Cell-type specific codon usage and differentiation.

Abstract

This paper presents evidence derived from the selective use of codons in ca. 40 eukaryotic genes (or messages derived from them) that codon usage is one of the most conserved features of messages for specific cell products. The theory has been developed by several investigations that the kinds of products of many if not most kinds of differentiated cells is determined by the pattern of translation abilities each cell possesses as it differentiates. A correllary of this thesis is that the groups of code words used for products of specific cell types, INDEPENDENTLY OF THE SPECIES INVOLVED, should exclude specific kinds of code words in one cell type and not another. To test this thesis, the specific frequency of codon usage and non-usage has been collated from the recently published literature and subjected to appropriate computer analysis. We find that: (1) certain codons are not used at all in any message for globins; (2) the pattern of codon usage is characteristic of specific products from specific embryonic derivatives, e.g. erythrocytes; (3) that certain code words are discriminated against generally in nearly all vertebrate cell messages evaluated; and (4) that the cell type from which a message is derived can be identified, at least in the case of 9 globin messages derived from species separated by millions of generations, purely on the basis of codon usage. From these studies it can be inferred that some evolutionary factor prevents the use of "forbidden" code words in specific kinds of cells. We propose that this factor derives from the fact that a "silent" mutation to a code word which is untranslatable by a differentiated cell will be lethal in the homozygous condition and that the thesis that so-called "codon restriction" is an important determinative factor in limiting what differentiating cells can synthesize in many kinds of developing cells explains the available evidence more adequately than alternative theories.