06. The Membrane and the Machine

Picture warm shallow water, or the plume of a hydrothermal vent, somewhere around four billion years ago. Fatty acid molecules, long carbon chains with a head that loves water and a tail that flees it, are crowding together in solution, and the moment enough of them gather, they curl spontaneously into hollow spheres. No gene directs this. No instruction exists yet to direct it. It is simply the shape this chemistry prefers once concentration and conditions allow it, and the hollow sphere it produces, called a vesicle, is the first rough draft of an inside separated from an outside.

Roy Black and Matthew Blosser's 2016 synthesis of the field, building on decades of work by pioneers such as Deamer, Luisi, Szostak and Rasmussen, describes these vesicles as more than passive bubbles. Built from single-chain fatty acids roughly ten to fifteen carbons long, they assemble across a wide range of conditions, a pH anywhere from about 6.5 to above 12, in water salted the way early oceans are thought to have been, with sodium, magnesium and calcium all present, and they form most readily around 70 degrees Celsius, scalding to the touch but unremarkable next to a hydrothermal source. Once formed, these vesicles actively bind and concentrate the raw building blocks of biology inside themselves: the individual letters that make up RNA, simple sugars, and, with somewhat less certainty, amino acids.

That concentrating effect matters more than it sounds. A self-copying molecule adrift in open water competes with nothing and protects nothing it makes; useful neighboring molecules simply drift away into the vastness of the ocean. Wrap the same chemistry inside a fatty membrane, and there is suddenly a boundary: an inside that can hoard its own building blocks, an outside that cannot immediately dilute them away. That boundary does real work for whatever chemistry it encloses, long before a single gene exists to build it on purpose. Every cell alive today still wears some descendant of that same boundary, a lipid bilayer: two layers of fatty molecules, tails pointing inward away from water, heads pointing outward into it. Laboratory versions of these same simple vesicles today can still encapsulate and protect short strands of RNA, reproducing in a dish the proposed first step from four billion years ago.

How much a membrane like this could accomplish entirely unaided, before any genetically encoded help arrived in the form of membrane proteins the cell built for itself, is still argued over, and so is how strongly amino acids specifically stick to such an early membrane compared to how well nucleobases and sugars do. The first step's chemistry is settled and uncontested. How many further steps it could take alone is not.

A second puzzle concerns which kind of membrane came first. The two great domains of single-celled life today build their membranes from chemically mirror-image parts. Archaea use branched hydrocarbon chains, called isoprenoids, linked by a tough ether bond to one mirror-image form of a glycerol backbone. Bacteria, and Eukarya, which inherited their membrane chemistry from bacteria-like ancestors, use straight fatty acid chains linked by an easier-to-break ester bond to the opposite mirror-image form of that same backbone. This split was long assumed to have opened up only after LUCA, since no living organism was ever found straddling both chemistries at once.

In 2021, Laura Villanueva, Bas Dutilh and colleagues narrowed that assumption without closing it. They found a complete, apparently functional archaeal-style pathway for building ether lipids hiding inside the genomes of ordinary bacteria belonging to the FCB superphylum, a large bacterial grouping named for three of its member lineages, Fibrobacteres, Chlorobi and Bacteroidetes, with the relevant genes switched on in wild environmental samples, not merely present and silent. Moving those bacterial genes into laboratory E. coli produced cells running a genuinely mixed membrane, archaeal chemistry and bacterial chemistry side by side in the same cell wall. That does not prove LUCA itself had such a mixed, double-handed membrane; no living organism confirms that directly. It does show the two chemistries are far less walled off from each other than the textbook divide implied.

The machine that never changed

Whatever LUCA's own membrane looked like, something running inside it was already reading the frozen, sixty-four-word code from two legs back and using it to build proteins. That something is the ribosome: a large molecular machine, built mostly out of RNA with a supporting cast of proteins, that takes a strand of messenger RNA and, three letters at a time, strings together the matching chain of amino acids.

The proof that this machine is itself a working fossil of the RNA world arrived as a single, decisive structure. In 2000, Poul Nissen, Jeffrey Hansen, Nenad Ban, Peter Moore and Thomas Steitz solved the atomic structure of the ribosome's large subunit, resolved down to 2.4 angstroms, a unit small enough to distinguish individual atoms, using material from Haloarcula marismortui, a salt-loving archaeon. Their results appeared as two companion papers in the same issue of Science: one from Ban's team describing the subunit's overall architecture, the other from Nissen's team zeroing in on the active site itself. At the exact site where two amino acids are stitched together, called the peptidyl transferase center, no protein sits close enough to touch the reaction; the active site is built entirely from two RNA molecules, the 23S and 5S ribosomal RNAs. Thomas Cech, reviewing the result in a commentary in the same issue, summed it up in five words: "the ribosome is a ribozyme," a piece of RNA doing a job usually reserved for a protein enzyme.

