Silk fibroin and hydroxypropyl cellulose composite injectable hydrogel-containing extracellular vesicles for myocardial infarction repair

Hydroxypropyl cellulose (HPC, 99%), sodium carbonate anhydrous (Na₂CO₃, AP), lithium bromide (LiBr, 99%), polyethylene glycol (PEG, Mn = 8000, 99%), dimethyl sulfoxide (DMSO, AR) and 1,2-distearoyl-sn-glycero-3- phosphoethanolamine-N-[folate (polyethylene glycol)] (DSPE-PEG-FA) were obtained from Aladdin Biochemical Technology Co., Ltd (Shanghai, China). HUVEC and AC16 cells were purchased from Shanghai Fu Heng BioLogy (Shanghai, China). The Silk fibroin (SF) was extracted from mulberry silkworms by a previous reported method [22]. The following animal use protocol was reviewed and approved by the Animal Ethics Committee of Zhejiang Sci-Tech University (No. 20230625-01) and Institutional Animal Care and Use Committee of Zhejiang Center of Laboratory Animals (ZACLA-IACUC-20040163). Reagents related to Sprague-Dawley rats and animal experiments were provided by Hangzhou Hao Ke Biotechnolgy Co, Ltd

The details of the isolation and characterization of EVs from AC16 cells (figure S1) were provided in Supporting Information (SI).

DSPE-PEG-FA-modified EVs were prepared via noncovalent interaction by mixture of DSPE-PEG-FA and EVs surface protein content with 1:1 weight ratio for 4 h at 37 °C. And the excess DSPE-PEG-FA was removed by centrifugation at 12,000 g. The incorporation of FA into EVs was confirmed by Fourier transform infrared spectroscopy (Nicolet Avatar 370, Thermo Scientific, USA) in the range of 400–4000 cm⁻¹.

We performed cell scratch assays using HUVEC cells to explore the effect of EVs on cell migration [23]. HUVEC cells were inoculated into a 12-well plate. Made a scratch in a 12-hole plate with a sterile 200 μL pipette tips. When the cell fusion reached 90% or more, the original medium was discarded. Then, the cells were grown in a serum-free medium containing EVs (20 μg/mL) for 24 h. After incubation for 24 h, images of cell migration in three fields were obtained by bright field optical microscopy. The closed area of cell scratches was calculated with the Image J software.

To prepare blank SF/HPC hydrogel, 5 wt% SF solution and 5 wt% HPC were mixed in a 1:1 volume ratio at 37 °C for 1 h. Rheological behaviors and viscosity of composite hydrogels were tested by rotational rheometer (MCR52, Austria). For the preparation of SF/HPC composite hydrogel-containing EVs, 12.5 μL EVs or FA-EVs (80 mg mL⁻¹) were pre-stirred with 1 mL SF/HPC mixed solution.

H9C2 cardiomyocyte cells was used to evaluate the biocompatibility of SF/HPC hydrogels. 150 mg SF/HPC hydrogel was to be sterilized by UV lamp for 30 min, and then immersed into 3 mL DMEM for 24 h to obtain the extraction solutions. Cardiomyocytes of H9C2 were inoculated at 1 × 10⁵ per well in 96-well plates and co-cultured with hydrogel extraction solutions of different concentrations (0, 5, 15, 30, and 50 mg mL⁻¹) for 24 h. Finally, the relative cell viability of H9C2 was evaluated by MTT assay.

Sprague-Dawley male rats (220–250 g) were selected for coronary artery ligation and intramuscular injection. The rats were anesthetized with 5% chloral hydrate (300 mg kg⁻¹) and the left anterior descending branch (LAD) coronary artery was permanently ligation with 6–0 polypropylene sutures [24].

And then all SD rats were randomly divided into five groups: Sham group (with the same above procedure but the left anterior descending coronary artery was not lapped), MI group (injection of 20 μL of PBS), SF/HPC group (injection of 20 μL of SF/HPC composite hydrogel), SF/HPC+EVs group (injection of 20 μL of SF/HPC hydrogel containing 20 μg AC16 cell-derived EVs) and SF/HPC+FA-EVs group (injection of 20 μL of SF/HPC composite hydrogel containing 20 μg FA-EVs). Intramuscular injection was performed immediately after animal model establishment.

