Leveraging Electrostatic Interactions for Drug Delivery to the Joint

Abstract

Arthritis is a debilitating joint disease with a high economic burden and prevalence. There are many challenges delivering therapeutics to the joint, including low bioavailability when administered systemically and low joint retention after intra-articular injection. Therefore, drug delivery systems such as nanoparticles, liposomes, dendrimers, and carrier proteins have been utilized to overcome some of these limitations. To enhance joint tissue localization and retention, there are opportunities to leverage electrostatic interactions between drug carriers and various tissues and cells. These opportunities, as they pertain to specific joint tissues, are explored in this review. Further, the impact that electrostatic interactions has on various drug delivery parameters, such as the formation of a protein corona, the uptake and cytotoxicity, and the biodistribution of the drug delivery systems, is discussed. Lastly, this review summarizes key findings from studies that have investigated the use of electrostatic interactions to increase targeting of specific joint tissues and limitations in preclinical investigations are identified. As more novel targets are discovered in treating arthritis, there will be a continued need to localize therapeutics to specific tissues for greater therapeutic outcomes and hence attention must be paid in designing the drug delivery systems.

Introduction

Arthritis, the most common knee joint disorder, affects 23% of all adults in the United States with annual direct medical costs of at least $140 billion.¹ Although osteoarthritis (OA) is the most common form of arthritis, it shares a common symptomology of pain, stiffness, and swelling as other types of arthritis, including rheumatoid arthritis (RA), gout, and lupus. These symptoms lead to work disability and other co-morbidities.² Clinically available interventions are being delivered by either systemic administration or direct intra-articular injections into the joint. Although there are disease-modifying drugs that can effectively treat some types of arthritis,³ many of the therapies suffer from common drug delivery limitations when delivered systemically, such as poor bioavailability to joint tissues and, consequently, insufficient dosages in target tissues.⁴ Even with direct injections to the joint, therapies experience rapid clearance from the joint via venules and lymphatic vessels located in the synovium and therefore do not reach their biological targets at sufficient levels.^5,6

As such, there has been significant preclinical investigation into leveraging passive or active targeting approaches with these intra-articular therapies to direct them toward specific tissues in the joint. Passive targeting utilizes the fundamental physicochemical properties of the therapy or its carrier such as size or surface charge to encourage greater, although nonspecific, interaction with the target tissue or cell. Active targeting approaches employ specific biochemical moieties to selectively bind to biological components of the target tissues' extracellular matrix (ECM) or cell surface. This includes modifying the surface functionality of therapeutics or their carriers to biologically interact with specific joint tissues or cells, enabling better targeting and increased joint retention.

Although a combination of active and passive targeting approaches will likely improve tissue targeting, passive targeting approaches offer less complexity in terms of minimizing chemical conjugation steps and, consequently, less costly scale-up of the manufacturing processes relative to active targeting modalities. Further, in some cases, regulatory classifications may be affected by the incorporation of biologic targeting moieties. These translational considerations may explain why the majority of U.S. Food and Drug Administration (FDA)-approved nanoparticle (NP)-based medicines achieve their effects through passive overactive targeting modalities.⁷ Lastly, active targeting may only lead to modest improvements in drug efficacy over passive targeting, as demonstrated in a review of cancer nanomedicine strategies,⁸ and hence might not justify the increased cost of production. In this manner, it would be prudent to closely examine passive targeting strategies such as the use of electrostatic interactions with biological targets.

The goal of this review is to highlight ways in which electrostatic interactions have been leveraged when designing various therapeutics or drug delivery carriers to achieve tissue-specific delivery within the joint. This review will include discussion of delivery parameters such as cell uptake and formation of protein corona that are influenced by the surface charge of therapeutics and their electrostatic interactions with tissues in the joint. It will also examine current investigations that have leveraged electrostatic interactions to target specific joint tissues.

Anatomy of a synovial joint

Arthritis impacts synovial joints in the body such as the knee joint, hip joint, and joints in the hand. These are characterized by the presence of a joint cavity filled with synovial fluid and contained within a fibrous capsule. The synovial joint contains several anatomical structures that are not present in fibrous or cartilaginous joints. The key components of the synovial joint include (1) the synovium, (2) articular cartilage, and (3) synovial fluid (Fig. 1).

FIG. 1.

Changes in joint tissues with progression of arthritis and changes in opportunities for targeting using electrostatic interactions in each of the major tissue types. Adapted from Brown et al.⁴⁷

The synovium is a highly vascularized, thin (20–40 μm) heterogeneous connective tissue that lines diarthrodial joints. Intercellular gaps of ∼0.1–5.5 mm exist at the synovial surface, under which a rich network of fenestrated and continuous capillaries are found.^9,10 As a result, the synovium is also the site at which transport of molecules in and out of the joint occurs. The synovium plays an important role in the secretion and turnover of synovial fluid, the lubricating fluid present in synovial joints.¹¹

Therapeutics designed for delivery to the synovium typically target one of the synoviocyte types: (1) Type A synovial macrophages, a minority population in normal joints, increase substantially under inflammatory conditions and produce pro-inflammatory mediators. These are the primary phagocytic cells of the joint that recognize and phagocytose foreign materials, possibly including drug delivery systems. (2) Type B fibroblast-like cells, the dominant population in normal synovium, primarily produce ECM components of the synovial fluid and synovium.¹² There are opportunities for electrostatic interactions to be leveraged when targeting the synovium, though these have not been investigated in preclinical investigations to date.

Articular cartilage is an avascular and aneural tissue that lines the articular surface of bones and allows for near-frictionless movement and shock absorption.^13,14 Cartilage comprises chondrocytes, the sole cell type residing in an organized network of collagen type II, a protein that is both unique to and abundant within this tissue. The collagen network has a mesh size of ∼50–60 nm^15,16 and is primarily filled with sulfated glycosaminoglycan-containing proteoglycans, with ∼20 nm spacing between side chains, which give the cartilage an anionic charge.^17,18 Targeting strategies for cartilage have leveraged these unique ECM characteristics, including passive targeting of nanomaterials to cartilage via electrostatic attraction to the anionic cartilage matrix^19,20 and active targeting strategies using binding moieties or agents specific to collagen type II.^21–23

Synovial fluid, the viscous non-Newtonian fluid within the joint space, is essential not only for shock absorption and lubrication but also for modulating the transport of various molecules to different tissues of the joint. Synovial fluid is a protein-rich mixture of plasma and hyaluronan synthesized by type B fibroblast-like synoviocytes.²⁴ The hyaluronan, a copolymer of N-acetyl-d-glucosamine and d-glucuronic acid, is a high-molecular-weight, negatively charged polysaccharide and major component of synovial fluid. Investigators have targeted the anionic charge of the synovial fluid with cationic carriers as a method to improve their residence time with the joint.^25–28

Other tissues in the joint such as the meniscus, fat pad, and various intra-articular ligaments (such as the anterior cruciate ligament) each contribute to joints in various ways.²⁹ Although the cartilage or the synovium are key targets for injectable therapeutics, these other tissues are also potential targets and often become depots for off-target accumulation.³⁰ For the purposes of this review, only the cartilage, synovium, and synovial fluid, tissues that are consistent across all synovial joints, will be examined for targeting of drug delivery systems using electrostatic interactions.

