Biochemical Pharmacology xxx (2017) xxx–xxx

Contents lists available at ScienceDirect

Biochemical Pharmacology
journal homepage: www.elsevier.com/locate/biochempharm

Commentary

Repurposing bacterial toxins for intracellular delivery of therapeutic

proteins
Greg L. Beilhartz a, Seiji N. Sugiman-Marangos a, Roman A. Melnyk a,b,⇑
a
Molecular Medicine Program, The Hospital for Sick Children Research Institute, 686 Bay Street, Toronto, ON M5G 0A4, Canada
b
Department of Biochemistry, University of Toronto, 1 King’s College Circle, Toronto, ON M5S 1A8, Canada

a r t i c l e i n f o a b s t r a c t

Article history: Despite enormous efforts, achieving efficacious levels of proteins inside mammalian cells remains one of
Received 9 March 2017 the greatest challenges in biologics-based drug discovery and development. The inability of proteins to
Accepted 7 April 2017 readily cross biological membranes precludes access to the wealth of intracellular targets and applica-
Available online xxxx
tions that lie within mammalian cells. Existing methods of delivery commonly suffer from an inability
to target specific cells and tissues, poor endosomal escape, and limited in vivo efficacy. The aim of the pre-
Keywords: sent commentary is to highlight the potential of certain classes of bacterial toxins, which naturally deliver
Bacterial toxins
a large protein into the cytosolic compartment of target cells after binding a host cell-surface receptor
Intracellular protein delivery
Biologics
with high affinity, as robust protein delivery platforms. We review the progress made in recent years
Protein therapeutics toward demonstrating the utility of these systems at delivering a wide variety of protein cargo, with spe-
Diphtheria toxin cial attention paid to three distinct toxin-based platforms. We contend that with recent advances in pro-
tein deimmunization strategies, bacterial toxins are poised to introduce biologics into the inner sanctum
of cells and treat a wealth of heretofore untreatable diseases with a new generation of therapeutics.
Ó 2017 Elsevier Inc. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00
2. Existing delivery platforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00
3. Toxin cell entry mechanisms: three distinct strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00
4. Delivery of proteinaceous cargo by AB toxins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00
5. Practical considerations for delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00
6. Re-directing toxins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00
6.1. Cellular tropism. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00
6.2. Tissue tropism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00
7. Future directions/challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00
7.1. Receptor targeting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00
7.2. Choice of cargo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00
7.3. Immunogenicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00
7.4. Platform optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00
8. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00

⇑ Corresponding author at: Research Institute, The Hospital for Sick Children, 686 Bay Street, Toronto, ON M5G 0A4, Canada.
E-mail address: roman.melnyk@sickkids.ca (R.A. Melnyk).

http://dx.doi.org/10.1016/j.bcp.2017.04.009
0006-2952/Ó 2017 Elsevier Inc. All rights reserved.

Please cite this article in press as: G.L. Beilhartz et al., Repurposing bacterial toxins for intracellular delivery of therapeutic proteins, Biochem. Pharmacol.
(2017), http://dx.doi.org/10.1016/j.bcp.2017.04.009
2 G.L. Beilhartz et al. / Biochemical Pharmacology xxx (2017) xxx–xxx
1. Introduction
In the past 30 years, since the approval of recombinant human
insulin, protein-based therapeutics have grown in prominence
and are now among the fastest-growing class of drugs [1]. In
2015, 13 new biologics were approved by the FDA, representing
29% of all approved drugs [2]. In the same year, seven of the top
10 best selling drugs were biologics, led by AbbVie’s Humira, which
brought in $14 billion (USD) in sales [3]. Therapeutic proteins and
peptides comprise 85% of the approximately 300 biologics that
are currently approved for clinical use [4]. Despite this dramatic
expansion in large-molecule therapeutics, the full potential of pro-
teins as drugs is constrained by their inability to cross biological
membranes and enter cells. As such, approximately 60% of all
human proteins – and the entire human genome – remain inacces-
sible to protein-based drugs [5]. Intracellular targets with a clear
therapeutic rationale, including protein-protein interactions, large
protein complexes, and important targets that are notoriously dif-
ficult to target with small molecules such as p53 and RAS [6,7], are
best suited to biologics-based therapies. Moreover, enzymes with
tremendous therapeutic potential including genome editing
machinery such as CRISPR/Cas9 or TALENs must be delivered as
nucleic acids or electroporated into cells, complicating their use
as therapeutics.
To solve this issue, several classes of protein carriers have been
explored to circumvent the barrier posed by the plasma mem-
brane, however, to date, no single platform has been identified that
is simultaneously efficient, safe, versatile and specific. We contend
that bacterial toxins, which bind specific cells and can efficiently
transport a broad spectrum of proteins to the cytosol have great
promise as platforms for delivering protein cargo into specific cells
in vitro and in vivo. Notably, the one class of biologics that enter
cells as part of their mechanism is immunotoxins, which use the
cell-penetrating machinery of bacterial toxins. The field of protein
delivery has greatly matured in the past several years, and novel
protein engineering techniques combined with improved methods
for definitively detecting cargo in the cytosol have set the stage for
bacterial toxins to emerge as a major platform to deliver the next
generation of biologics.
2. Existing delivery platforms
Cell-penetrating peptides (CPPs) have been used to deliver pro-
teins of various sizes, as well as diverse nucleic acids and even lipo-
somes and nanoparticles [8,9]. As a group, CPPs generally show
varying efficacy based on the CPP used, the cell type to be targeted,
and even buffer conditions [10]. The mechanism by which CPPs
enter cells has not been fully elucidated, however there is evidence
that CPPs can either directly cross the plasma membrane, or gain
entry via endocytosis and subsequent escape from endosomes
[11]. One or both of these mechanisms may be used depending
on concentration, temperature, and the choice of counter ion
[11]. Engineered supercharged proteins such as +36 GFP are highly
positively charged and appear to transport protein and nucleic acid
cargo into cells by a mechanism that is similar to CPPs [12]. While
generally effective in vitro, CPPs suffer from a lack of cell/tissue
specificity, which contributes to variable efficacy in vivo [13,14].
Virus-like particles (VLPs) are a loosely-defined group of self-
assembling structures based on viral proteins. VLPs of diverse
structures can be assembled from various mammalian, plant,
microbial and insect viruses, and are replication deficient, as they
lack any genetic material [15]. VLPs are able to encapsulate protein
cargo fused with viral proteins, or express them on their surface as
is often done in the case of VLP-based vaccines [16]. It has been
demonstrated that certain enveloped VLPs (HSV-1, HIV-1, etc.)
Please cite this article in press as: G.L. Beilhartz et al., Repurposing bacterial toxins for intracellular delivery of therapeutic proteins, Biochem. Pharmacol.
(2017), http://dx.doi.org/10.1016/j.bcp.2017.04.009
deliver their encapsidated cargo to the cytosol by direct membrane
fusion, whereas other non-enveloped viruses have complex endo-
somal escape mechanisms that are not completely understood
[17,18]. While VLPs are perfectly suited for vaccine development,
their strong induction of both humoral and cellular immune
responses likely restricts their in vivo use [15].
Liposomes and nanoparticles are broadly similar carrier sys-
tems that have been explored for delivery. Liposomes and
nanoparticles typically do not have built-in mechanisms for endo-
somal escape, although attempts have been made to decorate the
surface of nanoparticles or liposomes with CPPs or fusogenic pep-
tides/lipids to increase cytosolic delivery [19,20]. While both
classes can be targeted to specific cells by displaying antibodies
or other targeting moieties on their surface, achieving high surface
densities can be challenging, and even when successful can result
in faster plasma clearance [21,22]. In vivo toxicity of nanoparticles
has been described depending on the nanomaterials involved; lipo-
somes generally do not show overt toxicity in animal models but
can be genotoxic even at low doses, as well as being immunogenic
to varying degrees based on the specific lipids used [23,24].
