Regular exposure to low dose of diarrheic shellfish toxins: Implications for the seafood consumer

Abstract

Diarrhetic shellfish toxins (DSTs) are marine biotoxins produced by the dinoflagellates Dinophysis and Prorocentrum that cause gastrointestinal illness in humans worldwide. While European legislation avoids the risk of acute poisoning, the toxicity associated with repeated dietary exposure to DSTs in low quantities remains incompletely characterized. To address knowledge gaps regarding long-term exposure, we evaluated the toxicity of okadaic acid (OA), dinophysistoxin-1 (DTX1) and dinophysistoxin-2 (DTX2) administered by voluntary ingestion to male and female mice. Toxins were administered for 28 days at an experimentally calibrated dose (90 μg/kg bw) modelling a repeated human dietary exposure scenario derived from regulatory thresholds. Water consumption progressively decreased in treated females; the OA and DTX2 group also exhibited an increased urine protein-to-creatinine ratio, and minor histological renal changes. Repeated administration of DTX1 resulted in the most frequent diarrhea episodes and severe macroscopic damage to the digestive tract. Conversely, DTX2 produced the mildest clinical profile and lowest tissue retention. Furthermore, fecal toxin patterns differed by toxin: OA showed high initial excretion, DTX2 reached the highest fecal levels, and DTX1 was scarcely detected. Our findings indicate that repeated exposure to sublethal doses of DSTs by voluntary ingestion is associated with analogue- and sex-dependent persistent gastrointestinal alterations and subtle organ sensitivity, suggesting potential cumulative effects, which warrants further evaluation within risk assessment frameworks.

Graphical abstract

Highlights

• •Subchronic OA and DTX1 exposure induces qualitative gastric and renal changes.
• •Voluntary feeding in mice models 28-day dietary DSTs.
• •Differential accumulation of DSTs supports refining current oral TEFs.
• •Findings provide in vivo subchronic data to inform risk assessment.

1. Introduction

Harmful algal blooms (HABs) represent a global threat to human health and ecosystems, with their frequency and severity being intensified by climate change and human activities. Diarrhetic shellfish toxins (DSTs), principally okadaic acid (OA) and its analogues dinophysistoxin-1 (DTX1) and dinophysistoxin-2 (DTX2), are produced by marine dinoflagellates of the genera Dinophysis and Prorocentrum (Yasumoto et al., 1978). These compounds are involved in many HABs documented worldwide and are among the most prevalent marine toxins in Europe (Botana et al., 2021; Bresnan et al., 2021; Emery et al., 2021; Fernández et al., 2019; James et al., 2010; Visciano et al., 2016; Young et al., 2019). Fish are a marine biological group susceptible to DSTs through their diet (Corriere et al., 2020). These toxins can induce behavioural and morphological changes and increased mortality (Rolton et al., 2022; Rountos et al., 2019). However, information regarding the chronic toxicity of DSTs in fish in aquatic environments remains scarce (Yuan et al., 2024).

Filter-feeding shellfish exhibit tolerance to DSTs and are the main vectors of these toxins (Lv et al., 2025). Bivalves bioaccumulate the toxins and, when consumed by humans, cause a gastrointestinal syndrome known as Diarrhetic Shellfish Poisoning (DSP) (Fu et al., 2019; Reguera et al., 2014). Clinically, DSP is characterized by rapid-onset nausea, vomiting and profuse diarrhea, with recovery typically within a few days (Ito et al., 2002; Murata et al., 1982). Several in vivo studies have demonstrated the acute short-term toxicity of these compounds with effects on the liver, intestine and kidney in mammals (Abal et al., 2018; Costas et al., 2022; Fernández et al., 2014; Rodríguez-Santos et al., 2025). Although DSP is generally self-limiting and non-lethal, regulatory limits have been adopted to protect consumers (O'Mahony, 2018). The European Union (EU) has set a maximum level of 160 μg OA equivalents/kg shellfish meat to minimize the potential health risks (European Commission, 2004; European Food Safety Authority, 2008). If this limit is exceeded, shellfish harvesting areas are closed and the marketing of these animals is prohibited (European Food Safety Authority, 2009).

While current regulations provide a margin of safety against acute human poisoning, attention must be directed toward long-term dietary exposure to DSTs. Significant knowledge gaps persist regarding their chronic toxicity (Food and Agriculture Organization & World Health Organization, 2005), as well as their potential tumour promoting properties (Fujiki et al., 2018; Valdiglesias et al., 2011) and associations with cancer (Cordier et al., 2000; Jiménez-Cárcamo et al., 2020; López-Rodas et al., 2006). Chronic exposure to low doses of DSTs in seafood consumers is environmentally plausible because these toxins persist in water for extended periods and edible shellfish may contain toxin levels below the legal limit (Aron et al., 2022; Blanco et al., 2018, 2019; Fernández et al., 2019; Pizarro et al., 2009). Toxicological data based on single dose acute studies are insufficient to assess potential risks associated with repeated dietary intake of shellfish. Therefore, chronic exposure has recently become a focus of in vivo studies.

Regulatory agencies have explicitly called for subchronic and chronic comparative studies of different analogues using harmonized methodologies to refine Toxic Equivalency Factors (TEFs) and guideline toxicity values (World Health Organization & Food and Agriculture Organization, 2016). Since the EU legislation implemented the analytical control of marine toxins in molluscs by liquid chromatography and mass spectrometric protocols (LC-MS) (European Commission, 2011), the challenge is translating individual toxin concentrations into toxicity to determine whether regulatory limits have been exceeded. LC-MS analyses provide accurate quantitation of the toxin analogues (Klijnstra et al., 2021), but they have widely dissimilar toxic potency and do not equally contribute to the overall toxicity of the seafood, that is why appropriate TEFs are required (Botana et al., 2017). TEFs are defined as the toxicity ratio of a compound within a chemical group that shares the same mode of action as a reference compound. For DSTs, TEFs express the potency of each analogue relative to OA, facilitating dose-equivalent calculations in mixed toxin exposures (Botana et al., 2017; Van den Berg et al., 2006). The European Food Safety Authority (EFSA) Panel established TEF values of OA = 1, DTX1 = 1 and DTX2 = 0.6 based on in vitro and acute intraperitoneal data (European Food Safety Authority, 2008), that differ from those proposed with the results of oral toxicity (Abal et al., 2017, 2018). In addition, further studies to compare repeated oral dosing of toxins by voluntary feeding are required to better reflect human dietary intake and simulate the subchronic and chronic risk.

