The prebiotic effect of all biomasses was assessed in vitro by evaluating their ability to support and enhance the growth of probiotic bacterial strains relevant to fish and human health. Six reference strains were used – Bacillus spizizenii ATCC6633, Bacillus subtilis ATCC6051, Pseudomonas chlororaphis subsp. piscium DSM21509, Bifidobacterium longum ATCC15708, Lactobacillus rhamnosus ATCC BAA-3227, and Lactobacillus acidophilus ATCC314 – along with Lactococcus raffinolactis SAFE_10 recovered from diseased O. mykiss. This final strain was included due to the described probiotic properties of this species. This panel comprises both Gram-positive and Gram-negative bacteria, including aerobic, facultative anaerobic, and anaerobic/microaerophilic strains, which represent a broad spectrum of phylogenetic and physiological diversity relevant to fish and human health. Each bacterial strain was cultured in Basal Mineral Salts (BMS) medium, composed of 3.0 g/L NaNO₃, 0.5 g/L KCl, 1.0 g/L KH₂PO₄, 0.5 g/L MgSO₄·7H₂O, 0.1 g/L NaCl, and 1 mL/L trace salt solution composed of 1.0 g FeSO₄·7H₂O, 0.1 g CuSO₄·5H₂O, and 0.1 g ZnSO₄·7H₂O dissolved in 100 mL of distilled water, supplemented with 1% (w/v) of each freeze-dried biomass individually. Bacterial growth was assessed under both liquid and solid cultures. All cultures were incubated for 96 h under strain-specific temperature and oxygen conditions (Supplementary Table S2). For B. longum ATCC15708, an obligate anaerobic strain, oxygen depletion was achieved using an anaerobic jar system (Thermo Scientific, United States). In liquid media, growth was monitored every 24 h by measuring the culture’s optical density at 600 nm (OD₆₀₀), starting from an initial inoculum standardized to OD₆₀₀ = 0.01. For solid media assays, growth was evaluated qualitatively by visual inspection of colony formation after 96 h of incubation. The growth-promoting effects of biomass supplementation were determined by comparing results with those of the unsupplemented BMS, based on increases in OD₆₀₀ for liquid cultures and the presence of colony development on solid media. The assays were conducted in duplicate.

The antimicrobial activity of the previously obtained extracts was evaluated using an agar disk diffusion assay, as previously described by Girão et al. (2024b). The extracts were tested against a panel of pathogens relevant to both aquaculture and human health, comprising a diverse range of Gram-positive, Gram-negative, and fungal microorganisms. Aquaculture-relevant strains included Edwardsiella tarda DSM30052, Aeromonas hydrophila DSM3018, Pseudomonas anguilliseptica DSM1211, Yersinia ruckeri ATCC29473, Listonella (Vibrio) anguillarum ATCC19264, Tenacibaculum maritimum ATCC43397, and Lactococcus garvieae DSM20684. The human-associated pathogens tested were Escherichia coli ATCC25922, Staphylococcus aureus ATCC29213, Salmonella enterica ATCC25241, and Candida albicans ATCC10231. Additionally, two strains isolated from the rainbow trout – Aeromonas sobria SAFE_07 and Aeromonas salmonicida subsp. salmonicida SAFE_15 – were also included in the screening. Each strain was grown in the appropriate culture conditions (Supplementary Table S3). For the assay, a suspension of each strain was prepared in the appropriate liquid medium, with its turbidity adjusted to 0.5 McFarland standard (OD₆₂₅ = 0.08–0.13) (Balouiri et al., 2016). The suspensions were used to evenly seed the surface of agar plates. Blank paper disks (6 mm in diameter) were then placed on the inoculated plates and loaded with 15 μL of each extract (1 mg/mL). Enrofloxacin (1 mg/mL; Sigma-Aldrich, MO, United States) and nystatin (1 mg/mL; Sigma-Aldrich) were used as positive controls for bacterial and fungal strains, respectively. DMSO was used as the negative control. Each extract was tested in two independent replicates. Antimicrobial activity was assessed by measuring the diameter (mm) of the inhibition zones formed around the disks.

The antioxidant activity of the extracts was evaluated using the DPPH (2,2-diphenyl-1-picrylhydrazyl) radical scavenging assay as described by Dulf et al. (2015). Briefly, 40 μL of each extract was mixed with 200 μL of a DPPH solution (0.02 mg/mL). The mixture was incubated at room temperature for 15 min, and the absorbance measured at 517 nm. The assay was performed on a microplate, in duplicate. The extracts obtained with dichloromethane were excluded from the assay due to their poor solubility in methanol, the solvent used for preparing the DPPH reagent.

