Biomaterial-targeted precision nanoparticle delivery to the injured spinal cord

Abstract

Drug delivery requires precision in timing, location, and dosage to achieve therapeutic benefits. Challenges in addressing all three of these critical criteria result in poor temporal dexterity, widespread accumulation and off-target effects, and high doses with the potential for toxicity. To address these challenges, we have developed the BiomatErial Accumulating Carriers for On-demand Nanotherapy (BEACON) platform that utilizes an implantable biomaterial to serve as a target for systemically delivered nanoparticles (NPs). With the BEACON system, administered NPs are conjugated with a ligand that has high affinity for a receptor in the implanted biomaterial. To test BEACON, an in vivo spinal cord injury (SCI) model was used as it provides an injury model where the three identified criteria can be tested as it is a dynamic and complicated injury model with no currently approved therapies. Through our work, we have demonstrated temporal dexterity in NP administration by injecting 6 days post-SCI, decreased off-target accumulation with a significant drop in liver accumulation, and retention of our NPs in the target biomaterial. The BEACON system can be applied broadly, beyond the nervous system, to improve systemically delivered NP accumulation at an implanted biomaterial target.

Statement of significance

Targeted drug delivery approaches have the potential to improve therapeutic regimens for patients on a case-by-case basis. Improved localization of a therapeutic to site of interest can result in increased efficacy and limit the need for repeat dosing. Unfortunately, targeted strategies can fall short when receptors on cells or tissues are too widespread or change over the course of disease or injury progression. The BEACON system developed herein eliminates the need to target a cell or tissue receptor by targeting an implantable biomaterial with location-controllable accumulation and sustained presentation over time. The targeting paradigm presented by BEACON is widely applicable throughout tissue engineering and regenerative medicine without the need to retool for each new application.

1. Introduction

Controlled drug release from implantable biomaterials has been a viable method to sustain local therapeutic delivery for a number of tissue engineering and regenerative medicine applications [1-3]. Biomaterials alone provide a platform for regeneration as they can recapitulate the native extracellular matrix, providing a microenvironment more hospitable for tissue regeneration [4-6]. Active local microenvironment modulation with pro-regenerative therapeutic delivery from a biomaterial can further promote tissue regrowth and lead to better recovery [7]. Local delivery strategies are advantageous as they ensure the loaded therapeutic is released directly at the site of interest with minimal consideration required for therapeutic trafficking, active pharmaceutical ingredient (API) stability, and off-target accumulation and toxicity. Biomaterials can be tuned to control loaded drug release rates to sustain required therapeutic levels, effectively creating a reservoir for drug release [8,9].

The advantages of local drug release do not come without their pitfalls. For every drug loaded into a biomaterial, optimization for release rate needs to be tuned to maintain therapeutic levels, and this is compounded when multiple drugs are loaded at once, requiring re-optimization for every drug and drug combination [8]. Release rates can be further tuned with combination therapies where a therapeutic can be loaded into a drug carrier, which is subsequently added to an implantable biomaterial rather than releasing drug directly [10-12] or in response to external stimuli like light, ultrasound, and magnetic/electric fields [13]. Depletion of the loaded drug is inevitable with limited flexibility for replenishing the reservoir or changing therapeutic release timelines after implantation. Reloading the reservoir with more drug or a new drug could require a second surgical procedure to remove the implanted biomaterial and replace with a new one, leading to further tissue damage. Injury pathophysiology and subsequent cellular response timelines can vary widely on a patient-to-patient basis [14], necessitating alterable drug accumulation and release profiles.

Conversely, systemic drug delivery provides significant temporal control over drug administration without the need to directly access the tissue providing freedom to change therapeutic regimen based on recovery with minimal invasiveness. API trafficking to the tissue of interest must be considered when designing a drug delivery system as in vivo conditions can compromise its stability and limit efficacy before having its intended therapeutic effect. Drugs can be loaded into carriers like polymeric nanoparticles (NPs), lipid NPs, liposomes, micelles, and dendrimers, all of which have been used in preclinical studies [15], to improve drug efficacy. Additionally, drug carriers can offer a sustained release profile [1] and a tunable circulation time [16], each contributing to improved therapeutic benefits. Drugs and drug carriers are, however, subject to clearance mechanisms like uptake by cells of the reticuloendothelial system and filtering via the liver and kidneys that can result in poor accumulation at the therapeutic target [17-19]. Administering elevated doses can ensure sufficient API reaches the desired target, but high doses can result in complications in off-target tissues where the API could potentially reach toxic levels, causing more damage and even death [20,21].

Targeting mechanisms in drug delivery have the capability to increase accumulation at tissue targets throughout the body. The enhanced permeability and retention (EPR) effect is particularly useful when delivering therapeutics for cancer treatments as growing tumors exhibit increased permeability of delivered drugs or drug carriers into the tissue and poor lymphatic drainage thus retaining drugs with minimized clearance [22,23]. EPR effect-based delivery has exhibited limited clinical results and can result in little of the administered dose reaching the solid tumor [18,24,25]. Active targeting drug carriers can be conjugated with targeting ligands that bind to receptors on cells or tissues of interest to improve delivery in cancer [26-28], vascular injury [29,30], and neurodegeneration [31,32] applications [33]. An increase in accumulation limits the need to deliver high dose therapeutics thus minimizing potential side effects, however, ligand-receptor pairs can be susceptible to tissue remodeling, changes in cell infiltration, and cell surface receptor saturation [34].

Targeting systems have been developed to anchor to the extracellular matrix to capture administered prodrug [35,36] or to target an implant as a refillable depot [37,38] limiting reliance on naturally occurring ligand-receptor pairs. In this work, we have developed the BiomatErial Accumulating Carriers for On-demand Nanotherapy (BEACON) system that utilizes NP-mediated targeting to an implantable, pro-regenerative biomaterial scaffold, serving as a platform technology that can be applied broadly without necessitating re-optimization for different target tissues. We utilized an acute lateral hemisection spinal cord injury (SCI) to serve as an ideal application as individual patient recovery varies on a case-by-case basis, thus necessitating the therapeutic dexterity the BEACON system offers [14,39]. Our developed BEACON platform engineers our previously used poly(ethylene glycol) (PEG) tubes [40-42] to serve as an implantable biomaterial that acts as a pro-regenerative scaffold and as a binding target for injected poly(lactic-co-glycolic acid) (PLGA)-PEG NPs. The PEG tubes have demonstrated good bio-compatibility, no exacerbation of immune response, and increased axon elongation with improved functional recovery, as shown in previous work [40-42], while the PLGA-PEG NPs were chosen for their in vivo safety [43,44] and prior demonstrated success in nervous system applications [31,32]. Moreover, PLGA-PEG NPs have progressed to Phase II clinical trials for cancer therapeutics (NCT02283320), demonstrating their biocompatibility and safety in humans. BEACON uses implantable PEG tubes modified to contain biotin as a receptor for streptavidin-functionalized targeting ligand NPs (SNPs), thus taking advantage of the biotin/streptavidin complex [45]. This work represents a significant step forward towards developing a comprehensive biomaterial-based platform for NP delivery. We demonstrate here a biomaterial-based approach to developing BEACON, a system to improve targeted NP accumulation and retention at an implanted biomaterial that will be utilized in future work to deliver therapeutic-loaded NPs.

