Engineering Botulinum Toxins to Improve and Expand Targeting and SNARE Cleavage Activity.

Botulinum neurotoxins (BoNTs) are highly successful protein therapeutics. Over 40 naturally occurring BoNTs have been described thus far and, of those, only 2 are commercially available for clinical use. Different members of the BoNT family present different biological properties but share a similar multi-domain structure at the molecular level. In nature, BoNTs are encoded by DNA in producing clostridial bacteria and, as such, are amenable to recombinant production through insertion of the coding DNA into other bacterial species. This, in turn, creates possibilities for protein engineering. Here, we review the production of BoNTs by the natural host and also recombinant production approaches utilised in the field. Applications of recombinant BoNT-production include the generation of BoNT-derived domain fragments, the creation of novel BoNTs with improved performance and enhanced therapeutic potential, as well as the advancement of BoNT vaccines. In this article, we discuss site directed mutagenesis, used to affect the biological properties of BoNTs, including approaches to alter their binding to neurons and to alter the specificity and kinetics of substrate cleavage. We also discuss the target secretion inhibitor (TSI) platform, in which the neuronal binding domain of BoNTs is substituted with an alternative cellular ligand to re-target the toxins to non-neuronal systems. Understanding and harnessing the potential of the biological diversity of natural BoNTs, together with the ability to engineer novel mutations and further changes to the protein structure, will provide the basis for increasing the scope of future BoNT-based therapeutics.

Structural analysis of Clostridium botulinum neurotoxin type D as a platform for the development of targeted secretion inhibitors.

The botulinum neurotoxin type D is one of seven highly potent toxins produced by Clostridium botulinum which inhibit neurotransmission at cholinergic nerve terminals. A functional fragment derived from the toxin, LHn, consisting of the catalytic and translocation domains, has been heralded as a platform for the development of targeted secretion inhibitors. These secretion inhibitors are aimed at retargeting the toxin towards a specific cell type to inhibit vesicular secretion. Here we report crystal structures of LHn from serotype D at 2.3 Å, and that of SXN101959 at 3.1 Å resolution. SXN101959, a derivative that combines LHn from serotype D with a fragment of the growth hormone releasing hormone, has previously revealed promising results in inhibiting growth hormone release in pituitary somatotrophs. These structures offer for the first time insights into the translocation domain interaction with the catalytic domain in serotype D. Furthermore, structural information from small-angle X-ray scattering of LHn/D is compared among serotypes A, B, and D. Taken together, these results demonstrate the robustness of the 'LHn fold' across serotypes and its use in engineering additional polypeptide components with added functionality. Our study demonstrates the suitability of botulinum neurotoxin, and serotype D in particular, as a basis for engineering novel secretion inhibitors.

Engineered botulinum neurotoxins as new therapeutics.

Botulinum neurotoxins (BoNTs) cause flaccid paralysis by inhibiting neurotransmission at cholinergic nerve terminals. Each BoNT consists of three domains that are essential for toxicity: the binding domain, the translocation domain, and the catalytic light-chain domain. BoNT modular architecture is associated with a multistep mechanism that culminates in the intracellular proteolysis of SNARE (soluble N-ethylmaleimide-sensitive-fusion-protein attachment protein receptor) proteins, which prevents synaptic vesicle exocytosis. As the most toxic proteins known, BoNTs have been extensively studied and are used as pharmaceutical agents to treat an increasing variety of disorders. This review summarizes the level of sophistication reached in BoNT engineering and highlights the diversity of approaches taken to utilize the modularity of the toxin. Improved efficiency and applicability have been achieved by direct mutagenesis and interserotype domain rearrangement. The scope of BoNT activity has been extended to nonneuronal cells and offers the basis for novel biomolecules in the treatment of secretion disorders.

Engineering botulinum neurotoxin domains for activation by toxin light chain.

