Processing of predicted substrates of fungal Kex2 proteinases from Candida albicans, C. glabrata, Saccharomyces cerevisiae and Pichia pastoris

Abstract

Background: Kexin-like proteinases are a subfamily of the subtilisin-like serine proteinases with multiple regulatory functions in eukaryotes. In the yeast Saccharomyces cerevisiae the Kex2 protein is biochemically well investigated, however, with the exception of a few well known proteins such as the alpha-pheromone precursors, killer toxin precursors and aspartic proteinase propeptides, very few substrates are known. Fungal kex2 deletion mutants display pleiotropic phenotypes that are thought to result from the failure to proteolytically activate such substrates.

Results: In this study we have aimed at providing an improved assembly of Kex2 target proteins to explain the phenotypes observed in fungal kex2 deletion mutants by in vitro digestion of recombinant substrates from Candida albicans and C. glabrata. We identified CaEce1, CA0365, one member of the Pry protein family and CaOps4-homolog proteins as novel Kex2 substrates.

Conclusion: Statistical analysis of the cleavage sites revealed extended subsite recognition of negatively charged residues in the P1', P2' and P4' positions, which is also reflected in construction of the respective binding pockets in the ScKex2 enzyme. Additionally, we provide evidence for the existence of structural constrains in potential substrates prohibiting proteolysis. Furthermore, by using purified Kex2 proteinases from S. cerevisiae, P. pastoris, C. albicans and C. glabrata, we show that while the substrate specificity is generally conserved between organisms, the proteinases are still distinct from each other and are likely to have additional unique substrate recognition.

Figure 1

Plasmid constructions for proteinase expression. (A) Schematic representation of the domain structure of fungal Kex2 proteins. Kex2 consists of a signal peptide, an autocatalytically removed pro-peptide, a catalytic domain, a structural P-domain, a transmembrane domain and finally a cytosolic domain containing sorting signals to the Golgi apparatus. (B) For expression of C. glabrata Kex2 the part of the gene fused with a C-terminal 6 × His tag was cloned into pPic3.5 using the BamHI and NotI restriction sites. (C) For expression of P. pastoris Kex2 the part of the gene fused with a C-terminal 6 × His tag was cloned into pPic3.5 using the SnaBI and NotI restriction sites. (D) For expression of C. albicans Kex2 the part of the gene fused with a C-terminal 6 × His tag was cloned into pCIp10 using the HinDIII and NheI restriction sites.

Figure 2

Activity testing of the purified Kex2 enzymes. The proteinases were tested with a proteinacious model substrate, the α-pheromone of S. cerevisiae. Digestion of the substrate protein (20 KDa) with the different proteinases resulted in the same expected pattern of products (11 and 2–3 KDa). Sc: S. cerevisiae Kex2, Ca: C. albicans Kex2, Cg: C. glabrata Kex2 and Pp: P. pastoris Kex2.

Figure 3

Activity testing of refolded CaPho11. CaPho11 was isolated from denatured inclusion bodies and refolded. To ensure its correct folding, activity testing using pNPP as a substrate was performed at different pH values. The enzyme was active and exhibited a maximum activity at pH 4.2–4.3.

Figure 4

Proteolytic digests of putative substrate proteins. (A) Recombinant substrate proteins were digested with each of the four proteinases. -: neg. control, Sc: S. cerevisiae Kex2, Ca: C. albicans Kex2, Cg: C. glabrata Kex2 and Pp: P. pastoris Kex2. Potential fragment sizes are given in kDa underneath the names (vertical bars: potential cleavage sites). All digests were visualized in silver stained gels, except CaSun41, where the N-terminal X-press epitope was detected in a Western blot. Proteins are digested at all major substrate sites found in the sequence and for most, intermediate products can be observed. Proteins not hydrolysed by Kex2 are not depicted (see text). Substrate CA0365 is not processed by ScKex2, but by all other proteinases, most efficiently by CaKex2.

Figure 5

Statistical sequence analysis of predicted Kex2 cleavage sites. (A) sequence logo of cleaved sites and (B) of non-cleaved sites. Position 5 of the logos corresponds to the P1 position in the substrate. Negatively charged residues (red) are overrepresented in processed substrate P1', P2' and P4' positions. Color key: red: negatively charged, blue: positively charged, black: apolar, green: polar.

Figure 6

Investigation of the three dimensional models of Kex2, furin and Kumamolisin for substrate binding properties. (A) Superimposed 3D coordinate sets for the three proteases reveal colocalization of the Kumamolisin propeptide residues with the predicted S1' and S2' binding pockets in furin and Kex2. A region identified for binding of the inhibitor Eglin-c (purple) [31] is not involved in binding of the propeptide. (B) and (C) A potential S4' binding pocket is identified which is terminated by H369 in Kex2 and E262 in furin. Numbering "a" through "E" refers to residues used in Table 2, which lists the respective binding pockets and references.

Figure 7

Sequence Alignment of fungal Kex2-like proteins. A protein sequence alignment of the residues involved in substrate specificity determination shows that the electrostatic properties of the binding regions are highly conserved. Red: positive charges, blue: negative charges, orange: polar residues, green: apolar residues, : Propeptide cleavage site, numbering "a" through "E" refers to residues used in Table 2, which lists the respective binding pockets and references.

Figure 8

Relevance of proper folding for proteolysis. (A) Sites readily cleaved in the native protein (nat) are cleaved less in heat denatured protein (denat). Shown are the results for CaEce1 and CA1873. (B) Potential sites not cleaved in the native protein are cleaved when exposed to the environment by fusion between GST and GFP (see text).

Figure 9

Schematic representation of polypeptide- and Ops4-like substrates. Kex2 cleavage sites are represented by vertical bars. SP: Signal peptide, GPI: potential GPI anchor attachment site. The proteins are digested at all sites found (see Figure 4 and Figure 8).