|Home | About | Journals | Submit | Contact Us | Français|
Ischemia-reperfusion is a major cause of acute kidney injury and inflammation has been well-established as a detrimental process in the pathogenesis of kidney ischemia reperfusion injury (IRI). The kidney has the ability to be preconditioned by a non-lethal period of ischemia, rendering the kidney refractory to further ischemia-induced dysfunction (ischemic preconditioning: IPC). Regulatory T (Treg) cells are lymphocytes that suppress immune responses. We hypothesized that IPC is partially mediated by Treg cells. To test this, a model of delayed IPC was used by subjecting mice to 24 min of bilateral renal IRI or sham surgery on day 0, then 28 min IRI on day 7 (IPC = IRI/IRI; non-IPC = Sham/IRI). IPC significantly inhibited the accumulation of neutrophils and macrophages, tubular necrosis and loss of kidney function induced by IRI. The initial 24 min IRI caused a significant increase in kidney CD4+CD25+FoxP3+ and CD4+CD25+IL-10+ Treg cells at 7 days of reperfusion. Use of a Treg cell-depleting antibody (PC61) in preconditioned mice reversed the effect of IPC on kidney neutrophil accumulation and partially inhibited the functional and histological protection of IPC. Adoptive transfer of Treg cells, prior to IR, in naïve mice, mimicked the protective and anti-inflammatory effects of IPC on the kidney. These results demonstrate that suppression of inflammation and a significant fraction of kidney protection, imparted by delayed IPC, is mediated by Treg cells.
Acute kidney injury (AKI) is associated with high morbidity and mortality and can predispose individuals to end stage renal disease (1). Ischemia-reperfusion injury (IRI) is a major cause of AKI. While the pathogenesis of IRI is multi-factorial, inflammation in the post-ischemic kidney has been well-established as a detrimental process in the pathogenesis of kidney IRI (2–7). The innate immune response to IRI involves the activation and accumulation of neutrophils and macrophages in the post-ischemic kidney, and depletion of neutrophils or macrophages in mouse models is sufficient to preserve kidney function after IRI (5–10).
Similar to the heart, the kidney has the ability to be preconditioned by a non-lethal period of ischemia, rendering the kidney refractory to further ischemia-induced dysfunction (ischemic preconditioning; IPC (11–14)). Kidney IPC leads to increased renal pro-survival signals: HSP 27 expression (14) and Akt phosphorylation (13), and decreased phosphorylation of cell death-promoting p38 and Jun kinase (14). Additionally, adenosine receptor activation or inhibition mimics or prevents the protective effect of IPC, respectively (12, 13). Important studies recently demonstrated that adoptive transfer of kidney lymphocytes from preconditioned mice to non-preconditioned recipients confers protection from kidney IRI (11). These findings suggest that accumulation of lymphocytes with protective properties in the post-ischemic kidney may mediate IPC.
Regulatory T (Treg) cells are anti-inflammatory lymphocytes that have recently been identified in normal mouse kidneys (15–17). Expression of the IL-2 receptor (CD25) and the transcription factor FoxP3 identify a highly suppressive subset of CD4+ Treg cells. These lymphocytes use multiple mechanisms to inhibit the function and proliferation of pro-inflammatory leukocytes, including production of IL-10 and/or TGF-β, cell contact-dependent inhibition or generation of extracellular adenosine (18–22). Another important property of Treg cells is their propensity to home to areas of ongoing inflammation (23, 24).
We hypothesized that IR would cause Treg cell accumulation in the kidney as a consequence of the associated inflammation, which in turn would protect the kidney from subsequent IR-induced inflammation and kidney injury.
C57Bl/6 mice underwent 24 min of bilateral renal ischemia (preconditioning) or sham surgery (non-preconditioned controls) on day 0 and then were allowed to recover for 7 days. On day 7, both groups of mice underwent 28 min bilateral renal ischemia. Twenty-four hr after the initial surgery, preconditioned mice exhibited a slight, but significant, decrease in renal function compared to non-preconditioned control mice, as measured by plasma creatinine (PCr) levels (Figure 1a). By 72 hr of reperfusion no differences were observed in PCr levels between groups (Figure 1a). Twenty four hr after 28 min IRI, non-preconditioned mice (Sham/IRI) had severely decreased renal function, whereas preconditioned mice (IRI/IRI) were completely protected (Figure 1a). The increase in outer medulla acute tubular necrosis (ATN) observed in non-preconditioned mice after 28 min IRI was significantly reduced by ischemic preconditioning (Figure 1b–e).
