Precise mapping of a snake venom phospholipase A₂ interaction with a human nicotinic acetylcholine receptor

Abstract

The identification of selective molecules that interact with specific human nicotinic acetylcholine receptor (nAChR) subtypes is essential for elucidating their structure and function. Animal venoms, particularly snake venoms, offer a diverse repertoire of bioactive molecules that may selectively bind to human nAChR subtypes. In this study, we cloned transcripts encoding human nAChR subunits and successfully expressed their extracellular domains (ECDs) in yeast cells. We further examined the interactions between nAChR subunits and snake venom toxins, identifying that snake Asp49-phospholipase A₂ (Asp49-1) strongly interacts with the human β2 and β4 subunits while exhibiting weaker interactions with α subunits. To delineate the specific regions of Asp49-1 involved in binding to nAChR β2, we analyzed its three-dimensional structure and dissected it into six fragments. Our results indicate that the N-terminal region of Asp49-1, comprising 29 amino acids, exhibits strong binding affinity to nAChR β2, whereas the C-terminal region interacts weakly. Similarly, to determine the interaction sites on the nAChR β2 subunit, we segmented its ECD into four fragments, expressed them in yeast, and identified the proximal N-terminal region as the primary binding site for Asp49-1. Based on these interaction regions, we proposed the three-dimensional complex structure of Asp49-1 and α4β2 nAChR subtype. These findings contribute to a deeper understanding of nAChR subtype-specific interactions and have significant implications for the development of novel therapeutic strategies targeting the α4β2 nAChR subtype, which plays a critical role in various pathological conditions, including neurodegenerative disorders.

1. Introduction

Acetylcholine receptors (AChRs), a member of the superfamily of pentameric ligand-gated ion channels (reviewed in Kalamida et al., 2007; Gharpure et al., 2020) expressed by both neuronal and non-neuronal cells throughout the body, are involved in a variety of fundamental physiological processes. Dysfunction of these receptors has been implicated in various human disorders, including Alzheimer’s disease, Parkinson’s diseases, schizophrenia, certain forms epilepsies, and nicotine addiction. According to their relative affinities for nicotine or muscarine, AChRs are grouped into nicotinic AChRs (nAChRs) and muscarinic AChRs (mAChRs) (reviewed in Albuquerque et al., 2009). The five subunits of nAChRs are selected from a pool of 17 homologous subunits (α1-10, β1-4, γ, δ and ε) and organized around a central pore (Fig. 1A), resulting in a variety of heteropentameric (e.g., α4β2) and homopentameric (e.g., α7) subtypes (Cooper et al., 1991; reviewed in Ho et al., 2020). Extensive studies on nAChRs have demonstrated that each nAChR subunit consists of: 1) a conserved extracellular large NH2-terminal domain (ECD, ~220 amino acids) that bears the ligand binding site for agonists and antagonists; 2) four small (15–20 amino acids) hydrophobic transmembrane domains (TMD, TM1-TM4); 3) a large cytoplasmic domain (ICD, 100–150 amino acids) which bears phosphorylation sites; and 4) the C-terminal hydrophilic extracellular segment (4–28 amino acids) (Kalamida et al., 2007; Albuquerque et al., 2009; Gharpure et al., 2019) (Fig. 1B). Based on the major site of expression, nAChRs were further divided into two groups: (a) the muscle-type nAChRs composed of α1β1γδ (fetal) or α1β1εδ (adult); and (b) the neuronal-type nAChRs consisting of α2-10 and β2-4 (McGehee et al., 1995; Karlin et al., 2002). All nAChR subunits share structural similarities, but the diverse combinations of these subunits forming distinct nAChR subtypes, coupled with the ability of their ECDs to bind a wide range of ligands, present significant challenges in structurally characterizing specific nAChRs and a precise understanding of the functions of individual nAChR subtype. Therefore, this urges for identifying selective molecules targeting specific nAChR subtype for investigation of their structure and function to facilitate successful therapeutic strategies for nAChR associated neurological disorders.

Fig. 1.

Structure of nAChR. A) Five subunits organized around a central pore (Kalamida et al., 2007). B) Three major domains of each subunit (Gharpure et al., 2019).

