14:00 - 15:30
Poster DGfN 2020
Experimentelle Nephrologie 5 (P186 - P194; LA39 - LA41)
Objective: Dehydration is a common clinical finding and frequent among the elderly or patients with chronic diarrhea. Acute kidney injury frequently develops as a result of a fluid deficit. There is growing evidence that recurrent dehydration can cause chronic kidney disease. The kidney’s response to fluid deprivation is incompletely understood. Having a gene expression atlas of the kidney’s reaction to fluid deprivation at single cell resolution might help to understand biological mechanisms but also to identify biomarkers and therapeutic targets.
Method: We performed single-cell RNA sequencing of dissociated mouse kidneys after 24 hours of water restriction (n=2) and control kidneys (n=2). We assigned cell type information based on known marker genes, and systematically analyzed gene expression differences between baseline and water-restricted animals within different cell types. We furthermore applied a computational approach to spatially sort cells based on gene expression similarities to investigate corticomedullary gene expression profiles.
Results: Our data show cell type-specific differential gene expression in all kidney tubule cells with the most prominent response in collecting duct principal cells (CD-PC). Pathways dysregulated in CD-PC included sodium and water reabsorption, immune system modulation and endoplasmatic reticulum (ER) stress. Pathway activation displayed regional cortico-medullary differences.
Conclusion: Fluid deprivation induces regional and cell type-specific responses in kidney cells. Genes and pathways identified by single-cell transcriptomics comprise biomarkers and therapeutic targets for dehydration-associated pathologies.
Objective: Polyamines are organic polycations that regulate many cell functions including proliferation and differentiation. We and others have previously shown that polyamines are necessary for branching morphogenesis of murine embryonic kidney explants. Evidence has been provided that similar signaling pathways are activated during kidney development and renal repair. Against this background, we examined the hypothesis that different types of renal damage cause identical changes in polyamine homeostasis. For this purpose, we analyzed the expression of genes involved in polyamine metabolism (Odc1, Sms, Srm, Aoc1, Sat1, Paox, Smox) in various animal models of acute and chronic kidney injury.
Method: The renal expression of polyamine regulating enzymes was analyzed by RT-qPCR and RNAScope in the following rodent models of acute and chronic kidney injury: 17 min of renal ischemia followed by 6 hours to 21 days of reperfusion; kidneys from 24 hours to 7 days post-transplantation; rhabdomyolysis-induced kidney injury at 24 h and 48 h after intramuscular injection of 50 % glycerol (0.05 ml per 10 g body weight); daily subcutaneous injections of cyclosporin A (60 mg kg-1) for 6 weeks; hypertensive nephropathy caused by administration of Angiotensin II (3.0 mg/kg per min) for 4 weeks; Streptozotocin-induced diabetes mellitus after 6 weeks.
Results: We identified a common gene expression pattern across the different renal pathophysiologies: Polyamine-degrading enzymes (Aoc1, Sat1) were up-regulated in a nephron-segment specific manner, whereas expression of enzymes involved in polyamine synthesis (Odc1, Sms) was reduced in acute and chronic kidney injury. Preliminary data using cells lines with and without targeted deletion of Srm, Aoc1, Sat1, Paox and Smox by Crispr/Cas9 suggest a role of hypoxia in the expression of some polyamine regulating enzymes.
Conclusion: Our findings indicate that inhibition of polyamine synthesis and stimulation of polyamine breakdown are common to various forms of acute and chronic kidney injury. We therefore conclude that disturbed polyamine homeostasis plays a role in renal damage with different etiologies. Moreover, the polyamine system should be further explored as a potential target for renoprotection.
Objective: Flow cytometric analysis of human urinary cells is a promising diagnostic tool for renal diseases such as Acute Kidney Injury, Renal Transplant Rejection and Lupus Nephritis. As today, due to urine’s potential effects on cell viability and staining quality we rely on fresh urine samples (< 6 hours after voiding) for flow cytometric analysis.
We hereby present a simple two-step preservation method for urine samples allowing delayed sample preparation and analysis.
