The present study: 1) provided further evidence of the existence of gut microbiota alterations in the schizophrenia-like model induced by Poly I:C, including reduced microbial diversity, consistent with findings in patients with schizophrenia; and 2) demonstrated that exposure to cannabinoids during adolescence, particularly following MIA challenge, significantly alters gut microbiota composition and diversity in adulthood.

In this context, our findings are consistent with clinical studies reporting a dysbiotic gut microbiota profile with marked reduction in beneficial butyrate-producing bacteria, such as Faecalibacterium and Roseburia (Li Z. et al., 2024). In addition, MIA animals in the present study exhibited a reduced abundance of Clostridium hiranonis and a complete absence of Akkermansia muciniphila, two taxa closely associated with gut barrier integrity and anti-inflammatory activity. Such alterations in microbial populations are often associated with increased inflammation and metabolic imbalance, which may contribute to impaired immune regulation and abnormal neurodevelopmental outcomes (Tsamakis et al., 2022). Furthermore, the differential impact of cannabinoids observed in this study mirrors the distinct neurobiological profiles of THC and CBD. While THC has been shown to disrupt key neural circuits including reward pathways and exacerbate psychotic-like symptoms (Fu et al., 2021), CBD may exert anti-inflammatory actions and promote the restoration of gut barrier integrity (Tao et al., 2025), potentially counteracting some of the deleterious effects of the MIA challenge and THC treatment, though direct functional validation was beyond the scope of the present study.

The observed reduction in bacterial diversity in MIA-offspring is in line with our previous work showing a reduced number of species and relative abundance distribution (Romero-Miguel et al., 2023) and reflects features of gut dysbiosis linked to neurodevelopmental disorders (Hsiao et al., 2013). Also, the influence of CBD on microbial richness appears to be highly context-dependent, as it leads to a reduction in bacterial richness in healthy controls but not in MIA-offspring. However, combined exposure to THC and CBD under MIA conditions results in a further exacerbation of richness loss, indicating potential synergistic or additive effects of cannabinoids in the presence of immune-mediated developmental perturbations. These findings highlight the complex interplay between cannabinoid signaling, immune activation and neurodevelopmental disturbances in shaping gut microbiota composition. Reduced microbial diversity and altered community composition, as shown by the Shannon and Chao indices, are indicative of reduced functional capacity of the microbial ecosystem, which could contribute to dysregulated gut-brain axis signaling and increased vulnerability to neuropsychiatric disorders later in life (Guo et al., 2025).

Our results demonstrate that, at the phylum level, Bacillota and Bacteroidota dominated the gut microbiota ecosystem across all experimental conditions, consistent with their established role as major constituents of the mammalian gut microbiota. Given their importance in maintaining microbial homeostasis, we evaluated the Bacillota/Bacteroidota (B/B) ratio, a commonly used indicator of gut microbial balance whose elevation is frequently been associated with neuropsychiatric disorders. Although some clinical studies in schizophrenia have reported an elevated B/B ratio, our analysis did not reveal a significant effect of the MIA model on this parameter, showing only a non-significant trend toward increase, nor any interaction between MIA and cannabinoids (Szeligowski et al., 2020; Xiang et al., 2022). This lack of effect is consistent with previous findings in prenatal immune activation models (Juckel et al., 2021; Romero-Miguel et al., 2023). In contrast, CBD treatment reduced the B/B ratio in both saline and MIA-offspring, whereas THC tended to increase it, suggesting that cannabinoids can modulate overall microbial community structure independently of the underlying pathological condition (Ibrahim et al., 2022; Chen et al., 2025).

