This is the submitted version of the journal article: Cámara, Elena; Monforte, Sergi; Albiol i Sala, Joan; [et al.]. «Deregulation of methanol metabolism reverts transcriptional limitations of recombinant Pichia pastoris (Komagataella spp) with multiple expression cassettes under control of the AOX1 promoter». Biotechnology and Bioengineering, Vol. 116, Issue 7 (July 2019), p. 1710-1720. DOI 10.1002/bit.26947 This version is available at https://ddd.uab.cat/record/266676 under the terms of the license Biotechnology & Bioengineering Deregulation of methanol metabolism reverts transcriptional limitations of recombinant Pichia pastoris (Komagataella spp) with multiple expression cassettes under control of the AOX1 promoter Journal: Biotechnology and Bioengineering Manuscript ID 18-849.R1 Fo Wiley - Manuscript type: Article Date Submitted by the n/a rP Author: Complete List of Authors: Cámara, Elena; Universitat Autonoma de Barcelona, Chemical, Biological and Environmental Engineering ee Monforte Mercado, Sergi; Universitat Autonoma de Barcelona, Chemical, Biological and Environmental Engineering Albiol, Joan; Universitat Autonoma de Barcelona, Chemical, Biological and Environmental Engineering rR Ferrer, Pau; Luxemboug Institute of Science and Technology, Environmental Research and Innovation; Universitat Autonoma de Barcelona, Departament d'Enginyeria Quimica, Biologica i Ambiental ev Pichia pastoris, AOX1 promoter, methanol metabolism, recombinant Key Words: protein production, Mxr1, Mit1 iew John Wiley & Sons Page 1 of 38 Biotechnology & Bioengineering 1 2 3 Deregulation of methanol metabolism reverts transcriptional limitations of recombinant 4 5 6 Pichia pastoris (Komagataella spp) with multiple expression cassettes under control of the 7 8 AOX1 promoter 9 10 11 12 Elena Cámara1, Sergi Monforte, Joan Albiol, Pau Ferrer2 13 14 15 16 17 Department of Chemical, Biological and Environmental Engineering, Universitat Autònoma de 18 19 20 Barcelona, 08193 Bellaterra (Cerdanyola del Vallès), Catalonia, Spain 21 Fo 22 23 24 rP 25 26 Corresponding author: 27 28 ee 29 30 Pau Ferrer 31 rR 32 33 Telephone: +(352) 2758885064; e-mail: pau.ferrer@list.lu 34 35 ev 36 37 38 iew 39 1 Current 40 affiliation: Department of Biology and Biological Engineering, Division of Industrial 41 42 Biotechnology, Chalmers University of Technology, Sweden 43 44 45 2Current affiliation: Luxembourg Institute of Science and Technology, Luxembourg 46 47 48 49 50 51 52 53 54 55 Running title: Deregulation of methanol metabolism reverts P. pastoris limitations 56 57 58 59 60 1 John Wiley & Sons Biotechnology & Bioengineering Page 2 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Fo 22 23 24 rP 25 26 27 28 ee 29 30 31 rR 32 33 34 35 ev 36 37 38 iew 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 2 John Wiley & Sons Page 3 of 38 Biotechnology & Bioengineering 1 2 3 Abstract 4 5 6 7 The methanol-regulated alcohol oxidase promoter (PAOX1) of Pichia pastoris (syn. 8 9 Komagataella spp.) is one of the strongest promoters for heterologous gene expression. 10 11 Although increasing the gene dosage is a common strategy to improve recombinant protein 12 13 productivities, P. pastoris strains harbouring more than two copies of a Rhizopus oryzae lipase 14 15 16 gene (ROL) have previously shown a decrease in cell growth, lipase production and substrate 17 18 consumption, as well as a significant transcriptional downregulation of methanol metabolism. 19 20 This pointed to a potential titration effect of key transcriptional factors Mxr1 and Mit1 21 Fo 22 23 regulating methanol metabolism caused by the insertion of multiple expression vectors. 24 rP 25 26 To proof this hypothesis, a set of strains carrying one and four copies of ROL (1C and 4C, 27 28 respectively) were engineered to co-express one or two copies of MXR1*, coding for a Mxr1 ee 29 30 variant insensitive to repression by 14-3-3 regulatory proteins, or one copy of MIT1. Small 31 rR 32 33 scale cultures revealed that growth, Rol productivity and methanol consumption were 34 35 improved in the 4C- MXR1* and MIT1 strains growing on methanol as sole carbon source, ev 36 37 whereas only a slight increase in productivity was observed for re-engineered 1C strains. We 38 iew 39 40 further verified the improved performance of these strains in glycerol/methanol-limited 41 42 chemostat cultures. 43 44 45 Keywords: Pichia pastoris (Komagataella spp.), Mxr1, Mit1, AOX1 promoter, methanol 46 47 metabolism, heterologous gene dosage, recombinant protein production 48 49 50 51 52 53 54 Introduction 55 56 57 The methylotrophic yeast Pichia pastoris (Komagataella spp.) is a widely used recombinant 58 59 protein host and one of the most preferred yeast system for the production of a wide range of 60 3 John Wiley & Sons Biotechnology & Bioengineering Page 4 of 38 1 2 3 proteins, ranging from biopharmaceuticals and industrial enzymes to membrane proteins and 4 5 6 complex nanostructures (Corchero et al., 2013; Bill, 2015; Byrne, 2015; Rabert et al., 2013; 7 8 Spohner et al., 2015). Key features, established tools and methodologies for protein production 9 10 in P. pastoris have been extensively reviewed (Ahmad et al., 2014; Cos et al., 2006;; Gasser et 11 12 13 al., 2013; Vogl et al., 2013; Yang and Zhang, 2018; Zahrl et al., 2017). One of the most 14 15 compelling factors in this host is the presence of the strong and tightly regulated promoter of 16 17 alcohol oxidase 1 (PAOX1), the most frequently used option to drive the expression of foreign 18 19 genes (Vogl and Glieder, 2013). Catabolite repression of PAOX1 and other methanol utilization 20 21 Fo 22 (Mut) genes by glucose and glycerol, among other C-sources, has been reported since long 23 24 time (Tschopp et al., 1987). However, the molecular mechanisms of PAOX1 transcriptional rP 25 26 regulation have been unveiled only over the past 10 years. Lin-Cereghino et al. (2006) 27 28 ee 29 identified Mxr1 (methanol expression regulator 1), a transcription factor (TF) with a zinc 30 31 finger DNA-binding domain, homologous to the TF Adr1 from Saccharomyces cerevisiae. rR 32 33 Mutations in this gene resulted in a strain unable to grow in methanol and oleate and induce 34 35 ev 36 the transcription of alcohol oxidase 1 (AOX1), dihydroxyacetone synthase (DAS), peroxin 8 37 38 (PEX8) and peroxin 14 (PEX14). Further investigations by Kranthi et al. (2009 and 2010) iew 39 40 revealed several Mxr1-binding sites in the promoter of PEX8, AOX1 and DAS. Subsequently, 41 42 Parua et al. (2012) characterized a 14-3-3 family protein (regulator of numerous biological 43 44 45 processes in many eukaryotes, Fu et al., 2000) in P. pastoris. Furthermore, these authors 46 47 described a highly conserved yeast 14-3-3 binding motif in Mxr1, revealing that the interaction 48 49 between both proteins was due to the phosphorylation of Mxr1 Ser215 in a carbon-source 50 51 52 dependent manner, leading to the repression of Mxr1-dependent genes. Interestingly, this study 53 54 suggested that 14-3-3 might be functioning at a post-DNA binding step. Recently, Mit1 55 56 (methanol-induced transcription factor), Rop (repressor of phosphoenolpyruvate 57 58 59 carboxykinase) and Trm1 have also been described as transcription factors responsible for 60 4 John Wiley & Sons Page 5 of 38 Biotechnology & Bioengineering 1 2 3 methanol regulation (Kumar and Rangarajan, 2012; Sahu et al., 2014; Wang et al., 2016). Wang 4 5 6 et al. (2016) suggested a transmission of the methanol induction signal among Mxr1, Mit1 and 7 8 Trm1 through a cascade, being Mxr1 the responsible for the derepression of PAOX1, whereas 9 10 methanol presence would induce Trm1 and, subsequently, Mit1 expression. Finally, it has been 11 12 13 demonstrated that Rop (Kumar and Rangarajan et al., 2012) and Nrg1 (Wang et al., 2015), are 14 15 repressors of methanol metabolism, competing with Mxr1 for the same binding sites in PAOX1. 16 17 New insights in this complex regulatory circuit can provide innovative strategies for 18 19 recombinant protein production using PAOX1. For instance, Vogl and co-workers et al. (2018) 20 21 Fo 22 have successfully overexpressed MXR1 and MIT1 for activation of PAOX1 under methanol-free 23 24 carbon-limiting conditions (Vogl et al., 2018). rP 25 26 27 In this context, our recent study (Cámara et al., 2017) led to supports the hypothesis that PAOX1 28 ee 29 30 faces a transcriptional limitation when P. pastoris strains expressing multiple copies of a 31 rR 32 Rhizopus oryzae lipase (Rol) enconding gene are expressed more than one copy of a lipase of 33 34 Rhizopus oryzae (Rol) encoding gene under the control of this promoter. This also PAOX1 leads 35 ev 36 to results in a transcriptional limitation of in the methanol assimilation capacity of such strains. 