Establishment of snake venom gland organoids from a novel family, Colubridae

29 Non-front-fanged snakes are abundant, diverse and represent approximately 70% of extant 30 snakes. However, there is limited knowledge about most species and their venoms, in part due 31 to the technical and welfare challenges associated with venom extraction, low venom yields, 32 and the lack of cellular models available. Organoids represent an excellent opportunity to 33 overcome these challenges. Here, we establish, for the first time, venom gland organoids from 34 snakes of the Colubridae family and demonstrate the in vitro production of toxins. 35

36 Venomous snakes can broadly be divided into front-fanged (Elapidae, Viperidae and 37 Atractaspididae) and non-front-fanged (Ahaetuliinae, Calamariidae, Colubrinae, Dipsadidae, 38 Grayiidae, Natricidae, Pseudoxenodontidae, Sibynophis, and Scaphiodontophis) species 39 (Figueroa et al., 2016; Pyron et al., 2013; Zaher et al., 2019). Front-fanged snakes are the most 40 medically important snakes, possessing enlarged, hollow maxillary teeth (fangs) and a venom 41 gland with an internal reservoir for venom storage (Mackessy, 2022). However, non-front-42 fanged snakes are more diverse and widespread, comprising approximately 70% of all extant 43 snake species (Mackessy and Saviola, 2016). 44 Non-front-fanged snakes possess enlarged or grooved solid rear maxillary teeth and a 45 Duvernoy’s venom gland (Mackessy, 2022). Superficially similar to the venom gland of front 46 .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 31, 2026. ; https://doi.org/10.64898/2026.03.27.714740doi: bioRxiv preprint 3 fanged snakes, the Duvernoy’s venom gland is also located in the temporal region of the upper 47 jaw. However, lacking an internal reservoir, venom is stored intracellularly and exocytosed into 48 secretory tubules that flow into luminal ducts. Unlike the venom glands of front-fanged snakes, 49 the Duvernoy’s venom gland is usually not surrounded by musculature that pressurise the 50 gland, and venom is more slowly secreted around the base of the fang during prolonged biting 51 (Kardong and Lavin-Murcio, 1993). 52 Understanding snake venom production, function and composition has mostly been focussed 53 on front-fanged snakes. In line with this, organoids – self-organising 3D structures that grow 54 from stem cells and produce organ-specific cell types through lineage differentiation (Clevers, 55 2016) – have previously been established from the venom glands of Elapidae and Viperidae. 56 Snake venom gland organoids have been shown to recapitulate the venom gland through the 57 production of functionally active toxins (Post et al., 2020). Here, we have established 58 Duvernoy’s venom glands organoids from a non-front-fanged snake, Boiga dendrophila, and 59 compare these organoids to those from Bitis arietans (Viperidae) venom gland. B. dendrophila, 60 commonly known as the Mangrove Catsnake, is a large non-front fanged snake that is 61 widespread across southeast Asia. B. dendrophila venom is primarily composed of prey-62 specific three-finger toxins (3FTxs) (Dashevsky et al., 2018; Mackessy and Saviola, 2016; 63 Modahl and Mackessy, 2019), such as Boigatoxin-A (Lumsden et al., 2005) and Denmotoxin B 64 (Pawlak et al., 2006). This new organoid model represents an excellent opportunity to study the 65 evolution of venom systems, venom production and further characterise venom glands and 66 toxins from rear-fanged snakes. 67

68 Venom gland tissue and venom 69 Bitis arietans venom was sourced from the on -site herpetarium at the Liverpool School of 70 Tropical Medicine (LSTM). The facility and its protocols for the husbandry of snakes are approved 71 .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 31, 2026. ; https://doi.org/10.64898/2026.03.27.714740doi: bioRxiv preprint 4 by the UK Home Office and the LSTM Animal Welfare and Ethical Review Boards. Boiga 72 dendrophila venom was kindly donated by Stephen Mackessy, University of Col orado. Boiga 73 venom was collected following an IACUC approved protocol (0902E-SM-MBirdsL-12) as detailed 74 by Hill & Mackessy with the use of 30 µg/g ketamine and 6.0 µg/g pilocarpine (Hill and Mackessy, 75 1997). All venoms were collected either prior to or in accordance with the Nagoya protocol. 