Pichia pastoris as an Expression Host: Promoters, Secretome and Methanol Control
What the host offers
A 2014 review in Applied Microbiology and Biotechnology describes P. pastoris as an established expression host used mainly for biopharmaceuticals and industrial enzymes, distinguished by growth to very high cell densities, strong and tightly regulated promoters, and the option to produce gram amounts of recombinant protein per litre of culture either inside the cell or secreted [1]. The same review is blunt about the limits: not every protein is produced or secreted at those titres, and yields are frequently much lower for complex proteins — hetero-oligomers, membrane-attached proteins, or proteins prone to proteolytic degradation [1]. Work in recent years has focused on improving these difficult cases through either protein engineering or engineering of the host itself, including expression vector design, host strain engineering and screening for high-level expression strains [1].
Before a genome sequence was available, strain and process development for P. pastoris relied largely on analogies to well-studied yeasts such as S. cerevisiae [5]. The 9.4 Mb genome of the type strain DSMZ 70382 was sequenced and its secretome and sugar transporters analysed [5]. The computationally predicted secretome consists of 88 open reading frames, but when the organism was grown on glucose only 20 proteins were actually secreted at detectable levels [5]. That is a purification argument: the authors present it as a major feature of the host, namely low contamination of the heterologous product with host cell protein when a glucose-based expression system is used [5].
The same study draws a conclusion that matters for broth handling. Because few proteins and no proteases are actually secreted on glucose, cell lysis becomes the relevant cause of proteolytic degradation of secreted products [5]. The transporter analysis also explains growth behaviour: the genome carries two homologues of S. cerevisiae low-affinity hexose transporters and two of high-affinity transporters found in other Crabtree-negative yeasts, and the dominance of the high-affinity type limits glucose uptake rates and therefore the overflow metabolism seen in S. cerevisiae [5]. Unusually, the genome contains four H+/glycerol transporters, which the authors link to high specific growth rates on glycerol and to the suitability of a glycerol/glucose fermentation strategy [5].
Promoters: the main control handle
Expression in this host is driven by promoters, and the literature treats promoter choice as the principal tuning parameter. A 2008 paper in Nucleic Acids Research notes that only a limited set of promoters was available, and that fine-tuning of expression is often needed to maximise both product yield and product quality [2]. The authors built a promoter library by deleting and duplicating putative transcription-factor-binding sites within the AOX1 promoter sequence, initially spanning roughly 6% to more than 160% of wild-type promoter activity [2]. Characterisation under repressing, derepressing and inducing conditions identified at least 12 cis-acting elements involved in high-level expression from this promoter, and short artificial promoter variants were then built by combining those elements with a basal promoter [2].
The AOX1 promoter is induced by methanol, and several groups have looked for ways around that. A 2013 paper in Microbial Cell Factories set out the industrial requirement plainly: in a large-scale fed-batch culture, repression is wanted during the batch phase on surplus glycerol, and expression should switch on in the feed phase under carbon limitation [7]. DNA microarray analysis of wild-type cells in glycerol batch and glucose fed-batch identified six novel glucose-limit inducible promoters, which were used to express intracellular eGFP; the strongest expression combined with strong repression in pre-culture came from promoters designated P(G1) and P(G6) [7]. Tested with a secreted protein, human serum albumin, a P(G1) clone with two gene copies reached about 230% of the biomass-specific titre of a P(GAP) clone with the same copy number, while P(G6) reached 39% [7]. Clones carrying eleven gene copies produced about 1.0 and 0.7 g/L under P(G1) and P(G6) respectively in equal fed-batch processes [7]. The gene controlled by P(G1) was identified as a high-affinity glucose transporter, named GTH1 [7]. A 2006 review covers the broader operational picture — different phenotypes under the PAOX promoter, alternative promoters including GAP, FLD and ICL, medium composition, fed-batch and continuous strategies, mixed substrates, on-line monitoring of methanol and control algorithms — with the stated aim not only of maximising yield or productivity but of obtaining a repeatable batch-to-batch product for a robust industrial process [4].
