Biomedical Engineering Reference
In-Depth Information
the two moieties. This conjugate was several orders of
magnitude more cytotoxic to Daudi B lymphoma cells
overexpressing CD22 (a sugar binding transmembrane pro-
tein) as compared to unbound Onconase
1
. It was also very
active in vivo, able to extend the life of mice with dissemi-
nated Daudi lymphoma up to 200%. As a result, this
conjugate has been considered to be potential therapy for
non-Hodkin lymphoma.
D' Alessio and coworkers [60] employed a single-chain
variable region (scFv) of the murine antibody to ErbB2
transmembrane receptor in order to target human pancreatic
RNase to breast, ovary and lung carcinomas. The targeting
moiety was genetically fused to the N-terminus of human
pancreatic RNase (effector moiety) through a six-residue
spacer. Like LL2-Onconase
1
conjugate, this fusion protein
named by the authors “immunoRNase,” was highly specific
and very toxic to target cells, whereas human pancreatic
RNase itself was inactive.
As it is discussed in Section 4, production of chemical
conjugates of RNases to antibodies (e.g., LL2-Onconase
1
),
especially their large-scale manufacturing, poses numerous
technical and regulatory challenges. Therefore, more atten-
tion has been recently paid to immunoRNase fusion pro-
teins, like the mentioned anti-ErbB2 targeted enzyme. The
following sections of this chapter are devoted to such
constructs.
Table 22.1 shows immunoRNase fusions designed and
characterized to date. They are listed in chronologic order
illustrating the progress made in the field since 1992.
22.2.1 Immuno-Fusion Proteins of Eosinophil-Derived
Endotoxin or Angiogenin
Eosinophil-derived neurotoxin (Edn, RNase 2) [69,70]
recently reviewed [71] and angiogenin (Ang) [72] are
members of the superfamily of RNase A [71,73]. Both
enzymes are not inherently cytotoxic but they acquire
cytotoxic activity when fused to antibodies or other ligands.
Then, they interact with internalizing antigens that are
overexpressed on cancer cell surface.
The early papers by Rybak, Youle, and coworkers [74,75]
had a great influence on the field of RNase targeting. The
authors constructed a chimeric gene encoding a heavy chain
of mouse/human antibody to the human transferrin receptor
(CD71) and angiogenin. Expression of the synthetic gene in
E12B5 cells producing an antitransferrin receptor light chain
[76] resulted in a very low yield of secreted CH2-Ang fusion
protein. The product was very active (Table 22.1) against
K562 human erythroleukemia cells and activity was blocked
by the parent murine E6 antitransferrin receptor antibody.
The individual components of the fusion protein, the anti-
body or angiogenin, did not affect K562 cells. Therefore, it
was clear that toxicity was acquired as a result of their
fusion. The unusually high activity of this fusion protein was
later ascribed to its ability to form dimers [63,64].
The same group of authors followed up with a fusion of
the single chain (scFv) of the same E6 antibody and another
RNase, Edn [77]. This time the antibody moiety was fused to
the carboxy terminus of the RNase. The enzyme and anti-
body domains were separated by a 13 amino acid residues
spacer (48-60 residues segment of the staphylococcal
protein A) of the following sequence: AKKLNDAQAPKSD
[78]. The V
L
and V
H
fragments of the scFv were connected
by a 14 amino acid residue linker. A six histidine affinity tag
was attached at the carboxyl end of the V
H
fragment. The
entire construct was expressed in Escherichia coli BL21
(DE3) cells using the T7 system. Most of the fusion protein
was found in inclusion bodies. After denaturation and rena-
turation it was purified bymetal chelate chromatography. The
fusion protein was active against cell lines expressing trans-
ferrin receptor, K562 and A431 human epidermoid carci-
noma (Table 22.1) but not toward a cell line devoid of the
antigen. Another Edn-scFv (E6) fusion protein constructed in
a similar way, except for a different V
L
-V
H
linker, was toxic
to human cell lines derived frommelanoma, renal, and breast
carcinomas with IC
50
values of 8, 1.2, and 3 nM, respectively
[79] (Table 22.1).
Angiogenin was also fused to E6 derived scFv [80]. The
authors used two different peptides linking the V
L
and V
H
fragments, EGKSSGSGSESKEF or (GGGGS)
3
. Ang-scFv
22.2 DEVELOPMENT OF IMMUNORNASE
FUSION PROTEINS AS BIOPHARMACEUTICALS
A fusion protein may be defined as a genetic combination of
originally separate proteins [61] expressed from a single
synthetic gene. Most immunoRNase fusions are single chain
engineered proteins comprising an antibody fragment able
to bind the target antigen (targeting domain) and an entire
ribonucleolytic enzyme (effector domain) combined by
recombinant DNA technology. The domains are usually
separated by a short “spacer” amino acid sequence that
allows for independent folding of the construct components
afterexpression of the synthetic gene in a host cell. Targeting
domain may be composed of two “subdomains” like V
H
and
V
L
fragments of the antibody connected by a “linker” amino
acid sequence segment of various lengths. Other additional
short sequence segments frequently built in RNase fusion
proteins are “tag” sequences that facilitate detection or
isolation and purification of these constructs.
So far, immunofusion proteins of seven RNases were
designed and characterized as potential biopharmaceuticals.
The enzymes studied were: angiogenin, eosinophil-derived
neurotoxin, bovine seminal plasma RNase, human pancre-
atic RNase, Onconase
1
, Rana pipiens liver RNase, and
barnase. These fusion proteins were in part discussed in
previous review articles [3,8,27,56,57,62-68].
Search WWH ::
Custom Search