Richard Dawkins told us genes are selfish and bodies are vehicles. Mart Krupovic and Eugene Koonin, writing in a February 2026 preprint, say: look one level deeper. The ribosome is the entity that all genes became addicted to.
The argument is built on numbers. In a growing E. coli cell, translation consumes at least 50% of the total energy budget. Ribosomal RNA constitutes 95–98% of all cellular RNA. A single bacterium contains 20,000–27,000 ribosomes; a mammalian cell, up to 10 million. The ribosome isn't the most important machine in the cell because cells decided to invest in it. It's the most important machine because everything else evolved to sustain it.
The Bowman quote anchors the inversion: “The function of a ribosome is to build more ribosomes.” Cellular growth rate scales linearly with ribosome concentration. The production time of each ribosome — about 7 minutes in bacteria — sets a hard physical limit on how fast a cell can divide. The cell doesn't constrain the ribosome. The ribosome constrains the cell.
Krupovic and Koonin trace this dominance to the origin of life. In the RNA world, the proto-ribosome (the peptidyltransferase center, a ribozyme that catalyzes peptide bond formation) was likely a mutualistic symbiont of RNA replicases. It produced peptides that enhanced replication. A useful partner. But as the translation system complexified — evolving tRNAs, aminoacyl-tRNA synthetases, the genetic code — replicators switched from ribozyme-based to protein-based replication. That switch was irreversible. Once your replication machinery is made of proteins, you cannot replicate without translation. The mutualist became a dependency. The dependency became an addiction.
The paper calls this an “addiction module,” and the term is precise. In molecular biology, addiction modules are toxin-antitoxin systems where the antitoxin must be continuously produced or the toxin kills the cell. The ribosome functions analogously: it must continuously translate or everything stops. No other cellular system has this property. Metabolism can reroute. Membranes can reseal. DNA repair can improvise. But translation is the single chokepoint between genotype and phenotype. Without it, genes are inert sequence.
The virus test sharpens the argument. No virus in the history of life has evolved its own ribosome. Giant viruses encode tRNAs, translation factors, even ribosomal proteins — partial translation components that let them tweak the host machinery. But autonomous translation from scratch? Never. The energetic and temporal costs are prohibitive: producing enough ribosomes to translate a genome takes longer than the infection cycle allows. So every virus, without exception, hijacks the host's ribosomes. Patrick Forterre proposed the distinction: cells are “ribosome-encoding organisms” and viruses are “capsid-encoding organisms.” The ribosome is the line between cellular life and everything that parasitizes it.
Here's what makes the inversion work structurally, not just rhetorically. Typically, we think of the ribosome serving the cell — the cell needs proteins, so it builds ribosomes. Krupovic and Koonin point out that the entire metabolic network can be reframed as existing to feed the ribosome. The cell synthesizes amino acids so the ribosome has substrates. The cell generates ATP so the ribosome has energy. The cell replicates DNA so the ribosome has templates to transcribe from. Every pathway traces back to sustaining translation. The paper notes that the “ribosome catastrophe” — the point where a cell would need more ribosomes than can physically fit inside it — sets an absolute upper limit on cell size. The ribosome doesn't just consume resources. It defines the geometry.
The multilevel selection framework resolves the apparent contradiction. The ribosome is both selfish (consuming the bulk of cellular resources) and mutualistic (providing all the proteins everything else needs). This is exactly the structure of evolutionary transitions: lower-level selfishness becomes nested within higher-level cooperation. Genes were selfish replicators that became organized into chromosomes. Chromosomes became organized into cells. Cells became organized into organisms. At each transition, the lower-level selfishness didn't disappear — it was constrained. The ribosome's selfishness was the original constraint that made all subsequent levels possible. It didn't build the house. But it owns it.
And it can't be evicted. Four billion years of evolution, and the ribosome's core structure — the PTC, the catalytic heart — is essentially unchanged. The accretion model shows that the ribosome grew by adding layers, like geological strata, with the ancient core preserved beneath. Those 95 universally conserved genes encoding translation components are the most stable feature of all cellular life. DNA replication uses three non-homologous polymerase families across bacteria, archaea, and eukaryotes — the replication machinery has been replaced multiple times. But the ribosome persists. The landlord stays. The tenants rotate.
One detail from the paper that I keep returning to: rRNA operons have been found on plasmids — mobile genetic elements that replicate independently of chromosomes. Bacteria maintaining their ribosomal genes on plasmids for hundreds of millions of years. The ribosome's information carriers adopting the same strategy as selfish genetic elements, spreading horizontally. The ribosome doesn't just dominate from within the chromosome. When the opportunity arises, its genes go mobile.
The conventional view sees the cell as the organism and the ribosome as its production facility. Krupovic and Koonin's view sees the ribosome as the organism and the cell as its survival strategy. They're careful to note that the distinction is mutualistic — “it is impossible to determine objectively whether the cell serves the ribosome or vice versa.” But the history of energy allocation, evolutionary stability, and viral dependency all point the same direction. The ribosome was there first, became indispensable first, and consumed more resources than any other component from the beginning.
Dawkins told us to stop looking at the organism and start looking at the gene. Krupovic and Koonin are saying: stop looking at the gene and start looking at what the gene cannot function without. The landlord doesn't need to own the building outright. It just needs to be the one thing no tenant can replace.
Krupovic, M., and E.V. Koonin. “The selfish ribosome.” arXiv:2602.23268 (February 2026).