Renal Dendritic Cells Stimulate IL-10 Production and Attenuate Nephrotoxic Nephritis

Activated Conventional DCs Accumulate in Kidneys of Mice With NTN

We have recently demonstrated that kidney DCs accumulated and relocated to inflamed glomeruli in NTN.⁸ To further characterize these DCs, we analyzed kidney single-cell suspensions from healthy and nephritic mice by flow cytometry. On day 7 after injection of nephrotoxic sheep serum (NSS), histologic and functional signs of kidney damage were evident in a dose-dependent fashion.³ At this time point, DCs from nephritic kidneys were increased in numbers and showed the CD11c⁺ CD11b⁺ phenotype of conventional DCs (Figure 1, A and B) that DCs from healthy kidneys displayed as well (Figure 1A).^8,9 Costaining for the subtype marker CD8α showed its expression on some DCs in NTN (Figure 1A). The cells expressing high levels of CD8α and little or no CD11c (Figure 1A) lacked MHC II expression (data not shown), and thus represented CD8⁺ T cells. CD11c⁺ B220⁺ plasmacytoid DCs were not evident, whereas some CD11c⁻ B220⁺ B lymphocytes were detected (Figure 1A). DCs from nephritic kidneys showed a more mature phenotype than those in healthy organs, as evidenced by higher expression of MHC II and costimulatory molecules such as CD80 (Figure 1C). Nephritic kidneys contained an additional CD11c⁺ DCs subset devoid of surface MHC II expression (Figure 1A), which represented immature DC precursors recently recruited from the blood as previously reported.⁸ Thus, conventional DCs of the CD8α⁻ and the CD8α⁺ subtype accumulated in NTN and exhibited signs of activation.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Conventional DCs accumulate and mature in nephritic kidneys. (A) On day 7 after induction of NTN, DCs were isolated from the kidneys of nephritic mice (NTN) or non-nephritic controls (without NTN) and analyzed by flow cytometry for expression of DC subtype markers. Contour plots show representative expression profiles of the DC marker, CD11c, versus CD11b, CD8α, B220, and MHC II. (B) Gives statistical analysis of the numbers of DCs bearing these subtypes markers from the experiment shown in panel A (NTN, black bars; non-nephritic controls: white bars, n = 4). (C) MHC II and CD80 mean fluorescence intensities (MFI) detected on CD11c⁺ cells from the kidneys of nephritic (black bars) and non-nephritic (white bars) mice. These results are representative for 3 individual experiments. *P < 0.05; **P < 0.01.

Kidney DCs From Nephritic Mice Induce IFNγ and IL-10 Secretion

We speculated that MHC II⁺ renal DCs might be able to affect the course of NTN by interacting with infiltrating CD4⁺ T cells, which are crucial effectors in NTN.^1–3 To address this hypothesis, we isolated renal DCs from nephritic mice, allowed them to endocytose a model antigen, ovalbumin (OVA), and subsequently cocultured them with OVA-specific CD4 T cells (OT-II cells). After 3 d, OT-II cell numbers had increased over those in cultures without antigen, indicating antigen-specific proliferation (Figure 2A). Kidney DCs from nephritic mice stimulated OT-II cell proliferation more efficiently than DCs from healthy mice (Figure 2A), demonstrating functional consequences of DC activation in NTN (Figure 1C). As expected, splenic DCs used as controls for competent DCs elicited a stronger proliferative response (Figure 2A). This response was identical in nephritic and control mice (Figure 2A), demonstrating that CD4⁺ T cell activation in NTN was kidney-specific.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

DCs from nephritic kidneys activate CD4⁺ T cells and induce IFNγ and IL-10 production. DCs from the kidneys (left graphs) or spleens (right graphs) of nephritic mice (black bars) or non-nephritic controls (white bars) were isolated, cultured with 1 mg/ml OVA or without OVA (without OVA) for 2 h, washed, and were subsequently co-incubated with 1 × 10⁵ OT-II cells. After 3 d, the number of viable OT-II cells was determined by flow cytometry (A), and the concentrations of IFNγ (B) and IL-10 (C) in the culture supernatants by ELISA. Shown are results typical of 4 independent experiments in groups of 3 or 4 mice. n.s, not significant (P ≥ 0.05); *P < 0.05.

