Strains, plasmids, and reagents

Escherichia coli TOP10 strain, Pichia pastoris X-33 strain, pPICZαA secretory expression vector, yeast nitrogen base (YNB), D-sorbitol, D-biotin and BCA protein concentration assay kit were purchased from Invitrogen (CA, USA). Tryptone and yeast extract were purchased from Oxoid (Hampshire, England). Polyethylene glycol (PEG) 3350 and lithium chloride (LiCl) were purchased from Sigma-Aldrich (MO, USA). Sonicated single stranded salmon sperm DNA was purchased from Genmed (MA, USA). Zeocin was purchased from Invivogen (CA, USA). Steady-Glo® Luciferase Assay System was purchased from Promega (WI, USA). Plasmid miniprep kit, DNA recovery/purification kit, pMD20-T cloning vectors, restriction endonucleases, DNA polymerases for PCR reaction and T4 DNA ligase were purchased from Takara (Guangzhou, China). Primers synthesis and DNA sequencing service were provided by Invitrogen (Shanghai, China). Escherichia coli TOP10 single colonies were selected on normal (for pMD20-T cloning) or low salt (for pPICZαA cloning) LB agar plates (1% tryptone, 0.5% yeast extract, 1% (normal) or 0.5% (low salt) NaCl and 1.5% agar) with corresponding antibiotics. Newly generated Pichia pastoris transformants were initially selected on YPD plates (2% peptone, 1% yeast extract, 2% dextrose, 2% agar) with 100 µg/ml zeocin, then on YPD plates with increasing doses of zeocin (from 200, 500 to 1000 µg/ml) to determine the copy number of integrants. For methanol induced expression, the P. pastoris recombinants were first grown in BMGY (1% yeast extract, 2% peptone, 100 mM potassium phosphate (pH 6.0), 1.34% YNB, 4×10^-5% biotin, 1% glycerol) medium to reach higher biomass and then induced in BMMY medium which contains the same ingredients as BMGY except the replacement of glycerol with methanol, as detailed in EasySelect^TM Pichia expression kit user manual.

Vectors construction

Venus coding DNA sequence (CDS), followed by kanamycin resistance gene, was amplified with primers (forward primers which covered the P1’ site were listed on Table 1 with the reverse primer of 5’-GGCTAGCGGCCGCAGACATGATAAGATACATTGATGAG-3’ which complements the 3’ terminus of kanamycin resistance gene) to place all 20 amino acids (codons were chosen according to the yeast preference [41]) at P1’ site to form a vector library. The 20 PCR amplification products were sub-cloned into pMD20-T plasmid and sequenced subsequently to ensure the sequences were correct. Then these plasmids were digested with SalI and NotI, the PCR product fragments were recovered from the agarose gel and inserted between the XhoI and NotI sites of pPICZαA, resulting in the Venus P1’ vector library (Figure 1B).

For construction of luciferase version of the library (Figure 1B), the counterpart vectors were digested with EcoRI and NotI, the plasmid backbone fragments were recovered. Luciferase CDS was amplified with 5’-GGAATTCCTCGAGATGGAAGACGCCAAAAACATAA-3’ as the forward primer and 5’-GGCGGCCGCTAGCACGGCGATCTTTCCGCCCTTC-3’ as the reverse primer and sub-cloned into pMD20-T plasmid and sequenced. Then the inserted plasmid was digested with EcoRI and NotI and the inserted fragments were recovered and ligated with vector backbone fragments originated from the Venus vector library.

For construction of other recombinant mammalian proteins version of the libraries (Figure 1B), the counterpart vectors were constructed just the same as the luciferase library, with respective gene-specific primers.

For plasmids to introduce additional Kex2 copies into yeast genome (Figure 1B), we took advantages of the zeocin resistance gene cassette from pPICZαA and chose S-Venus and G-Venus to demonstrate the utility of such approach. First, the zeocin resistance gene cassette on the pPICZαA backbone was replaced with a pair of SfiI sites through mutation PCR with 5’-GGCCATTACGGCCAAGCTTGGCCAGGGCGGCCCACGTCCGACGGCGGCCCACGG-3’ as the forward primer and 5’-GGCCGCCCTGGCCAAGCTTGGCCGTAATGGCCGGTTTAGTTCCTCACCTTGTCG-3’ as the reverse primer. Second, the P. pastoris Kex2 CDS was cloned from yeast genome with 5’-CGGATCCACCATGTATTTGCCAGCACTTCGCTTAGC-3’ as the forward primer and 5’-GCTCGAGCAATGCCGCACGTTTGGGATGTTCATTAG-3’ as the reverse primer, sequenced and inserted into a plasmid to fuse the flag tag to C-terminus of the CDS and SfiI sites at both ends of this cassette. After elimination the endogenous SacI site with synonymous point mutation, Kex2 CDS was then subcloned into the Zeo(R) deleted S-Venus and G-Venus between SfiI sites. Finally, Zeo(R) was reintroduced into these plasmids at BamHI site to the 3’ end of AOX1 transcription terminator.

Pichia pastoris transformation and selection of transformants

The yeast expression library vectors were linearized by SacI digestion and transformed into P. pastoris X-33 with the lithium chloride transformation method described in EasySelect^TM Pichia expression kit user manual (Invitrogen). Transformants were initially grown on YPD plates supplemented with 100 µg/ml zeocin. After integration of the plasmid into the yeast genome was confirmed by colony PCR, resulting colonies were transferred to YPD plates with 200, 500 and 1000 µg/ml zeocin for determination of the copy number of integrants. The subsequent comparisons of secreted proteins were only made between transformants with approximately the same copy numbers as determined by the same concentration range of drug resistance against zeocin.

Pichia pastoris cultivation and methanol induced expression

The experimental protocol from EasySelect^TM Pichia expression kit user manual to express recombinant Pichia pastorish has been followed. Briefly, the selected colonies were initially cultivated (shaking vigorously at 250 rpm) in BMGY medium at 28-30°C until the value of OD₆₀₀ reached 2 approximately. After centrifugation and removal of BMGY, cell pellets were re-suspended in BMMY to an OD₆₀₀ of 1 to induce expression. The volume of the culture should be no more than 10-30% of the total well/tube/flask volume to ensure sufficient aeration. Methanol was added to a final concentration of 1% every 24 hours to maintain induction. Yeast culture media were sampled and assayed every 24 hours. Small-scale cultivation and expression using 96-deep-well plates (Bel-Art Scienceware, NJ, USA) were carried out as previously described [42,43] whenever high-throughput screening of the secretory productivity was needed.

Fluorescence and luminescence assays

Before these assays, 100 µl yeast culture medium from each sample was measured for OD₆₀₀ and the values were used to normalize fluorescence or luminescence data. For Venus fluorescence assay, the cells were sprung down and 80 µl supernatant was added to each well of 96-well solid bottom black plates (CulturPlate^TM-96, PerkinElmer, MA, USA) and the signal measured under an excitation spectrum of 515 nm and an emission spectrum of 528 nm. For luminescence assay, 50 µl supernatant was added with equal volume of Steady-Glo® reagent in each well of 96-well solid bottom white plates (CulturPlate^TM-96, PerkinElmer, MA, USA) and the signal measured on a Veritas ^TMmicroplateluminometer from Turner Biosystems (CA, USA).

Western blotting analysis

The yeast cell lysis and cellular total protein extraction for SDS-PAGE and western blotting were carried out as described previously [44]. The yeast culture supernatants were also collected, TCA precipitated and sampled when needed. The samples were subjected to SDS-PAGE after protein concentration determination with the BCA Protein Quantitative Analysis Kit (Shenergy Biocolor, Shanghai, China). The resulting SDS-PAGE gels were subsequently subjected to either Coomassie staining (R-250) or western blotting. For western blotting, after PVDF (Millipore) electro-transfer, the membranes were blocked, incubated with antibodies, washed and developed on Fuji medical X-ray film (Fujifilm, Tokyo, Japan) for photographing and analysis. The gray-scale intensity values were calculated by ImageJ.

Large-scale fermentation and purification

Scale-up expression was carried out in a 2 L baffled flask. P. pastoris strain harboring the most productive P1’-stem cell factor (V-SCF, Table 2) was cultivated in 500 ml BMGY at 28-30°C with constant vigorous shaking till the value of OD₆₀₀ reached 8.0-12.0. Cells were pelleted and re-suspended in 250 ml BMMY, then cultured for 96 hours. Methanol was added in the medium to a final concentration of 1% every 24 hours to maintain induction. Yeast culture was centrifuged; the supernatant was collected, sampled and assayed every 24 hours. The protein concentration of the supernatant was determined by Bradford assays; the supernatant collected was dialyzed against 2 L extract buffer (50 mM Tris-HCl, 200 mM NaCl, 20 mM imidazole, pH 8.0) at 4°C overnight, filtrated with 0.22 µm filter, loaded onto a 5 ml HisTrap^TM HP column (GE Healthcare, Piscataway, NJ) and washed with extract buffer. The protein was eluted with a gradient of 20-500 mM imidazole. The purified protein was sent for MS analysis for identification.

Statistical analysis

Each fluorescent/luminescent value with an error bar was presented as mean±SEM representing the average of 5-6 individual colonies for the same vector with the same antibiotic resistance range for zeocin. P values were calculated with t-test (Student’s t-test); the actual P values of all statistical significances were given on the corresponding figure legends.

Vector Construction and Yeast Transformation

The 20 yeast recombinant Venus/luciferase expression vectors were generated by PCR amplification and TA-cloning of Venus-Kana(R) fragments with all 20 AAs at P1’ site into the original yeast secretory expression vector pPICZαA (Figure 1B). Both the Venus coding sequence (CDS) and P1’ site were confirmed by DNA sequencing (data not shown). The engineered pPICZαA vectors were transformed into the Pichia pastoris strain X-33. The resultant yeast colonies grown on YPD plates containing antibiotic zeocin (100 µg/ml) were subjected to colony PCR to confirm transformation (data not shown), and transferred to YPD plates containing 200, 500 and 1,000 µg/ml zeocin to empirically estimate the copy number of the integrants. On average, 20-30 colonies derived from each vector were tested for their integrated copy number, from which 5-6 colonies for the given vector, with zeocin resistance ranging 200-500 and 500-1,000 µg/ml, respectively, were selected for subsequent analysis of their levels of secretion.

Cultivation and Fluorescence/luminescence Assays

Cultivations were sampled every 24 hours till 120 hours post methanol induction. The supernatants of the yeast cultures were assayed for fluorescence (Venus) or luminescence (luciferase). As shown in Figure 2, different P1’ amino acids rendered considerable differences on the recombinant protein levels in supernatants. In particular, the highest levels in Venus library were achieved by S-Venus (serine at the P1’ site), which were nearly 13 folds higher than that of Y-Venus, the lowest ones (Figure 2A, B). In the luciferase library, the highest N-luciferase produced approximately four folds as much compared to the lowest C-luciferase (Figure 2C, D). Our western result (Figure 3) of luciferase library supported the luminescent measurements since the A, D, K, N, S-luciferase appeared to possess more secretory productivity over the others while the C, L, Q, W-luciferase did not produce much at all. This result indicated that for different proteins different P1’ site is optimal.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 2. Fluorescence/luminescence measurements of Venus/luciferase libraries with variable P1’ site.

Relative fluorescence/luminescence units (RFU/RLU) were normalized to the OD₆₀₀ of the corresponding cultures and then analyzed and compared with the measurements of recombinant yeast strains with the same range of zeocin resistance. pPICZαA: empty vector control; single letters: the single-letter codes of the corresponding P1’ residues. (A) Venus library with zeocin resistance ranging 500-1,000 µg/ml and (B) 200-500 µg/ml; (C) luciferase library with zeocin resistance ranging 500-1,000 µg/ml and (D) 200-500 µg/ml. Each value representing the average of 5-6 individual colonies for each vector with the same resistance range for zeocin was presented as mean±SEM.

https://doi.org/10.1371/journal.pone.0075347.g002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 3. Luciferase protein determination with western blotting analysis of luciferase library.

(A) A typical western result of TCA precipitated yeast culture supernatants of luciferase library with variable P1’ site, all tested recombinant yeast strains were randomly picked with zeocin resistance ranging 500-1,000 µg/ml, 5 µg of supernatant proteins were loaded for the SDS-PAGE and (B) the western gray-scale intensity of luciferase library. pPICZαA: empty vector control; single letters: the single-letter codes of the corresponding P1’ residues.

https://doi.org/10.1371/journal.pone.0075347.g003

Optimization of Expression for Mammalian Proteins

To achieve the objective of increasing the yields of mammalian proteins with medical relevance, several such proteins were expressed using this methodology. The results showed that the yields of all the recombinant proteins tested could be enhanced by P1’ site replacement (Table 2). The most productive stem cell factor (V-SCF) shown in Table 2 enable us to express and obtain substantial amount of recombinant SCF through a scale-up yeast fermentation and protein purification (Figure 4). The identity of the harvested recombinant protein was confirmed to be SCF by Mass Spectrometry analysis (Figure 4), and its SCF bioactivity was confirmed by tests on hematopoietic stem cells (data not shown). SCF is a well-known cytokine that plays an important role in hematopoiesis, spermatogenesis and melanogenesis [45-53], which may be used along with other cytokines to culture hematopoietic stem cells and hematopoietic progenitors [45,51]. The cultivation of such stem cells would provide sufficient hematopoietic progenitor cells in clinical bone marrow transplantation in the treatment of leukemia and other diseases. Our example (Figure 4) demonstrated that protein of great interests could be efficiently produced in large amounts with the help of our system, especially those with less secretory productivity by traditional means. As another proof of principle, we have designed a degenerative library at the P1’ site for secretory expression of previously poorly expressed human tissue plasminogen activator (tPA) and interleukin 4 (data not shown).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 4. Scale-up expression, purification and mass spectrometric analysis of SCF (about 20 KDa).

(A) Time course for the induction of SCF in 12% SDS-PAGE of TCA precipitated yeast culture supernatant, stained with Coomassie R-250; (B) the corresponding time course showing western blotting result; (C) Ni²⁺-HisTrap^TM elution profile of recombinant SCF, the bound protein was eluted with a gradient of 20-500 mM imidazole; and (D) Coomassie-stained 12% SDS-PAGE of collected fractions 1-9, 4.8 mg purified SCF in 6 ml was obtained from 400 ml yeast culture supernatant; (E) MALDI TOF/TOF^TM MS analysis of the purified recombinant SCF.

https://doi.org/10.1371/journal.pone.0075347.g004

Genomic Integration of Additional Kex2 Copies

Finally, we investigated whether additional copies of Kex2 in P. pastoris hosts could further increase the secretory productivity. To this end, the Venus expression constructs with/without additional Kex2 copies were chosen, cultivated, sampled and expression levels determined. Western blotting analysis showed that introduction of additional Kex2 copies greatly increased the Kex2 expression in the yeast host cells (Figure 5A). The fluorescence intensity in yeast culture media was significantly elevated upon addition of more Kex2 copies in that G-Venus-kex2 nearly doubling the productivity (Figure 5B). Taken together, our results clearly demonstrated that the feasibility of achieving high levels of recombinant secretory proteins in P. pastoris by optimizing P1’ site and increased Kex2 copies.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 5. Integration of additional Kex2 copies and the resulting influence on Venus fluorescence measurements: (A) western blotting (protein loads: 30 μg/well) showing the expression of integrated additional Kex2 copies (with flag tag), P values determined by t-tests of and S, G (Zeo 200-500 µg/ml) are *: 0.0022 and **0.0015 respectively; and (B) comparisons of Venus fluorescence measurements (normalized to the OD₆₀₀ of the corresponding cultivation) between recombinant strains with zeocin resistance ranging from 200-500 µg/ml.

Each value representing the average of 5-6 individual colonies for each vector with the same resistance range for zeocin was presented as mean±SEM.

https://doi.org/10.1371/journal.pone.0075347.g005