Cardiac function was measured by echocardiography (VINNO 6, Suzhou, China) on the 28th day after surgery. The rats were anesthetized and then two-dimensional echocardiography of the short axis of the left ventricle was obtained through the anterior and posterior walls of the left ventricle at the level of the tip of the papillary muscle below the mitral valve. Left ventricular end-diastolic diameter (LVIDd) and left ventricular end-systolic diameter (LVIDs) were measured by echocardiography software, and left ventricular ejection fraction (LVEF) and left ventricular shortening fraction (LVFS) were calculated.

The rats were killed on 28th day. Then, the hearts were excised and fixed with 4% paraformaldehyde for 24–48 h. Then, the heart tissue samples were embedded in paraffin wax to prepare tissue sections (5 μm thickness). Sections were stained with Masson's trichromatic reagent and Sirius scarlet, and examined by biological microscope. Vascular factor was characterized by direct immunostaining sections. Infarct size and length were analyzed by Image J.

RNA was extracted from heart tissue with Trizol reagent (HKR022, Hangzhou Hao Ke Biotechnolgy Co, Ltd, Hangzhou) and then reverse-transcribed into complementary DNA (cDNA) with Prime Script™RT reagent (HKR06, Hangzhou Hao Ke Biotechnolgy Co, Ltd, Hangzhou). GAPDH was used as an internal control. Real-time PCR was performed using the ABI Step One Plus Real Time PCR System (AE341, TransGen Biotech, Beijing). Three mRNA of von Willebrand factor (VWF), transforming growth factor (TGFβ1), and sarcoplasmic reticulum Ca²⁺-ATPase (Serca2a) were analyzed.

All experimental data are expressed as the mean ± standard deviation (SD). Statistically significant differences were analyzed using t-test. When *p < 0.05, **p < 0.01, ***p<0.001 the difference was statistically significant. And ns, no significant difference. A p-value<0.05 was considered statistically significant.

EVs can be secreted by a variety of cells, including cardiomyocytes, fibroblasts, stem cells, etc They play a pivotal role in the regulation of myocardial fibrosis, angiogenesis and other pathological changes after MI [25]. Usually, EVs refer to cell-derived multimolecular assemblies in the nanometer to micrometer range, including vesicular and nonvesicular entities, which act as signaling molecules involved in the regulation of several physiological activities. They contain proteins, lipids, and nucleic acids (mRNAs, miRNAs, etc) with important functions [26, 27]. Various bioactive components in EVs regulate and participate in biological physiological activities. Therefore, EVs have great potential as a new noncellular therapies for the treatment of MI. Previous studies have reported that EVs are membrane particles with diameters ranging from 30 nm to 10 μm [28–30]. The surface of EVs is enrich with a variety of proteins, including CD9, CD63, CD81, and CD82, which are considered to be the signature proteins of EVs [31]. In this study, we used PEG co-precipitation method to isolate EVs from AC16 cell culture medium because of its simplicity (figure 1(a)). In addition, the fact that it essentially does not disrupt the morphology and function of EVs [32, 33]. To purify the EVs, the cell culture medium is collected and centrifuged at 3,000 rpm for 30 min to remove dead cells and cell. Then, the supernatant after centrifugation is filtrated by a 0.22 μm filter to remove the larger vesicles and proteins. And then the isolated EVs were analyzed by TEM, dynamic light scattering analysis (DLS), and Western blotting (WB). Figure 1(b) shows the TEM image of isolated EVs, exhibiting the round-shaped vesicles with the membrane bounded. The ionic strength of the EVs is about 8 × 10^–5 mmol L⁻¹ when detected the size distribution of EVs. We extracted EVs mainly distributed between 30 and 150 nm, in agreement with previously reported results. DLS results indicated that the average particle size of EVs was 100.44 ± 4.49 nm, and the polydispersity index (PDI) was 0.328 ling relatively uniform distribution of EVs particle size (figure 1(c)). Meanwhile, WB analysis demonstrated high expression of CD9 and CD81 (figure 1(d)). The results indicate that EVs are successfully isolated from AC16 cells. In addition, the protein concentration in AC16 cell-derived EVs is 14.25 ± 2.34 μg mL⁻¹ (figure S2 and S3). All these results indicated that the AC16 cell-derived EVs are successfully obtained in this study.