76 Snake venom glands used in this study , specifically Bitis arietans (captive-bred and Nigerian 77 locality; n=2) and Boiga dendrophila (captive-bred, n=2), came from snakes in the LSTM 78 herpetarium. Snakes were euthanised by overdose of anaesthesia and cervical dislocation. 79 Snake venom glands were dissected out as described in Puschhof et al. (Puschhof et al., 2021). 80 Snake venom gland organoids from Naja haje (captive-bred) were supplied by Jens Puschhof and 81 Yorick Post, Hubrecht Institute 82 Organoid establishment 83 Organoids were setup using methods similar to those detailed in Puschhof et al. (2021). In brief, 84 venom glands were either dissected from freshly euthanized snakes or collected from tissue 85 stored in liquid nitrogen in 90% (v/v) fetal bovine serum , 10% (v/v) DMSO. Under sterile 86 conditions, connective tissue was removed and gland cut into small pieces. Tissue was treated 87 using am antibiotic cocktail media (DMEM, 50 µg/ml gentamicin, 2.5 µg/ml ciprofloxacin, 20 µM 88 erythromycin, 10 nM azithromycin dehydrate, 100 µg/ml primocin) and digested using 89 collagenase media solution (Advanced DMEM, 1 mg/ml collagenase, 10 µM Y-27632) at 32 °C, 90 5% CO2 for up to 1 h. Digested structures were plated at a density of approximately 100 per 15 µl 91 in Cultrex basement membrane extract growth factor reduced type 2 (R&D Systems, 3533-001-92 02) and overlayed with expansion media ( Advanced DMEM, 1x GlutaMAX, 1x HEPES, 100 U/ml 93 penicillin-streptomycin, 100 µg/ml Primocin, 1.25 mM N-acetyl-L-cysteine, 1x B27 supplement, 94 10 mM nicotinamide, 1 µM prostaglandin E2, 2% (v/v) Noggin-conditioned media (IPA 95 Therapeutics N002), 2% (v/v) R-spondin-conditioned media (IPA Therapeutics R002), 50 ng/ml 96 .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 31, 2026. ; https://doi.org/10.64898/2026.03.27.714740doi: bioRxiv preprint 5 human epidermal growth factor, 0.5 µM A83 -91, 10 nM human gastrin-1, 100 ng/ml human 97 fibroblast growth factor -10) containing 10 µM Y -27632, 50 µg/ml gentamicin, 2.5 µg/ml 98 ciprofloxacin, 20 µM erythromycin and 10 nM azithromycin dehydrate. Expansion media was 99 changed every other day, and organoids were passaged by mechanical dissociation with a 27G 100 needle after 7 to 14 days. 101 Organoid differentiation and lysate harvest 102 Organoids were passaged at least once prior to differentiation. Organoids were grown in 103 expansion medium for 3 days then overlayed with differentiation medium (advanced DMEM, 1x 104 GlutaMAX, 1x HEPES, 100 U/ml penicillin-streptomycin, 100 µg/ml Primocin, 1.25 mM N-acetyl-105 L-cysteine, 1x B27 supplement, 10 mM nicotinamide, 1 µM prostaglandin E2) for 7 days with 106 media changed every other day. Organoid lysates were then harvested by sonication for 10s and 107 protein concentration measured by the Qubit Protein Assay (Thermofisher Scientific) following 108 manufactures instructions. 109 Organoid staining 110 Organoids were harvested with Cultrex Organoid Harvesting Solution (R&D Systems, 3700-100-111 01). Active caspase-3 activity was determined using CellEvent Caspase-3/7 Gren detection 112 reagent (Thermofisher Scientific, C10423). Organoids for hematoxylin and eosin (H&E) and 113 periodic acid-Schiff (PAS) staining were resuspended in HistoGel (Fisher Scientific, 12006679) 114 and staining outsourced to Liverpool Shared Research Facility, University of Liverpool. 115 Western blot 116 Western blot was undertaken as detailed in Dawson et al with 1.6 μg of organoid lysates and 117 0.64 μg venom (Dawson et al., 2024). Antivenom used for the detection of B. arietans toxins was 118 EchiTAb-Plus-ICP (lot 7131123PALF, Expiry: 11/2028), a polyspecific antivenom raised against a 119 mixture of African snakes including B. arietans. Neuro Polyvalent Snake Antivenom (NPAV, lot 120 NP00515, Expiry: 29/9/2020) – an antivenom targeting neurotoxic venoms from Naja kaouthia, 121 .