The human neuroblastoma cell line SH-SY5Y was maintained at 37°C in a humidified 5% CO₂ incubator in Roswell Park Memorial Institute (RPMI) medium supplemented with 15% foetal bovine serum (FBS) and 2 mM L-glutamine. Replicating cells were sub-cultured every 3–4 days in a 1:5 ratio using 0.25% trypsin/EDTA.

Crude venom milked from male and female specimens of P. nigriventer was kindly provided by Prof. Marcus Vinicius Gomez from the Institute of Teaching and Research of Santa Casa de Belo Horizonte, Belo Horizonte, Brazil. Venom (lyophilised, 1 mg) was dissolved in 100 μL Milli-Q water containing 0.05% trifluoroacetic acid (TFA) (Auspep, VIC, AU) and 5% acetonitrile (ACN) and centrifuged at 5,000 × g for 10 min to remove particulates. Venom was fractionated by reversed-phase high performance liquid chromatography (RP-HPLC) using a C18 column (Vydac 4.6 mm × 250 mm, 5 μm, Grace Discovery Sciences, United States) with a gradient of solvent B (90% ACN in 0.045% TFA) in solvent A (0.05% TFA). The gradient was 5% B for 5 min, followed by 20%–40% solvent B over 60 min at a flow rate 0.7 mL min⁻¹. Peaks were collected every minute, with fraction 1 eluted between 1 and 2 min and so on for the other fractions. Venom fractions were lyophilised before storage at –20°C.

Venom fractions were screened for neuroactivity at human (h) Na_V, Ca_V1, Ca_V2 and the α7 subtype of the human nicotinic acetylcholine receptor (nAChR-α7) as previously described (Cardoso et al., 2015). Briefly, SH-SY5Y cells were plated at 40,000 cells per well in 384-well flat clear-bottom black plates (Corning, NY, United States) and cultured at 37°C in a humidified 5% CO₂ incubator for 48 h. Cells were loaded with 20 μL per well Calcium 4 dye (Molecular Devices) reconstituted in assay buffer containing (in mM) 140 NaCl, 11.5 glucose, 5.9 KCl, 1.4 MgCl₂, 1.2 NaH₂PO₄, 5 NaHCO₃, 1.8 CaCl₂ and 10 HEPES pH 7.4 and incubated for 30 min at 37°C in a humidified 5% CO₂ incubator. For the hCa_V1 assay, the dye was supplemented with 1 μM ω-conotoxin-CVIF (CVIF) to inhibit Ca_V2, and in the hCav2 assay the dye was supplemented with 10 μM nifedipine to inhibit Ca_V1. For the nAChR-α7 assay, the dye was supplemented with PNU-120596 (Sigma-Aldrich), a positive allosteric modulator of nAChR-α7. Venom fractions were assayed in singleton for each ion channel tested. Fluorescence responses were recorded using excitation at 470–495 nm and emission at 515–575 nm for 10 s to set the baseline, then 300 s after addition of 10% venom fraction serial diluted at 1, 1:10, and 1:100, and for a further 300 s after addition of 50 μM veratridine for hNa_V, 90 mM KCl and 5 mM CaCl₂ for hCa_V, and 30 μM choline for nAChR-α7.

Venom fractions eluting between 10 and 45 min on RP-HPLC were analysed by mass spectrometry to investigate the masses and primary structures of their peptide components. Native mass determinations were carried out with 20% of each fraction dried by vacuum centrifuge and resuspended in 20 μL 1% formic acid (FA), followed by analysis using by liquid chromatography/tandem mass spectrometry (LC-MS/MS). For identification of primary structures, 20% of each peptide fraction was reduced and alkylated by adding 40 μL of reagent composed of 4.875 mL ACN, 4.5 mL ultrapure water, 0.5 mL 1M ammonium carbonate pH 11.0, 100 μL 2-iodoethanol and 25 μL triethylphosphine, and incubating for 1 h at 37°C. Samples were speed dried in a vacuum centrifuge, and digested with 40 ng/μL trypsin in 50 mM ammonium bicarbonate pH 8.0 and 10% ACN overnight at room temperature. Trypsin was inactivated by adding 50 μL solution containing 50% acetonitrile and 5% formic acid (FA), dried in speed vacuum centrifuge, and resuspended in 1% formic acid.

