Aminated dendritic MSNs (NH2-MSNs, cat no. R-SG-89NH2), 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC, cat no. LP-R4-057), cholesterol, 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy(polyethylene glycol)-2000] (DSPE-PEG2000, cat no. R-H54510), DSPE-PEG2000-RVG29 (cat no. R9985), Cy5.5 (cat no. D10061), and M-PEG-lipid (cat. no. LP-R4-18) were purchased from Ruixi Biotechnology Co., Ltd. (Xi’an, Shaanxi, China). Three siRNA (cat no. 15258) sequences targeting circHIPK2 (circHIPK2 siRNA) and a non-homologous negative control (siNC) without sequence homology to the target gene, along with fluorescein amidite (FAM)-labeled siRNA, were obtained from Qingke Biotechnology Co., Ltd. (Beijing, China). The siRNA-mate Plus transfection kit was acquired from GenePharma Co., Ltd. (cat no. G04026, Shanghai, China). Chloroform (CHCl₃ , cat no. CX1060) and other analytical-grade reagents were sourced from Chongqing Chuandong Chemical Group Co., Ltd. (Chongqing, China). Fetal bovine serum (FBS) was acquired from Lanzhou Minhai Bio-Engineering Co.Ltd. (cat no. SA211.02, Lanzhou, Gansu, China). Dulbecco’s Modified Eagle Medium/Nutrient Mixture F-12 (DMEM/F12, cat no. 11320033), and 0.25% trypsin-ethylenediaminetetraacetic acid (EDTA, cat no. 25200056) were purchased from Gibco BRL (Grand Island, NY, USA). The Cell Counting Kit-8 (CCK-8) was obtained from Boster Biological Technology Co., Ltd. (cat no. AR1199, Wuhan, Hubei, China). Trizol (cat no. RR036A), PrimeScript RT reagent Kit (cat no. RR092S), and SYBR Green PCR Master Mix (cat no.RR820A) were supplied by Takara Bio, Inc. (Shiga City, Shiga Prefecture, Japan). Quantitative real-time polymerase chain reaction (qPCR) primers (cat no. TSE20240906-029) were synthesized by ZerKcorp Biotech (Xi’an, Shaanxi, China). Agarose powder (cat no. GC205013), SerRed Nucleic Acid Stain (10,000$\times$, water-soluble, cat no. G3606), phosphate-buffered saline (PBS, cat no. G4202), Tris-buffered saline with Tween 20 (TBST, cat no. G2157), Tris-glycine sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) high-resolution rapid electrophoresis (cat no. G2081)/transferring buffer(cat no. G2164), paraformaldehyde fixative (cat no. G1101), hematoxylin and eosin (H&E) staining solutions (cat no. G1076), and cell culture consumables were purchased from Servicebio (Wuhan, Hubei, China). C57BL/6 (C57) neonatal mice were obtained from the Lanzhou Veterinary Research Institute, Chinese Academy of Agricultural Sciences (Lanzhou, Gansu, China). Glial fibrillary acidic protein (GFAP) antibody (Abcam, cat no. ab7260, Shanghai, China; 1:10,000 dilution), IL-1$\beta$ antibody (Proteintech, Wuhan, Hubei, China; 1:5000 dilution), and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) antibody (Proteintech, cat no. 60004-1-Ig, Wuhan, Hubei, China; 1:5000 dilution) were used for immunoblotting.

Specific siRNAs were designed to silence the expression of circHIPK2. The efficacy of circHIPK2 silencing was evaluated using quantitative reverse transcription PCR (qRT-PCR) and western blotting. A total of four groups were established: three experimental groups transfected with each siRNA targeting circHIPK2 (circHIPK2-1, circHIPK2-2, and circHIPK2-3) and one control group without siRNA transfection. The siRNA-mate Plus transfection reagent was used to deliver the siRNAs into astrocytes. The siRNA demonstrating the most efficient silencing effect was selected for subsequent experiments (Table 1).

