Between 2022 and 2025, 65 MG strains isolated from commercial broiler and layer flocks exhibiting clinical symptoms consistent with Mycoplasma infection across various regions of Tunisia were analyzed in this study (Table 1). Cultivation was performed in Frey’s broth, which was enriched with heat-inactivated fetal bovine serum, L-arginine, beta-nicotinamide adenine dinucleotide (β-NAD), phenol red (as a pH indicator), yeast extract, and glucose. To reduce bacterial contamination, penicillin was added to the medium. Inoculated cultures were incubated at 37 °C under microaerophilic and humidified conditions. Bacterial growth was assessed by changes in pH and media turbidity. Three successive rounds of cloning from individual colonies were conducted to ensure the purity of the isolates and genetic uniformity. Molecular identification of MG strains was confirmed by PCR using primers specific to the pMGA gene, a reliable genetic marker for MG, as described elsewhere (Mardassi et al., 2005).

The Spirulina platensis used in this study was obtained from a monitored aquaculture facility operated by Eden Life,¹ located in the Kettana oasis between Gabès and Mareth (Southern Tunisia) and situated in an ecologically protected zone, free from industrial and urban contaminants. Characterized by an arid climate, abundant sunlight, and stable temperatures, the local environment creates favorable conditions for Spirulina cultivation. The microclimate provided by the surrounding palm and pomegranate trees further enhances growth stability. The cyanobacteria were cultured in freshwater without synthetic chemicals. The biomass was collected after the cultivation phase, carefully dried under controlled conditions to retain its bioactive components, and then finely milled into a powder. It was stored under appropriate conditions pending use.

The antimicrobial susceptibility of MG to conventionnal antibiotics used in poultry production was evaluated, focusing on tetracyclines (doxycycline, oxytetracycline) and macrolides (tilmicosin, tylosin, Aivlosin). Stock solutions of the antibiotics were prepared and stored following the manufacturers’ instructions. Regarding the MG suspension, the bacterial titer was first determined by colony counting on solid Frey’s medium, followed by tenfold serial dilutions to obtain a final inoculum of 10⁵ CFU/mL and the susceptibility was determined by the broth microdilution method (Hannan, 2000). Briefly, 96-well microtiter plates were set up for each MG strain, with five rows containing serial two-fold dilutions of each antibiotic, ranging from 32 to 0.03 μg/mL, in Frey’s broth medium and a bacterial suspension (10⁵ CFU/mL) was added to each well. The remaining three rows served as controls: a growth control without antibiotic, a sterility control with no inoculum, and a quality control with the reference strain, MG S6 (ATCC 15302). The plates were incubated at 37 °C under the appropriate conditions and monitored daily for growth. The minimum inhibitory concentration (MIC) was defined as the lowest concentration that completely inhibited visible bacterial growth as described (Hannan, 2000), using interpretive criteria. All tests were performed in duplicates.

The broth microdilution method was used to determine the MIC of Spirulina platensis against MG (diluted to 10⁵ CFU/mL), with DMSO (100% v/v) as the solvent. Spirulina powder was initially prepared at concentrations of 2, 4, and 8 mg/mL. Each solution was then sonicated at 64 °C for 15 min, followed by incubation at 56 °C for 30 min to ensure complete solubilization and optimize the release of bioactive compounds. This process maximized DMSO solubility and antibacterial activity, and the resulting formulation was selected for further analysis to ensure consistent efficacy while maintaining the integrity of Spirulina’s active components. A starting Spirulina concentration of 40 mg/mL was added to the first well, and then the solution was serially diluted twofold to 1.953 μg/mL. Control wells included a growth control without Spirulina, a sterility control, and a solvent control (2.5% DMSO), all prepared at the same concentrations. Plates were incubated at 37 °C, and bacterial growth was assessed by observing a color change in the medium, using phenol red as a pH indicator. The MIC was recorded as having the lowest concentration of Spirulina that visibly inhibited MG growth. All assays were performed in duplicate. The MIC for the Spirulina extract was determined in duplicate for each isolate, with a third replicate performed when results were discordant (>1 two-fold dilution difference).

The cytotoxic effect of the Spirulina platensis extract was evaluated on RAW 264.7 murine macrophage cells using the MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] assay. Cells were seeded in 96-well plates at a density of 2.0 × 10⁴ cells per well (2.0 × 10⁵ cells/mL in 100 μL of complete medium) and allowed to adhere for 24 h at 37 °C in a humidified 5% CO₂ atmosphere. Serial two-fold dilutions of the extract in culture medium, ranging from 4,000 μg/mL to 125 μg/mL, were then added to the wells. After a 24-h incubation, 10 μL of MTT solution (5 mg/mL in PBS) was added to each well, resulting in a final concentration of 0.5 mg/mL. The plates were incubated for 4 h to allow formazan crystals to form. The medium was then carefully aspirated, and the formazan crystals were dissolved in 100 μL of DMSO. Absorbance was measured at 590 nm using a microplate reader, with a reference wavelength of 620–650 nm to correct for non-specific absorption.

