A Sensitive Cell-Based Assay for Testing Potency of Botulinum Neurotoxin Type A

Botulinum neurotoxin type A (BoNT/A) is a widely used biopharmaceutic for the treatment of neurological diseases and aesthetic medicine, allowing months-long paralysis of target muscles and glands. Large numbers of mice are used for multiple botulinum applications including batch release potency testing, antitoxin testing, countermeasure development and basic research. The mouse bioassay (MBA) has historically been the industry gold-standard in the botulinum field and is still heavily used for commercial product testing. BoNT/A intoxication causes severe suffering and application-specific, non-animal alternatives are urgently needed. It is widely accepted, that a cell-based assay (CBA) is the only way to faithfully replicate all the physiological steps of botulinum intoxication; comprising neuronal binding, internalization, endosomal escape, and cleavage of synaptosomal-associated protein of 25 kDa (SNAP25). However, it has not been straightforward to develop these assays and there are only a limited number of CBA currently in use. This is in part, due to the fact that very few cell lines have the appropriate levels of sensitivity to BoNT/A. In this study we have identified that LAN5 cells, a human neuroblastoma derived cell line, are sensitive to BoNT/A and can be engineered to express a recombinant NanoLuc luciferase tagged SNAP25 reporter molecule. On intoxication, the reporter molecule is cleaved and releases a NanoLuc-SNAP25 fragment which can be specifically captured on a 96-well plate for quantitative luminometry. Importantly, we demonstrate this new cell-based assay exhibits sensitivity comparable to the MBA.


