Enhanced biological activity of recombinant human interferon alpha produced in Pichia pastoris using a codon-optimized synthetic cDNA.
Most of the 600 biopharmaceuticals currently on the market are recombinant proteins. A variety of strategies are being used by designers and manufacturers of biopharmaceuticals to optimize production yield. One such method is the use of recombinant protein production systems such as those involving the bacterium Escherichia coli and the yeasts Saccharomyces cerevisiae and Pichia pastoris  . Early in the drug development phase, optimization strategies use genetic engineering and molecular biology techniques, whereas in later development stages, they use approaches such as high cell density cultivation, manipulation of the culture environment, downstream processing and engineering of the fermentation process .
Gene optimization strategies are based on the codon bias found in almost all species. This bias exists among the 61 amino acid codons found in mRNA molecules, and the levels of aminoacyl-tRNAs seem to correlate with the frequency of codon usage [3,4].
Therefore, one would expect to observe the impaired translation of an abundant mRNA species containing an excess of rare codons with low aminoacyl-tRNA levels. Such a situation arises after the initiation of the transcription of cloned heterologous genes in E. coli and yeast [5, 6]. Codons that are common in some species may be rare in others; therefore, expression can be limited by the amount of available aminoacyl-tRNA in the host cell. The premature termination of translation can occur when specific aminoacyl-tRNAs are depleted, and transcription can be terminated if the DNA has a high proportion of A and T bases [5, 6].
The E. coli and P. pastoris expression systems are currently the most widely-used systems for the large-scale production of various recombinant proteins of pharmaceutical interest. Recent developments with respect to the P.pastoris system have impacted not only the protein expression levels that can be achieved but also the quality of the heterologous proteins produced [7, 8]. When expressed in yeast, however, the codons of a human gene may not be optimal for the high expression of recombinant protein. Gene optimization strategies call for alterations in codon-usage and an increase in the proportion of G and C bases to improve expression levels. The consensus sequence that initiates translation also needs to be optimized.
Gene optimization strategies have been used successfully to express several human genes in Escherichia coli and P. pastoris including huIFN[alpha] [9,10] The expression of the hepatitis B virus e antigen using a cDNA with a codon bias reflecting that of P. pastoris, resulted in 5 times higher protein expression levels . The codon optimization of the human DNA sequence encoding glucocerebrosidase, a protein used to treat Gaucher disease, resulted in a 10.6-fold increase in the amount of protein expressed in P. pastoris . In another report, the optimization of a mycobacterial cDNA enhanced the Immunogenicity of a DNA Vaccine Encoding for antigen Ag85B . While the use of gene optimization to improve production yield is well documented, there are reports showing the effects of synonymous codon change on the biological activity of proteins [14,15,16].
Here, we describe a gene optimization method in P. pastoris that improved not only the production yield but also the biological activity of recombinant human interferon [alpha]2a (huIFN[alpha]2a). The yields achieved with partially optimized cDNAs were approximately 1.5 to 1.7 times higher than that achieved with the wild type cDNA, while a fully synthetic (FS) cDNA displayed a 3-fold higher yield and a 2.85-fold higher specific activity than the wild type cDNA.
Materials and Methods
Strains and plasmids
The Top10fF' strain of E. coli ([recA.sup.-], [endA.sup.-]) (Invitrogen, Groningen, Netherlands) was used as the host strain for cloning experiments. The P.pastoris host strain was [Mut.sup.S] KM71H (aox1::ARG4). The pPICZ[alpha]A plasmid (Invitrogen, Groningen, Netherlands) was used for expression in P. pastoris.
Composition of bacterial medium
Low-salt (< 90 mM) Luria-Bertani (LB) medium containing 1% tryptone, 0.5% NaCl, 0.5% yeast extract and Zeocin at a final concentration of 25 [micro]g/ml was used at a pH of 7.5.
Composition of media for shake flask production
Yeast extract-peptone-dextrose (YPD) medium was used containing 2% peptone, 1% yeast extract and 2% dextrose. To make YPDS medium, YPD was supplemented with 1 M sorbitol and Zeocin at a final concentration of 100 [micro]g/ml. Buffered minimal glycerol-complex medium (BMGY) was prepared using 2% peptone, 1% yeast extract, 1% glycerol, 1.34% yeast nitrogen base (YNB) with ammonium sulfate but without amino acids and biotin in 100-mM potassium phosphate buffer at pH 6.0. Buffered minimal methanol-complex medium (BMMY) had the same composition as BMGY medium, except 1% methanol was used instead of glycerol.
