An injectable, electrostatically interacting drug depot for the treatment of rheumatoid arthritis

Accepted Manuscript An injectable, electrostatically interacting drug depot for the treatment of rheumatoid arthritis Ji Hoon Park, Seung Hun Park, Hye Yun Lee, Jin Woo Lee, Bo Keun Lee, Bun Yeoul Lee, Jae Ho Kim, Moon Suk Kim PII:

S0142-9612(17)30701-9

DOI:

10.1016/j.biomaterials.2017.10.055

Reference:

JBMT 18337

To appear in:

Biomaterials

Received Date: 26 August 2017 Revised Date:

26 October 2017

Accepted Date: 26 October 2017

Please cite this article as: Park JH, Park SH, Lee HY, Lee JW, Lee BK, Lee BY, Kim JH, Kim MS, An injectable, electrostatically interacting drug depot for the treatment of rheumatoid arthritis, Biomaterials (2017), doi: 10.1016/j.biomaterials.2017.10.055. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

An injectable, electrostatically interacting drug depot for the treatment of

RI PT

rheumatoid arthritis

Ji Hoon Park, Seung Hun Park, Hye Yun Lee, Jin Woo Lee, Bo Keun Lee, Bun Yeoul Lee, Jae Ho Kim,

M AN U

SC

Moon Suk Kim*

TE D

Department of Molecular Science and Technology, Ajou University, Suwon 443-759, Korea

CORRESPONDING AUTHOR

AC C

EP

E-mail address: [email protected], Tel: 82-31-219-2608, Fax: 82-31-219-3931

ACCEPTED MANUSCRIPT ABSTRACT To the best of our knowledge, no studies have yet examined the electrostatic interaction of polyelectrolytes with electrolyte drugs for the treatment of rheumatoid arthritis (RA). Here, an

RI PT

injectable, electrostatically interacting drug depot is described. We prepared methoxy polyethylene glycol-b-(poly(ε-caprolactone)-ran-poly(L-lactic acid) (MC) diblock copolymers with a carboxylic acid group (MC-C) at the pendant position. MC-C was polyelectrolytes that exhibited negative zeta

SC

potentials. Sulfasalazine [Sul(-)] and minocycline [Min(+)], electrolyte RA drugs, exhibited negative and positive zeta potentials, respectively. The electrolyte RA drugs were loaded into the polyelectrolyte

M AN U

MC-C solution to prepare injectable, electrostatically interacting depot formulations. The formulation with an attractive electrostatic interaction [Min(+)-MC-C] exhibited gradual release of Min(+) from the MC-C depot over an extended period and suppressed the growth of inflammatory RAW 264.7 cells without affecting synovial cells. Mature chondrocytes were observed after H&E and safranin O staining of the cartilage of Min(+)-MC-C intra-articularly injected RA-induced rats. In comparison with other

TE D

formulations, Min(+)-MC-C induced the suppression of the expression of pro-inflammatory proteins TNF-α and IL-1β in the articular knee joint, which resulted in the amelioration of RA. In conclusion, an injectable, electrostatically interacting depot formulation administered through intra-articular injection

AC C

EP

successfully provided almost complete amelioration of RA.

KEYWORDS: electrostatic interaction, polyelectrolytes, depot, rheumatoid arthritis, intra-articular injection

ACCEPTED MANUSCRIPT

Introduction Injectable hydrogels have been widely utilized as a drug depot in drug delivery systems because their

RI PT

handling at room temperature is straightforward, they can be injected as a solution, and they easily form a drug depot; additionally, some have good structural integrity in vivo as a drug depot [1,2].

Recently, we prepared [(methoxy)polyethylene glycol (MPEG)-b-(poly(ε-caprolactone) (PCL)-ran-

SC

poly(L-lactic acid) (PLLA)] (MC) diblock copolymers derivatized with a carboxylic acid group (MC-C) at the pendant position on the polyester chains and examined their potential as injectable hydrogel

M AN U

candidates [3,4]. The main driving force for gelation of MC diblock copolymers was considered as the hydrophobic interaction among hydrophobic polyester segments. Additionally, the functional anionic (carboxylic acid) group on the MC diblock copolymers intra- and/or inter-molecularly stabilize or destabilize electrostatic interaction between anionic groups. The results showed that MC-C could act polyelectrolyte carriers for injectable hydrogels.