Introduction
Botulinum neurotoxins (BoNTs) belong to a family of AB toxins produced by anaerobic bacteria of the genus Clostridium, which cause the deadly disease Botulism with the main symptom being neuromuscular paralysis (Erbguth and Naumann, 1999;Schiavo et al., 2000;Montecucco and Molgó, 2005).BoNTs are typically expressed as a single-chain precursor protein, which is later cleaved into two polypeptide chains.These chains comprise a 100 kD heavy chain housing the receptor-binding and translocation domains, and a 50 kD light chain with metalloprotease activity.The light chain cleaves neuronal Soluble Nethylmaleimide-sensitive attachment protein receptors (SNARE), ultimately impeding vesicle fusion and neurotransmitter release (Lacy et al., 1998;Chaddock and Marks, 2006;Binz and Rummel, 2009).There are seven well characterized BoNT serotypes, A-G.BoNT/A, C, and E proteolyse SNAP-25, while BoNT/B, D, F, and G cleave vesicle-associated membrane proteins (VAMPs) also known as synaptobrevins (Lacy and Stevens, 1999;Schiavo et al., 2000;Rummel et al., 2004;Antonucci et al., 2008;Rossetto and Montecucco, 2008;Binz et al., 2010).In order to reach the intraneuronal SNAREs, BoNTs first bind neuronal gangliosides and then a synaptic vesicle protein (synaptotagmin or SV2) on the presynaptic membrane surface, allowing it to be endocytosed into a vesicle (Montecucco and Schiavo, 1994;Binz and Rummel, 2009).Once the toxin is internalized, the vesicle acidifies and the disulphide joining the botulinum heavy and light chain is reduced, the translocation domain changes conformation to form a putative protein transduction channel, allowing the SNARE protease to translocate into the cytosol (Koriazova and Montal, 2003;Puhar et al., 2004;Pirazzini et al., 2013).
Due to potential variations in the biological activity of each batch of Botulinum toxin produced by manufacturers, the toxin must undergo rigorous potency testing before it can be sold, with even more extensive testing during new product development.Shelf-life associated stability testing is also required.The "gold standard" test for Botulinum toxin activity is the MBA, where animals are injected with toxin and the potency is measured by the animal's death caused by muscle paralysis and asphyxiation.This assay has many serious disadvantages including significant ethical concerns and high levels of variability which means a large number of laboratory animals must be used (Sesardic et al., 2003;Taylor et al., 2019).Therefore, more precise replacement assays that follow the principles of the 3Rs (reduction, replacement, and refinement of animal use in research) are urgently needed.Several major manufacturers have developed CBAs for BoNT/A batch release (Adler et al., 2010).However, the expected reduction in animal use has not been as dramatic as one would have expected.This in part caused by the fact that CBAs have not been approved for use in all of the manufacturing steps (Taylor et al., 2019) and that animals are still being used to: (i) periodically confirm the performance of the cell-based assays by cross-testing in animals; (ii) reference standard qualification (a three-yearly regulatory requirement) and (iii) to cover temporary cell-based assay failures (information obtained from the EU ALURES database).So even with the existing cell-based assays being adopted, it is estimated that in Europe alone, 400,000 animals are being used per year for Botulinum testing (Taylor et al., 2019).Unfortunately, this number is likely to increase further as new manufacturers enter the market (Fonfria et al., 2018) that do not have access to CBAs.Therefore, alternative CBAs are clearly needed to reduce the high numbers of animals currently being used.
A significant challenge in the development of a CBAs is the identification of cell lines which are sensitive to BoNT/A as most cell lines are insensitive to Botulinum toxins (Pellett, 2013), which can in part be explained by the fact that BoNTs bind only mature neurons with well differentiated axons and nerve terminals.At present, there are only two neuroblastoma derived cell lines being used for the testing of Botulinum based therapeutics -SiMa cells and Neuro-2a cells.SiMa cells (Marini et al., 1999) are used in a patent protected cell-based potency assay for Botox ® where BoNT/A-cleaved SNAP-25 is detected using sandwich enzyme-linked immunosorbent assay (ELISA) (Fernández-Salas et al., 2012).Biosentinel and Ipsen (Atapattu et al., 2013;Fonfria et al., 2023) have developed the BoCell ™ fluorescence ratiometric assay which uses murine Neuro-2a cells, however it is several orders of magnitude less sensitive for BoNT/A than the mouse LD50 lethality bioassay.In this assay, SNAP25 is modified with two fluorescent proteins and upon exposure to BoNT/A the reporter molecule is cleaved and the levels of FRET between the reporter molecules is reduced (Dong et al., 2004;Fonfria et al., 2023).
In parallel to these studies, our group has made use of SiMa cells to develop a research grade CBA for measuring BoNT/B potency.This assay measures the Botulinum-triggered proteolysis of VAMP molecules using a luminescent reaction (Rust et al., 2017).While developing this assay we were also investigating the potential of alternate neuroblastoma derived cell lines for detecting other Botulinum toxins.In this study, we present a novel assay for measuring BoNT/A potency based on human LAN5 neuroblastoma cells and an engineered SNAP25 reporter molecule which is tagged with NanoLuc luciferase.Intoxication of the cells results in cleavage of the reporter molecule which can subsequently be measured using a cleavage specific antibody immobilized on a 96 well plate.This assay is simple to use and has a comparable level of sensitivity to the MBA, fulfilling existing criteria for animal-free potency testing of BoNT/A.
In both cases the Hemagglutinin tag (HA) was located between the NanoLuc and SNAP25, as shown schematically in Fig 2A.
LAN5 cells were transduced using a gamma-retroviral system based on the moloney murine leukemia virus (Swift et al., 1999).LAN5 cells were seeded at 0.75x10 6 cells per well in 6-well plates 2 days prior to transduction.VSV-G pseudo typed viral particles were produced in HEK293 cells as described in (Williams and Peden, 2023) and added to LAN5 cells overnight in the presence of polybrene (8 μg/mL).Transduction efficiency for the SNAP25 reporter constructs was determined by fluorescent microscopy using both eGFP expression and anti-HA staining (Cat No: 901502, 1/400).Stable expression of the reporter constructs was achieved by selecting transduced cells with 1.5 μg/ml puromycin (Thermo Fisher Scientific).The NanoLuc-SNAP25 expressing LAN5 cells were subsequently transduced with either FLAG-tagged SV2A or V5-tagged SV2C and selected with 0.2 mg/mL Hygromycin B (Invitrogen).Diagrams of the SV2 overexpression vectors (Fig. S3 1 ) and SNAP25 reporter constructs (Fig. 2A) were prepared with SnapGene® Viewer software version 7.1.1(from Dotmatics; available at snapgene.com).
Differentiation and toxin treatment LAN5 cells were seeded at a density of 2x10 5 cells/well on in-house generated laminin coated (10 μg/mL laminin -Sigma) 48-well plates and differentiated in 600 μL of Neurobasal media (Gibco), 2X B-27 Supplement (Gibco), 1mM HEPES, pH 7.2 (Fisher), 1X GlutaMAX (Gibco), 1% Penicillin/Streptomycin (Gibco), and 10 μM all-trans-retinoic acid (Sigma) for 72 h.After 3 days, half of the media was replaced with fresh differentiation media containing different concentrations of Botulinum toxin.The cells were then incubated for a further three days to allow for the uptake of the toxin and intracellular cleavage of the reporter molecule.To test the specificity of the cleavage specific SNAP25 antibody, we obtained cortical neurons which were surplus to requirement from the Grierson group.Wild type primary cortical neurons were dissected from C57BL/6 mouse embryos and seeded onto 16 mm coverslips (VWR, Cat No: 631-0152) in 12-well plates that were pre-coated with Poly-L-Lysine at a concentration of 1.25x10 5 cells/mL.Cells were grown in Neurobasal Ⓡ Medium (Gibco) supplemented with 2X B-27 Supplement (Gibco), 2 mM L-Glutamine, and 1% Penicillin/Streptomycin for 6 days, following a 3 day BoNT/A (100 pM) treatment.Cells were then fixed with Paraformaldehyde (PFA) and stained for the detection of cleaved SNAP25 (Fig. S3 1 ).

