Photoresponsive polymersomes for nanoencapsulation of multiple cargoes as a potential theranostic strategy

Abstract

Inspired by the cell's membrane architecture, self-assembling amphiphilic copolymers in polymersomes can form biomimetic, compartmentalized, bilayered, and versatile structures through supramolecular interactions, enabling the simultaneous co-encapsulation of hydrophilic and hydrophobic cargo. This approach protects cargo from the surrounding media and modulates cargo release via stimuli-responsive mechanisms, such as light. This work reports on a photosensitive polymersome derived from an amphiphilic random copolymer based on poly(ethylene-alt-maleic anhydride) and a 2-nitrobenzyl alcohol light-responsive moiety. Fourier-transform infrared spectroscopy, magnetic nuclear resonance spectroscopy, and thermal analysis were used to characterize the resulting amphiphilic copolymer. UV-light-responsive polymersomes were successfully assembled with a size of 80.38 ± 1.57 nm, a ζ-potential of −50.9 ± 0.8 mV, and a bilayer thickness of 3.5 ± 1.2 nm, as confirmed by cryo- and transmission-electron microscopy. Moreover, it assembled biotinylated polymersomes with similar physicochemical properties for the targeted delivery of cargo to cancer cells. It encapsulated 5-fluorouracil (5-FU) and rhodamine-B (Rh-B) into polymersomes with high encapsulation efficiency and loading capacity as cargo models of different natures, and gold nanoparticles and magnetic nanoparticles/5-FU as a potential theranostic strategy. Polymersomes demonstrated high biocompatibility, and the encapsulated 5-FU exerted cytotoxicity after 24 h of treatment following 5 minutes of UV-triggered cargo release, positioning them as stimuli-responsive nanosystems for electromagnetic irradiation-triggered drug delivery.

---

Introduction

Polymersomes are self-assembled supramolecular organizations that result from intermolecular interactions and mimic the functions of permeable cell membranes, constituting an exquisite example of a biomimetic nanostructure.^1–3 In this sense, supramolecular chemistry and its intricate assemblies have drawn inspiration from biological systems and functions, as the cellular bilayer originates from the self-assembly of biological components, such as lipids. Therefore, mimicking natural processes, such as cell compartmentalization, is a helpful strategy for co-encapsulating hydrophilic and hydrophobic cargo in their core and bilayer membrane, among many other applications⁴ Beyond the compartmentalized structures' ability to protect therapeutic cargo and deliver it specifically, their functionality is enabled by the inclusion of specific moieties that respond to external stimuli, such as light, which enable “on-demand” cargo release.⁵ The o-nitrobenzyl derivatives (ONB) have been employed as a photolinker in photo-triggered polymeric nanostructures, generating irreversible photoactivation due to efficient cleavable molecules with good leaving groups.⁵ Since the first use of 2-nitrobenzyl alcohol moieties as photolabile groups in polymer chemistry,⁶ numerous strategies have been reported for photo-stimulated cargo release.^7–11 ONB photocleaves upon one-photon absorption of about 310–340 nm with subsequent transient intermediates from fragmentation reactions disturbing the bilayer membrane of the respective polymersomes, causing disruption and reorganization into smaller architectures with the subsequent release of the loaded components.^9,12–14 Therefore, the photorelease destabilization mechanisms can be attributed to: (i) variations in the solubility and swelling behavior as a consequence of crosslinking processes,¹¹ (ii) the partial degradation of polymer conformation and solubilization, destabilizing the structure,¹⁴ (iii) the disintegration of the polymeric structure when the ONB group is located in the polymer backbone,¹⁵ among others. In addition to the photo-mediated release mechanism of polymersomes containing ONB, the integration of inorganic materials, such as upconversion particles, gold nanoparticles (AuNPs), and magnetic nanoparticles (MNPs), has been explored to enhance polymersome properties. These inorganic materials offer complementary features, including upconversion, photothermal effects,¹⁵ magnetic-triggered cargo release,¹⁶ and enhanced imaging contrast.^16,17 As a result, organic/inorganic nanoassemblies combine the biocompatibility, biodegradability, and chemical versatility of organic materials with the enhanced magnetic, electronic, and optical properties of inorganic materials, offering higher complexity but better performance.^18,19 Moreover, this integration enables the synergistic enhancement of organic and inorganic constituents in a unique system, with potential applications in diagnosis and cancer therapy. Furthermore, cutting-edge polymeric-inorganic nanoassemblies have been increasingly used to develop theranostic strategies that integrate tumor-targeted diagnosis, imaging, and therapy into a single system. For tumor-targeted diagnosis and therapy, a site-oriented approach is essential. Surface functionalization of polymersomes plays a key role in improving the effectiveness and efficiency of targeted therapeutic approaches. Functionalization involves selecting suitable linkers and conjugation methods that depend on the polymers and ligands.²⁰ Furthermore, the successful targeting of polymersomes toward cancer cells or tissues relies on particle size, surface coverage, and surrounding compatibility, among other factors. For example, cancer cells overexpress biotin receptors, leading to increased biotin uptake in rapidly growing tumor cells compared to normal cells. Therefore, this and many other ligands have been used for tumor targeting in the design of theranostic strategies that utilize different cargo.^21,22 Herein, we report the synthesis and characterization of poly(ethylene-alt-maleic anhydride) and 2-nitrobenzyl alcohol (PEMA-r-NBA) amphiphilic copolymers via a one-step nucleophilic substitution reaction, which modifies the PEMA backbone with the NBA light-responsive moiety. Scheme S1 illustrates the amphiphilic random copolymer PEMA-r-NBA, with an appropriate hydrophilic/hydrophobic ratio, self-assembled into polymersomes with a photo-stimulated response upon UV exposure. We successfully encapsulated hydrophilic cargos (i.e., 5-fluorouracil (5-FU), rhodamine-B (Rh-B), AuNPs) and hydrophobic iron oxide magnetic nanoparticles (Fe₃O₄ MNPs), demonstrating the versatility of polymersomes for potential theranostic applications. Although different cargoes have been encapsulated into photosensitive polymersomes elsewhere,^23–25 this is the first time cargo encapsulation into PEMA-r-NBA-based polymersomes. Then, it triggered the subsequent release of 5-FU by time-dependent UV irradiation. 5-FU and Fe₃O₄ MNPs were then co-encapsulated into the polymersomes, confirming the versatility and dual encapsulation capabilities of the vesicle-like structures. Moreover, biotin functionalization before self-assembly broadened the scope of a targeted cancer drug delivery system. Finally, the biocompatibility of bare and biotin-functionalized polymersomes was investigated in vitro using healthy and colon cancer cells, demonstrating their biocompatibility. Overall, the encapsulation of various cargos suggests the potential of polymersomes for thermodynamic and photodynamic therapy and imaging, which aligns with theranostic approaches.

Experimental section

Supplementary information (SI) details the reagents, solutions, and synthesis, assembly, and characterization of amphiphilic photosensitive copolymers.

Results and discussion

Physicochemical characterization of PEMA-r-NBA and biotin functionalized-PEMA-r-NBA copolymers

