Cell Clusters Research Articles

CRACD loss induces neuroendocrine cell plasticity of lung adenocarcinoma

Tumor cell plasticity contributes to intratumoral heterogeneity and therapy resistance. Through cell plasticity, some lung adenocarcinoma (LUAD) cells transform into neuroendocrine (NE) tumor cells. However, the mechanisms of NE cell plasticity remain unclear. CRACD (capping protein inhibiting regulator of actin dynamics), a capping protein inhibitor, is frequently inactivated in cancers. CRACD knockout (KO) is sufficient to de-repress NE-related gene expression in the pulmonary epithelium and LUAD cells. In LUAD mouse models, Cracd KO increases intratumoral heterogeneity with NE gene expression. Single-cell transcriptomic analysis showed that Cracd KO-induced NE cell plasticity is associated with cell de-differentiation and stemness-related pathway activation. The single-cell transcriptomic analysis of LUAD patient tumors recapitulates that the distinct LUAD NE cell cluster expressing NE genes is co-enriched with impaired actin remodeling. This study reveals the crucial role of CRACD in restricting NE cell plasticity that induces cell de-differentiation of LUAD.

Coarsened Mixtures of Hierarchical Skew Normal Kernels for Flow and Mass Cytometry Analyses

Cytometry is the standard multi-parameter assay for measuring single cell phenotype and functionality. It is commonly used for quantifying the relative frequencies of cell subsets in blood and disaggregated tissues. A typical analysis of cytometry data involves cell classification—that is, the identification of cell subgroups in the sample—and comparisons of the cell subgroups across samples or conditions. While modern experiments often necessitate the collection and processing of samples in multiple batches, analysis of cytometry data across batches is challenging because differences across samples may occur due to either true biological variation or technical reasons such as antibody lot effects or instrument optics across batches. Thus a critical step in comparative analyses of multi-sample cytometry data—yet missing in existing automated methods for analyzing such data—is cross-sample calibration, whose goal is to align corresponding cell subsets across multiple samples in the presence of technical variations, so that biological variations can be meaningfully compared. We introduce a Bayesian nonparametric hierarchical modeling approach for accomplishing both calibration and cell classification simultaneously in a unified probabilistic manner. Three important features of our method make it particularly effective for analyzing multi-sample cytometry data: a nonparametric mixture avoids prespecifying the number of cell clusters; a hierarchical skew normal kernel that allows flexibility in the shapes of the cell subsets and cross-sample variation in their locations; and finally the “coarsening” strategy makes inference robust to departures from the model not captured by the skew normal kernels. We demonstrate the merits of our approach in simulated examples and carry out a case study in the analysis of a multi-sample cytometry data set. We provide an R package for our method.

Identification of Cancer Stem Cell-related Gene by Single-cell and Machine Learning Predicts Immune Status, Chemotherapy Drug, and Prognosis in Lung Adenocarcinoma.

This study aimed to identify the molecular type and prognostic model of lung adenocarcinoma (LUAD) based on cancer stem cell-related genes. Studies have shown that cancer stem cells (CSC) are involved in the development, recurrence, metastasis, and drug resistance of tumors. The clinical information and RNA-seq of LUAD were obtained from the TCGA database. scRNA dataset GSE131907 and 5 GSE datasets were downloaded from the GEO database. Molecular subtypes were identified by ConsensusClusterPlus. A CSC-related prognostic signature was then constructed via univariate Cox and LASSO Cox-regression analysis. A scRNA-seq GSE131907 dataset was employed to obtain 11 cell clusters, among which, 173 differentially expressed genes in CSC were identified. Moreover, the CSC score and mRNAsi were higher in tumor samples. 18 of 173 genes were survival time-associated genes in both the TCGA-LUDA dataset and the GSE dataset. Next, two molecular subtypes (namely, CSC1 and CSC2) were identified based on 18 survival-related CSC genes with distinct immune profiles and noticeably different prognoses as well as differences in the sensitivity of chemotherapy drugs. 8 genes were used to build a prognostic model in the TCGA-LUAD dataset. High-risk patients faced worse survival than those with a low risk. The robust predictive ability of the risk score was validated by the time-dependent ROC curve revealed as well as the GSE dataset. TIDE analysis showed a higher sensitivity of patients in the low group to immunotherapy. This study has revealed the effect of CSC on the heterogeneity of LUAD, and created an 8 genes prognosis model that can be potentially valuable for predicting the prognosis of LUAD and response to immunotherapy.