TE D

Polyelectrolytes are polymers with ionic charged units [5]. The mixing of cationic and anionic polyelectrolytes can create reversible attractive electrostatic interactions. However, polyelectrolytes with the same charge induce repulsive electrostatic interactions. The extent of the attractive or repulsive

AC C

[6].

EP

electrostatic interactions is directly proportional to the magnitude of each charge in the polyelectrolytes

There are several types of drugs that possess neutral, positive, or negative properties and have commercially available grades for the treatment of several diseases [7]. If drugs possess positive or negative charges, they can act as electrolytes. Thus, the drugs with positive or negative charges can induce electrostatic interactions with the polyelectrolytes [8-12]. If the polyelectrolyte can form a drug depot, a mixture of the polyelectrolyte and electrolyte drug can be prepared as an injectable formulation in a homogeneous solution and the formulation can be injected in vivo. Therefore, the cationic or anionic electrolyte drugs inside the polyelectrolyte depot are expected

ACCEPTED MANUSCRIPT to generate attractive or repulsive electrostatic interactions with anionic polyelectrolytes, based on charge (Fig. 1a). Although few studies have been devoted to the comprehension of drug depot formation via the

RI PT

hydrophobic interaction among hydrophobic polyester segments as well as electrostatic interaction between polyelectrolytes [13-15], in view of the above, it would be highly interesting to explore the effects of the intra- and inter-molecular hydrophobic aggregations by polyelectrolytes or the attractive or

SC

repulsive electrostatic interaction between polyelectrolytes depots and electrolyte drugs; to the best of our knowledge, this has received little or no attention in the literature.

M AN U

Based on these considerations, we established the first goal of this work to be an examination of the attractive or repulsive electrostatic interactions between polyelectrolytes and electrolyte drugs, and then to examine the formation of injectable, electrostatically interacting drug depots formed by polyelectrolytes and electrolyte drugs.

Drug depots have been developed for various drugs as long-acting forms in injected sites. Thus, to

TE D

elucidate the in vivo disease applicability of the currently prepared injectable electrostatically interacting drug depot formulation, we selected rheumatoid arthritis (RA) as a chronic and almost incurable symmetric disease induced by a complex multifactorial pathogenesis. Although the cause of RA is not

AC C

articular joints [16].

EP

clear, it severely affects articular joints and bone and eventually leads to the progressive destruction of

Several chemical or biological disease modifying anti-rheumatic drugs (DMARDs) are used to reduce pain, decrease inflammation, and prevent the joint damage that results from RA progression [17,18]. Sulfasalazine [Sul(-)], which possesses negative groups in its molecular structure, has modest effectiveness in RA treatment [19]. It is recommended for medical use as a first-line RA treatment. Minocycline [Min(+)] is also used as an orally administered agent to treat RA [19] and is classified as a long-acting drug. Min(+) possess positive groups in its molecular structure.

ACCEPTED MANUSCRIPT However, repeated oral treatment with RA drug alone, including Sul(-) and Min(+), over an extended period may cause significant side effects, such as headache, upset stomach, vomiting, liver problems, and kidney problems [20]. These limitations of Sul(-) and Min(+) alone treatment can be overcome by

RI PT

alternative administration with little side effects. As a potential treatment method to minimize adverse and side effects, the direct intra-articular injection of RA drugs has been attempted because it can maximize the local treatment effects at the

SC

injected articular joint site [21]. Nonetheless, because there is a rapid clearance of RA drug from the injected articular joint, repeated intra-articular injection of RA drugs is required to maintain therapeutic

M AN U

drug concentrations in the articular joint for prolonged periods [22-25]. To maintain the therapeutic RA drug concentrations, we recently reported the creation of a local drug depot in the injected articular joint site as an alternative administration form [26,27]. The local drug depots in the injected articular joint can allow a rationale for the treatment of RA joints through the sustained release of RA drugs for prolonged periods after intra-articular injection.

TE D

Therefore, based on the above, we hypothesized that an injectable, electrostatically interacting depot formulation of the polyelectrolyte, MC-C and the electrolyte drugs, Sul(-) and Min(+), could be applied to produce a drug depot in the injected articular joint site for RA repair. To the best of our knowledge,

EP

there are no previous studies that describe the intra-articular injection of electrostatic interactions

AC C

between polyelectrolytes and electrolyte drugs and the sustained release of electrolyte drugs from the drug depot for the treatment of RA (Fig. 1b). Therefore, the objectives of the current study were: (1) to evaluate whether the prepared injectable electrostatically interacting depot formulation can be injected into the articular joint to produce a drug depot; and (2) to determine whether the electrostatically interacting depot formed in the injected articular joint site produced a significant improvement through the sustained release of RA drugs in RA joints.