Method: Our method consists of a simple procedure where we add liquid MOPS buffer and a powdery formaldehyde-releasing agent (Imidazolidinyl urea) to a urine specimen prior to storing it for later processing.
We have performed flow cytometric analysis on urine samples from patients with diverse renal pathologies – including Acute Kidney Injury, Renal Transplant Rejection and Lupus Nephritis – and compared both staining quality and cell numbers of fresh samples (processing < 6 hours after voiding) and samples preserved by our method (processing after 1 day, 3 days and 6 days).
We have analyzed T Cells (CD3+CD4+ or CD3+CD8+, n = 15 after 1 day, n = 10 after 3 and 6 days each), Macrophages (CD14+CD36+, n = 15 after 1 day, n = 10 after 3 and 6 days each) and Tubular Epithelial Cells (Cytokeratin+EpCAM+ or Cytokeratin+CD10+CD13+, n = 9 after 1 day, n = 10 after 3 and 6 days each). For comparing cell counts of samples treated with our preservation method to a reference sample we split every specimen equally. We have compared preserved samples stored for 1 day to fresh samples, whereas preserved samples stored for 3 or 6 days were compared to preserved samples stored for 1 day.
Results: With this preservation method we can perform delayed processing of urine samples without compromising the quality of the analysis. Even after storing samples for 6 days staining quality remains good, still allowing for an efficient gating.
When comparing fresh urine to samples after 1 day of preservation we have found almost identical cell counts and a linear correlation of the cell counts. This also applies when comparing samples after 3 or 6 days of preservation to samples preserved for 1 day.
Conclusion: Our method is an easy-to-use approach allowing a delayed sample processing of human urinary cells without impairing staining quality or altering cell counts. This could help overcome the logistical hurdles of analysing urinary cells and potentially enables centralized and standardised analysis facilitating multicenter studies as well as diagnostic use in a broad spectrum of clinics and outpatient care.
Objective: Decisive for podocyte morphology and homeostasis during health and disease is a specialized and regulated organization of the actin cytoskeleton. In this context, Alpha-actinin4 (ACTN4) has been shown to play a central role in cell architecture and function. Mutations in the ACTN4 gene are associated with focal segmental glomerulosclerosis (FSGS). Here, gene panel sequencing of genes associated with rare kidney diseases was performed in a pediatric patient presenting with steroid resistant nephrotic syndrome (SRNS) and FSGS. Therein, a potential disease causing, de novo variant of ACTN4 was identified, which was not found in available genome or exome databases. Aim of this study was to characterize this variant and elucidate its pathogenic potential for podocyte homeostasis and pathogenesis of FSGS also in comparison with previously reported pathogenic ACTN4 variants.
Method: To elucidate possible toxic effects of the newly identified ACTN4 variant, we employed the genetic toolbox of Drosophila. The fly holds podocyte-equivalent cells called nephrocytes, which are responsible for filtration and detoxification of the hemolymph. Cell-specific genetic manipulation enabled us to analyze the RNAi-mediated knock down of Actinin, the single fly homologue, in nephrocytes and its impact on the cells’ morphology and function. Rescue experiments with the novel human ACTN4 variant will now give indication about toxic effects of the newly identified mutation when compared to a wildtype variant as well as previously described pathogenic variants of ACTN4.
Results: Knockdown of Drosophila Actinin in nephrocytes leads to a severe functional phenotype, as the capacity of hemolymph filtration is diminished by up to 50%. Morphologically, we could show that localization of the ZO-1 homolog Polychaetoid, which is associated with the nephrocyte diaphragm (ND), is severely perturbed. High resolution microscopy of the nephrocyte’s surface even suggests an overall reduction of NDs. First rescue experiments with the wildtype variant of human ACTN4 led to a partial rescue of both, functional and morphological phenotypes seen in Actinin knock down.
Conclusion: Our results underline the importance of Actinin in nephrocyte architecture and function. Capacity of wildtype human ACTN4 in rescuing the Actinin knock down associated phenotypes indicates that the two proteins are interchangeable and underlines the fly model’s suitability. Further experiments will now be performed to elucidate the toxicity of the novel ACTN4 variant also in comparison to previously described pathogenic mutations.
