Synthesis and Evaluation of a Well-defined HPMA Copolymer-Dexamethasone Conjugate for Effective Treatment of Rheumatoid Arthritis

Materials

S,S′-Bis(α,α′-dimethyl-α″-acetic acid)-trithiocarbonate (28), N,N-dioctadecyl-N′,N′-bis(2-hydroxyethyl)-1,3-propanediamine (LA) (29), HPMA (30), N-methacryloylaminopropyl fluorescein thiourea (MA-FITC) (31) and N-methacryloylglycylglycine (MA-Gly-Gly-OH) (32) were prepared as described previously. Sephadex LH-20 resin was obtained from GE Healthcare (Piscataway, NJ, USA). Mycobacterium tuberculosis H37Ra (heat-killed, desiccated) was obtained from VWR International (West Chester, PA, USA). Paraffin oil (low viscosity, Bayol F) was obtained from Crescent Chemical Company, Inc. (Islandia, NY, USA). Dexamethasone (Dex) was purchased from Hawkins, Inc. (Minneapolis, MN, USA). All solvents and other reagents if not specified were purchased from Fisher Scientific (Pittsburgh, PA, USA) and used without further purification.

Instruments

¹H and ¹³C NMR spectra were recorded on a 500 MHz NMR spectrometer (Varian, Palo Alto, CA, USA). The mass spectrum analyses were performed with a LC/MS/MS system composed of an ACQUITY Ultra Performance LC (UPLC) system (Waters, Milford, MA, USA) and a mass spectrometer of Sciex 4000 Q TRAP with an ESI source (Applied Biosystems, Toronto, Canada). All chromatographic separations were performed with a C₁₈ column (Waters, 100×2.1 mm, 1.7 μm). The mobile phase consisted of 5% ACN in MeOH (mobile phase A) and 7.5 mM ammonium acetate, pH 6 (mobile phase B). The LC flow consisted of 45% mobile phase A and 55% mobile phase B at a total flow rate of 0.3 mL/min. The weight average molecular weight (M_w) and number average molecular weight (M_n) of copolymers were determined by size exclusion chromatography (SEC) using the ÄKTA FPLC system (GE HealthCare) equipped with UV and RI (KNAUER, Berlin, Germany) detectors. SEC measurements were performed on Superdex 200 column (HR 10/30) with phosphate-buffered saline (pH = 7.3) as the eluent. HPMA homopolymer (PHPMA) samples with narrow polydispersity were used as calibration standards. HPLC analyses were performed on an Agilent 1100 HPLC system (Agilent Technologies, Inc., Santa Clara, CA, USA) with a reverse phase C₁₈ column (Agilent, 4.6×250 mm, 5 μm). Bone mineral density (BMD) was measured with a pDEXA^® Sabre™ X-ray bone densitometer (Norland Medical System, Inc., Fort Atkinson, WI, USA)

Synthesis of N-methacryloylglycylglycyl hydrazide (MA-Gly-Gly-NHNH₂, Scheme 1)

N-Methacryloylglycylglycine (MA-Gly-Gly-OH, 0.8 g, 4 mmol) was suspended in ethanol (15 mL) at 0°C. Small amount of inhibitor (tert-octyl pyrocatechine) was added into the solution to prevent polymerization. N,N′-Dicyclohexylcarbodiimide (DCC, 0.9 g, 4.4 mmol) in ethanol (5 mL) was added into the reaction solution. The reaction mixture was stirred for 2 h at 0°C and 2 h at room temperature and then filtered to remove dicyclohexylurea (DCU) at 0°C. Hydrazine hydrate (0.4 mL) in ethanol (5 mL) was then added into the filtrate and the solution was stirred for 4 h at room temperature. When crystalline material started to precipitate, hexane (25 mL) was added and the final product was filtered and washed with ethanol-hexane (1:1, v/v). Yield: 60 %.

¹H NMR (d₆-DMSO) δ (ppm): 8.94 (s, 1 H), 8.15 (t, J = 5.9 Hz, 1 H), 8.07 (t, J = 5.9 Hz, 1 H), 5.74 (s, 1 H), 5.38, (s, 1 H), 4.20 (s, 2 H), 3.76 (d, J = 5.9 Hz, 2 H), 3.67 (d, J = 5.9 Hz, 2 H), 1.88 (s, 3 H); ¹³C NMR (d₆-DMSO) δ (ppm): 169.5, 168.3, 167.9, 139.6, 120.0, 42.7, 41.1, 18.8.

Synthesis of pH-sensitive, Dex-containing monomer (MA-Gly-Gly-NHN=Dex, Scheme 1)

N-Methacryloylglycylglycyl hydrazide (200 mg, 1 mmol) and Dex (390 mg, 1 mmol) were dissolved in methanol (9 mL). Acetic acid (0.5 mL) was added to the reaction solution as a catalyst. The solution was purged with Argon and stirred for 3 days at room temperature in a sealed ampule. After evaporation of the reaction solvent, the product was purified by flash column chromatography (chloroform/ethanol = 8:1, v/v). Yield: 30 %.

