Molecular Immunology 44 (2007) 2355–2361
A novel domain antibody rationally designed against TNF-␣ using
variable region of human heavy chain antibody as scaffolds to
display antagonistic peptides
Weisong Qin
a,1
, Jiannan Feng
b,1
, Yan Li
b
, Zhou Lin
b
, Beifen Shen
b,∗
a
Research Institute of Nephrology, Jinling Hospital, Nanjing, PR China
b
Institute of Basic Medical Sciences, Beijing, PR China
Received 30 August 2006; received in revised form 14 October 2006; accepted 20 October 2006
Available online 27 November 2006
Abstract
Neutralizing of TNF-␣ has been proved effective in treatment of some autoimmune diseases, e.g. rheumatoid arthritis and Crohn’s disease.
Low molecular weight synthetic peptides can mimic the binding sites of TNF-␣ receptors and block the activity of TNF-␣. In order to stabilize
the conformation, increase the affinity and bioactivity, in this study, heavy chain variable region of human antibody was used as a scaffold to
simultaneously display three peptides, which were designed on the interaction between TNF-␣ and it’s neutralizing monoclonal antibody. On the
basis of the structural character and physical–chemical property of the families of seven kinds of heavy chain variable regions (VH) in human
antibodies, the fifth type of VH was screened as scaffold to display the antagonist peptide. Based on the computer-guided molecular design method,
a novel domain antibody against TNF-␣ (named as ATD5) was designed as TNF-␣ antagonist. The theoretical study showed that ATD5 was
more stable than displayed antagonist peptide. The binding activity with TNF-␣ was higher than free peptides. After expression and purification
in Escherichia coli, ATD5 could bind directly with TNF-␣ and inhibit the binding of TNF-␣ to its two receptors, TNFR1 and TNFR2. ATD5
could also reduce the TNF-␣-mediated cytotoxicity and inhibit TNF-␣-meditated caspase activation on L929 cells in a dose dependent manner.
The activity of ATD5 was significantly stronger than three peptides displayed by ATD5. This study provides a novel strategy for the development
of new TNF-␣ inhibitors. This study demonstrates that it is possible to screen potential antagonists of TNF-␣ using in vitro analysis systems in
combination with the computer-aided modeling method.
© 2006 Elsevier Ltd. All rights reserved.
Keywords: TNF-␣ inhibitor; Domain antibody; Variable region; Molecular modeling
1. Introduction
Tumour necrois factor-␣ (TNF-␣) is a multifunctional cyto-
kine that plays a central role in infection, inflammation and
immune regulation (Beutler and Cerami, 1989; Smyth and
Johnstone, 2000). TNF-␣ exerts biological effects through
two TNF receptors (TNFR): TNFR1 (55-kDa) and TNFR2
(75-kDa) (Brockhaus et al., 1990). While normal serum level of
TNF-␣ is very important to immune homeostasis, inappropriate
production of TNF-␣ has been implicated in the pathogenesis of
∗
Corresponding author at: Institute of Basic Medical Sciences, P.O. Box
130(3), Taiping Road, Beijing 100850, PR China. Tel.: +86 10 66931325;
fax: +86 10 68159436.
E-mail address: shenbf@mx.cei.gov.cn (B. Shen).
1
These authors contributed equally to this work.
a wide spectrum of human diseases, including sepsis, cerebral
malaria, diabetes, cancer, osteoporosis, allograft rejection, and
autoimmune diseases such as multiple sclerosis, rheumatoid
arthritis, and inflammatory bowel diseases, etc. (Kollias et al.,
1999; Lipsky et al., 2000). Blocking the activity of excessive
TNF-␣ has become an effective therapeutic strategy (Palladino
et al., 2003).
Three protein-based TNF-␣ antagonists have been approved
by the US FDA and used in clinic successfully: infliximab
(Mouse-human chimeric anti-human TNF-␣ monoclonal
antibody, registerer as Remicade), adalimumab (Fully human
anti-human TNF-␣ monoclonal antibody, Humira) and etaner-
cept (fusion protein of soluble 75 kDa TNF-␣ receptors with
Fc fragment of human IgG, registered as Enbrel) (Moreland et
al., 1997; Maini et al., 1999; Leonardi et al., 2003). However,
these large macromolecules are expensive due to the costly
0161-5890/$ – see front matter © 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.molimm.2006.10.016