Increasing gene dosage greatly enhances recombinant expression of aquaporins in Pichia pastoris
- 9.1k Downloads
When performing functional and structural studies, large quantities of pure protein are desired. Most membrane proteins are however not abundantly expressed in their native tissues, which in general rules out purification from natural sources. Heterologous expression, especially of eukaryotic membrane proteins, has also proven to be challenging. The development of expression systems in insect cells and yeasts has resulted in an increase in successful overexpression of eukaryotic proteins. High yields of membrane protein from such hosts are however not guaranteed and several, to a large extent unexplored, factors may influence recombinant expression levels. In this report we have used four isoforms of aquaporins to systematically investigate parameters that may affect protein yield when overexpressing membrane proteins in the yeast Pichia pastoris.
By comparing clones carrying a single gene copy, we show a remarkable variation in recombinant protein expression between isoforms and that the poor expression observed for one of the isoforms could only in part be explained by reduced transcript levels. Furthermore, we show that heterologous expression levels of all four aquaporin isoforms strongly respond to an increase in recombinant gene dosage, independent of the amount of protein expressed from a single gene copy. We also demonstrate that the increased expression does not appear to compromise the protein folding and the membrane localisation.
We report a convenient and robust method based on qPCR to determine recombinant gene dosage. The method is generic for all constructs based on the pPICZ vectors and offers an inexpensive, quick and reliable means of characterising recombinant P. pastoris clones. By using this method we show that: (1) heterologous expression of all aquaporins investigated respond strongly to an increase in recombinant gene dosage (2) expression from a single recombinant gene copy varies in an isoform dependent manner (3) the poor expression observed for AtSIP1;1 is mainly caused by posttranscriptional limitations. The protein folding and membrane localisation seems to be unaffected by increased expression levels. Thus a screen for elevated gene dosage can routinely be performed for identification of P. pastoris clones with high expression levels of aquaporins and other classes of membrane proteins.
KeywordsPichia pastoris aquaporins Major Intrinsic Proteins qPCR
Most membrane proteins are not abundantly expressed in their native tissues making it hard to recover enough material for functional and structural studies. In addition several isoforms with similar biochemical properties are often co-expressed which further complicates the purification process. Hence, efficient heterologous systems for overexpression are important for biochemical characterisation of membrane proteins, which is also reflected by the increase in membrane protein structures derived from heterologously expressed material over the last decade [1, 2]. Escherichia coli is still the most common host for membrane protein overexpression but is often not efficient when expressing eukaryotic membrane proteins . Yeasts have emerged as alternative hosts for expression, having the benefits of being inexpensive and easy to manipulate in addition to having the ability to grow to high cell densities . Furthermore, they possess the eukaryotic machinery for post translational modifications  and the eukaryotic translocon complex  that might be required for synthesis of a correctly folded and fully functional membrane protein.
The use of the methylotrophic yeast Pichia pastoris for heterologous protein production was developed in the 1980s  and it has proven to be a successful host when overexpressing soluble proteins  as well as several different classes of membrane proteins, including a peptide transporter , ion channels [10, 11], ion pumps [12, 13], G-protein coupled receptors [3, 14, 15, 16], ABC transporters  and glucose transporters [18, 19]. Due to its ability to grow to high cell densities and to its exceptionally strong and inducible AOX1 promoter , P. pastoris has become a potent alternative to the more studied baker's yeast Saccharomyces cerevisiae as a eukaryotic tool for recombinant protein production.