Toxins
Commercially produced and purified, 150 kDa BoNT/A with a specific activity of 2.7x10 8 mouse LD50/mg was purchased from Metabiologics, United States, (Lot No: A042519-01).Commercial complex-associated BoNT/A, Botox Ⓡ , (Allergan, Lot No: C2982 C3F) was also used at a final concentration of 0.5 U/mL.Purified, 150 kDa BoNT/C (0.6 mg/mL) was a kind gift from our collaborator Thomas Binz.The toxins were handled in accordance with guidelines agreed by the University of Sheffield Health and Safety Committee.On completion of experiments, BoNT treated samples were inactivated by an acid/detergent mixture followed by incineration.
Fluorescence microscopy LAN5 cells were seeded on 8-well glass culture slides (Corning) (1x10 5 cells/well) precoated with poly-D-lysine (Merck, P6407) and laminin (10 μg/mL).After differentiation, and in some instances treatment with toxins, the cells were fixed with 4% PFA (Biolegend) for 15 min at room temperature.The PFA was quenched with 0.1 M Glycine for 5 min and the cells permeabilized with staining buffer (PBS, 0.1% Saponin, and 10% FBS) for 10 min on ice.The cells were incubated overnight at 4 °C with the indicated primary antibodies diluted in the staining solution.For detecting the expression of the SNAP25 reporter constructs, a mouse anti-HA antibody (Cat no 901502, 1/400) was used.For detecting expression of the SV2A and SV2C constructs, rabbit anti-V5 (Cell Signalling Technologies, 1/250) and mouse anti-FLAG M2 (produced in house by Genentech Inc.) antibodies were used.In addition, their expression was also confirmed using antibodies specific to SV2A (SYSY, Cat No: 119011, 1/100) and SV2C (SYSY, Cat No: 119203, 1/500) (Fig. S1 1 ).For cleaved SNAP25 detection, an inhouse rabbit polyclonal antibody (made by immunising rabbits with a synthesized peptide of , 1/5000) (Andreou et al., 2021) was used.For GT1b detection we used a mouse monoclonal primary antibody (Merck, Cat No: MAB5608, 1/500).The next day, cells were washed three times using the staining solution and incubated with goat anti mouse and/or rabbit secondary antibodies labelled with either AlexaFluor 488 or 594 (Invitrogen, Cat No: A11012, A11005, A11029 -1/1000 dilution) for 1 h at room temperature.The cells were washed three times with the staining solution and counterstained with DAPI (Invitrogen, 2μg/mL).Images were acquired at the Wolfson Light Microscopy Facility (University of Sheffield) with a Nikon A1 confocal laser microscopy and NIS Elements C imaging software (version 4.6) or with a digital fluorescence microscope (Leica Microsystems) at 40X or 63X magnification.Images were merged and handled using Image J2 (version 2.9.30/1.54f).(See all the listed antibodies for detailed information in Table S1 1 ).

NanoLuc based ELISA
After botulinum toxin treatment, the LAN5 cells were lysed by incubation with cell extraction buffer (PBS, 0.5% Triton X-100 and 1X Roche MiniEDTA protease inhibitor) for 15 min on a microplate shaker (600 rpm).The samples were then transferred into pre-chilled microcentrifuge tubes and centrifuged at 14000 rpm for 15 min at 4 °C.The supernatant was then collected and diluted in PBS and stored at -80 °C until required.For the one-step ELISA, Protein A coated 96-well plates (Thermo Fisher Scientific, UK) were incubated with 50 μL of antibody (in-house generated BoNT/A cleaved SNAP25 antibody, 1/50 dilution (0.44 mg/mL)) diluted in blocking solution (PBS, 1% w/v BSA) overnight at +4°C with gentle shaking.After antibody incubation, the wells were washed (PBS, 0.05% Tween) 3 times for 5 min on an orbital shaker.The samples were added to each well (50 μL) and incubated for 90 min at room temperature on a shaker.The samples were removed, and the wells washed 3 times.The buffer was then removed, and the plate incubated with 50 μL NanoGlo solution (2 μL of the substrate mixed with 48 μL of the buffer, Promega) for 5 min in the dark.Luminescence was read on Fluoroskan Ascent FL plate reader (Lab systems).

Statistical analysis
Graph Pad Prism was used for the statistical analysis of the ELISA data (Prism 9 for macOS, GraphPad SoftwareInc., La Jolla, CA, USA).The difference between Botox ® samples kept at room temperature was analyzed using One-way ANOVA.Statistical analysis to evaluate Botulinum sensitivity in the ELISA was assessed using Unpaired t-test.Two-way ANOVA was used to analyze the early sensitivity detection of BoNT/A treated cells.When experiments have been performed in duplicate, error bars show the mean and experimental range.When three independent experiments have been performed the mean for the data has been plotted and the error bars show SD.For the statistical analysis, cells were seeded into three separate wells.Each well was treated independently, and the lysate from each well was collected and analyzed by ELISA.For all graphs, calculated statistical significance has been noted as follows: * p≤0.05, * * p≤ 0.01, ***p<0.001,****p<0.0001.