Fortunately, among the vast array of natural molecules, animal venom toxins provide a unique tool for studying the structure and function of nAChR. Animal venom toxins, such as snake venom toxins, have been fine-tuned through millions of years of natural selection to exhibit high specificity and selectivity towards specific targets, particularly the receptor proteins (e.g., nAChR) in the nervous system of prey organisms, for the quick paralysis and capture of their prey. Snake venom represents a cocktail of biologically active toxins (proteins and peptides) produced by specialized glands that are evolutionarily related to salivary glands (Kochva, 1987). Toxins constitute the major portion of venom’s dry mass (>90 %) (Heise et al., 1995; Fry et al., 2004). Extensive studies show that the venom of a single snake species contains approximately 100 different toxins, but venom composition varies interspecifically and intraspecifically, and even ontogenetically (Aragόn et al., 1981; Alape-Girón et al., 2008); therefore, the venoms of 4177 species of snakes provide a virtually untapped reservoir of bioactive molecules with highly diverse sequences and structures (Uetz et al., 1996). Snake venom toxins have been instrumental in discovering some members of the nAChR family and understanding their structures and functions. One of the principle neurotoxic components of snake venom is a protein family termed three-finger toxin (3FTx), such as α-Bungarotoxin (α-Bgtx) isolated from the venom of snake (Bungarus multicinctus) which was discovered over sixty years ago (Chang et al., 1963) and is still the most widely used and valuable inhibitor in deciphering the different steps of the gating cycle of nAChRs (Dellisanti et al., 2007; Rahman et al., 2020; Noviello et al., 2021). However, among the nAChRs, only the α7, α9/α10, and muscle subtypes bind α-Bgtx (Dellisanti et al., 2007; Zouridakis et al., 2014; daCosta et al., 2015; Noviello et al., 2021). Thus, it is essential to identify different molecules targeting the other nAChR subtypes for the investigation of their structure and function. Among the millions of snake venom toxins, the secretory phospholipase A₂s (PLA₂s) are key components of many snake venoms and have been extensively studied. Snake secretory PLA₂s are small proteins comprising approximately 120 amino acids and typically contain seven pairs of disulfide bonds. The higher disulfide bond content in venom PLA₂ provides the advantages of being very structurally stable; however, it also presents challenges in heterologous expression and purification due to its aggregation in inclusion bodies. There are two major types of secreted snake venom PLA₂: Asp49-PLA₂ and Lys49-PLA₂. Asp49-PLA₂s are enzymatic, whereas Lys49-PLA₂s are non-enzymatic. The primary function of Asp49-PLA₂ is to hydrolyze phospholipids in cell membranes, thereby releasing fatty acids and lysophospholipids. In addition, snake venom PLA₂s contribute to the overall toxicity of snake venoms. They can induce local effects, including pain, swelling, and tissue necrosis at the envenomation site, as well as systemic effects such as coagulopathies, neurotoxicity, myotoxicity, and cardiotoxicity (Kini et al., 1987; Murakami et al., 2002; Burke et al., 2009).

In this research, we utilized our existing venom gland cDNA library from snake (Crotalus atrox) for identifying venom toxins selectively interacting with specific human nAChR subtypes. The interaction detections demonstrated that two venom toxins including one of the Asp49-PLA₂s named Asp49-1, and Cysteine-Rich Secretory Proteins (CRISPs) weakly interact with nAChR α subunits. Furthermore, among the β subunits, the interaction detection shows that Asp49-1 strongly interacts with β2 and β4 subunits. Since β2-related nAChR subtypes, such as α4β2, are abundantly expressed in the brain and play a crucial role in various physiological processes, this study elucidates the interaction between Asp49-1 and the nAChR β2 subunit using an integrated approach combining biochemical, molecular biology, and computational techniques. Specifically, we systematically dissected both the β2 subunit and Asp49-1 into multiple peptide segments. Through rigorous screening, we identified that the proximal N-terminus of the β2 subunit exhibits a strong interaction with the N-terminus of Asp49-1, while its interaction with the C-terminus of Asp49-1 is comparatively weaker. The results obtained from this study provide a unique tool for systematically probing structural, biophysical, and pharmacological properties of α4β4 nAChRs subtype, and allow us to compare the properties of nAChR subtypes. More importantly, the identified venom peptide targeting α4β2 nAChR will provide a potential model for developing alternative therapeutic strategies to traditional approaches to nAChR drug discovery.

2. Materials and methods

2.1. Cloning nAChR and venom toxin transcripts

2.1.1. Human nAChR subunits

To obtain the transcripts encoding human nAChR subunits, human brain and muscle total RNAs were purchased from Takara Bio Inc. (Takara Bio, USA), and reverse transcribed into cDNA. Based on the transcript sequences of human nAChR subunits available in the NCBI database, a multiple DNA sequence alignment was generated. Subsequently, based on the conserved regions, PCR primers were designed (Table S1) to amplify the full-length nAChR subunit transcripts. RT-PCR was performed using Phusion High-Fidelity DNA Polymerase (Thermo-Fisher Scientific, USA) with the following setup: after initial denaturation of DNA at 98 °C for 30 s, 35 cycle amplification were performed, including denaturation at 98 °C for 10 s, anneal at 50–60 °C (depending on the primers) for 30 s, and extension at 72 °C for 1 min, followed by further extension for 10 min at 72 °C. PCR products, visualized in 1 % agarose gel by staining with SYBR Safe DNA gel stain (ThermoFisher Scientific, USA), were excised and purified using the GENECLEAN kit (ThermoFisher Scientific, USA). The purified PCR product was ligated into the pJET 1.2 vector (ThermoFisher Scientific, USA) and transformed into DH5α bacterial competent cells. Plasmid DNAs were isolated from bacterial cells using the GenElute Plasmid Miniprep Kit (Sigma-Aldrich, USA) and sequenced in our lab using Applied Biosystems SeqStudio Genetic Analyzer (ThermoFisher Scientific, USA).