Construction of the recombinant pPICZ[alpha]A/rWHuIFN[alpha]2 expression vector
The human interferon [alpha]2b cDNA was cloned as described by Rabhi-Essafi et al., 2007. The cDNA was amplified using the F/IFN[alpha]2 forward primer designed to introduce an EcoRI site at the 5' end of the gene (Table 1) and the R/IFN[alpha]2 reverse primer designed to introduce a NotI site and a TGA stop codon at the 3' end of the gene (Table 1). The 498-bp RT-PCR product was purified and cut with EcoRI and NotI, then inserted into the 3.6-kb plasmid pPICZ[alpha]A (Invitrogen, Groningen, the Netherlands) to generate the pPICZ[alpha]A/rhuIFN[alpha]2a expression vector. Transformed E. coli Top10F' ([recA.sup.-]/[endA.sup.-]) clones were selected on low-salt LB medium with 25 [micro]g/ml Zeocin. The pPICZ[alpha]A/rhuIFN[alpha]2a recombinant plasmid containing the cDNA sequence encoding human IFN[alpha]2 was isolated by colony PCR using the 5' F/IFN[alpha]2a forward and 3' R/IFN[alpha]2a reverse primers according to procedures described by Ausubel et al., 2002 and by restriction analysis using the BglII restriction enzyme as recommended by the manufacturer (Amersham Biosciences, Athens, Greece). Finally, the nucleotide sequences of the positive clones were confirmed by DNA sequencing using an ABI PRISM 377 DNA sequencer (Perkin Elmer Applied Biosystems) and the Aox1/F and Aox1/R primers from the EasySelect Pichia expression kit (Invitrogen).
Design of P. pastoris codon-optimized human interferon [alpha]2 cDNA clones (WS, SW and FS)
The wild-type IFNa2a nucleotide sequence was analyzed with Graphical Codon Usage Analyzer software, which uses the nearest-neighbor method, available at http://www.geneart.degcua.schoedl.de/ or from ExPASy Proteomics tools at www.Swissprot.org (Fig. 1).
[FIGURE 1 OMITTED]
The human IFN[alpha]2a sequence was used as a template to design the first huIFN[alpha]2a synthetic fragment by PCR assembly using the synthetic primers Fa1, Fb1 and Fc1 (Table 1) and two external primers, 5' forward FfgsI and 3' reverse RfgsI. The Fa2, Fb2 and Fc2 primers and two external primers, 5' forward FfgsII and 3' reverse RfgsII, were used to generate the second synthetic human IFN[alpha]2 cDNA fragment.
The purified 200-bp DNA fragment corresponding to the first huIFN[alpha]2a synthetic fragment was cloned into the recombinant pGX4T1/rhuWIFN[alpha]2a plasmid (Fig. 2A), which was previously cut at the 5' EcoRI and 3' BglII restriction sites, generating a plasmid containing only the 48 bp corresponding to the 16 amino acids of the huIFN[alpha]2a C-terminus. According to the codon bias of P. pastoris, the 16 C-terminus amino acids of huIFN[alpha]2a do not contain any rare codons. Isolation of the recombinant P1 plasmid (pGEX4T1/huIFN[alpha]2a, containing the first synthetic fragment) was performed by restriction analysis using BglII. Finally, sequences of the positive clones were confirmed by DNA sequencing using the ABI PRISM 377 DNA sequencer (Perkin Elmer Applied Biosystems, Palo Alto, CA ,USA) and the 5' pGEX primer.
[FIGURE 2A OMITTED]
The second 267-bp huIFN[alpha]2a synthetic fragment was introduced into the P1 recombinant plasmid cut with BglII. Isolation of the P2 recombinant plasmid clone in pGEX4T1, which contained the huIFNa2a synthetic fragment 2 inserted in the correct orientation, was performed by colony PCR using the 5' FfgsI forward and 3' RfgsII reverse primers (Table 1) and by restriction analysis using EcoRI and XhoI as recommended by the manufacturer (Amersham Biosciences, Athens, Greece). Full-length synthetic huIFN[alpha]2a was cut with EcoRI and NotI and inserted into the pPICZ[alpha]A plasmid to generate the pPICZ[alpha]A/FS-huIFN[alpha]2a expression vector. Transformed E. Coli Top10F' ([recA.sup.-]/[endA.sup.-]) clones were selected on low-salt LB medium with 25-[micro]g/ml Zeocin. Isolation of the pPICZ[alpha]A/synthetic huIFN[alpha]2 recombinant plasmid containing the cDNA sequence encoding synthetic human IFN[alpha]2a was performed by colony PCR using the 5' FfgsI forward and 3' RIFN[alpha]2 reverse primers (Table 1) as described in reference  and by restriction analysis using EcoRI and NotI as recommended by the manufacturer (Amersham Biosciences, Athens, Greece). Finally, the nucleotide sequences of positive clones were confirmed by DNA sequencing using the Aox1/F and Aox1/R primers from the EasySelect Pichia expression kit (Invitrogen, Groningen, the Netherlands).
