Cross-Tissue Single-Nucleus Mapping and Disease Genes

DownLoad

Summary

Gaining insights into gene functionality and regulation in balance and disease necessitates discerning the cellular and tissue settings where genes manifest. The team harnessed four single-nucleus RNA sequencing techniques on an assortment of eight cryogenically preserved, archived tissue types derived from 16 donors across 25 samples, culminating in a cross-tissue atlas comprising 209,126 nuclear profiles. These were integrated across tissues, donors, and laboratory procedures through a conditional variational autoencoder. Utilizing this resultant cross-tissue atlas, the researchers underscored both shared and tissue-specific attributes of tissue-resident cellular populations, pinpointed cell types potentially influencing the neuromuscular, metabolic, and immune facets of monogenic diseases and the biological processes implicated in their pathogenesis, and determined cell types and gene assemblies potentially underlying the disease mechanisms for complex traits investigated through genome-wide association studies.

Fig.1 Graphical abstract. (Eraslan, 2023)

Research Criteria

Owing to the multifaceted and distinct characteristics of disease-linked variants, their systematic correlation to cellular entities and molecular operations mandates a comprehensive examination across an array of tissues and subjects. Preceding cell atlases predominantly hinged on fresh specimens from a solitary organ or tissue. The employment of single-nucleus RNA sequencing (snRNA-seq) to frozen, archived tissue facilitates the capture of cell types that endure beyond dissociation across extensive tissues. Advanced machine learning methodologies can amalgamate data from various individuals and tissues by mitigating batch discrepancies whilst safeguarding biological variation.

Fig.2 Experimental design. (Eraslan, 2023)

Sample Type

Human breast, esophagus, heart, lung, muscle, and prostate tissues.

Result—Annotating Cross-tissue Atlases Recovers a Wide Range of Cell Types, Including Difficult-to-profile and Rare Cell Subsets

The researchers synthesized data from diverse samples and methodologies using a conditional variational autoencoder (cVAE), a tool specifically designed to adjust for multiple sources of expression variation such as individual-, sex-, and protocol-specific effects, whilst maintaining tissue- and cell-type-specific variation. The effectiveness of the cVAE was validated against other data integration methods, providing comparable or superior results, thereby setting a benchmark for future integration endeavors. The study's cross-tissue atlas featured 43 broad cell classes, encompassing both shared and tissue-specific cell types. The atlas successfully captured profiles from cell classes typically challenging to profile by dissociation-based scRNA-seq, including adipocytes, skeletal muscle myonuclei, and cardiac myonuclei. Additionally, the research illuminated the potential of cross-tissue and cross-sample integration for delineating multiple rare cell subsets, such as Schwann cells, neuroendocrine cells, and interstitial cells of Cajal, all of which can contribute to various pathologies. These findings underscore the importance of profiling these rare cells in human tissues for the advancement of disease studies.

Fig.3 An atlas of cross-tissue snRNA-seq in eight archived, frozen adult human tissues. (Eraslan, 2023)

Result—snRNA-seq Protocols are Effective Across Tissues and Correspond to scRNA-seq Protocols

The research team evaluated the efficacy of four nucleus extraction and profiling protocols (CST, NST, TST, and EZ) across eight tissue samples using various quality control metrics, and compared cell diversity and proportions concerning snRNA-seq, scRNA-seq, and bulk RNA-seq datasets. The EZ protocol displayed inferior performance across all tissues by several metrics, including a lower total number of nuclei captured and higher levels of ambient RNA. The extraction protocols also demonstrated variance in the proportion of nuclei recovered from each cell type, with the TST protocol yielding the highest cell-type diversity on average across tissues. The study accentuates the importance of selecting protocols matching the scientific goal, tissue type, and complexity. When comparing their protocols to other snRNA-seq studies, the researchers found agreement in broad cell types, however, there were differences in proportions. They also compared snRNA-seq data to fresh-tissue scRNA-seq data, confirming the accuracy of their annotations. Despite the overall similarity in cell-type intrinsic profiles of protein-coding genes between the two, some divergences were observed. Notably, the team's snRNA-seq captured a relatively lower proportion of lymphocytes, suggesting that scRNA-seq may oversample immune cells.

Fig.4 Cell-type diversity and cell-intrinsic profiles are consistent between snRNA-seq and scRNA-seq. (Eraslan, 2023)

Creative Biolabs' Service

Single Nuclei RNA Sequencing Service

At Creative Biolabs, we present an assortment of exceptional single-cell nuclei RNA sequencing services, specifically designed to bolster biomedical research worldwide. Our offerings seamlessly integrate with intricate tissues like the brain, heart, and kidney, and even with scarce cryopreserved samples. This compatibility paves the way for an in-depth investigation of tumor cell diversity and disease-causing mechanisms at the granularity of the individual cell.

Learn more

At Creative Biolabs, we provide a pioneering single nuclei RNA sequencing service that allows a detailed inspection of gene expression at the level of individual nuclei. This innovative technique illuminates cellular diversity and function, providing an essential understanding of complex biological processes. Our method leverages top-tier technology and techniques to guarantee superior results and can accommodate a wide array of sample types, including rare and cryopreserved specimens. Our proficient team is dedicated to delivering efficient and streamlined service to propel your scientific explorations.

For any information, please contact us.

Reference

1. Eraslan, G.; et al. Single-nucleus cross-tissue molecular reference maps toward understanding disease gene function. Science. 2023, 376(6594): eabl4290.