The Role for T Cell Repertoire/Antigen-Specific Interactions in Experimental Kidney Ischemia Reperfusion Injury¹

Renal ischemia reperfusion injury (IRI)⁴ is a common cause of acute renal failure/acute kidney injury. Renal IRI worsens both early and late transplant kidney outcomes (1). T cells have recently been demonstrated to directly participate in early tissue injury and dysfunction in IRI in various organs including kidney (2, 3), liver (4), lungs (5), and brain (6). T cells are recruited early during the pathogenesis of renal IRI and these infiltrating cells secrete significant levels of proinflammatory cytokines (7). Depletion of T cells to very low levels leads to protection against renal IRI (8). Moderate protection observed in athymic (nu/nu) mice is reversible upon adoptive transfer of CD4 but not CD8 cells before ischemia and this protection is IFN-γ and CD28 dependent (9). STAT6 deficiency exacerbates renal IRI, whereas STAT4 deficiency has a protective effect on both renal (10) and liver (11) IRI, suggesting pathogenic and protective roles of Th1 and Th2 cells, respectively. CXCR3-mediated Th1 T cell recruitment to kidneys participates in renal ischemic injury (12). Although deficiency of T cells or B cells (13) is protective against injury after renal ischemia, RAG1-deficient mice that lack both T and B cells are usually not protected from kidney IRI and can paradoxically have worse injury (14, 15). These findings demonstrate the intricate nature of role of T cells in renal IRI.

The mechanisms by which T cells exert their pathogenic effect in ischemic kidneys could be either Ag dependent or non-Ag dependent (16). Non-Ag-dependent activation of T cells through cytokines (TNF-α, IL-6, etc.) secreted by macrophages and dendritic cells, chemokines (RANTES, MIP-1), oxygen free radicals, or complement could play an important role in ischemic injury to kidneys (17, 18, 19, 20). TLRs TLR2 and TLR4 expression is up-regulated on infiltrating leukocytes and tubular epithelial cells after IRI (21, 22, 23). Activation of TLRs helps in maturation of dendritic cells, which in turn activate T cells. The role of T cell activation in IRI through the Ag-TCR-mediated classical pathway has some support from recent data (7, 24). Mice deficient in TCR-αβ are significantly protected against IRI and have decreased levels of kidney TNF-α and IL-6 (25). We therefore hypothesized that pathogenic effect of T cells in renal IRI is influenced by diversity in the TCR repertoire and that restricting this repertoire would modify pathogenesis of renal IRI.

In this study, we used genetically engineered DO11.10-transgenic mice with their TCRs directed against chicken OVA peptide 323–339 (26, 27). Athymic nu/nu mice that are relatively resistant to renal IRI were reconstituted with T cells from either DO11.10 or wild-type (WT) mice without immunization and subsequently subjected to renal ischemia. We found that, although DO11.10 T cells can effectively infiltrate into ischemic kidneys, they failed to restore the kidney injury in nu/nu mice compared with WT T cells, indicating a diverse TCR repertoire in naive T cells is required for IRI. This resistance to renal IRI was associated with reduced IFN-γ secretion by DO11.10 T cells. T cell activation by immunization with OVA in CFA (OVA/CFA) reversed the protective phenotype of DO11.10 mice, demonstrating that once activated through Ag-TCR, a diverse TCR repertoire is no longer required for their pathogenic effect on IRI. Thus, TCR diversity as well as T cell activation are important mechanisms that underlie the T cell’s role in kidney IRI.

Transgenic DO11.10 mice (C.Cg-Tg(DO11.10)10Dlo/J) and WT BALB/cJ control mice were purchased from The Jackson Laboratory. DO11.10 mice constitutively express genes for T cell receptors that specifically recognize OVA. RAG2-deficient DO11.10 (BALB/c-(Tg) DO11.10-(KO) RAG2) mice were purchased from Taconic Farms. T cell-deficient nu/nu mice bred on a BALB/c background (CBy.Cg-Foxn1nu/J) were purchased from The Jackson Laboratory. Mice were housed under pathogen-free conditions according to established institutional protocols from the Institutional Animal Care and Use Committees of Johns Hopkins University.