2.1.2. Snake venom toxins

Snake venom toxin transcripts, including three PLA₂s (Asp49-1, Asp49-2, Lys49), three-finger toxin (3FTx) and Cysteine-Rich Secretory Protein (CRISP) transcripts, were obtained from the venom gland cDNA library of snake (Crotalus atrox), which was generated in our lab (Vega et al., 2024).

2.2. Yeast two-hybrid analysis

To detect venom toxin-nAChR subunit interactions, yeast two-hybrid (Y2H) analysis was conducted following the Matchmaker Gold Yeast Two-Hybrid System User manual (Takara Bio, USA). Specifically, venom toxin and nAChR subunit transcripts were amplified by PCR using transcript-specific primers incorporated with an EcoRI site for the forward primers and a BamHI site for the reverse primers. PCR was conducted using Phusion High-Fidelity DNA Polymerase as described previously. The vector plasmid DNAs, pGADT7 AD (pAD) and pGBKT7 (pBD), were digested with restriction enzymes (EcoRI and BamHI). Vectors and amplified transcripts digested with EcoRI and BamHI, were run on a 1 % agarose gel, and purified. The ligation was performed with T4 DNA ligase (Biolabs, USA) and the constructs were transformed into DH5α competent cells. All recombinant DNAs were sequenced to confirm that the junctions were in-frame with the vector promoters. The verified pAD-toxin transcripts and pBD-nAChR subunit transcripts were subsequently co-transformed into the yeast strain Y2HGold using the LiAc transformation method and spread on SD-Leu/-Trp (DDO) plates for growth at 30 °C. After 2–3 days, the clones grown on DDO plates were selected, diluted in ddH₂O, and spotted on DDO, TDO (SD-Leu/- Trp/-His/+ x-alpha-Gal) and QDO (SD-Leu/-Trp/-His-Ade) plates.

2.3. Heterologous gene expression and affinity purification

To obtain recombinantly expressed human nAChR subunits and venom toxins, the extracellular domains (ECDs) of nAChR subunit or mature toxin transcripts were amplified by PCR using gene-specific primers incorporated with BamHI site for forward primers and XbaI or HindIII restriction sites (depending on the gene sequence) for reverse primers. The purified PCR products were ligated into the corresponding sites (BamHI -XbaI or HindIII) of the yeast expression vector pEG(KT) (Pierce et al., 2009), a Glutathione S-transferase (GST)-tagged vector kindly provided by Dr. Derek C. Prosser from Virgnia Commonwealth University. The constructs were then transformed into the yeast strain BJ2168 (ATCC # 208277) using the LiAc transformation method and spread on SD-Uracil plates for growth at 30 °C. Yeast colonies grown on SD-Uracil were further spotted on SD-Leu plates to identify colonies containing high numbers of plasmids for subsequent large-volume incubation. Several identical colonies from the SD-Leu plate were selected, inoculated into 100 mL of base medium (SD-U + 2 % glucose), and incubated overnight at room temperature. The overnight growth medium was then inoculated into an induction medium (yeast extract 10 g/L, peptone 20 g/L, 2 % galactose, 0.016 % glucose) to an OD600 = 0.1 and incubated with shaking at room temperature for 72 h. Yeast cells were collected by centrifugation at 2950×g for 5 min, then sonicated in 1 x PBS buffer (1.37M NaCL, 100 mM Na₂HPO4, 27 mM KCL, 18 mM KH₂PO4, pH 7.4). Disrupted yeast cells were pelleted by centrifugation at 2950×g for 10 min, and the supernatant was subjected to an EconoFit Profinity GST Column (Bio-Rad, USA) using an Econo Pump (Bio-Rad, USA) followed by washing with 1 x PBS buffer. The GST-fusion proteins were eluted by Glutathione Elution Buffer (0.154 % L-Glutathione reduced in 50 mM Tris-HCL, pH 8.0). The quantity and quality of GST-fusion proteins were evaluated on 12 % SDS-PAGE, and the protein concentration was estimated using the BCA Protein Assay (Thermo-Fisher Scientific, USA)

