Design and Synthesis of Peptides from Phoneutria nigriventer δ-Ctenitoxin-Pn2a for Antivenom Production

The venom toxin δ-ctenitoxin-Pn2a of the spider Phoneutria nigriventer can cause severe envenomation in humans. Furthermore, the cystine-knot motif of δ-ctenitoxin-Pn2a provides exceptional stability, thereby hampering immune response activation. Here we identified epitope G34YFWIAWYKLANCKK48 from δ-ctenitoxin-Pn2a through the Immune Epitope Database Analysis Resource and used it to design antigenic peptides. The Cys residue was replaced by α-aminobutyric acid (Abu) to prevent disulfide bond formation. To increase the immunogenicity of these molecules, branched and N-palmitoylated versions were synthesized. Ac-GYFWIAWYKLAN-Abu-KKG-NH2 (A), Palm-GYFWIAWYKLAN-Abu-KKG-NH2 (B) and (Ac-GYFWIAWYKLAN-Abu-KK)2-KG-NH2 (C) were prepared by solid-phase synthesis and their identity was confirmed by ESI–MS. They were then studied by RP-HPLC and all the chromatograms obtained showed only one main peak. Cytotoxicity was evaluated on the murine macrophage cell line RAW 264.7 using the 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide assay in the presence of increasing doses of each peptide (0.25–10.0 µM). Peptide A did not show cytotoxicity between 0.25 and 10.0 µM, while B and C did at concentrations equal or over 0.5 and 10.0 respectively. The cellular distribution of NF-κB was examined by immunofluorescence after exposing macrophages to 0.5 µM of each peptide. Early activation was observed for all three peptides, thereby indicating that they are promising immunogens for antivenom production. Nevertheless, in vivo tests are still required to assess their immunogenic capacity and whether the antibodies generated can confer protection against the venom.


