Protein trafficking, ergosterol biosynthesis and membrane physics impact recombinant protein secretion in Pichia pastoris

Abstract

Background: The increasing availability of 'omics' databases provide important platforms for yeast engineering strategies since they offer a lot of information on the physiology of the cells under diverse growth conditions, including environmental stresses. Notably, only a few of these approaches have considered a performance under recombinant protein production conditions. Recently, we have identified a beneficial effect of low oxygen availability on the expression of a human Fab fragment in Pichia pastoris. Transcriptional analysis and data mining allowed for the selection of potential targets for strain improvement. A first selection of these candidates has been evaluated as recombinant protein secretion enhancers.

Results: Based on previous transcriptomics analyses, we selected 8 genes for co-expression in the P. pastoris strain already secreting a recombinant Fab fragment. Notably, WSC4 (which is involved in trafficking through the ER) has been identified as a novel potential target gene for strain improvement, with up to a 1.2-fold increase of product yield in shake flask cultures. A further transcriptomics-based strategy to modify the yeast secretion system was focused on the ergosterol pathway, an aerobic process strongly affected by oxygen depletion. By specifically partially inhibiting ergosterol synthesis with the antifungal agent fluconazole (inhibiting Erg11p), we tried to mimic the hypoxic conditions, in which the cellular ergosterol content was significantly decreased. This strategy led to an improved Fab yield (2-fold) without impairing cellular growth. Since ergosterol shortage provokes alterations in the plasma membrane composition, an important role of this cellular structure in protein secretion is suggested. This hypothesis was additionally supported by the fact that the addition of non-ionic surfactants also enhanced Fab secretion.

Conclusions: The current study presents a systems biotechnology-based strategy for the engineering of the industrially important yeast P. pastoris combining the use of host specific DNA microarray technologies and physiological studies under well defined environmental conditions. Such studies allowed for the identification of novel targets related with protein trafficking and ergosterol biosynthesis for improved recombinant protein production. Nevertheless, further studies will be required to elucidate the precise mechanisms whereby membrane biogenesis and composition impact on protein secretion in P. pastoris.

Figure 1

24-well plate screening of co-expressing strains. Eight individual clones of each target gene and of a control strain containing the empty vector were used for a small scale screening in 24-well plates. All cultures started with the same initial OD₆₀₀ of 0.1 in BMD medium containing 2% (w/v) glucose. After 24 hours, Fab titers and wet cell weight were calculated to determine the product yield (mg_Fab g_WCW^-1) at a given cultivation point (24 h). The yield ratios relative to the control strain (value = 1) are illustrated. Error bars indicate the standard error of the means.

Figure 2

Shake flask screening of clones co-expressing ERO1 and WSC4. Relative product yields of ERO1 (A) and WSC4 (B) co-expressing P. pastoris clones (4 individual clones from each target gene) secreting the recombinant Fab antibody. The data were normalized to the control culture values, and error bars indicate the standard error of the means.

Figure 3

Schematic view of the ergosterol pathway and fluconazole inhibition. Outline of the post-squalene ergosterol biosynthetic pathway illustrating the step inhibited by fluconazole. Dashed arrows indicate no specification of intermediates, and red arrows highlight the genes that are upregulated in the Fab strain in hypoxic conditions.

Figure 4

Fluconazole screening. A: Determination of the fluconazole working concentrations. Growth curves of the recombinant P. pastoris strain X-33 pGAPZαA Fab2F5 grown in BMD medium and in the presence of 0 - 2 μg ml^-1 fluconazole. B: Fluconazole screening in shake flask cultures. Fab productivities in different concentrations of fluconazole (0.2 - 1.0 μg ml^-1) were normalized to the values obtained from non-treated cells. The relative Fab productivities for each concentration are demonstrated, error bars indicate the standard error of the means.

Figure 5

Effect of Tweens and Triton X-100 on Fab yield in shake flasks. The effect of the non-ionic surfactants Tween 20 (T-20), Tween 80 (T-80) and Triton X-100 (TX-100) on the Fab productivity in P. pastoris X-33 pGAPZαA Fab2F5 is demonstrated. The mean ratios of treated samples normalized to the untreated control samples are illustrated, and error bars indicate standard errors of the means.