The immunomodulatory properties of the extracts were evaluated in vitro using the RAW 264.7 macrophage cell line (American Type Culture Collection, ATCC). Both anti-inflammatory and pro-inflammatory activities were assessed by quantifying nitric oxide (NO) production, as described by Regueiras et al., 2022 (Regueiras et al., 2021), while cytotoxic effects were determined using the MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] assay, as described by Girão et al. (2019). RAW 264.7 cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) (Gibco, Thermo Fisher Scientific, Waltham, MA, United States), supplemented with 10% (v/v) fetal bovine serum (Biochrom, Berlin, Germany), 1% (v/v) antibiotics including 100 mg/L streptomycin, 100 IU/mL penicillin (Biochrom, Berlin, Germany), and 0.1% (v/v) amphotericin B (GE Healthcare, Little Chalfont, UK). Cells were seeded in 96-well plates (3.5 × 10⁵ cells/well) and allowed to adhere overnight. Cultures were maintained at 37 °C in a humidified atmosphere containing 5% CO₂. For the assessment of anti-inflammatory activity, cells in each well were co-exposed to the extracts (15 μg/mL, dissolved in DMSO ≥99.9%.) and lipopolysaccharide (LPS; 1 μg/mL; Sigma-Aldrich, St. Louis, MO, United States). After 24 h of incubation, 75 μL of culture supernatant was transferred to a new 96-well plate and mixed with 75 μL of Griess reagent, composed of 10 mg/mL sulfanilamide and 1 mg/mL N-(1-naphthyl)ethylenediamine dihydrochloride in 2% phosphoric acid, to quantify nitric oxide (NO) production. Following a 10-min incubation in the dark, the absorbance was measured at 562 nm. Results were expressed as the percentage of NO production relative to the LPS-stimulated control (DMSO + LPS). Pro-inflammatory activity was evaluated under the same conditions but without LPS. Results were expressed as the percentage of NO production relative to the solvent control DMSO. The remaining cells from the initial plates, both those used for anti- and pro-inflammatory screening, were used to assess cell viability after extract exposure using the MTT assay. A fresh MTT solution (0.5 mg/mL in DMEM) was prepared, and 100 μL was added to each well. After a 45-min incubation at 37 °C, the medium was carefully removed, and formazan crystals were dissolved in 100 μL of dimethyl sulfoxide (DMSO). Finally, absorbance was measured at 570 nm, and cell viability was expressed as a percentage relative to the solvent control (DMSO). The assays were performed in two independent biological replicates, each conducted in technical triplicate. Data from immunomodulatory assays were tested for significant differences compared to the most appropriate control.

All quantitative data are presented as mean ± standard deviation, unless otherwise stated. For immunomodulatory, cytotoxicity and antioxidant assays, differences among treatments were analyzed using one-way analysis of variance (ANOVA) for normally distributed data, followed by Dunnett’s post hoc test to compare each treatment with the appropriate control. When data did not meet the assumptions of parametric analysis after transformation, the non-parametric Kruskal–Wallis test was applied, followed by Dunn’s multiple comparison test. A confidence level of 95% was considered in all statistical analyses. The significance level was set at p < 0.05 for all tests. All analyses were performed using SPSS (v27, IBM, NewYork, NY, United States).