2. Methods

2.1. Biotin-PEG implantable target tube fabrication

Hydrogel implants were fabricated as previously described with slight modifications to create our target tubes [42]. Briefly, 20% w/v 8-arm PEG-maleimide (PEG-MAL, 20 kDa; JenKem, Plano, TX) was mixed at a 9:1 (or 3:1) ratio with biotin-PEG-MAL (7.5 kDa; JenKem, Plano, TX) and cross-linked with 5 mM slow-degrading, plasmin-sensitive YKND cross-linking peptide (Ac-GCYKNDGCYKNDCG; Genscript, Piscataway, NJ) [46] to form microspheres through a water/oil emulsion. The biotin-PEG-YKND solution was homogenized in silicone oil (Fisher, Hampton, NH) with 2% v/v Tween-20 (Sigma, St. Louis, MO) at a speed for 4000 rpm for 1 min. Microspheres were rinsed by centrifugation three times. Irgacure 2959 photoinitiator dissolved in N-vinylpyrrolidinone (660 mg mL⁻¹ ; Sigma) was added to the microspheres at a final concentration of 1% w/v. The resulting microspheres were then packed into polydimethylsiloxane (PDMS, Dow Corning, Midland, MI) molds to generate biotin-PEG tubes (approximate OD, 700 μm; ID, 250 μm) and exposed to ultraviolet light for 3 min to initiate free radical polymerization (Fig. 1A). Tube dimensions were evaluated by averaging 10 evenly-spaced measurements from 5 tube samples at 10x magnification using an EVOS XL Core Microscope (Life Technologies, Carlsbad, CA, USA). The outer diameter of the tube was chosen so that 4 tubes would fit aligned into the lesion space created during in vivo studies, described later. The inner diameter was kept consistent with previous bridging strategies that have demonstrated axonal elongation in mouse models [42,47-50]. Tubes were rinsed three times, visually checked to ensure a hollow lumen, cut to size, dehydrated, and stored at −80 °C until use. Non-targeted, PEG only implants were fabricated in the same way only using 20% w/v 8-arm PEG-MAL.

Fig. 1.

Biotin-PEG tube fabrication. (A) Schematic depicting the fabrication of implantable biotin-PEG tubes via a microsphere intermediate that are packed into a mold to make the final tube structure. (B) Sample tube imaged longitudinally via light microscopy. SEM images show (C) tube lumen and (D) macroporosity of tube body. Fabricated gels had no significant differences in (E) microsphere diameter or (F) storage modulus between PEG and biotin-PEG. (G) Biotin was detectable in the biotin-PEG tubes over 12 weeks at 37 °C (red dashed line = expected concentration). Data are presented as mean ± SEM, n = 3 per condition, ***p <0.001, 200 μm (B) and 50 μm (C,D) scale bars.

2.2. Biotin-PEG implantable target tube characterization

Fabricated tubes were imaged using a JEOL JSM 6010PLUS/LA Analytical Scanning Electron Microscope (SEM; JEOL Ltd., Tokyo, Japan). Acquisition conditions included a 10 mm working distance, a spot size of 50, and voltage of 5 kV. Hydrogel microsphere diameter was determined via Mie scattering using a Microtrac S3500 (Microtrac, Montgomeryville, PA). Size distribution measurements were performed in triplicate and analyzed via Microtrac Flex (Microtrac) software. Biotin-PEG (9:1 and 3:1 ratios) and PEG only hydrogel discs were fabricated (D = 8 mm, H = 0.5 mm) and used for small angle oscillatory shear rheological characterization, performed using a Discovery HR-2 Rheometer (TA Instruments, New Castle, DE). Each disk was loaded onto an 8 mm flat plate with a gap height set to 0.5 mm and frequency and strain set at 1 Hz and 1%, respectively, chosen after performing frequency and strain sweeps. To determine storage (G’) and loss (G”) moduli, hydrogel discs were subjected to a time sweep with the described parameters for 5 min. All experiments were carried out at 37 °C to simulate body temperature. Biotin concentration of fabricated gels was determined using the Pierce Fluorescence Biotin Quantitation Kit (Thermo Fisher Scientific, Waltham, MA). Fabricated gels were degraded in trypsin (Thermo Fisher Scientific) diluted 10x from 0.25% stock with shaking at 37 °C overnight. After full sample degradation, biotin was quantified following the manufacturer’s instructions using a DTX 880 Multimode Detector fluorescent plate reader (Beckman Coulter, Brea, CA). Biotin stability was assessed by incubating samples at 37 °C for 1, 2, 4, 8, 10, and 12 weeks and similarly measuring biotin concentration.

2.3. Targeting SNP fabrication

Blank NPs (BNPs) were fabricated as previously described using nanoprecipitation [26,27] with additional conjugation steps to make targeting SNPs (Fig. 2A) and non-targeting control NPs (CNPs). PLGA-PEG-COOH (10 kDa, 2 kDa; Nanosoft Polymers, Winston-Salem, NC) was suspended at 5 mg mL⁻¹ in N,N-Dimethylformamide (DMF; Sigma) and added dropwise to 10 mL distilled water with constant stirring at 900 rpm and allowed to spin for 2 h at room temperature. The NPs were then washed three times and collected using Amicon Ultra-15 Centrifugal Filter Units (100 kDa cutoff; Sigma). After collection, NPs were suspended at 5 mg mL⁻¹ in distilled water and incubated with 200 mM EDC (Thermo Fisher Scientific) and 50 mM Sulfo-NHS (Thermo Fisher Scientific) for 15 min at room temperature with gentle shaking. Fluorescent streptavidin (AF488, −555, or −680 Streptavidin; Thermo Fisher Scientific) at 2% loading with respect to polymer mass, as previously reported [27], was subsequently reacted overnight with gentle shaking at room temperature with the activated NPs to form targeting SNPs. After overnight reaction, NPs were washed three times and collected using Amicon filtration, resuspended at 10 mg mL⁻¹ , and stored at 4 °C. Non-targeting CNPs were similarly fabricated, however, an aminated fluorophore (AlphaFluor 488-amine or AlphaFluor 680-amine; AAT Bioquest, Sunnyvale, CA) was conjugated to the activated NPs to allow for subsequent detection in place of the fluorescent streptavidin. After reacting overnight, non-targeted CNPs were collected, suspended, and stored the same as SNPs.

