Each tumor contains malignant cells that differ in genotype, phenotype, and in their interactions with the tumor micro-environment (TME). This results in distinct integrated cellular states that govern intra-tumor heterogeneity (ITH), a central challenge of cancer therapeutics. Dozens of recent studies have begun to describe ITH by single cell RNA-seq, but each study typically profiledonly a small number of tumors and provided a narrow view of transcriptional ITH. Here, we curate, annotate and integrate the data from 77 different studies to reveal the patterns of ITH across 1,163 tumor samples covering 24 tumor types. Focusing on the malignant cells, we find thousands of transcriptional ITH programs that can be described by 41 consensus meta-programs (MPs), each consisting of dozens of genes that are coordinately upregulated in subpopulations of cells within many different tumors. The MPs cover diverse cellular processes and differ in their cancer-type distribution. General MPs associated with processes such as cell cycle and stress vary within most tumors, while context-specific MPs reflect the unique biology of particular cancer types, often resembling developmental cell types and suggesting the co-existence of variable differentiation states within tumors. Some of the MPs are further associated with overall tumor proliferation or immune state, highlighting their potential clinical significance. Based on functional similarities among MPs, we propose a set of 11 hallmarks that together account for the majority of observed ITH programs. Given the breadth and scope of the investigated cohort, the MPs and hallmarks described here reflect the first comprehensive pan-cancer description of transcriptional ITH.

Generation of periodic patterns is fundamental to the differentiation of multiple tissues during development. How such patterns form robustly is still unclear. The Drosophila eye comprises ∼750 units, whose crystalline order is set during differentiation of the eye imaginal disc: an activation wave sweeping across the disc is coupled to lateral inhibition, sequentially selecting pro-neural cells. Using mathematical modelling, here we show that this template-based lateral inhibition is highly sensitive to spatial variations in biochemical parameters and cell sizes. We reveal the basis of this sensitivity, and suggest that it can be overcome by assuming a short-range diffusible activator. Clonal experiments identify Scabrous, a previously implicated inhibitor, as the predicted activator. Our results reveal the mechanism by which periodic patterning in the fly eye is stabilized against spatial variations, highlighting how the need to maintain robustness shapes the design of patterning circuits.

Tissue stem cells produce a constant flux of differentiated cells with distinct proportions. Here, we show that stem cells in colonic crypts differentiate early to form precisely 1:3 ratio of secretory to absorptive cells. This precision is surprising, as there are only eight stem cells making irreversible fate decisions, and so large stochastic effects of this small pool should have yielded much larger noise in cell proportions. We use single molecule FISH, lineage‐tracing mice and simulations to identify the homeostatic mechanisms facilitating robust proportions. We find that Delta‐Notch lateral inhibition operates in a restricted spatial zone to reduce initial noise in cell proportions. Increased dwell time and dispersive migration of secretory cells further averages additional variability added during progenitor divisions and breaks up continuous patches of same‐fate cells. These noise‐reducing mechanisms resolve the trade‐off between early commitment and robust differentiation and ensure spatially uniform spread of secretory cells. Our findings may apply to other cases where small progenitor pools expand to give rise to precise tissue cell proportions.

No abstract