Identification of the CD8+ T-cell Related Signature for Predicting the Prognosis of Gastric Cancer Based on Integrated Analysis of Bulk and Single-cell RNA Sequencing Data.

The infiltration of CD8+ T cells in the tumor microenvironment is associated with better survival and immunotherapy response. However, their roles in gastric cancer have not been explored so far. In here, the profiles of GC gene expression were collected from The Cancer Genome Atlas database. Single-cell transcriptomic data originated from GSE134520. Cell clustering, annotation, and CD8+ T-cell differential genes were from the TISCH database. We determined 896 CD8+ T-cell differential genes by scRNA-seq analysis. After integrating immune-related genes, 174 overlapping genes were obtained and a novel risk model was subsequently built. The performance of CD8+ T-cell-associated gene signature was assessed in the training and external validation sets. The gene signature showed independent risk factors of overall survival for GC. A quantitative nomogram was built to enhance the clinical efficacy of this signature. Furthermore, low-risk individuals showed higher mutation status, higher immune checkpoint expression, low Tumour Immune Dysfunction and Exclusion (TIDE) scores, and higher IPS-PD-1 combined IPS-CTLA4 scores, indicating a greater response to immunotherapy. In addition, analysis of IMvigor210 immunotherapy cohort demonstrated that low-risk individuals had a favorable response to prognosis and immunotherapy. In conclusion, we generated a CD8+ T-cell-related signature that can serve as a promising tool for personalized prognosis prediction and guiding decisions regarding immunotherapy in GC patients.

Single-cell transcriptome analysis reveals the regulatory functions of islet exocrine cells after short-time obesogenic diet.

This study aims to investigate the functions of exocrine islet cell subtypes in the early stage of obesity induced by high-fat diet (HFD), which is accompanied with deterioration of the systemic insulin response and islet subpopulation abnormalities. In this study, we analyzed published islet single-cell RNA sequencing (scRNA-seq) datasets from the early stage induced by HFD feeding. Bioinformatics tools such as findMarkers, Cellchat, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways, and Gene Ontology (GO) terms were applied to identify the different functions of exocrine cell clusters. A total of 26 cell clusters were obtained were identified from this dietary intervention model. Most proportions of cell subtypes were consistent between high-fat diet (HFD) and low-fat diet (LFD) groups, except for partial endocrine islet clusters and exocrine clusters. Most differentiated expression of genes in the HFD group was found in exocrine cluster. And we also found that the cell-cell interactions between ductal and endothelial cells were reduced in the HFD group, with the significant alteration in C17 (ductal) cluster. By further analyzing the co-expression regulatory network of transcription in the C17 cluster, we speculate that differentially expressed transcription factors affected the function of duct cells by affecting the expression of related genes in intercellular interaction networks, thereby promoting insulin resistance (IR) development. Our results provide a reference for the function and regulatory mechanisms of exocrine cells in the obesity induced by HFD and probably influence the process of following insulin resistance.

Single-cell transcriptome reveals the heterogeneity of malignant ductal cells and the prognostic value of REG4 and SPINK1 in primary pancreatic ductal adenocarcinoma

Background Pancreatic ductal adenocarcinoma (PDAC) is one of the leading causes of cancer-related deaths, with very limited therapeutic options available. This study aims to comprehensively depict the heterogeneity and identify prognostic targets for PDAC with single-cell RNA sequencing (scRNA-seq) analysis. Methods ScRNA-seq analysis was performed on 16 primary PDAC and three adjacent lesions. A series of analytical methods were applied for analysis in cell clustering, gene profiling, lineage trajectory analysis and cell-to-cell interactions. In vitro experiments including colony formation, wound healing and sphere formation assay were performed to assess the role of makers. Results A total of 32,480 cells were clustered into six major populations, among which the ductal cell cluster expressing high copy number variants (CNVs) was defined as malignant cells. Malignant cells were further subtyped into five subgroups which exhibited specific features in immunologic and metabolic activities. Pseudotime trajectory analysis indicated that components of various oncogenic pathways were differentially expressed along tumor progression. Furthermore, intensive substantial crosstalk between ductal cells and stromal cells was identified. Finally, genes (REG4 and SPINK1) screened out of differentially expressed genes (DEGs) were upregulated in PDAC cell lines. Silencing either of them significantly impaired proliferation, invasion, migration and stemness of PDAC cells. Conclusions Our findings offer a valuable resource for deciphering the heterogeneity of malignant ductal cells in PDAC. REG4 and SPINK1 are expected to be promising targets for PDAC therapy.