2.4. Enzyme linked immunosorbent assay

We conducted an enzyme-linked immunosorbent assay (ELISA) assay based on the protocol developed by Slutzki et al., 2012, with some major modifications. Specifically: 1) Immobilization, 100 μL/well of 10 μg/mL of synthesized peptides, either from Asp49-1 or from nAChR β2 in coating buffer, pH 7.4 (ThermoFisher Scientific, USA) and Coating buffer alone (background) were immobilized in clear Flat-bottom immune 96-well plates (ThermoFisher Scientific, USA) at 4 °C for 48 h. This was followed by washing in 1xPBS buffer three times at 10-min intervals. 2) Blocking, the peptide-coated plates were blocked with 150 μL of Super blocking buffer (ThermoFisher Scientific, USA) for 1 h at room temperature, followed by quick rinsing with 1 x PBS buffer. 3) Binding protein application, concentration gradients of the binding protein (GST-fusion proteins or supernatant of cell lysate) (100 μL/well), diluted in 1 x PBS buffer, were then applied to the coated plates. The plates were sealed with MicroAmp Clear Adhesive Film (ThermoFisher Scientific, USA) and incubated at 4 °C 48 h, followed by washing three times in 1 x PBS buffer at 10-min intervals. 4) Antibody application, 100 μL/well of the primary antibody (Anti-GST Tag antibody (Rabbit) polyclonal) (ThermoFisher Scientific, USA) diluted to 1:10,000 in 5 % BSA/1 x TBST solution was applied. This was followed by washing in 1 x TBST three times at 10-min intervals. Subsequently, 100 μL/well of the secondary antibody (Goat anti-rabbit IgG conjugated with HRP) was applied, followed by washing in 1 x TBST three times at 10-min intervals. 5) Substrate addition, 100 μL of 1-Step Ultra TMB-ELISA Substrate Solution (ThermoFisher Scientific, USA) was added to the wells and incubated at room temperature in the dark until the color turned blue (approximately 30–60 min). Then, 100 μL of 0.5 N Sulfuric acid (ThermoFisher Scientific, USA) was added and the plates were read in a UV–Vis Microplate Reader (Accuris MR9610, Accuris Instruments, USA) at a wavelength of 450 nm. 6) Data analysis, data analysis was carried out using GraphPad Prism (10.2.3).

3. Results

3.1. Transcripts encoding human nAChR subunits and snake toxins

Among the reported human nAChR subunits (α1, α2, α3, α4, α5, α6, α7, α9, α10, β1, β2, β3, β4, γ, δ, and ε), we obtained full-length transcripts for all except α5, α10, and ε. Fig. S1 shows some of the amplified full-length nAChR subunit transcripts, ranging from 1.5 to 2 kb, generated by RT-PCR using various primer pairs (Table S1). To facilitate comparison, we generated an amino acid sequence alignment of the extracellular domains (~210 amino acids) for all obtained nAChR subunits (Fig. S2). All obtained human nAChR subunits exhibit the conserved disulfide bond characteristic of the Cys-loop receptor superfamily, while all α subunits possess a second disulfide bond between adjacent cysteines within the neurotransmitter-binding pocket (Loop C) (Karlin, 2002).

Additionally, we identified three unique snake venom PLA₂ transcripts from the venom gland cDNA library of Crotalus atrox: two enzymatic PLA₂s (Asp49-1 and Asp49-2) with aspartic acid residue at position 49, and one PLA₂-like protein (Lys49) with lysine at position 49. After sequencing 288 cDNA clones, we identified 11 identical sequences for Asp49-2, 6 for Lys49, and only 2 for Asp49-1. Fig. S3 presents the translated full-length mature amino acid sequence of these three PLA₂ transcripts. Additionally, we identified the toxin transcripts encoding three-finger toxin (3FTx) and cysteine-rich secretory protein (CRISP).

3.2. Snake venom PLA₂ interacts with human nAChR subunits

To investigate the interactions between snake venom toxins and human nAChR subunits, we initially screened three human nAChR subunits (α1, α3, and α7) against five snake venom toxins including 3FTx, Lys49, Asp49-1, Asp49-2 and CRISP. It is important to note that the amino acid sequence of 3FTx from C. atrox differs from that of α-Bungarotoxins (α-Bgx) in Bungarus multicinctus venom. The yeast two-hybrid (Y2H) screen results (Fig. S4) show the interactions between Asp49-1, Lys49, and CRISP with all three nAChR α subunits. CRISP also exhibited a slight interaction with the empty pBD vector (negative control); therefore, we did not pursue this interaction further. Interestingly despite Asp49-1 and Asp49-2 being enzymatic PLA₂s and Asp49-2 being the predominant PLA₂ transcript in the glands of C. atrox, Asp49-2 failed to interact with α1, α3, and α7 subunits. Also, this 3FTx doesn’t interact with α subunits. Consequently, we focused our subsequent analysis on Asp49-1. After cloning the rest of the human nAChR subunits, we detected the interactions between Asp49-1, Lys49 and 3FTx with nAChR subunits (α1, α3, α4, α7, α9, β1, β2, β3, and β4) and observed that Asp49-1, as well as Lys49, interacted with α9, β2, and β4 subunits. However, Lys49 exhibited weak binding, whereas Asp49-1 displayed strong interactions (Fig. 2). These conclusions were further corroborated by subsequent ELISA assay. Additionally, it is important to note that α9 occasionally interacted with empty pAD vector in repeated Y2H experiments, leading us to focus on β2 and β4.