The nitrobenzyl (NB) moieties, such as ONB and NBA derivatives, are among the most well-known photolabile groups employed in synthetic organic chemistry and light-induced polymeric materials.²⁶ The light irradiation over NB moieties at a specific wavelength triggered a series of photochemical reactions initiated by reactive intermediates, which subsequently led to cleavage reactions. NB groups can be cleaved at wavelengths around 300–365 nm within minutes of irradiation exposure, depending on the intensity of the light source,²⁷ thereby constituting a versatile moiety for photoprotection and photo-triggered reactions in polymer-based materials. Herein, it synthesized NBA-based copolymers via the one-step nucleophilic substitution of alcohol-containing molecules in the PEMA backbone (Scheme 1). This reaction was made possible by the high reactivity of PEMA, which is attributed to its two structural units: succinic anhydride and ethane. While succinic anhydride exhibits high reactivity through nucleophilic substitution, ethane groups between adjacent anhydride rings reduce steric constraints and boost cyclic anhydride reactivity. As a result, the reaction between PEMA and NBA does not require the use of coupling agents. It simplifies purification processes and increases reaction efficiency, which are limited only by the equilibrium conditions governing water generation as a byproduct. Furthermore, the presence of multiple succinic anhydride units per polymer chain allows the chemical linkage of multiple side moieties with different chemical functionalities.²⁸ The successful nucleophilic substitution reaction of NBA in the PEMA copolymer (Fig. S1A) was confirmed by ¹H NMR. In Fig. S1A, the chemical shift at δ 2.30–1.19 ppm (m, 4H, A, B) is related to the two pairs of protons in each –CH₂– of the PEMA backbone, whereas the signal at δ 3.05 ppm (m, 2H, C, D) is consistent with the two protons bonded to the two –CH– of the anhydride ring.²⁸ For PEMA-r-NBA copolymer (Fig. S1B), the signal at δ 12.39 ppm (s, 1H, E, F, G) corresponds to the proton of the –COOH group after the anhydride ring aperture by substitution reaction or hydrolysis. The chemical shifts at δ 8.45–7.22 ppm (m, 4H, I, J, K) are related to the protons in the aromatic NBA ring and at δ 5.38 ppm (m, 2H, H), to the two protons in the –CH₂– of the formed ester in the nitrobenzyl-ester group.^10,28 According to eqn (S1), the degree of substitution was 27% for NBA in the PEMA-r-NBA copolymer. Therefore, the NBA reacted with succinic anhydride units to produce ethane-(2-nitrobenzylsuccinate) units. Additionally, the production of water as a byproduct enables the hydrolysis of succinic anhydride, yielding ethane units (succinic acid). Furthermore, 1H NMR confirmed the functionalization of the amphiphilic polymer with the biotin linker. Hence, in Fig. S1C, the chemical shifts at δ 6.4 ppm (m, 2H, B') correspond to two protons related to the protons in the two NH–C=O–NH groups of the biotin-ligand heterocycle. The chemical shifts between δ 4.0–4.5 ppm (m, 2H, A') correspond to the protons of the two –CH of the biotin-ligand heterocycle, while the chemical shift at δ 2.30–1.19 ppm (s, 18H, M, N, O, P, Q, S, T, U, W) correspond to the protons in each –CH₂– of the linker, which overlap with protons A and B of the PEMA backbone. It confirms the presence of biotin in the biotin-functionalized PEMA-r-NBA copolymer.²⁹ The substitution degree from biotin-linker functionalization was 3.1%, according to eqn (S2). Similar results were obtained by Quinchia et al., with a 3.5% substitution degree of the biotin-linker in a PEMA backbone modified with aminoazobenzene and functionalized with a hydrophilic amine-PEG3-biotin moiety.³⁰ Due to the random distribution along the copolymeric backbone, a dispersity in the molecular weight and density of the side moieties is expected for both copolymers. The attenuated Total Reflectance Fourier-transform infrared spectroscopy (ATR FT-IR) identified copolymer functional groups, providing a relevant approach related to the copolymer modification,³¹ as was evidenced in the ATR FT-IR spectra for PEMA (a), NBA (b), and PEMA-r-NBA copolymer (c) (Fig. S1Da–c). In Fig. S1D-a, the PEMA polymer presented two bands at 2938 and 2872 cm⁻¹ corresponding to the asymmetric and symmetric stretching vibrations of C_sp³–H bond from the methyl group. Also, a strong double stretching band belonging to the C=O bond from the anhydride appeared centered at 1845 cm⁻¹ and 1770 cm⁻¹.^32,33 In addition, typical bands corresponding to the cyclic compound's deformation vibrations are also evident in the 900–650 cm⁻¹ region. Regarding the NBA compound (Fig. S1D-b), a strong and wide band corresponding to the O–H tension vibration appeared centered at 3302 cm⁻¹. In addition, two strong bands related to the C–NO₂ and C–O stretching vibrations were evident at 1512 and 1040 cm⁻¹, respectively.³⁴ Fig. S1D-c shows the typical vibrations from the carboxylic acid and ester groups. The strong band at 1717 cm⁻¹ corresponds to the C=O bond from the COOH group, whereas the band centered at 1620 cm⁻¹ of the C=O bond is related to the ester formed after the nucleophilic substitution reaction between PEMA and NBA polymers.³⁵ Thermogravimetric analysis (TGA) was employed to investigate the thermal stability of the copolymer via degradation processes, including main-chain and side-group scissions, elimination reactions, depolymerization, and crosslinking.³⁵ The thermal behavior of PEMA shows two main thermal events at 298.41 and 433.18 °C (Fig. S1E-a). It has been reported that the first process involves the decomposition of the anhydride cycle, whereas the second involves the degradation of the remaining PEMA chain components.^35–37 The NBA thermogram (Fig. S1E-b) is consistent with previous reports, indicating that the main mass loss occurs at the minimum in the DTG curve, corresponding to 97.67% weight loss due to sample degradation.³⁸ The obtained PEMA-r-NBA polymer exhibits a thermal behavior that combines the performance of the two precursors, from which three significant weight losses were determined (Fig. S1E-c). The first thermal event at 122.64 °C indicates a loss of water due to the dehydration of the two adjacent carboxylic acids, leading to the formation of cyclic anhydrides.²⁸ The second weight loss at 249.88 °C appeared from the thermal decomposition of the pendant 2-nitrobenzyl ester group, whereas the third thermal event after 400 °C was due to the decomposition of aliphatic bridged chains.^39–41 Thermal parameters, including T_onset, T_max, and residual mass at 600 °C, are summarized in Table S1.

Scheme 1 (A) Schematic representation of the amphiphilic random copolymer PEMA-r-NBA and the resulting self-assembly into polymersomes. The scheme is not to scale. (B) Reaction scheme of the one-step nucleophilic substitution of NBA in the PEMA backbone for the synthesis of PEMA-r-NBA. “n” represents the degree of substitution of NBA, deriving the ethane-(2-nitrobenzylsuccinate) units, and “x–n” represents the residual ethane-(succinic acid) units obtained by hydrolysis of the non-reactant succinic anhydride. (C) Photoreaction of ortho-nitro benzyl ester from PEMA-r-NBA amphiphilic copolymer and the resulting leaving groups (nitrobenzaldehyde and –COOH) after UV irradiation. Adapted from Romano et al.⁴³

