Clot‐targeted magnetic hyperthermia permeabilizes blood clots to make them more susceptible to thrombolysis

Abstract Background Thrombolysis is a frontline treatment for stroke, which involves the application of tissue plasminogen activator (tPA) to trigger endogenous clot‐degradation pathways. However, it is only effective within 4.5 h of symptom onset because of clot contraction preventing tPA permeation into the clot. Magnetic hyperthermia (MH) mediated by tumor‐targeted magnetic nanoparticles is used to treat cancer by using local heat generation to trigger apoptosis of cancer cells. Objectives To develop clot‐targeting magnetic nanoparticles to deliver MH to the surface of human blood clots, and to assess whether this can improve the efficacy of thrombolysis of contracted blood clots. Methods Clot‐targeting magnetic nanoparticles were developed by functionalizing iron oxide nanoparticles with an antibody recognizing activated integrin αIIbβ3 (PAC‐1). The magnetic properties of the PAC‐1‐tagged magnetic nanoparticles were characterized and optimized to deliver clot‐targeted MH. Results Clot‐targeted MH increases the efficacy of tPA‐mediated thrombolysis in contracted human blood clots, leading to a reduction in clot weight. MH increases the permeability of the clots to tPA, facilitating their breakdown. Scanning electron microscopy reveals that this effect is elicited through enhanced fibrin breakdown and triggering the disruption of red blood cells on the surface of the clot. Importantly, endothelial cells viability in a three‐dimensional blood vessel model is unaffected by exposure to MH. Conclusions This study demonstrates that clot‐targeted MH can enhance the thrombolysis of contracted human blood clots and can be safely applied to enhance the timeframe in which thrombolysis is effective.


| INTRODUC TI ON
Ischemic strokes are a major cause of death and disability worldwide. 1,2 Clinical interventions aim to rapidly reestablish blood flow through the blocked vessel to minimize tissue damage. Current frontline treatment includes intravenous injection of the thrombolytic agent, tissue plasminogen activator (tPA), which enzymatically breaks down the fibrin scaffold holding together the blood clot to facilitate recanalization of the affected vessel. These thrombolytic drugs can be administered in out-of-hospital or nonspecialist clinical settings by any appropriately trained health care provider. 3,4 However, their use is currently limited by the short time window in which it is effective because clots become resistant to thrombolysis between 3 and 4.5 h after the onset of symptoms. 3,4 The reduced efficacy of thrombolysis at this stage is because of strong compaction of the blood clot, which impairs tPA diffusion into the thrombus. 3,4 Alternatively, mechanical thrombectomy can be used to surgically remove the blood clot and reopen the vessels in patients up to 24 h after symptom onset, leading to improved patient outcomes. 5 Yet these surgical procedures require transfer to specialist hospital units, and time for patient assessment before surgery, which prolongs the delay in restoring blood flow. Therefore, optimizing current thrombolytic techniques to make them safer and more efficient could improve functional outcomes in patients.
Magnetic hyperthermia (MH) uses magnetic iron oxide nanoparticles (IONPs) to generate heat upon exposure to an alternating magnetic field (AMF). By functionalizing with peptides and antibodies that targeted cancer cells, IONPs has been used to generate a highly localized heating response that selectively trigger apoptosis and necrosis in cancer cells, while leaving the healthy tissues surrounding the tumor intact. 6 Initial clinical trials using MH to treat glioblastoma multiform 7 or prostate cancer have yielded promising results. 8 Additionally, current clinical trials include the NoCanTher EU Project, 9 that has recently commenced. More importantly, companies like MagForce AG already received a European CE Certificate to use MH hyperthermia for glioblastoma treatment. 10,11 This same company is in process of obtaining a similar license to use MH for treatment of prostate cancer in the United States. 12 Those examples illustrate that MH is an emerging cancer therapy that will likely be used in clinical practice within the next decade.
Previous studies have demonstrated that heating platelets and red blood cells (erythrocytes) can trigger apoptosis 13 or eryptosis, respectively. 14 Triggering cell death in this way could clear cellular material from the compacted thrombi and thus improve the permeability of tPA into the clot, enabling the effectiveness of thrombolytic therapies to be extended over a wider timeframe. Imaging studies in patients who are experiencing an acute ischemic stroke has demonstrated that increased thrombus permeability was associated with a greater chance of rapid unblocking of the blood vessel (recanalization) after tPA treatment. 15 Additionally, tPA activity has been shown to be enhanced by hyperthermic temperatures, 16 suggesting that localized heating of the surface of the thrombi by IONPs could also act to optimize the enzymatic activity of tPA in the vicinity of the thrombi. Recent studies have demonstrated that it is possible to use functionalized magnetic nanoparticles that specifically bind to activated platelets as contrast agents for radiological imaging of thrombi. [17][18][19][20] However, no previous studies have examined the effect of MH on human blood clots.
Thrombi from ischemic stroke patients have been shown to possess a fibrinolytic-resistant outer shell rich in densely packed platelets. 21 Therefore, using an antibody that targets activated human platelets would provide an effective method for localizing the IONPs specifically to the offending thrombi. In this study, we fabricated IONPs labeled with the commercially available PAC-1 antibody that specifically binds to the active form of the human platelet fibrinogen receptor, integrin αIIbβ3. Because this is only found on activated human platelets, it allows specific targeting of the IONPs to the surface of the blood clot. Here, we have tested the ability of the PAC-1-functionalised IONPs (f-IONPs) to specifically deliver MH to the surface of human blood clots, and whether this could be used to enhance the thrombolytic action of tPA on contracted ex vivo-generated human blood clots.