Zoom In Zoom Out Reset image size

The folate receptor is widely present on the cell membrane surface of various cells, and it is a cell-targeting ligand with high affinity [34]. FA-EVs can bind to folate receptors on the surface of the cell membrane, increasing the effective amount of folate that reaches the cells. To enable the obtained EVs to be effectively delivered into myocardial tissue, they were modified by DSPE-PEG-FA. And the excess DSPE-PEG-FA was removed by centrifugation. To discriminate whether folate receptor (DSPE-PEG-FA) in-targeted with EVs, FT-IR analysis of EVs and FA-EVs were performed. Figure 1(e) and S4 show the FT-IR spectra of AC16 cell-derived EVs before and after modification of DSPE-PEG-FA. Two new peaks located at 1411 and 1602 cm⁻¹ can be observed in FA-EVs which are contribuated to benzoic vibrations and aromatic ring stretching vibrations of FA [35].

We examined the ability of AC16 cell-derived EVs to regulate cell migration by performing cell scratch assays with HUVEC cells. Figure 1(f) shows the representative images of HUVEC cells migration under stimulation of AC16 cell-derived EVs within the dose of 20 μg/mL for 0 and 24 h. In the case of HUVEC cells without any treatment, the scratch area is decreased from ∼85.67 ± 1.41% at 0 h to ∼65 ± 1.36% at 24 h. The HUVEC cells show the greater cell migration after cultured by EVs. The scratch area can be decreased from ∼76.87 ± 3.07% at 0 h to ∼42.25 ± 2.99% at 24 h, approximately 1.5-fold compared with blank group. The migration capacity of the HUVEC cells is significantly higher after cultured by FA-EVs, and the scratch area can be decreased from ∼82.54 ± 2.26% at 0 h to ∼31.39 ± 1.43% at 24 h (figure 1(g)). Therefore, the folate modified AC16 cell-derived EVs exhibit a greater influence on cell migration after 24 h (p<0.001) compared to blank and AC16 cell-derived EVs.

The SF/HPC composite hydrogels were prepared by blending SF and HPC for gelling at 37 °C for 1 h (figure 2(a)). The obtained SF/HPC composite hydrogels show a light yellow transparent state with an injectable property (figure 2(b)). More importantly, SF is a natural protein which can be used to mimic the function of the extracellular matrix (ECM); while, HPC is a polysaccharide with abundant hydroxyl groups and is a commonly used material for biological tissue engineering. In addition, random coiled SF molecules in the hydrogel are aligned to form β-sheets during the gelation process that result in contraction of the composite hydrogels [36] (figure 2(c)). We measured the dynamic modulus of SF/HPC composite hydrogels as a function of frequency at 37 °C and the results are shown in figure 2(d). The obtained SF/HPC composite hydrogels exhibit a viscoelastic solid (gel) behaviour, where the storage modulus (G') is higher than loss modulus (G''). In addition, the G' of the as-prepared hydrogel is ∼4–6 KPa, which is similar to the softness of biological tissues and cells [37–39]. The viscosity test shows a thixotropy behavior with shear thinning, which is suitable for injected tape casting (figure 2(e)). After incorporation of EVs into hydrogels, the encapsulated EVs still can be maintained their bioactivity due to their humidity microenvironment, and they will be released slowly as the degradation of the matrix of hydrogels [12, 13]. The released EVs can be internalized by endocytosis or fused directly with the cell membrane, releasing the contents into the target cells [40, 41]. Therefore, the obtained SF/HPC composite hydrogels not only provides a carrier platform for delivery of EVs, but also serves as a mechanical support for the damaged myocardium.