Twenty four hr after the second surgery, accumulation of neutrophils, macrophages, CD4+ T cells and B cells was assessed by flow cytometry. Neutrophil (CD45+7-AAD−CD11b+GR-1high) and macrophage (CD45+7-AAD−F4/80intLy6C+) numbers were markedly increased in the kidney after 28 min IRI in non-preconditioned mice (Sham/IRI, Figure 2a,b). The increase in innate inflammatory leukocytes was significantly attenuated by ischemic preconditioning (IRI/IRI, Figure 2a,b). In contrast, the number of B cells (CD45+7-AAD−B220+CD19+) in kidneys was significantly increased in the preconditioned mice vs. non-preconditioned mice, after 28 min IRI (B cells per gram kidney: 66,000±13,000 vs. 150,000±38,000, n=7, P<0.05). The number of CD4+ T cells (CD45+7-AAD−CD4+SSClow) in kidneys tended to be higher in the preconditioned mice vs. non-preconditioned mice, after 28 min IRI, although this difference was not statistically significant (CD4+ T cells per gram kidney: 106,000±32,000 vs. 172,000±36,000, n=10, P>0.05). Immunohistochemical analysis of frozen kidney sections using an antibody that recognizes neutrophils and recently emigrated monocytes (7/4; (25, 26)) confirmed the protective effect of ischemic preconditioning on innate leukocyte accumulation after IRI (Figure 2c,d,e).
Based on the previously demonstrated ability of Treg cells to traffic to areas of inflammation (23, 24), we hypothesized that kidney IRI would cause accumulation of Treg cells in the kidney. To test this, we measured the number of Treg cells in kidneys from mice at various time points after either sham surgery or 24 min bilateral renal ischemia by flow cytometry. Compared to sham-operated mice, we observed an approximately 3-fold increase in the number of CD4+CD25+FoxP3+ Treg cells in the kidney 7 days after IRI (Figure 3). The accumulation of Treg cells appeared to be kidney-specific, because no difference in the number of these lymphocytes was observed in the spleen. The observed increase in kidney Treg cell number was delayed and transient as no increase was noted at 3 days or 14 days after IRI (Figure 3). This is consistent with the finding that preconditioning was incomplete at day 3. Plasma creatinine was 0.16±0.05 vs. 0.33±0.03 mg/dl for sham/sham vs IRI/IRI, respectively (P<0.05) when preconditioned mice underwent 28 min IRI at day 3, and 0.17±0.03 vs. 0.26±0.02 mg/dl for sham/sham vs IRI/IRI, respectively (P=not significant) when preconditioned mice underwent 28 min IRI at day 7. To investigate the location of the accumulating Treg cells in the kidney, immunohistochemical analysis was performed with antibodies to CD4 and FoxP3. This revealed that CD4+FoxP3+ Tregs were observed at similar frequencies inside kidney vascular structures in both sham operated and post-ischemic kidneys (data not shown), but Treg cells were observed in the outer medulla interstitium only in post-ischemic kidneys (Figure 3b,c,d). To determine whether preconditioning also induces phenotypic changes in Treg cells, intracellular cytokine staining was performed on kidney CD4+CD25+ Treg cells for IL-10. Compared to Treg cells from non-preconditioned mouse kidneys, a greater percentage of Treg cells in kidneys of preconditioned mice expressed IL-10 (Figure 4). This increase in percentage translates into a greater than 8-fold increase in the number of IL-10+ Treg cells per kidney in preconditioned mice.