The assay was performed in three independent biological replicates (using separate cell passages and extract preparations). For each biological replicate, all extract concentrations and controls were tested in quadruplicate technical replicates. Cell viability for each condition was calculated as the mean of the technical replicates, and this mean value was used as a single data point (n = 1) per biological replicate for statistical analysis. The Spirulina concentration that maintained macrophage viability above 90% while retaining antibacterial activity was selected for subsequent antimicrobial testing.

The overall integrated methodology, from the cultivation of Spirulina platensis to the bioactivity assessment of its extract against Mycoplasma gallisepticum, is summarized in Figure 1.

All experiments were performed in duplicate, and the data were analyzed using RStudio (version 4.4.1). A two-way ANOVA was conducted to evaluate the effect of solvent type and concentration on the total phenolic content of Spirulina platensis and its antimicrobial activity. A one-way ANOVA was used to assess the other variables and to compare the effects of different concentrations of Spirulina platensis across various solvents with the positive controls. When results were significant (p < 0.05), the least significant difference (LSD) test was applied for pairwise comparisons. Pairwise associations between the log₁₀-transformed MIC values of the six antimicrobial agents were examined. After the Shapiro–Wilk test confirmed non-normal distributions for all agents (p < 0.05), we conducted a dual analytical approach: (1) Pearson’s product–moment correlation to quantify linear relationships, and (2) Spearman’s rank correlation as a complementary, distribution-free sensitivity analysis. This two-method strategy was implemented to ensure that any reported associations were not artifacts of distributional assumptions. Principal Component Analysis (PCA) was also performed as a multivariate approach to examine correlations between quantitative and qualitative variables. PCA reduces data complexity by transforming correlated variables into a set of uncorrelated components based on their covariance structure, enabling the identification of variables with the greatest explanatory power.

Using the broth microdilution method, in vitro testing of the 64 field isolates and the ATCC reference strain revealed a wide range of MICs (Table 2). Based on established susceptibility breakpoints for MG, doxycycline demonstrated the highest efficacy, with all isolates (100%) classified as sensitive, followed closely by oxytetracycline, which exhibited a 98% sensitivity rate. On the other hand, the highest resistance levels were observed with tilmicosin (87.5%) and tylosin (68.75%). Aivlosin displayed a mixed susceptibility profile, with 60.93% of isolates categorized as intermediate, 29.68% as sensitive, and 9.37% as resistant, suggesting a shifting trend in macrolide susceptibility, including Aivlosin.

The established MIC values of Spirulina platensis against MG isolates, ranging from 3.9 to 1,000 μg/mL, indicated broad-spectrum antibacterial activity (Table 2). The principle of MIC determination via broth microdilution is illustrated schematically in Figure 2, which depicts the transition from growth inhibition to bacterial growth across a serial dilution of the extract. Most of the tested strains (>65%) were inhibited at concentrations of 250 μg/mL or less, with several isolates showing notable sensitivity at low MICs (3.9 μg/mL and 15.62 μg/mL). These findings emphasize the significant antimicrobial potential of Spirulina, particularly against strains resistant to conventional antibiotics.

The distribution of MIC values for each treatment group was assessed using the Shapiro–Wilk test prior to comparative analysis. The results indicated that MIC values were not normally distributed across all treatments, with p-values < 0.05 in each case, including doxycycline (W = 0.757), Spirulina platensis (W = 0.678), and tylosin (W = 0.463), confirming significant deviation from normality. As a result, the nonparametric Kruskal-Wallis rank-sum test was used to compare MIC distributions among the six antimicrobial agents. Results from Figure 3 showed a highly significant difference in MIC values among the treatments (χ² = 270.89, df = 5, p < 2.2 × 10⁻¹⁶), with a very large effect size (ε² = 0.85, 95% CI [0.80, 0.89]). This effect size reflects the proportion of variability in MIC values explained by treatment group, without implying differences in absolute antimicrobial potency, which is indicated by the MIC magnitudes.

Following the significant Kruskal-Wallis result, a Dunn’s post hoc test with Bonferroni correction was conducted to identify specific pairwise differences between treatments. The results, summarized in Table 3, confirmed distinct antimicrobial profiles among the agents tested.

While significant differences in MIC magnitude were observed between Spirulina platensis and conventional antibiotics, with higher MIC values indicating lower in vitro potency for Spirulina, its potential advantages may relate to factors other than potency, such as its distinct activity profile or reported safety characteristics. Specifically, Spirulina platensis showed highly significant differences compared to doxycycline (p.adj < 0.001, rᵣᵦ = −0.95, 95% CI [−0.97, −0.92]), oxytetracycline (p.adj < 0.001, rᵣᵦ = −0.88, 95% CI [−0.92, −0.82]) and Aivlosin (p.adj < 0.001, rᵣᵦ = −0.91, 95% CI [−0.94, −0.86]).

