Ana Letícia Vanz

Secretory recombinant protein production with Pichia (syn. Komagataella) pastoris is commonly associated with the induction of an unfolded protein response (UPR) usually apparent through increased intracellular levels of endoplasmic reticulum (ER) resident chaperones such as Kar2/Bip. During methanol-induced secretory production of an insulin precursor (IP) under industrially relevant fed-batch conditions the initially high level of intracellular Kar2/Bip after batch growth on glycerol unexpectedly declined in the following methanol fed-batch phase misleadingly suggesting that IP production had a low impact on UPR activation. Analysis of the protein production independent level of Kar2/Bip revealed that high Kar2/Bip levels were reached in the exponential growth phase of glycerol batch cultures followed by a strong decline of Kar2/Bip during entry into stationary phase. Ultra-structural cell morphology studies revealed autophagic processes (e.g. ER phagy) at the end of the glycerol batch phase most likely responsible for the degradation of ER resident chaperones such as Kar2/Bip. The pre-induction level of Kar2/Bip did not affect the IP secretion efficiency in the subsequent methanol-induced IP production phase. During growth on methanol intracellular Kar2/Bip levels declined in IP producing and non-producing host cells. However, extracellular accumulation of Kar2/Bip was observed in IP-producing cultures but not in non-producing controls. Most importantly, the majority of the extracellular Kar2/Bip accumulated in the culture supernatant of IP producing cells as truncated protein (approx. 35 kDa). Rapid growth leads to higher basal levels of the major UPR marker protein Kar2/Bip independent of recombinant protein production. Entry into stationary phase or slower growth on poorer substrate, e.g. methanol, leads to a lower basal Kar2/Bip level. Methanol-induced secretory IP production elicits a strong UPR activation which counteracts the reduced UPR during slow growth on methanol. The major ER chaperone Kar2/Bip is found together with recombinant IP in the culture medium where full-length Kar2/Bip accumulates in addition to large amounts of truncated Kar2/Bip. Thus, for judging UPR activating properties of the produced protein it is important to additionally analyze the medium not only for intact Kar2/Bip but also for truncated versions of this UPR reporter protein.

Pichia pastoris is a popular yeast preferably employed for secretory protein production. Secretion is not always efficient and endoplasmic retention of proteins with aberrant folding properties, or when produced at exaggerated rates, can occur. In these cases production usually leads to an unfolded protein response (UPR) and the induction of the endoplasmic reticulum associated degradation (ERAD). P. pastoris is nowadays also an established host for secretory insulin precursor (IP) production, though little is known about the impact of IP production on the host cell physiology, in particular under industrially relevant production conditions. Here, we evaluate the cellular response to aox1 promoter-controlled, secretory IP production in controlled fed-batch processes using a proteome profiling approach. Cells were first grown in a batch procedure using a defined medium with a high glycerol concentration. After glycerol depletion IP production was initiated by methanol addition which was kept constant through continuous methanol feeding. The most prominent changes of the intracellular proteome after the onset of methanol feeding were related to the enzymes of central carbon metabolism. In particular, the enzymes of the methanol dissimilatory pathway - virtually absent in the glycerol batch phase - dominated the proteome during the methanol fed-batch phase. Unexpectedly, a strong decrease of UPR and ERAD related proteins was also observed during methanol-induced IP production. Compared to non-producing control strains grown under identical conditions the UPR down-regulation was less pronounced indicating that IP production elicits a detectable but non prominent UPR response which is repressed by the general culture condition-dependent UPR down-regulation after the shift from glycerol to methanol. The passage of IP through the secretory pathway using an optimized IP vector and growing the strain at fed-batch conditions with a high initial glycerol concentration does not impose a significant burden on the secretory machinery even under conditions leading to an extracellular accumulation of ~ 3 g L^-1 IP. The glycerol batch pre-induction culture conditions are associated with a high constitutive - recombinant protein production independent - induction of the UPR and ERAD pathways probably preconditioning the cells for effective IP secretion in the methanol fed-batch phase.

Background: Biopharmaceutical drugs are mainly recombinant proteins produced by biotechnological tools. The patents of many biopharmaceuticals have expired, and biosimilars are thus currently being developed. Human granulocyte colony stimulating factor (hG-CSF) is a hematopoietic cytokine that acts on cells of the neutrophil lineage causing proliferation and differentiation of committed precursor cells and activation of mature neutrophils. Recombinant hG-CSF has been produced in genetically engineered Escherichia coli (Filgrastim) and successfully used to treat cancer patients suffering from chemotherapy-induced neutropenia. Filgrastim is a 175 amino acid protein, containing an extra N-terminal methionine, which is needed for expression in E. coli. Here we describe a simple and low-cost process that is amenable to scaling-up for the production and purification of homogeneous and active recombinant hG-CSF expressed in E. coli cells. Results: Here we describe cloning of the human granulocyte colony-stimulating factor coding DNA sequence, protein expression in E. coli BL21(DE3) host cells in the absence of isopropyl-β-D-thiogalactopyranoside (IPTG) induction, efficient isolation and solubilization of inclusion bodies by a multi-step washing procedure, and a purification protocol using a single cationic exchange column. Characterization of homogeneous rhG-CSF by size exclusion and reverse phase chromatography showed similar yields to the standard. The immunoassay and N-terminal sequencing confirmed the identity of rhG-CSF. The biological activity assay, in vivo, showed an equivalent biological effect (109.4%) to the standard reference rhG-CSF. The homogeneous rhG-CSF protein yield was 3.2 mg of bioactive protein per liter of cell culture. Conclusion: The recombinant protein expression in the absence of IPTG induction is advantageous since cost is reduced, and the protein purification protocol using a single chromatographic step should reduce cost even further for large scale production. The physicochemical, immunological and biological analyses showed that this protocol can be useful to develop therapeutic bioproducts. In summary, the combination of different experimental strategies presented here allowed an efficient and cost-effective protocol for rhG-CSF production. These data may be of interest to biopharmaceutical companies interested in developing biosimilars and healthcare community.