Introduction
Spiders belonging to the genus Phoneutria, also known as banana or wandering spiders, are among the most venomous spiders of medical significance for humans in the world. Phoneutria nigriventer, which is found in South America-mainly in Brazil and northern Argentina-is the species with the greatest geographical distribution and clinical relevance (Bucaretchi et al 2018). This large spider has developed a highly efficient and specific venom for its natural prey, principally insects. Although its venom contains many enzymes, peptides, and low molecular weight compounds, it comprises mainly Cys-rich peptide neurotoxins that have significant effects on the sodium channels of insects (Penaforte et al. 2000;Diniz et al. 2018;Peigneur et al. 2018). Although most of these neurotoxins are insect-specific, P. nigriventer bites can cause medically relevant envenomation in humans, mainly as a result of the presence of δ-ctenitoxin-Pn2a (UniProt: P29425) and δ-ctenitoxin-Pn2c (UniProt: O76199). These two neurotoxins, also called PnTx2-6 and PnTx2-5 respectively, correspond to the semi-purified toxic fraction of the venom, namely PhTx2 (Araújo et al. 1993). Although present in low proportions compared to other components of the venom, these two neurotoxins are the most toxic ones for humans (Diniz et al. 2018;Matavel et al. 2009). They bind to mammalian sodium channels and inhibit their voltagegated inactivation, and they are thus the main molecules responsible for envenomation symptoms. δ-ctenitoxin-Pn2a and -Pn2c have a bioactive surface composed of the same positively charged residues that surround hydrophobic residues. Given the additional presence of Tyr35 and Trp37 in the hydrophobic patch of δ-ctenitoxin-Pn2a, this neurotoxin has six times higher affinity for the sodium channel than δ-ctenitoxin-Pn2c (Matavel et al. 2009).
Severe cases of envenomation usually occur in children because of their low body mass and must be treated with a specific antivenom, produced as described in 1894 (Phisalix and Bertrand 1894;Calmette 1894). This method involves repeatedly injecting P. nigriventer venom into large mammalians, mainly horses, to obtain immunoglobulins, which are then purified from plasma to make the antiserum. These antibodies recognize, bind, and neutralize P. nigriventer toxins (Bochner 2016). However, capturing this solitary and aggressive spider and extracting its venom is extremely challenging and dangerous. Furthermore, only a small amount of venom is obtained (generally by electrical stimulation) from each specimen (Lucas 2015). This costly and complex process, without any sustainable benefit for manufacturers, discourages the pharmaceutical industry from producing specific antivenoms for P. nigriventer. Consequently, these products are commonly provided by non-profit governmental organizations. However, the low budgets of such organizations impede them from meeting the national and regional demands, especially in those countries where spider bites are frequent. P. nigriventer antivenom is produced by the Butantan Institute in Brazil and the "Dr. Carlos G. Malbrán" National Administration of Laboratories and Institutes of Health in Argentina. As demonstrated by de Roodt et al. (2017), these two products can neutralize the venom of the Argentinean P. nigriventer, despite being produced using specimens from different parts of South America. P. nigriventer antivenom is used to treat envenomation by all species of the genus (Bucaretchi et al. 2018). Recently, Fernandes et al. (2022) demonstrated an overall similarity of the venom components between different species of the Phoneutria genus, with high sequence homology between the toxins. However, to date, the antigenic cross-reactivity of venoms of distinct Phoneutria sp. has not been studied.
P. nigriventer venom is a complex mixture, where only minor components, namely δ-ctenitoxin-Pn2a and -Pn2c, are toxic to humans. These small and stable Cys-rich components are present in low proportion in comparison to others that are much less toxic to humans. Furthermore, they induce fewer antibodies than other larger molecules in the venom, such as enzymes. Hence, there is an unbalanced antibody content in the resulting antivenoms and therefore higher doses are necessary to achieve the desired effect. The high content of heterologous proteins administered to patients increases the risk of allergic adverse reactions, such as anaphylaxis and serum sickness (Isbister and Fan 2011).
An alternative approach to obtain a higher proportion of antibodies against the most dangerous toxins would involve immunizing mammalians with pure synthetic toxins. These toxins can be obtained by solid-phase peptide synthesis (SPPS) or by expression in heterologous systems (Bermúdez-Méndez et al. 2018). However, δ-ctenitoxin-Pn2a and δ-ctenitoxin-Pn2c have five disulfide bridges and, like most venom toxins in spiders, they contain the inhibitor cystine knot structural motif (ICK or knottin). Hence, their synthetic or biosynthetic folding is complicated and large-scale production processes are costly (Clement et al. 2015;Saez et al. 2017;Torres et al. 2010;Tsuda et al. 2015). Furthermore, as previously described (Maillère et al. 1995;Moore et al. 2013;Kimura 2021), knottins show exceptional stability and therefore poor immunogenicity, thereby hampering their digestion by lysosomal/endosomal enzymes in the antigenpresenting cells of the immune system. This fragmentation is critical for further epitope presentation to CD4 + T helper cells, a key step in initiating the adaptive immune response (Hilligan and Ronchese 2020).
Alternative synthetic antigens containing one or more epitopes from the natural toxin can be designed to enhance the immune response (Camperi et al. 2020). Peptides of 10-20 amino acids are optimal antigens and can be easily produced by SPPS (Merrifield 1963). However, they are often weakly immunogenic and so are usually coupled to a carrier protein. Therefore, animals immunized with peptide-carrier conjugates also produce anti-carrier antibodies (Hancock and O'Reilly 2005). Other more straightforward and less costly strategies to effectively increase peptide immunogenicity include palmitoylation (Hopp 1984;Hamley 2021), dimerization, or multimerization of peptides (Tam 1988;Joshi et al. 2013;Marastoni et al. 2000). These modified peptides usually show more stability than venoms or peptide-protein conjugates and are easier to prepare and store for future use.
Given the costly and complicated process of P. nigriventer antivenom production and the high human toxicity of δ-ctenitoxin-Pn2a, here we focused on identifying the epitopes of this neurotoxin and on designing, synthesizing, and characterizing immunogenic peptides. Moreover, the cytotoxicity and capacity of each synthesized peptide to activate macrophages in vitro were evaluated.

Epitope Identification
The MHC-II Binding Prediction tool from the Immune Epitope Database (IEDB) Analysis Resource (http:// tools. iedb. org/ mhcii/) was used to identify linear class II T epitopes (CD4 + T cell-specific epitopes) of δ-ctenitoxin-Pn2a neurotoxin (Wang et al. 2008(Wang et al. , 2010, following the website instructions. 15-mer epitopes were chosen to be predicted and all the alleles from human and mouse provided were selected. Other mammalians, such as horses, were not an option in this tool. These predictions were made on 03/02/2020 and repeated on 05/19/2022.