Targeted secretion inhibitors (TSI) are a new class of biopharmaceuticals designed from a botulinum neurotoxin protein scaffold. The backbone consists of the 50-kDa endopeptidase light chain and translocation domain (N-terminal portion of the heavy chain), lacks neuronal toxicity, but retains the ability to target cytoplasmic soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) proteins. TSI are produced as single-chain proteins and then cleaved post-translationally to generate functional heterodimers. Precise proteolytic cleavage is essential to activate the protein to a dichain form. TSI are themselves highly specific proteases. We have exploited this activity to create self-activating enzymes by replacing the native proteolytic site with a substrate SNARE peptide for the TSI protease. We have also created cross-activating backbones. By replacing the proteolytic activation site in one backbone with the substrate SNARE peptide for another serotype, controlled activation is achieved. SNARE peptides encompassing the whole of the coiled-coil region enabled complete activation and assembly of the dichain backbone. These engineered TSI backbones are capable of translocating their enzymatic domains to target intracellular SNARE proteins. They are also investigative tools with which to further the understanding of endopeptidase activity of light chain in SNARE interactions.

Structures of engineered Clostridium botulinum neurotoxin derivatives.

Targeted secretion inhibitors (TSIs) are a new class of engineered biopharmaceutical molecules derived from the botulinum neurotoxins (BoNTs). They consist of the metalloprotease light chain (LC) and translocation domain (Hn) of BoNT; they thus lack the native toxicity towards motor neurons but are able to target soluble N-ethylmaleimide-sensitive fusion protein attachment receptor (SNARE) proteins. These functional fragment (LHn) derivatives are expressed as single-chain proteins and require post-translational activation into di-chain molecules for function. A range of BoNT derivatives have been produced to demonstrate the successful use of engineered SNARE substrate peptides at the LC-Hn interface that gives these molecules self-activating capabilities. Alternatively, recognition sites for specific exoproteases can be engineered to allow controlled activation. Here, the crystal structures of three LHn derivatives are reported between 2.7 and 3.0 Å resolution. Two of these molecules are derivatives of serotype A that contain a SNARE peptide. Additionally, a third structure corresponds to LHn serotype B that includes peptide linkers at the exoprotease activation site. In all three cases the added engineered segments could not be modelled owing to disorder. However, these structures highlight the strong interactions holding the LHn fold together despite the inclusion of significant polypeptide sequences at the LC-Hn interface.

Engineering toxins for 21st century therapies.

'Engineering Toxins for 21st Century Therapies' (9-10 September 2010) was part of the Royal Society International Seminar series held at the Kavli International Centre, UK. Participants were assembled from a range of disciplines (academic, industry, regulatory, public health) to discuss the future potential of toxin-based therapies. The meeting explored how the current structural and mechanistic knowledge of toxins could be used to engineer future toxin-based therapies. To date, significant progress has been made in the design of novel recombinant biologics based on domains of natural toxins, engineered to exhibit advantageous properties. The meeting concluded, firstly that future product development vitally required the appropriate combination of creativity and innovation that can come from the academic, biotechnology and pharma sectors. Second, that continued investigation into understanding the basic science of the toxins and their targets was essential in order to develop new opportunities for the existing products and to create new products with enhanced properties. Finally, it was concluded that the clinical potential for development of novel biologics based on toxin domains was evident.

Structure and activity of a functional derivative of Clostridium botulinum neurotoxin B.

Botulinum neurotoxins (BoNTs) cause flaccid paralysis by inhibiting neurotransmission at cholinergic nerve terminals. BoNTs consist of three essential domains for toxicity: the cell binding domain (Hc), the translocation domain (Hn) and the catalytic domain (LC). A functional derivative (LHn) of the parent neurotoxin B composed of Hn and LC domains was recombinantly produced and characterised. LHn/B crystallographic structure at 2.8Å resolution is reported. The catalytic activity of LHn/B towards recombinant human VAMP was analysed by substrate cleavage assay and showed a higher specificity for VAMP-1, -2 compared to VAMP-3. LHn/B also showed measurable activity in living spinal cord neurons. Despite lacking the Hc (cell-targeting) domain, LHn/B retained the capacity to internalize and cleave intracellular VAMP-1 and -2 when added to the cells at high concentration. These activities of the LHn/B fragment demonstrate the utility of engineered botulinum neurotoxin fragments as analytical tools to study the mechanisms of action of BoNT neurotoxins and of SNARE proteins.