Fig. 2.

Interaction detection between snake venom toxins and human nAChR subunits using yeast two-hybrid assay. Clones grown on DDO (SD-Leu/-Trp) medium indicated successful co-transformation of both the “bait” (such as nAChR subunits in the pBD vector) and the “prey” (such as venom toxins in the pAD vector) into the same yeast cells. Clones grown on TDO (SD-Leu/-Trp/-His/+x-a-Gal) medium displayed weak interactions, whereas those on QDO (SD-Leu/-Trp/-His/-Ade) medium exhibited strong interactions between the “prey” and “bait” proteins. The empty pAD and pBD vectors served as negative controls.

In addition, given the widespread expression of α4β2 nAChR receptor in human brain and their involvement in various physiological process (McGranahan et al., 2011; Absalom et al., 2013; Mazzaferro et al., 2016; Morales-Perez et al., 2016; Wilkerson et al., 2021), including nicotine addiction (Picciotto et al., 1998; Tapper et al., 2004), we focused on the interaction between the β2 subunit and Asp49-1.

3.3. Heterologous expression and peptide synthesis

To further verify the interaction between nAChR subunits and venom Asp49-1 using different methods, we recombinantly expressed nAChR subunits in yeast cells and synthesized Asp49-1 peptides. we subcloned the extracellular domains (~210 amino acids, around 25 kDa) of the nAChR subunits α3, α4, α7, α9, β2 and β4 into pEG(KT) vector under the control of the GAL1/10 promoter and successfully expressed in an induction medium containing 2 % galactose. Fig. 3 shows the successfully expressed α3, α4, β2, and β4, but some recombinant proteins were aggregated in inclusion bodies (in cell pellets). The expressed nAChR subunits, such as β2 and β4, were purified using GST affinity purification, yielding 10.5 μg for β2 and 8.4 μg for β4 from each of 200 mL of yeast culture. Additionally, we successfully expressed pEG-Asp49-1 in yeast cells; however, most of the recombinantly expressed pEG-Asp49-1 were aggregated into inclusion bodies. Therefore, based on the three-dimensional structure of Asp49-1 (Fig. 4A), and structure comparison between three PLA₂ proteins (Asp49-1, Asp49-2 and Lys49), we initially synthesized three peptides (P1, P56, and P104) (Fig. 4A & B) for Asp49-1.

Fig. 3.

Expression of nAChR subunits in yeast cells. The extracellular domains (ECDs) of nAChR subunits (e.g., α3, α4, β2, and β4) were successfully expressed in yeast cells. The lower panel displays a Coomassie blue-stained SDS-PAGE. The upper panel shows a Western blot analysis using a primary anti-GST antibody, revealing the detection of all GST fusion nAChR subunits, each with a molecular mass of 50 kDa (GST: 25 kDa; ECD: 25 kDa).

Fig. 4.

The interactions between nAChR subunits and Asp49-1 peptides. A) The 3D structure of snake PLA₂ (Asp49-1) simulated by AlphaFold2 shows the locations of three fragments (P1, P56 and P104). B) The linear amino acid sequence of Asp49-1, highlighting the three synthesized peptides (P1, P56 and P104). C) Summary of interaction, the β2 and β4 subunits exhibit strong interactions with the Asp49-1 (P1) compared to the empty vector pEG (negative control), α3, α4, α7, and α9 subunits. The concentration values at approximately 80 % saturation of cell lysates containing GST-nAChR subunits were used to generate this summary.