Solid-Phase Peptide Synthesis
The linear epitope predicted as having the highest score, G 34 YFWIAWYKLANCKK 48 , was selected for future studies. Peptides Ac-GYFWIAWYKLAN-Abu-KKG-NH 2 , Palm-GYFWIAWYKLAN-Abu-KKG-NH 2 and (Ac-GYFWIAWYKLAN-Abu-KK) 2 -KG-NH 2 were designed from this epitope and synthesized by SPPS using Fmoc/tBu chemistry (Carpino and Han 1970) and the Rink-Amide-MBHA resin. For each coupling, a three-fold excess of each Fmoc-protected amino acid with 3 eq. of Oxyma Pure was dissolved in the smallest possible volume of DMF and then 3 eq. of DIC was added for activation. Afterwards, the mixture was added to the solid-phase resin in a polypropylene column fitted with a polyethylene filter. The reaction was incubated with slow agitation for 45 min at room temperature. After monitoring the coupling completeness of each peptide using the Kaiser Test (Kaiser et al. 1970), Fmoc was removed with 20% piperidine in DMF (2 × 5 min). To overcome aggregation, 0.4 M LiCl in DMF was used as solvent instead of DMF in coupling and deprotection steps. To avoid peptide dimerization, α-aminobutyric acid (Abu) was coupled instead of Cys. N-terminal acetylation was achieved after peptide elongation by adding Ac 2 O (10 eq.) and DIC (10 eq.) in CH 2 Cl 2 and incubating for 30 min. N-terminal palmytoylation was achieved by adding palmitic acid (3 eq.) together with TBTU (3 eq.) and DIPEA (4 eq.) and incubating the mixture overnight at room temperature. For peptide dimerization, Fmoc-Lys(Fmoc)-OH was used. After coupling and Fmoc removal, peptide elongation from the α and ε Lys amino groups was achieved by duplicating the equivalents of reagents in each coupling. Side chain protecting groups were removed by treatment with TFA/TIS/ H 2 O/DODT (92.5:2.5:2.5:2.5) for 2 h. Subsequently, each peptide was precipitated with cold diethyl ether, and yield was determined by gravimetry.

Peptide Analysis by Electrospray Ionization Mass Spectrometry (ESI MS)
Peptides were analyzed by direct injection in a Thermo-Scientific, Q-Exactive Orbitrap at the Centro de Estudios Químicos y Biológicos por Espectrometría de Masa (CEQUIBIEM), Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires. Peptide ionization was performed by electrospray (voltage 2.5-3.5 kV). The spectra obtained were evaluated using the FreeStyle program, version 1.5. Each monoisotopic mass was obtained with the deconvolution function Xtract.

Peptide Analysis by High-Performance Liquid Chromatography (HPLC)
The synthesized peptides were studied by RP-HPLC in a Shimadzu LC-20AT instrument with a photodiode array UV-Vis detector (SPDM20A) using a XSelect Peptide CSH reverse-phase C18 column (130 Å, 3.5 μm, 4.6 mm × 50 mm). Each chromatography was performed at 50 °C to assure a good resolution (Mant et al. 2007) at a flow rate of 1.6 mL/min with linear gradients using Solvent A (0.045% TFA in H 2 O) and Solvent B (0.036% TFA in acetonitrile).

Cell Cultures
The murine macrophage cell line RAW 264.7 was obtained from the American Type Culture Collection (ATCC). Cells were transferred as per ATCC instructions, re-suspended in supplemented culture medium (RPMI-1640 with 10% FBS, 100 mg/ml of streptomycin, and 100 UI/ml of penicillin to inhibit bacterial contamination), and incubated at 37 °C with 5% CO 2 and high humidity. Cells (approximately 70-80% confluent) were subcultured every 2-3 days.

Cell Viability Assay
Cells (15,000 cells/200 µL) were depleted of serum for 4 h before incubating with or without each peptide (concentrations from 0.25 to 10 µM) in medium supplemented with 0.5% FBS for 24 h. The 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay (Promega Corp., Madison, WI, USA) was then used to evaluate the activity of the mitochondria respiratory chain as an indicator of the number of viable cells. All experiments were performed in triplicate and results were expressed as mean ± SD. Oneway ANOVA and Dunnett's test were performed with Graph Pad Prism 8 software, San Diego, California, USA. Values of p < 0.05 were considered significantly different from controls.

