Variable Expression of Programmed Cell Death Protein 1-Ligand 1 in Kidneys Independent of Immune Checkpoint Inhibition

Context Due to recent advantages in cancer therapy, immune checkpoint inhibitors (ICIs) are new classes of drugs targeting programmed cell death protein 1 (PD-1) or its ligand programmed cell death protein 1-ligand 1 (PD-L1) used in many cancer therapies. Acute interstitial nephritis (AIN) is a potential and deleterious immune-related adverse events (irAE) in the kidney observed in patients receiving ICIs and the most common biopsy-proven diagnosis in patients who develop acute kidney injury (AKI). Based on previous reports, AIN in patients receiving ICIs is associated with tubular positivity for PD-L1, implicating that PD-L1 positivity reflects susceptibility to develop renal complications with these agents. It remains unclear if PD-L1 positivity is acquired specifically during ICI therapy or expressed independently in the kidney. Methods PD-L1 was analyzed in experimental mouse models of ischemia-reperfusion injury (IRI), folic acid-induced nephropathy (FAN), unilateral ureteral obstruction (UUO), and nephrotoxic serum nephritis (NTN) by immunostaining, SDS-PAGE, and subsequent immunoblotting. In addition, we included a total number of 87 human kidney samples (six renal biopsies with AIN related to ICI therapy, 13 nephrectomy control kidneys, and 68 ICI-naïve renal biopsies with various underlying kidney diseases to describe PD-L1 expression. Results We here report distinct PD-L1 expression in renal compartments in multiple murine models of kidney injury and human cases with various underlying kidney diseases, including ICI-related AIN and renal pathologies independent of ICI therapy. PD-L1 is frequently expressed in various renal pathologies independent of ICI therapy and could potentially be a pre-requisit for susceptibility to develop AKI and deleterious immune-related AIN. In addition, we provide evidence that tubular PD-L1 positivity in the kidney is associated with detection of urinary PD-L1+ tubular epithelial cells. Conclusion Our study implicates that PD-L1 is frequently expressed in various renal pathologies independent of ICI therapy and could potentially be a pre-requisit for susceptibility to develop AKI and deleterious immune-related AIN. Because non-invasive detection of PD-L1+ cells in corresponding urine samples correlates with intrarenal PD-L1 positivity, it is attractive to speculate that further non-invasive detection of PD-L1+ cells may identify patients at risk for ICI-related AIN.


INTRODUCTION
Due to recent advantages in cancer therapy, immune checkpoint inhibitors (ICIs) are new classes of drugs targeting programmed cell death protein 1 (PD-1) or its ligand programmed cell death protein 1-ligand 1 (PD-L1, synonym CD274, B7 homolog 1) and are used in many cancer therapies (1). By blocking negative costimulatory pathways, ICIs allow T cells to remain activated thereby enhancing the anti-tumoral immune response. ICIs are now approved for the use in melanoma, small cell and non-small cell lung cancer, renal cell carcinoma and urothelial carcinoma, among others (1). However, despite benefits with respect to progression-free and overall survival, upregulation of the immune system has been associated with a wide spectrum of systemic immune-related adverse events (irAE) (2). Acute interstitial nephritis (AIN) is a potential and deleterious irAE in the kidney observed in patients receiving ICIs and is the most common biopsy-proven diagnosis in patients developing acute kidney injury (AKI) (3). AIN in patients receiving ICIs is associated with tubular positivity for PD-L1, suggesting that PD-L1 positivity reflects susceptibility to develop renal complications with these agents (3,4). Additionally to AIN, there is increasing recognition that glomerular disease and renal vasculitis are associated with ICI-related kidney injury with reports describing both, nephrotic and nephritic clinical presentations (5,6). These observations reveal distinct roles of PD-L1 among different renal compartments dependending on its presence or absence. In the context of ICI-related AKI, PD-L1 expression in distinct renal compartments has not yet been analyzed in human kidneys thus far. Furthermore, it remains still elusive if PD-L1 positivity is acquired specifically during ICI therapy or expressed independently in the kidney. Therefore, we here aimed to compare compartment-specific expression of PD-L1 among ICI-related AIN and ICI-naïve renal pathologies.

