A fertilized egg contains two genomes — one from each parent — that haven't started producing anything yet. For decades, the assumption was simple: the DNA sits in a relatively disorganized state until the embryo “switches on” its genes at zygotic genome activation (ZGA). Organization was downstream of activation. First you decide to read; then you arrange the books.
Researchers using Pico-C, a new chromatin capture technique sensitive enough to work in single cells (Nature Genetics, 2026), found the opposite. Before ZGA — before any gene is read — the DNA is already folded into complex loops and domains. Regulatory elements are positioned near their target genes. Enhancers are already in reach of the promoters they'll eventually activate. The 3D architecture of the genome is assembled in advance, like a stage set before the curtain rises.
This inverts the assumed direction. The standard model treated gene activation as the event and chromatin organization as its consequence — you start reading, and the reading creates structure. The data shows structure first, reading second. But this isn't merely “readiness” (having the right tools before the job). The spatial arrangement is the regulatory logic. Which enhancer reaches which promoter depends on how the DNA is folded. The loops aren't containers for regulation — they're the regulation itself, expressed in geometry rather than sequence.
A gene's sequence says what protein to build. Its position in 3D space — which loops it sits in, which regulatory elements are within interaction distance — says when and where to build it. The fertilized egg's genome arrives with both layers of information already encoded: the linear sequence (inherited from parents) and the spatial architecture (assembled pre-activation). The reading order is written in folds, not in letters.
The deeper implication is that the genome has two information channels operating on different substrates. The sequence channel is digital — A, T, G, C — and well understood. The spatial channel is geometric — loop sizes, domain boundaries, enhancer-promoter distances — and was invisible until techniques like Pico-C could resolve it in single cells. The previous assumption that early embryonic chromatin was disorganized wasn't based on evidence of disorder. It was based on the absence of tools to detect order. The structure was always there. The resolution wasn't.
What the embryo inherits, then, is not a set of instructions waiting to be organized but a set of instructions already organized into the order they'll be executed. The execution plan is physical. The genome doesn't compute its own reading order at runtime — it arrives pre-compiled.