Critical reflections on synthetic gene design for recombinant protein expression

Available online at www.sciencedirect.com

ScienceDirect Critical reflections on synthetic gene design for recombinant protein expression Annabel HA Parret1, Hu¨seyin Besir2 and Rob Meijers Gene synthesis enables the exploitation of the degeneracy of the genetic code to boost expression of recombinant protein targets for structural studies. This has created new opportunities to obtain structural information on proteins that are normally present in low abundance. Unfortunately, synthetic gene expression occasionally leads to insoluble or misfolded proteins. This could be remedied by recent insights in the effect of codon usage on translation initiation and elongation. In this review, we discuss the interplay between optimal gene and vector design to enhance expression in a particular host and highlight the benefits and potential pitfalls associated with protein expression from synthetic genes. Addresses 1 European Molecular Biology Laboratory, Hamburg Unit, Notkestrasse 85, 22607 Hamburg, Germany 2 Protein Expression and Purification Core Facility, European Molecular Biology Laboratory, Meyerhofstrasse 1, 69117 Heidelberg, Germany Corresponding author: Meijers, Rob ([email protected])

Current Opinion in Structural Biology 2016, 38:155–162 This review comes from a themed issue on New constructs and expression of proteins Edited by Rob Meijers and Anastassis Perrakis

http://dx.doi.org/10.1016/j.sbi.2016.07.004 0959-440/# 2016 Published by Elsevier Ltd.

Introduction The efficient overexpression of proteins is a prerequisite for successful structure determination. Structural genomics studies on a wide range of soluble protein targets over more than a decade have shown that more than 40% of the constructs failed due to limited expression (Figure 1). The high failure rate is a result of the use of standardized approaches, and can serve as a baseline for a structure determination project on a specific target. To alleviate these ominous statistics, a number of approaches have been adopted which address recombinant protein expression. The use of parallel cloning strategies employing different construct lengths and purification tags can increase success rate to a certain extent [1]. Alternatively, larger amounts of protein can be obtained by scaling up expression using large-scale fermentation [2]. However, large-scale fermentation often requires optimization to www.sciencedirect.com

1

address problems of plasmid loss, oxygen deprivation and pH fluctuation [3]. A more cost-effective approach is the optimization of the protein source, either by opting for a protein species of high natural abundance, or by using an efficient recombinant expression system. Traditionally, selecting protein homologues from species that can be purified from natural source has been rather successful [4–8]. Nevertheless, there is an increasing need to obtain the structure of a protein from a particular species, in order to assess species-specific functional aspects. Structures of human or murine proteins involved in the immune or nervous system can help to relate the effect of specific point mutants or posttranslational modifications. Similarly, structural analysis of virulent proteins from bacterial pathogens can reveal species-specific details important for drug design. In this review, we focus on the opportunities that gene synthesis provides to optimize recombinant protein expression of specific protein targets.

Application of synthetic genes for protein production The use of synthetic genes for the production of proteins for structural characterization is widespread. In general, gene synthesis may expedite the production of a series of constructs with engineered restriction sites. It is often preferred to the use of cDNA, which can contain several isoforms or splice variants. For target proteins identified from metagenomes or extremophiles, gene synthesis may provide the only route to protein expression, because cDNA is not available. Protein expression from codonoptimized genes has been especially successful for proteins involved in cell signalling whose native codon sequence can be strongly regulated, such as Irisin [9], Parkin [10] and Netrin [11]. Codon optimization has also uncovered hidden levels of gene regulation that are encoded in codon bias. Codon optimization of the FRQ protein involved in the regulation of circadian rhythms resulted in a loss of function, although expression levels were increased [12]. Codon optimization also revealed a hidden quality control mechanism encoded in the gene of the cystic fibrosis transmembrane conductance regulator (CFTR), an anion channel membrane protein [13]. Similarly, optimization of the CTP1L endolysin gene led to the discovery of a secondary translation site that produced a truncated protein. The truncated protein forms a 1:1 complex with the full length protein, and lysis activity is substantially reduced when the secondary translation site is removed by codon optimization Current Opinion in Structural Biology 2016, 38:155–162

156 New constructs and expression of proteins

Figure 1

2004

10 0%

35000

10 0%

40000

2016

60 %

25000 20000

51 %

15000

%

19 % 23 %

*

3%

4%

5000

3%

10000

17

No. of targets

30000

0 Cloned

Expressed

Purified

Experiment PDB deposit

Milestones Current Opinion in Structural Biology

A comparison of the total number of targets achieved per milestone along a structure genomics pipeline as reported by TargetDB in 2004 [104] and as of 15th of March 2016 (http://targetdb.rcsb.org/ metrics/). The recent data are compiled from soluble protein targets from three consortia (JCSG, NESG and NYSGRC). The attrition rate from clone to purified target has changed little over a decade and hovers around 75%, indicating that the early stages of a structure determination project are most crucial for successful completion. *The definition of the milestone ‘Experiment’ has changed from ‘crystals’ to ‘X-ray/NMR/EM studies performed for structural studies’. As a result, there is a large discrepancy between crystals obtained (4% in 2004) and experiments done on the sample (17%), irrespective of the outcome.

[14]. Toll-like receptors (TLRs) contain leucine-rich domains, and it was shown through targeted leucine codon optimization that transcription of TLRs is regulated by leucine codons to balance the abundance of certain receptors [15]. It is clear that gene synthesis may provide even unexpected benefits, but there are also numerous reports were it leads to reduced protein yield [16,17].