ACCEPTED MANUSCRIPT 2. Experimental Section 2.1. Materials Methoxy polyethylene glycol (MPEG, number average molecular weight (Mn) = 750 g/mol), stannous

RI PT

octoate, ε-caprolactone (CL), 4-dimethylaminopyridine (DMAP), triethylamine (TEA), Sul(-), Min(+), nitric acid, fast green solution, and safranin-O (SO) solution were purchased from Sigma-Aldrich (St. Louis, MO, USA). CL was distilled over CaH2 under reduced pressure. Twenty-four-well Transwell

SC

plates were purchased from Corning (Lowell, MA, USA). Bovine type II collagen and complete Freund’s adjuvant were purchased from Chondrex (Redmond, WA, USA). Synovial cells and

M AN U

Leibovitz’s L15 medium were purchased from ATCC (Manassas, VA, USA). RAW 264.7 cells were purchased from the Korean Cell Line Bank (Seoul, Korea). Dulbecco’s modified Eagle’s medium (DMEM), alamarBlue, mouse TNF-α, IL-1β ELISA Kit, and ProLong gold Antifade Reagent with 4′,6diamidino-2-phenylindole (DAPI) were purchased from Life Technologies (Grand Island, NY, USA). Sodium citrate buffer was purchased from Yakuri Pure Chemicals (Kyoto, Japan). Mayer’s hematoxylin

TE D

solution and mounting medium were purchased from Muto Pure Chemicals (Tokyo, Japan). Bovine serum albumin (BSA) was purchased from Millipore (Kankakee, IL, USA). Rabbit anti-rat TNF-α

1

AC C

2.2. Characterization

EP

primary antibody was purchased from Abcam (Cambridge, MA, USA).

H NMR spectra were measured using a Varian Mercury Plus 400 (Varian) with (CD3)2C=O or

CDCl3 in the presence of TMS as an internal standard. The molecular-weight distributions of MC and MC-C were measured at 40 °C using a YL-Clarity GPC system (YL 9170 RI detector) equipped with three columns (Shodex K-802, K-803, and K-804 polystyrene gel columns). For this measurement, CHCl3 was used as the eluent, at a flow rate of 1.0 mL/min, and polystyrene was used for calibration. The melting temperature (Tm) and the heat of fusion (∆Hm) of MC and MC-C solutions (20 wt%) were

ACCEPTED MANUSCRIPT determined using differential scanning calorimetry (DSC; Q1000, TA Instruments, Germany) performed from 10 to 60 °C at a heating rate of 5 °C/min in a nitrogen atmosphere. The heat of fusion per gram of

RI PT

MC and MC-C was calculated according to the area under the peak.

2.3. Synthesis of MPEG-b-(PCL-ran-PLA) diblock copolymer (MC)

MC diblock copolymer (750–2810/150 g/mol) with a PCL/PLA ratio of 95:5 was prepared as

M AN U

and the measured values were: C, 60.9; H, 8.9.

SC

previously reported [28]. The elemental analysis of the calculated values for MC were: C, 60.9; H, 8.7;

2.4. Synthesis of MPEG-b-(PCL-ran-poly(3-benzyloxy-L-lactide)) diblock copolymer (MC-Bz) For polymerization, all glasses were dried by hot heating in a vacuum, flushed, and handled under fresh dry nitrogen. 3-Benzyloxymethyl-6-methyl-1,4-dioxane-2,5-dione (fLA) was prepared by using a previously reported method [3]. To prepare the MC-Bz copolymer with a PCL/PfLA ratio of 95:5,

TE D

MPEG (1.51 g, 2.02 mmol) in toluene (80 mL) was azeotropically distilled to completely remove the water and then distilled to yield a final toluene volume of 50 mL. CL (4.34 g, 38.0 mmol) and fLA (0.50 g, 2.0 mmol) were added to the toluene and then 0.2 mL of a 0.1 M solution of stannous octoate in dried

EP

toluene was added to the MPEG solution under a nitrogen atmosphere at room temperature. The reaction