LAN5 cells are a new Botulinum-sensitive cell line
A significant challenge in developing cell-based assays (CBA) for detecting botulinum toxins is the identification of cell lines which are sensitive to these toxins as the majority of them do not express all the machinery required for intoxication, including toxin receptors and final target substrates.Botulinum receptors such as gangliosides and synaptic terminal specific proteins are only expressed at the appropriate levels in neuronal cells (Bajjalieh et al., 1992;Ledeen and Wu, 2018).For BoNT/A intoxication, SV2 and GT1b ganglioside are the main receptors shown to be important for this process, whereas BoNT/C utilizes dual gangliosides (Stenmark et al., 2008;Karalewitz et al., 2012;Fonfria et al., 2018).LAN5 cells are well characterized neuroblastoma derived cell lines which can be differentiated to have neuronallike characteristics (Reddy et al., 1991;Guglielmi et al., 2014) (Fig. 1A), thus we decided to evaluate whether they express the machinery required for intoxication.To test this, we blotted the cells for SV2 (Fig. 1B) and performed immunostaining for the GT1b ganglioside (Fig. 1C).In undifferentiated cells, we detected a weak band for SV2 at the correct molecular weight which became more prominent upon differentiation.In addition, we also observed staining for GT1b at the plasma membrane in the differentiated cells suggesting that this cell line may be sensitive to Botulinum toxins.To directly test this hypothesis, the cells were differentiated for 72 h and incubated with either BoNT/A or BoNT/C (a SNAP25 and Syntaxin cleaving toxin) and the levels of SNAP25 cleavage determined by immunoblotting using a polyclonal antibody which detects full length SNAP25 as well as the form of the protein which is missing its C-terminus (BoNT/A and BoNT/C remove either 7 or 8 amino acids respectively) (Lu, 2015;Pirazzini et al., 2017).In the toxin-treated samples, there was a clear shift in molecular weight for SNAP25 suggesting that the protein has been cleaved by the toxins.In addition, in the BoNT/C treated cells, the levels of Syntaxin were reduced in a concentration dependent manner suggesting that it has been cleaved by the toxin (Fig. 1D).Considering the current dominance of BoNT/A products, we proceeded to engineer LAN5 cells to express a SNAP-25 reporter molecule, the key target of BoNT/A.

NanoLuciferase tagged SNAP25 is cleaved by BoNT/A
The potency of BoNT/A can be measured using immunoblotting.However, this approach is time consuming, labor intensive and is not suitable for testing large numbers of samples.Therefore, we aimed to develop an assay to measure the activity of BoNT/A by allowing SNARE cleavage to be quantified using a plate reader-based approach.We have previously shown that NanoLuc tagged VAMP2 is cleaved by BoNT/B and can be used in a one-step ELISA for measuring toxin potency (Rust et al., 2017).Thus, we decided to take a similar approach where we tagged SNAP25 with NanoLuc (Fig. 2A).As it is unclear whether the tag would impact SNAP25 cleavage we tagged human SNAP25 either at its N or C-terminus.To reduce the chance that NanoLuc interferes with SNAP25 cleavage we used a HA tag as a linker between the two proteins.This tag also allows for the immunological detection of the reporter construct.The reporter constructs were transduced into the LAN5 cells and stable cell lines generated.Immunocytochemistry using an anti-HA antibody confirmed expression of both SNAP25 constructs in LAN5 cells (Fig. 2B).The engineered cells were then differentiated and incubated with BoNT/A for 72 h and the levels and molecular weight of both endogenous and tagged SNAP25 determined by immunoblotting with a SNAP25 antibody.When the NanoLuc reporter was linked to the C-terminus of SNAP25 (SNAP25-NanoLuc), we observed a concentration dependent loss of the reporter construct in the BoNT/A treated samples suggesting that the NanoLuc reporter was being cleaved from SNAP25 (Fig. 2C, left panel).However, when the NanoLuc was linked to the N-terminus of SNAP25 (NanoLuc-SNAP25) no obvious change in molecular weight was observed (Fig. 2C, right panel), possibly due to the fact that BoNT/A only cleaves 7 amino acids from the construct.
To unequivocally determine if the NanoLuc-SNAP25 reporter construct was being cleaved by BoNT/A, we made use of an in-house generated polyclonal antibody that selectively detects cleaved SNAP25.This antibody was generated by immunizing rabbits with a short peptide of TRIDEANQ) which mimics the epitope which is formed when SNAP25 is cleaved by BoNT/A (Yadirgi et al., 2017;von Berg et al., 2019).This antibody was able to detect the cleavage of both endogenous SNAP25 and NanoLuc-SNAP25 in equal measure suggesting that the toxin could cleave the fusion protein without hindrance (Fig. 2D).Furthermore, this cleavage specific antibody also worked by immunocytochemistry and could be used to detect the action of BoNT/A in both the LAN5 reporter cells (Fig. 2E) as well as mouse cortical neurons (See Fig. S3 1 ).