All strains, except B. longum ATCC 15708, exhibited growth in at least one biomass-enriched substrate, both in liquid and solid media, while growth in unsupplemented BMS was negligible, confirming that the biomass provided the sole carbon and nutrient sources (Figure 2; Supplementary Table S4). B. spizizenii ATCC 6633 showed consistent growth in response to nearly all biomass substrates. The highest OD₆₀₀ values at 96 h were observed with P. ostreatus fruit bodies supplemented with 12.5% frass (1.688), followed by 5% (1.655) and 10% (1.469). Scenedesmus spp. (OD₆₀₀ = 1.322) also strongly supported bacterial proliferation. Among all tested strains, P. chlororaphis subsp. piscium DSM 21509 showed the strongest and most consistent growth across all tested prebiotic strains. At 96 h, five biomass samples – P. ostreatus fruit body enriched with 5 to 15% frass – supported maximum bacterial growth of this strain (OD₆₀₀ = 3.000). Other high-performing substrates with top-ranking probiotic strains growth promoters included T. molitor (OD₆₀₀ = 2.555) and Scenedesmus spp. (OD₆₀₀ = 1.863). For all other biomasses, final OD₆₀₀ values still ranged from 0.430 to 1.636, demonstrating the strain’s broad metabolic versatility and its ability to utilize a wide spectrum of biomass-derived compounds for growth. L. raffinolactis SAFE_10 exhibited limited growth across all tested substrates, with OD₆₀₀ values never exceeding 0.5 at 96 h, indicating that the tested substrates showed low prebiotic responsiveness for this strain under the tested conditions. B. subtilis ATCC 6051 showed moderate to strong growth across several biomass substrates, with varying patterns over time. The highest OD₆₀₀ at 96 h was observed with P. ostreatus fruit body enriched with 15% frass (OD = 1.544), followed by 12.5% (1.455) and 12% (1.412). P. tricornutum also induced a strong and progressive growth response, yielding an OD₆₀₀ of 1.333, the same as that obtained with P. ostreatus fruit body enriched with 5% frass. The invertebrate-derived biomasses, T. molitor and E. fetida, yielded similar OD₆₀₀ values at 96 h (1.255 and 1.166, respectively). L. rhamnosus ATCC BAA-3227 showed moderate and substrate-dependent growth. At 96 h, the best-performing substrate was P. ostreatus stems enriched with 15% frass, which supported the highest growth (OD₆₀₀ = 1.355), followed by 10% (1.220) and 7.5% (1.163). Lower growth was observed in response to invertebrate, microalgae, and plant biomasses, with OD₆₀₀ values never exceeding 0.25 at 96 h. This strain therefore displayed a selective substrate preference, responding best to fungal stem-derived biomasses enriched with T. molitor frass, while showing negligible utilization of other biomass types. L. acidophilus ATCC 314 displayed limited growth across most biomass substrates, with a few notable exceptions. The highest OD₆₀₀ at 96 h was observed with E. fetida (1.032), followed by Scenedesmus spp. (0.989) and T. molitor (0.745). With all the other biomasses, OD₆₀₀ values remained at or near baseline throughout the incubation period (Figure 2). These results indicate a narrow substrate preference, responding specifically to invertebrate- and microalgae-based biomasses. In summary, the evaluated biomasses demonstrated distinct and selective prebiotic effects on a diverse range of probiotic strains relevant to both fish and human health. While most strains showed enhanced growth with at least one biomass substrate, the degree and specificity of growth varied notably among strains. P. chlororaphis subsp. piscium DSM 21509 exhibited the broadest metabolic versatility, thriving on nearly all biomass types, particularly P. ostreatus enriched with T. molitor frass. In contrast, strains like L. raffinolactis SAFE_10 and L. acidophilus ATCC 314 showed limited responsiveness, reflecting narrower substrate preferences. Prior studies on the same species tested in our work have reported prebiotic-like properties or the ability to modulate beneficial microbes: mushroom polysaccharides, especially β-glucans from P. ostreatus, have been shown to enhance the growth of lactic acid bacteria and bifidobacteria in vitro and to modulate gut communities in vivo (Yu et al., 2023); microalgae/diatoms such as P. tricornutum can shift microbiota composition and increase short-chain fatty acids in animals (Stiefvatter et al., 2022), and microalgal extracts can stimulate probiotic growth and confer protective effects in vitro (Carballo et al., 2019); invertebrate-derived materials, such as T. molitor and E. fetida contain chitin and chitosan, which consistently exhibit prebiotic activity for Lactobacillaceae and Bifidobacteriaceae (Lopez-Santamarina et al., 2020; Zacharis et al., 2024); duckweeds (Lemna spp.) can serve as substrates for probiotic Bacillus propagation and have been explored as fermentable feeds (Mahoney et al., 2021). Overall, fungal biomasses, invertebrate- and microalgae-derived substrates showed the most consistent prebiotic-like stimulation, highlighting their potential for targeted probiotic support and biotechnological applications.

