Protein Hypersaline Adaptation: Insight from Amino Acids with Machine Learning Algorithms

Abstract

Traditional bioinformatics methods performed systematic comparison between the halophilic proteins and their non-halophilic homologues, to investigate the features related to hypersaline adaptation. Therefore, proposing some quantitative models to explain the sequence-characteristic relationship of halophilic proteins might shed new light on haloadaptation and help to design new biocatalysts adapt to high salt concentration. Five machine learning algorithm, including three linear and two non-linear methods were used to discriminate halophilic and their non-halophilic counterparts and the prediction accuracy was encouraging. The best prediction reliability for halophilic proteins was achieved by artificial neural network and support vector machine and reached 80 %, for non-halophilic proteins, it was achieved by linear regression and reached 100 %. Besides, the linear models have captured some clues for protein halo-stability. Among them, lower frequency of Ser in halophilic protein has not been report before.

A novel salt-tolerant chitobiosidase discovered by genetic screening of a metagenomic library derived from chitin-amended agricultural soil

“…Analyzing the amino acid composition of 53D1 according to Bolhuis et al ( 2008 ), the chitobiosidase showed features of halophilic organism-derived proteins, including: (1) an increased number of acidic residues (Glu, Asp—14.4 % in 53D1, vs the average value of 10.9 % in E. coli proteome), of Ser and Thr (14.4 vs 11.1 %), and of small hydrophobic residues (Ala, Gly and Val—29.6 vs 23.7 %); (2) a decreased number of Lys (2.5 vs 4.7 %) and of large hydrophobic residues (Ile, Leu, Phe and Met—18.9 % vs 23.5 %) (Bolhuis et al 2008 ; Reed et al 2013 ; Oren and Mana 2002 ). That 53D1 chitobiosidase has features of a protein with hypersaline adaptation was confirmed by applying the models (principal components analysis, PCA; partial least squares regression, PLSR; linear regression, LR) developed by Zhang and Ge ( 2013 ). The abundance of acidic amino acids, the prevalence of Asp and His (abundant in halophilic proteins), in contrast to Phe and Ile (scarce in halophilic proteins), and finally the low content of Cys, Ile and Gly vs the high content of Asp, His, Gln, Glu and Arg confirmed that the 53D1 chitobiosidase is a halophilic protein.…”

Protein Adaptations in Archaeal Extremophiles

“…Bioinformatics analysis of halophilic proteins has shown that their sequences also consistently contain less serine. Serine is good at interacting with water but not at competing with charged ions, so it is thought that serine is less useful for proteins at high salt concentrations [85]. An alternative to increased water binding would be that the acidic residues on halophilic proteins bind hydrated cations which would maintain a shell of hydration around the protein [8, 79, 83, 86–88].…”

“…They propose that the lower hydrophobic contact in the core may counterbalance the increased strength of hydrophobic interactions in high salt concentrations [91]. Most halophilic proteins contain less of the large, aromatic hydrophobic amino acids [85]. In the homology-modeled structure of halophilic dihydrofolate reductase, there was a decrease in the number of large hydrophobic amino acids, and a reduction of the enzyme core was observed [86].…”

Structural and Mechanistic Insights into the Improvement of the Halotolerance of a Marine Microbial Esterase by Increasing Intra- and Interdomain Hydrophobic Interactions

“…Besides the surface electrostatic interactions, the hydrophobic interactions of halophilic proteins have been evolved for haloadaption. Compared to nonhalophilic homologs, halophilic proteins contain a decreased number of large hydrophobic amino acid residues and have reduced hydrophobic interactions on the surface and in the hydrophobic cores (29,30). Weakening of these interactions may prevent the protein from aggregation and/or inactivation in hypersaline environments (30,31).…”