The development and characterization of a quantitative CBA for measuring BoNT/A potency
To generate a cell-based assay for measuring BoNT/A potency two main components are required: a sensitive cell line and a method for measuring SNAP25 cleavage.In Figure 1 we demonstrated that LAN5 cells are sensitive to the toxin and in Figure 2 that they can be engineered to express an enzymatic reporter construct (NanoLuc-SNAP25) which is cleaved by BoNT/A and the resulting NanoLuc-SNAP25 fragment can be detected using a cleavage specific antibody.Thus, when all of these components are put together it should be possible to make a highly quantitative CBA for measuring BoNT/A activity (Fig. 3A).In this assay, a cleavage specific antibody is immobilized onto a plate and used to capture cleaved NanoLuc-SNAP25 and the amount of bound reporter molecules directly measured using a luminescence reaction.NanoLuc-based approaches are very sensitive and require fewer washing and incubation steps compared to a standard sandwich ELISA, potentially reducing the amount of time which would need to be spent on this assay (Boute et al., 2016).To test this approach, we immobilized the SNAP25 cleavage specific antibody onto a Protein A coated 96-well plate and incubated the plate with cell extracts generated from NanoLuc-SNAP25 reporter cells treated with or without 1 nM BoNT/A for 72 h (Fig. 3B, red bars).A robust luminescent signal was observed in the BoNT/A treated sample compared to the untreated control.In parallel, we also tested the specificity of the assay by treating the cells with 1 nM BoNT/C (Fig. 3B, red bars).As BoNT/C cleaves SNAP25 at a distinct site compared The HA tag acts as a linker between the NanoLuc and SNAP25.The GFP and puromycin-N-acetyltransferase genes are expressed as part of the same mRNA as the NanoLuc reporter, providing complementary approaches for selecting the expression of the reporter construct.The residues in SNAP25 cleaved by BoNT/A and C have been indicated.B. To confirm the reporter constructs were expressed in the transduced cells, the cells were fixed and stained with anti-HA tag antibody.Non-transduced LAN5 cells were used as negative control for the anti-HA staining.Scale bar 40 µm.C. The reporter cells were differentiated for 3 days and incubated with the indicated concentrations of BoNT/A for a further 72 h.Immunoblots were then prepared from the treated cells and probed with a polyclonal SNAP25 antibody which detects both endogenous (lower panels) and tagged SNAP25 (upper panels).Syntaxin was used as a loading control for the BoNT/A treated samples.D. To confirm whether the NanoLuc tagged SNAP25 reporter construct is being cleaved by BoNT/A, an immunoblot was prepared from cells treated with either 1 or 10 pM BoNT/A for 72 h.The blot was probed with the cleavage specific SNAP-25 antibody.Cleavage specific bands were observed at the predicted molecular weights for both endogenous and NanoLuc tagged SNAP25 (no signal was detected in the untreated samples).An antibody raised against Syntaxin served as a loading control.E. To further validate the specificity of the cleavage specific antibody and to determine if antibody could be used for immunofluorescence microscopy, LAN5 cells were treated with 1 nM BoNT/A, fixed, and stained with the antibody.Scale bar 20 µm. to BoNT/A, the antibody we have generated should be unable to bind the cleaved NanoLuc-SNAP25 reporter molecule.Only background levels of signal were observed with this treatment confirming the specificity of the cleavage specific antibody.As an additional control, we also treated SNAP25-NanoLuc cells with 1 nM BoNT/A and C (Fig. 3B, blue bars).Cleavage of this reporter molecule should only produce background levels of signal as the cleavage specific antibody should be incapable of binding the SNAP25-NanoLuc fragment as it does not contain the epitope it recognises.As predicted only background levels of signal were observed with these treatments.Taken together, these results demonstrate that the NanoLuc-SNAP25 cells, when used in combination with the cleavage specific antibody can be used to effectively detect BoNT/A activity.
To correlate the sensitivity of the CBA to the mouse lethality tests, we applied BoNT/A with known LD50 value (Metabiologics, USA) to the NanoLuc-SNAP25 cells and measured luciferase activity.BoNT/A activity was detectable in our ELISA at the lowest dose tested of 0.12 MLD50/well, which corresponds to 10 fM BoNT/A (Fig. 3C).We also investigated whether BoNT/A activity can be measured in a patient-ready lyophilised pharmaceutical product, Botox®, which contains high concentrations of human serum albumin and salts (Fig. 3D).The one-step ELISA method can reliably detect a single unit of Botox® suggesting that it can tolerate the excipients added to these products.To further explore the utility of the assay, commercial Botox® was resuspended in 0.9% NaCl to a concentration of 0.5 U/mL and was left at 20ºC for several days which could potentially cause protein degradation and/or aggregation, as it is above the recommended storage temperature for reconstituted Botox® (Walker and Dayan, 2014) (Fig. 3E).The cell-based assay was able to detect a gradual loss of BoNT/A activity over the 7-day period.Interestingly, the activity of the toxin was maintained up to 4-days at room temperature suggesting a robust protein structure of this bacterial neurotoxin.