¹H NMR (d₆-DMSO) δ (ppm): 10.84 (s, 0.5 H), 10.49 (s, 0.5 H), 8.19 (t, J = 6.3 Hz, 1 H), 8.10 (t, J = 6.3 Hz, 0.5 H), 7.88 (t, J = 6.3 Hz, 0.5 H), 6.78 (s, 0.5 H), 6.68 (s, 0.5 H), 6.45 (dd, J₁ = 15.6 Hz, J₂ = 10.2 Hz, 1 H), 6.24 (dd, J₁ = 24.9 Hz, J₂ = 10.2 Hz, 1 H), 5.73 (s, 1 H), 5.37 (s, 1 H), 5.12 (br, 1 H), 4.92 (s, 1 H), 4.66 (br, 1 H), 4.48 (dd, J₁ = 19.0 Hz, J₂ = 5.4 Hz, 1 H), 4.09 (m, 3 H), 3.86 (m, 1 H), 3.77 (m, 2 H), 2.93 (m, 1 H), 2.61 (m, 1 H), 2.19 (m, 4 H), 1.87 (s, 3 H), 1.72 (m, 1 H), 1.60 (q, 1 H, 10.6 Hz), 1.41 (s, 3 H), 1.20 (m, 3 H), 0.84 (s, 3 H), 0.77 (d, 3 H, J = 6.8 Hz); ¹³C NMR (d₆-DMSO) δ (ppm): 211.4, 170.6, 169.5 (d, syn and anti conformation), 167.8 (d, syn and anti conformation), 165.6, 157.2 (d, syn and anti conformation), 143.1, 139.7 (d, syn and anti conformation), 139.0, 126.5, 119.9 (d, syn and anti conformation), 100.8 (d, J_CF = 173 Hz), 90.3, 70.1 (d, J_CF = 37 Hz), 66.5, 47.6, 47.3 (d, J_CF = 22 Hz), 43.6, 42.5 (d, syn and anti conformation), 40.6 (d, syn and anti conformation), 36.0, 35.1, 34.0 (d, J_CF = 20 Hz), 32.2, 31.0, 27.5, 24.4 (d, J_CF = 6 Hz), 18.7, 16.8, 15.5. The syn/anti diastereomers were chromatographically separated using LC/MS/MS with chromatography conditions stated in the Instruments section. Mass Spectra (negative ion ESI) for both isomers showed the molecular ion [M-H]^- at 587.3, which confirms their monoisotopic mass of 588.3.

Synthesis of HPMA copolymer-Dex conjugate (P-Dex) via RAFT copolymerization (Scheme 2)

HPMA (444 mg, 3.1 mmol) and MA-Gly-Gly-NHN=Dex (128 mg, 0.22 mmol) were dissolved in methanol/DMF (6:1, v/v), with 2,2′-azobisisobutyronitrile (AIBN, 2.7 mg, 0.016 mmol) as initiator and S,S′-bis(α, α′-dimethyl-α″-acetic acid)-trithiocarbonate as RAFT agent. Trace amount of MA-FITC was also added into the copolymerization solution to afford the final product a fluorescent tag for easy detection in purification. The solution was purged with Argon and polymerized at 50°C for 2 days. The resulting polymer was first purified on a LH-20 column to remove the unreacted low molecular weight compounds, and then dialyzed. The molecular weight cutoff of the dialysis tubing is 25 kDa of globular protein. The polymer solution was lyophilized to obtain the final P-Dex. Yield: 250 mg.

To quantify Dex loading in P-Dex, it was hydrolyzed in 0.1 N HCl (1 mg/mL) overnight. The resulting solution was neutralized and analyzed with HPLC. Mobile phase, acetonitrile/water = 2/3; Detection, UV 240 nm; Flow rate, 1 mL/min; Injection volume: 10 μL. The analyses were performed in triplicate. The mean value and standard deviation were obtained with Microsoft Excel.

In vitro Dex release from P-Dex

P-Dex (2 mg/mL) was dissolved in acetate buffer (0.01 M with 0.15 M NaCl, pH 5.0) or phosphate buffer (0.01 M with 0.15 M NaCl, pH 7.4) and incubated at 37°C. At selected time intervals, the P-Dex solution (0.3 mL) was withdrawn and neutralized for HPLC analysis. The analysis of each sample was performed in triplicate. The mean value and standard deviation were obtained with Microsoft Excel.