The intensification of HAB conditions due to climate change may increase the presence of low concentrations of DSTs in shellfish. Therefore, there is an urgent need to understand the health impacts associated with the regular exposure to low levels of these toxins. To address these gaps, we conducted a 28-day in vivo repeated-dose oral toxicity study with OA, DTX1 and DTX2 at an experimentally calibrated rodent exposure intended to model a high-intake human dietary scenario based on the European regulatory limit. To better simulate human exposure to DSTs, administration was by a voluntary ingestion procedure (Rodríguez-Santos et al., 2024). Daily clinical observations, body weight and food intake were monitored throughout the study. After necropsy, a comprehensive assessment of organ weights, haematology, serum chemistry and histopathology were performed. By comparing OA, DTX1 and DTX2 under identical conditions, this study aims to clarify their relative subchronic toxicity, addressing the data requirement established by FAO/WHO and CODEX expert panels regarding data needs in DSTs risk assessment (Food and Agriculture Organization and World Health Organization, 2026; World Health Organization & Food and Agriculture Organization, 2016).

2. Materials and methods

2.1. Toxins

The toxins OA (CRM-02-OA lot 22-001), DTX1 (CRM-00-DTX1 lot 21-001) and DTX2 (CRM-02-DTX2 lot 19-001) with purity >98% were provided from Laboratorio CIFGA S.A. (Lugo, Spain). Stock solutions were prepared in ethanol and stored at −80 °C. Certified reference materials (CRMs) from Laboratorio CIFGA S.A. were used for quantification.

2.2. Animals

Four-week-old Swiss (ICR background) male and female mice (mean weight 25 g and 20 g, respectively; individual variation ±20%) were purchased from the Centre for Experimental Biomedicine (CEBEGA) and housed in the Experimental Animal Facility (Code: AE-LU-002) of the University of Santiago de Compostela. Animals were maintained under standard conditions (22 ± 2 °C, 60-70% humidity, 12 h light/dark cycles) with water and pelleted food ad libitum. A total of 40 mice were randomly assigned to four experimental groups (n = 10 per group: n = 5 males and n = 5 females): OA-administered, DTX1-administered, DTX2-administered or control (vehicle).

2.3. In vivo experimental procedure

Procedures were carried out in accordance with the Organisation for Economic Cooperation and Development (OECD) Guideline No. 407, European (EU directive 2010/63/EU) and Spanish legislation (Real Decreto 53/2013, Decreto 296/2008) and were approved by the Institutional Animal Care Committee of the University of Santiago de Compostela (01/22/LU-002).

Toxin solutions were freshly prepared in 0.9% saline. Mice received a daily sublethal dose of toxin (90 μg/kg bw) via voluntary ingestion of a 100 mg piece of bread containing 30 μL of toxin solution. This experimental dose of 90 μg/kg bw was selected as an experimentally calibrated rodent exposure intended to model a high-intake human dietary scenario based on the European regulatory limit (160 μg OA eq/kg shellfish meat), as previously described in our acute voluntary ingestion model (Rodríguez-Santos et al., 2024). Specifically, assuming a standard large human portion of 400 g of bivalve molluscs containing DSTs at the legal threshold (total intake of 64 μg OA equivalents) for a 70 kg adult, the human exposure equals approximately 0.91 μg/kg bw. Applying a standard 100-fold uncertainty factor (a 10-fold factor for inter-species extrapolation and a 10-fold factor for human intra-species variability) yields an experimental target dose of approximately 90 μg/kg bw in mice, representing a conservative exposure paradigm to evaluate subchronic toxicological outcomes. Five days before the experiment, mice were habituated to bread ingestion as was previously described (Rodríguez-Santos et al., 2024).

During the 28-day repeated exposure study, mice received bread containing vehicle (control), OA, DTX1 or DTX2. Mice were separated into individual cages during dosing to verify complete ingestion before return to group cages with food and water ad libitum.

Clinical signs were recorded daily. Specific symptoms and diarrhea episodes were monitored for 2 h post-ingestion of DSTs. Diarrhea severity was evaluated with a stool consistency scale: 0 (normal), 1 (slightly loose), 2 (unformed loose), 3 (diarrhea), and 4 (persistent watery diarrhea with repeated episodes in less than 20 min). Non-specific clinical signs (orbital tightening, apathy, lordosis, piloerection, ear position, cyanosis, tremors, nose bulge, dyspnea, and hunched posture) were scored based on presence and severity, using a scale ranging from 0 to 2: 0 (absence), 1 (moderate or intermediate presence), and 2 (obvious presence).

Animals were individually placed in metabolic cages starting the first day of treatment (day 0) and weekly thereafter to record food and water intake, as well as feces and urine production.

2.4. Necropsy and samples collection

After 28 days, mice were euthanized by CO₂ inhalation. Mice were subjected to a detailed necropsy that included observation and collection of thoracic (heart and lungs), abdominal organs (liver, spleen, kidneys, stomach and intestines) and brain. Gastrointestinal toxicity was quantified using the gastrointestinal index (scale 0-6) based on macroscopic damage and fluid accumulation in stomach, and intestine, as previously described (Rodríguez-Santos et al., 2025).

The gastrointestinal tract was emptied. Intestinal contents were used for DST quantification by UHPLC-MS/MS and for the detection of proteins associated with inflammatory responses by Western Blot. All organs were weighed post-mortem and frozen at −20 °C.