Characterization of empty bare polymersomes

The successful assembly of control nanopolymersomes (CNP) from PEMA-r-NBA was first screened through the measurement of the hydrodynamic diameter (D_H) with a mean diameter of 80.38 ± 1.57 nm and a dispersity (Đ) of 0.30 (Table S2). The negative ζ-potential around −50 mV confirmed the presence of hydrophilic carboxylic acid (–COOH) and carboxylate (–COO⁻) groups on the surface of the CNPs, as expected (Scheme S1), and correlates with the high stability of the photosensitive CNP by electrostatic repulsion.⁴² The successful obtention of quasispheric NPs was confirmed via negatively stained transmission electron microscopy (TEM) (Fig. 1A). Single particles were majoritarian identified with a D_TEM of 44.7 ± 1.2 nm (Fig. 1B, n = 120), but a few aggregated vesicles were also evident as a product of the hydrogen bond and ion-dipole between –COOH and COO⁻ groups on the CNP surface. These results were also corroborated by analysis of the CNP using bright-field scanning transmission electron microscopy (STEM) micrographs, which indicated a D_STEM of 42.5 ± 1.1 nm (n = 400), with a log-normal distribution (Fig. 1C and D). The D_H from dynamic light scattering (DLS) analysis was higher than that from DTEM and DSTEM due to the typical solvated corona of Particles in an aqueous dispersion and the absence of sonication before measurement (aggregation state),⁴⁴ as well as the lack of an appropriate autocorrelated radius-dependent scattering intensity. In this sense, particles with a smaller radius (R) than the wavelength of radiation (2R < λ/10) scatter in the Rayleigh limit where scattered intensity scales by particle radius to the sixth power (I ∼ R⁶).^44,45 The membrane was evident in TEM micrographs (Fig. 1E and inset), with an average thickness of 3.5 ± 1.2 nm (Fig. 1F, n = 120). This result aligns with other reported polymersomes, which have a bilayer thickness ranging from 2 to 50 nm, depending on the polymer's molecular weight or chain length.⁴⁶ The obtained membrane thickness was comparable with an amphiphilic triblock photoresponsive self-reducible polymersomes incorporating a photo-locked o-nitrobenzyl and L-cystine into the same polymer “unit”, with an average wall thickness of around 7.0 nm.⁴⁷ Moreover, cryo-TEM analysis employed in our system also confirmed the existence of bilayered polymeric vesicles smaller than 100 nm (Fig. S2). The stability of polymersomes is affected by the media under different conditions. For example, the media may impact the shape, size, and permeability of polymersomes.⁴⁸ Therefore, it studied the structural changes of different assembled polymersomes in various media to gain an insight into their properties, especially by mimicking the physiological environment and envisioning different applications. Hence, considering one of the three properties of the Helfrich model of curvature,⁴⁹ namely mean surface curvature, spontaneous curvature, and bending rigidity, the last one is related and governed by the chemical features of the polymersome membrane, such as polymer chain length and composition, and also by the surrounding features. Then, it investigated polymersome stability in terms of size distribution across various media, including buffer solutions (as shown in Fig. 1G), and summarized the results in Table S2. The DLS was measured after 15 min of stabilization in each medium. As previously mentioned, the –COOH and –COO⁻ of CNP interact with surrounding water via hydrogen bonds and ion-dipole interactions in aqueous media with a mid-acidic condition (pH 3.0–4.0). When lowering the pH to 1.4 in 0.5 M citric acid solution (CA, pH outside the buffer pH range 2.2–6.5), the mean D_H of CNPs increased around six-fold from 80.38 to 491.20 nm. This behavior can be attributed to the protonation of the –COO group, resulting in the formation of the –COOH group. Hence, COOH groups on CA could interact with the –COOH groups of CNP, forming hydrogen bonds and increasing CNP D_Hvia swelling. A previous study employed hyperbranched polyesters with succinic anhydride carboxyl-terminated vesicles to evaluate the pH responsiveness of the structures. It showed a similar effect on particle size that increased 10-fold after decreasing the pH from 3.0 to 2.0.⁵⁰ This behavior was also reported with different polymers with hydrophilic groups, such as the pH-induced “breathing” poly(ethylene oxide)₄₅-block-polyesterene₁₃₀-block-poly(2-diethyl aminoethyl-methacrylate)₁₂₀ vesicles when pH shifted from 10.4 to 3.5. The structures were protonated and hydrated at low pH, thus swelling. The effect of low pH was also associated with disruption of the vesicle wall, resulting in the highest permeability.⁵¹ Regarding the stability of the CNP in 0.1 M citrate buffer (CB) pH 4.0, it was found that the CNPs size increased moderately (112.30 nm) as compared to H₂O, as shown in Table S2. Unlike the non-buffered 0.5 M CA pH 1.4, where citric acid is 97.9% protonated (calculated using the protonation bundle from ChemAxon), promoting acid–base reactions between CA and the –COOH and –COO– groups on the surface of CNPs with a possible pH change, the presence of both protonated and deprotonated species in the 0.1 M CB pH 4.0, within the buffer solution pH range (3.0–6.2) minimizes the pH change by acid–base reactions and enhances the stability of the CNPs.⁵² Considering the performance of CNP in 1X PB pH 5.4 (0.01 M phosphate) and 1X PBS pH 7.4 (0.137 M NaCl, 0.0027 M KCl, and 0.01 M phosphate), the change in D_H with 1X PB pH 5.4 was negligible, keeping the size of CNP almost the same (82.66 nm) as the CNP in water (80.38 nm). Meanwhile, in 1X PBS pH 7.4, the size change was most significant (268.50 nm) due to the chloride ions (Cl⁻) and the increase in pH, which led to swelling. In this context, Cl⁻ ions could interact with the partially positively charged –CH₂⁵³ group adjacent to the ester bond in the PEMA-r-NBA copolymer via ion-dipole interactions, increasing the D_H. In addition, weakly hydrated Cl⁻ ions act as chaotropic ions in the Hofmeister series, contributing to a “salting-in” effect.^54–56 Hence, Cl⁻ ions act as destabilizing agents, increasing osmotic pressure and leading to the notorious increase in D_H.³⁰ This behavior was smaller, as confirmed by solutions containing only Cl⁻ ions as anions, i.e., 0.1 M NaCl and KCl (Fig. 1G and Table S2). Therefore, it is suggested that the absence of chlorides, combined with the presence of highly hydrated phosphate ions in 1X PB at pH 5.4, could maintain the polymersomes' size and keep the copolymeric chains joined,^49,59 thereby stabilizing over time and preventing the early release of encapsulated cargo. The effect of salts on polymersome stability is crucial for further analyzing the impact of pH. Considering the pK_a of COOH to be around 4.0–5.0, it is coincidentally noted that CNPs in the media with a COOH pK_a in this range had a similar size to the original size in H₂O. Thus, as shown in Table S2, a similar size trend, observed with 0.1 M CB at pH 4.0, was also evident in 1X PBS at pH 5.4. Furthermore, although 1X PBS pH 7.4 is a widely employed medium that simulates physiological conditions for biological model applications, the results evidenced a significant change in size, as mentioned before. Despite being closer to the CNP in H₂O, the size changed from 80.38 to 108.20 nm in 1X PBS pH 5.4. This behavior could imply an unexpected cargo release due to osmotic pressure and swelling of the polymeric structure induced by the buffer components. In this sense, results from 1X PBS pH 7.4 and 1X PB pH 5.4 suggest that they could be used in applications requiring rapid cargo release. It evaluated CNP's size and dispersity stability in H₂O, pH 3.0–4.0, 1X PB pH 5.4, and 1X PBS pH 7.4 over time. Fig. 1H–J presents stability monitoring during six weeks at 4 °C. In the case of H₂O, the size evidenced a low variation over time, oscillating between 80.38 and 71.75 nm, in agreement with Li et al.⁵⁷ A similar uniform behavior concerning the Đ parameter was evident except for the fourth week, when an unexpected increase in the Đ tendency was attributed to an unreliable measurement (blank point with blue periphery), so it was not considered in the fitting spline-line. In the case of 1X PB pH 5.4, the observed performance was similar to that of H₂O, exhibiting high stability until the sixth week, when an increase from 77.18 nm (in the fifth week) to 89.63 nm was observed. Concerning the Đ value, the evolution from week 0 to 6 increased from 0.27 to 0.33. As previously discussed, in 1X PBS pH 7.4, with respect to size changes, a significant increase in CNP size was evident, reaching a maximum of 289.00 nm and a Đ of about 0.38 after six weeks of storage. Therefore, water and 1X PB pH 5.4 media had negligible effects on the CNP D_H and Đ for approximately 5–6 weeks due to non-destructive osmotic and/or ionic stabilizing interactions. Moreover, the hydrodynamic radius (R_H), the radius of gyration (R_g), and their ratio (R_H/R_g) were evaluated for CNP in three media, pH 3.0–4.0 H₂O, 1X PB pH 5.4, and 1X PBS pH 7.4, after the sixth week. Hence, the R_H was determined by DLS, yielding particle sizes of 40.19, 41.33, and 164.20 nm, respectively. Additionally, R_g was estimated from the slope of the K_c/R_θ and sin²(θ/2) axis in the Guinier plots (Fig. 1L and M), resulting in 40.72, 42.16, and 169.98 nm in size in H₂O pH 3.0–4.0, 1X PB pH 5.4, and 1X PBS pH 7.4, respectively. The (R_H/R_g) ratio for CNPs in H₂O pH 3.0–4.0, 1X PB pH 5.4, and 1X PBS pH 7.4 was 1.0 for the three media, confirming a consistent polymersome morphology.⁵⁸

Fig. 1 (A) Negative-stained TEM micrograph of CNP in H₂O pH 3.0–4.0, at 15 000X. (B) CNP size distribution from TEM analysis (n = 200). (C) Negative-stained STEM micrograph of CNP in H₂O, pH 3.0–4.0, at 200 000X. (D) CNP size distribution from STEM analysis (n = 400). (E) Negative-stained TEM micrograph of CNP in H₂O, pH 3.0–4.0, at 97 000X, and the contrast of the hydrophobic membrane (inset). (F) Average membrane thickness distribution of CNP from TEM analysis (n = 120). (G) Stability of CNP in different media (CA: citric acid solution, CB: citrate buffer, PB: phosphate buffer containing Na₂HPO₄ and KH₂PO₄, PBS: phosphate buffer saline containing Na₂HPO₄, KH₂PO₄, NaCl, and KCl). Stability evaluation of CNP based on D_H and Đ measurements for six weeks in (H) H₂O pH 3.0–4.0, (I) 1X PB pH 5.4, and (J) 1X PBS pH 7.4. Guinier plots of CNP in (K) H₂O, pH 3.0–4.0, (L) 1X PB, pH 5.4, and (M) 1X PBS, pH 7.4.

UV-light stimulated photolysis

As previously reported, the expected mechanism of the nitrobenzyl–ester bond photocleavage involves an oxygen-transfer reaction of the aromatic nitro compound, which has a C–H bond in the ortho position. Furthermore, the reaction products result from the replacement of the C–H and nitro groups with C–OH and NO functionalities, respectively.^26,59,60 The final products correspond to 2-nitrosobenzaldehyde and carboxylic acid groups.⁶⁰ In the proposed system, the NBA moiety constitutes the photosensitive hydrophobic group responsible for the photo rupture of the nitrobenzyl ester bond, leading to polymersome destabilization and subsequent cargo release (Scheme 1). The kinetics of PEMA-r-NBA cleavage were determined by monitoring the products by ultraviolet-visible (UV-vis) spectroscopy. Fig. 2 exhibits the absorbance spectra of CNP from 0 to 300 s of UV irradiation at λ = 365 nm in H₂O pH 3.0–4.0, 1X PB pH 5.4, and 1X PBS pH 7.4 (Fig. 2A–C), showing the decrease of the NBA absorbance band at ∼265 nm and the subsequent irradiation time-dependent band centered at ∼310 nm corresponding to the formation of 2-nitrosobenzaldehyde subproduct. Therefore, two isosbestic points were observed at approximately 250 and 280 nm. The one at 250 nm decreased, while the one at 280 nm increased. Also, an increase in the maximum absorbance at 310 nm has been previously reported as evidence of photocleavage. Fig. S5A–C shows the extent of photorupture increases as a function of UV irradiation time. For example, CNP showed up to 95, 71, and 52% photocleavage in H₂O at pH 3.0–4.0, 1X PB at pH 5.4, and 1X PBS at pH 7.4, respectively. Therefore, as Cl⁻ ions of 1X PBS 7.4 interact with the copolymer chain, it is hypothesized that they provide a “stacking” effect over the structures in comparison to H₂O pH 3.0–4.0 and 1X PB pH 5.4. The photo-rupture is expected in the nitrobenzyl ester bond that connects the hydrophilic backbone to the NBA hydrophobic aromatic ring. In this context, the CNP size was measured in the three media at various irradiation times, as presented in Fig. 2D–F. An increase in the D_H was evident for the CNP across the three media, being more notable for pH 3.0–4.0 H₂O and 1X PB pH 5.4 than for 1X PBS pH 7.4 from 0 s to 300 s of UV irradiation. However, in comparison to water with a D_H increase of 150.00 nm between 0 and 300 s of UV irradiation, it is proposed that salts present in 1X PB pH 5.4 and 1X PBS pH 7.4 contributed to maintaining a moderate increase of the particle size through an attenuating effect evidenced with the lower D_H evolution of 70.00 and 22.00 nm in 1X PB pH 5.4 and 1X PBS Ph 7.4, respectively. Therefore, the hypothesized “stacking” effect across the structures explains the stabilization of CNP size and dispersity. In contrast, pH 3.0–4.0 H₂O and 1X PB 5.4, as chloride-free media, allow the reconfiguration of copolymeric chains under UV light until they reach an apparent stabilized size. The TEM micrographs in Fig. 2H and I show the resulting structures after exposure to two extreme UV times, i.e., 15 and 30 min, in water at pH 3.0–4.0. The results indicate an increase in polymersome diameter from D_TEM 44.6 ± 1.1 to 105.1 ± 1.2 nm and membrane thickness from 3.5 ± 1.2 to 19.1 ± 2.8 nm (n = 40) for CNP upon 30 min UV irradiation. The results are consistent regarding the structures of the polymersomes after photoirradiation. For example, it has been demonstrated that photocleavable polymersomes are based on PCL-PEG diblock copolymers bearing cleavable 2-nitrobenzyl alanine at the hydrophobic/hydrophilic interface. Therefore, hollow structures were maintained, and hydrophobic shells were thickened after irradiation.⁸ Moreover, Fig. 2H–I shows broken vesicles and shapeless aggregates, in agreement with previous authors, who found a similar behavior after 25 min of UV irradiation (λ = 365 nm), but tiny micelles and aggregates after 90 min of UV irradiation.¹⁴ Our polymersomes agree with the reported “linker-type” photoresponsive polymersomes based on NBA. First, the photocleavable moiety is introduced as a hydrophilic–hydrophobic block linker.⁶¹ Once the polymersomes are photo-stimulated, they can maintain their hollow conformation and even increase their size. However, they eventually collapse, thicken, and break down into smaller micelles due to polymersome decomposition and the formation of charged groups from photolysis.

Fig. 2 Photoresponsive behavior of CNP from 0 to 300 s of UV irradiation at 365 nm in (A) pH 3.0–4.0 H₂O, (B) 1X PB pH 5.4, and (C) 1X PBS pH 7.4, with the appearance of the characteristic band from the 2-nitrosobenzaldehyde subproduct after photolysis around 310 nm. D_H variation as a function of irradiation time in (D) pH 3.0–4.0 H₂O, (E) 1X PB pH 5.4, and (F) 1X PBS pH 7.4. (G) Schematic representation of the CNP after photolysis with disrupted membranes and smaller structures, such as micelles, resulting from the reorganization of photoruptured subproducts. Negative-stained TEM micrographs of CNPs regarding the remaining disrupted bilayers and degraded structures after (H) 15 and (I) 30 min UV irradiation in pH 3.0–4.0 H₂O. 5-FU release extent from CNP as drug delivery upon exposure to 365 nm UV light irradiation for 5 and 15 min in (J) pH 3.0–4.0 H₂O, and (K) 1X PB, pH 5.4. The fitting models corresponded to a Boltzmann fit (R-squared: 0.99 for UV 5 and 0.98 for UV 15) for pH 3.0–4.0 H₂O, and to a Boltzmann and LangmuirEXT1 fit (R-squared: 0.99 for UV 5 and 0.98 for UV 15) for 1X PB at pH 5.4. The models were related to a Logistic (R-squared: 0.85) and a Gram-Charlier (GCAS) fitting (R-squared: 0.77) for UV 0 in H₂O (pH 3.0–4.0) and 1X PB (pH 5.4) media, respectively.

Potential of cargo-encapsulated photosensitive polymersomes for UV-light stimulated release and theranostics

The successful encapsulation of Rh-B as a positively charged molecule and 5-FU as a neutral-charged anticancer agent was confirmed. Table S3 shows the properties of Rh-B P at 2.0 and 8.0% and 5-FU P at 10.0 and 20.0% (nominal percent). Whereas the size and dispersity of encapsulated Rh-B P into polymersomes increased for the two extents, the size of encapsulated 5-FU was almost the same as CNP's. The dispersity and ζ-potential of loaded polymersomes were close to those of the CNP. Additionally, both extents of the hydrophilic Rh-B dye encapsulated within the polymersomes' centric cavities exhibit similar encapsulation efficiencies of over 21%. However, a significant improvement in loading capacity (LC) was observed from Rh-B (2.0%) P to Rh-B (8.0%) P, increasing from 0.84 to 3.08% LC. The encapsulation efficiency (EE) and LC of 5-FU into polymersomes at two extents are also shown in Table S3. The 14.20% EE obtained for 5-FUP (20.0%) encapsulated into polymersomes was consistent with that from 5-FU encapsulated into block copolymers containing tertiary amine and caged carboxyl comonomers with a loading efficiency of up to 8.7%.¹² Overall, EE and LC of cargo encapsulated in photosensitive polymersomes increased as the encapsulation process began with more cargo. However, the EE and LC of loaded 5-FU were lower than those of loaded Rh-B. Yet, please note that 5-FU-loaded NPs were smaller than Rh-B-loaded NPs. Furthermore, Rh-B is positively charged due to the =N + –(CH₂CH₃)₂ group, which favors electrostatic interactions with –COOH and –COO– groups from the polymersomes, thereby increasing the loading extent compared to neutral 5-FU. In this sense, electrostatic interactions are hypothesized to enhance dye retention both inside and outside the polymersomes. The 5-Fluorouracil (20.0%) polymersomes (5-FU-P) were evaluated to quantify the 5-FU release as a function of UV (λ = 365 nm) exposure time, aiming to study the UV-stimulated release of hydrophilic cargo. NB derivatives may cause irreversible structural changes after light irradiation.⁶² Thus, releasing small molecules depending on the position of the NB moiety, i.e., either as a side chain or as part of the polymer backbone.⁶³ Therefore, the 2-nitrobenzyl ester bond in the PEMA-r-NBA copolymer is expected to undergo photorupture after UV irradiation in the ultraviolet region at 365 nm. Hence, it studied the 5-FU release following its absorption after passing through the dialysis membrane, as detailed in the SI, upon exposure to 0-, 5-, and 15-min UV irradiation at the same intensity, as shown in Fig. 2J and K. The irradiation times of 5 and 15 min for the release of the hydrophilic molecule from the aqueous compartment of the polymersome were selected as the maximum exposure times that showed photorupture (Fig. 2), within the overall duration range of a superficial radiation therapy (SRT) session.^64–66 In this context, a maximum of 73% 5-FU was released after 120 min in pH 3.0–4.0 H₂O (Fig. 2J) and 97% in 1X PB pH 5.4 (Fig. 2K) upon both 5- and 15-min UV-irradiation time—5-FU under light stimuli, monitored by changes in the maximum absorbance intensity in the media. Time 0 min is the instant at which the irradiation ends. Therefore, 5-FU reached a maximum release of 10.0–15.0% in water and 1X PB at pH 5.4 without irradiation (UV 0 min), suggesting the influence of the media during the release process. Therefore, UV exposure accelerated the release of 5-FU. Thus, in the release process, the photorupture phenomenon involves various stages. Hence, upon UV light stimulation, the ONB chromophore is excited to a higher-energy state, followed by electron transfer, rearrangement, and release of the subproduct. Therefore, the destabilization process occurs in various stages, leading from a moderate, initially rapid release to an increasingly photolytic rate and subsequent incremental drug release due to swelling and partial degradation of the polymer chains.⁶⁷ These results suggest a potential application of SRT for treating basal and squamous cell carcinomas, depending on exposure time and light source wavelength. AuNPs and Fe₃O₄ MNPs were encapsulated into polymersomes as potential photothermal and magnetic contrast agents, respectively, to demonstrate a proof-of-concept theranostic approach. Fig. 3A presents a scheme of the expected localization of hydrophilic AuNPs and hydrophobic MNPs in the core and bilayer membrane of the polymersome, respectively. The successful encapsulation of the AuNPs in the hydrophilic core was confirmed in Fig. 3B with a slightly clear polymeric structure (polymersome) surrounding the contrasting AuNPs (D_p ave 16.6 ± 1.3 nm (n = 32), Fig. S6) while maintaining approximately the original CNP size, 42.5 ± 1.6 nm (n = 120). In this sense, the polymersome-gold nanoparticle encapsulation ratio was about 1 : 1 due to the large size of the gold nanoparticles relative to the polymersome. It has been widely reported that AuNPs show a photothermal effect. Gold nanostructures exhibit effective local heating upon excitation of surface plasmon oscillations, making them appealing photothermal agents for cancer therapy. Regarding medical applications, gold nanostructures have been used to induce hyperthermia and destroy cancer cells, as well as to trigger drug release from temperature-sensitive carriers.^68–70 Hence, photothermal therapy is a potential non-invasive cancer treatment that combines nanomedicine with laser therapy. Hyperthermia is a medical treatment in which body tissues are subjected to a slightly higher temperature than normal to damage and destroy cancer cells.⁷¹ In this sense, the effective encapsulation of AuNPs into polymersomes is a suitable alternative for dual-therapy systems, providing a simultaneous structure for photothermal agents and active chemotherapeutic components, as previously reported.^72,73 Furthermore, the hydrophobic MNPs (D_p ave 7.72 ± 1.28 nm (n = 500), Fig. S7) were encapsulated into the hydrophobic membrane as expected (Fig. 3C). The sample was characterized by TEM before interaction with a 1T magnet, resulting in some localized MNPs spilling out of the polymeric structure. However, a general distribution of MNPs across the membrane was evident, as confirmed by higher-magnification images (Fig. 3D).

Fig. 3 Potential of cargo-encapsulated polymersomes for theranostics. (A) Schematic representation of the hydrophilic AuNPs and hydrophobic Fe₃O₄ MNPs localized in the core and bilayer membrane of the polymersome, respectively. TEM micrographs of (B) AuNP encapsulated into a polymersome core at 19 500X, and MNPs encapsulated into polymersomes with a tendency to be accommodated into the hydrophobic bilayer at (C) 9900X and (D) 71 000X. Magnetic susceptibility of (E) MNPs, (F) MNPs/5-FU P, and (G) the visual appearance of MNPs/5-FU P after four days of interaction with a 1.0 T magnet. Visual appearance and schematic representation of the lack of interaction between f-CNP, the strept-MBs, and the biotin-mediated interaction forming the strep-HRP-f-CNP–MBs magnetoconjugate (f-CNP). (H) The absorbance at 652 nm of buffer media (1X PB pH 5.4), non-functionalized CNP (C−), AuNPs coupled to a probe with a biotin-terminal strand as a positive control (C+), and (I) the incremental extents of biotin-linker in the modified copolymer-based polymersomes.

Herein, the membrane was elongated to accommodate the MNPs inside with a subsequent increase in the membrane diameter of about 14 nm from 3.5 ± 1.2 to 18.3 ± 1.3 nm for empty CNP and MNPs P (Fig. 3D). In this sense, magnetic hysteresis measurements (Fig. 3E and F) also evaluated the magnetic saturation of the MNPs alone (Fig. 3E) and when co-encapsulated with 5-FU into the polymersomes (MNPs/5-FU P) (Fig. 3F). The results showed magnetic saturation of about 46.0 and 24.0 emu g⁻¹ for the MNPs and MNPs/5-FU P, respectively, with no magnetic remanence. The lower saturation magnetization of the MNPs compared to reported systems, around 80.0–90.0 emu g⁻¹, is attributed to oleic acid physically adsorbed on the MNPs.⁷⁴ The coercivity field values analyzed from magnetization (M) vs. coercivity (H) curves for MNPs and MNPs/5-FU P were also negligible (Fig. 3E and F inset). Magnetic polymersomes have been previously assembled by encapsulating the MNP within the polymersome membrane. In this context, remote switching has previously been reported by controlling the electric field via electroporation of the polymeric structure. Magnetic polymersomes have been evaluated and shown excellent magnetic resonance (MR) contrast performance, making them potential theranostic delivery vehicles.¹⁶ In our system, the visual effect of a magnetic field from a 1 T magnet on the MNPs P was confirmed after a partial separation of MNPs/5-FU P (Fig. 3G inset) from the aqueous media after 4 days of interaction (Fig. 3G). These results suggest that the MNPs/5-FU P presented a superparamagnetic-like behavior despite decreased magnetic saturation compared to the MNPs alone due to an isolating-like effect from the polymeric vesicle.^75,76 A complementary analysis using Mössbauer spectroscopy is required to elucidate this behavior in depth. However, the obtained results surpass the magnetization levels of typical magnetic nanocomposites relevant to MR imaging, ranging from 1.0 to 10.0 emu g⁻¹ (USPIONs), when an external magnetic field gradient is applied.⁸¹ Other studies have also coined the term “magnetopolymersomes” to refer to vesicles that encapsulate MNPs and enable their concentration using large magnetic fields. Concerning AuNPs and magnetic nanoparticles, the proof-of-concept demonstrated the polymersome platform's versatility for encapsulating both hydrophobic and hydrophilic inorganic nanoparticles. However, despite the promising results obtained with encapsulating inorganic AuNPs and MNPs, it is important to note that UV irradiation has a low tissue penetration depth (<0.1 cm) due to its absorption by the skin, blood, and tissues.⁷⁷ In contrast, NIR irradiation displays negligible phototoxicity and can penetrate more deeply into biological tissue (from 0.1 cm up to 10 cm). NIR enhanced penetration capacity is related to hemoglobin, water, and lipids, which have low absorption in the NIR region.⁷⁸ Therefore, UCNPs, exhibiting an ascendant conversion of photons, constitute a suitable strategy for absorbing NIR light and emitting UV light in their surroundings. In this sense, the encapsulation of inorganic UCNPs will be a promising alternative, overcoming the limited penetration rate of UV light.^79,80 On the other hand, a biotinylated group permeable to the cell membrane was linked to the amphiphilic polymer during PEMA-r-NBA synthesis as a prior functionalization step, with four increasing extents of the biotinylated linker serving as the tumor-targeting moiety. Thus, it was expected that the NBA and biotin linker would randomly react with succinic anhydride units to produce ethane-(succinic acid), ethane-(o-nitrobenzylsuccinate), and ethane-(o-(biotin)-succinamic acid) units. Once the functionalized copolymers were assembled into biotin-functionalized polymersomes (f-CNP), their size, dispersity, and surface charge were analyzed. The sizes for 2.5, 5.0, 10.0, and 15.0% of nominal biotin functionalization were 119.40, 94.04, 106.10, and 122.40 nm, respectively, with dispersities around 0.3, and a ζ-potential between −51.90 and −59.85 ± 7.99 mV. The results suggest a slight increase in the D_H relative to the CNP, whereas the dispersity and ζ-potential remained similar to those of the bare CNP. Moreover, it implemented a homemade biotin surface test through the colorimetric detection of the streptavidin–biotin interaction to demonstrate the presence of biotin ligands on polymersome surfaces (Fig. 3H). Fig. 3I shows an increase in signal regarding the maximum absorbance of the bioconjugate at 652 nm as the biotin linker content increases. Therefore, it demonstrated the proper functionalization of polymersomes with the biotin linker.

