Discussion
We demonstrated a significant improvement in the efficacy of TeNT in vaccinated mice when administered alongside an inactivated decoy toxin. We propose that the decoy is acting as an immunological smokescreen, binding to protective antibodies, thus preventing them from neutralising some of the active TeNT. The is similar to a method of invasion employed by some bacteria and parasites (Wilson et al. , 2003). This effectively is analogous to reducing the concentration of circulating TeNT-neutralising antibodies. Furthermore, the decoy molecules were shown to be inert, retaining no residual, tetany-inducing activity. This is, to our knowledge, the first report of directly improving the activity of a tetanus toxin dose in vaccinated animals. The greatest increases in activity were observed at lower doses of TeNT. When TeNT was administered alone to vaccinated animals, we observed symptoms of tetany started at a dose of 250 ng with all mice reaching the trial endpoint at a dose of 4 µg. When administered alongside 100 µg of equimolar decoy solution, we observed symptoms at doses of TeNT as low as 0.2 ng, with a gradual increase in the responses observed as TeNT doses were increased to 250 ng, with all mice reaching the trial endpoint at a dose of 500 ng. TeNT administered alone exhibited a dose range of 16-fold from first detectable symptoms to all animals reaching the trial endpoint. This range increased to 2,500-fold when TeNT was administered with 100 µg of decoy molecules. These results suggest, that when combined with the decoy molecules, the effective activity of TeNT was greatly improved, and the likelihood of overdose was greatly reduced.
Tetanus toxin could be a tremendously impactful biological therapeutic for the treatment of neuromuscular disorders. It could revolutionise the treatment of disorders such as OSA (affecting approximately 5% of the population with significant health and economic burdens (Benjafieldet al. , 2019)), and spinal cord injury, and additionally has cosmetic and veterinary applications (Conduit et al. , 2007, Sasseet al. , 2005, Hesse et al. , 2020). Furthermore, the TTc domain has great potential as a non-disruptive mechanism for drug delivery across the barriers associated with the CNS (Benn et al. , 2005, Figueiredo et al. , 1997, Fishman et al. , 1990, Francis et al. , 2004, Moreno-Igoa et al. , 2010). The therapeutic and cosmetic success of botulinum toxin demonstrates how powerful a substance that can control muscle tone can be (Patil et al. , 2016, Pirazzini et al. , 2017, Shapira and Benhar, 2010, Dressler, 2012).
Human vaccination against tetanus introduces challenges to the successful development of therapeutic tetanus toxin. Protective immunity is induced by vaccination with tetanus toxoid, an inactivated form of tetanus toxin, which leads to a highly effective and long-lasting, antibody mediated response, directed against the toxin, rendering low-doses ineffective. The effect of the vaccination is long-lasting with a reported 14-year serum half-life of circulating antibodies (Borella-Venturini et al. , 2017, Mayer et al. , 2002). BoNT has been successfully applied as a therapeutic and cosmetic, because although it is the most potent neurotoxin known, it can be used safely in extremely low doses, to induce a mild, localised response (Chen, 2012, Dressler, 2012, Johnson, 1999). Overdoses can occur, but these result in paralysis of the local muscle group rather than a systemic issue (Taban and Perry, 2006). With careful administration, TeNT could be safely used to control local muscle tone, when the opposite effect to Botox is required - at least in a non-vaccinated individual.
Vaccination rates against Tetanus are high, but not complete. This must be considered when approaching tetanus toxin therapeutics, as a non-vaccinated person, or non-responder to vaccination, would be far more susceptible to the toxin than someone with a highly effective immune response. In 2009, Fishman et. al. demonstrated that localised tetany could be induced in a vaccinated population of mice (Fishman et al. , 2009). However, there was a relatively narrow dosage window between non-response and severe signs of tetany, and significant variability between individual animals; possibly caused by variability in serum antibody titre. Although the induction of a localised response in vaccinated animals was promising, the study demonstrated the need for a more sophisticated approach.