Fig. 2.

PLGA-PEG-Streptavidin NP (SNP) fabrication. (A) Schematic depicting the nanoprecipitation technique used for SNP fabrication. (B) Sample targeting SNPs imaged via TEM. (C) DLS data shows a smaller Z_avg diameter and higher zeta potential for blank unconjugated NPs (BNPs) compared to AF680 non-targeting control NPs (CNPs) and AF680 targeting SNPs. (D) Targeting SNPs are stable when stored at 4 °C but increase in Z_avg diameter and zeta potential at 37 °C. (E) NP parameter table depicting all NP condition parameters acquired. Data are presented as mean ± SEM, n = 3–6 per condition, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001, 20 nm scale bar (B).

2.4. Targeting SNP characterization

All fabricated NP conditions were analyzed via dynamic light scattering (DLS) using a Zetasizer Nano ZS (Malvern Panalytical, Westborough, MA). After formation, NPs were diluted in distilled water to 0.1 mg mL⁻¹ and the Z_average diameter, polydispersity index (PDI), and zeta potential were obtained from three independent measurements. NP morphology was analyzed via transmission electron microscopy (TEM) imaging by diluting to 0.05 mg mL⁻¹ and mixing with a 4% solution of uranyl acetate (Sigma). NP solutions were filtered with a 0.45 μm filter and ~20 μL were dropped on a copper grid and dried overnight at room temperature. TEM images were acquired using a JEOL JEM-1400 TEM (JEOL Ltd.). NP stability analyzed via DLS was tracked when stored at both 4 °C and 37 °C. Z_average diameter, PDI, and zeta potential measurements were taken daily for 1 week post-fabrication and continued twice weekly thereafter until 3 weeks. Similarly, SNP stability in serum was studied by suspending SNPs in 10% serum [51] and measuring Z_average diameter, PDI, and zeta potential at 0, 0.5, 1, 2, 4, 6, and 8 h and subsequently once daily until 5 days of incubation. Fluorescent streptavidin conjugation onto SNPs was assessed via standard curve comparing fabricated SNPs to known fluorescent streptavidin standards via fluorescent plate reader.

2.5. In vitro beacon characterization

SNP binding to tubes was visualized by incubating biotin-PEG (all remaining experiments are described using the 9:1 ratio) tubes with 1 mg mL⁻¹ AF555-SNPs for 1 hour at room temperature. Tubes were washed with phosphate buffered saline (PBS) and allowed to dry before imaging on a Nikon Eclipse Ti inverted microscope (Nikon, Tokyo, Japan) using a 10x dry objective. SNP binding to biotin-PEG hydrogels was assessed in vitro by incubating biotin-PEG or PEG only gels with 5 mg mL⁻¹ AF488-SNPs or AF488-CNPs. Gels with NPs were incubated at 37 °C with gentle shaking for 1 hour at which point gels were centrifuged and supernatant containing free NPs was collected and analyzed via fluorescent plate reader with fluorescent NP standard solutions to quantify NP loss relative to the initial amount added. For retention studies, aspirated supernatant was replaced with fresh PBS, and this process was repeated daily for 5 days. An additional experiment to model NP fouling was performed using mouse serum collected via terminal cardiac puncture. NPs were suspended at 5 mg mL⁻¹ in serum and allowed to shake for 2 h at 37 °C. After 2 h, the NP/serum mix was added to PEG and biotin-PEG gels and incubated for an additional hour at 37 °C and NP binding was assessed as previously described. Similarly, fouling on the tubes themselves was simulated using fetal bovine serum (FBS) (Thermo Fisher Scientific). In this experiment, tubes were incubated with FBS for 2 h, washed 3 times, and subsequently incubated with NPs. Samples were collected 1 hour after NP incubation and binding was assessed as previously described. SNP saturation of the biotin-PEG tubes was determined with elevating concentrations of AF488-SNPs up to 30 mg mL⁻¹ , 3 times the concentration administered in subsequent in vivo experiments. Biotin-PEG tubes were incubated with 0, 1, 2.5, 5, 10, 20, and 30 mg mL⁻¹ SNPs for 1 hour at 37 °C. After incubation, tubes were washed with PBS 3 times to remove unbound SNPs. Bound SNPs were freed from the tubes by degrading the tubes overnight in trypsin (Thermo Fisher Scientific) diluted 10x from stock. Fluorescent intensity of the degraded material with free NPs was quantified via fluorescent plate reader. Sequential binding was also probed by first incubating a dose of AF488-SNPs with tubes and rinsing with PBS 3 times and subsequently incubating with a dose of AF555-SNPs. SNP binding of each wavelength was assessed via fluorescence plate reader as described.

2.6. In vivo SNP pharmacokinetics

All animal work was performed with prior approval and in accordance with the Institutional Animal Care and Use Committee (IACUC) guidelines at the University of Miami. Cadmium Qdot^™ 705 ITK^™ Amino (PEG) Quantum Dots were purchased from Thermo Fisher Scientific. Quantum dots were subsequently coated with PLGA to improve hydrophobicity. To assess administered SNP pharmacokinetics, SNPs were loaded with PLGA-coated quantum dots. Quantum Dot SNPs (QD-SNPs) were injected intravenously via lateral tail vein in C57BL/6 mice at a dose of 5 mg/kg with respect to the QDs [31]. QD dose was determined using inductively coupled plasma mass spectrometry (ICP-MS) with an Agilent 7900 ICP-MS instrument (Agilent Technologies, Santa Clara, CA). A pre-injection blood sample was taken with subsequent blood sampling at 1, 3, 6, 12, 24, and 48 h post-injection. Cadmium levels in the blood plasma were determined using ICP-MS.