3.4. Interaction verification between venom Asp49-1 and nAChR subunits

The yeast two-hybrid (Y2H) method is widely used methodology to detect protein-protein interactions (PPI); however, the Y2H is occasionally criticized for generating false positives; therefore, the Y2H results usually need to be further confirmed by different methods. Here we employed ELISA to verify the interaction of nAChR subunits with Asp49-1. Three synthesized Asp49-1 peptides (P1, P56, and P104) were coated into 96-well plates and probed with the yeast cell lysate containing pEG-nAChR subunits (concentration, 15 μg/μL) by serial of dilutions. Fig. S5 shows ELISA results. The GST tag only (pEG) has no interaction with three Asp49-1 peptides (P1, P56 and P104), while α subunits including α3, α4, α7 and α9 have weak interactions with three Asp49-1 peptides. Importantly, both β4 and β2 strongly interact Asp49-1P1 (P1). We coated the 96-well plates with 100 μL of peptide in different concentration (5 μg/mL, 10 μg/mL, and 20 μg/mL) and observed that there is no difference between these concentrations, therefore, we chose 10 μg/mL for all future experiments. Meanwhile a binding protein concentration (e.g., the cell lysate of pEG-nAChR subunits) of about 80 % of the saturation value was chosen for all subsequent experiments. For example, based on this criterium, we obtained Fig. 4C, which shows that, compared to α subunits, β2 and β4 have a significant interaction with the Asp49 P1 peptide. The ELISA result further verified the Y2H results that both β2 and β4 strongly interact with Asp49-1, and the interaction between α9 and Asp49-1 is not stable in Y2H analysis due to occasionally interacting with empty vector.

To further investigate whether additional regions of Asp49-1 contribute to its interactions, we synthesized four additional peptides: Asp49b, P16, P37, and P83 (Fig. 5). Together with the previously synthesized peptides P1, P56 and P104, these peptides span the entire Asp49-1 sequence from the N-terminus to the C-terminus of Asp49-1. Asp49b and P16 encompass the P1 region (Fig. 5, amino acid sequence). ELISA assays using yeast cell lysates containing the GST-β2 fusion protein (pEG vector) revealed that the β2 subunit strongly interacts with the N-terminal P1 peptide and weakly binds to the C-terminal P104 peptide (Fig. 5A). Interestingly, when the P1 peptide was further divided into Asp49b and P16, neither peptide alone exhibited interaction with β2, suggesting that amino acids from both Asp49b and P16 are required for binding. To quantify these interactions, we utilized protein concentrations at approximately 80 % saturation of cell lysates containing GST-nAChR subunits (Fig. 5A), demonstrating that the P1 peptide exhibits significantly stronger interaction with the β2 subunit compared to other Asp49-1 peptides (Fig. 5B).

Fig. 5.

The interactions between nAChR β2 subunit and Asp49-1 peptides. A) Interaction detection between nAChR β2 subunit and Asp49-1 peptides, nAChR β subunit strongly interacts with N-terminus of Asp49-1 (P1, underlined in the amino acid sequence) while weakly interacting with the C-terminus of Asp49-1 (P104). Yeast cell lysate containing pEG-β2 was used to detect the interaction with Asp49-1 peptides coated on 96-well plates by ELISA. B) A summary of interaction, using the binding protein concentration values at approximately 80 % of saturation (A), the P1 peptide exhibits strong interactions with the nAChR β2 subunit when compared with the other peptides of Asp49-1.

3.5. Dissection of nAChR β2 subunit for interaction detection

To determine which extracellular domain (ECD) loop(s) of the β2 subunit interact with the P1 peptide of Asp49-1, the β2 ECD was segmented into four fragments: F(I-C), F(K-Y), F(K-L), and F(V-Y) (Fig. S6A and B). These fragments were subcloned into the pEG yeast expression vector and successfully expressed. The same ELISA procedure described previously was employed, wherein the P1 peptide was immobilized on 96-well plates and probed with yeast cell lysates containing GST-F(I-C), GST-F(K-Y), GST-F(K-L), or GST-F(V-Y). The ELISA results revealed that only F(I-C), which encompasses Loop D and Loop A, exhibited a strong binding to the P1 peptide (Fig. S6C).

To further pinpoint the specific region(s) within F(I-C) involved in this interaction, four peptides (βP26, β2P46, β2P82, and β2P133) were synthesized based on the three-dimensional structure of the nAChR β2 subunit (Fig. 6A–C). These synthesized β2 peptides were immobilized on 96-well plates and probed with yeast lysates containing GST-Asp49-1. The ELISA results demonstrated that Asp49-1 strongly interacted with β2P46 and β2P82 regions, whereas no interaction was observed with βP26 or β2P133 (Fig. 6D).

Fig. 6.

Dissection of nAChR β2 subunit. A) The ECD of the β2 subunit in the α4β2 nAChR subtype (PDB ID 6CNJ) was divided into four synthetic fragments: β2P26 (green), β2P46(red), β2P82(blue), and β2P133(orange). B) 3D structure of β2 subunit, with fragments color-coded as in A. C) Amino acid sequence of β2 subunit in different color corresponding to those in A and B. D) ELISA assay: the four β2 subunit fragments were synthesized, immobilized on a plate and assessed for interaction with yeast cell lysate containing GST-Asp49-P1.