The antimicrobial activity of the library of crude extracts generated from the biomasses was assessed against a panel of fish and human bacterial pathogens using the disc diffusion assay. Growth inhibition zones were recorded for each extract-strain combination (Figure 3). Of the 63 extracts screened, 20.6% (n = 13) exhibited activity against at least one pathogen, with inhibition zones ranging from 9 to 14 mm in diameter. The solvent control (DMSO) did not affect the proliferation of any bacterial strain tested, confirming that the observed inhibitory effects were attributable solely to biomass-derived compounds. Overall, extracts demonstrated greater efficacy against fish-associated pathogens, with measurable inhibition observed in six of the nine tested strains. In contrast, activity against human-associated pathogens was limited, with inhibition detected only for S. aureus ATCC 29213 exposed to the aqueous P. tricornutum extract. Differences in efficacy were observed among biomass types: microalgae extracts were the most active, inhibiting five distinct pathogens, followed by invertebrate-derived extracts (which inhibited three pathogens), and fungal and aquatic plant extracts (inhibiting two pathogens). Pathogen susceptibility varied, with T. maritimum ATCC 43397 being the most sensitive, followed by L. anguillarum ATCC 19264 and A. sobria SAFE_07, each inhibited by multiple biomass-solvent combinations. Notably, all active microalgal extracts inhibited L. anguillarum, whereas no other biomass demonstrated activity against this strain. The observed differences between fish- and human-associated pathogens suggest a degree of target specificity. Furthermore, solvent-dependent activity patterns highlight the importance of employing a broad range of solvents to recover both polar and non-polar bioactive compounds. Most of the extracts exhibiting antimicrobial activity were obtained using dichloromethane, followed by the aqueous extracts, and then those recovered using a 1:1 acetone/methanol mixture. These results align with previously reported antimicrobial properties of the tested biomass types. Fungal extracts, particularly from P. ostreatus, have been shown to inhibit both Gram-positive and Gram-negative bacteria (Gashaw et al., 2020; Yakobi et al., 2023). Although most studies to date have focused on human clinically relevant pathogens, our recent publication demonstrated for the first time that extracts from P. ostreatus partially cultivated on mealworm frass exhibit antimicrobial activity against T. maritimum (Hilali et al., 2025), a major marine aquaculture pathogen, supporting their use as sustainable functional ingredients for fish health management. Invertebrate-derived extracts, specifically those from T. molitor, rich in chitin and chitosan, and antimicrobial enzymes such as lysozymes and peptides, have demonstrated efficacy against a wide range of Aeromonas, Vibrio, Pseudomonas, and Staphylococcus species (Ansari and Sitaram, 2011; Hwang et al., 2022; Liu et al., 2004). To our knowledge, this is the first report of E. fetida derived extracts displaying inhibitory activity against T. maritimum. Microalgae are recognized as prolific producers of a wide range of bioactive metabolites. Among these, Scenedesmus spp. (Dantas et al., 2019; Zaharieva et al., 2022) and P. tricornutum (Desbois et al., 2008; Desbois et al., 2009) have been reported to yield extracts with antibacterial activity against relevant pathogens, including multidrug-resistant strains. In this work, we expand the repertoire of susceptible microbes by demonstrating inhibition of T. maritimum, A. sobria, L. garviae, and L. anguillarum, all major bacterial pathogens in aquaculture. Aquatic plants, such as L. minor, have been recognized for their antimicrobial potential. Early studies demonstrated broad-spectrum antibacterial activity of duckweed extracts against human-associated pathogens, including Staphylococcus, Bacillus, Citrobacter, and Neisseria species (Gülçin et al., 2010). Later work confirmed activity against Pseudomonas fluorescens and highlighted efficacy across various solvent extracts (González-Renteria et al., 2020). However, data targeting aquaculture pathogens remain sparse. In our study, we demonstrate for the first time that L. minor extracts inhibit A. hydrophila and T. maritimum, two critical fish pathogens, thereby expanding the known antimicrobial spectrum of duckweed and strengthening its promise as a sustainable agent for managing bacterial diseases in aquaculture systems. Altogether, our findings identify specific biomass–solvent combinations as promising candidates for the development of functional aquaculture feed ingredients, particularly relevant in production settings where effective control of bacterial pathogens is critical.

