Multiple egg defenses arose early in the history of apple snails preceding carotenoid-bound perivitellins

Apple snails ( Pomacea) lay egg masses above the waterline that are protected by complex biochemical defenses, yet current knowledge is largely derived from invasive and phylogenetically derived clades. Here, we investigate the early evolution of these defenses characterizing the eggs of Pomacea sordida , a representative of Flagellata, the most basal clade of the genus. We analyzed the biochemical composition of eggs and assessed defensive traits previously described in derived lineages. Eggs contained predominantly carbohydrates (~17,4% dw), followed by proteins (~13,9% dw) and low lipid levels, indicating carbohydrate-based embryonic energetics as in most gastropods. Carotenoid content was lowest in P. sordida and increased among more derived species. Unlike all Pomacea species studied, P. sordida eggs lacked the dominant carotenoid-binding perivitellin-1 (PV1), a trait associated with the lowest carotenoid content reported in the genus. Nonetheless, mass spectrometry revealed multiple PV1 subunits, representing the first evidence of a carotenoid-free PV1 in Pomacea . Despite lacking the dominant carotenoid-binding form of PV1, other components of the multilayered defense system were found in P. sordida eggs. Several proteins (22–66 kDa) were resistant to simulated gastrointestinal and fungal digestion, and PVF has strong protease inhibitors of animal, bacterial, and fungal origin (40-87% inhibition). Additionally, lectin-like activity agglutinated erythrocytes and both Gram-positive and Gram-negative bacteria. PV2-like neuro-enterotoxin was detected, compatible with the finding that the egg extracts were lethal to mice when injected and, when administered orally, induced marked but reversible morphological alterations in the murine intestine, including reduced absorptive surface, villus edema, and epithelial degeneration. Overall, these results demonstrate that key components of Pomacea egg-defense repertoire are already present in a basal lineage, suggesting that the evolution of the protein-based defense system observed in derived clades arose early in the history of the genus and later diversified into distinct clade-specific defensive strategies.

Eating and being eaten are central forces shaping ecological interactions and the evolution of defensive strategies. Predator-prey interactions are major drivers of defensive adaptation (Dawkins and Krebs, 1979; Vermeij, 1994). Eggs are especially vulnerable to predation because they are nutrient-rich and immobile (Ricklefs, 1969; Refsnider and Janzen, 2010), leading many taxa to evolve physical and biochemical defenses. In this context, apple snails of the genus Pomacea are unique among freshwater gastropods. Although these amphibious snails lay egg masses outside the water that are rich in macronutrients such as carbohydrates and proteins, eggs are rarely preyed upon (Hayes et al., 2015). This apparent paradox is explained by the evolution of a suite of biochemical defenses that protect embryos from predators, pathogens, and the environmental stresses associated with aerial deposition (Hayes et al., 2015). Unlike basal ampullariid genera that lay unpigmented gelatinous egg masses underwater, Pomacea experienced a major evolutionary and ecological transition to aerial egg deposition (Hayes et al., 2015). This shift is thought to have facilitated dispersal and diversification of the genus (Hayes et al., 2009a) generally attributed to the avoidance of aquatic egg predators, oxygen availability, incubation temperature, mechanical damage and cannibalism avoidance (Turner, 1998). Phylogenetically, Pomacea comprises four major clades that exhibit distinct evolutionary arrangements of biochemical egg defenses, from the most basal to the derived: Flagellata, with only a four species with restricted distributions, Effusa, Bridgesii, and the more derived and speciose Canaliculata (Figure 1). The latter includes the worldwide invasive species P. canaliculata (Lamarck, 1822) and P. maculata Perry, 1810 (Hayes et al., 2009b). Aerial deposition exposes eggs to solar radiation, high temperatures, desiccation, and terrestrial predators. To protect embryos from these stressors, Pomacea eggs contain a perivitelline fluid (PVF) that surrounds the embryo, as in most gastropods. In derived clades, PVF functions not only as embryonic nourishment but also as a complex biochemical defense that includes mainly polysaccharides and proteins called perivitelins (Garin et al., 1996; Heras et al., 1998, 2008; Dreon et al., 2006, 2010, 2014; Ituarte et al., 2008, 2012, 2018, 2019; Giglio et al., 2016; Pasquevich et al., 2017; Brola et al., 2020, 2021).