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 31, 2026. ; https://doi.org/10.64898/2026.03.27.714740doi: bioRxiv preprint 6 Ophiophagus hannah, Bungarus candidus and B. fasciatus – was used for the detection of B. 122 dendrophila toxins. Antivenoms were standardised to 10mg/ml and then diluted 1: 1,000. As in 123 Dawson et al., Horse IgG (H&L) Antibody DyLightTM800 Conjugated (Cambridge Bioscience, 124 608445002, 1:15,000) was utilised as the secondary antibody. 125 Activity assays 126 Organoids protein lysates (200 µg/ml) and venom (100 µg/ml – 10,000 µg/ml) were analysed for 127 snake venom metalloproteinase (SVMP) and phospholipase A2 (PLA2) activity. SVMP activity 128 was determined in vitro using a quenched fluorogenic substrate (ES010 Mca-KPLGL-Dpa-AR-129 NH2, R&D Bio­systems) as detailed previously (Menzies et al., 2022). PLA2 activity was 130 determined in vitro using the secretory phospholipase A2 assay kit from Abcam (AB133089) 131 following manufacture instructions. Inhibition of muscle-type nicotinic acetylcholine receptor 132 (nAChR) by B. dendrophila organoid protein lysates (60 µg/ml) and venom (0.1 µg/ml - 600 133 µg/ml) was determined using a membrane potential dye on TE671 cells expressing the γ-134 subunit containing muscle-type nAChR, as detailed in Patel et al. (Patel et al., 2023). Alpha-135 bungarotoxin (α-BgTx), isolated from Bungarus multicinctus venom was purchased from 136 Invitrogen (B1601) and used as a positive control at 90nM in the nAChR inhibition assay. 137 Statistical analysis 138 Data visualisation and statistical analysis was undertaken GraphPad Prism 10.2.1. Normality 139 was assessed by Shapiro-Wilk test and visualisation with Q-Q plots, and parametric or non-140 parametric statistical tests were undertaken as appropriate. Differences in SVMP activity were 141 assessed by comparing area under the curve. Differences in PLA2 activity were assessed by 142 comparing the final time points (60 min). Differences in inhibition of activated nAChR were 143 assessed by normalization to control acetylcholine (ACh)-induced nAChR activation. 144 .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 31, 2026. ; https://doi.org/10.64898/2026.03.27.714740doi: bioRxiv preprint 7

and Discussion 145 Establishment of Colubridae organoids 146 Boiga dendrophila venom gland was digested, resulting in the release of small structures 147 (Figure 1A, day 0) that generated proliferating organoids (Figure 1A, day 2 - 7). For comparisons 148 to front-fanged snakes, Bitis arietans venom glands organoids were also established. Digestion 149 of B. arietans venom glands resulted in the release of structures larger than that of B. 150 dendrophila (Figure 1A, day 0) which formed organoids that expanded over time (Figure 1A, day 151 2 - 7). Expansion of both B. dendrophila and B. arietans organoids was challenging, and 152 therefore Naja haje (Elapidae) organoids were cultured as a control (Figure 1A). These 153 organoids were non-granular and expanded rapidly. In accordance, the only published 154 accounts of long-term culture of snake venom gland organoids are those established from 155 Elapidae (Mackessy, 2022; Puschhof et al., 2021). 156 We then assessed whether B. dendrophila organoids could be established from tissue stored in 157 liquid nitrogen, because freshly excised snake venom gland tissue is not readily available to 158 many researchers. Establishment of B. dendrophila organoids from stored tissue samples, 159 utilising the same method as on fresh tissue, was unsuccessful (Supplementary Figure 1). 160 Histology of Colubridae organoids 161 Organoid morphology and cellular organization was assessed by H&E and PAS staining and 162 compared to native tissue. Organoids showed a high level of similarity to native tissue. The 163 majority of B. dendrophila organoids were lobular in shape composed of thick swirls of cells 164 around a small lumen (Figure 1B), with smaller cavities and thicker cellular linings than B. 165 arietans organoids. Some B. dendrophila organoids had a more spherical morphology 166 (Supplementary Figure 2). In organoids from both species, PAS-positive granules and cells in 167 the luminal lining confirmed the presence of mucosecretory cells, akin to tubules in the tissue. 