Transplantation-Induced Ischemia-Reperfusion Injury Modulates Antigen Presentation by Donor Renal CD11c⁺F4/80⁺ Macrophages through IL-1R8 Regulation

Significance Statement

Renal macrophages are key cells in controlling processes related to inflammation or repair after ischemia-reperfusion injury. Although macrophages from a donor kidney could also guide adaptive immune responses against renal tissue by virtue of their ability to act as antigen-presenting cells, data are lacking on whether donor-derived renal macrophages can function in this manner after being subjected to transplant-induced ischemia-reperfusion injury. The authors demonstrate in mice that such injury is sufficient to dampen donor renal macrophages’ ability to present antigens, skewing them toward a proreparative phenotype. Donor renal macrophages lacking IL-1R8 failed to orchestrate tissue repair, indicating that IL-1R8 is a key regulator of this shift. IL-1R8 thus represents a pathway that merits exploration in terms of modulating responses against autoantigens and alloantigens after kidney transplant.

Visual Abstract

Abstract

Background

In donor kidneys subjected to ischemia-reperfusion injury during kidney transplant, phagocytes coexpressing the F4/80 and CD11c molecules mediate proinflammatory responses and trigger adaptive immunity in transplantation through antigen presentation. After injury, however, resident renal macrophages coexpressing these surface markers acquire a proreparative phenotype, which is pivotal in controlling inflammation and fibrosis. No data are currently available regarding the effects of transplant-induced ischemia-reperfusion injury on the ability of donor-derived resident renal macrophages to act as professional antigen-presenting cells.

Methods

We evaluated the phenotype and function of intragraft CD11c⁺F4/80⁺ renal macrophages after cold ischemia. We also assessed the modifications of donor renal macrophages after reversible ischemia-reperfusion injury in a mouse model of congeneic renal transplantation. To investigate the role played by IL-1R8, we conducted in vitro and in vivo studies comparing cells and grafts from wild-type and IL-R8–deficient donors.

Results

Cold ischemia and reversible ischemia-reperfusion injury dampened antigen presentation by renal macrophages, skewed their polarization toward the M₂ phenotype, and increased surface expression of IL-1R8, diminishing activation mediated by toll-like receptor 4. Ischemic IL-1R8–deficient donor renal macrophages acquired an M₁ phenotype, effectively induced IFNγ and IL-17 responses, and failed to orchestrate tissue repair, resulting in severe graft fibrosis and aberrant humoral immune responses.

Conclusions

IL-1R8 is a key regulator of donor renal macrophage functions after ischemia-reperfusion injury, crucial to guiding the phenotype and antigen-presenting role of these cells. It may therefore represent an intriguing pathway to explore with respect to modulating responses against autoantigens and alloantigens after kidney transplant.

The ischemia-reperfusion injury (I/RI) that inevitably occurs during kidney transplantation represents an essential trigger for the innate immune system, which plays a pivotal role in the development of renal tissue damage and fibrosis.¹ Antigen presentation by cells of the innate immune system initiates and propagates the activation of adaptive immunity, precipitating the effector T cell allogeneic response toward nonself HLA molecules and favoring autoimmunity against cryptic self-antigens.^2,3

Early innate immune responses after I/RI include effects mediated by the mononuclear-phagocyte system, including professional antigen-presenting cells such as macrophages (MФs) and dendritic cells (DCs), which are aimed at sensing danger and maintaining tissue homeostasis.⁴ In animal models, I/RI causes a massive recruitment of monocytes into the kidney,⁵ which is followed by maturation of monocyte-derived DCs⁶ and by a rapid increase in the frequency of intragraft MФs.⁷ These effects are further potentiated by allogeneic antigen pressure from the graft, with increasing numbers of infiltrating DCs⁸ and induction of MФ-mediated inflammation and cytotoxicity,⁹ which concur in causing graft damage and transplant rejection both directly and through effector T cell activation.

Most of renal macrophages (rMФs) coexpress the F4/80 and CD11c molecules, surface markers originally described as exclusive of MФs or DCs, respectively.^10,11 F4/80⁺CD11c⁺ MФs from kidneys subjected to I/RI have been shown to upregulate class II MHC and costimulatory molecules, thereby increasing their antigen-presentation potential, and to secrete proinflammatory cytokines such as TNFα.^12,13 Consistently, clodronate-induced depletion of rMФs before I/RI induction has been shown to reduce the extent of injury.¹⁴ However, the same depletive agent increased the severity of renal lesions when administered during the I/RI recovery phase,⁷ suggesting that different subsets of MФs may mediate opposing functions. Indeed, MФs during the initial phase of I/RI display a proinflammatory polarization (M₁), which is gradually replaced over time by a proreparative phenotype (M₂).^15,16

Until recently, the majority of studies on rMФs did not distinguish between tissue-resident ones (i.e., donor-derived MФs transplanted along the graft) and those deriving from cells infiltrating de novo after I/RI (i.e., recipient cells), whose functions may substantially differ. Indeed, recipient monocytes that infiltrate the kidney at reperfusion have been shown to differentiate into proinflammatory MФs or mono-DCs.^8,17,18 On the other hand, mounting evidence suggests that tissue-resident rMФs, self-renewing cells derived from embryonic yolk sac or fetal liver progenitors,¹⁹ acquire a proreparative M₂ phenotype during ischemic injury, supporting angiogenesis, producing anti-inflammatory cytokines and counteracting TGFβ-induced fibrosis.^20,21

To date, however, no study has evaluated the effect of I/RI on the phenotype and ability of resident rMФs to act as professional antigen-presenting cells. Because antigen-dependent T cell activation plays a central role in mediating I/RI^22,23 and allo-/autoimmunity after kidney transplantation, we tested the hypothesis that I/RI would modulate the phenotype of donor rMФs and their ability to process and present antigens to T cells.

Methods

Animals

Mouse strains used for the experiments included C57BL/6, C57BL/6-Ly5.1, Balb/c, C57BL/6-Tg(TcraTcrb)1100Mjb/J (C57BL/6 mice possessing CD8⁺ T cells all specific for ovalbumin, OVA, peptide 254–267-H-2K^b complex; OT-1 mice), and C57BL/6-Tg(TcraTcrb)425Cjb/J (C57BL/6 mice possessing CD4⁺ T cells all specific for OVA peptide 323–339-H-2A^b complex; OT-2 mice), all purchased from Charles River Laboratories (Calco, Italy); and C57BL/6-Tg(CAG-OVA)916Jen/J, obtained from Jackson Laboratory. C57BL/6-IL-1R8^−/− mice were kindly provided by A. Mantovani and C. Garlanda.²⁴ In all experiments, mice were 8–10 weeks old. All C57BL/6 mice used expressed the CD45.2 variant except the C57BL/6-Ly5.1 mice, the recipients in transplant experiments, which expressed the CD45.1 variant.