To address the role of Treg cells in the protection afforded by kidney ischemic preconditioning, we utilized a monoclonal antibody to CD25 which is highly expressed on Tregs. We and others have used this antibody (PC61) to specifically decrease CD4+CD25+FoxP3+ Treg cells in C57Bl/6 mice (17, 27, 28). On day 2 after the initial 24 min IRI, preconditioned mice were injected with PC61 or control IgG (300μg, i.v.; Figure 5a). To test the effectiveness and specificity of this treatment, some antibody treated, preconditioned mice were sacrificed at day 7 after the initial preconditioning surgery and kidney Treg cell numbers, as well as other leukocyte subsets from kidney and spleen, were measured by flow cytometry. Similar to our previous experience with PC61 (17), we did not detect any difference, between preconditioned mice injected with control IgG or PC61, in kidney neutrophils, macrophages or dendritic cells, spleen B cells, CD4+ T or CD8+ T cells, NK or NKT cells measured at 5 days after injection (Day 7 of the preconditioning protocol, Figure 5a, data not shown). We did detect a significant decrease (~50%) in kidney Treg cell number in PC61-treated preconditioned mice compared to IgG-treated preconditioned mice (Figure 5b). As previously reported, PC61 also significantly reduced spleen Treg cell numbers by approximately 50% ((17); data not shown). Injection of control IgG had no effect on the neutrophil and macrophage accumulation induced by 28 min IRI in non-preconditioned mice nor the inhibitory effect of IPC on the inflammatory leukocyte accumulation (Figure 6a,b). In contrast, decreasing the number of Treg cells in preconditioned mice inhibited the ability of IPC to prevent neutrophil and macrophage accumulation after IRI (Figure 6a,b).
Treatment with PC61 had no effect on the recovery from the initial preconditioning surgery when measured at day 7. Renal function in non-preconditioned or preconditioned mice treated with either IgG or PC61 was no different on day 7, prior to the induction of 28 min bilateral renal ischemia (Figure 7a). Similarly, renal histology was not significantly different on day 7 between preconditioned mice treated with IgG or PC61; outer medulla ATN scores in IgG vs. PC61 were 1.2 ± 0.4 and 0.7 ± 0.1, respectively, P>0.05). In contrast, 24 hr after 28 min ischemia (Day 8) the non-preconditioned IgG-treated mice (Sham/IRI-IgG) exhibited a severe decline in renal function, whereas preconditioned IgG-treated mice (IRI/IRI-IgG) retained normal renal function, as measured by PCr levels (Figure 7a). Partial Treg cell depletion, in preconditioned mice (IRI/IRI-PC61), significantly inhibited the protective effect of IPC on renal function after IRI (Figure 7a). Histological analysis of outer medulla ATN was consistent with functional observations, revealing a partial block of the protective effect of IPC on ATN by decreasing Treg cell numbers in preconditioned mice (Figure 7b,c). Outer medulla ATN scores in preconditioned mice treated with IgG vs. PC61 were 1.0 ± 0.2 and 2.6 ± 0.3, respectively, P<0.05.
To test the ability of increasing Treg numbers in mice to protect the kidney from IRI, we performed adoptive transfer of 1×105 or 1×106 Treg cells (CD4+CD25+) or 1×106 non-Treg lymphocytes (CD4+CD25−) isolated from WT C57Bl/6 mice to naïve WT C57Bl/6 recipients 18 hr prior to 28 min bilateral renal ischemia. The freshly isolated Treg cells highly expressed FoxP3, as previously described ((17); data not shown). The mice that received 1×106 CD4+CD25− non-Treg lymphocytes prior to IRI exhibited a loss of kidney function (Figure 8a) and marked accumulation of neutrophils and macrophages (Figure 9) compared to sham-operated controls. Adoptive transfer of 1×105 Treg cells partially inhibited neutrophil accumulation and preserved renal function and 1×106 Treg cells completely blocked the loss of function and neutrophil and macrophage accumulation induced by IRI (Figures 8 and and9).9). Adoptive transfer of Treg cells had a dose-dependent protective effect on acute tubular necrosis (Figure 8b,c,d,e). Outer medulla 7/4 staining confirmed the flow cytometry results demonstrating dose-dependent protection from IRI-induced inflammation in the mice receiving Treg cells (Figure 9c,d,e).