Overexpression of SV2A enhances the performance of the CBA for BoNT/A detection
Various studies have elucidated that the SV2 family members serve as critical protein receptors for BoNT/A, mediating its internalization into neuronal cells (Dong et al., 2006;Mahrhold et al., 2006;Yao et al., 2016).To determine if SV2 overexpression can improve the performance of the assay, NanoLuc-SNAP25 cells were transduced with either SV2A or SV2C (tagged with FLAG or V5, respectively).The transduced cell populations showed staining for FLAG and V5 indicating that the constructs are being expressed (Fig. 4A).To determine if the transduced cells were overexpressing the recombinant receptors we stained the cells with antibodies against endogenous SV2A and SV2C (see Fig. S2 1 ).The levels of fluorescent signal for SV2A and SV2C were more pronounced in the transduced cells compared to that of the parental NanoLuc-SNAP25 cells.The transduced cells were then treated with varying concentrations of BoNT/A and compared to the parental population of NanoLuc-SNAP25 cells which only express endogenous levels of SV2.The cells overexpressing SV2A produced more signal at all BoNT/A concentrations tested compared to either the parental or SV2C transduced cells (Fig. 4B, at 1pM, 13-fold improvement in signal when compared to non-treated sample).To determine if SV2A overexpression also improved the amount of signal produced at the lowest concentrations of BoNT/A a more extensive serial dilution of the toxin was performed.Overexpression of SV2A increased the levels of signal detected at the lowest concentrations of toxin by approximately 2-fold (Fig. 4C).
To determine if SV2A overexpression could reduce the amount of time BoNT/A needs to be incubated with the cells to generate a robust signal we performed a time course experiment where the cells were incubated with BoNT/A for either 24 or 48 h (Fig. 4D).At both time points, the SV2A overexpressing cells produced significantly more signal suggesting that this approach may be useful for shortening the duration of the assay.

Discussion
The aim of this study was to develop a new research grade CBA for measuring the biological activity of BoNT/A.To achieve this we have identified a new neuroblastoma derived cell line which is sensitive to BoNT/A, generated a recombinant reporter molecule for measuring SNAP25 cleavage and used this as the basis of a quantitative One-Step ELISA assay.The CBA we have developed demonstrates comparable levels of sensitivity (0.01 pM/0.12U/well in ELISA) to the MBA (with a range of 0.6 to 2.5 mouse LD50 unit) (Pathe- Neuschäfer-Rube et al., 2015).We have demonstrated that the assay is robust and can be used to measure BoNT/A activity in commercial preparations which contain high levels of salt and stabilizers.We have also shown that the performance of the assay can be further enhanced by engineering the LAN5 cells to overexpress the receptor for BoNT/A.This platform will be useful for those studying the basic biology of Botulinum toxins, testing antitoxins and developing countermeasures targeting BoNT/A activity.In the longer term, with further development, this assay could be used in batch release potency testing (see limitations section).
The most important finding in this study is the identification that the LAN5 cells show similar levels of sensitivity as the MBA for BoNT/A (Wictome et al., 1999;Ferreira et al., 2004).LAN5 cells are relatively well characterized neuroblastoma derived cell lines (Reddy et al., 1991;Guglielmi et al., 2014) which are freely available to the research community.LAN5 cells can be genetically engineered using viral transduction and free of any restrictive intellectual property.We are not the first to investigate the utility of neuroblastoma derived cell lines for Botulinum toxin research and many studies have shown that this type of cell model are sensitive to toxins (Lee et al., 2008;Whitemarsh et al., 2012;Hong et al., 2016;Rasetti-Escargueil and Popoff, 2022), and two of the current CBAs which are in commercial use are also based on neuroblastoma derived cell lines (Fernández-Salas et al., 2012;Fonfria et al., 2023).LAN5 cells show comparable, if not better levels of sensitivity than several of these models and have the added advantage of being human derived (in SiMA cells: EC50~1-0.4U/well) (Fernández-Salas et al., 2012).One significant challenge when using neuroblastoma derived cell lines is their slow doubling time.We have found that LAN5 cells are relatively slow growing and only double approximately every 100 h which is longer than that of SiMa cells, which double between 48-100 h (unpublished observations).Another challenge when growing and differentiating neuroblastoma cells is their propensity to become clumpy.We have found that LAN5 cells are less prone to this problem compared to SiMA cells (unpublished observation).Thus, the LAN5 cells will be a useful alternative to the neuroblastoma derived cell lines that are currently in use.At present, it is unclear whether LAN5 cells are sensitive to other Botulinum toxins.However, based on their expression of Synaptotagmin it is likely that they will also show some sensitivity to VAMP cleaving toxins such as BoNT/B.
In the future, it will be of interest to determine how the LAN5 cells perform in relation to human induced pluripotent stem cell (hiPSCs) derived neurons.This model is currently used by one manufacturer and has been extensively studied by several research groups (Pellett et al., 2015(Pellett et al., , 2019;;Schenke et al., 2020).A recent study by (Cotter et al., 2023) has used IPSC derived motor neurons to measure BoNT/A activity using a Split-Luciferase based system.This approach does not require the use of a cleavage specific antibody and showed very high levels of sensitivity (lowest detectable dose of toxin 0.1 fM).However, a major drawback of this approach is the fact that motor neurons cannot be maintained in continuous cultures so must be genetically modified every time an assay needs to be performed, potentially limiting the utility of this approach.

