Injectable Shear-Thinning Hydrogels with Sclerosing and Matrix Metalloproteinase Modulatory Properties for the Treatment of Vascular Malformations

Abstract

Sac embolization of abdominal aortic aneurysms (AAAs) remains clinically limited by endoleak recurrences. These recurrences are correlated with recanalization due to the presence of endothelial lining and matrix metalloproteinases (MMPs)-mediated aneurysm progression. This study incorporated doxycycline (DOX), a well-known sclerosant and MMPs inhibitor, into a shear-thinning biomaterial (STB)-based vascular embolizing hydrogel. The addition of DOX was expected to improve embolizing efficacy while preventing endoleaks by inhibiting MMP activity and promoting endothelial removal. The results showed that STBs containing 4.5% w/w silicate nanoplatelet and 0.3% w/v of DOX were injectable and had a 2-fold increase in storage modulus compared to those without DOX. STB-DOX hydrogels also reduced clotting time by 33% compared to untreated blood. The burst release of DOX from the hydrogels showed sclerosing effects after 6 h in an ex vivo pig aorta model. Sustained release of DOX from hydrogels on endothelial cells showed MMP inhibition (ca. an order of magnitude larger than control groups) after 7 days. The hydrogels successfully occluded a patient-derived abdominal aneurysm model at physiological blood pressures and flow rates. The sclerosing and MMP inhibition characteristics in the engineered multifunctional STB-DOX hydrogels may provide promising opportunities for the efficient embolization of aneurysms in blood vessels.

Graphical Abstract

Combining occlusive, sclerosing, and MMP inhibitory properties enabled the development of a Gelatin-Laponite (STB) hydrogel loaded with Doxycycline (DOX) for abdominal aortic aneurysm (AAA) sac embolization and endothelial ablation. Validated in a patient-derived aneurysm model, STB-DOX gels showed optimal injectability, strength, and aneurysm sac embolization. In vitro, MMP inhibition and endothelial removal were observed.

1. Introduction

An abdominal aortic aneurysm (AAA) is a life-threatening condition caused by a localized dilatation of the abdominal aorta due to a weakening of the aortic wall ^[1]. AAA is the second most common aortic disease after atherosclerosis and the ninth leading cause of mortality, causing more than 200,000 deaths annually worldwide ^[2]. Endovascular aneurysm repair (EVAR) is an effective and less invasive alternative treatment for AAA compared to conventional surgery due to its lower postoperative mortality, shorter hospitalization, and faster recovery ^[3]. The procedure involves the insertion of a folded and compressed stent graft (SG) within a delivery sheath through the lumen of the access vessel ^[4]. The presence of blood flow perfusing the aneurysm outside the SG, known as endoleak, reduces the efficiency of EVAR and expands the aneurysm sac. Sac embolization is a standard method to prevent endoleak following EVAR by injecting an embolic agent to block the blood flow. The endothelial lining of an aneurysm has anti-thrombotic properties that promote endoleak recurrence ^[5], which results in abnormal vessel flow, further contributing to recanalization ^{[5a, 6]}. To date, commercially available embolizing materials have several limitations, including high cost, frequent recurrence, and difficult handling ^[7]. In addition, commercialized embolic agents are largely dependent on their occlusive properties and do not prevent the recanalization process nor address the underlying pathophysiology of aneurysm progression.

Previous studies have shown a significant correlation between chronic inflammation and the underlying pathophysiology of aneurysm growth. Chronic inflammation results in excessive matrix metalloproteinase (MMP) activity, which leads to progressive degeneration of the artery wall ^[8]. MMPs (MMP-2 and MMP-9), a class of zinc endopeptidases, trigger extracellular matrix (ECM) breakdown in AAAs by factors released during the inflammatory responses (Figure 1A) ^[9]. The most frequent proteinase expressed by human AAA tissue explants in vitro is MMP-9, which is actively generated by macrophages at the site of endovascular damage. Furthermore, higher levels of MMP-9 in plasma are associated with aneurysm growth/expansion ^[9]. MMP-2 also plays an important role in the enzymatic degradation of ECM in the aneurysm wall and is present in large amounts in aortic aneurysms ^[9b].

Figure 1. Mechanisms of aortic aneurysm development in arteries.

(A) Localization of the aortic aneurysms and schematics of healthy and dysfunctional aortic endothelium. An aortic aneurysm is formed by released MMPs from the endothelial cells. MMPs, in turn, degrade elastin fibers in the stroma, weakening the aortic wall and causing aneurysm formation. (B) schematic representation of two-stage release of DOX from STB hydrogel to induce Endothelial removal and MMP inactivation.

Doxycycline (DOX) is a tetracycline antibiotic licensed by the US Food & Drug Administration (FDA) and Health Canada for MMP suppression and endothelial denudation ^[10]. Endothelial denudation of aneurysms can reduce the persistence of endoleaks ^{[5a, 8a, 11]} and can improve angiographic effects by impeding recanalization after coil blockage in arteries. Local injection of DOX can cause endothelium injury, leading to local hemostasis and enabling occlusion ^[12]. Therefore, local delivery of DOX to the aneurysm site has the potential to effectively alleviate aortic expansion in the endoleak treatment of EVAR through endothelial denudation and MMP inhibition (Figure 1B) ^[8a].

In our previous study, we developed an injectable shear-thinning biomaterial (STB) based on silicate nanoplatelets (SNs) and gelatin (Gel-SN) nanocomposites for minimally invasive endovascular embolization ^[13] wherein the hemostatic function of SNs and hydrogel cohesion was enhanced via sodium phytate (Phyt) additives ^[14]. The resulting hydrogels showed great potential as an injectable matrix for integration with microcatheters in EVAR. These effects are critical in hydrogel stability and successful embolization for blockage of vasculatures in vivo ^[13]. In the present study, we aim to develop a sclerosing and MMP-inhibiting STB to promote endothelium denudation for more stable and efficient occlusion in treating AAAs using minimally invasive platforms. For this purpose, DOX additives are incorporated into the injectable shear-thinning Gel-SN biomaterials. The charge interactions introduced by DOX strengthen the cohesion and embolization efficacy of hydrogels. These charges enable additional hemostatic effects and a more facile injectability through catheters for minimally invasive delivery. Furthermore, the burst release of DOX additives provides endothelial ablation while its sustained release promotes MMP inhibition, leading synergistically to a durable and stable occlusion. Thus, the hydrogels display significant promise as an injectable platform for treating endoleaks and preventing aneurysm progression.