168 .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 31, 2026. ; https://doi.org/10.64898/2026.03.27.714740doi: bioRxiv preprint 8 Further analysis, such as single cell RNA sequencing and comparisons of transmission electron 169 micrographs, will be required to compare cell types (Mackessy, 2022). 170 Organoids were then differentiated, resulting in reduced organoid expansion. No visible 171 changes in B. dendrophila organoid morphology were observed under light microscope (Figure 172 2C). However, differentiation of B. arietans organoids appeared to induce some cell death. This 173 was confirmed by active caspase-3 staining (Supplementary Figure 3). In addition, some 174 extracellular matrix domes were completely degraded upon differentiation of B. arietans 175 organoids, resulting in organoids becoming two-dimensional (Supplementary Figure 4). It is 176 tempting to speculate that this could be due to SVMPs degrading the basement membrane 177 extract (BME). Consistent with this hypothesis, BME breakdown has been observed in cultures 178 of Crotalus atrox (Viperidae) venom gland organoids, but not Crotalus scutulatus venom gland 179 organoids, which are species with SVMP-rich and SVMP-poor venom, respectively (Dr Timothy 180 A. Peterson, University of Maryland, personal communications). 181 Venom toxin content in organoids lysates 182 Next it was assessed whether B. dendrophila and B. arietans venom gland organoids produced 183 venom similar in composition to living snakes (Figure 2A-B). Western blotting of B. arietans 184 venom revealed EchiTAb-Plus-ICP immunoreactive proteins at approximately 10 - 15 kDa and 185 40 - 70 kDa, consistent with the molecular weights of SVMPs and PLA2s (Bartlett et al., 2025; 186 Howard, 1975). The low molecular weight band (10 - 15k Da) was observed in both 187 differentiated and undifferentiated B. arietans organoid lysates, whilst a band at ~70 kDa was 188 only observed in undifferentiated organoids. These findings were also observed in organoids 189 lysates from B. arietans with Nigerian locality (Supplementary Figure 5), suggesting the 190 production of SVMPs and PLA2s in undifferentiated B. arietans organoids. 191 Due to the lack of Boiga spp. antivenom, Neuro Polyvalent Snake Antivenom (NPAV) – an 192 antivenom raised against Asian species with high 3FTx content – was utilised in an attempt to 193 .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 31, 2026. ; https://doi.org/10.64898/2026.03.27.714740doi: bioRxiv preprint 9 detect B. dendrophila toxins. NPAV had cross reactivity with B. dendrophila venom, but only low 194 immunoreactivity towards 3FTxs (10 - 15 kDa) (Supplementary Figure 6). Immunoreactive 195 proteins at 15 - 25 kDa, 55 kDa, 55 - 100 kDa and 100 - 130 kDa were seen for the 196 undifferentiated organoids and B. dendrophila venom. Bands at 55 to 100 kDa were detected in 197 differentiated B. dendrophila organoids (Figure 2A). Using information from published accounts 198 of Boiga venom composition, we speculate the 20 kDa organoid toxins detected to be CRISPs 199 (Dashevsky et al., 2018) and the ~55 kDa toxin to be SVMPs (Pla et al., 2018). The two toxins in 200 B. dendrohila venom at 55 - 100 and 100 - 130 kDa have been detected previously but remain 201 uncharacterised (Pla et al., 2018), preventing us from identifying these toxins in the organoids. 202 We were unable to confirm the presence of 3FTxs (toxins smaller than 15 kDa, typically 6 – 8.1 203 kDa) in organoids, potentially due to the low reactivity NPAV against B. dendrophila 3FTxs. 204 These findings demonstrate the capacity of non-front fanged snake venom gland organoids to 205 produce toxins that reflect the protein composition of whole venom. However, top-down 206 proteomics and transcriptomics analysis will be required to validate the presence of specific 207 toxins and further understand the differences between differentiated and undifferentiated 208 venom gland organoids. 