TL;DR
This paper reveals that individual cell clones, even within the same cell type, can develop and maintain distinct, heritable gene expression patterns. These differences, driven by heritable epigenetic states like chromatin accessibility, contribute significantly to cellular diversity and have implications for understanding development, aging, and disease.
What Problem Does This Paper Address?
The prevailing view is that variability in gene expression between cells of the same type is primarily due to random fluctuations (stochastic transcription) or temporary cell states. This paper challenges that by investigating whether long-term, heritable differences in gene expression can arise between clonal populations of cells. The authors hypothesize that such differences could be a significant, yet underappreciated, source of cellular diversity, impacting how tissues function and respond to various processes.
Methods
The researchers employed a multi-pronged approach:
- Single-cell RNA sequencing (scRNA-seq): This was the core technology, used to profile the gene expression of thousands of individual cells from clonal populations. They utilized both Smart-seq2 and Smart-seq3 protocols for high-sensitivity, full-transcript sequencing.
- TCR sequencing: T cell receptor (TCR) sequences were used as genetic barcodes to identify and track clonally expanded populations.
- ATAC-seq: This technique was used to map accessible regions of chromatin, providing insights into epigenetic regulation and its link to gene expression differences between clones.
- Machine Learning (SVM classifier): A linear Support Vector Machine (SVM) was trained to predict the clonal origin of single cells based on their gene expression signatures, demonstrating the distinctness of these clonal profiles.
- Principal Component Analysis (PCA) and UMAP: These dimensionality reduction techniques were used to visualize and analyze the complex gene expression data, revealing patterns of clonal separation and differentiation states.
- Genomic Copy Number Variation (CNV) analysis: This was performed to rule out genetic mutations as the primary driver of heritable gene expression differences.
Datasets
The study utilized two main biological systems:
- Human Lymphocytes:
- Clonally expanded CD8+ T cells from healthy donors after vaccination with yellow fever virus vaccine (YFV-17D).
- Data collected at various time points post-vaccination (e.g., Day 180, Day 1286).
- Samples included both in vivo (circulating blood) and in vitro expanded clones.
- Mouse Central Nervous System (CNS) Cells:
- Barcoded cells from the somatosensory cortex of mouse brains at postnatal day 14, following in utero lentiviral barcode library injection.
- Data collected from two mouse brains.
Key Insights and Ideas
- Clonal Identity as a Source of Diversity: The most striking finding is that cells within the same type can have distinct, heritable transcriptional "personalities" based on their clonal origin. This adds a new layer to our understanding of cellular heterogeneity beyond stochasticity.
- Epigenetics as the Driver: Heritable differences in gene expression are strongly linked to variations in chromatin accessibility, suggesting that epigenetic modifications play a crucial role in maintaining these clonal states.
- Long-Term Memory: These heritable transcriptional states can persist for extended periods in vivo (over a year), indicating a stable form of cellular memory.
- Two Layers of Heritable Traits: The study identified two types of heritable clonal traits: (1) broad differentiation biases involving coordinated gene blocks, and (2) more subtle, sometimes sporadic, gene expression differences within clones.
- Implications for Tissue Homeostasis and Disease: The existence of such heritable diversity has significant implications for how tissues maintain homeostasis, age, and respond to disease, potentially influencing clonal competition and selective pressures.
Key Results
- Human Lymphocytes:
- Identified hundreds of clonally variable genes between T cell clones from the same donor.
- Machine learning classifiers could predict clonal identity with high accuracy (80-100%) based on gene expression alone.
- Clonally related cells exhibited biases along a differentiation continuum that persisted for years.
- Mouse Brain Cells:
- Found interclonally differentially expressed genes across multiple cell types (neurons, astrocytes, oligodendrocytes, immune cells).
- The number of differentially expressed genes correlated with the number of clones and cells per clone.
- Demonstrated that clonally heritable gene expression patterns are present in diverse cell types and species.
- Chromatin Accessibility:
- 62% of differentially accessible chromatin regions (CREs) were located near genes showing interclonal expression differences.
- Correlated CRE variability with transcriptional differences, identifying thousands of CRE-gene pairs with correlated activity.
Limitations and Future Work
- Limitations:
- The study acknowledges that while genetic variation is ruled out as the primary driver, subtle genetic differences or allele-specific regulation might still play a minor role.
- The in vitro expansion of T cells might introduce some artifacts, though the in vivo data largely supports the findings.
- The mouse brain dataset had smaller clone sizes compared to the human T cell data, potentially limiting the detection of subtle differences in some cell types.
- Future Work:
- Investigating the precise epigenetic mechanisms that establish and maintain these heritable states.
- Exploring how these clonal differences impact tissue function, immune responses, and disease progression in more detail.
- Understanding the evolutionary implications of this heritable cellular diversity in multicellular organisms.
- Further dissecting the "two layers" of heritable traits and their distinct roles.
Summary updated