To generate chimeric clones with a half synthetic and half wild-type IFN[alpha] nucleotide sequence (Fig. 2D), the pGEX4T1/SynrhuIFN[alpha]2a synthetic full-length construct (P2 vector) was used as a template as follows. For the SW clone, the 267-bp DNA fragment corresponding to the second rhuIFN[alpha]2 wild-type fragment was cut with BglII and inserted into the pGEX4T1/SynrhuIFN[alpha]2a synthetic full-length recombinant vector (P2 vector) cut with the same restriction enzyme. For the WS chimeric clone (Fig. 2D), the second 267-bp huSynIFN[alpha]2 synthetic fragment was cut with BglII and cloned into the pGEX4T1/rWhuIFN[alpha]2a wild-type full-length vector cut with the same restriction enzyme.
[FIGURE 2D OMITTED]
Shake flask production of secreted rhuIFN[alpha]2
A single colony of pPICZ[alpha]A/rHuIFN[alpha]2 [Mut.sup.S] KM71H P.pastoris transformants was used to inoculate 50 ml of BMGY medium in a 500-ml flask and then incubated at 30[degrees]C with shaking at 250 rpm for 24 h. Cells were pelleted by centrifugation at 3,000 g for 10 min, resuspended in 5 ml of BMMY medium containing 0.5% methanol and incubated at 30[degrees]C with shaking at 250 rpm. This induction was repeated every 24 hours at a concentration of 1% methanol directly after addition. After 48 h of induction, cells were harvested by centrifugation at 10,000 g for 20 min at 4[degrees]C, and the supernatant was collected and stored at -20[degrees]C for protein expression analysis.
Electroporation of pPICZ[alpha]A/rhuIFN[alpha]2 cDNAs into P. pastoris [Mut.sup.S] KM71H
Recombinant pPICZ[alpha]A/huIFN[alpha]2a expression vectors were propagated in Top10F' ([recA.sup.-][endA.sup.-]) E. coli in the presence of 25-[micro]g/ml Zeocin. Plasmids were isolated from the transformed E. coli clones using a Qiagen plasmid miniprep kit. (Heidelberg, Germany)
Each of the pPICZ[alpha]A-modified huIFN[alpha]2 plasmids (FS, WS and SW) was cut according to the manufacturer's instructions using SacI, which does not cut within the modified IFN[alpha]2 cDNA. pPICZ[alpha]A containing the wild-type rW-huIFN[alpha]2 was cut with BstX1. The cut fragments were purified by phenol/chloroform/IsoAmyl Alcohol purification following standard protocols. Ten micrograms of linearized, recombinant plasmid DNA was used to transform the KM71H (aox1::ARG4) P. pastoris strain by electroporation (Bio-Rad Gene Pulser, 1500 volts charging voltage, 25-[micro]F capacitance, and 200 ohms resistance) as described in the [EasySelect.sup.TM] Pichia Expression Kit manual (Invitrogen). After 3 days of incubation at 30[degrees]C on solid YPDS medium containing Zeocin at 100 [micro]g/ml, 20 KM71H [Mut.sup.S] transformants were retained for further study.
Analysis of rhuIFN[alpha]2 protein expression
Culture supernatants were analyzed by SDS-PAGE. Electrophoresis was performed in a 15% SDS-polyacrylamide gel that was then stained with Coomassie Brilliant Blue. Recombinant fusion proteins were specifically detected by western blot analysis using an ECL kit (Amersham Biosciences, Athens, Greece) according to the manufacturer's instructions with a 1:400 dilution of an anti-human IFN[alpha] monoclonal antibody (Endogen Searchlight, U.S.) followed by a peroxidase-conjugated anti-goat/sheep IgG monoclonal secondary antibody (Sigma Aldrich, Germany). Image J software was used to compare hurIFN[alpha]2a protein expression after codon optimization. Finally, the concentration of rWT-huIFN[alpha]2a compared to synthetic rhuIFN[alpha]2 expressed from several Pichia pastoris recombinant clones was determined by a quantitative ELISA developed in-house. Dilution series containing 0 to 570 pg of HPLC-purified soluble IFN[alpha]2 produced in our laboratory were included in each assay to generate a standard curve. Recombinant proteins were detected using a biotin-labeled anti-human IFN[alpha] monoclonal antibody (ENDOGEN Searchlight, US) and a colorimetric detection system with a streptavidin-horseradish peroxidase (HRP) conjugates (Amersham Biosciences, Athens, Greece).