A number of factors can potentially affect the outcome of heterologous protein expression in P. pastoris, including the properties of the nucleotide sequence (e.g. codon composition, AT-content and secondary structure), mode of sequence insertion into the genome, choice of host strain, mode of expression, methanol utilisation phenotype and cultivation conditions [8, 20]. It has however been stated that when expressing protein from a single recombinant gene copy, the transcript level is most likely to be limiting in the protein production. It is therefore suggested that insertion of multiple recombinant gene copies should be aimed for in order to obtain maximal protein yields . When using a P. pastoris expression vector carrying the Sh ble gene which encodes a protein conferring resistance towards the antibiotic zeocin, a selection on varying antibiotic concentrations can be applied, as described in the Invitrogen Easy Select™ Pichia expression kit manual . The level of antibiotic resistance will reflect the number of plasmids that have been incorporated into the P. pastoris genome, that is, the recombinant gene dosage. Numerous studies have shown that increased number of inserted sequences can indeed boost the heterologous expression of soluble proteins [23, 24, 25, 26, 27, 28, 29, 30, 31, 32]. Concerning membrane proteins, the correlation between increased recombinant gene dosage and protein expression is less well studied and the results obtained are contradictory. There is one report where increased gene dosage, as estimated by semi quantitative DNA dot blotting, rendered significantly increased recombinant expression of a Na+, K+ -ATPase . In two other expression studies on GPCR:s, limited  or no  effect of increasing recombinant gene dosage was seen, suggesting that other factors were limiting the steady state expression levels.
Another class of membrane proteins that, in addition to the ones previously mentioned, has been successfully overexpressed in P. pastoris is the Membrane Intrinsic Proteins (MIPs), also called aquaporins [34, 35, 36, 37, 38]. Members of this superfamily of integral membrane proteins are present in nearly all living organisms . Aquaporins are proteins with a number of functions, whereof the most significant one is the facilitation of water transport over the membranes of the cell and the maintenance of water homeostasis in the organism . Functional and structural studies have led to a broadening of the knowledge base around these proteins . However, since many questions regarding the aquaporin function and structure remain unanswered and more studies are needed to complete the picture, convenient methods for routinely obtaining high expression levels of recombinant aquaporins in P. pastoris are desirable. Moreover, since the aquaporins are integral, they can serve as model proteins for other membrane spanning proteins and methods developed based on recombinant aquaporin expression might hence be applicable also in a broader context.
In this study, we systematically investigate how recombinant gene dosage influences expression levels of both mammalian (HsAQP5 and HsAQP8) and plant (SoPIP1;2 and AtSIP1;1) aquaporins in P. pastoris. We show that increased copy number leads to elevated heterologous expression of all the four members of the aquaporin family studied. Interestingly this includes AtSIP1;1, for which expression appears to be restricted due to impaired transcription and translation. A screen for multiple copy integrants should hence be a part of the routine optimisation strategy when overexpressing aquaporins in this host. We also present a method for fast, cheap and reliable determination of integrated plasmid copy number using qPCR. This method makes it possible to easily characterise the transformants and to correlate expression levels with gene dosage which is valuable when trying to find the P. pastoris clones with optimal characteristics for heterologous protein expression.
Analysis of copy number by qPCR
To be able to correlate the recombinant protein expression level with the number of inserted gene copies a method based on qPCR was developed. The method takes advantage of the fact that the AOX1 TT region, that is part of the pPICZB plasmid, is incorporated in the P. pastoris genome together with the gene to be expressed. Quantification of the AOX1 TT region will hence give an indirect measurement of the recombinant gene dosage, independent of which protein is to be overexpressed.
Comparison of single copy aquaporin clones
Quantification of AtSIP1;1 mRNA and modification of the consensus start sequence
Since the AtSIP1;1 construct was cloned with restriction enzymes different from those used for the other three aquaporins, it differed from the consensus start sequence  in one nucleotide (CAAAATGTCT → GAAAATGTCT). To investigate whether this deviation was causing the low AtSIP1;1 expression levels, a new construct carrying the consensus start sequence was made and protein expression of a clone selected on 100 μg/ml was compared with that of the AtSIP1;1:100:5 clone. Expression increased slightly but was still poor as compared with that of the single copy HsAQP5 clone (HsAQP5:100:5) that was included as a reference (data not shown).