2.7. Spinal cord injury surgeries

A C5 lateral (left) hemisection SCI was created in adult Balb/c female mice aged 6–8 weeks. Mice were anesthetized with 2% isoflurane and provided preemptive local pain management (1 mg kg⁻¹ sustained release bupivacaine). Anesthesia was confirmed via toe pinch at which point a 2 cm incision was made to the skin to facilitate a C5 laminectomy. A 1.15 mm lateral hemisection was excised from the left side of the spinal cord to provide a discrete injury region. Four dried biotin-PEG or PEG only implants were implanted into the injury site individually, and once in the lesion they rehydrated into their full shape. Additional controls included an “SCI” only condition where an injury was performed with no subsequent biomaterial implantation and a “Sham” condition where no injury was performed. After injury and implantation, all muscles were sutured and skin was stapled. Mice were immediately provided post-operative antibiotics (2.5 mg kg⁻¹ enrofloxacin; once daily) and supportive hydration (0.05 mL g ^{− 1} ; once daily). Bladders were expressed twice daily, and staples were removed after 6 days. All mice were allowed free access to water and an alfalfa-free chow diet, chosen to minimize background signal for in vivo imaging. Surgical controls were put in place to limit lesion size variance, including the order of incisions made to limit the effects of swelling to cut lines, measuring the distance between rostral and caudal cuts, and verifying the absence of bruising to the contralateral tissue. Exclusion criteria include any deviations to the surgical controls, as well as any variance to the recovery timeline, including an inability to ambulate by post-operative day 3. No mice met these exclusion criteria for this study.

2.8. In vivo imaging of nanoparticle biodistribution

On day 6 post-injury, all mice were injected with targeting SNPs, non-targeting CNPs, or saline as a no-injection control. Mice were lightly anesthetized with 2% isoflurane and restrained following IACUC guidelines to minimize distress. AF680-SNPs or AF680-CNPs were injected at 10 mg mL⁻¹ in 200 μL of sterile saline via lateral tail vein. On day 7 post-injury, mice were imaged using the In Vivo Imaging System (IVIS; PerkinElmer, Waltham, MA) to assess SNP and CNP biodistribution. To acquire in vivo images, mice were anesthetized with 2% isoflurane and transferred to the IVIS chamber under continuous isoflurane. In vivo images were captured with each acquisition containing a mouse that was not injected with fluorescent NPs to serve as a background correction. After in vivo imaging, mice were euthanized and spinal cords, livers, spleens, kidneys, lungs, hearts, and brains were harvested and imaged ex vivo. Isolated spinal cords were kept in their spinal columns for ex vivo imaging to mitigate the risk of the implanted biomaterial tubes freeing from the tissue. A subset of mice were not euthanized after in vivo imaging and were subsequently imaged via IVIS daily for the ensuing 5 days to assess the SNP and CNP retention at the implanted tubes. Similarly, each acquisition contained a mouse with no fluorescent SNPs/CNPs to allow for background correction. All images acquired via IVIS were analyzed and quantified using Living Image (PerkinElmer) software to evaluate total flux based on fluorescent signal.

An additional cohort of mice was used for a chronic time point study. All surgical procedures were followed the same, and post-operatively antibiotics and supportive hydration were provided for 2 weeks and 5 days post-injury, respectively. Imaging occurred 6 months post-injury, and mice were similarly injected with SNPs or saline one day prior to imaging. In vivo images were acquired as previously described.

2.9. Immunohistochemistry

Following ex vivo imaging, spinal cords were isolated from their spinal columns. All isolated organs were flash frozen and cryosectioned longitudinally in 18 μm sections. Samples were fixed and incubated with streptavidin to detect unbound biotin binding sites (1:1000, Thermo Fisher Scientific). Additional tissue was fixed, permeabilized, and incubated overnight with rat anti-F4/80 (1:200, Abcam, Cambridge, United Kingdom) with subsequent species-specific fluorescent antibodies at 1:100 0. Hoechst 33342 (Life Technologies) was used as a counterstain in tissue sections. Immunostained tissue sections were imaged using a Nikon Eclipse Ti inverted microscope (Nikon) using a 10x dry objective and a Zeiss LSM 880 (Zeiss, Oberkochen, Germany) using an 63x oil objective.

2.10. Statistics

Data normality was assessed using a Shapiro-Wilk normality test with an α value of 0.05, which determined parametric statistical tests were appropriate for our analyses. Single and multiple comparison pairs were analyzed using a one-way or two-way ANOVA with Tukey post-hoc test. All statistics test significance using an α value of 0.05. For all graphs, * denotes p < 0.05, ** denotes p < 0.01, *** denotes p < 0.001, and **** denotes p < 0.0001, unless otherwise specified in the figure caption. ROUT tests were performed with a false discovery rate of 1% to identify any outliers in data. Linear regressions are calculated by minimizing the sum of least squares. All values are reported as mean ± standard error of the mean (SEM), unless otherwise specified. Prism 7 (GraphPad Software, La Jolla, CA) software was used for all data analysis.

3. Results

3.1. Biotin-PEG tube formation is controlled through a multi-stage annealing process with sustained biotin availability

For the BEACON platform, the previously developed PEG tubes [42] were modified to include biotin-PEG-MAL, with biotin serving as the target receptor on our biomaterial (Fig. 1A). Tube dimensions were measured to have an outer diameter of 732 ± 4 μm and an inner diameter of 258 ± 7 μm. A clear lumen, as indicated by the dark middle (Fig. 1B) and opening on the end (Fig. 1C), was observed in the fabricated tubes, and the tube bodies were macroporous along their length (Fig. 1D). The microsphere intermediate was formed via water/oil emulsion, and 9:1 biotin-PEG microspheres (50 ± 4 μm) did not exhibit a significant difference in diameter compared to PEG only microspheres (45 ± 1 μm) (Figs. 1E, S1A-D). Sample microspheres are shown for morphological characterization in Fig. S1E,F. For mechanical properties, 3:1 and 9:1 biotin-PEG hydrogels were compared to PEG gels. The 3:1 gels had a significantly lower storage modulus (193 ± 8 Pa) compared to both the 9:1 (399 ± 28 Pa) and PEG only (406 ± 57 Pa) hydrogels. No significant differences were observed between the 9:1 and PEG hydrogels (Figs. 1F, S2A-D). For the remainder of the manuscript, biotin-PEG will be used to refer to the 9:1 ratio. Biotin availability over 12 weeks was probed in vitro, and no significant changes in biotin concentration were detected in the biotin-PEG tube conditions (Fig. 1G).

3.2. Targeting SNPs are fabricated via nanoprecipitation with subsequent streptavidin conjugation

PLGA-PEG NPs are utilized in the BEACON system to form targeting SNPs with a nanoprecipitation technique (Fig. 2A). Following streptavidin conjugation, SNPs were characterized via TEM (Fig. 2B) and DLS (Figs. 2C, S3) with a Z_average diameter of 71 ± 2 nm and zeta potential of −28 ± 2 mV. Control conditions included non-targeting, fluorophore only CNPs and blank BNPs with Z_average diameters of 68 ± 5 nm and 51 ± 4 nm and zeta potentials of −27 ± 6 mV and −12 ± 7 mV, respectively. No significant differences were observed between the SNPs and CNPs for either Z_average diameter or zeta potential, and both were significantly greater in Z_average diameter and significantly lower in zeta potential compared to the BNPs. All assessed NP parameters can be found in Fig. 2E. SNP (Fig. 2D) and CNP (Fig. S4) stability was evaluated at 4 and 37 °C over the course of 21 days. SNPs and CNPs incubated at 37 °C more quickly increased in Z_average diameter and zeta potential compared to the same condition at 4 °C. SNPs incubated in 10% serum at 37 °C demonstrated stability over the first 24 h after which Z_average diameter, PDI, and zeta potential increased significantly each day (Fig. S5).