Procedures involving animals and their care were conducted in conformity with the current laws, regulations, and policies governing the care and use of laboratory animals, including the Italian Governing Law (D.lgs 26/2014; Authorization n.19/2008-A, issued March 6, 2008 by Ministry of Health), the National Institutes of Health Guide for the Care and Use of Laboratory Animals (2011 edition), and European Union directives and guidelines (EEC Council Directive 2010/63/UE).

Experimental protocols were approved by the local Institutional Animal Care and Use Committee at Istituto di Ricerche Farmacologiche Mario Negri Istituto di Ricovero e Cura a Carattere Scientifico (IRCCS). Animals were housed in the Institute’s animal care facilities and regularly checked by a certified veterinarian, responsible for health monitoring, animal welfare supervision, experimental protocols, and procedure revision.

In vitro Cold Ischemia Experiments

For cold ischemia (CI) experiments, kidneys were harvested from C57BL/6 or IL-1R8^−/− C57BL/6 mice and subjected to 16 hours CI time in Ringer lactate solution. The contralateral kidney was not subjected to CI and served as internal control for every experiment. After isolation, the phenotype and immunostimulatory activity of renal phagocytes were evaluated. The experimental design has been summarized in Supplemental Figure 1A.

Reversible Post-Transplant I/RI Model

The left kidney from either C57BL/6 or IL-1R8^−/−C57BL/6 mouse (both strains expressing the CD45.2 variant) was isolated, flushed with ice-cold heparinized saline, and harvested. The kidney graft was subjected to variable degrees of CI and transplanted into a congeneic C57BL/6-Ly5.1 recipient mouse (expressing the CD45.1 variant). Recipient mice underwent left-sided nephrectomy and orthotopic implant. The right naïve kidney was then removed at the end of the procedure. Renal graft function was monitored by measurement of BUN levels with a reflotron test (Roche Molecular Biochemical). Recipient mice were euthanized at 1, 3 or 4, 7–10, or 30 days post-transplant, and kidneys were processed for all of the evaluations. The full experimental design has been summarized in Supplemental Figure 1B.

Preparation of Renal Mononuclear Phagocytes

Renal mononuclear phagocytes were isolated from kidneys as previously described.²⁵ Briefly, kidneys were finely minced and then digested with collagenase (Roche Diagnostic) and DNAse I; thereafter, tubular fragments from digested kidneys were removed by filtration. Vitality, assessed by Trypan blue dye exclusion, was on average >95%. For phenotypic analyses, cells were stained with fluorochrome-labeled antibodies as described below.

For mixed-lymphocyte reaction (MLR) experiments, CD11c⁺ cells were enriched using anti-CD11c–specific mAb-labeled microbeads (clone N418, MiniMacs; Miltenyi Biotec); magnetic bead separation was carried out according to manufacturer’s instructions. Purity was typically 85%–90% of total cells, and >95% of the purified renal CD11c⁺ cells coexpressed F4/80 (Supplemental Figure 2, A and B). In selected experiments, CD11c⁺F4/80⁻ or CD11c⁻F4/80⁺ cells were also isolated through positive and negative selection using anti-CD11c mAb-labeled microbeads and APC-conjugated F4/80 mAb (clone Cl:A3–1; Bio-Rad) followed by anti-APC antibody–labeled microbeads.

In transplant experiments, cells obtained after digestion and tubular fragments removal were stained with PE-conjugated anti-CD45.1 antibody followed by specific anti-PE antibody–labeled microbeads to exclude recipient cells. CD45.1⁻ cells were then incubated with anti-CD11c mAb–labeled microbeads. As shown in Supplemental Figure 2C, purified CD45.1⁻CD11c⁺ cells were F4/80⁺ (>95%) and CD45.2⁺ (>99%).

CD11c⁺F4/80⁺ cells isolated from either naïve nonischemic or ischemic or post-transplant kidneys were then analyzed by flow cytometry and used as stimulators in MLR experiments with either allogeneic (Balb/c) or syngeneic (OT-1 or OT-2) T cells.

Multicolor Flow Cytometry

Phenotype analyses were performed on a LSR Fortessa X-20 flow cytometer (BD Biosciences). Total cells obtained after kidney digestion or microbead-enriched cells, as appropriate, were resuspended in FACS buffer (PBS/FBS 2%) and stained for 30 minutes with fluorochrome-conjugated anti-CD45.1 PE (A20; BD Biosciences), anti-CD45.1 AF488 (A20; BioLegend), anti-CD45.2 PE (104; BioLegend), anti-CD45.2 FITC (104; BD Biosciences), anti-CD45.2 BV510 (104; BD Biosciences), anti-F4/80 APC (CI:A3–1; Bio-Rad), anti-CD11c APC-Cy7 (N418; Tonbo), anti-class II MHC I-A/I-E FITC (2G9; BD Biosciences), anti–Ki-67 PE (SolA15; Thermo Fisher), anti-CD206 BV421 (CO68CZ; BD Biosciences), or anti-CD38 FITC (90; eBioscience). After washing, cells were fixed and permeabilized (Intracellular Fixation & Permeabilization Kit; eBioscience) according to the manufacturer’s protocol and stained for 30 minutes with anti-IRF4 AF-488 (3E4; BioLegend) and anti-IRF8 PE (V3GYWCH; Thermo Fisher), or with anti-Ki-67 PE (SolA15; Thermo Fisher). Secondary staining for primary unlabeled antibodies, including anti-CX3CR1 (RBS11; Pantec), anti-erythropoietin receptor (EpoR, M-20; Santa Cruz Biotechnology), anti-toll-like receptor 4 (TLR4, M-300; Santa Cruz Biotechnology), and anti-CD280 (PA5–50956; Thermo Fisher), was performed with FITC-conjugated or PE-conjugated anti-rabbit IgG (Abcam). Similarly, FITC-conjugated anti-goat IgG (Jackson ImmunoResearch Laboratories) was used as secondary antibody after anti–IL-1R8 (Santa Cruz Biotechnology) staining. Multidimensional data were analyzed with FlowJo (FlowJo, Ashland, OR). Counting beads (True Count 123 Beads; eBioscience) were added to the sample before FACS analysis in experiments requiring absolute quantification.

Mixed-Lymphocyte Reaction and ELISpot Assays

CD11c⁺F4/80⁺ cells isolated from nonischemic, post-CI, or post-transplant kidneys were used as stimulators in mixed-lymphocyte reaction (MLR) experiments. Responder T cells were obtained by homogenization of lymph nodes of Balb/c (for the assessment of direct antigen presentation) or OT-2/OT-1 mice (for the assessment of indirect antigen presentation and crosspresentation). Renal CD11c⁺F4/80⁺ cells were used as stimulators in all experiments (CD11c⁺F4/80⁻ and CD11c⁻F4/80⁺ cells were also used when appropriate). Responder cells were resuspended in RPMI/20% FBS (10⁶ cells/ml) and cocultured with renal CD11c⁺F4/80⁺ cells (10:1 responder-to-stimulator ratio) in 96-well U-bottom plates at 37°C in 5% CO₂ for 56 hours, in the presence or absence of LPS (O111:B4 20 μg/ml; Sigma-Aldrich). At the end of the coculture, most of the CD11c⁺F4/80⁺ cells remained adherent to the plates, whereas responder cells were harvested after gentle mixing and added to the ELISpot plate for 16 additional hours at 37°C in 5% CO₂. In MLR experiments with OT-1/OT-2–derived responders, renal CD11c⁺F4/80⁺ cells were exposed to sterile OVA protein (Sigma-Aldrich) either in vitro (100 μg/ml for the whole MLR duration) or in vivo (by intravenous OVA infusion, 1 mg/500 μl saline/mouse, 1 hour before kidney harvesting, as previously described²⁶). As positive controls, renal CD11c⁺F4/80⁺ cells obtained from CAG-OVA mice were used as stimulators.

