NOVEL APPLICATIONS OF BISPECIFIC DART PROTEINS - Fusion Protein Technologies for Biopharmaceuticals: Applications and Challenges - page 522

36

NOVEL APPLICATIONS OF BISPECIFIC DART 1 PROTEINS

S YD J OHNSON ,B HASWATI B ARAT ,H UA W. L I ,R ALPH F. A LDERSON ,P AUL A. M OORE ,A ND E ZIO B ONVINI

MacroGenics, Inc., Rockville, MD, USA

36.1 Introduction
36.2 DART 1 proteins
36.3 Application of DART 1 to cross-link inhibitory and acti-
vating receptors
36.4 Application of bispecific antibodies in oncology
36.5 U-DART concept for screening DART 1 candidate targets
and mabs
36.6 U-DART concept for applications in autoimmune and
inflammatory disease
36.7 Conclusions and future perspectives
References

molecule then could be purified from the multiple HL HL

combinations that were produced. Another nonrecombinant
method chemically conjugated the Fab 0 portion of one mAb
to the Fab 0 of a second mAb by coupling the fragments at
their hinge region to obtain a bispecific F(ab 0 )2 molecule
[6,7].
The dual targeting concept in oncology has also been
explored for regulating cell processes and increasing target
specificity. Bispecific antibodies have been developed that
bind to two different targets on the same cell to simulta-
neously modulate two different signaling pathways (e.g.,
EGFR and IGF-1R [8] or two distinct receptors for the same
ligand family (e.g., VEGFRI and VEGFR2) [9] or to
enhance the specificity and avidity for targeted tumor cells
(e.g., HER2 and HER3) [9,10]. For these approaches, effec-
tor cell recruitment and function may or may not be desir-
able, depending on whether the antibody exerts direct
inhibitory effects, whether an Fc-mediated engagement
further contribute to this inhibitory activity, or effector
functions such as antibody-dependent cellular cytotoxicity
(ADCC) are beneficial.
There are two general strategies for generating recombi-
nant bispecific antibodies: fragment-based or immuno-
globulin-like (Ig-like). Fragment-based strategies often
employ one or more single-chain variable domains
(scFv), [11,12] where a 12-16 amino acid linker tethers
each respective V L -V H (or V H -V L ) pair and a second linker
links two scFv units on the same molecule. An alternate
arrangement referred to as a diabody [13], uses shorter
linkers between V L and V H (or vice-versa) such that the
scFv domain cannot be formed but instead the interaction
with a second chain creates a dimeric protein with two
variable domains.

36.1 INTRODUCTION

Bispecific antibodies, in the simplest sense, combine two
different naturally or synthetically developed antigen-bind-
ing domains in a nonnatural manner in order to create a
single molecule that can bind two different targets or two
distinct epitopes on the same target. Such recombinant
proteins can provide novel solutions for treating various
diseases.
The most fully developed use of bispecific antibodies is
for the redirected killing of tumors cells by immune effector
cells. Early examples that entered clinical trials coupled a
tumor-targeting antibody arm with an arm recognizing
either CD3 on T cells [1], CD16 on natural killer (NK)
cells or macrophages, [2,3] or CD64 on macrophages [4].
These early bispecific antibodies were generally derived by
fusing two hybridomas forming a “quadroma” cell line that
expressed the heavy (H) and light (L) chain of each individ-
ual monoclonal antibody (mAb) [5]. The desired bispecific

In the latter case, the Fv from

þ þ

two respective mAbs A and B (V L -A V H -B and V L -B