Treatment of AIA rats with P-Dex

Male Lewis rats (175 to 200 g) were obtained from Charles River Laboratories, Inc. (Wilmington, MA, USA) and allowed to acclimate for at least 1 week. To induce arthritis, Mycobacterium tuberculosis H37Ra (1 mg) and LA (5 mg) were mixed in paraffin oil (100 μL), sonicated and injected subcutaneously into the base of the tail (33). The progression of the joint inflammation was monitored daily. Special care was given to the rats as the inflammation developed to ensure access to water and food. All animal experiments were performed according to a protocol approved by the University of Nebraska Medical Center Institutional Animal Care and Use Committee and adhered to Principles of Laboratory Animal Care (National Institutes of Health publication 85–23, revised in 1985).

Rats with established arthritis were selected and randomly assigned into three groups, P-Dex, free Dex, and saline, with 6-7 AIA rats per group. Six healthy, untreated rats were also included as a control group. On the 14^th day post arthritis induction, P-Dex (100 mg/kg, [Dex]_P-Dex = 100 mg/g of P-Dex) was given intravenously to one group of AIA rats. An equivalent dose of free Dex (10 mg/kg of free Dex in the form of water-soluble dexamethasone sodium phosphate) was divided into four aliquots (2.5 mg/kg of free Dex) and was administered (i.p.) to the second group of AIA rats on days 14, 15, 16, and 17. Saline was given similarly to the third group of AIA rats. The clinical measurements of ankle joints were performed daily. On day 24, all animals were euthanized. The hind limbs were dissected at the knee joint. The BMDs (right leg) of the region from distal tibia to the phalanges of the foot were measured by peripheral dual energy x-ray absorptiometry (pDEXA).

Clinical measurements

The clinical parameters measured during the treatment included articular index (AI) score and ankle diameter. AI scores were taken for each hind paw by the same observer from day 8 to 24 and the sum of the score from each animal was recorded. The AI scoring for arthritis was performed using a 0-4 scale, where 0 = no signs of swelling or erythema, 1 = slight swelling and/or erythema, 2 = low-to-moderate edema and signs involving the tarsals (proximal part of the hind paw), 3 = pronounced edema with limited use of the joint and signs extending to the metatarsals, 4 = excessive edema with joint rigidity and severe signs involving the entire hind paw. Clipper measurements of ankle diameter (medial to lateral) were taken every day from day 8 to 24 using a digital caliper (World Prescision Instruments, Inc. Saraspta, FL, USA).

Histological analysis

The hind limbs were isolated and fixed with buffered formalin (10 %) for 3 days and decalcified with 12.5% (v/v) HCl for at least 2 more days. The decalcification solution was replaced with fresh solution every 24 hours. When decalcification was complete, the ankle joint was transected along the longitudinal plane to give approximately equal halves. Each half joint was then embedded in paraffin. Sections (5 μm thickness) from each ankle joint were cut approximately 200 μm apart and stained using a standard hematoxylin and eosin (H & E) staining method. The joints were histologically scored according to a grading system modified from the literature (34). The histological changes of joints were graded on the following parameters: synovial cell lining hyperplasia (0-2); villous hyperplasia (0-3); mononuclear cell infiltration (0-3); polymorphnuclear leukocytes infiltration in periarticular soft tissue (0-3); cellular infiltration and bone erosion at the distal tibia (0-2) and cellular infiltration of cartilage (0-1) (Table 1). The histology score was then recorded for each ankle joint and summarized for each animal by two independent examiners (JT and KF), who were blinded to the treatment the animal received.

Statistical methods

One-way analysis of variance (ANOVA) was performed followed by a post hoc test (Tukey-Kramer) for multiple comparisons using Instant Biostatistics (Version 3.0, GraphPad Software, La Jolla, CA, USA). A value of P < 0.05 was considered statistically significant.