PVF defenses in species from the Bridgesii and Canaliculata clades include perivitellins resistant to gastrointestinal digestion, broad-spectrum protease inhibitors, toxic dietary lectins, and, in the Canaliculata clade, a neuro-enterotoxin, collectively preventing digestion of egg proteins (Giglio et al., 2016; Ituarte et al., 2019). The most abundant of these perivitellins is perivitellin-1 (PV1), a highly stable, oligomeric carotenoprotein that fulfills multiple nutritional and defensive roles, including carotenoid stabilization and photoprotection, aposematic warning coloration, nutritional support to the embryo, and antinutritive resistance to digestion by predators. In some species of the Bridgesii and Flagellata clades, PV1 additionally displays lectin activity, which in Bridgesii causes intestinal alterations in rodents and exhibits antibacterial properties (Heras et al., 1998; Dreon et al., 2007; Ituarte et al., 2008, 2012, 2018; Dreon et al., 2014; Pasquevich et al., 2014, 2017, 2024; Brola et al., 2020, 2026). In addition to PV1, species from the Canaliculata clade express perivitellin-2 (PV2), an entero-neurotoxin that induces morphological changes in the rodent intestine upon ingestion and is lethal upon injection (Heras et al., 2008; Dreon et al., 2013; Giglio et al., 2020). Despite this growing body of knowledge, research is dominated by derived clades, particularly Canaliculata species, leaving biochemical defenses of basal clades largely unexplored (Pasquevich et al., 2024). This limits our ability to understand the evolutionary origins of Pomacea ’s unparallel defenses. The Flagellata clade, which includes P. sordida (Swainson, 1823), P. intermedia (Férussac in Quoy & Gaimard, 1825), P. patula (Reeve, 1856), and P. flagellata (Say, 1829) (Barbosa et al., 2022) is of particular interest because it represents the earliest diverging branch of the genus and may retain ancestral or transitional defensive states.

Here, we address this knowledge gap by characterizing the biochemical composition and egg defense system of Pomacea sordida, endemic to Rio de Janeiro, Brazil (Barbosa et al., 2022). We analyze egg macronutrient composition and investigate key defensive traits previously reported in derived clades, including protein resistance to digestion, protease inhibition, lectin activity, and oral and intraperitoneal toxicity to mice. By situating these findings in an evolutionary framework, we evaluate whether the complex biochemical defenses observed in derived clades emerged gradually or were already established early in the evolution of the genus.

Ethical statement

The Institutional Committee approved all animal procedures (protocol number P01-01-2016) for the Care and Use of Laboratory Animals (CICUAL) of the Faculty of Medicine, National University of La Plata (UNLP, Argentina). The experiments were conducted and adhered to the Guide for the Care and Use of Laboratory Animals (National Research Council 2010).

Animal collection and maintenance

Adult Pomacea sordida specimens were collected in the city of Rio de Janeiro, Brazil (S22°56’27.6” W 43°23’27.6”) whereas Pomacea canaliculata and P. scalaris were collected in the Palermo Lakes, Buenos Aires, Argentina (34°33’26.3”S 58°26’03.0”W) and kept in aquaria under laboratory conditions. Pomacea patula catemacensis eggs were obtained from stocks preserved from a previous study (Pasquevich et al., 2024) collected from Lake Catemaco, Veracruz, Mexico (18°24′14.75″N, 95°04′52.80″W). Egg clutches were collected within 24 hours of laying and stored at -20°C.

All egg masses contain embryos at or before the morula stage. Voucher specimens were deposited in the Mollusk Collection of the Instituto Oswaldo Cruz (CMIOC 14814). Female BALB/c mice, (13–16 g), were obtained from the Experimental Animal Facility, School of Veterinary Sciences, UNLP, and used for toxicity and intestinal histopathology assays. The same mouse strain employed in previous work was employed to facilitate comparisons among experiments.

Sample preparation

For egg composition analyses, 3 egg masses from different females were homogenized separately in 20 mM Tris-HCl buffer, pH 7.4, at a 1:3 w/v ratio. For functional analyses, three independent pools of 3 egg masses from different females were homogenized, and the egg soluble fraction was obtained according to Pasquevich et al. (2014). Briefly, homogenates were centrifuged at 10,000 xg and 4°C for 20min in an Avanti J-E centrifuge (Beckman), and the supernatant was centrifuged at 100,000 xg and 4°C for 50min in an L8–70 ultracentrifuge (Beckman). The resulting supernatant, the egg-soluble fraction defined as egg PVF, was stored at -70°C until use. Biochemical composition of the egg was performed from a homogenate of the total egg mass, as detailed below.

Biochemical composition of eggs

Dry mass and ashes

To determine the dry mass and ash content, pre-weighed egg masses (n=3) were sequentially heated at 100°C for 24h and at 550°C for 5h, and the products of each step were weighed.

Polysaccharides

Polysaccharide quantification was carried out using the phenol-sulfuric acid method in microplates, as described by Masuko et al. (2005). Briefly, in a 96-well flat-bottom plate, 50 µL of egg homogenate (n=3) was added, then 150 µL of sulfuric acid was added quickly, followed immediately by 30 µL of 5% phenol in water. The mixture was incubated in a thermo-heated bath at 90 °C, allowed to cool to room temperature, and the absorbance was measured at 490 nm in a plate reader (CLARIOstar Plus BMG LABTECH). The standard curve was performed with glucose (0.5, 1, 1.5, 2 µg/µL).

