ESCFD: Probabilistic Flow Diffusion Model for Accelerated High-Quality Single-Cell RNA-seq Data Synthesis

This study aimed to comprehensively characterize the relationship between glycated hemoglobin (HbA1c) and intervertebral disc degeneration (IVDD) by integrating population-level epidemiological profiling, Mendelian randomization (MR) analysis, multi-omics data, and single-cell transcriptomics. The objectives included clarifying the causal effect of HbA1c on IVDD, identifying key HbA1c-related genes, elucidating their cellular and molecular roles in disc degeneration, and exploring potential therapeutic targets. Global Burden of Disease (GBD) 2021 database was used to evaluate the incidence rate of diabetes mellitus (DM) and low back pain (LBP) across the entire global population. To assess the causal effect of HbA1c on IVDD, MR approache was conducted using summary statistics from large genome-wide association studies. Subsequently, integrating cis-eQTL information, bulk RNA sequencing, and single-cell RNA sequencing were performed to identify HbA1c-related genes involved in disc degeneration, and explore their roles along pseudotime differentiation trajectories. Potential associations between hub genes and IVDD, as well as candidate therapeutic compounds targeting these genes, were further evaluated through database screening and molecular docking. The findings from GBD indicated a correlation between diabetes and low back pain in both 1990 and 2021. The results of MR analysis revealed a causal relationship between HbA1c levels and IVDD. After integrated analysis of the eQTLGen database and RNA-seq data, we identified five hub HRGs, including STAT5A, CDKN1C, TRPS1, NME4, and OASL, which have been implicated in multiple biological processes and pathways associated with IVDD. Moreover, the scRNA-seq analysis showed that hub HRGs were associated with specific cell subpopulations in human nucleus pulposus (NP) tissues, such as inflamed chondrocytes, fibrochondrocytes, chondrocytes, and calcifying chondrocytes. Additionally, hub HbA1c related genes (HRGs) have been found to be associated with several DDDs such as back pain, osteoporosis, and spinal stenosis. Meanwhile, 12 drugs, including resveratrol, dexamethasone, and simvastatin, were identified that have the potential to target hub HRGs. This study revealed a causal effect of HbA1c on IVDD and identified five key HbA1c-related genes (STAT5A, CDKN1C, TRPS1, NME4, and OASL) involved in disc degeneration, providing potential targets for future research.

Single-cell RNA sequencing (scRNA-seq) technology produces an unprecedented resolution at the level of a unique cell, raising great hopes in medicine. Nevertheless, scRNA-seq data suffer from high variations due to the experimental conditions, called batch effects, preventing any aggregated downstream analysis. Adversarial Information Factorization provides a robust batch-effect correction method that does not rely on prior knowledge of the cell types nor a specific normalization strategy while being adapted to any downstream analysis task. It compares to and even outperforms state-of-the-art methods in several scenarios: low signal-to-noise ratio, batch-specific cell types with few cells, and a multi-batches dataset with imbalanced batches and batch-specific cell types. Moreover, it best preserves the relative gene expression between cell types, yielding superior differential expression analysis results. Finally, in a more complex setting of a Leukemia cohort, our method preserved most of the underlying biological information for each patient while aligning the batches, improving the clustering metrics in the aggregated dataset.

The mechanism of abdominal aortic aneurysm (AAA) has not been fully elucidated. In this study, we aimed to map the cellular heterogeneity, molecular alteration, and functional transformation of angiotensin (Ang) II-induced AAA in mice based on single-cell RNA sequencing (sc-RNA seq) technology. sc-RNA seq was performed on suprarenal abdominal aorta tissue from male Apoe-/- C57BL/6 mice of Ang II-induced AAA and shame models to determine the heterogeneity and phenotypic transformation of all cells. Immunohistochemistry was used to determine the pathophysiological characteristics of AAA. The single-cell trajectory was performed to predict the differentiation of fibroblasts. Finally ligand-receptor analysis was used to evaluate intercellular communication between fibroblasts and smooth muscle cells (SMCs). More than 27,000 cells were isolated and 25 clusters representing 8 types of cells were identified, including fibroblasts, macrophages, endothelial cells, SMCs, T lymphocytes, B lymphocytes, granulocytes, and natural killer cells. During AAA progression, the function and phenotype of different type cells altered separately, including activation of inflammatory cells, alternations of macrophage polarization, phenotypic transformation of vascular smooth muscle cells, and endothelial to mesenchymal transformation. The alterations of fibroblasts were the most conspicuous. Single-cell trajectory revealed the critical reprogramming genes of fibroblasts mainly enriched in regulation of immune system. Finally, the ligand-receptor analysis confirmed that disorder of collagen metabolism led by fibroblasts was one of the most prominent characteristics of Ang II-induced AAA. Our study revealed the cellular heterogeneity of Ang II-induced AAA. Fibroblasts may play a critical role in Ang II-induced AAA progression according to multiple biological functions, including immune regulation and extracellular matrix metabolic balance. Our study may provide us with a different perspective on the etiology and pathogenesis of AAA.