Pathogenesis of arthritis

Although the various types of arthritis have different etiologies, once initiated, their progression is often driven by a vicious cycle of pathological changes in multiple joint tissues (Fig. 1). In OA, the cartilage releases products of proteoglycan degradation and damage-associated molecular patterns (DAMPs) into the joint space that activate immune cells in the synovium.^31,32 Once activated, the type A synovial macrophages secrete pro-inflammatory cytokines, such as interleukin-1 (IL-1), IL-6, and tumor necrosis factor alpha (TNF-α),³³ which are transported through the synovial fluid and interact with the resident chondrocytes within cartilage. These soluble signals cause chondrocytes to accelerate the production of catabolic enzymes such as matrix metalloproteinases (MMPs)-1, -3, and -13 and a disintegrin and metalloproteinase with thrombospondin motifs (ADAMTS)-4 and -5 (aggrecanases-1 and -2).^34–36

A positive feedback loop is thus initiated as these proteinases degrade the cartilage matrix and release more DAMPs, thereby propagating inflammatory cascades in the synovium and sustaining chronic low-grade inflammation. Coupled with the inherent limited self-renewal capabilities of cartilage,³⁷ this environment yields conditions in which cartilage ECM breakdown outweighs repair or synthesis. As disease progresses, this pathological cycle is sustained and ultimately results in cartilage degradation, subchondral bone outgrowth, synovitis, disruption to ligament molecular structures, meniscal damage, and alternation of joint mechanics.³⁸ This culminates in clinical symptoms such as pain and disability for the patient.

On the other hand, in an auto-immune type of arthritis such as RA, the pathology typically involves both genetic and environmental factors, resulting in a triggered autoimmune response that can affect multiple joints. Both adaptive immune cells, such as B cells and T cells, and innate immune cells, such as the type A synovial macrophages, are key players in the development of immunological events in RA. During the initial stages of the disease, there is activation of antigen-presenting cells such as type A synovial macrophages, which then elicits specific humoral cellular activation. After trafficking to the synovium via synovial vasculature, the B cells will add to the inflammatory process through antibody production and T cells lead to macrophage activation and the overproduction of inflammatory cytokines such as IL-1 and TNF-α.^39,40

The presence of the activated T cells and B cells results in increased production of cytokines and chemokines, leading to a feedback loop for additional T cell, macrophage, and B cell interactions. The upregulated production of TNF-α and IL-1β induces synovial cells to produce tissue-degrading MMPs. In addition, these cytokines stimulate the proliferation of type B fibroblast-like synoviocytes, forming a pannus that is capable of invading cartilage and bone, leading to joint destruction.^41,42 Synovial fibroblasts stimulate the production of vascular endothelial growth factor leading to angiogenesis, which further perpetuates the inflammatory process by recruiting more inflammatory leukocytes.⁴³ In addition, the increased levels of TNF-α in the joint stimulate the maturation of osteoclasts, which results in erosion of the subchondral bone. The persistent inflammation that occurs in RA leads to systemic complications and comorbidities, which finally result in handicap and social dysfunction.

Overall, RA is characterized by a much more aggressive immune response that has a greater systemic impact than OA, which comprises a low-grade chronic inflammatory process localized to specific joints. Accordingly, therapeutics intended for more systemic conditions such as RA might need to be administered systemically whereas localized conditions such as OA could benefit from local injections of therapeutics.

Drug delivery for arthritis

Depending on the type of arthritis, the therapeutic targets can be varied. In RA, current clinical agents for treatment include corticosteroids, nonsteroidal anti-inflammatory drugs (NSAIDs), and disease-modifying antirheumatic drugs⁴⁴; whereas in OA, corticosteroids, hyaluronate injections, and NSAIDs including analgesics are typically prescribed in the clinic.⁴ NSAIDs and many anti-inflammatories are taken orally whereas therapeutics with poor bioavailability, stability, or systemic toxicity are injected intra-articularly. If delivered systemically, soluble substances enter the joint space via the capillary network of the highly vascularized sub-synovium. With this administration route, there can be off-target prolonged systemic exposure to these drugs, which can increase the risk of gastrointestinal complications such as gastric ulcerations and gastrointestinal bleeding as well as thrombotic cardiovascular complications for patients.⁴⁵

Conversely, direct intra-articular injections are commonly used in the clinical treatment of arthritis to overcome off-target systemic effects. With these direct injections, therapeutics tend to have immediate contact with joint tissues, having bypassed the physiological barriers associated with systemic entry into the joint. This route has the advantage of increased local drug concentration at the site of action, which leads to an overall decreased drug dose and minimized side effects.

However, this route of administration also comes with its limitations, including injection-site reactions in certain individuals and infection.⁴⁶ As previously mentioned, another major limitation of intra-articular injections is often the limited residence time of the therapeutic within the joint space after injection.⁴⁷ The unique barriers posed by each joint tissue to localization and uptake of therapeutics are also often overlooked when designing drugs for intra-articular delivery.⁴⁷ For instance, cartilage, a target of interest in joint preservation, comprises a spare population of cells within a dense avascular ECM. When faced with such tissue barriers, drugs are unable to reach their intra-tissue target sites at optimal doses for clinical efficiency. In this manner, the design of drugs for joint delivery needs to accommodate these additional requirements for joint tissue targeting.

To increase drug retention time and reduce drug clearance from joints, a multitude of controlled release drug delivery systems, including NPs, microspheres, liposomes, dendrimers, and hydrogels, have been developed. These sustained release systems for intra-articular injection have demonstrated effective conservation within the joint cavity.⁴⁸ They also improve the pharmacological activities of therapeutic drugs and overcome problems such as insufficient dosage in target pathological tissues, limited solubility, drug aggregation, low bioavailability, and poor biodistribution within the joint.

Herein, the term “drug delivery system” will be used to refer to any system that is either conjugated to or used to encapsulate a therapeutic, with the purpose of improving the pharmacological activity of that therapeutic through increased interaction between the therapeutic and its target cell or tissue type. These drug delivery systems will be examined in the next sections to determine the ways in which they have utilized electrostatic interactions to target specific joint tissues.

Often in the literature, studies interchangeably use the terms “surface charge density,” “surface potential,” “net charge,” “charge,” “zeta potential,” and “fixed charge density.” To understand the differences between these terms, it is important to delve into the electrostatic interactions at an ionic level. Positive counter-ions first attach to the negatively charged particle or vice versa, forming a rigid layer called the “Stern layer.” The particle continues to attract more counter-ions but now these counter-ions are being repelled by other counter-ions in the vicinity and by the Stern layer itself. This second region is called the diffuse layer and together, both layers are referred to as the double layer, which moves with the particle in a given medium. Surface potential is the electrical potential between the surface of the particle and any point in the suspending liquid. The zeta potential is the electrical potential at the slip plane where the surrounding liquid and double layer meet. Surface charge density is a theoretical measure of how much electric charge is accumulated over a surface. Most studies typically refer to the zeta potential of a particle when referring to its surface charge, though in principle these metrics are not synonymous.

Drug Delivery Systems That Leverage Charge for Delivery to the Joint

Due to poor drug solubility and tissue distribution of the drug within the joint, there has been significant investigation into designing drug delivery systems to improve on these limitations. Passive targeting strategies have been exploited by investigators to localize therapeutics to target tissues or cells. For instance, an FDA-approved formulation Zilretta^® utilizes the size properties of microspheres to enable greater retention in the joint and extended release treatment of the active ingredient from within the microsphere for arthritis treatment after intra-articular injection.⁴⁹

However, nanomaterials may be more beneficial than micron-sized carriers when utilizing passive targeting strategies. For instance, the smaller sizes of nanomaterials enable them to penetrate barriers in tissues such as the cartilage and localize intracellularly. Nanomaterials may also be advantageous over micron-sized materials when utilizing surface charge properties due to their high surface area to volume ratio. This property provides them with a greater reactive interface between themselves and the local environment or target tissue, enabling greater electrostatic interactions. This section will examine the ways in which the surface charge of drug delivery systems such as NPs, liposomes, dendrimers, and protein carriers have been manipulated to improve tissue targeting and retention within the joint.

Polymeric NPs

Polymeric NPs have been used to address the current challenges of intra-articular drug delivery. Enhanced intra-articular retention has been observed by polymeric NPs ranging from 10 to 900 nm.⁴⁷ They can be engineered to harness optimal targeting of drugs to a specific cell and/or tissue type in the joint. Moreover, NPs can be designed with enhanced drug-loading capacity, allowing improved pharmacokinetics, safe and effective drug delivery, and increased bioavailability of a therapeutic given their ability to penetrate ECM and cell barriers.⁵⁰ Compared with the free drugs, drug-loaded NP carriers present several advantages, including improved delivery of insoluble therapeutics, lower systemic side effects, protection of drug degradation, controlled release of the drug, and promotion of the drug transport across the cell membranes.^51,52

The surface charge of NPs can be manipulated by surface modification with charged species^53–56 or by deposition of polymers to neutralize existing surface charges.^57,58 For instance, a recent investigation varied the deposition of an amine-containing polymer, poly(allyl)amine, on poly(lactic glycolic) acid (PLGA) NPs to manipulate their overall electric charge and demonstrated greater retention of cationic NPs in healthy cartilage tissues ex vivo over their neutral counterparts.⁵⁵ Other investigations have also delved into the formulation of charge-reversal NPs that, in response to a specific stimuli, release their drug payload. These stimuli can be either endogenous factors (changes in pH, redox gradients, or enzyme concentration) or exogenous factors (light or thermo-stimulation).⁵⁹ In this manner, the surface charge of polymeric NPs has been manipulated with the objective of increasing interaction with target tissues.

Liposomes

Liposomes are mainly composed of phospholipids that are amphiphilic molecules that have a hydrophilic head and hydrophobic chains. In a unilamellar vesicle, there is just one lipid bilayer and they enclose an aqueous core, hence making them ideal for encapsulating hydrophilic drugs, whereas multilamellar vesicles have multiple lipid bilayers and preferentially entrap lipophilic drugs within the lipid bilayer. Liposomes provide sustained release through dissolution of the liposome and slow solubilization of the drug.⁶⁰ Liposomes have improved the stability and prolonged the biological half-life of drugs when delivering therapies to inflamed joints.^4,43,61 However, they have proven to be sensitive to mechanical loading, which can be problematic in orthopedic applications.⁶²

Liposomes are attractive drug delivery systems for encapsulation of hydrophobic corticosteroids.^63,64 For instance, a liposomal corticosteroid formulation containing dexamethasone-21-palmitate, Lipotalon^®, is available in Germany as the only intra-articular liposomal product for treating arthritis clinically.⁶⁵ Nanocourt^®, a long-circulating liposomal formulation encapsulating prednisolone, has demonstrated safety and a faster, more pronounced decrease in RA symptoms compared with reference medication (pulse intramuscular methylprednisolone) in Phase 2 trials.⁶⁶

In other cases, liposomal formulations, which have been investigated for treatments in other diseases, have proven to increase the bioactivity of the encapsulated drugs that would be relevant to arthritis applications. For instance, although peripheral vascular disease is the main indication for Liprostin^®, this therapy has proven to improve the therapeutic index of prostaglandin E-1, a suppressor of acute and chronic inflammation that is commonly used for treatment of arthritis in preclinical investigations.⁶⁷ This formulation demonstrated suppression of the histopathological joint changes in RA preclinical models.⁶⁷ The surface charge of liposomes can be manipulated based on the ratios of lipids and polymers used to create them to modulate their interactions with target tissues.

Dendrimers

Dendrimers are globular, monodisperse macromolecules in which all bonds emerge radially from a central focal point or core with a regular branching pattern and with repeat units that each form a branch point. The branched nature of dendrimers leads these molecules to have a large surface area to size ratio, which, in turn, enables greater interactions with cells and tissues.⁶⁸

By manipulating the types of polymers conjugated to the branch points, dendrimers can have tunable surface charges.⁶⁹ In a recent investigation, amine terminal polyamidoamine dendrimers were end functionalized with variable molar ratios of poly(ethylene glycol) (PEG) to control surface charge and vary uptake into OA cartilage ex vivo.²⁰ After intra-articular injection, the cationic dendrimer enhanced the joint residence time of the therapeutic being tested, insulin growth factor-1 (IGF-1), by 10-fold compared with free IGF-1 and prevented cartilage degeneration and osteophyte formation.²⁰ In this manner, the tunable surface charges of dendrimers can be utilized to increase intra-articular retention of therapies, possibly leading to better therapeutic outcomes.

Proteins and peptide-based carriers

Proteins such as albumin can be used as carriers for therapies to prolong their circulation time.⁷⁰ There are several factors that influence the electric charge of a protein: First, the net charge of the protein is affected by the pH of its surrounding environment and can become more positively or negatively charged due to the gain or loss of protons. Therefore, proteins are often associated with an isoelectric point, the pH at which it carries no net electrical charge. This is especially important in biological environments such as the joint capsule wherein pH of the synovial fluid has been demonstrated to decrease in inflammatory joints.⁷¹ Second, the amino acids that compose these proteins all contain the same backbone of an acidic and basic group but each amino acid also contains a unique functional group. Alongside the backbone, this group renders the overall net charge to the protein. These unique amino acids can also render specific tertiary structures to the overall peptide, contributing to the overall charge.⁷²

Amino acids in cell-penetrating peptides have been manipulated to enhance electrostatic interactions between them and cell membranes in cancer applications to enable better tumor cell selectivity for therapies.^73–76 Similarly within the joint, the proportion of lysine or arginine amino acids has been manipulated to vary charges of cationic peptide carriers and, consequently, vary interactions with anionic cartilage.⁷⁷ Moreover, highly positive charged glycoproteins such as Avidin⁷⁸ have been used to carry macromolecules into anionic cartilage tissues to a larger extent than less positive or neutral counterparts. In this manner, proteins or peptide-based carriers can be used to leverage electrostatic interactions with joint tissues and increase retention.

