Synthetic nanoscale electrostatic particles as growth factor carriers for cartilage repair

Abstract The efficient transport of biological therapeutic materials to target tissues within the body is critical to their efficacy. In cartilage tissue, the lack of blood vessels prevents the entry of systemically administered drugs at therapeutic levels. Within the articulating joint complex, the dense and highly charged extracellular matrix (ECM) hinders the transport of locally administered therapeutic molecules. Consequently, cartilage injury is difficult to treat and frequently results in debilitating osteoarthritis. Here we show a generalizable approach in which the electrostatic assembly of synthetic polypeptides and a protein, insulin‐like growth factor‐1 (IGF‐1), can be used as an early interventional therapy to treat injury to the cartilage. We demonstrated that poly(glutamic acid) and poly(arginine) associated with the IGF‐1 via electrostatic interactions, forming a net charged nanoscale polyelectrolyte complex (nanoplex). We observed that the nanoplex diffused into cartilage plugs in vitro and stimulated ECM production. In vivo, we monitored the transport, retention and therapeutic efficacy of the nanoplex in an established rat model of cartilage injury. A single therapeutic dose, when administered within 48 hr of the injury, conferred protection against cartilage degradation and controlled interleukin‐1 mediated inflammation. IGF‐1 contained in the nanoplex was detected in the joint space for up to 4 weeks following administration and retained bioactivity. The results indicate the potential of this approach as an early intervention therapy following joint injury to delay or even entirely prevent the onset of osteoarthritis.

relief after the manifestation of the disease and include high molecular weight hyaluronic acid, glucocorticoids, and nonsteroidal antiinflammatory drugs (NSAIDs). These therapies have no inherent cartilage regenerative properties. Steroids and NSAIDs may even suppress chondrogenesis if administered in repeated high doses. Since cartilage tissue is avascular, therapeutic agents must be delivered locally to intra-tissue targets through the extracellular matrix (ECM) via intraarticular (IA) injections. While local delivery mitigates systemic toxicity, all therapeutic drugs delivered through IA injections are cleared rapidly, with a clearance half-life that is weakly proportional to the molecular weight and spans from a few minutes to a day. 3,4 Hence, pain-relief mediated by the drugs is transient, if it occurs at all, and often insufficiently effective. 5 This timescale is also unsuitable for resolving inflammation and stimulating repair of cartilage, which can require the sustained presence of large molecules, such as growth factors, at therapeutic levels.
The use of growth factors as therapeutics to reverse or inhibit cartilage degradation has been an important focus in cartilage research.
By delivering agents that promote cartilage repair, it may be possible to create a treatment that not only alleviates pain for extended periods, but can also result in full, functional joint recovery. To this effect, the use of pro-anabolic growth factors has been explored as a means to stimulate cells to increase chondrogenesis via mitogenic activity and to produce cartilage matrix. The roles of the transforming growth factorbeta 6 (TGF-b) growth factor family, insulin-like growth factor-1 7 (IGF-1), and fibroblast growth factors 7,8 (FGF-2/FGF-18) in signaling chondrocytes to produce cartilage ECM are well known. In our own work, we have examined the pro-anabolic effects of IGF-1. 9,10 Other studies have shown that chondrocytes produce IGF-1 that can stimulate ECM biosynthesis in native cartilage and tissue engineered constructs in a dose dependent manner. 11 A key barrier to the clinical use of growth factors for cartilage repair is the delivery of therapeutically relevant amounts of growth factors to target cells over appropriate time scales to induce tissue regeneration. 12 Although it would be most desirable to deliver growth factors directly to chondrocytes for the greatest and most efficacious results, the diffusion of free growth factors from the joint space to the deep zones of the cartilage is hindered and inefficient. 13 Transport through cartilage is regulated by tightly packed collagen fibrils and highly charged proteoglycans of the ECM, which account for 60-70% and 15-35% of the dry weight of the tissue, respectively. 14 The combination of size and charge limitations makes the cartilage matrix a formidable barrier to the entry of therapeutic molecules and carriers.
The direct bolus injection of growth factors results in the rapid clearance of the growth factor out of the joint space before it can be effective as a therapeutic. Thus, supraphysiologic amounts, well above the therapeutically relevant levels, must be introduced to induce a biological effect, which can lead to safety concerns. Encapsulation of growth factors and localized delivery can potentially address many of these issues. However, the incorporation of growth factors into common polymeric drug carriers can lead to exposure to undesirable processing conditions such as temperature and solvent that often denatures the protein. Cytotoxic byproducts may also be created when the polymer itself degrades in confined environments of the body.
We hypothesized that by creating depots of IGF-1 within the cartilage matrix, IGF-1 can be presented for long periods of time and provide sustained signaling to promote matrix biosynthesis, chondrogenesis, and protection against cartilage degradation. In this report, we tested this hypothesis by using a growth factor carrier assembled using charge interactions between the IGF-1 and synthetic polypeptides, negatively charged poly(glutamic acid) (pGlu) and positively charged poly(arginine) (pArg), into a charged nanoscale polyelectrolyte complex (nanoplex). We observed that the IGF-1 retained bioactivity when incorporated in the nanoplex. In a rodent model for articular cartilage damage, we observed that when the IGF-1 was delivered from the nanoplex, it remained at a therapeutically relevant level within the joint space over a period of 4 weeks, protected the cartilage from degradation and mitigated joint inflammation by controlling interleukin-1 (IL-1) levels following injury.