Ischemic preconditioning is a powerful technique to protect organs, including the kidney, from ischemia-reperfusion injury. In the current study we have demonstrated that delayed kidney IPC specifically inhibits the innate inflammation (kidney neutrophil and macrophage accumulation) associated with kidney IRI. We also report the novel finding that mild kidney IRI induces a delayed increase in the number of FoxP3+ and IL-10+ Treg cells in the kidney, such that at the time of the subsequent more severe renal IRI, preconditioned mice have 3-fold more FoxP3+ and 8-fold more IL-10+ Treg cells in the kidney than non-preconditioned mice. In support of a protective role for Tregs in IPC, partial Treg depletion reversed the anti-inflammatory effects of delayed IPC and significantly inhibited the functional and histological protection. Finally, adoptive transfer of Treg cells, prior to IRI in naïve mice mimicked ischemic preconditioning in terms of blocking inflammation and preventing renal injury.
There are two methods to induce ischemic preconditioning in the kidney. One type is induced by repeated short ischemic times followed by periods of reperfusion (i.e. 4 cycles of 5 min ischemia and 5 min reperfusion). This type of preconditioning partially protects the kidney from 30 min of ischemia immediately after the last preconditioning cycle and at 24 hr after the last cycle, but not at 6 hours after preconditioning (29). Delayed kidney IPC involves a longer single initial ischemic time (i.e. >20 min) followed by a recovery period of 6 to 8 days (11, 14). After this procedure, mice are completely protected from subsequent ischemic injury for greater than one week and partially protected for up to twelve weeks (14). We used a model of delayed ischemic preconditioning that resulted in complete functional protection of the kidney from severe ischemic injury 7 days after the initial preconditioning surgery. This model of IPC markedly decreased the infiltration of neutrophils and macrophages in the post-ischemic kidney, which are both known to cause renal injury (5–10).
A report from Ascon et al. (30) demonstrated that adoptive transfer of lymphocytes from kidneys of preconditioned mice conferred protection against renal IRI in naïve T cell-deficient mice, suggesting that renal IRI causes kidney accumulation of lymphocytes with a protective (potentially anti-inflammatory) phenotype. Characterization of the specific lymphocyte (T or B cell) subset(s) responsible for the protection was not reported. We hypothesized that the highly immunosuppressive CD4+CD25+ Treg cells (expressing FoxP3+ and/or IL-10) could be important mediators of kidney ischemic preconditioning for several reasons. First, we have recently shown that CD4+CD25+FoxP3+ Treg cells are an intrinsic protective mechanism against kidney IRI in naïve mice, which directly and potently block the innate immune response to IRI (17). Second, WT, but not IL-10 KO, Treg cells could suppress IRI in lymphocyte-deficient RAG-1 KO mice (17). Furthermore, the trafficking and accumulation of Treg cells in areas of ongoing inflammation has been documented in other models (23, 24). We found that the initial ischemic insult in our preconditioning model resulted in a kidney-specific increase in the number of kidney CD4+CD25+FoxP3+ Treg cells and the number and percentage of CD4+CD25+ Treg cells expressing IL-10 in the kidney. Our findings are in agreement with a recent report from Gandolfo et al. (16) demonstrating renal accumulation of Treg cells after more severe IRI. The preconditioned mice, which were completely protected from ischemic injury, possessed significantly more kidney Treg cells than non-preconditioned mice that were susceptible to IRI. In addition, the localization of the Treg cells in the preconditioned kidneys was distinct, in that Tregs were consistently observed in the outer medulla interstitium (which is also the area of the most extensive inflammation and tubular necrosis following kidney IRI). In summary, the preconditioned kidney has a larger population of anti-inflammatory Treg cells and they are located in the most ischemia-susceptible region of the kidney, the outer medulla.
To determine the role of Treg cells in kidney IPC, the Treg cell-depleting antibody, PC61 was administered to preconditioned mice, resulting in a ~50% reduction in Treg cell numbers, compared to control IgG-treated preconditioned mice, at the time of the second ischemic insult. Decreasing Treg cell numbers, in preconditioned mice, caused inhibition of the anti-inflammatory effect of IPC and partially blocked the functional and histological protection. These results reveal an important role for Treg cells in kidney IPC. The finding that mice are not completely protected from IRI-induced dysfunction at 3 days after the initial preconditioning surgery (prior to kidney Treg cell accumulation) adds additional support for the contribution of Treg cells to the protection of IPC.
