Every time a cell divides, it must distribute its chromosomes equally between two daughter cells. The structure that makes this possible is the centromere — the region of each chromosome where the spindle apparatus attaches. Get the centromere wrong and chromosomes missegregate. Cells die, or worse, they don't die and become cancerous. The centromere is among the most critical structures in the genome.
In most organisms, centromeres are large, repetitive, and complex — stretching across hundreds of thousands of base pairs of tandem repeat sequences. Brewer's yeast (Saccharomyces cerevisiae) is the dramatic exception. Its centromeres are “point” centromeres: tiny, precisely defined DNA sequences, among the smallest and most streamlined in any eukaryote. Yeast centromeres were the first to be isolated and sequenced, in the early 1980s by Clarke and Carbon. For four decades, no one could explain how such minimal structures evolved from the elaborate centromeres found in related organisms.
Haase, Boeke, and Musacchio (Nature, 2026) found the answer in the genome's least reputable residents. By sequencing centromeres across the Saccharomycodales — a broader order of yeasts related to brewer's yeast — they identified intermediate forms they call “proto-point” centromeres. These intermediate structures bridge the gap between the large repeat-rich centromeres of more distant relatives and the tiny point centromeres of S. cerevisiae. The DNA at these intermediate centromeres traces to LTR retrotransposons — “jumping genes” that copy and paste themselves throughout the genome, the prototypical selfish genetic elements.
Retrotransposons are genomic parasites. They replicate at the expense of the host genome, inserting copies of themselves into new locations, occasionally disrupting genes, generally contributing nothing. They are the textbook example of DNA that exists because it is good at existing, not because it benefits the organism. Genomes spend energy suppressing them.
And yet. The centromere — the structure the genome cannot survive without — was built from retrotransposon material. The parasitic DNA that the genome tries to silence provided the raw material that evolution reshaped into the genome's most essential anchor point. The intermediate species show the transition in progress: centromeres that still contain recognizable retrotransposon fragments, halfway between the ancestral repeat-rich form and the derived minimalist form.
The structural insight is not just that parasites can be co-opted. It is that the most indispensable structures can originate from the most dispensable material. The centromere did not evolve from something important. It evolved from something the genome was actively trying to get rid of. The selection pressure that shaped the modern point centromere was not “build something essential from essential parts” but “repurpose what's available, including the junk.”
This connects to the nuclear envelope itself. The viral eukaryogenesis hypothesis proposes that the cell nucleus — the defining structure of all complex life — may have originated as a viral replication factory. A parasite built a compartment; the compartment outlasted the parasite and became the frame that made eukaryotic life possible. Now, within that nucleus, the anchor point for chromosome segregation turns out to derive from another class of parasite entirely.
The genome's most critical infrastructure, at two different scales, was built by its invaders.