209 Total protein staining showed the presence of many additional bands in the organoid lysates 210 that were not present in venom collected from either B. arietans or B. dendrophila. This is 211 because venom was harvested from organoids by sonication, resulting in the release of 212 intracellular non-toxin proteins. It would be interesting to explore alternative methods of 213 organoid toxin harvest by mechanical disruption such as freeze-thaw and repeat pipetting with 214 a narrow tip, in an attempt to reduce non-toxin protein content. 215 Venom toxin activity in organoids lysates 216 Next, we assessed if the toxins produced by venom gland organoids were functional. In 217 accordance with the SVMP activity detected in B. arietans venom (100 - 1,000 µg/ml, ANOVA, p 218 .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 31, 2026. ; https://doi.org/10.64898/2026.03.27.714740doi: bioRxiv preprint 10 < 0.001, Figure 2C) and B. dendrophila venom (1,000 µg/ml - 10,000 µg/ml, ANOVA, p < 0.05, 219 Figure 2D), undifferentiated B. arietans and B. dendrophila organoid lysates displayed 220 significant SVMP activity (ANOVA, p > 0.05, Figure 2E-F), consistent with the lack of antivenom-222 reactive proteins at ~55kDa in these samples. 223 In accordance with previous studies (Broaders and Ryan, 1997; Hill and Mackessy, 2000; 224 Howard, 1975), PLA2 activity was observed in B. arietans venom (100 - 1,000 µg/ml, ANOVA, p < 225 0.001, Figure 2G) and B. dendrophila venom (1,000 - 10,000 µg/ml, ANOVA, p < 0.05, Figure 2H). 226 Again, undifferentiated organoid lysates had PLA2 activity (ANOVA, p 0.05, Figure 2I-J). 228 B. dendrophila venom displayed dose-dependent α-neurotoxin activity; inhibiting ACh-induced 229 activation of the muscle-type nicotinic acetylcholine receptor (nAChR) with an IC50 of 0.86 230 µg/ml (Figure 2K). Interestingly, this is in the same order of magnitude as Dendroaspis viridis 231 venom (IC50 of 0.59 µg/ml) (Patel et al., 2023) and represent the first report of B. dendrophila 232 venom or toxin inhibiting a human nAChR. Previous studies have examined the chick or mouse 233 nAChR (Lumsden et al., 2004; Lumsden et al., 2005; Pawlak et al., 2006). Dosing with 234 differentiated and undifferentiated organoids slightly decreased ACh-induced nAChR 235 activation, however, this was not significant (ANOVA, p > 0.05, Figure 2L). This is potentially due 236 to the lower concentration used compared to the PLA2 and SVMP assay, as samples had to be 237 diluted to obtain the larger volume required for this assay. It would have been interesting to 238 repeat experiments with higher concentrations of organoid lysates. However, the limited 239 expansion of the organoids prevented this and highlights the need for further studies optimising 240 the sub-culture of this family of organoids. 241 These findings hint towards the production of functionally active toxins in organoids, however, it 242 is important to note the presence of intracellular non-toxin proteins in organoid lysates that 243 .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 31, 2026. ; https://doi.org/10.64898/2026.03.27.714740doi: bioRxiv preprint 11 could interfere with assay results. It will be important to isolate and test specific organoid 244 toxins by HPLC to test this hypothesis. 245 In conclusion, this is the first establishment of snake venom gland organoids from non-front-246 fanged snakes. Venom extraction from this group of snakes is challenging, providing very 247 limited yields and requiring extraction chemicals which contaminate the venom (Damm et al., 248 2025). This research takes a step towards utilising organoids as a tractable and renewable 249 platform for venom production and functional studies. Boiga spp. organoids specifically are a 250 valuable model to investigate evolutionary adaptations due to their prey-specific toxins, and 251 serve as a comparative baseline for the more specialised venom glands of front-fanged snakes. 252 Non-front-fanged organoids provide a powerful platform to investigate many aspects of snake 253 venom glands, from establishing cell types and lineages, to identifying drivers of venom 254 diversity, to studying venom synthesis and secretion (Mackessy, 2022). 255

