%MinMax: A versatile tool for calculating and comparing synonymous codon usage and its impact on protein folding

Abstract

Most amino acids can be encoded by more than one synonymous codon, but these are rarely used with equal frequency. In many coding sequences the usage patterns of rare versus common synonymous codons is nonrandom and under selection. Moreover, synonymous substitutions that alter these patterns can have a substantial impact on the folding efficiency of the encoded protein. This has ignited broad interest in exploring synonymous codon usage patterns. For many protein chemists, biophysicists and structural biologists, the primary motivation for codon analysis is identifying and preserving usage patterns most likely to impact high-yield production of functional proteins. Here we describe the core functions and new features of %MinMax, a codon usage calculator freely available as a web-based portal and downloadable script (http://www.codons.org). %MinMax evaluates the relative usage frequencies of the synonymous codons used to encode a protein sequence of interest and compares these results to a rigorous null model. Crucially, for analyzing codon usage in common host organisms %MinMax requires only the coding sequence as input; with a user-input codon frequency table, %MinMax can be used to evaluate synonymous codon usage patterns for any coding sequence from any fully sequenced genome. %MinMax makes no assumptions regarding the impact of transfer ribonucleic acid concentrations or other molecular-level interactions on translation rates, yet its output is sufficient to predict the effects of synonymous codon substitutions on cotranslational folding mechanisms. A simple calculation included within %MinMax can be used to harmonize codon usage frequencies for heterologous gene expression.

Keywords: biotechnology; codon harmonization; codon optimization; cotranslational protein folding; heterologous gene expression; protein aggregation; ribosome.

Figure 1

Synonymous codon substitutions can alter both total yield (gene expression) and folding yield. In this example, synonymous rare codons (red) reduce translation rate and enable the nascent protein to achieve a conformation that is kinetically inaccessible when the same sequence is translated faster using more common codons (green). Faster translation leads to more rapid appearance of the nascent protein C‐terminus, which interferes with the folding of the N‐terminus, leading to misfolding and aggregation. Hence while the total yield of protein is higher with common codons, the folding yield is lower. Note, however, that for some native structure topologies, faster translation at specific positions could instead enhance folding yield (see text).

Figure 2

Relative codon usage of the human muscle protein α‐actinin, calculated using %MinMax. (A) When translated using H. sapiens codon usage frequencies (black), codon usage across nearly the entire coding sequence is more common than average. In contrast, translation using Escherichia coli codon frequencies (brown) significantly alters the pattern of codon usage. This pattern can largely be restored by synonymous codon changes selected via a simple harmonization procedure included within %MinMax algorithm (green; see text). The average of 200 RRTs is shown as a grey line; the grey shading indicates %MinMax values within two standard deviations from this average. %MinMax values were averaged over a sliding window of 17 codons. For clarity, only the first 250 aa of α‐actinin are shown. (B) Ribbon diagram representing the first 250 aa from the crystal structure of full‐length α‐actinin (PDB:4D1E).32

Figure 3

Impact of calculating relative versus absolute codon usage. (A) Cys is a rare amino acid; hence in an absolute sense both Cys codons UGU and UGC are rare in all organisms. (B) In contrast, relative usage of UGU and UGC vary dramatically between organisms.