Correct Identification of the Core‐Shell Structure of Cell Membrane‐Coated Polymeric Nanoparticles

Abstract

Transmission electron microscopy (TEM) observations of negatively stained cell membrane (CM)‐coated polymeric nanoparticles (NPs) reveal a characteristic core‐shell structure. However, negative staining agents can create artifacts that complicate the determination of the actual NP structure. Herein, it is demonstrated with various bare polymeric core NPs, such as poly(lactic‐co‐glycolic acid) (PLGA), poly(ethylene glycol) methyl ether‐block‐PLGA, and poly(caprolactone), that certain observed core‐shell structures are actually artifacts caused by the staining process. To address this issue, fluorescence quenching was applied to quantify the proportion of fully coated NPs and statistical TEM analysis was used to identify and differentiate whether the observed core‐shell structures of CM‐coated PLGA (CM−PLGA) NPs are due to artifacts or to the CM coating. Integrated shells in TEM images of negatively stained CM−PLGA NPs are identified as artifacts. The present results challenge current understanding of the structure of CM‐coated polymeric NPs and encourage researchers to use the proposed characterization approach to avoid misinterpretations.

The core‐shell structure of cell membrane‐coated PLGA NPs (poly(lactic‐co‐glycolic acid nanoparticles) observed in some negatively stained transmission electron microscopy (TEM) samples is an artifact caused by the staining of the original PLGA NPs and not due to the actual cell membrane coating.

Introduction

Biomimetic nanoparticles (NPs) with functional cell membrane (CM) coatings have emerged as a promising therapeutic platform to provide NPs with desirable biological properties. ^[1] In recent years, a wide range of NPs has been explored and treated with CM coating, including poly(lactic‐co‐glycolic acid) (PLGA), ^[2] liposomes, ^[3] porous silica, ^[4] gold, ^[5] iron oxide, ^[6] metal‐organic frameworks (MOFs), ^[7] and upconversion NPs. ^[8] Among them, PLGA has been approved by the United States Food and Drug Administration (FDA) and is the polymer most widely used to produce CM‐coated NPs ^[9] due to its excellent biocompatibility, biodegradability, and high drug loading capacity. ^[10] Examples include red blood CM‐coated anisotropic PLGA NPs for detoxification, ^[11] genetically engineered CM‐camouflaged PLGA NPs capable of targeting inflammation, ^[12] and platelet membrane‐cloaked PLGA NPs to enhance cancer immunotherapy. ^[13] It is widely accepted that when these core NPs and CM materials are subjected to the disrupting force of extrusion or ultrasonic energy, they are spontaneously transformed into an integrated core‐shell nanostructure. ^[14] This idea is primarily based on the results of transmission electron microscopy (TEM) images of negatively stained CM‐coated PLGA (CM−PLGA) NPs. ^[15] However, it is challenged by the fact that the negatively stained bare PLGA NPs also have a characteristic core‐shell structure. ^[16] Negative staining method involves salt solutions of strongly electron scattering heavy‐metal compounds (e.g., uranyl acetate and phosphotungstic acid) deposited onto the drying sample to effectively preserve the sample by formation of a mold of the sample and artificially enhance the contrast for visualization of the sample. ^[17] Because the air‐drying negative staining method is prone to introduce artifacts, such as flattening of specimens or partial staining, ^[18] mistakes can occur during the interpretation of TEM images obtained through negative staining. ^[19] These limitations inspired us to explore whether polymeric NPs are truly coated with CMs, as TEM observations with negative staining might suggest. Several other techniques, such as dynamic light scattering (DLS), sodium dodecyl sulfate−polyacrylamide gel electrophoresis (SDS−PAGE), and fluorescence co‐localization,[ ^15d , ²⁰ ] have been applied to confirm the presence of CMs on PLGA NPs, but to date the precise structure of CM−PLGA NPs remains uncertain.

In the present work, we demonstrate that the core‐shell nanostructure observed in TEM images of certain negatively stained bare polymeric NPs, including PLGA, poly(ethylene glycol) methyl ether‐block‐PLGA (PEG−PLGA), and poly(caprolactone) (PCL) that are commonly used as core NPs for CM coating, is an artifact caused by the staining itself. Having confirmed successful CM coating on PLGA NPs by conventional techniques, we used our previously developed fluorescence quenching method ^[21] and TEM analysis combined with Triton X‐100 (TX‐100) treatment to further identify the structure of CM−PLGA NPs. We clarify the origin of the possible misinterpretation and identify the specific features of CM−PLGA NPs. Based on our results and TEM images, we conclude that most of the previously reported core‐shell structures of negatively stained CM−PLGA NPs are actually artificial structures caused by the sample staining, thereby leading to misleading assumptions about the success of the CM coating.

