Design of an improved universal signal peptide based on the α-factor mating secretion signal for enzyme production in yeast

Abstract

Saccharomyces cerevisiae plays an important role in the heterologous expression of an array of proteins due to its easy manipulation, low requirements and ability for protein post-translational modifications. The implementation of the preproleader secretion signal of the α-factor mating pheromone from this yeast contributes to increase the production yields by targeting the foreign protein to the extracellular environment. The use of this signal peptide combined with enzyme-directed evolution allowed us to achieve the otherwise difficult functional expression of fungal laccases in S. cerevisiae, obtaining different evolved α-factor preproleader sequences that enhance laccase secretion. However, the design of a universal signal peptide to enhance the production of heterologous proteins in S. cerevisiae is a pending challenge. We describe here the optimisation of the α-factor preproleader to improve recombinant enzyme production in S. cerevisiae through two parallel engineering strategies: a bottom-up design over the native α-factor preproleader (α_nat) and a top-down design over the fittest evolved signal peptide obtained in our lab (α_9H2 leader). The goal was to analyse the effect of mutations accumulated in the signal sequence throughout iterations of directed evolution, or of other reported mutations, and their possible epistatic interactions. Both approaches agreed in the positive synergism of four mutations (Aα9D, Aα20T, Lα42S, Dα83E) contained in the final optimised leader (α_OPT), which notably enhanced the secretion of several fungal oxidoreductases and hydrolases. Additionally, we suggest a guideline to further drive the heterologous production of a particular enzyme based on combinatorial saturation mutagenesis of positions 86th and 87th of the α_OPT leader fused to the target protein.

Keywords: Directed evolution; Enzyme heterologous expression; Saccharomyces cerevisiae; Signal peptide; Synthetic design; Α-factor preproleader.

Fig. 1

Scheme of MFα1 gene from S. cerevisiae and signal peptides. a The α-factor preproleader consists of a pre-region, a pro-region (with three N-glycosylation sites) and the first spacer (with KEX2 and STE13 cleavage sites). The signal peptide is followed by four α-factor gene copies separated by spacers (C) of different length. b The α_nat leader used in this study containing Lα42S and Dα83E mutations from pPICZα (Invitrogen) and the extra Glu-Phe residues after the spacer (light orange). c Evolved α_9H2 leader with Aα9D, Aα20T, Qα32H, Fα48S, Sα58G, Gα62R, Aα87T mutations (dark purple), as well as Lα42S, Dα83E mutations (light purple) and extra Glu-Phe (light orange)

Fig. 2

Laccase activities detected in S. cerevisiae microcultures expressing either PK2 (grey bars) or ApL (white bars) fused to the different single-mutated α leaders. Laccase activities were normalized to that of the corresponding parent type α_nat-PK2 or α_nat-ApL (red line). Error bars correspond to the error propagation of ten replicates of each parent type or individual mutant. Asterisks highlight significant differences between individual mutants and parent types according to Tukey's range test (95% confidence)

Fig. 3

Laccase activities detected in S. cerevisiae microcultures expressing either PK2 (a, c) or ApL (b, d). fused to individual, double and quadruple α-preproleader mutants (a, b), or to the best mutated α leaders (α_A9D,A20T and α_{A9D,A20T,T24S}) and the products of recombination with the second best (α_E86G,A87T) (c, d). Secreted activities were normalized to those of the corresponding parent types: α_nat-PK2 or α_nat-ApL (red line). Error bars correspond to the error propagation of ten replicates of each parent type or mutant. Asterisks indicate the highest laccase activities according to Tukey's range test (95% confidence)

Fig. 4

Top-down strategy over α_9H2 leader. a Scheme summarizing the three cycles of top-down design of α_9H2 leader directed to improve laccase secretion by removing possible non-beneficial mutations. The removed mutations are highlighted in each α leader sequence (SP1-SP8); colour codes correspond to those shown in Figure 1. b Laccase activities detected in S. cerevisiae microcultures expressing either PK2 (grey bars) or ApL (white bars) fused to the different reverted α_9H2 mutants (SP1-SP8). Laccase activities were normalized to that of the corresponding parent type α_nat-PK2 or α_nat-ApL (red line). Error bars correspond to the error propagation of ten replicates of each parent type or mutant. One asterisk indicates significant differences respecting the parent type and two asterisks highlight the clone with significant highest activity among all, according to Tukey's range test (95% confidence)

Fig. 5

Flask production of PK2 (a) and ApL (b) laccases by S. cerevisiae with the best α-factor leaders obtained in the bottom-up (α_A9D,A20T) and top-down (α_{A9D,A20T,A87T}) designing strategies compared with α_nat and α_9H2 leaders_. Laccase activity (U/L) was measured with ABTS pH 3. Error bars indicate standard derivation of three flask replicates

Fig. 6

Enzyme production by S. cerevisiae cultured in 96-well plates using α_nat, α_9H2 or α_OPT leaders. Secreted enzymatic activities of laccases (PK2, ApL, PeL, and PcL), aryl-alcohol oxidase (AAO), peroxidase (VP), β-glucosidases (BGL2 and BGL3) and sterol esterase (OPE) are indicated as fold improvements with respect to the activities obtained with α_nat leader. Error correspond to the error propagation of ten replicates of each construction (with α_nat, α_9H2 or α_OPT)

Fig. 7

a Percentages of clones with higher, lower or parent-like activities of mutant libraries obtained upon mutation of positions 86th and 87th of the spacer region (black) and on the 2nd and 3rd positions of NXT/S sequence of the N-glycosylation sites 57 (purple) and 67 (cyan) of α_OPT leader for secretion of laccases PK2 and ApL (interval of Confidence of 95%). b Activity landscapes of the aforementioned CSM86/87, CSM-NGly58/59 and CSM-NGly68/69 mutant libraries of α_OPT leader fused to laccases PK2 or ApL. The activities of the clones are shown as relative to the laccase activities obtained with α_OPT leader (depicted as 1); the interval of confidence of the CSM86/87 assay is indicated with dashed lines. c Secretion improvements for PK2 (top) and ApL (bottom) obtained throughout α-factor preproleader engineering, from α_nat to α_OPT mutated in 86/87