Crystal structure of a catalytically active, non-toxic endopeptidase derivative of Clostridium botulinum toxin A.

Botulinum neurotoxins (BoNTs) modulate cholinergic nerve terminals to result in neurotransmitter blockade. BoNTs consists of catalytic (LC), translocation (Hn) and cell-binding domains (Hc). The binding function of the Hc domain is essential for BoNTs to bind the neuronal cell membrane, therefore, removal of the Hc domain results in a product that retains the endopeptidase activity of the LC but is non-toxic. Thus, a molecule consisting of LC and Hn domains of BoNTs, termed LHn, is a suitable molecule for engineering novel therapeutics. The structure of LHA at 2.6 A reported here provides an understanding of the structural implications and challenges of engineering therapeutic molecules that combine functional properties of LHn of BoNTs with specific ligand partners to target different cell types.

Botulinum neurotoxin A and neurotoxin E cleavage products of synaptosome-associated protein of 25 kd exhibit distinct actions on pancreatic islet beta-cell Kv2.1 channel gating.

OBJECTIVES: Synaptosome-associated protein of 25 kd (SNAP-25) regulates pancreatic islet beta-cell-delayed rectifier K channels (Kv2.1) in addition to insulin exocytosis. Botulinum neurotoxin A (BoNT/A) and E (BoNT/E) cleavage and presumed deletion of SNAP-25 have been used to examine SNAP-25 function. We hypothesized that proteolytic products of SNAP-25 (206 amino acids) resulting from BoNT/A and BoNT/E cleavage, SNAP-25(1-197) and SNAP-25(1-180), have independent actions on beta-cell Kv gating. METHODS: We examined by confocal microscopy and immunoblotting BoNT/A and BoNT/E cleavage of SNAP-25 to these N-terminal fragments, and the consequent effects of these BoNTs and SNAP-25 fragments on Kv currents in rat beta cells and MIN6 cells by patch clamp electrophysiology. RESULTS: Confocal microscopy and immunoblotting showed that MIN6 cells transfected with BoNT/A or BoNT/E generated SNAP-25(1-197) and SNAP-25(1-180) fragments that were retained in the cytosol. Both BoNTs caused increased rate of channel activation and slowed channel inactivation, mimicked by these SNAP-25 fragments, but not full-length SNAP-25. These SNAP-25 fragments potentiated tetraethylammonium block of beta-cell Kv currents. CONCLUSIONS: BoNT/A or BoNT/E treatment of beta cells generates N-terminal SNAP-25 fragments that are retained in beta cells to directly influence Kv channel gating in a manner distinct from full-length SNAP-25, contributing to overall actions of these BoNTs on insulin secretion.

Preparation of specifically activatable endopeptidase derivatives of Clostridium botulinum toxins type A, B, and C and their applications.

Clostridium botulinum neurotoxins are potently toxic proteins of 150 kDa with specific endopeptidase activity for SNARE proteins involved in vesicle docking and release. Following treatment with trypsin, a fragment of botulinum neurotoxin serotype A that lacks the C-terminal domain responsible for neuronal cell binding, but retains full catalytic activity, can be obtained. Known as the LH(N) fragment, we report the development of a recombinant expression and purification scheme for the isolation of comparable fragments of neurotoxin serotypes B and C. Expressed as maltose-binding protein fusions, both have specific proteolytic sites present between the fusion tag and the light chain to facilitate removal of the fusion, and between the light chain endopeptidase and the H(N) translocation domains to facilitate activation of the single polypeptide. We have also used this approach to prepare a new variant of LH(N)/A with a specific activation site that avoids the need to use trypsin. All three LH(N)s are enzymatically active and are of low toxicity. The production of specifically activatable LH(N)/A, LH(N)/B, and LH(N)/C extends the opportunities for exploitation of neurotoxin fragments. The potential utility of these fragments is discussed.

Novel application of an in vitro technique to the detection and quantification of botulinum neurotoxin antibodies.