Objective: Prolyl Hydroxylase Domain Enzymes (PHDs) mark the hypoxia-inducible factor 1α (HIF1α) for proteosomal degradation via hydroxylation of prolyl residues within its oxygen-dependent degradation (ODD) domain. Pharmacological PHD inhibitors have nephroprotective effects in experimental kidney diseases and are currently evaluated in phase 3 clinical trials for the treatment of renal anemia. Empirically testing the inhibitory properties of novel PHD inhibitors provides key data for characterizing these substances and further understandung of their effects and potential applications.
Method: The human kidney cell line HEK293 is cultured and transfected with a fusion protein construct consisting of the ODD domain and Luciferase (ODD-Luc). The cells are incubated with several potential PHD inhibitors at various concentrations, stabilising the ODD-Luc construct. PHD inhibition is quantitatively assessed using a high-throughput luciferase assay and HIF stabilisation is verified by Western blotting. Using a ranfe of concentrations, the IC50 value is calculated. Well-established PHD inhibitors are used as reference to estimate a relative inhibition efficiency of novel PHD inhibitors.
Results: So far, four pyridinedicarboxylic acid derivates have been tested for their PHD inhibition properties. The lowest determined IC50 value is at 36µM. The position of the two carboxylic residues has an influence on IC50 values. Furthermore, the PHD inhibition property of a substance is dependent on cellular uptake, which can be modified by esterification.
Conclusion: High-throughput luciferase assays provide a quick and reliable way of testing substances for PHD inhibition. The presented method is a cost and time efficient alternative of Western blot testing the HIF1α stability with the added advantage of quantitation of the IC50 value. The present method will be applied to other cell types in future experiments to investigate cell type specific effects of PHD inhibitors.
Objective: CRISPR/Cas9 has revolutionized genetic disruption and identification of disease-causing genes in Xenopus tropicalis. Nevertheless, phenotyping remains a resource-intensive, and potentially biased process. Here, we adapt recent advances in computer vision, leveraging the U-NET convoluted deep neural network (CNN) towards rapid and unbiased phenotyping of embryonic kidney development in X. tropicalis and to segment embryonic kidneys in mouse in situ sections.
Method: We trained the U-NET network for automated computer-vision-driven segmentation of lectin-stained kidneys in X. tropicalis embryos. To recognize different measures of developmental defects we trained using images showing: normal kidneys, cystic kidneys, hypoplastic kidneys and renal angenesis. A second network was trained to recognize embryonic kidneys on in situ hybridization images of E13.5 mouse sections (genePaint.org). Different parameters (e.g. model learning rate) were adapted to achieve maximum predictive performance of the CNN. Next, we use CNN segmentations to extract features (e.g. kidney size) as distinct parameters inputting into a Python/R pipeline for statistical evaluation and data visualization.
Results: To assess quality of segmentation, we compared the CNN predictions to manually annotated ground truth reporting, among others, accuracy and F1 scores. Both models achieved a reasonably good F1 accuracy, which is currently being improved. Using minimal post-processing filters, embryonic kidneys could be segmented and features could be quantitatively assessed. Segmentation of 100 lectin-stained kidney images takes a human observer around four hours, but three minutes using our CNN approach.
Conclusion: The constant advances of high content screening platforms, the increasing output of clinical genomics in disease-causing variants and the relative ease by which these can be investigated in X. tropicalis (CRISPR/Cas9, overexpression) and other models fuels generation of large image datasets. Unbiased quantitative phenotyping of these disease models remains a bottleneck. Employing CNN for automated scoring of abnormalities occurring during embryonic kidney development in various model systems will aid rapid genotype-phenotype correlation studies.