Spleens were collected from DO11.10 or BALB/cJ WT donor mice. Splenic cells were collected by centrifugation, and RBC were removed using cell lysis buffer (eBioscience). Mouse T cell enrichment columns (R&D Systems) were used to isolate purified populations of T cells for adoptive transfer according to the manufacturer’s specifications. Depletion of CD25⁺ cells was performed using biotin-labeled rat anti-mouse CD25⁺ (clone 7D4; BD Biosciences), anti-biotin microbeads, and LD columns (Miltenyi Biotec). Approximately 15 × 10⁶ enriched cells were suspended in isotonic saline and injected i.p. into each athymic nu/nu mouse. IRI was induced in the mice 3 wk after transfer. The percentage of CD3⁺ cells within enriched cell population ranged between 85 and 92% and >99% of CD25⁺ cells were depleted as measured by flow cytometry.

The model is well established and described previously (9). Mice were anesthetized i.p. using 75 mg/kg sodium pentobarbital. Abdominal incisions were made and the renal pedicles were dissected gently. A microvascular clamp was placed on each renal pedicle. nu/nu mice underwent 30 min of renal pedicle clamping, whereas BALB/c mice underwent 35 min. We tested BALB/c mice kidneys after 30 min of ischemia. In our hands, we could not produce equivalent kidney damage in these experiments as in C57BL/6 mice after 30 min of ischemia. On the other hand, 35 min of ischemia have caused a consistent rise in serum creatinine (SCr) in BALB/c mice. Key controls were used in each study to correct for differences in ischemia time. During the procedure, animals were kept well hydrated with saline and kept at a constant temperature (37°C). After the clamps were removed, the wounds were sutured and the animals were allowed to recover. All surgical methods have been approved by the Institutional Animal Care and Use Committee.

Blood samples were obtained from the tail vein at 0, 24, 48, and 72 h after ischemia. SCr was measured in mg/dl using Cobas Mira plus autoanalyzer (Roche).

Animals were sacrificed at 72 h after ischemia and kidneys were harvested. Tissue sections were fixed with 10% formalin followed by paraffin embedding and then stained with H&E. Tissue sections were scored in a blinded fashion to evaluate the percentage of tubular damage in cortex and medulla as previously described (28).

Seventy-two hours after ischemia, mice were anesthetized, exsanguinated, and their kidneys and spleens were harvested for lymphocyte isolation (7). Decapsulated kidneys were suspended in complete RPMI 1640 medium and disrupted mechanically using Stomacher 80 Biomaster (Sewart). Samples were strained, washed, and resuspended in 36% Percoll (Amersham Pharmacia). These samples were then gently laid over 72% Percoll and centrifuged at 1000 × g for 30 min at room temperature. Lymphocytes were collected from Percoll interface, washed twice, and counted using trypan blue exclusion on a hemocytometer.

Isolated KMNC were incubated with anti-CD16/CD32 FcR-blocking Ab (BD Biosciences) and subsequently stained with the following Abs: anti-CD3-allophycocyanin, anti-CD4-PerCP, and KJ1-26-PE (all from BD Biosciences). Cells were fixed with 1% paraformaldehyde and analyzed using a FACSCalibur instrument (BD Biosciences) and CellQuest software (BD Biosciences).

One million KMNC or splenocytes were suspended in RPMI 1640 medium with 10% FBS, l-glutamine, and penicillin/streptomycin, incubated at 37°C in the presence of PMA (5 ng/ml), ionomycin (500 ng/ml; both Sigma-Aldrich), and monensin (BD Biosciences). After a 5-h culture, cells were washed and stained for surface markers using anti-CD4-PerCP and anti-CD3-allophycocyanin Abs. These cells were then permeabilized in Cytofix/Cytoperm solution (BD Biosciences) for 20 min and washed twice with perm/wash buffer. Cells were then incubated with anti-TNF-α-FITC and anti-IFN-γ-PE Abs for 20 min, washed with perm/wash buffer. and analyzed with a FACSCalibur instrument (BD Biosciences).

KMNC were pretreated with FcR-blocking Ab and anti-CD16/32 and surface stained with anti-CD25-FITC, anti-CD4-PerCP, and anti-CD3-allophycocyanin as described above. Cells were then permeabilized with freshly prepared fixation/permeabilization reagent as per the manufacturer’s instructions (eBioscience) and subsequently stained with anti-Foxp3-PE Ab. Samples were analyzed using a FACSCalibur instrument (BD Biosciences) and CellQuest software (BD Biosciences).