To assess T cell proliferation, 1 μCi ³H-thymidine was added for the last 16 hours of a standard MLR; thymidine incorporation was measured on washed cells using a scintillation β-counter and results were expressed as counts per minute. To detect the formation of IFN-γ⁺ or IL-17⁺ effector T cell clones, cells were placed in 96-well ELISpot plates (Millipore) precoated with capture anti-IFNγ (551881, ELISpot; BD Biosciences) or anti–IL-17 (3521–2H; Mabtech) antibodies. All assays were conducted according to the manufacturer’s instructions. The resulting spots were counted on a computer-assisted Immunospot image analyzer (Aelvis ELISPOT Scanner System) and results were expressed as the mean value of spots/100,000 cells after subtracting spots in negative controls (usually 0–2 spots).

Histology

Histologic analyses were performed on sections of Dubosq-Brazil fixed paraffin-embedded tissues stained with periodic acid–Schiff, hematoxylin and eosin, or Masson trichrome, using standard methods previously described.^27–29 Tubular damage (atrophy, casts, and necrosis) as well as interstitial fibrotic changes were graded from 0 to 4 (0: no changes; 1: changes affecting <25% of the sample; 2: changes affecting 25%–50% of the sample; 3: changes affecting 50%–75% of the sample; 4: changes affecting >75% of the sample). Biopsies were analyzed by a pathologist blinded to experimental groups.

Immunoperoxidase Analysis of Nitro-Tyrosine and Collagen III

Formalin (for nitro-tyrosine) or Dubosq-Brazil (for collagen III) fixed, paraffin-embedded, 3-µm-thick sections were treated with citrate buffer (10 mM citric acid, pH 6) for antigen retrieval. Sections were incubated with primary antibodies (nitro-tyrosine 1:2000; Millipore; collagen III 1:250; Abcam) overnight at 4°C. Sections were then incubated with a biotinylated secondary antibody (goat anti-rabbit IgG, 1:200 dilution; Vector Laboratories) and signals were developed with DAB-Nickel (Vector Laboratories). Semiquantitative scores (0: absent; 1: faint; 2: moderate; 3: intense) were calculated as a weighted mean.^28,29 At least 20 nonoverlapping fields (×400 magnification) for each section were examined by two blinded investigators.

Immunofluorescence Analysis of B Cells and IgG Deposition

Kidney frozen sections (3 μm) were fixed with acetone, and antigen retrieval was performed with citrate buffer (10 mM citric acid, pH 6). Intragraft B cells were detected by indirect immunofluorescence using a rat anti-CD20 mAb (SA275A11; BioLegend), followed by Cy3-conjugated goat anti-rat IgG (Jackson ImmunoResearch Laboratories). The number of total single positive cells were counted in at least ten randomly selected high-power fields for each animal. For IgG staining, sections were incubated with AF488-conjugated goat anti-mouse IgG antibody (Life Technologies); at least 20 nonoverlapping fields (×400 magnification) for each section were analyzed and semiquantitative score (0: absent; 1: faint; 2: moderate; 3: intense) was calculated as a weighted mean.^28,29

Quantitative Real-Time PCR

Total RNA was extracted from frozen kidneys by homogenization followed by TRIzol extraction (Invitrogen) or RNA isolation kit (Macherey-Nagel Nucleospin). RNA was treated with DNase and reverse transcribed to cDNA with the Superscript II RT enzyme kit (Invitrogen). We performed quantitative real-time PCR on Applied Biosystems 7300 Real-Time PCR System with Power SYBR Green Master Mix and primers specific for sequences of mouse Gapdh (forward: 5′-tca​tcc​ctg​cat​cca​ctg​gt-3′; reverse: 5′-ctg​gga​tga​cct​tgc​cca​c-3′, 50 nM), inducible nitric oxide synthase (iNOS, forward: 5′-aag​agg​aaa​agg​aca​tta​ac-3′; reverse: 5′-ctt​agg​gtc​atc​ttg​tat​tg-3′, 600 nM), TNF-α (forward: 5′-cag​acc​ctc​aca​ctc​aga​tca​tct​t-3′; reverse: 5′-cca​ctt​ggt​ggt​ttg​cta​cga-3′, 300 nM), collagen-1(forward: 5′-gag​cgg​aga​gta​ctg​gat​cg-3′; reverse: 5′-tac​tcg​aac​ggg​aat​cca​tc-3′, 300 nM), neutrophil gelatinase-associated lipocalin (NGAL, forward: 5′-atg​tca​cct​cca​tcc​tgg​tca-3′; reverse: 5′-ctg​tgc​ata​ttt​ccc​aga​gtg​aac-3′, 300 nM), and Ym1 chitinase like protein 3, Chil3, forward: 5′-aga​agc​aat​cct​gaa​gac​acc​at-3′; reverse: 5′-ttc​tat​tgg​cct​gtc​ctt​agc​c-3′, 300 nM). Gapdh served as housekeeping gene. We used the ΔΔ threshold cycle technique to calculate cDNA content in each sample using the cDNA expression in kidney from naive mice as reference (calibrator). Melting curve analysis showed a single dissociation peak for all gene PCR products, confirming the specificity of the reactions. No amplification was found in control reactions without cDNA.

Statistical Analyses

Continuous variables from experimental data were reported as mean±SD or median (interquartile range), as appropriate. Normality of distributions was visually inspected by box-plot representation and formally assessed with the D’Agostino–Pearson or Shapiro–Wilk tests. When indicated, data were log-transformed before analysis.

Between-group differences were analyzed by t test (with or without Welch correction) or Mann–Whitney U test, whereas comparisons of three or more groups were performed with ANOVA or Kruskal–Wallis test, as appropriate; homoscedasticity was tested with the Brown–Forsythe method, and the appropriate correction was applied in case of nonhomogenous variances. Appropriate follow-up tests (Holm–Sidak, Dunnet, or Dunn multiple comparisons tests) were performed in case of significant differences. Paired comparisons were carried out by means of paired t test or Wilcoxon matched-pairs signed-rank test, as appropriate. Statistically significant differences were assumed at 5% level of probability. All analyses were performed with GraphPad Prism version 8 (GraphPad, La Jolla, CA).