2.5. Histopathology

To support the toxicological analysis, a qualitative histopathological evaluation was also conducted. After euthanasia, mice were perfused intracardially with saline solution followed by buffered 10% formalin. Samples of heart, lung, brain, spleen, pancreas, liver, kidney, stomach and small and large intestine were fixed in buffered 10% formalin (24 h, 4 °C), embedded in paraffin and sectioned at 3–4 μm. Sections were mounted onto silanized slides, dried overnight at 37 °C and stained with haematoxylin and eosin (H&E). Gastric mucosal thickness was quantified at ten different points per mouse using the software ImageJ (Fiji).

2.6. Toxin analysis in biological samples

Solid samples (0.1 g of homogenized organs, feces, and intestinal contents) were extracted three times with methanol following as previously described (Louzao et al., 2021), and recently applied for voluntary ingestion studies (Rodríguez-Santos et al., 2024). Supernatants were evaporated to dryness, reconstituted in 100 μL of methanol and proteins were precipitated by adding 40 μL of methanol and 10 μL of a 10% trichloroacetic acid. After adding 50 μL of acetonitrile, extracts were filtered (0.2 μm) and centrifuged (1400 × g, 10 min) prior to UPLC-MS/MS analysis.

For fluid matrices, urine (100 μL) was first homogenized with 40 μL of methanol, 10 μL of a 10% trichloroacetic acid solution were added and the mixture was vortexed again to precipitate the proteins. Finally, 50 μL of acetonitrile were added; samples were then filtered (0.2 μm) and centrifuged (14500 × g, 10 min) and analyzed by UPLC-MS/MS.

Blood (300 μL) was extracted with 75% methanol (1200 μL), vortexed for 1 min, centrifuged (1400 × g, 10 min) and purified by filtration using Amicon Ultra 0.5 mL filters (Ultracel 3K, regenerated cellulose membrane 3 kDa NMWL; 11334 × g, 30 min). The ultrafiltrate was evaporated, reconstituted in 300 μL of methanol, filtered (0.2 μm) and centrifuged (14500 × g, 10 min) and analyzed by UPLC-MS/MS.

DSTs (OA, DTX1 and DTX2) were analyzed using an Acquity UPLC BEH C18 column (Waters, USA), coupled to a Xevo TQ MS mass spectrometer (Waters, UK), operating in multiple reaction monitoring (MRM) mode with electrospray ionization (ESI-). Toxins were quantified against CRMs provided by Laboratorio CIFGA S.A.

2.7. Urine analysis

Urinalysis was performed weekly (weeks 0–4) using 250 μL samples. Urinary specific gravity was determined via refractometer. Chemical parameters (pH and protein) were measured using IDEXX UA dipsticks (USA). The pH pad allowed readings within 5 and 9. The protein parameter on the dipstick was evaluated to screen for proteinuria. Results were recorded according to the semi-quantitative colorimetric scale provided by the manufacturer, ranging from negative (−) to markedly proteinuric samples (3+). The grading scale corresponded to the following approximate concentrations: trace, 1+ (30 mg/dL), 2+ (100 mg/dL) and 3+ (500 mg/dL).

Urinary sediment was examined in all samples, and those showing inactive sediment with proteinuria ≥ 30 mg/dL were further analyzed for the urine protein-to-creatinine ratio. The analysis was performed by dry-slide biochemistry (Catalyst One Analyzer, IDEXX, USA).

2.8. Western Blot analysis

Cyclophilin A (CypA) and cyclophilin C (CypC) were qualitatively evaluated in intestinal contents as inflammation biomarkers. Intestinal contents were weighed and diluted in PBS with protease/phosphatase inhibitors (1:10 w/v), homogenized by vortexing and centrifuged (3100 rpm, 30 min, 4 °C). Supernatants were filtered (0.22 μm) and further processed using a centrifugal ultrafiltration device (Amicon Ultra, Sigma-Aldrich). Two molecular weight cut-offs were used sequentially—50 kDa and 10 kDa— to isolate the intermediate molecular weight fraction.

Proteins were quantified using a Direct Detect system (Merck, Germany). Because intestinal luminal content lacks a stable constitutive housekeeping protein (such as β-actin or GAPDH), sample loading was strictly normalized by loading an equal amount of total protein (20 μg) per lane across all samples. Diluted proteins (20 μg in 30 μL of PBS) were separated on 4–20% polyacrylamide and transferred onto PVDF membranes using a Trans-Blot semi dry system. Molecular weights were determined using the protein size Precision Plus Protein Standard Kaleidoscope marker (Bio-Rad, Spain). Membranes were blocked (0.5% BSA) and incubated with anti-PPIA (CypA, 1:1000) or anti-PPIC (CypC, 1:1000) antibodies (Elabscience, lot numbers DM07TFV25480 and DL05ZR885111 respectively). Incubations were carried out using SNAP i.d. protein detection system (Merck). Bands were analyzed using the Supersignal West Pico/Femto substrates. Analysis of protein bands was performed using the Diversity GeneSnap software (Syngene, UK). To ensure technical and biological robustness, each sample was analyzed in duplicate. The results are representative of six biological replicates per treatment group (n = 3 per sex), showing consistent protein expression patterns across all analyzed samples.

2.9. Statistical analysis

Data were analyzed using R Studio (v. 4.3.1) and GraphPad Prism (v. 5.0). Weekly Area Under the Curve (AUC) for clinical symptoms and stool consistency (Week 1: days 0–6; Week 2: days 7–13; Week 3: days 14–20; Week 4: days 21–27) was calculated by the linear trapezoidal rule using DescTools package (v. 0.99.50) to integrate daily observations into a single robust value.

Physiological parameters measured repeatedly over time (water and food intake, fecal and urine output, and weekly fecal toxin excretion), were evaluated using three-way repeated measures Linear Mixed-effects Models (LMMs; lme4 v. 1.3-31 and car v. 3.1-3 packages) to assess the influence of Toxin Treatment (Control, OA, DTX1, DTX2), Sex, and Week (0–4) as fixed factors, including all way interactions, and “Mouse ID” as a random effect. Significant interactions involving sex were evaluated using sex-specific two-way LMMs. Single time point parameters (such as relative organ weights and final tissue toxin concentrations) were assessed via two-way ANOVA with Treatment and Sex as factors.

