FUSION PROTEINS WITH TOXIC ACTIVITY - Fusion Protein Technologies for Biopharmaceuticals: Applications and Challenges - page 257

many positive examples, this approach suffers from manu-
facturing difficulties and cost because often, the toxic
component reduces expression levels. Furthermore, toxic
proteins of nonhuman origin will cause an immune response
that over time will eliminate the fusion proteins through
neutralizing antibodies. All these disadvantages can be
circumvented when instead of large molecules small pep-
tides are used. For this purpose, toxic peptides are needed,
which can be identified from nature or artificially designed
and synthetically produced. Let us focus on natural peptides
found in insects, arthropods, spiders, reptiles, or amphibians.
Recently, insect toxins such as melittin became exploited
as anti-proliferative agents. Melittin is the major component
of bee venom. It is expressed as a 70 amino acid long
promelittin, which is maturated through proteolytic removal
of the amino terminus, resulting in a 26 residue long
amphiphilic active peptide. This small size is one of the
advantages of melittin. The peptide has a strong anti-prolif-
erative effect in many cancer cell lines, which is probably
based on the inhibition of calmodulin, the Akt pathway, and
the stimulation of phospholipase A2. Furthermore, melittin
can also disrupt cell membranes. The ability of proteolytic
activation has been utilized to generate a number of targeted
fusion proteins. One construct targeting tumor cells express-
ing integrin a v b 3 contained uPA cleavable linker and suc-
cessfully inhibited proliferation of uPA expressing cells
[98]. In a different approach, melittin was connected through
a protease-sensitive linker with a latency peptide of TGF- b
that shielded the cytotoxic activity. The proteolytic cleavage
of the linker by tumor-associated proteases released free
melittin that killed the respective cancer cells [99].
An example for an artificial peptide is a lytic peptide
consisting of lysine and leucine repeats that selectively
destroys cancer cells by disrupting the cell membrane.
Targeting was enabled by fusion to a second peptide that
binds to the transferrin receptor (TfR). Overexpression of
TfR is typical for a number of cancer cell types. Exposure of
12 cancer cell lines to this synthetic 32 AA fusion peptide
showed a clear dose-dependent response with an IC 50 of
4-9 m M. Cell killing was a function of TfR expression level.
Normal cell lines did not show cell death up to a concentra-
tion of 40 m M. Membrane disruption was fast with more
than 80% dead cells after 60min [100]. The same toxic
peptide was used in a similar approach, now fused to an
EGFR targeting peptide. This fusion peptide did also kill
cancer cells that were resistant to tyrosine kinase inhibitors.
The observed IC 50 values ranged from 6 to 12 m M. In these
experiments, it could be verified that the mechanism of cell
killing is apoptosis. In vivo administration of this fusion
peptide drastically reduced tumor size in three different
xenograft models while having no negative influence on
animal health, in general. The synthetic peptide could not be
degraded by serum proteases owing to its unique diastereo-
metric conformation [101].

17.8 CONCLUSIONS AND FUTURE
PERSPECTIVES

Specific and long-lasting eradication of malignant cells,
which is the holy grail of targeted therapy, has not been
achieved yet. But in the last four decades, since the success-
ful introduction of antibodies and their derivates, many new
weapons have been added to the arsenal. Fusion proteins can
contribute heavily in generating novel combinations of
targeting moieties and toxic molecules. From a targeting
perspective, the choice is between well-established specific
natural ligands and novel designer molecules based on
engineered scaffolds. Similarly, the eliminating function
can be derived from natural toxic proteins or factors that
trigger cell death either through their endogenous function
or by attracting cells from the immune system specialized in
eradicating malignant cells. Fusion proteins give new
options by allowing novel combinations of well-proven
molecules or integrating novel activities.
Despite the enormous progress and the huge body of
knowledge, a number of issues still remain to be solved.
The highly potent targeted therapies based on natural toxins
that frequently require only a single molecule to reach the
cytosol to efficiently kill a cell still need improvement
with regard to immunogenicity to allow several therapeutic
cycles without eliciting neutralizing antibodies. On the other
hand, replacing foreign proteins with human enzymes is
limited by the required escape from the endosomes.
Very promising seems the approach to attract immune cells
by fusion proteins containing leukocyte specific ligands or
Fc domains interacting with their specific receptors. Immune
therapy might also provide long-term protection from
recurring and relapsing cancers or metastases by creating
memory cells. This concept is one of the driving forces that
led to the invention of bi- or multi-specific antibodies that
are explained later in this topic in part IIIb. Good results have
been obtained by addressing hematological diseases because
single cells are much easier to reach by large molecules than
the hidden cells in the center of a solid tumor. Probably,
combination therapies mixing small molecule drugs with
engineered therapeutic fusion proteins could be a solution in
the near future. Alternatively, smaller variants with tailored
tissue penetration capability and optimized half-life can be
engineered.

REFERENCES

1. Strebhardt K, Ullrich A. (2008) Paul Ehrlich's magic bullet
concept: 100 years of progress. Nat. Rev. Cancer 8(6),
473-480.
2. Cizeau J, Torres MGP, Cowling SG, Stibbard S, Premsukh A,
Entwistle J, et al. (2011) Fusogenics: a recombinant immu-
notoxin-based screening platform to select internalizing