Production of monoclonal antibodies against botulinum neurotoxin in Nicotiana benthamiana

ABSTRACT

Botulism is a fatal neurologic disease caused by the botulinum toxin (BoNT) produced by Clostridium botulinum. It is a rare but highly toxic disease with symptoms, such as cramps, nausea, vomiting, diarrhea, dysphagia, respiratory failure, muscle weakness, and even death. Currently, two types of antitoxin are used: equine-derived heptavalent antitoxin and human-derived immunoglobulin (BabyBIG®). However, heptavalent treatment may result in hypersensitivity, whereas BabyBIG®, has a low yield. The present study focused on the development of three anti-BoNT monoclonal antibodies (mAbs), 1B18, C25, and M2, in Nicotiana benthamiana. The plant-expressed mAbs were purified and examined for size, purity and integrity by SDS-PAGE, western blotting and size-exclusion chromatography. Analysis showed that plant-produced anti-BoNT mAbs can fully assemble in plants, can be purified in a single purification step, and mostly remain as monomeric proteins. The efficiency of anti-BoNT mAbs binding to BoNT/A and B was then tested. Plant-produced 1B18 retained its ability to recognize both mBoNT/A1 and ciBoNT/B1. At the same time, the binding specificities of two other mAbs were determined: C25 for mBoNT/A1 and M2 for ciBoNT/B1. In conclusion, our results confirm the use of plants as an alternative platform for the production of anti-BoNT mAbs. This plant-based technology will serve as a versatile system for the development botulism immunotherapeutics.

KEYWORDS:

• Botulism
• botulinum neurotoxin
• Clostridium botulinum
• monoclonal antibody
• Nicotiana benthamiana

Introduction

Botulism is a neuroparalytic disease caused by the systemic effects of botulinum neurotoxin (BoNT), which is produced by gram-positive bacterium Clostridium botulinum. BoNT is one of the most potent toxins known, with seven main serotypes (A, B, C, D, E, F, G) discovered¹, and a new serotype H lately identified.² Botulism outbreaks in humans have been linked to BoNTs types A, B, and E, and, in rare cases, F and G. On the contrary, BoNT types C and D cause animal botulism.³ Those exposed to BoNTs often experience initial gastrointestinal symptoms such as cramps, nausea, vomiting, and diarrhea, which appear within 2–3 h of exposure; neurologic manifestations include ocular disturbances (blurred vision), dysphagia, followed by respiratory failure, muscle weakness, and death.⁴ Botulism can be classified into three types: foodborne, wound, and infant botulism. Foodborne botulism is caused by consuming foods contaminated with C. botulinum, and it most commonly occurs when canned foods are improperly preserved or stored under vacuum conditions. BoNT serotype A (BoNT/A) is the most toxic substance responsible for human botulism, with an estimated lethal dose (in a 70 kg person) of 70 µg by oral administration, 0.09–0.15 µg by intravenous injection, and 0.70–0.90 µg by inhalation of spores.⁵ Thus, BoNT/A was designated as category A toxic, with the potential to be used in terrorist attacks and biological warfare, posing a serious threat.⁶ BoNT is made up of a heavy chain (100 kDa) and a light chain (50 kDa) connected by a disulfide bond.⁷ The mechanism by which BoNT functions begins with the heavy chain binding to extracellular receptor cholinergic nerve terminals, and then the light chain cleaves the SNARE proteins, preventing the formation of synaptic fusion complex,^8,9 thereby inhibiting acetylcholine release, and causing muscle weakness.¹⁰ However, botulinum antitoxin can inhibit neurotoxin activity. Previous research has shown that intravenous administration of botulinum antitoxin within the first 24 h after exposure can reduce the effects of BoNT in vivo.¹¹