Post-hoc pairwise comparisons were conducted using Dunnett's against controls and Tukey's HSD test to identify differences among all treatment groups, sexes, and time points (emmeans v. 1.8.4-1 package), adjusting p-values for multiple comparisons. Significance was set at p < 0.05.

For urinary parameters, qualitative results were quantified. 'Negative' protein levels were assigned a value of 0 mg/dL and 'trace' levels 15 mg/dL. Where protein levels were negative and urine protein-to-creatinine ratio could not be calculated, a value of 0 was assigned to establish a non-proteinuric baseline.

3. Results

3.1. Clinical signs and specific symptoms of toxicity

Repeated administration of DSTs by voluntary feeding over 28 days resulted in a range of persistent non-specific clinical signs and gastrointestinal alterations, with distinct profiles depending on the toxin and sex-dimorphic response. Detailed statistical analyses of cumulative toxicity, including non-specific symptom frequency and severity as well as feces scores are detailed in Table S.3, providing specific inter-treatment and inter-sex comparisons.

Regarding the number of non-specific symptoms, DTX1 administration produced the greatest cumulative effect in mice (Fig. 1, top). This result was statistically significant in males during weeks 2 and 3 (p < 0.01 and p < 0.05, respectively), while females exhibited consistent higher symptoms count relative to controls throughout the first three weeks. Similarly, males administered with OA showed a significantly higher number of non-specific symptoms than the control group during the same period. A sex-dimorphic response was observed in week 2, with males displaying more symptoms than females after OA administration (†p = 0.016). For DTX2, the cumulative number of symptoms only reached statistical significance during week 4 in both males (p = 0.009) and females (p = 0.020).

Fig. 1.

Analysis of cumulative non-specific symptoms in mice during the 28-day study. Data are expressed as the AUC for the total number of symptoms (top) and clinical score (bottom) in male and female mice for each toxin group along the time. Results are presented as Mean ± SEM for each weekly interval. Statistical significance compared to the control group is indicated by (*p < 0.05, **p < 0.01, ***p < 0.001) according to Dunnett's post-hoc test. Significant differences between males and females within the same treatment group according to Mann-Whitney U post-hoc test are indicated by the dagger symbol (†p < 0.05).

The clinical score followed a similar pattern (Fig. 1, bottom). DTX1 effects were most pronounced in males during week 2 (p = 0.0008), with cumulative clinical scores significant in both sexes compared to controls throughout weeks 1–3. In the OA group, the cumulative clinical score was significant versus controls in males (weeks 1 and 3), and in females (weeks 1 and 4). A significant sex difference was noticed again in week 2 (†p = 0.021). DTX2 showed significant differences from the control only during the final week. The daily progression of both number and score of non-specific symptoms in male and female mice for each toxin group is detailed in Figure S.1.

In contrast to non-specific symptoms, the fecal consistency alterations were more prominent in females (Fig. 2). DTX1 produced the highest cumulative fecal scores, which were significant in females compared to controls from weeks 2 to 4, whereas in males, significance was reached only in week 4 (p = 0.0025). In the OA group, the diarrheic impact was significant in females versus control in week 2 but maintained a stable trend of soft stools throughout the study. Conversely, DTX2 did not induce any significant differences in cumulative fecal scores in either sex compared to controls. Daily progression of diarrhea severity is shown in Figure S.2.

Fig. 2.

Analysis of cumulative feces score during the 28-day study. Data represents the weekly AUC derived from fecal scores in female and male mice administered with OA, DTX1, or DTX2. Results are expressed as Mean ± SEM. Asterisks indicate statistical significance compared to the control group (*p < 0.05, **p < 0.01).

3.2. Physiological parameters and gastrointestinal findings

Fig. 3 illustrates the progression of relative body weight in male and female mice over the 28-day exposure period. No significant differences in weight gain were observed among the different treatment groups in either sex. Only male mice treated with DTX2 exhibited isolated statistically significant difference compared to the control group on day 27 (p < 0.05). However, this difference did not follow a consistent pattern over time.

Fig. 3.

Changes in body weight (%) of male (A) and female (B) mice that received OA, DTX1 or DTX2 for 28 days. Data are expressed as Mean ± SEM. * (p < 0.05).

Food consumption, feces production, water intake and urine production are illustrated in Fig. 4(A–D). Statistical parameters, including Mean Difference, 95% CI, t-ratio and p-values are detailed in Tables S.4-S.7. Regarding food consumption (Table S4), linear mixed-effects model analysis identified a significant main effect of the toxin (p = 0.0193), but no significant effects were found for sex (p = 0.69) or week of exposure (p = 0.52). While male mice showed a main effect of the toxin (p = 0.044) and females showed a slight downward trend under DTX1 treatment, post-hoc Dunnett's tests revealed no significant differences for any specific toxin compared to the control at any individual time point. In contrast, female food intake remained largely stable, only with a significant effect related to the week of exposure (p = 0.033). Similarly, no significant changes in fecal weight were detected for any toxin group or sex throughout the study (Table S.5).

Fig. 4.

Food consumption (A), water intake (C), feces (B) and urine (D) output in mice exposed to OA, DTX1 or DTX2 at 90 μg/kg recorded over a 28-day period. Values expressed as mean ± SEM for each group. Measurements were taken once a week throughout the study (day 0, 7, 14, 21 and 28). Statistically significant differences (p < 0.05) compared to controls are represented with *.

