We are creating a unified UKRI website that brings together the existing research council, Innovate UK and Research England websites.
If you would like to be involved in its development let us know.

Site search
Back to blog

Behind the picture: An early model of the ribosome

by Guest Author on 12 Mar 2014

The MRC National Institute for Medical Research (NIMR) turns 100 this year and its rich history is fertile ground for looking back on past discoveries. This 1960s molecular model represents an important point in the evolution of thinking about the structure of the ribosome. Katherine Nightingale spoke to Dr Bob Cox about constructing the model ― and getting a few giants of biomedical research to sign it too. 

Bob Cox's ribosome model, with Crick's signature second from right on the top row

Bob Cox’s ribosome model, with Crick’s signature second from right on the top row

Look closely at one of these polystyrene balls and you’ll find the autograph of one of science’s most celebrated sons, Francis Crick. As well as being part of the duo that discovered the structure of DNA, Crick also proposed the “central dogma” of molecular biology: that DNA makes RNA makes protein. Fitting then, that his signature is here on an early model of the ribosome, the molecular machine that makes proteins.

Ribosomes are cellular factories made of RNA and protein which ‘translate’ the genetic code into the corresponding amino acid code, specific to each protein. They are large and complex molecules, made up of around 50 proteins divided into two subunits. They were discovered in 1955, though they didn’t get their name until 1958.

This model, produced by NIMR researcher Bob Cox in 1969, was the first attempt to model ribosome structure in detail. Until then, only blurry microscope pictures had been available.

Researchers knew that ribosomes were made of proteins and RNA. The only other similar structures known at the time were viruses, which are DNA or RNA packaged inside an envelope of uniform proteins. But Bob knew that ribosomes were made up of lots of different types of proteins. “The question was how could all these proteins fit together with RNA to make a molecule that made some kind of functional sense?”

In Bob’s model the main structure is RNA ― represented by twisting copper wire ― bound to a range of proteins (the white balls). The small tags attached to the wire stand in for the RNA’s nucleotides. The distinctive folds in the wire form specific binding sites for proteins, and the arrangement means that the RNA is accessible to play its role in protein synthesis.

“Because each ball represents a different protein, I had a bit of fun and asked people to sign them,” says Bob. “They would laugh and say ‘one day it’ll be worth a lot of money.’” Other signatures on the model include giants of biomedical research such as Peter Medawar, John Kendrew, Arthur Kornberg and Marianne Grunberg-Manago.

The model was made in the NIMR’s workshop by Frank Doré. “He deserves considerable credit for putting the model together,” says Bob. The larger subunit has around 3,000 tags and the smaller subunit 1,500. Bob himself made an earlier version from dowling and wire.

The detailed structure of the ribosome wasn’t solved until the early 2000s, for which MRC scientist Venki Ramakrishnan won the Nobel Prize for Chemistry in 2009. Like Bob’s model, it shows two irregular shaped subunits, each with distinctive protuberances. It also shows interactions between distant segments of RNA, as well as the short range interactions shown in the model.

But for Bob, his model represents a good first stab at a structure, just a decade after the ribosome was first named.

“I would talk about [the structure] at meetings and take the model with me and no one laughed at it, so there was a feeling that though it might not be right it was a good start.”

Katherine Nightingale


The model is wrong in many fundamental ways and really is worse than the “blurry” em pictures that are dismissed here. Those blurry EM pictures turned out to be remarkably correct, although of limited detail.

This shows the dangers of trying to model something complex without sufficient data. DNA was vastly simpler by comparison, yet even that required data to constrain the model.

Venki Ramakrishnan

author avatar by Venki Ramakrishnan on 12-Aug-2014 11:19

Replying to Venki Ramakrishnan

I am posting this response on behalf of Robert Cox.

Katherine Nightingale
MRC Digital Content Editor

Dr. Ramakrishnan has disparaged work done when he was a teenager without having taken the trouble to read the relevant publications. Moreover, he appears to overlook the fact that ribosomes range widely in size and shape; that is, E.coli ribosomes are not representative of ribosomes in general. Ribosomes were identified as a distinct class of ribonucleoproteins in 1958 (Roberts, 1958). In 1960 I joined H. R. V. Arnstein to study the biosynthesis of rabbit haemoglobin. My first task was to develop a reliable method for the isolation of RNA (Cox and Arnstein, 1963; Cox 1968a). I chose to use the chaotrope 4 M guanidinium chloride for this purpose because RNAase is inactive in this reagent. (the stronger chaotrope 4M guanidinium isothiocyanate is currently the method of choice). Thus we were able to isolate undegraded RNA species from rabbit polyribosomes. This facility enabled us to study the properties of the RNA components of rabbit ribosomes both in situ and after isolation. Electron micrographs of rabbit ribosomes were obtained by my colleague Robin Valentine (Cox 1969).

Other ribonucleoproteins were known at the time. There were the plant viruses such as Tobacco Mosaic Virus and Turnip Yellow Mosaic Virus. Their structures were established by X-ray analysis (Caspar and Klug, 1962). In each case the RNA compound was found to lie within a protective coat of identical protein subunits Properties of viral RNA species were reviewed (see Cox, 1968b). I showed (Cox, 1969) that rabbit ribosomes could not have a similar structure. For this reason I was led to consider the question of a possible structure of rabbit ribosomes. I summarized the available data for rabbit ribosomes by constructing a model representing the RNA and protein components that was consistent with the e/m data of Robin Valentine (Cox and Bonanou, 1969). The model was well received. The view of my colleagues was that the model was a useful guide for students and non-specialists to visualize a complicated structure such as the ribosome and its role in protein synthesis (Cox 1970a; 1970b). By request, the model was displayed at the 500th Meeting of the Biochemical Society (December 1969) and five years later at the Royal Institution.

Colleagues were invited to sign a white sphere representing a protein subunit to show that the proteins were almost all different. The signatures included Francis Crick, Sir John Kendrew and Sir Peter Medawar.

Robert (Bob) Cox


Caspar, D.L.D. & Klug, A (1962). Cold Spring Harbour Symposium. Quant Biol. 27, 1 –

Cox, R.A. & Arnstein, H.R.V. (1963). The Isolation, Characterisation and Acid-Base Properties of Ribonucleic Acid from Rabbit-Reticulocyte Ribosomes. Biochem. J. 89, 574-585

Cox, R.A. (1968a). The use of guanidinium chloride in the isolation of nucleic acids. Methods in Enzymology, 12B, 120-129

Cox, R.A. (1968b). Macromolecular Structure of Ribonucleic Acids. Quarterly Reviews of the Chemical Society. 22, 499-526

Cox, R.A. (1969). The effect of pancreatic nuclease on rabbit reticulocyte ribosomes and its interpretation in terms of ribosome structure. Biochem. J. 114, 753-767

Cox, R.A. and Bonanou, S. (1969). A possible structure of the rabbit reticulocyte ribosome. An exercise in model building. Biochem. J., 114, 769-774

Cox, R.A., (1970a). The ribosome-decoder of genetic information. Science Journal, 6, 56-60

Cox, R.A. (1970b). Ribosome Structure and protein biosynthesis. Image Roche, International Edition No. 39, 2-7.

Roberts, R.B. (1958). In Microsomal Particles and protein synthesis. Ed. by R. B. Roberts. Pergamon Press, London. P viii.

author avatar by Katherine Nightingale on 13-Oct-2014 12:25

Leave a reply

You may use basic HTML in your comments. Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.


From category

Share this: