Preparation of the NLC dispersions produced two formulations with distinct visual appearances under the non-homogenized and homogenized conditions (Elkhateeb et al., 2023). The non-homogenized dispersion exhibited lower visual uniformity than the homogenized dispersion. The homogenized dispersion appeared more visually uniform than the non-homogenized sample (Kumari et al., 2026).

Morphological characterization of the prepared NLC dispersions was performed using scanning electron microscopy (SEM). The SEM micrographs revealed particulate structures with irregular morphology and a tendency to form aggregates rather than discrete, uniformly separated spherical nanoparticles (Figure 4).

The presence of particle clusters and uneven surface features suggests that agglomeration may have occurred during sample preparation, drying, or imaging, which is commonly observed in electron microscopy analysis of colloidal lipid-based systems (Bresch et al., 2022). These observations are consistent with the PSA results, which indicated the presence of larger particle populations or aggregates in the dispersion. Therefore, the SEM results support the formation of particulate structures in the prepared dispersion, but they do not yet provide sufficient evidence to confirm a fully uniform and physically stable NLC system. Further optimization of the formulation process and additional stability evaluations, including zeta potential measurement and storage stability testing, are required to confirm colloidal stability and reduce aggregation (Chauhan et al., 2020).

Candidate ligands for molecular docking against STMN1 (stathmin) were curated from Dr. Duke’s Phytochemical and Ethnobotanical Database and matched with 3D structures from PubChem prior to docking simulations to evaluate binding affinity and identify interacting residues ((Duke, 2016); (Trott & Olson, 2010)). A phosphate group was incorporated into the STMN1 model because STMN1 is regulated by phosphorylation, and phosphorylation at key serine residues  is known to modulate stathmin’s microtubule-related activity in cancer-relevant contexts (Liu et al., 2023). The ligand collection and the STMN1 model are illustrated in Figure 5, whereas the docked poses within the binding pocket and their interaction representations together with the phosphorylated STMN1 visualization are shown in Figure 6.

Docking analysis showed that the control ligand GDP exhibited the strongest binding affinity (−10.6 kcal/mol), whereas among the tested compounds, sinapic acid (−7.5 kcal/mol) and ferulic acid (−7.4 kcal/mol) demonstrated higher affinities than protocatechuic acid (−6.6kcal/mol) (Table 1). Recurrent interactions involving residues such as GLN11, CYS12, SER138, GLY142, and THR143 suggested a relatively conserved binding region within STMN1 (Zeng et al., 2024). Overall, sinapic acid and ferulic acid emerged as the most promising candidate ligands, although their predicted binding affinities remained lower than that of the control ligand GDP ((Papagiouvannis et al., 2024); (Theodosis-Nobelos et al., 2023))

Peritoneal macrophages isolated from BALB/c mice produced a cell suspension suitable for microscopy-based analysis. The coverslip preparation provided a denser and more clearly distinguishable cell distribution than the non-coverslip condition (Figure 7). Hemocytometer counting showed a cell density of 1.89 × 10⁶ cells/mL, confirming that the isolated cells were adequate for subsequent morphological observation and the latex-bead phagocytosis assay(Green & Sambrook, 2019).

Microscopic examination of 10% Giemsa-stained macrophages at 400× magnification clearly distinguished the nuclei and cytoplasm, enabling reliable cell identification (Velez-Hoyos & Jimenez-tobon, 2022). Representative micrographs (Figure 8) showed a dose-dependent increase in latex bead uptake, with the highest bead density observed in the 10 µL group. The observed cell size (approximately 15–26 µm) was consistent with the typical morphology of macrophages (Fakoya et al., 2025).

Phagocytic capacity differed among treatment groups. The control (0 µL), 5 µL, and 10 µL groups exhibited phagocytic capacities of 33.5%, 37%, and 91%, respectively (Figure 9). Games–Howell analysis showed no significant difference between the 0 and 5 µL groups (p = 0.143), whereas significant differences were observed between the 0 and 10 µL groups and between the 5 and 10 µL groups (both p < 0.001) (Table 2).

