Harnessing toxins


Syntaxin has developed molecular strategies that enable the pharmacological activity of microbial proteins to be transferred to novel recombinant protein therapeutics. Chief scientific officer Keith Foster explains how the resulting products could help in the treatment of severe, chronic diseases

Syntaxin has developed molecular strategies that enable the pharmacological activity of microbial proteins to be transferred to novel recombinant protein therapeutics. Chief scientific officer Keith Foster explains how the resulting products could help in the treatment of severe, chronic diseases

Pathogenic micro-organisms have evolved molecules, many of them proteins, that can modulate host-cell function, often by affecting cell signalling in some way. Known as effector molecules, these can be used to identify molecules and processes within a cell that could be targets for therapeutic intervention or that can act as tools to dissect individual components of cell-signalling pathways.

In addition to their use as research tools, the microbial effector molecules themselves, or components from them, may form the basis of therapeutic agents. The pharmacological properties of many of these microbial effectors, which result from millions of years of evolutionary selection, are the envy of many traditional small molecules drug discovery programmes in terms of both potency and target specificity.

The effector may either be a component of a traditional toxin that contains other domains that enable the molecule to bind to and be internalised into the target cell, or it could be what is effectively an isolated toxin domain that is "injected" into eukaryotic cells (the complex cells found in animals, plant and fungi) by a bacterial secretion system. Type III (TTSS) and type IV (TFSS) secretion systems, for example, form channels through which the effector protein can be trafficked across the bacterial membrane and "injected" into the host cell cytoplasm.1

Whether the effector protein is a component of a bacterial toxin or is an isolated molecule, it can pro-vide the pharmacologically active component of a therapeutic protein, in which other protein components are engineered to provide target-cell selectivity and enable delivery. This "molecular toolbox" approach to generating novel therapeutics has been extensively researched by the scientists at Syntaxin, a newly formed biopharmaceutical company that designs and develops bacterial-based protein therapeutics for neurological, respiratory and metabolic diseases, particularly making use of the effector domains of the clostridial neurtoxins.

Clostridal neurotoxins

Clostridial neurotoxins (CloNTs), both the botulinum neurotoxins (BoNTs) and tetanus neurotoxin (TeNT), the causative agents of botulism and of tetanus, respectively, are among the most potent acute lethal toxins known. Despite their extreme toxicity, BoNTs are used to treat a range of chronic neuromuscular conditions and other conditions involving hyperactivity of the peripheral nervous system. 2 The clinical use of BoNT in a neuromuscular disorder was first described by Alan Scott, who in 1980 reported that injection of small quantities of toxin, far lower than those that cause systemic toxicity, into the extraocular muscles produced a correction in strabismus (squint).

Since this pioneering report, the clinical use of BoNT has increased dramatically and a number of pharmaceutical preparations of BoNT are now available as licenced clinical products. BoNTs have been reported to be effective in over 100 different clinical conditions, such as hyperhydrosis (excessive sweating), drooling and certain pain states. A major advantage of BoNTs in clinical use is their prolonged duration of action: the enzymes remain stable within the cells, and can have an effect for several months.

All CloNTs share a common structure, consisting of a heavy chain (HC) of approximately 100 kDa co-valently joined by a single disulphide bond to a light chain (LC) of approximately 50 kDa.2-3 The HC consists of two domains, each of approximately 50 kDa.3 The C-terminal domain (HC) is required for high affinity neuronal binding, while the N-terminal domain (HN) is proposed to be involved in translocating the protein through the membrane and into the target neuron.

The LC is the effector domain, being a zinc-dependent endopeptidase that cleaves a protein es-sential to synaptic vesicle docking and fusion, thereby inhibiting neurotransmitter release. CloNTs are produced as single chain proteins that require proteolytic cleavage to generate the functional di-chain toxin molecule.

CloNTs inhibit neurosecretion by selective prote-olytic cleavage of one of three SNARE (soluble NSF accessory protein receptor) proteins: Syntaxin, SNAP-25 or synaptobrevin (otherwise known as vesicle associated membrane protein or VAMP). These proteins form a complex essential for synaptic vesicle docking and fusion at the pre-synaptic membrane.4

There are eight immunologically distinct serotypes of CloNT: seven BoNT serotypes, named A through to G (BoNT/A - BoNT/G) and one TeNT, and each of these, with the exception of BoNT/C1, cleaves just one SNARE protein at a single peptide bond that is specific to that neurotoxin. BoNT/C1 is unique amongst the CloNTs in having two substrate proteins, both SNAP-25 and syntaxin. The cleavage of SNARE proteins within nerve cells by CloNTs prevents the formation of a functional SNARE complex and so inhibits neurosecretion.

The SNARE complex is not unique to neurosecretion. It represents a universal mechanism for vesicle fusion and secretion in eukaryotic cells.4 This means that the endopeptidase activity of CloNTs is potentially capable of SNARE protein cleavage and inhibition of vesicle fusion and secretion in a wide range of cell types, not just neuronal cells. The neuronal selectivity of clostridial endopeptidase action results from the specificity of the binding domain of the neurotoxin. Replacing the native binding domain with a ligand possessing different cellular selectivity would create a novel engineered protein with the potential to selectively inhibit vesicle fusion events and secretion in alternative target cells.

Engineering molecules

The crystal structures of BoNT/A, BoNT/B and TeNT HC have all been solved and show distinct structural domains, each one corresponding to one of the three steps in the neurotoxins" mechanism of action: binding, translocation and catalytic activity.5-6 The three domains are arranged sequentially in a linear fashion with the translocation domain in the middle, and there are no interactions between the binding and catalytic domains.