Spermatogenesis, a key part of the spermiation process, is regulated by a combination of key cells, such as primordial germ cells, spermatogonial stem cells, and somatic cells, such as Sertoli cells. Abnormal spermatogenesis can lead to azoospermia, testicular tumors, and other diseases related to male infertility. The application of single-cell RNA sequencing (scRNA-seq) technology in male reproduction is gradually increasing with its unique insight into deep mining and analysis. The data cover different periods of neonatal, prepubertal, pubertal, and adult stages. Different types of male infertility diseases including obstructive and non-obstructive azoospermia (NOA), Klinefelter Syndrome (KS), Sertoli Cell Only Syndrome (SCOS), and testicular tumors are also covered. We briefly review the principles and application of scRNA-seq and summarize the research results and application directions in spermatogenesis in different periods and pathological states. Moreover, we discuss the challenges of applying this technology in male reproduction and the prospects of combining it with other technologies.

Tumor progression in colorectal cancer (CRC) is driven by a subpopulation of cells with tumor-initiating cell (TIC) activity. In the past, bulk experiments provided strong evidence for functional heterogeneity of the CRC TIC compartment with distinct types of TICs driving CRC progression. However, in these retrospective experiments direct assignment of functional states to individual functionally relevant cell types in CRC has not been possible. We here asked whether functional states of individual cells can be assigned to specific CRC cell subpopulations. Therefore, 4,663 single-cell profiles were generated by single-cell RNA-sequencing (scRNA-seq) of patient-derived CRC spheroid cultures (n=12). Non-negative matrix factorization identified 13 gene expression modules linked to cell states or cell types that resemble differentiation states of normal intestinal epithelial cells. To evaluate their presence in CRC patient tumors, the identified transcriptional programs were applied to publicly available expression data (TCGA cohort; n=459). Clustering of samples based on the contribution of individual expression signatures revealed six patient clusters with significantly different progression-free survival (p=0.043) and a trend towards different overall survival (p=0.12), indicating a relevance for patient outcome. Comparing cell state meta-signature expression in cell type-associated subpopulations revealed differences in functional states, i.e. proliferative and metabolic activity. Interestingly, stem-like cells showed significantly increased oxidative phosphorylation (OXPHOS) and decreased glycolysis scores, suggesting a link between metabolic states and cell differentiation. scRNA-seq of patient-derived organoid and xenograft models as well as primary tumors revealed similar trends. To assess whether metabolic states are linked to TIC activity, CRC spheroid cells were sorted according to mitochondrial membrane potential (MMP), and sphere forming cell (SFC) as well as TIC frequency were assessed by limiting dilutions. Strikingly, MMPhigh cells exhibited a pronounced increase of SFC frequencies in vitro compared to MMPlow cells in 80% of patients (n=5). In addition, TIC frequencies were strongly increased in the MMPhigh compared to the MMPlow subpopulation in vivo (n=2; 1/46,535 vs. 1/211,305 and 1/249 vs. 1/2,089), demonstrating that TIC activity relies on OXPHOS. In line with this, pretreatment with the OXPHOS inhibitor CCCP decreased SFC frequencies (n=3), indicating targetable metabolic vulnerability of CRC spheroid cells. In summary, we here show that transcriptional heterogeneity identifies functional states during CRC TIC differentiation. Targeting context-dependent regulation of tumor cell differentiation might unravel novel vulnerabilities for therapeutic intervention in human CRC. Citation Format: Martina K. Zowada, Stephan M. Tirier, Sebastian M. Dieter, Teresa Krieger, Ava Oberlack, Robert L. Chua, Mario Huerta, Foo Wei Ten, Karin Laaber, Jeongbin Park, Katharina Jechow, Torsten Müller, Mathias Kalxdorf, Mark Kriegsmann, Katharina Kriegsmann, Friederike Herbst, Jeroen Krijgsveld, Martin Schneider, Roland Eils, Hanno Glimm, Christian Conrad, Claudia R. Ball. Transcriptional heterogeneity identifies functional states of tumor-initiating cell differentiation in human colorectal cancer [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 1491.