The macrophage phagocytic index exhibited a sharper change than phagocytic capacity, capturing not only “how many cells are active” but also “how intensively” the participating cells internalize particles. The macrophage phagocytic index increased from 0.63 in the control group (0 µL) to 0.77 at 5 µL and rose markedly to 4.50 at 10 µL (Figure 10), indicating a dose-dependent enhancement in phagocytic activity. Games–Howell post hoc analysis confirmed significant differences among all treatment groups (p < 0.05) (Table 3). Although the 5 µL treatment produced only a modest increase, the 10 µL dose resulted in a substantially greater enhancement of phagocytic activity.

Candidate ligands curated from phytochemical databases and literature sources were docked to STMN1, and the resulting binding affinities and interaction profiles were evaluated as a first-pass in silico screen of predicted STMN1 interaction (Trott & Olson, 2010). Docking outcomes showed that the GDP control ligand produced the strongest predicted affinity (−10.6 kcal/mol) and formed multiple short-range contacts within the binding pocket, serving as a benchmark for comparison (Table 1). Among the tested phenolic acids, sinapic acid (−7.5 kcal/mol) and ferulic acid (−7.4 kcal/mol) showed the most favorable predicted binding affinities, whereas protocatechuic acid (−6.6 kcal/mol) showed weaker predicted binding (Table 1). Although the candidate ligands did not surpass the control, the affinity ordering suggests a possible trend in which substituted phenolic acids with more extensive functionalization may show more favorable predicted interactions within the STMN1 pocket under the present docking conditions (Zeng et al., 2024).

Interaction mapping further supports this ranking by highlighting recurring residues and plausible contact chemistry. Several residues appeared repeatedly across ligand–STMN1 complexes (e.g., SER138, GLY142, THR143, and GLN11/CYS12), suggesting that multiple ligands may occupy a relatively consistent predicted binding region of STMN1 (Table 1). Collectively, the docking results indicate that sinapic acid and ferulic acid showed the most favorable predicted interactions among the tested candidates and therefore represent the strongest in silico candidates for further STMN1-oriented evaluation in this study. At the same time, docking remains a predictive approach; recent evaluations emphasize that scoring functions and sampling limitations require cautious interpretation and that experimental validation is needed before concluding biochemical inhibition (Agu et al., 2023).

In vitro testing was conducted to evaluate whether the synthesized NLC dispersion could modulate macrophage phagocytosis as an early functional readout related to innate immune response in the tumor microenvironment. Peritoneal macrophages isolated from BALB/c mice produced a cell suspension suitable for microscopy-based assays, and the coverslip preparation enabled clearer observation than the non-coverslip condition  (Figure 7) ((Jesus et al., 2022); (Khatua et al., 2022)). Giemsa-stained micrographs showed distinguishable nuclei and cytoplasm, as well as latex beads associated with macrophages across treatment groups, supporting the feasibility of quantitative phagocytosis analysis (Figure 8) (Jesus et al., 2022)).

The biological response should be considered in relation to the physicochemical characteristics of the synthesized NLC system. Visual observation showed differences between homogenized and non-homogenized preparations (Figure 1), while PSA/DLS analysis indicated particle-size variation and moderate dispersity (Figure 2 and Figure 3). SEM observation also revealed irregular particulate structures with aggregation tendency (Figure 4). These findings suggest that the NLC system delivered to macrophages consisted of lipid-based particulate entities whose dispersion state may influence cell-particle interaction, membrane contact, and bead uptake response.

Quantitative analysis showed that the 10 µL treatment produced the strongest enhancement of macrophage phagocytic activity. Phagocytic capacity increased from 33.5% (0 µL) to 37% (5 µL) and 91% (10 µL), with significant differences between 10 µL and both lower groups (p < 0.001), while the 0 µL and 5 µL groups did not differ significantly (Table 2). The phagocytic index also increased from 0.63 to 0.77 and 4.50, respectively, with all pairwise comparisons showing significant differences (p < 0.05) (Table 3). These results indicate that the 10 µL NLC treatment increased both the proportion of active macrophages and the intensity of bead internalization, providing evidence of enhanced macrophage functional activity ((Riezk et al., 2025); (Ueno et al., 2021)).

However, these findings should not be interpreted as direct evidence of anti-metastatic activity or M2 macrophage inhibition, since macrophage polarization markers such as CD206, CD163, Arg-1, IL-10, iNOS, and CD86 were not assessed. Further studies using marker-based assays and macrophage-cancer cell co-culture or migration assays are required to confirm phenotype-specific and anti-metastatic relevance ((Lim et al., 2024); (Jeitler et al., 2024)).