Regarding water intake, male mice showed no differences across all groups, whereas females exhibited a reduction from day 14 onwards. Specifically, OA-treated females consumed significantly less water during weeks 2, 3 and 4; similarly, significant reductions were observed in DTX1 and DTX2 female groups during week 3. Furthermore, a significant difference was observed between the OA and DTX2 groups in females in week 4 (p = 0.0206; Table S.6). Concerning urine output, no significant differences were found in male mice, although a trend towards lower urine weight was recorded in toxin-treated females, likely correlating with reduced water intake. However, this decrease did not reach statistical significance relative to the control group or between toxins. Detailed p-values for all comparisons are provided in Tables S6-S.7.

Urinary pH remained stable throughout the study, with values within the physiological range (Table S.1). In toxin-treated females, urinary specific gravity values showed a minor decrease compared to controls (Table S.2). Regarding urine protein-to-creatinine ratio, males displayed higher values than females, although no significant differences were detected. In female mice, OA- and DTX2-treated groups showed significantly higher urine protein-to-creatinine ratio values compared with controls (Fig. S.3).

At 28 days, macroscopic signs showed that DTX1 induced the greatest gastrointestinal disturbances, particularly in the males (Fig. 5). For OA and DTX2, the gastrointestinal index was lower in both sexes, with mean values close to 1. Post-hoc comparisons showed no significant differences between DST-treated groups and controls in males (p = 0.0541) or females (p = 0.2820).

Fig. 5.

Gastrointestinal index of mice administered with OA, DTX1 or DTX2 (90 μg DST/kg bw) versus control during 28-day exposure period, represented as mean ± SEM. Statistical comparisons between toxin-treated groups were performed using two-way ANOVA (assessing Treatment × Sex) followed by Dunnett post-hoc test against controls. No significant differences were detected between toxins and control within each sex (p > 0.05). Exact p-values were p = 0.0541 for males and p = 0.2820 for females. Control animals exhibited a GI score of 0 (zero variance).

Relative organ weights were summarized in Table 1. Female control mice displayed higher relative organ weights than males. Male mice administered with OA for 28 days exhibited a significant increase in relative heart weight compared to the control group (p < 0.05). However, upward trends in relative spleen weight for OA and DTX2, as well as in stomach and small intestine weights for DTX1-treated females, did not reach statistical significance.

Table 1.

Organ to body weight ratio (expressed as % of body weight) of mice recorded 28 days after treatment with 90 μg DST/kg bw (OA, DTX1 or DTX2). Data are expressed as mean ± SEM. Statistical significance was determined by Two-way ANOVA followed by Tukey's HSD post-hoc test. Asterisks indicate significant differences compared to the respective control group (p < 0.05).

	Sex	Control	90 μg/kg bw
			OA	DTX1	DTX2
Heart	M	0.51 ± 0.03	0.61 ± 0.03*	0.52 ± 0.03	0.60 ± 0.03
	F	0.54 ± 0.02	0.58 ± 0.05	0.63 ± 0.02	0.56 ± 0.03
Lungs	M	0.67 ± 0.04	0.59 ± 0.01	0.62 ± 0.05	0.64 ± 0.04
	F	0.77 ± 0.03	0.72 ± 0.10	0.85 ± 0.03	0.82 ± 0.03
Liver	M	4.53 ± 0.23	4.78 ± 0.02	4.77 ± 0.13	5.33 ± 0.24
	F	4.83 ± 0.15	5.20 ± 0.35	5.09 ± 0.11	4.56 ± 0.12
Stomach	M	1.01 ± 0.03	0.95 ± 0.06	1.07 ± 0.09	0.85 ± 0.04
	F	1.19 ± 0.02	1.16 ± 0.08	1.32 ± 0.05	1.10 ± 0.07
Spleen	M	0.26 ± 0.01	0.27 ± 0.03	0.22 ± 0.02	0.27 ± 0.06
	F	0.35 ± 0.02	0.43 ± 0.02	0.34 ± 0.04	0.41 ± 0.04
Small intestine	M	4.15 ± 0.10	3.22 ± 0.48	4.37 ± 0.26	3.90 ± 0.04
	F	4.50 ± 0.20	4.14 ± 0.22	4.83 ± 0.02	4.15 ± 0.01
Large intestine	M	1.39 ± 0.07	1.21 ± 0.10	1.37 ± 0.11	1.26 ± 0.06
	F	1.58 ± 0.11	1.43 ± 0.13	1.81 ± 0.21	1.52 ± 0.06
Kidneys	M	2.08 ± 0.14	2.12 ± 0.09	1.95 ± 0.15	1.88 ± 0.08
	F	1.58 ± 0.06	1.53 ± 0.12	1.65 ± 0.07	1.68 ± 0.04
Brain	M	1.27 ± 0.05	1.06 ± 0.06	1.34 ± 0.09	1.22 ± 0.03
	F	1.64 ± 0.09	1.35 ± 0.10	1.61 ± 0.04	1.63 ± 0.05

3.3. Toxin excretion and accumulation

After 28 days, DSTs were recovered in higher amounts in feces than in the intestinal contents. Fecal levels varied by toxin, sex, and time (Fig. 6A). Males administered with OA exhibited the highest initial excretion followed by a gradual decline over the four-week period, while females showed stable concentrations. Post-hoc analysis identified week 2 as a critical divergence point for differential excretion. During this period, DTX1 was minimally detected in the feces of both sexes compared to other toxins. In males, DTX2 reached the highest fecal concentrations overall, being significantly higher than both DTX1 and OA. In females, DTX1 levels were significantly lower than those of both OA and DTX2. Statistical values of fecal toxin excretion are shown in Table S.9.

Fig. 6.

DST concentration in feces collected weekly (A) and organs (B) after repeated 28-day voluntary consumption, represented as Mean ± SEM. (A) The statistical significance of toxin fecal excretion was determined using a Linear Mixed-effects Model (Treatment × Time × Sex with Mouse ID as a random factor), followed by a Tukey's post-hoc test (p < 0.05). Different letters (a, b) indicate significant differences between treatment groups within the same week and sex. (B) DST concentrations in peripheral organs and gastrointestinal tissues at day 28. Toxin levels remained near or below the limit of quantification (LOQ) in most analyzed tissues, with no statistically significant differences observed between toxins (p > 0.05, two-way ANOVA).