Results

CD11c⁺F4/80⁺ MФs are the Most Abundant Renal Mononuclear-Phagocytes and Induce IFNγ and IL-17 Responses

Analysis of CD11c and F4/80 expression on renal leukocytes showed that CD11c⁺F4/80⁺ mononuclear-phagocytes represented the main renal leukocyte population (527,247±147,712 cells/kidney, n=10 mice), accounting for 56±16% of the total CD45⁺ renal cells (Figure 1A). Renal CD11c⁻F4/80⁺ and CD11c⁺F4/80⁻ subsets were comparatively smaller, accounting for 6±3% (75,296±62,055 cells/kidney) and 4±2% of CD45⁺ cells (36,226±18,007 cells/kidney), respectively (Figure 1A).

Figure 1.

Phenotype and immunostimulatory activity of renal mononuclear phagocytes. (A) Representative contour plot for CD11c (APC-Cy7) and F4/80 (APC) expression on pregated CD45⁺ renal single cells. (B) Representative contour plots for IRF4 (FITC) and IRF8 (PE) staining on pregated CD11c⁻F4/80⁺CD45⁺, CD11c⁺F4/80⁺CD45⁺, and CD11c⁺F4/80⁻CD45⁺ renal single cells. (C) Frequency of IFN-γ⁺ and IL-17⁺ alloreactive Balb/c T cells by ELISpot after 2 days of exposure to CD11c⁻F4/80⁺, CD11c⁺F4/80⁺, or CD11c⁺F4/80⁻ cells isolated from C57/BL6 kidneys. Mean±SD of three replicates from one representative experiment, *P<0.05. (D) Frequency of IFN-γ⁺ and IL-17⁺ alloreactive Balb/c T cells by ELISpot after 2 days of exposure to renal CD11c⁺F4/80⁺ cells either unstimulated or LPS-stimulated; *P<0.05. (E and F) Frequency of IFN-γ⁺ (E) and IL-17⁺ (F) OVA-specific OT2-CD4⁺ (T_H1 or T_H17) or OT1-CD8⁺ (T_C1 or T_C17) T cell clones detected by ELISpot after 2 days of exposure to renal CD11c⁺F4/80⁺ cells (OVA-transgenic or wild-type). Wild-type renal CD11c⁺F4/80⁺ cells were exposed or not to OVA, either in vitro (100 μg/ml) during MLR or in vivo by intravenous infusion (1mg/500 μl per mouse). OVA-exposed (both in vitro and in vivo) renal CD11c⁺F4/80⁺ cells were either stimulated with LPS or left unstimulated during the MLR. Values are mean±SD, n=3–4 independent experiments for each condition; *P<0.05. ND, not detectable.

Coexpression of F4/80 and CD11c has sparked the debate on the denomination of renal phagocytes.^10,26,30 Recently, Guilliams et al. identified tissue MФs as IRF8^LOIRF4^LO cells, whereas conventional DCs subtype 1 and subtype 2 were IRF8^HIIRF4⁻ and IRF8⁻IRF4^HI, respectively.³¹ Accordingly, in our experiments, IRF8 and IRF4 expression identified CD11c⁺F4/80⁻ cells as renal DCs, comprising both IRF8^HI conventional DC subtype 1 and IRF4^HI conventional DC subtype 2 (Figure 1B), whereas both CD11c⁻F4/80⁺ and CD11c⁺F4/80⁺ cells coexpressed low levels of IRF8 and IRF4, consistent with an MФ phenotype (Figure 1B, Supplemental Figure 3). Additional phenotypic analyses showed that the majority of CD11c⁺F4/80⁺ renal mononuclear-phagocytes also coexpressed class II MHC, TLR4, CX3CR1, and EpoR (Supplemental Figure 4A).

To evaluate the ability of CD11c⁺F4/80⁺ rMФs to perform direct antigen presentation we carried out allogeneic MLRs using Balb/c T cells as responders. rMФs promoted the expansion of both IFNγ- and IL-17–producing alloreactive T cell clones whereas formation of syngeneic T cell clones was virtually absent (0–2 C57BL6 T cell clones, either IFNγ- or IL-17–producing). Compared with other renal mononuclear-phagocytes, these cells tended to induce a larger number of clones, although comparisons reached statistical significance only for IFNγ assays (Figure 1C).

Because TLR4 engagement by damage-associated patterns plays a central role in the response to I/RI,³² we tested the immunostimulatory activity of CD11c⁺F4/80⁺ rMФs after engagement of TLR4 by LPS. LPS-stimulated rMФs generated significantly higher numbers of both IFNγ⁺ and IL-17⁺ alloreactive T cell clones than unstimulated cells (Figure 1D).

To assess the potential of CD11c⁺F4/80⁺ rMФs to perform indirect antigen presentation and crosspresentation, we evaluated the activation of OVA-specific CD4⁺/ CD8⁺ T cells by rMФs pulsed with OVA, both in vitro and in vivo. CD11c⁺F4/80⁺ rMФs were able to both present and crosspresent soluble antigens, as shown by the expansion of IFNγ⁺ and IL-17⁺ OVA-specific T cell clones (Figure 1, E and F) and by T cell proliferation (Supplemental Figure 4B). Of note, the number of T_H1, T_C1, T_H17, and T_C17 clones induced by OVA-pulsed CD11c⁺F4/80⁺ rMФs were comparable with those generated by OVA-tg CD11c⁺F4/80⁺ rMФs, regardless of whether stimulators were exposed to OVA in vitro during the MLR or in vivo by intravenous OVA infusion²⁶ (Figure 1, E and F).

Stimulation with LPS significantly increased the ability of OVA-pulsed CD11c⁺F4/80⁺ rMФs to generate OVA-specific IFNγ⁺ T_H1 and T_C1 clones (Figure 1E). Conversely, LPS stimulation did not modify the number of OVA-specific IL-17⁺ T_H17 and T_C17 clones induced by these cells (Figure 1F).

CI Does Not Reduce the Number of CD11c⁺F4/80⁺ rMФs but Induces an M₂ Polarization Shift

To evaluate how CI affects rMФs, CD11c⁺F4/80⁺ cells were isolated from C57/BL6 mouse whole kidneys subjected to 16 hours in vitro CI and compared with cells isolated from contralateral nonischemic kidneys. The number of CD11c⁺F4/80⁻ DCs and CD11c⁻F4/80⁺ MФs were almost halved after CI, whereas the amount of vital CD11c⁺F4/80⁺ rMФs did not significantly change (Figure 2A).

Figure 2.

