How cell families keep their identity through evolution

Recognising a neuron and identifying its precise cell type are different tasks. A September study asks how broad cellular identities survive evolutionary diversification.[1]

Researchers compared gene activity and accessible DNA in about 90,000 nuclei from three flatworm species. Neural networks learnt to predict accessibility from DNA sequence, revealing recurrent patterns called motifs.[1]

DNA has a physical arrangement

The genome is packaged. DNA wraps around histone proteins, and these units fold into chromatin. That arrangement allows a long molecule to fit inside a nucleus while retaining regions that can open for copying or transcription.[3]

A promoter illustrates why access matters. Proteins bind this stretch of DNA to begin making an RNA copy of a gene. The promoter can also work with other regulatory regions called enhancers. Understanding gene control therefore involves the sequences where these proteins act as well as the genes themselves.[4]

ATAC-seq measures an aspect of that physical availability. An enzyme inserts sequencing adapters preferentially into unprotected DNA. Reading the tagged fragments reveals where the enzyme could reach. Their lengths add another clue: fragments shorter than DNA normally protected by a nucleosome help distinguish exposed stretches from the DNA packaged around those protein spools.[2]

The method’s developers describe accessibility, protein binding and nucleosome positions as related layers of information. A DNA sequence supplies the letters; an accessibility experiment supplies evidence about their physical setting in the sampled cells. Combining those views gives a sequence model something biologically informative to predict.[2]

Shared patterns, changing combinations

Models recognised family-level accessibility across species but lost specificity for individual neural cell types. Mouse and zebrafish embryo analyses showed a similar pattern. Shared motifs persisted more consistently than their combinations.[1]

The authors propose “collective maintenance”: many shared motifs sustain family identity while their combinations evolve. This is a model-supported explanation of regulatory flexibility.[1]

Report an error or suggest a correction