Introduction: As cancer forms, single cells acquire advantageous changes that make them fitter than their neighbors and allow them to grow more effectively. This process generates a population of cells that are all derived from the same parent cell. This population, which we refer to as a tumor clone, will contain the same mutations as the initiating cell. These mutations can be tracked through genetic lineage tracing to determine which cells are clonally related in an early tumor.
Genetic lineage tracing can use somatic mutations in genomic or mitochondrial DNA to infer clonal relationships; however, somatic mutation tracking in genomic DNA at a single cell resolution remains low-throughput and does not capture cell state in the same cell. Here, we focus on genetic lineage tracing with mitochondrial mutations in RNA sequencing libraries to retrospectively trace cellular lineage and study cell state changes without the need for artificial markers (1,2). Current mitochondrial lineage tracing techniques require tissue dissociation and therefore, prevents analysis of cells within the tissue architecture or lineage relationships between adjacent cells. To address this gap, we developed a sequencing-based clonal tracing technology to study in situ cellular relationships in human tissue. We applied our technology to study human Barrett’s esophagus, a common form of metaplasia that is a precursor lesion for esophageal adenocarcinoma, to analyze transcriptional differences and visualize clonal organization in premalignant gastroesophageal tissue.
Materials and
Methods: Our approach combines high-resolution spatial transcriptomic sequencing with Slide-Seq and mitochondrial mutation lineage tracing on the same sample. We sectioned endoscopic biopsies onto DNA barcoded beads to introduce spatial barcodes into mature mRNA (3). We processed spatial transcriptomic data using the Slide-seqV2 pipeline for single-cell spatial gene expression analysis. For cell type annotation, robust cell type decomposition (RCTD) was used to predict cell type proportions for each barcoded bead using a single-cell RNA-sequencing (scRNA-seq) reference containing transcriptomic data for cell types of interest (4).
To achieve high coverage of mitochondrial transcripts, we then performed a PCR side reaction with primers tiling the mitochondrial genome to generate a library that preserves sequences from the spatial array (5). Spatial cDNA libraries were enriched by more than 10-fold for mitochondrial transcripts before sequencing with Illumina NextSeq500/550. Mitochondrial transcripts were analyzed for SNVs using maegatk. Estimated variant allele frequencies (VAFs) were mapped onto matched spatial coordinates in R to visualize clonal architecture at a single-cell resolution. Finally, we compared our approach to bulk detection of mitochondrial mutations in genomic data from an adjacent slide. We confirmed the presence of the mutations captured in the RNA libraries by detection in the bulk genomic DNA libraries, indicating that we captured real variants in the mitochondrial DNA.
Results, Conclusions, and Discussions: We applied our method to analyze two gastroesophageal pinch biopsies for gene expression and lineage relationships. Paired spatial transcriptomics and lineage tracing projected variant allele frequencies onto matched spatial coordinates to reveal clonal organization. We detected numerous non-germline lineages following enrichment to support the polyclonality of nondysplastic Barrett’s esophagus tissue (Figure 1A). We found two spatially-restricted lineages (Moran’s I values of 0.302 and 0.260) in nondysplastic Barrett’s esophagus, each localized to a single gland (Figure 1B). We next used RCTD to predict cell type proportions for each spatially-barcoded bead. To do so, we used a scRNA-seq reference containing expression data for eight gastroesophageal tissue types (6). RCTD classified the nondysplastic lineages as primarily composed of Barrett’s esophagus cells (Figure 1C). Although broadly classified as Barrett’s esophagus tissue, cell type annotation by marker gene expression showed spatial lineages 15777_G>C and 3054_G>C had distinct cell type compositions. This suggests our lineage tracing technology can identify unique stem cell compartments from which disease progenitors may originate (Figure 1D).
We also identified two spatially restricted lineages (Moran’s I values of 0.93 and 0.95) in gastric tissue adjacent to high grade dysplasia (Figure 1E). Each lineage contained both Barrett’s esophagus and gastric cell types as determined by RCTD (Figure 1F). We validated these findings with whole exome sequencing to confirm agreement between the heteroplasmic allele frequencies estimated by mitochondrial variant enrichment with bulk allele frequency estimates. Overall, we present a new method for genetic lineage tracing by mitochondrial mutation detection in spatial transcriptomic libraries. This tool will be broadly useful for studying clonal dynamics in tissues and disease, enabling the formation of spatially-aware phylogenetic trees and prediction of the transcriptional drivers of clonal expansion.
Acknowledgements (Optional):
References:
1. Ludwig, L. S. et.al. Lineage Tracing in Humans Enabled by Mitochondrial Mutations and Single-Cell Genomics. Cell, 176(6). (2019) 2. Hayakawa,Y.,Nakagawa,H.,Rustgi,A.K.,Que,J.&Wang,T.C.Stem cells and origins of cancer in the upper gastrointestinal tract. Cell Stem Cell 28, 1343–1361. (2021) 3. Rodriques, S. G., Stickels, R. R., Goeva, A., Martin, C. A., Murray, E., Vanderburg, C. R., Welch, J., Chen, L. M., Chen, F., & Macosko, E. Z.. Slide-seq: A scalable technology for measuring genome-wide expression at high spatial resolution. Science (New York, N.Y.), 363(6434), 1463–1467. (2019) 4. Cable, D. M., Murray, E., Zou, L. S., Goeva, A., Macosko, E. Z., Chen, F., & Irizarry, R. A. Robust decomposition of cell type mixtures in spatial transcriptomics. Nature biotechnology, 40(4), 517–526. (2022) 5. Miller, T. E., Lareau, C. et al. Mitochondrial variant enrichment from high-throughput single-cell RNA sequencing resolves clonal populations. Nature biotechnology, 40(7), 1030–1034. (2022) 6. Gier, R. A. et al. Clonal cell states link Barrett’s esophagus and esophageal adenocarcinoma. bioRxiv (2023)