LC-MS/MS samples were loaded onto a 150 mm × 0.1 mm Zorbax 300SB-C18 column (Agilent, Santa Clara, CA, United States) on a Shimadzu Nano LC system with the outflow coupled to a SCIEX 5600 Triple TOF (Framingham, MA, United States) mass spectrometer equipped with a Turbo V ion source. Peptides were eluted using a 30 min gradient of 1%–40% solvent B (90% ACN/0.1% FA) in solvent A (0.1% FA) at a flow rate of 0.2 mL/min. For MS1 scans, m/z was set between 350 and 2,200. Precursor ions with m/z 350–1,500, charge of +2 to +5, and signals with >100 counts/s (excluding isotopes within 2 Da) were selected for fragmentation, and MS2 scans were collected over a range of 80–1,500 m/z. Scans were obtained with an accumulation time of 250 ms and a cycle of 4 s.

A database of possible peptide sequences produced in P. nigriventer venom glands was compiled using a published venom-gland transcriptome (Diniz et al., 2018), from which open reading frames (ORFs) longer than 30 amino acids were identified and translated by TransDecoder. A list of 200 common MS contaminants was added to the translated ORFs, which was used as a sequence database to compare to mass spectral data using the Paragon algorithm in Protein Pilot 2.2 software (AB SCIEX). We report only peptides for which more than two tryptic fragments were detected with >95% confidence, or where one tryptic fragment was detected, and a secretion signal peptide was predicted by SignalP5.0.

Venom peptides identified in this study were selected based on their cysteine-rich scaffold and bioactivity, and their three-dimensional (3D) structure were predicted using the AlphaFold 2 algorithm (Jumper et al., 2021). All 3D structures displayed were from unrelaxed models ranked 1 for each peptide prediction. 3D structures were visualised and analysed using PyMol (Pymol, 2023).

Fluorescence traces from singletons were evaluated using the Maximum-Minimum or Area Under the Curve values generated after addition of ion channel activator. Data were normalised against the negative control (PSS buffer control) and positive control (ion channel activator) for each assay and corrected using the response over baseline from 1 to 5 s. No statistical analyses were required in this study.

Fractionation of 1 mg of P. nigriventer (Figure 2A) crude venom using RP-HPLC produced numerous peaks eluting between 20% and 40% solvent B, and fractions eluting between 11 and 45 min were selected for pharmacological analysis (Figure 2B). Screening using the SH-SY5Y neuroblastoma cell line revealed strong modulation of voltage-gated ion channels including both inhibition or enhancement of ion channel activity (Figure 2C). Venom fractions eluting between 18 and 34 min showed strong inhibition of Ca_V and Na_V activity, while fractions eluting between 41 and 45 min strongly activated Ca_V2 channels (Figure 2C, top panel). At a dilution of 1:10, these inhibitory effects persisted for both Na_V and Ca_V2 channels for fractions eluting at 19–20 min and 26–34 min and was absent for Ca_V1 channels (Figure 2C, middle panel). Fractions eluting from 21 to 25 min showed a clear preference for inhibiting only Ca_V2 channels (Figure 2C). Interestingly, at 1:10 dilution, channel activity enhancement was stronger on Na_V channels compared to Ca_V2 channels, suggesting potential concentration-dependent synergistic effects of venom peptides modulating both Na_V and Ca_V2 channels. At the highest venom dilution of 1:100, persistent inhibition of Na_V channel was observed for fraction 20 (F20), while the remaining inhibitory fractions preferentially inhibited only Ca_V2 channels (Figure 2C, bottom panel). Channel enhancement persisted for Na_V channels in fractions eluting from 41 to 45 min. No potent activity was observed against nAChR-α7 at any venom concentration tested. Overall, inhibitory activity was primarily observed for fractions eluting at shorter retention times (i.e., more hydrophilic compounds), while strong ion channel activation was induced by more hydrophobic peptides with longer RP-HPLC retention times.