Each of the above protein delivery systems possess one or more
of the ideal attributes, however, no single system has the full com-
plement of the features that would be desired in an idealized deliv-
ery platform. Bacterial toxins, which have evolved the ability to
bind host cell receptors, and subsequently escape from endosomes
to the cytosol of mammalian cells, are replete with features that
make them attractive candidates for delivery vectors. Anthrax
toxin (B. anthracis), exotoxin A (P. aeruginosa) and diphtheria toxin
(C. diphtheriae) are three different bacterial toxins that use three
distinct strategies to deliver their catalytic cargo into cells. The
modular structure of these toxins, with three exchangeable func-
tional domains, makes them particularly amenable to protein engi-
neering. The major perceived limitation of using toxins in humans
is the immune response, however recent work discussed herein
suggests that this is manageable and no longer a hurdle to
development.
3. Toxin cell entry mechanisms: three distinct strategies
Bacterial exotoxins are secreted proteins that are designed to
specifically bind, enter, and damage host tissue. The so-called
‘‘AB class” of toxins are of interest to the field of protein delivery,
as they are autonomous delivery vectors; that is, AB toxins possess
all the components necessary to gain access to the cytosol inde-
pendent of any other bacterial machinery. Members of this class
all minimally consist of an enzymatic A moiety and a B moiety that
both binds a cell-surface receptor and mediates entry into cells.
The A and B fragments are typically linked by a furin-cleavable lin-
ker and a disulfide bond [25]. AB toxins bind to a cell surface recep-
tor, which triggers endocytosis, transporting the toxin into the
early endosome for sorting. From here, some toxins are trafficked
to the endoplasmic reticulum (ER) where they undergo retrograde
translocation, while others escape directly from the endosome
(Fig. 1).
Among the toxins that can escape directly from endosomes, two
distinct classes of AB toxins – exemplified by anthrax toxin and
diphtheria toxin – are observed. In the case of anthrax toxin-like
platforms, the A and B fragments are independent entities that
associate non-covalently upon oligomerization of the B-
fragments, whereas for diphtheria toxin-like platforms, the A and
B fragments are encoded on the same polypeptide. The transloca-
tion pores created by these two types of toxins are also distinct.
Whereas the anthrax toxin pore spans the endosomal membrane
as a rigid b-barrel, diphtheria toxin creates flexible a-helical pores
across the membrane. Importantly, the properties of these two
 G.L. Beilhartz et al. / Biochemical Pharmacology xxx (2017) xxx–xxx
types of pores have great influence on the types of cargo that can
be transported by these toxins (vide infra).
4. Delivery of proteinaceous cargo by AB toxins
Anthrax toxin is perhaps the best characterized delivery plat-
form, having been demonstrated to co-deliver over 30 different
molecular cargo fused to the carboxy-terminus of LFN – the
amino-terminal 263-amino acid fragment of the anthrax toxin A-
domain (lethal factor) that recognizes the anthrax toxin pore
[26]. Examples of cargo include peptides, other toxin effector pro-
teins, non-canonical peptides, antibody-like molecules, and
oligonucleotides [27–29]. Not all cargos are delivered equally, as
a comparison of anthrax toxin and Pseudomonas exotoxin A
showed that Pseudomonas exotoxin A could deliver up to three
consecutive designed Ankyrin repeat proteins (DARPins) with
equivalent efficacies, whereas anthrax toxin could only deliver a
single Ankyrin domain (TM = 60 °C) [30]. Due to the rigid nature
of the anthrax toxin pore, and the lack of external factors driving
unfolding and translocation of cargo on the endosomal side, the
two more stably folded cargos (TM > 90 °C) could not be translo-
cated by anthrax. This inability to deliver stably folded cargo rep-
resents the principal weakness of the anthrax platform, and
precludes delivery of tight folders, proteins containing internal
disulfide bonds, proteins stabilized by bound ligands, and cyclic
peptides [26,31,32]. Pseudomonas exotoxin A does not appear to
have the same limitations as anthrax toxin when it comes to stably
folded cargo, presumably due to the presence of ER chaperones
that facilitate unfolding and disulfide bond reduction. Likewise,
diphtheria toxin has been demonstrated to efficiently translocate
a wide variety of protein cargos including the highly thermo-
and pH-stable fluorophore mCherry [33].
5. Practical considerations for delivery
The quantity of delivered cargo is one of the most important
variables in defining the efficiency of a delivery platform. Each
stage of the delivery process influences the amount of payload that
reaches the cytosol: receptor binding, internalization, furin cleav-
age, trafficking, translocation and disulfide reduction and release.
The biological effect of cargo is fundamentally tied to the amount
reaching the cytosol. Overexpression of prospective cargo in cell
culture is the only positive control available to demonstrate biolog-
ical response, making the interpretation of results ambiguous. Is
the protein making it to the cytosol in sufficient quantities? How
much protein is needed to elicit the same response as constitutive
overexpression? Unfortunately, one of the major impediments in
the field of protein delivery is lack of standardized methods for
quantifying cytosolic cargo. Microscopy of fluorescent cargo is a
strategy that can be used to visualize entry; however, great care
needs to be taken to account for signal coming from the vast num-
ber of endosomes that can confound the interpretation of the
amount of cargo reaching the cytosol. While delivery of therapeu-
tic cargo is normally done in the context of a non-toxic variant of
the toxin delivery platform (i.e., the catalytic A-domain is mutated
to be inactive), a commonly used technique, particularly among
those that work with bacterial toxins, is the inhibition of protein
synthesis by the wildtype diphtheria toxin A-domain fused to the
cargo protein under investigation. Toxicity as a proximal readout
is a sensitive indicator of cytosolic delivery of cargo due to the
extremely potent nature of DT-A, however, it is less useful for
quantifying the number of cargo molecules delivered [33]. Meth-
ods based on cell fractionation using detergents and western blot-
ting have been used by many [28,34], however, these methods are
semi-quantitative and can be technically difficult, given the fine
Please cite this article in press as: G.L. Beilhartz et al., Repurposing bacterial toxins for intracellular delivery of therapeutic proteins, Biochem. Pharmacol.
(2017), http://dx.doi.org/10.1016/j.bcp.2017.04.009
3
Fig. 1. Cellular entry mechanisms of bacterial toxins. Diphtheria toxin, anthrax
toxin and Pseudomonas exotoxin A bind to their specific receptors through their
receptor-binding domains (R), resulting in endocytosis and furin cleavage [83,84].
In contrast, anthrax toxin is furin-processed on the cell surface, which is required
for oligomerization of the B fragment and subsequent endocytosis [85]. The anthrax
toxin A-domain non-covalently associates with the B fragment heptamer, whereas
for diphtheria toxin and Pseudomonas exotoxin A, the A and B fragments remain
linked via a disulfide bridge after furin cleavage. Anthrax and diphtheria toxins are
trafficked to late endosomes, and undergo a conformational change and membrane
insertion of the transmembrane (T) domain in response to the low endosomal pH
[86–88]. Both toxins then translocate their catalytic A domains through the pore
and into the cytosol [89,90]. The reducing environment of the cytosol allows for the
release of the disulfide-linked diphtheria toxin A domain [34]. Pseudomonas
exotoxin A avoids the endolysosomal pathway by being trafficked to the ER via the
trans-Golgi network [91]. In this case, the disulfide linkage is broken by protein
disulfide isomerase (PDI) [92], allowing the A domain to be recognized and
unfolded by ER chaperones [93]. It is thought that the A domain is retrotranslocated
through the sec61 translocon into the cytosol [94].
line between solubilizing plasma and endosomal membranes
[35]. The method described by Verdumen et al. [30] shows promise
due to its versatility, as it can be easily applied to any protein cargo
through the addition of a 15-residue avi-tag. Techniques limited by
the semi-quantitative nature of western blotting can often be
improved by mass spectrometry if quantification is critical.
The need for a reliable quantification method was illustrated
previously with anthrax toxin [28]. Despite the efficient delivery
of several antibody-like cargos as demonstrated by protein synthe-
sis inhibition and western blot, the authors noted that the intracel-
lular effects were quite modest, even at the highest concentrations
tested. For example, HA4-7c12 is a tandem monobody that binds
with nanomolar affinity to the SH2 domain of the oncoprotein
Bcr-Abl [36]. While HA4-7c12 was shown to strongly inhibit Bcr-
Abl activity and induce apoptosis when transiently expressed in
K562 cells, treatment of cells with high doses of LFn-HA4-7c12
only elicited a moderate effect.
It seems unlikely that increasing binding affinity further would
drastically improve the effect, which raises the questions of how
much monobody is being delivered, and how much is required?
While this result demonstrates the proof-of-concept, it raises
another important aspect of protein delivery in general. Namely,
is it feasible to stoichiometrically target highly abundant intracel-
lular proteins? The answer to this question remains unclear, but it
would seem a high bar to set for any exogenous protein therapeu-
tic. Attempting to overtake host gene expression by delivering sto-
ichiometric amounts of exogenous protein would likely require
both a very high serum drug concentration, and an extremely
4 G.L. Beilhartz et al. / Biochemical Pharmacology xxx (2017) xxx–xxx
efficient delivery mechanism. Fortunately, enzymes as biologic
drugs can massively leverage their effects beyond simple target
binding, more than compensating for their lower intracellular
doses. Though the development of protein therapeutics has been
so far focused on antibodies and stoichiometric binding of a target
protein, the untapped potential of biologics lies in the delivery of
powerful therapeutic enzymes.
AB toxins all evolved highly efficient mechanisms to translocate
their cytotoxic cargo, however the cargo enzymes themselves are
also extremely potent. eEF-2, the target of both diphtheria toxin
and Pseudomonas exotoxin A, is estimated to be present on the
order of 107 molecules/cell [37], yet it is often claimed that a single
molecule of diphtheria toxin A-domain is sufficient to kill a cell
[38]. While the number of molecules that reach the cytosol during
an actual intoxication event is undoubtedly far greater, it does
underscore the power of signal amplification inherent to enzy-
matic cargo. For example, assuming the volume of a mammalian
cell to be 1–10 pL, the intracellular concentration represented
by a single molecule per cell is 10–100 fM [39]. Indeed, the best
examples of non-native cargo that have been delivered by bacterial
toxins with demonstrable biological effects are enzymes, such as
bacterial effector domains [28,40], b-lactamase [41], DHFR [32],
and a-amylase [33]. Regardless of the delivery system used, effort
spent optimizing the properties and potency of enzyme cargo may
be equally or even more beneficial than increasing the number of
delivered enzyme molecules.
6. Re-directing toxins
6.1. Cellular tropism
One of the principal advantages of toxin-based platforms is the
specificity gained by targeting a unique cell-surface receptor.
Moreover, due to their modular structure, both Pseudomonas exo-
toxin A [30,42] and diphtheria toxin [43,44] can be re-targeted
by replacing their receptor binding domains with a ligand (natural
or engineered) to a different receptor. Recent efforts have been
made to similarly re-target the B-moiety of anthrax toxin [30,45],
although this is complicated by the fact that removal of the
receptor-binding domain has consequences for proper PA folding.
The most notable examples of successful retargeting of toxins
come from the immunotoxin field. Many have recognized and
exploited the ability of diphtheria toxin to tolerate the replacement
of its C-terminal receptor-binding domain with cancer-targeted
antibodies and ligands. The FDA-approved Ontak (Denileukin difti-
tox) is a re-targeted diphtheria toxin variant with the cytokine IL-2
fused to the C-terminus of the DT1-389 fragment. This re-targets the
toxic A-fragment to cells expressing the high-affinity IL-2 receptor
(CD25, CD122 and CD132 trimer), which is often overexpressed on
leukemia cells. Similarly, DT1-389 fused to epidermal growth factor
(EGF) was developed for the targeting of EGFR-overexpressing
glioblastoma cells, and showed synergistic effects with IL13 fused
to Pseudomonas exotoxin (PE-IL13) [46,47]. Diphtheria toxin fused
with tandem single-chain antibodies against CD3 (Resimmune) has
shown efficacy in patients with cutaneous T-cell lymphoma [48].
In a further improvement on the same design, a bispecific
immunotoxin was developed by the fusion of tandem single-
chain antibodies targeting different cell surface receptors, in this
case CD19 and CD22 [49]. This construct has shown activity in pre-
clinical models and is entering clinical trials. Pseudomonas exotoxin
A-based immunotoxins are also in the clinic, including moxetu-
momab pasudotox which re-targets Pseudomonas exotoxin A with
an anti-CD22 antibody fragment and LMB-2, which targets CD25-
expressing cells [50,51]. Although diphtheria toxin can accept a
wide variety of targeting domains on its C-terminus, it must be
Please cite this article in press as: G.L. Beilhartz et al., Repurposing bacterial toxins for intracellular delivery of therapeutic proteins, Biochem. Pharmacol.
(2017), http://dx.doi.org/10.1016/j.bcp.2017.04.009
noted that not all of these moieties will be effective and some opti-
mization may be required (vida infra).
6.2. Tissue tropism
In addition to determining cell type specificity, the targeting
moiety also determines tissue tropism. This is particularly relevant
for biologics developed to treat tissue-specific diseases such as
osteosarcomas or neurological diseases such as Alzheimer’s,
Parkinson’s or lysosomal storage diseases (LSDs). LSDs result from
a mutation(s) that typically result in a non-functional protein
involved in the degradation of lysosome contents. These diseases
are commonly treated by enzyme replacement therapy (ERT),
where a recombinant version of the missing protein is injected into
patients. Recombinant enzymes enter cells via the mannose-6-
phosphate receptor (M6PR) and are subsequently trafficked to
lysosomes [52]. Efficacy is limited by M6PR as it is either absent
or poorly expressed in several important tissue types including
bone, muscle and the brain vasculature [53]. The absence of
M6PR on the blood-brain-barrier (BBB) is particularly problematic
as 75% of LSDs manifest with neurological symptoms [54]. The
BBB presents a formidable barrier to delivery of biologics into the
brain, as anything larger than small molecules are unable to diffuse
across.
Some receptors have the potential to be exploited for brain pen-
etration including receptors for low-density lipoprotein, transfer-
rin and insulin. The use of antibodies or antibody-like molecules
targeting these proteins for receptor-mediated transcytosis is an
area of active research [55–57]. Remarkably, diphtheria has been
shown to cross the BBB naturally [58–60], via HBEGF-mediated
transcytosis of brain endothelial cells. Current methods of deliver-
ing biologics into the brain involve either intrathecal or intracere-
broventricular injection directly into the brain [61,62]. The ability
of systemically administered protein drugs to cross the BBB and
access the neuronal cytosol would be a marked improvement on
current techniques.
