The Proteome of African Spitting and Non‐Spitting Cobra Venoms and Cytotoxicity Against Pancreatic Cancer Cells

ABSTRACT

African cobra (Naja spp.) venom contains toxins dominated by proteins and peptides with inter‐ and intra‐specific variations. There are several FDA‐approved drugs from snake venom toxins from other regions, including South America and Asia. Profiling the proteomes of medically important African cobra venoms from different locations will aid in developing more effective anticancer agents. The venoms of spitting cobras ( Naja pallida and Naja nigricincta woodi ) and non‐spitting cobras of the Uraeus subgenus ( Naja anchietae , Naja annulifera, and Naja nivea ) were fractionated by reverse phase‐high performance liquid chromatography (RP‐HPLC). Using label‐free LC–MS/MS, the venom toxins were identified and grouped into families based on their relative abundance. Venom cytotoxicity of both crude and fractionated samples was tested in pancreatic carcinoma cell lines (MIA PaCa‐2) using the Alamar Blue assay. Cell viability analysis revealed a cytotoxic effect of spitting cobra venoms against MIA PaCa‐2 cell lines compared to normal MRC‐5 cells. Conversely, venoms of non‐spitting cobras showed no cytotoxic activity against MIA PaCa‐2 cells. Selected RP‐HPLC venom fractions from the spitting cobras revealed that N. pallida Fraction 6 and N. n. woodi Fraction 9 at a minimal level were cytotoxic against MIA PaCa‐2 cells. LC–MS/MS data showed that while N. pallida Fraction 6 was dominated by basic phospholipase 2 CM‐III and Cytotoxin 2, N. n. woodi Fraction 9 was dominated by basic phospholipase 2 CM‐III, basic phospholipase 2 CM‐II and Cytotoxin 3. These fractions will be purified and studied to determine the mechanisms behind the underlying cytotoxicity against MIA PaCa‐2 cells.

Short abstract

African cobra venoms contain several compounds, mostly protein and peptide toxins. Profiling of the venom proteome is key to understanding the toxin compositions, which will improve the development of anticancer therapeutics. LC‐MS/MS proteomics of RP‐HPLC venom fractions showed distinct proteome profiles of the spitting and non‐spitting cobras. Cell viability assays showed spitting cobra venom and fractions, but not non‐spitting cobras, induced reduced cell proliferation of MIA PaCa‐2 cell lines, suggesting further studies as possible anticancer agents.

1. Introduction

African cobras (Naja spp.) belong to the Elapidae snake family, which are among the most widespread medically important snakes and contributors to many snakebite envenomation (SBE) cases in sub‐Saharan Africa (Benjamin et al. 2020; Gutiérrez et al. 2017). The venom of these snakes, among other toxins, is dominated by three‐finger toxins (3FTXs) and causes pathologies including tissue necrosis, cardiotoxicity, and neurotoxicity (Offor et al. 2022). The three subgenera of African cobras include Afronaja (African spitting cobras), Boulengerina and Uraeus (African non‐spitting cobras), with similar and different characteristics (Wallach et al. 2009; Wüster et al. 2018).

Data on available Naja spp. venom proteomes reveal that while the Afronaja and Boulengerina subgenera are dominated by the 3FTX and phospholipase A2 (PLA₂) toxin families, the Uraeus subgenus is dominated by 3FTX with little or no PLA₂ (Offor and Piater 2024a). Snakebite envenomation is a neglected tropical disease that affects mainly people from tropical and subtropical regions across the globe, mainly in Africa, Asia, and South America (Gutiérrez et al. 2017; WHO 2019). In sub‐Saharan Africa, there are an estimated 1 million snakebites and these include 268,471 cases of envenomation, 12,290 deaths, 14,766 amputations, and 55,332 cases of post‐traumatic stress disorder (Halilu et al. 2019). Recent reports suggest that snake envenomation data from sub‐Saharan African countries such as South Africa and Mozambique are underestimated due to challenges, including victims not reporting to hospitals, and instead using traditional healers (Berg et al. 2024; Farooq et al. 2022; Lermer et al. 2023). Thus far, antivenom is the main current treatment for SBE.

Snake venom contains several mixtures of compounds dominated by proteins and peptides that contribute to toxicity. Venom toxins can be explored for possible therapeutic application in the treatment of diseases, including cancer. To this end, snake venom toxins have been applied in the development of several antihaemorrhagic and antihypertensive drugs approved by the U.S. Food and Drug Administration (FDA), such as Tirofiban and Captopril, respectively (Cushman and Ondetti 1991).