Fluorescent traces measured upon addition of venom fractions revealed an increase in intracellular calcium ([Ca²⁺]_i), suggesting that these venom peptides can activate closed channels as well as enhance the responses of these channels opened using pharmacological intervention (Figures 2D–G). This was observed for Ca_V responses in the presence of 1 μM CVIF (Ca_V2 inhibitor, Figure 2D) and 10 μM nifedipine (Ca_V1 inhibitor, Figure 2E). In the absence of Ca_V inhibitors, these [Ca²⁺]_i responses resemble the levels of Ca_V1 responses in Figure 2D as observed for F40–F45 applied in the Na_V channels assay (Figure 2F). The activities of inhibitory fractions were mostly free from initial [Ca²⁺]_i responses upon venom addition, except for weak inhibitors observed in F19 for Na_V and F40 for nAChR-α7 (Figures 2F, G).

The venom of P. nigriventer has been extensively characterised in terms of composition and bioactivity (Diniz et al., 2018; Peigneur et al., 2018), including neuronal ion channel activity and proteomics, but not by using a combined approach. In this study, by combining these approaches, we were able to rapidly identify 58 peptides and proteins in the venom. Due to the complexity of previous nomenclature for P. nigriventer venom peptides, we refer to them here using both the rational nomenclature developed for spider toxins (King et al., 2008) and an identifying number (e.g., PN367) that is linked to a sequence and a list of previously used names in Supplementary Table S1. Of the 58 identified amino acid sequences, only eight (15%) are peptides that have had their bioactivity reported in previous studies (Figure 3A, Supplementary Table S1) (Peigneur et al., 2018). These included the known neuroactive components μ-CNTX-Pn1a (Tx1) (Diniz et al., 2006; Martin-Moutot et al., 2006), κ-CNTX-Pn1a (Tx3-1, PhK_V) (Kushmerick et al., 1999; Almeida et al., 2011), ω-CNTX-Pn1a (Tx3-2) (Cordeiro Mdo et al., 1993), Γ-CNTX-Pn1a [Tx4(5-5)] (Paiva et al., 2016), δ-CNTX-Pn1a [Tx4(6-1)] (de Lima et al., 2002; Emerich et al., 2016), δ-CNTX-Pn2c (Tx2-5a) (Yonamine et al., 2004), ω-CNTX-Pn4a (Tx3-6) (Cardoso et al., 2003; Vieira et al., 2005) and ω-CNTX Pn3a (Tx3-4) (Dos Santos et al., 2002) (Figure 3B). Even among these eight peptides, only a few venom peptides have had their molecular pharmacology characterized in detail (Peigneur et al., 2018), or their activities confirmed using recombinant peptides (Diniz et al., 2006; Paiva et al., 2016; Garcia Mendes et al., 2021).