Strain phenotype and the methanol pathway
Where methanol induction is used, the strain's methanol utilisation phenotype changes the outcome. In batch cultivations with methanol pulses, a MutS strain was found to be superior to a Mut+ strain in both volumetric productivity and efficiency for expressing recombinant horseradish peroxidase C1A [6]. Co-overexpressing key enzymes of the methanol utilisation pathway in MutS strains did not change stoichiometric yields, but significantly changed specific growth rate, specific substrate uptake rate and specific productivity [6]. Co-overexpression of dihydroxyacetone synthase gave a 2- to 3-fold more efficient conversion of methanol into product but reduced volumetric productivity, while co-overexpression of formaldehyde dehydrogenase gave a 2-fold more efficient conversion with at least similar volumetric productivity — the authors call the latter a valuable strategy [6].
Strain construction itself has been made more predictable. Deleting the P. pastoris KU70 homologue substantially increased targeted gene replacement efficiency, reaching over 90% targeting at the HIS4 and ADE1 integration loci with only 250 bp of flanking homologous DNA [3]. The deletion strain was more sensitive to UV than wild type, but no lethality, severe growth retardation or loss of gene copy number was detected over repeated rounds of cultivation and induction of heterologous protein production [3].
Secretion stress, and the limits of the Saccharomyces analogy
Host cells respond to recombinant protein production with metabolic and intrinsic stresses, including the unfolded protein response (UPR) [8]. Transcriptional monitoring in P. pastoris found that most folding and secretion related genes were regulated similarly to S. cerevisiae, while general stress response genes were differentially regulated [8]. Secretion of an antibody Fab fragment induced UPR target genes, though less strongly than overproduction of the UPR transcription factor Hac1p [8]. Overexpression of S. cerevisiae protein disulfide isomerase (PDI1) raised Fab secretion rates 1.9-fold without relieving UPR stress [8]. Lowering cultivation temperature from 25 °C to 20 °C increased the specific product secretion rate 1.4-fold in chemostat culture even though transcript levels of the Fab light and heavy chain genes fell significantly; some folding-related genes were down-regulated at the lower temperature while ER-associated degradation and secretory transport components were enhanced [8]. The authors' conclusion is that temperature affects productivity in a way that cannot be explained by thermodynamics alone, and that hosts must be analysed under real production conditions rather than extrapolated from model organisms [8].
What to take from it
For anyone specifying a yeast process, the evidence points to a short list of questions. Which Mut phenotype, since it changes productivity and substrate efficiency independently of yield [6]. The literature's own warning is the useful one: conclusions drawn from a model organism such as S. cerevisiae do not transfer wholesale, and the host should be characterised under the conditions it will actually be run in [8].
References
- Ahmad M, Hirz M, Pichler H, Schwab H. Protein expression in Pichia pastoris: recent achievements and perspectives for heterologous protein production — Applied microbiology and biotechnology, 2014-04-18.
- Hartner FS, Ruth C, Langenegger D, Johnson SN, Hyka P, Lin-Cereghino GP, Lin-Cereghino J, Kovar K, Cregg JM, Glieder A. Promoter library designed for fine-tuned gene expression in Pichia pastoris — Nucleic acids research, 2008-06-06.
- Näätsaari L, Mistlberger B, Ruth C, Hajek T, Hartner FS, Glieder A. Deletion of the Pichia pastoris KU70 homologue facilitates platform strain generation for gene expression and synthetic biology — PloS one, 2012-06-29.
- Cos O, Ramón R, Montesinos JL, Valero F. Operational strategies, monitoring and control of heterologous protein production in the methylotrophic yeast Pichia pastoris under different promoters: a review — Microbial cell factories, 2006-04-06.
- Mattanovich D, Graf A, Stadlmann J, Dragosits M, Redl A, Maurer M, Kleinheinz M, Sauer M, Altmann F, Gasser B. Genome, secretome and glucose transport highlight unique features of the protein production host Pichia pastoris — Microbial cell factories, 2009-06-02.
- Krainer FW, Dietzsch C, Hajek T, Herwig C, Spadiut O, Glieder A. Recombinant protein expression in Pichia pastoris strains with an engineered methanol utilization pathway — Microbial cell factories, 2012-02-13.
- Prielhofer R, Maurer M, Klein J, Wenger J, Kiziak C, Gasser B, Mattanovich D. Induction without methanol: novel regulated promoters enable high-level expression in Pichia pastoris — Microbial cell factories, 2013-01-24.
- Gasser B, Maurer M, Rautio J, Sauer M, Bhattacharyya A, Saloheimo M, Penttilä M, Mattanovich D. Monitoring of transcriptional regulation in Pichia pastoris under protein production conditions — BMC genomics, 2007-06-19.