CI injury modifies the phenotype and the immunostimulatory activity of renal CD11c⁺F4/80⁺ MΦs. (A) Cell counts of CD11c⁺F4/80⁺CD45⁺, CD11c⁺F4/80⁻CD45⁺, and CD11c⁻F4/80⁺CD45⁺ renal single cells from kidneys subjected or not to 16 hours of CI. Mean±SD, n=4 independent experiments; *P<0.05. (B) Log-fold change of median fluorescence intensity (MFI) of different markers on CD11c⁺F4/80⁺ rMΦs from kidneys subjected to CI compared with nonischemic kidneys. Mean±SD, n=3–4 independent experiments for each molecule; *P<0.05 versus pre-CI MFI. (C) T cell proliferation (³H-thymidine incorporation) of alloreactive Balb/c T cells after 2 days of exposure to CD11c⁺F4/80⁺ rMΦs isolated from kidneys subjected or not to 16 hours of CI. Values are the three replicates of a representative experiment; *P<0.05. (D) Representative images and results summary of IFN-γ⁺ and IL-17⁺ alloreactive Balb/c T cell frequency assessed by ELISpot after 2 days of exposure to renal CD11c⁺F4/80⁺ rMΦs isolated from kidneys subjected or not to 16 hours CI and used either unstimulated or LPS-stimulated. Mean±SD, n=7 independent experiments for IFN-γ⁺ and n=4 independent experiments for IL-17⁺; *P<0.05. (E and F) Frequency of IFN-γ⁺ (E) and IL-17⁺ (F) OVA-specific OT2-CD4⁺ or OT1-CD8⁺ T cells by ELISpot after 2 days of exposure to CD11c⁺F4/80⁺ rMΦs isolated from kidneys subjected or not to 16 hours CI. Renal CD11c⁺F4/80⁺ rMΦs were exposed to OVA in vivo by intravenous infusion and used either unstimulated or LPS-stimulated in vitro during MLR. Mean±SD, n=3–5 independent experiments for each condition; *P<0.05.

Class II MHC expression was significantly decreased in CD11c⁺F4/80⁺ rMФs isolated from ischemic kidneys, whereas expression of CX3CR1, EpoR, and TLR4 increased after CI. Furthermore, IRF4 showed a trend toward upregulation whereas IRF8 expression significantly decreased in rMФs exposed to CI, consistent with an M₂ polarization shift (Figure 2B, Supplemental Figure 5).

CD11c⁺F4/80⁺ rMФs isolated from ischemic kidneys failed to stimulate alloreactive T cell proliferation and showed a significantly reduced ability to induce both IFNγ⁺ and IL-17⁺ alloreactive T cell clones (Figure 2, C and D). Notwithstanding the increase in TLR4 expression, either small or no increment in the number of IFNγ⁺ and IL-17⁺ alloreactive T cell clones was observed when T cells were cocultured with LPS-stimulated ischemic CD11c⁺F4/80⁺ rMФs (Figure 2D).

Compared with in vivo OVA-pulsed nonischemic cells, ischemic CD11c⁺F4/80⁺ rMФs generated lower numbers of IFNγ⁺ T cell clones (Figure 2E), with a similar, albeit nonsignficant, trend observed also for IL-17⁺ clones (Figure 2F). The ability of in vivo OVA-pulsed ischemic rMФs to induce IFNγ and IL-17 responses did not substantially change after in vitro LPS stimulation (Figure 2, E and F).

rMФs Subjected to CI Show a Blunted Response to TLR4 Engagement associated with IL-1R8 Upregulation

In search of mediators involved in the blunted response of ischemic CD11c⁺F4/80⁺ rMФs to LPS stimulation, we evaluated the expression of IL-1R8, a negative TLR4/IL-1R regulator that limits renal I/RI-induced inflammation and T cell responses against kidney allografts.^33,34 FACS analysis revealed that, along with augmented expression of TLR4, the percentage of IL-1R8⁺ rMФs increased after CI (Figure 3A, Supplemental Figure 6A).

Figure 3.

IL-1R8 expression in CD11c⁺F4/80⁺ rMΦs after CI. (A) Percentages of IL-1R8⁺ cells on CD11c⁺F4/80⁺ rMΦs isolated from kidneys subjected or not to 16 hours of CI. Values from n=3 independent experiments; *P<0.05. (B) Frequency of IFN-γ⁺ and IL-17⁺ alloreactive Balb/c T cells by ELISpot after 2 days of exposure to CD11c⁺F4/80⁺ rMΦs isolated from IL-1R8^−/− C57/BL6 kidneys subjected or not to 16 hours of CI and used either unstimulated or LPS-stimulated. Mean±SD, n=4 independent experiments for each condition; *P<0.05.

To establish whether IL-1R8 played a role in regulating the immunostimulatory activity of ischemic renal CD11c⁺F4/80⁺ rMФs, we repeated our experiments with cells from IL-1R8^−/− mice. At variance with cells from wild-type animals, LPS-stimulated ischemic IL-1R8^−/−CD11c⁺F4/80⁺ rMФs induced the formation of significantly higher numbers of IFNγ⁺ and IL-17⁺ alloreactive T cell clones compared with unstimulated cells (Figure 3B).

Donor CD11c⁺F4/80⁺ rMФs Self-Renew within the Kidney Graft after Transplantation-Induced I/RI

To investigate whether the effects of CI that we observed in vitro would also extend to I/RI in vivo, we decided to set up a reproducible model of reversible post-transplant I/RI, testing three different lengths of CI time (5, 25, and 60 minutes). Mice receiving a kidney graft subjected to 60 minutes of CI showed a severe graft dysfunction as early as 1 day post-transplant and none of them reached day 7 post-transplant, whereas recipients of a kidney graft that had undergone 5 minutes of CI did not show graft dysfunction and consistently survived until the end of the study (30 days post-transplant) (Supplemental Figure 6B). Mice receiving a kidney graft subjected to 25 minutes of CI displayed a transient renal dysfunction at day 1 post-transplant, recovered within 7–10 days, and 80% were alive at day 30 post-transplant. Therefore, all subsequent transplants were preceded by a 25-minute CI time.

To study the fate and function of donor CD11c⁺F4/80⁺ rMФs in transplantation-induced I/RI, kidneys from CD45.2⁺ donors were transplanted into congeneic CD45.1⁺ recipient mice, thus allowing the distinction between donor renal-resident cells and infiltrating cells from the recipient.

The number of renal donor CD45.2⁺ leukocytes decreased the day after transplantation, but the number of donor CD11c⁺F4/80⁺ rMФs did not change. At 3 days post-transplant, donor rMФs diminished but subsequently recovered to pretransplant values, virtually accounting for the whole CD45.2⁺ leukocyte population within the kidney graft at 7–10 days post-transplant (Figure 4A, Supplemental Figure 7A). Similar results were observed when these experiments were repeated using CD45.2⁺IL-1R8^−/− mice as kidney donors (Figure 4B, Supplemental Figure 7B). The percentage of Ki67⁺ donor rMФs increased at 1–3 days post-transplant (Figure 4C), suggesting that proliferation was likely the main driver of the subsequent recovery of intragraft donor rMФs observed at 7–10 days post-transplant.

Figure 4.