Furthermore, snake venom toxins families including disintegrin (DIS), L‐amino acid oxidase (LAAO), 3FTX, and snake venom metalloproteinase (SVMP), have shown anticancer properties against selected cancer cell lines compared to normal cell lines (Offor and Piater 2024b). For example, the venom myotoxic peptides Asp49 PLA₂ (MTX‐1) and Lys49 PLA₂ (MTX‐2) of Bothrops brazili induced cytotoxicity against human T‐cell leukaemia (JURKAT) cell lines (Costa et al. 2008), while LAAO from Crotalus adamanteus venom showed antitumour activities against CAOV3 and OVCAR3 ovarian cancer cells (Xiong et al. 2022). To our knowledge, only PLA₂ crotoxin has passed Phase 1 clinical trials against various cancer cells (Almeida et al. 2023; Cura et al. 2002). This highlights the need for more anticancer research, especially with snake venoms endemic to sub‐Saharan Africa. In this regard, the 2022 Global Cancer Observatory (GLOBOCAN) ranked pancreatic cancer as the seventh and thirteenth cause of cancer‐related death globally and in Africa, respectively, with an approximately 5‐year survival rate of 9% (Globocan 2022; Lippi and Mattiuzzi 2020; Zhu et al. 2018). This study aims to profile the proteome of the venoms of two African spitting cobras ( N. pallida and N. n. woodi) and three non‐spitting cobra venoms ( N. anchietae , N. annulifera, and N. nivea ), and additionally investigate the cytotoxicity against pancreatic MIA PaCa‐2 cancer cell lines.

2. Materials and Methods

2.1. Venom

Lyophilized venom samples from two spitting cobras N. pallida (Tanzania), N. n. woodi (South Africa), and three non‐spitting cobras N. anchietae (Botswana), N. annulifera (South Africa) and N. nivea (Cape Town, South Africa) were supplied (SA Venom Suppliers) and stored at −20°C until use.

2.2. Reverse‐Phase High‐Performance Liquid Chromatography (RP‐HPLC)

Fractionation was performed using decomplexation strategy with minor modifications (Tan et al. 2019). Approximately 3 mg of each venom was dissolved in 200 μL 0.1% trifluoroacetic acid (TFA) in Milli‐Q H₂O. The solution was centrifuged at 10,000 xg, 4°C, for 10 min. The supernatant (25 μL injection volume for each run) was subjected to RP‐HPLC at a flow rate of 1 mL/min in the Agilent 1200 series purification system (Agilent Technologies Inc. USA) (Laustsen et al. 2015). The fractionation was carried out using a C18 column Lichrosphere RP18‐5 encapped 25 cm × 4.6 mm, 5 μm (Merck, Germany). The column was pre‐equilibrated with Eluent A (0.1% TFA in Milli‐Q H₂O) and eluted with Eluent B (0.1% TFA in acetonitrile (ACN)) with a linear gradient (5% B for 10 min, 5%–15% B over 20 min, 15%–45% B over 120 min, and 45%–70% B over 20 min). Absorbance was monitored at 215 nm, and peak fractions were collected by a fraction collector, lyophilized, and stored at −20°C until use.

2.3. Sodium Dodecyl Sulphate‐Polyacrylamide Gel Electrophoresis (SDS‐PAGE)

Lyophilized venom fractions from the RP‐HPLC fractionation were redissolved with Milli‐Q H₂O and analyzed with 15%‐SDS/PAGE at constant 20 mA and 300 V for approximately 1 h under reducing conditions (Laemmli 1970; Tan et al. 2019). A protein ladder was electrophoresed alongside the samples to gauge the molecular weight of the peptides. The gel was stained by Fairbanks with Coomassie Brilliant Blue R‐250 (Fairbanks et al. 1971).

2.4. Label‐Free LC–MS/MS Peptide Identification

2.4.1. Sample Preparation and LC–MS/MS Data Acquisition

Lyophilized RP‐HPLC fractions, and further pooled fractions thereof due to lower concentrations, were dissolved in Milli‐Q H₂0 and sent to the Centre for Proteomic and Genomic Research (CPGR; Cape Town, South Africa) for sample clean‐up and in‐solution label‐free LC–MS/MS analysis. Approximately 10 μg of protein samples were reduced with 10 mM dithiothreitol (DTT) (Sigma D9779, USA), alkylated with 15 mM iodoacetamide (IAA) (Sigma 16125, USA) at 37°C for 30 min. The pH of the samples was adjusted to approximately pH 8. Trypsin (Pierce 90058, USA) and LysC (Pierce 90307, USA) prepared in 50 mM triethylammonium bicarbonate (TEAB) (Sigma T7408, USA) were added in a ratio of 1:100 and 1:500 total protein, respectively. Samples were dried overnight by vacuum centrifugation and dissolved in 20 μL of loading buffer containing 2% ACN (Burdick and Jackson, Germany) and 0.1% formic acid (FA) (Sigma 56302, USA).

LC–MS was performed on a Q‐Exactive quadrupole‐orbitrap MS (Thermo Fisher Scientific, USA) coupled with a Dionex Ultimate 3000 nano‐IPLC system. Raw data were acquired with Xcalibur v4.1.31.9, Chromeleon v6.8 (SR13), Orbitrap MS v2.9 (build 2926) and Thermo Foundations 3.1 (SP4). Dissolved peptides in ACN and FA (approximately 400 ng) were loaded on a C18 trap column (PepMap100, 9027905000, 300 μm × 5 mm × 5 μm). The sample was washed for 3 min before switching the valve following elution of peptides on the analytical column. Chromatographic separation was performed using a ReproSil‐Pur 120 C‐18‐AQ column (DrMaisch, r119.aq.n150.075, 75 μm × 15 cm × 1.9 μm). Solvent A was composed of LC water (Burdick and Jackson BJLC365, USA) and 0.1% FA, while solvent B contained ACN and 0.1% FA. The multi‐step gradient for peptide separation was generated at 300 nL/min: 2%–5% solvent B for 1.5 min, 5%–18% solvent B for 50 min, 18%–30% solvent B for 1 min, and 30%–80% solvent B for 4 min. The gradient was held for 5 min at 80% solvent B, then returned to 2% solvent B for 5 min. The MS was operated in positive ion mode with a capillary temperature of 120°C and applied electrospray voltage of 1.95 kV. Data acquisition was carried out with Proxeon stainless steel emitters (Thermo Fisher TFES523, USA).

