Fig. 1

Profiling the cartography of the cancer ecosystem. (A) Hematoxylin and eosin (H&E) and chromogen-based immunohistochemical stained images of a primary cutaneous melanoma, showing the result of a machine learning algorithm capable of detecting image features that distinguish tumor from stromal and necrotic regions, and quantify the number of cancer cells expressing key biomarkers, as well as counts of tumor-infiltrating immune cells. (B) Single-cell isolation approaches include high-throughput microfluidic technologies and laser capture microdissection (LCM), adapted from [3]. (C) Technologies are now able to dissect single tumor cells at previously unattainable resolutions exemplified by approaches to analyze the genome, transcriptome and epigenome, adapted from [4]. (D) Gene expression patterns can also be spatially resolved across microenvironmental conditions. Schematic workflow of spatial transcriptomic analyses using fluorescence and Visium spatial technologies. Schematic of the RNAscope assay. After the tissue is permabilized to allow probe access, target RNA-specific oligonucleotide probes (Z) are hybridized to the RNA target template. Once adjacent target (Z) probe-pairs hybridize to the RNA template, the preamplifier binds to the Z probe, followed by the binding multiple amplification molecules. Each probe is conjugated to a different fluorophore which are detected using a fluorescent microscope. Workflow adapted from [5]. Shown H&E of a primary acral melanoma and the associated 4-plex RNAscope smFISH image, staining for 4 positive control genes. In the 10X Visium platform, tissue samples are sectioned and aligned onto the capture areas of the gene expression slides. Once the tissue is permeabilized, mRNA are hybridized to the capture probes and converted into a spatially barcoded cDNA library by reverse transcription. Sequencing the library then allows the cDNA barcodes to associate to a slide coordinate within the original tissue, mapping the gene expression profile to its spatial localization. Original image adapted with permission from 10x Genomics graphic. Shown H&E of a T3 primary lung adenocarcinoma and the associated spatially resolved gene expression plot, showing 6 mRNA clusters