7. Future directions/challenges
Bacterial toxins are among the best hopes for delivery of thera-
peutic proteins to the cytosol of human cells and the treatment of
many diseases and conditions. Their efficacy and versatility allow
for the conjugation of an immense variety of possible cargo. Direct
comparison of the delivery efficiencies of anthrax toxin and a CPP
revealed that the anthrax-based system was three orders of magni-
tude more efficient when delivering various antibody-like scaffolds
[28]. Diphtheria toxin has similarly been shown to be superior to
CPPs at delivering enzymes into cells [33]. There are points to keep
in mind, however, when designing an intracellular therapeutic
with bacterial toxins.
7.1. Receptor targeting
An important consideration for changing the receptor speci-
ficity for toxins is problems that may arise from receptor ago-
nism/antagonism. Furthermore, it should not be assumed that
the novel receptor will be internalized in response to binding, or
that trafficking of the new receptor would be the same as the
native one [63]. For example, HER2 (a common target in cancer
therapy due to its overexpression in many cancer types) has a
low rate of internalization in addition to a rapid recycling to the
plasma membrane from early endosomes [64]. This might suggest
that it would make a poor target for receptor-mediated endocyto-
sis of a toxin-based therapeutic, yet Affitoxin (an Pseudomonas exo-
toxin A-based immunotoxin) appears to be reasonably potent on
 G.L. Beilhartz et al. / Biochemical Pharmacology xxx (2017) xxx–xxx
HER2 overexpressing cells, suggesting that rate of internalization is
not necessarily a key attribute of a ‘‘good” ligand-receptor interac-
tion [42]. So, how to best interrogate a novel receptor? To answer
this question, a synthetic antibody library was screened to identify
92 HER2-specific single-chain Fvs (scFv) [65]. These were non-
covalently associated with a truncated form of Pseudomonas exo-
toxin A to form either monovalent (one scFV) or bivalent (tandem
scFvs) immunotoxins. They showed that bivalent arrangements
were more potent than monovalent ones for reasons beyond avid-
ity, as the effect varied among the scFvs. Interestingly, they also
found that binding affinities and kinetics of the scFvs were not cor-
related with cytotoxicity. Instead, the specific epitope and paratope
of the scFvs were deemed to be particularly important, as they
relate to the pH-dependence of binding and the downstream traf-
ficking of the toxin [65]. This unexpected result could be very use-
ful when re-targeting toxins for protein delivery.
Another way of changing the cell specificity of diphtheria toxin
is by replacing the furin cleavage site with a recognition sequence
for a cell-specific protease. Abi-Habib et al. replaced the furin site
in a diphtheria toxin-based immunotoxin with a sequence cleav-
able by urokinase plasminogen activator (uPA), a technique origi-
nally used for re-targeting anthrax toxin [66,67]. UPA is a serine
protease that is overexpressed in leukemias and other tumors,
and the new immunotoxin displayed greater toxicity to cells
expressing uPA.
Cell specificity can also be attained by the choice of cytosolic
target. The oncogenic bcr-abl fusion protein is a result of a chromo-
somal translocation in chronic myelogenous leukemia, among
others. Anthrax toxin was modified to deliver an LFN-monobody
targeting the bcr-abl fusion protein [28]. While the fusion toxin
itself is not cell specific, malignant cells will be disproportionally
sensitive to the monobody cargo. Combining some or all the above
techniques may confer exquisite target specificity to these protein
delivery vehicles, including the next generation of cancer
immunotoxins.
7.2. Choice of cargo
Translocation is typically thought to be the rate-limiting step in
toxin delivery. In immunotoxin applications where the cargo is a
cytotoxic enzyme, cell specificity outweighs efficiency, as only a
small quantity of cargo is required to kill cells. Other cargo may
require that more molecules get delivered to elicit the desired
response, as intracellular target proteins vary greatly in abundance
from many millions of copies, to a dozen or fewer [39]. As such,
acknowledgement of the limitations of bacterial toxins is required
to determine the most appropriate intracellular targets, cargo, and
applications. For example, as mentioned above, anthrax toxin is
not able to deliver cargo that cannot completely unfold on its
own, while Pseudomonas exotoxin A and diphtheria toxin do not
have the same limitation. While Pseudomonas exotoxin A (and pre-
sumably attached cargo) is unfolded by chaperones and isomerases
prior to translocation from the ER, diphtheria toxin appears to be
able to translocate cargo in a folded state [33]. Except for enzymes
involved in redox pathways, cytosolic proteins rarely contain disul-
fide bonds [68], so care should be taken with cargo proteins that
contain internal disulfide bonds, as the reducing environment of
the cytosol is not conducive to their proper folding. In all likeli-
hood, the cargo with the greatest chance of achieving a therapeutic
effect at the lowest doses is likely to be highly active enzymatic
cargo that is easily folded and stable in the mammalian cytosol.
7.3. Immunogenicity
The development of neutralizing antibodies against the toxin
carrier is the biggest hurdle to their development as delivery vec-
Please cite this article in press as: G.L. Beilhartz et al., Repurposing bacterial toxins for intracellular delivery of therapeutic proteins, Biochem. Pharmacol.
(2017), http://dx.doi.org/10.1016/j.bcp.2017.04.009
5
tors. Many applications for protein delivery require multiple, even
lifetime injections. Clearly, a host immune response is to be
avoided. Toxin-based protein carriers are not alone in this, as the
FDA has suggested that drug developers proactively address the
issue of immunogenicity and the rise of anti-drug antibodies
(ADAs) when it comes to biologics [69]. Fortunately, recent
advances hold promise that the immunogenicity problem is not
as severe as it once was. In the case of the immunotoxin denileukin
diftitox, although ADA titers increased after treatment, they had no
effect on clinical outcome [70]. The lack of a neutralizing antibody
response has been also observed for Pseudomonas exotoxin A-
based immunotoxins, and appears to be a feature of hematological
malignancies [50]. While this can be beneficial for the treatment of
blood cancers, the repeated use of diphtheria toxin as a protein
delivery system for solid-tumor immunotoxins or ERT would
require reducing the immunogenicity of diphtheria toxin itself.
Several examples of successful deimmunization involving the
identification of T and/or B cell epitopes, and subsequent mutage-
nesis to remove them while retaining protein function are now
available [71,72]. Several groups have successfully developed
deimmunized proteins, including diphtheria toxin while retaining
catalytic activity [73,74]. Instead of masking the toxin from the
immune system, another method is to induce immune tolerance
through induction of regulatory T cells (Tregs) [75]. Recently it
has been shown that the formation of ADAs can be prevented by
co-administering rapamycin-containing nanoparticles with an
antigen. This induces tolerogenic dendritic cells, increases Tregs,
and reduces B cell activation [76]. This is a promising approach,
as it promotes immune tolerance of the antigen, rather than
attempting to hide it from the immune system. In a similar vein,
the pre-administration of pentostatin and cyclophosphamide,
which depletes host T and B cells was shown to delay the emer-
gence of neutralizing antibodies against the Pseudomonas exotoxin
A-based immunotoxin SS1P, while improving outcomes in a
mesothelioma trial [77]. While a host immune response to biolog-
ics can reduce their effectiveness, and in severe cases be life-
threatening, it need not be an insurmountable hurdle to clinical
success.
7.4. Platform optimization
While bacterial toxins have evolved the ability to target specific
cells and escape from endosomes into the cytosol, this ability is
uniquely suited to perform its natural function, i.e. to kill cells. This
evolutionary optimization of toxins as dealers of cell death can be
ideal for targeted immunotoxins, however it would appear to be
less suited to other applications. There are several areas where tox-
ins can be improved from a protein delivery standpoint. As dis-
cussed previously, replacing the receptor binding moiety, or
developing a bivalent binding domain can decrease off-target
effects and improve the therapeutic index. The development of
delivery systems that can deliver multiple copies of an effector in
single polypeptide, either by incorporation of release domains or
homo-fusions of cargo enzymes, may increase the effectiveness
of the drug by delivering multiple cargos per receptor binding
event. It is also desirable to improve the efficiency of toxin translo-
cation, however an incomplete understanding of the mechanism of
translocation hinders a rational approach. Nonetheless, there are
several mechanistic steps that are required for translocation that
lend themselves to optimization, namely furin cleavage, disulfide
bond reduction and membrane insertion/disruption.
