An engineered CRISPR/Cas9 mouse line for simultaneous readout of lineage histories and gene expression profiles in single cells

Sarah Bowling, Duluxan Sritharan, Fernando G. Osorio, Maximilian Nguyen, Priscilla Cheung, Alejo Rodriguez-Fraticelli, Sachin Patel, Yuko Fujiwara, Bin E. Li, Stuart H. Orkin, Sahand Hormoz, Fernando D. CamargobioRxiv 2019 · DOI

TL;DR

This paper introduces CARLIN, a novel CRISPR-based mouse model that allows simultaneous, inducible lineage tracing and gene expression profiling in single cells. This system generates a high diversity of unique DNA barcodes, enabling researchers to reconstruct complex cellular histories and identify functional gene signatures in vivo, particularly in the context of stem cell biology and hematopoietic regeneration.

What Problem Does This Paper Address?

Understanding how cells develop and differentiate over time (lineage tracing) is crucial for many biological questions, especially in stem cell research. Traditional methods, like using dyes or fluorescent markers, are limited in their ability to track many cells simultaneously and provide only lineage information. While newer methods using DNA barcodes have emerged, they often require cells to be extracted and manipulated, or they only capture lineage information without gene expression data. This leaves a gap in our ability to simultaneously track cell ancestry and understand the molecular mechanisms driving their behavior in their natural environment. The paper aims to bridge this gap by providing a system that can record both lineage history and gene expression in individual cells within a living organism.

Methods

The core innovation is the CARLIN (CRISPR Array Repair LINeage tracing) mouse line. This system leverages CRISPR/Cas9 gene editing to create unique DNA "barcodes" within cells. Here's how it works:

  • CRISPR-Cas9 for Barcoding: The system uses CRISPR/Cas9 to induce double-strand DNA breaks in a specific array of target sites within the cell's genome.
  • Inducible Editing: Cas9 expression is controlled by doxycycline (Dox), meaning barcoding can be initiated at specific time points during development or adulthood. This allows for temporal control over lineage labeling.
  • High Diversity Barcodes: The CARLIN array contains multiple target sites, and the repair of the DNA breaks by the cell's natural repair mechanisms (non-homologous end joining) generates a vast diversity of unique DNA sequences (alleles). The authors estimate up to 44,000 distinct alleles can be generated.
  • Transcribed Barcodes: Crucially, these barcodes are transcribed into RNA, allowing them to be read out alongside the cell's transcriptome using single-cell RNA sequencing.
  • Sequential Labeling: The system supports sequential pulses of Dox, enabling the accumulation of multiple layers of lineage information and the construction of more complex phylogenetic trees.
  • Bioinformatic Pipeline: A novel pipeline was developed to analyze the CARLIN alleles from sequencing data, accounting for their location and variations.

Datasets

The study primarily uses data generated from CARLIN mice and CARLIN-derived embryonic stem (ES) cells.

  • CARLIN ES Cells: Used for initial validation of inducible editing, dose-response experiments, and testing sequential labeling capabilities.
  • CARLIN Mice Tissues: RNA was collected from various tissues (granulocytes, skin, brain, heart, liver, intestine, colon, lung, bone marrow HSCs, MPPs, B-cells) of CARLIN mice at different developmental stages and after perturbation (5-FU treatment).
  • Single-cell RNA Sequencing Data: Generated from pooled cells from multiple bones and from various tissues to capture both lineage barcodes and full transcriptional profiles.

No external benchmark datasets were explicitly mentioned for comparison, but the study establishes a new benchmark for simultaneous lineage and transcriptomic profiling in vivo.

Key Insights and Ideas

  • Simultaneous Lineage and Transcriptomics: The ability to get both lineage history and gene expression from the same cell in vivo is a major leap, allowing researchers to connect "who is who" with "what are they doing."
  • Inducible Control is Key: The Dox-inducible nature of CARLIN provides unprecedented temporal control over lineage labeling, allowing for precise timing of barcoding events.
  • High Barcode Diversity: The generation of up to 44,000 unique barcodes is sufficient for tracking complex clonal dynamics in various tissues and over extended periods.
  • Uncovering Clonal Bottlenecks: CARLIN revealed a previously unappreciated clonal bottleneck in hematopoietic stem cell (HSC) regeneration after injury, suggesting that only a small subset of HSCs are responsible for repopulating the blood system.
  • Identifying "Childless" HSCs: The study identified HSC clones that appear to have no detectable progeny, potentially representing a novel type of dormant or specialized HSC.
  • Tissue Phylogeny: Multiple pulses of CARLIN labeling allowed for the reconstruction of tissue phylogenetic relationships, revealing how different tissues are related based on shared cellular ancestry.

