This presentation by Dr. Fay Chen from the Broad Institute and Harvard introduces advanced spatial genomics tools, primarily focusing on Slide-seq and Slide-tags, to unravel cellular communities and tissue organization in health and disease, with a particular emphasis on cancer. The talk begins by contextualizing the evolution of genomics towards understanding spatial relationships of gene expression. It details Slide-seq, a high-resolution spatial transcriptomics method using barcoded beads, and its extension, Slide-DNA-seq, for concurrent DNA and RNA capture. Dr. Chen then presents Slide-tags, a novel approach that spatially barcodes individual nuclei before dissociating them for standard single-cell sequencing, thereby achieving high-quality single-cell resolution with retained spatial context across various modalities (RNA, DNA, TCR, chromatin accessibility). The utility of Slide-tags is powerfully demonstrated through a study of metastatic melanoma, revealing how tumor clonal heterogeneity dictates immune microenvironments, identifies recurrent cellular "niches," and influences gene expression programs. The presentation concludes by discussing future advancements, including image-free spatial reconstruction directly from sequencing data, enabling unprecedented scalability for whole-organ analysis, and the integration of diverse genomic and epigenomic measurements.
Understanding Cellular Communities and Tissues with Spatial Genomics
0:03 Introduction to Dr. Fay Chen: Anoja Pereira introduces Dr. Fay Chen, highlighting his pioneering work in expansion microscopy and Slide-seq for spatial genomics.
1:58 Speaker's Introduction & Disclosures: Dr. Chen apologizes for his remote presence and discloses his prior role as co-founder of Curio Biosciences, a company formed to disseminate Slide-seq and Slide-tags reagents, which has since been acquired by Takara.
2:46 Revolution in Genomics: Dr. Chen discusses the current revolution in genomics, emphasizing the ability to understand molecular states of cells directly within tissues, moving beyond traditional pathology.
5:00 Cellular Programs & Multicellular Circuits: He explains that cells have intrinsic molecular programs (e.g., identity, function, cell division) influenced by multicellular circuits within tissues, which ultimately govern tissue function and structure.
6:45 Cancer as a Case Study: Cancer is presented as an ideal application for spatial genomics due to its diverse cell types and states, which are heavily influenced by the microenvironment and intrinsic genomic alterations.
9:08 Development of Spatial Genomics Tools: Dr. Chen's lab focuses on two main technology types: spatial capture (Slide-seq) and direct in-situ sequencing.
10:32 Slide-seq Technology: Slide-seq, a high-resolution spatial transcriptomics technology, uses barcoded beads deposited on cover glass to capture mRNA, providing spatial localization at 10-micron resolution.
12:12 Early Slide-seq Applications: Initial experiments with Slide-seq on mouse hippocampus demonstrated the assignment of molecular identities to cells based on their transcriptomes in spatial context.
13:30 Slide-DNA-seq: The technology was extended to Slide-DNA-seq, capturing DNA alongside RNA to study genomic alterations (e.g., chromosomal copy number variations) within the tissue.
17:15 Explaining Gene Expression Variance: Using Slide-seq data from human melanoma, the team performed variance decomposition to understand how gene expression in tumor cells is influenced by intrinsic clonal identity versus extrinsic factors like local tumor cell density.
21:00 Overcoming Cell Mixture Challenges: Dr. Chen acknowledges the "cell mixture" problem in capture-based methods like Slide-seq, where pixels may contain RNA from multiple cells.
23:19 Introducing Slide-tags: To address the mixture problem and integrate single-cell advantages, Slide-tags was developed. This method spatially barcodes cell nuclei before dissociation and single-cell sequencing.
23:49 Slide-tags Mechanism: Photocleavable barcodes from an array diffuse into tissue nuclei, which are then isolated and profiled using standard single-cell platforms (e.g., 10x Genomics), providing high-quality single-cell data with spatial coordinates.
26:10 Quality Preservation & Recovery: Slide-tags does not affect single-nucleus RNA-seq quality, maintaining cell type proportions and molecule counts. Initial recovery was about 10%, improved to 25-30% of nuclei from tissue.
27:42 Versatility of Slide-tags: Slide-tags is compatible with various single-nucleus sequencing modalities, including RNA expression, TCR sequencing, copy number inference, and chromatin accessibility (ATAC-seq).
28:32 Melanoma Metastasis Study with Slide-tags: A study on human lymph node metastasis post-checkpoint blockade showed spatially segregated tumor subclones with distinct transcriptional states and unique cellular environments.
29:50 Linking Genetic Differences to Spatial States: Copy number variation analysis confirmed genetic differences driving spatially distinct tumor cell compartments.
30:49 Neighborhood Analysis & Immune Interactions: Analysis revealed different cellular environments for tumor subclones, with some enriched for CD8 T-cells and others depleted.
31:36 TCR Clonality & Tumor Subclones: TCR sequencing demonstrated specific T-cell receptor clones associating with particular tumor subclones, suggesting the tumor lineage significantly shapes immune interactions.
33:08 Melanoma Differentiation Phenotype: The study linked tumor subclones to changes in differentiation (e.g., mesenchymal-like transition), correlating with T-cell presence or depletion.
35:45 Summary of Melanoma Findings: Slide-tags revealed significant transcriptional and clonal heterogeneity in melanoma, specific TCR clone associations with tumor subclones, and the influence of intrinsic tumor properties on immune interactions.
36:44 Scaling Up & Recurrent Niches: The lab collected a large Slide-tags dataset (300k nuclei, 27 biopsies) to identify recurrent cellular "niches" or neighborhoods (e.g., immune effector hubs, myeloid/stromal, tertiary lymphoid structures).
39:05 Impact of Neighborhoods on Gene Expression: Linear regression was used to determine how cellular neighborhoods affect gene expression programs within malignant cells (e.g., interferon response associated with cytotoxic T-cell neighborhoods).
45:41 Defining Extracellular Signaling Environments: A new computational approach defines recurrent extracellular signaling programs (soluble ligands) that mediate communication between cell neighborhoods and intrinsic gene expression states.
47:52 Implications of Spatial Data: Spatial data enables understanding recurrent spatial neighborhoods, context-specific gene expression, extracellular signaling, and is crucial for learning rules of tissue organization, therapy development, and tissue engineering.
49:18 Slide-tags Technology Summary: Slide-tags offers single-cell resolution with spatial context, compatibility with existing single-cell workflows, and applicability to nearly every single-cell experiment type.
49:57 Future Directions: Spatial DNA Sequencing: Progress on Slide-tags for DNA sequencing provides high genome coverage per cell, enabling SNV calling, mutational signature analysis, and inference of local lineage information.
55:01 Future Directions: Scaling Throughput & Image-Free Reconstruction: A major bottleneck, array manufacturing, is addressed by a novel method to reconstruct spatial information without imaging, using only sequencing data, allowing for large-scale arrays (up to 7cm for human brain slabs).
1:06:42 Conclusion & Outlook: The decreasing cost of sequencing will make spatial measurements ubiquitous, enabling analysis of whole human organs, multi-omic (genomic/epigenomic) profiling, and efforts to infer dynamic information from static snapshots.
1:07:42 Q&A Highlights: Discussions cover integrating TMB/driver mutation status, the potential for using circulating biomarkers to reflect specific microenvironments, probe-based enrichment for targeted sequencing, and practical challenges of tissue transportation and FFPE tissue compatibility.