The highest DSTs levels were found in gastrointestinal tissues on day 28 (Fig. 6B), predominantly in the small and large intestines. In OA and DTX1 groups, the small intestine had higher toxin levels than the large intestine, but it was in the contents of the large intestine where more toxin was detected. Also, females generally showed higher toxin levels than males. Nevertheless, mice administered with DTX2 displayed comparable tissue retention profiles in both sexes. Low tissue concentrations of OA and DTX2 were found in the liver and the stomach of male and female mice while DTX1 was only detected in the stomach of female mice but not in the liver.

Interestingly, the pattern of high fecal excretion but low organ retention observed with DTX2 contrasts with DTX1, which exhibited lower fecal levels yet proportionally greater intestinal retention. The relatively low fecal excretion of DTX1 did not correlate with increased accumulation in peripheral organs (for instance heart, lungs or brain) where toxin remained below the limit of quantification (<LOQ).

3.4. Evaluation of intestinal cyclophilins

Fig. 7 illustrates the qualitative protein profiles of CypA and CypC detected in intestinal contents. A distinct band pattern for CypA was observed in the large intestine contents of mice that received OA and DTX2, and in males administered with DTX1. Notably, detectable CypA and CypC bands were exclusively observed in the small intestinal contents of mice treated with DTX1. In contrast, CypC was not detected in the intestinal contents of animals administered with DTX2 across all replicates.

Fig. 7.

Western Blot analysis of CypA (left) and CypC (right) in mice intestinal contents after 28-day exposure to OA, DTX1 or DTX2. Representative blots showing 1: Small intestine content, control mice. 2: Large intestine content, control mice. 3: Small intestine content, mice administered with DSTs. 4: Large intestine content, mice administered with DSTs.

3.5. Histopathological analysis

Representative liver sections (Fig. 8) showed mild, qualitative histological alterations in mice administered with DTX1 (E–F), particularly in females, characterized by dilated sinusoids and erythrocyte extravasation without leukocyte infiltration or necrosis. Hepatocyte morphology remained mostly intact. Liver sections from mice that received OA and DTX2 (C–D and G–H) were similar to the controls (A–B), which exhibited normal liver architecture, characterized by well-organized hepatic cords and sinusoids, with no evident signs of damage or inflammation.

Fig. 8.

Representative light micrographs of H&E staining of liver sections in samples from male (left) and female mice (right). Controls (A-B) and mice administered with OA (C-D), DTX1 (E-F) and DTX2 (G-H). Scale bars: 50 μm. Note dilated sinusoids (black arrows) and erythrocyte extravasation (red arrows).

Qualitative evaluation of H&E-stained kidney sections revealed that daily DSTs administration for 28 days caused minor, localized alterations in renal architecture compared to controls (Fig. 9). These morphological findings are most notable in representative samples from mice that received OA or DTX1 (C, E, F) and include vascular congestion (red arrows) and tubular damage, as well as increased Bowman's space (black arrows). These alterations were not observed in control animals or in DTX2-treated males (A, B, G), while DTX2-treated females only exhibited mild, isolated vascular congestion (H) without glomerular changes.

Fig. 9.

Representative H&E staining of kidney sections in samples from male (left) and female mice (right). Control (A-B) and mice administered with OA (C-D), DTX1 (E-F) and DTX2 (G-H). Scale bars: 50 μm. Red arrows denote vascular congestion and black arrows an increased Bowman's space.

At gastric level, Fig. 10 shows the mucosal thickness of the stomach in male and female mice after 28 days of exposure to DSTs. A significant reduction in mucosal thickness was observed in both sexes administered with OA and DTX1 compared to controls. In the case of DTX2, this reduction was only significant in male mice. Dunnett's Multiple Comparisons of gastric mucosal thickness are shown in Table S.8.

Fig. 10.

Stomach from controls or mice after 28-day exposure to DSTs. Mucosal thickness in the stomach of male and female mice expressed as Mean ± SEM (graph above). Significant differences compared to the control group are represented with *** (p < 0.0001). Detailed statistical comparisons, including p-values and 95% confidence intervals, are provided in Table S8. Representative H&E staining of stomach tissue sections (images down) Scale bars: 100 μm. A-D: Male mice. E-H: Female mice. A,E: Control; B,F: OA exposed mice; C,G: DTX1 exposed mice; D,H: DTX2 exposed mice.

4. Discussion

This study provides a comparative assessment of the 28-day repeated oral exposure to DSTs administered via voluntary ingestion in male and female Swiss mice. Regulatory agencies have emphasized that long-term effects of these DSTs remain insufficiently characterized, preventing the establishment of a tolerable daily intake (Botana, 2012; World Health Organization & Food and Agriculture Organization, 2016). Addressing this gap, we selected an experimental dose of 90 μg/kg bw, calibrated to model a high-intake human dietary scenario derived from the European regulatory limit of 160 μg OA eq/kg shellfish meat (European Food Safety Authority, 2008). We previously established a voluntary ingestion model using bread to deliver DSTs (Rodríguez-Santos et al., 2024, 2025). This novel, non-invasive and welfare-friendly experimental approach was selected in this work to evaluate the subchronic effects of DSTs. This refined, voluntary feeding method mimics natural human exposure while ensuring that the observed pathological responses are toxicologically driven, rather than artifacts of the administration procedure (Landes et al., 2026).

The observation of persistent non-specific clinical signs and gastrointestinal effects including diarrhea, indicates cumulative toxicity, with different profiles depending on the toxin and sex. A higher number of symptoms was observed in males exposed to all three toxins, reaching statistical significance in mice administered with OA and DTX1. Furthermore, DTX1 produced the most severe clinical picture, characterized by the highest frequency and intensity of clinical signs and diarrhea. This finding is consistent with acute oral toxicity studies in mice, which identified DTX1 as more lethal than OA (Abal et al., 2018), but contrast with recent reports describing similar intestinal effects for both toxins (Park et al., 2023). These discrepancies may be attributable to the stress related to the repeated gavage administration, or differences in clinical scoring systems in previous studies (Botana, 2012; Landes et al., 2026).