Agarose gel electrophoresis was performed to evaluate the siRNA loading capacity of NH₂-MSNs. At room temperature, siRNA (218 µg/mL) and NH₂-MSNs (5 mg/mL) were mixed at mass ratios (1:0, 1:10, 1:20, 1:25, 1:30, w/w), followed by incubation for 30 min. After incubation, 6$\times$ DNA loading buffer was added to the mixture. Samples were separated on a 3% agarose gel at 100 V for 20 min. The gel was visualized using a gel imaging system (version S2101237, ChampGel6000, Beijing, China) to observe band shifts of NH₂-MSN-siRNA complexes. The optimal binding ratio was determined to guide the preparation of M-R@siC-NPs.

Liposomes capable of crossing the BBB were prepared by mixing DPPC, cholesterol, M-PEG-lipid, and DSPE-PEG2000-RVG29 at a mass ratio of 3:1:1:1, dissolving the mixture in chloroform, and evaporating the solvent under vacuum to form a lipid film. The film was hydrated with 4 mL PBS and sonicated for 20 min to obtain a nanoemulsion. The emulsion was then emulsified under ice-cold conditions with 125 W ultrasonication for 8 min, followed by centrifugation at 6100 $\times$g for 10 min at 4 °C (repeated three times). The resulting product was purified via dialysis using a 100-nm cellulose membrane to obtain the liposomes.

For nanoparticle assembly, circHIPK2 siRNA and NH₂-MSNs were mixed at a mass ratio of 1:30, incubated at room temperature for 10 min, and electrostatically bound to form NH₂-MSN-siRNA complexes. The liposomes were then combined with NH₂-MSN-siRNA at a 1:1 mass ratio, incubated at room temperature for 10 min, and subjected to ultrasonication (3 min pulses with 5-s intervals). After centrifugation at 6100 $\times$g for 10 min, the precipitate was resuspended and stored at 4 °C.

As controls, target-free LNPs (siC-NPs) were prepared using the same method but omitting DSPE-PEG2000-RVG29 and M-PEG-lipid. The full name of siC-NPs is non-targeted modified circHIPK2 siRNA-lipid nanoparticles, which serve as non-targeted control nanoparticles in this study. Their core carrier components are consistent with those of M-R@siC-NPs (mannose/RVG29 peptide dual-modified circHIPK2 siRNA-lipid nanoparticles), with the only difference being the lack of two targeted ligands, namely M-PEG-lipid and DSPE-PEG2000-RVG29. Both siC-NPs and M-R@siC-NPs carry circHIPK2 siRNA3. Fluorescently labeled Cy5.5-conjugated targeting (Cy5.5-M-R@siC-NPs) and non-targeting nanoparticles (Cy5.5-siC-NPs) were also synthesized for fluorescence imaging.

The Malvern Nano analyzer (cat no. ZEN-3600, Zeta-Sizer, Malvern Instruments, Malvern, UK) was used to measure particle size and surface zeta potential. For sample preparation, 10 µL of nanoparticle solution was spotted onto a 400-mesh carbon-coated copper grid, air-dried at room temperature, and imaged using transmission electron microscopy (TEM, cat no. HT7800, Hitachi, Tokyo Metropolis, Japan) to observe the morphology, structure, and size of NH₂-MSNs and M-R@siC-NPs.

The colloidal stability of M-R@siC-NPs was evaluated by monitoring hydrodynamic particle size changes via The Malvern Nano analyzer over 14 days (0, 2, 4, 6, 8, 10, 12, and 14 days). Additionally, M-R@siC-NPs suspensions were separately incubated in DMEM/F12 cell culture medium, PBS (pH 7.4), pure FBS, and artificial cerebrospinal fluid (aCSF) at ambient temperature (25 °C). Morphological changes, including macroscopic aggregation or precipitation, were observed and recorded daily for 14 days. This temperature represents the typical ambient temperature used for simulating the storage of nanoformulations and in vitro experiments, and it is also consistent with the specifications on room temperature conditions for long-term stability testing as outlined in the “Guidelines for the Stability Testing of Pharmaceutical Preparations” of the Chinese Pharmacopoeia (25 °C $\pm$ 2 °C).