Zoom In Zoom Out Reset image size

Natural biopolymers have excellent compatibility with a wide range of cell types [42, 43]. In this study, we co-cultured an extract of SF/HPC composite hydrogel with H9C2 cells and determined cell viability by MTT assay (figure 2(f)). The cell viability of all hydrogel samples with ranging from 5 to 50 mg mL⁻¹ are higher than that of blank sample, indicating the hydrogel matrix has the function for promoting cell proliferation.

We also evaluated the repair effect of SF/HPC composite hydrogel on myocardial infarction by in vivo experiments. We purchased rats from the animal experimentation center, and after acclimatizing them to their environment by feeding them for a week, we then created a myocardial infarction model. We injected different hydrogels into the myocardium of the infarcted area of rats immediately after the animal model was established (figure 3(a)). The Sham group (normal heart) and MI group were taken as the control group. On 28th day of the animal model's establishment, we assessed its cardiac function by echocardiographic measurements. As showed in figure 3(b), the echocardiographic results show an increase in the left ventricular (LV) free wall diameter and the presence of significant systolic waves in the SF/HPC+FA-EVs group, suggesting improved neointimal contraction. Figure 3(c) shows that the left ventricular ejection fraction (LVEF) of SF/HPC, SF/HPC+EVs, and SF/HPC+FA-EVs groups when compared to the Sham and MI group. Sham group shows the highest LVEF at 93.17 ± 4.10% and MI group with lowest LVEF at 43.01 ± 3.40%. A slight enhancement, approximately at 52.54 ± 4.04%, can be obtained after only injection of SF/HPC hydrogels. However, after introduction of EVs into composite hydrogels, the LVEF can be increased to 59.14 ± 1.50%. And it can be further enhanced to 71.06 ± 8.66% for the composite after incorporating of folate modified EVs. These results indicate that compared to the MI, SF/HPC, and SF/HPC+EVs groups, the SF/HPC+FA-EVs group has a significantly improved ability to pump blood. Figure 3(d) shows the results of left ventricular shortening scores (LVFS) for each group. The highest and the lowest LVFS values are detected in the Sham (62.70 ± 9.60%) and MI (18.67 ± 1.82%) groups. The results show that the LVFS values of the SF/HPC and SF/HPC+EVs groups are significantly improved compared with the MI group, reaching 23.56 ± 2.20% and 27.66 ± 0.79%. Especially, the SF/HPC+FA-EVs group is raised to a more extent, being 44.77 ± 6.79% in comparison with SF/HPC and SF/HPC-EVs groups. Figures 3e and f show the lowest left ventricular

Zoom In Zoom Out Reset image size

internal diameter at diastole (LVIDd) and left ventricular internal diameter at systole (LVIDs) values for all groups. The LVIDd and LVIDs values in the SF/HPC+FA-EVs group are significantly lower than those in the MI group and near those in the Sham group. The aforementioned numerical findings suggest that SF/HPC composite hydrogels offer structural support to the damaged area and act as conveyors for EVs, thereby providing effective relief to left ventricular dilatation and enhancing cardiac functioning [44–46].