Animals
All experimental animal studies were performed with the approval of the Institutional Animal Care and Use Committee of the Beth Israel Deaconess Medical Center and the University Medical Center Göttingen. Experimental protocols are detailed below.

Ischemia-Reperfusion Injury
Eight to 12 weeks old C57BL/6 mice were anesthetized with isoflurane inhalation, and analgesia was performed by subcutaneous buprenorphine injection. The left renal hilus was separated from the surrounding tissues and ischemia was performed for 35 min at 25°C ambient room temperature. Mice were sacrificed at indicated time points after ischemic insult as described previously (7).

Folic Acid-Induced Nephropathy
Kidney injury was induced with a single intraperitoneal injection of folic acid (250 mg/kg body weight in PBS) in CD1 mice. Mice were sacrificed 96 days after iinjection.

Unilateral Ureteral Obstruction
Eight to twelve weeks old CD57BL/6 mice were anesthetized with isoflurane inhalation, analgesia was performed by subcutaneous buprenorphine injection. The ureter was separated from the surrounding tissues and two ligatures were placed about 5 mm apart in the upper two-thirds of the ureter of the left kidney to obtain reliable obstruction. Mice were sacrificed at indicated time points after ureter ligation as described previously (7).

Human Kidney Specimens
A total number of 87 kidney samples (six renal biopsies with AIN related to ICI therapy, 13 nephrectomy control kidneys and 68 ICI-naïve renal biopsies with various underlying kidney diseases) at the University Medical Center Göttingen were included. The use of parts of kidney biopsies for research purposes was approved by the Ethics Committee of the University Medical Center Göttingen (22/2/14). All patients consented to the use of routinely collected samples and data as part of their regular medical care.

Immunohistochemistry
Formalin-fixed, paraffin-embedded kidney sections were deparaffinized in xylene and rehydrated in ethanol containing distilled water. Tissue sections were stained using antibodies against PD-L1 (ab205921, Abcam, Cambridge, UK), labeling was performed using NovolinkTM Polymer Detection System (Leica Biosystems, Wetzlar, Germany) according to the manufacturer's protocol. Nuclear counterstain was performed by using Mayer's Hematoxylin Solution (Sigma, St. Louis, USA).

Western Blot Analyses
Total kidney lysates were homogenized in NP40 lysis buffer (Invitrogen, Carlsbad, USA) supplemented with protease cocktail inhibitor (Roche, Basel, Switzerland). After sonication, proteins were resolved by Tris-acetate-SDS acrylamide gel electrophoresis and transferred onto PVDF membranes (Invitrogen, Carlsbad, USA) and blocked in 5% dry milk in TBS-T (TBS pH 7.6, 0.1% Tween-20). After incubation with respective primary antibodies against PD-L1 (ab238697, Abcam, Cambridge, UK) and b-actin (ab8227, Abcam, Cambridge, UK) were used following incubation with secondary HRP-conjugated antibodies (Dako, Glostrup, Denmark). Luminescence was detected on an X-ray film using chemiluminescent substrate (Cell Signaling, Danvers, USA). Individual lanes represent biological replicates and densitometry was performed using ImageJ software (National Institute of Health, Bethesda, USA), PD-L1 band density was quantified relative to b-actin.

Flow Cytometry
Antibody staining followed standard protocols for antibody staining of cells in suspension. Within 1 h after collection, urine samples were centrifuged at 1,000 rpm at 4°C for 8 min. For analysis of urinary cells and KARPAS 299, cell suspensions were washed and resuspended in PBS and stained using antibodies against PD-L1 (329708, Biolegend, San Diego, USA). Cells were incubated with antibody cocktails for 15 min at 4°C in the dark and gated according to positive and isotype negative controls.

Statistical Methods
Variables were tested for normal distribution using Shapiro-Wilk test. Non-normally distributed continuous variables are expressed as median and interquartile range (IQR), categorical variables are presented as frequency and percentage. Statistical comparisons were not formally powered or prespecified. For group comparisons, the Mann-Whitney U-test was used to determine differences between median values. Spearman correlations were visualized by heatmaps reflecting mean values of Spearman's D, asterisks indicate p<0.05. Data analyses were performed with GraphPad Prism (version 8.4.0 for MacOS, GraphPad Software, San Diego, California, USA).

