PLGA (Resomer^® 504 H, 50:50 lactide: glycolide, acid terminal, MW 40.000 Da) was purchased by Sigma-Aldrich Co. Poly Vinyl Alcohol (PVA), Ethyl acetate (EtOAc), Capsaicin, trehalose, Phosphate Buffer Saline (PBS) Tween 20 and Dimethyl sulfoxide (DMSO) were purchased from Sigma-Aldrich Co. (Merk KGaA, St. Louis, MO, United States). Water, acetonitrile and trifluoroacetic acid (TFA) LC-MS grade were from (ROMIL Ltd, Cambridge, UK). Deionized water (18.2 MΩcm) was obtained from a Milli-Q system (Merck Millipore, St. Louis MO, United States).

Cap-loaded PLGA-nanoparticles (PLGA-Cap) were synthesized using the single emulsion solvent evaporation method. The oil phase was prepared by mixing 1 mg of cap (33 uL of the 30 mg/mL stock solution in ethyl acetate) and 5 mg of PLGA (500 μL, 10 mg/mL, dissolved in ethyl acetate) following 2 min of sonication. The oil phase was added to 5 mL of an aqueous PVA solution (0.5% w/v) and the phases were emulsified by sonication with a microtip (Sonifier™ SFX150, Branson Ultrasonics, Emerson Electric Co, St. Louis, MO, United States). Different ultrasonication protocols were tested in order to optimize the synthesis. In particular, time of activation and, percentage of amplitude were combined testing 18 possible combinations. After the ultrasonication step, the nanoparticle solution was left to stir overnight at room temperature. Finally, the nanoparticles were collected and washed twice with milli-Q water by centrifugation at 20,000 rcf and 4°C for 1 hour. Finally, the particles were freeze-dried by Alpha 1-2 LD (Christ, Memmingen, Germany) adding 1.25% w/v of trehalose as cryoprotectant. Empty PLGA particles (PLGA-N) were prepared with the same synthetic procedure. The process yield was calculated after the freeze-drying process as the ratio of the collected NPs to the starting raw materials (Cap and polymers, cryoprotectant).

The hydrodynamic size of particles, the zeta potential, and Poly Dispersion Index (PDI) were measured by DLS (Malvern Zetasizer Nano ZS instrument, 633 nm laser, dispersion angle of 173°) at fixed temperature (25°C) using a material refractive index (RI) of 1.59 with a dispersant RI of 1.33. All the formulations were diluted in milli-Q water (0.3 mg/mL) and measurements were performed with an equilibration time of 120 s for a total of 5 runs.

PLGA particles (Naked and Cap-loaded) and free Cap were analysed by Fourier Transform Infrared Spectroscopy (FTIR). Spectral analysis was carried out in the Spectrum 3 spectrometer (Perkin-Elmer, Inc. Waltham, MA, USA) equipped with a total attenuated reflectance accessory (UATR). The spectra were collected by performing 10 scans with a resolution of 4 cm⁻¹ in the region of 650–4,000 cm⁻¹.

Images of PLGA particles (Naked and Cap-loaded) were collected by using the highly automated Thermo Scientific™ PhenomPharos™ G2 FEG-SEM with the scanning transmission electron microscopy (STEM) detector. Freeze dried particles were solubilized in water, and 5 µL of the solution was deposited on special TEM copper grids with a carbon coated Formvar. After evaporation of the solution, the grid was washed with deionized water, and loaded on the STEM sample holder.

The RP-HPLC analysis of Cap was made by a cromatographic apparatus UltiMate 3,000 DIonex (Thermo-Fisher Scientific) equipped with a C18 BioBasic ™ column (50 × 2,1 mm, 5 μm), thermostat control at 37°C and UV detection at 227 nm. The elutions were performed at 0.2 mL/ min from the mobile phases A (trifluoroacetic acid 0.08% v/v in water) and B (trifluoroacetic acid 0.08% in acetonitrile) with the following gradient referred to the percentage of B: from 30% to 80% in 15 min; from 80% to 90% in 1 min; 90% from 16 to 21 min; from 90% to 30% in 1 min; 30% for 8 min. In this condition, the retention time of CAP was about 8.4 min. The CAP concentrations were calculated in relation to the calibration curve created by analyzing solutions between 1 and 80 μg/mL (Figure 1).

