complexes. In female NZM2328 mice, the d3tx effect represents an accelerated natural disease, with progression of the glomerular disease process as a whole, leading to high prevalence of both proliferative GN and chronic GN.
D3tx induces or enhances extrarenal autoimmune disease in NZM2328 mice
In addition to GN, d3tx NZM2328 mice developed autoimmune disease that affected other organs (Fig. 6). Prostatitis was detected in 73% of d3tx males. Thyroiditis of greater severity and Ab to thyroid Ags of high frequency were detected in d3tx mice, with greater severity in females (data not shown). Similarly, inflammation of the salivary glands (sialoadenitis) and lacrimal glands (dacryoadenitis), detectable in many sham tx mice, was significantly more frequent and more severe in d3tx mice (Fig. 6). The tissue inflammation, first detected at 20 wk, reached high incidence at 30 wk when disease incidence was recorded (Fig. 7).
CD25+ T cells from young adult NZM2328 donors suppressed d3tx-induced autoimmune disease of prostate, thyroid, and lacrimal glands, but not proliferative GN or sialoadenitis
To determine the capacity of CD25+ T cells to suppress autoimmune disease in d3tx NZM2328 mice, CD25+ T cells were obtained from 6- to 8-wk-old untreated NZM2328 donors that exhibited no detectable renal disease or serum dsDNA Abs. The cells were transferred into 7- to 10-day-old d3tx NZM2328 recipients, and each recipient’s renal pathology was evaluated at 20 wk. A second cohort was killed at 30 wk to study the pathology of organs other than the kidneys.
Transfer of 106 CD25+ T cells to d3tx NZM2328 mice completely suppressed prostatitis and significantly reduced the severity of inflammation in thyroid and lacrimal gland (Fig. 7). In contrast, the transfer of CD25+ T cells had no observable effect on the proliferative GN at 20 wk (Fig. 8). In both males and females, the extent of proliferative GN in d3tx mice that received CD25+ T cells was not reduced compared with that in d3tx mice. The disease was also comparable to that in another group of d3tx mice given purified CD25– T cells (106/mouse) at the same time as CD25+ T cells (data not shown). Similar to GN, sialoadenitis in d3tx mice was not affected by CD25+ T cell transfer (Fig. 7). Therefore, CD25+ T cells from NZM2328 mice, which strongly prevented the development of prostatitis, thyroiditis, and dacryoadenitis in d3tx mice, had little or no influence on the accelerated development of lupus GN and sialoadenitis in NZM2328 mice.
Progression from acute to chronic GN in NZM2328 mice depends on gender
To determine the fate of proliferative GN in d3tx mice, we studied a cohort of d3tx mice at 30 wk and compared their data to those obtained at 20 wk. A dramatic difference was seen in the disease progression between male and female mice (Fig. 10A). Many d3tx or sham-tx female NZM2328 mice developed severe chronic GN by 30 wk. In contrast, only rare d3tx NZM2328 males progressed to chronic GN; indeed, most male d3tx mice were completely free of renal histopathology at 30 wk. Interestingly, despite the absence of renal pathology, d3tx male NZM2328 mice continued to have high levels of serum dsDNA Abs (data not shown), and the renal glomeruli continued to harbor large quantities of IgG immune complexes (Fig. 10B) with C3 deposition (data not shown). Thus, the progression of lupus GN evident in d3tx NZM2328 mice was strongly influenced by gender, but this was independent of the dsDNA Ab response.
Discussion
Mice tx between days 3–5 of life develop organ-specific autoimmune disease. The organ susceptibility (stomach, thyroid, ovary, prostate, and lacrimal gland) is mouse strain dependent (BALB/c, C3H, and A/J, respectively) (24). Evidence for the mechanisms of d3tx disease induction suggest three major factors: 1) a differential ontogeny of autoreactive T effectors and inhibitory T regulators (19), 2) depletion of thymus-derived CD25+ Treg cells, and 3) d3tx-induced lymphopenia and the resultant homeostatic proliferation (17). In this study we have developed an experimental model of accelerated lupus GN in lupus-prone NZM2328 mice by d3tx. With this model we have investigated the in vivo function of thymus-derived CD4+CD25+ T cells in the regulation of spontaneous systemic autoimmune disease.
The systemic autoimmunity in NZM2328 mice is not associated with a global deficiency in functional CD25+ Treg cells. CD25+ T cells purified from 2- and 5-mo-old NZM2328 females were found to efficiently prevent the proliferation of Con A-stimulated CD25– T cells (data not shown). Coculture of purified CD25+ from 2- and 5-mo-old NZM2328 females with CD25– T cells in a 1:1 ratio was able to inhibit proliferation of CD25– T cells by 96 and 95%, respectively. Moreover, the CD25+ Treg cells of NZM2328 mice were able to suppress the development of organ-specific autoimmune disease of the prostate, thyroid, and lacrimal glands, diseases that also developed in d3tx NZM2328 mice, but were unrelated to the lupus manifestation. In contrast, infusion of CD25+ T cells did not influence the development of lupus GN. Detection of global Treg abnormality is the focus of most clinical research on Treg cells in human autoimmune diseases.
Given the general integrity of Treg in NZM2328 mice, the finding that they failed to suppress the d3tx-accelerated GN is of considerable interest. This result is consistent with the hypothesis that abnormal immunoregulation in lupus-prone NZM2328 mice is due to a selective deficiency of Treg cells that control the development of autoimmune GN and sialoadenitis in these animals. Defective Ag-specific Treg cells may thus contribute to the pathogenesis of lupus GN.
There is clear evidence for Ag-specific regulation of other autoimmune diseases. In d3tx mice and rats, autoimmune prostatitis and autoimmune thyroiditis suppression were more efficient if the CD4+ T cells were derived from donors that harbored the target Ag (25, 26). Induction of insulin-dependent diabetes mellitus by diabetogenic T cells in lymphopenic hosts could be prevented by cotransfer of large numbers (10–20 million) of Treg cells from naive NOD compared with only 2 million islet Ag-specific Treg cells when cotransferred with diabetogenic T cells (reviewed in Ref. 27). Recent studies show that exposure to ovarian or prostate Ags in vivo resulted in enhanced regulatory capacity of CD25+ T cells in prevention of oophoritis or prostatitis, respectively (E. S. Samy and K. S. K. Tung, unpublished observations; Y. Y. Seitady, K. Ohno, and K. S. K. Tung, unpublished observations). Thus, efficient regulation of autoimmune disease by CD25+ regulatory T cells is determined by Ag-specific suppression.
D3tx in NZM2328 mice has offered the opportunity to compare the same set of polyclonal Treg cells in the suppression of different autoimmune diseases that occur in the same individual. We currently cannot rule out the possibility that failure in suppression of GN and sialoadenitis may be related to the cell dose used in suppression; namely, if more cells were transferred to the d3tx recipients at different ages, perhaps disease suppression would be detectable. Nevertheless, our finding is consistent with the conclusion that with a single experimental protocol, the CD25+ T cells from lupus-prone mice exhibited a differential threshold of suppression for organ-specific autoimmune diseases vs lupus GN and sialoadenitis.
SLE is a complex systemic autoimmune disease, and the pathogenesis of tissue injury is not completely understood. For the pathogenesis of lupus GN, the emphasis has continued to focus on the role of autoantibodies and the attendant glomerular immune complexes that invariably accompany lupus GN, although recent studies have begun to address T cell-mediated immunity independent of Th cells for autoantibody response (28). In d3tx NZM2328 male mice, accelerated anti-dsDNA Ab and immune complex deposition was not accompanied by accelerated end-stage kidney disease. Secondly, infusion of CD25+ Treg cells in d3tx mice had little or no effect on accelerated lupus GN, but strongly suppressed the early spontaneous dsDNA autoantibody response in NZM2328 mice. These findings point to the dissociation between autoantibody production and target organ injury in NZM2328 mice and raise questions about the pathogenic requirement of dsDNA autoantibody in lupus GN. This is supported by genetic studies in NZM2328 mice (12, 29).
