Interleukin-2 Protects Lupus-prone Mice from Multiple End-organ Damage by Limiting CD4⁻CD8⁻ Interleukin-17-producing T-cells

Mice

Female MRL/MpJ (stock no 000486), MRL/MpJ-Tnfrsf6lpr (MRL/lpr, stock no 000485) mice were purchased from the Jackson Laboratories. MRL/lpr-Foxp3^gfp mice were generated by backcrossing the Foxp3^gfp reporter transgene cassette (C57BL/6) onto the MRL/lpr genetic background for more than 10 generations. Mice were maintained in the specific pyrogen free animal facility of Beth Israel Deaconess Medical Center. All animal procedures in the study were approved by the Institutional Animal Care and Use Committee (IACUC) of Beth Israel Deaconess Medical Center.

Recombinant-AAV vaccination

Full-length cDNA encoding murine IL-2 subcloned into an AAV-Tet-on vector plasmid as previously described(33). Briefly, The AAV-Tet-on vector contains a bidirectional promoter with a tetracycline response element flanked by mini-CMV promoters that regulate expression of IL-2 and reverse transcriptional transactivator gene. MRL/lpr and/or MRL/lpr-Foxp3^gfp mice were vaccinated with 5 × 10¹⁰ viral particles (v.p.) of rAAV-Tet-on-IL-2 in contralateral hind limb muscles using an insulin syringe. Twenty-four hours after rAAV injection, mice were fed ad libitum chow containing 200 mg/kg doxycycline (BioServ) for induction of IL-2 or control diet. Blood samples were collected from tails and urine samples were collected using individual metabolic cages periodically.

Flow Cytometry

Splenocytes were isolated as previously described(23). PBS-perfused kidneys and lungs were minced and digested with collagenase IV (100 μg/ml) for 30 minutes at 37 °C to prepare single cells.

Isolated cells were stained with the following antibodies specific for: TCRβ (H57-597, BioLegend), CD3ε (145-2C11, eBiosciences), CD4(GK1.5, BioLegend), CD8α (53-6.7, eBiosciences), B220 (RA3-6B2, BioLegend), CD25(PC61, BioLegend), I-A/I-E (M5/114.15.2, BioLegend), Nkp46 (29A1.4, BioLegend) for 30min at 4 °C. For intracellular staining, cells were stimulated with 20 ng/ml of phorbol-myristate acetate (PMA, Sigma) and 1 mg/ml of ionomycin(Sigma) for 4 hours, washed and stained with TCRβ, CD4, CD8, I-A/I-E. BD cytofix/cytoperm plus with Golgi stop staining kits (BD Biosciences) were used according to the manufacturer’s protocol. Antibodies of IFN-γ (XMG1.2, BioLegend) and IL-17A (TC11-18H10.1, BioLegend) were used for detecting intracellular cytokines.

Histological Analyses and Immunofluorescence

Kidneys, lungs and skin biopsies were fixed in 10% formaldehyde. Tissue sections were stained with H&E or Periodic acid-Schiff (PAS) and visualized by Nikon. Histological scores of skins and kidneys were evaluated as described(34, 35). Kidneys and skins were scored from 6 mice in each group. At least 20 glomeruli and 10 tubular and vessel area in low power field are scored and calculated an average in each tissue section. Briefly, severity of glomeluronephritis, interstitial nephritis and perivascular infiltrates were graded in a semiquantitative (0–3+) manner. For skin, the degree of acanthosis, none (0) to marked thickened dermis (27); hyperkeratosis, none (0) to markedly increased keratin (2); inflammation, sparse (0) to heavy lymphocytic infiltrates (27); fibrosis, dermal collagen with normal (0) to markedly thickened (2); vessels, normal (0) to diffuse dilated (2); ulcer, absence (0) ore presence (1). For immunofluorescence, frozen kidney sections (5 μm thickness) were fixed in acetone and blocked with 1% normal goat serum in PBS. Diluted FITC-anti-mouse IgG (Jackson Immunoresearch) and FITC-anti-mouse C3(36) antibodies were incubated overnight at 4 °C, and staining visualized with a Nikon C1 confocal microscope.