Proteins

Total egg protein was determined using the method of Markwell et al. (1978), with bovine serum albumin (BSA) (Sigma-Aldrich) as a standard. Starting with 200 uL of egg homogenate (n=3). Absorbance readings were collected on an Agilent 8453 UV/Vis diode array spectrophotometer (Agilent Technologies).

Lipids

Lipids were extracted from homogenates following the method of Bligh and Dyer, 1959). Egg homogenates (n=3) were mixed for 1min with methanol–chloroform–water (1:1:0.9, v:v:v), and. The lipid fraction was evaporated under a nitrogen atmosphere and determined gravimetrically on a microbalance Mettler M5.

Extraction and quantification of carotenoids

Carotenoids from P. sordida, P. patula, P. scalaris and P. canaliculata egg homogenates (n=3) were extracted using the acetone method used previously for egg carotenoproteins of Pomacea spp (Dreon et al., 2007; Ituarte et al., 2008; Pasquevich et al., 2014; Pavía et al., 2026). Absorbance spectrum was acquired using an Agilent 8453 UV-visible spectrophotometer. The total carotenoid content per gram of egg was calculated as previously (Pavía et al., 2026). The molar extinction coefficient of β-cryptoxanthin was used for P. sordida, P. patula, and P. scalaris, as β-cryptoxanthin is the predominant carotenoid in P. scalaris (Brola et al., 2026) and the absorption spectra of P. sordida and P. patula were nearly identical to that of P. scalaris, further supporting the use of the same extinction coefficient for the three species. In contrast, the molar extinction coefficient of astaxanthin was used for P. canaliculata, where astaxanthin is the principal carotenoid associated with the PV1 carotenoprotein (Brola et al., 2026).

Protein features

Total PVF proteins were quantified following the Bradford (1976) assay, using BSA as standard. The PVF protein profile was analyzed by Native and denaturing gel electrophoresis. Native PAGE was performed on a 4-20% gradient gel on a Mini-Protean III system (Bio-Rad) (15 μg total protein). Molecular mass standards were run in parallel. Gels were stained with Coomassie Brilliant Blue R-250. Band intensities were quantified with ImageJ (Schindelin et al., 2012). Density gradient ultracentrifugation in NaBr δ = 1.28g ml -1 of the PVF (n=3) was performed at 207,000 xg for 22h at 10°C in a Beckman L8M centrifuge. Fractions (200 μL) were sequentially collected from the top of the tube and analyzed at 280 nm. A tube containing 20 mM Tris-HCl was used as a blank for density calculations instead of the sample. The refractive index of the blank gradients was determined with a refractometer (Bausch and Lomb, New York, NY, USA) and converted to density using tabulated values (Orr et al., 1991).

Immunoblotting

To evaluate the presence of proteins cross-reactive with PcPV2 antibodies in the Flagellata clade, PVFs from P. sordida, P. patula, P. scalaris, and P. canaliculata (positive control) were analyzed with validated anti-PcPV2 sera (Dreon et al., 2002). PVFs (10 μg total protein) and a molecular weight marker (Dual Color, Bio-Rad) were separated by SDS-PAGE on 12% gels and transferred to nitrocellulose membranes (Amersham) in a Mini Transblot Cell (Bio-Rad) using 25 mM Tris-HCl, 192 mM glycine, 20% (v/v) methanol, and a pH 8.3 buffer. After blocking for 2h at room temperature with 3% (w/v) nonfat dry milk in PBS-Tween, the membranes were incubated overnight at 4°C with polyclonal antibodies against PcPV2 (1:3000 dilution) in 3% (w/v) nonfat dry milk in PBS-Tween. Specific antigens were detected after incubation for 2h at room temperature with a 1:5000 dilution of goat anti-rabbit IgG horseradish peroxidase conjugate (Bio-Rad, cat. no. 172-1019) in 3% (w/v) nonfat dry milk in PBS-Tween. Immunoreactivity was visualized by electrochemiluminescence on a Chemi-Doc MP Imaging System (Bio-Rad).

Affinity chromatography

For the identification of PV1 by mass spectrometry the PVF (n=3) was enriched for PV1 using affinity chromatography on a Hitrap™ NHS-activated HP column (GE Healthcare 71-7006–00 AX) prepared with anti-PpaPV1 antibodies (Pasquevich et al., 2024) according to the column’s instructions. Briefly, 10 mL of PVF from three independent egg clutches (n=3) was passed through the affinity column.