The prognostic genes model of breast cancer drug resistance based on single-cell sequencing analysis and transcriptome analysis

Breast cancer (BC) represents a multifaceted malignancy, with escalating incidence and mortality rates annually. Chemotherapy stands as an indispensable approach for treating breast cancer, yet drug resistance poses a formidable challenge. Through transcriptome data analysis, we have identified two sets of genes exhibiting differential expression in this context. Furthermore, we have confirmed the overlap between these genes and those associated with exosomes, which were subsequently validated in cell lines. The investigation screened the identified genes to determine prognostic markers for BC and utilized them to formulate a prognostic model. The disparities in prognosis and immunity between the high- and low-risk groups were validated using the test dataset. We have discerned different BC subtypes based on the expression levels of prognostic genes in BC samples. Variations in prognosis, immunity, and drug sensitivity among distinct subtypes were examined. Leveraging data from single-cell sequencing and prognostic gene expression, the AUCell algorithm was employed to score individual cell clusters and analyze the pathways implicated in high-scoring groups. Prognostic genes (CCT4, CXCL13, MTDH, PSMD2, and RAB27A) were subsewoquently validated using RT-qPCR. Consequently, we have established a model for predicting prognosis in breast cancer that hinges on drug resistance and ERGs. Furthermore, we have evaluated the prognostic value of this model. The genes identified as prognostic markers can now serve as a reference for precise treatment of this condition.

A reference profile of N-glycosylation for human kidney and the identification of cell-cell interactions between parietal epithelial cells and capillary endothelial cells by single-cell glycosylation-sequencing.

Background N-glycosylation is one of the most common post-translational modifications in humans, and these alterations are associated with kidney diseases. Methods A novel technological approach, single-cell N-acetyllactosamine sequencing (scLacNAc-seq), was applied to simultaneously detect N-glycosylation expression and the transcriptome at single-cell resolution in three human kidney tissues from zero-time biopsy. Cell clusters, glycation abundance in each cell cluster, functional enrichment analysis, cell-cell crosstalk, and Pseudotime analysis were applied. Results Using scLacNAc-seq, 24,247 cells and 22 cell clusters were identified, and N-glycan abundance in each cell was obtained. Transcriptome analysis revealed a close connection between capillary endothelial cells (CapECs) and parietal epithelial cells (PECs). PECs and CapECs communicate with each other through several pairs of ligand receptors (e.g., TGFB1-EGFR, GRN-EGFR, TIMP1-FGFR2, VEGFB-FLT1, ANGPT2-TEK, and GRN-TNFRSF1A). Finally, a regulatory network of cell-cell crosstalk between PECs and CapECs was constructed, which is involved in cell development. Conclusions We here, for the first time, constructed the glycosylation profile of 22 cell clusters in the human kidney from time-zero biopsy. Moreover, cell-cell communication between PECs and CapECs through the ligand-receptor system may play a crucial regulatory role in cell proliferation.

Integrated analyses reveal the diagnostic and predictive values of COL5A2 and association with immune environment in Crohn's disease.

The pathogenesis of Crohn's disease (CD) involves abnormal immune cell infiltration and dysregulated immune response. Therefore, thorough research on immune cell abnormalities in CD is crucial for improved treatment of this disease. Single-cell RNA sequencing (scRNA-seq) and bulk RNA-seq data of CD were obtained from the Gene Expression Omnibus (GEO) database. Cell-type identification by estimating relative subsets of RNA transcripts (CIBERSORT), weighted gene co-expression network analysis (WGCNA), protein-protein interaction (PPI) networks evaluated the proportion of immune infiltrating cells, constructed co-expression network and identified key genes, respectively. Based on the dataset (GSE134809), 15 cell clusters were defined and labeled as different cell types. Among the 11 modules, the yellow module had the closest relationship with plasma cells (cluster 5). Confirmed using RNA sequencing and IHC assay, the expression of COL5A2 in CD samples was higher than that in control samples. Furthermore, the COL5A2 protein expression remarkably decreased in the group of patients who responded to anti-tumor necrosis factor (TNF) treatments, compared to the non-response group. The comprehensive analyses described here provided novel insight into the landscape of CD-associated immune environment. In addition, COL5A2 were identified as potential diagnostic indicators for CD, as well as promising predictive markers for CD patients.