Adapting codon usage of target genes using codon optimization algorithms Exploiting gene synthesis for transgene expression enables researchers to influence the multitude of variables controlling protein yield. The biased usage of synonymous codons has a strong effect on protein expression [18]. Species-specific differences in this so-called codon bias can result in tRNA depletion, which slows down protein translation. This effect can be in part compensated by using an expression host that expresses rare tRNAs [16], however this is often insufficient [19]. An extensive collection of software tools (Table 1) are available to analyze codon usage and identify unwanted features that have been identified to hamper expression (Table 2). Gene design tools enable researchers to Current Opinion in Structural Biology 2016, 38:155–162

optimize the sequence of their gene of interest (GOI) [20]. A common strategy to manipulate codon bias to maximize protein translation is based on the ‘codon adaptation index’ (CAI) [21]. CAI is a species-specific index for codon frequency based on a set of highly expressing genes. Individual genes can then be analyzed, comparing the actual codons used to a fully optimized gene that contains only the most frequently used codons. It should be noted that CAI maximization does not necessarily correlate with high protein yield [22–25]. In general, the best results using CAI optimization are obtained when a subset of less frequent codons is replaced, after which a secondary list of criteria should be optimized involving DNA and RNA sequence elements that can negatively influence expression of the GOI (Table 2). It appears some features of codon bias are universal for all expression systems, whereas others are specific for prokaryotic or eukaryotic expression. Codon optimization algorithms will generally allow adjustment of several of these undesirable sequence properties simultaneously [20,26–28], although the weight given to each sequence parameter varies between different algorithms. It is important to note that proprietary CAI-based codon optimization algorithms adjust a subset of rare codons in a random manner. As a result, gene optimization is ambiguous, since the replacement of rare codons is not based on empirical data. Species-specific gene optimization routinely involves the use of the species-relevant CAI index, taking into account the codon frequency of the particular species. Other factors that affect for example eukaryotic protein expression are most often not considered, because the influence on expression levels is poorly understood.

Recent insights into the relationship between mRNA, protein expression and protein folding It is obvious that codon-mediated translational control is not yet well understood, and further insight will benefit gene design for protein expression. Recent studies indicate that rare codons may play different roles, such as safeguarding mRNA structure [29], depending on their positioning within the gene. A large scale study on bacterial genes expression in Escherichia coli under a common promotor system revealed that mRNA stability correlates with expression levels [30]. However, there is a difference in the effect between the ‘head’ of the gene (covering approximately 16 codons) and the rest of the gene. In the ‘head’, the mRNA structure is more important, whereas the rest of the gene is rather affected by codon-mediated translation elongation efficiency. This phenomenon is likely a universal property for both prokaryotic and eukaryotic protein expression [31]. Cell-free expression studies have shown that universal translation initation tags can be introduced upstream of the GOI that boost expression both in bacterial and eukaryotic systems [32,33]. To boost protein expression, it may be sufficient www.sciencedirect.com

The use of synthetic genes for protein expression Parret, Besir and Meijers 157

Table 1 Various tools to evaluate and optimize DNA sequences Name

Application Codon usage database

Web URL

Reference

http://www.kazusa.or.jp/codon/

[68]

Codon bias database

CBDB

http://homepages.luc.edu/cputonti/cbdb/

[69]

Analysis of codon usage

RaCC CAIcaI CAI calculator CodonO codonW GCUA INCA

http://nihserver.mbi.ucla.edu/RACC/ http://genomes.urv.cat/CAIcal/ http://www.umbc.edu/codon/cai/cais.php http://sysbio.cvm.msstate.edu/CodonO/ http://codonw.sourceforge.net/ http://gcua.schoedl.de/ http://bioinfo.hr/research/inca/ http://www.codons.org/index.html http://www.cmbl.uga.edu/software/codon_usage.html http://www.bioinformatics.org/sms2/codon_usage.html

Gene design tools

a

b

CodonOpt COOL b DNAWorksb D-Tailor c EuGenec GeneDesign b Gene Designerc GeneOptimizerb

JCat b mRNA Optimizerc OPTIMIZER b OptimumGeneb Synthetic Gene Designer b Visual Gene Developerc

https://eu.idtdna.com/CodonOpt http://cool.syncti.org/ https://hpcwebapps.cit.nih.gov/dnaworks/ https://sourceforge.net/projects/dtailor/ http://bioinformatics.ua.pt/eugene/ http://www.genedesign.org https://www.dna20.com/resources/genedesigner https://www.thermofisher.com/de/de/home/lifescience/cloning/gene-synthesis/geneart-genesynthesis/geneoptimizer.html http://www.jcat.de/ http://bioinformatics.ua.pt/software/mRNA-optimiser http://genomes.urv.es/OPTIMIZER/ http://www.genscript.com/cgi-bin/tools/rare_codon_analysis http://userpages.umbc.edu/wug1/codon/sgd/ http://visualgenedeveloper.net/index.html

[70] [71] [72] [73] [74] [38] [75] [76] [77] [78] [79] [80] [81] [82]

[83] [84] [85] [86] [87]

a

A very useful and comprehensive comparison of functionality and easiness of use of gene design tools for codon optimization is described by Gould et al. [26]. b Web-based tool. c Stand-alone software.