In the present study, additional experimental evidence was obtained that dissociated acute from chronic lupus GN. At 20 wk, d3tx results in an accelerated onset of acute GN in males normally resistant to GN. Significantly, in contrast to d3tx females, this acute GN does not progress to chronic GN. Thus, acute GN represents a checkpoint between inflammation and disease or loss of function that is influenced by gender. The lack of progression to chronic GN in d3tx male NZM2328 mice is similar to the progression from peri-insulitis to insulitis and diabetes mellitus in NOD mice (30) and is relevant to the human disease. Indeed, a report by Pollock and Pirani (31) identified a significant number of lupus patients with a histopathological diagnosis of mild glomerulitis and failed to progress over an 8-year follow-up period.
The findings in d3tx NZM2328 mice have raised important questions fundamental to the pathogenetic mechanism of lupus GN and sialoadenitis and suggest a future direction for the investigation of experimental and human lupus.
Acknowledgments
We are grateful to Dr. Shu Man Fu (University of Virginia, Charlottesville, VA PI:P50AR45222) for support and critical comments, and to Sharon Mangawang for expert technical assistance.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by National Institutes of Health Grants R01AI41236, AR45222, and K01DK063065.
2 Address correspondence and reprint requests to Dr. Harini Bagavant, Division of Rheumatology, Department of Internal Medicine, Health Sciences Center, Box 800412, University of Virginia, Charlottesville, VA 22908. E-mail address: hb5u@virginia.edu
3 Abbreviations used in this paper: Treg cell, CD25+ regulatory T cell; d3tx, tx on day 3 of life; GN, glomerulonephritis; NZM, New Zealand Mixed; tx, thymectomy; SLE, systemic lupus erythematosus; tx, thymectomy, thymectomized.
Received for publication March 18, 2005. Accepted for publication May 13, 2005.
References
Piccirillo, C. A., E. M. Shevach. 2004. Naturally-occurring CD4+CD25+ immunoregulatory T cells: central players in the arena of peripheral tolerance. Semin. Immunol. 16: 81-88.
Sakaguchi, S.. 2004. Naturally arising CD4+ regulatory T cells for immunologic self-tolerance and negative control of immune responses. Annu. Rev. Immunol. 22: 531-562.
Bassuny, W. M., K. Ihara, Y. Sasaki, R. Kuromaru, H. Kohno, N. Matsuura, T. Hara. 2003. A functional polymorphism in the promoter/enhancer region of the FOXP3/scurfin gene associated with type 1 diabetes. Immunogenetics 55: 149-156.(Harini Bagavant2 and Kenn)
The development of organ-specific autoimmune diseases in mice thymectomized on day 3 of life (d3tx mice) can be prevented by transferring CD4+CD25+ T cells from syngeneic, normal adult mice. Using a d3tx model, we asked whether CD4+CD25+ T cell deficiency contributes to glomerulonephritis (GN) in lupus-prone mice. New Zealand Mixed 2328 (NZM2328) mice spontaneously develop autoantibodies to dsDNA and female-dominant, fatal GN. After d3tx, both male and female NZM2328 mice developed 1) accelerated dsDNA autoantibody response, 2) early onset and severe proliferative GN with massive mesangial immune complexes, and 3) autoimmune disease of the thyroid, lacrimal gland, and salivary gland. The d3tx male mice also developed autoimmune prostatitis. The transfer of CD25+ cells from 6-wk-old asymptomatic NZM2328 donors effectively suppressed dsDNA autoantibody and the development of autoimmune diseases, with the exception of proliferative lupus GN and sialoadenitis. This finding indicates that NZM2328 lupus mice have a selective deficiency in T cells that regulates the development of lupus GN and sialoadenitis. After d3tx, the proliferative GN of female mice progressed to fatal GN, but largely regressed in the male, thereby revealing a checkpoint in lupus GN progression that depends on gender.