3.3. SNPs bind and are retained in biotin-PEG tubes in vitro

SNP binding to biotin-PEG tubes (Fig. 3A) was assessed in vitro. Binding was observed qualitatively and imaged (Fig. 3B), and SNP concentrations up to 30 mg mL⁻¹ (3 times the concentration delivered for in vivo studies) were incubated with tubes, and a linear increase with increasing SNP concentrations was observed (r² = 0.9690) indicating no saturation (Fig. 3C). Additionally, biotin-PEG tubes incubated with sequential doses of AF488-SNPs and AF555-SNPs demonstrated binding of both doses whereas control conditions either showed no binding or only signal of the administered wavelength (Fig. S6A). Both wavelengths were also observed qualitatively when bound to biotin-PEG tubes (Fig. S6B). NP binding and retention to biotin-PEG tubes was evaluated in vitro over the course of 5 days as a percent of the initial dose administered. After incubating for 1 hour, 77 ± 0.4% of the administered NPs were retained in the biotin-PEG + SNPs samples compared to 15 ± 4% (PEG + SNPs), 18 ± 3% (biotin-PEG + CNPs), and 25 ± 2% (PEG + CNPs) for all other conditions (Fig. 3D). On the day 5 time point, 60 ± 0.3% of the administered SNPs remained in the biotin-PEG tubes while all other conditions had negligible NP concentrations remaining. To further characterize retention, the NP release rate from the amount initially bound to the tubes was quantified, and by day 5, only 23 ± 0.3% of the SNPs bound to biotin-PEG tubes was released while control conditions released at significantly quicker rates (Fig. 3E). Additional experiments were performed to better mimic in vivo conditions by first incubating the SNPs in serum to simulate NP fouling. No significant difference was observed between SNPs incubated with and without serum, and both remained significantly greater than control conditions (Figs. 3F, S7). Tube fouling was also simulated by incubating the tubes in FBS prior to adding NP conditions. Tubes fouled with FBS still experienced a significant increase in SNP binding compared to control conditions (Fig. S8).

Fig. 3.

SNPs bind to biotin-PEG tubes in vitro. (A) Schematic depicting SNPs bind to fabricated biotin-PEG beacons. (B) Biotin-PEG tube incubated with SNPs. (C) Biotin-PEG sites are not saturated with increasing NP concentrations (D) SNPs bind to biotin-PEG beacons and are retained at significantly higher levels with a (E) significantly slower release rate. (F) Simulated NP fouling did not significantly change SNP binding to biotin-PEG beacons. Data are presented as mean ± SEM, n = 3 per condition, **p<0.01, ***p<0.0 01, ****p<0.0001 200 μm scale bar (B).

3.4. Injected SNPs circulate in blood in vivo

QD-SNPs were injected intravenously, and blood samples were subsequently collected at 1, 3, 6, 12, 24, and 48 h after injection (Fig. S9A). DLS measurements for QD-SNPs indicate a Z_average diameter of 74 ± 1 nm, PDI of 0.22, and zeta potential of −25 ± 1 mV (Fig. S9B). Collected blood samples were analyzed via ICP-MS to detect cadmium levels in the blood corresponding to QD-SNP concentration (Fig. S9C). Parameters assessed from the pharmacokinetic studies can be found in the table in Fig. S9D where a t_1/2 7.4 h was observed with a peak at 1 hour post-injection, a C_max of 281 ± 92 ng/mL, and AUC_max of 1920 ± 450 ng*hr/mL.

3.5. SNPs bind to implantable tubes in vivo with BEACON

Four biotin-PEG tubes were implanted into a C5 lateral left hemisection injury on day 0, and SNPs were injected intravenously on day 6 with subsequent analysis via IVIS and immunohistochemistry (IHC) on day 7 (Fig. 4A). For IVIS analysis, a no injection control mouse was used to eliminate background signal. Accumulation was observed in all injury/implant conditions receiving SNPs (Fig. 4B) and CNPs (Fig. S10A). Mice receiving biotin-PEG tubes with an SNP injection exhibited a significant increase in SNP accumulation compared to SCI (injury but no biomaterial implant) and Sham (no injury and no biomaterial implant) control conditions.

Fig. 4.

NP accumulation at the targeting depot. (A) Intervention timeline following C5 hemisection. (B) IVIS imaging demonstrates accumulation of SNPs in all conditions receiving injections. (C) SNP accumulation quantification demonstrates a significant increase in biotin-PEG conditions. (D) Targeted SNP accumulation is increased compared to non-targeted CNP accumulation. Data are presented as mean ± SEM, n = 5–11 per condition, *p<0.05, ***p<0.001.

This was also observed qualitatively with IHC where SNPs (red) were not seen in the SCI condition (Fig. 5A-D) but were visibly bound to biotin-PEG tubes (green) implanted in the injury site (Fig. 5E-H). Infiltrating macrophages (F4/80, green) also demonstrated uptake of SNPs (red) accumulated in the biotin-PEG tubes in the injury (Fig. S11). The average signal observed in mice with biotin-PEG tubes was greater than those with PEG only tubes, but no significant difference was observed (p = 0.18) (Fig. 4C). No significant differences were observed across all implant/injury conditions in mice that received CNPs (Fig. S10B). A significant increase in biotin-PEG + SNPs compared to biotin-PEG + CNPs was observed (Fig. 4D). No other significant differences were observed when comparing other SNP to CNP groups for each implant/injury condition. An additional small cohort of mice were used in a chronic study where SNPs were injected 6 months post-injury and implantation to probe whether or not SNPs could accumulate in biotin-PEG tubes. A non-significant increase (p = 0.25) of SNP accumulation was observed in the biotin-PEG group compared to the sham surgery controls (Fig. S12A-B).

Fig. 5.

SNPs (red) accumulate at the biotin-PEG tubes (green) in vivo. Mice receiving an SCI with no implant (A-D) had minimal SNP accumulation while mice receiving biotin-PEG (E-H) tubes demonstrated SNP accumulation in the implanted tubes. Nuclei (DAPI, blue), injury region denoted by white dashed line, 200 and 50 (inset) μm scale bars.