Donor CD11c⁺F4/80⁺ rMΦs cell count and phenotype in transplantation-induced I/RI. (A and B) Cell counts of total donor cells (CD45.2⁺), donor rMΦs (CD45.2⁺CD11c⁺F4/80⁺), total recipient cells (CD45.1⁺), and recipient rMΦs (CD45.1⁺CD11c⁺F4/80⁺) in wild-type (IL-1R8^+/+) and IL-1R8^−/− kidney grafts pretransplant (pre-tx) and at days 1, 3–4, and 7–10 post-transplant. Mean±SD, n=3–7 mice at each time point; *P<0.05 versus pre-tx, ND: not detectable. (C) Percentage of proliferating (Ki67⁺) cells on donor rMΦs in wild-type kidney grafts pretransplant and at days 1–3 and 7–10 post-transplant. Mean±SD, n=2–3 mice at each time point; *P<0.05 versus pre-tx. (D and E) Post-transplant log-fold change in median fluorescence intensity (MFI) of selected molecules in donor rMΦs compared with pretransplant values, both in wild-type (IL-1R8^+/+) and IL-1R8^−/− kidney grafts at days 1–3 (D) and 3–7 (E) post-transplantation. Mean±SD, n=2–6 mice for each molecule at each time point; *P<0.05. (F) iNOS, TNFα, and Ym1 mRNA expression in wild-type (IL-1R8^+/+) and IL-1R8^−/− kidney grafts at day 7 post-transplant. Mean±SD, n=3–4 mice; *P<0.05. (G) Semiquantitative scores (0–3) and representative images of nitro-tyrosine staining in wild-type (IL-1R8^+/+) and IL-1R8^−/− kidney grafts at day 7 post-transplant. *P<0.05. Original magnification, ×400.

Recipient CD45.1⁺ leukocytes started to infiltrate the ischemic kidney as early as the day after transplantation and increased up to 7–10 days post-transplant in both wild-type and IL-1R8^−/− kidney grafts (Figure 4, A and B, Supplemental Figure 7, A and B). Of note, recipient CD11c⁺F4/80⁺CD45.1⁺ cells were almost absent or very few at each time point post-transplant.

On the other hand, donor CD11c⁺F4/80⁻CD45.2⁺ DCs and CD11c⁻F4/80⁺CD45.2⁺ MФs decreased over time post-transplant and were gradually replaced by infiltrating recipient CD11c⁺F4/80⁻CD45.1⁺ and CD11c⁻F4/80⁺CD45.1⁺ cells, both in wild-type and IL-1R8^−/− kidney grafts (Supplemental Figure 8A). Notably, more than 85% of pretransplant renal donor CD11c⁺F4/80⁻CD45.2⁺ DCs were found in the spleen of recipient mice at 3–4 days post-transplant (42,168±37,256 cells/spleen, n=4). On the other hand, <5% of pretransplant donor rMФs were found in the spleen of recipient mice at 3–4 days post-transplant (CD11c⁺F4/80⁺CD45.2⁺ rMФs: 28,812±18,646; CD11c⁻F4/80⁺CD45.2⁺ MФs: 3047±2557, cells/spleen, n=4).

IL-1R8 Deficiency Skews Donor CD11c⁺F4/80⁺ rMФs toward the M₁ Phenotype after Post-Transplant I/RI

I/RI was associated with an early (1–3 days post-transplant) downregulation of class II MHC along with an increased expression of TLR4 and EpoR in wild-type donor CD11c⁺F4/80⁺ rMФs, which was not modified by IL-1R8 knockout (Figure 4D). On the other hand, CX3CR1 expression increased in wild-type donor rMФs, but did not change appreciably in IL-1R8^−/− grafts. We also observed a statistically significant difference in IRF4 and IRF8 expression between wild-type and IL-1R8^−/− rMФs after transplantation (Figure 4D). This difference was much more pronounced for IRF8, suggesting an M₁ polarization shift in donor IL-1R8^−/− CD11c⁺F4/80⁺ rMФs. Consistent with these findings, the expression of the M1 marker CD38³⁵ was higher in donor IL-1R8^−/− CD11c⁺F4/80⁺ rMФs compared with wild-type (Figure 4D).

Over a broader time frame (3–7 days post-transplant), CD11c⁺F4/80⁺ rMФs from IL-1R8^−/− donors failed to upregulate the expression of the M₂ polarization markers CD206 and CD280, which were instead increased in wild-type donor rMФs (Figure 4E). In line with the persistence of an M₁-skewed phenotype within IL-1R8^−/− kidney grafts, real-time PCR analysis of renal tissue documented lower levels of Ym1, higher levels of TNF-α, and higher expression of iNOS mRNA compared with wild-type grafts (Figure 4F). Consistently, nitro-tyrosine staining showed a significantly more severe tissue oxidative stress in IL-1R8^−/− kidney grafts (Figure 4G).

IL-1R8^−/− Donor CD11c⁺F4/80⁺ rMФs Fail to Brake the Activation of Adaptive Immunity in Transplantation-Induced I/RI

Donor CD11c⁺F4/80⁺ rMФs isolated from wild-type kidney grafts at post-transplant days 1–3 and used as activators in an allogeneic MLR demonstrated an impaired ability to induce IFNγ⁺ and IL-17⁺ alloreactive T cell clones. IFNγ⁺ clone induction did not significantly recover after LPS stimulation, whereas IL-17⁺ responses were significantly albeit marginally increased. On the other hand, donor IL-1R8^−/−CD11c⁺F4/80⁺ rMФs isolated at the same time points induced IFNγ⁺ and IL-17⁺ alloreactive T cell clones to an extent comparable with cells isolated from normal kidneys. In addition, this immunostimulatory ability significantly increased after LPS stimulation (Figure 5, A and B).

Figure 5.

Donor IL-1R8^−/− CD11c⁺F4/80⁺ rMΦs fail to brake adaptive immunity activation after transplantation-induced I/RI. (A and B) Frequency of IFN-γ⁺ (A) and IL-17⁺ (B) alloreactive Balb/c T cells by ELISpot after 2 days of exposure to rMΦs isolated from wild-type (IL-1R8^+/+) or IL-1R8^−/− kidneys (unstimulated and LPS-stimulated) before transplant (pre-tx) and at days 1–3 post-transplant. Mean±SD of five to ten replicates from two to seven mice at each time point. (C–F) Frequency of IFN-γ⁺ (C–E) and IL-17⁺ (D–F) OVA-specific OT2-CD4⁺ or OT1-CD8⁺ T cell clones after 2 days of exposure to rMΦs isolated from wild-type (IL-1R8^+/+) or IL-1R8^−/− kidneys (unstimulated and LPS-stimulated) before transplant (pre-tx) and at days 1–3 post-transplant. Mean±SD of three to eight replicates from two to three mice at each time point; *P<0.05 versus pre-tx unstimulated; ^#P<0.05 versus pre-tx LPS-stimulated; **P<0.05. ns, not significant.