37 38 iew 39 Because of MXR1 is constitutively expressed at low levels (Lin-Cereghino et al., 2006), our 40 41 observations suggested that the presence of multiple PAOX1 copies resulted in an insufficient 42 43 number of Mxr1 molecules to fully induce the expression transcription of methanol-regulated 44 45 46 genes in multi-copy expression strains, including the recombinant ROL. Coherently with this 47 48 hypothesis, Takagi et al. (2009) reported an increase in recombinant protein production due to 49 50 MXR1 overexpression. In addition, MIT1-deficient strains have marginal AOX1 mRNA when 51 52 53 exposed to methanol, highlighting Mit1 as an essential TF for growth on methanol (Wang et 54 55 al., 2016). 56 57 58 In this study, we have co-expressed a deregulated MXR1 variant, MXR1* (Parua et al., 2012), 59 60 and MIT1 in a series of P. pastoris strains harbouring one and four copies of ROL (named 1C 5 John Wiley & Sons Biotechnology & Bioengineering Page 6 of 38 1 2 3 and 4C, respectively, Cámara et al., 2016). Mxr1* contains a mutation in the Ser215 that 4 5 6 completely prevents the binding of the 14-3-3 proteins to this TF, including any background 7 8 binding activity in the presence of methanol, thereby ensuring a constant action of Mxr1 9 10 independent of carbon source (Parua et al., 2012). We hypothesised that, by expressing the 11 12 13 mutant form of MXR1 we would reinsure its full activity as transcriptional activator of PAOX1. 14 15 The effect of MXR1* and MIT1 co-overexpression on cell growth and Rol productivity of the 16 17 re-engineered 1C and 4C strains was tested in shake flasks. Furthermore, a comparative 18 19 transcriptional analysis of 8 key genes involved in different relevant cellular processes between 20 21 Fo 22 the reference strain 4C and two strains with two additional copies of Mxr1* and one additional 23 24 copy of Mit1, respectively, was performed in chemostat cultures. rP 25 26 27 Methods and materials 28 ee 29 30 Strains and plasmids 31 rR 32 33 P. pastoris strains carrying 1 and 4 copies of the lipase of Rhizopus oryzae (named 1C and 4C 34 35 ev 36 strains, respectively) were constructed previously by Cámara et al. (2016), and together with 37 38 the non-producing strain X-33 (named 0C), were used as a reference strains. These two strains iew 39 40 were used as starting strains for this study. pGAPHA, a pGAPZA-derived plasmid in which 41 42 43 the Zeocin resistance marker is replaced by the Hygromycin resistance one (Adelantado, 2016), 44 45 was used as MXR1* and MIT1 expression vector. 46 47 48 Construction of expression vector pGAPHA_MXR1* and pGAPHA_MIT1 49 50 51 The mutant MXR1* gene, with the Ser215 replaced by an Ala, was amplified by PCR from the 52 53 54 recombinant plasmid pG213 (a gift from E.T. Young, University of Washington, USA) using 55 56 MXR1*_PmlI and MXR1*_KpnI primers (listed in Supplementary file I) and the Q5®Hot 57 58 Start High-Fidelity DNA polymerase (New England Biolabs, Ipswich, MA, USA). After 59 60 6 John Wiley & Sons Page 7 of 38 Biotechnology & Bioengineering 1 2 3 amplification, PCR product was sequenced using MXR1*seq_1-5 primers (Supplementary file 4 5 6 I). Afterwards, MXR1* was cloned into the pGAPHA vector between the PmlI and KpnI sites, 7 8 under the control of the GAP promoter (PGAP). The resulting plasmid, pGAPHA_MXR1* 9 10 (Figure 1A), was sequenced using the primers pGAP_fw, MXR1*_int_rev and MXR1*seq_5- 11 12 13 6. Plasmid and PCR product sequencing was carried out by the Servei de Genòmica i 14 15 Bioinformàtica of the UAB. 16 17 18 MIT1 gene was amplified by PCR from P. pastoris genomic DNA using MIT1 fw(inf) and 19 20 MIT1 rev(inf) primers (Supplementary file I). pGAPHA vector was amplified using pGAPHA 21 Fo 22 23 fw and pGAPHA rev primers (Supplementary file I). Then, pGAPHA_MIT1 was constructed 24 by using Infusion cloning (Clontech Laboratories, Becton Dickinson, New Jersey, USA) and rP 25 26 27 sequenced using Mit1Seq1, Mit1Seq2, Mit1Seq3 and Mit1Seq4 primers (Supplementary file 28 ee 29 30 I). 31 rR 32 33 Transformation of P. pastoris and clone selection 34 35 ev 36 Competent cells were prepared following the protocol described by Cregg (2007). 100 ng of 37 38 XbaI-linearized pGAPHA_MXR1* and AvrII-linearized pGAPHA_MIT1 were introduced iew 39 40 into P. pastoris competent cells by electroporation, using a Gene PulserXcell™ Electroporation 41 42 43 System (Bio-Rad, Hercules, CA, USA). Instruments settings were 1500 V, 25 µF, and 200 Ω. 44 45 Subsequently, transformants were selected on Yeast Extract Peptone Dextrose (YPD) plates 46 47 (1% (w/v) yeast extract, 2% (w/v) peptone, 2% (w/v) dextrose) containing 250 µg/mL 48 49 50 Hygromycin (Invivogen, San Diego, CA, USA). Positive clones were confirmed by PCR (after 51 52 genomic DNA extraction with Wizard® Genomic DNA Purification Kit, Promega, Madison, 53 54 WI, USA), by means of the amplification of MXR1*and MIT1 insert using the pGAP and 55 56 MXR1*_int and pGAP and MIT1_int primers, respectively (Supplementary file I). 57 58 59 60 Growth conditions 7 John Wiley & Sons Biotechnology & Bioengineering Page 8 of 38 1 2 3 Shake flask cultures 4 5 6 7 The P. pastoris shake flask cultures were performed in triplicate as described in Cámara et al., 8 9 2016. Briefly, strains were first grown in 0.5 L-volume baffled Erlenmeyer flasks containing 10 11 50 mL of fresh Buffered Minimal Glycerol (BMG) medium. After 20 h, cells were harvested, 12 13 and used to inoculate 0.25-L volume baffled flasks with membrane screw caps (Duran Group, 14 15 16 Mainz, Germany) with 35 mL of Buffered Minimal Methanol (BMM) medium and further 17 18 incubated for 70 h. One pulse of pure methanol to a final concentration of 0.5% v/v was added 19 20 to the cultures each 24 h. All media were supplemented with Hygromycin (150 µg/mL). 21 Fo 22 23 Chemostat cultures 24 rP 25 26 27 Chemostat cultures of selected strains were carried out at a working volume of 1 L in a 1.5-L 28 ee 29 vessel bioreactor (Biostat B Plus, Sartorius AG, Göttingen, Germany) as described in Cámara 30 31 et al., 2016. Briefly, cells were grown under carbon-limited conditions at a dilution rate (D) of rR 32 33 0.1 h-1 by feeding a defined growth medium containing a glycerol/methanol mixture 34 35 ev 36 (60%/40%, w/w) as a carbon source. The bioreactor off-gasses were passed through a 37 38 condenser cooled at 4°C by means of the cryostat Ecoline Star edition RE 106 (Lauda Dr. R. iew 39 40 Wobser GmbH & Co. KG, Lauda-Königshofen, Germany), further dried with two silica gel 41 42 43 columns and subsequently analysed with mass spectrometer Omnistar TM GSD 300 02 44 45 (Balzers Instruments, Balzers, Liechtenstein). Samples were taken after a minimum of five 46 47 residence times, once the steady state conditions were reached (Jordà et al., 2013). The initial 48 49 50 batch medium contained: 19.95 g/L glycerol, 0.9 g/L citric acid, 6.3 g/L (NH4)2HPO4, 0.01 g/L 51 52 CaCl2, 0.45 g/L KCl, 0.25 g/L MgSO4·7H2O, 1 mL/L Biotin (0.2 g/L; Sigma, Munich, 53 54 Germany), 0.3 mL antifoam (Glanapon, Bussetti & Co GmbH, Wien, Austria) and 2.3 mL/L 55 56 PTM1 trace salts stock solution. Chemostat medium contained: 15 g/L glycerol, 10 g/L 57 58 59 methanol, 0.42 g/L citric acid, 2.18 g/L (NH4)2HPO4, 5.5 mg/L CaCl2·2H2O, 0.85 g/L KCl, 60 8 John Wiley & Sons Page 9 of 38 Biotechnology & Bioengineering 1 2 3 0.32 g/L MgSO4·7H2O, 0.5 mL Biotin (0.2 g/L), 50 µL antifoam (Glanapon) and 0.8 mL PTM1 4 5 6 trace salts stock solution. The pH was set to 5.0 with 25% HCl in both media. PTM1 trace salts 7 8 stock solution was the same as previously described by Baumann et al. (2008). 9 10 Analytical procedures 11 12 13 Biomass determination 14 15 16 Cell biomass was monitored by measuring the optical density at 600nm (OD600). For cellular 17 18 dry weight, the method previously reported by Jordà et al. (2012) was used. All determinations 19 20 were performed in triplicate. 21 Fo 22 23 Lipase activity assay 24 rP 25 26 The lipolytic activity determination was performed in triplicate samples using the Lipase 27 28 colorimetric assay (kit 1821792 from Roche Diagnostics), as previously described by Resina ee 29 30 et al. (2004). 31 rR 32 33 Metabolite quantification 34 35 ev 36 Glycerol, methanol and other potential extracellular compounds in the shake flask and 37 38 chemostat cultures were analysed by HPLC as previously reported by Cámara et al. (2016). iew 39 40 Analyses were performed in duplicate for each independent experiment. 