3.6. The BEACON system decreases SNP liver accumulation

Whole animal biodistribution was evaluated on day 7 with ex vivo IVIS analysis of spinal cords, livers, kidneys, spleens, brains, lungs, and hearts. Similar to the in vivo analysis, a significant increase in SNP accumulation was observed for mice receiving biotin-PEG tubes compared to SCI and Sham controls, and an increase that was not significant was observed in biotin-PEG mice compared to PEG mice (p = 0.08) (Fig. 6A-B). Qualitatively, accumulation was observed in the liver at high levels with accumulation to a lower extent observed in the kidneys, spleen, and lungs. Minimal accumulation was observed in the brain and heart (Figs. 6C-D, S13A-B). A significant decrease in liver accumulation was observed in PEG and biotin-PEG conditions compared to the sham control (Fig. 6D).

Fig. 6.

Targeting SNP biodistribution. (A) Isolated cords are left in spinal columns and imaged ex vivo using IVIS. (B) Quantification of accumulated SNPs in spinal cords showed an increase in Biotin-PEG conditions. (C) Sample organs isolated from mouse receiving biotin-PEG tubes and SNPs. (D) Quantification of accumulated SNPs in isolated organs of interest. Data are presented as mean ± SEM, n = 4–13 per condition, *p<0.05. Liv, liver; Kid, kidney; Spl, spleen; Br, brain; Lu, lungs; He, heart.

3.7. SNPs are retained longer in biotin-PEG tubes in vivo with BEACON

As previously described, injuries and implants were performed on day 0 and SNPs were administered on day 6. For retention studies, IVIS imaging was repeated for 5 days following SNP injection (Fig. 7A). Accumulation was observed on day 1 of IVIS imaging in both biotin-PEG and PEG conditions (Fig. 7B), and an increased in SNP signal was observed on day 2 for the biotin-PEG mice (Fig. 7C-D). By day 3, an observable drop in SNP signal was observed in PEG mice, and on day 4 a significant decrease was observed in total signal (Fig. 7C) and percent of initial bound dose remaining (Fig. 7D). For the biotin-PEG mice, only a significant drop was observed on day 5 in total signal, but no significant differences were observed comparing the initial time point to the final time point for bound dose remaining.

Fig. 7.

Targeting SNPs are retained in the biotin-PEG depot. (A) Intervention and imaging timeline following C5 hemisection. (B) IVIS imaging daily for 5 days following SNP injection. (C) IVIS signal quantified for 5 days following SNP injection. (D) Bound dose percent remaining for 5 days following SNP injection. Data are presented as mean ± SEM, n = 6 per condition, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

4. Discussion

Targeted drug delivery has great potential for improving therapeutic approaches to a number of injuries and diseases across tissue engineering and regenerative medicine [33]. With that in mind, targeting in central nervous system (CNS) injuries remains difficult as most targeting strategies rely on conjugating ligands to nanocarriers that have high specificity for receptors on cells or tissues. In CNS injuries, like SCI, common targets are either ubiquitously expressed throughout the entire nervous system or are highly susceptible to remodeling as the injury progresses. To ameliorate this design challenge, we have developed the BEACON targeting system that reframes how we approach targeted drug delivery. Rather than targeting a cell or tissue, we modified our implantable PEG tubes to serve as our target for NP accumulation. In this fashion, we are able to increase our targeting longevity and specificity by tuning our biomaterial to be present throughout injury progression and having a targeting receptor that is only present on our biomaterial.

The unmodified PEG tubes have previously demonstrated neuroregenerative and immune benefits following SCI where they have limited secondary injury progression and successfully modulated the immune response [40-42], but additional biochemical cues will further improve tube-mediated regeneration. As a vehicle for stem cell and gene therapies, regeneration was improved [40,41], but local release from the tubes lacked the temporal dexterity afforded by BEACON. For BEACON, we engineered the tubes to contain biotin, serving as a receptor moiety for injected SNPs to target. The SNPs could also be preloaded onto the tubes prior to implantation to serve as local release platform, and this has been done by other groups [52]. When preloading, however, that would mean the SNPs and the tubes would be delivered at the same time, losing temporal dexterity similar to direct therapeutic release from the tubes. Further, once implanted, a therapeutic regimen cannot easily be altered, thereby limiting the adaptability. By injecting the SNPs systemically to target the tubes, more control is afforded over therapeutic dosing.

With BEACON, it was important to maintain the physical properties of the blank PEG tubes. Using a microsphere intermediate when fabricating the tubes creates a macroporous network with cell-sized pores allowing for crucial cell infiltration that aids in tissue repair in vivo [50,53-57]. A tube without the pores or pores on a smaller scale would lose this interconnectivity, severely limiting the potential for cell infiltration and repair. The formed pores are largely attributable to the size of the microsphere intermediates formed during scaffold fabrication. Additionally, the plasmin-sensitive crosslinks have been shown to degrade slowly over time [46], and this is advantageous in the injury to allow for further tissue infiltration through the biomaterial as repair occurs. It is expected that cell infiltration will not be altered by the presence of biotin as average microsphere diameters did not change between blank PEG and biotin-PEG conditions. Scaffold stiffness is a crucial design consideration of implants for SCI therapies as the mechanical properties of the substrate can significantly alter neural repair. Softer biomaterial scaffolds that better mimic the native extracellular matrix (ECM) by matching CNS mechanical properties (<1000 Pa) have shown improved neural elongation [58-60], while stiffer gels have increased reactive glial cell activation [59]. The PEG tubes alone have a G’ of 406 ± 57 Pa, well within the expected range of CNS tissue, and have previously demonstrated robust axonal elongation [40-42]. The 9:1 biotin-PEG tube G’ was not significantly different, indicating the inclusion of biotin at this concentration did not alter mechanical properties, whereas the 3:1 concentration significantly decreased the measured G’. The tubes remained within range of mechanical properties of elongation-permissive substrates while maintaining stable biotin concentrations over 12 weeks indicating biotin availability over time, an important consideration for eventual work in chronic SCI delivery. A small study of chronic delivery was probed at 6 months post-tube implantation. Accumulation was observed in the biotin-PEG group, highlighting that the tubes had yet to be fully degraded and still actively participated in targeting.

In future applications, it could be advantageous to explore options other than the implantable biomaterial tubes. Implanting the tubes is easy to do when a patient is having surgery following SCI to remove any remaining debris or decompress the tissue, and this approach has been taken in clinical trials for SCI-based implants (NCT03105882, NCT02138110, NCT03762655). The implantation approach, however, could potentially limit BEACON’s widespread use in applications where an implant is not needed. It is unlikely that a physician would recommend an additional surgery to solely implant the biomaterial target. To mitigate the invasiveness of implantation, an injectable biomaterial could be utilized in place of the tubes. We are currently investigating the feasibility of targeting our microsphere network when injected rather than implanted post-SCI.