Biological assays

In vitro cytotoxicity of empty (CNP) and functionalized (f-CNP) NPs. The biocompatibility and cytotoxic effects of CNP, f-CNP, unfunctionalized 5-FUP NPs (5-FUP), and functionalized 5-FUP NPs (f-5FUP) were evaluated using an MTT assay with three cell lines: one healthy (CHOK1), one colorectal cancer cell line (SW420), and one metastatic colon cancer cell line (SW620). After 24 h of treatment, both empty CNP and f-CNP showed no cytotoxicity in the three cell lines at any evaluated concentration, as previously reported by other authors.⁸² According to Anderson-Darling and Shapiro-Wilk tests, the data have a normal distribution (p > 0.05), as shown in Fig. 4. In CHOK1 cells, there was no statistically significant difference in viability compared to the control cells at concentrations up to 300 µg mL⁻¹, equivalent to the copolymer concentrations for CNP and f-CNP, except at 150 µg mL⁻¹ (p > 0.05). However, the decreasing nutrient concentration due to the increasing water content as the concentration of polymersomes increased exerted a slight effect on the viability of the SW480 and SW620 cell lines (p < 0.05) compared with the control cells and the solvent control. Yet, they were viable in all polymersome concentrations. Therefore, it is suggested that the polymersome system is safe for cells and that dilution of the culture media has an apparent effect, decreasing nutrient availability.⁸³ However, all the treatments presented a viability over 70%, in accordance with ISO 10993-583, confirming the empty vesicles were not cytotoxic to the cells after 24 h of exposure, in agreement with the literature. For example, non-cytotoxicity has also been observed with 0.1 mg mL⁻¹ PEMA electrospun fibers on mesenchymal cells after 24 h.⁷⁷ Cell viability of MCF-7 cells reached over 80% by treatment with up to 400 µg mL⁻¹ UV-responsive o-nitrobenzyl molecule-based liposomes. Furthermore, maleic anhydride copolymers, including PEMA, have been used as platforms for molecular biosurface engineering.⁸⁴

Fig. 4 Cytotoxicity test of empty CNP (yellow) and f-CNP (green) interacting with (A) healthy CHOK1 and (B) SW480 and (C) SW620 colon cancer cell lines after 24 h of treatment. (CC: control cell, SC: solvent control - ultrapure water, and C+: 10.0% v/v of DMSO in water). All tests were performed twice by quintupling. The black line indicates that all treatments had a cell viability above 70%. (*p = 0.05, **p = 0.01, and ***p = 0.001).

Photo-induced anticancer activity of f-5FUP

The photo-induced anticancer activity of the f-CNP, 5-FUP, and f-5FUP treatments was evaluated after 24 h at 0 and 5 min of UV-light exposure, as shown in Fig. 5. According to the ANOVA test, the treatments had a significant effect on cell viability (p < 0.05). The test was applied to compare pairs of treatments with or without light exposure for the three cell lines. A negligible effect of UV light was observed for the three cell lines, compared to the control cells. Also, a significant statistical effect was observed for f-CNP (p < 0.05) in CHOK1 cells upon UV irradiation, suggesting that photorupture subproducts, such as nitrosobenzaldehyde, may be toxic to cells.⁸⁵

Fig. 5 Cytotoxicity test of SW480 and SW620 colon cancer cell lines and healthy CHOK1 cell line after 24 h of f-CP, 5-FUP, and f-5FUP treatment with 0 (WUV) or 5 min (UV) of UV light exposure. All tests were accomplished twice by quintupling (*p = 0.05, **p = 0.01A and ***p = 0.001).

Viability of the CHOK1 cell line decreased to 30% for WUV-5-FUP, indicating the higher susceptibility of the healthy cell line to the 5-FU anticancer agent than the SW480 and SW620 cancer cell lines. The effect of 5-FU on CHOK1 is mediated by RNA cytotoxicity; i.e., 5-FU exerts its cytotoxicity in CHOK1 by incorporating into RNA, altering processes such as splicing and translation.⁸⁶ The results also indicated a relatively lower toxic effect of WUV-5-FUP and UV-5FUP in SW480 and SW620 cancer lines compared to that in healthy cells, which might be related to the resistant behavior typical of cancer cell lines. In previous studies, both cell lines showed resistance to 5-FU at concentrations ranging from 0.13 to 100 µM.⁸⁷ Additionally, the IC₅₀ of 5-FU for both SW480 and SW620 was defined in previous studies, with values of 10.3 and 446.6 µM for SW480 and SW620, respectively, indicating a higher resistance in the metastatic cell line (SW620).⁸⁸

Moreover, the results showed that the presence of the biotin linker in the formulations plays a key role in reducing 5-FU toxicity in the three cell lines before UV irradiation.²⁰ The higher cell viabilities for each cell line corresponded to NPs functionalized with biotin. It may be related to the fact that biotin enhances cell proliferation by regulating the transcription and expression of various proteins.^89,90 Furthermore, it is noteworthy that WUV-f-5F shows higher cell viability than WUV-5FU across the three cell lines. This behaviour might be related to the slowdown in cargo release caused by the biotin-coated NPs. It is hypothesized that the presence of the hydrophobic biotin linker exposed on the outer surface of the polymersomes forms an additional layer that hinders the release of the hydrophilic 5-FU. Therefore, some specific targeted strategies, such as the functionalization of nanoparticles with the epithelial cell adhesion molecule (EpCAM) gene,⁹¹ the cell surface protein mucin 1 (MUC1),⁹² or the epidermal growth factor receptor (EGFR),⁹³ need to be explored to provide more efficient routes for targeting cancer cell lines. A significant effect of light exposure (p < 0.05) was observed for UV-f-CNP and UV-f-5FUP in both SW480 and SW620, with decreased cell viability in the UV-exposed treatments compared to the same treatments without UV exposure (WUV-f-CNP and WUV-f-5FUP). These results reinforce the notion that the photorupture subproducts may be toxic to cell lines; therefore, it would be mandatory to site-direct the treatment specifically to cancer cells. Although the UV-5-FUP did not show a reduction in cell viability compared to WUV-5-FUP in the three cell lines, cell viability was lower compared to the control cells. Furthermore, both treatments had a dramatic effect on CHOK1 cell viability, demonstrating greater susceptibility to 5-FU chemotherapy than colon cancer cells.⁹⁴ Remarkably, the UV-f-5FUP treatment led to a significant decrease in the cell viability for SW480 and SW620 (p <0.05) in comparison to the concomitant counterpart without UV exposure, (but not in the CHOK1 cell line), demonstrating the synergy between the 5-FU and the photo-rupture subproducts and highlighting the need for a ligand at the outermost NP surface to direct the treatment to the cancer cells specifically. Overall, the results revealed that the effect of 5-FU was worse on the healthy cell line than on the cancer lines due to the sensitivity of CHOK1. The system is light-sensitive, as demonstrated by decreased cell viability for UV-f-CNP in the three cell lines due to photorupture subproducts, and specific targeting strategies are needed to site-direct the treatments. Furthermore, additional studies are required to elucidate the mechanisms of polymersome endocytosis.

Conclusions

A novel amphiphilic random copolymer was synthesized using a simple substitution reaction, and the resultant amphiphilic PEMA-r-NBA copolymer was successfully used to assemble photosensitive nanopolymersomes. It studied the stability, physicochemical performance, and photoresponsive features of the resultant polymersomes. Two hydrophilic cargoes, Rh-B dye and 5-FU, were encapsulated with an EE of up to 21.84 and 14.20%, respectively. Polymersome-encapsulated 5-FU was stimulated by UV laser irradiation, reaching a maximum of 97% cargo release induced by the photo-rupture and reorganization of nanopolymersomes in 1X PB pH 5.4. The results also revealed that hydrophobic MNPs and hydrophilic AuNPs were encapsulated in the corresponding polymersomes' hydrophobic and hydrophilic regions, opening the path for theranostic applications. The biotin ligand was successfully anchored to the copolymeric backbone at the polymersome's outermost surface, demonstrating the polymersome's amenability for functionalization. The system is a versatile, compartmentalized nanostructure synthesized via a novel one-step copolymerization process, exhibiting tunable behavior in response to the surrounding media. It can encapsulate hydrophilic charged or neutral cargos and inorganic NPs, with potential theranostic applications. Overall, the results revealed the innocuous effects of UV alone and of the polymersome nanocarriers without UV irradiation on the three cell lines evaluated. In addition, the UV mechanism was effective, offering opportunities for photo-stimulated drug delivery. However, several factors need to be studied in depth to exploit the proposed nanosystem fully. For instance, the effect of 5-FU encapsulated into polymersomes over longer times, the impact of photo-triggered reaction subproducts that could be used synergistically with chemotherapeutics, and the functionalization of NPs for site-directed therapeutics against cancer cell lines.