41 42 43 44 Flow cytometry analysis 45 46 47 Cell counting and viability was measured using the methods previously described by Cámara 48 49 et al. (2016). To determine the cell population size, the Flow Cytometry Size Calibration Kit 50 51 (Thermo Fisher Scientific, Waltham, MA, USA) was used. Briefly, microspheres suspensions 52 53 54 of 1 µm, 2 µm, 4 µm, 6 µm, 10 µm and 15 µm of diameter were sonicated for 5 s at 50 W with 55 56 the ultrasonic processor VC-5 (Vibracell, Sonics & Materials, Newtown, CT, USA) before 57 58 being subjected to analysis. After that, different thresholds were established for each range in 59 60 9 John Wiley & Sons Biotechnology & Bioengineering Page 10 of 38 1 2 3 the analysis software to define the interval size of the processed samples. Experiments were 4 5 6 performed using the Guava EasyCyte Mini cytometer (Millipore, Jaffrey, NH, USA), with a 7 8 488 nm Argon laser. An amount of 5,000 cells were measured per analysis at a flow rate of 9 10 0.59 mL/s. Viability assays were carried out in duplicates, whereas cell counting and cell size 11 12 13 determination were performed in triplicate for each shake flask and chemostat culture. 14 15 16 Gene copy number determination by digital droplet PCR (ddPCR) 17 18 19 Gene dosage quantification of ROL, MXR1 and MIT1 was analysed by ddPCR using the method 20 21 previously described (Cámara et al., 2016), using the reference gene β-actin (ACT1) as Fo 22 23 endogenous control for data normalization and the set of primers for MXR1, MIT1, ACT1 and 24 rP 25 26 ROL genes described in Supplementary file I for DNA amplification. To verify the ROL and 27 28 MXR1 gene dosage stability, biomass samples of the chemostat cultures were taken after five ee 29 30 residence times, once the steady state was achieved. Reagents for ddPCR were purchased to 31 rR 32 33 Bio-Rad (Hercules, CA, USA), whereas primers were synthesized by Sigma Aldrich (St Louis, 34 35 MO, USA). ev 36 37 38 Measurement of transcriptional levels by ddPCR iew 39 40 41 For transcript quantification, 5 mL of collected pellets at the end of the shake flask culture were 42 43 44 previously treated with phenol (5% v/v) and further stored at -80°C. Total RNA extraction was 45 46 performed using the RNeasy Mini Kit (Qiagen, Hilden, Germany) whereas cDNA was 47 48 synthesized by using the iScript™ cDNA Synthesis Kit (Bio-Rad) following the 49 50 manufacturer’s instructions. RNA quality was tested by measuring the ratio of absorbance at 51 52 53 260 and 280 nm using the Nanodrop 1000 (Thermo Fisher Scientific). ddPCR reactions 54 55 contained 10 µL of QX200™ ddPCR™ EvaGreen Supermix, 200 nM of forward primer, 100 56 57 nM of reverse primer, 0.4 ng of cDNA and the required amount of DNAse/RNAse-free water 58 59 60 up to 20 µL of final volume. Reactions were incubated at 95°C for 10 min, followed by a 10 John Wiley & Sons Page 11 of 38 Biotechnology & Bioengineering 1 2 3 denaturation (94°C, 30 s) and an annealing/extension step (60.2°C, 1 min for the KAR2 and 4 5 6 ROL primers; 56.5°C for the rest) during 40 cycles. Positive droplets of each PCR reaction 7 8 were normalized in relation to ACT1 signal. Primer list is shown in Supplementary File I. 9 10 11 Statistics 12 13 14 Data are expressed as the mean ± standard deviation (SD). Statistical analyses of the data were 15 16 17 performed using the unpaired Student’s t-test using Microsoft’s Excel software. A P-value 18 19 lower than 0.05 was considered statistically significant. 20 21 Fo 22 Results and Discussion 23 24 rP 25 Construction and characterization of Rol-producing strains co-expressing MXR1* or MIT1 26 27 28 In order to increase the Mxr1 levels in the Rol-producing strains, as well as ensuring a constant ee 29 30 31 activation of Mxr1-dependent promoters under induction conditions, an expression vector rR 32 33 harboring MXR1* (encoding for the Mxr1 variant carrying the mutation in the position Ser215) 34 35 was introduced in the Rol-producing reference strains 1C and 4C. Because previous studies ev 36 37 had reported a lethal effect of MXR1 overexpression under the control of inducible promoters 38 iew 39 40 (PAOX1 and PPEX8, Lin-Cereghino et al., 2006), but permissive growth when using constitutive 41 42 promoters (Takagi et al., 2012; Sahu and Rangarajan, 2015), PGAP was selected as the promoter 43 44 to drive MXR1* expression. Five and eleven MXR1* transformants were isolated for 1C and 45 46 47 4C strains, respectively. All 1C-MXR1* and 3 out of 11 for the 4C-MXR1* clones resulted to 48 49 have the MXR1* expression cassette integrated in their genome. Intriguingly, only three clones 50 51 of the 1C-MXR1*strain (1C1, 1C2 and 1C3) were capable of growing in liquid media. We 52 53 54 further determined the total MXR1 (MXR1+MXR1*) copy number for each clone by ddPCR 55 56 (Figure 1B), revealing a population with one MXR1* copy (1C2-MXR1* and 1C3-MXR1* 57 58 1C_MXR1*a and 1C_MXR1*b derived from the 1C strain, and 4C1-MXR1* and 4C2-MXR1* 59 60 11 John Wiley & Sons Biotechnology & Bioengineering Page 12 of 38 1 2 3 4C_MXR1*a and 4C_MXR1*b derived from the 4C strain), or two copies of MXR1* (1C1- 4 5 6 2MXR1* and 4C6-2MXR1*) 1C_2MXR1* and 4C_2MXR1*), demonstrating that PGAP was a 7 8 suitable option to express MXR1*. The fact that only clones with low gene dosages of MXR1* 9 10 were capable of growth in liquid media could be due to the deleterious effect of higher gene 11 12 13 dosages of this gene, i.e. suggesting that clones that can only grow on agar plates might 14 15 harboring a higher number of two copies of MXR1* copies can only grow on agar plates. 16 17 18 Concerning MIT1 co-overexpression, no fitness-related problems were observed in agar plates 19 20 and liquid medium growth steps during the isolation of these strains, allowing to isolate a series 21 Fo 22 23 of 1C- and 4C Rol-producing clones containing one extra copy of MIT1. 24 rP 25 26 Overexpression of MXR1* or MIT1 increases Rol production levels and reverts detrimental 27 28 effect of ROL multi-copy expression on methanol assimilation ee 29 30 31 As previously reported (Cámara et al., 2016; Cámara et al., 2017), P. pastoris strains carrying rR 32 33 more than one copy of a PAOX1-ROL expression cassette exhibit a reduced growth and methanol 34 35 ev 36 assimilation capacity. A first series of shake flask cultivations was carried out to test the effect 37 38 of MXR1* and MIT1 co-overexpression on cell growth, methanol consumption and Rol iew 39 40 production after 70 h of cultivation. As expected, biomass levels of the reference 4C-Rol- 41 42 43 producing strain were 3-fold lower than those observed for its 1C counterpart (Figure 2A). In 44 45 addition, significant levels of residual methanol were detected at the end of 4C strain culture, 46 47 while no substrate accumulation was observed in the 1C strain case (Figure 2C). Moreover, 48 49 50 while virtually all cells in the non-producer strain (0C) culture were viable (97%), a decrease 51 52 of 20% in viability was observed in 1C and 4C cultures (Supplementary File II2). ROL 53 54 overexpression further impacted cell size: whereas a slight (but significant) change in 55 56 population size profile was observed between strains 0C and 1C, a remarkable increase of 57 58 59 60 12 John Wiley & Sons Page 13 of 38 Biotechnology & Bioengineering 1 2 3 larger cell fractions (6 µm and ≥10 µm) was measured for the 4C strain, compared with the 4 5 6 other reference strains (Supplementary File II). 7 8 9 Co-expression of MXR1* or MIT1 in 1C strains had either no or slightly negative impact on 10 11 cell growth compared to the reference 1C strain, also showing no residual methanol at the end 12 13 of the culture. In contrast, the average biomass levels of both MXR1* and MIT1 series of 4C 14 15 16 clones presented 1.73- and 1.75-fold increase, respectively. Consistently, these clones also 17 18 showed significantly lower residual methanol levels at the end of the cultivation (2.12- and 19 20 3.48-fold reduction, respectively). Nonetheless, MXR1* strains still showed a 10-20% of non- 21 Fo 22 23 viable cells at the end of cultivations, and this value was even higher (around 40% non-viable 24 cells) in the case of the 4C1-MXR1* _MXR1*a strain. rP 25 26 27 28 As previously reported by Cámara et al. (2017), the increase in ROL gene dosage led to a higher ee 29 30 lipase specific productivity in the 4C strain compared to the 1C strain. Interestingly, for all the 31 rR 32 33 1C-MXR1* strains, Rol production was between two- and three-fold higher than the 1C control 34 35 strain (Figure 2B). However, Rol production in 1C-MIT1 strains was comparable to the ev 36 37 reference strain, or even slightly (not significantly) reduced (Figure 3B). Concerning the re- 38 iew 39 40 engineered 4C strains, 4C-MXR1 strains showed an average 1.2-fold increase and 4C-MIT1 41 42 strains a 2-fold increase. Although the average Rol production in MXR1* clones was not 43 44 significantly improved, a MXR1 clone showed a 1.8-fold increase. Notably, this clone was 45 46 determined to have 2 extra copies of the MXR1* gene instead of one (clone 4C6-2MXR1*, 47 48 49 Figure 1). As shown in Figure 2C and 3C, this increase in the lipase production was 50 51 concomitant with a sharp decrease in the residual methanol in the medium, reflecting a 52 53 methanol consumption rate similar to the 1C strains. In fact, strains with higher methanol 54 55 56 consumption rate (0C and 1C) presented an average population size lower than those with 57 58 reduced methanol assimilation capacities (4C and 4C1-MXR1*a). Notably, significant 59 60 differences were also observed in terms of cell size distribution between the strains 4C2- 13 John Wiley & Sons Biotechnology & Bioengineering Page 14 of 38 1 2 3 MXR1*b and 4C6-2MXR1* referred to the 4C strain control (Supplementary file II Figure 2B). 