The methylotrophic yeast P. pastoris has, during the last decades, increased in popularity as a eukaryotic host for recombinant protein expression . Numerous reports have described successful overexpression of soluble proteins as well as of membrane proteins in this system, but sufficient protein yields have not always been obtained. Several methods have been described that could be used to improve the outcome of expression trials, among which an optimisation of recombinant gene dosage has proven to be one of the most potent ones . When it comes to the aquaporin family of membrane proteins, optimisation of the nucleotide sequence both regarding codon composition and AT content  and controlled growth in bioreactors  have been reported to improve the yields of heterologous protein. However, in no reports has a systematic examination of the effect of gene dosage on recombinant aquaporin expression in P. pastoris been done and results obtained for other membrane proteins [12, 15, 33] are based on less precise DNA quantification methods. Neither has transcript levels for different genes with the same promoter and copy number been compared. A better understanding of how these factors may affect the yield will facilitate the design of future membrane protein expression experiments.
As for many other proteins expressed in P. pastoris [23, 29, 30, 32], recombinant aquaporin expression was shown to be intimately connected to the level of resistance towards the antibiotic. In general clones selected at 500 or 1000 μg zeocin/mL are more likely to express high levels of recombinant protein as compared with the ones selected at 100 μg/mL. In this study we chose to perform a two step antibiotic selection in order to take full advantage of all the clones generated in each transformation. By doing a first selection at 100 μg zeocin/mL all recombinant clones were recovered and these were then subjected to higher zeocin concentrations. The benefit of this method is that the risk of losing potential high expressing clones by subjecting them to lethal levels of antibiotic is eliminated. The drawback is that several clones from the same origin may be recovered during the second selection event, as possibly seen for the SoPIP1;2 clones (discussed below).
The large scale membrane preparation from a clone expressing high levels of one of the aquaporins (HsAQP8) showed that the folding and the targeting to the membrane seem to be unaffected by the increased work load that is put on the protein production and the membrane insertion machinery. Comparison with a clone with low expression level showed that the majority of the protein ends up in the membrane fraction in both cases and, most importantly, that the membrane fraction from the highly expressing clone contains substantially larger amounts of recombinant protein (Figure 2). Some degradation seems to occur and this could probably be avoided by optimisation of the cell breakage procedure or by expression in a protease deficient strain . When combining the high amount of correctly folded and targeted protein in multi-copy transformants with the high cell densities obtained in bioreactor cultures, exceptional membrane protein yields could be achieved. This was seen for one multi-copy HsAQP5 clone with high levels of recombinant protein expression, where 145 mg of functional protein was purified from 1 L fermenter culture (unpublished data). Protein from this clone was also used for crystallisation and rendered well ordered crystals diffracting to 3.1 Å [34, 42], confirming the high quality of the expressed protein.
When copy number was determined it was confirmed that the variation in antibiotic tolerance was due to differences in recombinant gene dosage among the different clones. Assuming that the Sh ble gene is incorporated at the same ratio as the AOX1 TT sequence it generally seems like the presence of 1 copy of the Sh ble zeocin resistance gene is the minimum requirement for growth at 100 μg zeocin/mL, 4 copies at 500 μg/mL and 9 copies at 1000 μg/mL, but clones with as many as 17 gene copies were recovered from plates with the highest antibiotic concentration. The results obtained for the aquaporin isoforms included in this study suggest a trend between elevated gene dosage and increased recombinant expression, also for SoPIP1;2 where, even though only intermediate copy numbers were observed, expression increased by a factor of 70. This is in accordance with the results obtained when expressing a Na+K+-ATPase , but in contradiction with results from a GPCR expression trial . Thus it seems like the multi copy strategy may not be the universal solution to all problems encountered when trying to over express MP:s, since there may be several factors limiting the recombinant protein production. One such bottleneck might be a limitation of the membranes to accommodate the expressed protein, resulting in a maximal upper limit for protein expression when gene dosage is further increased. Indeed, in our study we observe a plateau in the average expression when increasing gene copy number as exemplified by HsAQP5. The average expression of the 500- and 1000-selected clones were determined to 0.69 and 0.58 arbitrary units, respectively when the corresponding average gene dosage increased from 6.4 and 10.5 respectively. Still, a 147 fold increase in HsAQP5 expression level was obtained utilising the multi copy strategy. This demonstrates the potential of this method.