Key Results

  • Inducible and Diverse Editing: CARLIN successfully generated a high diversity of edited CARLIN alleles (up to 44,000 estimated) in an inducible manner in ES cells and various adult mouse tissues.
  • Dose and Duration Dependence: Editing efficiency and allele diversity were directly correlated with the dose and duration of doxycycline induction.
  • Sequential Labeling for Complex Histories: Sequential Dox pulses allowed for multi-level lineage reconstruction, demonstrating the potential for building hierarchical histories.
  • In Vivo Barcoding: CARLIN successfully barcoded cells in vivo across multiple tissues, with editing rates ranging from 31% to 88% in most tissues.
  • Hematopoietic Regeneration Bottleneck: In response to 5-FU treatment, CARLIN revealed a significant reduction in clonal diversity and identified a small number of highly active HSC clones responsible for blood replenishment.
  • Bias in Bone Colonization: Analysis of fetal liver-derived HSCs showed significant biases in their representation across different bones, suggesting niche-dependent expansion or colonization.
  • Identification of Gene Signatures: The system enabled the identification of potential gene signatures associated with "active" HSC states, including genes like Plac8.

Limitations and Future Work

  • Editing Efficiency: While CARLIN can generate diverse barcodes, the fraction of cells edited (16-74%) and captured (32-63%) could be further improved. Future work could focus on optimizing Dox timing/dose, using more highly expressed loci, or incorporating additional CARLIN arrays.
  • Barcode Diversity for Whole Tissues: While 44,000 alleles are sufficient for many applications, higher diversity might be needed for analyzing very large or complex adult tissues. Modifications like using homing guide RNAs or combining with Cre-based tracing could increase diversity.
  • Potential for PCR/Sequencing Errors: The authors acknowledge that errors in PCR/sequencing could lead to loss of reads, impacting accuracy.
  • Future Applications: The authors suggest extending CARLIN to use environmentally-sensitive promoters for stimuli other than doxycycline (e.g., pathogen exposure, nutrient intake) and applying it to other complex biological systems. The identification of novel gene signatures for HSC states also opens doors for further functional studies.

Summary updated

Abstract

Abstract Tracing the lineage history of cells is key to answering diverse and fundamental questions in biology. Particularly in the context of stem cell biology, analysis of single cell lineages in their native state has elucidated novel fates and highlighted heterogeneity of function. Coupling of such ancestry information with other molecular readouts represents an important goal in the field. Here, we describe the CARLIN (for C RISPR A rray R epair LIN eage tracing) mouse line and corresponding analysis tools that can be used to simultaneously interrogate the lineage and transcriptomic information of single cells in vivo. This model exploits CRISPR technology to generate up to 44,000 transcribed barcodes in an inducible fashion at any point during development or adulthood, is compatible with sequential barcoding, and is fully genetically defined. We have used CARLIN to identify intrinsic biases in the activity of fetal liver hematopoietic stem cell (HSC) clones and to uncover a previously unappreciated clonal bottleneck in the response of HSCs to injury. CARLIN also allows the unbiased identification of transcriptional signatures based on in vivo stem cell function without a need for markers or cell sorting.
BibTeX
@misc{bowling2019engineered,
  author = {Sarah Bowling and Duluxan Sritharan and Fernando G. Osorio and Maximilian Nguyen and Priscilla Cheung and Alejo Rodriguez-Fraticelli and Sachin Patel and Yuko Fujiwara and Bin E. Li and Stuart H. Orkin and Sahand Hormoz and Fernando D. Camargo},
  title = {{An engineered CRISPR/Cas9 mouse line for simultaneous readout of lineage histories and gene expression profiles in single cells}},
  year = {2019},
  doi = {10.1101/797597},
  howpublished = {bioRxiv},
}
Figure