Viewed Large Fraction of DN T Cells in “Normal” Sections of Human Histologically

BASIC RESEARCH www.jasn.org

Double-Negative ab T Cells Are Early Responders to AKI and Are Found in Human Kidney

† † Maria N. Martina,* Sanjeev Noel, Ankit Saxena,* Samatha Bandapalle, Richa Majithia,* ‡ † Chunfa Jie, Lois J. Arend,* Mohamad E. Allaf,§ Hamid Rabb, and Abdel Rahim A. Hamad*

† *Department of Pathology, Department of Medicine, and §Department of Urology, Johns Hopkins University, Baltimore, Maryland, and ‡Department of Surgery, Northwestern University Feinberg School of Medicine, Chicago, Illinois

ABSTRACT Ischemia-reperfusion injury (IRI) is a major cause of AKI, and previous studies established important roles for conventional CD4+ T cells, natural killer T cells, and CD4+CD25+FoxP3+ Tregs in AKI pathogenesis. We 2 2 recently identified CD4 CD8 (double-negative; DN) T cells as an important subset of ab T cell receptor– positive cells residing in mouse kidney. However, little is known about the pathophysiologic functions of kidney DN T cells. In this study, we phenotypically and functionally characterized murine kidney DN T cells in the steady state and in response to IRI. Unlike CD4+ and CD8+ T cells, DN T cells in the steady state expressed high levels of CD69, CD28, and CD40L; differentially expressed IL-27 and IL-10 anti-inflammatory cytokines; spontaneously proliferated at a very high rate; and suppressed in vitro proliferation of activated CD4+ T cells. Within the first 3–24 hours after IRI, kidney DN T cells expanded significantly and upregulated expression of IL-10. In adoptive transfer experiments, DN T cells significantly protected recipients from AKI by an IL-10–dependent mechanism. DN T cells also made up a large fraction of the T cell compartment in human kidneys. Our results indicate that DN T cells are an important subset of the resident ab+ T cell population in the mammalian kidney and are early responders to AKI that have anti- inflammatory properties.

J Am Soc Nephrol 27: 1113–1123, 2016. doi: 10.1681/ASN.2014121214

AKI is associated with high mortality in native Hence, very little is known about their roles in nor- kidneys, increased morbidity in allografts, and very mal and ischemic kidneys. high health care costs. Among the most common In the present study, we phenotypically and causes of AKI in both native and transplanted functionally characterized murine kidney DN T kidneys is ischemia-reperfusion injury (IRI). Both cells both in the steady state and in response to IRI. innate and adaptive immune cells contribute to the DN T cells also made up a large fraction of the T cell disease pathogenesis.1 Dendritic cells have also compartment in human kidneys. Our results in- been implicated in AKI.2,3 dicate that DN T cells are an important subset of During the last two decade, the focus has been on the resident ab+ T cell population in the mamma- conventional T cells, leading to important advances lian kidney and are early responders to AKI with in understanding contributions of specificsubsets anti-inflammatory properties that can improve AKI in promoting and suppressing AKI.4–6 Efforts to understand the role of nonconventional T cells in IRI7 have been growing. As a consequence of Received December 11, 2014. Accepted July 15, 2015. 2 2 such efforts, we identified CD4 CD8 (double- Published online ahead of print. Publication date available at negative; DN) cells as an unconventional and un- www.jasn.org. + expected component of ab T cell receptor (TCR) Correspondence: Dr. Hamid Rabb, Ross 965, Johns Hopkins cells in the normal mouse kidney.8 DN ab+ T cells University, 720 Rutland Avenue, Baltimore, MD 21205. Email: are rare in secondary lymphoid organs and are [email protected] + among the least studied subset of ab Tcells.9 Copyright © 2016 by the American Society of Nephrology