Objective: Amyloidogenic light chain (AL) amyloidosis is a systemic protein misfolding and deposition disease, which most commonly affects kidney, heart or both. This organ tropism seems to depend on the synergistic effects of light chain (LC) sequence alterations, posttranslational modifications and organ environment. We aim to replicate and elucidate the kidney and heart tropism of AL amyloidosis in vitro and characterise interactions of kidney and heart cells with patient-derived and recombinantly expressed LCs and variable domains (VLs) compared to the corresponding germline peptides.
Method: Here, a combination of cellular assays was used to determine the effects of LCs and VLs on proliferation, cell metabolism and overall cell survival in renal cells. In detail, primary mesangial and proximal tubular cells isolated from rat and human kidneys were incubated with 15µg/ml or 30µg/ml recombinant patient LC or VL. Metabolic activity was measured after 24h by a resazurin reduction assay. Regarding cell proliferation, we analysed immediate effects of LC and VL treatment for 8h on in BrdU-labelling assays. After 72h of coincubation the overall cell death rate was detected by live/dead flow cytometry assay.
Results: So far, recombinant LC and VL from one patient with AL deposits in heart and kidney and the corresponding germline λ LC and VL were tested. The cell death rate was not increased after 72h co-incubation with patient-derived LC compared, while both germline and patient-derived VL seem to be slightly toxic for rat mesangial cells. Interestingly, 30µg/ml patient-derived LC enhances redox activity of mitochondrial reductase in all but human mesangial cells, indicating oxidative stress; VL treatment increased metabolic activity independent of subtype or concentration. Additionally, cells displayed an increase in proliferative activity after addition of LC or VL, which was up to 70% stronger in patient compared to germline LC-treated rat mesangial cells. In cells treated with 30µg/ml VL proliferation was comparable to untreated controls and in tubular cells even marginally inhibited.
Conclusion: In conclusion, we were able to establish a set of assays to replicate organ tropism of patient-derived LCs and VLs, which allows a characterisation of LC toxicity with regards to metabolic activity and cell death rate. This assay system revealed cell type- and species-specific differences in kidney cells and can now be used to compare the toxicity of a different patient LCs, fibrils and VL in cells from other AL amyloidosis-affected organs, such as the heart.
Objective: We demonstrated earlier that renal afferent pathways combine very likely "classical" neural signal transduction to the central nervous system and a substance P (SP) dependent mechanism to control sympathetic activity. SP content of afferent sensory neurons is known to mediate neurogenic inflammation upon release. We tested the hypothesis that alterations in SP dependent mechanisms of renal innervation contribute to experimental nephritis.
Method: Nephritis was induced by OX-7 antibodies in rats, six days later instrumented for recording of blood pressure (BP), heart rate (HR), drug administration; intrarenal administration (IRA) of the TRPV1 agonist capsaicin to stimulate afferent renal nerve pathways containing SP and implantation of electrodes for renal sympathetic nerve recordings (RSNA). FACS assessed the presence of the SP receptor NK-1 on renal immune cells.
Results: IRA capsaicin decreased RSNA from 62.4±5.1 mV*sec to 21.6±1.5 mV*sec (*p<0.05) in controls, a response impaired in nephritis. Suppressed RSNA in nephritic rats and controls transiently but completely recovered after systemic administration of a neurokin 1 (NK1-R) blocker. NK-1 receptors occurred mainly on CD11+ dendritic cells (DCs). An enhanced frequency of CD11c+NK1R+ cell, NK-1 receptor+ macrophages and DCs were assessed in nephritis. Administration of the NK-1R antagonist aprepitant during nephritis reduced CD11c+NK1R+ cells, macrophage infiltration, renal expression of chemokines and markers of sclerosis
Conclusion: Hence, SP promoted renal inflammation by weakening sympathoinhibitory mechanisms while at the same time substance SP released intrarenally from afferent nerve fibers aggravated immunological processes i.e. by the recruitment of DCs.
