Biomedical Engineering Reference
In-Depth Information
production of bsAB that is essentially a hybrid of two
hybridomas. Although scientifically very elegant, it became
apparent that production in industrial scale is not feasible
because of the complexity of species produced [12]. Other
early approaches included chemical crosslinking [13,14],
tandem fusions of scFv or V-domains [15,16] or their fusion
to the C-terminus of heavy chains [8]. Up to now, only one
bsAB (Catumaxomab, [17]) has been approved in Europe for
treatment of malignant ascites. Catumaxomab is a rat/mouse
chimeric IgG2a displaying a high level of immunogenicity.
Other bsABs in advanced clinical development such as
Blinatumomab are based on tandem fusions of scFvs and
therefore do not display mAB effector functions or favorable
pharmacokinetic properties such as long serum half-life
in humans.
A human heterodimeric Fc fusion retaining long serum
half-life and mAB-like effector functions is the ideal vehicle
for engineering therapeutic bsABs. Genentech's “knobs-
into-holes” technology [18] offered a first technical solution
by introducing different complementary mutations in the
CH3 domain that favor heterodimerization.
Recently, Davis et al. [19] designed SEED (strand-
exchange engineered domains), a novel bsAB platform
based on the engineering of the CH3 domain. SEED is
based on a rational design strategy that combines structur-
ally related and conserved sequences of human IgG and IgA
CH3 domains. By alternating sequences derived from both
CH3 domains, two distinct SEED protein chains were
created (designated AG and GA) creating asymmetric inter-
faces with modified b -sheet structures that preferentially
heterodimerize (Figure 37.1). As CH3 domains drive the
heterodimerization of immunoglobulin heavy chains, AG
and GA chains expressed in the context of antibody heavy
chains can be utilized to produce monovalent and mono-
specific or bivalent and bispecific antibody-like molecules.
In the development and design of the SEED molecule,
special care was taken that the resulting bispecific molecule
should bear the same biological properties as the parental
immunoglobulin G (IgG) molecule. This includes long
serum half-life, antibody-dependent cellular cytotoxicity
(ADCC) and complement-dependent cytotoxicity (CDC),
as well as excellent biophysical properties such as high
protein homogeneity, protein A binding, and thermal stabil-
ity. The task was made more complicated by the fact that IgA
does not have all the same properties of IgG. Consequently,
the final molecule was not simply IgG on one half and IgA
on the other, but required some judicious addition of IgG
sequences to both halves to ensure properties such as protein
A and FcRn binding on each side of the protein. The
resulting SEED molecule then required extensive testing
in its earliest phases to guarantee that the reengineering
did not disrupt any of the critical biological properties
enumerated above. In addition, it was important to evaluate
the manufacturability of
FIGURE 37.1 SEED design produces an asymmetric hetero-
dimer protein scaffold. Diagram of the crystal structure determined
at 2.5 A of the AG/GA SEED heterodimer protein. The common
axis of symmetry is denoted with a dashed line through the strand-
exchange crossover segments (black). Sequence segments predom-
inantly derived from IgA are shown on the right in dark grey and
those derived from IgG on the left in light grey. The sequence on a
single chain alternates between IgA and IgG such that in the folded
tertiary structure one half of the strands and interface are derived
from IgA (right side) and the other half from IgG (left side).
expression levels, ease of purification, thermal stability,
and long-term storage potential. Only a SEED that passed
all of these tests would be ready for incorporation into the
therapeutic pipeline.
37.2 TECHNICAL ASPECTS
37.2.1 Workflow
For studies to assess the SEED design and SEED proteins,
we established a standardized workflow (Figure 37.2A) to
produce and characterize the purity and heterodimer content
of SEED proteins before further downstream biophysical
and biological experiments. For these assessments, SEED-
based proteins were compared to IgG1 Fc-based proteins to
test if SEED retains the desirable properties of therapeutic
antibodies. Proteins were produced by transient transfection
of 293 or Chinese hamster ovary (CHO) cells, followed by
protein A-purification from culture medium and then pre-
parative size-exclusion chromatography (SEC). Productivity
of proteins was measured as the yield of protein A-purified
protein obtained per millliter of cell culture medium. The
this new molecule,
including
Search WWH ::




Custom Search