Botulism is currently treated with heptavalent botulinum antitoxin derived from horses.¹² Nonetheless, the use of hyperimmune serum from animals often results in low specific activity, batch-to-batch variability, disease risk due to pathogen transmission, and the potential for life-threatening hypersensitive immune responses.¹³ These issues can be addressed through genetic engineering techniques such as the production of recombinant monoclonal antibodies (mAbs), which reduces the risk of pathogen hypersensitivity.¹⁴ Yet, producing mAbs in animal cells is costly due to the need for a large, expensive bioreactor and limited cultivation space, resulting in the inability to produce enough to meet demand. In a previous report, the cost of goods sold for mAbs production using animal cell culture platform was estimated to be around $232 USD per gram, while the plant-based platform showed a cost of goods sold of around $99 USD per gram.¹⁵ These findings demonstrate that the plant expression system reduces cost of goods sold by more than 50% when compared to the animal cell system.¹⁶ For this reason, rather than using bioreactor tanks, plant-based technology has been introduced to reduce manufacturing costs and is regarded as a safe alternative. Plant systems are effective and more economical than traditional methods due to their ease of cultivation, low or no animal pathogen contamination, high production level, and low cost.¹⁷ Furthermore, plants for pharmaceutical production have the advantage of being easily scalable without incurring high costs.¹⁸ Several studies have reported the use of plant expression systems to produce mAbs and antibody fragments, as well as recombinant proteins.^19–23 In particular, Nicotiana benthamiana has been widely used in plant virology owing to its low nicotine content, nonfood status, and high biomass production.^24,25 However, post-translational modifications, such as glycosylation, have been shown to have a significant impact on the functionality of plant-derived proteins.^26–28 Plant N-glycans primarily consist of β1,2-xylose (Xyl) and α1,3-fucose (Fuc) linked to the mannosylchitobiosyl core region. Meanwhile, when proteins contain an ER retrieval motif KDEL sequence, glycoproteins can be retained or returned to the ER, and such proteins carry typical oligomannose-type residues.^29–31 However, some studies have shown that retaining proteins in the ER increase production levels.^27,32 Williams et al.³³ found that glycosylation has an impact on the pharmacodynamics (PD) of mAbs, indicating that mAb activity is dependent on glycosylation. Another study by Koide et al.³⁴ found that deglycosylase rabbit antibodies lost their ability to activate antibody-dependent cell-mediated cytotoxicity (ADCC) and complement binding. Nose and Wigzell,³⁵ on the other hand, revealed that glycosylation had profound effects on Fc receptor binding and complement activation, but had little effect on antigen binding. In this study, we aimed to generate mAbs targeting BoNT from N. benthamiana, which can solve the problem of expensive antitoxins by utilizing a low-cost scalable plant platform. The plant-produced anti-BoNT mAbs could be a viable and safe immune therapeutic candidate for botulism treatment.

Materials and methods

Construct design for gene synthesis

The amino acid sequences of anti-BoNT mAbs (1B18, C25, and M2) were retrieved^36–38 and codon optimized using Invitrogen GeneArt® Gene Synthesis (Thermo Scientific, USA) for expression in N. benthamiana. The heavy chain (HC) and light chain (LC) genes were commercially synthesized and fused to the constant domains of human IgG1 gamma chain and kappa chain, respectively. Additionally, a murine signal peptide sequence was carried at the N-terminus, as well as a SEKDEL tag at the C-terminus of HC (Figure 1). The full-length anti-BoNT HC (Figure 1a) and LC (Figure 1b) were cloned into pBY2eK geminiviral vector after digesting with XbaI and SacI (BioLabs, Massachusetts, USA), and the recombinant pBY2eK-HC and pBY2eK-LC were then transferred into Escherichia coli (DH10B strain) by heat shock method. The screening of transformed colonies was carried out by colony PCR with specific primer pairs, as shown in Table 1. Positive clones were cultured in Luria Bertani (LB) media supplemented with 50 mg/L kanamycin and incubated overnight at 37°C. Plasmids were extracted using a DNA-spin plasmid DNA Purification kit (iNtRon Biotechnology, Korea) and then transferred into Agrobacterium tumefaciens (GV3101 strain) by electroporation.

Figure 1. Schematic illustration of anti-BoNT heavy chain (a) and light chain (b) genes in pBy2eK plant expression vector.

Table 1. Primer sequences used for colony PCR.

Primer	Sequence
Forward primer:	GGA GAG GAC CTC GAG AAA C
Reverse primer:	GCT TTG CAT TCT TGA CAT C

Meanwhile, recombinant BoNT/A and B proteins were generated by overlapping PCR using direct primers listed in Table 2. The BoNT/A (GenBank accession number: ABS38337) and BoNT/B (NCBI: WP_012291519.1) sequences were engineered to contain point mutations of E²²⁴A, R³⁶³A, Y³⁶⁶F for mutated BoNT/A (M-BoNT/A1) and H²³⁰A, E²³¹A, H²³⁴A for catalytically inactive BoNT/B (ciBoNT/B1). The resulting M-BoNT/A1 and ciBoNT/B1 proteins were inactive versions of native toxins. Moreover, a polyhistidine tag (8× His tag) was fused at the C-terminus of recombinant BoNT/A and B. All proceeding steps, including cloning into pBY2eK expression vector, bacterial transformation, colony screening and plasmid extraction, were performed as described above.