Although Treg cell-deficient mice, FoxP3-deficient Scurfy mice, are available, they develop extensive autoimmune disease and early death (31) precluding their use in our studies. Furthermore, our goal was to test the effectiveness of IPC in kidneys that contained the basal number of Treg cells found in non-preconditioned mice. We were able to create these conditions in two ways: 1) use of preconditioned mice 3 days after the initial surgery and prior to the increase in kidney Treg cell numbers, and 2) by treating preconditioned mice with PC61 to reduce the number of kidney Treg cells at 7 days after the initial surgery. Both methods resulted in partial reversal of the functional protection of IPC.
Other immune cells may participate in kidney preconditioning based upon studies that have demonstrated their role in kidney IRI or in actual preconditioning. B cells do accumulate in the post ischemic kidney (9) and may have a pathogenic role in kidney IRI (3). However no studies have examined their role in preconditioning. Although macrophage infiltration may contribute to the recovery following kidney injury they do not contribute to ischemic preconditioning in kidneys (32). To our knowledge there are no studies demonstrating a role for neutrophils in ischemic preconditioning in kidneys. Further studies that focus on immune cells may reveal additional mechanisms of ischemic preconditioning.
While our results demonstrate that Treg cells are important mediators of the anti-inflammatory action of IPC, other mechanisms protect the preconditioned kidney, even in the presence of extensive inflammatory leukocyte accumulation. In delayed kidney IPC, several non-immune mechanisms have previously been described including upregulation of the renal pro-survival signals HSP 27 expression (14) and Akt phosphorylation (13), and decreased phosphorylation of cell death promoting p38 and Jun kinase (14). These changes presumably occur in the epithelial and endothelial cells of the preconditioned kidney and make them more resistant to ischemia and inflammatory mediator (reactive oxygen species, proteases, etc.)-induced toxicity, leading to the substantial residual protection observed in preconditioned mice treated with PC61.
Our goal in studying IPC is to discover novel and less invasive strategies to protect the kidney from IRI. As a first step in this direction we assessed the effectiveness of adoptive transfer of Treg cells to a naïve WT recipient prior to kidney IRI. We found that, in contrast to an equivalent number of non-Treg CD4+ lymphocytes, Treg cells could completely inhibit inflammation and renal dysfunction induced by severe bilateral renal ischemia. These results complement our previous finding that Treg cells, but not non-Treg lymphocytes, can inhibit renal neutrophil accumulation and IRI in lymphocyte-deficient RAG-1 KO mice (17). Therefore, in the presence or absence of other leukocytes, Treg cells suppress the innate inflammation and subsequent loss of function induced by kidney ischemia-reperfusion. Our findings suggest that adoptive transfer or pharmacologic induction of Treg cells may represent less invasive methods to protect organs from ischemic insult.
The source of the renal Treg cells, accumulating after IRI, is not currently known. At least 2 different explanations are possible, which are not mutually exclusive. First, existing Treg cells from the circulation could be recruited to the post-ischemic kidney by chemokines. The CCR2 chemokine receptor directs Tregs towards arthritic joints (33). Monocyte chemoattractant protein-1 (MCP-1/CCL-2) is the major ligand for CCR2 and the expression of this chemokine is highly upregulated in the kidney after IRI (34, 35). A second potential source of the Treg cells could be kidney non-Treg lymphocytes, activated by T cell receptor stimulation in the presence of TGF-β, which has been shown to induce FoxP3 expression and a suppressive phenotype (36). While TGF-β is produced in the kidney after ischemia (37), so is IL-6 (38), and the combination of these 2 cytokines promotes the induction of Th17 cells rather than Treg cells (39). However, the time course and the intra-kidney differences in concentrations of these cytokines may still permit the induction of Treg cells in the post-ischemic kidney. The origin of Treg cells accumulating after IRI is under investigation and may lead to more specific ways to mimic IPC in different organs.
In summary, IPC markedly inhibits inflammation, acute tubular necrosis and loss of function induced by ischemia-reperfusion. We demonstrate that the initial ischemic insult (preconditioning) causes a kidney-specific increase in Treg cells, such that the preconditioned kidney has several-fold more Treg cells than non-preconditioned kidneys. Partial depletion of Treg cells in preconditioned mice restores the accumulation of neutrophils and macrophages, induced by IRI, and partially reverses functional and histological protection. Finally, increasing Treg cell numbers in WT mice, by adoptive transfer, mimics the anti-inflammatory and renal protective effects of IPC. Taken together these results suggest that strategies designed to increase regulatory T cells may represent a novel therapy for ischemia-reperfusion injury.