Molecular mechanism for the interaction of MOG1 with the intracellular Loop II of cardiac sodium channel Nav1.5 and its role in Brugada syndrome

Abstract Background SCN5A encodes the cardiac sodium channel a-subunit Nav1.5 with an essential role in generation and conduction of cardiac action potential. Mutations in SCN5A cause Brugada syndrome (BrS) and long QT syndrome (LQTS). The function of Nav1.5, including its cell membrane trafficking and clustering in caveolae, is regulated by MOG1. We recently showed that gene therapy with AAV9-MOG1 reversed disease phenotypes of BrS, dilated cardiomyopathy, cardiac conduction disease and sick sinus syndrome associated with SCN5A mutations. MOG1 is a small chaperon that interacts with multiple sites of Nav1.5, including intracellular Loop I between transmembrane domain I and II and Loop II between domain II and III. We previously characterized the interaction between MOG1 and the intracellular Loop I of Nav1.5, and defined amino acids D24, E36, D44, E53, and E101A of MOG1 as critical residues for its interaction with Nav1.5 Loop I. However, the mechanistic understanding of the interaction between MOG1 and intracellular Loop II of Nav1.5 is needed. Purpose The aim of this study is to perform systematic mutational and electrophysiological studies to identify the key structural elements involved in the interaction between MOG1 and Nav1.5 Loop II and determine the significance of this interaction to the pathogenesis of BrS. Methods Patch-clamping was used to record sodium currents. Glutathione S-transferase (GST) pull-down were used to characterize protein-protein interactions. Results Nav1.5 Loop II contains 765 amino acids, and the domain between amino acids 1157 and 1200 was shown to be the interaction domain for MOG1. Six serial deletions were created, including LoopII△940-982, LoopII△983-1025, LoopII△1026-1068, LoopII△1069-1112, LoopII△1113-1156, and LoopII△1157-1200. GST-pull-down showed that only LoopII△1157-1200 failed to interact with MOG1. MOG1 enhanced sodium current densities all Nav1.5 deletion mutants except for LoopII△1157-1200. Further deletions by every 10-11 amino acids defined the MOG1-interacting domain to 1190-1200. Point mutation analysis defined the MOG1-Nav1.5 Loop II interaction domain to an 8-amino-acid domain of R1193-Y1200 of Nav1.5 Loop II. Mutations R1193W, L1194M and Y1199S associated with BrS reduced the interaction between MOG1 and Nav1.5. MOG1 enhanced sodium current densities from wild type Nav1.5, but not from mutant Nav1.5 with mutations R1193W, L1194M and Y1199S. Conclusions This study identifies the critical elements of Nav1.5 Loop II for interaction with MOG1, and reveals the molecular mechanism by which Nav1.5 mutations R1193W, L1194M and Y1199S cause BrS. Mutations R1193W, L1194M and Y1199S weaken the interaction between Nav1.5 and MOG1, and reduce MOG1-enhanced cardiac sodium current densities.

The role of elasticity on adhesion and clustering of neurons on soft surfaces

The question of whether material stiffness enhances cell adhesion and clustering is still open to debate. Results from the literature are seemingly contradictory, with some reports illustrating that adhesion increases with surface stiffness and others suggesting that the performance of a system of cells is curbed by high values of elasticity. To address the role of elasticity as a regulator in neuronal cell adhesion and clustering, we investigated the topological characteristics of networks of neurons on polydimethylsiloxane (PDMS) surfaces - with values of elasticity (E) varying in the 0.55–2.65 MPa range. Results illustrate that, as elasticity increases, the number of neurons adhering on the surface decreases. Notably, the small-world coefficient – a topological measure of networks – also decreases. Numerical simulations and functional multi-calcium imaging experiments further indicated that the activity of neuronal cells on soft surfaces improves for decreasing E. Experimental findings are supported by a mathematical model, that explains adhesion and clustering of cells on soft materials as a function of few parameters - including the Young’s modulus and roughness of the material. Overall, results indicate that – in the considered elasticity interval – increasing the compliance of a material improves adhesion, improves clustering, and enhances communication of neurons.

Single-cell RNA analysis of chromodomain-encoding genes in mesenchymal stromal cells of the mouse dental pulp.