The encapsulation efficiency (EE) and the drug loading (DL) of the PLGA-NPs was assessed via RP-HPLC. Specifically, the particles were dissolved in DMSO at concentration of 5 mg/mL. The solution was centrifuged at 16,000 rcf for 15 min at room temperature, therefore, 5 µL of the supernatant were analysed as described in the previous paragraph.

The encapsulation efficiency (EE) was calculated as: E E % = C i − C f C i ​ * ​ 100 where Ci is the theoretical initial concentration of Cap added and Cf is the free Cap concentration recovered in the supernatant after centrifugation.

The percentage of drug loaded (DL) was calculated as follows: D L % = W e W p ​ * ​ 100 where We is the weight of the encapsulated cap and Wp is the weight of the particles.

The dialysis bag methodology was implemented for assessing the in vitro release study following prior published methodologies (Caputo et al., 2023a; 2023b). Briefly, 1 mg of PLGA-NPs loaded with CAP were added to 100 uL of release buffer (phosphate buffer pH 7.4 supplemented with Tween 20 0.3% w/v). The particle solution was loaded into a microdialysis filters (Pierce™ 96-well Microdialysis Plate, cut-off 3.5 K MWCO), and immersed in 1 mL of the release buffer. The solutions were kept in continuous agitation at 37°C, and at specific time points the external solution was fully recovered and replaced with 1 mL of fresh release buffer. The release test was run for 7 days. Finally, the particles loaded into the microfilter were collected and dissolved in DMSO to quantify the residual molecules. All the recovered aliquots were analysed using RP-HPLC. The results were expressed as mean values +/− standard deviation of three independent measurements.

The uptake of PLGA nanoparticles loaded with the Coumarin 6dye (C6 Loaded PLGA) by HepG2 cells was assessed by Confocal Laser Scanning Microscopy (CLSM) (STELLARIS 8 - Leica Microsystems, Wetzlar, Germany). HepG2 cells were seeded on plate at density of 8 × 10⁴/cm² 24 h before stimulation. At each time point (1 h and 3 h), slides were washed with PBS and fixed with 4% Paraformaldehyde (PFA). The CLSM is equippend with a white light laser tunable between 440 and 790 nm for the excitation, while, a HC PL APO CS2 40x/1.10 water immersion objective was used for image acquisition. Emission signals were acquired with Power HyD detectors. The system was controlled using Leica Application Suite (LAS) X v4.3 software (Leica Microsystems, Wetzlar, Germany). The excitation wavelengths and detection windows for Coumarin 6-loaded nanoparticles were as follows: Hoechst 3,342 (excitation at 405 nm; detection range 420–455 nm) and Coumarin 6 (excitation at 464 nm; detection range 480–550 nm).

All the data shown in this study were presented as the mean ± SD or STDERR of three distinct investigations (unless otherwise indicated). The standard deviation (SD) of the mean is shown by the error bars. Data were statistically analyzed using Origin 2018 (OriginLab, United States) and GraphPad Prism 5.0 (GraphPad Software, Inc., CA, United States).

PLGA nanoparticles were synthesised by single emulsion solvent evaporation method. The different parameters such as polymers concentration, the nature of the Cap and homogenization settings have been investigated in detail to obtain nanoparticles with the best characteristics in terms of dimension, yield, PDI and Cap loading.