Biological activity of rhIFN[alpha]2
The biological activity of recombinant huIFN[alpha]2a was determined using an antiviral assay as described in reference . This assay is based on the ability of huIFN[alpha]2a to inhibit the cytopathic effects of encephalomyocarditis virus (EMCV) on the glioblastoma cell line 2D9. HEK 293P cell lines stably transfected with an IFN-inducible promoter sequence (ISRE) linked to the secreted alkaline phosphatase (SEAP) gene were used to perform a reporter gene assay. One unit of activity was defined as the amount of recombinant hIFN[alpha]2a required to produce antiviral activity equivalent to that of 1 IU of the huIFN[alpha]2 reference standard (code: 95/566; Division of Immunobiology; National Institute for Biological Standards and Control,[NIBSC] Potters Bar, UK). IFN potency values were calculated using an in-house parallel line displacement program with the IFN 95/566 standard as the primary calibrator. Potency values were statistically analyzed using Prism software developed at NIBSC. For each recombinant IFN preparation, the assay was performed in duplicate.
The design and construction of modified synthetic codon-optimized human interferon [alpha]2a cDNAs
The design strategy for the fully synthetic clone (FS-huIFN[alpha]2)
The sequence of the wild-type human IFN[alpha]2 cDNA (WT-IFN[alpha]2a) was analyzed using Graphical Codon Usage Analyzer software, which uses the nearest-neighbor method. This analysis allowed the identification of codons in the native IFN[alpha] sequence rarely used by P. pastoris, equaling 25% of the total number of codons (Fig. 1).
To construct a fully synthetic human IFN[alpha]2a cDNA, we designed a cloning approach based on synthetic overlapping primers and a PCR assembly strategy. The detailed steps of the cloning strategy are shown in Fig. 2A, B and C.
[FIGURE B OMITTED]
[FIGURE 2C OMITTED]
Data from the codon sequence analysis were used to design synthetic primers Fa1, Fb1, and Fc1 and to generate the first huIFNa2a synthetic fragment (fragment 1) by PCR assembly (Fig. 2A).
Primers Fa2, Fb2, and Fc2 were used to generate the second synthetic human IFN[alpha]2 cDNA fragment (fragment 2) (Fig. 2B). Gene synthesis and fragment assembly were carried out using the PCR method described in Fig. 2A and B. The purified 186-bp DNA fragment corresponding to the first huIFN[alpha]2 synthetic fragment was cloned into pGX4T1/rhuIFN[alpha]2a (wild type cDNA sequence) previously cut at the 5' EcoRI and 3' BglII restriction sites to generate a plasmid containing only the 48 bp corresponding to the 16 residues at the huIFN[alpha]2 COOH terminus. Based on the codon usage bias of Pichia pastoris, the coding sequence for these 16 residues does not contain rare codons. The isolation of the P1 recombinant plasmid (pGEX4T1 containing the synthetic fragment corresponding to the C-terminal half of human IFN[alpha]2) (Fig. 2A) was performed by restriction digestion with BglII. Ten clones integrated the first synthetic fragment. Finally, the sequences of the positive clones were confirmed by DNA sequencing.
The second 268-bp huIFN[alpha]2a synthetic fragment obtained by PCR assembly was introduced into the P1 recombinant plasmid cut with BglII. The recombinant clones that had integrated this fragment in the correct orientation were identified by colony PCR using the 5' F/fgsI and 3' R/fgsII primers as described by Ausubel et al., 2000. The integrity of the full-length rhuSynIFN[alpha]2a sequence was checked by DNA sequencing and restriction analysis using EcoRI and XhoI. Full-length synthetic huSynIFN[alpha]2a was cut with EcoRI and NotI and inserted into the pPICZ[alpha]A plasmid (Invitrogen) to generate the pPICZ[alpha]A/rSynHuIFN[alpha]2a expression vector (Fig. 2C).
Chimeric clones: A half synthetic, half wild type (SW and WS) design strategy
The SW chimera was designed to have an optimized synthetic N-terminal half sequence (up to codon [TTC.sup.192]) and a wild-type C-terminal half sequence (Fig. 2D). Clone SW was obtained by cloning the SW chimeric sequence between the EcoRI and NotI restriction sites in the pPICZ[alpha]A plasmid (Invitrogen), creating the pPICZ[alpha]A/rSWhuIFN[alpha]2a expression vector. The WS chimera was designed to have a wild-type N-terminal half sequence (up to codon [TTC.sup.192]) and a synthetic C-terminal half sequence (Fig. 2D) and was obtained by cloning the WS chimeric sequence between the EcoRI and NotI restriction sites in the pPICZ[alpha]A plasmid, creating the pPICZ[alpha]A/rWShuIFN[alpha]2a expression vector.