Immunofluorescence Studies
RAW 264.7 cells (150,000 cell/ml) cultured on coverslips and treated for 4 h with or without 0.5 µM of each peptide were immunoassayed. Briefly, the cultures were washed three times with phosphate buffer solution (PBS), fixed with ice methanol for 10 min at − 20 °C, and rinsed with acetone for 5 min to permeate membranes. Non-specific binding sites were blocked by treatment with 3% bovine serum albumin in PBS for 45 min at room temperature. To detect the translocation of nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) from the cytoplasm to the nucleus, cells were exposed overnight to mouse anti-NF-kB monoclonal antibody (SC-8414, Santa Cruz Biotechnology, INC) (1:100) at 4 °C in a humidified chamber and then rinsed with PBS and incubated with Alexa Fluor-488 goat anti-mouse antibody (Molecular Probes. Life Technology) (1:200) for 2 h at room temperature. A Nikon Microscope (Eclipse Ti, Nikon Instruments, Melville, NY, USA) was used to take the fluorescence microphotographs using the Micrometrics SE Premium software. These images were then processed with Bioimage XD 1.0 software. Images were taken with the acquisition software FluoView version 3.3 and analyzed using ImageJ (NIH, Maryland, USA) and Image-Pro plus version 4.5 (Media Cybernetics, Rockville, MD, USA).

Results and Discussion
A schematic representation of δ-ctenitoxin-Pn2a is shown in Fig. 1. Disulfide bridges between Cys3(I)-Cys17(IV) and Cys10(II)-Cys23(V) conform a macrocycle crossed by the bridge between Cys16(III) and Cys31(VI), forming a knot-like structure (inhibitor cystine knot or knottin) (Faddeev and Niemi 1997). Two more disulfide bridges between Cys14-Cys46 and Cys25-Cys29 add further stability to the molecule (Matavel et al. 2009). The neurotoxin sequence in FASTA format in the IEDB tool revealed that the linear class II T epitopes predicted as having the highest affinity (the smallest numbered percentile rank, http:// tools. iedb. org/ mhcii/ help/) were in its C-terminal region. Of these, C-terminal sequence G 34 YFWIAWYKLANCKK 48 was selected for immunogen design because high scores were obtained with this epitope when using both the human and mouse alleles provided by the IEDB tool. Also, this fragment does not form part of the Cys-knot (Matavel et al. 2009), thus making it more accessible for subsequent recognition by antibodies. Furthermore, given that this neurotoxin sequence contains most of the residues that interact with sodium channels (Matavel et al. 2009), those antibodies that recognize this sequence could neutralize the toxin.
The acetylated and amidated linear epitope analog N-terminus, Ac-GYFWIAWYKLAN-Abu-KKG-NH 2 , was synthesized, where the Cys residue was replaced by Abu (ethylglycine) to avoid uncontrolled dimerization. Although Ser and Abu are isosteres of Cys, Abu was chosen because Ser introduces a hydrophilic hydroxyl group that is not present in the original sequence (Gregory et al 2019). Also, the N-terminus palmitoylated analog Palm-GYFWIAWYKLAN-Abu-KKG-NH 2 , as well as the branched peptides (Ac-GYFWIAWYKLAN-Abu-KK) 2 -KG-NH 2 were prepared, as palmitoylation and peptide dimerization strategies have been described as effective to increase peptide immunogenicity (Hopp 1984;Hamley 2021;Marastoni et al. 2000;Sheard et al. 2022). The peptide yield exceeded 90% in all cases, as determined by gravimetry. Table 1 shows the peptides synthesized, together with their corresponding molecular weight, monoisotopic mass, and isotopic mass distribution. All peptides were first analyzed by direct injection ESI-MS, and monoisotopic masses were obtained with the deconvolution function Xtract. In all cases, the main signals in the mass spectra corresponded to the synthesized peptide ions, along with a few much smaller signals, possibly related to low levels of impurities (Fig. 2). The main impurities observed in each MS spectra were M + 44 in linear peptides and M + 44 and M + 88 in dimer peptides. These are a very common impurities in SPPS when using Fmoc-Trp(Boc)-OH as the building block. They derive from the incomplete decarboxylation of Trp(CO 2 H) intermediate during final cleavage, which then easily hydrolyzes when the peptides are dissolved in water solutions (Chan and White 2009).
Next, the peptides were evaluated by RP-HPLC. The chromatograms showed a main peak corresponding to the synthesized peptide, along with other much smaller peaks associated with low levels of impurities (Fig. 3a-c).
Macrophages recognize invading pathogens and are key to orchestrating an appropriate immune response (Locati et al 2020). Therefore, peptide cytotoxicity was evaluated in vitro on murine macrophage cells by MTT assays after treating them with increasing concentrations of each peptide. According to ISO 10993-5, a reduction in cell viability of more than 30% indicates cytotoxicity (Standard 2009). The results demonstrated that none of the peptides assayed affected cell viability at 0.25 µM after 24 h of incubation (Fig. 4). When cells were treated with peptide Ac-GYFWIAWYKLAN-Abu-KKG-NH 2 at concentrations between 0.5 and 10.0 µM, a significant decrease in viability was observed (p < 0.01-p < 0.0001). Nevertheless, this decrease was small, not exceeding 30% at any of the concentrations assayed (Fig. 4a). This result is consistent with the low toxicity of the highly similar peptide Ac-GERRQYFWIAWYKLANSKK-NH 2 , which was also designed from δ-ctenitoxin-Pn2a to evaluate its capacity of potentiate the erectile function (Silva et al. 2015). Likewise, Palm-GYFWIAWYKLAN-Abu-KKG-NH 2 caused  a significant decrease in viability (p < 0.05-p < 0.0001) between 0.5 and 10.0 µM but exceeded 30% at 5.0 and 10.0 µM and therefore exerted a cytotoxic effect (Fig. 4b). On the other hand, the branched peptide (Ac-GYFWIAWYKLAN-Abu-KK) 2 -KG-NH 2 did not cause a significant reduction in viability between 0.5 and 1.0 µM and caused a significant decrease only at 5.0 and 10.0 µM (p < 0.05 and p < 0.001 respectively) but exceeded 30% only at 10 µM  . 4c). These observations imply that, although peptide palmitoylation and ramification are useful strategies to boost immunogenicity, in this case, they also increased cell cytotoxicity. This should be taken into consideration when dosing the peptide for its use as an immunogen. The transcriptional factor NF-κB plays a critical role in macrophage activation by increasing the expression of various cytokines that are essential for the immune response (Lee et al. 2016). When immunogens activate cells, NF-κB migrates from the cytosol to the nucleus, where it upregulates cytokine gene expression (Livolsi et al. 2001). To determine the capacity of the peptides to activate macrophages, we performed immunofluorescence studies to evaluate NF-κB cellular distribution in the RAW 264.7 cell line. According to the cell viability results and the ISO 10993-5 considerations about cytotoxicity previously described, cells were incubated with 0.5 µM of each peptide for 4 h, to assure macrophage activation without cell toxicity. As shown in Fig. 5a-a i , in resting macrophages, the transcription factor was restricted to the cytosol, while after LPS treatment, which was used as a positive control, NF-κB was localized both in the cytosol and nucleus ( Fig. 5b-b i ). Of note, immunofluorescence was observed in the nucleus of the cells in the presence of all the peptides assayed (Fig. 5c-c i , d-d i , e-ei), thereby suggesting that these molecules promote an immune response. Ongoing experiments to analyze the cytokines produced in the presence of the different peptides are required to identify macrophage phenotypes.