Myocardial infarction causes myocardial fibrosis, where collagenous tissue replaces normal myocardial tissue, resulting in impairment of heart function [38, 47]. Prevention of deleterious myocardial fibrosis and remodeling is critical for myocardial infarction repair. The improvement of adverse remodeling and myocardial fibrosis in hearts with myocardial infarction was evaluated by Masson trichrome staining. As shown in figure 4(a), Masson trichrome staining shows that the anterior wall of the left ventricle in the sham operated group is red, with a regular shape and a natural thickness. It represents normal heart muscle tissue. In the MI group, the anterior wall of the left ventricle is composed mainly of blue fibrous tissue with a small amount of red normal myocardial tissue. This is because the infarcted area is superseded by collagen deposits (blue). There are more myocardial tissues with red staining in the composite hydrogel implantation groups (SF/HPC and SF/HPC+EVs). In particular, the fibrotic area is decreased sharply in the SF/HPC+FA-EVs group when compared to the SF/HPC and SF/HPC-EVs groups. Figure 4(b) shows the quantitative analysis results of the LV infarcted area. Sham group shows the lowest LV infarcted area at 2.24 ± 0.21%. The best reparative effect is achieved in the SF/HPC+FA-EVs group and the LV infarcted area is decreased from 32.99 ± 1.51% (MI group) to 5.49 ± 0.26% compared with 25.21 ± 2.48% and 18.85 ± 1.11% for SF/HPC and SF/HPC+EVs groups. The LV collagen area results of all groups are shown in figure 4(c). The highest LV collagen area (25.77 ± 2.29%) is detected in MI group due to the occurred severe fibrosis. All composite hydrogel therapy groups have significantly reduced LV collagen area as opposed to the MI group, and the best cardiac repair result is SF/HPC+FA-EVs group with minimal LV collagen area (5.86 ± 0.50%). This may be due to the good compatibility of the hydrogel with the organism, as well as the ability of AC16 cell-derived EVs to promote angiogenesis.

Zoom In Zoom Out Reset image size

As we all known, TGFβ1 initiates myofibroblast transformation and myocardial fibrosis and causes damage to the heat [48, 49]. We examined the expression of TGFβ1 gene in rat heart tissue by RT-PCR assay. As revealed in figure 4(d), the expression level of the TGFβ1 gene can be effectively down-regulated in the SF/HPC+FA-EVs group, which shows that it significantly inhibits myocardial fibrosis. In general, the use of free EVs alone has limited effect on the repair of myocardial infarction. However, a composite hydrogel system can be constructed by encapsulating targeted EVs in a hydrogel. This system can significantly promote cell proliferation and inhibit the process of myocardial fibrosis, providing a good platform for myocardial tissue repair.

Immunofluorescence staining and RT-PCR were used to detect myocardial markers including platelet-endothelial cell adhesion molecule-1 (CD31), vascular hemophilia factor (VWF), and sarcoplasmic reticulum calcium ATPase 2a (Serca2a) protein after four weeks. The expression levels of CD31 can be used to demonstrate the presence of vascular endothelial cell tissue. Figure 5(a) shows the representative immunofluorescence staining images of CD31. Compared with the MI group, the CD31 level expression is significantly upregulated in the composite hydrogels containing EVs, especially in the SF/HPC+FA-EVs-treated group [50, 51]. Figure 5(b) shows the CD31 expression levels in all groups. According to our experiments, the expression of CD31 in the SF/HPC group is about ∼1.8-fold than that in the MI group. This indicates that the hydrogel we have prepared has a certain promoting effect on the cell proliferation when it is used alone, which may be related to the SF component in the hydrogel. After incorporation of EVs into composite hydrogel, the CD31 expression can be increased to ∼2.5-fold in SF/HPC-EVs group and further improve to ∼3.3-fold in SF/HPC+FA-EVs group. Apparently, the expression of CD31 is at the highest expression level in the SF/HPC+FA-EVs group, it may be attributed to its better promotion of cell proliferation and the potential biological function of the hydrogel matrix combined with the multifunctional targeted EVs.

Zoom In Zoom Out Reset image size

Vascular hemophilic factor (VWF) promotes platelet aggregation and adhesion, which has an impact on the development and prognosis of coronary artery disease (CAD) [52]. Figure 5(c) shows the expression levels of VWF for all groups. VWF can be effectively up-regulated expression in all the MI hearts compared with normal hearts in sham group. Among them, the highest VWF expression level can be detected in SF/HPC+FA-EVs group, which may be attributed to the sustaining release of EVs from the hydrogels. Upregulation of VWF expression can accelerate cardiac function recovery. Serca2a is a key protein that maintains myocardial Ca²⁺ homeostasis [53]. Compared to the MI group, Serca2a mRNA is up-regulated in the SF/HPC and SF/HPC-EVs groups (P<0.001). Serca2a mRNA expression in the SF/HPC+FA-EVs group increases to ∼1.8-fold of the MI group, indicating an enhanced myocardial contractility.