RESULTS

Physicochemical profile of CUR-loaded NLCs

Dynamic light scattering analysis showed that CUR incorporation increased the mean particle size of the NLC formulation. CUR-NLCs had a mean diameter of 215.6 ± 2.6 nm compared with 177.0 ± 3.2 nm for the unloaded NLC formulation (Table 1). CUR-NLCs also exhibited a more homogeneous size distribution, as indicated by a lower polydispersity index (0.145 ± 0.003) than blank NLCs (0.269 ± 0.034). Surface charge analysis showed moderately negative zeta potentials of −17.5 ± 0.78 mV for CUR-NLCs and −23.12 ± 0.9 mV for blank NLCs. CUR incorporation into the lipid matrix was highly efficient, with an encapsulation efficiency of 98.79 ± 1.51% and a loading capacity of 2.82% (w/w).

Formulation | NLC | CUR-NLC
Size (nm) | 177.0 ± 3.2 | 215.6 ± 2.6
Polydispersity index | 0.269 ± 0.034 | 0.145 ± 0.003
Zeta potential (mV) | −23.12 ± 0.9 | −17.5 ± 0.78
Encapsulation efficiency (%) | ND | 98.79 ± 1.51
Loading capacity (%) | ND | 2.82 ± 0.04

Morphological assessment using transmission electron microscopy confirmed the formation of well-defined, spherical nanoparticles with a uniform appearance (Figures 1A and 1 B). The images indicated nanoscale particles in the expected size range of approximately 200 nm, consistent with dynamic light scattering measurements. These observations were interpreted qualitatively, considering possible effects of sample preparation.

Cumulative release profiles of CUR are shown in Figure 1C. CUR-NLCs exhibited an initial release phase followed by sustained release over 48 h, indicating prolonged release kinetics compared with free CUR, which showed a less sustained release pattern over the same period. Independent-samples t-test analysis demonstrated significant differences in cumulative release between free CUR and CUR-NLC formulations at multiple time points, particularly from 2 to 24 h (p < 0.05). The greatest differences were observed from 4 to 8 h (t = −16.678 to −21.768, p < 0.001), during which CUR-NLCs exhibited markedly greater cumulative release than free CUR. No statistically significant difference was detected at 48 h (p > 0.05).

Figure 1: Morphology and release profile of curcumin encapsulated in nanostructured lipid carriers (CUR-NLCs). (A and B) Transmission electron microscopy images showing the spherical morphology and uniform distribution of CUR-NLCs at scale bars of 200 nm and 100 nm, respectively. (C) In vitro cumulative release profiles of free CUR and CUR-NLCs under simulated physiological conditions over 48 h. CUR-NLCs exhibited an initial release phase followed by sustained-release behavior, indicating efficient encapsulation and controlled-release characteristics. Data are presented as mean ± standard error (n = 3). Statistical comparisons between free CUR and CUR-NLC formulations at corresponding time points were performed using independent-samples t-tests. Significant differences in cumulative release were observed at 0.5 h (p < 0.05) and from 2 to 24 h (p < 0.001), whereas no significant differences were detected at 1 h and 48 h (p > 0.05).

Functional group analysis

Figure 2 illustrates the Fourier-transform infrared spectra of free CUR and CUR-NLCs, and the major absorption peaks and functional group assignments are summarized in Supplementary Table S1. In the CUR-NLC formulation, the characteristic O–H stretching vibration of CUR, observed near 3508 cm⁻¹, appeared shifted and showed lower band intensity. The C=O stretching vibration at approximately 1736 cm⁻¹ also shifted after incorporation into NLCs. Peaks associated with C=C stretching at approximately 1600–1625 cm⁻¹ and aromatic ring vibrations appeared broader and less intense in CUR-NLCs. Additional changes were observed in the 1300–1400 cm⁻¹ region, corresponding to C–O stretching and C–H bending vibrations.