AC C

solution was polymerized for 24 h at 130 °C and then poured into a mixture of n-hexane and ethyl ether (v/v = 4/1) to precipitate the MC-Bz copolymer. The precipitated MC-Bz copolymer was filtered and dried in a vacuum to yield 5.36 g (1.65 mmol, 91%) of colorless MC-Bz copolymer. The molecular weights of the PCL and PfLA segments in MC-Bz copolymer were determined by 1H-NMR analysis. The ratios of the PCL and PfLA segments were determined through the comparison of the total methylene protons in PCL at δ = 2.3 ppm and the phenyl proton signals of PfLA at δ = 7.2–7.4 ppm with the total methyl protons in MPEG at δ = 3.38 ppm as a standard of 750 g/mol.

ACCEPTED MANUSCRIPT 2.5. Synthesis of MPEG-b-(PCL-ran-poly(3-hydroxyl-L-lactide)) (MC-OH) The MC-Bz copolymer (5 g, 1.55 mmol) was dissolved in anhydrous THF (300 mL). Subsequently, 10% w/w (2 g) of Pd/C (palladium, 10 wt% (dry basis) on activated carbon (50% water w/w, Degussa

RI PT

type E101 NE/W)) was added to the MC-Bz copolymer solution. The suspension was stirred under a hydrogen atmosphere for 12 h. The reaction suspension was filtered through a celite filter to remove the reaction catalyst. The solution was concentrated by rotary evaporation and dried in a vacuum to yield 4.3

SC

g (1.4 mmol, 90%) of colorless MC-OH copolymer with a PCL/PLA-OH ratio of 95:5.

M AN U

2.6. Synthesis of MPEG-b-(PCL-ran-poly(3-carboxyl-L-lactide)) (MC-COOH) To prepare MC-C, MC-OH (2 g, 0.67 mmol) was dissolved in 150 mL dioxane and succinic anhydride (0.38 g, 3.36 mmol) was added to the MC-OH solution. Solutions of DMAP (0.41 g, 3.35 mmol) and TEA (0.30 g, 3.35 mmol) dissolved in 20 mL of dioxane were added into the MC-OH reaction solution. The reaction solution was stirred at room temperature for 24 h under a nitrogen

TE D

atmosphere and then poured into a mixture of n-hexane and ethyl ether (v/v = 1:1) to precipitate the MC-C copolymer. The precipitated copolymer was dried in a vacuum to give a produce the MC-C copolymer with a PCL/PLA-COOH ratio of 95:5 (1.81 g, 0.53 mmol, 87%). The elemental analysis of

EP

the calculated values for MC-C were C, 59.8; H, 8.4; and the measured values were: C, 59.7; H, 8.9.

AC C

Ethylene oxide was used to sterilize MC and MC-C for use in the subsequent RA animal experiments.

2.7. Zeta potential

The zeta potentials of the MC and MC-C copolymers, Sul(-), and Min(+) were measured by dynamic light scattering (DLS, ELSZ-1000; Otsuka Electronics, Osaka, Japan). The zeta potential measurements for drug alone (1, 5 and 10 mg/mL), copolymers alone (10, 50, 100 and 200 mg/mL), and drug-loaded copolymers (drug:copolymer ratio = 5 mg/mL:200 mg/mL) (n = 3) were performed by dissolving the test

ACCEPTED MANUSCRIPT compounds in DW at 80 °C. All zeta potentials were measured three times and scanning was repeated four times.

RI PT

2.8. Determination of sol-gel phase transition times The MC and MC-C copolymers were dissolved at 80 °C in DW to obtain concentrations of 200 mg/mL copolymer hydrogel and 5 mg/mL Sul(-) and Min(+). The time taken to form homogeneous

SC

opaque emulsions of MC and MC-C, and drug-loaded MC and MC-C copolymers was determined and defined as the solubilization time. After 48 h at 4 °C, the homogeneous opaque emulsions in the vials

M AN U

were gently stirred at room temperature and the vials were immersed in a 37 °C water bath. When the vials were maintained at 37 °C, the time taken by the emulsions to stop exhibiting flow was measured and defined as the gelation time.