A panel of cytokines was measured in kidney samples taken at 72 h after ischemia after sacrifice. A Bio-Plex Protein Array system (Bio-Rad) was used according to the instructions of the manufacturer. This is a multiplexed, particle-based, flow cytometric assay, which utilizes anti-cytokine mAbs linked to microspheres incorporating distinct proportions of two fluorescent dyes. Our assay was customized to detect and quantify IL-2, IL-10, GM-CSF, and IFN-γ. For each cytokine, eight standards were used ranging from 2 to 3200 pg/ml and the minimal detectable dose was <10 pg/ml.

DO11.10 or BALB/c mice were injected s.c. with OVA_323–339 (100 μg/mouse) emulsified in CFA. Control mice received CFA alone. Seven days later, all mice underwent renal ischemia surgery.

Data are expressed as mean ± SE. Comparison of group means was performed using a two-way ANOVA or unpaired Student’s t test. A value of p < 0.05 was considered significant.

We tested whether T cells from TCR-transgenic DO11.10 mice traffic differently to ischemic kidneys. Enriched T cells from spleens of DO11.10 or WT mice were adoptively transferred i.p. to nu/nu mice. Three weeks later, these mice underwent renal IRI. At 72 h after ischemia, both recipient groups expressed 3–4% CD4⁺CD3⁺ T cells among total KMNC, 3.52 ± 0.43% and 3.27 ± 0.40% in WT to nu/nu and DO11.10 to nu/nu, respectively (Fig. 1). There was no statistically significant difference between the two groups (p > 0.05). Blood and spleen showed similar percentage of T cells (data not shown). The proportion of OVA-specific T cells in nu/nu recipients transferred with DO11.10 T cells was comparable to that of donors (75.23% ± 4.94). This implies that Ag specificity of T cells does not modify trafficking to ischemic kidney.

Athymic nu/nu mice lack T cells and have decreased kidney IRI injury compared with strain-matched controls. This protective effect is reversed after adoptive transfer of T cells (9). Whereas nu/nu mice that received WT T cells showed an increase in SCr levels, those that received T cells from DO11.10 mice had significantly less kidney dysfunction (2.18 ± 0.34 vs 1.34 ± 0.81 of SCr. (mg/dl) on day 2, p < 0.05; Fig. 2). WT T cell recipients had higher histological injury than those that received T cells from DO11.10 mice (Fig. 3, A and B). On day 3, tubular epithelial damage in medulla in WT to nu/nu mice was significantly higher than that in DO11.10 to nu/nu mice (24 ± 5.6% vs 11 ± 1.4%, p < 0.05). This result supports the previous report by our group (9) that full expression of renal IRI requires the presence of T cells in kidney and that WT T cells with a diverse TCR repertoire lead to worse injury compared with T cells with the limited TCR repertoire of DO11.10 T cells (Figs. 1 and 2).

We have previously shown that T cells infiltrating ischemic kidneys have increased ability to produce proinflammatory cytokines (7). One mechanism of attenuated ischemic renal injury in recipients of DO11.10 T cells could be reduced production of proinflammatory cytokines. To assess the involvement of these cytokines, we measured intracellular cytokines in T cells after renal IRI. Adoptively transferred T cells from BALB/c and DO11.10 mice were similar in their ability to produce IFN-γ and TNF-α in spleen. However, those that trafficked to kidney showed different proinflammatory phenotypes. IFN-γ production in kidney-infiltrating T cells was lower in nu/nu mice transferred with DO11.10 T cells (2.12 ± 0.40%) compared with that in mice transferred with WT T cells (5.8 ± 1.34%, p < 0.05; Fig. 4,A). The numbers of CD3⁺CD4⁺ T cells in kidney after 72 h of IRI were 0.97 ± 0.35 × 10⁴ cells in the WT to nu/nu and 0.83 ± 0.32 × 10⁴ cells in the DO11.10 to nu/nu group, and they were not different significantly (n = 6/group; p > 0.05). Therefore, a lower proportion of IFN-γ-producing T cells in nu/nu mice transferred with DO11.10 T cells is not due to differences in absolute number of T cells infiltrated in injured kidney between two groups. TNF-α production was not different statistically in experimental (16.02 ± 5.05%) and control groups (13.99 ± 5.11%; n = 6/group, p > 0.1; Fig. 4 B).

IL-10 and IFN-γ were up-regulated in ischemic kidneys and are thought to participate in recovery and injury in renal IRI, respectively (7, 29, 30). To test whether the limited TCR repertoire of DO11.10 T cells in nu/nu mice led to immune deviation by shifting cytokine patterns, we determined expression of IL-2, IL-10, GM-CSF, and IFN-γ at the protein level in the whole kidney by using a multiplex assay. There was no difference in IL-2 and GM-CSF (p > 0.1). IFN-γ and IL-10 levels were slightly higher in DO11.10 T cell recipient kidneys (9.5 ± 2.9 and 9.2 ± 1.6 pg/mg, respectively) compared with controls (7.8 ± 2.8 and 6.3 ± 2.9 pg/mg, respectively), but it was not statistically significant (p > 0.1; Fig. 5). Cells other than CD4⁺ T cells may contribute to the production of IFN-γ, since the number of CD4⁺ T cells were not different in the injured kidney between experimental and control mice (Fig. 4).

Increased infiltration of CD4⁺CD25⁺Foxp3⁺ Treg cells in DO11.10 T cell recipients could be responsible for attenuated proinflammatory cytokine response and for less injury. We therefore measured the proportion of Treg cells among kidney-infiltrating T cells of nu/nu mice that received DO11.10 or WT T cells using flow cytometry. The percentage of Treg cells in DO11.10 T cells transferred mice was not substantially different from that in kidney as well as spleen of WT T cell-transferred mice (p > 0.1, Fig. 6). DO11.10 T cell-transferred nu/nu mice were not deficient in Treg cells both in kidney and spleen compared with WT T cell recipients. This suggests that the number of Treg cells is not the major factor altering attenuated renal IRI responses in DO11.10 T cell-transferred nu/nu mice.

To directly examine whether Ag-specific activation of T cells is required for kidney ischemia injury, we immunized DO11.10 mice and BALB/c mice with OVA/CFA or CFA alone. Seven days later, mice were subjected to renal IRI and the SCr level was measured. Without immunization, DO11.10-transgenic mice had significantly lower levels of SCr after ischemia than WT mice (at 24 h, 0.99 ± 0.18 vs 2.25 ± 0.45 mg/dl, p < 0.05; Fig. 7,A). Compared with DO11.10 mice either untreated (0.99 ± 0.18 mg/dl) or treated with CFA alone (1.33 ± 0.27 mg/dl), OVA/CFA-immunized DO11.10 mice (2.69 ± 0.41 mg/dl) had higher SCr levels at 24 h after IRI (p < 0.01 or p < 0.05, respectively; Fig. 7,B). There was no significant difference in SCr between unimmunized and immunized WT mice (p > 0.1; Fig. 7, A and B). The SCr level of OVA/CFA-immunized DO11.10 mice increased to that of WT mice () (Fig. 7, A and B), suggesting that Ag-specific T cell activation through TCR is important for renal IRI. Activated T cells were detected in kidney on day 3 of IRI. The proportion of CD25-expressing CD4⁺ T cells was significantly higher in OVA/CFA-immunized DO11.10 mice (19.77 ± 1.55%; p < 0.05) compared with that of the CFA-treated group (7.46 ± 0.35%), OVA/CFA-immunized WT mice (9.13 ± 0.77%), or CFA-treated WT (6.15 ± 0.67%; Fig. 7,C). In contrast to the result in naive mice where diverse TCR repertoires were required to induce renal IRI (Fig. 2), a limited TCR diversity of OVA-specific T cells in DO11.10 mice was sufficient to induce renal injuries after immunization with OVA/CFA.

To further investigate the role of activated T cells in renal IRI, we isolated naive T cells by depleting the T cell preparations of CD25-expressing cells before adoptive transfer into nu/nu mice. SCr levels were similar in WT to nu/nu mice regardless of CD25 depletion (p > 0.05; Fig. 8,A). However, it was substantially higher in the early phase of renal IRI (day 1) in recipients of naive T cells from DO11.10 mice compared with that in recipients of DO11.10 whole T cells (1.00 ± 0.33 vs 0.43 ± 0.19, p < 0.05; Fig. 8,A). There was no difference in SCr on day 3 of renal IRI. This result suggests that CD25⁺ T cells play an important role in the early phase of renal IRI in the presence of a limited TCR repertoire of T cells in kidney. CD3⁺ T cells in splenocytes were 20–30% and enriched up to 70–83% (Fig. 8,B). CD25⁺CD4⁺ T cells were 7–9% in splenocytes or CD3-enriched populations in naive mice. CD25 depletion was >99% verified by FACS analysis (Fig. 8 C). CD25-expressing cells include both activated T cells and also CD4⁺CD25⁺ Treg cells, and the CD25-depletion procedure likely removed Treg cells from the T cell preparation (31). Indeed, 90% of CD4⁺CD25⁺ cells in naive mouse spleen are shown to coexpress FoxP3, a marker of Treg cells (32). Therefore, it is difficult to determine which cells of these two populations mainly contribute to the increase of SCr on day 1 in recipients of DO11.10 naive T cells.

To examine the role of CD25⁺ T cells in renal IRI in the presence of T cells having a limited TCR repertoire, we used RAG2-deficient DO11.10 mice, which have only one clonotype TCR specificity and not other T cells, including Treg cells (33). However, DO11.10 mice have Treg cells and CD4⁺ T cells with other TCR repertoires than OVA-specific TCR (34). T cells were prepared from RAG2-deficient DO11.10 or DO11.10 mice and adoptively transferred into nu/nu mice 3 wk before IRI. The SCr level was significantly higher in RAG2-deficient DO11.10 to nu/nu (2.3 ± 0.78 mg/dl) on day 2 compared with DO11.10 to nu/nu (1.34 ± 0.81 mg/dl, p < 0.05; Fig. 8,D). Moreover, the survival rate at day 3 was 20% in RAG2-deficient DO11.10 to nu/nu and 100% in DO11.10 to nu/nu mice (data not shown). There were no CD8⁺ T cells or CD25⁺ T cells in the preparation from RAG2-deficient DO11.10 (Fig. 8 E).

In this study, using OVA-specific TCR-transgenic mouse DO11.10, we report that TCR diversity is important for participation of T cells in kidney IRI. To address the question whether there was differential recruitment of T cells to kidneys due to lack of TCR diversity, we measured the percentage of T cells among total KMNC. We found no difference in proportion of CD4⁺ T cells in KMNC of nu/nu mouse adoptively transferred with DO11.10 T cells or WT T cells. However, IFN-γ production in kidney-infiltrating DO11.10 T cells was decreased compared with that in WT T cells. This indicates that relative protection from ischemic injury, found in DO11.10 T cells transferred nu/nu mice, was due to a limited TCR repertoire and altered function of T cells rather than their decreased recruitment to kidneys. We also demonstrate that Ag-specific activation of T cells through OVA-specific TCR renders DO11.10 mice susceptible to renal IRI. Thus, after Ag-specific activation, the TCR repertoire becomes less important for kidney IRI than T cell function.