Experimental Kidney Injury Induces Programmed Cell Death Protein 1-Ligand 1 Expression in Different Renal Compartments
We first analyzed PD-L1 expression in various experimental mouse models of acute kidney injury. By using ischemiareperfusion injury (IRI) as a model of experimental AKI, we found PD-L1 induced in early injury at day 3 and 7, whereas renal recovery at day 10 after ischemia resulted in normalization of PD-L1 expression ( Figures 1A, B and Supplementary Figure  1A Figure 1B), revealing that damage to distinct compartments of the kidney results in PD-L1 induction. Among different renal compartments, tubular PD-L1 is present in experimental models of acute tubular damage (IRI and FAN), whereas glomerular positivity for PD-L1 was most prominent in a model of glomerular injury (NTN, Figure 1E). In summary, we provide evidence that PD-L1 is induced in injured kidneys independently of ICI therapy. Moreover, renal PD-L1 induction depends on the type of kidney injury leading to compartment-specific PD-L1 upregulation.
Programmed Cell Death Protein 1-Ligand 1 Localizes to Different Compartments in Immune Checkpoint Inhibitor-Related Acute Kidney Injury and Immune Checkpoint Inhibitor-Naïve Kidneys Based on our observation that distinct experimental models of tubular and glomerular injury correlate with PD-L1 induction in diseased kidneys, we next included a total number of 87 kidney specimens (6 renal biopsies with AIN related to ICI therapy, 13 nephrectomy control kidneys and 68 ICI-naïve renal biopsies with various underlying kidney diseases, Supplementary Table 1). While all nephrectomy control kidneys were negative, PD-L1 expression was found in all cases of ICI-related AKI as reported previously (Figures 2A-C and Supplementary Figure 2A Table 2). PD-L1 positivity was most prominent in diabetic nephropathy (DN), ANCAassociated vasculitis (AAV) and lupus nephritis (LN), but also focal-segmental glomerulosclerosis (FSGS), acute tubulointerstitial nephritis (AIN) and IgA vasculitis ( Figure 2C). Among different renal compartments, qualitative histological analyses revealed that tubular PD-L1 was present in all renal pathologies positive for PD-L1, including ICI-related AIN and ICI-naïve cases of kidney disease (Figures 2D, E). In contrast, glomerular PD-L1 positivity was only found in ICI-related AIN, DN, AAV and LN ( Figure 2E). Endothelial PD-L1 expresion was least and only observed in ICI-related AIN, DN and AAV ( Figure  2E). Comparison of ICI-naïve renal pathologies revealed that only tubular PD-L1 positivity was found in the majority of cases, followed by PD-L1 positivity within all renal compartments ( Figure 2F). We did not detect isolated endothelial PD-L1, positivity was associated with either glomerular, tubular PD-L1 or both ( Figure 2F). Correlation analysis with clinical and laboratory findings among ICI-naïve renal pathologies revealed that renal PD-L1 upregulation among all different compartments was consistently associated with elevated C-reactive protein (CRP, Figure 2G and Supplementary Table 3). Since elevated levels of CRP have been described previously as an early marker to predict irAE in context of ICI therapy, we next analyzed the CRP course among aforementioned cases of ICI-related AIN (8). Baseline and CRP course until AKI development was available in four cases of ICI-related AIN showing no correlation with onset of AKI related to ICI therapy (Supplementary Figure 2B), limiting its usability as kidney-specific marker for ICI-related AIN. In summary, we found that PD-L1 positivity is present in all cases of ICI-related AIN and 27.9% of ICI-naïve cases, localized to different renal compartments. Interestingly, elevated levels of CRP were associated with intrarenal PD-L1 positivity in ICI-naïve renal pathologies, whereas no correlation was observed with AKI related to ICI therapy.