| Iron oxide synthesis, functionalization, and characterization
Iron oxide nanoparticles were synthesized using alkaline precipitation of iron salts through a modified version of the protocol described by Massart et al. 22,23 IONPs were coated with citrate using the protocol described by Campelj et al. 24,25 As previously described, PAC-1 antibody was covalently attached to the citrated surface by carbodiimide activation of its carboxyl group to create the f-IONPs. 26 The f-IONPs were found to be stable when dispersed into human platelet-poor plasma (PPP) from three different donors for up to 1 h ( Figure S1).
Details on nanoparticle fabrication, functionalization, and characterization are provided in the Supporting Information.

| Blood collection and generation of ex vivo-derived blood clots
This study was approved by Keele University Research Ethics Committee (project reference MH-200154, ERP2335). Blood was collected by venipuncture from healthy volunteers who had given written Essentials • Thrombus contraction reduces the efficacy of thrombolysis by impeding Activase® permeability.
• Platelet-targeted magnetic nanoparticles were used to deliver localized heating to blood clots.
• Clot-targeted magnetic hyperthermia increased the permeability of contracted human blood clots.
• Magnetic hyperthermia can act as an adjuvant to thrombolysis, without damaging the endothelium. informed consent. Blood was mixed either with 1 part blood: 9 parts 3.8% [w v −1 ] sodium citrate solution. Platelet-rich plasma (PRP) was isolated by centrifuging whole blood at 700 g for 8 min, followed by treatment with 100 μM aspirin and 0.1 U ml −1 apyrase. Platelet-poor plasma was prepared by centrifuging PRP at 350 g for 20 min, and the supernatant was collected. In some experiments, platelets were fluorescently labeled by collecting whole blood mixed with sodium citrate containing

| f-IONPs binding to activated platelets and ex vivo-derived blood clots
Functionalized IONPs binding to activated platelets was assessed using a microplate-based assay using goat anti-mouse Alexa Fluor 488-labeled anti-IgM mu chain antibody (Abcam) to detect the presence of f-IONPs. This is described fully in the supplementary materials.
Binding of f-IONPs to ex vivo-derived blood clots was assessed using alternating current (AC) susceptibility measurements and confocal imaging. AC susceptometry is a technique for measuring a magnetic property of materials called complex magnetic susceptibility (χ AC ). Because χ AC is a complex number, it is divided into a real A 150μl sample of the supernatant was extracted and diluted in a further 400 μl HBS. Absorbance of this supernatant was measured at 415 nm using a plate reader according to a previously described method. 30 The remaining thrombi was then extracted from the glass tubes, washed with 10 ml HBS, and weighed on a microbalance.

| Scanning electron microscopy of whole blood thrombi
Following their respective treatments, clots were fixed with White's buffer containing 2.5% [v v −1 ] glutaraldehyde for 2 h at 4°C and prepared according to OTOTO protocol 31 (see Supporting Information).
The clots were imaged using a Hitachi S4500 scanning electron microscope (SEM) operating at 5 KV. Before SEM visualization, samples were randomized by assigning a code and analyzed afterwards by an independent assessor of the team.

| Clot permeability
Twenty microliters of DiOC 6 -labeled PRP was recalcified with 20 mM and emission wavelengths of 490 to 520 and 590 to 620 nm. Threedimensional (3D) images were reconstructed, and mean slice dextran fluorescence was measured using ImageJ software.
Preparation and viability assessment of 3D endothelial cell cultures human umbilical vein endothelial cells (HUVECs) were cultured on 3 mg ml −1 type I collagen hydrogels as previously described. 32 The hydrogels were cultured for 3 days before being used in experiments. period. The hydrogels were then stained using a live-dead cell double staining kit II (Thermofisher Ltd). The viability of the 3D endothelial cell culture was then assessed using a FV300 confocal Olympus microscope using excitation wavelengths of 473 nm and 543 nm, and emission wavelengths of 490 to 520 nm for calcein and 590 to 620 nm for ethidium homodimer III, respectively. Cell viability was quantified in five different regions of the gel by using ImageJ software relating the proportion of live versus dead cells in each hydrogel.

| Statistical analysis
Blood from two to five donors were used in each experiment. n signifies independently tested platelet samples. Values are reported as mean ± standard error of the mean (SErM), except for IONPs physicochemical and morphological characterization, which are mean ± standard deviation (SD). Statistical significance was assessed by one-way anova for repeated measurements with a post hoc Tukey test using OriginLab Pro 8.5 software. A p < .05 value was considered statistically significant.

| PAC1-functionalised IONPs bind to ex vivogenerated blood clots
To develop clot-targeted magnetic nanoparticles, citrate-coated IONPs were functionalized with PAC-1 antibody (illustrated schematically in Figure 1A). The production of IONPs was assessed using transmission electron microscopy ( Figure 1B and Figure S1). To assess whether the PAC-1 antibody on the surface of the f-IONPs retained its functionality, we assessed binding of f-IONPs to resting and activated platelets using the Alexa Fluor488-labeled anti-IgM antibody, demonstrating an increase of fluorescence and therefore biding of nanoparticles to the well plate only when activated platelets are present (Figure 2A). To assess whether the f-IONPs were also able to effectively label ex vivo-derived human blood clots, samples were incubated with f-IONPs and their magnetic signature detected by AC susceptibility measurements. AC magnetic susceptibility measurements enable to detect the immobilization of magnetic nanoparticles in biological entities through monitoring the imaginary part (χ″) of complex magnetic susceptibility (χ AC ) in f-IONPs (see Section 2). As shown in Figure 2B, F-IONPs dispersed in PPP exhibited a prominent Brownian relaxation of the f-IONPs, as indicated by the wide curve with a peak around 100 Hz in the imaginary part of the susceptibility (χ″, Figure 2B, black arrow). This indicates that f-IONPs can rotate freely in PPP. 27,28,33 In contrast, f-IONPs incubated with blood clots in PPP showed a significant reduction in the imaginary part of the susceptibility signal (χ″, Figure 2B To be effective at delivering MH to a blood clot, the f-IONPs must be able to generate significant heat when stimulated with an AMF after any functionalization procedure. Because magnetic properties strongly determine f-IONPs heating capacity, AC magnetization loops were measured before and after nanoparticle functionalization. It was found that functionalization did not significantly modify the magnetic properties of IONPs under AMF ( Figure 2D).
To determine the AMF conditions required for optimal heat generation, the specific absorption rate (SAR), was measured in samples contained in a polystyrene-shielded glass vials ( Figure 2E  The highest SAR was found when using an AMF at 306 kHz and 30 mT (1.3 ± 0.1 kW g Fe −1 ; Figure 2F). These AMF parameters were then used to assess the ability of different concentrations of f-IONPs (0.021, 0.089, and 0.33 g Fe L −1 ) to heat PPP prewarmed to 37°C when exposed to this AMF for 30 min. As shown in Figure 2G, the f-IONPs were able to increase the temperature in a dose-dependent manner. The highest concentration of 0.33 g Fe L −1 f-IONPs was able to induce hyperthermic temperatures within 10 min of AMF application. Thus, this f-IONP concentration was initially used to assess the effect of clot-targeted MH on the clot breakdown.

| Clot-targeted MH enhances the thrombolytic action of tPA on contracted blood clots
To test whether clot-targeted MH could enhance the thrombolytic activity of tPA on contracted thrombi, whole human blood was stimulated to clot by addition of ADP, and allowed to contract fully for visibly seen to be more disrupted, with a larger sedimentation of released red blood cells observed in the plasma below the clot when compared to those treated with tPA alone ( Figure 3A). These results qualitatively suggested an accelerated degradation of the blood clots when MH was used in combination with tPa. To quantitatively assess these initial visual observations, the clots were retrieved and weighed on a microbalance. Additionally, absorbance readings of plasma samples were used to assess the release of red blood cells from the clots as well as hemoglobin from damaged red blood cells using a previously described method. 30 Plasma samples from clots treated with both tPA and MH exhibited significantly higher values of absorbance at 415 nm (275 ± 33% of control) than those treated with tPA alone (213 ± 23% of control; n = 15; p < .05; Figure 3B). These data demonstrate that there was greater release of red blood cells and hemoglobin from the clots into the surrounding plasma. This enhanced dissolution of the clots was also observed in the clot weights. MH reduced the weight of tPA + MH-treated clots to 75.6 ± 2.7% of untreated control clots, whereas those treated with tPA alone only reduced to 81.3 ± 3.6% (n = 15, p < .05; Figure 3C). These data demonstrate that MH enhances the thrombolytic activity of tPA on contracted human blood clots.
Clot-targeted MH treatment elicits both localized heating responses at the surface of the clot, as well as macroscopic heating of the surrounding solution. Therefore, to assess whether the enhanced thrombolysis observed was due to the macroscopic heating of the surrounding solution experiments were also performed to assess the effect of heating samples to the macroscopic temperatures elicited by MH seen in Figure 2G

| Thrombus-targeted MH increases permeability to tPA
To examine if MH treatment was able to enhance the permeability of blood clots to tPA, ex vivo-derived clots were fabricated in the same manner as before except using PRP samples instead of whole blood.

| Clot-targeted MH enhances tPA-mediated fibrin breakdown and triggers erythrocyte disruption on the surface of contracted blood clots
Scanning electron microscopy was used to assess how MH was assisting tPA in triggering breakdown of the contracted blood clots.
Contracted blood clots were treated with either tPA alone, MH alone or both in combination before fixation and imaging by SEM.
An untreated clot was also used a control. As expected, treatment of tPA alone caused a loss of fibrin on the surface of the blood clot when compared with the untreated control ( Figure 5A,C). Clots exposed to MH in the absence of tPA were observed to have less red blood cells present on the surface 14 ( Figure 5B). However, fibrin fibers appeared to be unaffected by MH alone. Last, combined treatment of tPA and MH appeared to remove almost all of the visible fibrin fibers from the surface suggesting that MH can enhance the thrombolytic activity of tPA ( Figure 5D1). Additionally, as with MH treatment alone, red blood cells were seen to be smaller, more spherical structures ( Figure 5D1) with a few cells showing membrane blebbing in some samples, potentially indicative of eryptosis 9 (red arrows in Figure 5D2). These results are consistent with previous studies that have shown that heating enhances the thrombolytic actions of tPA 16 and triggers eryptosis. 14 However, we cannot rule out the possibility that the change in erythrocyte morphology observed is due to heat-induced necrotic damage of these cells. These data also indicate that the localized heating of the clot does not adversely impact fibrin structure or cause it denaturation to reduce its susceptibility to fibrinolysis. The results presented here suggest that a local heating effect induced by MH at the surface of the clot could elicit this adjuvant effect on thrombolysis both by directly increasing the fibrinolytic activity of tPA, 11 as well as enhancing tPA diffusion into the clot through disruption of erythrocytes on the clot surface.

| MH using an optimized dose of f-IONPs has no significant effect on the viability of the surrounding endothelium
Because MH could elicit unintended thermal damage to the surrounding blood vessel, we next sought to identify the minimal MH

F I G U R E 4 Legend on next page
F I G U R E 5 Clot-targeted MH enhances tPA-mediated fibrin breakdown and triggers greater red blood cell release from the surface of highly contracted ex vivo-generated blood clots. Representative SEM images of ex vivo whole blood clots surfaces of (A) nontreated or (B) treated with 0.33 g Fe L −1 f-IONPs mediated MH, (C) 2 μg tPa, and (D1,2) a combined treatment with the same amount of tPa and 0.33 g Fe L −1 f-IONPs mediated MH for 30 min. The control group was held at a temperature of 37°C for the same period as the MH treatment. Image D1 was obtained from the same donor as figures A-C and demonstrates the presence of spherical erythrocytes and a loss of fibrin on the surface of these clots. Image D2 is taken from another donor and shows the presence of erythrocytes with membrane blebbing (red arrows). AMF conditions: 306 kHz and 30 mT. n = 3. 185.5 ± 15.4% of control, respectively; n = 11; Figure 6B). Although the MH elicited by both f-IONP doses also tended to enhance the tPA-mediated reduction in clot weight, this was only statistically significant for the higher dose of 0.089 g Fe L −1 ( Figure 6C). These data suggest that this lower dose can still enhance the thrombolytic action of tPA while maintaining the temperature of the surrounding plasma below 42°C. This is lower than the 45°C elicited in our earlier experiments, as well as previous studies using untargeted magnetic nanocubes functionalized with tPA to enhance dissolution of murine blood clots (49°C). 35 These data therefore demonstrate that surface- HUVEC culture. The HUVEC-coated surface was exposed to a 500μl cell culture sample containing either no f-IONPs or 0.089 g Fe L −1 f-IONPs, and incubated for 30 min in either the presence or absence of the same MH treatment previously used ( Figure 7A). HUVEC cell viability was then assessed using live/dead cell staining to assess whether MH could damage the endothelial lining ( Figure 7B). As shown in Figure 7D, HUVEC hydrogels treated with MH mediated by 0.089 g Fe L −1 f-IONPs showed no significant difference in cell viability (90.9 ± 1.4%, n = 8) in comparison with HUVECs exposed to media containing no f-IONPs and not subjected to an AMF (93.2 ± 1.1%, n = 8). Additionally, exposure to either the AMF or f-IONPs alone did not produce a significant reduction in cell viability (95.2 ± 0.9% and 93.5 ± 1.4% cell viability, respectively; n = 8). Additionally, similar tests performed 24 h after each treatment lead to insignificant differences between the control (94.7 ± 1% of cell viability; n = 8; Figure 7E) and samples treated with AMF, f-IONPs, or the combination of AMF and f-IONPs (95.8 ± 0.6; 93.8 ± 1.8 and 95.7 ± 0.6 of cell viability, respectively; n = 8; Figure 7E). These data therefore indicate that the macroscopic heating of PPP could enhance thrombolysis without impacting on the viability of the nearby endothelium.

