Referee 1's Review:

**********
N.B. These are the comments made by the referee when reviewing an earlier version of this paper. Prior to publication, the manuscript has been revised in light of these comments and to address other editorial requirements.
*********

The authors show that it is important to maintain translational 'pauses' when expressing heterologous proteins in E. coli. They replicate these slowly translated regions of mRNA by identifying these regions of slow translation and replacing with synonymous codons that have the same usage frequencies in the heterologous expression host.

The expression of four P. falciparum proteins is discussed in this current work. Two of these proteins (MSP1.42 FMP003 and LSA-NRC H) have already been discussed in other publications (references 26 and 39). The algorithm described in this work was applied in these two other papers. It is not clear what new concepts/principles are brought forth in this work, it appears to be a more detailed description of a method applied in references 26 and 39.

Other comments:

1. In 'Results', section 'Single codon application of codon harmonization', the authors discuss the analyses of the I141 conversion mutant. They also mention the efforts to analyze a Y151 mutant, but they were not able to obtain that construct. Unless there is a good reason for not obtaining this mutation (which is not mentioned), this part can be left out of the manuscript.

2. The authors mention that expression levels of the FMP010 construct exceed wild type level expression by at least 100 fold (based on gel densitometry). How much was recovered after purification in terms of mg/g wet cells?

3. How much more protein was expressed of the codon harmonized LSA-NRC construct? Only a gel is shown (Fig. 6), but no quantitation given.

4. It would be helpful if the authors could present a table that shows the various constructs, where the mutations lie, and how much they improved the expression levels.