Furin cleavage is absolutely required for release of the cargo
into the cytosol, however in certain instances high receptor density
and low furin expression can make cleavage rate-limiting. While
furin cleaves at the consensus RXXR sequence, optimization of
the surrounding sequence can enhance the efficiency of furin
6 G.L. Beilhartz et al. / Biochemical Pharmacology xxx (2017) xxx–xxx
cleavage [78]. This can improve delivery efficiency, as it has been
shown that ‘‘pre-nicking” diphtheria toxin and Pseudomonas exo-
toxin A with furin results in increased toxicity [79]. Similarly, the
intracellular disulfide bond that links the diphtheria toxin A
domain to the translocation domain has been shown to be exposed
only at low pH, and that reduction of this bond by as yet undiscov-
ered mechanisms is the rate-limiting step in diphtheria toxin
intoxication [34].
Despite several studies, the sequence determinants of optimal
membrane insertion/translocation are not well understood for
many toxins, and perhaps an unbiased approach to this problem
is warranted. The incorporation of CPPs into the translocating
polypeptide, for example, may facilitate toxin translocation,
despite the differences in mechanisms. The hydrophobicity of the
transmembrane helices involved in pore formation varies between
toxins, the significance of which is not yet understood, although
the pattern of hydrophobicity is similar across many disparate
types of pore-forming toxins [80].
Certain small molecules have been shown to enhance endoso-
mal release [81,82]. Lysosomotropic agents such as chloroquine
are weak bases that accumulate in endolysosomal compartments
and become protonated. At low doses, they inhibit endosome mat-
uration by raising the lumenal pH, but at high concentrations they
cause an accumulation of counter ions, followed by osmotic swel-
ling and endosome rupture. This has been labeled the ‘‘proton-
sponge effect”. Other methods that cause endosomal rupture,
including photochemical disruption, tend to have mechanism-
based toxicities. Amphipathic peptides derived from viral
sequences can increase endosomal escape, possibly by endosomal
lysis or vesicle fusion [81]. These techniques are likely most useful
in an in vitro setting, as co-dosing of these agents with bacterial
toxins or other delivery systems would be complex.
8. Conclusions
The rapid emergence of biologics as potent therapeutics has ini-
tiated a race to develop platforms to deliver these diverse mole-
cules into cells. While antibodies and their analogs are currently
enjoying unquestioned dominance in the biologics market, the real
potential of proteins as drugs is stifled by the cellular membrane of
human cells. This barrier eliminates almost two thirds of the
human proteome and the entire genome from therapeutic inter-
vention by biologics. The use of bacterial toxins for intracellular
delivery remains an underappreciated approach to gaining entry
to this target space. While the delivery of proteins into specific
cells and tissues is not a trivial task, the available technology and
techniques have progressed to the point where engineering and
examining bacterial toxins as effective delivery vehicles is possible.
We propose that bacterial toxins represent a promising platform
for delivery of intracellular protein therapeutics that should be
explored and considered further.
Acknowledgements
R.A.M. was supported by Canadian Institutes of Health Research
(CIHR) Operating Grant, and a Brain Canada Multi-Investigator
Research Initiative Grant.
References
[1] B. Leader, Q.J. Baca, D.E. Golan, Protein therapeutics: a summary and
pharmacological classification, Nat. Rev. Drug Discov. 7 (1) (2008) 21–39.
[2] FDA, Novel drugs 2015 Summary, in: C.o.D.E.a. Research (Ed.) 2016.
[3] PharmaCompass, Top drugs by sales revenue in 2015: Who sold the biggest
blockbuster drugs?, 2016. http://www.pharmacompass.com/radio-compass-
blog/top-drugs-by-sales-revenue-in-2015-who-sold-the-biggest-blockbuster-
drugs. (Accessed February 23, 2017 2017).
Please cite this article in press as: G.L. Beilhartz et al., Repurposing bacterial toxins for intracellular delivery of therapeutic proteins, Biochem. Pharmacol.
(2017), http://dx.doi.org/10.1016/j.bcp.2017.04.009
[4] R.P. Evens, Pharma success in product development-does biotechnology
change the paradigm in product development and attrition, AAPS J. 18 (1)
(2016) 281–285.
[5] M. Uhlen, L. Fagerberg, B.M. Hallstrom, C. Lindskog, P. Oksvold, A. Mardinoglu,
A. Sivertsson, C. Kampf, E. Sjostedt, A. Asplund, I. Olsson, K. Edlund, E.
Lundberg, S. Navani, C.A. Szigyarto, J. Odeberg, D. Djureinovic, J.O. Takanen, S.
Hober, T. Alm, P.H. Edqvist, H. Berling, H. Tegel, J. Mulder, J. Rockberg, P.
Nilsson, J.M. Schwenk, M. Hamsten, K. von Feilitzen, M. Forsberg, L. Persson, F.
Johansson, M. Zwahlen, G. von Heijne, J. Nielsen, F. Ponten, Proteomics. Tissue-
based map of the human proteome, Science 347 (6220) (2015) 1260419.
[6] A.D. Cox, S.W. Fesik, A.C. Kimmelman, J. Luo, C.J. Der, Drugging the
undruggable RAS: mission possible? Nat. Rev. Drug Discov. 13 (11) (2014)
828–851.
[7] K.H. Khoo, C.S. Verma, D.P. Lane, Drugging the p53 pathway: understanding
the route to clinical efficacy, Nat. Rev. Drug Discov. 13 (3) (2014) 217–236.
[8] T. Lehto, K. Ezzat, M.J. Wood, S. El Andaloussi, Peptides for nucleic acid
delivery, Adv. Drug Deliv. Rev. 106(Pt A) (2016) 172–182.
[9] H. Gao, Q. Zhang, Z. Yu, Q. He, Cell-penetrating peptide-based intelligent
liposomal systems for enhanced drug delivery, Curr. Pharm. Biotechnol. 15 (3)
(2014) 210–219.
[10] C. Bechara, S. Sagan, Cell-penetrating peptides: 20 years later, where do we
stand? FEBS Lett. 587 (12) (2013) 1693–1702.
[11] T. Takeuchi, S. Futaki, Current understanding of direct translocation of
arginine-rich cell-penetrating peptides and its internalization mechanisms,
Chem. Pharm. Bull. (Tokyo) 64 (10) (2016) 1431–1437.
[12] B.R. McNaughton, J.J. Cronican, D.B. Thompson, D.R. Liu, Mammalian cell
penetration, siRNA transfection, and DNA transfection by supercharged
proteins, Proc. Natl. Acad. Sci. U.S.A. 106 (15) (2009) 6111–6116.
[13] I. Martin, M. Teixido, E. Giralt, Building cell selectivity into CPP-mediated
strategies, Pharmaceuticals (Basel) 3 (5) (2010) 1456–1490.
[14] U. Niesner, C. Halin, L. Lozzi, M. Gunthert, P. Neri, H. Wunderli-Allenspach, L.
Zardi, D. Neri, Quantitation of the tumor-targeting properties of antibody
fragments conjugated to cell-permeating HIV-1 TAT peptides, Bioconjug.
Chem. 13 (4) (2002) 729–736.
[15] M. Zdanowicz, J. Chroboczek, Virus-like particles as drug delivery vectors, Acta
Biochim. Pol. 63 (3) (2016) 469–473.