The selection of the pPICZ[alpha]A/rSW-huIFN[alpha]2 and pPICZ[alpha]A/rWS-huIFN[alpha]2a recombinant plasmids (SW and WS, respectively) containing the cDNA sequence encoding the human IFN[alpha]2a chimera was performed by DNA sequencing using the 5' AOXF and the 3' AOXR primers, respectively.
Isolation of clones producing high protein levels
To isolate clones producing high protein levels, we used a rapid and direct method based on growth on increasing concentrations (500, 1000 and 2000 [micro]g/ml) of Zeocin . We assumed that clones resisting the highest concentration of antibiotic [2000 [micro]g/ml of Zeocin] have similar copy number of IFNa cDNA as consensually admitted by the community of P.pastoris experts [5, 18,19, 20]. The clones producing the highest protein levels were selected by the visual comparison of the band intensities of the culture supernatants subjected to either SDS-PAGE and Coomassie Brilliant Blue staining (Fig. 3A) or western blot analysis (Fig. 3B). Three rWT-huIFN[alpha]2[alpha] clones 7, 13 and 14 were able to grow on plates with 2,000 [micro]g/ml Zeocin. These clones were identified as the highest producing clones in comparison to 12 other clones that were resistant to a high concentration of Zeocin [clones 1, 2, 3, 4, 6, 7, 9, 10, 12, 13, 14, and 20]. One of the three clones producing the highest rFW-huIFN[alpha]2 protein levels, clone 7, was selected to be used as a wild-type huIFN[alpha]2 expression reference clone. Clones expressing the highest levels of human IFN[alpha]2a protein from the fully synthetic FS, SW and WS chimeras were selected in a similar fashion. For the clones producing high protein levels, expression levels correlated with Zeocin resistance.
[FIGURE 3 OMITTED]
To analyze the effect of gene dosage on the protein expression level of rhuIFN[alpha]2a, a total of nine recombinant P. pastoris clones were picked from plates containing 2,000-[micro]g/ml Zeocin. These clones included three fully synthetic FS-rhuIFN[alpha]2a clones ([10.sup.S], [11.sup.S], and [12.sup.S]), three SW rhuIFN[alpha]2a clones ([4.sup.SW], [11.sup.SW], and [18.sup.SW]) and three WS rhuIFN[alpha]2a clones ([8.sup.WS], [14.sup.WS], and [20.sup.WS]).
Clones [C8.sup.S] and [C15.sup.S] were only resistant to 100 [micro]g/ml Zeocin and showed the lowest expression levels. Clone [C8.sup.S], one of the clones with the lowest protein expression, was used as an internal ELISA control. A visual comparison of the band intensities was performed by a western blot analysis (Fig. 4) of the culture supernatants from FS-rhuIFN[alpha]2a (clones [10.sup.S] and [11.sup.S]), SW rhuIFNa2a (clones [11.sup.SW] and [18.sup.SW]) and WS rhuIFN[alpha]2a (clones [8.sup.WS] and [20.sup.WS]).
[FIGURE 4 OMITTED]
Quantitative analysis of the effects of gene modifications on rhuIFN[alpha]2 expression
We evaluated the effects of codon optimization on protein expression levels using the clones with the highest protein expression from each category (WT, FS, WS and SW). The levels of rhuIFN[alpha]2a protein expression were determined using a quantitative ELISA developed in-house (Fig. 5). HPLC-purified, soluble IFN[alpha]2a produced in our laboratory was used to generate a standard curve. In addition, two internal ELISA controls were used. Western blot analysis (Fig. 6A) was performed for rhuIFN[alpha]2a protein expressed from the wild-type cDNA (clone [C7.sup.W]) and the optimized cDNAs (FS, WS and SW) and analyzed by quantitative densitometry using ImageJ software (Fig. 6B).
[FIGURE 5 OMITTED]
[FIGURE 6 OMITTED]
Both the quantitative ELISA and ImageJ western blot analysis showed the same significant improvement in the rhuIFN[alpha]2 expression level after optimizing the cDNA sequence for the codon bias of P. pastoris. The yields from the WS and SW clones were 1.5 and 1.7 times higher, respectively, than the yield from the wild-type clone, and the FS clone gave a 3-fold higher yield (750 mg/L) than the wild-type clone.
Biological activity of soluble huIFNa2 expressed in a P. pastoris host
The biological activity of each purified recombinant huIFN[alpha]2a protein (rWThuIFN[alpha]2a and rSynhuIFN[alpha]2a) produced in P. pastoris was determined using antiviral and reporter gene assays . The two recombinant IFN preparations were calibrated against the WHO IFN[alpha] international standard (code: 95/566), .
The results of the reporter gene assay fit a sigmoidal dose-response curve with IFN concentration (the log of the reciprocal of the IFN dilution) plotted against absorbance. Using the linear portion of the curve, a parallel line displacement program determined the concentration of interferon in a sample by comparing the responses for the test and reference solutions using statistical methods for parallel line assays. Prism software was used to calculate and compare the standard error values of the IFN (95/966) standard preparation and recombinant huIFNalpha2 for each experiment. The results consistently showed that compared to the IFN WHO international standard (a specific activity of 1.4x[10.sup.8] IU/mg (Meager, 2002), the recombinant IFN[alpha]2a produced from the fully synthetic clone (FS) had an average specific activity of 4.48 x [10.sup.8] IU/mg (+/- 0.140), and the recombinant IFN produced from the native sequence (FW) had an average specific activity of 1.57x[10.sup.8] IU/mg (+/- 0.214). The specific activity of the rSynhuIFN[alpha]2a clone was 2.85 times higher than that of the rWThuIFN[alpha]2a clone and 3.20 times higher than that of the WHO IFN[alpha] international standard, (p [less than or equal to] 0.05).
Several strategies have been developed to increase the production of recombinant proteins in different host expression systems, particularly for P. pastoris and E. coli [5, 9, 19, and 20]. The concept that organisms display a non-random pattern of synonymous codon-usage has been confirmed by the explosion of sequence data available from recent genome sequencing projects. Indeed, all organisms investigated to date have shown a general bias toward a subset of the 61 possible sense codons. Although there are several hypotheses to explain the origin of this bias, a model involving a selection for translational efficiency has been well-supported in prokaryotes, unicellular eukaryotes, and, to a lesser extent, insects [21, 22, and 23]. The optimization of coding sequences toward the codon bias of the host has led to an increase in heterologous protein production in a variety of host cell types [24, 25, and 26]. However the effect of such codon optimization strategy on the biological activity of recombinant human IFNa, has not been investigated previously.
In our study, we used cDNA codon optimization to improve the yield of rhuIFN[alpha] in P. pastoris. We adjusted the codons in the human interferon a2a cDNA according to the codon bias of P. pastoris and examined the effects on the biological activity of the recombinant protein.
We designed and constructed four IFN[alpha]2a cDNA, one identical to the native sequence [wild-type], one fully synthetic sequence [FS] adjusted to the Pichia pastoris codon bias and two partially adjusted chimeric clones (SW and WS). In the SW clone, the synthetic covered half of the IFN coding sequence starting at the 5' end, whereas in the WS clone, the synthetic sequence covered the coding sequence starting at codon [TTC.sup.198] and extending to the stop codon. The recombinant IFN[alpha] produced from the different clones had the same molecular weight as that produced from the native human cDNA sequence and was specifically recognized by a monoclonal anti-human IFN[alpha] antibody by western blot analysis. The latter result indicates that the antigenic determinants of the protein were conserved.
The comparisons of the production yields showed that the yields of the partially modified IFN[alpha] clones (SW and WS) were 1.5 to 2 times higher than that of the wild-type clone. In addition, the fully synthetic clone gave a yield that was 3 times higher (750 mg/L) than that of the wild-type clone. Compared to data in the literature, a 3-fold improvement in expression may not appear noteworthy, probably due to the prior optimization of the sequence used to initiate translation and the relatively high amount (250 mg) of IFN[alpha]2a produced by the native cDNA in shake flasks. However, these results represent a significant improvement in expression, especially considering that further improvement may be achieved using optimization strategies at later stages. The biological activity of these rhuIFN2[alpha]a preparations was determined and compared to the WHO IFN[alpha] international standard. Surprisingly, the IFN[alpha] produced from the fully synthetic cDNA had consistently shown a higher specific activity (4.48 x [10.sup.8] IU/mg) than that of the IFN[alpha] produced from the native cDNA (1.570 x [10.sup.8] IU/mg). Thus, within the accuracy limits of the biological activity test, codon optimization improved not only the production yield but also the specific activity of IFN. Additional data regarding the efficiency of transcription and translation and their eventual effects on rIFN[alpha] folding are needed, however, to further examine this difference in specific activity. Both high structural similarity between a recombinant protein and its native counterpart and high biological activity are of paramount importance in the approval of biosimilar drugs (or follow-up proteins) by the pharmaceutical regulatory authorities.
Here, we demonstrated that using an interferon a cDNA sequence that fully complies with the P. pastoris codon-usage can improve the production yield and, more importantly, the biological activity of recombinant human interferon a. This biotechnology-based strategy can be applied in the design and manufacturing of other recombinant protein-based biopharmaceuticals to lower the production costs and thus the prices of this important class of new drugs.
List of abbreviations P. pastoris Pichia pastoris E. coli Escherichia coli IFN[alpha Interferon alpha huIFN[alpha]= human interferon alpha rhuIFN[alpha]= recombinant human interferon alpha YPD Yeast extract-peptone-dextrose medium BMGY minimal glycerol-complex medium BMMY buffered minimal methanol-complex medium rhuWIFN[alpha]2 recombinant human interferon alpha 2 wild-type cDNA sequence huSynIFN[alpha]2 human synthetic interferon alpha 2 SW 5' half synthetic-3' half wild-type human interferon alpha 2 cDNA sequence WS 5' half wild-type-3' half synthetic human interferon alpha 2 cDNA sequence FS fully synthetic human interferon alpha 2 cDNA sequence WT Wild-type human interferon alpha 2 cDNA sequence W Wild type S Synthetic Aox1 Alcohol oxidase 1 EMCV encephalomyocarditis virus ISRE IFN-inducible promoter sequence SEAP gene secreted alkaline phosphatase gene
We thank Dr Anthony Meager, Immunobiology Division, National Institute for Biological Standards and Control, Potters Bar, UK, for his help in analyzing the biological activity of rhIFN[alpha]2.
The detailed IFN production process, for the work reported in this paper is described in the International patent No: WO 2007/099462 A2.
 Schmidt, F.R., 2004, "Recombinant expression systems in the pharmaceutical industry," App, Microb. Biotech., 65, (4), pp.363-72.
 Graumann, K.P., and Premstaller, A., 2006, "Manufacturing of recombinant therapeutic proteins in microbiological system," Biotech J., 1, pp.164-86.
 Shah, P., Gilchrist, M.A. 2010, "Effect of correlated tRNA abundances on translation errors and evolution of codon usage bias," PLoS Genetics, 6 (9) pp. e1001128
 Shardiwal, R.K., Sohrab, S.S., 2010, "A more elaborative way to check codon quality: an open source program," Inter. J. of Bioinfor. Res. App., 6 (3), pp.223-229.
 Daly, R., Hearn, M.T.W., 2005, "Expression of heterologous protein in Pichia pastoris: a useful experimental tool in protein engineering and production," J. Mol. Recog., 18 (2), pp.119-138.
 Sreekrishna, K., "et al1", 1997, "Strategies for optimal synthesis and secretion of heterologous proteins in the methylotrophic yeast Pichia pastoris. Gene 190, pp.55-62.
 Benabdesselem, C., "et al", 2006, "Enhanced Patient Serum Immunoreactivity to Recombinant Mycobacterium tuberculosis CFP32 Produced in the Yeast Pichia pastoris Compared to Escherichia coli and Its Potential for Serodiagnosis of tuberculosis," J. Clin. Microb., 44, (9), pp.3086-3093.
 Benabdesselem, C., "et al", 2007, "High Level Expression of Recombinant Mycobacterium tuberculosis Culture Filtrate Protein CFP32 in Pichia pastoris. Mol. Biotech., 35, pp.41-49.
 Rabhi-Essafi I., "et al", 2007, "A strategy for high-level expression of soluble and functional human interferon alpha as a GST-fusion protein in E.coli," Prot. Eng. Des. Sel. 20(5), pp.201-9.
 Retnoningrum, D.S., "et al", 2010, "Construction of synthetic open reading frame encoding human interferon alpha 2b for high expression in Escherichia coli and characterization of its gene product," J. Biotech., 145(2), pp.193-8.
 Li, Z., "et al ", 2008, "Optimization of the expression of hepatitis B virus e gene in Pichia pastoris and immunological characterization of the product" J. Biotech,. 138 (1), pp.1-8.
 Sinclair, G., Choy, F.Y.M., 2002, "Synonymous codon usage bias and the expression of Human gluccocerebrosidase in the methylotophic yeast, Pichia pastoris," Prot Expr. Purif., 26, pp.96-105.
 Ko H.J., "et al", 2005, "Optimization of codon usage enhances the immunogenicity of a DNA vaccine encoding mycobacterial antigen Ag85B," Infect and Immun., 73(.9), pp.5666-5674.
 Komar, A.A., "et al", 1999, "Synonymous codon substitutions affect ribosome traffic and protein folding during in vitro translation. FEBS Lett., 462, pp.387-391.