Table 2. Primer sequences used to generate mutated BoNT/A and B sequences.

Primer	Sequence
Forward primer: E^224A	GAC TCT TGC TCA CGC TCT TAT TCA TG
Reverse primer: E^224A	CAT GAA TAA GAG CGT GAG CAA GAG TC
Forward primer: R^363A, Y^366F	GCT GAA CGC TAA GAC CTT CCT CAACTT CG
Reverse primer: R^363A, Y^366F	CGA AGT TGA GGA AGG TCTTAG CGT TCA GC
Forward primer: H^230A, E^231A, H^234A	CTT ATG GCT GCC CTT ATT GCT GTG CTG C
Reverse primer: H^230A, E^231A, H^234A	GCA GCA CAG CAA TAA GGG CAG CCA TAA G
Forward primer: XbaI-SP2	CCT CTA GAA TGG CTA ACA AGC ACC TCT CAT TGT CTC TCT TCC TTG TGC TCC TTG GTC TTT CTG CTT CTC TTG C
Reverse primer: SacI-KD	CTC TAG AAC AAT GGG CTG GTC C

Expression of anti-BoNT/A mAbs in Nicotiana benthamiana

The A. tumefaciens containing anti-BoNT HC and LC genes were individually cultivated in LB broth supplemented with 50 mg/L of kanamycin, gentamycin (AppliChem, Darmstadt, Germany), and rifampicin (Olon active pharmaceutical, Italy) for overnight at 28°C. The cultures were diluted with infiltration buffer (10 mM 2-N-morpholino-ethanesulfonic acid (MES) and 10 mM MgSO₄, pH 5.5) to give a final optical density at 600 nm (OD₆₀₀) of 0.2. The bacteria containing HC and LC vectors of anti-BoNT mAbs were injected at a ratio of 1:1 into 4–6-week-old N. benthamiana leaves using syringe agroinfiltration. Infiltrated tobacco plants were collected on 2-, 4-, 6-, 8-, and 10-days post-infiltration (dpi) in order to examine the antibody expression levels. The crude proteins were extracted with 1× PBS (137 mM NaCl, 2.7 mM KCl, 4.3 mM Na₂HPO₄, 1.47 mM KH₂PO₄, pH 7.4), and the protein yield was quantified by sandwich ELISA. Then, anti-BoNT mAb production was scaled-up using vacuum agroinfiltration, and transfected leaves were harvested at optimal dpi.

To quantify mAbs by ELISA, a 96-well plate was coated with goat polyclonal anti-human IgG Fc-specific antibody (ab97221, Abcam, United Kingdom) diluted to 2 µg/ml and incubated overnight at 4°C. Coated plates were washed twice with 1× PBS-T (0.05% Tween 20 in 1× PBS) and once with 1× PBS before adding 5% (w/v) skim milk blocking buffer. Plates were incubated for 1 h at 37°C followed by washing three times with 1× PBS-T. Anti-BoNT mAbs 1B18 and M2 were diluted 500×, while C25 was diluted 1,000×. Human IgG1 kappa isotype control (ab206198, Abcam, United Kingdom) was used as the standard and was serially diluted two-fold (starting at 125 ng/mL). The samples and standard were added to the wells and incubated for 2 h at 37°C. Captured mAbs were detected using HRP-conjugated anti-human kappa antibody (2060–05, Southernbiotech, USA) diluted 1:2,500. Color development was performed using 3,3,’5,5’-tetramethylbenzine (TMB) substrate (Promega, USA) and stopped using 1 M H₂SO₄. Optical density was measured at 450 nm (OD₄₅₀) using a microplate reader (Thermo Scientific, USA).

Protein extraction and purification

Leaves from infiltrated plants were collected at 4 dpi, stored frozen, and then used for extraction with ice-cold 1× PBS in a 1:2 (w/v) ratio. Frozen leaves were homogenized by grinding with a blender (OTTO King Glass, Thailand) for 5 min at room temperature. Crude extracts were clarified by centrifugation (13,000 rpm, 45 min), and the supernatant was loaded onto an affinity chromatography column. The plant-produced anti-BoNT mAbs were purified using a protein A resin column, washed with 1× PBS, and eluted with 0.1 M sodium citrate (9.38 mM C₆H₉Na₃O₉, 90.62 mM C₆H₈O₇, pH 3). The purified proteins were immediately neutralized with 1.5 M tris-HCl at pH 8.8.

