Fibrin‐specific poly(N‐isopropylacrylamide) nanogels for targeted delivery of tissue‐type plasminogen activator to treat thrombotic complications are well tolerated in vivo

Abstract Targeted drug delivery for maintaining blood fluidity can reduce the risks associated with systemic anticoagulants that can lead to off‐target bleeding. Recently, there has been much interest in targeted delivery of tissue‐type plasminogen activator (tPA) for treating thrombotic complications. The work presented here characterizes a fibrin‐specific nanogel (FSN) design for targeted delivery of tPA to treat thrombotic complications. Fibrin binding and clot degradation were characterized in vitro, and animal models of thrombosis were used to examine nanogel effects on coagulation parameters. In vitro assays showed tPA‐FSNs attach to fibrin in a dose‐dependent manner independent of tPA loading. In animal models of thrombosis, including an electrolytic injury to monitor clot properties in real time, and a lipopolysaccharide‐induced disseminated intravascular coagulation (DIC) animal model, tPA‐FSNs modulated fibrin/fibrinogen and platelet incorporation into clots and at optimized dosing could recover consumptive coagulopathy in DIC. Distribution of unloaded and tPA‐loaded FSNs showed potential clearance of tPA‐FSNs after 24 h, although unloaded FSNs may be retained at sites of fibrin deposits. Maximum tolerated dose studies showed tPA‐FSNs have minimal toxicity up to 20 times the optimized therapeutic dose. Overall, these studies demonstrate the therapeutic efficacy of targeted fibrinolysis for systemic microthrombi and begin to evaluate key translational parameters for tPA‐FSN therapeutics, including optimal tPA‐FSN dosage in a DIC rodent model and safety of intravenous tPA‐FSN therapeutics.


| INTRODUCTION
Maintaining blood fluidity and hemostatic balance is critically important in the vascular system. Initiation of the coagulation cascade can occur as a result of vascular injury, upon which tissue factor-bearing cells come in contact with blood. [1][2][3] Additionally, in cases of severe sepsis or systemic inflammatory conditions, the coagulation cascade can also become activated through tissue factor-mediated pathways. [4][5][6] While inhibitory pathways and mediators, including fibrinolysis, should normally regulate the overactivation of the coagulation cascade, in conditions like sepsis, inflammation, or trauma, an imbalance of coagulation and fibrinolysis may occur, leading to thrombotic complications.
The activation of the coagulation cascade in vivo culminates in fibrin polymerization, initiated by the serine protease thrombin.
Thrombin activates soluble fibrinogen and promotes its polymerization into an insoluble fibrin mesh. 7 Platelets, also activated by thrombin, incorporate in the forming fibrin mesh creating an immobile clot within the vasculature. Plasmin plays a central role in clot degradation when it becomes activated from fibrin-bound plasminogen. Plasminogen activators, such as tissue-type plasminogen activator (tPA), are released into the bloodstream and initiate this process upon binding to fibrin, activating plasminogen, which is transformed to plasmin and cleaves fibrin. 8 Targeting fibrin for directed thrombolysis therapies represents an optimal strategy because it mimics the body's natural regulation of coagulation.
Dysregulation of clotting or fibrinolysis can result in serious conditions, therefore thrombolytic therapies must be carefully managed. Recombinant tPA is currently approved by the Food and Drug Administration for use in ischemic stroke patients; however, intravenous delivery of tPA can lead to off-target bleeding. Therefore, there has been much interest in targeted delivery through various strategies, presented in several review articles. [9][10][11][12][13] In ischemic stroke patients, where targeted delivery of tPA has recently been investigated, approaches include ultrasound-triggered targeting of thrombolysis with tPA 14 and magnetic targeting of tPA-loaded microrods or nanoparticles. 15,16 Additionally, in events such as myocardial infarction, fibrin-targeted tPA delivery in conjunction with antifibrotic agent drug delivery has been investigated, showing enhanced cardiac function, potentially due to targeted fibrinolysis of fibrin deposition in injured heart tissue leading to a synergistic effect with antifibrotic agents. 17 For venous thrombus therapy, tPA delivery through magnetic targeting and ultrasound stimulation has shown success, 18 and in initial in vitro experimentation and computational simulations, fibrinogen-mimicking nanovesicles carrying tPA show particle binding to activated platelets and effective clot lysis. 19 Targeted delivery of tPA could also be promising for more complex disorders such as disseminated intravascular coagulation (DIC). DIC is a secondary disorder to conditions such as sepsis, trauma, or systemic inflammation and leads to excessive thrombin generation creating microvasculature thrombi and eventual clotting factor consumption and hemorrhage. [20][21][22][23][24][25] In DIC, treatment can be directly opposing; for example, plasma/ platelet transfusions are often administered in cases of bleeding, or anticoagulant/fibrinolytic therapies are often administered in cases of thrombosis complications. Further complicating the management of DIC, systemic administration of therapeutics can have deleterious off-target effects exacerbating the hemostatic imbalance of the disorder. [26][27][28] Therefore, a targeted approach to deliver thrombolytic agents, such as tPA, to aberrant thrombi also represents a promising treatment strategy for DIC.
To that end, the development of targeted fibrinolytic therapies can play a major role in improving therapeutic options for a variety of thrombotic complications. Fibrin-targeting therapies offer an attractive strategy to direct treatment to sites of adverse thrombi in the vasculature. Additionally, delivery of plasminogen activators, such as tPA, may help to regulate the overactivation of the coagulation cascade if administered in cases where bleeding phenotypes are not present. To that end, we developed fibrin-specific nanogels (FSNs), comprised of core-shell microgels synthesized using precipitation polymerization reactions to make cores containing 90% poly(Nisopropylacrylamide) (pNIPAM) and 10% N,N 0 -methylenebis(acrylamide) (BIS), and shells containing 93% pNIPAM, 2% BIS, and 5% acrylic acid (AAc). Prior to conjugation with any fibrin-specific element, these particles are referred to as "core-shell nanogels," and once conjugated to fragment E fibrin antibody, these particles are referred to as FSNs. FSN carriers are then loaded with tPA to create tPA-FSNs.
The overall object in this study is to examine tPA-FSNs action on the dissolution of thrombi in vitro and in vivo and to assess the particles tolerance in vivo. It is hypothesized that tPA-FSNs work by targeting fibrinolytic capabilities directly to fibrin formations that could help in the dissolution of adverse thrombi and correction of consumptive coagulopathy in circumstances such as DIC, and that tPA-FSNs are well tolerated in the body even at higher than therapeutic doses. Therefore, tPA-FSN action was studied in vitro and in vivo to examine fibrin-binding capabilities, degradation with clots in platelet-rich and platelet-poor plasma, the dissolution of thrombi in a DIC rodent model, and coagulation parameters following treatment. Tolerance of tPA-FSNs at higher than therapeutic doses was studied in healthy mice by monitoring weight, food consumption, serum chemistry, and hematology parameters.

