LM-GVP: A Generalizable Deep Learning Framework for Protein Property Prediction from Sequence and Structure

Proteins perform many essential functions in biological systems and can be successfully developed as bio-therapeutics. It is invaluable to be able to predict their properties based on a proposed sequence and structure. In this study, we developed a novel generalizable deep learning framework, LM-GVP, composed of a protein Language Model (LM) and Graph Neural Network (GNN) to leverage information from both 1D amino acid sequences and 3D structures of proteins. Our approach outperformed the state-of-the-art protein LMs on a variety of property prediction tasks including fluorescence, protease stability, and protein functions from Gene Ontology (GO). We also illustrated insights into how a GNN prediction head can guide the protein LM to better leverage structural information. We envision that our deep learning framework will be generalizable to many protein property prediction problems to greatly accelerate protein engineering and drug development.

Design, Synthesis, and Antitumor Activities Study of Stapled A4K14-Citropin 1.1 Peptides

A4K14-citropin 1.1 is a structurally optimized derivative derived from amphibians' skin secreta peptide Citropin, which exhibits broad biological activities. However, the application of A4K14-citropin 1.1 as a cancer therapeutic is restricted by its structural flexibility. In this study, a series of all-hydrocarbon stapled peptides derivatives of A4K14-citropin 1.1 were designed and synthesized, and their chemical and biological characteristics were also investigated. Among them, A4K14-citropin 1.1-Sp1 and A4K14-citropin 1.1-Sp4 displayed improved helicity levels, greater protease stability, and increased antitumor activity compared with the original peptide, which establishes them as promising lead compounds for novel cancer therapeutics development. These results revealed the important influence of all-hydrocarbon stapling side chain on the secondary structure, hydrolase stability, and biological activity of A4K14-citropin 1.1.

Constrained Peptides Mimic a Viral Suppressor of RNA Silencing

The structure-based design of constrained alpha-helical peptides derived from the viral suppressor of RNA silencing TAV2b is described. We observe that the introduction of two inter-side chain crosslinks provides peptides with increased alpha-helicity and protease stability. One of these modified peptides (B3) shows high affinity for different double-stranded RNA structures including a palindromic siRNA as well as microRNA-21 and its precursor pre-miR-21. Notably, B3 binding to pre-miR-21 inhibits Dicer processing in a biochemical assay. As a further characteristic this peptide also exhibits cellular entry. <br>

Constrained Peptides Mimic a Viral Suppressor of RNA Silencing

The structure-based design of constrained alpha-helical peptides derived from the viral suppressor of RNA silencing TAV2b is described. We observe that the introduction of two inter-side chain crosslinks provides peptides with increased alpha-helicity and protease stability. One of these modified peptides (B3) shows high affinity for different double-stranded RNA structures including a palindromic siRNA as well as microRNA-21 and its precursor pre-miR-21. Notably, B3 binding to pre-miR-21 inhibits Dicer processing in a biochemical assay. As a further characteristic this peptide also exhibits cellular entry. <br>

The chloroplast localization of protease mediated by the potato rbcS signal peptide and its improvement for construction of photorespiratory bypasses

Location of the proteases would affect on protease stability and photorespiratory bypass pathway, while it is unsolved. Potato rbcS signal peptide was analyzed and constructed into the protease for study of their localization site. The tartronate semialdehyde reductase (EcTSR) proteins could be accurately and efficiently located in chloroplast only when this signal peptide was extended to 80 amino acids. The signal peptide would help malate synthase (CmMS) locate to the surface of chloroplast, to form granules on the outer membrane of chloroplast. The whole spectrum scanning showed that these proteins could enter chloroplast. A signal peptide named PCS1 (Peptide of self-cleavage site 1) carrying a self-cleavage site was designed, and sixteen amino acids from the blue pigment precursor protein of chloroplast positioning signal of Silene pratensis were added to the C-terminal of PCS1. Transient expression, Western blot analysis and full-spectrum scanning showed that PCS1 could locate the EcTSR to the chloroplast, after the removal of the signal peptide.

Synthesis and anti-osteoporosis activity of novel Teriparatide glycosylation derivatives

N-Acetyl glucosamine glycosylation of Teriparatide achieves enhanced helicity, greater protease stability and increased osteoblast differentiation promoting ability.