 Kimchi-Sarfaty, C., "et al", 2007, "A "Silent" Polymorphism in the MDR 1 Gene Changes Substrate Specificity," Science, 315, pp.525-528.
 Sauna, Z.E., "et al", 2007, "Silent Polymorphisms Speak: How They Affect Pharmacogenomics and the Treatment of Cancer," Cancer. Res., 67(20), pp.9609-9612.
 Ausubel, F.M., "et al", 2002, "Current protocols in Molecular Biology," Wiley Interscience ISBN 978-0-471-50338-5.
 Meager, A., 2002, "Biological assays for interferons," J. Immun. Meth., 26, pp.121-136
 Higgins, D.R., "et al", 1998, "Small vectors for expression based on dominant drug resistance with direct multicopy selection," Meth. Mol. Bio., 103, pp.41-53.
 Cregg, J.M., "et al", 2000, "Recombinant protein expression in Pichia pastoris," Mol. Biotech., 16(1), pp.23-52.
 Sorensen, H.P., Mortensen, K.K., 2005, "Soluble expression of recombinant proteins in the cytoplasm of Escherichia coli," J. Biotech., 115, pp.113-128.
 Bartolome, C., Charlesworth, B., 2006," Genet., 174 (4), pp.2033-44.
 Vicario, S., "et al", 2008, "Molecular Biology Evolution. 25 (11), pp.2269-77.
 Supek, F., Muc, T., 2010, "On relevance of codon usage to expression of synthetic and natural genes in Escherichia coli.," Genet., 185 (3), pp.1129-34.
 Cid-Arregui, A., "et al", 2003. "A synthetic E7 gene of human papillomavirus type 16 that yields enhanced expression of the protein in mammalian cells and is useful for DNA immunization studies. J. Virol., 77 (8), pp.4928-37.
 Lee S.F., "et al", 2009, "," Microbio., 155, pp.3581-8.
 Rosano, G.L., Ceccarelli, E.A., 2009, "Rare codon content affects the solubility of recombinant proteins in a codon bias-adjusted Escherichia coli strain," Micro. Cell Fact. 8 (41), pp.86-94
Imen Rabhi (1) and Dahmani M. Fathallah (1,2) *
(1) Molecular Biotechnology Group, Institute Pasteur, Tunis, Tunisia
(2) Biotechnology Program, King Fahd Chair for Biotechnology, Arabian Gulf University, Manama, Bahrain
* Corresponding Author
E-mail: Dahmani M. Fathallah: email@example.com, firstname.lastname@example.org
E-mail: Imen Rabhi: email@example.com, firstname.lastname@example.org
Table 1: Primers used for the construction of optimized human interferon alpha cDNA sequences. Primer Sequence (5' end.......3' end) designation 1. FIFN[alpha]2b TGGAATTCTGTGATCTGCCTCAAACCCA 2. RIFN[alpha]2b ATTCTGCGGCCGCTCATTCCTTACTTCTTAAACTTTC 3. Fa1 TGGAATTCTGTGATTTGCCTCAAACCCACTCCTTGGGT TCCAGAAGAACCTTGATGTTGTTGGCACAGATGAGAA AAATCTCTTTGTTCTCC 4. Fb1 AAACTCCTCCTGTGGAAATCCAAAGTCATGTCTGTCCTTCAA GCAGGAGAACAAAGAGAT 5. Fc1 CCACAGGAGGAGTTTGGCAACCAGTTCCAAAAGG CTGAAACCATCCCTGTCTTGCATGAGATGATCCAGCAGATCTTC 6. FfgsI TGGAATTCTGTGATTTGCCTCA 7. RfgsI GAAGATCTGCTGGATCATCTC 8. Fa2 CAGATCTTCAATTTGTTCTCCTCCACAAAGGACTCTTCTG CTGCTTGGGATGAGACCTTGTTGGACAAATTCTACACTGA ATTGTACCAGCAGTTGAATGAC 9. Fb2 GTATTTTCTAACAGCCAAAATGGAGTCCTCCTTCATCAA TGGAGTCTCTGTAACACCAACACCCTGGATAACACAGGC TTCCAAGTCATTCAACTGCTG 10. Fc2 GCTGTTAGAAAATACTTCCAAAGAATCACTTTGTATTT GAAAGAGAAGAAATACTCCCCTTGTGCCTGGGAGGTTG TCAGAGCAGAAATCATGAGATCT 11. FfgsII TGAAGATCTTCAATTTGTTCTCCAC 12. RfgsI TGAAGATCTCATGATTTCTGCTCTGA