Introduction
The CD4+CD25+ regulatory T cell (Treg)3 subset has been established as a powerful regulator of the host immune response to foreign and self Ags (1, 2). Adoptive transfer of Treg cells from a normal host silences the autoreactive T cell response and prevents or abrogates autoimmune disease development. Study of the spontaneous fatal human immune dysregulation, polyendocrinopathy, enteropathy, X-linked syndrome, and the fatal lymphoproliferative disorder in scurfy mice has provided compelling evidence that autoimmune disease can result from functional mutations of the forkhead/winged helix transcription factor, which is essential for Treg development and function (3, 4, 5). This exciting finding begs the question of whether deficiency in Treg cells also contributes to the pathogenesis of the more common human autoimmune diseases, such as systemic lupus erythematosus (SLE), multiple sclerosis, and rheumatoid arthritis. Indeed, recent studies have described a deficiency of in vitro CD4+CD25+ T cell function in patients with multiple sclerosis and rheumatoid arthritis (6, 7). Reduced numbers of CD25+ T cells in the peripheral blood of SLE patients have also been reported (8). However, the clinical studies are not able to evaluate the disease-suppressing capacity of Treg cells in vivo. We therefore turned to murine models of SLE to determine whether Treg cells in these mice are deficient in their capacity to inhibit spontaneous systemic autoimmune disease.
SLE is a complex multisystemic autoimmune disease, characterized by circulating Ab to nuclear and cytoplasmic Ags (9). Lupus glomerulonephritis (GN) affects 50% of lupus patients and is associated with the deposition of immune complexes and complement components in the glomeruli, with infiltration of inflammatory cells and, in severe disease, glomerular and interstitial fibrosis, tubular atrophy, and renal failure (10). The lupus-prone New Zealand Mixed 2328 (NZM2328) mouse, a recombinant inbred strain that originated from the crosses among New Zealand Black and New Zealand White mice and their progenies also develops GN (11). NZM2328 mice have autoantibodies to nuclear Ags including dsDNA. At 12 mo of age, 72% of females and 25% of males develop severe proteinuria with lupus GN (12).
Previous studies have explored the presence of Treg cells for the dsDNA autoantibody response. Seo et al. (13) showed that the transfer of Treg cells into mice expressing anti-dsDNA BCRs abrogated the maturation process of transgenic B cells into Ab-forming cells. La Cava et al. (14) showed that Treg cells from (New Zealand Black x New Zealand White) F1 mice injected with a tolerogenic peptide from the CDR3 region of an anti-dsDNA Ab prevented B cell synthesis of dsDNA Ab. In these studies the question of whether CD25+ regulatory T cells can suppress lupus GN was not investigated.
In the present study we used the classical approach to induce autoimmune disease by thymectomy (tx) on day 3 of life (d3tx) (15, 16). Although the mechanism of d3tx-induced autoimmune diseases is not fully understood, the current belief is that d3tx results in preferential depletion of CD4+CD25+ Treg cells relative to the autoreactive CD25– T effector population. The resultant state of regulatory and effector T cell imbalance is also exaggerated by homeostatic T cell expansion in the profoundly lymphopenic d3tx mice (17, 18). Regardless of the precise mechanism of disease induction, all d3tx-induced, organ-specific autoimmune diseases in nonlupus mice are readily prevented by early infusion of CD4+CD25+ T cells from normal syngeneic adults (19, 20).
In this study we have investigated the capacity of CD25+ regulatory T cells from lupus-prone NZM2328 mice to suppress autoimmune disease and autoantibody response in d3tx NZM2328 lupus mice. The study was possible because of the observation, described below, that d3tx NZM2328 mice developed severe proliferative GN of early onset, sialoadenitis, and accelerated autoantibody response to dsDNA. As an important control, we studied the capacity of the same population of CD25+ T cells from NZM2328 mice to suppress other organ-specific autoimmune diseases unrelated to the lupus syndrome.
Materials and Methods
Mice
A colony of NZM2328 has been maintained and housed under specific pathogen-free conditions at the University of Virginia Animal care facility. All mice used were generated in this facility. The animal housing and experimental protocols were performed in accordance with National Institutes of Health guidelines. Male and female NZM2328 mice were thymectomized (tx) on days 2–4 after birth using established protocols and aseptic technique (21). Briefly, 2- to 4-day-old mice were anesthetized using hypothermia, and the sternum was exposed through a midline incision. The sternum and ribs were cut, and the thymus was aspirated using a Pasteur pipette. The wound was closed using surgical glue. At the time of death, tissue from the superior mediastinum was collected and studied by histology for the presence of any residual thymus. Mice with residual thymus were excluded from the study. Sham-tx mice underwent surgery on days 2–4 after birth without removal of the thymus.