Syntheses of MA-Gly-Gly-NHN=Dex and P-Dex

As the first step in synthesis of MA-Gly-Gly-NHN=Dex, MA-Gly-Gly-NHNH₂ must be obtained. MA-Gly-Gly-OH was reacted with ethanol to form an ethyl ester. It was then hydrazinolyzed to obtain MA-Gly-Gly-NHNH₂ with about 60 % of yield. This route was proved to be very simple with easy purification workup. To obtain the final hydrazone-containing Dex monomer, Dex was initially reacted with MA-Gly-Gly-NHNH₂ in DMF using acetic acid or HCl as a catalyst. The reaction was not successful. Switching solvent to methanol greatly improved the efficency of reaction. After flash column chromatography, the final product was obtained with 30 % of yield. The structures of MA-Gly-Gly-NHN=Dex (syn/anti diastereomers, ratio ≈ 1:1, Scheme 3) were first revealed by ¹H and ¹³C NMR spectra. They were then confirmed with LC/MS/MS analysis. Two peaks were found in the LC spectrum. They both yielded a strong molecular ion [M-H]^- with a mass of 587.3 in the negative ionization mode, which agrees with the monoisotopic mass of 588.3 for MA-Gly-Gly-NHN=Dex. Using MS/MS analysis, the fragmentation patterns of the diastereomers were studied. The most abundant fragments were at 557.1 due to the loss of a hydroxymethyl group and at 180.0 due to the breakage of the bond between MA-Gly-Gly-NH and N=Dex with additional loss of a H₂O molecule. No attempt was made to separate the diastereomers. Using AIBN as the initiator and S,S′-bis(α, α′-dimethyl-α″-acetic acid)-trithiocarbonate as the RAFT agent, MA-Gly-Gly-NHN=Dex was copolymerized with HPMA. After purification with LH-20 column and dialysis, the FPLC analysis results showed that the weight average molecular weight (M_w) of P-Dex is 34 kDa with a polydispersity index (PDI) of 1.34, which is narrower than a regular free radical polymerization of HPMA (PDI = 1.5-1.6). The Dex content in the P-Dex was determined as 100 mg/g of P-Dex. Conversion of MA-Gly-Gly-NHN=Dex in the copolymerization is about 50 %. When the feed-in ratio of Dex monomer changed, the convention ratio was kept at about 50 %. The HPLC analysis showed that 99.2 % of Dex content was covalently conjugated to the polymer. The 0.8 % free Dex detected in the purified P-Dex may come from the dialysis step. This conjugate was used for in vitro Dex release study and in vivo treatment study of the AIA rats.

In vitro Dex release from P-Dex

The in vitro Dex release was studied by incubating P-Dex in pH 5.0 and pH 7.4 at 37°C (Figure 1). Under neutral condition, there is almost no Dex release from P-Dex; while at pH 5.0, Dex was gradually released from the conjugate. During the course of the experiment (14 days), P-Dex demonstrated a zero-order release at pH 5.0. The release rate of Dex from P-Dex is about 1% of the loaded drug per day.

Clinical evaluations of AIA rats treated with P-Dex, free Dex and saline

The therapeutic effect of P-Dex was evaluated on AIA rats in comparison with free Dex, saline and healthy controls. On day 8 to 9 post arthritis induction, the disease started with the onset of mild swelling and increased ankle diameter (Figures 2 & 3). The ankle joint inflammation continues to worsen, and by day 13 to 14, the swelling started to plateau. All treatments were initiated on day 14. The response to treatments by both free Dex and P-Dex were immediate and very significant. As shown in Figures 2 & 3, ankle size and AI scores for these two groups were greatly reduced on day 15. The animals were more mobile and active. Upon cessation of the free Dex treatment on day 18, inflammatory flare was observed, with a rapid decrease in mobility. By the end of the study on day 24, the animals that had been treated with free Dex were indistinguishable from those in the saline group in terms of ankle diameter and AI scores. Contrary to the free Dex treatment, single P-Dex treatment strongly suppressed the ankle joint inflammation during the entire course of the treatment. It was maintained at the end of the study (euthanasia on day 24).

BMD assessment

The measurement of BMD from distal tibia to the phalanges of the foot of all animals was performed at the end of the study (Figure 4). One-way analysis of variance (ANOVA) produced a P value of less than 0.01, considered significant. The saline group was found to have the lowest mean BMD values of 0.156, followed by free Dex group (BMD = 0.165) and P-Dex group (BMD = 0.175). The healthy control group showed the highest mean BMD value of 0.178 among the four groups. From the pDEXA images, we found that the major bone destruction of arthritic joints happened at distal tibia, with minor BMD reduction at carpal and metacarpal bones.

Histological evaluation of ankle joints

As shown in Figure 5, the saline group was found with the highest histological score of 12.8, indicating severe inflammation and joint destruction. It was followed by free Dex group with a slightly lower score of 9.3. P-Dex treatment, on the other hand, shows the lowest histology score of 1.3, which is very close to the score of 0 of the healthy reference group. One-way analysis of variance (ANOVA) produced a P value of less than 0.0001, considered significant. Figure 6 composes of photomicrographs of representative H & E sections from each group. All animals from the saline treated group showed the highest scores in each parameter of histological grading. Severe bone destruction of the distal tibia and cartilage erosion of the joint are present. In the free Dex treatment group, synovial cell lining and villous hyperplasia are moderate in most of the cases. Polymorphonuclear leukocyte infiltration into periarticular soft tissue fluctuated between mild and moderate levels. Bone and cartilage destruction was present in all cases. Compared to these two groups of animals, P-Dex treatment demonstrates the best anti-inflammatory effect with profound bone and cartilage protection. Polymorphonuclear leukocyte infiltration into periarticular soft tissue, cellular infiltration of cartilage, cellular infiltration and bone destruction at the distal tibia are absent in most cases, which is similar to the healthy control group. Thin rim of synovial cell lining with mildly increased cellular activity was observed, suggesting the presence of minor synovitis.