Within the phylum Bacillota, we observed significant modulations in families such as Clostridiaceae and Clostridiales Family XVI, the latter being notably increased by combined THC and CBD treatment in Saline-offspring. These trends were largely preserved at the genus and species levels, where THC selectively enriched Clostridium hiranonis, Roseburia and Oscillibacter in MIA-offspring. These microorganisms play an important role (directly or indirectly) in short-chain fatty acid (SCFA) production and energy metabolism (Zhang et al., 2022). They are also prolific butyrate producers (Li et al., 2022; Ye et al., 2023), conferring antioxidant and anti-inflammatory effects through mechanisms such as suppression of lipid peroxidation (MDA) (Vincent et al., 2013) and inhibition of NF-κB-dependent transcription of TNF-α and IL-6 (Arpaia and Rudensky, 2014; Kristiansson et al., 2021). Similarly, Oscillospiraceae abundance increased following CBD treatment in MIA-offspring. Given the well-established anti-inflammatory properties of this family, which is largely attributed to the production of SCFA, particularly butyrate production, this finding may reflect a microbiota-mediated mechanism through which CBD contributes to the restoration of this taxon in MIA-offspring.

Specific taxonomic shifts were also observed at the family level within the Bacteroidota phylum, which is fundamental for immune homeostasis and polysaccharide metabolism (Yoo et al., 2020; Afzaal et al., 2022; Kim et al., 2024). For instance, CBD treatment increased the relative abundance of Prevotellaceae in both Saline and MIA-offspring, whereas THC treatment showed the opposite pattern. This CBD-induced shift has previously been linked to reduced SCFA production and increased oxidative stress in rodent models of neuroinflammation (Mostafavi Abdolmaleky and Zhou, 2024; Graf et al., 2025). Prevotellaceae is commonly found in the human gut microbiota and is typically associated with diets rich in complex carbohydrates and fibre. However, while clinical reports on Prevotellaceae in schizophrenia remain heterogeneous (Su et al., 2025; Tang et al., 2025), our data suggest that cannabinoids can differentially modulate this family. Moreover, CBD also induced a marked upregulation of Flavobacteriaceae, a taxon presumed to produce enzymes involved in oxidative regulation and maintenance of redox balance (Zhou et al., 2024b), which may contribute to a healthier microbial profile, as its increase has been associated with symptom remission in schizophrenia (Romero-Miguel et al., 2023).

At the species level, THC enriched Parabacteroides distasonis in MIA-offspring, a taxon associated with immune regulation via aryl-hydrocarbon receptor activation, a pathway implicated in the attenuation of oxidative stress (Abreu and Abreu et al., 2021). In contrast, CBD, either alone or in combination with THC, reduced Parabacteroides and Parabacteroides distasonis abundance, suggesting a shift toward a microbial profile with potentially diminished anti-inflammatory capacity. These findings underscore the distinct and cannabinoid-specific effects of THC and CBD on gut microbial composition. Furthermore, Akkermansia muciniphila emerged as an important contributor to group discrimination in the multivariate analyses. This observation is particularly relevant, as A. muciniphila is essential for maintaining gut barrier integrity. Its absence in untreated MIA-offspring reflects an increased susceptibility to gut barrier dysfunction and inflammation, whereas the partial restoration observed following cannabinoid exposure suggests these compounds might help mitigate some of the microbiota alterations associated with the MIA phenotype (Khalili et al., 2024).

Beyond the dominant phyla, significant modulations were also observed in less abundant taxa, including Pseudomonadota, Nitrospinota and Deferribacterota. Both CBD and THC increased the relative abundance of Nitrospinota, a phylum involved in nitrogen cycling, suggesting a cannabinoid-driven modulation of microbial nitrogen metabolism that is independent of early immune activation (Lücker et al., 2010; Daims et al., 2015). In contrast, Deferribacterota abundance was notably reduced in MIA-offspring and was undetectable in animals exposed to either THC and CBD alone, suggesting that this phylum is highly sensitive to early-life inflammatory insults and cannabinoid exposure. Interestingly, its persistence in animals receiving the combined THC + CBD treatment may indicate a differential response to co-administration of both cannabinoids, highlighting a potentially distinct modulatory effect of the combined treatment on this microbial taxon (Rooks and Garrett, 2016; Borrego-Ruiz and Borrego, 2025).