Quantitative real-time PCR analysis

Total mRNA was purified from muscle, lung and kidney using TRIZol (Invitrogen). Messenger RNA from splenocyte was purified using RNeasy Mini Kit (QIAGEN). Complementary DNA (cDNA) was synthesized using cDNA EcoDry Premix (Clontech) for PCR amplification. Taqman primers were purchased from Applied Biosystems for Gapdh (Mm99999915_g1), Il17a (Mm00439619_m1), IFN-γ (Mm01168134_m1), TNFα (Mm00443260_g1), MCP1 (mm00441242_m1), CCL20 (Mm001268754_m1), CXCL11(Mm00444662_m1) and IL8(Mm04208136_m1). Gene expression was assessed by comparative CT method.

T Cell Labeling and Transfer experiment

CD3⁺ T cells were isolated from spleen and lymph nodes of MRL/lpr mice at 12 weeks of age using MACS pan T cell isolation kit (Miltenyi Biotec), then labeled with CellTrace Violet (Life Technologies) or carboxyfluorescein succinimidyl ester (CFSE) according to the manufacture’s protocol. In brief, cells were washed and resuspended to 10⁷ cells/ml in pre-warmed PBS containing 5 μM CellTrace Violet or 5μM CFSE and incubated for 20 minutes at 37 °C, then washed three times with RPMI containing 10% FCS and resuspended with PBS. Aliquots of 1.5 × 10⁸ stained cells were injected i.v. into MRL/MpJ mice at 12 weeks of age or MRL/lpr mice at the age of 14 weeks. For IL-2 treatment, mice were injected with 4 μg/body of mouse rIL-2 (BioLegend) twice a day for 5 days. Fluorescence-positive cells in spleens were counted and analyzed by flow cytometry. For in vitro IL-2 treatment, Isolated DN T cells were labeled with 1 μM CFSE and 5 × 10⁶ cells were incubated with or without 5 × 10⁶ of spleen cells for 5 days in the presence of indicated concentrations of IL-2.

Treatment with IL-2/anti-IL-2 monoclonal antibody complex

1.5 μg of mouse rIL-2 and 15 μg of anti-IL-2 monoclonal antibody clone S4B6-1 and the same amount of IL-2 and JES6-1A12 (BioXcell) were mixed and incubated at room temperature for 30 minutes, respectively. MRL/lpr mice at the age of 14 weeks were received i.p. injection of 15 μg of isotype control rat IgG, IL-2/S4B6-1 and IL-2/JES6-1 complexes every other day for 2 weeks.

ELISA

IL-2, IFN-γ and IL-17 were measured in mice sera using Ready-set-go ELISA system (eBiosciences). Anti-dsDNA antibody was detected using mouse anti-ds DNA IgG ELISA kit (Alpha Diagnostic International).

Statistical analyses

Unpaired two-tailed Student t test and Mann Whitney tests were used for analyzing the differences between two groups. For time-series experiments, paired t-test or two-way ANOVA with Bonferoni’s post-test were used. Differences between three data sets were analyzed by one-way ANOVA. A p value of < 0.05 was considered significant. Statistical analyses were performed in GraphPad Prism 5.01 software.

IL-17-producing DN T cells increase with advancing age in MRL/lpr mice

We previously reported that CD3⁺CD4⁻CD8⁻ DN T cells from patients with SLE and lupus-prone mice produce IL-17. Similarly, a significant increase in the percentage of DN versus CD4⁺ T cells that have the potential to produce IL-17 (IL-17-DN T cells) was detected in 19 week-old MRL/lpr mice, whereas the frequency of IFN-γ⁺ DN T cells was reduced relative to IFN-γ⁺ CD4⁺ T cells (CD4 vs DN of IFN-γ-positive cell percentage; 14.8 ± 1.33 vs 1.51 ± 0.0338 %, p<0.0001, IL-17; 1.37 ± 0.134 vs 2.85 ± 0.320 %, p= 0.0028, Figure 1A). Furthermore, the total number of splenic IL-17-DN T cells rather than IL-17-CD4⁺ T cells increased markedly in an age-dependent manner (5.34 ± 0.52 × 10⁵ for 15 weeks and 20.1 ± 3.67 × 10⁵ for 19 weeks, p= 0.0434, Figure 1B). Interestingly, the number of IFN-γ-DN T cells also increased in aged mice (4.86 ± 0.05 × 10⁵ for 15 weeks and 11.6 ± 1.72 × 10⁵ for 19 weeks, p= 0.0455); in contrast the number of IFN-γ-CD4⁺ T cell remained constant during disease development. These results indicate that DN T cells represent the preferential source of IL-17 in aged MRL/lpr mice and could contribute to disease development.