Diarrhea was observed through the first week after repeated DTX1 and OA administration indicating cumulative toxin effects. For both toxins, the diarrheic impact became significant only in females from the second week onward, revealing greater female susceptibility to the specific and subchronic gastrointestinal effects of OA and DTX1. The rapid onset of diarrhea is consistent with previous reports of acute gastrointestinal disturbances following DST exposure (Ito et al., 2002; Louzao et al., 2021; Rodríguez-Santos et al., 2025; Tubaro et al., 2003). The sustained diarrheic stools observed for DTX1 from day 7 to 28 further supports the higher subchronic toxicity of this analogue as was previously suggested by acute studies (Abal et al., 2018; Suzuki and Okada, 2018). In contrast, the absence of significant alterations in fecal consistency in the DTX2 group throughout the study reinforces its profile as the mildest analogue with minimal impact on intestinal water secretion (Abal et al., 2018; Aune et al., 2012; Rodríguez-Santos et al., 2024; Tubaro et al., 2003).

Unlike acute high-dose studies, repeated exposure to the subchronic dose of 90 μg/kg had no substantial impact on body weight progression or overall food intake in either sex (Abal et al., 2018; Rodríguez-Santos et al., 2024; Sosa et al., 2022). While an initial weight loss that disappears after repeated high-dose OA has been reported previously (Park et al., 2023), our results show a more stable physiological profile. Notably, although the statistical analysis of weekly food consumption in males revealed a significant main effect of DSTs, this did not translate into specific differences during post-hoc comparisons. This suggests that, while DSTs exposure may have a subtle, systemic effect on male feeding behaviour at this dose, the magnitude of the effect was insufficient to reach statistical significance at individual time points. By contrast, female food consumption was unaffected by DSTs, which highlights another aspect of the sex-specific responses observed throughout this study. Regarding water intake, a clear sexual dimorphism was observed. No changes were detected in males, whereas a reduction in water consumption was observed in females from day 14 onwards, particularly in mice receiving OA. This physiological stress is further supported by a trend toward lower urine weight in treated females, higher urine protein-to-creatinine ratio values in females that received OA and lower urinary specific gravity compared to controls, suggesting sex-dependent renal vulnerability. While these functional trends and minor observational kidney variations, such as vascular congestion and increased Bowman's space, suggest potential renal stress, the evidence is qualitative. Rather than demonstrating definitive tubular dysfunction, these findings suggest renal sensitivity and require further detailed quantitative histological investigations.

The inability to concentrate urine despite lower fluid intake may reflect subtle functional perturbations in renal handling rather than established tubular pathology. This pattern suggests that DSTs might elicit a mild, sex-dependent tissue sensitivity affecting fluid balance mechanisms, with a more pronounced response in females. The histopathological kidney lesions observed after 28 days of voluntary DSTs administration further reinforce the presence of minor renal alterations. Repeated exposure to OA and DTX1 led to architectural changes, including vascular congestion, tubular damage and an increase in Bowman's space, whereas DTX2-treated mice exhibited an essentially preserved renal morphology with only mild, isolated vascular congestion in females. Similar morphological changes, including glomerular retraction, have been described in subchronic studies of other marine toxins, suggesting that DSTs may similarly disrupt the glomerular filtration barrier (Raposo-García et al., 2022; Sosa et al., 2022). In contrast, no histopathological alterations were reported after 7-day oral gavage administration of OA and DTX1, although increases in kidney injury biomarkers were detected (Park et al., 2023). In the liver, daily DTX1 exposure caused mild vascular alterations, including sinusoidal dilatation and erythrocyte extravasation. These results suggest that repeated exposure to OA and DTX1 at an experimentally calibrated dose derived from regulatory thresholds ultimately leads to qualitative renal histological alterations and subtle variations in the hepatic microvasculature. This is consistent with trends observed in acute studies at higher doses (Abal et al., 2018; Ito et al., 2002; Sosa et al., 2022), but not detected in 7-day studies with low doses, where cumulative toxic stress may not yet have reached critical thresholds (Park et al., 2023).

Repeated voluntary administration of DSTs for 28 days was associated with persistent macroscopic and microscopic alterations throughout the digestive tract, demonstrating local gastrointestinal sensitivity under repeated exposure. Prolonged contact between DTX1 and OA and the gastrointestinal mucosa resulted in localised damage, organ dilatation and increased fluid accumulation contributing to the persistent diarrhea. These findings are consistent with previous acute studies using higher DSTs doses (Aune et al., 2012; Costas et al., 2022; Rodríguez-Santos et al., 2025; Sosa et al., 2022; Terao et al., 1986; Tubaro et al., 2003). Microscopically, gastric alterations associated with hypersecretion (Ito et al., 2002; Reale et al., 2019; Wang et al., 2012), were evaluated by measuring gastric mucosal thickness. A significant reduction in gastric mucosal thickness was detected in mice receiving OA and DTX1, and in males administered with DTX2. This mucosal thinning may reflect chronic epithelial alteration as previously described in rats (Huang et al., 2022).