[16] N. Kushnir, S.J. Streatfield, V. Yusibov, Virus-like particles as a highly efficient
vaccine platform: diversity of targets and production systems and advances in
clinical development, Vaccine 31 (1) (2012) 58–83.
[17] C.E. Ashley, E.C. Carnes, G.K. Phillips, P.N. Durfee, M.D. Buley, C.A. Lino, D.P.
Padilla, B. Phillips, M.B. Carter, C.L. Willman, C.J. Brinker, C. Caldeira Jdo, B.
Chackerian, W. Wharton, D.S. Peabody, Cell-specific delivery of diverse
cargos by bacteriophage MS2 virus-like particles, ACS Nano 5 (7) (2011)
5729–5745.
[18] S.J. Kaczmarczyk, K. Sitaraman, H.A. Young, S.H. Hughes, D.K. Chatterjee,
Protein delivery using engineered virus-like particles, Proc. Natl. Acad. Sci. U.S.
A. 108 (41) (2011) 16998–17003.
[19] A.M. Koch, F. Reynolds, H.P. Merkle, R. Weissleder, L. Josephson, Transport of
surface-modified nanoparticles through cell monolayers, ChemBioChem 6 (2)
(2005) 337–345.
[20] J. Heyes, L. Palmer, K. Bremner, I. MacLachlan, Cationic lipid saturation
influences intracellular delivery of encapsulated nucleic acids, J. Control.
Release 107 (2) (2005) 276–287.
[21] T.M. Allen, P.R. Cullis, Liposomal drug delivery systems: from concept to
clinical applications, Adv. Drug Deliv. Rev. 65 (1) (2013) 36–48.
[22] P.P. Deshpande, S. Biswas, V.P. Torchilin, Current trends in the use of liposomes
for tumor targeting, Nanomedicine (Lond) 8 (9) (2013) 1509–1528.
[23] W.H. De Jong, P.J. Borm, Drug delivery and nanoparticles: applications and
hazards, Int. J. Nanomed. 3 (2) (2008) 133–149.
[24] J.B. Szebeni, Adverse immune effects of liposomes: complement activation,
immunogenicity and immune suppression, in: D. Peer (Ed.), Harnessing
biomaterials for Nanomedicine: Preparation, toxicity and applications, Pan
Stanford Publishing, 2009.
[25] K.B. Knudsen, H. Northeved, P.E. Kumar, A. Permin, T. Gjetting, T.L. Andresen, S.
Larsen, K.M. Wegener, J. Lykkesfeldt, K. Jantzen, S. Loft, P. Moller, M.
Roursgaard, In vivo toxicity of cationic micelles and liposomes,
Nanomedicine 11 (2) (2015) 467–477.
[26] J.J. Mekalanos, R.J. Collier, W.R. Romig, Enzymic activity of cholera toxin. II.
Relationships to proteolytic processing, disulfide bond reduction, and subunit
composition, J. Biol. Chem. 254 (13) (1979) 5855–5861.
[27] A.E. Rabideau, B.L. Pentelute, Delivery of non-native cargo into mammalian
cells using anthrax lethal toxin, ACS Chem. Biol. 11 (6) (2016) 1490–1501.
[28] P.D. Dyer, T.R. Shepherd, A.S. Gollings, S.A. Shorter, M.A. Gorringe-Pattrick, C.K.
Tang, B.N. Cattoz, L. Baillie, P.C. Griffiths, S.C. Richardson, Disarmed anthrax
toxin delivers antisense oligonucleotides and siRNA with high efficiency and
low toxicity, J. Control Release 220(Pt A) (2015) 316–328.
[29] X. Liao, A.E. Rabideau, B.L. Pentelute, Delivery of antibody mimics into
mammalian cells via anthrax toxin protective antigen, ChemBioChem 15 (16)
(2014) 2458–2466.
[30] J.C. Milne, S.R. Blanke, P.C. Hanna, R.J. Collier, Protective antigen-binding
domain of anthrax lethal factor mediates translocation of a heterologous
protein fused to its amino- or carboxy-terminus, Mol. Microbiol. 15 (4) (1995)
661–666.
[31] W.P. Verdurmen, M. Luginbuhl, A. Honegger, A. Pluckthun, Efficient cell-
specific uptake of binding proteins into the cytoplasm through engineered
modular transport systems, J. Control. Release 200 (2015) 13–22.
 G.L. Beilhartz et al. / Biochemical Pharmacology xxx (2017) xxx–xxx
[32] I. Zornetta, L. Brandi, B. Janowiak, F. Dal Molin, F. Tonello, R.J. Collier, C.
Montecucco, Imaging the cell entry of the anthrax oedema and lethal toxins
with fluorescent protein chimeras, Cell. Microbiol. 12 (10) (2010) 1435–1445.
[33] J. Wesche, J.L. Elliott, P.O. Falnes, S. Olsnes, R.J. Collier, Characterization of
membrane translocation by anthrax protective antigen, Biochemistry 37 (45)
(1998) 15737–15746.
[34] A. Auger, M. Park, F. Nitschke, L.M. Minassian, G.L. Beilhartz, B.A. Minassian, R.
A. Melnyk, Efficient delivery of structurally diverse protein cargo into
mammalian cells by a bacterial toxin, Mol. Pharm. 12 (8) (2015) 2962–2971.
[35] E. Papini, R. Rappuoli, M. Murgia, C. Montecucco, Cell penetration of diphtheria
toxin. Reduction of the interchain disulfide bridge is the rate-limiting step of
translocation in the cytosol, J. Biol. Chem. 268 (3) (1993) 1567–1574.
[36] J. Niklas, A. Melnyk, Y. Yuan, E. Heinzle, Selective permeabilization for the
high-throughput measurement of compartmented enzyme activities in
mammalian cells, Anal. Biochem. 416 (2) (2011) 218–227.
[37] F. Grebien, O. Hantschel, J. Wojcik, I. Kaupe, B. Kovacic, A.M. Wyrzucki, G.D.
Gish, S. Cerny-Reiterer, A. Koide, H. Beug, T. Pawson, P. Valent, S. Koide, G.
Superti-Furga, Targeting the SH2-kinase interface in Bcr-Abl inhibits
leukemogenesis, Cell 147 (2) (2011) 306–319.
[38] T. Geiger, A. Wehner, C. Schaab, J. Cox, M. Mann, Comparative proteomic
analysis of eleven common cell lines reveals ubiquitous but varying expression
of most proteins, Mol. Cell. Proteomics 11 (3) (2012). M111 014050.
[39] M. Yamaizumi, E. Mekada, T. Uchida, Y. Okada, One molecule of diphtheria
toxin fragment A introduced into a cell can kill the cell, Cell 15 (1) (1978) 245–
250.
[40] R. Milo, P. Jorgensen, U. Moran, G. Weber, M. Springer, BioNumbers–the
database of key numbers in molecular and cell biology, Nucleic Acids Res. 38
(2010) D750–D753 (Database issue).
[41] I. Antic, M. Biancucci, K.J. Satchell, Cytotoxicity of the Vibrio vulnificus MARTX
toxin effector DUF5 is linked to the C2A subdomain, Proteins 82 (10) (2014)
2643–2656.
[42] H. Hu, S.H. Leppla, Anthrax toxin uptake by primary immune cells as
determined with a lethal factor-beta-lactamase fusion protein, PLoS ONE 4
(11) (2009) e7946.
[43] R. Zielinski, I. Lyakhov, A. Jacobs, O. Chertov, G. Kramer-Marek, N. Francella, A.