Detection of Clostridium botulinum neurotoxin (BoNT) neutralising antibodies is currently achieved using the mouse lethality assay (MLA). This technique has provided the majority of the data for vaccine development and, with the increasing use of BoNT as a therapeutic agent, the MLA is the assay of choice to evaluate 'non-responder' antisera. However, the MLA is semi-quantitative and has an animal consumption rate that raises ethical concerns. The development of an alternative is therefore desirable. Here, we describe an in vitro neuronal release assay that may represent such an alternative in terms of both its sensitivity and ability to produce quantitative data. Initially recognised in the course of assessing a novel vaccine candidate, the suitability of this assay has been further explored using an International standard. The results support the conclusion that the detection of neutralising antibodies in human sera should be attempted using this method.

Crystal structures of Erythrina cristagalli lectin with bound N-linked oligosaccharide and lactose.

Erythrina cristagalli lectin (ECL) is a galactose-specific legume lectin. Although its biological function in the legume is unknown, ECL exhibits hemagglutinating activity in vitro and is mitogenic for T lymphocytes. In addition, it has been recently shown that ECL forms a novel conjugate when coupled to a catalytically active derivative of the type A neurotoxin from Clostridium botulinum, thus providing a therapeutic potential. ECL is biologically active as a dimer in which each protomer contains a functional carbohydrate-combining site. The crystal structure of native ECL was recently reported in complex with lactose and 2'-fucosyllactose. ECL protomers adopt the legume lectin fold but form non-canonical dimers via the handshake motif as was previously observed for Erythrina corallodendron lectin. Here we report the crystal structures of native and recombinant forms of the lectin in three new crystal forms, both unliganded and in complex with lactose. For the first time, the detailed structure of the glycosylated hexasaccharide for native ECL has been elucidated. The structure also shows that in the crystal lattice the glycosylation site and the carbohydrate binding site are involved in intermolecular contacts through water-mediated interactions.

Retargeted clostridial endopeptidases: inhibition of nociceptive neurotransmitter release in vitro, and antinociceptive activity in in vivo models of pain.

Clostridial neurotoxins potently and specifically inhibit neurotransmitter release in defined cell types. Previously reported data have demonstrated that the catalytically active LH(N) endopeptidase fragment of botulinum neurotoxin type A (termed LH(N)/A) can be retargeted to a range of cell types in vitro to lead to inhibition of secretion of a range of transmitters. Here, we report the synthesis of endopeptidase conjugates with in vitro selectivity for nociceptive afferents compared to spinal neurons. Chemical conjugates prepared between Erythrina cristagalli lectin and LH(N)/A are assessed in vitro and in in vivo models of pain. Chemical conjugates prepared between E. cristagalli lectin and either natively sourced LH(N)/A, or recombinant LH(N)/A purified from Escherichia coli are assessed, and equivalence of the recombinant material is demonstrated. The duration of action of inhibition of neurotransmitter release by the conjugate in vitro is also assessed and is comparable to that observed with Clostridium botulinum neurotoxin. Selectivity of targeting and therapeutic potential have been confirmed by in vivo electrophysiology studies. Furthermore, the analgesic properties of the conjugate have been assessed in in vivo models of pain and extended duration effects observed. These data provide proof of principle for the concept of retargeted clostridial endopeptidases as novel analgesics.

Isolation of the gene and large-scale expression and purification of recombinant Erythrina cristagalli lectin.

Using polymerase chain reaction, the coding sequence for Erythrina cristagalli lectin (ECL) has been cloned and expressed in Escherichia coli. The amplified DNA sequence of ECL is highly homologous to that previously reported for Erythrina corallodendron lectin (ECorL), confirming the absence of introns in the ECL gene. The polypeptide sequences of ECL and ECorL have been compared and five amino acids have been identified that differentiate the two proteins. Recombinant E. cristagalli lectin (recECL) was expressed in E. coli from a genomic clone encoding the mature E. cristagalli lectin gene. Constitutive expression localised recombinant protein in inclusion bodies, which were solubilised, and recECL, subsequently refolded and purified by lactose affinity chromatography. Significant advantages were observed for purification from inclusion bodies rather than from a clone optimised to express soluble protein. A large-scale purification scheme has been developed that can prepare functional recECL from inclusion bodies with a yield of 870 mg/l culture. By the range of characterisation methods employed in this study, it has been demonstrated that recECL is functionally equivalent to native ECL obtained from the E. cristagalli plant. In addition, characterisation of the binding of radiolabelled recECL to cultured dorsal root ganglia demonstrated that recECL binds to a single pool of receptors.

