Minimal genomes and synthetic cells

Two programmes converge on the same question from opposite directions. The top-down approach removes genes from a living organism until nothing further can go without killing it. The bottom-up approach assembles a cell-like system from purified components and asks what must be added before it is alive. Neither has finished, and the gap between them is the honest measure of how well the field understands what a cell is.

What the reduced genome established

The reference result is JCVI-syn3.0, derived from Mycoplasma mycoides — an organism chosen because it already has a very small genome and no cell wall. Its genome was chemically synthesised, assembled in yeast and transplanted into a recipient cell whose own genome was removed, so the synthetic chromosome alone directed the resulting organism. Reduction proceeded by rounds of design, build and test, guided by transposon mutagenesis to identify which genes tolerated disruption. The result retains 473 genes and is a self-replicating cell.

The finding that matters is not the number. Around a fifth of the retained genes had no assigned function — they could not be deleted without loss of viability, and nobody could say what they do. That is a direct statement about the limits of annotation: sequence similarity, the workhorse of functional inference, fails for a substantial fraction of the genes a bacterium cannot live without. Subsequent work has assigned functions to some of them, and the general point stands.

The programme also clarified essentiality itself. A quasi-essential gene is one whose loss is not lethal but severely slows growth; removing many of them individually is possible while removing them together is not. Essentiality is therefore relative to a medium, a temperature and a tolerated growth rate — a minimal genome is minimal for a specified environment, and a richer medium permits a smaller one. Syn3.0 also grew slowly and divided into irregular shapes until a small set of genes was restored, which showed that normal division depends on genes that a viability screen calls dispensable.

Bottom-up, and what is still missing

The synthetic-cell effort encapsulates transcription and translation machinery in lipid vesicles, typically using reconstituted systems of purified components. Such vesicles express proteins, and constructed systems have driven membrane growth and division. What has not been demonstrated is the closed loop: a vesicle that synthesises its own machinery, including its own ribosomes, and divides indefinitely. Ribosome self-assembly outside a cell remains the acknowledged bottleneck.

Why anyone wants this

A reduced chassis is the response to the burden and unpredictability described on the synthetic biology page: fewer genes means fewer unmodelled interactions and less metabolism competing with the intended pathway. The counterweight is robustness — a stripped organism has no reserve for stress, which is a containment argument as much as a manufacturing limitation.

This page does not cover the genome-writing chemistry itself, nor the editing tools; those are treated separately.