Limitations of the study
While our CBA demonstrates significant promise, it is important to acknowledge the potential limitations of this study.For example, this assay has only been tested in-house and has not been transferred to other research groups for further validation.The experiments we have described in this paper have mainly focused on developing and exploring the utility of the CBA, it will be important in the future to perform a larger study with enough samples to calculate the assay's Inter/Intra coefficient of variation.This information will be critical if the assay is to be used for batch release potency testing.We have not determined the long-term stability and performance of the NanoLuc-SNAP25 reporter cells.In the experiments performed for this manuscript we have only used cells up to passage 18 so it is possible that at higher passages the cells may lose expression of the reporter construct, ability to differentiate and take up toxins.We have not assessed the robustness of the assay in relation to how changes in media composition impact the growth and differentiation of the LAN5 cells.For example, LAN5 cells are currently grown in FCS, which can be a significant source of assay variability especially if it needs to be sourced from a different manufacturer.Thus, in the longer term, it will be advisable to adapt the LAN5 cells to grow in chemically defined media.The assay uses a rabbit polyclonal antibody to measure SNAP25 cleavage which may hamper the adoption of the assay as polyclonal sera can show significant batch to batch variation and can be hard to remake.Thus, a key priority for the future will be to make a recombinant capture antibody which can be used to detect cleaved SNAP25.The assay is specific to BoNT/A so it is unlikely that it will be suitable for testing environmental and clinical matrices that can contain several different serotypes and material which is incompatible with sterile cell culture (Beske et al., 2016;Thirunavukkarasu et al., 2018).Finally, we have not fully explored the intellectual property associated with generating cleavage specific antibodies and SNAP25 reporter molecules which are used in the CBA.Thus, the assay's use may be constrained to non-commercial applications.
In summary, we have demonstrated that LAN5 cells are a useful cell line for measuring the biological activity of BoNT/A and when engineered to express a NanoLuc tagged SNARE proteins can be the basis of a highly sensitive and quantitative CBA.LAN5 cells show similar levels of sensitivity to the mouse bioassay and if more broadly adopted can potentially reduce the numbers of mice being used in Botulinum toxin research, antitoxin testing and the development of novel counter measures.In the longer term, it may be possible to further develop this platform to help reduce the numbers of animals being used during potency testing in the manufacture of medicines which contain BoNT/A.

Fig. 1 :
Fig. 1: LAN5 cells exhibit a high sensitivity to BoNT/A A. A timeline showing the differentiation and treatment periods used to detect BoNT activity.Brightfield images of differentiated and undifferentiated LAN5 cells (no toxin treatment).Neurites can be observed in both differentiated and undifferentiated cells (red arrows).Scale bar: 50 µm.B. LAN5 cells were differentiated as indicated in the timeline and blotted for SV2.The levels of SV2 increase upon differentiation.Undif: Undifferentiated, Dif: Differentiated cells.β-actin was used as a loading control.C. Differentiated LAN5 cells were stained with an anti-GT1b antibody (green).Differentiated LAN5 cells express GT1b.Nuclei were stained with DAPI (blue).Scale bar 20 µm.D. LAN5 cells were differentiated as described in Fig. 1A and incubated with the indicated concentrations of BoNT/A and C. The cells were then harvested for immunoblotting and the membranes probed with the indicated antibodies.Differentiated LAN5 cells are sensitive to BoNT/A (1 pM) and C (100 pM).BoNT/C also cleaves syntaxin as seen by the reduction in its levels.

