Poor man’s ribo-seq: circa 1968

Nowadays, every molecular biologist has heard about ribosome profiling and how useful it is to determine the distribution of ribosomes on mRNA molecules in cells.This technology, first reported in 2009, is widely employed and has proved incredibly powerful for genome-wide surveys of which mRNAs are being translated and can often define the exact reading frame that is being used [1].However, the concept of ribosome profiling goes back much further.Nearly 60 years ago, the Journal of Molecular Biology, at the time the leading journal for molecular science studies, published what one might consider to be the first ribosome profiling experiment [2].Of course, in 1968, no nucleotide sequences were available, and only very few protein sequences were known.In consequence, ribosome distribution on an mRNA molecule could not be determined by modern ribosome footprinting methods, where the positions of ribosomes are inferred by sequencing the ribosome-protected mRNA fragments.Back in the 1960s, an important model system for studying mRNA translation was the rabbit reticulocyte, an immature, anucleate red blood cell whose main function is to synthesize the and chains of globin and assemble them together with haem into the haemoglobin 2 2 (haem) 4 tetramer.In the final two to three days after the reticulocyte emerges from the bone marrow and before it matures into a red cell, greater than 90% of the proteins made by a reticulocyte are globin chains.By 1966, the sequences of the two rabbit globin chains had been established by direct protein sequencing [3,4].The availability of the and chain globin sequences permitted the prediction and identification of all of the (15) and (15) chain globin peptides that can be generated by trypsin digestion.A 1966 study in which human bone marrow-derived reticulocytes were pulse-labelled with 14 C-labelled lysine for 3 min showed that tryptic peptides mapping between position 90 and the C-terminus in both the and chains derived from completed globin chains exhibited a rapid increase in relative specific activity [5].This led the authors to postulate that a control point exists on the globin mRNAs where ribosome transit is paused or slowed in the vicinity of codon 90 to give time for the nascent globin chains to fold into a specific conformation that allows haem insertion (His87 is a key coordination group for haem); only when haem is coordinated with the nascent chain would translation speed up again [6].Such a mechanism could preclude wasteful synthesis of globin chains when haem levels are low.In 1967, Tim Hunt and Alan Munro, my colleagues in the Department of Biochemistry at the University of Cambridge, and I set out to test this hypothesis more directly by determining the distribution of ribosomes on the and globin mRNAs in vivo.The plan was to isolate nascent globin chains from polysomes purified from lysates of reticulocytes that had been pulselabelled with a radioactive amino acid, and then measure the relative levels of each of the globin tryptic peptides derived from the nascent and chains.As first demonstrated by Dintzis for globin chains [7], proteins are translated from N-to C-terminus.This means that each nascent and globin chain longer than approximately 10 residues (the N-terminal peptides of the and chains are 7 and 8 residues, respectively) will contain their N-terminal tryptic peptides, increasingly longer nascent chains will contain additional peptides from the middle of the and chains, and only the longest globin chains, just prior to their release from the terminating ribosome, will contain the most C-terminal peptides.If ribosomes were uniformly distributed on