2. Results and Discussion

2.1. Characterization of injectable hydrogels

Injectable STBs based on Gel-SN (containing 1.5% w/v gelatin and 4.5% w/v SNs) hydrogels are versatile platforms for minimally invasive endovascular embolization due to their robust structure and hemostatic function ^{[13, 14]}. As a combinatorial strategy, DOX was incorporated into the Gel-SN hydrogels (Gel-SN-DOX α hydrogels where α represents the DOX content in % w/v) to improve embolizing efficiency through aneurysm endothelial ablation and MMP activity inhibition. As shown in Figure 2A, the surface charge distribution of SNs is anisotropic with positive edges and negative surfaces. Likewise, DOX in neutral pH is a zwitterion consisting of positive and negatively charged functional groups ^[15]. Thereby, DOX can interact with SNs electrostatically via the positively charged NH₃⁺ groups and negatively charged carboxylic groups present in DOX interacting with the negative surfaces and positive edges of SNs, respectively. To confirm the effect of DOX additives on hydrogels’ net charges, the zeta potential of the hydrogels was measured (Figure 2B). Our results showed that SNs had a dominant negative surface charge of −25 mV while DOX showed a dominant positive charge of +19 mV. Introducing DOX to SNs resulted in a significant decrease in the zeta potential of DOX, indicating possible electrostatic interaction between the two moieties. When mixed with gelatin, the zeta potential of Gel-SN-DOX increased due to the introduction of more positive charges by gelatin. Based on previous studies, the additional intermolecular electrostatic interactions introduced by DOX improve the hydrogel integrity and mechanical properties ^[14].

Figure 2. STBs characterization.

(A) The illustrative presentation of Gel-SN-DOX preparation procedure. The Gel-SN-DOX α hydrogels (where α represents the DOX content in % w/v) contain 1.5 w/v gelatin and 4.5% w/v SNs. (B) Zeta potential of the biomaterial composites dissolved in DI water. (C) Viscosity measurements of the hydrogels. (D) Time sweep tests with alternating strain amplitudes. (E) Schematic of the injection force measurement setup. (F) Injection force response of Gel-SN. (G) Results of injection force at 33.33 mm/min. (H) SEM images of the hydrogels (scale bar: 100 μm, magnification: 500X). One-way analysis of variance (ANOVA) with Tukey post hoc comparisons was performed to determine statistically significant differences (****p < 0.0001, ***p < 0.001, **p < 0.05, *p < 0.01, and ns denotes nonsignificant, n=3 per group).

The impacts of DOX-triggered interactions on the mechanical characteristics of Gel-SN hydrogels were studied using rheological analyses. According to Figure 2C, the viscosity of all the hydrogels decreased by increasing the shear rate. This effect can be attributed to the electrostatic network of SN which is mechanically sensitive to shear stresses. Although a lower concentration of DOX (0.1% w/v) did not affect the viscosity of Gel-SN, further addition of DOX (0.3% w/v) significantly increased the viscosity from ~300,000 to ~700,000 Pa.s. Viscosity dropped more rapidly with the shear rate at higher DOX concentrations of 0.5% or 1% w/v (Figure S1A).

Figure 2D demonstrates the impacts of shear-thinning and rheological restoration in time sweeps with alternate cycles containing high (100%) and low strains (0.01%). All formulations were able to regain their mechanical strength at low strain periods after the breakage of the hydrogel network at high strain intervals due to the reintegration of lost ionic interactions. Gel-SN-DOX 0.3% demonstrated higher G’ than Gel-SN controls at low strain amplitudes (G’≈4000 Pa for Gel-SN-DOX 0.3 compared to G’≈2000 Pa for Gel-SN controls), which we ascribe to the more potent electrostatic attractions and a more cohesive network with the incorporation of charged DOX compounds. In contrast, at larger strain amplitudes, the broken network of Gel-SN-DOX 0.3 was significantly softer than Gel-SN (G’≈5 Pa for Gel-SN-DOX 0.3 compared to G’≈50 Pa for Gel-SN controls), suggesting an easier injectability (Figure 2E–G). However, further addition of DOX up to 0.5% and 1% w/v (Figure S1B) reversed this trend. STB-DOX 0.5% and STB-DOX 1% were considerably softer than Gel-SN at larger strain amplitudes (Figure 2D and Figure S1B), which corresponds to lower loads for injection (injection force≈7 N for STB-DOX 0.5) ^[16]. The amount of SN was optimized (4.5% w/w) based on higher storage moduli and higher cohesiveness with easier injection force. Increasing the amount of SN to 6% w/w increased the viscosity and storage moduli but also considerably increased the injection force and subsequently blocked the catheters (Figure S1C, D).

Injectability is a fundamental requirement for the minimally invasive application of embolic agents ^{[13, 14, 16]}. STBs are appealing because they soften when shear-inducing injection forces are applied, making them easier to flow and for surgeons to inject ^[17]. The experimental setup depicted in Figure 2E was utilized to characterize the injectability of hydrogels. Figure 2F displays a typical injection force measurement of hydrogels running through a 2.8F catheter. The injection force rapidly increases until the electrostatic connections in the SN structure are destroyed with the maximal force followed by a plateau related to the hydrogel dynamic flow ^[13]. A parametric analysis was conducted to comprehend the injection force for various injection rates, outlet kinds, and hydrogel compositions. The required force for hydrogel injection through a needle (18G and 23 G) and catheter (2.8F) at a 33.33 mm/min injection rate (100 mm/min is shown in Figure S2A) was assessed and presented in Figure 2G. The needle gauges and injection rates were selected in accordance with the standard needle and syringe-catheter setup used in clinical settings ^{[13, 14]}. Based on the obtained results, a more considerable injection force was needed to inject the hydrogels at larger flow rates (Figure S2) and smaller diameters of the outlet (Figure 2G) ^{[13, 14]}. Additionally, incorporating 0.3% w/v DOX into the gel (Gel-SN-DOX 0.3) reduced the force (13 ± 0.4 N) needed to inject Gel-SN (32.5 ± 1.14 N), due to the material’s greater potential to flow under shear and thanks to enhancing electrostatic interaction.

Scanning Electron Microscopy (SEM) was conducted to characterize the microscale structural properties of freeze-dried hydrogels (Figure 2H). SEM analysis showed that the pore size of the STB gel in the absence of DOX exhibits a notable degree of heterogeneity, spanning a broad range from 20 to 150 μm (Figure S3). However, when DOX is introduced into the STB gel, the pore size distribution becomes more homogeneous, clustering mainly in the range of 20 to 50 μm. This result suggests that DOX influences the STB gel’s pore structure, resulting in more uniform and narrow pore size distribution. The pore-size homogenization effect could be explained by the electrostatic interaction between the STB gel and DOX, as indicated by the zeta potentials of the composites (Figure 2B). Due to the distinct charge properties of STB (anionic) and DOX (cationic) molecules, these components undergo electrostatic interactions, resulting in consistency of pore structure and controlled pore size.

2.2. Hemostatic function and hemocompatibility of STBs

Blood clot formation around the hydrogels is a key factor for maintaining stability in the aneurysm sac, as it prevents biomaterial disintegration and endoleaks ^[14]. The hemostatic effects of DOX supplements in Gel-SN-based platforms are examined using an in vitro coagulation test (Figure 3A, B) ^[18]. According to Figure 3B, the STB and STB-DOX hydrogels significantly reduced the clot formation time compared to untreated blood control. This is attributed to the electrostatic charges of the SNs activating the coagulation cascade ^[19]. Specifically, the STB gel without DOX significantly reduced the blood clotting time by 15% relative to the blood controls. Additionally, the incorporation of DOX (0.3% and 0.5% w/v), further decreased the clotting time by 33% (i.e., 14 min compared to 21 min blood control), with no significant difference between 0.3% and 0.5% w/v DOX content.

Figure 3. Hemostatic effects and hemocompatibility of STBs.

(A) Results of in vitro clotting time assay. (B) The effects of DOX additives on the relative decrease in clotting time (%). (C) Images of the hemolysis test. (D) Quantification of blood hemolysis after exposure to STB hydrogels, normalized to DI water treatment. For the quantification results, ((B) and (D)), one-way analysis of variance (ANOVA) with Tukey post hoc comparisons was performed to determine statistically significant differences (****p < 0.0001, and ns denotes nonsignificant, n=3 per group).

The improved hemostatic function with DOX additives supports the role of electrostatic charge interactions in promoting blood coagulation, which is critical for biomaterial stability in the aneurysm sac and eventually in successful occlusion.

Hemolysis experiments were performed to investigate the hemocompatibility of the hydrogels ^[20]. The STB hydrogels without DOX and with low doses of DOX (0.1% and 0.3% w/v) showed a hemolysis percentage below the 5% standard safe limit ^[21]. The hemolysis values for the STB gel, STB-DOX 0.1%, and STB-DOX 0.3% were 4.1 % ± 0.0, 4.1 % ± 0.1, and 4.4 % ± 0.1, respectively. However, the addition of DOX at 0.5% w/v, increased the hemolysis percentage to above 5% (6.1 % ± 0.1). Thus, the STB-DOX hydrogels with 0.3% w/v DOX content represent a suitable formulation for subsequent optimizations.

2.3. Swelling and degradation

Excess swelling can cause tissue compression; thus, hydrogel water uptake should be minimal to avoid damage to the surrounding arterial tissues ^[14].For this purpose, the swelling ratio (%) of hydrogels in response to incubation in PBS was monitored over time (Figure 4A). The results suggest that all STB conditions reached a maximum swelling of ~5% to ~15% within the first 20 min and stayed at its equilibrium for the remaining duration of the study. Incorporation of DOX did not significantly affect the swelling behavior of Gel-SN hydrogels, suggesting that the amounts of DOX added to hydrogels do not increase the risk of tissue compression.

Figure 4. In vitro physical and biological characterization of STBs.

(A) The swelling kinetics of hydrogels. (B) Degradation study of STBs in terms of mass remaining in response to incubation time in PBS. (C) Release of DOX from the hydrogels in PBS. (D) MMP2/9 inhibition in HUVECs after treatment with STB hydrogel formulations. (E) HUVEC viability after the treatment with STBs for up to 72 h. (F) Live-dead imaging of the HUVECs treated with STBs for up to 7 days. One-way analysis of variance (ANOVA) with Tukey post hoc comparisons were performed to determine statistically significant differences (****p < 0.0001, ***p < 0.001, **p < 0.05, and *p < 0.01, n=3 per group). Scale bar = 100 μm.

To examine the degradability of composite hydrogels, STBs were incubated in PBS at 37 °C for up to 175 h (Figure 4B). The degradation study was performed similarly in citrated human plasma to evaluate the effect of MMPs and relevant physiological enzymes on degradation rate of the hydrogels ^[22](Figure S4A). The results showed that all STB formulations were stable with minimal degradation both in PBS and plasma across the tested period (Figure S4B). No significant differences were observed in the total remaining mass between Gel-SN and Gel-SN-DOX hydrogels. The slow degradation of the STBs is advantageous for the long-term occlusion of the aneurysm sac, minimizing the risk of recanalization and reintervention ^[11].

2.4. In vitro cytocompatibility and MMP inhibition via DOX release

The release profile of DOX from the STBs was examined at 37 °C in PBS (Figure 4C) and citrated human plasma (Figure S5A). All the formulations exhibited a burst release in ~6 h, followed by a sustained release with a decayed release rate over 72 h. Both PBS and citrated human plasma showed similar DOX release pattern with no significant difference until day 5 (Figure S5B). This release profile makes the STBs a good fit for endovascular embolization as (i) the short-term burst release phase facilitates removal of the endothelial cells, and (ii) the long-term sustained release enables the MMP inhibition process.

The effect of DOX concentrations on endothelial cell and macrophages activity was assessed via MMP activity and cellular cytotoxicity (Figure 4D, E and Figure S6). The effects of DOX released from STBs on the suppression of MMP-2 and MMP-9 expressed by human umbilical vein endothelial cells (HUVECs) and mouse macrophage cell line (ATCC RAW 264.7) were investigated by a commercial MMP kit (Figure 4D, Figure S6). The results of endothelial’ s MMP activity were presented as the ratio of inhibition relative to no treatment conditions. It was found that the addition of 0.3% w/v DOX boosted MMP inhibiting effects after 7 days. This agrees with the DOX release results, where a plateau was observed after 72 h.

The MMP inhibition analysis of the RAW mouse macrophage cell line showed that the highest inhibition is observed in the STB-DOX0.3 group. Day 1 and 4 eluates showed significant differences between STB, STB-DOX0.1, and STB-DOX03. A trend of increased MMP inhibition was observed with increasing DOX concentrations (Figure S6). The data from this study indicates that DOX has the ability to inhibit MMPs regardless of the types of cells that are producing them.

Cell viability of HUVECs was assessed via PrestoBlue and live-dead assays. The PrestoBlue assay was conducted on HUVECs in wells treated with eluates from STBs, and the results were normalized to a no treatment group to obtain cell viability (Figure 4E). Cell viability was measured for samples that eluted the drug for 24, 48, and 72 h. As the DOX concentration increased, cell viability dropped significantly. STB-DOX 0.5% showed less than 50% viability across the measurements. Live-dead staining results confirmed the results of PrestoBlue tests (Figure 4F). As the DOX concentration increased to 0.5% w/v, more dead cells (red) appeared compared to the live cells (green). Hence, according to the release, cell viability, and MMP inhibition results, the STB-DOX 0.1 and 0.3 groups were selected for further evaluation.

2.5. Benchmark occlusive properties of Gel-SN-DOX composite STB

The ability to impede blood flow is a crucial consideration because SN-DOX gels are designed to embolize blood arteries and aneurysms; however, this ability cannot be deduced from rheometry data. By simulating the in vivo circulation patterns of an endoleak: blood flow, pressure, viscosity, and endoleak anatomy ^[23] an in vitro bench test was designed to assess embolization capabilities. This model did not include a cellular component since it was primarily designed to evaluate the mechanical properties of our gels. The efficiency of Gel-SN-DOX hydrogels in occluding blood flow was tested using an in vitro bench test, presented in Figure 5A. An endoleak was reproduced in vitro by a specific phantom design based on the typical size of the aneurysms (diameter: 44mm, length: 60 mm), presenting a fusiform aneurysm (ELASTRAT Sàrl (Geneva, Switzerland)). The phantom is installed around a 3D-printed tube that imitates a stent graft, and the circulation passes between the phantom and said tube.

Figure 5: Benchmark test diagram and results of the injectable STB hydrogels.

A- Schematic representation of the endoleak bench test. B- A typical flow rate changes in the system while the gel is injected through the aneurysm sac and blocks the model. C- the embolization success rate for each formulation. The data are plotted for n=10 per group.

In brief, the phantom is part of a closed circuit fed by a pulsative pump (Shelley medical imaging technologies (Ontario, Canada)) and the solution passing through the circuit is blood-mimicking (Shelley medical imaging technologies (Ontario, Canada)). The pump pushes the blood-mimicking solution with a pressure close to the average systemic pressure of the human body (200–220 mmHg) ^[24], and the flow is monitored after embolization of the lumbar arteries. Our main goal is to fill the aneurysm sac with the embolizing agent to stop endoleak. The embolizing agent was delivered using a clinical catheter introduced through the aneurysm sac. As a result, it would be expected that the flow through the lumbar arteries is blocked. The proximal flow value after the endoleak embolization is the first endpoint, with an optimal endpoint being a flow of zero; thus, there should be no flow in the endoleak after the procedure. As a secondary endpoint, migration of the gel after embolization would be an essential point to be noted. An optimal outcome is a migration distance of 0 cm starting from the end of the lumbar arteries. The stages of hydrogel injection and simultaneous flow changes in the lumbar arteries are shown as diagrams in Figure 5B. Blood flow circulation in the system was set as the flows in the aneurysm model after stent insertion reach about 15–20 ml/min, corresponding to the endoleak flow in an actual aneurysm ^[25]. When the injection of the hydrogel into the aneurysm sac was initiated, the flow decreased gradually for up to 8–10 min. After 10 minutes post embolization, gel breakage or fluid leakage, shown by flow increment in lumbar arteries, was considered a failed embolization. In contrast, a constant zero flow at the sac was considered a successful embolization. After each trial, the phantom was opened, cleaned, and reused. The success rates of embolization were compared between various formulations in Figure 5C. STB-DOX 0.3 hydrogel presented a higher embolization success rate compared to other formulations. This could be explained by the higher electrostatic intermolecular interaction in STB-DOX 0.3 hydrogels, which considerably improved STBs’ occlusion efficacy and embolization success rate by 50%.

2.6. Effect of Gel-SN-DOX composite STB on endothelial phenotype and morphology

This study investigated the effects of STB-DOX treatment on various proteins secreted from HUVECs using immunocytochemistry. The rationale for conducting this experiment was to understand better how this treatment affects the cellular and molecular mechanisms involved in maintaining the integrity and function of blood vessels. By analyzing the expression of proteins such as CD31, VE-cadherin, ZO-1, VEGFR2, and VWF, the study aimed to gain insight into the potential consequences of STB-DOX treatment on endothelial cell function, angiogenesis, and hemostasis (Figure 6).

Figure 6: The immunocytochemistry experiments conducted on HUVECs with different treatments.

The HUVECs were either untreated (control) or treated with STB gel, STB-DOX 0.1, and STB-DOX 0.3 eluates for days 1, 2, and 4. The cells were probed with antibodies against VE-cadherin, CD31, VEGFR2, VWF, and Zo-1 to analyze the effects of the treatments by each hydrogel (n=3 per group). (A) The merged fluorescence images in (a) left column represents VE-Cadherin (red), CD31 (green), and nucleus (blue), while (b) central column represents merged fluorescence images of VEGFR2 (red), VWF (green) and nucleus (blue). The (c) right-most column represents merge fluorescence images of Zo-1 (red), CD31 (green), and nucleus (blue). Please refer to Figures S7, S8, and S9 for individual images. Scale bar – 200 μm. (B) The fluorescence intensity of respective protein expression, labeled on the Y-axis, was quantified and plotted for untreated cells, STB gel-treated cells, and STB-DOX 0.1 and 0.3 treated cells. The quantification data is color-coded, with grey representing untreated cells, black representing STB gel-treated cells, blue representing STB-DOX 0.1 treated cells, and green representing STB-DOX 0.3 treated cells. The quantification was performed on 3 random images for 20 cells per group, and the ANOVA test was used to compare the groups. Asterisks (*) indicate the significance level, with ****p < 0.0001, ***p < 0.001, **p < 0.05, and *p < 0.01. Day 1, day 2, and day 4 quantification are represented as (i), (ii), and (iii), respectively.

CD31 is a transmembrane protein expressed on the surface of endothelial cells ^[26]. We observed that treatment with STB gel led to an upregulation of CD31 protein expression on days 1 and 2 of treatment. However, treatment with DOX at concentrations of 0.1 and 0.3 resulted in further downregulating of CD31 expression, reducing it to the level observed in untreated samples. To investigate the impact of these treatments on the structural integrity of blood vasculatures, we further quantified the expression of VE-cadherin, a protein involved in preserving the structural integrity of blood vasculatures ^[27], on CD31-positive cells using ImageJ software (Figure 6Aa and 6Ba). The results showed a downregulation of VE-cadherin at a DOX concentration of 0.3, which could be associated with the destabilization of blood vessels and the development of inflammation ^[27]. In addition to observing the changes in CD31 and VE-cadherin expression, we also investigated the effects of STB DOX treatment on ZO-1 expression in CD31-positive HUVECs (Figure 6Ac and 6Bc). ZO-1 is a protein critical for forming and maintaining tight junctions between cells, which help regulate the passage of substances between blood vessels and surrounding tissues ^[28]. We observed the downregulation of ZO-1 in CD31-positive HUVECs with increasing STB-DOX treatment concentrations. This downregulation could affect the formation and function of tight junctions in these cells and have consequences for the integrity of the endothelial barrier.

VEGFR2 (vascular endothelial growth factor receptor 2) is a transmembrane receptor tyrosine kinase expressed on the surface of HUVECs ^[29]. VEGFR2 plays a crucial role in regulating angiogenesis, a function of new blood vessel formation from existing ones. This receptor is activated by binding to specific ligands, such as vascular endothelial growth factor (VEGF), which initiates intracellular signaling cascades leading to the activation of downstream effectors, including cell proliferation, survival, and migration ^[29]. We observed that increasing concentrations of STB-DOX downregulate the expression of VEGFR2 in HUVECs (Figure 6Ab and 6Bb), suggesting that the signaling pathway is compromised. This downregulation could have implications for the proliferation and survival of endothelial cells and the promotion of angiogenesis. VWF (Von Willebrand Factor), a large glycoprotein, is generated and released by endothelial cells, including HUVECs. VWF has a key role in hemostasis, which is the process that prevents blood loss after vascular injury. VWF has several functions in the vasculature, including promoting platelet adhesion to sites of vascular injury, serving as a carrier for factor VIII (a clotting factor), and contributing to maintaining of blood vessel integrity ^[30]. VWF is produced as a large precursor protein called pre-pro-VWF, which is processed into the mature form by proteolysis and glycosylation. The mature form of VWF is a large multimeric protein stored in specialized granules called Weibel-Palade bodies in endothelial cells. Upon activation, VWF is rapidly released from Weibel-Palade bodies and enters the bloodstream, where it binds to platelets and other proteins involved in hemostasis ^[31]. Downregulation of VWF in HUVECs due to STB-DOX treatment could indicate that the drug affects the expression or activity of this protein in these cells (Figure 6Ab and 6Bb). This could have several consequences, including impairing the ability of HUVECs to respond to vascular injury, reducing platelet adhesion and aggregation, and potentially affecting the maintenance of blood vessel integrity.

Overall, the study’s results suggest that STB-DOX treatment may have adverse effects on the expression and activity of these critical proteins in HUVECs, which could impact the health and function of blood vessels.

2.7. Ex vivo assessment of embolization and vascular endothelium denudation

The embolization and vascular endothelium ablation efficiency of Gel-SN-DOX were tested in an ex vivo setting. Porcine aorta samples were tied in one end and injected with Gel-SN-DOX to achieve luminal occupancy. After 6h embolization, samples were fixed, sectioned, and stained for endothelial coverage (Figure 7A). An untreated aorta was studied as a negative control for embolization, and STB gel as a control for DOX-induced vascular endothelium ablation efficiency. Hematoxylin and eosin staining of the untreated aortic sections showed the endothelial cell layer (Figure 7B, denoted by *, hematoxylin positive endothelial nuclei blue), observed at the luminal side of the artery, over the tunica media (elastin fibers: pink, smooth muscle nuclei blue).

Figure 7: Ex vivo tests of the injectable hydrogel for endothelial ablation.

A- The porcine aorta was prepared for the injection, clamping one side before the gel injection (n=3 per group). Lower panel: Porcine aorta after embolization and fixation in formalin. B- Hematoxylin and eosin (H&E) staining of the porcine aorta cross-sections. C- Immunohistochemical (IHC) staining and confocal laser scanning microscopy of the aortic cross-sections with CD31 (blue: DAPI, red: Actin, green: CD31).

The STB-DOX 0.3 group showed complete removal of endothelial cells. The complete ablation is the combinational effect of the shear forces and the ablative effect of DOX in a higher dose. After this initial observation of endothelial ablation with hematoxylin and eosin-stained slides, we performed immunohistochemistry on the sections (Figure 7C). Sections were stained for endothelial marker CD31, phalloidin for cell cytoskeleton, and DAPI for nuclei. Two layers of the aortic wall, tunica intima and tunica media, are discernible in the untreated aortic sections. A green and blue signal on the luminal side showed endothelial cells, and a red and blue signal showed elastin fibers and smooth muscle cells in the tunica media. We observed gradual reduction in the green signal with increasing DOX concentration through our STB-DOX formulation (STB-DOX 0.1, STB-DOX 0.3 and STB-DOX 0.5 groups) which indicate the removal of endothelial layer.

STB-DOX 0.3 showed little to no CD31 signal, proving near complete ablation of the endothelial layer after embolization. One of the common failures in AAA embolization is recanalization after the embolization ^[32]. Lerouge et al. showed that the ablation of the aorta’s endothelial layer significantly decreased the aneurysm sac’s recanalization in an animal model ^[33]. In another animal study, embolization of the arterial lining showed thrombosis in EVAR long-term ^[5a]. Our study’s successive ablation of the endothelial lining shows great promise for the long-term success of the proposed injectable embolizing and ablative Gel-SN-DOX composition for AAA repair using the EVAR method. DOX has been a candidate drug for AAA treatment in systemic administration because it was shown to have endothelial remodeling properties in vitro ^[34]. Unfortunately, its effects on endothelial remodeling in vitro did not translate to in vivo through systemic administration ^[35]. On the other hand, local use could mimic the in vitro observations of previous reports ^[36]. Here, we showed that STB gels loaded with DOX have ablative properties (Figure 7) as well as MMP inhibitory effects (Figure 4) and showed endothelial modulation (Figure 6).

3. Conclusions

In conclusion, our study presented a novel approach to treating abdominal aortic aneurysms using STBs and DOX. Incorporating DOX into an injectable physical hydrogel composed of gelatin and silicate nanoplatelets resulted in a dual purpose of endothelial ablation and MMPs inhibition, with the subsequent extended slow-release stage being beneficial for preserving MMPs inhibition. The STB-DOX hydrogels demonstrated robust mechanical properties and could obstruct tubular structures under physiological pressure. The hydrogels also effectively inhibited MMPs, promoted sclerosing effects, and successfully occluded a patient-derived aneurysm sac model at physiological blood pressures and flow rates. The study also revealed that STB-DOX treatment affects the expression of several key proteins involved in the maintenance of endothelial cell function, angiogenesis, and hemostasis, including CD31, VE-cadherin, ZO-1, VEGFR2, and VWF. These results suggest that these injectable, shear-thinning hydrogels have exciting potential for blood vessel and abdominal aortic aneurysm embolization.

Future research could expand upon the potential of these hydrogels for other applications and optimize their composition and properties for other medical conditions. One potential avenue for investigation is the efficacy of STB-DOX hydrogels in treating aneurysms in different locations in the body, such as cerebral and thoracic aortic and peripheral artery aneurysms. Long-term studies on the safety and efficacy of these hydrogels should be conducted in animal models and clinical trials. These studies should evaluate the potential for adverse effects, such as inflammation or toxicity, as well as the long-term effectiveness of the hydrogels in preventing aneurysm recurrence and improving patient outcomes. Overall, the development of these injectable shear-thinning hydrogels incorporating DOX represents a promising advancement in targeted therapy for vascular conditions.

4. Experimental Section

Hydrogel Preparation

In accordance with earlier methods, stock solutions of 9% w/v silicate nanoplatelets (Laponite XLG, BYK) and 18% w/v gelatin (G1890, Sigma) were made using Milli Q deionized (DI) water (1). Hydrogels were synthesized by mixing gelatin stock and silicate nanoplatelet stock with DI water to attain final concentrations of 1.5% w/v for gelatin and 4.5% w/v for silicate nanoplatelets. This mixing process was carried out using a speed mixer (DAC 150.1 FV-K, FlackTek) at 3000 rpm for a duration of 15 minutes. For samples containing doxycycline hyclate (98% (HPLC), Sigma), the appropriate amount of DOX (0.1, 0.3 or 0.5% w/v) was initially dissolved in the DI water fraction before combining it with gelatin and silicate nanoplatelets. The initial and final concentrations of the hydrogels’ components are outlined in Table S1.

Zeta Potential Measurements

Hydrogel samples (n=3 per group) were dissolved in DI water at 2 mg/ml. Zeta potential was measured at room temperature using Zetasizer Nano ZS, Malvern analytical.

Scanning Electron Microscopy (SEM)

Hydrogel samples (n=3 per group) were freeze-dried for 48 h and their surface was gold coated using South Bay Technology sputter coated. The samples were loaded on an SEM plate using double-sided carbon tape. The SEM images were captured using a scanning microscope (Supra 40VP, Zeiss), with a high resolution at an accelerating voltage of 12 kV and magnification of 500X. Three distinct SEM images of STB, STB-DOX0.1, STB-DOX0.3, and STB-DOX0.5 hydrogels were analyzed using ImageJ software. The pore size of each sample was measured and subsequently plotted as a pore size distribution.

Determination of Swelling

The following formula was used to calculate the percentage of hydrogels’ swelling ratio:

(W_s-W_d) ×100/W_d. The W_s is the weight of the swollen hydrogel and W_d is the initial weight of the wet hydrogel. The prepared hydrogels were submerged in Dulbecco’s phosphate-buffered saline (PBS [21600–010, Gibco]) at 37 °C for predetermined time periods. After soaking, enlarged samples were withdrawn from PBS and immediately weighed. All data were averaged over all replicates (n=3 per group), and the experiments were performed at least five times.

Determination of Degradation

Weighed material samples (200 mg, n=3 per group) were incubated in a 24-well plate either with 1.5 mL PBS or human plasma to mimic the pathological condition at 37°C. Samples were taken out of the media and weighed at predetermined intervals up to 175 h. Material samples were completely dried before being weighed by first being frozen at −80°C for 24 h and then freeze-dried for 48 h. After that, samples were put back into fresh media.

Rheological Testing

To evaluate shear stress, viscosity, and storage modulus, a rheometer (MCR 302, Anton Paar) with a sandblasted parallel plate geometry (25 mm diameter) was employed. Tests were conducted on hydrogels (n=3 per group) at 25°C after bringing all samples to room temperature for equilibration. At a temperature of 25 °C, frequency sweeps from 0.001 to 100 Hz at 1% strain and shear rate sweeps from 0.001 to 100 s⁻¹ at 10 points per decade were carried out. Oscillatory stress sweeps were performed from 0.01 to 100 Pa at 1 Hz. Strain sweeps from 0.01 to 100% were carried out at 1 Hz. Shear-thinning test was done at 1 Hz with a 100% strain applied outside the linear viscoelastic region, then 1% strain for 5 min each cycle to evaluate the gel recoverability.

Injectability Testing

To test injectability of experimental groups, hydrogels (n=3 per group) were placed into a medical syringe (1 mL)-needles (18G and 23G) and a syringe (1 mL)-catheter (2.8F) system. For catheter trials, the catheter tip was submerged in a PBS bath at 37°C. A lower tensile grip held the syringe in place while an upper compression platen depressed the syringe plunger. Two compression rates were tested (33.33 and 100 mm/min). Mechanical testing was performed on an Instron 5943 equipped with a 1 kN load cell. The injection force load was recorded with Bluehill v3 software.

DOX Release Test

To study DOX release rate, STB-DOX gels (5 mL, n=3 per group) were produced, given five minutes to gel, and placed in the 50 ml tube filled either with 15 ml of PBS or citrated plasma at 37 °C. Continuous stirring at a speed of 30 rpm was used to track the dissolution of DOX for a week. At various time intervals, sample aliquots were taken out and subjected to fluorescence analysis at 395 nm. The quantity of released DOX at each time point was determined using a calibrated curve generated from 0–50 mg/mL. The Korsemeyer-Peppas formula was used to calculate the drug release kinetics:

$$\raisebox{1ex}{$M_t$}\!\left/ \!\raisebox{-1ex}{$M_{\mathrm{\infty }}$}\right.=k.t.n$$

M_t/M_∞ is the quantity of medication released at time t, k is a constant describing the dosage form’s structural and geometrical features, and n is the diffusion exponent indicating the release mechanism.

Embolization Test

Figure 5A illustrates the method, which entails using a syringe pump to inject a blood mimicking fluid (consists of 5 μm diameter nylon scattering particles suspended in a fluid base of water, glycerol, dextran and surfactant) into silicon tubes attached to an abdominal aneurysm phantom (Elastrate, Switzerland). The circulation runs between the phantom and the 3D-printed tube, which mimics a stent graft, where the phantom is mounted. The phantom, in a nutshell, is a component of a closed circuit fed by a pulsating pump. A pulsating pump gradually pushed the solution through the circuit, creating a blood-like solution from various substances. To obtain an average systemic pressure of 100 mmHg, the pressure of the system is monitored via a pressure probe located close to the endoleak. A temperature probe monitors the system’s temperature to ensure that the liquid maintains its 37°C temperature. Following lumbar artery embolization, the flow is finally monitored. Our primary objective is to completely fill the endoleak with the embolizing agent using a catheter inserted through the aneurysm sac. The maximum pressure applied to the gel is 220 mmHg, which is higher than the naturally occurring blood systolic pressure. The only exceptions are when the gel ruptures or enables liquid to pass through, which causes a sudden flow to be recorded by the flowmeter and a drop in pressure that is continually recorded by a computer. Final data are provided for n=10 per group including the impact of gel composition on embolization success rate.

MMP Inhibition Assay

MMP 2/9 Inhibition Assay was performed on HUVECs cultured in 12 well tissue culture plates. ATCC vascular cell basal medium (PCS-100-030) with the addition of Endothelial cell growth kit-VEGF (PCS-100–041) was used for culturing HUVECs. Initially, 100,000 cells were seeded into the tissue culture well plates (without additional protein coating) and incubated for 4 h to attach. The cell adhesion was verified using an inverted microscope, before starting experimentation. Cell media was changed with eluates (2ml per well) from the injectable hydrogels (500 mg of gel in 5 ml media) after 4 and 7 days of release. Cells were incubated for 24 h before MMP 2/9 assay. The assay used was Innozyme Gelatinase assay (EMD Millipore CBA003) according to the manufacturer’s protocol. In summary, the cells were incubated in the assay reagent for 3 h, and afterward, the liquid was transferred to a 96-well plate and fluorescence was read (Exc. max: ~325 nm; Emi. max.: ~393 nm). Samples were prepared in triplicate. The negative control was untreated cells, positive controls were free DOX and MMP2/9 substrate. Samples were read (n=3 per group). using Thermo Scientific Varioskan 6.

Additionally, macrophage RAW 264.7 Cells were cultured using DMEM high glucose supplemented with Penicillin-Streptomycin (1% V/V) and FBS (10% V/V). After reaching confluency, the cells were detached using Trypsin and counted. A 200000 cells/ well cell density was employed to prepare 24 well plates. Eluates of STB, STB-DOX0.1, and STB-DOX0.3 gels were prepared by releasing the drug from gels (500 mg of gel in 5 ml media) for 1,4 and 6 days. After overnight incubation of cells for cell adhesion, the experiment was started. The eluates were added to the well plates (n=3) and incubated with the cells for 24 h. The media was collected from the wells and used for MMP inhibition measurements. MMP Activity Assay Kit (Abcam ab112146) was used to detect the MMP activity. Media samples (50 μl) collected from the wells were incubated at 37 °C for 30 mins in 96 well plates, and MMP green substrate was added (50 μl) to the wells. As the blank assay buffer was used. Media was used as a control to normalize the results to baseline growth media MMP activity. After incubation for 1 h at room temperature, the plate was read using a spectrofluorometer (Varioskan LUX, Thermo Fischer). Samples were measured with Ex/Em=490/525 nm. The blank was subtracted from all samples. Results were normalized to growth media readings and presented as MMP inhibition % compared to cells in growth media. Statistical analysis was performed using Prism 9 software with Two-Way ANOVA with Tukey’s post-hoc test.

Cytotoxicity Analysis

The HUVECs were cultured in Dulbecco’s Modified Eagle Medium (DMEM). HUVECs were implanted and allowed to develop on hydrogel-coated slides for five days. After incubation, slides were stained using the PrestoBlue reagent in a Live-Dead cytotoxicity assay. Samples were seen with a fluorescence microscope (Keyence, BZ-X710). To analyze data quantitatively, total cells in five non-overlapping spots were counted at 100x objective using ImageJ software. The viability percentage is assessed (mean of n=3 per group ± SD) by dividing the number of living cells by total cell counts.

Hemolysis Testing

Human whole blood containing sodium citrate anticoagulant (Zenbio, USA) was used to test the hemocompatibility of STB and STB-DOX hydrogels. Blood samples were diluted 50x using saline solution (0.9% (w/v)). Hydrogels (1 ml, n=3 per group) were placed into centrifuge tubes, and diluted blood (1 ml) was added to each sample. The tubes containing diluted blood with DI water and saline served as positive and negative controls, respectively. The samples were placed in a shaker incubator (Labline Instruments) at 37 °C for 2 h, followed by centrifugation at 2000 rpm for 15 minutes (Sorvall Legend X1R, ThermoFisher Scientific). Supernatant absorbance at 545 nm was measured using a microplate reader (Varioskan LUX, ThermoFisher Scientific). A particular equation was used to determine the hemolysis %: Hemolysis (%) = (A_s-A_neg)/A_pos × 100%. Where A_s is the sample absorbance, A_neg is the saline absorbance, and A_pos is the DI water absorbance with diluted blood.Clotting Time Assay

The mixture of citrated human blood and 0.1 M CaCl₂ at a 9:1 ratio vortexed for 10 sec to reverse coagulation. STB hydrogels (n=3 per group) were loaded into syringes and injected (200 μl) into a 48-well plate. Blood (200 μl) was added to each well, and at the certain time points, the wells were rinsed with saline (0.9% w/v) to remove the soluble blood. The wells containing blood without samples served as control. The clotting time was determined when a uniform clot was formed on the surface of the samples.

Endothelial functional evaluation using cell morphology and immunocytochemistry

HUVECs were used for the evaluation of endothelial function. ATCC vascular cell basal medium (PCS-100–030) was used with the addition of Endothelial cell growth kit-VEGF (PCS-100–041) for culturing HUVECs. Cells at a density of 50,000 per well were seeded into 24-well tissue culture plates and incubated for 4 h to adhere. After confirming cell adhesion by an inverted phase-contrast microscope, eluates released from the hydrogels (STB, STB-DOX0.1, STB-DOX0.3 (500 mg of gel in 5 ml media), n=3 per group) for 1, 4, and 7 days were added to the wells and cultured for 48 h. To perform immunocytochemistry, the treated cells were first fixed using 4% paraformaldehyde. The cells were then permeabilized using a 0.1% Triton X-100 to allow the antibodies to penetrate the cells. The cells were blocked with a blocking buffer, 1% BSA, to reduce the nonspecific binding of the antibodies. Primary antibodies specific to the VE-cadherin polyclonal antibody (Invitrogen Catalog # 36–1900), Alexa Fluor^® 488 Anti-CD31 (Abcam ab215911), anti-VEGF Receptor 2 antibody (Abcam ab5473), anti-VWF antibody (Abcam ab201336), and Anti-ZO1 tight junction protein antibody (Abcam ab216880) was then added to the cells in group and incubated at 4 °C temperature overnight. After washing away the unbound primary antibodies, fluorescence tagged secondary antibodies namely goat anti-mouse IgG H&L Alexa Fluor^® 488 (Abcam ab150113) and goat anti-rabbit IgG H&L Alexa Fluor^™ 594 (Invitrogen Catalog # A-11012) were added to the cells and incubated at room temperature for 2 h. The cells were then washed again and mounted on slides using an anti-fade mounting medium with a DAPI to visualize the cell nuclei. Finally, the cells are observed using a fluorescence microscope Echo Revolution, and images are captured and analyzed using ImageJ software.

Ex vivo deendothelialization

Porcine aorta samples were obtained from a third party (Sierra Corp). Samples were trimmed to the same length (3cm), and the lumen was injected with STB containing 0, 0.1, 0.3, and 0.5% w/v of DOX (n=3 per group). An aortic sample was not treated and served as the negative control. All the samples were incubated in RPMI media supplemented with an antibiotic cocktail, FBS (20% v/v), HEPES buffer (25.03 mM), and L-glutamine for 3 h at 37 °C. Subsequently, all the samples were fixed in neutral-buffered formalin (10%) and proceeded with paraffin block preparation. The tissue blocks were cut (8 μm thick), deparaffinized, rehydrated in a series of alcohol, and stained with hematoxylin (Leica Biosystems) and eosin (Sigma) (H&E). Bright-field images were captured using Echo revolution microscope.

For immunofluorescence staining, Alexa Fluor^® 488 mouse monoclonal [JC/70A] antibody to CD31 (Abcam, 215911) was used as an endothelial marker. The samples were counter-stained with Alexa Fluor 594 Phalloidin (Fisher Scientific, USA) and DAPI (Sigma-Aldrich). Finally, the sections were observed using fluorescence Echo revolution microscope.

Statistical Analysis

Statistical analysis of the experimental result was performed using GraphPad Prism software (Version 9.3.0, USA). Except when otherwise specified, all measurements were made at least in triplicate (n=3). These numbers show the mean and standard deviation. One-way analysis of variance (ANOVA) with Tukey post hoc comparisons was used to determine the statistical significance of the difference between the means. Statistically significant differences were denoted as ∗ and specific p values are mentioned in the figure’s legends wherever necessary.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.