Most of the identified sequences in this study (74%) represent peptides with unexplored bioactivity; 38 (65%) of the 43 peptides identified have cysteine-rich scaffolds typical of spider-venom peptides (Figure 3C). Some of these venom peptides, such as PN367 and PN363, have a type I scaffold (Diniz et al., 2018) and are predicted by Alphafold 2 to fold into cystine-knot scaffolds typical of spider-venom peptides (King and Hardy, 2013) (Figure 4). Scaffolds II-VIII either form elaborated cystine-knot folds with extra disulphide bonds, or alternative structures such as for scaffolds III and IV (Figure 4). Novel peptides with high identity with other toxins and not previously described in P. nigriventer venom included: PN367 displaying identity with a Agelena orientalis venom peptide; PN369 displaying identity with a Lycosa singoriensis venom peptide, and PN365 displaying scaffold III and identity with another Lycosa singoriensis venom peptide (Supplementary Table S1). Additional disulphide-rich scaffolds present in P. nigriventer venom include three peptides predicted by the algorithm HMMER to form a thyroglobulin type 1 repeat domain (E < e⁻¹⁷ in each case), one of which has been previously reported as U24-CNTX-Pn1a; peptide PN370 which displays high identity with a peptide found in venom of the scorpion Scorpiops jendeki and is predicted by the algorithm HMMER to form into a trypsin-inhibitor-like cysteine-rich domain (E < 2e⁻¹³); and the peptide PN376 that is predicted by HMMER to form a fungal protease inhibitor domain (E < e⁻⁶) (Supplementary Table S1). Additional new scaffolds identified in this study were named following the previous suggested nomenclature (Diniz et al., 2018) as X (CXCC motif, 12 Cys residues: −C−C−C−C−CXCC−C−C−C−C−C−), XI (12 Cys residues: −C−C−C−CXC−CXC−C−CXC−C−C−), XII (11 Cys residues: −C−C−CXC−CXC−C−C−CXC−C) and XIII (10 Cys residues: −C−C−C−C−C−C−CXC−C−C−), and include the peptides PN376, PN372, PN373 and PN375, and PN370, respectively.

Only 9% of the identified sequences were peptides with two or fewer Cys residues (Supplementary Table S1). F17 contained a peptide (PN361) matching a C-terminally amidated peptide precursor from Araneus ventricosus identified in a genomic study (Kono et al., 2019). This precursor has 70% sequence identify with the prohormone-1 like precursor from the honeybee Apis mellifera (UniProt P85798) which is believed to be cleaved to form three short peptides with neuronal activity. Another short peptide, PN366 identified in F18 and F28–F30, matches a neuropeptide in the sea slug Aplysia californica (UniProt P06518). Larger proteins were also detected in some fractions; for example, F18 and F31 contained a fragment at 58% and 70% total fraction components, respectively, matching a zinc metalloprotease from the nematode Caenorhabditis elegans (UniProt 55112) which contains a peptidase family M12A domain.

The cysteine-rich scaffolds of venom peptides identified in this study were compared to the classification previously proposed for P. nigriventer venom peptides (Diniz et al., 2018) (Figures 3C, 4). Peptides in fractions displaying inhibitory properties corresponded to scaffolds I, II, IV, V and VIII, as well as unnamed scaffolds, while peptides in fractions with activation properties comprised mostly the scaffold V. All of these scaffolds are inhibitor cystine knot motifs, except for scaffold IV which had the highest level in F26 represented by the peptide PN350.

Neuroactive peptides with greater hydrophilicity (i.e., those with short RP-HPLC retention times) showed pharmacological properties reminiscent of known spider-derived μ-toxins (F17 and F18) and ω-toxins (i.e., inhibition of Ca_V1 and Ca_V2 channels by F19 and F20) (Figures 2C, 5A). Major components driving those bioactivities were the pharmacologically characterised peptides μ-CNTX-Pn1a, ω-CNTX-Pn1a and ω-CNTX-Pn3a, as well as additional peptides with unknown activity (Figure 4A). As the hydrophilicity of the peptides decrease (i.e., peptides with long RP-HPLC retention times), persistent Ca_V2 inhibition was observed with maximum inhibitory activity in F26 and F27, and with the additional peptide ω-CNTX-Pn4a detected in F24 (Figures 2C, 3B, 5B). Interestingly, venom peptides characterized as K_V modulators, such as κ-CNTX-Pn1a, were detected in fractions displaying strong inhibition of calcium influx with potential μ- and ω-pharmacology (fractions 26 and 27); it was not clear if the observed bioactivity was associated to the modulation of K_V channels, or to other unexplored peptides in these fractions.