4. Discussion

Venomous animals utilize their toxin proteins and peptides as highly specialized biochemical weapons to quickly immobilize and digest prey. Over millions of years of natural selection, these toxins have evolved to precisely target neural components, such as ion channel receptors, leading to rapid paralysis. Among various neural receptor proteins, including calcium, potassium, sodium, and nAChRs, we selected nAChRs due to their large extracellular domains (~210 amino acids), which offer extensive binding sites for venom toxins. We hypothesize that, upon envenomation, snake toxins interact with nAChRs in the nervous system, triggering rapid immobilization.

Compared to homopentameric nAChRs, such as the α7 nAChR, the study of heteropentameric nAChRs, particularly α4β2, presents significant challenges due to their structural complexity, difficulties in expression and purification, functional variability, and technical obstacles in structural determination. However, the α4β2 subtype is the most abundant subtype in the human brain and serves as a key mediator in various neurological processes, including cognition, consciousness, and nicotine addiction (Lester et al., 2009; Son et al., 2009; Weltzin et al., 2016). Moreover, β2 subunit-containing nAChRs play a crucial role in the inhibitory effects of SAZ-A on body mass and food intake and interact with the melanocortin system (Dezfuli et al., 2016), making them a promising target for obesity treatment. Despite their significance, no ligand has been identified that selectively binds to nAChR β subunits. Therefore, identifying a molecule with high specificity for the β2 subunit is essential for advancing the structural and functional characterization of the α4β2 receptor subtype.

Various methodologies are available for detecting protein-protein interactions (PPIs), including Y2H, co-immunoprecipitation (Co-IP), pull-down assays, surface plasmon resonance (SPR), X-ray crystallography, nuclear magnetic resonance (NMR), and ELISA. In this study, we employed Y2H system in combination with ELISA as a complementary method to investigate the interactions between snake venom toxins and human nAChR subunits. We initially selected Y2H due to its ability to detect interactions without requiring purified proteins, a significant advantage given the challenges associated with purifying membrane-bound human nAChRs and recombinantly expressed venom toxin aggregation. In our preliminary screening, we investigated the interactions between snake venom neurotoxins (e.g., PLA₂s, CRISP, and 3FTx) and nAChR α subunits using the Y2H system. Among the three PLA₂ variants, two enzymatic (Asp49-1 and Asp49-2) and one non-enzymatic (Lys49), only Asp49-1exhibited weak interactions with the α1, α3, and α7 nAChR subunits (Fig. S4). Additionally, CRISP was also found to interact with these subunits (Fig. S4). In contrast, Asp49-2 did not interact with any nAChR subunits despite its higher transcript abundance in venom glands and close phylogenetic relationship to Asp49-1. In the subsequent analysis focusing on PLA₂s and incorporating additional nAChR subunits, we surprisingly observed that β2 and β4 exhibited significantly stronger interactions with Asp49-1 than the α subunits (Fig. 2). In snake PLA₂ enzymes, the enzymatic sites are conserved; however, the neuro-toxic sites may differ (Rouault et al., 2006; Tsai et al., 2011; Dennis et al., 2011; Oliveira et al., 2018). Asp49-1 and Asp49-2 share 54.1 % amino acid sequence similarity across their 122-residue length. The N-terminal regions (residues 1–29) share only 41.4 % similarity, indicating a higher degree of sequence divergence in this region, which may contribute to distinct neurotoxic effects. Additionally, given that β2-containing nAChR subtypes, such as α4β2, are highly expressed in the central nervous system and other physiological regions, we focused our investigation on the β2 subunit. However, the recombinant expression of Asp49-1 and β2 ECD primarily resulted in aggregation within inclusion bodies (Fig. 3), posing significant challenges for purification. Consequently, rather than relying on purified recombinant proteins, we conducted ELISA assays using cell lysates containing GST-fusion proteins (e. g., GST-β2 ECD or GST-Asp49-1) obtained following centrifugation of sonicated cells. We also assessed the affinity-purified recombinant proteins from large-scale cultures (e.g., 1 L) in ELISA experiments. The primary distinction between lysate-derived and purified proteins in ELISA is the signal intensity. Therefore, we used high concentrations (>15 μg/μL) of cell lysate for most ELISA assays.

To precisely identify the binding site(s) of Asp49-1(the potential neurotoxic regions) and the nAChR β2 subunit involved in complex formation, we systematically analyzed their 3D structures and dissected Asp49-1 and β2 into six and four fragments, respectively. Through multiple rounds of ELISA analysis, we identified peptide fragment P1 of Asp49-1 as the key region interacting strongly with two specific regions of the β2 subunit, β2P46 and β2P82 (Figs. 5 and 6). Our findings are seemingly aligned with the previous studies that venom toxins often involve interactions with receptors (e.g., N-type) or ion channels (e.g., mAChR) in the body, and the N-terminus of venom toxin proteins, seems to play a key role in these interactions, especially in how the enzyme binds to and affects nerve cells (Liu et al., 1995; Nicolas et al., 1997; Huang et al., 2008). Based on our current findings and the predicted electrostatic interactions between Asp49 P1 and β2P46 & β2P82, we modeled the PLA₂₋ α4β2 complex structure (Fig. 7) using HADDOCK (Honorato et al., 2021, 2024). In contrast to the vurtoxin-nAChR complex modeled by Vulfius et al. (2014) using the HEX program, in which the snake PLA₂ (vurtoxin) binds at the α–β, α–δ, and α–γ subunit interfaces of the Torpedo nAChR, with its active site oriented toward the lipid bilayer and not occupying the classical agonist or competitive antagonist binding site, our predicted complex structure indicates that the snake PLA₂ interacts with the proximal N-terminus of the ECD of the β2 subunit, and possibly with the α4 subunits. Further refinement of our model will be conducted upon identification of the specific amino acid residues mediating the interaction.