Significant differences with moderate to large effect sizes were also observed between several conventional antibiotics. In fact, doxycycline demonstrated significantly lower MICs than oxytetracycline (p.adj < 0.001, rᵣᵦ = −0.55, 95% CI [−0.65, −0.42]), while tilmicosin and tylosin were found to have distinct activity levels (p.adj < 0.01, rᵣᵦ = 0.45, 95% CI [0.20, 0.65]).

Overall, these results indicate that the antimicrobial agents tested, including Spirulina platensis, differed significantly in their activity against M. gallisepticum. The distinct profile of Spirulina platensis, characterized by its significant separation from conventional antibiotics, supports its role as a unique natural alternative worthy of further investigation (Table 3).

Associations between the MIC values (log₁₀-transformed) of the tested antimicrobial agents were assessed. Given the non-normal distribution of the data, both Pearson and Spearman rank correlation analyses were performed. The results revealed consistent patterns between the two methods for most comparisons (Figure 6; Table 4).

A strong, significant positive association was confirmed between tilmicosin and tylosin (Pearson’s r = 0.647, 95% CI [0.477, 0.770]; Spearman’s ρ = 0.661, p < 0.001), consistent with their shared macrolide class. A moderate positive association was also observed between doxycycline and Aivlosin (Pearson’s r = 0.554, 95% CI [0.357, 0.704]; Spearman’s ρ = 0.568, p < 0.001), suggesting potential cross-resistance.

In contrast, Spirulina platensis showed negligible or weak correlations with conventional antibiotics. Most associations were non-significant in both analyses. For example, the correlation between Spirulina platensis and tylosin was negligible (Pearson’s r = −0.009, 95% CI [−0.255, 0.237]; Spearman’s ρ = −0.025, p = 0.845). The strong concordance between Pearson and Spearman coefficients for key antibiotic pairs (e.g., tilmicosin-tylosin, doxycycline-Aivlosin) confirms that the observed correlation structure is robust and not an artifact of distributional assumptions (Table 4).

The PCA was performed to explore phenotypic variability in antimicrobial susceptibility among macrolide-resistant MG isolates. The PCA biplot of MIC data revealed three distinct clusters, each reflecting specific antimicrobial response patterns (Figure 7). Group 1 included isolates with low MICs across all treatments, indicating high susceptibility. Conversely, Groups 2 and 3 contained isolates with moderate and high MICs, respectively, likely representing macrolide-resistant profiles. The relationships among the antimicrobial agents, illustrated in the PCA correlation circle (Figure 8), are shown in Figure 8. Spirulina platensis was positioned orthogonally to tylosin and tilmicosin, indicating statistical independence (r ≈ 0) and suggesting a mechanism different from that of conventional antibiotics. In contrast, doxycycline, oxytetracycline, Aivlosin, and tilmicosin clustered closely together, forming inter-vector angles of less than 45°, indicating strong positive correlations (r > 0.7, p < 0.001) and suggesting potential shared resistance mechanisms.

The individuals’ factor map shows the distribution of isolates in the PCA space (Figure 9), with color intensity indicating cos² values, which represent the degree of contribution of the first two principal components of each isolate. Isolates positioned at the edge of the plot, especially isolates #3, #27, and #28, contributed most significantly to the overall variance and therefore correspond to highly resistant or phenotypically distinct strains. Additionally, Dimension 1 accounted for 30.2% of the variance, while Dimension 2 explained 23.4%, highlighting the complexity and diversity of resistance profiles within the population. Overall, these findings highlight the distinct antimicrobial activity of Spirulina platensis and support its potential as a complementary or exploratory approach in the management of Mycoplasma gallisepticum infections, rather than as an alternative to conventional antibiotics.

The cytotoxicity of the Spirulina platensis extract was assessed on RAW 264.7 cells using the MTT assay. Data are presented as the mean ± standard deviation of three independent biological replicates, each performed with four technical replicates. Over the tested concentration range (125 to 4,000 μg/mL), the extract showed no significant cytotoxic effect, with cell viability consistently exceeding 90% at all concentrations compared to the untreated control (p > 0.05, one-way ANOVA) (Table 5). The effect size for the overall ANOVA was negligible (η² < 0.01). A concentration of 2,000 μg/mL was thus established as a reliable non-toxic threshold in this model.

The calculated selectivity index (SI) values, which ranged from 2 to 512.82 across the 64 Mycoplasma gallisepticum isolates, indicated considerable variation in the therapeutic window. A post hoc power analysis for the key comparison of cell viability at 4000 μg/mL versus the control, based on the observed effect size and variance, confirmed a statistical power > 80% (alpha = 0.05), supporting the robustness of the non-toxicity conclusion (Table 5).

These results confirm the high biocompatibility of Spirulina platensis and its safety at concentrations relevant for antimicrobial testing. The absence of cytotoxicity supports its use in subsequent antibacterial evaluations and underscores its potential as a safe, natural antimicrobial agent.