2.9. Viscosity measurements

TE D

Each block copolymer (0.5 g) was individually dissolved in 5-mL vials with distilled water to make 20-wt% concentrations and stored at 4 °C. The drug-loaded copolymers (final concentration of drug:copolymer ratio = 5 mg/mL:200 mg/mL) were individually dissolved and stored at 4 °C. After 48

EP

h, the viscosity of copolymer solution and drug-loaded copolymer solutions were examined by using a

AC C

rheometer (MCR 102, Anton Paar, Ostfildern, Germany) with a politer temperature-controlled bottom plate and a 25.0-mm parallel plate measuring system. All measurements were conducted with a gap length of 0.3 mm at an oscillating frequency of 1 Hz and 0.1% of the oscillating strain, from 10–60 °C in increments of 4 °C/min. The viscosity (η), storage modulus (G’), loss modulus (G’’), and phase angle (tan δ) were calculated by the instrument's software.

2.10. In vitro drug release

ACCEPTED MANUSCRIPT Drug-loaded copolymers (drug:copolymer ratio = 5 mg/mL:200 mg/mL) were loaded in 5-mL vials and incubated at 37 °C to form drug-loaded copolymer hydrogels. The vial with drug-loaded copolymer hydrogels were immersed in 4 mL PBS and shaken at 100 rpm and 37 °C for up to 40 days. For each

RI PT

experiment, 1 mL of solution was extracted from the vials at predetermined time intervals and 1 mL fresh PBS was immediately added into the vials to restore the volume. The amount of Sul(-) and Min(+) was analyzed by using a high-performance liquid chromatography (HPLC) system (Agilent 1200 series,

SC

Waldbronn, Germany) with detection at 360 nm for Sul(-) and 265 nm for Min(+) on the Hypersil C18 column (4.6 mm × 250 mm, 5 µm). For Sul(-), the mobile phase consisted of methanol and ammonium

M AN U

acetate (10 mM, pH 7.0 with acetic acid) (48:52, v/v); for Min(+), sodium phosphate (100 mM) and acetonitrile (85:15, v/v) were used. The mobile phase was eluted at a flow rate of 1.0 mL/min. Three independent release experiments were performed for each drug-loaded copolymer composition. The amount of drug released from the copolymer was calculated through a comparison with standard

TE D

calibration curves prepared using known concentrations of Sul(-) and Min(+).

2.11. Inflammatory tests using synovial cells and RAW 264.7 cells Synovial cells (3 × 104 cells/well) or RAW 264.7 (1 × 104 cells/well) were seeded in the lower

EP

chambers of a 24-well Transwell plates and cultured in DMEM supplemented with 10% FBS and 1%

AC C

penicillin-streptomycin (PS) for 3 days at 37 °C in a humidified incubator containing 5% CO2. To determine drug concentrations, the cell proliferation of synovial cells or RAW 264.7 in 1 mL culture medium containing 1, 3, 5, 7 and 10 mg/mL of Sul(-) or Min(+) alone (or at 1, 4, 7, 10, 14, 18 and 21 days for 5 mg/mL of Sul(-) or Min(+) alone) was added to well plate and measured after 24 h by using the alamarBlue assay (Life Technologies, Camarillo, CA, USA) (below method). MC alone (200 mg/mL) and MC-C alone (200 mg/mL) individually was added to well plate and incubated at 37 °C to form MC and MC-C hydrogel depot without drug. The synovial cells or RAW 264.7 was individually added on the MC and MC-C hydrogel depot. After 1, 4, 7, 10, 14, 18 and 21 days, cell proliferation of

ACCEPTED MANUSCRIPT synovial and RAW 264.7 cells was determined by using the alamarBlue assay. Sul(-)-MC (5 mg/mL Sul(-) in 200 mg/mL MC), Min(+)-MC (5 mg/mL Min(+) in 200 mg/mL MC), Sul(-)-MC-C (5 mg/mL Sul(-) in 200 mg/mL MC-C), and Min(+)-MC-C (5 mg/mL Min(+) in 200

RI PT

mg/mL MC-C) were individually added to the upper chambers of the Transwell plates and maintained at 37 °C to form each drug-loaded hydrogel depot. The synovial cells or RAW 264.7 was individually added on the bottom chambers of each Transwell plate and maintained at 37 °C in a humidified

SC

incubator for 21 days. The formulations containing only Sul(-) and Min(+) were changed only once after 1 day and the culture medium without Sul(-) and Min(+) was changed every 3 days. For the hydrogel

M AN U

formulations, the culture medium was changed every 3 days. The cell proliferation was determined by using the alamarBlue assay. One milliliter of the mixture of alamarBlue solution (0.1 mL) and DMEM (0.9 mL) was added to the 24-well Transwell plates and incubated for 4 h. Aliquots of the alamarBlue solution from the 24-well Transwell plates were transferred to a 96-well plate. The absorbance of the solution was measured at 570 nm and 600 nm by using an ELISA micro-plate reader (VersaMax,

TE D

Molecular Devices, Sunnyvale, CA, USA). All experiments were individually performed three times and the results were presented as mean ± standard deviation (SD).

EP

2.12. Treatment of RA-induced rats

AC C

The Institutional Animal Experiment Committee of the School of Medicine of Ajou University approved all protocols used in this study (Approval No. 2015-0047). Experiments that involved the treatment and induction of RA were conducted in accordance with the approved guidelines. For the preparation of the RA animal models, bovine type II collagen and complete Freund’s adjuvant were mixed at a ratio of 1:1 for 30 min. The mixed solution was then incubated to remove air-bubbles at room temperature. Then, the mixed solution (250 µL) was injected into the tails of 4-week-old male Lewis rats. After 3 weeks, the RA-induced rats (RA rats) showed the evidence of edema and erythema in the tarsals, ankle joints, and feet.

ACCEPTED MANUSCRIPT The rats were randomly divided into six experimental groups. The injectable formulations of Sul(-) and Min(+) only (5 mg/mL), Sul(-)-MC, Min(+)-MC, Sul(-)-MC-C, and Min(+)-MC-C were individually and intra-articularly injected into the articular knee joint of the rats with RA by using an 26-

RI PT

G needle. The rats were individually housed for experimental periods.

2.13. Articular index score and ankle diameter measurements of RA rats

SC

The hind paws of drug formulation-treated RA rats and normal rats were observed and measured to determine ankle diameter and articular index (AI) score at 1, 2, 3, 4, 5 and 6 weeks. Ankle diameter was

M AN U

measured in three rats by using Vernier calipers. AI was blind tested and scored as follows: no edema or erythema (score 0), slight edema and/or erythema (+ 1), moderate edema and shackles (+ 2), edema with limited use of joints and symptoms of enlargement to the metatarsal (+ 3), and excessive swelling as a serious symptom associated with joint ankyloses and total hind paw (+ 4) [25]. All ankle diameters and

standard deviation (SD).

2.14. Histology

TE D

AI scores were independently measured three times and the results were presented as the mean ±

EP

The drug formulation-treated RA rats and normal rats were individually sacrificed at predefined times

AC C

of 1, 2, 4, and 6 weeks. Then, articular knee joints were extracted from individual rats with RA. The removed articular knee joints were fixed in 10% formalin for 4 days, demineralized for 2 days by 6% nitric acid, and dehydrated in 100% ethanol. The fixed tissues were embedded in paraffin and then cut into 7-µm slices.

For H&E staining, the deparaffinized slides were hydrated using 100%, 90%, 80%, 70%, and 60% of ethyl alcohol in a regular sequence. Then, the slides were washed in running tap water and stained with hematoxylin and eosin for 3 min, respectively. Thereafter, the stained slides were fixed and mounted with mounting medium (Muto Pure Chemicals, Tokyo, Japan).

ACCEPTED MANUSCRIPT For safranin O (SO) staining, the hydrated slides were stained with Mayer’s hematoxylin solution for 5 min and rinsed in DW for 10 min. The slides were then stained with 0.002% fast green solution for 30 s and rinsed quickly with 1% acetic acid solution. Finally, the slides were placed in a 0.1% SO solution

RI PT

for 6 min, dehydrated, and fixed with mounting medium. For TNF-α staining, the hydrated slides were dipped in 10 mM sodium citrate buffer solution, heated to 120–130 °C for 20 min to recover the antigen, washed with PBS, and blocked with 10% BSA in PBS

SC

at 37 °C for 1 h. The slides were subsequently incubated at 4 °C overnight with a 1:100 dilution of rabbit anti-rat TNF-α primary antibody, rinsed with PBS, and incubated with a 1:200 dilution of Alexa

M AN U

Fluor 594-conjugated secondary antibodies for 2 h in the dark at room temperature. Finally, the slides were mounted with ProLong® Gold Antifade Reagent.