Monitoring and autopsy
The mice were monitored for proteinuria, and tail blood was collected at monthly intervals. At 20 or 30 wk of age, the mice were killed, and a complete autopsy was performed. Organs (kidneys, salivary glands, lacrimal glands, thyroid, ovary, testes, and prostate) were collected in Bouins’ fixative and processed for histopathology. One kidney was snap frozen in liquid nitrogen and used for immunofluorescence staining and Ab elution studies.
Histopathology, GN severity index and organ specific autoimmunity. Four-micron sections were cut from Bouins-fixed, paraffin-embedded tissues and stained with H&E. An observer blinded to the experimental design evaluated histopathology. The severity of acute GN was graded based on the percentage of glomeruli involved, the severity of inflammatory cell infiltration, and proliferative changes in mesangial and/or peripheral regions. Glomeruli were measured at their maximum diameter using an ocular micrometer, and numbers of nuclei in each glomerulus were counted. The data for glomerular size and cellularity presented are the mean of readings from 10 glomeruli for each mouse. Chronic GN was scored based on glomerular fibrosis, interstitial fibrosis, and tubular atrophy. Each of the above-mentioned changes was scored from 0 to 4, with 0 indicating no pathology and 4 being maximum pathology. The severity index represents a cumulative score. Severity of organ-specific disease was scored based on the presence of inflammatory infiltrates in different tissues, with 0 as normal, 1–3 indicating increasing severity of infiltration, and 4 associated with destruction of normal architecture.
Estimation of anti-dsDNA IgG and total IgG by ELISA
Ninety-six-well plates were coated with streptavidin (1 mg/ml), followed by biotinylated plasmid DNA as previously described (22). After blocking with 3% BSA in PBS, the plates were incubated with sera at a 1/100 dilution. Bound Abs were detected with HRP-conjugated goat anti-mouse IgG, followed by o-phenylene diamine and hydrogen peroxide in citrate phosphate buffer, pH 5.0. The reaction was stopped with 2.5 N sulfuric acid, and the OD was read at 490 nM in an ELISA reader. An mAb to dsDNA (R4A; gift from Dr. B. Diamond, Albert Einstein College of Medicine, Bronx, NY) was used as the standard. IgG levels in sera and kidney eluates were estimated using a sandwich ELISA as previously described (21). Purified mouse IgG at different concentrations (4–500 ng/ml) was used to construct a standard curve.
Detection of renal immune complexes by direct immunofluorescence
Immune complexes in renal glomeruli were detected by direct immunofluorescence (21). Five-micron sections of snap-frozen kidney were fixed in ice-cold acetone for 10 min. The sections were rinsed with PBS and blocked with normal goat serum in 3% BSA/PBS (1/10, v/v) for 20 min. The sections were then incubated with FITC-conjugated goat anti-mouse IgG (Southern Biotechnology Associates; 1/50 dilution) for 45 min. After two rinses in PBS, the tissue sections were mounted with Vectashield mounting fluid (Vector Laboratories), and fluorescence intensity was graded from 0 to 4 in a blinded fashion.
Acid elution of Igs from kidney
Igs were eluted from the kidney as previously described (23). Briefly, each individual frozen kidney was weighed and homogenized in cold PBS on ice. The homogenate was centrifuged at 1200 x g for 10 min at 4°C. The pellets were washed by resuspension in cold PBS, followed by vortexing and centrifugation to remove serum proteins. The washing was repeated until the OD280 of the supernatant was <0.05. The pellet was then suspended in 0.1 M glycine with 1% BSA, pH 2.8 (10 ml/g kidney), and mixed gently at 4°C for 20 min. The suspension was centrifuged, and supernatant was immediately neutralized with 1 M Tris base.
Purification of CD25+ and CD25– T cells
Lymph nodes were harvested from NZM2328 mice and enriched for T cells by negative selection on T cell enrichment columns (R&D Systems). The cells were incubated with biotinylated 7D4 Ab to CD25 (BD Pharmingen). After washing, the cells were incubated with streptavidin-PE (Rockland), followed by anti-PE magnetic beads (Miltenyi Biotec). The cells were washed and then passed over a magnetic column using the double-positive selection mode in an autoMACS cell separation system (Miltenyi Biotec). The unlabeled CD25– cells were collected in the flow-through fraction. The double-positive selected CD25+ fraction of >90% purity was used for passive transfer experiments as well as in vitro assays.