The antioxidant activity of the biomass derived aqueous and acetone/methanol (1:1) extracts was evaluated by their capacity to scavenge DPPH (2,2′-diphenyl-1-picrylhydrazyl) free radicals (Table 2). The extracts obtained with dichloromethane were excluded from the assay due to their poor solubility in methanol. Values ranged from less than 1 μmol TE g⁻¹ to more than 13 μmol TE g⁻¹ depending on the biomass type and extraction solvent. Aquatic plants, duckweed and watercress, showed the highest activities in aqueous extracts (13.4 and 8.7 μmol TE g⁻¹, respectively). These results may indicate a high content of hydrophilic antioxidants such as phenolic compounds, flavonoids, and vitamin C analogues (Martínez-Sánchez et al., 2008; Muller et al., 2025; Takács et al., 2025), which are known to enhance systemic antioxidant capacity in fish (Bešlo et al., 2023). Microalgae exhibited a contrasting antioxidant activity profile, where aqueous P. tricornutum extracts showed reduced activity (0.6 μmol TE g⁻¹) but the respective acetone/methanol extract showed comparatively higher activity (6.0 μmol TE g⁻¹), which may be due to its carotenoid content, such as fucoxanthin (Neumann et al., 2019). These lipophilic compounds are readily absorbed and deposited in fish tissues, where they can act as chain-breaking antioxidants and protect cellular components from oxidative damage (Xiao et al., 2020). Fungal biomass, particularly aqueous extracts of P. ostreatus fruiting body, showed low antioxidant activity (1.6–3.2 μmol TE g⁻¹), while the aqueous extracts of stems and acetone/methanol extracts of both fruit bodies and stems showed a consistently lower antioxidant capacity. These results indicate a low abundance of hydrophilic compounds, which could contribute to the free radical scavenging capacity when included in aquafeeds. In the same regard, the studied invertebrate extracts showed the lowest values (around 1–2 μmol TE g⁻¹), suggesting that while they represent a sustainable protein source, their role as a source of antioxidant compounds may be limited. Altogether, these results suggest that biomass sources differ not only in the magnitude of antioxidant activity but also in the type of compounds they provide. Aquatic plants showed a significantly higher antioxidant capacity, whereas microalgae showed more moderate antioxidant activity, depending on the polar nature of the extracted compounds. Given the negative impact of oxidative stress on fish growth, immunity, and overall welfare, incorporating antioxidant-rich biomasses as functional ingredients into aquafeeds could reduce reliance on synthetic supplements and enhance fish resilience (Hu et al., 2025).

The immunomodulatory potential of the biomass-derived crude extract library was evaluated based on their ability to modulate NO production in RAW 264.7 macrophages, either in the absence (pro-inflammatory assay) or presence (anti-inflammatory assay) of LPS, as well as their impact on cell viability (cytotoxicity assay) (Figure 4). Significant pro-inflammatory activity (p < 0.001), measured as an increase in NO production relative to basal levels (solvent control, DMSO), was detected for aqueous extracts from fungi (P. ostreatus stems, 2.5%), microalgae (P. tricornutum and Scenedesmus spp.), and aquatic plants (L. minor and N. officinale), with increases ranging from 10 to >40%. Given their detrimental potential for aquaculture nutrition, these extracts were removed from the valorization pipeline. Significant anti-inflammatory activity, indicated by reduced NO production in LPS-stimulated cells relative to the DMSO + LPS control, was observed across a broader spectrum of biomass types, including fungal extracts (P. ostreatus stems 10–15%, fruit bodies 0–15%), microalgae (Scenedesmus spp.), annelid (E. fetida), and aquatic plant (N. officinale), with reductions between 20 and 50%. Importantly, several extracts with strong anti-inflammatory activity, namely P. ostreatus stems (10%), fruit bodies (0–2.5%), Scenedesmus spp., and N. officinale, did not cause any significant reduction in RAW 264.7 cell viability, demonstrating that their NO-lowering effects were not attributable to cytotoxicity. All the remaining ones were considered as “false positives” and omitted from further consideration. These non-cytotoxic, anti-inflammatory profiles are particularly promising as functional feed ingredients, offering bioactive properties that may help modulate inflammatory responses in aquaculture species without compromising cell health. These results are consistent with previous reports showing that edible mushrooms, including P. ostreatus, and microalgae, such as P. tricornutum, can modulate macrophage NO production and inflammatory cytokines through pathways including NF-κB and iNOS, without inducing cytotoxicity (Jedinak et al., 2011; Llauradó Maury et al., 2021; Riccio and Lauritano, 2019). In contrast, the invertebrate-derived E. fetida extract, despite showing anti-inflammatory activity, markedly reduced macrophage viability (<20%), indicating strong cytotoxic potential and limiting its suitability for feed applications. Taken together, these findings identify P. ostreatus fruit bodies 0–2.5% and stems 10%, Scenedesmus spp., and N. officinale as the most promising biomass sources for the development of novel aquafeed ingredients with targeted anti-inflammatory activity and minimal cytotoxic risk.