4 5 6 More specifically, a 20% increase was detected for the 4µm-cell population in the 4C-MXR1* 7 8 strains, while the 6µm-cell fraction was reduced around 15%, altogether resulting in a cell 9 10 distribution size similar to the 1C strain control, thus indicating a possible reversion of the 11 12 13 detrimental effects of ROL overexpression on the methanol assimilation capacity of the 4C 14 15 strain by means of the expression of MXR1*. 16 17 18 Chemostat cultures reveal a shift in the preference change of C-source consumption pattern 19 20 depending on the strain 21 Fo 22 23 To validate the results obtained in the shake flask cultures, the 4C reference strain, a 4C strain 24 rP 25 26 expressing 2 copies of MRX1* and a representative clone of a 4C strain overexpressing MIT1 27 28 (4C6-2MXR1* and 4C-MIT1, respectively) were further grown in chemostat cultures at a D of ee 29 30 0.09 h-1 using a mixed glycerol:methanol feed (60/40% v/v), i.e. under conditions analogous to 31 rR 32 33 our previous studies (Cámara et al., 2016; Cámara et al., 2017). In terms of extracellular lipase 34 35 activity, a 30% higher volumetric productivity was reached for the 4C6-2MXR1* strain in ev 36 37 comparison to 4C, corresponding to a slight but significant increase of 10% in terms of specific 38 iew 39 40 lipase productivity (Table I). Unexpectedly, a reduction of 13% and 28% in extracellular lipase 41 42 activity levels and specific lipase productivity, respectively, was observed for the 4C-MIT1 43 44 strain compared to the reference 4C strain. This might be due to higher tendency to foaming 45 46 observed for this strain under the tested chemostat conditions, since gas-liquid interphases can 47 48 49 influence lipases adsorption and activity. Nonetheless, the cultivations confirmed the positive 50 51 physiological impact of both MXR1* and MIT1 overexpression in the 4C strain. In particular, 52 53 biomass levels were about 30% and 20% higher than the 4C strain for 4C6-2MXR1* and 4C- 54 55 56 MIT1, respectively, also reflected by an increase of 15% in the biomass/substrate yield in both 57 58 cases (Table I). Under the established culture conditions, chemostats were are carbon limited 59 60 when growing the reference strains 0C and 1C, but residual methanol was is observed in the 14 John Wiley & Sons Page 15 of 38 Biotechnology & Bioengineering 1 2 3 fermentation broth when growing multi-copy strains (i.e. 4C strain). Notably In this 4 5 6 experiment, a substantial change was observed regarding the specific C-source consumption 7 8 rate between when comparing 4C vs and the 4C6-2MXR1* and 4C-MIT strains: Despite 9 10 glycerol concentration in chemostat cultures of all 4C strains was below detection limit, MXR*- 11 12 13 and MIT1-overexpressing 4C strains showed significantly higher biomass yields. 14 15 Consequently, specific glycerol consumption decreased about a 30%, from 1.96 mmol g-1 16 17 DCW h-1 19.59 for 4C to 1.52 mmol g-1 DCW h-1 for the 4C6-2MXR1* strain and 1.48 mmol 18 19 g-1 DCW h-1 for the 4C-MIT1 strain. Such decrease, because despite that no glycerol 20 21 Fo 22 accumulation was observed in the bioreactor, the increase in the biomass levels modified the 23 24 assimilation rate. In contrast, the specific methanol consumption rate appeared to increased for rP 25 26 both modified strains, from 1.44 to 1.71 and 1.70 mmol g-1 DCW h-1, when comparing the 4C 27 28 ee 29 with the 4C6-2MXR1* and 4C-MIT1 strain, respectively. Nevertheless, some residual 30 31 methanol (3.7 g/L) was still observed in the cultivation broths ure media (not observed in rR 32 33 analogous conditions for the reference strains 0C and 1C), which may pose indicate the 34 35 ev 36 possibility of further improvement of the methanol assimilation capacity, deserving further 37 38 studies. Interestingly, this might also indicate a variation in the C-source preference of the iew 39 40 MXR*- and MIT1-overexpressing 4C strains may reflect pointing out a lower metabolic 41 42 constraints of in the a methanol metabolism with lower constrains that could be also reflected 43 44 45 at transcriptional level. 46 47 48 Transcriptional analysis reveals an upregulation of methanol metabolism key genes in the 49 50 4C6-2MXR1* and 4C-MIT1 strains 51 52 53 To further evaluate if the physiological changes observed in the strains 4C6-2MXR1* and 4C- 54 55 56 MIT compared to the reference strain 4C (i.e. higher biomass levels, increase in the methanol 57 58 specific consumption rate), corresponded to a transcriptional alteration (or deregulation) of the 59 60 methanol metabolic pathway, several key genes of this route were analyzed in cells grown in 15 John Wiley & Sons Biotechnology & Bioengineering Page 16 of 38 1 2 3 chemostat cultures (Figure 3). Specifically, transcriptional levels of AOX1, the gene encoding 4 5 6 the enzyme alcohol oxidase I were 2.3-fold and 2.1-fold higher for 4C6-2MXR1* and 4C-MIT1 7 8 than for the reference strain 4C. Similarly, DAS1 and FDH1, encoding for key enzymes of the 9 10 methanol assimilatory and dissimilatory pathways (Figure 4) were also upregulated 3 and 1.9- 11 12 13 fold in the strain 4C6-2MXR1*, and 1.75 and 1.52-fold in the 4C-MIT1 strain. This was 14 15 coherent with the increased methanol assimilation capacity of these strains observed in both 16 17 shake flasks and chemostat cultures. In relation to ROL transcriptional levels were increased 18 19 by, a 5-fold and 3.2-fold increase was observed with in the 4C6-2MXR1* strain and a 3.2-fold 20 21 Fo 22 with 4C-MIT1 strain, respectively (Figure 3). Interestingly, this rise in mRNA did not result in 23 24 a similar increment in extracellular lipase production for 4C6-2MXR1*, where the specific rP 25 26 lipase productivity was only a 10% higher than the reference strain for 4C6-2MXR1* and a 27 28 ee 29 30% for 4C-MIT1 (Table I), pointing out to a translational (Cámara et al 2017) and/or 30 31 posttranslational limitation (Resina et al., 2007; Resina et al., 2009). Moreover, increased rR 32 33 foaming occurring during 4C-MIT1 chemostats could also partially explain the lower specific 34 35 ev 36 product production. In addition, as our previous transcriptomic study revealed a high impact of 37 38 the overexpression of ROL on the peroxisomal pentose phosphate regeneration pathway (PPP) iew 39 40 and peroxisome biogenesis in ROL multi-copy strains (Cámara et al., 2017), the mRNA levels 41 42 of fructose 1,6-bisphosphate aldolase (FBA1-2) and peroxine 6 (PEX6) were selected as 43 44 45 markers of these pathways. In the case of FBA1-2, an increase of 40% and 36% was detected 46 47 for 4C6-2MXR1* and 4C-MIT1 strains, respectively, supporting the role of the peroxisomal 48 49 PPP in methanol assimilation (Ruβmayer et al., 2015). Regarding PEX6 (coding for an ATPase 50 51 52 involved in the peroxisomal protein import and crucial for peroxisomal biogenesis, Saffian et 53 54 al., 2012), no significant regulation was detected comparing both strains. However, to 55 56 completely discard any regulation of peroxisomal biogenesis it would be interesting to further 57 58 59 60 16 John Wiley & Sons Page 17 of 38 Biotechnology & Bioengineering 1 2 3 analyze whether there were any changes in the expression of PEX8 and PEX14, also coding for 4 5 6 peroxines implied in the protein import machinery and directly regulated by MXR1. 