to modify the codon usage at the ‘head’ of the gene in order to optimize the mRNA structure for efficient translation initiation, and leave the codon sequence of the rest of the gene intact. In prokaryotes, translation of a transcript is initiated before the transcription process is complete, due to the recruitment of ribosomes to the newly synthesized mRNA. In genes which are not highly expressed, the translation efficiency is not necessarily correlated with protein synthesis rates, as ribosome profiling studies have shown [34–36]. These studies point to another level of regulation, where codon usage affects protein folding and post-translational modifications. A bioinformatics analysis in E. coli revealed that putative translational attenuation sites are widespread, and were identified in about 60% of protein coding sequences [37]. In this study, Li et al. proposed that these internal Shine–Dalgarno sequences are the main determinants of translation elongation rates in bacteria allowing for translational pausing where necessary. www.sciencedirect.com

Bioinformatic analyses indicate that rare codon clusters are found in a wide variety of genes, especially in the 50 and 30 termini [38–40]. Such clusters are thought to slowdown the ribosome during translation elongation in order to allow co-translational protein folding [41–43]. Slow elongation is especially important in case of membrane protein targeting [44,45,46]. As a consequence, modifying the codons involved in the regulation of translation speed can culminate in protein misfolding. In one particular case, the replacement of frequently used codons by rare codons at domain boundaries resulted in an improved solubility of an enzyme produced in E. coli due to slower translation rate at these sites [47]. To circumvent deleterious effects of codon usage during cotranslational protein folding, Angov et al. proposed a socalled codon harmonization strategy where rare codons are positioned where domain boundaries were predicted [48,49]. When compared to conventional codon optimization, this strategy resulted in improved expression level and soluble protein yield when overexpressing optimized genes from the eukaryotic parasite Plasmodium falciparum Current Opinion in Structural Biology 2016, 38:155–162

158 New constructs and expression of proteins

Table 2 Codon sequence features potentially influencing protein expression levels Sequence feature

Possible effect

General features G + C content Very high G + C content (>70%) Very low G + C content (<30%) mRNA secondary structure Global Near RBS Repetitive sequences Rare codons Global

References [15,22,88,89]

mRNA secondary structure formation; can slow down or inhibit translation Can slow down transcription elongation [31,53,56,90–92] Can promote mRNA stability Reduction of translational initiation Inhibition of translation through formation of mRNA stem–loops

[16]

Can result in translational pausing, tRNA depletion resulting in low protein yield, mRNA degradation Promotes accurate co-translational folding Shorter protein (if in frame); mixture of two proteins; incorrect protein if out of frame

[27,39,93,94]

Bacterial features Chi-site stretches Alternative transcriptional initiation sites RNase E sites SD-like RBS sequences

Recombination hotspots; DNA instability Leaderless transcripts; shorter or incorrect protein Can effect mRNA stability Can cause translational pausing (ribosome stalling)

[95] [96–98] [31,99] [37]

Eukaryotic features High CpG content Internal TATA-boxes Cryptic splice sites Potential polyA sites

Can result in incorrect transcription initiation Can result in incorrect transcription initiation Missplicing of mRNA; incorrect protein Can induce premature termination of translation

[88,100]

At protein domain boundaries Alternative start codons

in E. coli. All these studies indicate that in gene synthesis, it may be important to conserve codon usage in some parts of the gene, to maintain the regulatory elements that affect folding and post-translational modifications.

The interface between the synthetic gene and the expression vector It is important to keep in mind that the mRNA produced to translate into protein is often not limited to the synthetized gene, but contains fragments derived from the expression vector. In fact, the challenge to optimize

[40–42,48] [14]

[101] [102,103]

recombinant protein expression involves three intersecting areas. First, the choice of expression host, second, the selection of the expression vector [50] and third, the GOI (Figure 2). Much research has been done to optimize expression vectors adapted to the expression host. However, the adaptation of the GOI towards the expression vector is often neglected. Most bacterial vector systems are used as plug-and-play, where the GOI is inserted downstream of a vector-encoded ribosome binding site (RBS). As part of the 50 untranslated region (50 UTR), the RBS controls the rate of ribosome recruitment as well as

Figure 2

GOI

insert

vector rt

ta

RNA pool

s al

ion

Protein yield

fin

af

e

ag

t ity

e as

e

ot

pr

sit

I

GO

*

5’ UTR

Vector

Host

S n RB tra AUG

start of transcript

t sla

10 bp

protein synthesis

integration Current Opinion in Structural Biology

Optimization triangle illustrating the key determinants in recombinant protein expression. The inset illustrates the anatomy of a typical bacterial overexpression vector belonging to the popular pET-series in the region surrounding the insertion site of the GOI. The asterisk indicates a potential scar resulting from the cloning strategy. Current Opinion in Structural Biology 2016, 38:155–162

