Patient-derived renal cell carcinoma organoids for personalized cancer therapy

Abstract

Background: Kidney cancer is one of the most common solid tumors. The advancement of human kidney cancer research and treatment has been hindered by a lack of research models that faithfully recapitulate the diversity of the disease.

Methods: We established an effective three-dimensional culture system for generating kidney cancer organoids from clinical renal cell carcinoma samples. Renal cell carcinoma (RCC) organoids were characterized by H&E staining, immunofluorescence, whole-exome sequencing, RNA sequencing and single-cell RNA sequencing. The use of RCC organoids in personalized cancer therapy was assessed by testing their responses to treatment drugs and chimeric antigen receptor T cells.

Results: Using this organoid culture system, 33 kidney cancer organoid lines from common kidney cancer subtypes, including clear cell renal cell carcinoma (ccRCC), papillary renal cell carcinoma (pRCC), and chromophobe renal cell carcinoma (chRCC), were generated. RCC organoids preserved the histological architectures, mutational landscapes, and transcriptional profile of the parental tumor tissues. Single-cell RNA-sequencing revealed inter- and intra-tumoral heterogeneity in RCC organoids. RCC organoids allowed for in vitro drug screening and provided a tool for assessing the efficacy of chimeric antigen receptor T cells.

Conclusions: Patient-derived RCC organoids are valuable pre-clinical models for academic research and personalized medicine.

Keywords: drug screening; organoids; personalized medicine; renal cell carcinoma.

FIGURE 1

Establishing a biobank of patient‐derived renal cell carcinoma (RCC) organoids: (A) overview of experimental design; (B) RCC organoid formation efficiency in basal medium (BM) and modified medium (with each component individually omitted from the BM), shown are bright‐field images of RCC organoids formed after 2 weeks of culture in indicated media. Scale bar, 100 μm; (C) pie chart showing the subtypes of established 33 RCC organoids in this study; (D) representative haematoxylin–eosin (H&E) staining images of RCC tumour tissue (top row) together with the bright‐field microscopy images (middle row) and H&E staining images (bottom row) of corresponding RCC organoids. Scale bar, 50 μm

FIGURE 2

Patient‐derived renal cell carcinoma (RCC) organoids preserve histopathological characteristics of parental tumours. Representative immunofluorescence staining images of paired RCC tumours and organoids (_T, tumour; _O, organoids) for alpha‐methylacyl CoA racemase (AMACR), Cytokeratin 7 (CK7), CD10, PAX2, E‐cadherin, CK8/18, vimentin and Ki‐67. Nuclei were stained with 4′,6‐diamidino‐2‐phenylindole (DAPI) (blue). Scale bar, 50 μm

FIGURE 3

Renal cell carcinoma (RCC) organoids recapitulate the genetic alterations in the parental tumours: (A) the somatic genomic landscape of 16 RCC organoid lines (_O) and the corresponding parental tumours (_T). The types of genetic alterations are indicated in the legend: (B) proportions of base substitutions in RCC organoids (_O) and parental tumours (_T); the six types of base substitutions are represented: (C) DNA copy number alterations in RCC organoids (_O) and tumour tissues (_T).

FIGURE 4

Transcriptomic analysis of renal cell carcinoma (RCC) organoids: (A) Heat map showed the differentially expressed genes between RCC tissues and organoids. Genes with |log₂FC| > 1 and adjust p < .05 were presented. A total of 1353 genes and 491 genes were presented in the upper and lower panel, respectively; (B and C) boxplot showed the top 10 significantly enriched pathways in RCC tissues (B) and organoids (C) using DEGs in DAVID database; (D) Uniform Manifold Approximation and Projection (UMAP) plot of the RNA sequencing (RNA‐seq) data of RCC organoids and tissues; (E) boxplot showed the gene expression correlation between tumour–organoid pairs, random tumour–organoid pairs or random tumour–tumour pairs; (F) UMAP plot of the RNA‐seq data from 16 RCC samples which successfully formed organoids and 10 RCC samples that failed to derive tumour organoids; (G and H) GSEA plot showed the enrichment of cancer‐associated pathways (G) and metabolism/adhesion‐associated pathways (H) between RCC tumours which successfully formed organoids and those failed to derive tumour organoids.

FIGURE 5

Analysis of cellular heterogeneity in clear cell renal cell carcinoma (ccRCC) organoids by single‐cell RNA sequencing: (A) tSNE plot of 14217 cells from 3 RCC organoid lines. Each dot represents one single cell coloured by cluster identity; (B) heat map showed the expression of marker genes for each subcluster calculated using roc algorithm in FindMarkers module; (C) heat map showed the enrichment of hallmark pathways in each subcluster calculated using VISION; (D) heat map of cell‐type‐specific ligand–receptor interactions inferred by CellPhoneDB. Circle size indicates the significance of interactions and circle colour indicates the mean expression of receptor and ligand genes for each pair; (E) Kaplan–Meier analysis of overall survival (OS) in TCGA cohorts separated by M5 signature using SingleR script

FIGURE 6

Drug screening in patient‐derived renal cell carcinoma (RCC) organoids: (A) heat map of logIC₅₀ values for 24 compounds tested on 16 RCC organoid lines; (B and C) dose–response curves for RCC organoids treated with the indicated chemotherapy drugs (B) and mTOR inhibitors (C). Each data point represents the mean of three biological replicates (organoids from different passages), with error bars representing ± standard error of the mean (SEM); (D) representative scatterplots of 1‐AUC (area under the curve) values for two biological replicates of the drug screening data, highlighting drugs (red) having an obvious inhibitory effect on viability (1‐AUC > .5 for both biological replicates) of indicated organoid lines; (E) representative scatterplots of 1‐AUC from drug screening data of paired drugs with the same nominal targets. Each data point represents three biological replicates, with error bars representing ± SEM

FIGURE 7

Modelling immunotherapy with coculture of renal cell carcinoma (RCC) organoids and chimeric antigen receptor (CAR)‐T cells: (A) The structure of CD70‐specific CAR; (B) the expression level of CD70 in one normal kidney tissue–organoid pair N‐10, and three RCC tumour–organoid pairs clear cell renal cell carcinoma (ccRCC)‐24, ccRCC‐25 and chromophobe renal cell carcinoma (chRCC)‐1 by immunohistochemistry. Scale bar, 50 μm; (C) quantification of the production of tumour necrosis factor (TNF)‐α and interferon (IFN)‐γ by ELISA at 2 days after coculture of RCC organoids or normal kidney organoids with CD70 CAR‐T cells, CD19 CAR‐T cells or control CAR‐T cells; (D) the level of cleaved‐caspase‐3 in RCC or normal kidney organoids after coculture with CD70 CAR‐T cells, CD19 CAR‐T cells or control CAR‐T cells for 2 days; (E) quantification of the percentage of CFSE (carboxyfluorescein succinimidyl ester)‐labelled T cells after incubation with RCC or normal kidney organoids for 3 days. Values represent mean ± standard error of the mean (SEM) (n = 3). ns, not significant. **p < .01; ***p < .001 by two‐tailed, unpaired t‐test