| Nanogel characterization
Fabrication of FSNs ( Figure 1a) first was characterized by the successful synthesis of core-shell (CS) nanogels. Following synthesis via precipitation polymerization reaction, core nanogels were found to have a hydrodynamic diameter of 161 ± 26 nm. Following shell addition, CS nanogels measure 238 ± 19 nm in diameter using NanoSight particle tracking analysis (Figure 1d). After conjugation of CS nanogels to fragment E antibody to create FSNs, particle size was similarly characterized, with a hydrodynamic diameter measuring 230 ± 53 nm, and after loading with tPA, measuring 244 ± 67 nm ( Figure S1). A polydispersity index (PDI) of core, CS, FSNs, and tPA-FSNs was determined as follows: 0.026, 0.0064, 0.053, and 0.075, respectively. Representative atomic force microscopy (AFM) images and height traces are shown and demonstrate a similar increase in particle size following shell addition (Figure 1b,c). Average AFM diameter and height measures from previously studied nanogels of the same formulation are 166 ± 16 nm and 17 ± 3 nm, respectively, for core nanogels, and 263 ± 28 nm and 29 ± 4 nm, respectively, for CS nanogels. 29  F I G U R E 1 Overview of tissue-type plasminogen activator-fibrin-specific nanogels (tPA-FSN) particle design and characterization. (a) Coreshell nanogel composition and schematic of tissue-type plasminogen activator (tPA) loading into FSNs. (b) Representative atomic force microscopy (AFM) images of core and core-shell nanogels. (c) Representative height traces of single particles (depicted with a red line through the center) from AFM images of core and core-shell nanogels. (d) Particle size distribution of hydrodynamic diameter measurements from the core and core-shell nanogels utilizing NanoSight particle tracking analysis software. At least 10 8 core and core-shell nanogels were tracked. (e) Fibrin binding assay of FSNs, tPA-FSNs, control sheep immunoglobulin G (CS-IgG) particles, and tPA-CS-IgG particles at 0, 0.001, 0.01, 0.05, 0.1, and 0.2 mg/ml concentrations on fibrin-coated wells or negative control 2% powdered milk in phosphate-buffered saline (MPBS) wells (n = 6-9/ group). Mean ± SD is shown. Data were analyzed via a two-way analysis of variance with a Tukey's post hoc test using a 95% confidence interval. a: p < 0.0001 b: p < 0.001 compared to IgG control particle types increase in fibrin binding potential, up to 0.8 mg/ml, was observed ( Figure S3).