To target the biotin-PEG tubes, SNPs were developed for beacon using a PLGA-PEG copolymer with subsequent streptavidin conjugation. When designing an NP-based drug delivery system, size, shape, and charge all play a crucial role impacting cellular uptake, circulation time, organ accumulation, and, importantly in the CNS, blood-brain/spinal cord barrier (BBB/BSCB) crossing [61-64]. For this study, the SNPs and CNPs had Z_average diameters (~70 nm) and zeta potentials (~ −30 mV) that were not statistically significantly different from each other, indicating the CNPs could serve as an appropriate negative NP control for non-targeted controls. In this work, the SNPs are delivered during an injury phase where the BSCB is still largely damaged, but eventually the BEACON system will be applied to chronic injuries where endogenous repair mechanisms and scarring will make it increasingly difficult for the SNPs to access the implanted tubes. Future applications could investigate the effect of different sized NPs on their accumulation and retention in the biotin-PEG tubes. Numerous studies have already demonstrated that small diameter NPs (<200 nm), like the SNPs, have an increased potential to pass the undisrupted BBB/BSCB compared to larger NPs [63,65]. With that in mind, it would be expected that early on in injury while the BSCB remains disrupted, there would be less of a noticeable effect of size on accumulation in the biotin-PEG tubes. As the injury progresses and the BSCB begins to repair, it is likely that larger NPs would no longer accumulate with as high of efficiency compared to their smaller counterparts. It should be noted, however, that while smaller NPs have demonstrated improved accumulation in the CNS, there is a functional lower limit that demonstrates as NP size decreases, so does the encapsulation efficiency of a loaded drug [66,67]. In future studies when a drug is to be loaded, it will be important to keep encapsulation efficiency in mind as we build on prior drug loading work with these base NPs by ourselves and others [26-29,31,32,68].

Additional targeting ligands, like the triphenylphosponium (TPP) cation, can be incorporated onto the particles as it has demonstrated efficacy at crossing the intact BBB [31,32]. Further, the negative zeta potential measured for the SNPs indicates a long potential circulation time, aiding in their targeting of the implantable biotin-PEG tubes as increased circulation increases the probability of reaching the tubes [62]. Negatively charged particles also have reduced non-specific uptake by the liver and the spleen [69] improving their trafficking to the tubes. Both Z_average diameter and zeta potential were stable when stored at 4 °C compared to 37 °C, as expected based on previous work [68]. Altogether, these data demonstrate that the SNPs can be stored or shipped safely, however, we use them within 3 days post-fabrication to ensure quality.

Administered SNP accumulation in the biotin-PEG tubes was evaluated 24 h following intravenous injection on day 6 post-SCI to allow sufficient circulation time and binding before data collection as PEGylated NPs have demonstrated improved circulation for at least 24 h post-injection [69,70]. For analysis, IVIS signal was normalized to each NP batch and condition to allow for comparison. The biotin-PEG + SNP combination was significantly increased over sham and SCI controls with SNPs and all injury/implant conditions with CNPs highlighting the requirement for both a targeted tube and streptavidin to be present for the BEACON system to function properly. Mice receiving biotin-PEG tubes with SNP injections had the greatest NP accumulation of all conditions tested, however, this was not significantly greater than mice receiving blank PEG tubes and SNPs. In vitro experiments demonstrated non-specific binding in control conditions was observed when SNPs were administered to blank PEG tubes. Further, when the tubes were incubated first in FBS to simulate potential fouling and blocking binding sites, the difference between PEG and biotin-PEG was decreased. With that in mind, it is possible that post-biotin-PEG tube implantation, adsorption on the tubes is impacting biotin binding site availability and limiting interaction with the SNPs to some degree and thus decreasing their binding potential, relative to the control biomaterial. Additionally, it is possible the injected SNPs are taken up by circulating immune cells that then carry the SNPs to the injury as demonstrated by the overlap in F4/80 and SNPs. Park et al . used PLGA-based NPs to reprogram immune cells to accumulate in an SCI where a non-targeted, implanted biomaterial bridge was used to guide tissue repair and promote functional recovery. The PLGA NPs were taken up by circulating macrophages that trafficked the NPs to the injury while modulating their polarization towards a pro-regenerative phenotype [71]. They report a 3-fold difference in NP accumulation in the group receiving NPs with the implant compared to their controls. A similar effect is observed in tumor targeting where macrophages migrate towards administered NPs and can carry them 2–5 times deeper into tumor tissues compared to diffusion alone [72].

Sequential dosing is another important consideration for BEACON as delivering multiple rounds of therapeutics would be advantageous. SNP accumulation in sequential doses has been investigated in vitro with no observable saturation of SNPs in the tubes, indicating rounds of therapeutics could be repeated as needed for individual patients. In future studies, sequential delivery of SNPs will be further investigated to determine if multiple rounds of therapeutics can be delivered. Initially, unloaded SNPs will be delivered in sequence, and this will be followed by studying sequential delivery of therapeutic-loaded SNPs. It has been shown previously that co-delivery of anti-inflammatory and neurotrophic factors have benefits in SCI repair [48]. Delivering a sequence of an anti-inflammatory loaded SNP first with subsequent neurotrophic factor loaded SNP second is of particular interest.

In targeted NP delivery for CNS injuries, a common strategy is to conjugate NPs with a short peptide containing the CAQK sequence [73-75]. CAQK selectively binds to chondroitin sulfate proteoglycans (CSPGs) that are upregulated post-injury [76]. NPs conjugated with CAQK have demonstrated improved targeting, but it should be noted that often these NPs are administered either immediately [75] or shortly after [73,74] the injury occurs. This is important to note because administering at this short time point indicates that the BBB/BSCB are still damaged resulting in increased permeability into the CNS. In our system, we administered our NPs 6 days after injury when the BSCB has had time to repair, thus making it more restrictive. Though it can be difficult to compare in a quantitative fashion, as NP accumulation is typically demonstrated as an IVIS signal that could be normalized differently, relative to other organ accumulation in each study saw similar accumulation of our NPs in the biomaterial at this time point compared to CAQK targeting NPs [74]. Additionally, it is important to note that CSPG upregulation is not sustained throughout the course of secondary injury as reactive astrocyte proliferation gradually stops [77], thus limiting the times at which CAQK NPs could be delivered whereas with BEACON, we can target the biomaterial at different time points without the need to reengineer for a new target. Last, other targeting papers rarely investigate retention of the NPs in the injury, but those that do have demonstrated improved delivery of their loaded drug [78]. Retention in the injury is an important consideration that must be taken into account if a therapeutic is going to have controlled release from the delivered NP, and we have demonstrated that our SNPs are retained over 5 days in our biomaterial target. Future work can tune therapeutic release to match this retention profile and more effectively deliver a dose of therapeutic-loaded SNPs.