We next examined the cytokines produced in these cocultures. DCs from nephritic kidneys induced fivefold increased levels of the Th1 cytokine IFNγ, which nevertheless were 10-fold lower than those accomplished by splenic DCs (Figure 2B). Strikingly, kidney DCs from nephritic mice induced fivefold increased amounts of IL-10 (Figure 2C). Splenic DCs from healthy and nephritic mice produced little of this cytokine indicating disease- and organ-specific IL-10 induction. No cytokine production was detected when DCs or T cells were cultured separately or without antigen (Figure 2, B and C). The requirement of antigen-specific T cell stimulation for IL-10 production implied that DCs did not secrete noticeable amounts of this cytokine themselves. Thus, DCs induced not only a typical Th1 cytokine, but also the prototypic immunosuppressive cytokine, IL-10, which has been shown to suppress NTN.^{24–28,30–32}

Endogenous T Cells From Nephritic Mice Produce IL-10

To investigate whether endogenous T cells also produced IL-10 in the kidney, we isolated these cells on day 5 after NTN induction. Nephritic kidneys contained more CD4⁺ T cells than kidney from non-nephritic controls (Figure 3A), consistent with their recruitment in NTN.⁸ When we stimulated equal numbers of renal T cells with anti-CD3 antibody in vitro and determined cytokine concentrations in the supernatant after 2 d, T cells from nephritic mice showed pronounced IL-10 production, whereas those from healthy mice produced 60% less of this cytokine (Figure 3B). IFNγ was produced in similar amounts regardless of nephritis (Figure 3B). Without anti-CD3 stimulation, no cytokine release was observed, indicating production by T cells (Figure 3B). These findings demonstrate that kidney T cells in NTN had been programmed for high IL-10 production on a per cell basis, whereas their ability to produce IFNγ was not altered.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Endogenous CD4⁺ T cells produce IL-10 and IFNγ in NTN. (A) On day 5 after induction of NTN, kidneys from nephritic mice and non-nephritis controls were taken and CD4⁺ T cell numbers were determined. (B and C) 1 × 10⁵ CD4⁺ T cells described in 3A were cultured in anti-CD3-coated 96 well-plates. After 40 h, IFNγ (B) and IL-10 (C) concentrations were determined by ELISA. Shown are results typical for 3 individual experiments. wo aCD3, without anti-CD3 antibody stimulation. n.s, not significant (P ≥ 0.05); **P < 0.01; ***P < 0.001.

Establishing and Characterization of an In Vivo Depletion System for Kidney DCs

Theoretically, the above cytokine profile might lead to either Th1-mediated aggravation or IL-10-mediated attenuation of NTN. To distinguish these possibilities, we established an experimental system to study the course of NTN in the absence of kidney DCs. We used transgenic CD11c-DTR/GFP mice³⁴ expressing the human diphtheria toxin receptor in CD11c⁺ DCs and green fluorescence protein (GFP). Injection of diphtheria toxin (DT) into these animals causes systemic depletion of CD11c⁺ DCs, which can be monitored by analysis for the loss of cells expressing GFP. Without DT injection, these mice did not show any detectable immunologic alterations to wild-type C57/BL6 control mice.³⁴ In addition, we did not observe histologic or functional differences when NTN was induced in nondepleted CD11c-DTR/GFP compared with nontransgenic control mice (data not shown).

As we intended to target kidney DCs in the effector phase of NTN, we designed a protocol that avoided effects of DC depletion on the autologous immune response against sheep immunoglobin (Ig), while at the same time ensured equal glomerular deposition of sheep Ig in DC-depleted and nondepleted animals. This was achieved by injecting DT on day 4 after NSS injection; thus, the DC-dependent induction phase of the autologous response and after glomerular sheep Ig deposition (Experimental protocol in Figure 4A). This resulted in 80% loss of CD11c⁺ DCs (n = 4 mice/group, P < 0.05), when kidneys were analyzed on day 5 (Figure 4B). Depletion was specific for DCs because intrarenal macrophages and lymphocytes (NK, CD4⁺, and CD8⁺ T cells) were not significantly changed (Figure 4B). In the spleen, DCs were reduced by 75% and macrophages by 33% (Figure 4C), consistent with previous reports showing that in the spleen of CD11c-DTR/GFP mice only DC and marginal zone macrophages were depleted.^34–36