256 Thank you to Gemma Charlesworth and Marie O’Brien at the University of Liverpool Shared 257 Research Histology Core Facility for undertaking the H&E and PAS staining (RRID:SCR_026606). 258 Thank you to Grant Hughes and Shaun Pennington for instrument support. 259 260 261 262 263

264 Bartlett, K. E., Westhorpe, A., Wilkinson, M. C. and Casewell, N. R. (2025). Snake venom 265 metalloproteinases from puff adder and saw-scaled viper venoms cause cytotoxic effects 266 in human keratinocytes. Toxins 17, 328. 267 .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 31, 2026. ; https://doi.org/10.64898/2026.03.27.714740doi: bioRxiv preprint 12 Broaders, M. and Ryan, M. F. (1997). Enzymatic properties of the Duvernoy’s secretion of 268 Blanding’s Tree snake (Boiga blanding) and the Mangrove Snake ((Boiga dendrophila). 269 Toxicon 35, 1148. 270 Clevers, H. (2016). Modeling development and disease with organoids. Cell 165, 1586–1597. 271 Damm, K., Hartwig, C., Vilcinskas, A., Mackessy, S. P., Lüddecke, T. and Damm, M. (2025). 272 Ketamine and metabolites in snake venom: effects of venom extraction and potential 273 impact on animal models. Sci. Rep. 15, 1–5. 274 Dashevsky, D., Debono, J., Rokyta, D., Nouwens, A., Josh, P. and Fry, B. G. (2018). Three-275 finger toxin diversification in the venoms of cat-eye snakes (Colubridae: Boiga). J. Mol. 276 Evol. 86, 531–545. 277 Dawson, C. A., Bartlett, K. E., Wilkinson, M. C., Ainsworth, S., Albulescu, L. O., Kazandijan, 278 T., Hall, S. R., Westhorpe, A., Clare, R., Wagstaff, S., et al. (2024). Intraspecific venom 279 variation in the medically important puff adder (Bitis arietans): Comparative venom gland 280 transcriptomics, in vitro venom activity and immunological recognition by antivenom. 281 PLoS Negl. Trop. Dis. 18, 1–22. 282 Figueroa, A., McKelvy, A. D., Grismer, L. L., Bell, C. D. and Lailvaux, S. P. (2016). A species-283 level phylogeny of extant snakes with description of a new colubrid subfamily and genus. 284 PLoS One 11, 1–31. 285 Hill, R. E. and Mackessy, S. P. (1997). Venom yields from several species of colubrid snakes 286 and differential effects of ketamine. Toxicon 35, 671–678. 287 Hill, R. E. and Mackessy, S. P. (2000). Characterization of venom (Duvernoy’s secretion) from 288 twelve species of colubrid snakes and partial sequence of four venom proteins. Toxicon 289 38, 1663–1687. 290 Howard, N. L. (1975). Phospholipase As from puff adder (Bitis arietans) venom. Toxicon 13, 21–291 30. 292 Kardong, K. V. and Lavin-Murcio, P. A. (1993). Venom delivery of snakes as high-pressure and 293 low-pressure systems. Copeia 1993, 644–650. 294 Lumsden, N. G., Fry, B. G., Manjunatha Kini, R. and Hodgson, W. C. (2004). In vitro 295 neuromuscular activity of ‘colubrid’ venoms: clinical and evolutionary implications. 296 Toxicon 43, 819–827. 297 Lumsden, N. G., Fry, B. G., Ventura, S., Manjunatha Kini, R. and Hodgson, W. C. (2005). 298 Pharmacological characterisation of a neurotoxin from the venom of Boiga dendrophila 299 (Mangrove catsnake). Toxicon 45, 329–334. 300 Mackessy, S. P. (2022). Venom production and secretion in reptiles. Journal of Experimental 301 Biology 225, 1–10. 302 Mackessy, S. P. and Saviola, A. J. (2016). Understanding biological roles of venoms among the 303 Caenophidia: the importance of rear-fanged snakes. Integr. Comp. Biol. 56, 1004–1021. 304 Menzies, S. K., Clare, R. H., Xie, C., Westhorpe, A., Hall, S. R., Edge, R. J., Alsolaiss, J., 305 Crittenden, E., Marriott, A. E., Harrison, R. A., et al. (2022). In vitro and in vivo preclinical 306 .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 31, 2026. ; https://doi.org/10.64898/2026.03.27.714740doi: bioRxiv preprint 13 venom inhibition assays identify metalloproteinase inhibiting drugs as potential future 307 treatments for snakebite envenoming by Dispholidus typus. Toxicon X 14, 1–10. 308 Modahl, C. M. and Mackessy, S. P. (2019). Venoms of rear-fanged snakes: new proteins and 309 novel activities. Front. Ecol. Evol. 7, 1–18. 310 Patel, R. N., Clare, R. H., Ledsgaard, L., Nys, M., Kool, J., Laustsen, A. H., Ulens, C. and 311 Casewell, N. R. (2023). An in vitro assay to investigate venom neurotoxin activity on 312 muscle-type nicotinic acetylcholine receptor activation and for the discovery of toxin-313 inhibitory molecules. Biochem. Pharmacol. 