Identification of PV1 by mass spectrometry analysis

Proteins (50 μg) of three independent PVF samples were loaded in a 12% SDS polyacrylamide gel for electrophoresis. The gel was cut after 1cm of the sample had run into it. Then, the gel slices were subjected to in-gel digestion, following a modification of the method described by Shevchenko et al. (2006). Briefly, gel slices were destained with a solution of 50% methanol/50 mM NH 4 HCO 3, washed with MilliQ, dried with 100% ACN, and re-hydrated in 100 mM NH 4 HCO 3. Dithiothreitol (10 mM) and iodoacetamide (55 mM) were sequentially applied to reduce and alkylate the disulfide bonds of proteins. The gel was digested with sequencing-grade trypsin (Promega) in 50 mM NH4HCO3. Peptides were purified by the StageTips method described by Rappsilber et al. (2007) and dried in a vacuum concentrator. All peptides were separated by online liquid chromatography on an EASY-nLC II nano-LC coupled with an LTQ Orbitrap Velos ETD ion trap (Thermo Scientific) in a data-dependent acquisition (DDA) mode. Data was analyzed by MaxQuant software (v2.6.7.0) using a database made with previously published PV1 sequences and visualized by Perseus (v2.1.3.0).

Perivitellin fluid spectrophotometric analysis

The UV-visible absorption spectra of PVF (n=3) were recorded between 350 and 650 nm at room temperature using an Agilent 8453 UV-visible spectrophotometer.

In vitro resistance to digestion

The resistance of P. sordida PVF to protease degradation was studied in vitro as previously described (Dreon et al., 2010; Pasquevich et al., 2017). For simulated gastrointestinal digestion, PVF protein (150 μg) in phosphate buffer was first incubated with pepsin at an enzyme: substrate ratio of 1:20 (w/w) at 37°C in simulated gastric fluid (SGF) containing 150 mM NaCl (pH 2.5). BSA was used as positive control. Aliquots of the sample and control were taken at 0 and 120min of incubation. The pH was raised to 8.5 with 150 mM Tris-Cl buffer to stop the reaction and samples stored at -20°C until processing. The remaining solutions were then incubated with trypsin to simulate the intestinal phase at an enzyme: substrate ratio of 1:3 (w/w) at 37°C in SGF. The pH of the solution was adjusted with 0.1mol l -1 NaOH, and 0.25mol l -1 taurocholic acid and 0.15mol l -1 Tris-HCl buffer (pH 8.5) were added. Aliquots of the sample and control were taken as in the gastric phase and stored at -20°C. Finally, all samples were boiled at 100°C for 10min and analyzed on a 12% SDS-PAGE to verify protein integrity.

Proteinase K digestion was performed following Pasquevich et al. (2017). The PVF (1 mg/mL) was incubated with proteinase K (1, 10 and 100 µg/mL) in 50mM Tris/HCl buffer (pH 8.0) with 10 mM CaCl2 at 37°C for 30min. The digestion was stopped by boiling in SDS sample buffer, and the products were analyzed by SDS-PAGE as above.

Resistance to SDS denaturation

Resistance to sodium dodecyl sulphate (SDS)-induced denaturation serves to identify proteins whose native conformations are kinetically trapped in a specific conformation because of an unusually high-unfolding barrier that results in very slow unfolding rates. The resistance to SDS was assayed following the procedure. Briefly, PVF in Laemmli sample buffer (pH 6.8) containing 1% SDS was either boiled for 10min or unheated before its analysis by 4–20% SDS-PAGE. The gels were then stained with Coomassie blue.

Protease inhibition

To assay the protease inhibition ability of PVF proteins, proteases were preincubated with PVF (n=3) (in 0.02 M Tris/HCl buffer pH 7.4) for 5min at 20°C before measuring activity. Incubations of the enzyme in 0.02 M Tris/HCl buffer, pH 7.4, without PVF proteins were included as controls. Bovine pancreatic trypsin (Sigma-Adrich, #T9935) was assayed with 0.025 mM N-benzoyl-L-arginine ethyl ester (BAEE) in 0.067 M phosphate buffer, pH 7.6 at 37°C, at an enzyme: sample ratio of 1:4.35 w/w, measuring absorbance at 253 nm (Schwert and Takenaka, 1955). Subtilisin from B. licheniformis (Sigma-Aldrich #P5380) was assayed with 0.025 mM BAEE in 0.067 M phosphate buffer, pH 7.6 at 50°C at an enzyme: sample ratio of 1:1 w/w (Schwert and Takenaka, 1955). Bovine pancreatic α-chymotrypsin (Sigma-Aldrich #C3142) was assayed with 1.18 mM N-benzoyl-L-tyrosine ethyl ester (BTEE) in 0.08 M Tris/HCl buffer pH 7.8, 2 M CaCl2 at 25°C in an enzyme: sample ratio 1:3.17 w/w, measuring absorbance at 256 nm (Wirnt and Bergmeyer 1974). Inhibition of porcine pancreatic elastase (Sigma-Aldrich #E1250) was determined with N-succinyl-Ala-Ala-Ala-Ala p-nitroanilide 4.4 mM in 0.1 M Tris/HCl buffer pH 8.0 at 25°C at an enzyme: sample ratio of 1: 1.98w/w, measuring absorbance at 410 nm (Bieth et al., 1974). All measurements were performed in triplicate using an Agilent 8453 spectrophotometer (Agilent Technologies).

To test the inhibition ability of proteinase K, PVF (final concentration 1 µg/µL) was incubated in 50 mM Tris-HCl buffer, pH 8 and CaCl 2 (10 M) with proteinase K (enzyme: substrate ratio 1:16 w/w) for 5min at 37°C, then BSA (final concentration 0.6 µg/µL) was added to the vial. Positive and negative controls of BSA and PVF were performed in parallel. After 30min., aliquots were taken and analyzed by SDS-PAGE.

Hemagglutination assays

PVF lectin activity was tested using rabbit red blood cells (RBC) obtained from the animal husbandry facility of UNLP. Erythrocytes were prepared as indicated in Ituarte et al. (2012). Two-fold serial dilutions of PVF (n=3) in PBS (50 μL) starting at 1.4 ± 0.13 mg/mL were incubated with an equal volume of RBC at 2% (v/v) in PBS in U-shaped microtiter plates at 37°C for 2h. Results are expressed as the lowest protein concentration showing hemagglutinating activity visible to the naked eye.

Antimicrobial activity and bacterial agglutination

PVF was tested on cultures of Escherichia coli (BL-21) and Staphylococcus pseudointermedius grown on solid media. On each plate, 50 μL of culture was spread on LB/agar in triplicate, and 20min later, 10 μL of serial PVF dilutions (1/1 to 1/8 v/v) were placed on inoculated plates, and inhibition zones were recorded after 18h at 37 °C. A sterile buffer was used as a negative control and penicillin-streptomycin as a positive control.

Microbial agglutination assays were performed by incubating E.coli and Staphylococcus pseudointermedius (OD600=1.2) with 10 μL of PVF serial dilutions in 96-well plates in triplicate, overnight at room temperature. Aggregation was evaluated by visual inspection and microscopy.

Lethality assay

To assess whether P. sordida eggs contain neurotoxins as reported in other Pomacea species (Heras et al., 2008; Giglio et al., 2016) mice (n=10 per group) were randomly allocated and intraperitoneally injected with 200 μL of PBS (control) or PVF (2.68 μg/μL). The behavior of the mice was observed for 96h, and behavioral changes or neurological signs were recorded. Special attention was paid to signs already reported for Pomacea neurotoxins, including weakness, lethargy, tachypnea, hirsute fur, abduction of the hind limbs preventing them from supporting their weight, and spastic movements of the tail muscles. All behavioral observations were performed by an observer blinded to the treatment allocation.

Effect of PVF oral administration on the murine intestine.

The effect of a diet containing PVF on the small intestinal epithelium was evaluated using female BALB/c mice. Groups of 10 animals (5–7 weeks old, 12.9 ± 1g) were randomly assigned to 24h control, 48h control and 48h treated (n = 3 per treated group and n=2 per control group) and placed in a room conditioned with a 12:12 l:d cycle, 22 ± 1˚C, 45-60% relative humidity. Treated groups received 200 μL of PVF (2.68 μg/μL) in PBS via oral gavage using an animal feeding needle (Popper and Sons, Inc). Controls received PBS only. At the corresponding time points, animals were euthanized by gradual CO 2 inhalation from a compressed cylinder (displacement rate of 20–30% chamber volume/min) in a closed chamber, avoiding prefilled containers or dry ice. Death was ensured by prolonged exposure and confirmed by cardiac arrest and the absence of the palpebral reflex. The first segment of the small intestine was excised, rinsed 6 times with PBS to remove food and unbound proteins, and fixed in 10% neutral formalin for histological examination. Histology was performed following that described in Dreon et al. (2014). In short, 24 hours after formalin fixation, the intestine tissue samples were dehydrated with ethanol and stored until processed. Samples were then completely dehydrated in 100% ethanol and embedded in paraffin wax. Sections (5–7 µm) were stained for general morphology analysis with hematoxylin and eosin. Twenty-five villi and crypts from small intestine mucosae were selected at random and their length and width measured. This data was employed to calculate mucosal absorptive surface area following the method of Kisielinski et al. (2002). Histological sections were evaluated by an observer blinded to the treatment groups.

Statistical analysis

Statistical analyses were performed using GraphPad version 10.2.3. The Shapiro-Wilk test was used for normality analysis. An unpaired t-test with Welch’s correction was used to protease inhibition data. The Kruskal-Wallis test was used to compare intestinal microvilli measurements, and the Brown-Forsythe and Welch ANOVA tests were used to compare carotenoid levels.

Biochemical composition

Dry mass represents 17.74% ± 2.40% of the total egg mass, while ash represents 6.37% ± 1.08% (m/m) of the wet mass, i.e., 35.79% ± 1.20% (m/m) of the dry mass of the egg (dm).