Fig. 7.

Predicted docking pose of venom PLA₂ and the human α4β2 nAChR subtype. A) side view of the docking pose, and B) top-to-bottom view of the docking pose. Three β2 subunits are depicted in green, two α4 subunits in orange, and Asp49 is highlighted in cyan.

Surprisingly, the Asp49-1 binding sites (β2P46 and β2P82) on the nAChR β2 subunit differ from the classical neurotransmitter binding pocket, which is located at the interface between an α and β subunit in heteropentameric nAChRs, and three loops (D, E and F) of the β2 complementary (–) side are involved in ligand interaction (Morales-Perez et al., 2016). For example, nicotine binds to the classical neurotransmitter site at the α-β interface, almost fully buried from solvent. High affinity ligands binding to the classic binding pocket are also related to N-glycosylation of α subunits. For instance, Psaridi-Linardaki et al. (2002) demonstrated that glycosylation affects α-Bgtx binding to some, but not all, nAChRs, and deglycosylation of the ECD of the human α1 subunit (expressed in Pichia pastoris) abolishes α-Bgtx binding. Additionally, α subunits, but not β, may require different subunits such as δ to form a heterodimer with a high affinity ligand-binding site (Shtrom et al., 1996). It is possible that the weak interaction observed between the α subunits and Asp49-1 in our study is due to glycosylation or the absence of other associated subunits. It is important to note that protein-protein interactions can be analyzed using either sequence-based approaches, which rely on the linear amino acid sequence, or structure-based approaches, which utilize the three-dimensional conformation of proteins. Each method has its own advantages and limitations. Structure-based techniques, such as cryo-electron microscopy (cryo-EM) or X-ray crystallography, are considered the gold standard for resolving complex protein structures due to their high spatial resolution. However, these techniques require large quantities of highly purified proteins and access to specialized instrumentation, which may not always be feasible. Therefore, in this study, we employed a sequence-based approach to investigate the interaction between venom PLA₂ and human nAChRs.

PLA₂s have been widely reported to modulate receptor and ion channel activity by altering membrane properties or through their metabolic products, such as arachidonic acid and lysophospholipids (Alix et al., 1997; Massicotte et al., 1992; Liu et al., 2001). Additionally, using a voltage-clamp technique, Vulfius et al. (2014) demonstrated that PLA₂s isolated from various snake venoms suppress acetylcholine- or cytisine-induced currents in Lymnaea stagnalis neurons. These findings suggest that Asp49–1 may employ a similar mechanism to inhibit ion flux through nAChRs. However, whether PLA₂s function as direct agonists or antagonists of ion channel receptors remains unclear. In future studies, we will generate a series of targeted mutations in the β2P46 and β2P82, as well as in the adjacent α4 subunit, to identify specific amino acids involved in their interaction. This will enable us to simulate a precise complex structure of α4β2 nAChR-PLA₂ using in HADDOCK method. Additionally, we will investigate the role of Asp49-1 in interacting with α4β2 subtype using calcium imaging approach and identify potential inhibitors capable of disrupting this interaction, which will provide novel insights into PLA₂-mediated modulation of neuronal receptors and contribute to the development of targeted therapeutic strategies.

5. Conclusion

The α4β2 nAChR is predominantly expressed in the central nervous system (CNS), with high concentrations in the cerebral cortex, hippocampus, and thalamus. This receptor subtype is a critical target for therapeutic interventions in neurodegenerative disorders. Additionally, the β2 subunit has been implicated in the regulation of food intake, further underscoring its physiological significance. In this study, we identified a snake PLA₂ isoform, Asp49-1, that exhibits a strong interaction with the human nAChR β2 and β4 subunits, while displaying weaker interactions with α subunits, including α1, α3, α4, and α7. Further structural analysis revealed that the proximal N-terminus of the β2 subunit interacts strongly with the N-terminal region (residues 1–29) of Asp49-1, with potential weak interactions observed at the C-terminal region of Asp49-1.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.toxicon.2025.108438.

Data availability

Data will be made available on request.