There has been significant study undertaken on the molecular characterisation of TeNT from the 1980s onwards; however, TeNT is a complicated molecule and the mechanics of its actions and interactions are still not completely understood (Gramlich et al. , 2013, Wanget al. , 2012, Pirazzini et al. , 2016, Rossetto et al. , 2011). In fact, an accurate structure of TeNT, derived from x-ray crystallography, was only solved in 2017 (Masuyer et al. , 2017). Tetanus toxin has a three sub-domain structure, split into two chains joined by a disulfide bond in the active form, very similar in structure and function to the botulinum toxins (Beise et al. , 1994, Eiselet al. , 1986a, Helting and Zwisler, 1977, Masuyer et al. , 2017, Masuyer et al. , 2011, Turton et al. , 2002). The main difference between tetanus and botulinum is that tetanus is transported to the CNS, where it disrupts the synaptic vesicle fusion complex in the inhibitory interneuron, thereby increasing the neurological signal to the corresponding muscle (Ovespian et al. , 2015, Lalli et al. , 2003, Bercsenyi et al. , 2013, Caleo and Schiavo, 2009, Restani et al. , 2012). Botulinum, conversely, acts in the peripheral nervous system to prevent the firing of the motor neuron synapse (Grumelli et al. , 2005, Lalli et al. , 2003, Patilet al. , 2016, Pirazzini et al. , 2017). It was therefore proposed that substitutions from botulinum to tetanus toxin, in regions other than that required for transport to the CNS, could yield a therapeutically viable version of the toxin with the activity of tetanus, but with reduced immunogenicity (Fürch et al. , 2007, Wang et al. , 2008, Wang et al. , 2012). The results were surprising though as substitution of the enzymatic region of BoNT to TeNT (thought to be analogous) resulted in a TeNT with the activity of a BoNT (Wang et al. , 2012). Furthermore, recent studies characterising the human and murine antibody response to TeNT have demonstrated protective that antibodies recognise multiple epitopes in the binding, transport and enzymatic regions of the toxin (Palermoet al. , 2017, Lavinder et al. , 2014, da Silva Antuneset al. , 2017, Yousefi et al. , 2014, Lukić et al. , 2015, Qazi et al. , 2006). There is also significant variation in the immunodominant epitopes between individual humans, and although mouse antibodies tend to recognize similar regions to humans, the specific sequences recognised are different. This leads to the conclusion that selectively mutating immunogenic regions would be unlikely to give the same results in mice and humans (Yousefi et al. , 2016). Due to these difficulties, we have focused on developing “sequence agnostic” approaches to improving the therapeutic potential of tetanus toxin.
We predict the dose of TeNT required to treat a condition such as OSA, or other conditions, is significantly less than the dose required to induce an observable physiological response. There is evidence to support this assumption reported in the treatment of OSA symptoms in a British bulldog (Sasse et al. , 2005), and the treatment of spinal cord injury in dogs (Hesse et al. , 2020). For the purpose of this study we compared doses required to induce observable tetany in vaccinated animals, as we were demonstrating the effectiveness of our approach on avoiding immunity, rather than analysing models of disease. Certainly, the clinically relevant dose of toxin would be significantly smaller than that required to induce physically observable signs of localised tetany, and by extension, systemic tetany. The questions therefore remain, does the decoy approach improve TeNT activity enough to treat disease without risking any sort of dangerous response at the same dose in a vaccinated or non-vaccinated individual? If not, can a simple serum ELISA be used to predict the safe dose for an individual? To find answers, studies could be performed in an animal model of OSA such as the British bulldog, as demonstrated by Sasse et. al.(2005).
It is necessary to consider that the decoy solution may boost the antibody response itself — a common problem with protein therapeutics in general (Dressler, 2002, Benecke, 2012, Naumann et al. , 2013). However, TeNT is an unusual case when it comes to protein therapeutics. While with other proteins, avoidance of the induction of immunity is paramount, for TeNT, that immunity already exists. Furthermore, the fact that this methodology could be applied to any protein, would suggest that it may be useful when long-term administration of a protein therapeutic has produced immunity that renders it ineffective, for example with Botox (Albrecht et al. , 2019, Hefter et al. , 2019, Dressler, 2002, Benecke, 2012). The decoy doses applied here, 5–100 µg, are quite high, but the dose of active TeNT required would be extremely low compared to some other commonly used protein therapeutics, so the combined therapeutic dose of TeNT and decoy could be kept within reasonable margins. For example, adalimumab is administered at a 40 mg dosage fortnightly for rheumatoid arthritis, plaque psoriasis, ankylosing spondylitis and Crohn’s disease, while interferon β1-a is administered at between 20 and 250 µg, either weekly or multiple times per week, for the treatment of multiple sclerosis (Naumann et al. , 2013). By contrast, we predict a TeNT therapeutic would be administered once every three to six months, at doses in the nanogram range: similar to Botox (Naumann et al. , 2013). The decoy solution tested was an equimolar solution of TTc R1225E W1288A and LCHn and was shown to have an inhibitory effect on TeNT activity in naïve mice. With further refinement, the inhibitory effect could be reduced or resolved, and the concentration of each component could be changed in such a way that the overall dose of decoy proteins required could be reduced.
TeNT has the potential to be a transformative protein therapeutic. Our results clearly demonstrate the potential of the decoy technology and support the idea that the barriers to developing a TeNT therapeutic are not insurmountable. Conditions that are either caused by low muscle tone or could be treated by increasing muscle tone, are numerous, often debilitating, and come with significant social and economic burdens. We believe that the technology presented here is the first major step towards realising the therapeutic potential of this powerful neurotoxin.