Treatment of MRL/lpr mice with IL-2-rAAV increases Treg and decreases DN T cells but does not affect autoantibody production

IL-2 has a short half-life in serum, which typically necessitates frequent and/or continuous injections to mediate a therapeutic effect. On the other hand, long-term administration and doses of IL-2 too high induce adverse effects including systemic activation of T cells and other immune effectors, and vascular or capillary leak syndrome (VLS) leading to severe pulmonary edema and volume depletion. To overcome these obstacles a rAAV vector encoding inducible IL-2 was employed(33). MRL/lpr mice were injected intramuscularly (i.m.) with 5 × 10¹⁰ viral particle (v.p.) of rAAV1-Tet-IL-2 and fed doxycycline (Dox)-containing chow for 4 weeks. The expression of IL-2 mRNA in the muscle and IL-2 protein in the serum was highest on day 7 after infection and gradually decreased after 14 and 28 days (Figure 2A,B). Notably, IL-2 mRNA was 3-fold higher even on day 28 as compared with that of anti-CD3/CD28-activated splenocytes and more than 5000-fold higher than control (day 0), indicating elevated and sustained expression of IL-2 in the muscle. Serum IL-2 concentration was significantly increased on day 7 (334.4 ± 35.7 pg/ml, p= 0.0119) and day 14 (66.5 ± 5.39 pg/ml, p= 0.0225) and showed no significance on day 28 (56.5 ± 12.5 pg/ml p= 0.0508) but still tended to be increased, as compared with controls (15.6 ± 9.13 pg/ml) (Figure 2B).

To test whether IL-2-rAAV therapy affected disease development, peripheral lymphocyte populations and serum autoantibody production were examined in MRL/lpr mice injected i.m. with IL-2-rAAV at 8 weeks of age and fed Dox. Consistent with a previous report, DN T cells were significantly decreased in the peripheral blood by IL-2-rAAV (percentages of DN T cells in total T cells of control vs IL-2: day 14; 26.9 ± 2.948 vs 13.78 ± 0.917. p= 0.0046, day 21; 33.83 ± 7.062 vs 18.40 ± 1.699. p= 0.0463, day 28; 51.83 ± 7.248 vs 26.95 ± 3.160. p= 0.0174, Figure 2C). Moreover, an increased percentage of CD4⁺CD25⁺Foxp3⁺ Treg was observed 14 days post infection and sustained through day 21 (percentage of Treg cells in CD4⁺ T cells of control vs IL-2: day 7; 4.91± 0.193 vs 6.22± 0.027, day 14; 4.13 ± 0.403 vs 7.89 ± 0.403, p<0.0001, day 21; 4.14 ± 0.517 vs 6.79 ± 0.451, p<0.001, Figure 2D). However, the titer of anti-dsDNA, the representative autoantibody in MRL/lpr mice, increased gradually with age and was comparable between control- and IL-2-rAAV-treated mice (Figure 2E). These results clearly demonstrate that IL-2 reduces the DN T cell pool and expands Treg without affecting autoreactive antibody responses in MRL/lpr mice.

IL-2-rAAV administration into MRL/lpr mice with established disease reduces DN T cells and cytokine production while expanding Treg

In considering the translational potential of IL-2 therapy, 12 week-old MRL/lpr-Foxp3^gfp mice, a time point at which disease is ongoing were treated with IL-2-rAAV. The Foxp3-GFP reporter cassette facilitates identification of Treg in the periphery and various organs. Eight weeks after treatment IL-2-treated mice exhibited reduced spleen weights (control vs IL-2: 0.9433 ± 0.1094 g vs 0.6117 ± 0.0686 g, p=0.028)(Figure 3A). The percentage of DN T cells among total T cells and the absolute number in the spleen were significantly reduced (Figure 3C) (cell number of control vs IL-2: 1.30 ± 0.197 × 10⁸ vs 0.590 ± 0.108 × 10⁸, p=0.0098). Treg (GFP⁺) were significantly increased in IL-2-treated mice (control vs IL-2: 2.54 ± 0.144 × 10⁶ vs 4.32 ± 0.566 × 10⁶, p=0.021) (Figure 3D). The fact that CD4⁺ and CD8⁺ T cell numbers were comparable between the control- and IL-2-treated groups indicated that the splenomegaly in MRL/lpr mice was mainly due to expansion of DN T cells. We also found that serum concentrations of IFN-γ (control vs IL-2: 976.8 ± 423.6 pg/ml vs 147.2 ± 20.40 pg/ml, p= 0.0359) and IL-17 (control vs IL-2: 3106 ± 418.1 pg/ml vs 1595 ± 423.8 pg/ml, p= 0.0294) were significantly decreased in IL-2-rAAV-treated mice (Figure 3B). These results suggest that systemic inflammatory responses were reduced by IL-2-rAAV treatment, at least in part, by the reduction of DN T cells.