For recombinant M-BoNT/A1 and ciBoNT/B1, frozen infiltrated leaves were extracted with a commercial blender in ice-cold imidazole (IMAC 5; 20 mM Tris-HCl pH 7.4, 50 mM NaCl, and 5 mM Imidazole) at a 1:2 (w/v) ratio. Recombinant proteins were purified in a column packed with Ni Sepharose^TM 6 Fast Flow (Cytiva, United Kingdom) resin. The column was then washed with washing buffer (IMAC 30; 20 mM Tris-HCl pH 7.4, 50 mM NaCl, and 30 mM Imidazole), and proteins were collected with elution buffer (IMAC 250; 20 mM Tris-HCl pH 7.4, 50 mM NaCl, and 250 mM Imidazole). Finally, plant-produced anti-BoNT mAbs and BoNT/A and B proteins were buffer-changed to 1× PBS using a 3.5 kDa dialysis membrane (Thermo Scientific, USA) and concentrated using Amicon® 50K centrifugal filters (Merck Millipore, Ireland). The purified proteins were filter-sterilized using sterile syringe filter 0.22 um (Filter-Bio, China) in an aseptic laminar flow cabinet (LaboGene, Denmark). The purity was determined by reduced and non-reduced SDS-PAGE.

SDS‐PAGE and western blot analyses

SDS-PAGE and immunoblotting were used to assess the size and integrity of purified mAbs and mutated BoNTs. Briefly, proteins were separated on a 4–15% polyacrylamide gel under reducing (125 mM Tris-HCl pH 6.8, 12% (w/v) SDS, 10% (v/v) glycerol, 22% (v/v) beta-mercaptoethanol, 0.001% (w/v) bromophenol blue) and non-reducing (125 mM Tris-HCl pH 6.8, 12% (w/v) SDS, 10% (v/v) glycerol, 0.001% (w/v) bromophenol blue) conditions. Proteins were either stained with InstantBlue® dye or transferred to nitrocellulose membrane. The membrane was blocked with 5% (w/v) skim milk to prevent nonspecific binding, and then incubated with HRP-conjugated anti-human IgG antibody (2040–05, Southernbiotech, USA) diluted 1:10,000 and HRP-conjugated anti-human kappa antibody diluted 1:2,500 for mAb detection or HRP-conjugated anti-His antibody (652503, Biolegend, USA) diluted 1:10,000 for mBoNT/A1 and ciBoNT/B1 detection. Finally, the probed mAbs and recombinant BoNTs were treated with enhanced chemiluminescence (ECL) reagent (Promega, USA) and exposed to X-ray film (Carestream, USA).

Size exclusion chromatography

Size-exclusion chromatography (SEC) was performed with Yarra SEC-3000 GFC column (300 mm × 7.8 mm, Phenomenex, USA) mounted on UHPLC system (Waters, USA) to further examine the purity of plant-produced anti-BoNT mAbs. The proteins were eluted with 0.1% sodium azide in 1× PBS buffer (0.1% w/v NaN₃, 100 mM Na₃PO₄, pH 6.8) at a flow rate of 0.3 mL/min. The ultraviolet absorbance was detected at a wavelength of 280 nm, and the peak area was integrated by Empower 3 software (Waters, USA).

In vitro binding ELISA

About 2 µg/ml of mBoNT/A1 and ciBoNT/B1 were coated on 96-well plates and incubated overnight at 4°C. Coated plates were washed two times with 1× PBS-T and once with 1× PBS. Plates were blocked with 5% (w/v) skim milk for 1 h at 37°C, then washed three times with 1× PBS-T. Anti-BoNT mAbs were three-fold serially diluted (starting at 40 µg/mL), added to the wells, and incubated for 2 h at 37°C. Bound mAbs were detected using HRP-conjugated anti-human kappa antibody diluted 1:2,500. Plates were further incubated for 2 h at 37°C before washing three times with 1× PBS-T. Binding was determined using TMB substrate, and the reaction was quenched using 1 M H₂SO₄. Optical density (OD₄₅₀) was measured using a microplate reader.