J Am Soc Nephrol 27: 1113–1123, 2016 ISSN : 1046-6673/2704-1113 1113 BASIC RESEARCH www.jasn.org outcome. Our findings are of potentially clinical relevance be- shared by DN T cells isolated from lymph node, which were cause we found DN T cells in the “normal” portion of kidneys cycling at rates similar to those of CD4 and CD8 T cells (Figure removed due to renal masses in humans. We conclude that DN 2C). Thus, using two independent approaches, we found that T cells are a prominent T cell subset in mouse and human kidney DN T cells were dividing at an exceptionally high rate in kidneys with potentially important immunoregulatory func- the steady state; this is a unique property not shared by kidney tions that can be harnessed to develop therapies for kidney CD4 and CD8 cells. diseases. Kidney DN T Cells Suppress CD4 T Cell Proliferation We and others have previously shown that DN T cells from RESULTS nonrenal sources, such as the periphery of gld mice, have T cell–suppressive functions.13,14 To determine whether kidney DN T Cells Are a Dominant Subset in Murine Kidneys DN T cells have suppressive functions, we assessed their ability with a Unique Phenotype to inhibit T cell proliferation using a standard T cell suppres- Using a modified protocol,10 we isolated and investigated the sion assay.15 We purified CD4 T cells, labeled them with car- phenotypic and functional characteristics of DN T cells. In boxyfluorescein succinimidyl ester (CFSE), and cultured them extensively flushed kidneys, DN T cells constituted 20%– alone or in the presence of kidney DN T cells. The cocultures 38% of ab+ T cells of normal mice (Figure 1A). In contrast, were activated with CD3/CD28 beads, and proliferation was DN T cells made up only 5%–11% of ab+ T cells in lymph determined after 7 days by measuring CFSE dilution (Figure node and spleen, respectively (Figure 1A) and were rare in pe- 3A) as previously described.13 The proliferation of CD4 T cells ripheral blood (1.060.2) (data not shown). To distinguish our was significantly inhibited in the presence of kidney DN T DN T cell population from NKT cells, we excluded CD1d/ cells, but not by the addition of a similar number of unlabeled PBS57 tetramer+ natural killer T (NKT) cells (5% of the kidney CD4 T cells (Figure 3B). These results demonstrate that kidney T cells8) from our analysis. Phenotypically, kidney DN T cells DN T cells have suppressor function. expressed significantly less CD62L and higher levels of CD44 and CD69 compared with their CD4 and CD8 counterparts. In DN T Cells Are Early Responders to IRI-Induced AKI addition, DN T cells expressed high levels of CD40L and CD28 Given the highly proliferative nature of kidney DN T cells in the co-stimulatory molecules (Figure 1, B and C). These results steady state, we sought to understand how they would respond confirm and extend our previous findings8 that DN T cells to alterations in their microenvironment caused by IRI. We make up a large component of ab+ T cells in the kidney and subjected mice to bilateral IRI and assessed the effect on DN T display an activated phenotype in the steady state. cell homeostasis compared with that of CD4 and CD8 T cells at three time points (3, 24, and 72 hours) (Figure 4A). DN T cells Kidney Resident DN T Cells Actively Proliferate during expanded significantly and became the dominant subpopula- the Steady State tion 3h after IRI. DN T cells maintained dominance until 24 The activated phenotype of DN T cells prompted us to assess hours after IRI, as indicated by their high frequency and whether they are actively proliferating in the steady state. We absolute numbers (Figure 4B). DN T cells then decreased sig- used two independent approaches to assess proliferation of nificantly to below their steady state in both frequency and kidney DN T cells and compare it to that of their CD4+ and CD8+ absolute numbers by 72 hours after IRI. This kinetics was dis- counterparts. In the first set of experiments, we assessed pro- tinct from those of CD4 and CD8 T cells, neither of which liferation by analyzing intracellular expression of the Ki67 anti- displayed significant alterations in their absolute numbers in gen. This nuclear protein is present in all phases of proliferating the current studies. However, their frequencies significantly cells but is absent in G0 phase and widely used as a proliferation decreased because of the significant increases of the absolute marker.11,12 There were more Ki67+ DN T cells (36%) in com- number of DN T cells. Thus, the significant increases in the parison to CD4 (5.6%) and CD8 (1%) subsets (Figure 2A), in- frequencies of DN T cells at 3 and 24 hours after IRI was not dicating steady-state proliferation of kidney DN, but not CD4 at the expenses of CD4 or CD8 T cells. Moreover, 3-hour posti- or CD8 T cells. In the second sets of experiments, we used the schemia DN T cells incorporated more BrdU than DN T cells bromodeoxyuridine (BrdU) incorporation assay as an alterna- at the steady state (Figure 4 C). These results show that DN tive approach to confirm this observation. We injected mice in T cells are highly sensitive to alterations in the renal microen- the experimental group with 1 mg of BrdU twice within a 24- vironment. In addition, they reveal that DNT cells respond to hour period, whereas mice in the control group received PBS. IRI in a potent innate-like manner, resulting in their pre- Kidney T cells were then isolated and the three subsets analyzed dominance during the first 24 hours. for BrdU incorporation by flow cytometry. Consistent with their high expression of Ki67, DN T cells incorporated BrdU at sig- Cytokine Production by Kidney Resident DN T Cells in nificantly higher rate than did the CD4 and CD8 T cells, all the Steady State and in Response to IRI isolated from same kidneys (P,0.0001) (Figure 2B). The high Cytokines are essential mediators of T cell effector functions. proliferative property of kidney DN T cells was not a property Whereas the proinflammatory cytokines such as IFN-g and

1114 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 1113–1123, 2016 www.jasn.org BASIC RESEARCH

Figure 1. DN T cells are the predominant T cell subtype in murine kidneys and have a unique phenotype. Steady-state DN T cell distribution in different tissue compartments, phenotypic characterization, and activation status in WT mice. (A) Left plots show representative data of normal WT mice. Lymphocytes were isolated from lymph nodes, spleen, and kidney; CD45+ cells were gated; and frequency of CD4+,CD8+, and DN T subsets among gated TCRab+ cells were analyzed by ﬂow cytometry. Right graph shows cumulative data from two independent experiments (n=4). (B) Histograms show expression of indicated markers in kidney DN, CD4, and CD8 T cell subsets. Data are from one of three experiments with similar results. (C) Graphs show mean ﬂuorescence intensity (MFI) of indicated markers. Data are from one of three experiments with similar results. Error bars represent SD. *P,0.05; **P,0.01; ****P,0.0001.