| Constructing the nanoplex
Solutions of IGF-1 (Biovision, Milpitas, CA), poly(L-glutamic acid) (Alamanda Polymers, Huntsville AL, MW 7.5 kDa) and poly(L-arginine) (Alamanda Polymers, Huntsville, AL MW 5.8 kDa) in ultrapure water were prepared. First, the polyGlu was added to the IGF-1 at a rate of 1 ml/min under high agitation. Next, the polyArg was added to the IGF-1/polyGlu mixture at a rate of 1 ml/min under high agitation. The final molar concentrations of the components were 1:5:5 (IGF-1: poly-Glu:polyArg) The resulting particles were stirred for 1 hr and passed through 0.22 lm syringe filters. The particles were dialyzed against water using a 10 k MWCO membrane (Spectrum Labs). Particle size and zeta potential were measured using a Zetasizer Nano ZS90 (Malvern).

| Cryo-TEM
Nanoplex samples were prepared and immediately processed for cryo-TEM imaging. The nanoplex samples were processed using a vitrification robot (Vitrobot, FEI) in which the relative humidity is maintained close to saturation. A 3-lL drop of the solution was placed on a carbon-coated lacey substrate supported by a TEM 300 mesh copper grid (Ted Pella). After automatic blotting, the grid was rapidly plunged into liquid ethane at its melting temperature. This resulted in a vitrified film. The vitrified specimen was then transferred under a liquid nitrogen environment to a cryo-holder (model 626, Gatan Inc., Warrendale, PA) into the electron microscope, Tecnai 20, Sphera (FEI), operating at 200 kV with a nominal underfocus of 2-4 lm. The working temperature was kept below 21758C, and the images were recorded on Gatan 794 MultiScan digital camera and processed with Digital Micrograph 3.6. Particle sizes were quantified using thresholding and particle size counting algorithms within ImageJ (NIH).

| BrdU testing for IGF-1 bioactivity
The ability of IGF-1 to induce proliferation of an immortalized chondrocyte cell line, CHON-001 (ATCC) was used to measure bioactivity. 5000 CHON-001 cells at low passage number (less than P4) were dispensed in each well of a 96 well plate (quantification) in 0.2 ml of growth medium (Dulbecco's Modified Eagle's Medium supplemented with 10% fetal bovine serum and 1% antibiotic solution containing penicillin and streptomycin). Cells were allowed to establish for 24 hr.
Nanoplex IGF-1 and free IGF-1 at matched concentrations were then added to the wells. Cells were allowed to proliferate for 24, 48, and 72 hr, without media change. At the end of each timepoint, cells were labeled with 10 lM bromodeoxyuridine (BrdU) for another 2 hr (Roche Diagnostics, Mannheim, Germany). Incorporation of BrdU into DNA was detected using chemiluminescence, following the manufacturer's protocol.