Consistent with results in the CI model, donor CD11c⁺F4/80⁺ rMФs isolated from wild-type kidney grafts at days 1–3 post-transplantation almost completely lost the capability to generate OVA-specific IFNγ⁺ T cell clones. Exposure of these phagocytes to LPS did not promote the induction of IFNγ⁺ CD4⁺ clones, but it had a small but significant effect on IFNγ⁺ CD8⁺ T cells (Figure 5, C and E). On the other hand, donor IL-1R8^−/− rMФs isolated at the same time point could induce OVA-specific IFNγ⁺ T cell clones to an extent comparable with those obtained from naïve kidneys. Of note, antigen presentation significantly increased after LPS stimulation in this setting (Figure 5, C and E). Independent of LPS stimulation, both wild-type and donor IL-1R8^−/−CD11c⁺F4/80⁺ rMФs isolated at days 1–3 post-transplant induced numbers of OVA-specific IL-17⁺ T cell clones comparable with those obtained with pretransplant rMФs (Figure 5, D and F).

IL-1R8^−/− Kidney Grafts Do Not Regain Tissue Homeostasis after Transplantation-Induced I/RI

Recipients of ischemic IL-1R8^−/− kidney grafts experienced worse graft dysfunction at day 1 post-transplant compared with mice receiving ischemic wild-type kidney grafts, and did not completely recover over the medium (7–10 days post-transplant) and long term (30 days post-transplant) (Figure 6A).

Figure 6.

IL-1R8^−/− kidney grafts do not correctly regain tissue homeostasis after I/RI. (A) BUN at several time points as a measure of renal function in mice receiving a wild-type (IL-1R8^+/+) or IL-1R8^−/− kidney graft subjected to 25 minutes of CI. Mean±SD, n=4–16 mice at each time point; *P<0.05 versus wild-type. (B) Tubulointerstitial injury scores (0–4) of kidney grafts at days 7–10 and day 30 post-transplant, mean±SD, n=3 mice at each time point; *P<0.05. Double arrows indicate the brush border, single arrows indicate the casts, arrowheads indicate cellular debris, and A indicates tubular atrophy. Original magnification, ×200. (C) Semiquantitative scores (0–3) of collagen III staining at day 30 post-transplant. Mean±SD, n=3 mice; *P<0.05. Original magnification, ×200 (×400 in the insert). (D) Fibrotic area scores (0–4) by Masson stain. Mean±SD, n=3 mice. Original magnification ×200. (E) Semiquantitative scores (0–3) of IgG deposition. Mean±SD, n=3 mice; *P<0.05. Original magnification, ×400 (green: IgG, red: wheat germ agglutinin).

Consistent with functional parameters, IL-1R8^−/− kidney grafts displayed a marginally higher expression of NGAL at day 3 post-transplant (Supplemental Figure 8B) and more severe tubular injury 7–10 days post-transplant, which failed to completely recover in the long term (Figure 6B). Persistent tubular damage suggested an improper tissue repair program within IL-1R8^−/− kidney grafts, a hypothesis that was also supported by enhanced collagen III deposition, interstitial fibrosis, and collagen I mRNA expression compared with wild-type kidney grafts at 30 days post-transplant (Figure 6, C and D, Supplemental Figure 8C).

Interestingly, we observed a significantly higher number of infiltrating B cells in perivascular and peritubular areas of IL-1R8^−/− kidney grafts compared with wild-type ones (20±3 versus 9±6 CD20⁺ cells/high power field, n=3 mice; P<0.05) at 30 days post-transplant, which was associated with a diffuse and intense IgG deposition in peritubular capillary areas, virtually absent in wild-type grafts (Figure 6E).

Discussion

Ischemia/reperfusion injury is an unavoidable trigger of sterile inflammation in kidney transplantation. Tissue-resident MФs are key cells in controlling inflammatory and repairing processes after I/RI.³⁶ However, because the ischemic kidney is a source of nonself and tissue-specific antigens, rMΦs could also guide deleterious adaptive immune responses toward the graft by virtue of their strategic localization at the abluminal side of peritubular capillaries, which grants them the ability to monitor transendothelial protein transport³⁷ and to effectively mediate antigen presentation, thus favoring allo- and autoimmunity against the renal tissue.

Our experiments show that rMΦs (CD45⁺CD11c⁺F4/80⁺IRF8^LOIRF4^LO cells) are indeed efficient antigen-presenting cells, able to perform both direct antigen presentation and indirect/crosspresentation of in vivo processed soluble antigens. These phagocytes are powerful inducers of IFNγ⁺ and IL-17⁺ effector T cells, and this activity is further potentiated by the engagement of TLR4, a receptor that binds several different danger molecules released during I/RI.³⁸

After prolonged CI, rMΦs demonstrated improved survival compared with other renal mononuclear-phagocytes. Ischemic rMΦs upregulated CX3CR1 and EpoR expression, factors that could be instrumental in promoting their viability. Indeed, signaling through CX3CR1 is crucial for the survival of arterial MΦs³⁹ and EpoR expression protects renal cells from ischemia-induced apoptosis.⁴⁰ In addition, rMΦs downmodulated MHCII expression after both in vitro CI and in vivo I/RI, consistent with the injury-induced transcriptional reprogramming toward a developmental state recently described by Lever et al.²¹

Most importantly, downregulation of IRF8 and upregulation of IRF4 suggest that ischemic rMΦs were rewired toward an M₂ phenotype after CI,^41,42 in line with recent evidence that IRF4 expression in rMФs is crucial for M₂ MФ polarization and for the restriction of T_H1 cytokine responses after I/RI.⁴³ Consistent with this hypothesis, ischemic rMΦs lost the capability to induce IFNγ and IL-17 T cell responses. Although CI elicited TLR4 upregulation in rMΦs, these phagocytes did not regain their original antigen-presenting ability after TLR4 engagement in vitro. Increased IL-1R8 expression in ischemic rMΦs and higher immunostimulatory activity of ischemic IL-1R8^−/− rMΦs after TLR4 engagement suggest that upregulation of IL-1R8, a negative TLR4 regulator, could be the way rMΦs counteract the excessive stimulation of adaptive immunity in the ischemic kidney.

To evaluate the phenotype and function of donor rMΦs after reversible I/RI, we used a model of congeneic renal transplantation. Compared with models of I/RI obtained by clamping the renal artery,^33,44 this method allowed us to discriminate between renal mononuclear-phagocytes from the donor and those derived from recipient monocytes infiltrating the graft at reperfusion.

The number of donor rMΦs decreased early after transplantation and recovered to pretransplant values within 7–10 days. Although we cannot rule out a possible contribution of differentiating donor CD11c⁻F4/80⁻ and/or CD11c⁻F4/80⁺ to rMΦs rebound, our results suggest that the recovery of rMΦs in the ischemic kidney relies in large part on their self-renewing capability.

Tissue MΦs that reside in organs such as intestine or heart largely depend on the infiltration of circulating monocyte precursors to renew after injury.^17,45 By contrast, very few intrarenal recipient CD45.1⁺ cells coexpressed CD11c and F4/80 at 7–10 days after transplantation, indicating that donor rMΦs did not require infiltration of circulating precursors to recover to pretransplant levels.

