Strains and cultivation
Escherichia coli DH10B (Invitrogen) was used for subcloning. It was routinely cultivated in petri dishes or shake flasks using LB media supplemented with 25 μg mL-1 Zeocin. A wild type Pichia pastoris strain CBS2612 which can grow on minimal media supplemented with biotin, was used for protein expression in this work.
The main culture for screenings was either done with YP or BM media and glucose feed beads (12 mm, Kuhner, CH) which provided the carbon source.
YP media contained 20 g L-1 peptone and 10 g L-1 yeast extract, which can be supplemented with 12.6 g glycerol or 20 g glucose to obtain YPG and YPD, respectively. For cultivation on plates, 5 g L-1 agar-agar was added to the liquid medium. BM media was based on YP, supplemented with 13.4 g L-1 yeast nitrogen base (Cat.No. 291940, Becton Dickinson, FR) with ammonium sulfate, 0.4 mg L-1 biotin and 100 mM potassium phosphate buffer pH 6.0.
Identification of novel inducible promoters
a) Bioreactor cultivations
Fermentations for the identification of promoter candidates were done in 3.5 L working volume bioreactors (Minifors, Infors, CH) in three biological replicates. Cells were grown for about 24 h in batch on glycerol medium, followed by an exponential feed phase on glycerol fed batch medium calculated as described by Resina et al.
 with a specific growth rate of μ= 0.1 h-1 and a substrate yield coefficient of YX/S of 0.5 g g-1. Sequentially, chemostat cultivation (D = μ = 0.1 h-1) with high density glucose medium was performed.
Glycerol batch medium contained per liter: 2 g citric acid monohydrate, 39.2 g glycerol, 20.8 g NH4H2PO4, 0.5 g MgSO4∙ 7H2O, 1.6 g KCl, 0.022 g CaCl2∙ 2H2O, 0.8 mg biotin and 4.6 mL PTM1 trace salts stock solution. HCl was added to set the pH to 5.0.
Glycerol fed-batch medium contained per liter: 632 g glycerol, 8 g MgSO4∙ 7H2O, 22 g KCl, and 0.058 g CaCl2∙ 2H2O.
High-density chemostat medium contained per liter: 2 g citric acid monohydrate, 99.42 g glucose monohydrate, 22 g NH4H2PO4, 1.3 g MgSO4∙ 7H2O, 3.4 g KCl, 0.02 g CaCl2∙ 2H2O, 0.4 mg biotin and 3.2 mL PTM1 trace salts stock solution. HCl was added to set the pH to 5.0.
PTM1 trace salts stock solution contained per liter: 6.0 g CuSO4∙ 5H2O, 0.08 g NaI, 3.36 g MnSO4∙ H2O, 0.2 g Na2MoO4∙ 2H2O, 0.02 g H3BO3, 0.82 g CoCl2, 20.0 g ZnCl2, 65.0 g FeSO4∙ 7H2O, 0.2 g biotin and 5.0 mL H2SO4 (95%-98%).
b) Microarray hybridization
RNA purification and sample preparation as well as microarray hybridization (in-house designed P. pastoris specific oligonucleotide arrays, AMAD-ID: 034821, 8x15K custom arrays, Agilent) and data analysis were done as described by Graf et al.
Characterization of promoter strength and regulation
Cloning and transformation was done using the in-house vector pPuzzle
, which contains a Zeocin resistance cassette for selection in both E. coli and yeast, an expression cassette for the gene of interest (GOI) consisting of a multiple cloning site and the S. cerevisiae CYC1 transcription terminator, and a locus for integration into the P. pastoris genome (3´ AOX1 region or rDNA locus). Promoter sequences (up to 1000 bps upstream of the start codon of their respective genes) were PCR-amplified from P. pastoris genomic DNA (primer sequences see Additional file
1: Table S1). The promoters were ligated into pPuzzle in front of the start codons of the model proteins, using the ApaI and the SbfI restriction sites of the multiple cloning site of the vector. Vectors expressing the respective model protein under control of PGAP were used as controls throughout the study. For the expression of heterodimeric HyHEL antibody Fab fragment (HyHEL Fab), the expression cassettes of light chain and Fab heavy chain (each under control of PG1) were combined into one vector (using the strategy described in
HSA was secreted by its native secretion leader, while for CpB and HyHEL Fab the S. cerevisiae alpha mating factor signal sequence was used. To avoid positional effects on reporter gene expression levels, genome integration of the expression plasmids was targeted to either the 3´flanking region of the AOX1 gene or the ribosomal DNA locus (rDNA, for multicopy integration) of P. pastoris, respectively.
Plasmids were linearized within the genome integration region prior to electroporation (2 kV, 4 ms, GenePulser, BioRad) into electrocompetent P. pastoris. Multicopy integration of HSA expressing clones was done as described by Marx et al.
 and selected at higher Zeocin concentrations (up to 1000 μg mL-1).
