High-level expression of heterologous proteins
Carsten-Peter Carstens Julie Bonnardel Anna Waesche
Codon bias is a significant obstacle for efficient expression of heterologous genes in E. coli hosts. In GC-rich genomes, such as mammals, rare arginine codons (AGG or AGA) and the proline codon (CCC) most frequently affect bacterial gene expression. For such situations, Stratagene introduces BL21-CodonPlus-RP cells* to make protein expression more reliable. This strain contains a ColE1-compatible plasmid that encodes extra copies of the argU and proL tRNA genes and is able to rescue expression of genes restricted by either AGG/ AGA codons or CCC codons.
The E. coli BL21 strains offer many advantages for expression of heterologous proteins. Derived from E. coli B, these strains naturally lack the Lon protease and are engineered to be deficient for the OmpT protease. The BL21(DE3) and the BL21(DE3)pLysS hosts permit induced expression of T7 RNA polymerase. When used in conjunction with pET-type vectors, the expressed T7 RNA polymerase provides significant levels of transcription of heterologous genes. Stratagenes BL21-Gold competent cells** are another member of the BL21 family of cells that feature two additional advantages: They are EndA1 deficient, which allows preparation of high-quality miniprep DNA, and they have the Hte phenotype, which confers increased transformation efficiency.1
Even with these many innovations, the expression of heterologous proteins in E. coli can be difficult. Most frequently, failed or insufficient expression of the heterologous gene is caused by the presence of codons that are rarely used in E. coli.2,3 Forced high-level expression of genes with rare codons for the E. coli host can lead to depletion of the endogenous pools of the corresponding tRNAs. This in turn, leads to slowing or abortion of the translation process and subsequent degradation of the mRNA.4 When the problem of codon bias causes frameshifts, codon skipping, or misincorporations, it may escape detection. More commonly, manifestations of codon bias include no or low protein expression and the presence of aborted translation products. Previously, the typical remedies for overcoming codon bias were to alter the codon specifications of a target gene by site-directed mutagenesis or to circumvent the problem by switching to a eukaryotic expression system.
To make it easier to resolve the codon bias problem, Stratagene released BL21-CodonPlus-RIL competent cells,5 which are derivatives of the BL21-Gold series. BL21-CodonPlus-RIL cells contain a ColE1-compatible vector with extra copies of the argU, ileY, and leuW tRNA genes. These tRNAs recognize arginine, isoleucine, and leucine codons (AGA/AGG, AUA, and CUA), respectively. Host cell expression of these tRNAs can lead to increased synthesis of recombinant proteins, as translation is no longer limited by the availability of tRNAs that recognize rare codons. However, using these codons is predominantly a problem in organisms with AT-rich genomes (Table 1). For organisms whose genomes are GC rich, the problematic codons for expression are the arginine codon AGG (recognized by the tRNA product of the argU gene) and the proline codon CCC (recognized by the tRNA product of the proL gene).
Hence, Stratagene designed BL21-CodonPlus-RP and BL21-CodonPlus(DE3)-RP competent cells to solve this dilemma. These cells include a ColE1-compatible expression plasmid that contains the tRNA genes argU and proL. The corresponding tRNAs recognize the arginine codons, AGA/AGG, and the proline codon, CCC, respectively.
We per formed a series of experiments to demonstrate the effectiveness of the BL21-CodonPlus strains in rescuing expression of genes that are affected by rare codon usage. We transformed BL21-CodonPlus(DE3)-RIL cells and BL21-Gold(DE3) cells with plasmids encoding the following genes regulated by the T7 RNA polymerase-responsive promoter (Figure 1): human cardiac troponin-T (hcTnT-wt) 5, yeast Hsp 104, and four genes of archael origin. For each recombinant gene, protein synthesis greatly increased in the BL21-CodonPlus(DE3)-RIL cells, as compared to BL21-Gold(DE3) cells. Expression of the control genes, which do not encode rare codons (l-phosphastase and c-Jun N-terminal kinase (JNK), was equivalent in BL21-CodonPlus(DE3)-RIL and BL21-Gold(DE3) cells. Moreover, extra copies of the tRNA genes did not cause any obvious deleterious effects to the host cells. Accordingly, BL21-CodonPlus cells can also be used as hosts for expressing genes that are expressed well in conventional E. coli strains.
To confirm the function of the argU and proL tRNAs in BL21-CodonPlus(DE3)-RP cells, we performed two test transformations (Figure 2). BL21-CodonPlus(DE3)-RP and BL21-Gold(DE3) cells were transformed with plasmids encoding the hcTnT-wt (argU dependent) or CBP-3xP-Cre (proL dependent) genes. The BL21-CodonPlus(DE3)-RP cells, but not the parental BL21-Gold(DE3) cells, restored expression of both genes. These data demonstrate the suitability of BL21-CodonPlus(DE3)-RP cells for high-level expression of heterologous genes restricted by the presence of AGG/AGA and CCC codons in conventional BL21 hosts.
Stratagenes BL21-CodonPlus competent cells resolve the problem of codon bias for expressing heterologous genes in E. coli. While the BL21-CodonPlus-RIL strain is designed for the expression of AT-rich genes, the new BL21-CodonPlus-RP cells specifically assist gene expression from GC-rich organisms. In both strains, extra copies of tRNA genes rescue expression of genes that would otherwise be restricted by the presence of rare codons; however, these additional genes cause no deleterious effects to the host cells. Consequently, BL21-CodonPlus-RIL and BL21-CodonPlus-RP competent cells are useful strains for standard expression in E. coli, even in the absence of a detected codon bias problem.
Jerpseth, M., et al. (1998) Strategies 11: 3-4.
Kane, J. (1995) Curr. Opin. Biotechnol. 6: 494-500.
Bonekamp, F., et al. (1985) Nuc. Acids Res. 13: 4113-4123.
Deana, A., Ehrlich, R., and Reiss, C. (1998) Nuc. Acids Res. 26: 4778-4782.
Carstens, C.-P. and Waesche, A. (1999) Strategies 12: 49-51.
Hu, X., et al. (1996) Protein Expr. Purif. 7: 289-293.
* Patents pending
** Patent pending