Influence of Charge on Various Delivery Parameters for the Joint

The surface charge of drug delivery systems can influence various parameters that are critical to drug delivery, including the formation of a protein corona around the delivery system, the uptake and cytocompatibility of the system with cells, and their biodistribution within the body when delivered systemically or within the joint when delivered intra-articularly. Mechanistic studies of the influence of electric charge on these various delivery parameters have been conducted in disease applications outside of arthritis, and the lessons learned from these can be extrapolated to drug delivery within the joint.

Influence of charge on protein corona formation

Most drug delivery systems with a net surface charge undergo the phenomenon of protein corona formation on contact with a biological environment, which can influence the fate of the system (Fig. 2). In physiological conditions, a protein corona is often formed on the surface of charged drug delivery systems. This occurs because biomolecules adsorb onto them, which lowers the surface energy by entropy-driven water molecule displacement and blocking hydrophobic parts.⁷⁹ Adsorption of proteins is fostered by several forces such as hydrogen bonds, Van der Waals interactions, as well as hydrophobic and electrostatics interactions.⁸⁰

FIG. 2.

Factors that influence protein corona formation and how protein corona affects the fate of carrier.

The formation of a protein corona is a dynamic process and corresponds strongly to the competitive binding of biomolecules at the surface of the drug delivery system. Plasma proteins compete for the occupation of the surface of the drug delivery systems, resulting in a sequential competitive adsorption, commonly known as the Vroman effect. This effect depends on numerous factors such as the time of exposure, plasma dilution, the pH and the temperature of the environment, the surface charge, and the specific surface chemistry of the delivery system.⁸¹

Although it is critical to investigate the formation and biological importance of the protein corona in the same environment within which the carrier is intended for,⁸² this can pose challenges in the setup of controlled in vitro investigations. For instance, therapeutics intended for the joint are often tested in serum and instead testing in synovial fluid would be ideal wherein the changes in the types of proteins present in the fluid and the concentration of macromolecules such as hyaluronic acid with arthritis would need to be simulated to capture disease condition.^83,84

Besides hydrophobic and hydrophilic interactions, the surface charge on these drug delivery systems has been shown to play a significant role in influencing the types of proteins that would adsorb and form the protein corona.⁸⁵ For instance, albumin is one of the most abundant proteins in serum and is associated with a slightly net negative charge.⁸⁶ As a result, albumin commonly interacts with positively charged carriers.^87,88 This could result in a decreased surface charge on these carriers and hence lower the ability for these to target anionic cartilage tissues if delivered systemically. On the other hand, albumin association with carriers has been shown to reduce clearance mechanisms, evade immune mechanisms, and improve system bioavailability.^89,90

The surface charge of the carrier can strongly influence the secondary structures of adsorbed proteins and, in turn, influence factors such as cellular uptake. For instance, similar-sized gold NPs with varying surface charges showed similar adsorption of bovine serum albumin (BSA), whereas cationic NPs exhibited greater cell uptake than their anionic counterparts, possibly suggesting the modification of the internal structure of the BSA by the NP charge.⁹¹ Another study has demonstrated that the cellular binding and uptake of anionic NPs, of a broad range of charges, is inhibited by free serum proteins in solution irrespective of the composition of the NPs.⁹²

Groups have investigated the usage of various polymers to minimize the adsorption of proteins onto surfaces of carriers and create stealth-like behavior by preventing opsonin interactions and subsequent phagocyte clearance.⁹³ Zwitterionic polymers, containing either zwitterionic groups or a mixture of anionic and cationic terminal groups, have shown high resistance to nonspecific protein adsorption and reduced complement activation.⁹⁴

On the other hand, polymers can be used to modify the surface of carriers to attract specific proteins. For instance, it has been demonstrated that the stealth-like behavior of carriers functionalized with PEG could be due to adsorption of specific proteins such as clusterin or apolipoprotein A4, not due to avoidance of a protein corona formation altogether.⁹⁵ In this manner, the degree of PEGylation and density of PEG on the surface of these carriers can lead to modifications in the protein corona and accordingly the interactions of the carriers with cells in biological environments.^96–99 Although PEG conjugation has led to various successes with providing stability and increasing circulation time, PEGylation could impact the activity of the active drug by changing its conformation, causing steric hindrance and altering electrostatic binding properties.¹⁰⁰ Further, the possibility of inducing production of anti-PEG IgM antibodies could lead to rapid elimination and enhanced hepatic uptake of a second dose of the PEGylated carrier.¹⁰¹

Within the joint, the composition of the protein corona that forms around injected carriers is heavily influenced by the synovial fluid. Although there are currently no published investigations on the protein corona formed in the joint on carriers, proteomic analysis on the synovial fluid itself and changes that occur with arthritis have been conducted.^84,102 These reveal that complement proteins in synovial fluid samples from inflammatory arthritic patients are relatively higher than normal.⁸⁴ Greater adsorption of these complement proteins to carriers could lead to greater recognition by immune cells in the synovium.

The protein corona that forms around carriers within the joint will impact any passive targeting strategies employed by the carrier. A recent investigation demonstrated that all formulations of PLGA or polystyrene NPs of various surface charges became anionic after a coating in synovial fluid, with cationic NPs undergoing a complete charge reversal.⁵⁶ This, in turn, led to decreased retention of the particles within healthy and OA mimic cartilage explants. In a cellular uptake study, incubation of NPs of different surface charges with synovial fluid changed their internalization by synoviocytes.⁵⁶ These changes are hypothesized to occur through adsorption of synovial fluid components on the particle surfaces to create a protein corona that might mask surface functional groups. They may also influence their trafficking to cells and tissue within the joint and their uptake.⁵⁶ In this manner, further consideration of the presence of synovial fluid as a potential influence on the electric charge of drug delivery systems is necessary when designing these systems for arthritis.

Influence of charge on cell uptake and cytotoxicity

The surface charge of drug delivery systems can influence the extent and mechanism of cellular uptake they undergo and, consequently, their cytotoxicity. For instance, cationic NPs have demonstrated entry into cells by using direct translocation, in addition to endocytic pathways, with a fast and high degree of uptake in a noncytotoxic manner when introduced at a low concentration.¹⁰³ Cells uptake positively charged gold particles,¹⁰⁴ superparamagnetic iron oxide particles,^105,106 hydroxylapatite,⁵³ PEG-polylactide,¹⁰⁷ chitosan,⁵⁷ and polystyrene^108,109 particles to a higher extent than their anionic counterparts. However, at higher concentrations, cationic surface charge for most carriers correlates not only with higher cellular uptake but also with greater cytotoxicity in nonphagocytic cells. This is because they cause more pronounced disruption of plasma-membrane integrity, stronger mitochondrial and lysosomal damage, and a higher number of autophagosomes than anionic carriers.¹¹⁰

There have also been clear distinctions in uptake of charged molecules when comparing nonphagocytic with phagocytic cells. Some studies have demonstrated that cationic NPs cause plasma-membrane disruption to a greater extent and apoptosis in nonphagocytic cells than phagocytic cells.^111,112 On the other hand, anionic NPs are better ingested and more cytotoxic in phagocytic cells.^111,112 Further, the presence of the carrier in serum also appears to reduce uptake in nonphagocytic cells but increases it in phagocytic cells.¹¹² Hence, during the design of drug delivery systems, when considering surface charge, the cellular target (whether phagocytic and nonphagocytic) needs to be considered. It is important to understand the tissue or cell target of interest, whether the phagocytic macrophages in the joint or the nonphagocytic chondrocytes in the cartilage and the ideal electrostatic interactions that need to be leveraged to target that cell type. However, there will certainly be some off-target uptake as well and it will be critical to understand the possible undesirable effects that might occur in those cases and mitigate them if needed.

Once taken up into the cell, intracellular trafficking of the carriers can vary depending on their net charge. The efficient escape from endo-lysosomal compartments has been frequently demonstrated by cationic polyplexes, in gene delivery applications.^113,114 There are several theories hypothesized for this phenomenon: The “proton-sponge” hypothesis claims particle escape from the endo-lysosomal compartment by capturing protons and buffering the endosomal pH, leading to the rupture of the endosome through osmotic swelling. A more recent theory postulates that the direct interaction of the cationic carriers with the endosomal membrane leads to its destabilization and subsequent carrier escape.¹¹⁵ A study has demonstrated that positively charged chitosan-based NPs could escape from the lysosome after being internalized and exhibit perinuclear localization, whereas the negatively and neutrally charged NPs preferred to remain localized with the lysosome.⁵⁷ In arthritic applications, it is important to determine whether the therapy needs to act within the cytosol or even traffic to the nucleus, as in the case of gene therapies for instance, and design its carrier accordingly to enable lysosomal escape.

Influence of charge on systemic biodistribution

Systemic delivery of therapeutics is often utilized in arthritic diseases wherein multiple joints or smaller joints are affected, making intra-articular injections less desirable or less feasible. When delivering therapeutics systemically, the biodistribution of the drug throughout the body is typically examined. There have been discrepancies in the outcomes of studies investigating the biodistribution of drug delivery systems with variable surface charges. It was previously mentioned that anionic particles tend to be preferentially taken up by phagocytic cells over cationic particles.^111,112

On the other hand, several studies have demonstrated that high-surface-charge carriers (whether negative or positive) tend to be recognized by the immune system and hence are cleared quickly. For instance, a study has demonstrated that when injected intravenously, high-surface-charged PEG-oligocholic acid-based micellar NPs, either positive or negative, tended to localize in the liver.¹¹⁶ This was attributed possibly to active phagocytosis by Kupffer cells in the liver, based on the nonspecific uptake of the high-surface-charged NPs by RAW 264.7 macrophages in vitro.¹¹⁶ Another investigation has also demonstrated that gold NPs, with zwitterionic material coating, show prolonged circulation time and enhanced tumor accumulation, whereas solely positively or negatively charged NPs are easily cleared by the reticuloendothelial system (RES) after either intraperitoneal or intravenous injection.¹¹⁷ In vitro studies have confirmed that there is no significant difference in phagocytosis between cationic and anionic surfaces when compared at a zeta potential of the same absolute value.¹¹⁸

In a study that investigated the role of surface charge of liposomes in their biodistribution after intravenous injection into a preclinical model of RA, both strongly anionic and cationic liposomes were easily adsorbed by complement proteins, engulfed by RES, and accumulated in the liver tissue.¹¹⁹ As previously mentioned, liposomes with high charge density were more easily recognized by complement proteins and consequently cleared by the RES in this study.¹¹⁹ In contrast, liposomes with slight positive charges were in circulation for a longer period, likely due to their interactions with serum proteins; whereas liposomes with slight negative charges accumulated in the RA paws to a greater extent.¹¹⁹

In yet another investigation utilizing a preclinical model of RA, a positively charged cystine-dense peptide demonstrated localization to cartilaginous tissues such as joint cartilage and intervertebral disks after systemic delivery (oral gavage).⁷² By using a linearized version of the peptide as a negative control, the investigators demonstrated that the disulfide-bonded tertiary structure and the associated distribution of positive charges of the peptide were necessary for joint accumulation.⁷² Peptide concentrations in the blood, muscle, and liver were lower than that in the knee; whereas the peptide concentration in the kidney was either higher or comparable to that in the knee at all timepoints.⁷² In this manner, there appears to be many variations in the patterns of systemic biodistribution of carriers based on their charges across studies.

Leveraging Charge to Target Specific Knee Joint Tissues

As previously mentioned, the knee joint comprises a variety of tissues, each playing its own individual role during the pathogenesis of the various arthritis arthropathies. Regardless of the type of arthritis, the pathological signals associated with progression transpire among the various joint tissues, including cartilage, synovium, and synovial fluid, resulting in extensive tissue cross-talk and a multifaceted disease progression. Electrostatic interactions between drug delivery systems and the tissue type targeted by the therapeutic have been leveraged in many preclinical investigations to increase appropriate localization of the therapeutic (Table 1 and Fig. 3).

Table 1.

Drug Delivery Systems That Leveraged Electrostatic Interactions to Target Pathological Tissues in Arthritis Models