www.sciencedirect.com

The use of synthetic genes for protein expression Parret, Besir and Meijers 159

the efficiency of translation initiation [51]. It is generally assumed that the mRNA sequence of the 50 UTR of an expression vector is optimized in order to maximize the translational efficiency of the GOI within a particular host. However, the mRNA structure of the initial 40 coding nucleotides can be crucial to obtain high protein expression levels, given that in general strong secondary structure (due to e.g. high GC content) inhibits efficient translation initiation [22,24,52–55,56,57]. An additional global observation is the relatively higher abundance of rare codons in the initial 30–50 codons [58]. The correlation between the presence of rare codons and weak mRNA structure in the 50 end of coding sequences is still under debate [55]. In typical bacterial vectors, the 50 end of the coding sequence (excluding the GOI) often encodes secretion signals, affinity tags, fusion proteins or linker remnants resulting from cloning, but the mRNA structure in this region is rarely scrutinized (Figure 2). This may explain why the expression of some proteins is strongly affected by the placement of a tag on either the N-terminus or Cterminus or by the length and sequence of the linker region between tag and first codon of the transgene. Nterminal tags may have evolved towards optimal mRNA structure, thus increasing expression of some proteins with a peculiar codon sequence at the start of the gene. For example, it has been shown that addition of a leader sequence can overcome the notoriously low expression of AT-rich genes in E. coli [59]. On the basis of the recent evidence on the importance of codon usage at the translation initiation site, the choice of any expression vector should include a review of this region. If the protein expression is suboptimal, it could be worthwhile to optimize elements involved in translation initiation upstream of the GOI and include these in gene synthesis. In eukaryotes, translation initiation is an intricate stepwise process orchestrated by multiple components. Extensive studies in Saccharomyces cerevisiae in particular revealed that key regulatory elements, i.e. translation initiation motifs and secondary structures in the 50 -UTR, modulate protein abundance [58,60–62]. One particularly successful approach to manipulate protein yield in mammalian cells has been the incorporation of viral introns and optimized Kozak sequences into the translation initiation sequence of the corresponding expression vectors [63]. Translation termination can also be promoted by the introduction of viral introns such as the post-transcriptional regulatory element derived from woodchuck hepatitis virus [64]. It is expected that further large-scale investigations into the regulation of eukaryotic protein translation will yield more elements that can boost protein expression.

Concluding remarks Economically attractive access to gene synthesis has enabled optimization of gene sequences in order to www.sciencedirect.com

improve recombinant protein expression. Hence, species-specific gene optimization using proprietary algorithms can increase protein yield. Unfortunately, predicting the output of the different optimization algorithms is still a major challenge, especially for membrane and secreted proteins, due to the many parameters influencing codon usage. It would be a great advance if commonly used codon optimization algorithms were improved in the future to incorporate the new insights obtained in recent years with respect to codon usage. This will probably involve segmentation of the gene, where each segment has different requirements for codon optimization, including segments for translation initiation, domain boundaries and sites for post-translational modifications. Optimized protein translation initiation could be addressed by vector design, rather than codon optimization of the gene of interest. Indeed, synonymous exchange of a small number of codons or introducing non-coding secondary structure elements at the 50 -end of a gene can already be sufficient for a significant increase in expression levels [65,66]. This can be combined with gene fragment synthesis, consisting of linear DNA fragments that can be used for direct cloning into expression vectors or as template for PCR amplification. In this respect, the Biobricks movement (http://biobricks.org/) is a promising initiative to deliver direct access to tools to make proteins from genetically optimized building blocks [67]. The expectation is that broad studies on the relationship between codon usage and tRNA abundance as well as mRNA structure and stability will lead to a set of robust rules for gene design. This will further enable efficient protein expression that is affordable for the larger molecular biology community.

Conflict of interest statement Nothing declared.

Acknowledgements The authors are very grateful to Xia Sheng (Genscript) and Michel Koch for providing valuable input into the manuscript. We would like to apologize to those authors whose work could not be cited here due to space restrictions.

References and recommended reading Papers of particular interest, published within the period of review, have been highlighted as:  of special interest  of outstanding interest 1.

Vincentelli R, Cimino A, Geerlof A, Kubo A, Satou Y, Cambillau C: High-throughput protein expression screening and purification in Escherichia coli. Methods Struct Proteom 2011, 55:65-72.

2.

Makowska-Grzyska M, Kim Y, Maltseva N, Li H, Zhou M, Joachimiak G, Babnigg G, Joachimiak A: Protein production for structural genomics using E. coli expression. Methods Mol Biol 2014, 1140:89-105. Current Opinion in Structural Biology 2016, 38:155–162

160 New constructs and expression of proteins

3.

Assenberg R, Wan PT, Geisse S, Mayr LM: Advances in recombinant protein expression for use in pharmaceutical research. Curr Opin Struct Biol 2013, 23:393-402.

21. Sharp PM, Li WH: The codon Adaptation Index — a measure of directional synonymous codon usage bias, and its potential applications. Nucleic Acids Res 1987, 15:1281-1295.

4.

Kendrew JC, Dickerson RE, Strandberg BE, Hart RG, Davies DR, Phillips DC, Shore VC: Structure of myoglobin: a threedimensional Fourier synthesis at 2 A˚ resolution. Nature 1960, 185:422-427.

22. Kudla G, Murray AW, Tollervey D, Plotkin JB: Coding-sequence determinants of gene expression in Escherichia coli. Science 2009, 324:255-258.

5.

Morth JP, Pedersen BP, Toustrup-Jensen MS, Sørensen TL-M, Petersen J, Andersen JP, Vilsen B, Nissen P: Crystal structure of the sodium–potassium pump. Nature 2007, 450:1043-1049.

6.

Yusupova G, Yusupov M: Ribosome biochemistry in crystal structure determination. RNA 2015, 21:771-773.

24. Allert M, Cox JC, Hellinga HW: Multifactorial determinants of protein expression in prokaryotic open reading frames. J Mol Biol 2010, 402:905-918.

