Spatial tech and the cortex

Spatial omics works like a street map for tissue: it measures genes in place instead of grinding cells up and losing their address. That is letting brain researchers test whether the cortex is organized by broad gradients, sharp borders, or both. (cell.com) In May 2025, a Nature study mapped more than 18 million single cells across eight human fetal cortical areas and seven developmental time points using multiplexed error-robust fluorescence in situ hybridization, or MERFISH. The authors reported a gradual anterior-to-posterior transition across most areas and a separate sharp border between primary and secondary visual cortex by gestational week 20. (nature.com) That result speaks to an old argument in cortical development. Reviews in Neuron describe the “protomap” model as cortex-intrinsic patterning and the “protocortex” model as stronger shaping by outside input such as thalamic connections. (cell.com) The gradient side of that debate got another data point this week. A Science paper published on April 16, 2026, reported an “opposing molecular gradient axis” in marmoset cortex, with one gradient radiating from allocortical and periallocortical regions and another from primary sensory areas. (science.org) The Science team said association cortex sits where those two gradients meet, and that the pattern lines up with gene expression, cell composition, cortical borders, thalamic gene expression, and thalamocortical connectivity. The paper also said the gradients are conserved across humans, macaques, marmosets, and mice, but are refined after birth. (science.org) That is the backdrop for the social-media shorthand “Pr-Al Axis.” The published paper describes a cortex-wide axis anchored at allocortical and primary sensory ends; the post’s label appears to be an informal summary of that broader opposing-gradient framework rather than the paper’s formal name. (science.org) The human fetal atlas does not reduce the cortex to a single rule. Its authors found one mode that looks like a smooth molecular slope across most of the cortex and another mode that looks like a step change at the V1-V2 boundary, a mix that cuts across the old either-or framing. (pmc.ncbi.nlm.nih.gov) Earlier work had already pointed in that direction. A 2021 Nature atlas of cortical arealization reported dynamic molecular signatures and referenced a posterior-high to anterior-low expression gradient in neocortex during development. (nature.com) What changed is the resolution. Reviews of spatial and spatiotemporal omics describe a shift from bulk readouts and dissociated single cells to methods that keep molecular measurements tied to exact positions in tissue, making it possible to compare gradients, borders, layers, and connections in the same anatomical frame. (nature.com) So the current picture of the cortex looks less like a patchwork of isolated modules and more like a map with slopes and edges. Spatial data are turning that metaphor into something researchers can measure, area by area and cell by cell. (science.org)