FUSION PROTEINS WITH TOXIC ACTIVITY - Fusion Protein Technologies for Biopharmaceuticals: Applications and Challenges - page 248

on the cell surface, the heparin binding epidermal growth
factor precursor, it is cleaved by a furin-like protease
localized on the surface and internalized through clathrin-
coated pits. After cleavage, the two fragments are still joined
through a single disulfide bridge. The lowering of the pH in
the endosome induces a conformational change of the
translocation domain, exposing hydrophobic residues that
enable membrane insertion and finally escape from the
endosome. In the cytosol, the catalytic domain transfers
adenosine diphosphate-ribose (ADP-ribose) to the eukary-
otic translation elongation factors 2 (EF2), thus inhibiting
protein synthesis and ultimately causing cell death. Since the
toxin has catalytic activity, a single molecule is sufficient to
kill a cell [13]. Various truncations were evaluated to
identify the optimal construct with regard to immunogenic-
ity, unspecific binding, and potency.
Ontak, also called denileukin diftitox, the second ever
approved fusion protein, is a recombinant immunotoxin
consisting of diphtheria toxin whose binding domain is
replaced by interleukin 2 (IL-2). The molecule binds to cells
expressing the high affinity IL-2 receptor that can be frequently
found on T-cell lymphoma [14]. After the success with Ontak,
several other DT-containing immunotoxins were evaluated in
preclinical and clinical studies, but today no study is active.
The exotoxin A of the Gram-negative, aerobic Pseudom-
onas aeruginosa bacterium contains 613 amino acid residues
and is composed of three domains. In principle, the same
components are present as in DT, however at different
positions. This time, the recognition domain is located at
the N-terminus (Ia), followed by the translocation domain
(II). At the C-terminus, the catalytic domain (III) can be
found. Removal of the carboxy-terminal lysine by a plasma
carboxypeptidase exposes the sequence REDL, which will
be later used for intracellular sorting. The toxin binds to low-
density lipoprotein receptor-related protein 1 (CD91) and is
internalized through clathrin-coated pits. After low pH
induced receptor dissociation and conformational change,
furin cleaves the toxin. Reduction of the only disulfide bridge
separates the binding domain from the still connected cata-
lytic and translocation domain that continues with traveling
to the trans-Golgi network. Binding of the exposed REDL to
the KDEL receptor sorts the fragment to the endoplasmic
reticulum. There the translocation domain mediates the
escape and brings the catalytic domain to the cytosol where
it inactivates EF2 in a similar manner as DT [15].
The majority of immunotoxins is nowadays based on PE.
Currently, most advanced in clinical Phase III is cintredekin
besudotox. This immunotoxin is a combination of interleu-
kin 13 (IL-13) with a de-immunized PE variant. It targets
cells expressing receptors for IL-13 in glioblastoma [16].
A different toxic mechanism is executed by anthrax
toxin(AnTx),whichissecretedbytheGram-positive,
spore-forming bacterium Bacillus anthrachis.AnTxis
composed of three large separately encoded proteins:

protective antigen (PA), lethal factor (LF), and edema
factor (EF). After binding of PA to its integrin-like cell
surface receptor, a 20-kDa fragment is cleaved off by furin,
co-localized at the surface. The still bound large fragment
of PA assembles into a pore-like homo-heptamer that
associates with three molecules of LF and/or EF. This
complex reaches the early endosome after clathrin-
mediated endocytosis. The lowering of the pH leads to a
conformational change of the pore-like structure into a real
pore through which LF and/or EF translocate to the cyto-
plasm. Both translocated proteins have toxic activity: LF
has a protease activity that degrades MEK kinases while
EF as adenylate cyclase increases cAMP concentration,
ultimately leading to cell death [17].
Anthrax toxin can be used to design novel binary toxins.
The first component comprises a modified protective antigen
in which the original furin cleavage motif is replaced by
another protease motif, for instance that of urokinase plas-
minogen activator (uPA). This motif was selected because
uPA is heavily expressed on the surface of lung cancer cells
and mostly absent on normal cells. The second part consists
of the ADP-ribosylation domain of Pseudomonas exotoxin
A fused to truncated LF to enable translocation through the
pore. A higher specificity is obtained by including a cancer
cell specific cleavage motif and by separating binding and
catalytic component to become active only when interacting
properly inside the cell [18]. This concept has been
employed in combination to other protease motifs (e.g.,
MMP2 and MPP9 [19]) or with other bacterial toxins (Shiga
toxin or diphtheria toxin [20]) fused to the truncated LF
domain. Engineering of the furin cleavage site contributes
also to an up to 20-fold improved circulating half-life. So far,
binary anthrax toxins have been very potent against a
number of tumor types in animal models.
A totally different approach is taken to apply botulinum
neurotoxin (BT) from the anaerobic, Gram-positive bacteria
of the genus Clostridium, specifically of strains of
botulinum. The toxin consists of a heavy and a light chain.
The heavy chain includes two domains: one to bind surface
molecules on neurons, another to enable translocation into
the cytosol. The light chain contains the catalytic domain,
which encodes for an endopeptidase that cleaves soluble N-
ethylmaleimide-sensitive factor attachment protein receptor
(SNARE) proteins, thus inhibiting vesicular secretion and
ultimately neurotransmission. Different BT serotypes cleave
different SNARE proteins. Heavy and light chains are
connected via noncovalent interactions and a covalent
disulfide bridge. The mechanism of BT activity starts
with binding of the heavy chain to polysialogangliosides
at the presynaptic terminal. After internalization and acidi-
fication of the endosome, the translocation domain forms a
pore through which the light chain escapes into the cytosol.
There the endopeptidases specifically cleave SNARE pro-
teins, preventing SNARE complex assembly and exocytosis.