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Efficiency and specificity of DC depletion in the kidney. (A) Experimental protocol to determine specificity of DC depletion. Four days after induction of NTN in CD11c-DTR/GFP mice (black columns) or wild-type controls (white columns), all mice were injected intraperitoneally with 4 ng DT per g body weight. (B and C) On day 5, kidney (B) or spleen (C) single-cell suspensions were analyzed by flow cytometry for total numbers of CD11c⁺ DCs, CD11c⁻ CD11b⁺ macrophages, CD4⁺ and CD8⁺ T cells, and for NK cells per organ. (D) Protocol to determine the effect of DC depletion on NTN. Four and 10 days after induction of NTN in CD11c-DTR/GFP mice (black columns) or wild-type controls (white columns), mice received 4 ng DT per g body weight. (E) On day 14, kidney single-cell suspensions were analyzed by flow cytometry for CD11c⁺ cells, GFP⁺ cells, and for co-expression of CD11b and MHC II. Contour plots show representative expression profiles of these markers. (F and G) Statistical analysis of the total numbers of intrarenal CD11c⁺ DCs, CD11c⁺ DCs expressing transgenic GFP, MHC II⁺ DCs, and CD11c⁻ CD11b⁺ macrophages (F), and of CD4⁺ T cells, CD8⁺ T cells and CD19⁺ B cells (G) on day 14. (H) Numbers of splenic CD11c⁺ DCs, CD4⁺ T cells on day 14. These results are representative for 2 individual experiments using groups of 4 mice. n.s, not significant (P ≥ 0.05); *P < 0.05; **P < 0.01.

A single DT injection was unlikely to continuously deplete DCs throughout the experiment because DCs in lymphatic organs were replenished within 2 to 4 d.³⁴ However, a second DC depletion after this short time was lethal for CD11c-DTR/GFP mice, as reported by others.³⁷ Nevertheless, we found that a second depletion after 1 wk was tolerated and resulted in 85% reduction of kidney DCs on day 14 (Experimental protocol in Figure 4D, results in Figure 4, E and F). Numbers of DCs in the spleen were unaltered on day 14 (Figure 4H), as opposed to their efficient depletion on day 5 (Figure 4B). Thus, replenishment of DCs in the kidney occurred slower than in the spleen, resulting in prolonged absence of renal DC.

Numbers of renal macrophages on day 14 were unchanged (Figure 4, E, lower row, and F), again demonstrating that the depletion protocol specifically ablated renal DCs. Renal CD4⁺ T cells were 40% to 50% fewer, but this was not statistically significant (Figure 4G). Renal CD8⁺ T cells were significantly reduced by 60% (Figure 4G). However, reduced CD8⁺ T cell numbers are unlikely to aggravate murine NTN, as this model has been shown to be independent of these cells.³⁸ A slight nonsignificant change was seen also in numbers of renal B cells, which were examined as a control-lymphocyte population (Figure 4G). Numbers of splenic CD4⁺ (Figure 4H) and CD8⁺ T cells (data not shown) were unaltered, indicating that the tendency toward reduced renal T cell numbers was a secondary effect of DC depletion, consistent with reports showing that T cells are not directly targeted in CD11c-DTR/GFP mice.^34,36

The DC reduction was also apparent when the MHC II⁺ DCs were enumerated (Figure 4F), implying that less DCs remained available for interaction with infiltrating CD4⁺ T cells. This was confirmed by immunohistochemistry, which showed a marked reduction of MHC II⁺ cells in kidneys of DT-injected CD11c-DTR/GFP mice, whereas many such cells were visible in the tubulo-interstitium of DT-injected wild-type mice (Figure 5, A and B). In particular, the periglomerular infiltration evident in nephritic mice was lost after DC depletion (Figure 5, C and D). MHC II staining was relatively specific for kidney DCs because most renal MHC II⁺ cells expressed the DC marker CD11c (Figure 1A), which in the murine kidney is a specific DC marker.⁸

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Histologic images. Kidney sections from DT-injected CD11c-DTR/GFP mice (B, D, F, H) and controls (A, C, E, G) from the protocol shown in Figure 4D (using 4 mg sheep Ig/g body weight) were taken on day 11 and stained for MHC II (A-D) or by PAS (E-F). A, B, E, and F: original magnification ×100; C, D, G, and H: original magnification ×200. Shown are representative sections illustrating the quantitative analyses shown in Figure 7, A through D.

