An opposing molecular gradient axis underlies primate cortical organization.

The principles organizing cellular diversity and connectivity in primate brains remain elusive. By integrating spatial transcriptomics, magnetic resonance imaging, and retrograde labeling in marmosets, we identified two opposing molecular gradients that undergo postnatal refinement, emanating from allocortices and primary sensory cortices, respectively. These gradients reconcile conflicting hypotheses on cortical expansion and characterize distinct cortical areas. Cortical gradients align with thalamic gene expression and thalamocortical projection patterns. At gradient intersections, the default mode network and frontal pole exhibited similar molecular features in humans and marmosets, despite species-specific differences in functional connectivity. Comparative analysis of gradient-related genes showed that marmoset and human auditory cortices are highly similar but differ from those of macaques, potentially reflecting complex vocalization. Together, these opposing gradients represent a fundamental organizing principle of the primate cortex. Editor's summary: The expansion of the cerebral cortex during primate evolution has resulted in remarkable complexity in the cellular organization and connectivity of this brain region, yet the fundamental principles governing the organization of the expanded cortex remain unclear. Huang et al. developed a single-cell whole-brain spatial transcriptome database, applied retrograde neuronal tracing, and used functional magnetic resonance imaging datasets to construct a three-dimensional integrated multimodal atlas of the entire marmoset brain. Two distinct molecular gradients emanated from allocortical and primary sensory regions, converging at association cortices. Cross-species comparisons revealed similarities and differences among marmosets, macaques, and humans. The study provides a comprehensive, multimodal characterization of primate cortical organization. —Mattia Maroso INTRODUCTION: The primate cerebral cortex has expanded into a complex mosaic of specialized areas and networks that support advanced cognition. However, the fundamental principles governing this complex cortical organization remain elusive, which has led to conflicting hypotheses regarding cortical expansion. The dual-origin hypothesis posits that the cortex expanded from two evolutionarily ancient allocortical regions and progressively differentiated toward the most specialized six-layered primary sensory areas (the koniocortex). Conversely, alternative theories propose that primary sensory areas served as the early anchors for cortical expansion. Resolving this conflict requires a holistic view that integrates gene expression, cellular architecture, and brain-wide connectivity. RATIONALE: The common marmoset (Callithrix jacchus) offers a solution to this challenge. It retains key primate brain characteristics yet has a smooth (lissencephalic) cortex, bypassing the analytical difficulties caused by the complex folding seen in the cortices of larger primates. Leveraging this advantage, we integrated whole-brain, single-cell–resolution spatial transcriptomics and single-nucleus RNA sequencing with magnetic resonance imaging (MRI) and neuronal tracing data. This multimodal synthesis allowed us to uncover the principles underlying cortical organization. RESULTS: Our analysis uncovered a fundamental axis of cortical organization defined by two opposing molecular gradients that capture the dominant patterns of cortical gene expression, cell composition, and hierarchy. One gradient radiates from allocortical and periallocortical regions (e.g., the piriform and entorhinal cortices), whereas the opposing gradient originates from primary sensory areas, with the association cortex residing at their intersection. Conserved across humans, macaques, marmosets, and mice, these gradients reconcile the conflicting hypotheses by demonstrating that both the allocortical and primary sensory areas act as anchors at opposite ends of a single organizational axis. Although present at birth, these gradients undergo prominent postnatal refinement, which suggests that they are actively shaped by sensory experience. Molecular shifts along these gradients underpin cortical parcellation, where sharp transitions in cell composition and gene expression align with cortical area borders and reveal previously unrecognized subdivisions. Notably, these cortical gradients are mirrored in thalamic gene expression patterns and align with thalamocortical connectivity. This molecular coupling between the cortex and thalamus is significantly stronger in marmosets than it is in mice, which highlights the evolutionary advances in primate-specific thalamocortical integration. At the gradient convergence zone, the putative default mode network (DMN) and the frontal pole (area 10) in marmosets share similar molecular signatures despite their weak functional connectivity, suggesting that this molecular identity evolved before the strong connectivity observed in humans. Finally, comparative analysis of gradient-related genes revealed that the marmoset auditory cortex more closely resembles the human auditory cortex compared with that of the macaque, likely reflecting shared neural mechanisms for complex vocal communication. CONCLUSION: Through multimodal analysis, we identify an opposing molecular gradient axis as a fundamental principle of primate cortical organization that resolves debates regarding cortical expansion. This axis links molecular profiles to anatomical and functional architecture, offering a precise biological basis for delineating cortical boundaries, elucidating cortical-subcortical relationships, characterizing functional networks, and identifying species-specific molecular specializations. Together, this work establishes the opposing gradient axis as a key organizational backbone and presents a foundational multimodal resource for understanding primate brain organization and evolution. Opposing molecular gradients of the cerebral cortex.: By integrating whole-brain spatial transcriptomics with MRI and neuronal tracing in marmosets, we revealed an opposing molecular gradient axis as a fundamental principle of cortical organization. Undergoing active postnatal refinement, these gradients serve as a key organizational backbone for delineating cortical boundaries, elucidating cortical-subcortical relationships, characterizing functional networks, and identifying species-specific molecular patterns. 3D, three-dimensional; PC1, principal component 1; Al, allocortical/periallocortical; Pr, primary sensory; P0, postnatal day 0; P32, postnatal day 32; P3M, postnatal 3 months. [ABSTRACT FROM AUTHOR]