| D ISCUSS I ON AND CON CLUS I ON S
The time window to successfully treat ischemic strokes using thrombolytic therapies is limited to 3 to 4.5 h. After this period, the blood clot becomes resistant to thrombolysis from the inability of tPA to penetrate the highly restricted intercellular spaces in the contracted clot. Both the platelets as well as the presence of polyhedrocytes (erythrocytes deformed by clot contraction) have been suggested to impair thrombolysis by preventing tPA penetration into the core of the clot. 4 Specific heating of the clot surface increased clot permeability both by directly increasing the fibrinolytic activity of tPA, as well as through disrupting erythrocytes on the surface to increase clot permeability to tPA. Interestingly, increased thrombus permeability was previously found to be predictive of successful early recanalization of obstructed blood vessels. 15 Although the targeted heat delivery elicited by the f-IONPs to the surface of the clot could disrupt its structure, the macroscopic heating of the surrounding plasma was crucially not sufficient to affect the viability of endothelial cells, indicating that this thrombus-targeted MH should not elicit adverse side effects through damaging the surrounding vasculature. The feasibility of using MH to induce localized heating in the brain for therapeutic applications has been previously demonstrated in humans. 36 In this case, a magnetic field generator was safely used to apply MH as an anticancer therapy that increased survival times in glioblastoma patients. The results reported here are the first to demonstrate the potential of clottargeted MH to enhance thrombolysis in humans and suggests that this could be used as a viable method to improve outcomes in stroke patients. Because the PAC-1 antibody is specific for human platelets, it is not possible to currently conduct experiments on in vivo thrombosis models. Additional work will be required to produce, optimize the magnetic properties, and validate the effectiveness of analogous JON/A-functionalized IONPs to elicit clot-targeted MH before can be conducted to confirm the in vivo efficacy and safety of this treatment. 37 The ability of MH to enhance the susceptibility of contracted blood clots to tPA, could provide us with a method to increase the time window in which clot lysis is effective. This is especially important for communities who do not have access to the specialist personnel or surgical facilities required to undertake thrombectomy. Thus, clot-targeted MH could significantly reduce the incidence of death and disability in people experiencing ischemic stroke globally.
Further development of the PAC-1-functionalized IONPs could provide additional benefits of this magnetic nanotechnology in the diagnosis and treatment of ischemic stroke. This includes using the nanoparticles to specifically target release of tPA at the clot, thus reducing the risk of intracranial hemorrhage from systemic application of this drug. 35 Additionally, further development of the composition and size of the magnetic nanoparticle core could allow these clot-targeted nanoparticles to be used as a contrast agent to facilitate delivery of the magnetic nanoparticles into occluded blood vessels 38 the imaging of blood clots, 39 or for magnetic filtration of labeled emboli to prevent these lodging in other brain regions or in the lungs, minimizing the secondary consequences of thrombolysis. 40 Last, we recently demonstrated citrate-coated magnetic nanoparticles such as these are able to slow platelet activation, which could further prevent additional thrombi forming. 41 Development of these multifunctional magnetic nanoparticles could provide a theranostic agent that is able to rapidly target, treat and facilitate the removal of thrombi from the cerebral circulation.