To verify our postulate that sparing DCs until day 4 ensured an unchanged systemic immune response against sheep Ig, we determined specific antibodies against sheep Ig in DC-depleted nephritic mice and controls on day 14. The titers of specific total Ig and IgG1, indicating Th2 responses, were identical, whereas specific IgG2a, indicative of Th1 responses, were somewhat diminished, albeit not significantly (Figure 6A). In addition, the cytokine production of splenic leukocytes in response to sheep Ig was similar (Figure 6B). Thus, our depletion protocol left the magnitude and the polarization of the splenic T cell response unchanged, implying that the systemic immune response against sheep Ig, and thus the autologous phase of NTN, had not been compromised, as intended.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

Systemic immunity against sheep Ig is not affected by DC depletion. (A) Serum samples from DT-injected CD11c-DTR/GFP mice (black bars) and controls (white bars) from the experimental protocol shown in Figure 4D were taken on day 14 and analyzed by ELISA for titers of total murine IgG, IgG1, and IgG2a levels specific for sheep IgG (mouse IgG/IgG1/IgG2a α sheep IgG). (B) 2 × 10⁵ splenocytes from the same mice were cocultured with 10 μg/ml sheep Ig. After 3 d, concentrations of IL-10 and TNFα and, in separate experiments, of IL-2 and IFNγ, in the culture supernatant were determined by ELISA. n.s, not significant.

Depletion of DCs in the Immune Effector Phase Aggravates NTN

We next examined the effect of kidney DC depletion on NTN by histology. In the nondepleted nephritic control group, glomeruli focally demonstrated segmental necrosis and extracapillary proliferation (Figure 5, E and G). The tubulo-interstitium showed focal tubular damage, as evidenced by tubuli with flattened epithelium and periodic acid-Schiff (PAS)-positive material in the tubular lumen (Figure 5, E and G). Some tubuli lacked luminal space and showed pronouncedly vacuolated cytoplasm of the epithelium. A sparse mononuclear cell infiltrate could be seen in the interstitium (Figure 5, E and G). DC depletion in DT-injected CD11c-DTR/GFP mice led to aggravated tubular damage as compared with DT-injected nontransgenic controls (Figures 5, F and H and 7A). In 2 of 5 experiments, increased extracapillary proliferation/necrosis was noted, whereas no significant differences were seen in the other three experiments (data not shown). No significant changes were seen in the scores for glomerular sclerosis and chronic tubulointerstitial damage on day 11 (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Figure 7.

Depletion of DCs aggravates kidney damage in NTN. (A-D) Groups of 6 to 8 CD11c-DTR/GFP mice (squares) or nontransgenic controls (triangles) that had received 2.5 mg nephritogenic sheep Ig/g body weight were injected with DT as shown in the protocol in Figure 4D. On day 10, mice were placed in metabolic cages and 24 h urine was collected for determining urinary protein excretion per creatinine (D and F). On the following day, kidneys were taken for histologic analysis of the acute tubulointerstitial damage (A) and the focal segmental glomerulosclerosis scores (B) and serum samples for calculating creatinine clearance (C). These results are representative for 4 of 5 experiments in groups of 3 to 4 mice. In the fifth experiment, differences were not statistically significant. (E and F) In 2 separate experiments depicted by open or closed symbols, mice received only 2.0 mg nephritogenic sheep Ig/g body weight, and histologic (E) and urine analysis (F) was performed on day 14. n.s, not significant (P ≥ 0.05); *P < 0.05; **P < 0.01.