Transfer of CD25+ T cells into d3tx mice
Seven- to 10-day-old d3tx mice were given a single injection of purified CD25+ T cells (106cells/mouse i.p.) Tail blood was collected from the recipient mice at 13, 15, 17, and 20 wk. The mice were killed at 20 wk, and kidneys were studied. A cohort of mice was killed at 30 wk and also studied for organ-specific disease.
Statistical analysis
Methods for statistical analysis include Student’s t test, Mann-Whitney U test, 2 analysis, and ANOVA using PRISM version 3.02 (GraphPad).
Results
D3tx accelerated dsDNA autoantibody production and renal glomerular immune complex deposition in male and female NZM2328 mice
D3tx NZM2328 male and female mice were studied for serum dsDNA Ab and serum IgG at 5, 9, 13, and 20 wk. Compared with sham-tx mice, the dsDNA Ab response of d3tx mice was greatly accelerated (Fig. 1). A significant increase in total serum IgG was also seen in d3tx mice (females, p = 0.04; males, p < 0.0001; data not shown). To control for hypergammaglobulinemia, the data are expressed as the dsDNA Ab titer per U of serum IgG in Fig. 1.
The d3tx male and female mice had comparable incidence of proliferative GN (87 vs 90%, respectively) of comparable severity (Fig. 4). However, in 43% of d3tx females, in addition to the proliferative GN described above, there was evidence of chronic GN, including glomerulosclerosis, epithelial cell crescents, interstitial inflammation, and atrophy of renal tubules (Figs. 3D, 4D, and 5C).
Therefore, d3tx greatly accelerated the development of lupus GN in NZM2328 mice. In male mice, the changes included exclusively an early-onset proliferative GN characterized by profound glomerular cell proliferation and accumulation of mesangial immune complexes. In female NZM2328 mice, the d3tx effect represents an accelerated natural disease, with progression of the glomerular disease process as a whole, leading to high prevalence of both proliferative GN and chronic GN.
D3tx induces or enhances extrarenal autoimmune disease in NZM2328 mice
In addition to GN, d3tx NZM2328 mice developed autoimmune disease that affected other organs (Fig. 6). Prostatitis was detected in 73% of d3tx males. Thyroiditis of greater severity and Ab to thyroid Ags of high frequency were detected in d3tx mice, with greater severity in females (data not shown). Similarly, inflammation of the salivary glands (sialoadenitis) and lacrimal glands (dacryoadenitis), detectable in many sham tx mice, was significantly more frequent and more severe in d3tx mice (Fig. 6). The tissue inflammation, first detected at 20 wk, reached high incidence at 30 wk when disease incidence was recorded (Fig. 7).
CD25+ T cells from young adult NZM2328 donors suppressed d3tx-induced autoimmune disease of prostate, thyroid, and lacrimal glands, but not proliferative GN or sialoadenitis
To determine the capacity of CD25+ T cells to suppress autoimmune disease in d3tx NZM2328 mice, CD25+ T cells were obtained from 6- to 8-wk-old untreated NZM2328 donors that exhibited no detectable renal disease or serum dsDNA Abs. The cells were transferred into 7- to 10-day-old d3tx NZM2328 recipients, and each recipient’s renal pathology was evaluated at 20 wk. A second cohort was killed at 30 wk to study the pathology of organs other than the kidneys.
Transfer of 106 CD25+ T cells to d3tx NZM2328 mice completely suppressed prostatitis and significantly reduced the severity of inflammation in thyroid and lacrimal gland (Fig. 7). In contrast, the transfer of CD25+ T cells had no observable effect on the proliferative GN at 20 wk (Fig. 8). In both males and females, the extent of proliferative GN in d3tx mice that received CD25+ T cells was not reduced compared with that in d3tx mice. The disease was also comparable to that in another group of d3tx mice given purified CD25– T cells (106/mouse) at the same time as CD25+ T cells (data not shown). Similar to GN, sialoadenitis in d3tx mice was not affected by CD25+ T cell transfer (Fig. 7). Therefore, CD25+ T cells from NZM2328 mice, which strongly prevented the development of prostatitis, thyroiditis, and dacryoadenitis in d3tx mice, had little or no influence on the accelerated development of lupus GN and sialoadenitis in NZM2328 mice.