Results and Discussion

First, bare PLGA NPs were fabricated using the one‐step nanoprecipitation method as described previously. ^[15b] DLS analysis established that the average hydrodynamic diameter (D_h) of the as‐prepared PLGA NPs was 138±1 nm (Figure 1a). To verify successful CM coating, negative staining of TEM samples is normally used to enhance the contrast and thereby aid observation of CMs on the surface of NPs. For comparison, we performed TEM imaging of the PLGA NPs with and without negative staining. In contrast to the TEM images without negative staining (Figures S1), the TEM images of negatively stained PLGA NPs with uranyl acetate (2 %, pH 4.4) revealed that most of the PLGA NPs (92 %) had a core‐shell structure despite the actual solid structure of the NPs (Figures 1b–e and Figures S2). The smallest NPs (8 %) still appeared correctly as NPs with the solid structure. Thus, the core‐shell structure of PLGA NPs can easily be observed incorrectly in TEM images with negatively stained samples. Because it is so easy to misinterpret negatively stained samples, ^[19] we further utilized cryogenic TEM (cryo‐TEM), a well‐known nearly artifact‐free technique, ^[22] to verify the native state of PLGA NPs. As illustrated in Figures 1f, g, the cryo‐TEM images of PLGA NPs did not show the core‐shell structure but clearly revealed a solid structure. These results indicate that the apparent core‐shell structure in the TEM images of negatively stained PLGA NPs may be an artifact caused by the staining and drying process. To gain more insight into the morphology of PLGA NPs, we applied surface characterization techniques including atomic force microscopy (AFM) and field‐emission scanning electron microscopy (FE‐SEM). The AFM images of PLGA NPs obtained under ambient conditions revealed a spherical shape and smooth surface (Figure 1 h), consistent with the FE‐SEM results of PLGA NPs (Figures S3). Of note, the corresponding height profiles reflected the flattened spheres characteristic of PLGA NPs (Figures 1i,j), probably caused by the AFM tip convolution of NPs ^[23] or the drying process of a spherical structure.

Figure 1.

Characterization of PLGA NPs. a) Size distribution of PLGA NPs determined by DLS. b–d) TEM images of negatively stained PLGA NPs with core‐shell (c) and solid (d) nanostructures. Scale bars, 100 nm in (b) and 50 nm in (c and d). e) Diameter distribution of PLGA NPs calculated from TEM data (n = 185 for the core‐shell structure; n = 16 for the solid structure). The inset shows the relative proportions of core‐shell and solid structures in the negatively stained TEM samples. f, g) Cryo‐TEM images of PLGA NPs at low magnification (f) and high magnification (g). Scale bars, 100 nm in (f) and 50 nm in (g). h, i) AFM images of PLGA NPs at low magnification (h) and high magnification (i). Scale bars, 200 nm in (h) and 100 nm in (i). j) Corresponding height profiles of the AFM image (i) along the white dashed line.

To test whether other stain chemicals could produce artifacts, the bare PLGA NPs were negatively stained by phosphotungstic acid (1 %, pH 6.7). Consistent with the negative staining results with the uranyl acetate, the staining core‐shell artifact was observed by using phosphotungstic acid as stain chemical (Figures S4), indicating that the formation of core‐shell artifact was independent on the staining chemical. Moreover, the change of staining time (i.e., 10 s, 30 s, 5 min, and 30 min) did not affect the shell thicknesses of staining artifacts (Figures S5). In view of the above findings, the reasons for artifacts are not absolutely clear. One possibility is that uranyl ions of the staining agent are partly adsorbed onto and partly penetrate into the porous surface layer of the polymeric NPs.[ ¹⁷ , ^18c , ²⁴ ] Besides using nanoprecipitation to prepare the PLGA NPs, we also examined the morphology of NPs prepared using the emulsion/evaporation method. ^[15d] The TEM images of negatively stained PLGA NPs also revealed a core‐shell structure (Figures S6). In summary, our finding that most of the solid PLGA NPs erroneously appeared to have a core‐shell structure in their TEM images after negative staining challenged us to correctly identify the CM coating on these PLGA NPs.

