Towards an arthritis flare-responsive drug delivery system

Abstract

Local delivery of therapeutics for the treatment of inflammatory arthritis (IA) is limited by short intra-articular half-lives. Since IA severity often fluctuates over time, a local drug delivery method that titrates drug release to arthritis activity would represent an attractive paradigm in IA therapy. Here we report the development of a hydrogel platform that exhibits disassembly and drug release controlled by the concentration of enzymes expressed during arthritis flares. In vitro, hydrogel loaded with triamcinolone acetonide (TA) releases drug on-demand upon exposure to enzymes or synovial fluid from patients with rheumatoid arthritis. In arthritic mice, hydrogel loaded with a fluorescent dye demonstrates flare-dependent disassembly measured as loss of fluorescence. Moreover, a single dose of TA-loaded hydrogel but not the equivalent dose of locally injected free TA reduces arthritis activity in the injected paw. Together, our data suggest flare-responsive hydrogel as a promising next-generation drug delivery approach for the treatment of IA.

Fig. 1

Triglycerol monostearate (TG-18) self-assembles to form arthritis flare-responsive hydrogel. a Schematic showing self-assembly of TG-18 to form hydrogel and encapsulation of triamcinolone acetonide (TA). TG-18 dissolves by heating in DMSO/water and self-assembles into lamellar structures upon cooling. Drugs like triamcinolone acetonide (TA) can be encapsulated within the hydrophobic core of the lamellae. b High-resolution scanning electron microscopy (HR-SEM) of TA-loaded TG-18 hydrogels. TG-18 lamellar structures extend to form higher order fibrous assemblies. The entangled fibers form the bulk injectable hydrogel that can be administered via intra-articular injection. c Disassembly of TG-18 hydrogel in response to flare-associated enzymes, including MMPs that are present in the inflamed joint

Fig. 2

TG-18 hydrogel has long-term hydrolytic and encapsulation stability in PBS and exhibits on-demand release of encapsulated TA. a In vitro release kinetics of TA from TG-18 hydrogel in PBS at 37 °C without or with esterase (T. lanuginosus lipase, 200 U/ml). Fresh enzyme was added at the indicated time points (arrows). ***P < 0.001 compared with PBS on day 50. b–d In vitro release kinetics of TA from TG-18 hydrogel in PBS at 37 °C without or with MMP-2 (1.5 µg/ml), MMP-3 (5 µg/ml), or MMP-9 (1 µg/ml). Fresh enzyme or enzyme+MMP inhibitor was added at the indicated time points (arrows). **P < 0.01 compared with PBS or MMP-2 (MMP-9)+inhibitor on day 30 and ****P < 0.0001 compared with PBS or MMP-3+inhibitor on day 25. e In vitro release kinetics of TA in response to synovial fluid from human rheumatoid arthritis joints (SF-RA) and synovial fluid from healthy human joints (SF-healthy). Two hundred microliters of fresh synovial fluid or synovial fluid + MMP inhibitor cocktail were added at the indicated time points (arrows). ****P < 0.0001 and ***P < 0.001 on day 30. P-values were determined by Student’s t-test with Welch’s correction in a and one-way Anova with Tukey’s post hoc analysis in b–e. Data are means ± SD of technical repeats (n = 3, experiments performed at least twice)

Fig. 3

TA-loaded hydrogel is biocompatible with cells from healthy and arthritic human joints. a–d Primary human chondrocytes or synoviocytes from healthy donors or RA patients were incubated in a 96-well transwell plate in medium, PBS, or in medium with 50 µl blank hydrogel (Blank gel), 50 µl TA-loaded TG-18 hydrogel (TA gel), DMSO (equivalent to 50 µl hydrogel) or free drug (Free TA, equivalent to 50 µl TA-loaded TG-18 hydrogel) added to the upper chamber of the transwell. After 24, 48, and 72 h of incubation, metabolic activity was determined by XTT assay. e Representative fluorescence microscopy images after live/dead staining of primary human chondrocytes or synoviocytes from healthy donors or RA patients incubated for 72 h in medium or medium with blank hydrogel (Blank gel) or TA-loaded TG-18 hydrogel (TA gel) added to the upper chamber of the transwell (Scale bar: 400 µm). Viable cells stain green with calcein-AM, whereas dead cells stain red with ethidium homodimer-1. Data in a–d are means ± SD of technical repeats (n = 3, experiment performed at least twice)