Drug delivery system	Surface charge	Size	Disease model	Target	Major findings	Ref.
Polymeric NPs
NPs made with both Eudragit RL100 and cationically substituted dextran	Positively charged (exact zeta potential not provided but investigators have previously confirmed the expected positive zeta potential of similar cationic NPs)	100–150 nm (depending on which dextran derivative is used)	Rats (healthy)	Synovial fluid	When the fluorescently labeled free tetrapeptide was injected intra-articularly, the fluorescence levels decreased to 23% of the initial concentration within 2 days. In contrast, 74% of the fluorescent tag incorporated into the NPs via an ester bond remained in the knee 7 days after injection.	27
Positive surface-charge PLGA/Eudragit RL NPs	+21.3 mV	170.1 nm	Mice (healthy)	Synovial fluid	Electrostatic interaction of the cationic NPs with the anionic HA polysaccharides, attributed to the rapid formation of micrometer-sized, ionically cross-linked clusters that localized the NPs on the surface and/or inside of the aggregates. In mice, the NPs were efficiently retained in the knee joint for more than 4 weeks after intra-articular injection.	26
PLGA and PS NPs with different surface functionalization	Before incubation with synovial fluid: PLGA PVA: −17.0 mV PLGA DMAB: +24.6 mV PS COOH: −16.7 to −45.4 mV (varied with size) PS NH2: +25.1 to +44.3 mV (varied with size) After incubation with synovial fluid: PLGA PVA: −16.5 mV PLGA DMAB: −10.9 mV PS COOH: −26.9 to −34.9 mV (varied with size) PS NH2: N/A (aggregated)	Before incubation with synovial fluid: PLGA PVA: 297.1 nm PLGA DMAB: 261.5 nm PS COOH: 58.9–202.2 nm PS NH2: 57.4–123.6 nm After incubation with synovial fluid: PLGA PVA: 292.1 nm PLGA DMAB: 300.6 nm PS COOH: 104.7–294.3 nm PS NH2: N/A (aggregated)	Bovine cartilage explants (healthy and enzymatically digested by collagenase type II)	Cartilage	Without incubation in synovial fluid, PLGA NPs surface-modified with a quaternary ammonium cation (DMAB) had the greatest retention within cartilage explants despite not having the highest surface charge of the formulations. Retention of the PLGA DMAB NP was diminished 2- to 2.9-fold in OA tissue and in the presence of synovial fluid.	56
PLGA NPs modified with poly(allylamine) hydrochloride (PAA), a primary amine-containing polymer	Untargeted: −1.2 mV Passive targeting: 22.7 mV Active targeting: 20.1 mV	Untargeted: 195.3 ± 14.8 nm Passive targeting: 172.4 ± 20.5 nm Active targeting: 180.2 ± 8.0 nm	Bovine cartilage explants (healthy and enzymatically digested by collagenase type II) Rats (Collagenase Model)	Cartilage	Untargeted NPs were released from healthy cartilage into the surrounding media after 24 h. Both passive and active NPs showed significantly improved NP retention relative to untargeted NPs in healthy cartilage. In OA cartilage, untargeted and passive NPs showed similar levels of retention whereas active NPs retention was 3.1-fold higher than the others. In rats, active NPs exhibited an increase in NP association with femoral cartilage from 12% in healthy knees to 21% of signal in OA knees after intra-articular injection. Femoral cartilage localization did not change for the passive NPs and significantly increased for untargeted NPs in OA.	55
Liposomes
Liposomes encapsulating human lactoferrin	Positively charged liposomes (exact zeta potential not provided)	Not provided	Mice (CIA model)	Synovial fluid	Free lactoferrin disappeared rapidly from the injected joint, with two-thirds lost by 2 h and 2% remaining at 24 h. Entrapment in positively charged liposomes enhanced retention time, with close to 50% still present after 6 h and 15% at 24 h.	28
Dendrimers
Amine terminal PAMAM dendrimers end functionalized with variable molar ratios of PEG (Gen 4 or 6) and conjugated to IGF-1	Gen 4: +42 mV (theoretical) Gen 6: +123 mV (theoretical)	Gen 4: 4.5 nm (14 kDa) Gen 6: 6.7 nm (58 kDa)	Bovine cartilage explants (healthy) Rat (ACLT and MMx Model)	Cartilage	The dendrimer-IGF-1 conjugate penetrated bovine cartilage of human thickness within 2 days. After intra-articular injection, the Gen 6 dendrimer enhanced IGF-1 joint residence time in rat OA knees by 10-fold for up to 30 days compared with free IGF-1 (3 days). It also rescued cartilage from degeneration and reduced osteophyte burden in the surgical OA rat model.	20
Proteins and peptide-based carriers
Avidin (highly positive charged glycoprotein)	+20 (theoretical based on structure) +6.2 (theoretical based on Donnan theory and equilibrium uptake data) +7.3 (reported)	7 nm, MW 66 kDa	Bovine cartilage disks (healthy or treated with either chondroitinase-ABC or trypsin)	Cartilage	Positively charged Avidin showed 400 × higher equilibrium uptake compared with its neutral counterpart. Avidin retained in explants 15x longer than neutral counterpart.	78
Avidin (both unlabeled and fluorescently labelled)	Avidin alone: pI 10.5	Avidin alone: MW 66 kDa (diameter ∼7 nm)	Rats (healthy) Bovine cartilage explants (healthy)	Cartilage	In vivo, Avidin stayed within the joint for 7 days after intra-articular injection with the highest concentration found in cartilage. In vitro, there was a positive correlation between tissue sGAG content and Avidin uptake in the cartilage.	19
Fluorescently labeled Avidin	Positively charged (exact zeta potential not provided)	Not provided	Rabbits (healthy)	Cartilage	The half-life of Avidin was 5–6 times shorter in thin rat cartilage as compared with thick medial tibial cartilage of rabbits after intra-articular injection indicating importance of cartilage thickness when evaluating retention.	138
Avidin conjugated to dexamethasone	Avidin alone: pI 10.5, net charge = +20	Avidin alone: 7 nm diameter	Bovine cartilage explants (treated with IL-1α)	Cartilage	A combination of slow and fast release Avidin-dexamethasone conjugates had greater suppression of IL-1-induced GAG loss versus a single dose of soluble dexamethasone.	140
Avidin conjugated to dexamethasone	Avidin alone: Positively charged (exact zeta potential not provided)	Not provided	Rabbit (ACLT model)	Cartilage	Avidin-dexamethasone penetrated the full cartilage thickness and was retained for at least 3 weeks after intra-articular injection. Avidin-dexamethasone suppressed joint swelling and catabolic gene expression, improved histological score of cell infiltration into synovium, and reduced osteophyte formation to a greater extent than free dexamethasone but both did not restore loss of cartilage stiffness.	142
Cartilage penetrating and binding cationic peptide carriers	Net charge between +7 to +20 (estimated using Donnan-Boltzmann)	MW ∼2.5 to 4 kDa	Bovine cartilage disks (healthy and treated with trypsin-EDTA phenol red for varying time lengths)	Cartilage	CPC uptake increased with increasing net charge up to +14 but dropped as charge increased further due to stronger binding interactions and decreased weak and reversible charge interactions, which hindered CPC penetrability through full tissue thickness and uptake. Arginine-rich CPCs bound more strongly with the aggrecan-GAGs compared with the lysine-rich CPCs because of short-range H-bond and hydrophobic interactions that further stabilize electrostatic binding.	77
Multi-arm Avidin conjugated to dexamethasone	Between +6 and +20	<10 nm hydrodynamic diameter	Bovine cartilage (healthy and treated with IL-1α)	Cartilage	A multi-arm construct of Avidin penetrated through healthy and arthritic cartilage with long residence time. The construct had 3 times greater uptake in healthy cartilage than OA mimic. A single low dose of the Avidin-dexamethasone conjugate suppressed GAG loss and cell death with the OA mimic tissue.	144
CDP conjugated to TAA (steroid)	Positively charged (analysis of electrostatic potential showed the surfaces are overwhelmingly positively charged, with only small negatively charged patches)	Not provided (CDPs are ∼40 amino acids long)	Healthy mice Human cartilage (healthy) Rats (CIA model)	Cartilage	After systemic administration (oral gavage), radiolabeled CDPs accumulated and persisted in cartilaginous tissues such as joint cartilage and IVDs in healthy mice. Using a linearized version of the CDP, investigators demonstrated that the distribution of the positive charges and disulfide-bonded tertiary structure of the CDP were necessary for cartilage targeting. Fluorescence labeled CDPs were observed to be diffused through the ECM of ex vivo human cartilage explants and less so close to lacunae of cells. Systemic delivery of a CDP with a labile linker to TAA reduced inflammation and limited systemic steroid exposure in a CIA model.	72
Others
Nanocomplex comprised positively charged PEG–TRAIL and negatively charged HA	−30 to +375 mV (depending on feed ratio of HA to PEG-TRAIL)	100–250 nm (depending on feed ratio of HA to PEG-TRAIL)	Mice (healthy and CIA model)	Synovial Fluid	As part of the nanocomplex, the TRAIL displayed a prolonged delivery in vivo. In a disease model, intra-articularly injected PEG-TRAIL nanocomplex led to greater protection of the joints from the inflammatory responses associated with RA and better clinical scores than the TRAIL complex alone.	25
Tantalum oxide (Ta2O5) NPs as contrast agents for cartilage imaging	By DLS, neutral: −1.14 mV Positive: +7.58 mV Negative: −19.07 mV	Neutral: 6.5 ± 0.5 nm Positive: 3.3 ± 0.7 nm Negative: 5.0 ± 0.3 nm	Murine ex vivo proximal tibial cartilage (healthy) Cadaveric OA defect in proximal MCP joint of a human index finger (healthy) Rats (healthy)	Cartilage	In ex vivo system, uptake of the cationic NP was faster and greater than the neutral and anionic NP. In cadaveric joints, the cationic NP showed penetration into defects of the proximal MCP joint. In vivo, the cationic NP exhibited substantially greater affinity for articular cartilage after intra-articular injection.	137
Cationic gadolinium-based MRI contrast agent Gd(DTPA)Lys2	Overall positive charge at physiological pH (exact value not provided)	Not provided	Bovine cartilage Plugs (healthy)	Cartilage	The technique that utilizes an anionic counterpart (dGEMRIC) demonstrated that the T1 values increased from the superficial to middle to deep zones corresponding to a decrease in concentration of gadopentetic acid. The technique utilizing the cationic agent (dGEMRIC+) revealed the opposite pattern, where the T1 values decreased with depth corresponding to increasing concentrations of the cationic contrast agent.	137
Nanoplex created by electrostatic assembly of IGF-1 with poly (glutamic acid) and poly (arginine)	Positively charged (0 to +140 mV by DLS for all molar ratios explored, charge for final formulation not provided)	16.6 ± 5.4 nm	Bovine cartilage explants (treated with IL-1α) Rat (ACLT and MMx Model)	Cartilage	The nanoplex-delivered IGF-1 and free IGF-1 had a similar distribution profile throughout the cross-sectional area of the bovine cartilage explant. In the rat OA model, after intra-articular injection, the IGF-1 was present in joint for 4 weeks when delivered by nanoplex versus a few days when delivered in free form.	141
PEG conjugated with different compounds (chitosan, methylene blue, or HEMA-Co-TMAP)	Charge density of ∼0.5–4.5 kDa−1	Not provided	Bovine cartilage explants (healthy)	Cartilage	The positive charge density of different compounds correlated with their accumulation in bovine cartilage.	143

ACLT, anterior cruciate ligament transection (OA model); CDP, cystine-dense peptide; CIA, collaged induced arthritis (RA model); COOH, carboxylic acid; CPC, cationic peptide carriers; dGEMRIC, delayed gadolinium-enhanced MRI of cartilage; DLS, dynamic light scattering; DMAB, didodecyldimethylammonium bromide; ECM, extracellular matrix; EDTA, ethylenediaminetetraacetic acid; GAGs, glycosaminoglycans; HA, hyaluronic acid; HEMA-Co-TMAP, hydroxyethylmethacrylate-co-trimethylammoniumpropylacrylamid; IGF-1, insulin-like growth factor-1; IL, interleukin; IVD, intervertebral discs; MCP, metacarpophalangeal; MMx, medial meniscectomy (OA model); MW, molecular weight; NH2, amine; NP, nanoparticle; OA, osteoarthritis; PAMAM, polyamidoamine; PEG, poly(ethylene glycol); pI, isoelectric point, the pH at which the net charge of the solute is neutral; PLGA, poly(lactic glycolic) acid; PS, polystyrene; PVA, poly(vinyl alcohol); RA, rheumatoid arthritis; sGAG, sulfated glycosaminoglycans; TAA, triamcinolone acetonide; TRAIL, tumor necrosis factor-related apoptosis-inducing ligand.

FIG. 3.

Opportunities in the joint to leverage electrostatic charges for drug delivery. Adapted from Brown et al.⁴⁷

Synovium

Although active targeting modalities such as tuftsin peptides,¹²⁰ IL-1Ra,¹²¹ vasoactive intestinal peptides,¹²² αvβ3-targeted RGD,^123,124 folic acid,^125–128 or macrophage-derived microvesicle proteins¹²⁹ have been used in various investigations to target cells in the arthritic synovium, there are minimal instances of electrostatic interactions being leveraged for targeting. However, electric charge has been manipulated in other cases to determine preferential targeting of key immune cells that are part of the arthritic synovium, such as phagocytic macrophages.

As previously mentioned, there has been ample discrepancy around the variations in clearance of cationic versus anionic systems by phagocytic cells in the body. Some studies have shown that anionic NPs are better ingested and more cytotoxic in phagocytic cells than cationic carriers,^111,112 whereas others have shown that the highly charged carriers (whether positive or negative) activate the complement system and become opsonized, leading to recognition and uptake by immune cells such as macrophages.^116,117,119

Multiple investigations in the 1980s reported that negatively charged lipids, such as phosphatidylserine and phosphatidylglycerol, are preferentially recognized by macrophages over neutral counterparts, such as phosphotidylcholine, and consequently they have been incorporated into liposomes for uptake enhancement in cancer applications.^130–133 However, this phenomenon might not occur due to the electrostatic interactions between lipids, such as phosphatidylserine, and macrophages but rather due to biological recognition by receptors specific to this lipid that is present on phagocytic cells.¹³⁴ Therefore, there is certainly some uncertainty about how to best use electrostatic interactions to target cells such as macrophages in the synovium. To date, there have not been any investigations that have used electrostatic interactions as a primary means of targeting the arthritic synovium and there appears to be a need for further study in this area.

Cartilage

The dense network of sulfated proteoglycans gives cartilage a bulk anionic charge, corresponding to a proteoglycan content or fixed charge density of −158 to −182 mM for human articular cartilage.¹³⁵ The utilization of electrostatic interactions to enhance drug transport into cartilage and enable sustained binding of drugs within the tissue's highly negatively charged ECM has been described in at least one comprehensive review.¹³⁶ Briefly, it has been demonstrated that by coupling drugs to positively charged carriers, greater binding with the negatively charged glycosaminoglycans will drive enhanced penetration into the cartilage tissue.^{19,20,72,78,137–144}

However, a higher surface charge does not necessarily correlate with increased uptake of positively charged species. A recent study demonstrated that penetration of a cationic peptide into cartilage increased with increasing net charge up to +14 but dropped as charge increased further due to stronger binding interactions with the ECM components and decreased weak and reversible charge interactions.⁷⁷ This hindered penetrability of the carrier through full tissue thickness.⁷⁷

Another study found that PLGA NPs surface-modified with a quaternary ammonium cation (didodecyldimethylammonium bromide [DMAB]) had the greatest retention within cartilage explants despite not having the highest surface charge of all formulations of NPs tested, including polyvinyl alcohol coated PLGA NPs, amine functionalized polystyrene NPs, and carboxylic acid functionalized polystyrene NPs.⁵⁶ Further, although the amine functionalized polystyrene NPs aggregated with synovial fluid, they did not aggregate visibly with a concentration of anionic hyaluronic acid similar to that in a healthy joint.⁵⁶ This indicates that properties such as surface chemistries and/or hydrophilicity could play a more dominant role than surface charge to influence the NP interaction with cartilage. Therefore, although surface charge is a key physicochemical property that can be leveraged to increase targeting of drug delivery systems to cartilage, it cannot be the sole consideration.

One must also consider that the charge density of cartilage is altered as arthritic disease progresses due to loss of proteoglycans, the highly sulfated and negatively charged molecules that contribute to the charge density of cartilage. In this manner, localization and retention for drug delivery modalities that leverage electrostatic interactions for targeting purposes could be compromised in disease conditions. A recent study demonstrated that the retention of anionic PLGA NPs was significantly greater than that of neutral NPs in healthy ex vivo cartilage explants. However, in an OA model, the retention of the neutral NPs and cationic NPs was comparable.⁵⁵ In this manner, the electrostatic interactions rendered between the cationic NPs and the anionic cartilage become inconsequential in the disease state. Therefore, investigators should consider how disease state can compromise the electrostatic interactions intended for carriers when designing them and testing them in preclinical models.

Synovial fluid

By manipulating the electric charge on various drug delivery platforms, electrostatic interactions with synovial fluid have been leveraged as a method to improve whole joint residence. In such cases, drug delivery systems were designed to ionically cross-link with endogenous and in some cases exogenous, hyaluronic acid in the synovial fluid.^26,27 In both these cases, nanomaterials were decorated with cationic Eudragit to form micrometer-sized gels through ionic interactions with the hyaluronic acid polysaccharides in the intra-articular space after injection. The signal from one such Eudragit NP-hyaluronate system was detectable even after 4 weeks postinjection, whereas free dye reached a near-zero fluorescence signal after 2 weeks in vivo,²⁶ illustrating successful application of tissue targeting as a method to prolong whole joint retention.

Another study has demonstrated that the intentional mixing of negatively charged hyaluronic acid onto a positively charged PEG-derivatized TNF-related apoptosis inducing ligand (TRAIL) nanocomplex led to greater therapeutic impact within an RA mouse model than control nanocomplex in phosphate-buffered saline due to a more sustained release of the therapy.²⁵ There was a substantial reduction of serum inflammatory cytokines and collagen-specific antibodies that are responsible for RA progression with the hyaluronic acid and TRAIL nanocomplex mixture.

However, in advanced arthritis, hyaluronic acid is decreased in the synovial fluid⁸³ and the viscosity and chondroprotective function are also reduced.¹⁴⁵ Therefore, this change in synovial fluid content should be taken into consideration when delivery systems are designed for intentional entrapment in the synovial fluid to sustain the joint residence of therapeutics.

Carriers could also be engineered to control the adsorption of specific synovial fluid components that favor interactions with target cells or tissues. This exploitation of adsorption of specific proteins to carriers to enhance trafficking to target cells has been done in other fields such as the usage of apolipoprotein A-I to cross the blood–brain barrier.¹⁴⁶ Given the substantial proteomic studies conducted on synovial fluid from arthritic joints,^84,102 these data can be used to design carriers for favorable adsorption of specific proteins to enable better targeting.

Conclusions

Arthritis, one of the most common chronic conditions in the United States, is hallmarked by inflammation and degradation of key joint tissues such as the cartilage. As the efficacy of systemic treatments is often hampered by bioavailability and stability issues, direct intra-articular injections of therapies into affected joints are typically conducted to treat arthritis. However, even with intra-articular injections, small molecules are rapidly cleared from the joint within hours via synovial vasculature and larger macromolecules within days via synovial lymphatics. As a result, high doses and/or frequent injections are required to observe therapeutic benefits. Drug delivery systems that encapsulate or are conjugated to therapies are being investigated to enable greater retention in the joint, controlled release within the joint space, and tissue-specific targeting.

Currently, with therapies for arthritis, size is often leveraged by drug delivery systems to increase their joint retention. For instance, nanocarriers are often used to clear more slowly than a free drug, thereby improving drug residence time and biodistribution in the joint. At the same time, nano-scale materials can penetrate ECM and cell barriers, thereby enabling intracellular release. Other disease applications such as cancer and solid tumor targeting have also been harnessing the advantages of this physicochemical property.

However, investigators are only beginning to utilize electrostatic interactions for delivery of therapies to pathological tissues in arthritis with the target tissues of interest in these investigations being cartilage and synovial fluid. Fields such as cancer have been leveraging electrostatic interactions between carriers such as cell-penetrating peptides and tumor cells to increase cellular uptake. Hence, knowledge about design criteria considerations and pitfalls to be avoided can be learned from other fields.

Although progress is being made toward utilizing electric charge to target drug delivery systems to joint tissues in arthritic conditions, the environment into which the systems are being introduced must be considered in the drug delivery design. pH of the biological fluid, the types of proteins present in the environment that could form a corona, and the pathological state of the tissue are just some of the factors that could alter the surface charge of the drug delivery system. In this manner, the charge of the system is certainly coupled to environmental factors and hence should be tested in vitro in relevant conditions before introduction in vivo. In preclinical animal models, the system needs to also be tested in arthritic conditions and not just healthy conditions, wherein biodistribution of the carrier between various joint tissues needs to be determined to study targeted drug delivery.

Nonetheless, drug delivery systems utilizing charge as a means of targeting arthritic tissues will continue being a promising platform, especially when combined with other design parameters such as size and targeting moieties, enabling emerging therapeutics to have increased joint retention and bioavailability and overall improved therapeutic efficacy.

Authorship Confirmation Statement

Both authors have contributed substantially to this article, including drafting and critically revising the article (S.K. and B.S.). All co-authors have reviewed and approved of the article before submission. This article has been submitted solely to this journal and is not published, in press, or submitted elsewhere.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

Research reported in this publication was supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) of the National Institutes of Health (NIH) under award number R01AR071335.