Progression from acute to chronic GN in NZM2328 mice depends on gender
To determine the fate of proliferative GN in d3tx mice, we studied a cohort of d3tx mice at 30 wk and compared their data to those obtained at 20 wk. A dramatic difference was seen in the disease progression between male and female mice (Fig. 10A). Many d3tx or sham-tx female NZM2328 mice developed severe chronic GN by 30 wk. In contrast, only rare d3tx NZM2328 males progressed to chronic GN; indeed, most male d3tx mice were completely free of renal histopathology at 30 wk. Interestingly, despite the absence of renal pathology, d3tx male NZM2328 mice continued to have high levels of serum dsDNA Abs (data not shown), and the renal glomeruli continued to harbor large quantities of IgG immune complexes (Fig. 10B) with C3 deposition (data not shown). Thus, the progression of lupus GN evident in d3tx NZM2328 mice was strongly influenced by gender, but this was independent of the dsDNA Ab response.
Discussion
Mice tx between days 3–5 of life develop organ-specific autoimmune disease. The organ susceptibility (stomach, thyroid, ovary, prostate, and lacrimal gland) is mouse strain dependent (BALB/c, C3H, and A/J, respectively) (24). Evidence for the mechanisms of d3tx disease induction suggest three major factors: 1) a differential ontogeny of autoreactive T effectors and inhibitory T regulators (19), 2) depletion of thymus-derived CD25+ Treg cells, and 3) d3tx-induced lymphopenia and the resultant homeostatic proliferation (17). In this study we have developed an experimental model of accelerated lupus GN in lupus-prone NZM2328 mice by d3tx. With this model we have investigated the in vivo function of thymus-derived CD4+CD25+ T cells in the regulation of spontaneous systemic autoimmune disease.
The systemic autoimmunity in NZM2328 mice is not associated with a global deficiency in functional CD25+ Treg cells. CD25+ T cells purified from 2- and 5-mo-old NZM2328 females were found to efficiently prevent the proliferation of Con A-stimulated CD25– T cells (data not shown). Coculture of purified CD25+ from 2- and 5-mo-old NZM2328 females with CD25– T cells in a 1:1 ratio was able to inhibit proliferation of CD25– T cells by 96 and 95%, respectively. Moreover, the CD25+ Treg cells of NZM2328 mice were able to suppress the development of organ-specific autoimmune disease of the prostate, thyroid, and lacrimal glands, diseases that also developed in d3tx NZM2328 mice, but were unrelated to the lupus manifestation. In contrast, infusion of CD25+ T cells did not influence the development of lupus GN. Detection of global Treg abnormality is the focus of most clinical research on Treg cells in human autoimmune diseases.
Given the general integrity of Treg in NZM2328 mice, the finding that they failed to suppress the d3tx-accelerated GN is of considerable interest. This result is consistent with the hypothesis that abnormal immunoregulation in lupus-prone NZM2328 mice is due to a selective deficiency of Treg cells that control the development of autoimmune GN and sialoadenitis in these animals. Defective Ag-specific Treg cells may thus contribute to the pathogenesis of lupus GN.
There is clear evidence for Ag-specific regulation of other autoimmune diseases. In d3tx mice and rats, autoimmune prostatitis and autoimmune thyroiditis suppression were more efficient if the CD4+ T cells were derived from donors that harbored the target Ag (25, 26). Induction of insulin-dependent diabetes mellitus by diabetogenic T cells in lymphopenic hosts could be prevented by cotransfer of large numbers (10–20 million) of Treg cells from naive NOD compared with only 2 million islet Ag-specific Treg cells when cotransferred with diabetogenic T cells (reviewed in Ref. 27). Recent studies show that exposure to ovarian or prostate Ags in vivo resulted in enhanced regulatory capacity of CD25+ T cells in prevention of oophoritis or prostatitis, respectively (E. S. Samy and K. S. K. Tung, unpublished observations; Y. Y. Seitady, K. Ohno, and K. S. K. Tung, unpublished observations). Thus, efficient regulation of autoimmune disease by CD25+ regulatory T cells is determined by Ag-specific suppression.