IL-2-rAAV suppresses skin and lung inflammation in MRL/lpr mice

The typical skin lesions of diseased MRL/lpr mice are characterized by dermal hyperplasia and marked inflammatory cell infiltrate. Alopecia on the back of the neck and partial shedding and scabbing of the ear which were observed in the control rAAV-treated mice were diminished significantly in IL-2-rAAV-treated mice (Figure 4A, left). As seen in Fig. 4A IL-2-rAAV significantly decreased skin injury (37) (control vs IL-2: acanthosis; 1.83 ± 0.477 vs 0.667 ± 0.333. p= 0.0364, hyperkeratosis; 1.17 ± 0.307 vs 0.500 ± 0.224. p= 0.0550, inflammation; 1.833 ± 0.477 vs 0.833 ± 0.307. p= 0.0543, fibrosis; 1.17 ± 0.307 vs 0.333 ± 0.211. p= 0.0247, vessel cellular infiltration; 1.00 ± 0.365 vs 0.167 ± 0.167. p= 0.0323, ulcer; 0.333 ± 0.211 vs 0.167 ± 0.167 p= 0.275, n=6, Figure 4A, right).

Lung inflammation is common in MRL/lpr mice and is manifested by peribronchiolar and perivascular mononuclear cell infiltration. Perivascular infiltration of mononuclear cells was readily seen in control but not IL-2-rAAV-treated mice (Figure 4B). CD45⁺ cells infiltrating the lungs were increased in control mice as compared to IL-2-rAAV treated mice. More than 80% of CD45⁺ leukocytes were TCRβ⁺ T cells, which in turn were substantially decreased in IL-2-rAAV treated mice (control vs IL-2: CD45⁺ 5.05 ± 0.97 × 10⁵ vs 2.68 ± 0.49 × 10⁵; p = 0.045, Total T 4.14 ± 0.82 × 10⁵ vs 1.86 ± 0.38 × 10⁵; p= 0.032, CD4⁺ 8.71 ± 2.76 × 10⁴ vs 3.52 ± 0.25 × 10⁴; p= 0.078, CD8⁺ 3.90 ± 1.35 × 10⁴ vs 1.60 ± 0.22 × 10⁴; p= 0.094, DN 2.53 ± 0.47 × 10⁵ vs 1.32 ± 0.45 × 10⁵; p= 0.046, Figure 4B, right). Because IL-17 is well-known inducer of inflammatory cytokines and chemokines in the lungs, expression of TNFα, MCP-1 and IL-8 in the lung tissues was assessed. Expression of all of these molecules were upregulated in control mice, but significantly reduced in the lungs of IL-2-treated animals (TNFα of control vs IL-2: p= 0.0806, MCP1; p= 0.0105, IL-8; p= 0.0434, Figure 4C), indicating that the inflammatory response in the lung is reduced by ectopic IL-2.

IL-2-rAAV treatment reduces kidney damage by reducing the inflammatory immune response