2.4.2. Data Analysis

Database search was performed using PMI‐Byonic‐com v3.8.13 Byonic Software (Protein Metrics, USA) and a Naja protein database sourced from UniProtKB (www.uniprot.org) dated June 19, 2023. The search parameters used included trypsin enzyme cutting lysin (K) and arginine (R) residues at the C‐terminal, fully specific at peptide termini, maximum number of missed cleavages was two, precursor tolerance 10 ppm and fragment tolerance 20 ppm. The fixed modification was carbamidomethyl (C), while variable modifications were deamidated (N, Q), and oxidation (M). The protein false discovery rate (FDR) cut‐off was 1%, with a maximum precursor mass of 10,000 and a threshold of 300 set for significance. The software generated unique peptides representing the total number of peptide spectrum matches (PSMs) for the proteins. A score plot and the mass error loading plot were generated to gauge possible variations, confidence and significance of proteins. A Byonic score of greater than 300 and log base 10 of protein p‐value greater than 1 were used as a threshold to determine the significance of differentially expressed proteins. The relative abundance of each venom protein was calculated using data from LC–MS/MS and RP‐HPLC, and presented in a pie chart.

2.4.3. Estimation of Protein Relative Abundance

The relative abundance of protein A was calculated as the ratio of the mean spectral intensity of protein A from a particular fraction to the total mean spectral intensity of all proteins in that fraction multiplied by the area under the curve (AUC) for the peak fraction in %.

$$\text{Relative abundance of protein}\mathrm{A}\left(\%\right)=\frac{\text{Mean spectral intensity of protein}\mathrm{A}\text{in fraction}}{\text{Total spectral intensity of}\mathrm{all}\text{proteins in the fraction}}\mathrm{x}\mathrm{AUC}\text{of the peak fraction}$$

2.5. Cell Viability Assay

MIA Paca‐2 cancer and MRC‐5 normal cell lines (donated by Prof. L. R. Motadi, University of Johannesburg, SA) were cultured in Dulbecco's modified Eagle's medium (DMEM) (Thermo Fisher Scientific, USA) supplemented with fetal calf serum (FCS) (Thermo Fisher Scientific, USA), 1% Amphotericin B (Thermo Fisher Scientific, USA), and 1% Pen‐Strep antibiotic (Penicillin Streptomycin) (Thermo Fisher Scientific, USA) for 24 h at 37°C, 5% CO₂ until 70–80 confluency. Cells were rinsed with 1X PBS (Thermo Fisher Scientific, USA) and detached with Trypsin (Thermo Fisher Scientific, USA). The final volume of cells (100 μL) containing 1 × 10⁴ cells/well was seeded on a 96‐well plate and incubated for 24 h. Media was discarded and replaced with 100 μL fresh media containing crude venom (1, 10, and 20 μg/mL) or Fractions (5, 10, 20, and 40 μg/mL) dissolved in 1X PBS and incubated for 24 h. Untreated cells received fresh media (100 μL), positive control cells were treated with 1% Etoposide (Sigma, USA) and negative control cells were treated with 0.1% PBS. Alamar Blue (10%) (Thermo Fisher Scientific, USA) was added to each well and incubated in the dark for 2 h under the same conditions. Fluorescence was measured at an excitation wavelength of 530/25 nm and emission wavelength of 590/25 nm on a GloMax Discover microplate reader (Promega, USA).

2.6. Data Analysis

Data analysis, including bar graph generation, one‐way ANOVA followed by Dunnett's post hoc test, and IC₅₀, was performed with Graph Pad Prism v8.4.3(686) software. Data from three independent replicates are presented as mean ± standard deviation (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001) where a p‐value less than 0.05 (p < 0.05) is significant to treatments that caused reduced cell viability compared to untreated control.

3. Results

3.1. RP‐HPLC Fractionation of Spitting and Non‐Spitting Cobra Venom Toxins

Fractionation of two spitting cobra venoms ( N. pallida and N. n. woodi) and three non‐spitting cobra venoms ( N. anchietae , N. annulifera, and N. nivea ) were performed with RP‐HPLC. The chromatograms generated from fractionation showed different venom elution profiles between 1 and 140 min. Most N. pallida venom proteins were eluted between 45 and 140 min (Figure 1A), N. n. woodi (50–140 min) (Figure 1C), N. anchietae (60–135 min) (Figure 2A), and N. annulifera (Figure 2C) and N. nivea (Figure 2F) between 45 and 135 min. The SDS‐PAGE analysis of the collected peak fractions showed that most of the venom toxins in the spitting and non‐spitting cobras were low molecular weight proteins (Figures 1 and 2).

FIGURE 1.