7.

Celie PHN, van Rossum-Fikkert SE, van Dijk WJ, Brejc K, Smit AB, Sixma TK: Nicotine and carbamylcholine binding to nicotinic acetylcholine receptors as studied in AChBP crystal structures. Neuron 2004, 41:907-914.

25. Menzella HG: Comparison of two codon optimization strategies to enhance recombinant protein production in Escherichia coli. Microb Cell Fact 2011, 10:15.

8.

Locher KP, Lee AT, Rees DC: The E. coli BtuCD structure: a framework for ABC transporter architecture and mechanism. Science 2002, 296:1091-1098.

26. Gould N, Hendy O, Papamichail D: Computational tools and algorithms for designing customized synthetic genes. Front Bioeng Biotechnol 2014, 2:41.

9.

Bostro¨m P, Wu J, Jedrychowski MP, Korde A, Ye L, Lo JC, Rasbach KA, Bostro¨m EA, Choi JH, Long JZ et al.: A PGC1adependent myokine that drives browning of white fat and thermogenesis. Nature 2012, 481:463-468.

10. Trempe J-F, Sauve´ V, Grenier K, Seirafi M, Tang MY, Me´nade M, Al-Abdul-Wahid S, Krett J, Wong K, Kozlov G et al.: Structure of parkin reveals mechanisms for ubiquitin ligase activation. Science 2013, 340:1451-1455. 11. Finci LI, Kru¨ger N, Sun X, Zhang J, Chegkazi M, Wu Y, Schenk G, Mertens HDT, Svergun DI, Zhang Y et al.: The crystal structure of netrin-1 in complex with DCC reveals the bifunctionality of netrin-1 as a guidance cue. Neuron 2014, 83:839-849. 12. Zhou M, Guo J, Cha J, Chae M, Chen S, Barral JM, Sachs MS, Liu Y: Non-optimal codon usage affects expression, structure  and function of clock protein FRQ. Nature 2013, 495:111-115. This paper describes how modification of codon usage in a circadian protein revealed how the use of specific codons regulates phosphorylation and structure stability. 13. Shah K, Cheng Y, Hahn B, Bridges R, Bradbury NA, Mueller DM: Synonymous codon usage affects the expression of wild type and F508del CFTR. J Mol Biol 2015, 427:1464-1479. 14. Dunne M, Leicht S, Krichel B, Mertens HDT, Thompson A, Krijgsveld J, Svergun DI, Go´mez-Torres N, Garde S, Uetrecht C et al.: Crystal structure of the CTP1L endolysin reveals how its activity is regulated by a secondary translation product. J Biol Chem 2016, 291:4882-4893. 15. Newman ZR, Young JM, Ingolia NT, Barton GM: Differences in codon bias and GC content contribute to the balanced expression of TLR7 and TLR9. Proc Natl Acad Sci U S A 2016, 113:E1362-E1371. 16. Gustafsson C, Govindarajan S, Minshull J: Codon bias and heterologous protein expression. Trends Biotechnol 2004, 22:346-353. 17. Wu G, Zheng Y, Qureshi I, Zin HT, Beck T, Bulka B, Freeland SJ: SGDB: a database of synthetic genes re-designed for optimizing protein over-expression. Nucleic Acids Res 2007, 35:D76-D79. 18. Quax TE, Claassens NJ, Soll D, van der Oost J: Codon bias as a  means to fine-tune gene expression. Mol Cell 2015, 59:149-161. Excellent review illustrating the different types of codon bias and its underlying principles. The authors describe the potential of codon bias to modulate gene expression and explore the challenges associated with exploiting codon bias as a tool for synthetic biology. 19. Rosano GL, Ceccarelli EA: Rare codon content affects the solubility of recombinant proteins in a codon bias-adjusted Escherichia coli strain. Microb Cell Fact 2009, 8:41. 20. Welch M, Villalobos A, Gustafsson C, Minshull J: Designing genes for successful protein expression. Methods Enzym 2011, 498:43-66. Current Opinion in Structural Biology 2016, 38:155–162

23. Welch M, Govindarajan S, Ness JE, Villalobos A, Gurney A, Minshull J, Gustafsson C: Design parameters to control synthetic gene expression in Escherichia coli. PLoS One 2009, 4:e7002.