TNF-a can cause tissue damage, the anti-inflammatory cyto- and analyzed each subset for IFN-g, IL-27, and IL-10, using kines such as IL-1016 and IL-2717,18 can prevent tissue damage normal kidney DN T cells as control. Global gene expression and maintain immune homeostasis. To determine the major analysis of kidney DN T cells showed that gene encoding IL- cytokines produced by kidney DN T cells and their modula- 10 increased 16-fold, whereas the one encoding IL-27 decreased tion by IRI, we performed global gene expression analysis of by about 2-fold (Figure 5C). The microarray data were vali- DN, CD4+,andCD8+ T cells in the steady state and after dated using quantitative PCR, which showed that expression ischemia-reperfusion (Martina et al. Manuscript in prepara- of IL-10 was significantly increased in DN T cells after IRI (6- tion). IL-27 and IL-10 were significantly expressed in DN T fold), whereas that of IL-27 decreased by about 0.44-fold and cells at the steady state, both at the mRNA and protein levels IFN-g decreasedby0.3-fold(Figure5D).Theresultswere (Figure 5, A and B). In addition, DN T cells expressed lower further confirmed at the protein expression level expression; levels of IL-17A at the protein level (Supplemental Figure 1). IL-10 increased (P=0.02)andIL-27decreased(P=0.003) after To further evaluate the modulation of IL-10 and IL-27 expres- 3 hours of IRI (Figure 5 E and Supplemental Figure 2). We sion by DN T cells in response to IRI, we subjected mice to IRI. measured intracellular cytokines after phorbol 12-myristate Three hours later we sorted kidney CD4, CD8, and DN T cells 13-acetate/ionomycin stimulation because unstimulated cells

J Am Soc Nephrol 27: 1113–1123, 2016 Double-Negative T Cells in Kidney 1115 BASIC RESEARCH www.jasn.org

migrated to the kidney (Supplemental Fig- ure 4). We adoptively transferred DN T cells into two groups of wild-type (WT) B6 mice, 24 hours before IRI. To assess theroleofIL-10,wetreatedmiceinone group with anti–IL-10R mAb; mice in the control group received isotype control antibody. As additional controls, we infused total T cells or no cells (PBS alone) in two groups of recipients and subjected all mice to the same experimental conditions as the mice injected with DN T cells. Mice were monitored for survival and kidney function for 72 hours after IRI with serum creatinine (SCr) measured every 24 hours. Mice that received DN T cells were significantly protected (P#0.01) from AKI in comparison to those that received only PBS or total T cells (Figure 6). There was no mortality in the 10 mice that received DN T cells before AKI. However, 2 of 10 mice died in the PBS group and 2 of 9 mice died in total T cells group. Furthermore, the protective effect of DN T cell transfer was signiﬁcantly abrogated by the co-injection of anti–IL-10R neutralizing mAb as indicated by SCr (6SEM) levels measured in mice that received DN T cells alone or with anti–IL-10R at 24 hours (0.760.2 versus 2.060.2; P#0.01), 48 hours (0.460.1 versus 2.460.4; P#0.001) and 72 hours (0.560.1 versus 2.060.6; P#0.001). These results Figure 2. Kidney resident DN T cells are rapidly proliferating in the steady state. (A) show that DN T cells are protective during Representative dot plots from one of two experiments showing Ki-67 expression by AKI and this protection is IL-10 dependent. gated kidney ab+ T cell subsets, isolated from normal WT mice (n=5 each experiment). (B) Histogram shows BrdU incorporation by kidney T cells from mice that were injected Kidney DN T Cell in Humans with BrdU 24 hours earlier. Graph shows cumulative data from three independent To evaluate whether DN T cells are also experiments (total n=5 mice). Each symbol represents average data per experiment. present in the human kidney, we isolated (C) Comparison BrdU incorporation of peripheral lymph nodes and kidney DN T cells. ab T cells from “normal” tissue of patients ****P,0.0001. with renal cell carcinoma undergoing total or partial nephrectomy, using our modiﬁed protocol for isolation of DN T cells from induced little or no detectable intracellular cytokines (Supple- nonlymphoid tissues.10 Flow cytometric analysis revealed mental Figure 3). that DN T cells were present in each of the examined biopsy specimens and their frequency varied from 18.3% to 61% of Adoptive Transfer of DN T Cells Prevents AKI ab T cells (Figure 7) similar to mouse kidneys. Demographic Given the in vitro ability of kidney DN T cells to suppress CD4 information for these cases (age, sex, diagnosis) and abTCR T cell proliferation (Figure 3) and their production of IL-10 levels are shown in Supplemental Figures 5 and 6, respectively. (Figure 5), we next determined whether DN T cells could inhibit IRI-induced AKI and whether any effect is IL-10 dependent. For these experiments, we used DN T cells isolated from DISCUSSION lymph nodes of gld donors because large numbers of cells were needed for successful adoptive transfer studies. Gld DN T cells, Whereas most efforts to understand the roles of ab+ T cells similar to kidney DN T cells, can suppress T cell proliferation in normal and ischemic kidneys have focused on the CD4, in vitro.13 In addition, adoptively transferred gld DN T cells CD8, CD4+CD25+FoxP3+, and NKT subsets, this study

1116 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 1113–1123, 2016 www.jasn.org BASIC RESEARCH