P. pastoris cells were first selected and cultivated in petri dishes on YPD agar and then inoculated in an YPG medium as pre-culture for screenings and fermentations. Antibiotic selection by Zeocin was applied on plates and in pre-culture at a concentration of 25 μg mL-1 or higher.
b) Expression screening
Expression of intracellular eGFP and the secreted proteins HSA, CpB and HyHEL Fab with the novel promoters in comparison to PGAP was evaluated in shake flask screenings. All screenings were performed at 25°C and with shaking at 180 rpm. Single colonies were inoculated in YPG medium with selection pressure (Zeocin) for pre-culture. After approximately 24 hours, the pre-culture was used to inoculate the main culture with an optical density (OD600) of 0.1 (for eGFP) or 1 (for HSA, CpB and HyHEL Fab) in 10 mL YP or BM medium, respectively. Glucose feed beads (12 mm, Kuhner, CH) were used to generate glucose-limiting growth conditions. Expression of eGFP was measured at the end of pre-culture and at 24 and 48 hours of the main culture. Culture supernatant of clones expressing secreted protein was harvested from the pre-culture and after 48 hours and cell density was determined by measuring wet cell weight or OD600.
c) Comparative analysis of P. pastoris promoter activity
In order to analyze relative transcription strength of the PG promoters at different glucose concentrations, a comparative promoter activity study using various glucose concentrations (ranging from 20 to 0.002 g L-1 glucose) was performed with eGFP expressing clones in 24-well plates (Cat. No. 7701–5110, Whatman, UK) covered with breath seal membranes (Cat. No. B-100, Excel Scientific, CA). Glucose concentrations of 20, 10, 5, 2.5, 1.25, 0.63, 0.31, 0.16, 0.08, 0.04, 0.02, 0.01, 0.005, 0.002 g L-1 were obtained by serial dilution in YP media, and represent the inital setpoints. The main culture was inoculated from YPG-Zeocin pre-culture with an OD600 of 0.01 and samples were taken after 5–6 hours and analyzed by flow cytometry.
d) Fed batch cultivation
All fed batch fermentations were done in 1.0 L working volume bioreactors (SR0700ODLS, DASGIP, DE). The dissolved oxygen was controlled at DO = 20% with the stirrer speed (400 – 1200 rpm). Aeration rate was 18 L h-1 air, the temperature was controlled at 25°C and the pH was controlled at 5.85 for HSA
 or pH 5.0 for the other proteins
 with addition of ammonium hydroxide (25%). To start the fermentation, 300 mL batch medium was sterile filtered into the fermenter and a P. pastoris clone was inoculated from an overnight pre-culture with a starting optical density (OD600) of 1. For the cultivation of clones expressing eGFP, the batch phase of approximately 25 h was followed by a fed batch phase with a feeding rate optimized according to
. HSA expressing strains were cultivated as described by Marx et al.
, where the batch phase was followed by a constant feed of 2 g h-1 fed batch medium for 100 h, Carboxypeptidase B and HyHEL Fab expressing clones were cultivated similarly. Samples were taken during batch and fed batch phase, and analyzed for expression.
Glycerol batch and glucose fed batch media for eGFP, HyHEL Fab and Carboxypeptidase B expressing clones were exactly as described in
, while for the production of HSA the media was described in
e) Chemostat cultivation
A strain expressing HSA (2 GCN) under control of PG1 was tested for its growth rate dependent expression behaviour in chemostat at different dilution rates (D = 0.03, 0.06, 0.10, 0.13)
For characterization of glucose uptake characteristics, the P. pastoris wild type strain and the strain deleted for PAS_chr1-3_0011 were cultivated in glucose limited chemostats at D = 0.05, 0.10, and 0.14 h-1. Samples were taken rapidly as described below.
a) Copy number determination with real-time PCR
Genomic DNA was isolated using the DNeasy Blood&Tissue Kit (Cat. No. 69504, Quiagen, DE). Gene copy numbers were determined with quantitative PCR using the SensiMix SYBR Kit (QT605-05, Bioline reagents, UK). The primers (supplementary Additional file 1: Table S1) and sample were mixed with the SensiMix and applied for real time analysis in a real-time PCR cycler (Rotor Gene, Qiagen, DE). All samples were analyzed in tri- or quadruplicates. Data analysis was performed with the two standard curve method of the Rotor Gene software. The actin gene ACT1 was used as calibrator.
b) Determination of protein expression levels
A plate reader (Infinite 200, Tecan, CH) was used to determine eGFP fluorescence in fermentation samples. Therefore, samples were diluted to an OD600 of 5 and fluorescence intensity was then related to the culture volume.
Expression of eGFP in screenings was analyzed by flow cytometry as described before
. Specific eGFP fluorescence referred to in this study is the fluorescence intensity related to the cell volume for each data point as described by Hohenblum et al.
. Then the geometric mean of the population´s specific fluorescence was normalized by subtracting background signal (of non-producing P. pastoris wild type cells) and related to expression under the control of PGAP.
For quantification of HSA in shake flask and fermentation supernatants, the Human Albumin ELISA Quantitation Set (Cat. No. E80-129, Bethyl Laboratories, TX) was used. The HSA standard was applied with a starting concentration of 400 ng mL−1. Dilution-, Blocking- and Washing buffer were based on TBS (50 mM Tris–HCl, 140 mM NaCl, pH 8.0) and completed with BSA (1% (w/v)) and/or Tween20 (0.05% (v/v)) accordingly.
HyHEL Fab was determined with ELISA as described previously
CpB was quantified using an enzymatic assay based on the cleavage of hippuryl-L-arginine (Cat. No. H2508, Sigma, MO). Generation of hippuric acid was monitored at its absorbance maximum of 254 nm. Prior to the measurement, the samples were desalted with Zeba Spin columns (Thermo Fisher Scientific, IL) and activated with trypsin (Cat. No. T8345, Sigma, MO).
c) Determination of residual glucose
The D-Glucose Assay - GOPOD-Format (Megazymes, IE) was used to determine residual glucose of chemostat samples. Supernatant sampling was done by pumping culture broth out of the bioreactor by producing an overpressure, and its direct sterile filtration using a vacuum filter unit (Cat. No. 5141178, Whatman, UK). Glucose-limited cultivations usually go along with very low residual glucose concentrations in the supernatant, so the manufacturer’s protocol was adapted for glucose concentrations from 10 to 100 mg L-1. Briefly, the ratio of reaction buffer to sample was changed from 30:1 to 3:1.