Stephen, R. Fisher, R. Blumenthal, J. Capala, Affitoxin–a novel recombinant,
HER2-specific, anticancer agent for targeted therapy of HER2-positive tumors,
J. Immunother. 32 (8) (2009) 817–825.
[44] D.P. Williams, K. Parker, P. Bacha, W. Bishai, M. Borowski, F. Genbauffe, T.B.
Strom, J.R. Murphy, Diphtheria toxin receptor binding domain substitution
with interleukin-2: genetic construction and properties of a diphtheria toxin-
related interleukin-2 fusion protein, Protein Eng. 1 (6) (1987) 493–498.
[45] L.F. Jean, J.R. Murphy, Diphtheria toxin receptor-binding domain substitution
with interleukin 6: genetic construction and interleukin 6 receptor-specific
action of a diphtheria toxin-related interleukin 6 fusion protein, Protein Eng. 4
(8) (1991) 989–994.
[46] A. Mechaly, A.J. McCluskey, R.J. Collier, Changing the receptor specificity of
anthrax toxin, MBio 3 (3) (2012).
[47] T.F. Liu, K.A. Cohen, J.G. Ramage, M.C. Willingham, A.M. Thorburn, A.E. Frankel,
A diphtheria toxin-epidermal growth factor fusion protein is cytotoxic to
human glioblastoma multiforme cells, Cancer Res. 63 (8) (2003) 1834–1837.
[48] T.F. Liu, M.C. Willingham, S.B. Tatter, K.A. Cohen, A.C. Lowe, A. Thorburn, A.E.
Frankel, Diphtheria toxin-epidermal growth factor fusion protein and
Pseudomonas exotoxin-interleukin 13 fusion protein exert synergistic
toxicity against human glioblastoma multiforme cells, Bioconjug. Chem. 14
(6) (2003) 1107–1114.
[49] A.E. Frankel, J.H. Woo, C. Ahn, F.M. Foss, M. Duvic, P.H. Neville, D.M. Neville,
Resimmune, an anti-CD3epsilon recombinant immunotoxin, induces durable
remissions in patients with cutaneous T-cell lymphoma, Haematologica 100
(6) (2015) 794–800.
[50] D.A. Vallera, H. Chen, A.R. Sicheneder, A. Panoskaltsis-Mortari, E.P. Taras,
Genetic alteration of a bispecific ligand-directed toxin targeting human CD19
and CD22 receptors resulting in improved efficacy against systemic B cell
malignancy, Leuk. Res. 33 (9) (2009) 1233–1242.
[51] R.J. Kreitman, I. Pastan, Antibody fusion proteins: anti-CD22 recombinant
immunotoxin moxetumomab pasudotox, Clin. Cancer Res. 17 (20) (2011)
6398–6405.
[52] R.J. Kreitman, W.H. Wilson, J.D. White, M. Stetler-Stevenson, E.S. Jaffe, S.
Giardina, T.A. Waldmann, I. Pastan, Phase I trial of recombinant immunotoxin
anti-Tac(Fv)-PE38 (LMB-2) in patients with hematologic malignancies, J. Clin.
Oncol. 18 (8) (2000) 1622–1636.
[53] W.S. Sly, H.D. Fischer, The phosphomannosyl recognition system for
intracellular and intercellular transport of lysosomal enzymes, J. Cell.
Biochem. 18 (1) (1982) 67–85.
[54] R.J. Desnick, E.H. Schuchman, Enzyme replacement therapy for lysosomal
diseases: lessons from 20 years of experience and remaining challenges, Annu.
Rev. Genomics Hum. Genet. 13 (2012) 307–335.
[55] D.J. Begley, C.C. Pontikis, M. Scarpa, Lysosomal storage diseases and the blood-
brain barrier, Curr. Pharm. Des. 14 (16) (2008) 1566–1580.
[56] J. Kreuter, D. Shamenkov, V. Petrov, P. Ramge, K. Cychutek, C. Koch-Brandt, R.
Alyautdin, Apolipoprotein-mediated transport of nanoparticle-bound drugs
across the blood-brain barrier, J. Drug Target. 10 (4) (2002) 317–325.
[57] M.J. Coloma, H.J. Lee, A. Kurihara, E.M. Landaw, R.J. Boado, S.L. Morrison, W.M.
Pardridge, Transport across the primate blood-brain barrier of a genetically
engineered chimeric monoclonal antibody to the human insulin receptor,
Pharm. Res. 17 (3) (2000) 266–274.
Please cite this article in press as: G.L. Beilhartz et al., Repurposing bacterial toxins for intracellular delivery of therapeutic proteins, Biochem. Pharmacol.
(2017), http://dx.doi.org/10.1016/j.bcp.2017.04.009
7
[58] R.J. Boado, Y. Zhang, Y. Wang, W.M. Pardridge, Engineering and expression of a
chimeric transferrin receptor monoclonal antibody for blood-brain barrier
delivery in the mouse, Biotechnol. Bioeng. 102 (4) (2009) 1251–1258.
[59] P.J. Gaillard, A. Brink, A.G. de Boer, Diphtheria toxin receptor-targeted brain
drug delivery, International Congress Series 1277(Drug Transport(ers) and the
Diseased Brain) (2005) 185–198.
[60] J.H. Han, S.A. Kushner, A.P. Yiu, H.L. Hsiang, T. Buch, A. Waisman, B. Bontempi,
R.L. Neve, P.W. Frankland, S.A. Josselyn, Selective erasure of a fear memory,
Science 323 (5920) (2009) 1492–1496.
[61] C.J. Wrobel, D.C. Wright, R.L. Dedrick, R.J. Youle, Diphtheria toxin effects on
brain-tumor xenografts. Implications for protein-based brain-tumor
chemotherapy, J. Neurosurg. 72 (6) (1990) 946–950.
[62] P. Calias, W.A. Banks, D. Begley, M. Scarpa, P. Dickson, Intrathecal delivery of
protein therapeutics to the brain: a critical reassessment, Pharmacol. Ther. 144
(2) (2014) 114–122.
[63] J.L. Cohen-Pfeffer, S. Gururangan, T. Lester, D.A. Lim, A.J. Shaywitz, M.
Westphal, I. Slavc, Intracerebroventricular delivery as a safe, long-term route
of drug administration, Pediatr. Neurol. 67 (2017) 23–35.
[64] R. Irannejad, N.G. Tsvetanova, B.T. Lobingier, M. von Zastrow, Effects of
endocytosis on receptor-mediated signaling, Curr. Opin. Cell Biol. 35 (2015)
137–143.
[65] V. Bertelsen, E. Stang, The mysterious ways of ErbB2/HER2 trafficking,
Membranes (Basel) 4 (3) (2014) 424–446.
[66] S.C. Hou, H.S. Chen, H.W. Lin, W.T. Chao, Y.S. Chen, C.Y. Fu, C.M. Yu, K.F. Huang,
A.H. Wang, A.S. Yang, High throughput cytotoxicity screening of anti-HER2
immunotoxins conjugated with antibody fragments from phage-displayed
synthetic antibody libraries, Sci. Rep. 6 (2016) 31878.
[67] R.J. Abi-Habib, S. Liu, T.H. Bugge, S.H. Leppla, A.E. Frankel, A urokinase-
activated recombinant diphtheria toxin targeting the granulocyte-macrophage
colony-stimulating factor receptor is selectively cytotoxic to human acute
myeloid leukemia blasts, Blood 104 (7) (2004) 2143–2148.
[68] S. Liu, T.H. Bugge, S.H. Leppla, Targeting of tumor cells by cell surface urokinase
plasminogen activator-dependent anthrax toxin, J. Biol. Chem. 276 (21) (2001)
17976–17984.