Objective: Patients with chronic kidney disease (CKD) have a markedly increased incidence of cardiovascular events and mortality compared with the age-matched general population. Among others, the high concentration of circulating uremic toxins in CKD patients leads to vascular inflammatory responses, thereby inducing endothelial dysfunction. During the last years, a correlation between endothelial (dys)function and the mechanical properties of endothelium (endothelial stiffness) could be shown. Here, we hypothesized that uremic serum, derived from CKD patients, changes the nanomechanical properties of endothelial cells, leading to endothelial dysfunction and vascular inflammation. In addition, the influence of CKD on mineralocorticoid receptor (MR) expression and the nitric oxygen (NO) bioavailability was studied, which is linked to endothelial cell function.
Method: We used primary human endothelial cells (HUVEC/HAEC) cultured with 10% pooled sera derived from either healthy or uremic patients undergoing dialysis. Also uremic sera from pre-dialytic children (CKD3 N=9, CKD4 N=8, CKD5 N=7) were employed. The mechanical stiffness of the endothelial cortex was monitored using Atomic Force Microscopy-based nanoindentation. Quantification of cortical actin was done with Phalloidin stainings. MR and eNOS (endothelial NO synthase) expression were analyzed via qPCR and the NO bioavailability was quantified (Griess test).
Results: Uremic serum significantly increased the mechanical stiffness of the endothelial cortex by 18% (Ctr 1.23±0.02 vs. uremic serum 1.45±0.03 pN/nm) which was abolished by application of the MR-antagonist spironolactone (1.17±0.02 pN/nm). This could be confirmed by 22% increase of cortical actin in uremic serum treated HUVEC. An incremental effect of uremic serum on mechanical stiffness corresponding to CKD stage could be shown (Ctr 0.99±0.01; CKD3 1.19±0.01; CKD4 1.22±0.01; CKD5 1.36±0.01 pN/nm). In addition, uremic serum lead to 4-fold MR expression compared to control serum. Uremic serum also significantly decreased eNOS expression and NO bioavailability by 40% and 37%, respectively.
Conclusion: Our data identify the endothelial cortex as an important target of uremic toxins with the MR as possible mediator of uremia-induced endothelial dysfunction. This finding support the proposed beneficial effects of MR inhibition during CKD. Importantly, different CKD stages could be discriminated, suggesting that uremic serum-mediated stiffening of the endothelial cortex might be used as a diagnostic tool for uremic toxins in the near future.
Objective: Autosomal-dominant polycystic kidney disease (ADPKD) is the most common monogenetic disease with a potential lethal outcome. Most patients develop end-stage renal disease (ESRD) during their lifetime and there is no effective pharmacological treatment. Although mutations to PKD1 or PKD2 have been shown to cause ADPKD two decades ago, the function of both genes still remains elusive. Modelling of PKD gene function in animals has expanded our knowledge about ADPKD. However, animal studies are time consuming, not amenable to high throughput screening, and provide limited mechanistic insights. A cellular readout for PKD gene function mimicking impaired tubular morphogenesis observed in ADPKD could overcome these roadblocks.
Method: Wildtype mIMCD-3 cells were seeded in a Matrigel-collagen-scaffold in the presence of hepatocyte growth factor (HGF) in 8-well chamber slides. Cells were incubated for 10 days. Within 10 days cells organized to complex structures (tubules and spheroids, respectively). Structures were imaged with a confocal microscope. Data analysis was performed by training and using a neural network (U-Net). Automated segmentation provided unbiased and rapid quantification.
Results: In a Matrigel-collagen-scaffold, mIMCD-3 cells organized into complex structures. Most of these structures displayed characteristics of tubules (length > width, apicobasal polarity, cilia and lumen).
Automated segmentation showed on average 34.3 tubules/well in wildtype mIMCD-3 cells. In contrast, isogenic PKD1-KO cells showed predominantly spheroids and significantly less tubules compared to wildtype (8.9 tubules/well, p < 0.0001). PKD1-Rescue cells showed 27.5 tubules/well, significantly more compared to PKD1-KO cells (p = 0.0028).