This structural arrangement facilitates the engineering of novel proteins with therapeutic potential, which combine the properties of particular domains of the CloNT molecules with other protein domains with alternative functions. Specifically, scientists at Syntaxin have been seeking to replace the binding domain of CloNTs with alternative ligands, to create engineered proteins able to inhibit vesicular trafficking in selected target cells. This "molecular toolbox" approach is shown schematically in Figure 1.

Initially, the novel proteins were engineered by generating a fragment of the neurotoxins known as the LHN fragment, which lacked the binding domain, and then chemically coupling peptide ligands to this fragment. LHN/A, for example, could be prepared by suitable limited proteolytic cleavage of BoNT/A.7

Unfortunately, preparation of the LHN fragment from other serotypes by protease treatment was found to be technically far more challenging, and there have been few reports of success. The recombinant LHNs were expressed in E. coli as single chain polypeptides and, to enable activation, the inter-domain region between the LC and HN domains was engineered to incorporate a recognition site for cleavage by a specific protease.8-9 As expected, the recombinant fragments were catalytically active but lacked the toxicity of the neurotoxin as they were unable to enter any cells.

A variety of cell-binding ligands were coupled to the LHN fragment using the heterobifunctional coupling agent SPDP. The recombinant proteins were able to enter both neuronal and non-neuronal cells, depending on the specificity of the targeting domain, and inhibited secretion via cleavage of the relevant substrate SNARE protein.10-11

While chemical coupling strategies proved that the concept worked in practice, the inherent heterogeneity of chemical conjugates, and the associated difficulty of developing a regulatory compliant process based on them, makes the method unsuitable as the basis for developing a pharmaceutical product. Fully recombinant expression approaches are the preferred route for development of a therapeutic protein, so the requirement was to develop a fully-recombinant chimera protein, incorporating the translocation and endopeptidase domains of a CloNT and a targeting ligand. Given the size and complexity of such a fusion protein, producing one as a functional, soluble recombinant protein is a challenging task.

Making use of 15-years of research expertise in engineering recombinant bacterial-based proteins, scientists at Syntaxin developed robust technologies to produce such recombinant fusion proteins, and have reported the successful production of a fully recombinant protein consisting of the LHN fragment of BoNT/C1 and epidermal growth factor, EGF-LHN/C.12

EGF-LHN/C is fully functional and inhibits secretion from target epithelial cells via cleavage of its substrate SNARE proteins in a receptor-dependent manner.

Syntaxin was recently awarded a collaborative r&d grant from the UK Department for Trade and Industry to develop enhanced methods for the bioprocessing of complex proteins in conjunction with the Advanced Centre for Biochemical Engineering (ACBE) at UCL and the Health Protection Agency This grant will further enhance Syntaxin's ability to engineer novel therapeutic proteins, for example these clostridial endopeptidase fusion proteins.

Therapeutic potential

The clinical potential of proteins that deliver clos-tridial endopeptidase activity into specified target cells was initially demonstrated in relation to chronic pain. A chemical conjugate of Erythrina cristagalli lectin and LHN/A, ECL-LHN/A, was shown to inhibit release of neurotransmitters of the pain pathway, substance P and glutamate, from embryonic dorsal root ganglion neurons in culture.13 This inhibition was maintained for at least 25 days following treatment, demonstrating that the conjugate retained the duration of action characteristic of the native neurotoxin. Intrathecal administration into the spinal cord of mice resulted in a prolonged withdrawal latency in a "hotplate" model of acute thermal pain.14 This effect was sustained for more than 30 days after administration of the conjugate. By contrast, morphine ceased to demonstrate analgesic activity in the same model within less than a day. ECL-LHN/A also inhibited inflammatory pain in a formalin model of imflammatory pain in rats. Again, the duration of analgesia was prolonged.13-14

Syntaxin has entered into an exclusive research and license arrangement with Allergan Inc. to design and develop novel clostridial endopeptidase-based proteins in the field of pain. Under the agreement, Allergan fund research efforts utilising Syntaxin's proprietary technology to engineer clostridial endopeptidase-based proteins that are specific for peripheral nociceptive nerve cells and thus have potential as novel long acting analgesic agents.

Potential for therapeutic action has also been demonstrated by the recombinant fusion protein EGF-LHN/C, which has been shown to inhibit mucin secretion from a human respiratory epithelial cell line.12 These data are encouraging for the development of new agents that will inhibit hypersecretion of mucus in the airways, thereby contributing to the treatment of a variety of respiratory conditions such as COPD and asthma. In addition to demonstrating that the EGF-LHN/C fusion protein is an effective agent for achieving inhibition of mucin release, these results are significant in confirming that the action of the clostridal endopeptidase can be harnessed to inhibit SNARE-dependent vesicle fusion and secretion occurring in non-neuronal cells.

The results reported here for ECL-LHN/A and EGF-LHN/C support the view that engineered proteins based upon clostridial endopeptidase enzymes can have discrete pharmacological effects on the function of specified target cells, and thus could be used as therapeutic proteins. The data also demonstrate that the novel engineered proteins have retained the duration of action of the native CloNTs, which is such a significant aspect of the clinical success of BoNT/A. The novel engineered clostridal endopeptidase fusion proteins would, therefore, be particularly suited to the treatment of chronic medical conditions.

In addition to pain and respiratory diseases involving excess mucus production (COPD, asthma and cystic fibrosis), Syntaxin is also exploring the potential of novel bacterial-based protein therapeutics in relation to a range of other chronic disease targets, including metabolic diseases. It is to be expected in the next few years that not only will the first clinical validation of this novel approach to therapeutic proteins become available, but also that the therapeutic potential of the technology will be established in a wider range of clinical opportunities.

Email this story