Figure 3. Kidney DN T cells suppress CD4+ T cell proliferation. (A) Histograms show CFSE dilution under the indicated culture conditions. Numbers indicate percentages of CD4+ that had proliferated, as indicated by CFSE dilution. (B) Graph shows cumulative results. Each symbol represents a single co-culture. Unst, unstimulated. ***P,0.001. demonstrates for the first time that kidney DN ab T cells are interactions with their microenvironment in a manner that major modulators of renal immune responses. We found that is different from that of CD4 and CD8 T cells. They could be DN T cells made up third of all ab+ T cells in the mouse kidney involved in recognition and sensing self-antigens that are not and that they expressed high levels of the CD69, CD28, and recognized by the conventional T cell subsets. DN T cells also CD40L co-stimulatory molecules. Furthermore, unlike CD4 express elevated levels of IL-27 and IL-10. Whereas IL-10 is and CD8 T cells, DN T cells divided at elevated rates in the one of the most potent anti-inflammatory cytokines steady state, uniquely expressed IL-27 and IL-10 and sup- known,19,20 both pro- and anti-inflammatory functions have pressed CD4 T cell proliferation in vitro. The frequency and been ascribed to IL-27.21 In addition, IL-27 can promote IL-10 absolute numbers of DN T cells rapidly increased within 3 production.17 Future studies should determine how DN T hours after IRI, and expansion of DN T cells was associated cells use IL-27 and IL-10 to maintain immune homeostasis with significant upregulation of IL-10 expression. The fre- in the kidney. The high expression levels of CD28 and CD40L, quency and absolute numbers of DN T cells returned to levels two of the major co-stimulatory molecules, could give DN T below normal by 72 hours after IRI. Consistent with their cells advantages in outcompeting conventional T cells for in- upregulation of IL-10 after IRI, DN T cells were protective teractions with local antigen presenting cells (APCs) and sub- during AKI by an IL-10–dependent pathway. These findings stantially influencing both cellular and humoral immune are likely to have clinical relevance because DN T cells were responses in the kidney. For example, CD40L/CD40 can also found in high numbers in human kidney samples. influence the ability of CD4 T cells to regulate activation, B DN T cells constitute 20%–38% of ab T cells in the normal cell class switching, and licensing of DCs22 to activate CD8 T mouse kidney, which is similar to the frequency of CD8 T cells. cells. Furthermore, it will be particularly important in the However, their physiologic role appears to differ from that of future to determine whether the lack of CD4 and CD8 co- CD8 and CD4 T cells. In support of this notion, DN T cells are receptors confers a unique ability that allows DN T cells to constitutively dividing in the steady state, suggesting active interact with nonprofessional APCs, including renal tubular

J Am Soc Nephrol 27: 1113–1123, 2016 Double-Negative T Cells in Kidney 1117 BASIC RESEARCH www.jasn.org

cells. Future studies should examine and distinguish between these possibilities. In addition, it will be important to determine the MHC restriction elements of kidney DN T cells and whether their proliferation is driven by cytokines or in response to autoantigens. The observation that DN T cells are the predominant early ab+ T cell responders to IRI by virtue of their rapid and innate-like expansioninnumberswithinhoursafter the injurious insult is another key finding of this study. Although we cannot formally rule out infiltration, the low frequency of DN T cells in peripheral blood make this possibility unlikely. In addition, during IRI, the reason underlying their loss is unknown but needs to be investigated in the future. Our results also identify kidney DN T cells as an unexpected and significant contributor of IL-10 in the steady state and at the early phase of IRI. IL-10 has been implicated in ameliorating renal tissue injury under different pathophysiologic conditions.27 Consistent with these reports, our results show that the ability of DN T cells to improve the course of ischemia-mediated AKI was IL-10 dependent. Thus, while kidney DN T cells can suppress CD4 T cell proliferation, peripheral gld DN T cells can suppress IRI- induced AKI. Unfortunately, difficulty in obtaining sufficient kidney DN T cells hampered our efforts to directly assess their suppressive function in vivo. However, gld mouse–derived DN T cells, produced in large numbers, also have in vitro suppressive function and readily migrated into the kidney after adoptive transfer (Supplemen- Figure 4. Rapid expansion of DN T cells in response to IRI. (A) Dot plots show frequencies of DN, CD4+, and CD8+ subsets of gated kidney ab T cells at baseline tal Figure 4). Furthermore, IL-10 and FasL (untreated mice) and indicated time points after IRI from one of three independent genes are closely localized on chromosome 2/2 experiments with similar results. (B) Graph shows pooled absolute numbers of DN, 1, precluding the generation of IL-10 / CD4+, and CD8+ T cells from the three experiments (n=10 per group) ***P,0.001. gld double knockout to determine Error bars represent SD. (C) Histogram shows BrdU incorporation by kidney DN T cells whether the role of IL-10 in regulating the from mice at baseline compared with 3 hours after IRI (n=3 mice). FMO, or fluores- suppressive ability of DN T cells is cell au- cence minus one. tonomous. Moreover, DN T cell–derived IL-10 could contribute to the induction of 2 epithelia. Our previous results show that CD4 , but not CD4+, other IL-10–producing cell types,19 including type 1 regulatory T cells are readily activated by staphylococcal superantigen cells and Breg cells, that could subsequently become major sources immobilized on the surface of specific mAbs23 and by the in- during the repair phase after contraction of DN T cells. Kidney testinal epithelial cell line Caco2.24 Expression of CD4+ in cells DN T cells are also a major source of IL-27, a member of the also contributes to the termination of expansion of chronically IL-10 family that has been implicated in IL-10 induction,17 CD4 T cells by predisposing them to activation-induced cell and IL-27–deficient mice develop exacerbated experimental death.25,26 Hence, the lack of co-receptors could potentially be crescentic GN.28 Interestingly, IRI did not enhance but rather vital in allowing continuous proliferation of kidney DN T decreased expression of IL-27 by DN T cells. This finding

1118 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 1113–1123, 2016 www.jasn.org BASIC RESEARCH

Figure 5. Kidney DN T cells express distinct cytokines. (A) Relative mRNA level of indicated cytokines by different T cell subsets measured in the steady state by real-time PCR compared with CD4 T cells (*P,0.05; ***P,0.001). (B) Protein level using intracellular staining. Kidney DN T cells expressed significantly higher levels of IL-27, intermediate expression of IL-10, and low expression of IFN-g in the steady state. (C) Heat map shows changes in cytokines and cytokine receptors in kidney DN T cells in response to ischemia- reperfusion. Arrows show IL-10 and IL-27 cytokines. (D) Relative mRNA level of indicated cytokines by purified DN T cells isolated 3 hours after IRI. The expression of IL-10 by DN T cells was significantly enhanced at 3 hours after IRI (6-fold) but was reduced for IL-27 (0.44-fold) and for IFN-g (0.003-fold). The real-time data were normalized with actin values and compared with DN T samples from normal (no-IRI) mice. (E) Intracellular protein level expression showed a significant increase for IL-10 (*P=0.02) and a decrease in IL-27 (**P=0.003) expression after 3 hours of IRI. Error bars indicate SEM. should be further explored in the future to evaluate its signifi- kidneys surgically excised because of adjacent renal carcino- cance given that IL-27 can exert proinflammatory functions.29 mas. However, we cannot exclude the possibility that the These findings have relevance to humans, as we found a seemingly normal sections may have abnormalities not viewed large fraction of DN T cells in “normal” sections of human histologically. Thus, given their high frequency and unique

J Am Soc Nephrol 27: 1113–1123, 2016 Double-Negative T Cells in Kidney 1119 BASIC RESEARCH www.jasn.org

Figure 5. Continued.

response to IRI in mice, studying mouse and human kidney Isolation of Kidney T Cells DN T cells will likely improve understanding of IRI, with pos- Our protocol for kidney and peripheral T cell isolation was described sible relevance to other organs. This study paves the way for before.10 Briefly, mice were anesthetized with intraperitoneal pento- future studies that will elucidate DN T cell functions and their barbital (75 mg/kg), underwent midline abdominal incisions, were role in autoimmunity and perhaps harness them as a novel exsanguinated, and had their renal pedicles dissected and kidneys therapeutic target. removed. Spleen, kidney draining lymph nodes, and inguinal lymph nodes were removed to use as control. Spleen and lymph node cells were treated with erythrolysis buffer (ammonium-chloride-potassium) CONCISE METHODS to remove red blood cells. Cells then were re-suspended and coun- ted; purity of cells was usually .94% of viable lymphocytes. The absolute number of lymphocytes was calculated by multiplying the Mice + Male C57BL/6J WT mice were bred under specific pathogen-free total number of CD45 cells by the percentage of each subset (de- fl conditions at the central animal facility of the Johns Hopkins termined by ow cytometry). University.Forallexperiments,malemice8–10 weeks of age were used in accordance with the Animal Care and Use Committee guidelines. Antibodies and Reagents All reagents used for media, PBS, and Percoll were obtained from Human Samples Sigma-Aldrich and BD Bioscience. The fluorochrome-conjugated Kidney samples were obtained from partial nephrectomies performed mAb to mouse antigens used for flow cytometry analysis were CD45 because of renal cell carcinoma. Samples were obtained and analyzed (30-F11) APC-Cy7, abTCR (H57–597) Pacific blue, CD8a (53–6.7) at different time points. The kidney tissue was digested according to FITC/PerCP, CD4 (RM4–5) APC/PerCP, CD28 (37.51) PE, CD69 our protocol for isolation of lymphocytes; stained with mAb anti– (1H.2F3) FITC, CD62-L (MEL-14) PECy7 (Biolegend), and CD44 CD45 (HI30) PerCP (Bio Legend), TCRab (IP26) FITC (eBio- (IM7) PE (BD Pharmingen). For intracellular staining mAb, the fol- science), CD8 (HIT8a) PE (eBioscience), and CD4 (OKT4) APC lowing were used: IL27 (MM27–7B1) PE (eBioscience), IL10 (JES5– (eBioscience); and analyzed by flow cytometry. The institutional re- 16E3) PE, IL-17 (TC11–18H10) BV421, and IFN-g (XMG1.2) PE view board approved the experiments. (BD Biosciences).

1120 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 1113–1123, 2016 www.jasn.org BASIC RESEARCH

immunoﬂuorescence staining was analyzed with LSR II using FACS Diva software (BD Biosciences) and analyzed using Flow Jo V10 software (Tristar Software). The lymphocytes were gated using forward- and side-scatter to exclude debris and dead cells, and at least 100,000 events acquired.

BrdU Proliferation Assay and Ki-67 Expression Toassess proliferation in kidney purified T cells, we performed a BrdU assay. Mice were injected intraperitoneally with 1 mg of BrdU in PBS twice within a 24-hour period, followed by analysis of different T cell subsets for BrdU incorporation using the BrdU flow kit protocol (BD Figure 6. Adoptive transfer of DN T cells improves course of AKI. Pharmingen) following manufacturer instructions. Cells were ana- SCr levels were determined at baseline and 24, 48, and 72 hours lyzed using LSR-II cytometry. We also analyzed different T cell subsets after renal ischemia. Adoptive transfer of DN T cells 24 hours for Ki-67 expression using anti–Ki-67 mAb (SolA15) PerCP-eFluor before IRI reduced SCr levels, whereas SCr levels increased in 710 by flow cytometry as described earlier. mice that received no cells (PBS) or total T cells. Results are reported as mean6SEM. **P#0.01 for DN T cell versus WT comparisons; ##P#0.01for DN T versus T cell comparisons. CFSE Proliferation Assay Weused standard functional suppression assay to analyze the ability of purified kidney DN T cells to suppress CD4 T cell proliferation as Intracellular Cytokine Staining previously described.13 We cultured freshly isolated, CFSE-labeled 2 Single-cell suspensions were prepared and stimulated with 5 ng/ml splenic CD4 T cells alone or in the presence of unlabeled CD4+CD25 T phorbol 12-myristate 13-acetate and 500 ng/ml ionomycin (Sigma- cells as a control, kidney DN T cells at ratios of 1:1 and 1:05, stim-

Aldrich) for 4 hours at 37°C in a 5% CO2 humidified atmosphere ulated with anti-CD3/CD28 beads (Dynabeads Mouse T activator; incubator in the presence of Golgi Plug (BD Biosciences). Surface Life Technologies). Proliferation was measured after 7 days by flow staining of stimulated cells was performed with mAb anti-CD45, cytometry. anti-CD8, anti-CD4, and anti-ab+TCR Pacific blue for 30 minutes at 4°C. Cells were then permeabilized with perm/wash solution fol- Mouse Renal Ischemia Model lowed by an additional 30 minutes of incubation with fluorochrome- An established model of renal ischemia-reperfusion in mice was used. conjugated monoclonal antibodies IL-27, IL-10, IL-17A, IFN-g,or Briefly, mice were anesthetized with an intraperitoneal injection of fluorescence minus one–matched control antibodies and analyzed by sodium pentobarbital (75 mg/kg). Following an abdominal medial an LSR II flow cytometer. incision, the renal pedicle was dissected, and a microvascular clamp (Roboz Surgical Instrument, Gaithersburg, MD) was placed on each Flow Cytometry Analysis renal pedicle for 30 minutes. During the procedure, animals were kept Lymphocytes were preincubated with anti-CD16/CD32 Fc-R for 10 well hydrated with warm saline and at a constant temperature (37°C). minutes to minimize nonspecific antibody binding. Cells were then After 30 minutes of ischemia, the clamps were removed and the incubated with various combinations of mAb for 25 minutes at 4°C, wounds were sutured. The animals were allowed to recover with washed twice with FACS buffer, and subsequently analyzed. Six-color free access to food and water.

Figure 7. Detection of DN T cells in human kidneys. Plots show percentages of T cell subsets from four different individuals who underwent partial or total nephrectomies secondary to renal cell carcinoma. CD45+ T cells were gated, and frequencies of DN, CD4+ and CD8+ subsets among gated ab+ TCR cells were determined. Numbers indicate frequency in each quadrant.

J Am Soc Nephrol 27: 1113–1123, 2016 Double-Negative T Cells in Kidney 1121 BASIC RESEARCH www.jasn.org

Adoptive Transfer and Assessment of Kidney Injury REFERENCES Highly purified (.97%) total T cells were isolated from spleens using the Pan T cell Isolation Kit (Miltenyi Biotec), whereas DN T cells 1. Rabb H, Daniels F, O’Donnell M, Haq M, Saba SR, Keane W, Tang WW: (.95% pure) were isolated from lymph nodes of gld mice as recently Pathophysiological role of T lymphocytes in renal ischemia-reperfusion described.30 We transferred (2.53106) DN T cells or conventional ab injury in mice. Am J Physiol Renal Physiol 279: F525–F531, 2000 2. Tadagavadi RK, Reeves WB: Renal dendritic cells ameliorate nephro- T cell into separate groups of 8-week-old C57BL/6 male mice (n=10 toxic acute kidney injury. JAmSocNephrol21: 53–63, 2010 per group), 24 hours before ischemia-reperfusion. Mice in the DN T 3. Dong X, Bachman LA, Miller MN, Nath KA, Griffin MD: Dendritic cells group were divided into two subgroups treated with 100 mgofanti– facilitate accumulation of IL-17 T cells in the kidney following acute IL-10R neutralizing mAb intraperitoneally (1B1.3A clone, BioXcell) renal obstruction. Kidney Int 74: 1294–1309, 2008 or isotype control antibody. The first dose was given 72 hours before 4. Burne MJ, Daniels F, El Ghandour A, Mauiyyedi S, Colvin RB, O’Donnell MP, Rabb H: Identification of the CD4(+) T cell as a major pathogenic surgery, with subsequent injections given 24 hours before and imme- factor in ischemic acute renal failure. J Clin Invest 108: 1283–1290, 2001 diately after the surgery. Blood samples were collected at 0, 24, 48, and 5. Gandolfo MT, Jang HR, Bagnasco SM, Ko GJ, Agreda P, Satpute SR, 72 hours after IRI to measure serum creatinine. Additionally, WT Crow MT, King LS, Rabb H: Foxp3+ regulatory T cells participate in mice that did not receive any cell type (PBS only) were used as control repair of ischemic acute kidney injury. Kidney Int 76: 717–729, 2009 to provide baseline information on IRI-induced AKI. Mortality rate 6. Kinsey GR, Sharma R, Huang L, Li L, Vergis AL, Ye H, Ju ST, Okusa MD: Regulatory T cells suppress innate immunity in kidney ischemia- was determined and survivors were euthanized after 72 hours. reperfusion injury. JAmSocNephrol20: 1744–1753, 2009 7. Li L, Huang L, Sung SS, Lobo PI, Brown MG, Gregg RK, Engelhard VH, Real-Time Quantitative PCR Okusa MD: NKT cell activation mediates neutrophil IFN-gamma production To measure IL-27, IL-10, and IFN-g, mRNA transcripts and total and renal ischemia-reperfusion injury. J Immunol 178: 5899–5911, 2007 RNA from sorted DN and CD4+ or CD8+ T cells (,95% pure) 8. Ascon DB, Ascon M, Satpute S, Lopez-Briones S, Racusen L, Colvin RB, Soloski MJ, Rabb H: Normal mouse kidneys contain activated and CD3 were reverse transcribed using a high-capacity cDNA synthesis kit +CD4- CD8- double-negative T lymphocytes with a distinct TCR rep- (Life Technologies). Forty nanograms of cDNA of each sample was ertoire. J Leukoc Biol 84: 1400–1409, 2008 used for each real-time PCR reaction using gene-specificprimerand 9. Martina MN, Noel S, Saxena A, Rabb H, Hamad AR: Double negative probe sets in the Quant Studio 12 K Flex real-time PCR system. Real- (DN) ab T cells: Misperception and overdue recognition. Immunol Cell time data were normalized with b-actin and analyzed as previously Biol 93: 305–310, 2015 10. Martina MN, Bandapalle S, Rabb H, Hamad AR: Isolation of double described31 and fold change in mRNA expression was calculated us- ab J Vis Exp ΔΔ negative T cells from the kidney. (87): 2014 ing the 2- CT method. 11. Gerdes J, Lemke H, Baisch H, Wacker HH, Schwab U, Stein H: Cell cycle analysis of a cell proliferation-associated human nuclear antigen de- Statistical Analyses fined by the monoclonal antibody Ki-67. JImmunol133: 1710–1715, Data were expressed as mean6SD or mean6SEM using Prism 6 1984 12. Hinkle B, Slepchenko B, Rolls MM, Walther TC, Stein PA, Mehlmann (GraphPad Software); n indicates the number of animals per group. LM, Ellenberg J, Terasaki M: Chromosomal association of Ran during Comparisons were drawn using a two-tailed t test. Comparisons be- meiotic and mitotic divisions. JCellSci115: 4685–4693, 2002 tween multiple groups were performed by a one-way ANOVA test 13. Hamad AR, Mohamood AS, Trujillo CJ, Huang CT, Yuan E, Schneck JP: followed by the Student-Newman-Keuls test where appropriate. Sta- B220+ double-negative T cells suppress polyclonal T cell activation tistical significance was determined as P,0.05. by a Fas-independent mechanism that involves inhibition of IL-2 production. JImmunol171: 2421–2426, 2003 14. Zhang ZX, Yang L, Young KJ, DuTemple B, Zhang L: Identification of a previously unknown antigen-specific regulatory T cell and its mecha- ACKNOWLEDGMENTS nism of suppression. Nat Med 6: 782–789, 2000 15. Tran DQ: In vitro suppression assay for functional assessment of human Methods Mol Biol – This work is supported by National Institutes of Health (NIH) 1R2- regulatory T cells. 979: 199 212, 2013 16. Saxena A, Khosraviani S, Noel S, Mohan D, Donner T, Hamad AR: 1AI095484-01 (to H.R. and A.R.H.) and by a generous gift of Interleukin-10 paradox: A potent immunoregulatory cytokine that has Mr. Rogelio Miro of Panama. We thank Drs. Ben Larman and Daniel been difficult to harness for immunotherapy. Cytokine 74: 27–34, 2015 Wansley from the Department of Pathology, Johns Hopkins Uni- 17. Awasthi A, Carrier Y, Peron JP, Bettelli E, Kamanaka M, Flavell RA, versity, for valuable suggestions and critical reading of the manu- Kuchroo VK, Oukka M, Weiner HL: A dominant function for interleukin fl script.WealsothankDrs. ChristianSchuetz,MathiasOelke,andChiu 27 in generating interleukin 10-producing anti-in ammatory T cells. Nat Immunol 8: 1380–1389, 2007 Yen-Ling from the Department of Pathology-Immunopathology at 18. Troy AE, Zaph C, Du Y, Taylor BC, Guild KJ, Hunter CA, Saris CJ, Artis D: fi Johns Hopkins University for helpful scienti c discussions, and IL-27 regulates homeostasis of the intestinal CD4+ effector T cell pool Wesley Howard and Robin Welsh of the Department of Biophysics and limits intestinal inflammation in a murine model of colitis. JIm- and Biophysical Chemistry and Immunology Graduate Training munol 183: 2037–2044, 2009 ’ Program, respectively at Johns Hopkins University for reading the 19. Saraiva M, O Garra A: The regulation of IL-10 production by immune cells. Nat Rev Immunol 10: 170–181, 2010 manuscript. 20. Ouyang W, Rutz S, Crellin NK, Valdez PA, Hymowitz SG: Regulation and functions of the IL-10 family of cytokines in inflammation and disease. Annu Rev Immunol 29: 71–109, 2011 DISCLOSURES 21. Bosmann M, Ward PA: Modulation of inflammation by interleukin-27. None. J Leukoc Biol 94: 1159–1165, 2013

