Intra/Extracellular Multi-Drug Delivery for Osteoarthritis

Osteoarthritis (OA), the most common form of arthritis, affects hundreds of millions of people worldwide and tens of millions of people within the United States. This disease is typically diagnosed only after extensive and irreparable damage to the joints. There are currently no clinically effective disease modifying drugs that can slow or stop disease progression. While cartilage degeneration is the hallmark of OA, there is increasing recognition that OA is a disease of the whole joint, and multiple joint tissues contribute to disease progression. Due to the complex pathogenic processes that involve interactions between different joint tissues, a successful disease modifying therapy will likely require treatment with multiple drugs, each having a different target. While several disease modifying drug candidates have shown promise in disease models both in vitro and in vivo, delivering these drugs effectively and with minimal side effects remains challenging. Cartilage does not have a blood supply, which decreases the efficacy of systemic drug administration methods. In intra-articular injections, drugs are directly injected into the affected joints. But any drugs injected into the joint are rapidly cleared out by the joint capsule over the span of a few hours to a day. As a result, there is limited drug penetration into cartilage. Frequent injections with high drug doses can overcome this challenge, but such a treatment would lead to undesirable systemic side effects and increase the risk of infections within the joint. Recent prior work in our lab has demonstrated that positively charged drug delivery carriers can bind to negatively charged extracellular matrix components in cartilage and thereby improve drug uptake and retention. In this thesis, we characterized the effect of varying the charge of drug delivery carriers on their uptake and penetration into human and bovine cartilage tissues and cells. We identified optimally charged carriers that can be used to deliver drugs to extracellular or intracellular targets. We successfully used these carriers to provide sustained and targeted delivery of growth factors to full-thickness human cartilage explants in vitro, and developed a mathematical model that predicts in vivo transport behavior in human knee joints. We further established an in vitro cartilage-synovium co-culture model that captures physiologically relevant tissue interactions that contribute to OA progression. We also tested drugs targeting inflammatory pathways in this co-culture model, and the results provide a starting point for developing a combination therapy of growth factors and anti-inflammatory drugs conjugated to optimally charged carriers. This thesis is organized as follows: Chapter 1 provides a broad overview of different diseases affecting cartilage, including osteoarthritis. In Chapter 2, we used engineered green fluorescent proteins (GFPs) with a range of net positive charges and surface charge distributions to characterize the effects of charge on carrier transport in cartilage. In both bovine and human cartilage, the uptake of GFPs into cartilage tissue explants decreased with increasing net charge. In contrast, cellular uptake of GFPs increased with increasing charge. Experiments with three neutrally charged GFP variants demonstrated that the surface charge distribution of the carrier also plays an important role in determining its transport properties. Based on the results of this study, we identified optimally charged GFP carriers for delivering drugs to extracellular matrix or cell-surface targets, as well as to intracellular targets. In Chapter 3, we tested the delivery of Insulin-like growth factor 1 (IGF-1), a pro-anabolic drug, using engineered GFP carriers. Since the target for IGF-1 is the extracellular domain of a cell-surface receptor, and a cationic GFP variant with a net charge of +9 (abbreviated as +9 GFP) was found to be optimal for extracellular targets, we designed, expressed and purified fusion proteins of IGF-1 with +9 GFP. Five fusion protein variants had flexible or rigid polypeptide linkers of different sizes connecting the IGF-1 and +9 GFP domains. A sixth fusion protein with no linker was also synthesized. Single doses of two of the fusion proteins had sustained IGF-1 bioactivity in both normal and cytokine-treated human cartilage explants for 7 to 10 days. These fusion proteins increased sGAG biosynthesis rates in normal cartilage and rescued the loss of sGAG biosynthesis in cytokine-treated cartilage, but could not rescue the increase in cumulative sGAG loss caused by cytokine treatments. These responses are consistent with the effects of free IGF1 in human cartilage explant cultures. However, the main difference is that free IGF-1 needs to be continuously replenished to achieve these effects, whereas a single dose at the start of the experiments was sufficient for the fusion proteins. All of the experimental work in this thesis was performed using in vitro cartilage explant cultures. In order to successfully translate these results to preclinical and clinical studies, it is important to predict the transport behavior of GFP carriers and carrier-drug conjugates once they are injected into the joints. In Chapter 4, we developed a mathematical transport model and fit model predictions to data from in vitro dynamic uptake experiments to estimate the transport properties of engineered GFPs and GFP-IGF-1 fusion proteins in cartilage. This model was then used to predict the concentration of GFPs and GFP-IGF fusion proteins in synovial fluid and inside human knee cartilage in an intact knee joint as a function of time after intra-articular injection. The model predicted that significant amounts of the injected molecules could quickly penetrate cartilage tissue before being cleared out by the joint capsule. These cartilage concentration predictions will enable the estimation of injection doses that can achieve appropriate drug doses within cartilage. Additionally, the model also predicts the amount of GFPs and GFP-IGF fusion proteins that are cleared out into the systemic circulation by the joint capsule. These predictions will be useful in assessing the likelihood of potential systemic side effects and estimating safe injection doses. In Chapter 5, we established an in vitro model of post-traumatic osteoarthritis (PTOA) in which cartilage and synovium explants were cultured together. In these experiments, the injury caused by cutting synovium explants triggered the release of large quantities of cytokines. In both human and bovine co-cultures, the cytokine levels were comparable to those observed clinically in the days and weeks following a traumatic joint injury. Cytokine and chemokine levels in co-culture decreased with time, which is also similar to clinical observations. Co-culture with synovium led to significant decreases in the sGAG biosynthesis rate, explant sGAG content, explant metabolic rate and cell viability in cartilage explants. There were no changes in the sGAG (sulfated glycosaminoglycan) biosynthesis rate or ex-plant sGAG content of human cartilage explants that were co-cultured with synovium over the 2-week duration of the experiments. In chapter 6, we tested toll-like receptor (TLR) inhibitors and MAPK pathway inhibitors in the human cartilage-synovium co-culture system. In experiments with tissues from two human donors, treatment with these inhibitors decreased the levels of cytokines and chemokines released by synovium. Future directions could include further characterization with multiple donor tissues (in order to account for donor-to-donor variability), and conjugating the inhibitors to optimally charged GFPs for combination therapy with GFP-IGF fusion proteins. Copies available exclusively from MIT Libraries, libraries.mit.edu/docs|docs@mit.edu