Fig. 4

TG-18 hydrogel disassembly correlates with arthritis severity. a DiR-loaded hydrogels were incubated in PBS without or with esterase (T. lanuginosus lipase, 2 or 200 U/ml). To quantify fluorescence signals at each time point, transwell inserts with hydrogel were temporarily removed from the plate and placed on a new plate for imaging using an in vivo imaging system (IVIS). Images of a representative well from each experimental group are shown. b, c Relative fluorescence curves (normalized to day 0) for hydrogels inclubated without or with esterase and their area under the curves (AUCs) (****P < 0.0001). d Experimental outline: Mice were injected with fluorescent dye-loaded hydrogel (4 µl) into the right hindpaw (RHP) on day −1. Arthritis was subsequently induced with two i.p. injections of 37.5 or 75 μl KBx/N serum on day 0 and day 2. Control animals received no serum. Every other day, animals were imaged using IVIS, arthritis severity was scored clinically, and paw swelling was measured with calipers. e, f Change in RHP thickness curves and their AUCs (*P < 0.05 and **P < 0.01). g IVIS images of a representative animal from each experimental group. h, i Relative fluorescence (normalized to day 0) measured over the RHP and AUCs (***P < 0.001 and ****P < 0.0001). Data in b, c are means ± SD of technical repeats (n = 6, experiment performed twice). P-values were determined using one-way Anova with Tukey’s post hoc analysis. Data in b, c are means ± SD of technical repeats (n = 6, experiment performed at least twice). Data in e, f and h, i are means ± SEM (n = 6 mice/group, experiment performed twice)

Fig. 5

Local delivery of TA encapsulated in TG-18 hydrogel improves therapeutic efficacy compared to free TA in a mouse model of IA. a Experimental outline: Arthritis was induced by i.p. injection of 37.5 µl K/BxN serum on day 0 and day 2. Immediately after the second dose of K/BxN serum, TA-loaded TG-18 hydrogel (TA gel, 20 mg TA/ml, 4 µl), free TA (20 mg TA/ml, 4 µl), or blank TG-18 hydrogel (Blank gel, 4 µl) were injected into the right hindpaw. Every other day, arthritis severity was scored clinically and paw swelling was measured with calipers. b, c RHP clinical score curves and their AUCs (***P < 0.001). d, e Change in RHP thickness curves and their AUCs (**P < 0.01). f, g Total clinical score curves and their AUCs (****P < 0.0001). h, i Change in total paw thickness curves and their AUCs (*P < 0.05). P-values were determined by one-way Anova with Tukey’s post hoc analysis. Data are means ± SEM (n = 18–20 mice per group, data were pooled from three independent experiments)

Fig. 6

Local delivery of TA-loaded TG-18 hydrogel demonstrates therapeutic efficacy in severe arthritis. a Experimental outline: Arthritis was induced by i.p. injection of 75 µl K/BxN serum on day 0 and day 2. Immediately after the second dose of K/BxN serum, animals were randomized into four groups. The right hindpaw of each mouse was injected with hydogel on day 2 and day 6. The (Blank gel→Blank gel) group received blank hydrogel (4 µl) twice, (TA gel→Blank gel) mice received TA-loaded TG-18 hydrogel (20 mg TA/ml, 4 µl) on day 2 and blank hydrogel on day 6, the (TA gel→TA gel) group received two doses of TA-loaded TG-18 hydrogel, and the (Blank gel→TA gel) group received blank hydrogel on day 2 and TA-loaded TG-18 hydrogel on day 6. Every other day, arthritis severity was scored clinically and paw swelling was measured with calipers. b, c RHP clinical score curves and their AUCs and d, e Change in RHP thickness curves and their AUCs for (Blank gel → Blank gel), (TA gel → Blank gel), and (TA gel → TA gel) (**P < 0.01, ***P < 0.001, ****P < 0.0001). f, g RHP clinical score curves and their AUCs and h, i Change in RHP paw thickness curves and their AUCs for (Blank gel → Blank gel) and (Blank gel → TA gel) (*P < 0.05, ****P < 0.0001). For c, e, P-values were determined by one-way Anova with Tukey’s post hoc analysis and for g, i, by Student’s t-test with Welch’s correction. Data are means ± SEM (n = 6 mice per group)