Male C57Bl/6 mice, 6 to 8 weeks of age weighing 20 to 25 grams, were obtained from Charles River Laboratories (Wilmington, MA). CD4+CD25+ Treg cells and CD4+CD25– non-Treg lymphocytes were isolated from WT C57Bl/6 mice using the Dynal (Carlsbad, CA) CD4+ negative selection kit and Miltenyi Biotec (Auburn, CA) CD25+ positive selection kit according as previously described (17). Cells were adoptively transferred by tail vein injection in 200 μl normal saline.
For the renal IRI protocol, both renal pedicles were exposed and cross-clamped for 24 or 28 min as previously described by our laboratory (7, 9, 17). All animal experiments were approved by the University of Virginia Institutional Animal Care and Use Committee.
Five color flow cytometry was utilized to determine the number and phenotype of leukocytes in the kidney. Counting beads (Caltag, Carlsbad, CA) were used as described previously (7, 9, 17) to determine the total number of CD45+ cells/g kidney. CD4+CD25+FoxP3+ Tregs were stained, fixed and permeabilized using the eBioscience (San Diego, CA) FoxP3 buffer set according to the manufacturer's protocol as previously described (17). Anti-mouse fluorophore-labeled antibodies and dead cell stains used have been described previously (17).
Intracellular cytokine staining for IL-10. Kidneys were harvested and digested with collagenase as previously described (7, 9, 17). The kidney cell suspension was enriched for mono-nuclear cells by layering the suspension (in 5ml ice-cold PBS) on top of 2.5mL Histopaque (Sigma) and centrifuging at 230 × g for 20 min. Cells residing at the interface of PBS and Histopaque were recovered and washed 2× with PBS. Cells were incubated 5 hr at 37°C in RPMI 1640 (Gibco) + 10% FBS, penicillin/streptomycin and 50μM β-mercaptoethanol at 2 million cells per 2 ml and treated with leukocyte activation cocktail (Invitrogen) according to the manufacturer's protocol. Extracellular staining was performed as described above for CD45, 7-AAD, CD4 and CD25, then cells were fixed and permeabilized using the BD Cytofix/perm kit according to the manufacturer's protocol and stained with 0.2μg per sample α-mouse IL-10 APC (Clone JES5-16E3; BD Biosciences).
Kidneys were fixed overnight in 1% PLP, incubated for 48h in 30% sucrose at 4°C, then embedded and frozen in Optimal Cutting Temperature compound (OCT, Ted Pella Inc., Redding, CA). Ten micron frozen kidney sections from each mouse were blocked with goat serum and 2.4G2, then labeled with Treg specific antibodies, CD4-PE and FoxP3-Alexa Fluor 647, or FITC-conjugated neutrophil/monocyte-specific antibody (clone 7/4; Cedarlane, Hornby, Ontario, Canada). Nuclei were visualized using DAPI. Specimens were mounted with ProLong Gold Antifade reagent (Molecular Probes, Eugene, OR) and examined using a Zeiss Axiovert 200 microscope with ApoTome.
The TIB-222 hybridoma was used by the Lymphocyte Culture Core Facility at the University of Virginia to produce the anti-mouse CD25 monoclonal antibody (clone PC61). Rat IgG isotype control, ImmunoPure Rat IgG was purchased from Pierce (Rockford, IL).
Comparisons of two treatment groups were made using an unpaired t-test to determine statistical significance unless a normality test failed, in which case the Mann-Whitney Rank Sum Test was performed. Analysis of variance (ANOVA) was performed for comparisons of 3 or more groups and the Holm-Sidak procedure used for pair-wise comparisons using the SigmaStat statistical software (San Jose, CA). A p-value of <0.05 was considered statistically significant.
We are grateful to Dr. Kenneth Tung and Hui Qiao (University of Virginia) for the use of the TIB-222 hybridoma and assistance with purification of the PC61 mAb. We also thank Drs. Peter Lobo, Rahul Sharma and the Okusa laboratory for helpful discussions and suggestions about this project.
This work was supported by NIH grants, DK56223, DK58413, DK62324, T32DK072922-01, F32DK083185 and PO1HL07336.