Abstract

Synonymous codon replacement can change protein structure and function, indicating that protein structure depends on DNA sequence. During heterologous protein expression, low expression or formation of insoluble aggregates may be attributable to differences in synonymous codon usage between expression and natural hosts. This discordance may be particularly important during translation of the domain boundaries (link/end segments) that separate elements of higher ordered structure. Within such regions, ribosomal progression slows as the ribosome encounters clusters of infrequently used codons that preferentially encode a subset of amino acids. To replicate the modulation of such localized translation rates during heterologous expression, we used known relationships between codon usage frequencies and secondary protein structure to develop an algorithm (“codon harmonization”) for identifying regions of slowly translated mRNA that are putatively associated with link/end segments. It then recommends synonymous replacement codons having usage frequencies in the heterologous expression host that are less than or equal to the usage frequencies of native codons in the native expression host. For protein regions other than these putative link/end segments, it recommends synonymous substitutions with codons having usage frequencies matched as nearly as possible to the native expression system. Previous application of this algorithm facilitated E. coli expression, manufacture and testing of two Plasmodium falciparum vaccine candidates. Here we describe the algorithm in detail and apply it to E. coli expression of three additional P. falciparum proteins. Expression of the “recoded” genes exceeded that of the native genes by 4- to 1,000-fold, representing levels suitable for vaccine manufacture. The proteins were soluble and reacted with a variety of functional conformation-specific mAbs suggesting that they were folded properly and had assumed native conformation. Codon harmonization may further provide a general strategy for improving the expression of soluble functional proteins during heterologous expression in hosts other than E. coli.

Highlights

  • Changes to protein structure and function that occur after synonymous codon replacement indicate that protein structure is DNA sequence dependent

  • I158 satisfies the criterion of being among these ten residues and fits the space parameter, its isoacceptor usage frequency in the native expression host is greater than 25%; this residue was not calculated to be part of the link/end segment

  • Following full codon harmonization of this segment for expression in E. coli, codon usage patterns for heterologous expression were in good agreement with those of the gene in its native host, e.g. with the usage rate for P153 increasing to 8% as compared to 11% in P. falciparum

Read more

Summary

Introduction

Changes to protein structure and function that occur after synonymous codon replacement indicate that protein structure is DNA sequence dependent. Synonymous codon substitutions that change codon usage frequencies from infrequent to frequent usage in regions of slow mRNA translation can deleteriously affect enzyme activity [1]. Contrary to conventional thinking, synonymous codon substitutions may not always be silent; changing codon usage frequency affects protein structure and function, and the frequency with which codons are used imparts vital information for the development of secondary and tertiary protein structure. Species-specific disparities in codon usage are frequently cited as the cause for failures in recombinant gene expression by heterologous expression hosts. Such failures include lack of expression, or the expression of protein that is non-functional or insoluble, or protein that is truncated owing to proteolysis or premature termination of translation [3,4,5]. Translation is not a steady state process, rather it occurs in pulses, as can be observed from ribosomal pausing [13] and even ribosome stacking, on specific stretches of mRNA [14]; these temporal changes in translational rate have been shown to depend on relative tRNA levels [15]

Methods
Results
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call