1122 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 1113–1123, 2016 www.jasn.org BASIC RESEARCH

22. Frentsch M, Arbach O, Kirchhoff D, Moewes B, Worm M, Rothe M, 28. Summers SA, Phoon RK, Ooi JD, Holdsworth SR, Kitching AR: The IL-27 Scheffold A, Thiel A: Direct access to CD4+ T cells specificfordefined receptor has biphasic effects in crescentic glomerulonephritis medi- antigens according to CD154 expression. Nat Med 11: 1118–1124, ated through Th1 responses. Am J Pathol 178: 580–590, 2011 2005 29. Hunter CA, Kastelein R: Interleukin-27: Balancing protective and 23. Hamad AR, Herman A, Marrack P, Kappler JW: Monoclonal antibodies pathological immunity. Immunity 37: 960–969, 2012 defining functional sites on the toxin superantigen staphylococcal en- 30. Mohamood AS, Bargatze D, Xiao Z, Jie C, Yagita H, Ruben D, Watson J, terotoxin B. JExpMed180: 615–621, 1994 Chakravarti S, Schneck JP, Hamad AR: Fas-mediated apoptosis regu- 24. Hamad AR, Marrack P, Kappler JW: Transcytosis of staphylococcal su- lates the composition of peripheral alphabeta T cell repertoire by perantigen toxins. JExpMed185: 1447–1454, 1997 constitutively purging out double negative T cells. PLoS One 3: e3465, 25. Hamad AR, O’Herrin SM, Lebowitz MS, Srikrishnan A, Bieler J, Schneck 2008 J, Pardoll D: Potent T cell activation with dimeric peptide-major his- 31. Mohamood AS, Trujillo CJ, Zheng D, Jie C, Murillo FM, Schneck JP, tocompatibility complex class II ligand: the role of CD4 coreceptor. J Hamad AR: Gld mutation of Fas ligand increases the frequency and up- Exp Med 188: 1633–1640, 1998 regulates cell survival genes in CD25+CD4+ TR cells. Int Immunol 18: 26. Viola A, Salio M, Tuosto L, Linkert S, Acuto O, Lanzavecchia A: Quan- 1265–1277, 2006 titative contribution of CD4 and CD8 to T cell antigen receptor serial triggering. JExpMed186: 1775–1779, 1997 27. Jang HR, Gandolfo MT, Ko GJ, Satpute SR, Racusen L, Rabb H: B cells limit repair after ischemic acute kidney injury. JAmSocNephrol21: This article contains supplemental material online at http://jasn.asnjournals. 654–665, 2010 org/lookup/suppl/doi:10.1681/ASN.2014121214/-/DCSupplemental.

J Am Soc Nephrol 27: 1113–1123, 2016 Double-Negative T Cells in Kidney 1123 CD4 CD8 DNT IL-17A

7.3±2.9 6.4±2.6 4.9±3

SSC 4 8 T D D IL17A IL17A-expressing cell (%) C C DN

Supplemental Figure 1: IL17A protein level using intracellular staining. Kidney DN T cells show low expression of IL17A in the steady state. CD4 CD8 DNT

IFN-y 2.62 3.62 3.04

IL-27

2.8 0.76 1.11

IL-10

5.17 6.72 13.3

Supplemental Figure 2: Intracellular cytokines expression in post IR kidneys. Representative FACS plots showing intracellular cytokines in CD4, CD8 and DN T cells isolated from post-ischemic kidneys. IL10 IL27

PMA No PMA CD4

2.57 2.29

CD8

4.58 4.68

DNT

7.38 9.85

Supplemental Figure 3: Effect of PMA treatment on intracellular cytokine levels. Representative FACS plots showing overlaid panels to demonstrate the effect of PMA treatment on intracellular IL10 and IL27 expression in CD4, CD8 and DN T cells. Recipient #1 Recipient #2 Recipient #3 Recipient #4 Recipient #5

PBS control 2% 7.8% 13% 11% 12% FCS

CFSE

Supplemental Figure 4: Homing of gld DN T cells into the kidney. B6 mice (n= 5) were injected with CFSE-labeled gld DN T cells or PBS and analyzed 24 h later for presence of CFSE-labeled cells in kidney tissue. Each dot plot represents one recipient. Case % DN % CD4 % CD8 % DP Age Gender Diagnosis

HYBRID TUMOR 1 18.3 25.8 50.9 5 63 M (RCC + ONCOCYTOMA) RCC (PAPILLARY) 2 37.2 30.2 29 3.59 69 M

RCC (CLEAR CELL 3 31 27.7 35 6.3 56 F TYPE ) 6 RCC, CLEAR CELL 4 61 15 17.7 68 M TYPE

Supplemental Figure 5: Demographics for human samples. Table shows percentage of human kidney T cell subsets age, gender and diagnosis for each case. 56% 58% 54% 50% SSC

TCR αβ

Supplemental Figure 6: TCR + expression in human kidney T cells. Plots show the expression level of TCR + in lymphocytes from human kidney samples.