Analysis of spitting cobra Naja pallida and Naja nigricincta woodi venoms subsequent to fractionation. Venom was fractionated by RP‐HPLC in chromatograms (A) N. pallida and (C) N. n. woodi. The peak fractions were analyzed with 15% SDS‐PAGE under reducing conditions for N. pallida (B) and N. n. woodi (D and E).

FIGURE 2.

Analysis of non‐spitting cobra Naja anchietae , Naja annulifera, and Naja nivea venoms subsequent to fractionation. Venom was fractionated by RP‐HPLC in chromatograms (A) N. anchietae , (C) N. annulifera , and (F) N. nivea . The peak fractions were analyzed with 15% SDS‐PAGE under reducing conditions for N. anchietae (B), N. annulifera (D and E), and N. nivea (G and H).

3.2. LC–MS/MS Proteome Profiling of Spitting and Non‐Spitting Cobra Venoms

Spitting and non‐spitting cobra venom fractions collected from the RP‐HPLC fractionation were subjected to in‐solution label‐free LC–MS/MS protein identification. Significant proteins were sorted according to their respective fractions and calculated relative abundance (Tables S1–S5). The proteins were further grouped into their families: N. pallida (12 families), N. n. woodi (8 families), N. anchietae (12 families), N. annulifera (12 families), and N. nivea (12 families) (Tables S6–S10) (Figures 3 and 4). Other low abundance toxins and proteins with no known toxins activities are shown in Tables S11–S15.

FIGURE 3.

Proteome analysis of the venoms of the spitting cobras Naja pallida (A) and Naja nigricincta woodi (B). The venom peak fractions from the RP‐HPLC were analyzed with label‐free LC–MS/MS. The relative abundance of the proteins was grouped into different families. Abbreviations: CTX, cytotoxin/cardiotoxin; SNTX, short neurotoxin; WNTX, weak toxin; LNTX, long neurotoxin; PLA₂, phospholipase A2; SVMP, snake venom metalloproteinase; KSPI, Kunitz‐type serine protease inhibitor; CRISP, cysteine‐rich secretory protein; NGF, nerve growth factor; CVF, cobra venom factor; ShKT, protein containing ShKT domain; 5’‐NUC, 5′‐nucleotidase. CTX, SNTX, and WNTX are part of the three‐finger toxins (3FTX) superfamily. Others represent toxins of very low abundance and proteins of unknown toxin function.

FIGURE 4.

Proteome analysis of the venoms of non‐spitting cobras Naja anchietae (A), Naja annulifera (B), and Naja nivea (C). The venom peak fractions from the RP‐HPLC were analyzed with label‐free LC–MS/MS. The relative abundance of the proteins was grouped into different families. Abbreviations: CTX, cytotoxin/cardiotoxin; SNTX, short neurotoxin; LNTX, long neurotoxin; WNTX, weak toxin; NTL, neurotoxin‐like protein; PLA₂, phospholipase A2; SVMP, snake venom metalloproteinase; KSPI, Kunitz‐type serine protease inhibitor; CRISP, cysteine‐rich secretory protein; NGF, nerve growth factor; CVF, cobra venom factor; ShKT, ShKT domain‐containing protein; 5’‐NUC, 5′‐nucleotidase; C3, complement C3; VES, vespryn; SCP, SCP domain‐containing protein. CTX, SNTX, LNTX, NTL, and WNTX are part of the three‐finger toxins (3FTX) superfamily. Others represent toxins of very low abundance and proteins of unknown toxin function.

Figure 3A,B show that the spitting cobras N. pallida and N. n. woodi venoms were dominated by 3FTXs and PLA₂ families. For the N. pallida , 3FTXs were the most dominant family with a relative abundance of 54.50% (Figure 3A). The subfamilies comprise short neurotoxins (SNTX, 3.32%), long neurotoxins (LNTX, 0.01%), weak neurotoxins (WNTX, 0.62%), and cytotoxins (CTX, 50.55%) of the total venom proteome. After 3FTX, the second most dominant toxin family is phospholipase A2 (PLA₂, 39.34%). Other important toxins identified from the venom of N. pallida include snake venom metalloproteinase (SVMP, 0.80%), cysteine‐rich secretory protein (CRISP, 0.16%), and Kunitz‐type serine protease inhibitor (KSPI, 0.09%). Similarly, 3FTXs were the most dominant family for the N. n. woodi venom (Figure 3B). This toxin comprised 57.03% of the total venom protein. The subfamilies were made up of SNTX (0.19%), LNTX (0.01%), WNTX (1.06%), and CTX (55.78%) of the total venom proteome. PLA₂ (31.40%) was the second most abundant protein family. Other toxins with notable abundance include SVMP (3.73%), NGF (0.51%), CRISP (0.49%), and CVF (0.22%).