Within the phylum Pseudomonadota, cannabinoid exposure appeared to limit the expansion of potentially opportunistic taxa such as Enterobacteriaceae and Pasteurellaceae, particularly in MIA-offspring. This effect is consistent with previous evidence indicating that SCFA-rich environments can suppress the growth of Pseudomonadota taxa members and mitigate the oxidative and inflammatory burden associated with their expansion (Sun et al., 2024). These microbial shifts may have important functional consequences. For instance, cannabinoid treatment decreased the abundance of Bilophila, a genus frequently elevated in patients with schizophrenia and linked to low-grade inflammation, gut dysfunction and cognitive impairment (Su et al., 2025; Ataei et al., 2026). Given that Bilophila has been shown to negatively impact butyrate-producing bacterial populations, its reduction, together with the enrichment of beneficial taxa, supports a shift toward a more anti-inflammatory and metabolically favorable gut environment. Collectively, these microbiota changes may be contributing to the attenuation of the MIA-induced inflammatory alterations by promoting a more balanced and protective intestinal ecosystem.

Finally, correlation analyses linking gut microbiota composition with ventricular volume and prefrontal cortex oxidative/inflammatory markers further support the taxonomic shifts described above. These associations are consistent with previous evidence showing that schizophrenia- and MIA-related phenotypes are accompanied by gut dysbiosis, immune activation, and altered redox balance, and that microbial alterations may contribute to neuroinflammatory processes and structural brain changes in vulnerable offspring (Nguyen et al., 2021; Zhou et al., 2024a). In relation to the MIA phenotype, Mycoplasmatota was associated with increased iNOS and reduced HO1 levels, a pattern consistent with the pro-inflammatory and impaired antioxidant status reported in schizophrenia and MIA models (Su et al., 2025). These findings align with the notion that MIA-induced dysbiosis may contribute to neuroinflammation while weakening antioxidant defence, thereby contributing to neurobiological alterations, including ventricular enlargement (Zhou et al., 2024a). THC exposure showed the most prominent microbiota-phenotype associations. In MIA-offspring, THC-related microbial changes were associated with ventricular volume, with positive correlations involving Bacillota and a negative correlation with Bacteroidota. In Saline-offspring, the anti-inflammatory Mycoplasmatota taxa was positively associated with iNOS, whereas pro-inflammatory taxa, such as Pasteurellaceae and Desulfovibrionaceae, showed negative associations. Rather than reflecting a uniform anti-inflammatory effect, this pattern suggests that THC reshapes specific host-microbiota relationships in a phenotype-dependent manner, consistent with previous reports showing that cannabinoid effects on the gut microbiota depend on the underlying inflammatory status (Ibrahim et al., 2022). Moreover, the negative associations between Mycoplasmatota and the antioxidant markers HO1 and NQO1 in MIA-offspring, further support its potential role as an indicator of inflammatory burden and impaired redox regulation (Su et al., 2025). Meanwhile, CBD treatment revealed a different pattern of associations. The relationship between Mycoplasmatota and ventricular volume varied according to the phenotype, being positive in Saline-offspring but negative in MIA-offspring, even in the presence of THC (MIA-THC-CBD group). In addition, in MIA-offspring, Nitrospinota and Bacteroidota were positively correlation with iNOS, whereas Mycoplasmatota showed a negative correlation with iNOS but a positive association with HO1. These findings suggest that CBD may differentially modulate inflammatory and antioxidant pathways, potentially promoting a reorganization of microbiota-host interactions distinct from that induced by THC (Ibrahim et al., 2022; Lamanna-Rama et al., 2024). Collectively, our results show that THC and CBD differentially modulate the gut-brain axis in a phenotype-dependent manner. While THC-induced enrichment of specific Bacillota taxa may counteract some MIA-associated inflammatory alterations, CBD appears to engage distinct inflammatory and antioxidant signaling pathways. Overall, these observations extend prior work by demonstrating that cannabinoid-induced microbiota changes are not uniform but depend on the neurodevelopmental background, and are linked to both structural and redox-related brain outcomes (Nguyen et al., 2021; Lamanna-Rama et al., 2024). They also highlight the complexity of cannabinoid-microbiota interactions and their potential relevance for neurodevelopmental disorders characterized by immune dysregulation and altered neuroinflammatory signaling.