Early mortality in MRL/lpr mice is often a result of renal failure due to inflammation. As expected severe mononuclear cell infiltration into the perivascular, tubulointerstitial and periglomerular areas of the kidney were observed in control-rAAV-treated mice (Figure 5A). In contrast, cellular infiltration in these areas was significantly reduced in IL-2-rAAV-treated mice. Notably, hypercellularity and expansion of the mesangial area in the glomeruli were significantly reduced in IL-2-rAAV-treated mice (glomeruli score of control vs IL-2: 1.47 ± 0.358 vs 0.587 ± 0.0926; p= 0.0092, tubulointerstitum 1.37 ± 0.405 vs 0.531 ± 0.140; p= 0.0477, perivascular 1.37 ± 0.295 vs 0.565 ± 0.100; p= 0.0172, Supplementary Figure 1A, Figure 5B), while C3 and IgG deposition were observed irrespective of the treatment group (Supplementary Figure 1A). The decrease of proteinuria in IL-2-rAAV-treated mice (urine albumin/ creatinine ratio on day 55 of control vs IL-2: 15.1 ± 7.74 × 10⁵ vs 1.44 ± 0.39 × 10⁵ mg/g; p <0.05) clearly reflected the attenuation of glomerulitis, tubulointerstitial and perivascular inflammation (Figure 5C). CD4⁺ and DN T cells predominated among CD45⁺ cells in the kidneys, and all subsets of T cells including Treg were decreased by IL-2 treatment (CD45⁺ cells of control vs IL-2: 11.8 ± 1.11 × 10⁵ vs 5.12 ± 0.318 × 10⁵; p= 0.0034, Total T cells 8.87 ± 1.00 × 10⁵ vs 4.27 ± 0.30 × 10⁵; p= 0.0045, CD8 1.47 ± 0.258 × 10⁵ vs 0.565 ± 0.119 × 10⁵; p= 0.019, CD4 3.53 ± 0.367 × 10⁵ vs 1.72 ± 0.225 × 10⁵; p= 0.0028, DN 3.85 ± 1.00 × 10⁵ vs 1.84 ± 0.143 × 10⁵, p= 0.0468, Treg 2.74 ± 0.014 × 10⁴ vs 1.39 ± 0.02 × 10⁴; p= 0.0063, Figure 5D). Gene expression levels of IFN-γ, IL-17, chemokines and chemokine ligands including CXCL9, IL-8, and CCL20 that promote recruitment of inflammatory cells were also reduced in the whole kidney of IL-2-rAAV-treated mice (IFN-γ; p= 0.0418, IL-17; p= 0.0366, CXCL9; p= 0.0304, IL-8; p= 0.0157, CCL20; p=0.0016, Figure 5E). These results demonstrate that IL-2-treatment decreases kidney injury by reducing the local recruitment of inflammatory cells.

IL-2 selectively decreases the DN T cell population by affecting its viability

The above results prompted us to consider whether IL-2 treatment controled DN T cell viability or generation in diseased mice. Accordingly, Calcein-violet labeled T cells from MRL/lpr mice were transferred to MRL/MpJ mice and rIL-2 was administrated for 5 days. Labeled T cells were not detected in the kidneys or the lungs regardless of IL-2 treatment (data not shown), indicating that T cell migration was not influenced by IL-2 treatment. Yet, the number of Calcein -Violet⁺ DN, but not CD4⁺ and CD8⁺ T cells, was found decreased in the spleens of the recipient mice (Figure 6A). This observation suggests that IL-2 negatively affects survival of DN T cells.

To address whether IL-2 has an influence on DN T cell viability, we cultured spleen cells for 5 days in the presence of IL-2 and survived CD4⁺, CD8⁺ and DN T cells were counted. The number of survived CD4⁺ and CD8⁺ T cells were sustained or slightly increased in high concentrations of IL-2. However, survived DN T cell number was dramatically decreased in low concentration (1.25 ng/ml) of IL-2 and IL-2-mediated decrease of DN T cell viability was concentration-dependent (Figure 6B). Next we examined whether IL-2 directly affects DN T cell survival, CFSE-labeled DN T cells alone were cultured or co-cultured with splenocytes in the presence of IL-2. We found that the viability of DN T cells was increased in the presence of splenocytes as compared with in the absence of them (Figure 6C). When DN T cells were singly cultured with IL-2, the number of live cells did not change or rather increased slightly in higher concentrations of IL-2. However, when co-cultured with splenocytes, IL-2 significantly induced DN T cell death in a dose-dependent manner (Figure 6C). We also co-cultured DN T cells with CD4⁺ T cells or with CD8⁺ T cells in the presence of IL-2 and the viability of DN T cells was decreased when co-cultured with CD8⁺ T cells but not with CD4⁺ T cells (data not shown and supplementary Figure 2B). These results suggest that IL-2 can act on non-DN T cell that drives DN T cell death. We checked the expression of IL-2 receptors by each subset of T cells. Neither IL-2 receptor alpha chain (CD25) nor beta chain (CD122) was detectable on DN T cells, whereas CD122 was highly expressed on CD8⁺ T cells. Moreover, phosphorylation of STAT5 by IL-2 was significantly obsereved in NK cells and CD8⁺ T cells but not in non-CD8⁺ T cells (data not shown and Supplementary Figure 2A). These results indicate that IL-2 indirectly affects DN T cell viability, possibly through modulating other lymphocyte function.