Differences among DSTs are further supported by their different toxicokinetic profiles and physiological effects following repeated exposure. DSTs accumulate along the digestive tract and are primarily eliminated via feces (Abal et al., 2018; Suzuki and Okada, 2018; Twiner et al., 2016). Although gastric toxin concentrations remained low, this observation aligns with previous reports indicating peak DSTs levels between 2 and 6 h post-ingestion, limiting long-term accumulation in this organ (Abal et al., 2018; Suzuki and Okada, 2018). The highest toxin concentrations were found in the small and large intestines. Females administered with OA and DTX1 exhibited slightly higher intestinal toxin levels than males. In contrast, DTX2 showed high fecal elimination accompanied by low intestinal tissue retention with no sex differences. Its poor gastrointestinal absorption and faster clearance are likely to limit its systemic bioavailability, as previously reported (Alves et al., 2019; Manita et al., 2017; Rodríguez-Santos et al., 2025). On the other hand, the disappearance of DTX1 marked by the lowest fecal excretion and proportionally greater intestinal uptake, indicates higher bioavailability and prolonged tissue contact, which is consistent with its greater toxicity (Louzao et al., 2021; Rodríguez-Santos et al., 2024). Mice receiving OA initially displayed high fecal excretion decreasing over time. Local intestinal action of DSTs is known to trigger hypersecretion and fluid accumulation, thereby altering gastrointestinal motility (Ito et al., 2002; Reale et al., 2019; Wang et al., 2012; Abal et al., 2018; Louzao et al., 2021). Furthermore, preliminary observational analysis of fecal samples suggested an upward trend in sodium and chloride excretion following repeated exposure to OA and DTX1, which is consistent with the secretory diarrhea pattern previously reported in acute toxicity studies (Costas et al., 2022). However, due to sample mass constraints following primary toxin quantification, these ion alterations remain strictly qualitative support for our clinical observations and require further dedicated investigation.

Diarrhea may result from multiple mechanisms, including increased secretion and motility, reduced absorption, and the activation of inflammatory responses (Rao, 2019). Recently, extracellular Cyps have emerged as novel inflammatory biomarkers (Chen et al., 2025; Gegunde et al., 2023; Alfonso et al., 2019). Cyps are evolutionarily conserved cytosolic proteins with peptidyl-prolyl cis-trans isomerase activity and include members such as CypA and CypC (Xue et al., 2018), which can be secreted into the extracellular space in response to different stimuli, acting as an autocrine and paracrine mediator that exacerbate inflammation (Chen et al., 2025). This study provides the first qualitative profiling of extracellular CypA and CypC in the intestinal contents following repeated voluntary exposure to DSTs. Western blot analysis revealed distinct band profiles, with CypA detected in the large intestinal contents of mice administered with OA and DTX2, as well as in DTX1-treated males. Notably, detectable CypA and CypC bands were exclusively observed in the small intestinal lumen of mice daily administered with DTX1, which may correlate with the longer intestinal residence time and slower excretion of this toxin compared to other DSTs (Louzao et al., 2021; Rodríguez-Santos et al., 2024, 2025). Elevated Cyp levels in body fluids are considered as an inflammatory response to injury (Hoffmann and Schiene-Fischer, 2014). In contrast, the absence of detectable CypC in intestinal contents of animals treated with DTX2 qualitatively aligns with its significantly milder macroscopic and clinical profile. Even though the action of DSTs is related to the neuro-intestinal connection (Costas et al., 2022; Louzao et al., 2015, 2021), these descriptive biomarker profiles suggest that localized mucosal irritation and cellular stress contribute to the subchronic gastrointestinal picture caused by OA and DTX1. Nevertheless, because luminal contents preclude standard housekeeping protein normalization and densitometric statistical testing, these findings should be regarded as descriptive qualitative support for macroscopic observations rather than definitive proof of quantified inflammation.

Several methodological aspects of the present study should be acknowledged. In accordance with standard 28-day oral toxicity protocols and animal welfare principles (OECD Guideline 407), the sample size was set at five animals per sex and group. While this sample size was appropriate for identifying clear, statistically significant variations in clinical scores, fluid intake over time, and fecal toxin excretion, a larger cohort would have increased the statistical power to detect subtle physiological variations. The study evaluated a single exposure level (90 μg/kg bw), which was specifically calibrated to model a high-intake human dietary scenario derived from the regulatory threshold. Therefore, a full dose–response relationship and baseline dose modelling could not be established. The 28-day endpoint represents a subchronic exposure window. Longer-term chronic studies (e.g. 90 days or lifetime exposure) are required to evaluate potential cumulative risks over longer periods and establish formal no-observed-adverse-effect levels (NOAELs). Although gastrointestinal endpoints, fluid balance and organ toxin concentrations were determined quantitatively, histopathological examinations of the liver and kidneys, and luminal cyclophilin detection, were qualitative and descriptive. While voluntary oral ingestion in mice closely replicates natural dietary exposure, direct extrapolation of rodent physiological responses and scaling to human consumer safety thresholds should be done with caution due to differences in metabolism, toxicokinetics and toxin receptor sensitivity between species.

This study fills critical knowledge gaps identified by regulatory agencies by providing the first comparative assessment of DSTs under subchronic conditions, relevant for food safety. Our findings show that 28 days of daily exposure by voluntary feeding to an experimentally calibrated dose of 90 μg/kg bw, modelling a high-intake human dietary scenario derived from regulatory thresholds, elicited prominent gastrointestinal disturbances, including diarrhea, and macroscopic alterations in the stomach and intestines, with subtle qualitative signs of renal and hepatic sensitivity. Different toxicological profiles were observed: DTX1 exhibited lower fecal excretion and proportionally greater intestinal tissue retention. This indicates prolonged mucosal exposure and higher tissue bioavailability, which is consistent with its pronounced local gastrointestinal toxicity, severe macroscopic damage, and the differential presence of luminal cyclophilins (CypA and CypC). Conversely, DTX2 displays rapid fecal elimination and minimal tissue accumulation, restricting its bioavailability and accounting for its significantly milder clinical and histological profile. The observed sexual dimorphism in clinical symptoms and water consumption, along with the distinct potency of DTX1, indicates that repeated dietary exposure warrants comprehensive toxicological consideration. While these observations alone do not establish a human tolerable daily intake or demonstrate that the current regulatory limit is inadequate, they provide valuable in vivo evidence to refine TEFs based on subchronic oral data and support future risk assessment frameworks for marine biotoxins.

Ethics declaration

This study was conducted in accordance with the ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines. This study was approved by the Institutional Animal Care Committee of the University of Santiago de Compostela. (Approval No. 01/22/Lu-002)

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Handling Editor: Dr. Yeonhwa Park

Appendix A. Supplementary data

The following is the Supplementary data to this article: