Signal Peptide Efficiency: From High-Throughput Data to Prediction and Explanation

Abstract

The passage of proteins across biological membranes via the general secretory (Sec) pathway is a universally conserved process with critical functions in cell physiology and important industrial applications. Proteins are directed into the Sec pathway by a signal peptide at their N-terminus. Estimating the impact of physicochemical signal peptide features on protein secretion levels has not been achieved so far, partially due to the extreme sequence variability of signal peptides. To elucidate relevant features of the signal peptide sequence that influence secretion efficiency, an evaluation of ∼12,000 different designed signal peptides was performed using a novel miniaturized high-throughput assay. The results were used to train a machine learning model, and a post-hoc explanation of the model is provided. By describing each signal peptide with a selection of 156 physicochemical features, it is now possible to both quantify feature importance and predict the protein secretion levels directed by each signal peptide. Our analyses allow the detection and explanation of the relevant signal peptide features influencing the efficiency of protein secretion, generating a versatile tool for the de novo design and in silico evaluation of signal peptides.

Protein-Specific Signal Peptides for Mammalian Vector Engineering

“…Accordingly, we sought to develop a tool that could be utilized to screen signal peptide performance in silico, to minimize the required in vitro testing space while maximizing protein expression. Although model-based tools have been created that can predict signal peptide performance in protein-partner specific contexts, this has only been achieved in bacterial systems . Utilizing the data obtained from screening our signal peptide panel in combination with three different molecules (i.e., Figure ), we attempted to build a model linking signal peptide performance to discrete protein sequence features.…”

“…As all recombinant protein products are secreted from the host cell factory, they have an absolute requirement to be paired with an appropriate signal peptide, a short N-terminal amino acid sequence that facilitates co-translational translocation of nascent polypeptides into the endoplasmic reticulum (ER). , Although signal peptides adhere to a generic three-domain structure, comprising a basic N-domain, a hydrophobic H-domain, and a slightly polar C-domain, the sequence features underpinning their performance are relatively poorly understood. Recent studies have begun to elucidate mechanistic design rules that govern whether a signal peptide will generally encode low or high ER translocation rates; however, (i) this work is predominantly in bacterial systems, , and (ii) there is a paucity of information regarding context-specific functionality, whereby individual signal peptide performance is highly variable dependent on the partner-protein used.…”

“…12,13 Although signal peptides adhere to a generic three-domain structure, comprising a basic N-domain, a hydrophobic H-domain, and a slightly polar C-domain, 14 the sequence features underpinning their performance are relatively poorly understood. Recent studies have begun to elucidate mechanistic design rules that govern whether a signal peptide will generally encode low or high ER translocation rates; 15 however, (i) this work is predominantly in bacterial systems, 16,17 and (ii) there is a paucity of information regarding context-specific functionality, whereby individual signal peptide performance is highly variable dependent on the partner-protein used. Previous work in CHO cells has identified bottlenecks in the secretory pathway as a limiting factor in production yields.…”

Protein-specific signal peptides for mammalian vector engineering

“…Accordingly, we sought to develop a tool that could be utilised to screen signal peptide performance in silico , to minimise the required in vitro testing space. While model-based tools have been created that can predict signal peptide performance in protein-partner specific contexts, this has only been achieved in bacterial systems [17]. Utilising the data obtained from screening our signal peptide panel in combination with three different molecules (i.e., Fig.1), we attempted to build a model linking signal peptide performance to discrete protein sequence features.…”

“…While signal peptides adhere to a generic three-domain structure, comprising a basic N-domain, a hydrophobic H-domain and a slightly polar C-domain [14], the sequence features underpinning their performance are relatively poorly understood. Recent studies have begun to elucidate mechanistic design rules that govern whether a signal peptide will generally encode low or high ER translocation rates [15], however i) this work is predominantly in bacterial systems [16, 17], and ii) there is a paucity of information regarding context-specific functionality, whereby individual signal peptide performance is highly variable dependent on the partner-protein used.…”