Results

Expression of anti-BoNT mAbs and recombinant BoNT/A and B in N. benthamiana

The A. tumefaciens containing the pBY2eK-HC and pBY2eK-LC vectors were transferred into N. benthamiana, and the leaves were harvested at 2, 4, 6, 8, and 10 dpi. Inoculated leaves developed necrosis, and the severity of the necrotic symptoms increased with the dpi (Figure 2a). In particular, necrosis was visible 6 days after agroinfiltration, with evident dryness and white spots at infiltrated leaf areas, indicating recombinant protein expression. The collected leaves were then homogenized, and the crude extracts were used to determine the antibody content by ELISA. Based on the results, anti-BoNT mAbs 1B18, C25, and M2 had the highest expression levels at 6 dpi, accumulating up to 54.5 µg/g, 144.4 µg/g, and 111.9 µg/g fresh weight, respectively (Figure 2b).

Figure 2. Day optimization of anti-BoNT mAbs expressed in N. benthamiana. (a) Development of necrosis in Agrobacterium-infected tobacco leaves on days 2, 4, 6, 8, and 10 after infiltration (dpi). MAb expression induced necrotic symptoms at the inoculation leaf spot. (b) ELISA was used to determine the mAb yield in crude extracts from 2, 4, 6, 8, and 10 dpi.

Furthermore, crude proteins were used for analyzing specific mAb expression via western blot. For HC detection, 1B18, C25, and M2 mAbs were detected with an HRP-conjugated anti-human gamma antibody, and the results revealed protein bands at approximately 150 kDa in non-reducing condition for the full-length mAbs (Figure 3a) and approximately 50 kDa in reducing condition for the HC fragments (Figure 3b). Meanwhile, for LC detection, the three anti-BoNT mAbs were detected using HRP-conjugated anti-human kappa antibody, and findings showed bands at approximately 150 kDa (Figure 3c) and 25 kDa (Figure 3d) for non-reduced mAbs and reduced LC fragments. Altogether, recombinant 1B18, C25, and M2 were produced in N. benthamiana and migrated at expected molecular masses in SDS-PAGE and immunoblots.

Figure 3. Transient expression of anti-BoNT mAbs in plants. Western blot analysis of 1B18, C25, and M2 to determine mAb expression using HRP-conjugated anti-human gamma and anti-human kappa IgGs under non-reducing (a,c) and reducing (b,d) conditions. Lanes 1, 2, and 3 represent 1B18, C25, and M2.

Plant crude extracts producing recombinant BoNT/A and B were also tested for protein expression by western blot, and the plant-produced BoNTs were detected with an HRP-conjugated anti-His tag antibody. Both the mBoNT/A1 and ciBoNT/B1 have bands at approximately 150 kDa under reducing condition, as predicted (Figure 4). These results indicate that recombinant BoNTs were successfully expressed in N. benthamiana and used as coat proteins for in vitro binding analysis.

Figure 4. Expression analysis of BoNT/A and B proteins. The expression of plant-produced mBoNT/a1 and ciBoNT/b1 were analyzed by western blot probed with anti-his-HRP antibody under reducing condition. Lane M as protein molecular weight marker; lanes 1 and 2 represent mBoNT/a1 and ciBoNT/b1.

Purification of anti-BoNT mAbs produced in N. benthamiana

The plant-expressed anti-BoNT 1B18, C25, and M2 were purified using protein A column, and elute fractions were evaluated for size and purity by reduced and non-reduced SDS-PAGE and western blotting (Figure 5). Under the non-reducing condition, all three mAbs appear as distinct bands around 150 kDa in the polyacrylamide gel (Figure 5a). On the other hand, reduced samples of mAbs showed HC of about 50 kDa and LC of about 25 kDa (Figure 5d). More importantly, based on visual inspection, the clear bands corresponding to 1B18, C25, and M2 indicated high purity for plant-produced mAbs. These SDS-PAGE data were also consistent with the western blot results. When the samples were probed with HRP-conjugated anti-human gamma chain (Figures 5b, e) or HRP-conjugated anti-human kappa chain (Figures 5c, f), bands of 150, 50, and 25 kDa were detected, corresponding to the full-length mAbs and their HC and LC fragments. SEC was also used to assess the purity and aggregation of plant-produced anti-BoNT mAbs. As demonstrated in Figure 6, a sharp peak (major peak) was observed for each mAb, which represented the affinity-purified 1B18, M2, and C25. Moreover, the presence of high molecular weight aggregated forms (minor peaks) that eluted prior to the main monomer peak was seen in all three mAbs, but no mAb fragments were found in the chromatogram. Based on our results, plant-derived 1B18 (Figure 6a), C25 (Figure 6b), and M2 (Figure 6c) largely displayed intact mAb with monomeric assembly, accounting for 89.14%, 95.76%, and 87.03%, respectively.

Figure 5. Purification of plant-produced anti-BoNT mAbs by protein a affinity chromatography. The purity of 1B18, C25, and M2 were analyzed by SDS-PAGE and stained with InstantBlue® dye under non-reducing (a) and reducing (d) conditions. Purified anti-BoNT mAbs were detected by Western blot probed either with anti-human gamma chain or anti-human kappa chain antibodies under non-reducing (b,c) and reducing (e,f). Lane M as protein molecular weight marker; lanes 1, 2, and 3 represent 1B18, C25, and M2.

Figure 6. Size exclusion profile of purified plant-produced anti-BoNT mAbs. Lanes a, b, and c represent 1B18, C25, and M2.

Additionally, the plant-produced mBoNT/A1 and ciBoNT/B1 were purified using a nickel column, and the affinity-purified proteins were examined for integrity by SDS-PAGE and immunoblotting. According to the SDS-PAGE results, the observed molecular weights of mBoNT/A1 and ciBoNT/B1 proteins were around 150 kDa (Figure 7a). Western blot analysis showed that anti-His antibody recognized both mBoNT/A1 (Figure 7b) and ciBoNT/N1 (Figure 7c) and were detected at expected sizes, which is consistent with the SDS-PAGE analysis.

Figure 7. Purification of plant-produced BoNT/A and B by nickel affinity chromatography. (a) SDS-PAGE analysis of purified plant-derived BoNTs under reducing condition. Lane M as protein molecular weight marker; lanes 1 and 2 represent mBoNT/a1 and ciBoNT/b1. Western blot analysis of mBoNT/a1 (b) and ciBoNT/b1 (c) probed with HRP-conjugated anti-his tag antibody. Lane M as protein molecular weight marker; C as the crude protein extracts; E as the elute fractions.

Binding of plant-produced anti-BoNT mAbs against BoNT proteins

BoNT binding was measured by ELISA, and inactive M-BoNT/A1 and ciBoNT/B1 proteins were produced in N. benthamiana as immobilized target antigens. Given that mBoNT/A1 and ciBoNT/B1 proteins are similar to the native neurotoxin,^9,39 they were used to coat the plates as relevant alternatives for highly toxic BoNTs. Three-fold serial dilutions of 1B18, C25, and M2 were captured on the mBoNT/A1 or ciBONT/B1 before HRP-conjugated anti-human kappa antibody was added for detection. Plant-produced anti-SARS-CoV-2 H4 (N) was used as the negative control. When plant-produced mAbs were captured on plant-derived mBoNT/A1, 1B18 bound more strongly than C25 and M2, but M2 bound very weakly (Figure 8a). Meanwhile, when plant-produced mAbs were captured on plant-derived ciBoNT/B1, M2 bound highest at low mAb concentrations, while 1B18 bound in a concentration-dependent manner. In contrast, C25 did not bind to ciBoNT/B1, which is comparable to the negative control (Figure 8b).

Figure 8. Binding affinity of plant-produced anti-BoNT mAbs to recombinant BoNT targets. ELISA binding of three-fold serially diluted 1B18, C25, and M2 to (a) plant-produced mBoNT/a1 and (b) plant-produced ciBoNT/b1 proteins. N: plant-based anti-SARS-CoV-2 H4 mAb.

Discussion

BoNT is the most poisonous toxin ever discovered,⁴⁰ with even trace amounts causing fatal neuroparalytic disease and death. Current treatment options include heptavalent botulinum antiserum¹² and human-derived botulism immunoglobulins (BabyBIG®, Botulism Immune Globulin Intravenous (Human), BIG-IV).⁴¹ However, antitoxins are difficult to access due to their high price; for example, the current cost for a single dose of Baby BIG® is $45,300 USD.⁴² In addition, the main concerns with these conventional botulinum antitoxins are hypersensitivity reactions caused by equine anti-BoNT serum¹³ and low yield limitation for BabyBIG®.⁵ Hence, developing a new class of antitoxins against BoNTs is essential for potential botulism therapies. Previous literatures reported the use of mAbs to block the activity of BoNT, which reduced the severity of neurotoxin in several animal models.^43–46 Anti-BoNT mAbs bind specifically to BoNTs, inhibiting the entry of toxin and hydrolysis of SNARE proteins. As a result, SNARE proteins can carry the neurotransmitter vesicles and mediate normal neurotransmitter release into the synaptic cleft.⁴⁷ As described, mAbs hold great promise as BoNT therapeutics and could thus be utilized in the treatment and prevention of botulism.³⁸ In order to study the potential of mAbs against BoNTs, alternative approaches to high-quality large-scale production have been explored. Prior studies described the advantages of plant-based protein technology for low cost of goods manufactured.⁴⁸ Likewise, the production scale in plants can be rapidly expanded. Plant-based mAb production was found to be economically competitive with that of CHO cells. A cost model data for mAb yields in plants, reaching 2.0 mg kg^{− 1},⁴⁹ and with filter capacities of 100–1000 L m^{− 2} to reduce consumable expenses,⁵⁰ indicated that the production costs for plant-derived mAbs can significantly decrease to below $108 USD per gram.⁵¹ In contrast, mAbs produced in CHO cells cost approximately $216 USD per gram.⁵² Plant systems are also single-use technologies employed to produce a single product prior to homogenization.^53,54 This, in turn, lowers the risk of product cross contamination in plants throughout upstream production, without requiring extra capital or increased consumables fees. As compared to multi-use fermenters for multiple protein products, this minimizes cleaning and validation processes, as well as associated labor and costs. Therefore, the inherent single-use property of plants reduces production costs, resulting in large cost savings at the production scale.^55–57 Moreover, plants provide the benefits of simple cultivation, where sterility is not necessary during production, as intact plants can rely on their natural immunity to protect themselves from pathogens.⁵⁸ However, unlike cell cultures, exposing a product to non-sterile and non-aseptic conditions may result in regulatory scrutiny.^59,60 In this work, the production of antibody-based BoNT antitoxins such as mAbs 1B18, C25, and M2 in N. benthamiana was investigated.

Here, we focused on three anti-BoNT mAbs: 1B18, C25, and M2. Previous research has shown that C25 and M2 are highly effective at blocking BoNT/A or B toxins in mouse models,^38,61 whereas 1B18 can capture both BoNT/A and B toxins.⁶² Since antibody cocktail therapy has been shown to improve the overall efficiency of mAbs,⁶¹ mAbs 1B18, C25, M2 were produced using a plant expression platform to test the potential of a three-mAb treatment in future studies. Anti-BoNT mAbs were transfected into tobacco leaves and examined for protein expression. After infiltration, necrosis was observed in Agrobacterium-injected leaves, which at the same indicated protein expression. As expected, non-infiltrated leaf areas showed no necrosis symptoms. The expression level of all mAbs increased at 6 dpi, and then decreased at 8 dpi, attaining its lowest level at 10 dpi.⁶³ The highest yield of recombinant anti-BoNT mAbs expressed in plants was obtained at 6 dpi, with up to 144.4 µg/g of C25, 111.9 µg/g of M2, and 54.5 µg/g of 1B18, respectively. An earlier study presented a comparison of yields between mAbs produced in a plant-based platform and those produced in conventional culture conditions.⁶⁴ For instance, the anti-PD-1 mAb produced in plants demonstrated a yield of 344 μg/g leaf fresh weight, whereas the CHO-cell-produced anti-PD-1 yielded only around 3–6 μg/mL of mAb.⁶⁵ Similarly, the production of recombinant mAbs was found to be faster in plants compared to other expression platforms.^66,67 Specific detection of mAbs was performed by western blot assay. Under non-reduced condition, anti-human gamma antibody reacted primarily with an approximately 150 kDa protein band. Meanwhile, proteins with molecular masses of approximately 50 kDa and 25 kDa reacted with anti-human gamma and kappa chains under reduced conditions with beta-mercaptoethanol. Multiple bands on the blot can be caused by loading too much lysate onto a gel, which leads the detection antibody to bind nonspecifically to proteins in abundance.⁶⁸ A fully assembled IgG has a molecular weight of 150 kDa; we thereby conclude the presence of two heavy chains and two light chains linked together by disulfide bonds.⁶⁹ The plant-produced anti-BoNT mAbs were purified by protein A affinity chromatography, and their purity was determined by means of SDS-PAGE and a subsequent immunoblot. Protein bands of 150, 50, and 25 kDa were detected, most likely corresponding to monomeric IgG and its reduced HC and LC forms.