Neuroactive peptides presenting more hydrophobic structures showed properties of μ and ω-peptides, but with preference for Ca_V2 channels as observed for fraction 34 in which the peptide Γ-Pn1a is the major component, consistent with its previously observed modulation of multiple cation channels (Paiva et al., 2016); and of δ-peptides as observed in fractions 41 to 45, in which major components included the peptides δ-Pn1a and δ-Pn2c (Figures 2C–F, and Figures 5C, D). Notably, the main components of some of the most neuroactive fractions are peptides with unexplored bioactivity, e.g., fraction 26 (Figures 2C, 4, 5).

Our approach allowed classification of P. nigriventer venom peptides into four major groups based on their bioactivity (Figure 6; Table 1). Group 1 is comprised of μ and ω peptides with scaffold type VIII and more hydrophilic properties as they eluted between F17 and F21. As representatives from this group, μ-CNTX-Pn1a and ω-CNTX-Pn3a have a potential “KR electrostatic trap”, a pharmacophore described in spider-venom peptides that modulate ion channels (Hu et al., 2021; Wisedchaisri et al., 2021), in their primary and tertiary structures (Figure 6A). This pharmacophore is likely composed of residues R61, K67, K70, K71, R74 and R75 in μ-CNTX-Pn1a and residues K54, K56, R59, K65, K70, R71, K73 and K74 in ω-CNTX-Pn3a. Within this group, the ω-CNTX-Pn3a homologue PN319 differs at three positions, making it an interesting candidate for further characterisation.

Group 2 comprises κ and ω peptides that eluted between F17 and F28, with scaffold types II and VII (Figure 6B). As representatives from this group, peptides κ-CNTX-Pn1a, ω-CNTX-Pn1a and ω-CNTX-Pn4a also contain a “KR trap” pharmacophore comprised of residues R20, K23, K34, K35 and K36 for ω-Pn1a; R21, K24, K35 and K36 for κ-Pn1a; and K42, R47, K48, K49, K51, K53 and K54 for ω-Pn4a. In this group, PN107 differs from κ-CNTX-Pn1a by only two residues and is an interesting peptide for further exploration.

Group 3 is comprised of more hydrophobic Γ peptides that eluted in F33–F36 and possess a type V scaffold (Figures 3C–D, 6C). It is represented by Γ-CNTX-Pn1a with a potential “KR trap” comprising residues K35, R41, K42 and K43. Although Γ peptides modulate N-methyl-D-aspartate (NMDA) glutamate receptors, Γ-CNTX-Pn1a has also been reported as a β-peptide that inhibits Na_V channels (Paiva et al., 2016), which agrees with the results from our high-throughput ion channels assays (Figures 2C–F, 3). Interestingly, Γ-CNTX-Pn1a predicted 3D structure formed a cyclic structure in which the N-terminal cysteine formed a disulfide bridge with C-terminal cysteine (Figures 4C, 6C). These same fractions contain other ICK peptides including PN003 and PN292 with scaffold types I and II, respectively; their pharmacological targets have not been explored but they likely contribute to the strong inhibition of Ca_V channels by F34 (Figures 2, 3, 4).

Group 4 is composed of very hydrophobic δ peptides that elute in F40–F45 and possess a type V scaffold (Figure 2B, 6D). It is represented by δ-CNTX-Pn1a with potential “KR trap” comprising residues K43, K44, and K45 (Figure 6D). In this group we also identified δ-CNTX-Pn2c which differs not only in primary structure but also in the scaffold V tertiary structure by presenting a non-cyclic structure compared to the cyclic structure predicted for δ-CNTX-Pn1a connected by the N- and C-terminal cysteines (Figures 4D, 6D). Beyond these known peptides, this group comprised interesting unexplored peptides such as PN032 and PN023 showing δ peptide domains and differing from Γ-CNTX-Pn1a by 12 and 11 residues, respectively.