FocusDrug Delivery

Catering to chondrocytes

See allHide authors and affiliations

Science Translational Medicine  28 Nov 2018:
Vol. 10, Issue 469, eaav7043
DOI: 10.1126/scitranslmed.aav7043


An innovative strategy for delivering drugs to chondrocytes in situ offers new avenues for treating osteoarthritis (Geiger et al., this issue).

In 1743, William Hunter complained that “ulcerated cartilage is universally allowed to be a very troublesome disease…that, when destroyed, it is never recovered” (1). Fast forward 275 years and nothing much has changed: Cartilage still does not regenerate and osteoarthritis (OA), the main cause of cartilage degeneration, is rampant. In this issue of Science Translational Medicine, Geiger et al. (2) describe a novel technology that offers considerable promise for reversing this miserable state of affairs.


OA is a disease in desperate need of better therapy. It is incurable, difficult to treat, and a source of much suffering for the estimated 26 million people in the United States and 200 million people worldwide who are afflicted. OA is already the leading cause of disability among the elderly, and its prevalence is rising—not only because populations are aging and body mass index is increasing but also because of unidentified environmental factors. Pain is a predominant symptom, and the lack of adequate analgesia for OA pain contributes to the opioid epidemic. OA also involves the progressive erosion of the structures of the joint, which occurs as a largely independent pathophysiological process. A drug that slowed the rate of joint destruction would be labeled a DMOAD (disease-modifying osteoarthritis drug); however, there are no U.S. Food and Drug Administration (FDA)–approved DMOADs. In the absence of such drugs, many patients’ OA progresses to the point of requiring joint replacement surgery, which makes a major contribution to the $185 billion estimated annual cost of OA in the United States.

OA is a disease of diarthrodial (synovial) joints whose basic anatomy consists of a synovial membrane–lined space within which the cartilaginous surfaces covering the underlying bone articulate during motion (Fig. 1). Lubrication by synovial fluid allows almost frictionless movement. OA is characterized by erosion of the articular cartilage, a unique tissue: It has no blood supply, lymphatics, or nerves, and the bulk of cartilage comprises extracellular matrix. To early histologists, the sparse population of isolated chondrocytes (cartilage cells) embedded within this abundant matrix appeared to be of little metabolic consequence, reinforcing the conventional view that OA was a condition caused by “wear and tear” of the cartilaginous surfaces as they rubbed against each other during locomotion. This misconception delayed research into pharmacologic approaches to treating OA. Only in the latter half of the 20th century did investigators begin to appreciate that dysregulated chondrocyte physiology played a key role in the initiation and progression of OA. This made the chondrocyte a central target for therapeutic manipulation, a trend that continues to intensify.

Fig. 1 Delivering drugs to chondrocytes in situ by intra-articular injection.

Although drugs can be easily injected into joints with OA, most materials within the joint space are rapidly removed by lymphatic drainage or by diffusion into the subsynovial capillaries. Penetration of the articular cartilage, where the chondrocytes reside, is restrained sterically and electrostatically by the high concentration of anionic (−) glycosaminoglycans (GAGs). Geiger et al. (2) developed polyamidoamine (PAMAM) dendrimer nanoparticles to penetrate the extracellular matrix of cartilage. These small particles have a tunable positive (+) surface charge, enabling them to bind reversibly to the anionic GAG chains, unlike larger, negatively charged particles. The kinetics of binding facilitate dendrimer transport through the cartilage to the chondrocytes, where they can deliver a therapeutic payload.


Drug delivery to chondrocytes is complicated by the lack of blood supply to cartilage. Many large molecules of potential therapeutic interest, particularly biologics, are restrained in their ability to diffuse from synovial capillaries (the closest vasculature) into the joint space. Because of this, intra-articular (into the joint) injection has become a popular means of drug delivery (3). This approach makes sense for treating a disease that affects a relatively small number of joints and has few, if any, important systemic sequelae. Intra-articular delivery, however, does not solve the additional problem that, once inside the joint, materials are rapidly removed by lymphatic drainage or diffusion into synovial capillaries. Moreover, even if the drug lingers inside the joint space, it still has to penetrate the cartilage; this is where the difficulty really begins.


Chondrocytes are sparsely distributed within a dense extracellular matrix comprising tightly packed proteoglycan molecules restrained within a collagenous mesh. Because the glycosaminoglycan (GAG) components of the proteoglycans are highly anionic and tightly compressed, cartilage has a high, fixed, negative charge density. Indeed, the potential difference at the interface between the synovial fluid and the surface of cartilage gives rise to a local electric field on the order of 108 V/m (at the nanoscale), which is about 100 times higher than the electric field of a lightning strike (4). This local Donnan interfacial potential drop sets up a tissue-scale equilibrium that moves cations (positive charges) into the cartilage and repels anions (negative charges) as a result of electrostatic interactions. The dense packing of cartilage proteoglycans also sterically hinders entry into cartilage, creating a small effective pore size (about 4 to 6 nm) (5). Shape is also a factor, with linear solutes having greater penetration. Particles larger than about 15 nm are sterically excluded from normal articular cartilage, and particles with a net negative charge are electrostatically excluded; conversely, positively charged particles diffuse more easily, in a charge-dependent manner. With this understanding, Bajpayee et al. (68) have been developing agents that can diffuse through the full thickness of cartilage, remain there, and deliver a therapeutic payload to chondrocytes.