IL-2 targeting to non-Treg cells decreases the number of DN T cells and leads to reduced lymphadenopathy

It is well known that administration of IL-2 both in humans and mice results in expansion of Treg. Treg suppress several functions of conventional T cells directly and/or indirectly. Activated T cells can be suppressed by molecules such as IL-10, TGFβ and IL-35, or CTLA-4, or can be killed directly by perforin-expressing Treg cells (38). IL-2 also contributes to the generation and proliferation of cytotoxic T lymphocytes (CTLs) and natural killer (10) cells. Functional impairment of cytotoxic lymphocytes in lupus has been reported and found to correlate with disease activity both in humans and mice (39, 40). Therefore, we examined the effect of IL-2/anti-IL-2-complexes that can be used to selectively target different immune effector subsets. We used IL-2 complexed with two different clones of anti-IL-2 monoclonal antibodies (mAbs), S4B6 and JES6-1. These two types of IL-2/mAb complexes have distinct biological activities. IL-2/S4B6-1 complex preferentially binds to memory CD8⁺ T cells and NK cells expressing high levels of CD122 and expands these cells in vivo. By contrast, IL-2/JES6-1 complex binds to CD25 and induces selective expansion of CD4⁺CD25⁺ Treg (41). Isolated T cells from MRL/lpr mice were labeled with CFSE, transferred to recipient MRL/lpr mice, which were then treated with IL-2/ anti-IL-2 mAb complexes every other day for 7 days. CD8⁺ T cells and CD4⁺Foxp3⁺ Treg were significantly increased in IL-2/S4B6 and IL-2/JES6-treated mice, respectively, compared to control rat IgG-treated mice (Figure 7A and Supplementary Figure 2B). Notably, the frequency (Supplementary Figure 2B) and number (cell number of rat IgG; 6.08 ± 0.39 × 10⁴, IL-2/JES6 6.32 ± 0.784 × 10⁴, IL-2/S4B6 3.00 ± 0.305 × 10⁴, p<0.05, Figure 7B) of DN T cells were significantly decreased only in IL-2/S4B6-treated mice, indicating that IL-2-delivered to cytotoxic lymphocytes affect the DN T cell population. Consistent with previous reports (25, 42), CFSE-labeled CD8⁺ T cells and CD4⁺Foxp3⁺ cells underwent significant proliferation upon transfer into IL-2/S4B6-treated and IL-2/JES6-treated mice, respectively (Supplementary Figure 2C). Notably, the frequency of proliferating DN T cells was significantly decreased in IL-2/S4B6-treated but not IL-2/JES6-treated mice (percentage of proliferating cells of rat IgG; 40.77 ± 1.78, IL-2/JES6 42.03 ± 0.584, IL-2/S4B6 31.4 ± 1.40, p<0.01, Figure 6C). These results suggest that the direct effects of IL-2 on cytotoxic lymphocytes results in inhibition of DN T cell expansion.

IL-2 delivery to cytotoxic lymphocytes reduces lymphadenopathy

DN T cells make up ~90% of the cells residing in the massive lymph nodes of diseased MRL/lpr mice. To determine if IL-2/S4B6 treatment reduced lymphadenopathy, 14 week-old MRL/lpr mice were treated with control rat-IgG, IL-2/JES6-1 or IL-2/S4B6-1, every other day for 3 weeks. No difference in the number of splenic DN T cells was seen in IL-2/JES6-1- and control rat-IgG-treated mice. On the other hand, CD8⁺ T cells and NK cells in spleens from IL-2/S4B6-treated mice were significantly increased and DN T cells decreased, although the total cell number was comparable in all groups (Figure 7C). However, size and weight of axillary and mesenteric lymph nodes were significantly reduced in IL-2/S4B6-treated mice as compared with control mice (lymph node weight of control vs IL-2/S4B6: axillary 0.290 ± 0.015 g vs 0.165 ± 0.035 g, p<0.05, mesenteric 1.003 ± 0.096 g vs 0.523 ± 0.048 g, p<0.05, Figure 7D). These results suggest that activated/expanded cytotoxic lymphocytes, but not Treg, lead to the reduction of DN T cells resulting in the amelioration of lymphadenopathy in MRL/lpr mice.