It is remarkable that from the 100 μg zeocin/mL selection, at least 9 out of 20 clones (45%) proved to harbour more than one recombinant gene copy. This frequency of multiple gene insertion events is notably higher than the expected 1-10% suggested in the EasySelect™ Pichia Expression Kit manual (Invitrogen). Possibly this high multi-copy frequency could have arisen during the two step zeocin selection due to posttransformational vector amplification which has been reported to occur when subjecting recombinant P. pastoris clones to gradually increasing concentrations of antibiotic . In that report posttransformational vector amplification upon recurring antibiotic selection lead to elevated gene dosage in 40% of the clones studied. That frequency is comparable with the 45% seen in this study.
One problem that is commonly encountered when expressing protein in P. pastoris is the occurrence of false positives, that is, clones that have the capacity to grow at high antibiotic levels but express little or no recombinant protein . False positives were also observed in this study. For the clones carrying the SoPIP1;2 construct, five out of the ten clones selected at the three highest zeocin levels (500 μg/mL, 1000 μg/mL and 2000 μg/mL) showed little or no recombinant protein expression. qPCR results showed that two of these clones (SoPIP1;2:1000:2 and SoPIP1;2:1000:3) carried only one copy of the SoPIP1;2 gene and three of them (SoPIP1;2:500:1, SoPIP1;2:2000:1 and SoPIP1;2:2000:2) were assigned copy numbers close to four. Probably the three latter false positives have a common origin in a clone with faulty heterologous expression. Since all SoPIP1;2 clones selected at 1000 or 2000 μg zeocin/mL were shown to harbour unexpectedly low numbers of integrated plasmid, the antibiotic selection for recovering multiple insertions seems to have failed. The only factor that differed in this selection as compared with the ones done for the other three constructs was the incubation time, which was increased from three to four days. Possibly there was degradation of zeocin over time or this procedure favoured the selection of mutations that increased resistance in alternative ways, not coupled to gene copy number. Regardless of the underlying mechanism it is apparent that extended incubation should be avoided in order to exclude the formation of false positives.
The results from the SoPIP1;2 selection emphasise the importance of evaluating the outcome of the antibiotic selection not only at the protein level, but also at the DNA level, to ensure that the selection is efficient and that satisfactory gene dosage is obtained. If, as in the case with SoPIP1;2, gene dosage turns out to be moderate one should consider redoing the transformation and selection in order to obtain clones with maximal gene dosage since they may be superior protein producers. The qPCR based method presented here offers a convenient and reliable means of characterizing recombinant P. pastoris clones and could be included in the standard protocols when screening for multi-copy transformants. The big advantage of this new method as compared with other qPCR based methods used to verify recombinant gene dosage in P. pastoris [32, 44] is its versatility. Since the primers used are designed to amplify a part of the vector sequence rather than the specific gene, it can be used to quantify any recombinant gene sequence sitting in an expression vector carrying the AOX1 TT sequence.
One of the aquaporins, AtSIP1;1, proved to express very low levels of recombinant protein as compared with the other aquaporin isoforms. Analysis of the AtSIP1;1 mRNA levels showed that AtSIP1;1 transcripts were not as abundant as HsAQP5 transcripts. Due to the presence of an EcoRI cleavage site within the AtSIP1;1 coding sequence, the EcoRI restriction enzyme could not, as for the other three aquaporin isoforms, be used when creating the AtSIP1;1 construct. It is possible that the deviating site of insertion into the expression vector had an impact on the mRNA levels. However, AtSIP1;1 mRNA levels were still considerably higher than these of actin. Recombinant protein expression could hence be expected to be lower than that of HsAQP5 but should still be readily detectable with a western blot, unless there are obstructions in the translation that could contribute to the poor AtSIP1;1 expression. To exclude the possible adverse effects on translation of a deviation in the nucleotide sequence proximate to the start codon, a new construct carrying the correct P. pastoris consensus start sequence was made. This modification did not lead to significant improvement of the recombinant protein expression, and hence that factor was excluded as being a major obstacle to efficient recombinant AtSIP1;1 expression.
Pairwise identity and similarity between aquaporin isoforms
In this study we present a method based on qPCR to determine recombinant gene copy number, that facilitates the validation of the antibiotic selection and the characterisation of recombinant P. pastoris clones. Employing this method we have systematically investigated factors that could influence recombinant aquaporin expression in P. pastoris. We conclude that expression of correctly folded and targeted aquaporins is strongly dependent on recombinant gene dosage and also that increasing the gene dosage could, to some extent, help improving poor expression caused by inefficiency at the transcriptional and translational levels.