Glioblastoma (GBM) is the most common and malignant type of primary brain tumor, and more effective treatment options are needed. Both inter- and intratumoral heterogeneity present major challenges to the application of targeted therapies in GBM. Therefore, precision medicine approaches to GBM would benefit significantly from the ability to predict drugs or drug combinations that target specific subpopulations of tumor cells. Model systems, such as adherent cell lines, neurospheres, patient-derived xenografts (PDXs), and patient-derived organoids, have been reported as platforms for drug screening and accessing drug responses. However, these models do not recapitulate the full heterogeneity of GBM tissue, lack key components of the tumor microenvironment or take weeks to months to establish, which limits the predictive power of drug response assays or delays clinical decision-making. To address these limitations, we are combining ex vivo slices of intact, patient-derived GBM tissue with single-cell RNA-seq (scRNA-seq) for small-scale drug screening and assessment of patient- and cell type-specific drug responses. We generated slices from both preclinical mouse glioma models and surgical specimens from GBM patients and showed that acute slices preserved both the tumor heterogeneity and tumor microenvironment observed in scRNA-seq of cells directly isolated from tumor tissue. To test drug responses, we treated tissue slices from GBM mouse models and five different patients with six drugs for 18hr. By performing scRNA-Seq and analyzing transcriptional profiles of treated and untreated control slices, we identified drug-induced transcriptional responses in specific subpopulations of tumor cells, patient-specific drug sensitivities, and drug effects conserved in both mouse and human tumors. The GBM tissue slices were generated immediately following surgical resection, and experiments were completed within 24 hours. With these features, our method is attractive for rapidly accessing cell type- and patient-specific drug responses and has potential for preclinical drug screening and guiding personalized treatment for GBM.

Alternative splicing is an essential mechanism for generating protein diversity by producing distinct isoforms from a single gene. Dysregulation of splicing that affects pancreatic function and immune tolerance has been linked to both types 1 and 2 diabetes. Next-generation sequencing technologies, with their short read lengths, are limited in their ability to accurately detect splice variants. Long-read sequencing technologies offer the potential to overcome these limitations by providing full-length transcript information; however, their application in single-cell RNA sequencing has been hindered by technical challenges, including insufficient read lengths and higher error rates. Furthermore, cell types that produce high levels of a single transcript, such as islet endocrine cells, can obscure identification of lower-abundance transcripts. In this study, we optimized a protocol for single-cell long-read sequencing in pancreatic islets to improve read length and transcript detection. Our findings demonstrate that 5′ library preparation protocols outperform 3′ protocols, resulting in better transcript identification. Furthermore, we show that targeted depletion of insulin transcripts enhances the detection of informative reads, highlighting the utility of transcript-depletion strategies. This optimized protocol enables isoform-specific gene expression analysis and reveals differential transcript usage across the various cell types in pancreatic islets. By leveraging this approach, we gain deeper insights into the transcriptomic complexity and cellular heterogeneity within pancreatic islets.Article HighlightsThis study addresses the limitations of current single-cell long-read RNA sequencing technologies in detecting full-length transcripts and isoform diversity, particularly in pancreatic islets.We demonstrate that optimizing single-cell library preparation protocols reproducibly enhances read length and transcript identification in pancreatic islets.Combined with targeted insulin depletion and extended reverse transcription, 5′ capture methods significantly improved read length and isoform detection compared with standard protocols, while maximizing the number of informative reads.These improvements yield longer reads in single-cell experiments, substantially enhancing transcript identification and enabling more accurate analysis of isoform diversity.

Global deaths attributed to gastric cancer (GC) are increasing, yet our understanding of it remains limited. Recently, single-cell RNA sequencing (scRNA-seq) and spatial transcriptomics have provided important insights into the dissection of metastasis-related biological processes in subpopulations of cells in the GC tumor microenvironment, especially intratumoral cellular heterogeneity and cell-cell interactions. In this review, we discuss the mechanisms underlying GC metastasis and potential strategies for developing upcoming immunotherapies by combining advances in scRNA-seq and spatial transcriptomics.

There is gaining popularity in the use of single-cell technology and analysis in studying the pathogenesis of abdominal aortic aneurysm (AAA). As there are no current pharmacologic therapies for impeding aneurysm growth or preventing AAA rupture, identifying key pathways involved in AAA formation is critical for the development of future therapies. Single-cell RNA sequencing (scRNA-seq) technology provides an unbiased and global view of transcriptomic characteristics within each of the major cell types in aneurysmal tissues. In this brief review, we examine the current literature utilizing scRNA-seq for the analysis of AAA and discuss trends and future utility of this technology.

Advances in single-cell RNA sequencing (scRNA-seq) technologies in the past 10 years have had a transformative effect on biomedical research, enabling the profiling and analysis of the transcriptomes of single cells at unprecedented resolution and throughput. Specifically, scRNA-seq has facilitated the identification of novel or rare cell types, the analysis of single-cell trajectory construction and stem or progenitor cell differentiation, and the comparison of healthy and disease-related tissues at single-cell resolution. These applications have been critical in advances in cardiovascular research in the past decade as evidenced by the generation of cell atlases of mammalian heart and blood vessels and the elucidation of mechanisms involved in cardiovascular development and stem or progenitor cell differentiation. In this Review, we summarize the currently available scRNA-seq technologies and analytical tools and discuss the latest findings using scRNA-seq that have substantially improved our knowledge on the development of the cardiovascular system and the mechanisms underlying cardiovascular diseases. Furthermore, we examine emerging strategies that integrate multimodal single-cell platforms, focusing on future applications in cardiovascular precision medicine that use single-cell omics approaches to characterize cell-specific responses to drugs or environmental stimuli and to develop effective patient-specific therapeutics.

Recent advances of single-cell RNA sequencing in toxicology research: Insight into hepatotoxicity and nephrotoxicity

Applications of single-cell and bulk RNA sequencing in onco-immunology

Single-cell RNA-sequencing is an invaluable research tool that allows for the investigation of gene expression in heterogeneous cancer cell populations in ways that bulk RNA-seq cannot. However, normal (i.e., non tumor) cells in cancer samples have the potential to confound the downstream analysis of single-cell RNA-seq data. Existing methods for identifying cancer and normal cells include copy number variation inference, marker-gene expression analysis, and expression-based clustering. This work aims to extend the existing approaches for identifying cancer cells in single-cell RNA-seq samples by incorporating variant calling and the identification of putative driver alterations. We found that putative driver alterations can be detected in single-cell RNA-seq data obtained with full-length transcript technologies and noticed that a subset of cells in tumor samples are enriched for putative driver alterations as compared to normal cells. Furthermore, we show that the number of putative driver alterations and inferred copy number variation are not correlated in all samples. Taken together, our findings suggest that augmenting existing cancer-cell filtering methods with variant calling and analysis can increase the number of tumor cells that can be confidently included in downstream analyses of single-cell full-length transcript RNA-seq datasets.

<div>Abstract<p>The development of single-cell RNA sequencing (scRNA-seq) technologies has greatly contributed to deciphering the tumor microenvironment (TME). An enormous amount of independent scRNA-seq studies have been published representing a valuable resource that provides opportunities for meta-analysis studies. However, the massive amount of biological information, the marked heterogeneity and variability between studies, and the technical challenges in processing heterogeneous datasets create major bottlenecks for the full exploitation of scRNA-seq data. We have developed IMMUcan scDB (<a href="https://immucanscdb.vital-it.ch" target="_blank">https://immucanscdb.vital-it.ch</a>), a fully integrated scRNA-seq database exclusively dedicated to human cancer and accessible to nonspecialists. IMMUcan scDB encompasses 144 datasets on 56 different cancer types, annotated in 50 fields containing precise clinical, technological, and biological information. A data processing pipeline was developed and organized in four steps: (i) data collection; (ii) data processing (quality control and sample integration); (iii) supervised cell annotation with a cell ontology classifier of the TME; and (iv) interface to analyze TME in a cancer type–specific or global manner. This framework was used to explore datasets across tumor locations in a gene-centric (<i>CXCL13</i>) and cell-centric (B cells) manner as well as to conduct meta-analysis studies such as ranking immune cell types and genes correlated to malignant transformation. This integrated, freely accessible, and user-friendly resource represents an unprecedented level of detailed annotation, offering vast possibilities for downstream exploitation of human cancer scRNA-seq data for discovery and validation studies.</p>Significance:<p>The IMMUcan scDB database is an accessible supportive tool to analyze and decipher tumor-associated single-cell RNA sequencing data, allowing researchers to maximally use this data to provide new insights into cancer biology.</p></div>

Advances in single-cell RNA sequencing (scRNA-seq) technologies allow researchers to analyze the genome-wide transcription profile and to solve biological problems at the individual-cell resolution. However, existing clustering methods on scRNA-seq suffer from high dropout rate and curse of dimensionality in the data. Here, we propose a novel pipeline, scBKAP, the cornerstone of which is a single-cell bisecting K-means clustering method based on an autoencoder network and a dimensionality reduction model MPDR. Specially, scBKAP utilizes an autoencoder network to reconstruct gene expression values from scRNA-seq data to alleviate the dropout issue, and the MPDR model composed of the M3Drop feature selection algorithm and the PHATE dimensionality reduction algorithm to reduce the dimensions of reconstructed data. The dimensionality-reduced data are then fed into the bisecting K-means clustering algorithm to identify the clusters of cells. Comprehensive experiments demonstrate scBKAP's superior performance over nine state-of-the-art single-cell clustering methods on 21 public scRNA-seq datasets and simulated datasets. The source codes and datasets are available at https://github.com/YuBinLab-QUST/scBKAP/ and https://doi.org/10.24433/CO.4592131.v1.