| Bovine cartilage explant harvest and culture
Three-millimeter diameter cartilage explants were biopsied from the femoropatellar grooves of young (1-2 weeks old) bovine knee joints (Research 87, Boylston, MA). These explants were trimmed to 1 mm thickness at the superficial zone of the cartilage. Explants were cultured in Dulbecco's Minimal Essential Medium plus 10% v/v fetal bovine serum, supplemented with pen/strep antibiotics, L-Proline, ascorbic acid, HEPES, sodium pyruvate, and non-essential amino acids.
Explants were cultured for 48 hr after harvest, prior to any experiments to allow the tissue to recover from the harvesting procedure.

| 35 SGAG biosynthesis assay
To assess cartilage matrix biosynthesis in response to treatment, cartilage explants were incubated with IGF-1 alone, IGF-1 in nanoplex form, or culture medium in the presence of 35 S. Cartilage proteoglycan (such as aggrecan) biosynthesis in this timeframe would incorporate 35 S radiolabel into the tissue, which was measured by digesting the tissue with proteinase K and assaying the tissue extract with a liquid scintillation counter (PerkinElmer). Sulfated proteoglycan biosynthesis rate in response to treatment is proportional to the amount of 35   for 4 days to simulate early effects of injury on matrix loss, a treatment that causes 20% GAG loss, predominantly from the superficial zone. 15 After thorough washing to remove phenol red containing media, explants were incubated with labeled IGF-1 or IGF-1 nanoplex in PBS for an additional 48 hr. Explants were again washed thoroughly in PBS to remove any weakly bound material. The 3-mm diameter, 1mm thick cylindrical explants were then sectioned along their long axis to yield thin sections 3 mm long and 1 mm wide. These sections were immediately imaged with a Nikon A1R confocal microscope to visualize the diffusion gradient of IGF-1 or nanoplex IGF-1 along the 1 mm thickness of the explant. was used to acquire and quantify the fluorescence in the same animal (ex/em 649/675). Images were taken periodically after injection until the signal was not detectable.

| Histology analysis
After euthanasia, the joints were excised, trimmed and fixed in 4% paraformaldehyde (PFA) for 48 hr and transferred to a 70% ethanol solution. Joints were partially decalcified for about 4 hr using a rapid decalcifying formic acid/hydrochloric acid mixture (Decalcifying Solution, VWR). The joint area was cut in cross-section with a razor blade and embedded in paraffin wax. Sections (5 mm) of the cross section were stained with Safranin-O and imaged using bright-field microscopy. Heat-induced epitope retrieval was used prior to applying immunohistochemical stains for detecting aggrecan and collagen. Anti-Aggrecan and anti-collagen antibody (Abcam) was used in a 1:200 dilution in Tris-buffered saline and with an HRP conjugate for detection.