Author contributions

Elisa Hernandez Becerra: conceptualization, methodology, formal analysis, investigation, data curation, writing – original draft. Jennifer Quinchia: conceptualization, methodology, formal analysis, investigation, data curation, writing – original draft. Maritza Londoño: methodology, formal analysis, and data curation. Marlon Osorio: methodology, technical support, review & editing. Giuseppe Battaglia and José Muñoz López: methodology and formal analysis. Cristina Castro: project administration, funding acquisition. Jahir Orozco: conceptualization, formal analysis, writing – review & editing, supervision, project administration, funding acquisition.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information: ¹H-NMR and FTIR spectra and assignments, TGA and DTG thermograms, intensity-average D_H distribution, UV-vis absorption spectra, (cryo)TEM and STEM micrographs, calibration curves, and cytotoxicity graphics are in the SI (PDF). See DOI: https://doi.org/10.1039/d5tb02112j.

Acknowledgements

The work has been funded by MINCIENCIAS, MINEDUCACIÓN, MINCIT, and ICETEX through the Program Ecosistema Científico Cod. FP44842-211-2018, project number 58536. J. O. thanks the support from The University of Antioquia and the Max Planck Society through the cooperation agreement 566-1, 2014. In addition, we thank the Ruta N complex and EPM for hosting the Max Planck Tandem Groups. We thank Prof. Daniel Ruiz-Molina from the Institut Català de Nanociència i Nanotecnologia (ICN2-CSIC) for the Cryo-TEM analysis.

References

1. C. LoPresti, H. Lomas, M. Massignani, T. Smart and G. Battaglia, J. Mater. Chem., 2009, 19, 3576–3590.
2. A. K. Sharma, P. Prasher, A. A. Aljabali, V. Mishra, H. Gandhi, S. Kumar, S. Mutalik, D. K. Chellappan, M. M. Tambuwala, K. Dua and D. N. Kapoor, Drug Delivery Transl. Res., 2020, 10, 1171–1190.
3. E. Mattia and S. Otto, Nat. Nanotechnol., 2015, 10, 111–119.
4. Z. Zhang, Z. Fan and J. Du, Precis. Med. Eng., 2024, 1, 100004.
5. E. Hernández Becerra, J. Quinchia, C. Castro and J. Orozco, Nanomaterials, 2022, 12, 836.
6. C. C. Petropoulos, J. Polym. Sci., Polym. Chem. Ed., 1977, 15, 1637–1644.
7. E. Cabane, V. Malinova and W. Meier, Macromol. Chem. Phys., 2010, 211, 1847–1856.
8. J. S. Katz, S. Zhong, B. G. Ricart, D. J. Pochan, D. A. Hammer and J. A. Burdick, J. Am. Chem. Soc., 2010, 132, 3654–3655.
9. S. Yamamoto, T. Yamada, G. Kubo, K. Sakurai, K. Yamaguchi and J. Nakanishi, Polymers, 2019, 11, 1254.
10. N. Doi, Y. Yamauchi, R. Ikegami, M. Kuzuya, Y. Sasai and S. Ichi Kondo, Polym. J., 2020, 52, 1375–1385.
11. W. Hou, R. Liu, S. Bi, Q. He, H. Wang and J. Gu, Molecules, 2020, 25, 5147.
12. X. Wang, C. Yao, G. Zhang and S. Liu, Nat. Commun., 2020, 11, 1–13.
13. Y. Zhou, R. Chen, H. Yang, C. Bao, J. Fan, C. Wang, Q. Lin and L. Zhu, J. Mater. Chem. B, 2020, 8, 727–735.
14. E. Cabane, V. Malinova, S. Menon, C. G. Palivan and W. Meier, Soft Matter, 2011, 7, 9167–9176.
15. W. Hou, R. Liu, S. Bi, Q. He, H. Wang and J. Gu, Molecules, 2020, 25, 542–552.
16. J. Bain, L. Ruiz-Pérez, A. J. Kennerley, S. P. Muench, R. Thompson, G. Battaglia and S. S. Staniland, Sci. Rep., 2015, 5, 1–8.
17. K. Yang, S. Zhang, J. He and Z. Nie, Nano Today, 2021, 36, 101046.
18. R. R. Arvizo, S. Bhattacharyya, R. A. Kudgus, K. Giri, R. Bhattacharya and P. Mukherjee, Chem. Soc. Rev., 2012, 41, 2943–2970.
19. P. Zrazhevskiy, M. Sena and X. Gao, Chem. Soc. Rev., 2010, 39, 4326–4354.
20. M. Fernández and J. Orozco, Polymers, 2021, 13, 2464.
21. M. Gao, F. Yu, C. Lv, J. Choo and L. Chen, Chem. Soc. Rev., 2017, 46, 2237–2271.
22. J. Chen, S. Chen, X. Zhao, L. V. Kuznetsova, S. S. Wong and I. Ojima, J. Am. Chem. Soc., 2008, 130, 16778–16785.
23. P. Mena-Giraldo, S. Pérez-Buitrago, M. Londoño-Berrío, I. C. Ortiz-Trujillo, L. M. Hoyos-Palacio and J. Orozco, Sci. Rep., 2020, 10, 1–12.
24. P. Mena-Giraldo and J. Orozco, Polymers, 2021, 13, 3920.
25. P. Mena-Giraldo and J. Orozco, ACS Appl. Mater. Interfaces, 2022, 14, 5897–5907.
26. P. Xiao, J. Zhang, J. Zhao and M. H. Stenzel, Prog. Polym. Sci., 2017, 74, 1–33.
27. H. Zhao, E. S. Sterner, E. B. Coughlin and P. Theato, Macromolecules, 2012, 45, 1723–1736.
28. D. W. Johnson, New Applications for Poly(ethylene-alt-maleic anhydride), PhD thesis, Durham University, 2010.
29. Z. Yuan, X. Yi, J. Zhang, S. Cheng, R. Zhuo and F. Li, Chem. Commun., 2013, 49, 801–803.
30. J. Quinchia, A. F. Cruz-Pacheco, D. Ruiz-Molina and J. Orozco, Chem. Eng. J., 2024, 500, 157165.
31. H. Ishida, Fourier Transform Infrared Characterization of Polymers, Springer, US, 1987.
32. R. A. Nyquist, in Interpreting Infrared, Raman, and Nuclear Magnetic Resonance Spectra, Academic Press, 2001, pp. 205–212.
33. B. L. Rivas, S. A. Pooley, E. Pereira, E. Montoya and R. Cid, J. Appl. Polym. Sci., 2006, 101, 2057–2061.
34. P. Larkin, in Infrared and Raman Spectroscopy, Elsevier, 2011, pp. 73–115.
35. F. Bettahar, F. Bekkar, L. Pérez-Álvarez, M. I. Ferahi, R. Meghabar, J. L. Vilas-Vilela and L. Ruiz-Rubio, Polymers, 2021, 13, 972.
36. G. Panzarasa, A. Osypova, G. Consolati, F. Quasso, G. Soliveri, J. Ribera and F. W. M. R. Schwarze, Nanomaterials, 2018, 8, 54.
37. M. H. El-Newehy, H. El-Hamshary, A. Alamri and S. S. Al-Deyab, Int. J. Polym. Mater. Polym. Biomater., 2014, 63, 563–575.
38. M. L. Calzada, A. González, R. Poyato and L. Pardo, J. Mater. Chem., 2003, 13, 1451–1457.
39. X. Zhang, M. Zhang, M. Wu, L. Yang, R. Liu, R. Zhang, T. Zhao, C. Song, G. Liu and Q. Zhu, Polymers, 2021, 13, 2392.
40. S. An, H. Kim, M. Kim and S. Kim, Langmuir, 2020, 36, 3046–3056.
41. L. Xiang Guo, J. Guan, B. Ping Lin and H. Yang, J. Cent. South Univ., 2015, 22, 3296–3301.
42. A. K. Kohli and H. O. Alpar, Int. J. Pharm., 2004, 275, 13–17.
43. A. Romano, I. Roppolo, E. Rossegger, S. Schlögl and M. Sangermano, Materials, 2020, 13, 1–26.
44. S. R. Aragón and R. Pecora, J. Chem. Phys., 1976, 64, 2395–2404.
45. B. K. Wilson and R. K. Prud'homme, J. Colloid Interface Sci., 2021, 604, 208–220.
46. F. Araste, A. Aliabadi, K. Abnous, S. M. Taghdisi, M. Ramezani and M. Alibolandi, J. Controlled Release, 2021, 330, 502–528.
47. C. Weng, H. Chen, T. Xu, Z. Li, X. Liu, M. Ding, Q. Zhang, H. Tan and Q. Fu, ACS Mater. Lett., 2020, 2, 602–609.
48. G. Liu, J. Tan, J. Cen, G. Zhang, J. Hu and S. Liu, Nat. Commun., 2022, 13, 1–12.
49. U. Seifert, K. Berndl and R. Lipowsky, Phys. Rev. A: At., Mol., Opt. Phys., 1991, 44, 1182–1202.
50. Z. Shi, Y. Zhou and D. Yan, Macromol. Rapid Commun., 2008, 29, 412–418.
51. S. Yu, T. Azzam, I. Rouiller and A. Eisenberg, J. Am. Chem. Soc., 2009, 131, 10557–10566.
52. R. A. Cox, Angew. Chem., Int. Ed., 2013, 52, 7638.
53. R. M. Perera, S. Gupta, T. Li, M. Bleuel, K. Hong and G. J. Schneider, Soft Matter, 2021, 17, 4452–4463.
54. B. A. Deyerle and Y. Zhang, Langmuir, 2011, 27, 9203–9210.
55. B. Kang, H. Tang, Z. Zhao and S. Song, ACS Omega, 2020, 5, 6229–6239.
56. Y. Men, W. Li, C. Lebleu, J. Sun and D. A. Wilson, Biomacromolecules, 2019, 23, 11.
57. S. Li, B. Byrne, J. E. Welsh and A. F. Palmer, Biotechnol. Prog., 2007, 23, 278–285.
58. L. K. E. A. Abdelmohsen, R. S. M. Rikken, P. C. M. Christianen, J. C. M. van Hest and D. A. Wilson, Polymer, 2016, 107, 445–449.
59. J. A. Barltrop, P. J. Plant and P. Schofield, Chem. Commun., 1966, 822–823.
60. A. P. Pelliccioli and J. Wirz, Photochem. Photobiol. Sci., 2002, 1, 441–458.
61. E. Cabane, Photoresponsive polymersomes as smart, triggerable nanocarriers, University of Basel, 2013.
62. J. A. Barltrop and P. Schofield, Tetrahedron Lett., 1962, 3, 697–699.
63. P. Xiao, J. Zhang, J. Zhao and M. H. Stenzel, Prog. Polym. Sci., 2017, 74, 1–33.
64. E. H. Kim, J. Korean Med. Assoc., 1963, 6, 142–146.
65. A. Sauré, J. Patrick, S. Tyldesley and M. L. Puterman, Eur. J. Oper Res., 2012, 223, 573–584.
66. E. B. Turner, Xray Tech.
67. Y. Liu, T. Wang and W. Wang, Chem. Soc. Rev., 2025, 54, 5792–5835.
68. J. F. Liao, W. T. Li, J. R. Peng, Q. Yang, H. Li, Y. Q. Wei, X. N. Zhang and Z. Y. Qian, Theranostics, 2015, 5, 345.
69. Z. Sun, J. Wang, B. Guo, S. Zhao, S. Miao, M. Xia, Z. Zhong and F. Meng, J. Controlled Release, 2025, 385, 113976.
70. C. Chen, G. Chu, W. He, Y. Liu, K. Dai, J. Valdez, A. Moores, P. Huang, Z. Wang, J. Jin, M. Guan, W. Jiang, Y. Mai, D. Ma, Y. Wang and Y. Zhou, Adv. Mater., 2023, 35, 2207950.
71. M. Hasannia, K. Abnous, S. M. Taghdisi, A. Sazgarnia, S. Nekooei, M. Ramezani and M. Alibolandi, J. Drug Delivery Sci. Technol., 2024, 94, 105526.
72. L. Wang, Y. Yuan, S. Lin, J. Huang, J. Dai, Q. Jiang, D. Cheng and X. Shuai, Biomaterials, 2016, 78, 40–49.
73. K. Zhu, G. Liu, G. Zhang, J. Hu and S. Liu, Macromolecules, 2018, 51, 8530–8538.
74. H. L. Ding, Y. X. Zhang, S. Wang, J. M. Xu, S. C. Xu and G. H. Li, Chem. Mater., 2012, 24, 4572–4580.
75. G. Barrera, P. Tiberto, P. Allia, B. Bonelli, S. Esposito, A. Marocco, M. Pansini and Y. Leterrier, Appl. Sci., 2019, 9, 212.
76. S. Mørup, EPL, 1994, 28, 671–676.
77. A. Gupta, P. Avci, T. Dai, Y.-Y. Huang and M. R. Hamblin, Adv. Wound Care, 2013, 2, 422.
78. X. Chen, Z. Lin, N. Cheng, Y. Mo, L. Lu, J. Hou, Z. Li, X. Nie, S. Gao and Q. Hua, Mater. Today Bio, 2025, 32, 101871.
79. F. Wang, D. Banerjee, Y. Liu, X. Chen and X. Liu, Analyst, 2010, 135, 1839–1854.
80. C. Wang, L. Cheng and Z. Liu, Theranostics, 2013, 3, 317.
81. C. Sanson, O. Diou, J. Thévenot, E. Ibarboure, A. Soum, A. Brulet, S. Miraux, E. Thiaudière, S. Tan, A. Brisson, V. Dupuis, O. Sandre and S. Lecommandoux, ACS Nano, 2011, 5, 1122–1140.
82. M. Tollemeto, S. Ursulska, P. L. W. Welzen, L. H. E. Thamdrup, A. Malakpour Permlid, Y. Li, G. Soufi, T. Patiño Padial, J. B. Christensen, L. Hagner Nielsen, J. van Hest and A. Boisen, Small, 2024, 20, 2403640.
83. N. Selzner, M. Selzner, R. Graf, U. Ungethuem, J. G. Fitz and P. A. Clavien, Cell Death Differ., 2004, 11, S172–S180.
84. T. Pompe, S. Zschoche, N. Herold, K. Salchert, M. F. Gouzy, C. Sperling and C. Werner, Biomacromolecules, 2003, 4, 1072–1079.
85. H. Saleem and U. Iqbal, Cureus, 2018, 10, e2163.
86. E. Willmore and B. W. Durkacz, Biochem. Pharmacol., 1993, 46, 205–211.
87. K. Bracht, A. M. Nicholls, Y. Liu and W. F. Bodmer, Br. J. Cancer, 2010, 103, 340–346.
88. E. Martínez, M. Osorio, C. Finkielstein, I. Ortíz, M. S. Peresin and C. Castro, Int. J. Biol. Macromol., 2022, 220, 802–815.
89. T. Salazar-Anzures, K. Pastén-Hidalgo, G. Sicilia-Argumedo, L. Riverón-Negrete, A. de J. Hernández-Vázquez and C. Fernanadez-Mejia, Toxicol. Appl. Pharmacol., 2021, 433, 115774.
90. C. E. Karachaliou and E. Livaniou, Int. J. Mol. Sci., 2024, 25, 6578.
91. S. Sarkar, P. Moitra, W. Duan and S. Bhattacharya, Adv. Healthcare Mater., 2024, 13, 2402259.
92. F. Corzana, A. Asín, A. Eguskiza, E. De Tomi, A. Martín-Carnicero, M. P. Martínez-Moral, V. Mangini, F. Papi, C. Bretón, P. Oroz, L. Lagartera, E. Jiménez-Moreno, A. Avenoza, J. H. Busto, C. Nativi, J. L. Asensio, R. Hurtado-Guerrero, J. M. Peregrina, G. Malerba, A. Martínez and R. Fiammengo, Angew. Chem., Int. Ed., 2024, 63, e202407131.
93. R. Djermane, C. Nieto, M. A. Vega and E. M. M. del Valle, Sci. Rep., 2024, 14, 1–14.
94. S. Azwar, H. F. Seow, M. Abdullah, M. F. Jabar and N. Mohtarrudin, Biology, 2021, 10, 854.

---