Figure 2: Fourier-transform infrared spectra of curcumin (CUR) and CUR encapsulated in nanostructured lipid carriers (CUR-NLCs). The bottom spectrum represents CUR, and the top spectrum represents CUR-loaded into NLCs. Characteristic shifts and intensity changes in functional group peaks support the interaction of CUR with the lipid matrix, indicating successful encapsulation within the CUR-NLC system.

Antibacterial activity of CUR formulations

Table 2 summarizes the antibacterial activity of CUR formulations against S. agalactiae isolates. CUR-NLCs produced significantly larger IZ values (14–15 mm) than free CUR (6 mm) across all tested isolates (p < 0.05). MIC values were 1000 ppm for both CUR and CUR-NLCs across all isolates, whereas MBC values were 2000 ppm for both formulations.

Parameters | Isolate | CUR | CUR-NLC
IZ (mm) | FPrA02 | 6 ± 0ᵃ | 15 ± 1ᵇ
 | FNA07 | 6 ± 0ᵃ | 14 ± 2ᵇ
 | ENC06 | 6 ± 0ᵃ | 15 ± 2ᵇ
MIC (ppm) | FPrA02 | 1000 | 1000
 | FNA07 | 1000 | 1000
 | ENC06 | 1000 | 1000
MBC (ppm) | FPrA02 | 2000 | 2000
 | FNA07 | 2000 | 2000
 | ENC06 | 2000 | 2000

Growth performance and feed utilization

Growth performance and feed utilization values measured at 30 and 60 days are summarized in Table 3. At day 30, fish fed the CUR-NLC-supplemented diet showed significantly higher FW and WG than the control and NLC groups (p < 0.05). SGR, FI, and PER were also significantly higher in the CUR-NLC group, whereas FCR did not differ significantly among treatments at this stage (p > 0.05).

Time/diets | FW | WG | SGR | FI | FCR | PER | HSI
30 days |  |  |  |  |  |  | 
C | 17.7 ± 1.1ᵃ | 9.1 ± 1.1ᵃ | 2.24 ± 0.19ᵃ | 0.53 ± 0.03ᵃ | 2.67 ± 0.35ᵃ | 0.30 ± 0.04ᵃ | 0.90 ± 0.14ᵃ
CUR | 21.5 ± 1.2ᵃᵇ | 12.9 ± 1.2ᵃ | 2.90 ± 0.18ᵇᶜ | 0.64 ± 0.04ᵃᵇ | 1.94 ± 0.18ᵃ | 0.43 ± 0.04ᵃᵇ | 1.84 ± 0.22ᵇ
CUR-NLC | 24.7 ± 1.5ᵇ | 16.1 ± 1.5ᵇ | 3.34 ± 0.20ᶜ | 0.74 ± 0.04ᵇ | 1.82 ± 0.20ᵃ | 0.54 ± 0.05ᵇ | 1.52 ± 0.17ᵃᵇ
NLC | 18.3 ± 0.7ᵃ | 9.7 ± 0.7ᵃ | 2.44 ± 0.13ᵃᵇ | 0.55 ± 0.02ᵃ | 2.08 ± 0.20ᵃ | 0.32 ± 0.02ᵃ | 1.41 ± 0.15ᵃᵇ
60 days |  |  |  |  |  |  | 
C | 37.9 ± 1.0ᵃ | 29.3 ± 1.0ᵃ | 2.45 ± 0.04ᵃ | 1.14 ± 0.03ᵃ | 1.76 ± 0.06ᵃᵇ | 0.97 ± 0.03ᵃ | 1.57 ± 0.09ᵃ
CUR | 41.1 ± 2.0ᵃ | 32.5 ± 2.0ᵃ | 2.55 ± 0.08ᵃ | 1.23 ± 0.06ᵃ | 1.96 ± 0.13ᵇ | 1.08 ± 0.07ᵃ | 2.06 ± 0.18ᵇ
CUR-NLC | 53.0 ± 1.6ᵇ | 44.4 ± 1.6ᵇ | 3.01 ± 0.05ᵇ | 1.59 ± 0.05ᵇ | 1.64 ± 0.05ᵃ | 1.48 ± 0.05ᵇ | 1.79 ± 0.13ᵃᵇ
NLC | 40.0 ± 1.1ᵃ | 31.4 ± 1.1ᵃ | 2.55 ± 0.04ᵃ | 1.20 ± 0.03ᵃ | 1.73 ± 0.06ᵃ | 1.05 ± 0.04ᵃ | 1.52 ± 0.06ᵃ

After 60 days of feeding, dietary differences became more pronounced. Fish receiving the CUR-NLC diet had significantly higher FW and WG than all other groups (p < 0.05), together with significantly greater FI and PER. SGR was also significantly higher in the CUR-NLC group than in the control and NLC groups (p < 0.05). In contrast, FCR remained statistically comparable among treatments, although numerically lower values were recorded in the CUR-NLC group.

Repeated-measures analysis of variance was performed to evaluate overall temporal and treatment effects. Time had a significant effect on FW, WG, FI, and PER (p < 0.05), indicating progressive changes over the experimental period. In contrast, SGR and FCR were not significantly affected by time (p > 0.05). Significant time × treatment interactions were observed for FW (p = 0.026), WG (p = 0.026), FI (p = 0.029), and PER (p = 0.023), indicating that the magnitude of temporal responses differed among dietary treatments. However, no significant interaction effects were detected for SGR or FCR (p > 0.05), indicating consistent temporal patterns across treatments for these parameters. No significant overall between-subject treatment effects were detected for FW, WG, SGR, FI, FCR, or PER in the repeated-measures model (p > 0.05). Detailed replicate tank-level datasets for growth performance, feed utilization, and HSI analyses are provided in Supplementary Table S2.

HSI response to dietary treatments

HSI values at 30 and 60 days are presented in Table 3. At day 30, HSI differed significantly among treatments, with the CUR group showing the highest value (p < 0.05). A similar pattern was observed at day 60, when fish fed CUR showed significantly higher HSI than the control group, whereas the CUR-NLC and NLC groups showed intermediate values. Repeated-measures analysis of variance confirmed significant effects of time (p = 0.014) and treatment (p = 0.007) on HSI, whereas the interaction between time and treatment was not significant (p = 0.310), indicating a consistent temporal trend across dietary treatments.

Length–weight relationship and growth pattern

Growth pattern analysis after 60 days of feeding demonstrated clear dietary effects on the length-weight relationship of Nile tilapia (Figures 3A–D). Estimation of the allometric growth coefficient revealed the highest b value in fish receiving the CUR-NLC diet (b = 3.278), indicating enhanced positive allometric growth. Lower b values were recorded in fish fed the CUR diet (3.045), followed by the control (2.684) and NLC (2.642) groups.

Regression analysis demonstrated strong relationships between body length and weight across all experimental diets. The CUR-NLC treatment showed the strongest linear relationship, with a correlation coefficient of 0.972 and a coefficient of determination of 0.945. Slightly lower but still strong associations were observed in the CUR group (r = 0.964; R² = 0.930), whereas the control (r = 0.951; R² = 0.905) and NLC (r = 0.947; R² = 0.897) groups showed comparatively weaker relationships. All fitted regression models were highly statistically significant (p < 0.001).

Assessment of Kn condition further supported the dietary effects on growth performance. Kn was highest in fish fed the CUR diet (1.214), followed by fish receiving CUR-NLC supplementation (1.067). In contrast, fish in the control (0.982) and NLC (0.974) groups showed lower Kn values, indicating comparatively reduced condition status. Individual fish length and body weight measurements used for regression and condition factor analyses are provided in Supplementary Table S3.

Post-challenge survival following S. agalactiae infection

Kaplan-Meier survival curves following S. agalactiae ENC06 challenge are presented in Figure 4A. Survival differed significantly among treatment groups according to the log-rank (Mantel-Cox) test (χ²(4) = 40.797, p < 0.01).