| R E S U LTS
By taking advantage of weak electrostatic charge interactions for the generation of self-assembled nanoscale carriers, we used low molecular weight synthetic polypeptides to package IGF-1. We tuned the ratio of the polypeptides, negatively charged pGlu and positively charged pArg, to induce self-assembly into organized and charged nanostructures.
The formation of nanoplexes was induced by the sequential addition of positively charged IGF-1, pGlu, and pArg (Figure 1a, b). The ratios of the components were varied across a broad range to explore the size and zeta potential of the nanoplex using dynamic light scattering (DLS) coupled with zeta potential analysis. Over the molar ratios that we explored, the number averaged hydrodynamic size and zeta potential of the resulting nanoplexes as measured by DLS ranged from 100 to 850 nm and 0 to 1140 mV, respectively (Figure 1c). The polydispersity index ranged from 0.1 to 0.25. To maximize penetration into the cartilage to form IGF-1 depots, we selected the nanoplex formulation with the lowest apparent hydrodynamic radius based on DLS, which consisted of 10% IGF-1 by mass. We used cryo-electron microscopy to more directly measure the nanoplex particle size and shape. (Figures 1d   and S1). Interestingly, cryo-EM suggested a substantially smaller size The nucleus (blue), membrane (magenta), pArg (red), and IGF-1 (green) have been stained. Yellow shows assembled nanoplex. (f) Bioactivity of IGF-1 as measured by BrdU proliferation assay of CHON-001 cells (n 5 6 per condition, **p < .01, *p < .05) (g) Rate of cartilage tissue sulfate incorporation in response to treatment with 20 mM nanoplex, 20 mM IGF-1 alone, or no treatment added to cartilage media consisting of DMEM, 10% FBS, non-essential amino acids, pyruvate, HEPES, ascorbate, and proline (n 5 10 per condition, **p < .01) populations of particle size. However, the cumulant method used for calculating the particle size applies greater weight to the larger scatterers. We thus conclude that the nanoparticle formulation consists predominantly of positively charged nanoplexes containing IGF-1 that are less than 20 nm in size, as well as some larger colloidal aggregates that are comprised of these smaller nanoplexes.
We assessed the ability of the nanoplex IGF-1 to interact with  Figure 2a). The nanoplex delivered IGF-1 and free IGF-1 had a similar distribution profile throughout the section (Figure 2b, c). We then investigated transport within injured cartilage by treating the explants with IL-1a prior to incubation with nanoplex. Both IGF-1 and pArg were labeled to analyze nanoplex disassembly within the tissue. pArg and IGF-1 co-localize as the nanoplex enters the tissue, but as it moves deeper, only IGF-1 is present (Figure 2d, e) We examined the efficacy of the nanoplex growth factor delivery system in a rat osteoarthritis model (Figure 3). We treated the animals via intra-articular injection of the nanoplexes at 48 hr after inducing To monitor the level of joint inflammation, we measured IL-1b levels in the synovial fluid of the injured joint and the contralateral joint, in which surgery was not performed. In untreated animals, IL-1b levels increased steadily over the course of the study (Figure 3c). When free IGF-1 was administered, there was an initial decrease in the IL-1b concentration in the synovial fluid, which gradually increased 2 weeks after IGF-1 injection. When treated with the nanoplex, the IL-1b production in the joint was suppressed at a low level, comparable to the uninjured joint, over the duration of the study. We also measured IGF-1 concentrations in the synovial fluid (Figure 3d). We observed the same trend as that of the IVIS fluorescence measurements. IGF-1 was detectable for several weeks longer in animals treated with the nanoplex IGF-1 compared with free IGF-1, which was undetectable after day 9