27. Gustafsson C, Minshull J, Govindarajan S, Ness J, Villalobos A, Welch M: Engineering genes for predictable protein expression. Protein Expr Purif 2012, 83:37-46. 28. Chung B, Lee D-Y: Computational codon optimization of synthetic gene for protein expression. BMC Syst Biol 2012, 6:134. 29. Presnyak V, Alhusaini N, Chen YH, Martin S, Morris N, Kline N, Olson S, Weinberg D, Baker KE, Graveley BR et al.: Codon optimality is a major determinant of mRNA stability. Cell 2015, 160:1111-1124. 30. Boe¨l G, Letso R, Neely H, Price WN, Wong K-H, Su M, Luff JD,  Valecha M, Everett JK, Acton TB et al.: Codon influence on protein expression in E. coli correlates with mRNA levels. Nature 2016, 529:358-363. Large-scale protein expression study combined with biochemical analysis of optimized synthetic genes. This extensive analysis in E. coli strongly suggests that codon content directly modulates two competing processes, mRNA stability and protein elongation. 31. Del Campo C, Bartholoma¨us A, Fedyunin I, Ignatova Z: Secondary structure across the bacterial transcriptome reveals versatile roles in mRNA regulation and function. PLoS Genet 2015, 11:e1005613. 32. Haberstock S, Roos C, Hoevels Y, Do¨tsch V, Schnapp G, Pautsch A, Bernhard F: A systematic approach to increase the efficiency of membrane protein production in cell-free expression systems. Protein Expr Purif 2012, 82:308-316. 33. Mureev S, Kovtun O, Nguyen UTT, Alexandrov K: Speciesindependent translational leaders facilitate cell-free expression. Nat Biotechnol 2009, 27:747-752. 34. Brar GA, Weissman JS: Ribosome profiling reveals the what, when, where and how of protein synthesis. Nat Rev Mol Cell Biol 2015, 16:651-664. 35. Ingolia NT, Ghaemmaghami S, Newman JRS, Weissman JS: Genome-wide analysis in vivo of translation with nucleotide resolution using ribosome profiling. Science 2009, 324:218-223. 36. Ingolia NT: Ribosome profiling: new views of translation, from single codons to genome scale. Nat Rev Genet 2014, 15:205-213. 37. Li G-W, Oh E, Weissman JS: The anti-Shine–Dalgarno sequence drives translational pausing and codon choice in bacteria. Nature 2012, 484:538-541. 38. Clarke TF, Clark PL: Rare codons cluster. PLoS One 2008, 3:e3412. 39. Clarke TF, Clark PL: Increased incidence of rare codon clusters at 50 and 30 gene termini: implications for function. BMC Genomics 2010, 11:118. 40. Chartier M, Gaudreault F, Najmanovich R: Large-scale analysis of conserved rare codon clusters suggests an involvement in www.sciencedirect.com

The use of synthetic genes for protein expression Parret, Besir and Meijers 161

co-translational molecular recognition events. Bioinforma Oxf Engl 2012, 28:1438-1445. 41. Komar AA: A pause for thought along the co-translational folding pathway. Trends Biochem Sci 2009, 34:16-24. 42. Zhang G, Hubalewska M, Ignatova Z: Transient ribosomal attenuation coordinates protein synthesis and cotranslational folding. Nat Struct Mol Biol 2009, 16:274-280. 43. Pechmann S, Frydman J: Evolutionary conservation of codon optimality reveals hidden signatures of cotranslational folding. Nat Struct Mol Biol 2013, 20:237-243. 44. Fluman N, Navon S, Bibi E, Pilpel Y: mRNA-programmed  translation pauses in the targeting of E. coli membrane proteins. eLife 2014:3. Relevant study underlying the importance of slow protein elongation for membrane protein targeting in E. coli. 45. Pechmann S, Chartron JW, Frydman J: Local slowdown of translation by nonoptimal codons promotes nascent-chain recognition by SRP in vivo. Nat Struct Mol Biol 2014, 21:1100-1105. 46. Nørholm MHH, Light S, Virkki MTI, Elofsson A, von Heijne G, Daley DO: Manipulating the genetic code for membrane protein production: what have we learnt so far? Biochim Biophys Acta BBA Biomembr 2012, 1818:1091-1096. 47. Hess A-K, Saffert P, Liebeton K, Ignatova Z: Optimization of translation profiles enhances protein expression and solubility. PLOS ONE 2015, 10:e0127039. 48. Angov E, Hillier CJ, Kincaid RL, Lyon JA: Heterologous protein expression is enhanced by harmonizing the codon usage frequencies of the target gene with those of the expression host. PLoS One 2008, 3:e2189. 49. Angov E, Legler PM, Mease RM: Adjustment of codon usage frequencies by codon harmonization improves protein expression and folding. In Heterologous Gene Expression in E. coli. Edited by Evans , Xu M-QQ.. Humana Press; 2011:1-13. 50. Sørensen HP, Mortensen KK: Advanced genetic strategies for recombinant protein expression in Escherichia coli. J Biotechnol 2005, 115:113-128. 51. Salis HM, Mirsky EA, Voigt CA: Automated design of synthetic ribosome binding sites to control protein expression. Nat Biotechnol 2009, 27:946-950. 52. Araujo PR, Yoon K, Ko D, Smith AD, Qiao M, Suresh U, Burns SC, Penalva LOF: Before it gets started: regulating translation at the 50 UTR. Comp Funct Genomics 2012, 2012:1-8. 53. Kozak M: Regulation of translation via mRNA structure in prokaryotes and eukaryotes. Gene 2005, 361:13-37. 54. Gu W, Zhou T, Wilke CO: A universal trend of reduced mRNA stability near the translation-initiation site in prokaryotes and eukaryotes. PLoS Comput Biol 2010, 6:e1000664. 55. Tuller T, Zur H: Multiple roles of the coding sequence 50 end in gene expression regulation. Nucleic Acids Res 2015, 43:13-28. 56. Goodman DB, Church GM, Kosuri S: Causes and effects of  N-terminal codon bias in bacterial genes. Science 2013, 342:475-479. This paper reports expression data of an extensive library of synthetic reporter constructs consisting of 14,000 combinations of promoters, ribosome binding sites and 11 N-terminal codons fused to GFP. The magnitude of this study allows for detailed dissection of how individual Nterminal codons effect expression yield of GFP fusions. The authors report an up to 14-fold increase in expression when using N-terminal rare codons. 57. Shah P, Ding Y, Niemczyk M, Kudla G, Plotkin JB: Rate-limiting steps in yeast protein translation. Cell 2013, 153:1589-1601. 58. Tuller T, Carmi A, Vestsigian K, Navon S, Dorfan Y, Zaborske J, Pan T, Dahan O, Furman I, Pilpel Y: An evolutionarily conserved mechanism for controlling the efficiency of protein translation. Cell 2010, 141:344-354. 59. Ishida M, Oshima T, Yutani K: Overexpression in Escherichia coli of the AT-rich trpA and trpB genes from the hyperthermophilic www.sciencedirect.com