Conclusion: We demonstrate the establishment of a novel 3D kidney tubule organoid assay. In appropriate extracellular matrix, mIMCD-3-cells develop into complex structures with key features of kidney tubules. Moreover, this assay allows to interrogate PKD gene function. PKD1-KO cells develop significantly less tubules compared to wildtype cells. The molecular mechanisms for impaired tubule formation in PKD1-KO cells will be determined and may reveal novel insights to PKD gene function. Furthermore, the establishment of a workflow for automated segmentation enables us to generate and analyze large data sets with high throughput.
Objective: The cold shock Y-box binding protein-1 (YB-1) orchestrates transcriptional and translational programs, coordinates tubular cell proliferation, differentiation, and stress responses and has been reported to regulate abundance of the preprorenin mRNA with potential effects for blood pressure.
Method: Here we seek to clarify YB-1’s role in the aforementioned settings by generating an inducible conditional whole body knockout of Ybx1. Ybx1 deficient animals age normally over 16 months without apparent defects. Our first intervention was a high salt diet (HSD, 4% NaCl in the diet and 1% NaCl in tap water).
Results: Contrary to our expectations, Ybx1-deficient mice showed no change in renin expression or blood pressure. High salt diet led to a significant proteinuria, amino aciduria and glucosuria in wildtype but not YB-1 knockout animals, which was not accompanied by changes of the podocyte slit diaphragm. The tubular cell composition was unaltered within the distal tubule and collecting duct, however, tubular cell phenotypes were strongly affected within the proximal tubulus and accompanied by glomerular tubularization. Sodium/glucose co-transporter-2 as well as megalin expression are downregulated by HSD in wild type animals. In YB-1 knockout animals megalin receptor and SGLT2 are markedly upregulated, whereas cubulin is downregulated.
Conclusion: Thus, we show that Ybx1 deletion does not affect blood pressure and renin secretion; however Ybx1 deficiency fundamentally alters proximal tubular cell phenotypes with skews reabsorption thresholds.
Hintergrund: Mediale Gefäßverkalkungen sind mit der hohen kardiovaskulären Mortalität bei chronischer Niereninsuffizienz assoziiert. Die Verkalkung ist ein aktiver Prozess, der durch Hyperphosphatämie und osteo-/chondrogene Transdifferenzierung von glatten Gefäßmuskelzellen gefördert wird. Diese Studie untersuchte die Rolle des sauren Sphingomyelinase (ASM)/Ceramidsystems bei der Aktivierung pro-kalzifizierender Signalwege in Gefäßmuskelzellen.
Methode: Versuche wurden an humanen und murinen Gefäßmuskelzellen, isoliert perfundierten Gefäßen sowie einem cholecalciferol-induzierten Verkalkungsmodell der Maus durchgeführt.
Ergebnisse: Supplementation von Ceramid oder Sphingomyelinase verstärkte die phosphatinduzierte osteo-/chondrogene Transdifferenzierung und Verkalkung von glatten Gefäßmuskelzellen. Blockade der Serum- und Glukokortikoid-induzierbaren Kinase 1 (SGK1) hemmte die pro-kalzifizierenden Effekte von Ceramid oder Sphingomyelinase. ASM-Defizienz in Gefäßmuskelzellen führte zu einer abgeschwächten osteo-/chondrogenen Transdifferenzierung und Verkalkung nach Phosphatexposition. ASM-Defizienz führte ebenso zu reduzierter Verkalkung in isoliert perfundierten Mausarterien und einem cholecalciferol-induzierten Verkalkungsmodell. Behandlung mit den funktionellen ASM-Inhibitoren Amitryptilin und Fendilin schwächte Gefäßverkalkung in vitro und in vivo ab.
Zusammenfassung: Das ASM/Ceramid System aktiviert über SGK1 pro-kalzifizierende Signalwege in Gefäßmuskelzellen und fördert dadurch die Gefäßverkalkung. Funktionelle ASM Inhibitoren könnten ein neues Konzept zur Reduktion medialer Gefäßverkalkungen darstellen.
Fragen und Antworten der Posterbewerter und Autoren
Nur Posterbewerter und Autoren dieser Postergruppe können Fragen/Antworten erstellen.
Sie müssen sich anmelden, um Kommentare hinzuzufügen.