Selection of multi-copy P. pastoris transformants and expression analysis
The cDNA of the different aquaporin isoforms (HsAQP5, HsAQP8, SoPIP1;2 and AtSIP1;1) was amplified with the primers listed in Additional file 1, Table S1. The respective PCR products were cloned into the pPICZB vector (Invitrogen) between the EcoRI and NotI sites (HsAQP5, HsAQP8 and SoPIP1;2) or the XhoI and NotI sites (AtSIP1;1) conferring the addition of a myc antibody epitope and His6-tag to the C-terminus of the amino acid sequence. The resulting plasmids were sequenced in order to confirm the correct nucleotide sequence.
The four constructs were linearised by SacI and complete linearisation was confirmed by agarose gel electrophoresis. 10 μg of linearised plasmid was transformed into competent wild type X-33 P. pastoris cells by electroporation according to the EasySelect™ Pichia Expression Kit Manual (Invitrogen).
Transformants were initially selected on YPD (1% w/v yeast extract, 2% w/v peptone, 2% w/v dextrose) agar plates containing 100 μg/mL Zeocin. After 3 days, transformants were resuspended in YPD medium that was added directly to the plates. Cells from the same transformation event were pooled and plated onto YPD agar plates containing 100, 500, 1000 or 2000 μg/mL Zeocin to select for clones with higher resistance level. Colonies that appeared within 3 days after the second plating event were believed to be true positives and 5 representatives from each construct and zeocin level were further analysed. For the HsAQP5, HsAQP8 and AtSIP1;1-constructs, the selection at 2000 μg zeocin/mL rendered no colonies. For the SoPIP1;2 construct no colonies had appeared on the plates containing the two highest antibiotic concentrations after three days, hence these were left to incubate for another 24 h where after colonies appeared and 3 clones from the 1000 μg/ml and 2 clones from the 2000 μg/mL-plates were further analysed. The selected clones were assigned clone IDs describing the isoform, the antibiotic selection level in μg/mL (bold) and the number of the clone (e. g. HsAQP5:100:1).
Small scale expression screen
To analyse the protein expression levels of the selected clones, a small scale protein expression screen was performed. To generate biomass, transformants were grown in 5 mL BMGY (1% w/v yeast extract, 2% w/v peptone, 100 mM potassium phosphate pH 6.0, 1.34% w/v yeast nitrogen base, 4 × 10-5% w/v biotin, 1% v/v glycerol) over night. Cells were harvested and resuspended in 5 mL BMMY (1% w/v yeast extract, 2% w/v peptone, 100 mM potassium phosphate pH 6.0, 1.34% w/v yeast nitrogen base, 4 × 10-5% w/v biotin, 0.5% v/v methanol) to an optical density at 600 nm (OD600) = 1. Induction was maintained for 72 h by addition of methanol to a final concentration of 0.5% v/v every 24 h. 3 h after the last methanol addition 20 OD600 units of cells from each clone were harvested by centrifugation and resuspended in 100 μL breaking buffer (50 mM NaPO4 pH 7.4, 1 mM EDTA, 5% v/v glycerol, 1 mM PMSF). Glass beads were added and cells were broken by vortexing 8 × 30 s with 30 s intervening cooling sessions on ice. Glass beads, unbroken cells and cell debris was removed by centrifugation at 18 000 g for 5 min at 4°C. To the crude extract obtained 3.33 × SDS loading buffer (250 mM Tris-HCl pH 6.8, 40% v/v glycerol, 8% w/v SDS, 2.37 M β-mercaptoethanol, 0.1% w/v Bromophenol Blue) was added for denaturation of the sample (30 min at 20°C). Samples were loaded onto 4-12% gradient SDS gels (NuPAGE®Novex® Bis-Tris 4-12% Midi Gels, Invitrogen) for separation of proteins and subsequently transferred to polyvinylidene difluoride (PVDF) membranes (Millipore). Recombinant proteins were visualised by immunodetection (Primary Ab; mouse anti-(H)4, QIAGEN and secondary Ab; polyclonal goat antimouse IgG HRP conjugate, Dako). Densitometric analyses of the western blots, to quantify total protein and degree of degradation for each clone, was performed using the Image J software .
Large scale expression and membrane preparation
A large scale expression and subsequent membrane preparation was made in order to analyse whether overexpression affected membrane localisation of heterologous protein. A clone with low (HsAQP8:100:2) and one with high (HsAQP8:1000:1) expression level were compared. The expression was performed in baffled flasks according to the Invitrogen EasySelect™ manual. Briefly, 100 mL BMGY cultures were grown at 28°C and with shaking at 220 rpm over night to OD600 = 5. Cells were harvested by centrifugation, resuspended in 500 mL BMMY to OD600 = 1 and cultured for 75 h at 28°C with shaking at 200 rpm. Recombinant protein expression was maintained by addition of methanol to a final concentration of 0.5% v/v every 24 h. 3 h after the last methanol addition, cells were harvested by centrifugation and resuspended in 250 mL cold breaking buffer. Cells from 1 L culture (18.6-24.6 g) were broken using a Bead Beater (BioSpec Products) for 10 × 30 s with 30 s intervening cooling sessions. Intact cells and cell debris was removed by centrifugation at 1400 g for 2 × 30 min with an intervening resuspension in breaking buffer to ensure efficient recovery of membranes. To collect the total membrane fraction, the 1400 g supernatant was centrifuged at 150 000 g for 2 h. The membrane pellet was resuspended in Buffer A (20 mM HEPES-NaOH pH 7.8, 50 mM NaCl, 10% v/v glycerol, 2 mM β-mercaptoethanol). All fractions recovered from the membrane preparation procedure were analysed for protein content by immunobloting (as above) after dilution to a normalised volume (corresponding to 270 mL/fraction).
Determination of copy number by qPCR
DNA extraction from P. pastoris
Genomic DNA from the 60 selected clones (HsAQP5:100:1-HsAQP5:1000:5; HsAQP8:100:1-HsAQP8:1000:5; SoPIP1;2:100:1-SoPIP1;2:2000:2 and AtSIP1;1:100:1-AtSIP1;1:1000:5) and from untransformed X-33 cells was extracted using the Easy-DNA™ Kit (Invitrogen). RNA was removed by incubation with RNase A (Fermentas) at 20°C over night. The DNA was quantified by absorbance measurements at 260 nm.
Oligonucleotide primer design
qPCRs were performed using a Corbett Research Rotor Gene thermal cycler (Montreal Biotech). Following pre-incubation at 95°C for 10 min the thermal cycler was programmed to perform 30-40 cycles of: 15 s at 95°C; 30 s at 60°C; 30 s at 72°C. Triplicate or quadruplicate samples of each template were analysed. Individual reactions were carried out in 20 μl volumes containing 10 μL 2 × Maxima™ SYBR® Green qPCR Master Mix (Fermentas), 0.25 μM fw- and rev- primers and 10 ng template DNA.
qPCR data analysis
Analysis of AtSIP1;1 and HsAQP5 mRNA-levels by qPCR
RNA extraction and cDNA synthesis
Total RNA was extracted from induced single copy Pichia clones carrying either the AtSIP1;1-construct (AtSIP1;1:100:2) or the HsAQP5 construct (HsAQP5:100:5) using the PureLink™ RNA Mini Kit (Invitrogen). As a control, RNA was also prepared from uninduced AtSIP1;1:100:2 and HsAQP5:100:5 cultures. Integrity of the RNA was analysed on a 2100 Bioanalyzer (Agilent Technologies). Samples corresponding to 1.6 μg total RNA were subjected to RNase free DNase (Promega) and cDNA was produced using the RevertAid™ H Minus First Strand cDNA Synthesis Kit (Fermentas) with the oligo dT-primers supplied in the kit.
Oligonucleotide primer design
Oligonucleotide primers (Additional file 2, Table S2) were designed using the PrimerPy software (http://code.google.com/p/oligobench/wiki/PrimerPy). Gene specific primers (HsAQP5_fw/HsAQP5_rev and AtSIP1;1_fw/AtSIP1;1_rev) were constructed to amplify the HsAQP5 and AtSIP1;1 cDNA respectively and for internal standardisation primers directed towards the P. pastoris actin1 cDNA (PpAct_fw/PpAct_rev) were designed.
qPCR was carried out as described above. 1 μL of the 20 μL cDNA reaction was used as template and samples were run as quadruplicates.
qPCR data analysis
This work was supported by grants from the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (Formas), the Swedish Research Council (VR) and the Research School of Pharmaceutical Sciences (FLÄK). The pLS15 plasmid was a kind gift from Dr. Sofia Möller.
- 12.Strugatsky D, Gottschalk KE, Goldshleger R, Bibi E, Karlish SJ: Expression of Na+,K+-ATPase in Pichia pastoris: analysis of wild type and D369N mutant proteins by Fe2+-catalyzed oxidative cleavage and molecular modeling. J Biol Chem. 2003, 278 (46): 46064-46073. 10.1074/jbc.M308303200.CrossRefGoogle Scholar
- 15.Weiss HM, Haase W, Michel H, Reilander H: Comparative biochemical and pharmacological characterization of the mouse 5HT5A 5-hydroxytryptamine receptor and the human beta2-adrenergic receptor produced in the methylotrophic yeast Pichia pastoris. Biochem J. 1998, 330 (Pt 3): 1137-1147.CrossRefGoogle Scholar
- 19.Tyagi NK, Goyal P, Kumar A, Pandey D, Siess W, Kinne RK: High-yield functional expression of human sodium/d-glucose cotransporter1 in Pichia pastoris and characterization of ligand-induced conformational changes as studied by tryptophan fluorescence. Biochemistry. 2005, 44 (47): 15514-15524. 10.1021/bi051377q.CrossRefGoogle Scholar
- 20.Sreekrishna K, Brankamp RG, Kropp KE, Blankenship DT, Tsay JT, Smith PL, Wierschke JD, Subramaniam A, Birkenberger LA: Strategies for optimal synthesis and secretion of heterologous proteins in the methylotrophic yeast Pichia pastoris. Gene. 1997, 190 (1): 55-62. 10.1016/S0378-1119(96)00672-5.CrossRefGoogle Scholar
- 21.Romanos M, Scorer C, Sreekrishna K, Clare J: The generation of multicopy recombinant strains. Methods Mol Biol. 1998, 103: 55-72.Google Scholar
- 22.Invitrogen: EasySelect Pichia Expression Kit,version G. 2005, [http://tools.invitrogen.com/content/sfs/manuals/easyselect_man.pdf]Google Scholar
- 26.Brierley RA: Secretion of recombinant human insulin-like growth factor I (IGF-I). Methods Mol Biol. 1998, 103: 149-177.Google Scholar
- 31.Clare JJ, Romanos MA, Rayment FB, Rowedder JE, Smith MA, Payne MM, Sreekrishna K, Henwood CA: Production of mouse epidermal growth factor in yeast: high-level secretion using Pichia pastoris strains containing multiple gene copies. Gene. 1991, 105 (2): 205-212. 10.1016/0378-1119(91)90152-2.CrossRefGoogle Scholar
- 42.Backmark A, Nyblom M, Tornroth-Horsefield S, Kosinska-Eriksson U, Norden K, Fellert M, Kjellbom P, Johanson U, Hedfalk K, Lindkvist-Petersson K, et al: Affinity tags can reduce merohedral twinning of membrane protein crystals. Acta Crystallogr D Biol Crystallogr. 2008, 64 (Pt 11): 1183-1186.CrossRefGoogle Scholar
- 47.Abramoff MD, Magelhaes PJ, Ram SJ: Image Processing with ImageJ. Biophotonics International. 2004, 11 (7): 36-42.Google Scholar
- 48.Rozen S, Skaletsky H: Primer3 on the www for general users and for biologist programmers. Methods Mol Biol. 2000, 132: 365-386.Google Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.