Format

Send to

Choose Destination
J Mol Biol. 2007 Jun 8;369(3):852-62. Epub 2007 Mar 14.

Analyzing the "degree of humanness" of antibody sequences.

Author information

1
Biomolecular Structure and Modelling Unit, Department of Biochemistry and Molecular Biology, University College London, Gower Street, London WC1E 6BT, UK.

Abstract

Genetically engineered mouse antibodies are now commonly in clinical use. However, their development is limited because the human immune system tends to regard them as foreign and this triggers an immune response. The solution is to make engineered antibodies appear more human. Here, we propose a method to assess the "degree of humanness" of antibody sequences providing a tool that may contribute to predictions of antigenicity. We analyzed sequences of antibodies belonging to various chains/classes in human and mouse. Our analysis of metrics based on percentage sequence identity between antibody sequences shows distinct differences between human and mouse sequences. Based on mean sequence identity and standard deviation, we calculated Z-scores for data sets of antibody sequences extracted from the Kabat database. We applied the analysis to a set of humanized and chimeric antibodies and to human germline sequences. We conclude that this approach may aid in the selection of more suitable mouse variable domains for antibody engineering to render them more human but in general, we find that typicality of a sequence compared with the expressed human repertoire is not well correlated with antigenicity. We have provided a Web server allowing humanness to be assigned for a sequence.

PMID:
17442342
DOI:
10.1016/j.jmb.2007.02.100
[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Elsevier Science
Loading ...
Support Center