Consistent with the more severe acute tubulointerstitial damage, creatinine clearance of DC-depleted mice was significantly reduced, indicating compromised kidney function (Figure 7). In addition, urinary protein concentration per creatinine was significantly higher in the depleted group (Figure 7B). These findings indicated that kidney DCs attenuated NTN.

The analysis above was performed on day 11, the day after the second depletion, because DC-depleted mice suffered acute kidney failure in these experiments and had to be killed. To exclude that the depletion maneuver as such caused changes similar to acute tubulointerstitial damage, we induced NTN with a 20% lower serum dose, which allowed analyzing the mice 4 d after the second depletion, on day 14. In 2 of 2 experiments, aggravated acute tubulointerstitial damage was seen (Figure 7E), whereas scores for extracapillary proliferation/necrosis, glomerular sclerosis, and chronic tubulointerstitial damage again showed no significant changes (data not shown). In one of the two experiments, proteinuria per creatinine was determined, showing in 3 of 4 DC-depleted mice higher values than in the 3 nondepleted control (Figure 7F), albeit not to a statistically significant extent.

Immunological Mechanisms With Protective Potential in NTN

To elucidate the mechanisms underlying DC-mediated protection against NTN, we enumerated nT_regs, which have previously been reported to attenuate NTN by effects mediated in the secondary lymphatics.²³ Indeed, CD25⁺ FoxP3⁺ cells in the largest lymphatic organ, the spleen, were more than 50% less frequent, albeit not statistically significant (Figure 8A). In the kidney, we found very few CD25⁺ FoxP3⁺ cells (Figure 8B), consistent with the previous study.²³ These numbers were unchanged after DC depletion (Figure 8B), arguing against an effect of DC depletion on intrarenal nT_reg.

• Download figure
• Open in new tab
• Download powerpoint

Figure 8.

Immunological mechanisms with protective potential in NTN. (A and B) DT was injected intraperitoneally on day 4 after induction of NTN in groups of 3 to 4 CD11c-DTR/GFP mice (black columns) or wild-type controls (white columns). On day 5, spleen (A) or kidney (B) single-cell suspensions were analyzed by flow cytometry for total numbers of CD4⁺ CD25⁺ and CD4⁺ CD25⁺ FoxP3⁺ T_reg. (C and D) CD11c⁺ DCs from the spleens (C) or kidneys (D) of day 5 nephritic mice (dashed line) or non-nephritic controls (gray areas, “without NTN”) were analyzed by flow cytometry for ICOS-L expression. n.s, not significant (P ≥ 0.05).

Finally, we speculated on mechanisms by which DCs might have induced IL-10 in CD4⁺ T cells. Previous work showed that DCs could do so by producing IL-10 themselves¹⁶ or by expressing inducible costimulatory molecule ligand (ICOS-L).^39,40 Our finding that CD4⁺ T cells cocultured with kidney DCs from nephritic mice did not secrete IL-10 unless stimulated with antigen (Figure 2B) implied that the DCs did not secrete this cytokine themselves, at least not in measurable amounts. In contrast, kidney DCs expressed about threefold more ICOS-L on their surface than splenic DCs (mean fluorescence intensities: spleen NTN, 1387; spleen without NTN, 1308; kidney NTN, 4451; kidney without NTN, 3675). The 20% increase of expression on kidney, but not splenic, DCs of nephritic mice (Figure 8, C and D) was consistent with the possibility that IL-10 induction may have occurred via ICOS-L.

Mice, Reagents, Nephrotoxic Nephritis Model

Female 6- to 8-wk-old C57/BL6 and CD11c-DTR/GFP mice³⁴ were bred and kept in specific-pathogen-free condition at the animal facility of University of Bonn. NSS was generated as described^4,52 by immunizing sheep with homogenized murine renal cortex in CFA, followed by monthly boosting doses in IFA. The amount of NSS per g body weight used in this study (2.0 or 2.5 mg Ig per g body weight) was within the linear range of a dose response.³ All animal studies have been approved by Institutional and Government Review Boards. All reagents were from Sigma-Aldrich (Steinheim, Germany), if not specified otherwise.