| tPA-FSNs degrade clots ex vivo made from human platelet-poor plasma and platelet-rich plasma
Clot structure of plasma clots shows enhanced fibrin network with Although, tPA-FSN treatment resulted in a lower peak relative intensity of fibrin/fibrinogen at the thrombus site, and a reduced initial slope, compared to control conditions (saline, FSNs, CS-IgG, and CS particles; Figure 3b). Similarly, while platelets accumulate at the site of thrombus for all groups, animals receiving tPA-FSNs reached a lower relative intensity of platelets incorporated compared to controls ( Figure 3c). FSNs and CS particles may enhance platelet recruitment due to their influence on platelet margination; however, future studies are needed to determine this mechanism. These data suggest that following the initiation of thrombus formation, tPA-FSNs can modulate the accumulation of fibrin and platelets, illustrating their potential efficacy in disease states of hypercoagulation.

| tPA-FSNs mitigate thrombi complications in a DIC rodent model at an optimal dose
To assess the ability of tPA-FSNs to mitigate thrombotic complica- (c) Immunohistochemistry (IHC) tissue sections of heart, lung, kidney, and liver from control animals and DIC animals treated with 0, 5, 10, and 20 mg/kg tPA-FSNs (fibrin = red, CD61 = green, DAPI = blue). (d) Quantification of IHC fibrin and platelet (CD61) presentation in tissue sections via ImageJ particle count analysis. For each organ, at least three quantified tissue sections were measured and averaged per animal. The mean of five animals per group ± SD is shown. Data were analyzed via a one-way analysis of variance with a Tukey's post hoc test using a 95% confidence interval. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 ( Figure 4a). While substantial fibrin deposition was also visualized at 5 mg/kg tPA-FSN dose in the DIC animals, at 10 and 20 mg/kg, the appearance of fibrin deposition appears to be reduced in the heart, lung, kidney, and liver. To quantify the decrease in thrombi, immunohistochemistry (IHC) images that were stained for fibrin and platelets (CD61) were imaged and quantified using ImageJ software for fibrin or platelet particle counts (Figure 4b,c). In control animals, minimal amounts of fibrin or platelets (CD61) were observed in the heart, lung, kidney, and liver. In DIC animals receiving 0 mg/kg tPA-FSNs, signifi- (c) Confocal microscopy images were taken, and fiber density quantification was conducted of clots polymerized from isolated platelet-poor plasma from control animals and DIC animals treated with 0, 5, 10, and 20 mg/kg tPA-FSNs. Clots from each animal were polymerized and three images per clot were quantified for fiber density and averaged. n = at least 5 animals/group. Mean ± SD is shown. Data were analyzed via a one-way analysis of variance with a Tukey's post hoc test using a 95% confidence interval. *p < 0.05, **p < 0.01, ****p < 0.0001

| Analysis of toxicity of tPA-FSNs in vivo at supratherapeutic doses
At doses, 5-20 times the optimized therapeutic dose of 10 mg/kg, administration in healthy mice shows no significant changes in animal weight between groups up to 5 days after injection. (Figure 7a). Additionally, no significant differences in food consumption between treatment groups were observed over 5 days (Figure 7b). Hematology parameters, including neutrophils, reticulocytes, white blood cells, red blood cells, monocytes, and platelet count (Table 1), measured at the study endpoint at 5 days postinjection show no significant differences between treatment groups and controls. Serum chemistry parameters, also measured at the study endpoint at 5 days postinjection, show significant decreases in blood urea nitrogen at 10 and 20 mg/kg compared to control animals, which could indicate effects on liver function; however, alanine aminotransferase, alkaline phosphatase, aspartate aminotransferase, total bilirubin, albumin, and total protein show no significant differences in the 10 and 20 mg/kg groups compared to control animals. Additionally, potassium levels at 50 mg/kg are significantly higher than controls, although the levels measured fall within the normal range. 30 The degree of hemolysis caused by nanogel exposure is shown in Figure S2. At the human equivalent dose (HED) of 0.01 mg/ml, and 100Â the HED of 1 mg/ml, CS nanogels, FSNs, and tPA-FSNs are all below 5% hemolysis. Note: After receiving 0, 50, 100, or 200 mg/kg tPA-FSNs, blood samples from each animal were analyzed for serum chemistry and hematology parameters, listed above. n = 4-5 animals/group depending on available sample volume. Mean ± SD is shown. Data were analyzed via a one-way analysis of variance with a Dunnett post hoc test using a 95% confidence interval. Abbreviations: ALT, alanine transaminase; ALP, alkaline phosphatase; AST, aspartate transaminase; BUN, blood urea nitrogen; HCT, hematocrit; HGB, hemoglobin; MCH, mean corpuscular hemoglobin; MCHC, mean corpuscular hemoglobin concentration; MCV, mean corpuscular volume; RBC, red blood cells; tPA-FSN, tissue-type plasminogen activator-fibrin-specific nanogel; WBC, white blood cells. isotype (Thermo Fisher Scientific) was conjugated to CS nanogels as a control for binding studies. Nanogel size was characterized via NanoSight particle tracking analysis (Malvern Panalytical) and AFM (Asylum Research). PDI was calculated from NanoSight particle tracking analysis using the following equation: PDI = (σ/d) 2 , where σ is the SD of particle distribution and d is the average hydrodynamic particle