7 8 9 Concerning the mRNA levels of native MXR1 and MIT1, no changes were observed between 10 11 both modified 4C strains and the reference 4C strain. Hence, the observed increases in total 12 13 (i.e. native+heterologous) MXR1 or MIT1 expression levels can be assigned to the transcription 14 15 16 factors encoding genes heterologously expressed under control of GAP promoter. Specifically, 17 18 the 4C6-2MXR1* and 4C-MIT1 strains showed an increase of 55% and 50% in total MXR1 19 20 and MIT1 transcript levels, respectively. Therefore, the deregulation of methanol metabolism 21 Fo 22 23 observed in these strains could be directly related to the significant increase in both 24 transcription factors mRNA levels. rP 25 26 27 28 Remarkably, no changes in the MXR1 mRNA levels of 4C6-2MXR1* were detected relative to ee 29 30 the 4C strain. The self-regulation of transcription factors has been extensively reviewed in 31 rR 32 33 prokaryotes and eukaryotes (MacPherson et al., 2006; Stekel and Jenkins, 2008), and even 34 35 described for Trm1 (Wang et al., 2016), a TF also involved in methanol metabolism regulation. ev 36 37 In this respect, our results hypothesis cannot totally exclude the possibility that points out to an 38 iew 39 40 auto regulatory feedback loop by Mxr1 to control the pool of available molecules occurs under 41 42 in these tested culture conditions. Despite this possibility, the significant upregulation detected 43 44 in the methanol-related genes could be explained by the fact that, even though the 45 46 transcriptional efficiency per copy (i.e. the ratio between mRNA levels and the number of 47 48 49 copies of a certain gene) of MXR1 decreased, the mutation introduced in a subpopulation of 50 51 molecules of Mxr1 molecules (e.g. ensuring a total derepression of PAOX1) was enough to trigger 52 53 an overexpression of theose monitored genes. Although Parua et al. (2012) demonstrated that 54 55 56 there was no 14-3-3 dependent inhibition of methanol utilization (Mut) genes in P. pastoris 57 58 cultures growing on with glycerol or glucose as a C-source, basal phosphorylation of Mxr1 59 60 (i.e. promoting the union of 14-3-3 proteins) wasere observed in the analysis published, even 17 John Wiley & Sons Biotechnology & Bioengineering Page 18 of 38 1 2 3 in fully inducible conditions (methanol as the sole C-source). Due to the low levels of Mxr1, 4 5 6 this basal phosphorylation might severely affect the transcriptional levels of methanol- 7 8 dependent genes in the multi-copy strains. Hence, In this context, our results strongly support 9 10 our initial hypothesis points that mutation in the Ser215 of Mxr1 (i.e. Mxr1*), which prevents 11 12 13 its union to of 14-3-3 proteins to Mxr1*, allows for the higher transcriptional levels of Mut 14 15 genes observed in the strains expressing this MXR1*, resulting in a better methanol assimilation 16 17 capacity. This is further reflected in the reduced Altogether, it could explain the decrease in 18 19 methanol residual levels observed in for the 4C6-2MXR1* chemostat cultures. strain. 20 21 Fo 22 23 Overall, the results described above allow us to confirm the main role of Mxr1 in PAOX1-driven 24 recombinant protein production in P. pastoris. Due to the constitutive low expression of MXR1 rP 25 26 27 (Lin-Cereghino et al., 2008; Vogl and Glieder, 2013), a limited number of Mxr1 molecules has 28 ee 29 30 multiple binding sites available in several genes (AOX1, DAS, PEX8, ACS1) (Kranthi et al., 31 rR 32 2009; Kranthi et al., 2010; Sahu and Rangarajan, 2015). In the case of the ROL single copy 33 34 strain, this Mxr1 pool can be sufficient to accomplish an appropriate methanol assimilation 35 ev 36 concomitant with a proper recombinant protein expression (Figure 5, top). However, in the 37 38 iew 39 case of recombinant ROL multi-copy strains (i.e. with an increased number of PAOX1 sequences) 40 41 the number of available molecules of Mxr1 might be insufficient to deal with the regulation of 42 43 all the Mxr1-dependent genes and thereby negatively affecting the corresponding 44 45 46 transcriptional levels, including ROL expression (Figure 5, middle). Hence, this assumption 47 48 could explain the plateau-like trend observed between gene copy number and mRNA levels of 49 50 recombinant genes under PAOX1 control in other studies (Liu et al., 2014; Sha et al., 2013; Huang 51 52 53 et al., 2014; Yang et al, 2016). Our results strongly suggest that MXR1* overexpression can at 54 55 least partially revert Mxr1 limiting levels in ROL multi copy strains (Figure 5, bottom). 56 57 Furthermore, the underlying mechanisms of PAOX1 repression in glycerol culture conditions 58 59 allowed for maintaining an inducible recombinant protein system, while taking advantage of 60 18 John Wiley & Sons Page 19 of 38 Biotechnology & Bioengineering 1 2 3 the MXR1* expression. Moreover, as Mit1 and Trm1 cooperate with Mxr1 in the regulation of 4 5 6 methanol metabolism (Wang et al. (2016)) and, considering that TRM1 is also expressed at 7 8 constitutive low levels as MXR1, the MXR1* overexpression effect might be further improved 9 10 by enhancing the Trm1 expression levels. 11 12 13 Finally, as a typical Unfolded Protein Response (UPR) marker gene, KAR2 mRNA levels were 14 15 16 also measured. Previous transcriptomic analysis of the reference 4C strain did not show a 17 18 dramatic impact of ROL overexpression on the folding and secretory pathways (Cámara et al., 19 20 2017). Interestingly Strikingly, the 4C6-2MXR1* strain showed an upregulation of about 50% 21 Fo 22 23 in KAR2 transcription compared to the reference strain, whereas no significant changes were 24 observed for the 4C-MIT1 strain no significant changes were observed in KAR2 mRNA levels. rP 25 26 27 This might point out at a possible direct induction of UPR by MXR1* overexpression or, 28 ee 29 30 indirectly, as a result of enhanced of ROL expression. Further analyses beyond the scope of 31 rR 32 this study would be necessary to investigate whether the folding machinery of the strain 4C6- 33 34 2MXR1* is the limiting step in the production of a functional lipase as a result of the increased 35 ev 36 ROL transcriptional levels. 37 38 iew 39 40 Conclusions 41 42 43 Knowledge about the underlying regulation mechanisms of the AOX1 promoter of P. pastoris 44 45 has increased significantly over past recent years, opening the door to develop subsequent 46 47 engineering strategies developments overcoming transcriptional limitations and enhance ing 48 49 50 the performance of PAOX1-based expression systems. 51 52 53 Notably, in this study we show that overexpression of one of the major transcription factors 54 55 strongly activating PAOX1 in P. pastoris (MXR1 or MIT1) was sufficient to reverse the 56 57 transcriptional attenuation of methanol metabolism caused by the insertion of multiple PAOX1- 58 59 60 based expression cassettes, further supporting our initial hypothesis that these transcription 19 John Wiley & Sons Biotechnology & Bioengineering Page 20 of 38 1 2 3 factors are limiting. Furthermore, partial rewiring of PAOX1 transcriptional circuits by 4 5 6 overproducing a deregulated form of Mxr1 can still preserve a basic output in terms of 7 8 regulation pattern (derepression under glycerol-limiting conditions and induction by methanol) 9 10 in strains harboring multiple PAOX1-based expression cassettes, making these engineered strains 11 12 13 suitable for bioreactor-scale cultivations based on mixed carbon source feeding strategies. 14 15 Notably, Vogl et al. 2018 have shown that, by overexpressing transcription factors such as 16 17 MIT1 in PAOX1 based expression strains, it was sufficient to transform such methanol-dependent 18 19 strains into a glucose/glycerol regulated system. This allowed developing fed-batch cultivation 20 21 Fo 22 strategies for regulated, methanol-free, recombinant protein production. Our study further 23 24 illustrates the potential of engineering regulatory circuits for enhanced protein production rP 25 26 processes. 27 28 ee 29 30 Finally, the proposed regulatory model of PAOX1 by Mxr1 and Mit1 in the multi-copy strains, 31 rR 32 brings light to better understand the transcriptional bottlenecks previously observed in several 33 34 studies on multi-copy strains using methanol-based systems. Further quantitative physiology 35 ev 36 studies of these new strains could shed light about the global impact of these modifications in 37 38 iew 39 P. pastoris and, in particular, to novel bottlenecks that may appear downstream transcription 40 41 still preventing a fully linear increase of secreted product as a function of heterologous gene 42 43 dosage. 44 45 46 Acknowledgements 47 48 49 This work was supported by the project CTQ2013-4239-R and CTQ2016-74959-R 50 51 (AEI/FEDER, UE) of the Spanish Ministry of Economy, Industry and Competitiveness, 2014- 52 53 SGR-452 and Reference Network in Biotechnology (XRB) (Generalitat de Catalunya). 54 55 56 57 58 59 60 20 John Wiley & Sons Page 21 of 38 Biotechnology & Bioengineering 1 2 3 REFERENCES 4 5 Adelantado, N. 2016. Lipidomics studies of recombinant Pichia pastoris for improved recombinant 6 7 protein secretion through cell engineering. PhD thesis, Universitat Autònoma de Barcelona. Available 8 9 at http://www.tesisenred.net/handle/10803/384229 10 11 Ahmad M, Hirz M, Pichler H, Schwab H. 2014. Protein expression in Pichia pastoris: Recent 12 13 achievements and perspectives for heterologous protein production. Appl. Microbiol. Biotechnol. 14 15 98:5301–5317. 16 17 Bill RM. 2015. Recombinant protein subunit vaccine synthesis in microbes: A role for yeast? J. Pharm. 18 19 Pharmacol. 67:319–328. 20 21 Byrne B. 2015. Pichia pastoris as an expression host for membrane protein structural biology. Curr. Fo 22 23 Opin. Struct. Biol. 32:9–17. 