Conclusions
Here we designed immunogenic peptides using the δ-ctenitoxin-Pn2a G 34 YFWIAWYKLANCKK 48 epitope. Cys isosteric substitution by Abu prevented disulfide bond formation. The acetyl and palmitoyl peptides, as well as the dimer versions of both, were synthesized and characterized. None of the peptides showed cytotoxicity between 0.25 and 1.00 µM. Furthermore, the cellular distribution of NF-κB reflected early macrophage activation in response to all the peptides assayed. Given this observation, these molecules emerge as promising candidates for in vivo evaluation as immunogens for antivenom production. These peptides can be produced at higher yields and more easily and safely than venom extraction from P. nigriventer. Also, synthetic peptides are easier to transport and store for long periods than crude venom, thereby facilitating logistical distribution. Furthermore, the antivenom produced would show higher specificity for the most toxic component of the venom. The results reported herein should serve to increase interest in the large-scale production of these peptides by the pharmaceutical industry.
Although δ-ctenitoxin-Pn2a is considered the most toxic component of P. nigriventer venom, antibodies against peptides from a single toxin may not be able to completely neutralize the toxicity of the venom. Therefore, peptides designed from other neurotoxins, also highly toxic for mammals like δ-ctenitoxin-Pn2c, may be necessary. In addition, in vivo tests are still necessary to assess the immunogenic Data show the mean ± SD (n = 3). Statistically significant difference vs. control according to One-way ANOVA: *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001 compared with the control group (0.00 μM) Page 9 of 11 25 capacity of these peptides, alone and in combination, and whether the antibodies generated can offer protection against the venom.