| DISCUSSION
diameter. Zeta potential was tested for core, CS, and FSN particles.
Conjugation efficiency of the fragment E fibrin antibody to CS nanogels was determined utilizing a known particle/ml concentration (NanoSight) and a CBQCA protein quantitation kit (Thermo Fisher Scientific C6667) to determine the amount of antibody per mg FSN, which taken together give the amount of antibody per FSN particle.
Based on the known amount of antibody and CS particles added to the conjugation reaction, a conjugation efficiency was calculated.
tPA (Sigma Aldrich) was loaded into FSNs using a rehydration "breathing in" technique where lyophilized nanogels were rehydrated in a tPA solution (29 μg/ml) at 20 mg/ml. 33,34 After agitation at 4 C for 24 h to allow for drug entanglement into the polymer network of the nanogels, tPA-FSNs were centrifuged to remove unloaded tPA, washed with ultrapure water, and lyophilized for future use.
To examine fibrin binding capabilities of the particles, fibrin attachment assays were conducted using a plate-based assay with samples of FSNs, tPA-FSNs, CS nanogels conjugated to isotype control antibodies (CS-IgG), and tPA-CS-IgG. Thin fibrin layers were deposited using a previously published method 35   for each organ for each animal. Images were quantified for fibrin and platelet particle counts using ImageJ particle analysis software.
To evaluate coagulation parameters, manual platelet counts were performed using a hemocytometer and PPP was obtained from whole blood via centrifugation at 2000Âg for 3 min, followed by a second centrifugation at 2000Âg for 10 min. Clots formed from obtained PPP for each animal were examined using confocal microscopy as described above. Clots were made with 0.5 U/ml thrombin, 0.05 mg/ml Alexa-Fluor 488 fibrinogen, and 5 mM CaCl 2 . At least, three z-stack images were taken per clot. Binary z-stack projections were quantified for fiber density (black pixels/white pixels).

| Hemolysis assay
Citrated whole blood was obtained from Zen Bio (Durham, NC). Erythrocytes were isolated by centrifuging whole blood at 1500 rpm for 10 min followed by three washes in sterile saline. Isolated red blood cells were diluted 1:1 with saline to obtain 50% hematocrit stock. Concentrations of 0.1 and 1 mg/ml CS, FSN, and tPA loaded FSNs were added to tubes containing 950 μl sterile saline and 50 μl of 50% hematocrit erythrocyte solution. Particle concentrations were chosen from the in vivo human equivalent dose (0.01 mg/mg) and 100Â that dosage (1.0 mg/ml). 42 Sterile saline was used as a 0% hemolytic solution (negative control) and 0.1% Triton X was used as a 100% hemolytic solution (positive control). Solutions were incubated on a shaker at 37 C for 1 h prior to removal of intact red blood cells by centrifuging at 10,000 rpm for 5 min. The supernatant was removed and absorbance was read at 540 nm. Percent hemolysis was calculated as follows: Hemolysis rate (%) = (A t À A nc )/(A pc À A nc ) Â 100, where A t = absorbance of test supernatant, A nc = absorbance of negative control supernatant, and A pc = absorbance of positive control supernatant. 43

| Statistical analysis
Statistical analysis was performed by using GraphPad Prism 8 (GraphPad, San Diego, CA). Fibrin binding, clot degradation assays, biodistribution studies, maximum tolerated dose, animal weight, and food consumption were analyzed via a two-way analysis of variance with a Tukey's post hoc test using a 95% confidence interval.
Dose-response IHC quantification, platelet count, and fiber density were analyzed via a one-way analysis of variance with a Tukey's post hoc test using a 95% confidence interval. Maximum tolerated dose hematology and chemistry parameters were analyzed via a one-way analysis of variance with a Dunnett post hoc test using a 95% confidence interval. Outlier tests were performed on all data sets before statistical analysis. All data are presented as average ± SD.

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