| DISCUSSION
We examined the long-term delivery of a therapeutic growth factor (IGF-1) within the cartilage tissue by packaging it into an offering the benefit of sustained, long-term therapy. We selected poly (glutamic acid) as the polyanion, which is a water-soluble, biodegradable, and edible biopolymer. It is clinically used in FDA-approved polymer-drug conjugate systems. 16 When used to deliver IGF-1, we observed stable binding of the poly(glutamic acid) and the IGF-1. Poly (arginine) has been widely examined in drug and gene delivery applications for its ability to transport molecules across cell membranes. 17,18 In our work, poly(arginine) packaged the poly(glutamic acid)-IGF-1 complex and conferred a net positive charge to the complex for transport through negatively charged cartilage ECM. The assembled nanoplex demonstrated co-localization with the membrane of chondrocytes (Figure 1e), which could be due to either nonspecific charge interaction with the membrane itself or specific interaction with the extracellular IGF-1 receptor. Electrostatic interactions of the nanoplex with the highly negatively charged cartilage matrix are anticipated to facilitate early penetration of the complex. These same interactions may also enable unpackaging of the complex in cartilage tissue. Both polypeptides used can be broken down over longer timeframes into their constituent amino acids and efficiently processed and cleared by the body.
While numerous growth factors stimulate chondrocytes to synthesize cartilage ECM, 19 IGF-1 is the main pro-anabolic growth factor for articular cartilage and is crucial in maintaining cartilage homeostasis and remodeling cartilage. 20 The role of IGF-1 in articular cartilage metabolism has been extensively investigated in both healthy and diseased joints. In our own work, we have demonstrated in vitro that IGF-1 stimulates proteoglycan production in a dose-dependent manner. 21 IGF-1 has been demonstrated to induce the production of hyaline-like cartilage with native mechanical properties, as opposed to fibrocartilage. 22 The negatively charged GAG chains of aggrecan play a unique role in determining the transport properties of the cartilage. 23 It has been estimated that the effective mesh size of these networks is on the 10 s of nanometers scale in healthy cartilage. 24 This matrix is highly anionic and increasingly impermeable to molecules much greater than the size of albumin (67 kDa), depending upon their charge and conformation. 25 The nanoplex construct, with a net positive charge and 10-40 nm size range for non-aggregated particles, is suitable for diffusion into and through this matrix; cartilage injury may enable diffusion of even larger molecules. 26 It is particularly notable that these nanoplex particles are soft, dynamic, charged assemblies capable of engaging with the cartilage based on electrostatics and undergoing changes in size, shape, and colloidal stability during diffusion.
To reduce the risk of infection and joint damage, intra-articular injections are administered infrequently, with intervals that can span weeks to months. 27 Hence, sustained delivery of the growth factor is key to the success of this strategy. Various intra-articular drug carriers based on hydrogels, liposomes, and polymeric micro and nanoparticles have been reported. [28][29][30][31][32] Recent in vivo studies with small nanoparticles have indicated the possibility of penetration into 50 micronthick mouse cartilage, with a multi-day residence time within the cartilage space. 33 Whereas many of these systems have demonstrated some level of efficacy, their applicability is limited to delivering small molecules and by long-term instability. Only recently, have delivery systems begun to harness charge interactions, which play a critical role in mediating transport through cartilage. 34,35 Electrostatic interactions between soluble IGF-1 and GAGs regulate transport and binding within cartilage. IGF-1 is a basic protein (pI 8.5) that partitions into cartilage on the basis of size and charge; in addition, IGF-1 binds to specific domains (IGF binding proteins) that are contained within the ECM. 36 The advantage of a nanoplex system is that it can transport the IGF-1 out of the high clearance rate joint space into the cartilage matrix to create depots that sustain IGF-1 release and signaling. The nanoplex has a high cargo capacity and densely packs the IGF-1, which comprises approximately 10% of the total mass of the particle and transports it through the cartilage tissue. Conventional bulk polymer delivery systems typically comprise about 1% of growth factor by mass, a fraction of which is bioavailable. 37 Importantly, we observed that the IGF-1 maintained its bioactivity in the nanoplex (Figure 1e-g).
In bovine cartilage explants, we observed that a single dose of the nanoplex was sufficient to induce comparable levels of GAG synthesis as free IGF-1. We note that the dose of IGF-1 used in these studies While efficient transport through cartilage is desirable, overcoming joint clearance mechanisms in vivo is also crucial to transport therapeutically relevant levels of IGF-1 into the cartilage at therapeutic timescales. Efficient lymphatic drainage rapidly removes materials from the synovial fluid resulting in an intra-articular dwell time of proteins on the order of several hours. 38 In inflamed joints, fluid flow to the lymphatics is enhanced resulting in even faster clearance of macromolecules. 39 Using fluorescence and ELISA measurements, we observed that when the IGF-1 was packaged in the nanoplex, its half-life in the joint significantly increased as measured by IVIS fluorescence. The IGF-1 clearance rate from the synovial fluid was constant (0.7 nM/day,  40,41 We measured the IL-1b concentration in the synovial fluid, which would be sampled continuously. We note that the concentrations in harvested whole knee joints and blood maybe higher in similar OA models. 42,43 Our prior work with human cartilage and that of others has indicated that an IGF-1mediated anabolic effect is the primary mechanism by which new matrix is synthesized. 15,19 However, it has also been demonstrated that IGF-1 can suppress IL-1b-induced NF-jB activation via inhibition of IjB-a kinase, which is an anti-catabolic mechanism. 44 The reduction in IL-1b levels in the synovial fluid are likely an indirect effect of suppressing inflammation.
In our prior work, we have observed quantifiable GAG loss as early as 48 hr after incubation of IL-1a with bovine cartilage plugs, 15,45 mainly from the superficial zone. Based on this understanding of the progression of OA, we administered the therapy within 48 hr after surgery, during the acute inflammation phase, to mimic an earlyintervention treatment. It is unlikely that there were any major macroscopic defects in the cartilage structure at this timepoint post-surgery, though tissue GAG loss has likely been initiated. The rat joint injury model used in the study has been well characterized, in which minor focal defects appear at around 1 week after injury and worsen into serious defects by 4 weeks. 45 In this study, the cartilage maintained its structure when IGF-1 was delivered from the nanoplex and was morphologically similar to uninjured cartilage. Safranin-O staining indicated that the cartilage tidemark was well defined, with an even distribution of proteoglycans. We observed greater loss of proteoglycans when only free IGF-1 was used. The untreated control resulted in nearcomplete loss of proteoglycan, with the tidemark moving close to the articulating surface. IHC indicated qualitatively that aggrecan content after ACLT with nanoplex treatment was more similar to that in the normal uninjured joint. Taken together, the observations suggest our treatment likely mitigates degeneration to some extent and induces new tissue formation to repair the tissue loss that could not be prevented over the time frame of the study.
The materials used in this study are readily translatable to clinical application. Both glutamic acid and arginine are on the FDA inactive excipient list. Recombinant human IGF-1 is a clinically used protein therapeutic for the treatment of growth failure and growth factor deficiency disorders and is manufactured for these indications. 46 In the nanoplex system, all materials are held together by electrostatic charge, and used without chemical or biological modification. All of these factors ease the clinical translation of this approach. In the standard rodent model of joint injury described in the study, we demonstrated that the nanoplex system delayed cartilage degradation in a mechanically destabilized joint complex. Our studies demonstrated the efficacy of the nanoplex system in preventing cartilage degradation over a period of 4 weeks post-injury. Long-term cartilage degradation is inevitable in the surgical model of mechanical joint destabilization in a rodent. We envision that the nanoplex system will supplement the standard pharmacological therapy that is administered after a significant joint injury, such as an ACL tear, which is well correlated with post-traumatic OA. Prior to a human clinical trial, we anticipate that the nanoplex system will need to be tested over a longer timescale in a larger animal OA model (e.g., sheep) in which the injury and intervention more closely mimics that of a human joint injury. These larger animal trials will also help determine the nanoplex dose range for therapeutic benefit in humans. In clinical trials, both short-term (alleviating inflammation) and long-term (preventing/delaying onset of OA) patient outcomes and data will help guide decisions for clinical use.

| C O NC LU S I O N S
Regenerative strategies to replace damaged cartilage are often used, but only after the clinical symptoms manifest, at which time it is often too late to rescue degraded cartilage tissue. This work demonstrates a strategy to potentially prevent cartilage degradation through the use of a charged carrier system to transport an anabolic protein into the cartilage for therapeutic efficacy. The nanoplex offers a platform for combining multiple complementary and synergistic growth factors to promote cartilage repair and a way to efficiently deliver potent biologics through the dense cartilage ECM barrier. The work highlights a potential interventional therapy for trauma-induced joint injuries that can potentially reverse cartilage damage and prevent progression to irreversible joint degradation.

ACKNOWLEDGMENTS
The authors acknowledge assistance from the Hope Babette Tang