The biochemical composition of P. sordida eggs was analyzed regarding the main energy sources: Eggs are mainly composed of carbohydrates (30.9 ± 4.9 mg polysaccharide/g egg; ~17,43% dm). Proteins are present in smaller amounts, 24.78 ± 5.4 mg protein/g egg, ~13,9% dm. While lipids represent the least abundant component (1.34 ± 0.15 mg lipids/g egg; ~0,76% dm). Carbohydrates are the most abundant energy reserve and are therefore likely to represent the principal energy source.

Carotenoid content in eggs of representative Pomacea

The fine structure of the carotenoid absorption spectrum in hexane is shown in Figure 2A. Whereas P. sordida, P. patula, and P. scalaris spectra are similar, the one of canaliculata clade differs. The egg carotenoid content increased progressively from the more basal species to the more derived ones, with values ​​of 1.33 ± 0.23, 3.85 ± 0.74, 7.68 ± 1.63, and 15.07 ± 4.01 µg carotenoids/g egg in P. sordida, P. patula, P. scalaris, and P. canaliculata, respectively (Figure 2B). Differences were significantly different among species except between P. scalaris and P. canaliculata, where a tendency for P. canaliculata to have more carotenoids is observed.

Protein profile of eggs and UV-visible spectrum of PVF

The qualitative protein profile, determined by electrophoresis and hydration densities of the PVF, showed multiple protein bands, but not the dominant high-molecular-weight protein typically found in other Pomacea species (Supplementary Figure S1). Likewise, proteins were distributed in a wide range of hydration densities in a NaBr density gradient, without the major peak characteristic of PV1 present in all species of Pomacea studied (Figure 3). Instead, a single broad peak is observed, composed of several proteins. In line with this, the UV-visible absorbance spectrum of P. sordida PVF displayed low absorbance in the visible range, with no prominent carotenoid-associated PV1 peak. This further suggests the absence of the major soluble carotenoprotein, distinguishing P. sordida from more derived species that rely on such proteins for antioxidant protection and egg pigmentation (Supplementary Figure S2).

Main perivitellins of Pomacea sordida

The PVF of P. sordida showed cross-reactivity with anti-PcPV2 antibodies (Supplementary Figure S3) indicating the presence of conserved PV2-like proteins in the basal Flagellata clade. Immunoreactive proteins containing both subunits, the 67 kDa MACPF domain subunit and the 31 kDa tachylectin domain subunit were found in P. sordida and P. patula but not in P. scalaris (Bridgesii clade). In both Flagellata species the tachylectin intensity showed weaker immunoreactivity than the MACPF-domain subunit, whereas both subunits displayed comparable immunoreactivity in P. canaliculata. Mass spectrometry of the PV1-enriched fraction of P. sordida identified five PV1 subunits with sequence similarity to homologous proteins from P. canaliculata, P. maculata, P. scalaris, and the basal ampullariid, Marisa cornuarietis (Supplementary Tables S1, S2).

Protein resistance to simulated digestion and SDS

The protease digestion assays showed the presence of PVF proteins resistant to digestion. Several remained intact after simulated gastrointestinal digestion and proteinase K treatment at 1 ug/ul. SDS-PAGE revealed protein bands ranging from ~22 to 66 kDa that resisted degradation (Figures 4A, B). On the other hand, in SDS resistance, fewer resistant proteins are shown compared to the other species studied (Supplementary Figure S4). This high stability is consistent with a potential antinutritional role, although functional assays (e.g., nutrient assimilation, feeding trials) are required to confirm this putative defensive function.

Protease inhibition

Protease inhibition assays showed that PVF significantly inhibits digestive proteases from animal, bacterial and fungal origins. Inhibition levels reached 86.9 ± 3.94% for trypsin, 78.0 ± 1.20% for α-chymotrypsin, 63.9 ± 2.20% for elastase and 40.7 ± 2.42% for Subtilisin A (Figure 4C). PVF also reduced proteinase K activity, as BSA was not completely digested (Figure 4D), highlighting the potential capacity of P. sordida PVF as a chemical defense against a wide range of proteases.

Lectin activity

Considering the lectin-like activity in eggs of other Pomacea species, we tested the agglutinating capacity of PVF from P. sordida. Positive hemagglutination using rabbit erythrocytes was observed up to a protein concentration of 1.36 ± 0.12 µg/µL, indicating the presence of active lectin-like proteins (Figures 5A–C). No bacterial growth inhibition was observed. However, the PVF also showed lectin-like agglutinating activity, capable of agglutinating gram-positive and gram-negative bacteria up to a concentration of 5.88 ± 2.88 µg/µL PVF (Figures 5D–I).

Egg toxicity

Intraperitoneal administration of P. sordida PVF (0.5 mg) caused 100% mortality in mice at 96h, preceded by the same neurological signs than previously reported for P. maculata PVF and purified PV2, such as weakness and lethargy, tachypnea, hirsute hair, and abduction of the hind limbs (Giglio et al., 2020b).