The siRNA concentration was quantified using a microvolume UV-Vis spectrophotometer (cat no. DENOVIX DS-11, DENOVIX, Wilmington, DE, USA) at 260 nm. Encapsulation efficiency (EE, %) was calculated via centrifugation (14,000 $\times$g, 30 min) to separate unencapsulated siRNA. The formula used to calculate EE was:

$$\text{EE(\%)}=\frac{\text{Total siRNA in NPs}-\text{Free siRNA}}{\text{Total siRNA in NPs}}\times 100$$

Drug-loading capacity (DLC, %) was calculated using the following formula:

$$\mathrm{DLC}(\%)=\frac{\mathrm{Total}\mathrm{siRNA}\text{in NPs}\times \text{EE(\%)}}{\text{Total lipid nanoparticle mass}}\times 100$$

C57BL/6 neonatal mice (specific pathogen-free grade) were purchased from Lanzhou Veterinary Research Institute, Chinese Academy of Agricultural Sciences. All histopathological evaluations were guided by faculty members specializing in pathology from the School of Basic Medical Sciences, Lanzhou University. The animal experimental protocols were approved by the Animal Ethics Committee of the Second Hospital of Lanzhou University (approval number: D2024-439).

Primary astrocytes were seeded into 96-well plates at a density of 1 $\times$ 10⁴ cells/mL (100 µL per well). Edge wells were filled with sterile PBS to minimize edge effects. Cells were allowed to adhere for 24 h at 37 °C in an incubator with 5% CO₂. Solutions containing PBS (control), siC-NPs, and M-R@siC-NPs were prepared at gradient concentrations (0, 20, 40, 60, 80, 100, and 120 µg/mL). Primary astrocytes were treated with the respective solutions, with three replicates per group, and incubated for an additional 24 h. Prior to measurement, the supernatant was aspirated, and 10 µL of CCK-8 reagent was added to each well followed by 90 µL of serum-free medium. The plates were incubated in the dark for 2 h. Cell viability was assessed using a microplate reader (cat no. ELX808, Bio Tek Instruments, Inc., Windsor, VT, USA) across four groups (normal control (NC), oxygen-glucose deprivation (OGD), siC-NPs, M-R@siC-NPs) to evaluate nanoparticle cytotoxicity. During the culture and identification of astrocytes, detailed descriptions have been provided regarding their cellular morphology and the expression of the specific marker GFAP, and the result of mycoplasma testing was confirmed to be negative.

Nanoparticle solutions (20, 40, 60, 80, and 100 µg/mL), PBS (negative control), or triple-distilled water (positive control) were added to 1.5-mL eppendorf tubes (1 mL per group). Blood was collected from the orbital venous plexus of healthy C57BL/6 mice using EDTA anticoagulant tubes. After mixing, the blood was centrifuged at 3000 $\times$g for 15 min at room temperature. The plasma was discarded, and 20 µL of red blood cell pellet was resuspended in the respective solutions. The mixtures were incubated at 37 °C for 4 h before repeated centrifugation at 3000 $\times$g for 15 min. The supernatant (100 µL) was transferred to a 96-well plate, and absorbance (optical density, OD) was measured at 540 nm using a microplate reader (Bio Tek Instruments, Inc.). Hemolysis rates were calculated using the standard formula:

$$\text{Hemolysis rate}(\%)=\frac{\mathrm{OD}(\text{sample})-\mathrm{OD}(\mathrm{PBS})}{\mathrm{OD}\left({\mathrm{ddH}}_2\mathrm{O}\right)-\mathrm{OD}(\mathrm{PBS})}\times 100\%$$