In this work, 18 different conditions of Ultrasonication were tested, changing the parameters of amplitude, sonication duration (Time ON) and numbers of repetitions. Since amplitude and time are two important parameters that need to be carefully considered to obtain nanoparticles with optimal properties, we tested different conditions to select the appropriate settings. (Yan et al., 2022). In particular, the amplitude provides the necessary force to break the interface between the two phases formed during the synthesis (Water and Oil), which are immiscible with each other, and allow the formation of particles.

The duration of sonication is also significant as it determines the amount of acoustic energy delivered to the suspension. In addition, setting a non-continuous pulse, or a delay between repeated activations (time ON and OFF) prevents the polymer from overheating. (Ruiz et al., 2022). All tested settings are listed in Table 1.

The optimization of the ultrasonic phase influences the properties of the particles in terms of size and PDI. Once all formulations were characterized by DLS, we report only the data of the particles that have a PDI <0.2 and a size that is in the ideal range of the order of 100 nm (Table 2). For this reason, the sample 1, 3 and 7 were excluded for PDI > 0.2 and sample 2 for size > 300 nm. Nanoparticle yield (Table 2) was assessed post-freeze-drying with 1.25% w/v trehalose. Small particle size hindered pellet formation during washing, causing losses in the supernatant and resulting in low yields. Consequently, samples 8, 9, and 15 were omitted from further analysis due to insufficient yield. The optimized synthesis process resulted in high-quality nanoparticles, characterized by diameters less than 120 nm and a highly uniform size distribution (PDI <0.2), as confirmed by DLS (Figure 2).

In particular, the PLGA-N nanoparticles have a diameter of 117 ± 3.16 nm and Zeta potential of −30.4 ± 0.81 while the PLGA-Cap have a diameter of 96.84 ± 3.95 nm and zeta potential of −26.8 ± 1.7 mV (Figure 2). The variation in the size of the loaded particles compared to the Naked ones is given by the hydrophobic interaction between the hydrophobic Cap molecule inside the core and the hydrophobic chain of polymer (Jäger et al., 2018). Our preparations stand out as they are significantly smaller than all other Cap loaded PLGA NPs reported in the literature. While previous studies have documented Z-average sizes ranging from 145–250 nm (Parashar et al., 2019; Fan et al., 2022; Malewicz et al., 2022) to as large as 3.9 µm (Kim et al., 2011) our samples achieve a markedly smaller size. This reduced dimension is a key advantage, as it directly influences crucial secondary properties such as targeting efficiency, degradation rate, and cellular uptake mechanisms (Sur et al., 2019). The stability of the PLGA-Cap nanoparticles was tested 7 months after production (see Supplementary Figure S1; Supplementary Table S1), indicating that no physicochemical modifications occurred during this long storage period.

Fourier transform infrared spectroscopy (FTIR) was performed on free capsaicin, PLGA-N NPs and PLGA-Cap NPs to further confirm the encapsulation capability of the proposed carrier. As shown in the Figure 3, for the Naked NPs is possible to observe: a sharp peak at 1750 cm^-1 due to the C=O stretching vibration of the ester group of the polymer; a very pronounced peak at 1,088 cm^-1 which corresponds to C-O-C; and a band between 2,850 and 3,000 given by the stretching of C-H. (Zhou et al., 2013). Compared to naked ones, PLGA-Cap NPs have an extra band at 3,324 cm⁻¹ (shifted compared to capsaicin alone which has it at 3,447 cm⁻¹) which refers to an amino group (N-H) and a hydroxyl group. Another amino group is observed at 1,516 cm⁻¹ that is characteristic of Capsaicin. (Mondal et al., 2019). A small peak at 1,629 cm⁻¹ was attributed to olefinic C=O stretch, C=C stretch, and amide II (Giri et al., 2017). All the FTIR results are reported in the Table 3. The presence of analogous bands in the spectra of both free Cap and our PLGA-Cap confirms the successful encapsulation achieved with our formulation.

In order to obtain information about the morphology of the Nanoparticles, STEM analysis was conducted. The results suggest that the PLGA-Cap had a spherical shape as shown in the Figure 4. Encapsulation Efficiency (EE) and the Drug Loading (DL) were quantified by RP-HPLC as describe in the Methods section. The developed method allows the analysis of up to 1 µg of Cap. The different preparations obtained were characterized to obtain the value of EE% and DL%. As shown in the table below (Table 4) the best results were obtained following the last three protocols (samples 16, 17 and 18- the one that were synthetized with 50% of Amplitude). By combining the main characteristics of the formulation such as size, PDI and EE%, we select the synthesis protocol used for sample 17 for all further particle preparations (EE% of 74.41 ± 0.12 and a DL% of 12.40 ± 0.02).

Once the sonication parameters have been chosen, we have optimized incubation, centrifuge conditions and washing times. Regarding the centrifuge condition, the initial centrifuge parameters were 10 min, 16,000 rcf at Room Temperature (RT); the optimization involves increasing the centrifugation speed and time with decreasing the temperature, 20,000 rcf and 4°C for 1 hour (as reported in the method section) in order to break down the particles better. So, we have achieved improved performance and encapsulation efficiency, obtaining an EE% of 95.91 ± 5.77 e DL% 16.23 ± 0.56. This result suggest that we have an improvement of 21% respect our first preparation, as it is shown by Figure 5A.

To the best of our knowledge, our preparation has the higher EE% for Cap loaded in PLGA NPs, compared to what is currently reported in the literature. Actually the Cap-loaded PLGA preparation with the highest EE% is 75% (Fan et al., 2022), so our preparation has an improvement of 21%. Other paper reported PLGA NPs loaded with Cap but the EE% is lower respect our results. As it is shown by the Table 5, in the other papers with Cap loaded in PLGA NPs the EE% does not exceed 50% (Kim et al., 2011; Parashar et al., 2019; Malewicz et al., 2022). Lipid carriers demonstrated a comparable encapsulation efficiency (EE%) to our findings, likely due to their inherent ability to facilitate the internalization of capsaicin, a hydrophilic molecule. However, their particle size was larger than that observed in our samples (Qi et al., 2021; Arunprasert et al., 2022). The importance to have small nanoparticles depends on the passive targeting mediated by the enhanced permeability and retention (EPR) mechanism witch allow the NPs penetration in the cancer cells (Zhou et al., 2017; Dutt et al., 2023).

With the same RP-HPLC method used for EE% estimation, the Cap released from our NPs was evaluated, as described in the methods section, by incubating the PLGA-Cap NPs in the release buffer solution (PBS at pH 7.4 with 0.3% w/v of Tween 20) at 37°C for a week. As seen in Figure 5B, the Cap has a fast release from PLGA-NPs, consistent with prior research on similar preparations. 64% is released by 8 h, 94.65% by 24 h, and near-total release (99.7%) by 48 h, indicating full release within 2 days. The kinetic of release of Cap from PLGA was observed to have the same rate also in other papers where in some cases it takes 24 h to reach 81% (Parashar et al., 2019), but at most 100% of the release is reached in 72 h (Kim et al., 2011; Malewicz et al., 2022).

We also evaluated the in vitro biological effect of our preparations on HepG2 cell line. First, we performed cell viability MTT assays in a range of concentrations of 100–300 μM, taking in account that reported calculated IC₅₀ for Cap at 24 h was very heterogeneous, ranging from 365 μM (Impheng et al., 2014) to 150 μM (Bort et al., 2017), depending on experimental conditions employed. In accordance with previous works, after seeding cells at density 2.9 × 10⁴/cm², we observed an IC₅₀ at 24 h of 265 ± 8 μM for free Cap. We also measured IC₅₀ at 24 h for PLGA-Cap and we found a value of 276 ± 17 μM. These findings indicated that free Cap and PLGA loaded Cap had similarly effects on cell viability. In the attempt to highlight eventual differential cell responses in a sub-toxic range of concentrations, we next performed MTT assays at a low cell density (1.6 × 10⁴/cm²) with Cap concentrations between 5 and 100 μg/mL for 24 h, 48 h, and 72 h (Figure 6). At 24 h, a slight but significant reduction of cell viability was registered only at 100 µM. At 48 h, we observed a slight increase of cell viability, more evident for free Cap. This unexpected finding might be due to a pro-proliferative effect of Cap at low doses, as already reported in other in vitro systems, when a stimulation longer than 24 h was applied (Malagarie-Cazenave et al., 2009; Zhai et al., 2018; Khonglim et al., 2024). At 72 h, the viability-reducing effect predominated: we observed that PLGA-Cap reduced cell viability with a similar trend with respect to free Cap, with a significant reduction in the concentration range 10–100 µM for PLGA-Cap and 5–100 µM for free Cap. On the whole, these data confirmed a similar behaviour of PLGA-Cap with respect to free Cap in exerting an inhibitory effect on HepG2 cell viability.

To better characterized the biological effect of PLGA-Cap preparations we observed nuclei morphology after treatments with doses already employed in the MTT assays, i.e. 10–100 μM, for 24 h, with the aim to observe early nuclear events that could be indicative of cell death induction. Nuclei of HepG2 cells treated for 24 h with PLGA-Cap or free Cap were stained with Hoescht and then visualized by fluorescence microscopy. We observed a slight increase of dysmorphic nuclei for both treatments compared to the respective vehicles (Figure 7), however this increase was significant in the range 25–100 µM for free Cap and in the range 10–100 µM for PLGA-Cap.

Considering that alteration of nuclear morphology together with nuclear fragmentation is an event that occurs in different cells death processes, including apoptosis (Ziegler and Groscurth, 2004), we investigated apoptosis induction by evaluating the level of proteins having a crucial role in apoptosis (Singh et al., 2019), in the same conditions employed for nuclear visualization. Thus, we analysed Bcl2 and Bax proteins, two markers involved in intrinsic pathway of apoptosis. In particular, the Bax/Bcl-2 ratio determines the cell fate in life or death in response to an apoptotic stimulus; an increased Bax/Bcl-2 ratio is central to induce cell death and loss of cell resistance to apoptosis (Vaskivuo et al., 2002). Western blot analyses revealed that in HepG2 cells, PLGA-Cap or free Cap were both able to slightly increase Bax level and reduce Bcl2 level. As a consequence, the Bax/Bcl-2 ratio increased for both treatments (Figure 8), suggesting that not only free Cap, but also PLGA-Cap similarly was inducing apoptosis.

To obtain an accurate quantitative measurement of apoptosis induction, we also evaluated the enzymatic activity of caspase 3, an enzyme crucial in apoptosis programmed cell death (Porter and Jänicke, 1999). Given the high sensibility of the method to detect caspase 3 activity, we were able to observe a significant caspase 3 activation after 6 h of treatment with both free Cap and PLGA-Cap. We registered an effect that was higher at low concentration of free Cap, 10 µM and 25 μM, respect to concentrations of 50 μM and 100 µM (Figure 9A). In the presence of PLGA-Cap, caspase 3 activity increased at 10 µM of PLGA-Cap and maintained the same activation level at higher concentrations (25, 50, and 100 µM) (Figure 9A). Furthermore, at 100 µM we detected a significant higher increase of caspase 3 activity of 40% in PLGA-Cap respect to cells treated with free Cap. To support data from caspase 3 activity assay, we analysed, by western blot, how the cleaved and active form of caspase 3 accumulated into HepG2 cells. In this case, the treatment was prolonged tor 48 h to better visualize the caspase 3 protein level. We observed an increase of cleaved caspase 3 protein level both with PLGA-Cap and free Cap (Figure 9B); at 100 µM this increase was significantly higher of 46.7% in samples treated with PLGA-Cap than in samples treated with free Cap (Figure 9B), according with caspase 3 activity. Finally, we were interested also to evaluate another desirable property of the encapsulated Cap as potential anti-cancer drug, i.e., the ability to modulate autophagy. Indeed, several works reported that Cap was able to induce autophagy in several cancer cell lines, including HepG2 cells, at concentrations around 100 μM (Chen et al., 2016). Our data (Supplementary Figure S3) suggested that, at the experimental conditions at which we observed apoptosis, we also registered a moderate autophagic response in PLGA-Cap-treated cells.

Considering that apoptotic induction in vitro by Cap is often mediated by ROS generation (Hacioglu, 2022; Zhang, R et al., 2008; Joung et al., 2007), we evaluated ROS level in presence of PLGA-Cap and free Cap. HepG2 were treated for 6 h with PLGA-Cap or free Cap and then ROS levels were analyzed by measure DCF fluorescence in Hepg2 cells stained with DCFDA. HepG2 increased DCFH-DA fluorescence accumulation as conseguence of ROS production in presence of both PLGA-Cap and free Cap (Figure 10); at the highest concentration tested (100 µM), ROS accumulation was significantly more induced in PLGA-Cap than in free Cap samples. In particular, we observed an increase of 58% at 100 µM of ROS levels in PLGA-Cap respect to free Cap samples. These findings indicated that an efficient ROS generation could be at the bases of apoptosis induction in HepG2-Cap-treated cells.

Very often, to improve the bioavailability, stability and pharmacokinetics of Cap, it is encapsulated in NPs of different nature and size. To assess the anticancer effects of these Cap formulations, their action on the apoptotic pathway involving Bax/Bcl2, caspase-3 expression and oxidative stress is frequently analysed (Merritt et al., 2022). By western blot it was found that Cap encapsulated in PLGA-PEG-NP increased caspase 3 levels in carcinoma lung cells isolated from rats (Parashar et al., 2019). Chitosan nanocapsules containing Cap also activated caspase 3 in T24 bladder cancer cells (von Palubitzki et al., 2020). Moreover, works by Elkholi et al. (E Elkholi, 2014) and Hazem et al. (Hazem et al., 2021) revealed that Cap loaded in trimethyl chitosan nanoparticles (TMC-NP) triggered apoptosis in HepG2 cells. Specifically, the researchers demonstrated that this formulation reduced the expression of the pro-survival gene Bcl-2 and increased that of the pro-apoptotic gene Bax. Furthermore, immunocytochemistry highlighted strong staining for caspase3 in cells treated with Cap-TMC-NP. Apoptosis was more evident when cells were treated with Cap-TMC-NP (100 µM) than with free Cap. (E Elkholi, 2014; Hazem et al., 2021). In HepG2 cells capsaicin loaded solid lipid nanoparticles (SLNs) induces apoptosis by ROS generation and loss mitochondrial membrane potential. (Kunjiappan et al., 2021). Overall, these results are in agreement with the biological effects induced by our preparation of PLGA-Cap, which increased the expression and activity of caspase 3, ROS generation and modulates the Bax/Bcl-2 ratio. Considering that PLGA nanoparticles have the ability to increase stability of loaded compounds and protect them from degradation (Rezvantalab et al., 2018; Yang et al., 2024), our results suggested that PLGA nano particles might have the ability to improve the effect of free Cap, possibly shielding Cap molecules from rapid cellular metabolization and inactivation processes.

Finally, to further confirm the cellular uptake and intracellular transport capabilities of our PLGA nanoparticles, we investigated the intracellular distribution of C6-loaded fluorescent PLGA particles in HepG2 cells using CLSM (Vanni S et al., 2025). Fluorescence microscopy revealed rapid and substantial internalization of the PLGA nanoparticles, evidenced by a strong green fluorescent signal distributed throughout the cytoplasm within just 1 hour of incubation (Figure 11; Supplementary Figure S2).