The chromodomain helicase DNA-binding (CHD) and chromobox (CBX) families of proteins play crucial roles in cell fate decisions, differentiation, and cell proliferation in a broad variety of tissues and cell types. CHD proteins are ATP-dependent epigenetic enzymes actively engaged in transcriptional regulation, DNA replication, and DNA damage repair, whereas CBX proteins are transcriptional repressors mainly involved in the formation of heterochromatin. The pleiotropic effects of CHD and CBX proteins are largely dependent on their versatility to interact with other key components of the epigenetic and transcriptional machinery. Although the function and regulatory modes of CHD and CBX factors are well established in many cell types, little is known about their roles during osteogenic differentiation. A single-cell RNA-sequencing analysis of the mouse incisor dental pulp revealed distinct spatiotemporal expression patterns of CHD- and CBX-encoding genes within different clusters of mesenchymal stromal cells (MSCs) representing various stages of osteogenic differentiation. Additionally, genes encoding interaction partners of CHD and CBX proteins, such as subunits of the trithorax-COMPASS and polycomb chromatin remodeling complexes, exhibited differential co-expression behaviors within MSC subpopulations. Thus, CHD- and CBX-encoding genes show partially overlapping but distinct expression patterns in MSCs, suggesting their differential roles in osteogenic cell fate decisions.

#1449 Maturation of tertiary lymphoid structures in progressive kidney damage due to chronic phosphate load

Abstract Background and Aims We have recently shown that a high phosphate diet (HPD) in mice leads to an increase in FGF23 levels and causes progressive tubular damage and interstitial fibrosis with increased accumulation of macrophages and concomitant perivascular immune cell aggregates in the corticomedullary zone. Such tertiary lymphoid structures (TLS) can develop as a result of chronic inflammation and have already been described in lupus nephritis and IgA nephropathy, among others. In our model, the question arose to what extent damage to the proximal tubule is temporally associated with the development and maturation of TLS and which factors favor TLS development. Method Wild-type mice were fed HPD for 1, 2, 3, 4 and 6 months and the kidneys of each were compared with those of mice on a normal phosphate diet (NPD). To evaluate the effect of high FGF23 ± phosphate, Fgf23 was overexpressed in mice on NPD and HPD using an adeno-associated virus (AAV-Fgf23), and hypophosphatemic Hyp mice were used. Results An increased tubular damage score was seen after only two months of HPD. In parallel, perivascular immune cell clusters developed in the corticomedullary zone and in the renal cortex, which were characterized by aggregates of CD45R+ B cells and CD3+ T cells, as well as LYVE1+ lymphatic vessels after three months of HPD and were accompanied by increased expression of venous markers (Plvap, Aplnr, EphB4) and cell adhesion molecules (Sell, Vcam1, Madcam1). Further analyses showed a time-dependent induction of the immunofibroblast-derived chemokines Cxcl13 and Ccl19, as well as the lymphotoxins Lta and Ltb, which are important for the differentiation of immunofibroblasts into follicular dendritic cells (FDC) and follicular reticular cells (FRC). Already after four months of HPD, proliferating B-cell clusters with FDCs, FRCs located in T-cell clusters, podoplanin+ networks, high endothelial venules, plasma cells and increased IgD synthesis indicated fully mature TLS, while tubular damage and fibrosis continued to increase until six months of HPD. Compared to AAV-Fgf23 mice on NPD, only AAV-Fgf23 animals on HPD developed tubular damage with fibrosis and fully mature TLS. Hypophosphatemic Hyp mice with high FGF23 showed neither signs of tubular damage nor the development of TLS. Conclusion The development of renal TLS due to chronic high phosphate load coincides with the development of tubular damage, with earlier maturation of TLS and progression of tubular damage.

#1914 Comprehensive single-cell sequencing and lipidomics identify diverse injury states and cholesterol dynamics of the PTs in AKI-to-CKD transition

Abstract Background and Aims The underlying cellular events and metabolic dysfunction of acute kidney injury (AKI) transformed to chronic kidney disease (CKD) is still largely unclear. Previous studies have suggested lipids and metabolites have a profound impact on acute and chronic inflammation. One of the main sites of renal lipid accumulation is the proximal tubule cells (PTs). Here we demonstrated the injury states and cholesterol dynamics of PTs and its potential function in AKI to CKD in unilateral ischemia-reperfusion injury (UIR) mice model. Method We employed single-cell combinatorial indexing lipidomics to analyze 15 mouse kidneys from a mouse model of failed repaired UIR at five time points. We identified cell types through unsupervised clustering analysis. Differential gene expression, enrichment analysis and main function scoring were applied to define functional heterogeneity and injury states of PTs. Next, pseudotime analysis of differentiation transitions was performed to create the diverse PT lineages. Finally, using SCENIC analysis, we identified the key transcription factors that drive the differentiation of PTs. Results We identified 13 main cell clusters and the dynamic changes in the process of AKI to CKD transition were analysed. A remarkable decrease in PTs after injury at day 1 followed with a recovery at day 3 post AKI were observed. Spatiotemporal profiling of the main cell types showed that the most injured areas in the acute stages of kidney injury are primarily concentrated in the outer stripe, specifically the region where the proximal straight tubules (PSTs) are located. In order to gain a higher resolution of the dynamics of PST subtypes during AKI to CKD, we performed sub-clustering analysis specifically on PSTs, partitioning them into 10 subclusters. Single-cell analysis identified maladaptive PSTs (PST-5) with the highest apoptosis and ferroptosis scores. Notably, pseudotime analysis revealed that PST-5 differentiated from PST-4, which has the highest lipid metabolism score. Indeed, among all the cells in the kidney, the PTs exhibited the most active metabolic reprogramming due to their differentiation and redifferentiation early after AKI. Lipid omics profiling reveals an abnormal elevation of the cholesterol metabolites 27-hydroxycholesterol (27-OH) and 25-hydroxycholesterol (25-OH) following AKI. Interestingly, the PST-4-specific expression of Cyp7b1 served as a key enzyme regulating the concentration of 27-OH. This gene has been found to be significantly absent after AKI. Moreover, the binding of 27HC, a natural ligand, to endogenous estrogen receptor alpha (ERα) leads to the upregulation of heme oxygenase 1 (Hmox1), thereby promoting ferroptosis. Finally, SCENIC analysis revealed that Cyp7b1-mediated accumulation of 27-HC is dependent on the transcription factor Atf3. Conclusion Our results indicate that Cyp7b1 deficiency confers ferroptosis, which are primarily dependent on the upregulation of cellular cholesterol homeostasis and 27-HC secretion, leading to the progression of AKI. These findings suggest that the Cyp7b1/Hmox1/Atf3/27-HC axis may serve as a potential therapeutic target for preventing the development of CKD.

#161 Single-Nucleus RNA-Seq of Human Kidney Biopsy Tissue Reveals Microstructure-Specific Similarities and Differences in Various Glomerulonephritis

Abstract Background and Aims Transcriptomic alterations in kidney diseases may lead to the discovery of novel biomarkers. Single-nucleus RNA-seq (SnSeq) is a valuable method for profiling the transcriptome of kidney-residing cells. Additional SnRNA-seq data is necessary, including cases of human biopsy-confirmed glomerulonephritis. Method We collected snap-frozen kidney biopsy tissues from 6 cases of IgA nephropathy (IgAN), 6 cases of minimal change disease, 6 cases of PLA2R-Ab positive membranous nephropathy, 3 cases of diabetic kidney disease, and 7 nephrectomy control cases. SnSeq was performed on the kidney tissues, and kidney cells were identified in the dataset through UMAP clustering. The proportions of each cell subcluster were compared between the study groups. Differentially expressed genes (DEGs) were identified. K-means clustering and Weighted Gene Co-expression Network Analysis (WGCNA) were performed. Results We successfully collected transcriptomic profiles of 55 615 cells from IgAN, 53 370 from minimal change disease, 53 495 from membranous nephropathy, 34 343 from diabetic kidney disease, and 66 161 from control cases. The SnSeq data successfully identified major kidney cell clusters, including a considerable number of glomerular cells such as podocytes, mesangial cells, endothelial cells, and others. DEGs were largely identified in podocytes and proximal tubule cells, as well as in other glomerular cells. K-means clustering identified diverse similarities and differences between the study groups for each cell type. We identified diverse cell type-specific but disease-common DEGs, particularly in the downregulated genes, suggesting that the studied diseases share similar destruction in kidney microstructure. In comparison, cell type-common but disease-specific DEGs were also identified, considering a common pathophysiologic mechanism may occur in total kidney cortex. Diseases that can feature nephrotic syndrome showed high expression of AVP receptors, suggesting the association with edema, or highly expressed renin-angiotensin-aldosterone pathway was noted in IgAN or diabetic kidney disease, which was not evident in minimal change disease or membranous nephropathy patients. The studied disease shared certain types of inflammatory pathways (e.g. interferon-alpha or TGF-beta) but some pathways were differently enriched according to the disease types. WGCNA analysis revealed gene modules that were highly correlated with age, proteinuria, and estimated GFR, regardless of the diagnosis subtype. Conclusion The current SnSeq data is, to date, the largest one, including biopsy-confirmed human kidney glomerulonephritis cases. The results identified disease-specific changes in specific cell types. Additionally, common transcriptomic alterations occurring following changes in estimated GFR, age, and proteinuria were investigated. The data may be a valuable resource for future research on glomerulonephritis biomarker development.