24 rP 25 26 Cámara E, Albiol J, Ferrer P. 2016. Droplet digital PCR-aided screening and characterization of Pichia 27 pastoris multiple gene copy strains. Biotechnol. Bioeng. 113:1542–1551. 28 ee 29 30 Cámara E, Landes N, Albiol J, Gasser B, Mattanovich D, Ferrer P. 2017. Increased dosage of AOX1 31 promoter-regulated expression cassettes leads to transcription attenuation of the methanol rR 32 33 metabolism in Pichia pastoris. Sci. Rep. 7:44302. 34 35 ev 36 Corchero JL, Gasser B, Resina D, Smith W, Parrilli E, Vázquez F, Abasolo I, Giuliani M, Jäntti J, Ferrer P, 37 Saloheimo M, Mattanovich D, Schwartz S, Tutino ML, Villaverde A. 2013. Unconventional microbial 38 iew 39 systems for the cost-efficient production of high-quality protein therapeutics. Biotechnol. Adv. 40 41 31:140–153. 42 43 Cregg J M. 2007. DNA-mediated transformation. Methods Mol. Biol. 389:27-42. 44 45 Cregg JM, Cereghino JL, Shi J, Higgins DR. 2000. Recombinant Protein Expression in Pichia pastoris. 46 47 Mol. Biotechnol. 16:23–52. 48 49 De Schutter K, Lin Y-C, Tiels P, Van Hecke A, Glinka S, Weber-Lehmann J, Rouzé P, Van de Peer Y, 50 51 Callewaert N. 2009. Genome sequence of the recombinant protein production host Pichia pastoris. 52 53 Nat. Biotechnol. 27:561–566. 54 55 Ferrer P, Albiol J. 2014. 13C-based metabolic flux analysis of recombinant Pichia Pastoris. Methods in 56 57 Molecular Biology. Methods Mol Biol. 1191:291–313. 58 59 60 21 John Wiley & Sons Biotechnology & Bioengineering Page 22 of 38 1 2 3 Fu H, Subramanian RR, Masters SC. 2000. 14-3-3 proteins: Structure, Function, and Regulation. Annu. 4 5 Rev. Pharmacol. Toxicol. 40:617–647. 6 7 Gasser B, Prielhofer R, Marx H, Maurer M, Nocon J, Steiger M, Puxbaum V, Sauer M, Mattanovich D. 8 9 2013. Pichia pastoris: protein production host and model organism for biomedical research. Future 10 11 Microbiol. 8:191–208. 12 13 Huang J, Xia J, Yang Z, Guan F, Cui D, Guan G, Jiang W, Li Y. 2014. Improved production of a 14 15 recombinant Rhizomucor miehei lipase expressed in Pichia pastoris and its application for conversion 16 17 of microalgae oil to biodiesel. Biotechnol. Biofuels 7:111. 18 19 Isidro IA, Portela RM, Clemente JJ, Cunha AE, Oliveira R. 2016. Hybrid metabolic flux analysis and 20 21 recombinant protein prediction in Pichia pastoris X-33 cultures expressing a single-chain antibody Fo 22 23 fragment. Bioprocess Biosyst. Eng. 39:1351-1363. 24 rP 25 Jordà J, De Jesus SS, Peltier S, Ferrer P, Albiol J. 2014a. Metabolic flux analysis of recombinant Pichia 26 27 pastoris growing on different glycerol/methanol mixtures by iterative fitting of NMR-derived 13C- 28 ee labelling data from proteinogenic amino acids. N. Biotechnol. 31:120–132. 29 30 31 Jordà J, Rojas HC, Carnicer M, Wahl A, Ferrer P, Albiol J. 2014b. Quantitative metabolomics and rR 32 33 instationary13C-metabolic flux analysis reveals impact of recombinant Protein Production on trehalose 34 and energy metabolism in Pichia pastoris. Metabolites 4:281–299. 35 ev 36 37 Jordà J, Jouhten P, Cámara E, Maaheimo H, Albiol J, Ferrer P. 2012. Metabolic flux profiling of 38 iew recombinant protein secreting Pichia pastoris growing on glucose:methanol mixtures. Microb. Cell 39 40 Fact. 11:57. 41 42 43 Kranthi BV, Kumar HR, Rangarajan PN. 2010. Identification of Mxr1p-binding sites in the promoters of 44 genes encoding dihydroxyacetone synthase and peroxin 8 of the methylotrophic yeast Pichia pastoris. 45 46 Yeast 27:705–711. 47 48 49 Kranthi BV, Kumar R, Kumar NV, Rao DN, Rangarajan PN. 2009. Identification of key DNA elements 50 involved in promoter recognition by Mxr1p, a master regulator of methanol utilization pathway in 51 52 Pichia pastoris. Biochim. Biophys. Acta - Gene Regul. Mech. 1789:460–468. 53 54 Kumar NV, Rangarajan PN. 2012. The zinc finger proteins Mxr1p and repressor of 55 56 phosphoenolpyruvate carboxykinase (ROP) have the same DNA binding specificity but regulate 57 58 methanol metabolism antagonistically in Pichia pastoris. J. Biol. Chem. 287:34465–34473. 59 60 22 John Wiley & Sons Page 23 of 38 Biotechnology & Bioengineering 1 2 3 Lin-Cereghino GP, Godfrey L, de la Cruz BJ, Johnson S, Khuongsathiene S, Tolstorukov I, Yan M, Lin- 4 5 Cereghino J, Veenhuis M, Subramani S, Cregg JM. 2006. Mxr1p, a key regulator of the methanol 6 7 utilization pathway and peroxisomal genes in Pichia pastoris. Mol. Cell. Biol. 26:883–897. 8 9 Liu H, Qin Y, Huang Y, Chen Y, Cong P, He Z. 2014. Direct evaluation of the effect of gene dosage on 10 11 secretion of protein from yeast Pichia pastoris by expressing EGFP. J. Microbiol. Biotechnol. 24:144– 12 151. 13 14 15 MacPherson S, Larochelle M, Turcotte B. 2006. A fungal family of transcriptional regulators: the zinc 16 17 cluster proteins. Microbiol. Mol. Biol. Rev. 70:583–604. 18 19 Mattanovich D, Graf A, Stadlmann J, Dragosits M, Redl A, Maurer M, Kleinheinz M, Sauer M, Altmann 20 21 F, Gasser B. 2009. Genome, secretome and glucose transport highlight unique features of the protein Fo 22 23 production host Pichia pastoris. Microb. Cell Fact. 8:29. 24 rP 25 Parua PK, Ryan PM, Trang K, Young ET. 2012. Pichia pastoris 14-3-3 regulates transcriptional activity 26 27 of the methanol inducible transcription factor Mxr1 by direct interaction. Mol. Microbiol. 85:282–298. 28 ee 29 Rabert C, Weinacker D, Pessoa A, Farías JG. 2013. Recombinant proteins for industrial uses: Utilization 30 31 of Pichia pastoris expression system. Brazilian J. Microbiol. 44:351–356. rR 32 33 Resina D, Bollók M, Khatri NK, Valero F, Neubauer P, Ferrer P. 2007. Transcriptional response of P. 34 35 ev pastoris in fed-batch cultivations to Rhizopus oryzae lipase production reveals UPR induction. Microb. 36 37 Cell Fact. 6:21. 38 iew 39 Resina D, Maurer M, Cos O, Arnau C, Carnicer M, Marx H, Gasser B, Valero F, Mattanovich D, Ferrer P. 40 41 2009. Engineering of bottlenecks in Rhizopus oryzae lipase production in Pichia pastoris using the 42 43 nitrogen source-regulated FLD1 promoter. N. Biotechnol. 25:396–403. 44 45 Resina D, Serrano A, Valero F, Ferrer P. 2004. Expression of a Rhizopus oryzae lipase in Pichia pastoris 46 47 under control of the nitrogen source-regulated formaldehyde dehydrogenase promoter. J. Biotechnol. 48 49 109:103–113. 50 51 Rußmayer H, Buchetics M, Gruber C, Valli M, Grillitsch K, Modarres G, Guerrasio R, Klavins K, Neubauer 52 53 S, Drexler H, Steiger M, Troyer C, Al Chalabi A, Krebiehl G, Sonntag D, Zellnig G, Daum G, Graf AB, 54 Altmann F, Koellensperger G, Hann S, Sauer M, Mattanovich D, Gasser B. 2015. Systems-level 55 56 organization of yeast methylotrophic lifestyle. BMC Biol. 13:80. 57 58 59 Saffian D, Grimm I, Girzalsky W, Erdmann R. 2012. ATP-dependent assembly of the heteromeric Pex1p- 60 Pex6p-complex of the peroxisomal matrix protein import machinery. J. Struct. Biol. 179:126–132. 23 John Wiley & Sons Biotechnology & Bioengineering Page 24 of 38 1 2 3 Sahu U, Krishna Rao K, Rangarajan PN. 2014. Trm1p, a Zn(II)2Cys6-type transcription factor, is essential 4 5 for the transcriptional activation of genes of methanol utilization pathway, in Pichia pastoris. Biochem. 6 7 Biophys. Res. Commun. 451:158–164. 8 9 Sha C, Yu XW, Lin NX, Zhang M, Xu Y. 2013. Enhancement of lipase r27RCL production in Pichia pastoris 10 11 by regulating gene dosage and co-expression with chaperone protein disulfide isomerase. Enzyme 12 Microb. Technol. 53:438–443. 13 14 15 Spohner SC, Müller H, Quitmann H, Czermak P. 2015. Expression of enzymes for the usage in food and 16 17 feed industry with Pichia pastoris. J. Biotechnol. 202:118–134. 18 19 Stekel DJ, Jenkins DJ. 2008. Strong negative self regulation of prokaryotic transcription factors 20 21 increases the intrinsic noise of protein expression. BMC Syst. Biol. 2:6. Fo 22 23 Takagi S, Tsutsumi N, Terui Y, Kong XY. 2009. Method for methanol independent induction from 24 rP 25 methanol inducible promoters in Pichia. U.S. Patent 8,236,528 B2, August 7, 2012. 26 27 28 Tschopp JF, Brust PF, Cregg JM, Stillman CA, Gingeras TR. 1987. Expression of the lacZ gene from two ee 29 methanol-regulated promoters in Pichia pastoris. Nucleic Acids Res. 15:3859–3876. 30 31 rR 32 Vogl T, Glieder A. 2013. Regulation of Pichia pastoris promoters and its consequences for protein 33 production. N. Biotechnol. 30:385–404. 34 35 ev 36 Vogl T, Hartner FS, Glieder A. 2013. New opportunities by synthetic biology for biopharmaceutical 37 38 production in Pichia pastoris. Curr. Opin. Biotechnol. 24:1094–1101. iew 39 40 Vogl T, Sturmberger L, Fauland PC, Hyden P, Fischer JE, Schmid C, Thallinger GG, Geier M, Glieder A. 41 42 2018. Methanol independent induction in Pichia pastoris by simple derepressed overexpression of 43 44 single transcription factors. Biotechnol. Bioeng. 115:1037-1050. 45 46 Yang Z, Zhang Z. 2018. Engineering strategies for enhanced production of protein and bio-products in 47 48 Pichia pastoris: A review. Biotechnol Adv. 36:182-195. 