Unlike the spitting cobra, the non‐spitting cobra venoms have 3FTX as the only dominant toxins protein family (Figure 4A–C). Naja anchietae venom was dominated by 3FTXs (84.39%) with subfamilies such as SNTX (0.14%), WNTX (6.3%), and CTX (83.62%) (Figure 4A). Other toxins with notable toxin activities include SVMP (1.96%), PLA₂ (0.31%), CRISP (0.94%), NGF (1.22%), and 5′‐NUC (1.75%). In comparison, Naja annulifera venom contains 86.69% of the 3FTXs family, which comprised SNTX (1.36%), LNTX (0.07%), WNTX (4.03%), NTL (0.13%), and CTX (81.10%) subfamilies. The venom contains other important toxins, which include SVMP (0.10%), CRISP (0.28%), KSPI (0.13%), NGF (0.12%), 5′‐NUC (0.11%), and PLA₂ (0.05%). N. nivea venom, like all cobras studied, was dominated by 3FTXs (82.72%) (Figure 4C). The subfamily also had CTX (70%) as the most dominat toxin, followed by WNTX (6.22%), LNTX (3.24%), SNTX (2.02%), and NTL (1.2%). Other key toxins identified from N. nivea venom include SVMP (2.59%), CRISP (1.11%), KSPI (2.97%), NGF (0.77%), 5′‐NUC (0.26%), PLA₂ (0.45%), and VES (0.18%).

3.3. Anticancer Activity of Crude Venom and RP‐HPLC Fractions

3.3.1. Cytotoxicity of Crude Venoms on MIA PaCa‐2 Cell Lines

The crude venom from two spitting cobra venoms ( N. pallida and N. n. woodi) and three non‐spitting cobras of Uraeus subgenus ( N. anchietae , N. annulifera, and N. nivea ) were tested against MIA PaCa‐2 cell lines to investigate the possible cytotoxicity activities using the Alamar Blue assay. Crude venom from N. pallida with (IC₅₀ = 8.66) and N. n. woodi (IC₅₀ = 30.38) induced a concentration‐dependent decrease in cell viability in MIA PaCa‐2 cell lines (Figure 5A,B) compared to the MRC‐5 normal cell lines (Figure 6A,B). There was more pronounced cytotoxicity effect of N. pallida crude venom on MIA PaCa‐2 cell lines compared to the N. n. woodi venom. Interestingly, normal MRC‐5 cell lines showed no significant decrease in cell viability when treated with crude venom from N. pallida and N. n. woodi at the highest tested concentration of 20 μg/mL (Figure 6A,B). Conversely, non‐spitting cobra venoms were not cytotoxic against MIA PaCa‐2 (Figure 5C–E) nor MRC‐5 normal cell lines, except for a slightly decreased cell proliferation observed in the latter with N. anchietae and N. annulifera treatments (Figure 6C,D). Since spitting cobra venoms were effective against MIA PaCa‐2 cancer cell lines with no observable effect on the cell viability of MRC‐5 normal cell lines, the RP‐HPLC fractions thereof were used for further testing.

FIGURE 5.

Cytotoxicity of crude venoms against MIA PaCa‐2 cell lines. Venoms from the spitting cobras ( N. pallida (A) and N. n. woodi (B)) and non‐spitting cobras ( Naja anchietae (C), Naja annulifera (D), and N. nivea (E)) were used to treat the cancer cells to determine the cytotoxicity effects. Data was normalized to 100% against the untreated cells. Etoposide (1% w/v) was used as a positive control, while 0.1% 1X PBS was used as a negative control. Data from three independent replicates are presented as mean ± standard deviation (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001) where a p‐value less than 0.05 (p < 0.05) is significant to treatments that caused reduced cell viability compared to untreated control.

FIGURE 6.

Cytotoxicity of crude venoms against MRC‐5 normal cell lines. Venoms from the spitting cobras ( N. pallida (A) and N. n. woodi (B)) and non‐spitting cobras ( Naja anchietae (C), Naja annulifera (D), and N. nivea (E)) were used to treat the normal cell lines to determine the possible cytotoxicity effects. Data was normalized to 100% against the untreated cells. Etoposide (1% w/v) was used as a positive control, while 0.1% 1X PBS was used as a negative control. Data from three independent replicates are presented as mean ± standard deviation (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001) where a p‐value less than 0.05 (p < 0.05) is significant to treatments that caused reduced cell viability compared to untreated control.

3.3.2. Cytotoxicity of RP‐HPLC Venom Fractions on the MIA PaCa‐2 Cell Lines

Cytotoxicity of the two most prominent peaks from the RP‐HPLC fractions of N. pallida (Fractions 6 and 8) and N. n. woodi (Fractions 7 and 9) venoms were tested against MIA PaCa‐2 cell lines. There was a reduction in MIA PaCa‐2 cell proliferation observed with the N. pallida venom Fraction 6 (IC₅₀ = 37.15) at the concentrations 5, 10, and 40 μg/mL (Figure 7A). Conversely, reduction of cell proliferation was not observed when MIA‐PaCa‐2 cell lines were treated with N. pallida venom Fraction 8 (Figure 7B). In contrast N. n. woodi Fraction 9 showed a slight reduction of cell proliferation at concentrations 5 and 10 μg/mL, while there was no significant cytotoxicity observed with N. n. woodi Fraction 7 (Figure 7C,D).

FIGURE 7.

Cytotoxicity of fractions from the spitting cobra venoms against MIA PaCa‐2 cell lines. Fractions from N. pallida (A and B) and N. n. woodi (C and D) were used to treat the cancer cells to determine the cytotoxicity effects. Data was normalized to 100% against the untreated cells. Etoposide (1% w/v) was used as a positive control, while 0.1% 1X PBS was used as a negative control. Data from three independent replicates are presented as mean ± standard deviation (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001) where a p‐value less than 0.05 (p < 0.05) is significant to treatments that caused reduced cell viability compared to untreated control.

4. Discussion

While the two spitting cobras, N. pallida and Naja nigricincta woodi belong to the Afronaja subgenus, the three non‐spitting cobras Naja anchietae , Naja annulifera, and N. nivea are members of the Uraeus subgenus (Offor and Piater 2024a; Trape et al. 2009; Wallach et al. 2009; Wüster et al. 2018). These cobras are part of the Elapidae snake family that cause many snakebite envenomations in sub‐Saharan Africa (Benjamin et al. 2020; Gutiérrez et al. 2017). The chromatograms from the fractionation of these cobra venoms and SDS‐PAGE analysis revealed that all venom proteins eluted within 140 min. These toxins were mostly of low molecular weight below 15 kDa and are mainly 3FTXs and PLA₂ protein families, which dominate cobra venom proteome (Nguyen et al. 2022; Petras et al. 2011).

From the LC–MS/MS analysis data, it was observed that the 3FTX protein family dominated both the spitting and non‐spitting cobra venoms. CTXs, a subfamily member of the 3FTXs, were the most abundant in all the five Naja snake species studied, which is in agreement with previous cobra proteome data (Nguyen et al. 2022; Offor and Piater 2024a). Also, there were varying relative abundance of other 3FTX subfamilies, including LNTX, WNTX, and SNTX, detected from the venoms of these cobras, highlighting venom toxin variations within snakes of the Naja genus. The major difference between the spitting cobras and non‐spitting cobras (Uraeus subgenus) venom was the low abundance of PLA₂ in the latter, while PLA₂ was the second most dominant toxin in the former. These reports agree with the available proteomic data on African spitting and non‐spitting cobra venoms (Nguyen et al. 2022; Offor and Piater 2024a). PLA₂ and 3FTXs are believed to be responsible for the toxic effects, especially local tissue damage or cytotoxicity induced by spitting cobra venoms (Petras et al. 2011). Neurotoxicity, a major characteristic of non‐spitting and spitting cobra venom, is driven by the activities of 3FTXs through interactions with nicotinic acetylcholine receptors (nAChRs) in the neuronmuscular junction, which leads to paralysis and respiratory failure (Williams et al. 2011).

This study highlighted several intra‐ and inter‐species variations in comparison to available proteome data, which may be attributed to the distinct geographical location of the snakes. For instance, there were changes in the relative abundance of 3FTX (54.55%) and PLA₂ (39.34%) from the Tanzanian N. pallida venom in this study compared to a Kenyan counterpart with 3FTX and PLA₂ relative abundance of 67.70% and 30.10%, respectively (Petras et al. 2011). Additionally, this study identified other low‐abundance toxins, including CRISP, LAAO, NGF, and 5′‐NUC from the Tanzanian N. pallida , that were not detected in the previous studies (Nguyen et al. 2022; Petras et al. 2011). The South African N. n. woodi spitting cobra venom in this study had a relative abundance of 3FTX (57.04%) and PLA₂ (31.40%) compared to N. n. nigricincta subspecies from the same country with 3FTX (74.57%) and PLA₂ (22.244%) (Nguyen et al. 2022).

For the non‐spitting cobras, a Namibian N. anchietae venom contained 3FTX (96.65%) and PLA₂ (0.16%) (Nguyen et al. 2022) compared to the Botswanan N. anchietae venom in this study with 3FTX (84.39%) and PLA₂ (0.31%). Here, CRISP, NGF, and 5′‐NUC were detected but not in the previous study. N. annulifera (South Africa) venom from this study contained 3FTX (86.69%) and PLA₂ (0.05%), which is comparable to those from the same snake species from Mozambique (3FTX—79.20% and PLA₂—2.70%) (Sánchez et al. 2021). In another study, PLA₂ was not detected in a Mozambican N. annulifera venom, establishing the variations among snake species sourced from the same country (Tan et al. 2020). The South African N. nivea in this study had 3FTX (82.72%) and PLA₂ (0.45%) compared to two previous experiments whereby PLA₂ was detected in only one venom, even when both snakes were from South Africa (Nguyen et al. 2022; Tan et al. 2022). There were similarities and differences observed in the detection and relative abundance of other toxins, including SVMP, LAAO, CRISP, NGF, CVF, 5’‐NUC, and KSPI, when comparing our data and those previously published from N. pallida (Petras et al. 2011; Nguyen et al. 2022), N. anchietae (Nguyen et al. 2022), N. annulifera (Tan et al. 2020; Sánchez et al. 2021; Nguyen et al. 2022), and N. nivea (Nguyen et al. 2022; Tan et al. 2022; Lüddecke et al. 2024; McFarlane and Pukala 2024).