Oral administration and effects on the murine midgut

Oral administration of P. sordida PVF induced significant but transient morphological alterations in the mouse small intestine. Twenty-four hours after administration, a significant reduction in the absorptive surface was observed compared with controls, whereas by 48h the absorptive surface had returned to control values (Figure 6A). Histological examination 24h after administration revealed shorter and wider villi, prominent villus capillaries, edema of the lamina propria, and mild inflammatory infiltration by macrophages and lymphocytes (Figures 6B, C, E–G). In addition, the epithelium at the tip of some villi showed degenerative changes (Figures 6E–G). However, 48h after treatment, the intestine exhibited signs of recovery toward normal morphology, and the absorptive surface evolving toward control values (Figures 6D, H). The villi were somewhat longer and narrower, approaching the morphology observed in controls. Nevertheless, edema and prominent villus capillaries remained evident in some villi, indicating that tissue recovery was not yet complete.

Egg composition and energy source in Pomacea sordida mirrors that of other gastropods

As in most gastropods, apple snail eggs have little yolk. Instead, their PVF is the main source of nutrients for both the embryos and newly hatched snails (Heras et al., 1998; Hayes et al., 2015). P. sordida eggs contain carbohydrates, the main macromolecular component, consistent with that observed in most gastropods, including other Pomacea species (Livingstone and de Zwaan, 1983; Heras et al., 1998; Giglio et al., 2016). Likewise, the primary polysaccharide is likely galactogen, a sugar exclusive to gastropods, which is not only a source of energy but may also contribute to embryo defense. It cannot be digested by animals (except gastropods) or bacterial glycosidases, protecting against predation and microbial degradation (Giglio et al., 2016). This composition aligns with other gastropod eggs energy metabolism, which is based on carbohydrates (Livingstone and de Zwaan, 1983; Heras et al., 1998; Giglio et al., 2016). The second most abundant component in eggs was proteins, although present at lower levels compared to other Pomacea eggs (Heras et al., 1998; Giglio et al., 2016). Compared with other Pomacea species, P. sordida eggs contain relatively lower protein levels, whereas derived Canaliculata species invest proportionally more protein in the PVF, consistent with the expansion of protein-based defensive functions reported for this clade. On the other hand, lipids are present in very low concentrations, just like the other species studied (Heras et al., 1998; Giglio et al., 2016). Regardless of the clade, and beyond its nutritive role, the PVF of all Pomacea studied contains an array of defensive strategies that are unparalleled among animals, protecting against predators, pathogens, and physical stress (Dreon et al., 2006; Heras et al., 2007; Hayes et al., 2015). P. sordida was not an exception, as discussed below.

Insights into the early evolution of PV1 and egg pigmentation in apple snails

The shift from aquatic to aerial oviposition, which has seldom occurred in animals, has been considered a key feature for the diversification and spread of Pomacea species (Hayes et al., 2009a). In this regard, in Pomacea species examined so far, PV1 is the dominant egg perivitellin and forms a highly abundant carotenoid-binding carotenoprotein that provides nourishment while contributing to egg defense (Heras et al., 1998; Dreon et al., 2002; Ituarte et al., 2008; Brola et al., 2020; Pasquevich et al., 2024). Unexpectedly, P. sordida lacked this abundant pigmented carotenoid-binding oligomeric complex characteristic of other Pomacea species, despite expressing at least five PV1 subunits. This is, to our knowledge, the first carotenoid-free PV1 reported in Ampullariidae and likely represents an ancestral condition preceding the evolution of abundant PV1 carotenoprotein complexes and conspicuous egg coloration. Although P. sordida eggs contain the lowest carotenoid levels reported for Pomacea, these pigments remain in the insoluble egg fraction rather than associated with PV1. Together, these observations suggest that egg carotenoid accumulation and the evolution of the PV1 carotenoprotein would initially have been independent processes that became associated later during Pomacea evolution.

Interestingly, there is a carotenoid-binding PV1 in the closely related P. patula (Pasquevich et al., 2024), suggesting that the association between PV1 and carotenoids arose early during the evolutionary transition from aquatic to aerial egg deposition. The presence of PV1 subunits in Marisa sp., a basal ampullariid genus laying unpigmented eggs (Ip et al., 2019, 2020), would suggest that PV1 predates the acquisition of carotenoid binding capacity, supporting the view that carotenoid association would be a derived trait. Ongoing studies of PV1 sequence evolution will help clarify these evolutionary relationships.