D3tx in NZM2328 mice has offered the opportunity to compare the same set of polyclonal Treg cells in the suppression of different autoimmune diseases that occur in the same individual. We currently cannot rule out the possibility that failure in suppression of GN and sialoadenitis may be related to the cell dose used in suppression; namely, if more cells were transferred to the d3tx recipients at different ages, perhaps disease suppression would be detectable. Nevertheless, our finding is consistent with the conclusion that with a single experimental protocol, the CD25+ T cells from lupus-prone mice exhibited a differential threshold of suppression for organ-specific autoimmune diseases vs lupus GN and sialoadenitis.
SLE is a complex systemic autoimmune disease, and the pathogenesis of tissue injury is not completely understood. For the pathogenesis of lupus GN, the emphasis has continued to focus on the role of autoantibodies and the attendant glomerular immune complexes that invariably accompany lupus GN, although recent studies have begun to address T cell-mediated immunity independent of Th cells for autoantibody response (28). In d3tx NZM2328 male mice, accelerated anti-dsDNA Ab and immune complex deposition was not accompanied by accelerated end-stage kidney disease. Secondly, infusion of CD25+ Treg cells in d3tx mice had little or no effect on accelerated lupus GN, but strongly suppressed the early spontaneous dsDNA autoantibody response in NZM2328 mice. These findings point to the dissociation between autoantibody production and target organ injury in NZM2328 mice and raise questions about the pathogenic requirement of dsDNA autoantibody in lupus GN. This is supported by genetic studies in NZM2328 mice (12, 29).
In the present study, additional experimental evidence was obtained that dissociated acute from chronic lupus GN. At 20 wk, d3tx results in an accelerated onset of acute GN in males normally resistant to GN. Significantly, in contrast to d3tx females, this acute GN does not progress to chronic GN. Thus, acute GN represents a checkpoint between inflammation and disease or loss of function that is influenced by gender. The lack of progression to chronic GN in d3tx male NZM2328 mice is similar to the progression from peri-insulitis to insulitis and diabetes mellitus in NOD mice (30) and is relevant to the human disease. Indeed, a report by Pollock and Pirani (31) identified a significant number of lupus patients with a histopathological diagnosis of mild glomerulitis and failed to progress over an 8-year follow-up period.
The findings in d3tx NZM2328 mice have raised important questions fundamental to the pathogenetic mechanism of lupus GN and sialoadenitis and suggest a future direction for the investigation of experimental and human lupus.
Acknowledgments
We are grateful to Dr. Shu Man Fu (University of Virginia, Charlottesville, VA PI:P50AR45222) for support and critical comments, and to Sharon Mangawang for expert technical assistance.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by National Institutes of Health Grants R01AI41236, AR45222, and K01DK063065.
2 Address correspondence and reprint requests to Dr. Harini Bagavant, Division of Rheumatology, Department of Internal Medicine, Health Sciences Center, Box 800412, University of Virginia, Charlottesville, VA 22908. E-mail address: hb5u@virginia.edu
3 Abbreviations used in this paper: Treg cell, CD25+ regulatory T cell; d3tx, tx on day 3 of life; GN, glomerulonephritis; NZM, New Zealand Mixed; tx, thymectomy; SLE, systemic lupus erythematosus; tx, thymectomy, thymectomized.
Received for publication March 18, 2005. Accepted for publication May 13, 2005.
References
Piccirillo, C. A., E. M. Shevach. 2004. Naturally-occurring CD4+CD25+ immunoregulatory T cells: central players in the arena of peripheral tolerance. Semin. Immunol. 16: 81-88.
Sakaguchi, S.. 2004. Naturally arising CD4+ regulatory T cells for immunologic self-tolerance and negative control of immune responses. Annu. Rev. Immunol. 22: 531-562.
Bassuny, W. M., K. Ihara, Y. Sasaki, R. Kuromaru, H. Kohno, N. Matsuura, T. Hara. 2003. A functional polymorphism in the promoter/enhancer region of the FOXP3/scurfin gene associated with type 1 diabetes. Immunogenetics 55: 149-156.(Harini Bagavant2 and Kenn)