Ever since the FDA approved the first venom peptide‐based drug, Captopril, for the treatment and management of hypertension cardiovascular disease, the search for snake venom‐based drugs has intensified (Cushman and Ondetti 1991). Crude venoms from cobras such as Naja kaouthia , N. annulifera , and Ophiophagus hannah induced cytotoxicity, apoptosis, and antimetastasis against PaTu 8988 t pancreatic tumour cells (Kerkkamp et al. 2018). Snake venom crotoxin has also shown promise in Phase 1 clinical trials towards cancer (Cura et al. 2002). Crotoxin was administered intramuscularly for 30 days to sufferers of breast cancer, gastrointestinal cancer, nonsmall cell lung cancer, squamous cervix carcinoma, prostate cancer, thyroid carcinoma, larynx carcinoma, bladder carcinoma, fallopian tube adenocarcinoma, head and neck cancer, low grade fibrosarcoma, Erwing's sarcoma, and liposarcoma (Cura et al. 2002). Besides some manageable conditions, crotoxin's therapeutic response was promising with the recommended dose of 0.18 mg/m² for Phase 2 clinical trial. Although several venom toxins, including 3FTX, PLA₂, DIS, and LAAO, have been shown to be effective, potent anticancer agents, more studies are needed to elucidate their mechanisms of action and possibly progress to clinical studies (Offor and Piater 2024b).

In this study, crude venoms and RP‐HPLC fractions of two spitting cobras ( N. pallida and N. n. woodi) were cytotoxic against MIA PaCa‐2 cell lines compared to the MRC‐5 normal cell lines in a concentration‐dependent manner. Conversely, the three non‐spitting cobras of Uraeus subgenus (N.anchietae, N. annulifera, and N. nivea ) were not cytotoxic against MIA PaCa‐2 cell lines, with minimal reduction in cell proliferation in the N.anchietae‐ and N. annulifera ‐ treated MRC‐5 normal cell lines. Crude venom from Naja naja oxiana has been shown to be cytotoxic against human breast cancer (MCF‐7), human hepatocellular carcinoma (HepG2), and human prostate carcinoma (DU145) cell lines compared to normal cell lines (MDCK) (Ebrahim et al. 2016).

The RP‐HPLC fractions from the non‐spitting cobra crude venoms were discontinued since the crude venoms were not cytotoxic against the MIA PaCa‐2 cell lines. Two RP‐HPLC fractions from each of N. pallida ( N. pallida Fractions 6 and 8) and N. n. woodi (N. n. woodi Fractions 7 and 9) venoms were tested against MIA PaCa‐2 cell lines. For N. pallida , only Fraction 6 showed significant cytotoxicity against MIA PaCa‐2 cells at concentrations of 5 and 40 μg/mL. Also, for N. n. woodi only Fraction 9 induced minimal reduction of cell proliferation against MIA PaCa‐2 cell lines at 5 and 10 μg/mL concentration. Further analysis of the LC–MS/MS data revealed that the most dominant toxins detected in N. pallida Fraction 6 are the basic phospholipase 2 CM‐III and Cytotoxin 2 with proteome relative abundance of 36.7% and 2.0%, respectively (Table S1). N. n. woodi Fraction 9 was dominated by basic phospholipase 2 CM‐III, basic phospholipase 2 CM‐II and Cytotoxin 3 with proteome relative abundance of 16.1%, 5.3%, and 2.5%, respectively (Table S2). Further purification and study of N. pallida Fraction 6 and possibly N. n. woodi Fraction 9 is needed to elucidate the main toxin component responsible for the observed cytotoxicity against MIA PaCa‐2 cell lines. Studies have shown anticancer activities of PLA₂ and cytotoxins, among others. For instance, a basic Asp49‐PLA₂ from Bungarus fasciatus induced cytotoxicity and cell death in MCF‐7 (breast cancer) and A549 (lung cancer) cells compared to noncancerous human epithelial HK cells (Tran et al. 2019). PLA₂ from Iranian Vipera raddei kurdestanica induced cytotoxicity and apoptosis in MCF‐7 and MDA‐MB‐231 breast cancer cell lines compared to the MCF‐10 a normal human mammary epithelial cells (Malekara et al. 2020). Crotoxin (F1 CTX), a heterodimeric PLA₂ isolated from Crotalus durissus terrificus induced cytotoxicity against pancreatic cancer cell lines (PSN‐1 and PANC‐1) compared to the normal cells (Muller et al. 2018). This toxin also showed high sensitivity against glioma cells (GAMG and HCB151) and induced DNA damage via H2AX activity in both cells. NN‐32, a 3FTX, has been shown to anticancer potential activities against A549 lung cancer cells through induction of apoptosis, alteration of mitochondrial membrane potential, and cell cycle arrest (Kurkute et al. 2023). Cytotoxins (CT1 and CT2) isolated from Naja oxiana and CT1 and CT3 from Naja haje induced cytotoxicity against A549 and promyelocytic leukaemia HL60 cells (Feofanov et al. 2005). We cannot rule out the possibility of non‐spitting cobra venom and/or PLA₂ toxin inducing cytotoxicity against other tissues. Notably, the non‐spitting cobra venomm used in this study ( N. anchietae , N. annulifera, and N. nivea ) were of the Uraeus subgenus, whose PLA₂ composition is mostly less than 1% of the venom composition (Offor and Piater 2024a). Conversely, the venom of non‐spitting cobras of the Boulengerina subgenus has a comparable PLA₂ composition with the spitting cobras (Afronaja subgenus). Thus, we are tempted to believe that the majority of cytotoxicity activity of these Uraeus non‐spitting cobra venom is mainly contributed by the 3FTXs rather than the PLA₂. For instance, while Egyptian non‐spitting cobra ( N. haje ) venom showed anticancer properties against lymphoblastic leukemia 1301 (El Hakim et al. 2011), F7 fraction of Moroccan N. haje venom showed anti‐proliferative effect against hepatocellular carcinoma cell lines when compared to normal cell lines (Lafnoune et al. 2021). A recent study of cytotoxin peptides isolated from N. anchietae and N. senegalenses showed inhibition proliferation of U87 glioblastoma cells (Boughanmi et al. 2024).