Histology

Light microscopy was performed on 3-μm paraffin sections of PFA-fixed tissue stained by periodic acid-Schiff, and Goldner Trichrome. Kidney damage was histologically scored as described previously.^53,54 In brief, extracapillary proliferation/necrosis and focal segmental glomerulosclerosis were defined as follows: 0, no; 0.5, <25%; 1, 26% to 50%; 2, 51% to 75%; 3, >75% of a glomerulus showing extracapillary proliferation/necrosis and sclerosis. Injury was evaluated in at least 50 glomeruli per sections. Acute tubulointerstitial damage was defined as follows: 0, no tubular damage, no interstitial edema; 0.5, thinning of the brush border; 1, thinning of the tubular epithelia; 2, denudation of the tubular basement membrane; 3, tubular necrosis. Chronic tubulointerstitial damage was defined as broadening of the basement membrane of tubuli with flattened epithelium, and interstitial matrix increase and/or fibrosis, and was evaluated as follows: 0.5, focal chronic damage; 1, diffuse chronic damage. Tubulointerstitial damage was judged in whole kidney sections including cortex and outer stripe of outer medulla. Tubulointerstitial damage scores were calculated as described.^53,54

Immunohistology for MHC II⁺ cells was performed by staining acetone-fixed cryosections of kidney tissue with biotinylated anti-murine IA^b (clone AF6–120.1) antibody (BD-Pharmingen, Heidelberg, Germany). Antibody binding was revealed with streptavidin-peroxidase and 3,3′-diaminobenzidine, followed by methyl green counterstaining.

Isolation of Murine DCs and CD4⁺ T Cells

Kidneys and spleens were digested with collagenase (Roche Diagnostic, Mannheim, Germany) and DNAse-I as described.⁸ Tubular fragments from digested kidneys were removed by sedimentation. CD11c⁺ DCs and CD4⁺ T cells were then enriched using nanobead-labeled specific monoclonal antibodies (clones N418 and GK1.5, respectively) (Miltenyi, Bergisch-Gladbach, Germany). Magnetic bead separation was done according to the manufacturer's instructions. Purity was usually 90% to 95% for DCs and >95% for CD4⁺ T cells. OT-II cells were isolated from spleen and lymph nodes of OT-II mice using nanobead-based kits. Purity was usually about 85%.

Flow Cytometry

After treatment with FcBlock (BD-Pharmingen), cells were stained for 15 min on ice with fluorochrome-labeled monoclonal antibodies against CD4 (clone GK1.5), CD8α (53 to 6.7), B220 (RA36B2), NK1.1, CD11b (M1/70), CD11c (HL3), I-a^b (AF6–120.1) (BD-Pharmingen), CD25 (PC61), FoxP3 (FJK-16s), and ICOS-L (IIK5.3) (Ebiosciences, Kranenburg, Germany). Dead cells were excluded with Hoechst 33342 or 7-AAD and analyzed with a FACScalibur (Becton Dickinson, CA).

In Vitro Cytokine Production

For determining cytokine induction by DC, 5 × 10⁴ DCs were cocultured with 1 mg/ml OVA in 100 μl DMEM (10% FCS) in 96-well plates for 2 h, then washed and coincubated with 1 × 10⁵ OT-II cells for 3 d. For determining cytokine production by renal T cells, 1 × 10⁵ renal CD4⁺ T cells were cultured for 42 h in 96-well plates coated with 10 μg/ml anti CD3 (clone 2c-11). For determining cytokine production in response to sheep-Ig, 2 × 10⁵ cells splenic cells were cultured for 72 h with 10 μg/ml sheep globulin. Cytokines in the supernatants were measured using commercial enzyme-linked immunosorbent assay (ELISA) kits (R&D Duosets, Wiesbaden-Nordenstadt, Germany).

Miscellaneous Assays

Serum levels of murine Ig specific for sheep Ig levels were determined by ELISA,³ proteinuria by Bradford assay (BioRad, Munich, Germany), and creatinine as described.⁸ For creatinine clearance, urine was collected over 16 to 24 h to obtain daily creatinine excretion.

Statistics

Results are expressed as mean ± SD. Comparisons were drawn using a two-tailed t test (Prism 4, Graphpad Software, San Diego, CA) except for histologic analysis, where the two-sided nonparametric Mann-Whitney U test was used.