Consistent with their phenotype, rMΦs did not leave the kidney after I/RI, as suggested by the extremely low number of CD11c⁺F4/80⁺ cells detected in the recipient spleen. At variance, most of renal donor CD11c⁺F4/80⁻ DCs migrated to the recipient spleen early after transplantation, and were fully replaced in the graft by recipient CD11c⁺F4/80⁻ DCs within 7–10 days post-transplant, as previously described in allogeneic kidney transplantation.⁶

Although IL-1R8 expression did not appreciably affect donor rMΦs survival and recovery, this molecule was crucial to guide their phenotype and antigen-presenting function. At variance with their wild-type counterparts, donor IL-1R8^−/− rMΦs strongly upregulated IRF8 and did not substantially modify their expression of the M₂-related markers CD206 and CD280, thus suggesting a polarization shift toward a proinflammatory M₁ phenotype. IRF8 upregulation in M₁ MΦs has been shown to mediate the transcription of iNOS and IL-12,⁴² with the latter favoring growth and maintenance of IFNγ⁺ and IL-17⁺ effector T cell clones.⁴⁶ Consistently, ischemic IL-1R8^−/− kidney grafts displayed higher iNOS and severe oxidative stress at 7 days after transplantation.

Donor rMΦs from IL-1R8^−/− grafts at early post-transplant time points failed to downregulate direct antigen presentation, inducing high numbers of alloreactive IFNγ⁺ and IL-17⁺ T cell clones. These observations are consistent with our previous results in a mouse model of spontaneously tolerant kidney allotransplantation, in which rejection of IL-1R8^−/− allogeneic kidney grafts was associated with higher numbers of intragraft IFNγ⁺ and IL-17⁺ cells.³⁴ Of note, exposure to IFNγ has been shown to polarize MΦs toward the M₁ phenotype,^41,47 thus induction of IFNγ responses by IL-1R8^−/− donor rMΦs could have in turn promoted and maintained an M₁ polarization in these cells.

Donor IL-1R8^−/− rMΦs also efficiently presented the OVA model antigen and generated large numbers of antigen-specific IFNγ⁺ CD4⁺ and CD8⁺ T cell clones, suggesting that these cells could potentially present novel/cryptic antigens released after I/R injury. IL-1R8^−/− congeneic kidney grafts at 30 days post-transplant displayed a significantly increased infiltration of B cells and IgG deposition in peritubular capillaries. These findings suggest that lack of IL-1R8 expression in rMΦs could expose the ischemic kidney to an increased risk of autoantigen-directed humoral responses, which have been increasingly recognized as a cause of acute and chronic rejection in renal transplant recipients.⁴⁸

Ongoing inflammation and enhanced immune responses in IL-1R8^−/− grafts resulted in dysregulated tissue repair and severe graft dysfunction, with incomplete recovery of tubular atrophy and a higher degree of collagen III deposition and interstitial fibrosis.

Overall, IL-1R8 upregulation in rMΦs protects the kidney from overshooting adaptive immunity after I/R injury, preventing uncontrolled inflammation, tissue fibrosis, and aberrant humoral responses (Supplemental Figure 9).

Following the theory of Matzinger and Kamala, who postulated that tissue-specific factors are responsible for the education of resident mononuclear-phagocytes⁴⁹ to avoid excessive adaptive immunity activation, and for the promotion of appropriate local immune response, IL-1R8 could be the kidney factor that guides balanced immune response toward the ischemic insult.

IL-37 administration, a recently discovered ligand of IL-1R8,⁵⁰ could theoretically harness these protective functions of rMΦs, especially in cases of severe I/RI owing to prolonged ischemia, which are associated with worse fibrosis and increased risk of humoral rejection.^51,52 Further studies will be needed to assess the relevance of this pathway in the setting of allogeneic transplantation.

Disclosures

Dr. Benigni reports personal fees and nonfinancial support from Inception Sciences Canada, Inc., personal fees and nonfinancial support from Janssen Research and Development, LLC, outside the submitted work. Prof. Remuzzi reports personal fees and nonfinancial support from Alexion Pharmaceuticals Inc, personal fees and nonfinancial support from Alnylam, personal fees and nonfinancial support from Boehringer Ingelheim, personal fees and nonfinancial support from Handok Inc, personal fees and nonfinancial support from Inception Sciences Canada, personal fees and nonfinancial support from Janssen Pharmaceutical, outside the submitted work. All of the remaining authors have nothing to disclose.

Supplemental Material

This article contains the following supplemental material online at http://jasn.asnjournals.org/lookup/suppl/doi:10.1681/ASN.2019080778/-/DCSupplemental.

Supplemental Figure 1. Schematic description of in vitro CI experiments (experimental design 1) and post-transplant I/RI experiments (experimental design 2).

Supplemental Figure 2. Expression of CD11c and F4/80 on renal cells before (A) or after (B) microbead enrichment. Expression of CD11c, F4/80, CD45.2, and CD45.1 on microbead-enriched renal cells obtained from kidney grafts (C).

Supplemental Figure 3. Gating strategy for FACS analysis of IRF4 and IRF8 expression in CD11c⁻F4/80⁺CD45⁺, CD11c⁺F4/80⁺CD45⁺, and CD11c⁺F4/80⁻CD45⁺ renal cells.

Supplemental Figure 4. Expression of MHCII, CX3CR1, EpoR, and TLR4 on CD11c⁺F4/80⁺CD45⁺ renal cells (A). IFNγ⁺ OVA-specific T cell clone formation and OVA-specific T cell proliferation of OT1 or OT2 T cells exposed to renal CD11c⁺F4/80⁺ MФs in the presence of OVA (B).

Supplemental Figure 5. Expression of MHCII, CX3CR1, EpoR, TLR4, IRF4, and IRF8 on CD11c⁺F4/80⁺CD45⁺ renal cells obtained from kidneys subjected or not to 16 hours CI.

Supplemental Figure 6. Expression of IL-1R8 on CD11c⁺F4/80⁺CD45⁺ cells obtained from kidneys subjected or not to 16 hours CI (A). BUN concentration in CD45.1⁺ C57/BL6 mice receiving a CD45.2⁺ C57/BL6 kidney graft subjected to 60, 5, or 25 minutes of CI (B).

Supplemental Figure 7. Donor CD45.2⁺ and recipient CD45.1⁺ cells in IL-1R8^+/+ (A) or IL-1R8^−/− (B) CD45.2⁺ kidneys transplanted into CD45.1⁺ recipient mice and harvested at pretransplant or post-transplant days 1, 4, and 7.

Supplemental Figure 8. Cell counts of donor and recipient CD45.2⁺CD11c⁻F4/80⁺ or CD45.2⁺CD11c⁺F4/80⁻ cells, in IL-1R8^+/+ or IL-1R8^−/− kidney grafts at pre- and post-transplant (A). Expression of NGAL (B) and Collagen-1 (C) in IL-1R8^+/+ and IL-1R8^−/− kidney grafts at day 3 (NGAL) or day 30 (Collagen-1) post-transplant.

Supplemental Figure 9. Schematic diagram of possible mechanisms through which IL-1R8 expression in renal rMΦs protects the kidney from overshooting adaptive immunity after I/RI, preventing uncontrolled inflammation, tissue fibrosis, and aberrant humoral responses.