49 50 Wang X, Menghao C, Shi L, Wang Q, Zhu J, Wang J, Zhou M, Zhou X, Zhang Y. 2015. PpNrg1 is a 51 52 transcriptional repressor for glucose and glycerol repression of AOX1 promoter in methylotrophic 53 54 yeast Pichia pastoris. Biotechnol Lett. 38:291-298. 55 56 Wang X, Wang Q, Wang J, Bai P, Shi L, Shen W, Zhou M, Zhou X, Zhang Y, Cai M. 2016. Mit1 57 58 transcription factor mediates methanol signaling and regulates Alcohol Oxidase 1 promoter in Pichia 59 pastoris. J. Biol. Chem.: 291:6245-61. 60 24 John Wiley & Sons Page 25 of 38 Biotechnology & Bioengineering 1 2 3 Yang H, Zhai C, Yu X, Li Z, Tang W, Liu Y, Ma X, Zhong X, Li G, Wu D, Ma L. 2016. High-level expression 4 5 of Proteinase K from Tritirachium album Limber in Pichia pastoris using multi-copy expression strains. 6 7 Protein Expr. Purif. 122:38–44. 8 9 Zahrl RJ, Peña DA, Mattanovich D, Gasser B. 2017. Sytems biotechnology for protein production in 10 11 Pichia pastoris. FEMS Yeast Res. 17:fox068. 12 13 14 15 16 17 18 19 20 21 Fo 22 23 24 rP 25 26 27 28 ee 29 30 31 rR 32 33 34 35 ev 36 37 38 iew 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 25 John Wiley & Sons Biotechnology & Bioengineering Page 26 of 38 1 2 3 FIGURE LEGENDS 4 5 6 7 Figure 1 Characterization of MXR1 copy number. Determination of total MXR1 8 9 (MXR1+MXR1*) copy number by ddPCR. Because only one copy of MXR1 is present in the 10 11 genome of P. pastoris (De Schutter et al., 2009; Mattanovich et al., 2009), clones with two or 12 13 three copies of MXR1 consequently harbor one or two copies of MXR1*, respectively. In bold, 14 15 16 clone nomenclature used in the MXR1* verification by PCR amplification. In brackets, final 17 18 nomenclature based on the results of MXR1* quantification by ddPCR. 19 20 21 Figure 2 Screening of MXR1* transformants in shake flask cultures. A) Biomass levels Fo 22 23 measured by OD600. B) Specific lipase activity in shake flask cultures. C) Residual methanol 24 rP 25 26 quantified by HPLC at the end of the cultures. Samples were taken after 70 h of culture. Values 27 28 for reference strains 1C and 4C are marked as horizontal thick black lines, whereas 1C and 4C ee 29 30 strains expressing MXR1* were represented as green and reddish plot boxes, respectively. 31 rR 32 33 Black dots inside plot boxes indicate values of individual clones of each population. 34 35 Corresponding confidence intervals are shown in the Supplementary file III. ev 36 37 38 Figure 3 Screening of MIT1 transformants in shake flask cultures. A) Biomass levels iew 39 40 measured by OD600. B) Specific lipase activity in shake flask cultures. C) Residual methanol 41 42 43 quantified by HPLC at the end of the cultures. Samples were taken after 70 h of culture. Values 44 45 for reference strains 1C and 4C are marked as horizontal thick black lines, whereas 1C and 4C 46 47 strains expressing MIT1 are represented as green and purple plot boxes, respectively. Black 48 49 50 dots inside plot boxes indicate values of individual clones of each population. Corresponding 51 52 confidence intervals are shown in the Supplementary file IV. 53 54 55 Figure 4 Transcriptional levels of selected genes. Comparative of mRNA levels of 4C (blue) 56 57 and 4C6-2MXR1* (maroon) strains. Relative amounts of mRNA were calculated comparing to 58 59 60 ACT1 expression levels of each strain. Thicker black lines represent the routes directly 26 John Wiley & Sons Page 27 of 38 Biotechnology & Bioengineering 1 2 3 controlled by measured genes. DHA, dihydroxyacetone; E4P, erythrose-4-phosphate; 4 5 6 HCOOH, formic acid; Form, formaldehyde; F6P, fructose-6-phosphate; GAP, glyceraldehyde- 7 8 3-phosphate; S1,7BP, sedoheptulose-1,7-bisphosphate; S7P, sedoheptulose-7-phosphate; GS- 9 10 CH2OH, S-(hydroxymethyl)gluthiatione, X5P, xylulose-5-phosphate; R5P, ribose-5- 11 12 13 phosphate; Rul5P, ribulose-5-phosphate; MXR1,methanol expression regulator 1; MIT1, 14 15 methanol-induced transcription factor 1; AOX1, alcohol oxidase 1; DAS1, dihydroxyacetone 16 17 synthase; FDH1, formate dehydrogenase 1; FBA1-2, fructose 1,6 –bisphosphate aldolase; 18 19 ROL, Rhizopus oryzae lipase; KAR2, endoplasmic reticulum chaperone BiP karyogamy gene 20 21 Fo 22 2; PEX6, peroxisomal biogenesis factor 6.*P< 0.05, **P< 0.01, comparing with 4C control 23 24 strain. rP 25 26 27 Figure 5 Regulatory model of MXR1 in PAOX1-driven recombinant protein production 28 ee 29 30 driven by PAOX1. Proposed regulatory model, where the presence of multiple TF binding sites 31 rR 32 contributes to a higher transcriptional levels of the gene as the number of attached TF increases, 33 34 due to a synergistic stimulation of gene expression. Top figure: Scenario for recombinant 35 ev 36 protein production in single copy strains. Under methanol-growth conditions, the number of 37 38 iew 39 limited molecules of Mxr1, which are present in a limited number, are available, which are 40 41 distributed among the multiple binding sites of the PAOX1 and PDAS, as well as among other 42 43 Mxr1-regulated promoters. Given the limited Mxr1 binding sites, methanol presence promotes 44 45 46 high level induction of PAOX1-driven recombinant expression. There may be also some basal 47 48 binding of Mxr1 molecules to PAOX1 in the presence of methanol. Middle figure: Scenario for 49 50 recombinant protein production in a 4C copy strain. As the number of Mxr1 binding sites 51 52 53 increases, the number of Mxr1 molecules available per PAOX1 binding site decreases, 54 55 resulting in reduced transcription levels of endogenous genes controlled by this transcription 56 57 factor Mxr1decreased. Bottom figure: Scenario for recombinant protein production in 4C 58 59 strain expressing Mxr1*. Overexpression of MXR1* leads to an increase in the total number 60 27 John Wiley & Sons Biotechnology & Bioengineering Page 28 of 38 1 2 3 of Mxr1 molecules. Moreover, replacement in the position Ser215 of Mxr1 by an Ala prevents 4 5 6 its phosphorylation and the subsequent union of 14-3-3 proteins. Consequently, the number of 7 8 occupied Mxr1 binding sites increases, with an important fraction of mutant Mxr1* molecules 9 10 also being bound, thereby obtaining Consequently, higher transcriptional levels of Mut 11 12 13 methanol related genes and ROL are obtained. Number of Mxr1 sites on the PAOX1 is depicted 14 15 according to Kranthi et al. (2009) and Kranthi et al. (2010). 16 17 18 Supplementary file I 1 List of primers used for cloning and droplet digital PCR. Tm, 19 20 melting temperature; %GC, percentage of guanosine+cytosine. 21 Fo 22 23 Supplementary file II 2 Flow cytometric analysis of screening samples. A) Viability of 24 rP 25 26 the cells at the end of the screening. B) Cell size distribution of yeast cells. 27 28 ee 29 Supplementary file III Screening of MXR1* transformants in shake flask cultures. A) 30 31 Biomass levels measured by OD600. B) Specific lipase activity in shake flask cultures. C) rR 32 33 Residual methanol quantified by HPLC at the end of the cultures. Samples were taken after 34 35 ev 36 70 h of culture. Values for reference strains 1C and 4C are marked as horizontal black lines. 37 38 Black dots indicate values of individual clones of each population. Error bars show an iew 39 40 interval with a 95% confidence level. 41 42 43 44 Supplementary file IV Screening of MIT1 transformants in shake flask cultures. A) 45 46 Biomass levels measured by OD600. B) Specific lipase activity in shake flask cultures. C) 47 48 Residual methanol quantified by HPLC at the end of the cultures. Samples were taken after 49 50 70 h of culture. Values for reference strains 1C and 4C are marked as horizontal black lines. 51 52 53 Black dots inside plot boxes indicate values of individual clones of each population. Error 54 55 bars show an interval with a 95% confidence level. 56 57 58 59 60 28 John Wiley & Sons Page 29 of 38 Biotechnology & Bioengineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Fo 20 21 22 23 rP 24 25 26 ee 27 28 29 rR 30 31 Figure 1. Characterization of MXR1 copy number. Determination of total MXR1 (MXR1+MXR1*) copy number 32 by ddPCR. Because only one copy of MXR1 is present in the genome of P. pastoris (De Schutter et al., 2009; Mattanovich et al., 2009), clones with two or three copies of MXR1 consequently harbor one or two copies of ev 33 MXR1*, respectively. In bold, clone nomenclature used in the MXR1* verification by PCR amplification. In 34 brackets, final nomenclature based on the results of MXR1* quantification by ddPCR. 35 36 iew 254x190mm (300 x 300 DPI) 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 John Wiley & Sons Biotechnology & Bioengineering Page 30 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Fo 20 21 22 23 rP 24 25 26 ee 27 28 29 rR 30 31 Figure 2. Screening of MXR1* transformants in shake flask cultures. A) Biomass levels measured by OD600. 