Understanding the structural details of APOBEC3-DNA interactions using graph-based representations

TY - JOUR

T1 - Understanding the structural details of APOBEC3-DNA interactions using graph-based representations

AU - Ng, J. C.F.

AU - Fraternali, F.

N1 - Publisher Copyright: © 2020 Copyright: Copyright 2021 Elsevier B.V., All rights reserved.

PY - 2020

Y1 - 2020

N2 - Human APOBEC3 (A3; apolipoprotein B mRNA editing catalytic polypeptide-like 3) is a family of seven enzymes involved in generating mutations in nascent reverse transcripts of many retroviruses, as well as the human genome in a range of cancer types. The structural details of the interaction between A3 proteins and DNA molecules are only available for a few family members. Here we use homology modelling techniques to address the difference in structural coverage of human A3 enzymes interacting with different DNA substrates. A3-DNA interfaces are represented as residue networks (“graphs”), based on which features at these interfaces are compared and quantified. We demonstrate that graph-based representations are effective in highlighting structural features of A3-DNA interfaces. By large-scale in silico mutagenesis of the bound DNA chain, we predicted the preference of substrate DNA sequence for multiple A3 domains. These data suggested that computational modelling approaches could contribute in the exploration of the structural basis for sequence specificity in A3 substrate selection, and demonstrated the utility of graph-based approaches in evaluating a large number of structural models generated in silico.

AB - Human APOBEC3 (A3; apolipoprotein B mRNA editing catalytic polypeptide-like 3) is a family of seven enzymes involved in generating mutations in nascent reverse transcripts of many retroviruses, as well as the human genome in a range of cancer types. The structural details of the interaction between A3 proteins and DNA molecules are only available for a few family members. Here we use homology modelling techniques to address the difference in structural coverage of human A3 enzymes interacting with different DNA substrates. A3-DNA interfaces are represented as residue networks (“graphs”), based on which features at these interfaces are compared and quantified. We demonstrate that graph-based representations are effective in highlighting structural features of A3-DNA interfaces. By large-scale in silico mutagenesis of the bound DNA chain, we predicted the preference of substrate DNA sequence for multiple A3 domains. These data suggested that computational modelling approaches could contribute in the exploration of the structural basis for sequence specificity in A3 substrate selection, and demonstrated the utility of graph-based approaches in evaluating a large number of structural models generated in silico.

KW - APOBEC3

KW - Protein structural networks

KW - Protein-DNA interaction

KW - Structural bioinformatics

UR - http://www.scopus.com/inward/record.url?scp=85105688986&partnerID=8YFLogxK

U2 - 10.1016/j.crstbi.2020.07.001

DO - 10.1016/j.crstbi.2020.07.001

M3 - Article

AN - SCOPUS:85105688986

SN - 2665-928X

VL - 2

SP - 130

EP - 143

JO - Current Research in Structural Biology

JF - Current Research in Structural Biology

ER -