Earlier research using the small cationic protein avidin as a model compound (6) not only confirmed these biophysical principles but also demonstrated a fortunate alignment between the binding properties of matrix GAGs and the requirements of a drug delivery system based on charge. Such a system needs to bind to matrix GAGs with sufficient strength to be drawn into and retained within the cartilage, but with an “off rate” sufficient to allow progress through the matrix, so that all the material does not accumulate at the surface of the cartilage and prevent further entry. Moreover, once inside the cartilage, it needs to engage the target chondrocytes, which will not happen if all of the material is bound tightly to matrix. Small cationic particles like avidin bind weakly to individual GAGs (KD ~ 150 μM), but the high GAG concentration in the cartilaginous matrix (NΤ ~ 3000 μM) forms a very efficient system for capturing the particles and then releasing them to downstream GAGs—and, ultimately, to chondrocytes (Fig. 1) (6). Application of these properties to drug delivery thus ingeniously converts the extracellular matrix of cartilage from a barrier into an accomplice when targeting chondrocytes.

The first demonstration of this principle used avidin bound to dexamethasone, a synthetic steroid injected into the joint to treat OA. Ex vivo data confirmed the efficiency of this type of delivery system and its superiority over free dexamethasone in protecting cartilage; a proof-of-principle study showed beneficial effects in a rabbit model of OA using a single intra-articular injection of an avidin-dexamethasone conjugate (7). However, there is concern that chronic application of intra-articular steroids inhibits chondrocytes’ matrix synthesis in human joints, thereby exacerbating erosion of cartilage in OA. Various growth factors are thus under investigation as alternative payloads; Geiger et al. (2) selected insulin-like growth factor 1 (IGF-1), which enhances matrix synthesis by chondrocytes and protects them from apoptosis.


Combining empirical findings with theoretical considerations, Bajpayee and Grodzinsky (8) and, more recently, Krishnan et al. (9) developed a mathematical model to explain the passage of solutes through the extracellular matrix of cartilage based on solute charge. With this model as a foundation, Geiger et al. (2) described the synthesis of designer nanoparticles based on polyamidoamine (PAMAM) dendrimers as a cartilage-penetrating delivery system. With a size of <10 nm, the nanoparticles are small enough to enter articular cartilage. They have 64 to 256 cationic primary amines that can be modified quantitatively with poly(ethylene glycol) (PEG) oligomers to tune their surface charge. These amines can also be conjugated to various drugs. Of considerable advantage to future clinical translation, the dendrimers are defined chemical entities that can be synthesized in large amounts and thoroughly analyzed by established methods. Moreover, PEG is already a component of marketed pharmaceuticals, and PAMAM dendrimers have been used safely in clinical trials.

In a series of screening experiments, Geiger et al. (2) confirmed that uptake into cartilage explants increased with the surface charge of the dendrimer; unfortunately, toxicity also increased. Selecting two candidates that gave adequate penetration into cartilage without reducing cell viability, they conjugated the dendrimers to IGF-1 without loss of the growth factor’s bioactivity. The conjugate sterically blocks the N terminus of IGF-1 with which inhibitory IGF-1 binding proteins (IGFBPs) interact. Thus, conjugating IGF-1 in this way not only delivers it to chondrocytes but also protects it from IGFBPs.

Increased retention of conjugated IGF-1 was confirmed in vivo by imaging rat knees after intra-articular injection. Free IGF-1 had an intra-articular half-life of just over 10 hours, whereas two different dendrimer constructs extended this to 1.1 and 4.2 days, respectively. The dendrimer responsible for the longer half-life was then evaluated in a particularly aggressive model of OA, which involves subjecting the rat knee to anterior cruciate ligament transection and medial meniscectomy. A single injection of the dendrimer–IGF-1 conjugate protected the articular cartilage from destruction over a 4-week period and reduced synovial inflammation, whereas free IGF-1 was ineffective. However, both free and conjugated IGF-1 reduced the osteophyte (bony growth, also known as bone spur) burden, although statistical significance was not achieved in the case of free IGF-1. There was no evidence of toxicity in kidney, liver, or lung, and blood chemistry remained normal.

These are highly encouraging data. With the possible exception of gene delivery using adeno-associated virus (10), there is no other drug delivery system that can influence the metabolism of chondrocytes in situ throughout the full thickness of articular cartilage in a sustained fashion. As Geiger et al. (2) point out, it is important to investigate dendrimer-based drug delivery in large animals where the thickness of articular cartilage approximates that of human joints. In the human hip and knee, for instance, this can be 1 to 2 mm, whereas the thickness of articular cartilage in rodent joints is one or two orders of magnitude less. These matters are not trivial when considering that diffusion-binding kinetics proceed as the square of cartilage thickness. With this flexible dendrimer platform technology in place, these investigators are in a good position to optimize the system, investigate additional payloads, scale up, perform detailed preclinical studies, and produce good manufacturing practice–grade material as a prelude to clinical trials in human and veterinary medicine.


Funding: C.H.E.’s research is funded by the U.S. Department of Defense (W81XWH-16-1-0540) and a gift from the Musculoskeletal Regeneration Partnership Fund by Mary Sue and Michael Shannon.
View Abstract

Navigate This Article