The implication runs deep. RNA had to be capable of stitching amino acids together before proteins existed to do that job any other way. The ribosome's catalytic heart is not a metaphor for the RNA world; it is a still-working piece of it. That structural fact has not been seriously disputed since 2000. What remains under active study is exactly how such a ribozyme first assembled, and how it later acquired its surrounding shell of protein support.

That shell has its own history, and the ribosome's shape seems to remember building it. In 2009, Konstantin Bokov and Sergey Steinberg mapped the internal folding pattern of the large ribosomal RNA subunit and found it organized in concentric layers, an ancient core wrapped in shell after shell added later, rather than as an arbitrary tangle of RNA. They concluded that the ribosome's origin "happened nearly four billion years ago," a figure reasoned out from structure rather than read off any rock. A later reanalysis, published in the Journal of Molecular Evolution under the title "Ancestral Insertions and Expansions of rRNA do not Support an Origin of the Ribosome in Its Peptidyl Transferase Center," pushed back on one detail, questioning whether that ancient core lines up exactly with the peptide-bond-forming site, without overturning the layered picture itself.

A separate 2015 study by Anton Petrov and colleagues built out a fuller six-phase growth sequence for the whole translation system: an ancestral RNA core first learned to fold correctly, then to catalyze a chemical reaction, then to pair up with a second subunit, then to co-evolve with its emerging protein partners, then to run on an energy-driven decoding cycle, and only last did today's proteinized outer surface take shape. Every ribosome alive still carries this same layered core; eukaryotic ribosomes, the kind found in cells with a nucleus, added two further and more recent layers of their own, long tentacle-like RNA extensions found nowhere else in the tree of life. Several such accretion models compete over the exact order of the layers, without disagreement on the broad direction: a small, ancient RNA core, growing outward into today's hybrid machine.

The reading step itself works through a division of labor between two kinds of molecule, as Miguel Angel Rubio Gomez and Michael Ibba's comprehensive 2020 review lays out. A family of enzymes called aminoacyl-tRNA synthetases, molecular matchmakers, one enzyme family shared across all of life, attach each of the twenty amino acids to its own correct transfer RNA, the adaptor molecule carrying the matching three-letter anticodon, checking and double-checking the pairing before release. Once that charging step is done correctly, the ribosome itself never rechecks the amino acid's chemical identity; it simply trusts the three-letter pairing between a codon on the messenger strand and an anticodon on the incoming transfer RNA, and forms the bond. Oddly, the matchmaker enzymes come in two structurally unrelated families, Class I and Class II, ten enzymes each, that independently evolved to perform the identical job, a split whose reason is still an open question, one that reaches back toward the same pre-LUCA code-building era from two legs back.

Put the two halves of the reading system side by side, RNA-only catalysis at the ribosome's core, protein-based precision at the charging step, and the whole machine reads as a fossil record in its own right: an RNA-world veteran wearing tools built by the protein world it helped bring into being. That same machine, structurally recognizable down to its oldest core, runs unmodified in every domain of life today. Two independent inventions of a machine this specific, this old and this exact would be a coincidence nobody studying it has ever needed to invoke, which makes the ribosome, if anything, an even stronger argument for one shared ancestor than the frozen codon table alone.

A membrane to hold a self apart from the world. A frozen code to write proteins by. A machine old enough to remember reading the first of them. By roughly four billion years ago, give or take the usual margins, the essential parts were all in place. What was not yet in place was scale, something to make energy abundant enough to build a far larger cell out of these same old parts. That arrives next, when one cell does something stranger than inventing any new molecule: it swallows another cell whole, and keeps it.

Overturned if: a ribosome's peptidyl transferase center were found built from protein rather than RNA in any lineage, breaking the claim that peptide-bond catalysis is a universal, inherited relic of the RNA world.

Sources

• 2016 · Black, R.A. & Blosser, M.C., "A Self-Assembled Aggregate Composed of a Fatty Acid Membrane and the Building Blocks of Biological Polymers Provides a First Step in the Emergence of Protocells" — Life 6:33
• 2021 · Villanueva, L. et al., "Bridging the membrane lipid divide: bacteria of the FCB group superphylum have the potential to synthesize archaeal ether lipids" — The ISME Journal 15:168–182
• 2000 · Nissen, P., Hansen, J., Ban, N., Moore, P.B., Steitz, T.A., "The structural basis of ribosome activity in peptide bond synthesis" — Science 289:920–930, PMID 10937990
• 2009 · Bokov, K. & Steinberg, S.V., "A hierarchical model for evolution of 23S ribosomal RNA" — Nature 457:977–980
• 2015 · Petrov, A.S. et al., "History of the Ribosome and the Origin of Translation" — PNAS 112:15396–15401, PMID 26621738
• 2020 · Rubio Gomez, M.A. & Ibba, M., "Aminoacyl-tRNA synthetases" — RNA 26:910–936, PMC7373986