Fig. 2 :
Fig. 2: NanoLuciferase tagged SNAP25 constructs (NanoLuc-SNAP25 and SNAP25-NanoLuc) are cleaved by BoNT/A A. Diagram showing the design of the N and C-terminal NanoLuc tagged SNAP25 reporter constructs and selection cassettes.The HA tag acts as a linker between the NanoLuc and SNAP25.The GFP and puromycin-N-acetyltransferase genes are expressed as part of the same mRNA as the NanoLuc reporter, providing complementary approaches for selecting the expression of the reporter construct.The residues in SNAP25 cleaved by BoNT/A and C have been indicated.B. To confirm the reporter constructs were expressed in the transduced cells, the cells were fixed and stained with anti-HA tag antibody.Non-transduced LAN5 cells were used as negative control for the anti-HA staining.Scale bar 40 µm.C. The reporter cells were differentiated for 3 days and incubated with the indicated concentrations of BoNT/A for a further 72 h.Immunoblots were then prepared from the treated cells and probed with a polyclonal SNAP25 antibody which detects both endogenous (lower panels) and tagged SNAP25 (upper panels).Syntaxin was used as a loading control for the BoNT/A treated samples.D. To confirm whether the NanoLuc tagged SNAP25 reporter construct is being cleaved by BoNT/A, an immunoblot was prepared from cells treated with either 1 or 10 pM BoNT/A for 72 h.The blot was probed with the cleavage specific SNAP-25 antibody.Cleavage specific bands were observed at the predicted molecular weights for both endogenous and NanoLuc tagged SNAP25 (no signal was detected in the untreated samples).An antibody raised against Syntaxin served as a loading control.E. To further validate the specificity of the cleavage specific antibody and to determine if antibody could be used for immunofluorescence microscopy, LAN5 cells were treated with 1 nM BoNT/A, fixed, and stained with the antibody.Scale bar 20 µm.

Fig. 3 :
Fig. 3: Characterization of the CBA for the detection of BoNT/A activity.A. Diagram outlining the key steps of the one-step ELISA for detecting cleaved NanoLuc-SNAP25.The cleavage specific SNAP25 antibody is immobilized on Protein A-coated 96-well plates and the amount of cleaved SNAP25 bound to the plate is measured using luciferase activity.Created with BioRender.com.B. LAN5 cells expressing either the N or C-terminal NanoLuc SNAP25 reporter constructs were differentiated for three days, incubated with 1nM BoNT/A or BoNT/C for 72 h and SNAP25 cleavage measured using the one-step ELISA.Specific signal was only observed with BoNT/A treated NanoLuc-SNAP25 reporter cells (red bars) (n=2 and error bars show experimental range).Control: Differentiated and untreated reporter cells.C. LAN5 cells expressing Nanoluc-SNAP25 were differentiated for 3 days and incubated with the indicated concentrations of BoNT/A for 72 h and SNAP25 cleavage measured using the one-step ELISA.The one-step ELISA allows detection of cleaved SNAP25 at femtomolar doses of BoNT/A (n=3, **p= 0.0015, ****p<0.0001,Unpaired t-test).D. LAN5 cells expressing NanoLuc-SNAP25 were differentiated for 3 days and incubated with the indicated units of Botox® 72 h and SNAP25 cleavage measured using the one-step ELISA.The assay can detect a single unit of Botox® (n=2 and error bars show experimental range).E. LAN5 cells expressing NanoLuc-SNAP25 were differentiated for 3 days and incubated with Botox® (15U) which had been left at 20 °C for increasing periods of time.SNAP25 cleavage was then measured using the one-step ELISA (n=3, *p<0.05,****p<0.0001,Oneway ANOVA).NTC: non-treated control.ALU: arbitrary luminescence units.

Fig. 4 :
Fig. 4: SV2A overexpression improves the performance of the CBA for detecting BoNT/A A. LAN5 cells expressing NanoLuc-SNAP25 were transduced with either SV2A or C and stained with either anti-FLAG tag (SV2Atransduced cells) or anti-V5 tag (SV2C-transduced cells).Nuclei were stained with DAPI.Only NanoLuc-SNAP25 expressing LAN5 cells were used as a negative control.Images were taken at 40X magnification (Scale bars: 20 µm).Both SV2 constructs were expressed in the transduced cells.B. NanoLuc-SNAP25 LAN5 cells overexpressing SV2A or SV2C were differentiated for 3 days, incubated with the indicated concentrations of BoNT/A for 72 h and SNAP25 cleavage measured using the one-step ELISA.The data was then plotted as a fold change of the parental NanoLuc-SNAP25 cells.SV2A overexpressing cells show an enhanced response to BoNT/A (n=2, Error bars show experimental range).C. SV2A overexpressing NanoLuc-SNAP25 LAN5 cells cells were differentiated for 3 days, incubated with the indicated concentrations of BoNT/A for 72 h and SNAP25 cleavage measured using the one-step ELISA.The SV2A overexpressing cells were capable of detecting 10 fM BoNT/A with signal which was approximately 2-fold greater than the parental cells (n=3, Mean (SD), ***p= 0.0002, ****p<0.0001,Unpaired t test).D. SV2A overexpressing NanoLuc-SNAP25 cells were differentiated for three days, incubated with 100 pM BoNT/A for either 24 or 48 h and SNAP25 cleavage measured using the one step ELISA (n=3, Mean (SD), ****p<0.0001,Two-way ANOVA).Overexpression of SV2A allows the robust detection of BoNT/A with shortened incubation periods.