archaeon Pyrococcus furiosus. FEMS Microbiol Lett 2002, 216:179-183. 60. Dvir S, Velten L, Sharon E, Zeevi D, Carey LB, Weinberger A, Segal E: Deciphering the rules by which 50 -UTR sequences affect protein expression in yeast. Proc Natl Acad Sci U S A 2013, 110:E2792-E2801. 61. Crook NC, Schmitz AC, Alper HS: Optimization of a yeast RNA interference system for controlling gene expression and enabling rapid metabolic engineering. ACS Synth Biol 2014, 3:307-313. 62. Redden H, Morse N, Alper HS: The synthetic biology toolbox for tuning gene expression in yeast. FEMS Yeast Res 2014 http:// dx.doi.org/10.1111/1567-1364.12188. 63. Backliwal G, Hildinger M, Chenuet S, Wulhfard S, De Jesus M, Wurm FM: Rational vector design and multi-pathway modulation of HEK 293E cells yield recombinant antibody titers exceeding 1 g/l by transient transfection under serumfree conditions. Nucleic Acids Res 2008, 36:e96. 64. Zufferey R, Donello JE, Trono D, Hope TJ: Woodchuck hepatitis virus posttranscriptional regulatory element enhances expression of transgenes delivered by retroviral vectors. J Virol 1999, 73:2886-2892. 65. Karimi Z, Nezafat N, Negahdaripour M, Berenjian A, Hemmati S, Ghasemi Y: The effect of rare codons following the ATG start codon on expression of human granulocyte-colony stimulating factor in Escherichia coli. Protein Expr Purif 2015, 114:108-114. 66. Cheong D-E, Ko K-C, Han Y, Jeon H-G, Sung BH, Kim G-J, Choi JH, Song JJ: Enhancing functional expression of heterologous proteins through random substitution of genetic codes in the 50 coding region. Biotechnol Bioeng 2015, 112:822-826. 67. Sleight SC, Sauro HM: Randomized BioBrick assembly: a novel DNA assembly method for randomizing and optimizing genetic circuits and metabolic pathways. ACS Synth Biol 2013, 2:506-518. 68. Nakamura Y, Gojobori T, Ikemura T: Codon usage tabulated from international DNA sequence databases: status for the year 2000. Nucleic Acids Res 2000, 28:292. 69. Hilterbrand A, Saelens J, Putonti C: CBDB: the codon bias database. BMC Bioinformatics 2012, 13:62. 70. Puigbo` P, Bravo IG, Garcia-Vallve S: CAIcal: a combined set of tools to assess codon usage adaptation. Biol Direct 2008, 3:38. 71. Wu G, Culley DE, Zhang W: Predicted highly expressed genes in the genomes of Streptomyces coelicolor and Streptomyces avermitilis and the implications for their metabolism. Microbiol Read Engl 2005, 151:2175-2187. 72. Angellotti MC, Bhuiyan SB, Chen G, Wan X-F: CodonO: codon usage bias analysis within and across genomes. Nucleic Acids Res 2007, 35:W132-W136. 73. Fuhrmann M, Hausherr A, Ferbitz L, Scho¨dl T, Heitzer M, Hegemann P: Monitoring dynamic expression of nuclear genes in Chlamydomonas reinhardtii by using a synthetic luciferase reporter gene. Plant Mol Biol 2004, 55:869-881. 74. Supek F, Vlahovicek K: INCA: synonymous codon usage analysis and clustering by means of self-organizing map. Bioinforma Oxf Engl 2004, 20:2329-2330. 75. Stothard P: The sequence manipulation suite: JavaScript programs for analyzing and formatting protein and DNA sequences. BioTechniques 2000, 28:1104. 76. Chin JX, Chung BK-S, Lee D-Y: Codon Optimization OnLine (COOL): a web-based multi-objective optimization platform for synthetic gene design. Bioinforma Oxf Engl 2014, 30:2210-2212. 77. Hoover DM: DNAWorks: an automated method for designing oligonucleotides for PCR-based gene synthesis. Nucleic Acids Res 2002, 30:43e. Current Opinion in Structural Biology 2016, 38:155–162