[69] N.J. Bulleid, Disulfide bond formation in the mammalian endoplasmic
reticulum, Cold Spring Harb. Perspect. Biol. 4 (11) (2012).
[70] USFAD Administration, Guidance for Industry: Immunogenicity Assessment
for Therapeutic Protein Products, in: O.o.M.P. Division of Medical Policy
Development, Center for Drug Evaluation and Research (CDER) (Ed.) 2014.
[71] F.M. Foss, DAB(389)IL-2 (ONTAK): a novel fusion toxin therapy for lymphoma,
in: Clin. Lymph. 1 (2) (2000) 110–116. discussion 117.
[72] K.E. Griswold, C. Bailey-Kellogg, Design and engineering of deimmunized
biotherapeutics, Curr. Opin. Struct. Biol. 39 (2016) 79–88.
[73] J.R. Schafer, B.M. Jesdale, J.A. George, N.M. Kouttab, A.S. De Groot, Prediction of
well-conserved HIV-1 ligands using a matrix-based algorithm, EpiMatrix,
Vaccine 16 (19) (1998) 1880–1884.
[74] A.S. Parker, Y. Choi, K.E. Griswold, C. Bailey-Kellogg, Structure-guided
deimmunization of therapeutic proteins, J. Comput. Biol. 20 (2) (2013) 152–
165.
[75] J.U. Schmohl, D. Todhunter, S. Oh, D.A. Vallera, Mutagenic deimmunization of
diphtheria toxin for use in biologic drug development, Toxins (Basel) 7 (10)
(2015) 4067–4082.
[76] A.S. De Groot, L. Moise, J.A. McMurry, E. Wambre, L. Van Overtvelt, P.
Moingeon, D.W. Scott, W. Martin, Activation of natural regulatory T cells by
IgG Fc-derived peptide ‘‘Tregitopes”, Blood 112 (8) (2008) 3303–3311.
[77] T.K. Kishimoto, J.D. Ferrari, R.A. LaMothe, P.N. Kolte, A.P. Griset, C. O’Neil, V.
Chan, E. Browning, A. Chalishazar, W. Kuhlman, F.N. Fu, N. Viseux, D.H.
Altreuter, L. Johnston, R.A. Maldonado, Improving the efficacy and safety of
biologic drugs with tolerogenic nanoparticles, Nat. Nanotechnol. 11 (10)
(2016) 890–899.
[78] R. Hassan, A.C. Miller, E. Sharon, A. Thomas, J.C. Reynolds, A. Ling, R.J. Kreitman,
M.M. Miettinen, S.M. Steinberg, D.H. Fowler, I. Pastan, Major cancer
regressions in mesothelioma after treatment with an anti-mesothelin
immunotoxin and immune suppression, Sci. Transl. Med. 5 (208) (2013)
208ra147.
[79] S.A. Shiryaev, A.V. Chernov, V.S. Golubkov, E.R. Thomsen, E. Chudin, M.S. Chee,
I.A. Kozlov, A.Y. Strongin, P. Cieplak, High-resolution analysis and functional
mapping of cleavage sites and substrate proteins of furin in the human
proteome, PLoS ONE 8 (1) (2013) e54290.
[80] M.F. Chiron, C.M. Fryling, D.J. FitzGerald, Cleavage of pseudomonas exotoxin
and diphtheria toxin by a furin-like enzyme prepared from beef liver, J. Biol.
Chem. 269 (27) (1994) 18167–18176.
[81] Z. Zhang, M. Park, J. Tam, A. Auger, G.L. Beilhartz, D.B. Lacy, R.A. Melnyk,
Translocation domain mutations affecting cellular toxicity identify the
Clostridium difficile toxin B pore, Proc. Natl. Acad. Sci. U.S.A. 111 (10) (2014)
3721–3726.
[82] A.K. Varkouhi, M. Scholte, G. Storm, H.J. Haisma, Endosomal escape pathways
for delivery of biologicals, J. Control. Release 151 (3) (2011) 220–228.
[83] D.S. D’Astolfo, R.J. Pagliero, A. Pras, W.R. Karthaus, H. Clevers, V. Prasad, R.J.
Lebbink, H. Rehmann, N. Geijsen, Efficient intracellular delivery of native
proteins, Cell 161 (3) (2015) 674–690.
[84] J.G. Naglich, J.E. Metherall, D.W. Russell, L. Eidels, Expression cloning of a
diphtheria toxin receptor: identity with a heparin-binding EGF-like growth
factor precursor, Cell 69 (6) (1992) 1051–1061.
[85] M.Z. Kounnas, R.E. Morris, M.R. Thompson, D.J. FitzGerald, D.K. Strickland, C.B.
Saelinger, The alpha 2-macroglobulin receptor/low density lipoprotein
8 G.L. Beilhartz et al. / Biochemical Pharmacology xxx (2017) xxx–xxx
receptor-related protein binds and internalizes Pseudomonas exotoxin A, J.
Biol. Chem. 267 (18) (1992) 12420–12423.
[86] D.B. Lacy, D.J. Wigelsworth, R.A. Melnyk, S.C. Harrison, R.J. Collier, Structure of
heptameric protective antigen bound to an anthrax toxin receptor: a role for
receptor in pH-dependent pore formation, Proc. Natl. Acad. Sci. U.S.A. 101 (36)
(2004) 13147–13151.
[87] S. Choe, M.J. Bennett, G. Fujii, P.M. Curmi, K.A. Kantardjieff, R.J. Collier, D.
Eisenberg, The crystal structure of diphtheria toxin, Nature 357 (6375) (1992)
216–222.
[88] J. Jiang, B.L. Pentelute, R.J. Collier, Z.H. Zhou, Atomic structure of anthrax
protective antigen pore elucidates toxin translocation, Nature 521 (7553)
(2015) 545–549.
[89] T.M. Koehler, R.J. Collier, Anthrax toxin protective antigen: low-pH-induced
hydrophobicity and channel formation in liposomes, Mol. Microbiol. 5 (6)
(1991) 1501–1506.
[90] E. Lemichez, M. Bomsel, G. Devilliers, J. vanderSpek, J.R. Murphy, E.V. Lukianov,
S. Olsnes, P. Boquet, Membrane translocation of diphtheria toxin fragment A
Please cite this article in press as: G.L. Beilhartz et al., Repurposing bacterial toxins for intracellular delivery of therapeutic proteins, Biochem. Pharmacol.
(2017), http://dx.doi.org/10.1016/j.bcp.2017.04.009
exploits early to late endosome trafficking machinery, Mol. Microbiol. 23 (3)
(1997) 445–457.
[91] B.A. Krantz, R.A. Melnyk, S. Zhang, S.J. Juris, D.B. Lacy, Z. Wu, A. Finkelstein, R.J.
Collier, A phenylalanine clamp catalyzes protein translocation through the
anthrax toxin pore, Science 309 (5735) (2005) 777–781.
[92] V.K. Chaudhary, Y. Jinno, D. FitzGerald, I. Pastan, Pseudomonas exotoxin
contains a specific sequence at the carboxyl terminus that is required for
cytotoxicity, Proc. Natl. Acad. Sci. U.S.A. 87 (1) (1990) 308–312.
[93] B. Tsai, C. Rodighiero, W.I. Lencer, T.A. Rapoport, Protein disulfide isomerase
acts as a redox-dependent chaperone to unfold cholera toxin, Cell 104 (6)
(2001) 937–948.
[94] S. Massey, H. Burress, M. Taylor, K.N. Nemec, S. Ray, D.B. Haslam, K. Teter,
Structural and functional interactions between the cholera toxin A1 subunit
and ERdj3/HEDJ, a chaperone of the endoplasmic reticulum, Infect. Immun. 79
(11) (2011) 4739–4747.