In addition to PLGA NPs, other polymeric NPs, such as PEG−PLGA NPs and PCL NPs, are commonly used as core materials for CM coating for cancer therapy. ^[25] Therefore, we examined the structures of PEG−PLGA NPs and PCL NPs obtained by the nanoprecipitation method (Figure 2). Consistent with the PLGA NPs, negatively stained TEM samples of PEG−PLGA NPs with D_h of 80±2 nm (Figure 2a) also displayed a core‐shell structure (Figures 2b, c). Specifically, the proportion of those with a core‐shell structure (average diameter: 65±10 nm) was 56 % (Figure 2d). Similarly, the core‐shell structure was observed for PCL NPs of D_h of 165±1 nm (Figures 2e–g), with a core‐shell structure proportion of 67 % (Figure 2 h). The variation in the proportions of core‐shell structures among these different polymeric NPs observed in the negatively stained TEM samples could be related to their different sizes and surface porosities. Together, these results confirm that the artifacts resulting from the negative staining of samples for TEM analysis of the commonly used polymeric NPs for CM coating are responsible for the erroneous assignment of the core‐shell structure.

Figure 2.

Characterization of PEG−PLGA NPs and PCL NPs. a) Size distribution of PEG−PLGA NPs determined by DLS. b, c) TEM images of negatively stained PEG−PLGA NPs at low magnification (b) and high magnification (c). Scale bars, 100 nm in (b) and 50 nm in (c). d) Diameter distribution of PEG−PLGA NPs calculated from TEM data (n = 105 for the core‐shell structure; n = 84 for the solid structure). The inset shows the relative proportions of core‐shell and solid structures in the negatively stained TEM samples. e) Size distribution of PCL NPs as determined by DLS. f, g) TEM images of negatively stained PCL NPs at low magnification (f) and high magnification (g). Scale bars, 100 nm in (f) and 50 nm in (g). h) Diameter distribution of PCL NPs calculated from TEM data (n = 91 for core‐shell structure; n = 45 for solid structure). The inset shows the relative proportions of core‐shell and solid structures in the negatively stained TEM samples.

Currently, the construction of CM‐coated polymeric NPs (taking PLGA as an example) involves three steps: (1) extracting CMs from the source cells, (2) forming CM‐derived vesicles, and (3) fusing CM vesicles with core PLGA NPs by co‐extrusion (Figure 3a). Here, mouse colon carcinoma (CT26) cells were selected as a model source cell line for the CM coating. The DLS results revealed that the coating of PLGA NPs with the extracted CMs caused a slight increase in the D_h of the PLGA NPs (from 137±2 to 153±4 nm; Figure 3b) and a change in zeta potential from −42±1 to −34±2 mV (Figure 3c), indicating the presence of CMs on the PLGA NPs. When the CM vesicles with shell thickness of about 10 nm (Figure 3d and Figure S7) were coated onto PLGA NPs, an integrated core‐shell structure for PLGA NPs and a membrane patch was clearly observed by TEM (Figure 3e). To further verify the co‐fusion of CMs and PLGA NPs, CT26 cells were incubated with CM−PLGA NPs for 6 h; the PLGA NPs and CM materials were labeled with fluorescein (FITC; green) and 1,1′‐dioctadecyl‐3,3,3′,3′‐tetramethylindocarbocyanine perchlorate (DiI; orange), respectively. Confocal laser scanning microscopy (CLSM) images confirmed that most of the green fluorescence signal emitted by the PLGA NPs was co‐localized with the orange fluorescence signal emitted by the CT26 CMs with the Pearson's correlation coefficient of 0.92 (Figure 3f), indicating successful formation of CM−PLGA NPs. Analysis by SDS−PAGE revealed that the protein profiles of the CM−PLGA NPs resembled those of the CM (Figure 3g), suggesting successful translocation of CM proteins onto the PLGA NPs. Moreover, to examine the colloidal stability of the PLGA NPs, the obtained PLGA NPs and CM−PLGA NPs were suspended in phosphate‐buffered saline (PBS), and the size of the NPs was monitored using DLS over a 48 h period at 37 °C. Figure 3 h shows that the size of the CM−PLGA NPs did not change remarkably compared with the bare PLGA NPs, confirming the improved stability of the CM−PLGA NPs. Taken together, these traditional characterization data provide strong evidence for the successful coating of CMs onto PLGA NPs, supporting the previously reported results for CM−PLGA NPs.[ ^15c , ^15d , ²⁶ ]

Figure 3.

Preparation and characterization of CM−PLGA NPs. a) Schematic illustration of the preparation of CM−PLGA NPs. Extracting cancer CMs and their subsequent coating onto the core PLGA NPs by the extrusion method. b, c) Hydrodynamic size (b) and zeta potential (c) of PLGA NPs, CM materials, and CM−PLGA NPs. Data represent mean ± standard deviation (s.d.; n = 3). d, e) Representative TEM images of negatively stained CM vesicles (d) and CM−PLGA NPs (e). The CM patch is indicated by a yellow arrow in (e). Scale bars, 50 nm. f) CLSM images showing the intracellular colocalization of the PLGA NPs (labeled with FITC; green) and CM materials (labeled with DiI; orange) after being internalized by CT26 cells. The CM−PLGA NPs were incubated with CT26 cells for 6 h. The cell nuclei were stained with 4′,6‐diamidino‐2‐phenylindole (DAPI; blue). Scale bar, 10 μm. g) Protein profiles of PLGA NPs, CM materials, and CM−PLGA NPs determined by SDS−PAGE. h) Colloidal stability of PLGA and CM−PLGA NPs incubated in PBS at 37 °C was evaluated by monitoring the particle size changes using DLS at various time points up to 48 h. Data represent mean ± s.d. (n = 3).

Observing CM coating in negatively stained TEM samples is a direct and simple identification approach because the CM thickness should be 5–10 nm.[ ⁸ , ²⁷ ] To explore whether the shell thickness due to the staining artifact varied in this thickness range, we synthesized a series of PLGA NPs with different diameters by tuning the ratios of solvent to water and ionic content during the nanoprecipitation procedure (Table S1). The DLS results (Figure 4a) revealed that the PLGA NPs had D_h of 82±1, 108±2, 135±4, and 234±3 nm, corresponding to the sizes of core NPs typically used for coating. Calculated from TEM images (Figure 4b and Figures S8), the shell thickness due to artifacts increased with increasing size (Figure 4c). Remarkably, the shell thicknesses of staining artifacts of the 82, 108, and 135 nm NPs were 5±1, 9±1, and 10±2 nm, respectively, which resembled the CM thickness. These results concerning the shell thickness of the staining artifact raised the question of how to correctly identify the CM coating.

Figure 4.

Identification of the CM−PLGA NPs. a) Size distributions of four sizes of PLGA NPs determined by DLS. b) TEM images of negatively stained PLGA NPs with different sizes (d=82, 108, 135, and 234 nm). Scale bars, 50 nm. c) Quantification of the shell thickness of staining artifacts in four differently sized PLGA NPs measured from negatively stained TEM samples (n = 20). d) Schematic representation of the LB−PLGA NPs. e) Mean diameters and zeta potentials of bare PLGA NPs, liposomes, and LB−PLGA NPs. Data represent mean ± s.d. (n = 3). f) TEM image of negatively stained LB−PLGA NPs. Scale bar, 100 nm. g) Quantification of the proportions of full CM coating for PLGA NPs, CM−PLGA NPs, and LB−PLGA NPs before and after addition of TX‐100. Data represent mean±s.d. (n =3). Significance was determined by one‐way ANOVA followed by post hoc Tukey test. **p  < 0.01. h) Schematic illustration of LB−PLGA NPs after addition of TX‐100. i) TEM images of negatively stained PLGA NPs, CM−PLGA NPs, and LB−PLGA NPs before and after addition of TX‐100. Scale bars, 100 nm.

To better identify the CM−PLGA NPs, 1,2‐dioleoyl‐sn‐glycero‐3‐phosphocholine (DOPC) lipid bilayer‐coated PLGA (LB−PLGA) NPs were prepared as the control group due to their previously reported lipid shell−polymer core structure (Figure 4d). ^[28] The D_h of the LB−PLGA NPs (153±2 nm) was slightly larger than that of the PLGA NPs (134±1 nm; Figure 4e), indicating successful surface coating with LB. Meanwhile, the zeta potential of the LB−PLGA NPs increased from −42±2 mV for the PLGA NPs to −23±1 mV due to the coating of the neutrally charged LB. The TEM imaging after negative staining revealed a single LB layer (∼5 nm) on the surface of the LB−PLGA NPs (Figure 4f), which was not observed on bare PLGA NPs (Figures S2), further confirming the presence of the LB around the PLGA NPs. Next, we measured the proportion of fully coated NPs in CM−PLGA NPs and LB−PLGA NPs using our recently developed fluorescence quenching assay, ^[21] which uses dithionite, a membrane‐impermeant reducing agent, to probe the integrity of the CM coating. Consistent with the previously reported conclusion of partial CM coating, ^[21] the proportion of fully coated CM−PLGA NPs was approximately 7 %, which was notably less than that of LB−PLGA NPs (∼49 %; Figure 4g). These results suggest that the complete shell layer of CM−PLGA NPs observed with the negatively stained TEM samples was due to artifacts caused by the staining of the original PLGA NPs rather than the CM coating. If the outside layer or the shell was the CM coating as described in previous reports,[ ² , ¹³ , ^15d , ²⁹ ] the proportion of fully coated NPs could not be so low when compared with the LB−PLGA NPs. Therefore, we added nonionic TX‐100, a surfactant commonly used in the solubilization of lipid membranes (Figure 4 h), ^[30] to test whether the measured proportion of fully coated NPs of CM−PLGA NPs/LB−PLGA NPs resulted from protection by the lipid membranes. Following the addition of TX‐100, the proportions of fully coated NPs for both CM−PLGA NPs and LB−PLGA NPs were close to that of bare PLGA NPs (Figure 4g), confirming that LB was the observed outside layer of the LB−PLGA NPs. To provide visualized evidence for this procedure, we carried out TEM imaging of negatively stained PLGA NPs, CM−PLGA NPs, and LB−PLGA NPs before and after the addition of TX‐100 (Figure 4i). The images clearly showed that upon addition of TX‐100, the core‐shell structure of these NPs derived from the original PLGA NPs was well preserved, while the lipid membranes of CM−PLGA NPs and LB−PLGA NPs were destroyed. Together, these findings confirm that the core‐shell structure of CM−PLGA NPs observed in some negatively stained TEM samples is an artifact caused by the staining of the original PLGA NPs and not because of the actual CM coating.

Conclusion

A comprehensive understanding of the structure of CM‐coated NPs is imperative to develop efficient biomimetic NPs for biomedical applications. When determining core‐shell structures by means of negatively stained TEM samples, it is vital to be aware that the technique has certain deficiencies that can lead to erroneous conclusions. This is especially true when polymeric NPs are the core that is CM‐coated. The staining agent itself can produce an artificial surface layer on top of the NPs, and this can only be distinguished with difficulty from the actual CM coating layer, especially when the dimensions of the actual surface layer and the artificial one due to the staining artifact are similar. Obviously, the porous surface layer of polymeric NPs is the main reason for the artifact: uranyl ions adsorb onto and penetrate into the surface, which appears as a dense surface layer in TEM imaging. The staining works differently with different types of core NPs, such as mesoporous SiO₂ NPs (Figures S9). Thus, conclusions regarding core‐shell structures must be evaluated on a case‐by‐case basis. In general, the smooth, even, and homogenous surface layers observed for CM−PLGA NPs cannot be caused by CMs because CMs always form an irregular and uneven surface structure. The results of the present study highlight the importance of combining methods to quantitatively analyze fully coated NPs and supplementary TEM analysis after the addition of TX‐100 to verify successful CM coating. The characterization approach applied in the present study will allow researchers to identify whether the observed surface layer corresponds to CMs or is an artifact.

In summary, we used combined imaging techniques of TEM, cryo‐TEM, SEM, and AFM to systematically investigate the morphology of bare polymeric NPs. We confirmed that some core‐shell structures apparent in negatively stained TEM samples are due to a staining artifact. Our results regarding the proportion of fully coated NPs provide strong evidence of partial coating for CM−PLGA NPs and will help establish guidelines to avoid misinterpreting artifacts as CM coating when examining CM‐coated polymeric NPs.

Experimental Section

Synthesis and characterization of PLGA NPs, PEG−PLGA NPs and PCL NPs: PLGA NPs were prepared by a one‐step nanoprecipitation method. ^[15b] Briefly, 10 mg PLGA was dissolved in 1 mL acetone and added dropwise into 3 mL of water. The mixture was stirred in open air for 2 h and then placed it in a vacuum oven to remove the residual organic solvent. The differently sized PLGA NPs (82, 108, 135, and 234 nm) were prepared by changing the polymer concentration, ratio of solvent to water and ionic content in the aqueous solution during the nanoprecipitation process (Table S1). For comparison, PLGA NPs were prepared by emulsion/solvent evaporation method as described previously. ^[15d] The PEG‐PLGA NPs were synthesized using nanoprecipitation method as same with the PLGA NPs.To obtain PCL NPs, 2.5 mg PCL was dissolved in 1 mL of acetone. The organic phase was then added dropwise into 3 mL water under vigorous stirring. Then, the organic solvent was removed by stirring in the open air overnight under room temperature.

The morphology of the PLGA NPs, PEG−PLGA NPs and PCL NPs was visualized using TEM (JEM‐2100F, JEM Ltd., Japan), in which samples were negatively stained with uranyl acetate. Briefly, a drop (10 μL) of the sample was placed onto a formvar and carbon coated glow discharged copper grid held by tweezer. After 2 min incubation, the excess solution was removed with filter paper. When the suspension was partially dried, the grid was washed four times by touching it to the surface of a drop of distilled water. Afterwards, the excess water was removed by touching the grid to a filter paper. A small drop (7 μL) of uranyl acetate (2 %, pH 4.4) or phosphotungstic acid (1 %, pH 6.7) dissolved in distilled water was then applied to the grid. After the predetermined staining time (i.e., 10 s, 30 s, 5 min, and 30 min), the excess stain was removed by touching the edge of grid to a filter paper, and the grid was dried at room temperature before TEM observation. Of note, in a typical experiment, unless stated otherwise, the uranyl acetate was used as the stain chemical and the staining time was 30 s. For the FE‐SEM imaging, the samples were sputter‐coated with gold and imaged using a Carl Zeiss Sigma HD|VP FE‐SEM.

Preparation of CM−PLGA NPs and LB−PLGA NPs: To obtain CM vesicles, the isolated membrane materials were physically extruded through a 400 nm polycarbonate membrane for 13 passes. The resulting membrane vesicles were then coated onto PLGA core NPs by coextruding vesicles and cores through a 400 nm polycarbonate membrane for at least 13 times to form CM−PLGA NPs. The LB−PLGA NPs were prepared according to the previously described protocol with minor modifications. ^[31] Briefly, the as‐prepared DOPC liposomes (2.5 mg/mL) and PLGA NPs were then mixed in equal volumes (usually 500 μL) and incubated at room temperature for 1 h with occasional agitation. The mixture was further purified by centrifugation to remove any excess lipids and finally dispersed in 1 mL of 20 mM HEPES buffer (pH=7.4).

Physical characterization of CM−PLGA NPs: The size distributions and zeta potential of prepared NPs were assessed by a DLS (Malvern Zetasizer Nano ZS). To examine the morphology, samples were placed onto the TEM grid, stained with uranyl acetate as described above, and imaged using a JEM‐2100F (JEM Ltd., Japan) microscope. To confirm the colocalization of PLGA NPs and CT26 CM, FITC and DiI dye were employed to label PLGA NPs and CM materials, respectively. The resulting dual‐fluorophore‐labeled CM−PLGA NPs were incubated with CT26 cells for 6 h and visualized using a CLSM (Zeiss LSM 800 Airyscan, Carl Zeiss, Jena, Germany). Nuclei of the cells were stained with DAPI (1 μg/mL in HBSS). As for the FITC and DiI dye colocalization, the Pearson's correlation coefficient was calculated using the ImageJ software according to the previously described protocol. ^[32] CM proteins were further analyzed using SDS−PAGE. The protein concentrations of CM and CM−PLGA NPs were measured with the BCA assay kit. After being denatured, 20 μg of each sample was added into a 10 % SDS−PAGE gel, ran at 150 V for 50 min, and then stained with Coomassie blue. Subsequently, the gel was washed by deionized water and imaged by camera. The test for colloidal stability of PLGA and CM−PLGA NPs were performed by suspending NPs in PBS and determining the size of NP by DLS over 48 h at 37 °C. In addition, the morphology of bare PLGA NPs, CM−PLGA NPs, and LB−PLGA NPs before and after addition of 1 % TX‐100 was examined by TEM (JEM‐2100F, JEM Ltd., Japan).

Conflict of interest

The authors declare no conflict of interest.

Supporting information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supporting Information

L. Liu, W. Yu, J. Seitsonen, W. Xu, V.-P. Lehto, Chem. Eur. J. 2022, 28, e202200947.

Data Availability Statement

The data that support the plots within this paper and other findings of this study are available from the corresponding author upon reasonable request.