Using signal peptide prediction with caution, a case study in Aspergillus niger xylanase

Abstract

Signal peptide (SP) prediction is used, but not known correct or not. The Aspergillus niger GH10 xylanase (XynB) SP is predicted to be 1–19 (Met₁–Ser₁₉) residue, differing from the Penicillium simplicissimum xylanase 1–25 residue SP. To determine the real SP, two types of XynB, XynΔ19 and XynΔ25, were constructed by respectively deleting the 19 (Met₁–Ser₁₉) or 25 (Met₁–Arg₂₅) residues. The XynΔ25 had 10 °C higher T_opt and 21.6-times longer thermostability than the XynΔ19 (46 vs. 36 °C and 47.6 vs. 2.2 m). When the kinetics were assayed, the XynΔ25 had ∼2.1-times higher V_max and higher binding-affinity for xylan than the XynΔ19 (53.7 vs. 25.3 μmol/ml/m and 2.43 vs. 2.96 mg/ml). Thus, the XynB real SP is the 1–25 and not predicted 1–19 residues. The extra six N-terminal residues (Glu₂₀Pro₂₁Ile₂₂Glu₂₃Pro₂₄Arg₂₅) drastically interfered with the XynΔ19 thermal activity, thermostability, and catalytic efficiency.

Introduction

Signal peptide (SP) usually sits at protein N-terminus and plays an important role in its maturation and translocation (http://en.wikipedia.org/wiki/G%C3%BCnter_Blobel), such as, secretion from cytoplasm to cell culture medium, transportation to Glogi apparatus or to endoplasmic reticulum. A database is built for better understanding of how SP functions [1].

In investigation of protein function, E. coli is easy and often used as host cells [2], [3], [4]. When a protein, especially, an eukaryotic one, is expressed in E. coli, the host cell usually cannot identify and delete the protein SP. It has to be deleted artificially through PCR; otherwise, SP would interfere with, or even worse, damage the protein function. A specific SP is usually selected and introduced to secret a recombinant protein from host cell. For example, the SP of Aureobasidium pullulans xylanase was introduced to secret the Penicillum citrinum xylanase into E. coli cell culture medium [2]. As for expression vectors in Pichia pastoris cell, the Saccharomyces cerevisiae α-mating factor is the generally used secretion signal [5]. Because of importance in analyzing protein function, SP is predicted by many webs [1], such as, signalP (http://www.cbs.dtu.dk/services/SignalP), SIG-Pred: Signal peptide prediction (http://www.bioinformatics.leeds.ac.uk/prot_analysis/Signal.html). These computational tools facilitate the investigation of protein function. However, the prediction result is usually not known whether correct or not.

Xylanase (EC 3.2.1.8) is important in biomass conversion, renewable energy production, feed stock, etc. [6], [7], [8]. Because of widely used in producing enzymes for fermentation industry, the Aspergillus niger CBS 513.88 genome was sequenced and annotated [9]. Based on the genome annotation, we cloned an A. niger xylanase (XynB). The enzyme belongs to GH10 family hydrolase. With an intention to investigate its property using E. coli cell, the SP should be identified and deleted artificially through PCR. According to the annotation (Swiss-prot: A2QFV7), the XynB SP was predicted to be the 1–19 residues (Met₁–Ser₁₉) [9]. The Penicillium simplicissimum xylanase (Swiss-prot: P56588) is another GH10 family enzyme; however, its SP is determined to be the 1–25 residues [10]. The two enzyme SP residues should be similar, because the XynB and the P. simplicissimum xylanase have 85.8% sequence identity. However, the predicted 1–19 (Met₁–Ser₁₉) residue is contrary to the probable 1–25 (Met₁–Arg₂₅) residue. With an intention to investigate which one is the real SP, two types of XynB, XynΔ19 and XynΔ25, are constructed by respectively deleting the first 19 or 25 residues. By determining and comparing properties of the two constructs, the 1–25 residue is proved to be the XynB real SP.