While MRC‐5 normal cells were derived from normal lung fibroblast cells (connective tissue), MIA PaCa‐2 cells were derived from pancreatic carcinoma (epithelial tissue). MIA PaCa‐2 cells have a higher proliferation rate, altered phospholipid membranes and signalling pathways for venom toxins interaction and hydrolysis compared to MRC‐5 normal cell lines. These differences could regulate the accessibility, selectivity, and activities of the snake venom and/or toxins (e.g., PLA₂). While we acknowledge that the differences in the source of these cells could affect the activities of the venom toxins, negative controls (PBS‐ and DMEM‐treated cells) and positive controls (etoposide‐treated cells) were used to compare the cell viability data obtained from the snake venom and/or toxin‐treated cells.

5. Conclusion

Snakebite envenomation causes morbidity and mortality across sub‐Saharan Africa. There are several FDA‐approved drugs from snake venom toxins from other regions, including South America and Asia. Hence, there is a need to profile venoms from medically important Africa Naja spp. from various locations for possible anticancer properties. In this study, LC–MS/MS venom profiles showed that the 3FTXs family (mostly CTX subfamily) was the most dominat toxin in both spitting ( N. pallida and N. n. woodi) and non‐spitting ( N. anchietae , N. annulifera, and N. nivea ) cobras. Additionally, PLA₂ was the second most dominant toxin in the spitting cobra venom, but has a very low relative abundance in the non‐spitting cobra venom, thus suggesting the notion that it plays little or no role in the envenomation caused by the non‐spitting cobra of Uraeus subgenus. Other notable toxin families identified in this study with varied relative abundance include SVMP, CRISP, KSPI, CVF, NGF, and 5’‐NUC. To our knowledge, this study importantly presents the first proteomic profiling data of N. n. woodi. Here, we established that the spitting cobra venoms were cytotoxic against MIA PaCa‐2 cell lines in comparison to the normal MRC‐5 cell lines at the maximum concentration tested. Conversely, the venom of non‐spitting cobras was not cytotoxic against the MIA PaCa‐2 cell lines. The study further implicated the toxins content of most N. pallida Fraction 6 (basic phospholipase 2 CM‐III and Cytotoxin 2) and N. n. woodi fraction 9 (basic phospholipase 2 CM‐III, basic phospholipase 2 CM‐II and Cytotoxin 3) as potential peptides responsible for the decreased cell proliferation observed in the MIA PaCa‐2 cell lines. As such, further studies are required to identify the specific venom peptide toxin responsible for these cytotoxic effects. Solving the structure of African cobra snake venom toxins (e.g., PLA₂ and 3FTXs) can help in determining cancer cell selectivity and penetration. Also, developing a stable toxin conjugate or delivery system to the cancer target is key for targeted anticancer therapy. In addition, the anticancer effect of these venom toxins should be further tested against other cancer cells.

Author Contributions

Benedict C. Offor: methodology, investigation, validation, data curation, formal analysis, writing the original draft. Beric Muller: venom supplier. Lesetja R. Motadi: methodology, resources, manuscript editing. Lizelle A. Piater: methodology, project administration, resources, funding acquisition, manuscript writing and editing. All authors read and approved the final manuscript.

Ethics Statement

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1 Identification of proteins from RP‐HPLC fractions of Naja pallida using label‐free LC–MS/MS.

Table S2 Identification of proteins from RP‐HPLC fractions of Naja nigricincta woodi using label‐free LC–MS/MS.

Table S3 Identification of proteins from RP‐HPLC fractions of Naja anchietae using label‐free LC–MS/MS.

Table S4 Identification of proteins from RP‐HPLC fractions of Naja annulifera using label‐free LC–MS/MS.

Table S5 Identification of proteins from RP‐HPLC fractions of Naja nivea using label‐free LC–MS/MS.

Table S6 Families of identified proteins from N. pallida venom.

Table S7 Families of identified proteins from N. nigricincta woodi venom.

Table S8 Families of identified proteins from N. anchietae venom.

Table S9 Families of identified proteins from N. annulifera venom.

Table S10 Families of identified proteins from N. nivea venom.

Table S11 Other proteins identified from N. pallida venom.

Table S12 Other proteins identified from N. nigricincta woodi venom.

Table S13 Other proteins identified from N. anchietae venom.

Table S14 Other proteins identified from N. annulifera venom.

Table S15 Other proteins identified from N. nivea venom.

Offor, B. , Muller B., Motadi L., and Piater L.. 2025. “The Proteome of African Spitting and Non‐Spitting Cobra Venoms and Cytotoxicity Against Pancreatic Cancer Cells.” Journal of Applied Toxicology 45, no. 10: 2055–2067. 10.1002/jat.4825.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.