It has been previously demonstrated that nu/nu mice are moderately resistant to renal IRI and this injury can be restored upon reconstitution with syngeneic T cells (9). Indeed, nu/nu mice reconstituted with T cells from WT controls had significantly higher SCr levels after ischemia compared with intact nu/nu mice (9). When T cells from DO11.10 mice were adoptively transferred to nu/nu mice, these mice failed to show reversal of protective phenotype. Thus, TCR-transgenic T cells are not as effective as WT T cells in augmenting the subdued kidney injury phenotype of T cell-deficient mice. Since we selectively transferred T cells from donor mice to T cell-deficient mice, the observed difference between TCR-transgenic and WT control T cell recipients can be purely attributed to the nature of T cells adoptively transferred. However, these results do not undermine roles of other immune parameters such as neutrophils, macrophages, or change in cytokine/chemokine milieu in renal IRI. Although infiltration of CD4⁺ T cells was similar in kidneys and spleens of nu/nu recipient mice of DO11.10 or WT T cells, fewer T cells in the kidney of DO11.10 T cell-transferred mice produced proinflammatory cytokine, IFN-γ, compared with those in WT T cell-transferred mice. Bio-Plex cytokine assay measures cytokine levels in whole kidney protein extract. We found that there was no statistically significant difference in levels of IL-2, IL-10, GM-CSF, and IFN-γ, although there was a trend toward increased IL-10 in DO11.10 T cell recipient kidneys. An apparent inconsistency between levels of IFN-γ by intracellular cytokine staining and Bio-Plex assay can be explained by the fact that the former measured the frequency of IFN-γ secreting CD4⁺ T cells alone, whereas the source of this cytokine in the latter could involve other cell type such as NK cells or CD8⁺ T cells. Taken together, these results indicate that nu/nu recipients of DO11.10 T cells have attenuated ischemic injury compared with their controls, likely due to a limited TCR repertoire. This attribute, even though it does not limit infiltration of T cells in ischemic kidneys, attenuates their proinflammatory cytokine responses. Therefore, without Ag-specific activation, TCR diversity plays an important role in renal IRI. We propose that T cells require TCR diversity to recognize yet unidentified Ag targets in ischemic kidneys after IRI. This TCR-mediated immune response is responsible, in part, for the full tissue injury after renal IRI.

The role of Treg cells in renal IRI is of considerable interest. Depletion of CD4⁺CD25⁺ Treg cells, starting 24 h after IRI, leads to increased damage at 3 and 10 days after ischemia (35). In this current study, we demonstrate the presence of CD4⁺CD25⁺Foxp3⁺ Treg cells in kidneys after ischemia and compared this in different experimental groups. The frequency of these cells was not different in DO11.10 mice or nu/nu mice reconstituted with DO11.10 T cells compared with their WT controls. This suggests that differential susceptibility to ischemic injury in nu/nu mice reconstituted with WT T cells or DO11.10 T cells was not due to altered kidney Treg trafficking.

T cells can be activated through an Ag-dependent or Ag-independent manner. Ag-independent activation of T cells could occur in renal IRI through cytokines, chemokines, and oxygen-derived free radicals or through complement system (36). Other studies indicate that Ag-dependent activation through classical signal 1 and signal 2 pathways is required for pathogenesis of ischemic injury. In one study, deficiency of TCR-αβ rendered mice resistant to ischemic injury (25). Another study demonstrated that blockade of the costimulatory pathway by CTLA-4 protects the kidneys during acute IRI (37). Irrespective of the mode of stimulation, once activated, T cells contribute to ischemic injury. Indeed, we found that DO11.10-transgenic mice once immunized with antigenic peptide OVA showed restoration of severe ischemic injury phenotype due to the activation of T cells.

Given that the D011.10 TCR transgenic is in some ways a “leaky” knockout, we studied T cell transfers from D011.10 TCR transgenics bred on a RAG2-deficient background (33, 34). T cell transfers from these mice actually enhanced renal dysfunction and mortality in recipient mice, similar to the worse course of IRI in RAG-deficient mice vs WT as previously published (14). From the point of TCR diversity, this result initially stands in contrast to the result in Fig. 2 that diverse TCR repertoires are required for renal IRI. However, the composition of T cell preparation was different between DO11.10 and RAG2-deficient DO11.10. RAG-deficient mice could well be lacking lymphocytes that play a protective role in kidney IRI, which overcomes the absence of deleterious lymphocytes. Indeed, Treg cells are functionally present in DO11.10 mice, but not in RAG2-deficient DO11.10 mice (33). Therefore, we could presumably exclude the possibility of proliferation and renal infiltration of preexisting Treg cells in RAG2-deficient DO11.10 to nu/nu mice.

In conclusion, our data demonstrate that the diversity of the TCR repertoire participates in the pathogenesis of kidney IRI, but even a limited repertoire is sufficient once T cells are Ag-specifically activated through TCR. Moreover, this study has demonstrated for the first time that Ag-specific activation of T cells plays a role in renal IRI. Newer therapeutic modalities to control specific T cell responses could be applicable to management of renal IRI.

We thank Yanfei Huang, Karl Womer, Douglas Linfert, Carolyn Feltes, and Mark Soloski for their valuable suggestions.

The authors have no financial conflict of interest.