NOVEL APPLICATIONS OF BISPECIFIC DART PROTEINS - Fusion Protein Technologies for Biopharmaceuticals: Applications and Challenges - page 526

migrated around 29 kDa, similar to that expected for an
antibody Fab fragment.
This HER2

cytotoxic T-cell lymphocyte assay using A498 renal cell
carcinoma as target cells and human PBMC as effector cells.
The result of this screen, shown in Figure 36.2A, indicated
that many of the fluoresceinated mAbs were able to redirect
lysis of A498 cells in the presence of TCR U-DART,
suggesting that the target was amenable for the development
of T-DART 1 candidates. Furthermore, this screening pro-
cess provided a preliminary ranking of relative potency. No
activation or killing by the U-DART with effector cells was
observed in the absence of a fluoresceinated partner mAb. In
contrast, the mouse mAbs, being mostly IgG1, showed little
to no ability to mediate ADCC. A few IgG2a mAbs showed
limited activity that was Fc domain dependent (hence, relied
on FcR-dependent effector mechanisms) and thus did not
reflect T-cell mediated killing, the mechanism of action
desired for T-DART 1 development. U-DART studies
were subsequently expanded to other tumor cell

TCR DART 1 protein potently redirected
cytotoxicity against HER2 positive target cells. Moreover,
it retained high level of activity even against targets that
expressed low surface HER2 density, with a HercepTest
score as

(Figure 36.1B, C), that were
ineffectively lysed by trastuzumab. Addition of the gran-
zyme B inhibitor, zAAD-CMK, or the perforin inhibitor,
concanamycin A, blocked the cell-killing activity, demon-
strating that this DART 1 functions through the granzyme
B/perforin pathway (Figure 36.1C). This DART 1 protein
also effectively limited establishment of HER2

low as 1 þ

tumor
xenografts in human PBMC-reconstituted NOD/SCID
mice (Figure 36.1D).

þ

lines

36.5 U-DART CONCEPT FOR SCREENING
DART 1 CANDIDATE TARGETS AND MABS

expressing the target of interest.
Together with immunohistochemistry on normal and
tumor tissue specimens, flow cytometry, epitope binning,
and affinity analyses, the U-DART screens helped
in selecting a subset of eight mAbs for further analysis.
U-DART dose titration studies revealed that mAb #2 and
mAb #6 consistently performed the best both in terms of
EC 50 and maximum lysis (Figure 36.2B). Consistent with
the U-DART screen, a T-DART 1 candidate constructed
with a humanized version of mAb #6 potently mediated
lysis of target cells (Figure 36.2C). Thus, the U-DART
screen contributes a relevant functional evaluation of large
mAb panels at an early stage, assisting, together with
other parameters,

Given the applicability of the redirected killing approach to
multiple cancer targets, a rapid screening method was
needed in order to be able to evaluate antibodies “off the
shelf” for use in the bispecific DART 1 format before
embarking on more extensive lead generation, human-
ization, and engineering efforts.
For this purpose, we constructed a DART 1 protein
whereby one arm targets a relevant molecule on the desired
effector cell population (e.g., CD3, TCR, CD16) and the
other arm is an Fv specific for fluorescein (derived from the
4-4-20mAb). These proteins are termed “Universal
DART 1 ” or U-DART. In a U-DART-based screening pro-
cess, the candidate antibody is simply conjugated with
fluorescein isothiocyanate (FITC) and the fluoresceinated
mAb then combined with the U-DART and evaluated for its
intended biological effect. Hundreds of mAbs to known or
unknown targets can be screened rapidly using this
approach. The screen can serve multiple purposes, including
prioritizing mAbs against known targets for conversion to T-
DART 1 and clinical development as well as screening of
mAbs against unknown targets for selection for use in target
identification. This screening platform works with both
monoclonal and polyclonal antibodies from commercial
and proprietary sources.
As an example of this approach, a panel of 34 fluorescei-
nated mAbs was assessed for their ability to redirect T-cell
lysis of tumor cells using a 4-4-20

in identifying leads for further

development.
A similar mAb screening process was also developed
using 4-4-20

CD16 U-DART for the functional screen of
Fc-receptor (e.g., NK-cell) mediated redirected killing of
tumorcells,andcanbeusedtoevaluatecandidatemAbs
for use in either a CD16-based DART 1 or alternatively an
IgG antibody product with an Fc enhanced for CD16A
binding.

36.6 U-DART CONCEPT FOR APPLICATIONS IN
AUTOIMMUNE AND INFLAMMATORY DISEASE

As outlined earlier, one approach to control inflammatory
diseases is based on the hypothesis that activation and
inhibitory receptors on the surface of the same cell can
be cross-linked using bispecific DART 1 proteins in order to
dampen the responsiveness of that cell to activation and the
amplification of downstream effects, such as cytokine
releases, antibody production, or histamine release, depend-
ing on the cell type and application. The CD32B

TCR U-DART (Figure
36.2A). These 34 murine mAbs were all reactive against a
proprietary target and derived from numerous different
hybridoma campaigns in which mice were immunized
with human fetal or tumor-derived stem-like cell lines.
The purified mAbs were conjugated with fluorescein (FL)
to a similar degree and then combined with the 4-4-20

CD79B
DART 1 protein capable of cross-linking the inhibitory
CD32B receptor and the activating CD79B component of TCR U-DART at a fixed concentration of each and used in a