Deep Batch Integration and Denoise of Single-Cell RNA-Seq Data.

Numerous single-cell transcriptomic datasets from identical tissues or cell lines are generated from different laboratories or single-cell RNA sequencing (scRNA-seq) protocols. The denoising of these datasets to eliminate batch effects is crucial for data integration, ensuring accurate interpretation and comprehensive analysis of biological questions. Although many scRNA-seq data integration methods exist, most are inefficient and/or not conducive to downstream analysis. Here, DeepBID, a novel deep learning-based method for batch effect correction, non-linear dimensionality reduction, embedding, and cell clustering concurrently, is introduced. DeepBID utilizes a negative binomial-based autoencoder with dual Kullback-Leibler divergence loss functions, aligning cell points from different batches within a consistent low-dimensional latent space and progressively mitigating batch effects through iterative clustering. Extensive validation on multiple-batch scRNA-seq datasets demonstrates that DeepBID surpasses existing tools in removing batch effects and achieving superior clustering accuracy. When integrating multiple scRNA-seq datasets from patients with Alzheimer's disease, DeepBID significantly improves cell clustering, effectively annotating unidentified cells, and detecting cell-specific differentially expressed genes.

Regulation of Fuzheng Huayu capsule on inhibiting the fibrosis-associated hepatocellular carcinogenesis

In the current study, bioinformatics analysis of the hepatocellular carcinoma (HCC) dataset was conducted with the hepatoprotective effect of the Fuzheng Huayu (FZHY) capsule against the diethylnitrosamine-induced HCC progression analyzed. Eight cell clusters were defined and tanshinone IIA, arachidonic acid, and quercetin, compounds of the FZHY capsule, inhibit HCC progression-related fibrosis by regulating the expression of PLAU and IGFBP3. Combined with the ameliorative effect of the FZHY capsule against liver dysfunctions and expression of PLAU and IGFBP3, our study confirmed the effect of the FZHY capsule on inhibiting the fibrosis-associated HCC progression via regulating the expression of PLAU and IGFBP3.

Targeting miR-181a/b in retinitis pigmentosa: implications for disease progression and therapy

BackgroundRetinitis pigmentosa (RP) is a genetically heterogeneous group of degenerative disorders causing progressive vision loss due to photoreceptor death. RP affects other retinal cells, including the retinal pigment epithelium (RPE). MicroRNAs (miRs) are implicated in RP pathogenesis, and downregulating miR-181a/b has shown therapeutic benefit in RP mouse models by improving mitochondrial function. This study investigates the expression profile of miR-181a/b in RPE cells and the neural retina during RP disease progression. We also evaluate how miR-181a/b downregulation, by knocking out miR-181a/b-1 cluster in RPE cells, confers therapeutic efficacy in an RP mouse model and explore the mechanisms underlying this process.ResultsOur findings reveal distinct expression profiles, with downregulated miR-181a/b in RPE cells suggesting a protective response and upregulated miR-181a/b in the neural retina indicating a role in disease progression. We found that miR-181a/b-2, encoded in a separate genomic cluster, compensates for miR-181a/b-1 ablation in RPE cells at late time points. The transient downregulation of miR-181a/b in RPE cells at post-natal week 6 (PW6) led to improved RPE morphology, retarded photoreceptor degeneration and decreased RPE aerobic glycolysis.ConclusionsOur study elucidates the underlying mechanisms associated with the therapeutic modulation of miR-181a/b, providing insights into the metabolic processes linked to its RPE-specific downregulation. Our data further highlights the impact of compensatory regulation between miR clusters with implications for the development of miR-based therapeutics.Graphical

Abstract A023: Stratification of bladder cancer patients with the use of flow cytometry is increasing the diagnostic potential of molecular classifiers