162 New constructs and expression of proteins

78. Guimaraes JC, Rocha M, Arkin AP, Cambray G: D-Tailor: automated analysis and design of DNA sequences. Bioinforma. Oxf. Engl. 2014 http://dx.doi.org/10.1093/bioinformatics/btt742. 79. Gaspar P, Oliveira JL, Frommlet J, Santos MAS, Moura G: EuGene: maximizing synthetic gene design for heterologous expression. Bioinforma Oxf Engl 2012, 28:2683-2684. 80. Richardson SM, Wheelan SJ, Yarrington RM, Boeke JD: GeneDesign: rapid, automated design of multikilobase synthetic genes. Genome Res 2006, 16:550-556. 81. Villalobos A, Ness JE, Gustafsson C, Minshull J, Govindarajan S: Gene Designer: a synthetic biology tool for constructing artificial DNA segments. BMC Bioinformatics 2006, 7:285. 82. Fath S, Bauer AP, Liss M, Spriestersbach A, Maertens B, Hahn P, Ludwig C, Scha¨fer F, Graf M, Wagner R: Multiparameter RNA and codon optimization: a standardized tool to assess and enhance autologous mammalian gene expression. PLoS ONE 2011, 6:e17596. 83. Grote A, Hiller K, Scheer M, Munch R, Nortemann B, Hempel DC, Jahn D: JCat: a novel tool to adapt codon usage of a target gene to its potential expression host. Nucleic Acids Res 2005, 33:W526-W531. 84. Gaspar P, Moura G, Santos MAS, Oliveira JL: mRNA secondary structure optimization using a correlated stem–loop prediction. Nucleic Acids Res 2013, 41:e73. 85. Puigbo P, Guzman E, Romeu A, Garcia-Vallve S: OPTIMIZER: a web server for optimizing the codon usage of DNA sequences. Nucleic Acids Res 2007, 35:W126-W131.

92. Pop C, Rouskin S, Ingolia NT, Han L, Phizicky EM, Weissman JS, Koller D: Causal signals between codon bias, mRNA structure, and the efficiency of translation and elongation. Mol Syst Biol 2014, 10:770. 93. Kane JF: Effects of rare codon clusters on high-level expression of heterologous proteins in Escherichia coli. Curr Opin Biotechnol 1995, 6:494-500. 94. Drummond DA, Wilke CO: Mistranslation-induced protein misfolding as a dominant constraint on coding-sequence evolution. Cell 2008, 134:341-352. 95. Taylor AF, Amundsen SK, Smith GR: Unexpected DNA contextdependence identifies a new determinant of Chi recombination hotspots. Nucleic Acids Res 2016 http:// dx.doi.org/10.1093/nar/gkw541. 96. Chenge J, Kavanagh ME, Driscoll MD, McLean KJ, Young DB, Cortes T, Matak-Vinkovic D, Levy CW, Rigby SEJ, Leys D et al.: Structural characterization of CYP144A1 — a cytochrome P450 enzyme expressed from alternative transcripts in Mycobacterium tuberculosis. Sci Rep 2016, 6:26628. 97. Shell SS, Wang J, Lapierre P, Mir M, Chase MR, Pyle MM, Gawande R, Ahmad R, Sarracino DA, Ioerger TR et al.: Leaderless transcripts and small proteins are common features of the mycobacterial translational landscape. PLoS Genet 2015, 11:e1005641. 98. Srivastava A, Gogoi P, Deka B, Goswami S, Kanaujia SP: In silico analysis of 50 -UTRs highlights the prevalence of Shine– Dalgarno and leaderless-dependent mechanisms of translation initiation in bacteria and archaea, respectively. J Theor Biol 2016, 402:54-61.

86. Wu G, Bashir-Bello N, Freeland SJ: The Synthetic Gene Designer: a flexible web platform to explore sequence manipulation for heterologous expression. Protein Expr Purif 2006, 47:441-445.

99. Mackie GA: RNase E: at the interface of bacterial RNA processing and decay. Nat Rev Microbiol 2012, 11:45-57.

87. Jung S-K, McDonald K: Visual gene developer: a fully programmable bioinformatics software for synthetic gene optimization. BMC Bioinformatics 2011, 12:340.

100. Jonkers I, Kwak H, Lis JT: Genome-wide dynamics of Pol II elongation and its interplay with promoter proximal pausing, chromatin, and exons. eLife 2014, 3:e02407.

88. Bauer AP, Leikam D, Krinner S, Notka F, Ludwig C, Langst G, Wagner R: The impact of intragenic CpG content on gene expression. Nucleic Acids Res 2010, 38:3891-3908.

101. Bukovac SW, Bagshaw RD, Rigat BA, Callahan JW, Clarke JTR, Mahuran DJ: Cryptic splice site in the complementary DNA of glucocerebrosidase causes inefficient expression. Anal Biochem 2008, 381:276-278.

89. Kudla G, Lipinski L, Caffin F, Helwak A, Zylicz M: High guanine and cytosine content increases mRNA levels in mammalian cells. PLoS Biol 2006, 4:e180.

102. Proudfoot NJ: Ending the message: poly(A) signals then and now. Genes Dev 2011, 25:1770-1782.

90. Kertesz M, Wan Y, Mazor E, Rinn JL, Nutter RC, Chang HY, Segal E: Genome-wide measurement of RNA secondary structure in yeast. Nature 2010, 467:103-107.

103. Pfarr DS, Rieser LA, Woychik RP, Rottman FM, Rosenberg M, Reff ME: Differential effects of polyadenylation regions on gene expression in mammalian cells. DNA Mary Ann Liebert Inc 1986, 5:115-122.

91. Bentele K, Saffert P, Rauscher R, Ignatova Z, Blu¨thgen N: Efficient translation initiation dictates codon usage at gene start. Mol Syst Biol 2013, 9:675.

Current Opinion in Structural Biology 2016, 38:155–162

104. Chayen NE: Turning protein crystallisation from an art into a science. Curr Opin Struct Biol 2004, 14:577-583.

www.sciencedirect.com