Corresponding cumulative mortality, survival, and RPS values are summarized in Table 4. The infected control group showed the highest mortality (62.2 ± 5.9%) and the lowest survival (37.8 ± 5.9%), which differed significantly from the CUR-NLC and negative control groups (p < 0.05). Fish fed the CUR-NLC diet showed the lowest mortality (11.1 ± 2.2%) and highest survival (88.9 ± 2.2%), with an RPS of 82.4 ± 2.8%. Similarly, the negative control group showed low mortality (13.3 ± 3.8%) and high survival (86.7 ± 3.8%), corresponding to an RPS of 77.5 ± 7.1%. Fish fed free CUR showed intermediate responses, with mortality of 33.3 ± 3.8% and survival of 66.7 ± 3.8%, resulting in an RPS of 44.3 ± 11.1%. In contrast, the blank NLC group showed mortality (55.6 ± 4.4%) and survival (44.4 ± 4.4%) values that were not statistically different from those of the infected control group (p > 0.05), with a low RPS of 9.3 ± 11.0%. Daily cumulative survival records for each replicate tank during the post-challenge period are presented in Supplementary Table S4.

Survival risk assessment based on Cox proportional hazards modeling

The Cox proportional hazards model indicated that survival outcomes differed significantly among dietary treatments following bacterial challenge (χ² = 38.960, df = 4, p < 0.001; Figure 4B). The cumulative hazard plot further showed that fish receiving CUR-NLC supplementation maintained the lowest relative mortality risk throughout the post-challenge period compared with all infected treatment groups.

Infected ControlCURCUR-NLCNLCNegative ControlcabAInfected ControlNLCaCURCUR-NLCNegative ControlcbBInfected ControlCURCUR-NLCNLCNegative ControlcabAInfected ControlNLCaCURCUR-NLCNegative ControlcbBInfected ControlCURCUR-NLCNLCNegative ControlcabAInfected ControlNLCaCURCUR-NLCNegative ControlcbBInfected ControlCURCUR-NLCNLCNegative ControlcabAInfected ControlNLCaCURCUR-NLCNegative ControlcbB

Figure 4: Disease resistance responses of Nile tilapia following experimental challenge with Streptococcus agalactiae ENC06. (A) Kaplan-Meier survival curves showing cumulative survival among dietary treatment groups during the 15-day post-challenge period. (B) Cumulative hazard (%) plots derived from Cox proportional hazards regression analysis illustrating relative mortality risk across treatments over time. Different lowercase superscripts indicate statistically significant differences among treatments (p < 0.05).

Treatments | Mortality (%) | Survival (%) | RPS (%)
Infected control | 62.2 ± 5.9ᵃ | 37.8 ± 5.9ᵃ | 0.0
Negative control | 13.3 ± 3.8ᶜ | 86.7 ± 3.8ᶜ | 77.5 ± 7.1ᶜᵈ
CUR | 33.3 ± 3.8ᵇ | 66.7 ± 3.8ᵇ | 44.3 ± 11.1ᵇ
CUR-NLC | 11.1 ± 2.2ᶜ | 88.9 ± 2.2ᶜ | 82.4 ± 2.8ᶜ
NLC | 55.6 ± 4.4ᵃ | 44.4 ± 4.4ᵃ | 9.3 ± 11.0ᵃ

When mortality risk was evaluated relative to the infected control group, fish receiving CUR showed a significantly reduced likelihood of death, corresponding to an estimated hazard reduction of 50% (Exp(B) = 0.500; 95% confidence interval: 0.266-0.940; p = 0.031). A stronger protective effect was observed with the CUR-NLC treatment, with a markedly lower mortality risk (Exp(B) = 0.144; 95% confidence interval: 0.055-0.374; p < 0.001). In contrast, administration of blank NLCs did not alter survival probability compared with the infected control group (Exp(B) = 0.958; p = 0.878). As expected, fish in the uninfected control group showed substantially reduced mortality risk throughout the observation period (Exp(B) = 0.174; 95% confidence interval: 0.072-0.422; p < 0.001).