32 B) Specific lipase activity in shake flask cultures. C) Residual methanol quantified by HPLC at the end of the cultures. Samples were taken after 70 h of culture. Values for reference strains 1C and 4C are marked as ev 33 34 horizontal thick black lines, whereas 1C and 4C strains expressing MXR1* were represented as green and reddish plot boxes, respectively. Black dots inside plot boxes indicate values of individual clones of each 35 population. Corresponding confidence intervals are shown in the Supplementary file III. 36 iew 37 254x190mm (300 x 300 DPI) 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 John Wiley & Sons Page 31 of 38 Biotechnology & Bioengineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Fo 20 21 22 23 rP 24 25 26 ee 27 28 29 rR 30 31 Figure 3. Screening of MIT1 transformants in shake flask cultures. A) Biomass levels measured by OD600. 32 B) Specific lipase activity in shake flask cultures. C) Residual methanol quantified by HPLC at the end of the cultures. Samples were taken after 70 h of culture. Values for reference strains 1C and 4C are marked as ev 33 34 horizontal thick black lines, whereas 1C and 4C strains expressing MIT1 are represented as green and purple plot boxes, respectively. Black dots inside plot boxes indicate values of individual clones of each population. 35 Corresponding confidence intervals are shown in the Supplementary file IV. 36 iew 37 254x190mm (300 x 300 DPI) 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 John Wiley & Sons Biotechnology & Bioengineering Page 32 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Fo 20 21 22 23 rP 24 25 26 ee 27 28 29 rR 30 31 32 ev 33 34 35 36 iew Figure 4 Transcriptional levels of selected genes. Comparative of mRNA levels of 4C (blue) and 4C6-2MXR1* 37 (maroon) strains. Relative amounts of mRNA were calculated comparing to ACT1 expression levels of each 38 strain. Thicker black lines represent the routes directly controlled by measured genes. DHA, 39 dihydroxyacetone; E4P, erythrose-4-phosphate; HCOOH, formic acid; Form, formaldehyde; F6P, fructose-6- 40 phosphate; GAP, glyceraldehyde-3-phosphate; S1,7BP, sedoheptulose-1,7-bisphosphate; S7P, sedoheptulose-7-phosphate; GS-CH2OH, S-(hydroxymethyl)gluthiatione, X5P, xylulose-5-phosphate; R5P, 41 ribose-5-phosphate; Rul5P, ribulose-5-phosphate; MXR1,methanol expression regulator 1; MIT1, methanol- 42 induced transcription factor 1; AOX1, alcohol oxidase 1; DAS1, dihydroxyacetone synthase; FDH1, formate 43 dehydrogenase 1; FBA1-2, fructose 1,6 –bisphosphate aldolase; ROL, Rhizopus oryzae lipase; KAR2, 44 endoplasmic reticulum chaperone BiP; PEX6, peroxisomal biogenesis factor 6.*P< 0.05, **P< 0.01, 45 comparing with 4C control strain. 46 47 75x69mm (300 x 300 DPI) 48 49 50 51 52 53 54 55 56 57 58 59 60 John Wiley & Sons Page 33 of 38 Biotechnology & Bioengineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Fo 20 21 22 23 rP 24 25 26 ee 27 28 29 rR 30 31 Figure 5. Regulatory model of MXR1 in PAOX1-driven recombinant protein production. Proposed regulatory 32 model, where the presence of multiple TF binding sites contributes to a higher transcriptional levels of the gene as the number of attached TF increases, due to a synergistic stimulation of gene expression. Top ev 33 figure: Scenario for recombinant protein production in single copy strains. Under methanol-growth 34 conditions, the molecules of Mxr1, which are present in a limited number, are distributed among the 35 multiple binding sites of the PAOX1 and PDAS, as well as other Mxr1-regulated promoters. Given the limited 36 iew Mxr1 binding sites, methanol presence promotes high level induction of PAOX1-driven recombinant 37 expression. There may be also some basal binding of Mxr1 molecules to PAOX1 in the presence of methanol. 38 Middle figure: Scenario for recombinant protein production in a 4C copy strain. As the number of Mxr1 39 binding sites increases, the number of Mxr1 molecules available per PAOX1 binding site decreases, resulting in 40 reduced transcription levels of endogenous genes controlled by this transcription factor. Bottom figure: 41 Scenario for recombinant protein production in 4C strain expressing Mxr1*. Overexpression of MXR1* leads to an increase in the total number of Mxr1 molecules. Moreover, replacement in the position Ser215 of Mxr1 42 by an Ala prevents its phosphorylation and the subsequent union of 14-3-3 proteins. Consequently, the 43 number of occupied Mxr1 binding sites increases, with an important fraction of mutant Mxr1* molecules also 44 being bound, thereby obtaining higher transcriptional levels of Mut genes and ROL. Number of Mxr1 sites on 45 the PAOX1 is depicted according to Kranthi et al. (2009) and Kranthi et al. (2010). 46 47 254x190mm (300 x 300 DPI) 48 49 50 51 52 53 54 55 56 57 58 59 60 John Wiley & Sons Biotechnology & Bioengineering Page 34 of 38 1 2 3 Table 1. Macroscopic growth parameters of P. pastoris strains growing on glycerol-methanol 4 in chemostat cultures 5 6 Biomass qs, glycerol qs, methanol Lipase 7 DCW Yx/s Lipase activity (g L ) -1 (mmol g-1 DCW (g C-mol )-1 (mmol g-1 (mmol g-1 activity (UA g-1 DCW) 8 h-1) DCW h-1) DCW h-1) (UA mL-1) 9 4C 8.64 ± 0.24 3.37 ± 0.17 13.66 ± 0.38 -1.96± 0.20 -1.44 ± 0.10 40.13 ± 0.02 4644.33 ± 129.04 10 11 4C6-2MXR1* 11.12 ± 0.11 3.37 ± 0.17 15.94 ± 0.16 -1.52± 0.20 -1.71 ± 0.20 57.57 ± 1.23 5177.23 ± 121.89 12 4C-1MIT1 10.32 ± 0.18 3.02 ± 0.12 16.00 ± 0.16 -1.48 ± 0.07 -1.70 ± 0.03 35.21 ± 0.97 3344.41 ± 286.65 13 DCW, dry cell weight, Biomass, specific biomass production rate, Yx/s biomass yield, qs, specific substrate 14 consumption rate. 15 16 17 18 19 20 21 Fo 22 23 24 rP 25 26 27 28 ee 29 30 31 rR 32 33 34 35 ev 36 37 38 iew 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 John Wiley & Sons Page 35 of 38 Biotechnology & Bioengineering Supplementary file 1. List of primers used for cloning and droplet digital PCR. 1 2 3 PRIMER SEQUENCE (5'-3') Tm %GC AMPLICON LENGHT (bp) 4 5 ROL probe CCGGTCACTCACTCGGTGGTGCA 75 65 - 6 ACT probe TCCGTATGGATCGGTGGTTCTATCCTCGCT 75 53 7 ROL fwd1 CCTGTCGTCCAAGAACAAC 62 52 164 8 ROL rev1 GAGGACCACCAACAGTGAAG 62 53 9 ACT1 fw TGTCCGGTGGTACTACTATGTTCC 65 50 10 199 ACT1 rev GATTCGTCGTACTCTTGCTTTGA 62 43 11 12 AOX1 fw GACATTCACGGTTTCGAAGG 61 40 78 13 AOX1 rev CCTCAAGAAGTCCTGGCAAAC 63 42 14 DAS1 fw TTGAACTGGGACGGAGTG 60 44 135 15 DAS1 rev CAAACCGTTGGCAATAGCAC 62 40 16 FAB1-2 fw CCCTTGGTTTGACGGAATG 60 42 17 97 FAB1-2 rev TTCCTCCGACAGGTCTAAC 60 42 18 19 FDH1 fw GGTGCTGGAAGAATTGG 57 52 135 Fo 20 FDH1 rev GACAGTGTCGACTCTTC 55 52 21 KAR2 fw GATGAAGTCGGGTCGTGTAC 61 55 110 22 KAR2 rev TCTTAGCAGCATCACCCAACC 62 50 23 rP MIT1 fw CGTGAATCTGCAACAACAGC 57 50 24 MIT1 rev CGGATCTGAACTGCCAGAG 58 58 151 25 26 MIT1 fw(inf) CGAGGAATTCACCGAAACGATGAGTACCGCAGCCCCAATC 71 55 ee 2710 27 MIT1 rev(inf) CATGTCTAAGGCTACAAACGATTTCTATTCTTCAACATTCCAG 67 37 TAG 28 MIT1 seq_1 GGGAAGTTGCATTGCTGAC 57 53 - 29 rR 30 MIT1 seq_2 GTCAGCAATGCAACTTCCC 56 53 - 31 MIT1 seq_3 CTCTGACAAGTTCTCATTGAG 55 43 - 32 MIT1 seq_4 GGAATTGAACAAACCCTTGACC 56 45 - ev 33 MXR1 fw TTCGCCCATTCTACTCG 57 52 34 MXR1 rev GGGCTTGACTCATTTCAG 57 52 101 35 pGAP_fw GATTATTGGAAACCACCAGAATCG 61 41 36 iew MXR1*_int_re 450 37 v AGTCTTGTTATGGCCGCATC 62 50 38 MXR1*_PmlI 39 fw GGCGACACGTGATGAGCAATCTACCCCCAAC 75 48 3490 40 MXR1*_KpnI CGGCAGGTACCCTAGACACCACCATCTAGTCGG 76 60 41 rev 42 MXR1* seq_1 CCGGCTGTATCTGGATTTAG 59 50 - 43 MXR1* seq_2 CTCTCGATGTGAACAGGAAC 59 50 - 44 MXR1* seq_3 GTGGCTACGATAGGTGCAG 61 57 - 45 MXR1* seq_4 CACAGTTGGGATAGGATG 56 50 - 46 MXR1* seq_5 CTTTCAGCTGCGGCATTTG 62 52 - 47 MXR1* seq_6 GCAAAATGGCATTCTGACATCC 62 47 - 48 49 pGAPHA fw GTTTGTAGCCTTAGACATGAC 55 43 3389 50 pGAPHA rev CGTTTCGGTGAATTCCTCGTTTC 58 48 51 PEX6 fw GGTTTGGATGTGGTCAAG 57 50 52 148 PEX6 rev TGGCTTTAGCAAGCAGAG 59 50 53 ROL fwd AAGTGGGACTGTGTCCAATG 61 50 54 158 ROL rev GCACTTCTGAAGGAGTTG 57 50 55 56 Tm, melting temperature; %GC, percentage of guanosine+cytosine. 57 58 59 60 John Wiley & Sons Biotechnology & Bioengineering Page 36 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Fo 20 21 22 23 rP 24 25 26 ee 27 28 29 rR 30 31 Supplementary file 2. Flow cytometric analysis of screening samples. A) Viability of the cells at the end of 32 the screening. B) Cell size distribution of yeast cells. ev 33 254x190mm (300 x 300 DPI) 34 35 36 iew 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 John Wiley & Sons Page 37 of 38 Biotechnology & Bioengineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Fo 20 21 22 23 rP 24 25 26 ee 27 28 29 rR 30 31 Suplementary file 3 32 254x190mm (300 x 300 DPI) ev 33 34 35 36 iew 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 John Wiley & Sons Biotechnology & Bioengineering Page 38 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Fo 20 21 22 23 rP 24 25 26 ee 27 28 29 rR 30 31 Supplementary file 4 32 254x190mm (300 x 300 DPI) ev 33 34 35 36 iew 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 John Wiley & Sons