Conjugating targeting molecules to NPs alters their physicochemical properties potentially decreasing binding, however. Ligand conjugation can increase NP corona formation, compared to unconjugated NPs that could block binding sites [79]. Streptavidin, the ligand of choice here, is a large protein (~60 kDa) [80], and large ligands have poor NP conjugation efficiency compared to smaller ligands [81], further limiting potential binding. When fabricating the SNPs, streptavidin is added in excess to allow for maximal binding, but steric hindrance between streptavidin molecules can potentially limit their conjugation to the NP surface. Additionally, steric hindrance could potentially limit the SNPs binding to the tubes, as the large ligand conjugated to the SNP might prevent the binding interaction between the SNP and the tubes. Decreased binding has been observed between streptavidin and biotin when biotin was attached to large ligands [82]. Moving forward, a smaller ligand could improve SNP binding to the tubes. Smaller biotin-binding peptides on the order of 2 kDa, ~30x smaller than streptavidin, have been identified [83,84], and their smaller size would limit steric hindrance and improve conjugation efficiency to the NPs. Future studies will probe conjugation efficiency, biomaterial binding and retention, and fouling with biotin-binding peptide conjugated NPs to further optimize the targeting and accumulation of the BEACON system.

We observed SNP accumulation to be on average the highest in the liver for all organs investigated. The liver is an important clearance organ heavily involved in NP circulation and processing by the body [85]. Notably, a significant decrease in liver accumulation was observed in animals receiving biotin-PEG and PEG implants compared to the sham surgery control. A significant reduction of liver accumulation could result in dramatic changes in toxicity as demonstrated via magnetic nanoparticles targeting an implanted stent in the carotid artery. Chorny et al. reported high uptake in the liver with their NPs, but also observed a significant therapeutic effect when using their targeting system while using a dose 3 orders of magnitude lower than the reported maximum tolerated dose for paclitaxel [86]. Limiting liver uptake of therapeutic NPs is an important design consideration when designing a drug delivery system susceptible to liver clearance. Liver clearance is often dependent on NP size and typically associated with larger NPs (>100 nm) [87], however smaller particles are still often highly processed by the liver [88]. In SCI, limiting liver uptake is especially important as glucocorticoids including methylprednisolone and dexamethasone can be used as a treatment option to reduce inflammation and promote repair [89,90]. In order to reach therapeutically relevant levels in the injured spinal cord, high-dose therapy is often required. High-dose pulses have been associated with liver toxicity and potentially inducing acute liver failure [91-93], however, high drug-loaded microspheres administered locally in the injured spinal cord promote significant functional recovery with minimal side effects [94]. Using the targeted system presented here we can similarly limit off-target accumulation, thereby reducing the potential risk of glucocorticoid-based treatment of SCI. Future work with the BEACON system will investigate glucocorticoid-loaded SNP delivery to the biotin-PEG tubes in SCI where immune modulation, axon elongation and remyelination, and functional recovery will be probed in addition to assessing liver toxicity compared to bolus glucocorticoid injection. The SNPs additionally will be able to extend the half-life beyond the quick clearance that is seen with glucocorticoids, as demonstrated by the pharmacokinetic results presented here. This will improve the safety of glucocorticoid-based therapies by limiting the need to redose with high concentration injections.

Similarly, the kidneys act as a clearance organ that is typically more associated with smaller NPs due to the physiology of the fenestrations of membrane layers [95-98]. NPs in the range of the SNPs in this study (~70 nm) have shown kidney accumulation likely due to degradation products that are small enough to pass through kidney membrane pores [95,96]. Compared to the liver, the kidneys had low SNP uptake, but on average for the organs tested, it was the second highest behind the liver. It should be noted that there was an increase in accumulation in the sham surgery condition compared to the other conditions as well. Other organs that can exhibit NP uptake post-systemic administration like the spleen [99] and lungs [100] had relatively low uptake, while there was only background signal in the brain and the heart.

In addition to the SNPs accumulating in the biotin-PEG tubes, it will be important for future applications of BEACON that they are retained in order to sufficiently release their therapeutic pay-load. In in vitro experiments, low binding and quick washout was observed when investigating NP retention in the absence of biotin, streptavidin, or both while the presence of both resulted in a dramatic increase in NP retention with slow unbinding from the tubes. For the in vivo work, IVIS imaging was performed daily for 5 days starting on day 7 post-SCI, to model the theoretical therapeutic release from PLGA-PEG-based NPs that persist on the order of days [29,31,32,68]. Longer retention at the biotin-PEG tubes vs the PEG only tubes importantly demonstrates that SNPs are actively binding to the present biotin, rather than passively accumulating via diffusion into the porous biomaterial. Altogether, this could require fewer doses of SNPs to be delivered thus improving the safety profile of the system by limiting complications attributable to off-target accumulation. Another contributor to improved retention was an increase in SNP accumulation on average on day 2 post-injection. Similar effects have been observed in other targeting systems where NP accumulation peaks between 1 and 3 days post-injection [37,38]. A 2 day differential between the targeted biotin-PEG with SNPs condition and the non-targeted PEG with SNPs was observed, highlighting the improvement of SNP delivery when using BEACON.

5. Conclusions

The work presented in this report details a targeted delivery strategy that eliminates the need to rely on tissue-specific biomarkers in favor of the BEACON platform, a targeted NP delivery paradigm that controls NP accumulation via biomaterial placement. A targeting system with controllable specificity based on biomaterial implant location improves the precision, longevity, and tailorability of targeted delivery by removing the reliance on targets that can change over the course of an injury or targets that can be largely expressed throughout an entire organ or organ system. In this work, the BEACON system utilizes the biotin/streptavidin complex, a high affinity interaction, by modifying our PEG tube implants to express biotin and serve as a receptor for targeted delivery of PLGA-PEG nanoparticles functionalized with streptavidin. An SCI model, characterized by a complex and dynamic injury microenvironment, was used in this work to demonstrate the utility of the BEACON system in a complex, dynamic injury environment. A significant increase in NP accumulation in the implanted biomaterial and a decrease in accumulation in the liver was observed. As the system is based on an implantable biomaterial target, it can broadly be applied to other disease or injury applications that would benefit from an implant and therapeutic delivery.