216, 1–13. 314 Pawlak, J., Mackessy, S. P., Fry, B. G., Bhatia, M., Mourier, G., Fruchart-Gaillard, C., 315 Servent, D., Ménez, R., Stura, E., Ménez, A., et al. (2006). Denmotoxin, a three-finger 316 toxin from the colubrid snake Boiga dendrophila (mangrove catsnake) with bird-specific 317 activity. Journal of Biological Chemistry 281, 29030–29041. 318 Pla, D., Petras, D., Saviola, A. J., Modahl, C. M., Sanz, L., Pérez, A., Juárez, E., Frietze, S., 319 Dorrestein, P. C., Mackessy, S. P., et al. (2018). Transcriptomics-guided bottom-up and 320 top-down venomics of neonate and adult specimens of the arboreal rear-fanged Brown 321 Treesnake, Boiga irregularis, from Guam. J. Proteomics 174, 71–84. 322 Post, Y., Puschhof, J., Beumer, J., Kerkkamp, H. M., de Bakker, M. A. G., Slagboom, J., de 323 Barbanson, B., Wevers, N. R., Spijkers, X. M., Olivier, T., et al. (2020). Snake venom 324 gland organoids. Cell 180, 233-247.e21. 325 Puschhof, J., Post, Y., Beumer, J., Kerkkamp, H. M., Bittenbinder, M., Vonk, F. J., Casewell, 326 N. R., Richardson, M. K. and Clevers, H. (2021). Derivation of snake venom gland 327 organoids for in vitro venom production. Nat. Protoc. 16, 1494–1510. 328 Pyron, R. A., Burbrink, F. T. and Wiens, J. J. (2013). A phylogeny and revised classification of 329 Squamata, including 4161 species of lizards and snakes. BMC Evol. Biol. 13, 1–53. 330 Zaher, H., Murphy, R. W., Arredondo, J. C., Graboski, R., Machado-Filho, P. R., Mahlow, K., 331 Montingelli, G. G., Quadros, A. B., Orlov, N. L., Wilkinson, M., et al. (2019). Large-scale 332 molecular phylogeny, morphology, divergence-time estimation, and the fossil record of 333 advanced caenophidian snakes (Squamata: Serpentes). PLoS One 14, 1–82. 334 335 .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 31, 2026. ; https://doi.org/10.64898/2026.03.27.714740doi: bioRxiv preprint Figure 1. Establishment, growth and differentiation of snake venom gland organoids. Tissue fragments were isolated from freshly excised tissue of snake venom gland of Boiga dendrophila and Bitis arietans and expanded for 7 days in proliferation media. Naja haje organoids were revived from liquid nitrogen and expanded. A) Bright field images were taken to monitor organoid formation and growth. B) B. dendrophila and B. arietans organoids were stained with H&E and PAS. C) Light microscopy of differentiated and undifferentiated B. dendrophila and B. arietans venom gland organoids . Day 0 Day 2 Day 5 Day 7 A C Differentiated organoids Undifferentiated organoids B. dendrophila B. arietans B B. arietansB. dendrophila H&E PAS H&E PAS Organoids Tissues B. arietansB. dendrophila N. haje .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 31, 2026. ; https://doi.org/10.64898/2026.03.27.714740doi: bioRxiv preprint Figure 2. Toxin composition and activity of B. dendrophila and B. arietans venom and venom gland organoids. A) Toxin composition of B. arietans organoids and venom were analysed by Western blot using antivenom EchiTAb-plus. B) Toxin composition of B. dendrophila organoids and venom were analysed by Western blot using antivenom NPAV. C-F) SVMP activity of B. arietans venom (C), B. arietans organoids (D), B. dendrophila venom (E), B. dendrophila organoids (F). G-J) PLA2 activity of B. arietans venom (G), B. arietans organoids (H) B. dendrophila venom (I) and B. dendrophila organoids (J). K-L) Nicotinic acetylcholine receptor inhibition by B. dendrophila venom (K) and organoids (L). * p< 0.05, ** p< 0.05, *** p<0.001 relative to control for SVMP and PLA2 assay, and relative to ACh for neurotoxicity assay. Abbreviations: 3FTx, three finger toxins; ACh, acetylcholine; BgTx, Bungarotoxin; CRISPs, cysteine-rich secretory proteins; Dif, differentiated organoids; PLA2, phospholipase 2; SVMP, snake venom metalloproteinases; Undif, undifferentiated organoids, CA * *** E G I D B F H J K L *** *** *** *** * *** *** *** *** * **** *** * * SVMP activity PLA2 activity nAChR antagonism .CC-BY 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 31, 2026. ; https://doi.org/10.64898/2026.03.27.714740doi: bioRxiv preprint