Kidney-Derived Cells Are Detectable in Corresponding Urine Samples Along With Urinary Programmed Cell Death Protein 1-Ligand 1 + Cells
Since intrarenal PD-L1 positivity could potentially be a prerequisit for susceptibility to deleterious immune-related AIN, we next aimed to directly detect urinary PD-L1 + cells in association with intrarenal PD-L1 positivity. Corresponding urine samples of kidney specimens positive for PD-L1 confirmed presence of kidney-derived cells including proximal (CD10 + ), distal tubular epithelial cells (EpCAM + ) and podocytes (WT1 + ) in addition to PD-L1 + cells within urinary sediment ( Figure 3A). To further validate these observations, we next established flow cytometry for PD-L1 + urinary cells. By using KARPAS 299 as established T cell lymphoma cell line positive for PD-L1, PD-L1 + cells were equally detectable in corresponding urine samples (Supplementary Figure 3A, B) (9). Direct comparison of PD-L1 abundance in the kidney and corresponding urine samples revealed that detection of urinary PD-L1 + cells significantly correlated with tubular, but not glomerular or endothelial PD-L1 positivity in the kidney (Figures 3B, C). This was further supported by histological observation of detached intratubular PD-L1 + epithelial cells in kidney specimens ( Figure 3D). In summary, we provide evidence that tubular PD-L1 positivity in the kidney correlates with detection of urinary PD-L1 + tubular epithelial cells, potentially allowing non-invasive measurement.

DISCUSSION
This is the first report of distinct PD-L1 expression in renal compartments of multiple murine models and human cases with various underlying kidney diseases, including ICI-related AIN and renal pathologies independent of ICI therapy. AIN is a deleterious immune-related adverse effect in the kidney observed in patients receiving ICIs and the most common biopsy-proven Since PD-L1 positivity has previously been experimentally connected with protection from injury in mice, our findings implicate that PD-L1 is equally induced as a protective, antiinflammatory response to injury in diseased human kidneys (13,14). Furthermore, we provide evidence that PD-L1 expression in the kidney is associated with elevated levels of CRP, in line with previous reports that elevated levels of CRP are an early marker to predict irAE in the context of ICI therapy (8). While levels of CRP did not correlate with AKI related to ICI therapy, renal PD-L1 positivity in the kidney correlates with detection of PD-L1 + urinary cells, thereby allowing non-invasive measurement. In addition, ICI therapy may be the only therapeutic option available to effective treat and maintain tumor remission in certain cancers. A key concern is the safety of ICI re-challenge after an episode of AKI. Based on previous reports, recurrence of ICI-related AKI occurred in about 23% of rechallenged patients, with a shorter latency period between the initial AKI episode and rechallenge (15). Therefore, it is also tempting to speculate that measurement of shedded PD-L1 + renal cells may also allow noninvasive monitoring of rechallenged patients to detect recurrence of ICI-related AKI. These assumptions have to be corroborated further.
In summary, our study implicates that PD-L1 is frequently expressed in various renal pathologies independent of ICI therapy and could potentially be a pre-requisit for susceptibility to develop AKI and deleterious immune-related AIN. Since non-invasive detection of PD-L1 + cells in corresponding urine samples correlates with intrarenal PD-L1 positivity, it is attractive to speculate that further non-invasive detection of PD-L1 + cells may identify patients at risk for ICI-related AIN.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the Ethics Committee of the University Medical Center Göttingen. The patients/participants provided their written informed consent to participate in this study. The animal study was reviewed and approved by the Institutional Animal Care and Use Committee of the Beth Israel Deaconess Medical Center and the University Medical Center Göttingen.

AUTHOR CONTRIBUTIONS
BT conceived the study, collected and analyzed the data, and cowrote the first draft. SH, SK, DT, and AG collected and analyzed the data. PK and MZ participated in the construction and editing of the manuscript. SH and SBK contributed equally as first coauthors. All authors contributed to the article and approved the submitted version.

FUNDING
BT was supported by the Research program, University Medical Center Göttingen (1403720 to BT). This work was also supported by grants from the Deutsche Forschungsgemeinschaft (ZE523/2-3 and SFB1002/D03 to MZ) and by equipment grant INST 1525/ 16-1. We also acknowledge support by the Open Access Publication Funds of the Göttingen University. The funding sources had no involvement in the design, collection, analysis, interpretation, writing, or decision to submit the article.