Botulinum and tetanus neurotoxins: structure, function and therapeutic utility.

The toxic products of the anaerobic bacteria Clostridium botulinum, Clostridium butyricum, Clostridium barati and Clostridium tetani are the causative agents of botulism and tetanus. The ability of botulinum neurotoxins to disrupt neurotransmission, often for prolonged periods, has been exploited for use in several medical applications and the toxins, as licensed pharmaceutical products, now represent the therapeutics of choice for the treatment for several neuromuscular conditions. Research into the structures and activities of botulinum and tetanus toxins has revealed features of these proteins that might be useful in the design of improved vaccines, effective inhibitors and novel biopharmaceuticals. Here, we discuss the relationships between structure, mechanism of action and therapeutic use.

Expression and purification of catalytically active, non-toxic endopeptidase derivatives of Clostridium botulinum toxin type A.

Clostridium botulinum neurotoxin type A is a potently toxic protein of 150 kDa with specific endopeptidase activity for the SNARE protein SNAP-25. Proteolytic cleavage of BoNT/A with trypsin leads to removal of the C-terminal domain responsible for neuronal cell binding. Removal of this domain result in a catalytically active, non-cell-binding derivative termed LH(N)/A. We have developed a purification scheme to prepare LH(N)/A essentially free of contaminating BoNT/A. LH(N)/A prepared by this scheme retains full enzymatic activity, is stable in solution, and is of low toxicity as demonstrated in a mouse toxicity assay. In addition, LH(N)/A has minimal effect on release of neurotransmitter from a primary cell culture model. Both the mouse bioassay and in vitro release assay suggest BoNT/A is present at less than 1 in 10(6) molecules of LH(N)/A. This represents a significant improvement on previously reported figures for LH(N)/A, and also the light chain domain, previously purified from BoNT/A. To complement the preparation of LH(N)/A from holotoxin, DNA encoding LH(N)/A has been introduced into Escherichia coli to facilitate expression of recombinant product. Expression and purification parameters have been developed to enable isolation of soluble, stable endopeptidase with a toxicity profile enhanced on that of LH(N)/A purified from BoNT/A. The recombinant-derived material has been used to prepare antisera that neutralise a BoNT/A challenge. The production of essentially BoNT/A-free LH(N)/A by two different methods and the possibilities for exploitation are discussed.

Inhibition of release of neurotransmitters from rat dorsal root ganglia by a novel conjugate of a Clostridium botulinum toxin A endopeptidase fragment and Erythrina cristagalli lectin.

Clostridial neurotoxins potently and specifically inhibit neurotransmitter release in defined cell types. Here we report that a catalytically active derivative (termed LH(N)/A) of the type A neurotoxin from Clostridium botulinum has been coupled to a lectin obtained from Erythrina cristagalli to form a novel conjugate. This conjugate exhibits an in vitro selectivity for nociceptive afferents compared with the anatomically adjacent spinal neurons, as assessed using in vitro primary neuronal culture systems to measure inhibition of release of neurotransmitters. Chemical conjugates prepared between E. cristagalli lectin and either natively sourced LH(N)/A or recombinant LH(N)/A purified from Escherichia coli are assessed, and equivalence of the recombinant material are demonstrated. Furthermore, the dependence of inhibition of neurotransmitter release on the cleavage of SNAP-25 is demonstrated through the use of an endopeptidase-deficient LH(N)/A conjugate variant. The duration of action of inhibition of neurotransmitter released by the conjugate in vitro is assessed and is comparable with that observed with Clostridium botulinum neurotoxin. Finally, in vivo electrophysiology shows that these in vitro actions have biological relevance in that sensory transmission from nociceptive afferents through the spinal cord is significantly attenuated. These data demonstrate that the potent endopeptidase activity of clostridial neurotoxins can be selectively retargeted to cells of interest and that inhibition of release of neurotransmitters from a neuronal population of therapeutic relevance to the treatment of pain can be achieved.