Abstract Recent research suggests the potential categorization of bladder cancer (BLCa) patients into molecular subtypes to aid in the selection of appropriate treatment strategies, be it conventional, targeted, or a combination thereof. However, the practical implementation of molecular classifiers has faced obstacles due to the intricate nature of gene regulation, tumor heterogeneity, and challenges related to tumor sampling caused by substantial stromal infiltration and tumor multifocality. Hence, there is a pressing need for more straightforward marker-based methodologies to assist in genomic and transcriptomic subtyping efforts. Our study addressed this gap by establishing and validating a panel of antibodies for application in flow cytometry (FCM) to enhance the stratification of BLCa patients. We curated a set of 17 FCM markers capable of delineating four cell clusters (epithelial, immune, endothelial, and stromal) and defining tumor classification (basal and luminal), immune infiltration, and hot-target expression on BLCa tissue samples. This FCM panel was deployed on ex vivo tissue samples treated with chemotherapy and targeted therapy agents. Initially tested on a cohort of 10 BLCa tissue samples, our validation of the FCM panel revealed correlations between marker expression and clinical-pathological features, such as the presence of metastasis and heightened immune and stromal infiltration. Notably, samples exhibiting lower claudin-4 levels displayed increased immune infiltration, while those with metastases at diagnosis showcased elevated basal marker expression. The translational potential of the FCM panel was demonstrated in ex vivo experiments where tissue slices from six patients were subjected to chemotherapy and targeted therapy agents, leading to observable shifts in marker expression post-treatment, particularly pronounced with chemotherapy. Noteworthy effects included reductions in epithelial cells expressing FGFR3 and EGFR following targeted therapy with erdafitinib or lapatinib, alongside alterations in the basal/luminal cell ratio indicated by increased CD44, CD90, and CD239 marker expressions. Our findings underscore the value of FCM in categorizing BLCa patients based on specific marker expression, supplementing conventional histological analyses to enhance cancer diagnosis, guide personalized therapeutic approaches, and monitor treatment efficacy. Citation Format: Martina Radic, Federico La Manna, Martina Minoli, Panagiotis Chouvardas, George N. Thalmann, Roland Seiler-Blarer, Marianna Kruithof-de Julio, Bernhard Kiss. Stratification of bladder cancer patients with the use of flow cytometry is increasing the diagnostic potential of molecular classifiers [abstract]. In: Proceedings of the AACR Special Conference on Bladder Cancer: Transforming the Field; 2024 May 17-20; Charlotte, NC. Philadelphia (PA): AACR; Clin Cancer Res 2024;30(10_Suppl):Abstract nr A023.

P39 Using single-cell transcriptomics to investigate the potential role of the human hair follicle dermal papilla in controlling hair fibre pigmentation

Abstract Introduction and aims The human dermal papilla (DP) is a population of cells located at the base of the hair follicle, which signals to surrounding matrix cells facilitating their incorporation into the layers of the hair fibre and the hair follicle. The DP controls the process of hair follicle induction, reflecting its critical role in orchestrating processes integral in hair growth. We hypothesize that distinct functional subgroups exist within the structure, which would mean that there are different types of DP cells within each follicle. Methods To begin testing this, we reanalysed the single-cell RNA sequencing (RNA-Seq) dataset of the human anagen hair follicle used in the 2022 study by Shim et al. To expand on this, we are now generating single-cell RNA-Seq datasets of patient-matched pigmented and nonpigmented hair follicles (n = 4 patients). These datasets will be used to explore heterogeneity in the human DP, and to investigate which subpopulations are potentially changed in the case of pigmentation loss. This will allow us to explore a potentially undiscovered role of the DP in controlling hair fibre pigmentation. Results We found four distinct subpopulations in the DP with statistically significant marker genes in the Shim dataset. This confirms transcriptional heterogeneity in the human DP, hinting at clusters of DP cells with specialized functions within each hair follicle. One subcluster, which expressed S100B, endothelin-3 and bone morphogenetic protein (BMP)4, was of particular interest and gene ontology analysis of the genes in this cluster revealed terms associated with skeletal development and BMP signalling. We hypothesize that this DP cluster regulates melanin transfer from bulbar melanocytes to neighbouring keratinocytes, owing to the known involvement of the mentioned genes in both melanogenesis and melanin transfer. Conclusions The human DP is a heterogeneic structure, with a potentially unexplored role in controlling hair fibre pigmentation.