Recombinant mBoNT/A1 and ciBoNT/B1 proteins were also transiently expressed in N. benthamiana and purified for use as coating antigens. The binding specificities of plant-produced 1B18, C25, and M2 against mBoNT/A1 and ciBoNT/B1 were determined by ELISA technique. Among these mAbs, 1B18 had the highest affinity for mBoNT/A1, followed by C25, and M2 had the lowest affinity. Meanwhile, plant-produced M2 was found to have the highest antigen-binding activity to ciBoNT/B1, followed by 1B18. C25 was unable to bind to ciBoNT/B1. Previous studies have shown that 1B18 has binding affinities for both BoNT/A and BoNT/B,⁷⁰ which is consistent with our findings. Similarly, C25 did not recognize BoNT/B but was specific for BoNT/A,⁷¹ and M2 did not recognize BoNT/A but specific for BoNT/B.³⁸ Thus, three mAbs are required to potently neutralize toxin. Other literature on neutralization of toxin revealed that C25 and M2 mAbs were highly effective in inhibiting the toxicity of BoNT in vivo. All mice were completely protected when given 100 µg of M2 mixed with 10 LD₅₀ of BoNT/B1,³⁸ and although 50 µg of C25 failed to protect mice against 20 LD₅₀ BoNT/A challenge, it did demonstrate to prolong the time of death.⁶¹ Tacket et al. discovered that 115 of 132 patients who received antitoxin experienced a significant decrease in mortality.¹¹ The administration of antitoxin within the first 24 h of symptom onset resulted in a 36% reduction in mortality. Furthermore, even when the antitoxin was administered after 24 h of symptom onset, the death rate decreased by 31% compared to untreated patients. On the other hand, despite the fact that no studies have reported 1B18 neutralizing effects, antigen-binding tests have revealed high efficiency for binding with both BoNT/A and B, supporting earlier research.³⁷ Mukherjee et al. developed six sheep-derived mAbs (SMabs) against BoNT/A1: 1G4, 2G11, 4F7, 5E2, 5F7, and 16F9. These SMabs were tested in a mouse model of botulism.⁷² Although each SMAb bound to the BoNT/A1 toxin, their specific-binding properties to native and recombinant BoNT/A1 varied. Out of the six SMabs, 2G11, 4F7 and 5E2 displayed effective binding to plate-bound BoNT/A1, whereas 1G4, 5F7 and 16F9 demonstrated minimal or no binding to insoluble forms of BoNT/A1. Interestingly, only 1G4, 5F7, and 16F9 were determined to contribute to protective efficacy. These findings show a lack of correlation between in vitro binding and in vivo protective efficacy, as 2G11, 4F7 and 5E2 SMabs bound strongly to plate-bound BoNT but did not bind soluble BoNT/A1, the form present in vivo. Hence, the study suggests that SMabs ability to bind soluble BoNT/A1 may correlate to their protective efficacy. Nonetheless, relevant references discovered that combining mAbs increases the effectiveness of neutralizing antibodies even more.^38,61 The in vivo neutralizing activities of plant-produced 1B18, C25, and M2, as well as the potential of combinatorial mAb treatment to further improve neutralization efficiency, will be explored in the future.

In summary, this study demonstrated that three anti-BoNT mAbs can be successfully produced in N. benthamiana with high yield and purity. Furthermore, plant-produced 1B18, C25, and M2 have high affinity and recognized recombinant BoNT/A and B, which will be useful in the development of plant-based therapeutics for botulism treatment.

Author contributions statement

Waranyoo Phoolcharoen conceived and designed the study. Kornchanok Sangprasat, Christine Joy I. Bulaon, Kaewta Rattanapisit, and Perawat Jirarojwattana performed construct design for gene synthesis. Kornchanok Sangprasat and Apidsada Wongwatanasin performed gene cloning. Kornchanok Sangprasat performed expression, purification experiments, SDS-PAGE, western blot analysis. Kornchanok Sangprasat and Theerakarn Srisangsung performed in vitro binding assay. Kornchanok Sangprasat interpreted the data. Kornchanok Sangprasat and Christine Joy I. Bulaon drafted the manuscript. All authors reviewed the manuscript.

Disclosure statement

Waranyoo Phoolcharoen is a founder/shareholder of Baiya Phytopharm Co., Ltd. Christine Joy I. Bulaon, Kaewta Rattanapisit, and Apidsada Wongwatanasin are employees of Baiya Phytopharm Co., Ltd. No potential conflict of interest was reported by the remaining authors.

Data availability statement

The data presented in this study are available from the corresponding author upon reasonable request.

Additional information

Funding

This study was supported by the National Vaccine Institute (NVI) of Thailand Science Research and Innovation, the Second Century Fund (C2F) scholarship of Chulalongkorn University, Thailand, the NSRF via the Program Management Unit for Human Resources & Institutional Development, Research, and Innovation [grant number B13F660137], and the 90th Anniversary of Chulalongkorn University, Rachadapisek Sompote Fund.