Defenses by lectin-like activity

All Pomacea eggs so far reported exhibit lectin-like activity associated with anti-predation defenses (Brola et al., 2020; Giglio et al., 2020a; Brola et al., 2026) and protection against microorganisms (Ituarte et al., 2018). The activity is associated with PV1s in P. patula (Flagellata clade), P. scalaris and P. diffusa (Bridgesii clade) (Ituarte et al., 2012, 2018; Brola et al., 2020; Pasquevich et al., 2024). In these species, PV1-associated lectin activity differs phylogenetically in their affinity and specificity (Pasquevich et al., 2024). In contrast, the canaliculata clade PV1 does not show agglutinating activity (Pasquevich et al., 2017), and a mild lectin activity is associated with PV2 toxin in P. canaliculata and P. maculata) (Ituarte et al., 2012; Dreon et al., 2013, 2014; Giglio et al., 2016, 2018). Likewise, Pomacea sordida PVF also showed moderate agglutinating activity against both erythrocytes and bacteria, but further work is needed to identify to which protein is ascribed. The results of this study are consistent with the observation that lectin-like activity appears to be widespread across the genus and probably also the family, as it has also been reported in egg extracts of other ampullariid snails, namely Pila ovata (Olivier, 1804) and Pomacea urceus (Müller, 1774) (Uhlenbruck et al., 1973; Baldo and Uhlenbruck, 1974). Whether PV2-like proteins may contribute to the observed toxicity is a question for future investigation.

Defenses by low-digestibility proteins and protease inhibitors

Pomacea eggs have some proteins that withstand digestion by proteases, the most abundant being PV1s. In addition, the second most abundant protein, PV2 toxin in the Canaliculata clade also resists digestion, reaching biologically active and adversely affecting the gut absorption of nutrients (Heras et al., 2008; Giglio et al., 2020a). P. sordida, also has several proteins that resist protease digestion. This makes them a poor quality food source for predators, which has been interpreted as an effective antinutritive defense in other Pomacea (Pasquevich et al., 2017, 2024; Brola et al., 2020).

The low digestibility of egg proteins is further enhanced by the inclusion, in several Pomacea species, of protease inhibitors in their eggs. These inhibitors act on both animal and bacterial proteases in P. canaliculata and P. maculata (Giglio et al., 2016; Mu et al., 2017; Ituarte et al., 2019; Ip et al., 2020). Likewise, P. sordida eggs also display broad-spectrum protease inhibitors against animal, bacterial, and fungal proteases. Together, these traits are likely to reduce egg digestibility, acting as a deterrent that adds another layer to Pomacea’s defense system against predators and pathogens.

Defense by adversely affecting the digestive system

The PVF of P. sordida is toxic to mice when injected and, under the ecologically relevant route of oral administration, induces reversible alterations in intestinal morphology. Ingestion of P. sordida eggs, like those of P. canaliculata and P. scalaris, reversibly decreases the absorptive surface of the small intestine in mice. Although the effect is similar, the proteins involved appear to differ among species, involving various types of perivitellin lectins (Heras et al., 2008; Dreon et al., 2014; Giglio et al., 2018; Ituarte et al., 2018). In P. sordida, molecules responsible for the gut alteration remain to be identified, though lectin-like proteins such as PV1 subunits and PV2-like immunoreactive proteins are plausible candidates and deserve further investigation. According to Liu et al. (2013) these intestinal alterations, in particular the decrease in the absorption surface and the morphology of the villi, could cause food to pass through the intestine more quickly, suggesting a reduction in nutrient absorption. This, prevents proper digestion, making it an unattractive food source for predators.

Evolutionary implications

The characterization of the egg defense in Pomacea sordida, an early-diverging species according to the most accepted phylogenetic hypothesis (Hayes et al., 2009a) provide insights into the evolutionary trajectory of this genus. As summarized in Figure 7, several defensive traits are already present in Flagellata, indicating that the complex biochemical defense repertoire was established early in Pomacea evolution. This highlights the evolutionary pressures faced by these organisms as soon as they conquered the terrestrial egg laying, necessitating robust mechanisms to enhance reproductive success in a hostile environment. However, it is particularly intriguing that even within Flagellata, certain species lack specific adaptive traits, such as the ability to effectively accumulate and transport carotenoids through an abundant protein. This absence suggests a divergence in evolutionary pathways and ecological strategies among species, leading to varying levels of biochemical resource allocation that may significantly influence reproductive outcomes.

As shown in this study Pomacea eggs biochemical arsenal would improve the capacity to deter predation and enhance reproductive success. These strategies were likely subject to further modification and specialization in more derived clades. As the phylogenetic relationships within the Flagellata clade remain incompletely resolved (Barbosa et al., 2022), the reconstruction of the evolutionary trajectory within the clade will benefit from broader taxonomic sampling. It is worth noting that the Flagellata clade comprises only four species, two of which have been characterized regarding their biochemical defenses.

This study provides the first evidence that core biochemical defense repertoire of the Pomacea egg defense system were already present in an early-diverging species, whereas highly expressed carotenoid-binding PV1 complexes and increased egg pigmentation represent later, lineage-specific innovations.

Further comparative analyses across additional Flagellata species in addition to ecological studies of predator responses will advance in this hypothesis.