In strict accordance with the guidelines on non-clinical safety studies of pharmaceuticals issued by the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH), these guidelines explicitly require that the potential effects of pharmaceuticals on major physiological functions (cardiovascular, respiratory, central nervous systems) be evaluated. Additionally, the study design and implementation refer to the Good Laboratory Practice for Non-Clinical Studies (GLP) promulgated by the National Medical Products Administration (NMPA) of China. Although this study does not belong to the anti-tumor field, the requirement in this guideline that nanopharmaceuticals should focus on carrier accumulation, organ-targeted distribution, and long-term toxicity is universally applicable. To evaluate the systemic toxicity of the nanoparticles, 10-day-old C57BL/6 neonatal mice were randomly divided into four groups (n = 3 per group, statistical analysis is conducted based on the complete set of sample data): a NC group, PBS group, siC-NPs group (80 µg/mL), and M-R@siC-NPs group (80 µg/mL). Nanoparticles were administered intravenously for 20 consecutive days. Blood samples were collected for complete blood count (CBC) analysis using a Mindray Veterinary Automated Hematology Analyzer (cat no. BC-2800vet, Mindray Biomedical Electronics Co., Ltd. Shenzhen, Guangdong, China). Serum was obtained via refrigerated centrifugation, and biochemical parameters were analyzed using an Automated Biochemical Analyzer (cat no. Chemray 240, Rayto Life Technology Co., Ltd. Shenzhen, Guangdong, China). Mice were anesthetized via intraperitoneal injection of 1% sodium pentobarbital at a dose of 50 mg/kg body weight. Anesthesia was confirmed when the mice exhibited no response to toe pinch and loss of corneal reflex. Subsequently, the mice were euthanized by cervical dislocation, in compliance with the Animal Ethics Guidelines of the Second Hospital of Lanzhou University. Death was verified by the absence of breathing and heartbeat for 5 consecutive minutes. After euthanasia, heart, liver, spleen, lung, kidney, and brain tissues were collected, fixed, and stained with H&E. Histopathological changes were quantitatively analyzed using the TissueFAXS Plus panoramic tissue cytometry system (cat no. TissueFAXS Plus, TissueGnostics GmbH, Vienna, Vienna State, Austria).

Primary astrocytes were seeded on confocal dishes at a density of 1 $\times$ 10⁵ cells/mL and allowed to adhere overnight in an incubator. Experimental groups included: a PBS group, Cy5.5-siC-NPs group (80 µg/mL), and Cy5.5-M-R@siC-NPs group (80 µg/mL). After 24 h in culture, Hoechst 33258 (cat no. BL804A, Biosharp Biotechnology Co., Ltd. Nanjing, Jiangsu, China) was used to stain cell nuclei. Fluorescence imaging of the cells was performed 10 min after staining using a laser scanning confocal microscope (version ZEN Blue 2.3, Zeiss LSM880, Jena, Germany).

Ten-day-old C57BL/6 neonatal mice were randomly divided into three groups (n = 9 per group): a PBS group, Cy5.5-siC-NPs group, and Cy5.5-M-R@siC-NPs group. The corresponding agents were intraperitoneally injected into each group. At time points of 0, 1, 2, 4, 6, 8, 10, 12, 24, and 48 h post-injection, in vivo fluorescence imaging was performed using a small-animal imaging system (Vie-works Smart-LF, Eschborn, Hessen, Germany) to monitor nanoparticle accumulation and assess targeting specificity. Mice were euthanized when the peak intracranial fluorescence intensity was achieved, and brain tissues were harvested for ex vivo imaging. Heart, liver, spleen, kidney, and lung tissues were also collected and imaged using the Vie-works Smart-LF small-animal live imaging system to observe the Cy5 fluorescence distribution in major organs [52]. Brain tissue sections were further processed with 4’,6-Diamidino-2-Phenylindole (DAPI, cat no. D1306, Thermo Fisher Scientific Inc. Waltham, MA, USA) staining for pathological examination.

Primary astrocytes were isolated from postnatal day 2–3 C57BL/6 neonatal mice anesthetized with isoflurane and disinfected with 75% ethanol. Whole brains were dissected under a biosafety cabinet, minced in pre-cooled Hank’s Buffered Salt Solution (HBSS) buffer, and digested with 0.25% trypsin-EDTA at 37 °C for 5 min. The digestion was halted by adding DMEM/F12 medium containing 10% FBS, followed by centrifugation at 1200 $\times$g for 5 min for the collection of cell pellets. Cells were resuspended and cultured at 37 °C in an incubator with 5% CO₂. After 3–4 days in culture, contaminating oligodendrocytes and microglia were removed by shaking at of the culture plates 260 rpm for 12 h using a THZ-98A thermomixer (cat no. THZ-98A, Shanghai Zhicheng Analytical Instrument Manufacturing Co., Ltd., Shanghai, China). High-purity astrocytes were characterized via GFAP immunofluorescence staining, according to the following procedure. Cells were fixed with 4% paraformaldehyde, permeabilized with 0.1% Triton X-100, incubated overnight with rabbit anti-GFAP primary antibody (1:10,000) at 4 °C, and labeled with Alexa Fluor 488-conjugated secondary antibody (1:1000, Invitrogen, Carlsbad, CA, USA) for 1 h at room temperature in the dark. Nuclei were counterstained with DAPI for 5 min. Stained samples were viewed using a Zeiss LSM880 laser scanning confocal microscope, and GFAP-positive cell proportions were quantified using ImageJ software (Version number: 1.54 g, NIH, Bethesda, MD, USA).

For the in vitro hypoxia–ischemia (HI) model, the cell culture medium was switched to serum-free medium 24 h before the cells were subjected to OGD. After two washes with PBS, the cells were incubated in serum-free and glucose-free medium and placed in a hypoxic chamber (1% O₂, 5% CO₂, 94% N2, cat no. Coy-HC-05, Coy Laboratory Products, Inc. Grass Lake, MI, USA) at 37 °C for 6 h. The medium was then exchanged with complete medium for 24–48 h of recovery culture before collection for further analysis.

For the in vivo HIBD model, 6–8-day-old C57BL/6 neonatal mice were anesthetized with 2% isoflurane. The left common carotid artery was permanently ligated with an 8-0 surgical suture. Post-surgery, mice were exposed to a hypoxic environment (8% O₂, 92% N₂, flow rate 1.5 L/min) in a chamber for 2 h. Successful establishment of the HIBD model was verified using laser speckle imaging. Mice were randomized to four groups: the sham group, HIBD group, siC-NPs group, and M-R@siC-NPs group. Brain morphology was observed at 3, 5, and 7 days post-surgery. Brain tissues were collected, sectioned, and subjected to H&E staining and immunofluorescence analysis for assessment of pathological changes.

After treatment, brain tissues and cells were washed 2–3 times with pre-cooled PBS. Protein lysates were prepared using radioimmunoprecipitation assay (RIPA) lysis buffer (Ailv Life, AIWB-012, Wuhan, Hubei, China) containing protease inhibitors, incubated on ice for 30 min, and centrifuged at 12,000 $\times$g for 20 min at 4 °C for the collection of supernatant. Total protein concentration was determined using the BCA assay ( PC0020, Solarbio,, Beijing, China) and normalized. Equal amounts of proteins were separated by 10% SDS-PAGE and transferred to polyvinylidene difluoride (PVDF) membranes (Millipore, ISEQ00010, Darmstadt, Hessen, Germany). Membranes were blocked with QuickBlock blocking buffer (Servicebio, G2052-500ML) for 10 min at room temperature, followed by overnight incubation at 4 °C with primary antibodies: GFAP (1:10,000), IL-1$\beta$ (Proteintech, 26048-1-AP; 1:5000), and GAPDH (Proteintech, CL594-60004; 1:5000). After washing with TBST, membranes were incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies (1:5000) at room temperature for 1 h. Protein bands were detected using a chemiluminescence imaging system (cat no. Tanon 5200 Multi, Shanghai Tanon Science & Technology Co., Ltd., Shanghai, China), and gray value quantification analysis was performed using ImageJ software.

Mice in the siC-NPs and M-R@siC-NPs groups (n = 12/group) received intraperitoneal injections of LNPs (80 µg/mL) for 5 consecutive days, followed by a 2-day pause, totaling four cycles over 28 days (100 µL per dose). Post-treatment, three mice per group were euthanized for histopathological analysis, including H&E staining and GFAP immunofluorescence staining to assess glial activation. Samples from parallel groups (n = 3/group) were used for western blot analysis to quantify GFAP and IL-1$\beta$ expression in brain tissues.

Statistical analysis was performed using GraphPad Prism 10.1.2 (GraphPad Software, Inc. La Jolla, CA, USA). All data are presented as mean $\pm$ standard deviation (SD). Differences among multiple groups were compared using one-way analysis of variance (ANOVA) followed by Tukey’s post-hoc test. Pairwise comparisons between two groups were performed using Student’s t-test. Values of p 0.05 indicated statistical significance.

In the development of functional nanoparticles for this study, the first challenge was to identify highly efficient circHIPK2 silencing sequences. For this purpose, primary astrocytes were transfected with three distinct circHIPK2 siRNAs (circHIPK2-1, circHIPK2-2, and circHIPK2-3). FAM-labeled siRNAs were utilized to track transfection efficiency (Fig. 1A). qRT-PCR results showed that compared with the Control group (without siRNA transfection), all three circHIPK2 siRNAs significantly reduced the expression level of circHIPK2 mRNA in primary astrocytes. Among them, circHIPK2 siRNA-3 exhibited the optimal silencing efficiency, mediating an 80.13% decrease in circHIPK2 mRNA expression. This efficiency was significantly higher than that of circHIPK2 siRNA-1 and circHIPK2 siRNA-2, with a statistically significant difference (p 0.0001), confirming that siRNA3 has the strongest ability to specifically recognize and degrade the target gene mRNA (Fig. 1D). We further evaluated the effect of the three siRNAs on GFAP expression via Western blot. The results demonstrated that the GFAP protein expression in the circHIPK2 siRNA-3 group was 40% lower than that in the Control group. This inhibitory effect was significantly more pronounced than those in the siRNA-1 and siRNA-2 groups (p 0.001), and was positively correlated with the high degradation efficiency at the mRNA level (Fig. 1B,C). These findings confirm that the silencing effect of siRNA-3 can be effectively transmitted to the protein expression level, forming a complete functional chain of gene silencing functional inhibition. circHIPK2 siRNA-3 is identified as the optimal sequence for silencing circHIPK2, with its gene sequences being 5^′-GCUUAGUCUUUGAGAUGUU-3^′ (sense strand) and 5^′-AACAUCUCAAAGACUAAGC-3^′ (antisense strand). Based on these findings, circHIPK2-3 was selected for use in the subsequent experiments.

Fig. 1.

Silencing efficiency of three different circHIPK2 siRNA sequences in astrocytes. (A) Cellular localization of FAM-labeled siRNAs (green fluorescence) at 48 h post-transfection, with nuclei counterstained by DAPI (blue). Scale bar = 100 µm. (B) Western blot analysis of glial fibrillary acidic protein (GFAP) protein expression across groups. GAPDH served as a loading control. (C) Quantitative analysis of GFAP protein levels normalized to GAPDH expression (n = 3). (D) Quantitative reverse transcription PCR (qRT-PCR) analysis of circHIPK2 mRNA expression (n = 3). Data are presented as mean $\pm$ SD. Statistical significance was determined by one-way ANOVA with Tukey’s post-hoc test (ns p $\ge$ 0.05; *p 0.05; ***p 0.001; ****p 0.0001 vs. control group). FAM, fluorescein amidite.

The physicochemical properties of the different synthesized nanoparticles are summarized in Fig. 2. Dynamic light scattering (DLS) analysis revealed that the hydrodynamic diameters of the NH₂-MSNs, NH₂-MSN-siRNA, and M-R@siC-NPs were 115.80 $\pm$ 9.65 nm, 118.40 $\pm$ 7.21 nm, and 134.78 $\pm$ 10.25 nm, respectively, with polydispersity index (PDI) values of 0.206 $\pm$ 0.041, 0.211 $\pm$ 0.019, and 0.205 $\pm$ 0.089, respectively (Fig. 2A). The narrow PDI values indicated homogeneous particle size distribution and high-quality synthesis. Zeta potential measurements further demonstrated surface charges of 20.00 $\pm$ 0.24 mV for NH₂-MSNs, –3.2 $\pm$ 0.86 mV for NH₂-MSN-siRNA, and 4.3 $\pm$ 0.48 mV for M-R@siC-NPs (Fig. 2B). The transition from positive to negative zeta potentials confirmed electrostatic binding between the positively charged NH₂-MSNs and negatively charged siRNA. TEM images revealed spherical morphologies for both NH₂-MSNs (Fig. 2C) and M-R@siC-NPs (Fig. 2D), with uniform diameters less than 200 nm and no apparent aggregation or fragmentation. The consistent particle size, comparable zeta potentials, and excellent colloidal stability of M-R@siC-NPs align with the requirements for biomedical applications. These physicochemical properties are critical for the performance of these nanoparticles in subsequent gene delivery and biosensing experiments, influencing parameters such as blood circulation time, cellular uptake, and biodistribution.

Fig. 2.

Physicochemical characterization of nanoparticles. (A) Hydrodynamic size distribution of different nanoparticle formulations (n = 3; data presented as mean $\pm$ SD). (B) Zeta potential analysis of nanoparticles (25 °C, dispersed in deionized water, n = 3). (C,D) TEM images of NH₂-MSNs (C) and M-R@siC-NPs (D), showcasing morphological features and dispersion homogeneity. Scale bar = 200 nm. TEM, transmission electron microscopy; NH2-MSNs, aminated dendritic MSNs; M-R@siC-NPs, mannose-RVG29@CircHIPK2siRNA-NPs .

To determine the optimal mass ratio for binding circHIPK2 siRNA (218 µg/mL) with NH₂-MSNs (5 mg/mL), agarose gel electrophoresis was performed. The results demonstrated that when the mass ratio of siRNA to NH₂-MSNs reached 1:30 or higher, the fluorescence intensity of the siRNA band decreased (Fig. 3A), indicating successful electrostatic binding between the positively charged NH₂-MSN and negatively charged siRNA. Thus, a 1:30 mass ratio was employed in the preparation of nanoparticles for subsequent experiments. To evaluate the encapsulation efficiency and drug loading capacity of the nanoparticles, ultramicroscopic UV-Vis spectrophotometry was used to quantify siRNA content (R² = 0.9983). The calculated encapsulation efficiency and drug loading capacity were 75.28% $\pm$ 0.06% and 15.63% $\pm$ 0.27%, respectively (Fig. 3B).

Fig. 3.

Characterization of siRNA encapsulation efficiency and drug loading capacity of nanoparticles. (A) Agarose gel electrophoresis analysis of circHIPK2 siRNA binding to amino-functionalized dendritic silica nanoparticles (NH₂-MSNs) at varying mass ratios (1:0, 1:10, 1:20, 1:25, and 1:30). Disappearance of the siRNA band intensity indicates successful loading. (B) Standard curve for siRNA quantification: absorbance at 260 nm was measured for gradient siRNA solutions, yielding a linear regression equation of Y = 0.002X + 0.005 (R² = 0.9983, n = 3; data presented as mean $\pm$ SD).

To evaluate the stability and biocompatibility of nanoparticles in physiological environments, the morphology and dispersion of nanoparticle suspensions in DMEM/F12, PBS, FBS, and aCSF were observed over 14 days. The results demonstrated that M-R@siC-NPs maintained a uniform dispersion without significant aggregation or precipitation in all tested media. Notably, no evident interaction between nanoparticles and serum proteins or culture medium components was observed in FBS and DMEM/F12, confirming effective surface modification (Fig. 4A–C). These findings indicate that M-R@siC-NPs exhibit excellent long-term stability, making them suitable for in vitro cell culture and in vivo CNS delivery applications. Additionally, DLS analysis revealed no significant fluctuations in the hydrodynamic diameter of M-R@siC-NPs throughout the 14-day study period (p $\ge$ 0.05, Fig. 4D). Particle size measurements at specific time points (day 0, 2, 4, 6, 8, 10, 12, and 14) were as follows: 134.78 $\pm$ 10.25 nm, 135.12 $\pm$ 9.87 nm, 136.54 $\pm$ 10.13 nm, 137.21 $\pm$ 9.92 nm, 136.89 $\pm$ 10.31 nm, 137.56 $\pm$ 10.05 nm, 138.12 $\pm$ 9.78 nm, and 138.45 $\pm$ 10.22 nm. These consistent measurements further confirm the structural integrity of M-R@siC-NPs under physiological conditions.

Fig. 4.

Physical stability of M-R@siC-NPs in various media. (A–C) Macroscopic appearance of undiluted original dispersions of M-R@siC-NPs stored in DMEM/F12 medium, PBS, FBS, and aCSF on days 1, 7, and 14 (stored at 25 °C in the dark). No visible aggregation or precipitation was observed in any medium, confirming the excellent colloidal stability of the M-R@siC-NPs. (D) Hydrodynamic size evolution of M-R@siC-NPs measured by DLS over 14 days (n = 3; data presented as mean $\pm$ SD). PBS, phosphate-buffered saline; FBS, fetal bovine serum; aCSF, artificial cerebrospinal fluid; DLS, dynamic light scattering.