The images of all stained slides were observed by using an Axio Imager A1 (Carl Zeiss Microimaging GmbH, Göttingen, Germany) equipped with Axiovision software (Rel. 4.8, Carl Zeiss Microimaging GmbH). The cartilage thickness and TNF-α positive area were calculated using ImageJ

positions for each specimen.

TE D

software (National Institutes of Health, Bethesda, MD, USA) from three different, randomly selected

EP

2.15. TNF-α and IL-1β measurements of RA rats

AC C

To measure the expression of TNF-α and IL-1β in the drug formulation-treated rats and the normal rats, the rats were individually sacrificed at predefined times of 1, 2, 4 and 6 weeks. The articular knee joints were extracted from individual RA rats. Each articular knee joint was homogenized in 50 mM Tris HCl (pH 7.4) with 0.5 mM dithiothreitol and proteinase inhibitor cocktail (10 g/mL) using an IKA 3420000 T 10 basic ULTRA-TURRAX disperser (IKA, China) at 25,000–30,000 rpm for 10 min and incubated at 37 °C for 15 min. The TNF-α and IL-1β concentrations in the articular knee joint protein extracts were determined by a rat TNF-α ELISA Ready-SET-GO! Kit (eBioscience) and a rat IL-1β ELISA Ready-SET-GO! Kit (eBioscience), respectively. A Coated Corning Costar 9018 ELISA plate

ACCEPTED MANUSCRIPT was sealed, incubated overnight at 4 °C, and washed three times with wash buffer (1× PBS, 0.05% Tween 20). To perform the assay, 200 µL 1× ELISA/ELISPOT dilution was added to each well and incubated at room temperature for 1 h. The wells were washed with wash buffer and 100 µL of the knee

RI PT

joint protein extracts was added to the each well and incubated overnight at 4 °C. Each well was incubated for 1 h with antibodies, washed again with wash buffer, and then incubated with the AvidinHRP for 30 min at room temperature. After incubation, each well was washed with wash buffer and

SC

treated with 1× TMB solution for 15 min. Stop solution (50 µL) was added to each well and the absorbance was measured at 450 nm. All experiments were independently repeated three times and the

M AN U

results were presented as the mean ± standard deviation (SD).

2.16. Statistical analysis

Cytotoxicity data in synovial cells and RAW 264.7 cells was obtained from n = 3 independent

TE D

experiments for each data point. AI scores and ankle diameters of rats with RA were obtained from independent experiments with n = 3 at 1–6 weeks. The thickness of cartilage in H&E staining, TNF-α positive area in TNF-α staining and immunosuppressive effect of TNF-α and IL-1β in experimental

EP

arthritis were obtained from n = 3 independent experiments at 1, 2, 4, and 6 weeks. The results were analyzed by using one-way analysis of variance (ANOVA) with Bonferroni’s post-hoc test and SPSS

AC C

12.0 software (SPSS Inc., Chicago, IL, USA).

3. Results

3.1. Preparation and characterization of injectable MC and MC-C diblock copolymers We previously reported that the MC diblock copolymer solution exhibited a sol-gel phase transition close to body temperature [28]. In this work, an MC diblock copolymer with PCL/PLA ratio of 95:5 was prepared with an MPEG chain with a MW of 750 g/mol and PCL/PLA with a MW of 3,000 g/mol.

ACCEPTED MANUSCRIPT Next, MC-Bz diblock copolymers were synthesized via the ring-opening polymerization of the CL and fLA in a 95:5 ratio by using MPEG as an initiator in the presence of stannous octoate. MC-Bz diblock copolymers were obtained with a total MW of 3,000 g/mol. Then, the benzyl pendant position

RI PT

on MC-Bz was removed by a quantitative deprotection reaction to produce MC-OH. The hydroxyl pendant group of MC-OH was reacted with glutaric anhydride to give MC with a carboxylic acid pendant group of 5 mol% (MC-C). The characterization of MC and MC-C were fully confirmed by

SC

elemental analysis and 1H-NMR (see Experimental section and Supplementary Information: Tab. S1,

M AN U

Fig. S1).

3.2. Phase transition of injectable MC and MC-C

The aqueous solutions of MC and MC-C were prepared as homogeneous white opaque emulsions (Supplementary Information: Fig. S2).

In the DSC measurement (Supplementary Information: Tab. S2), the MC solution exhibited a

TE D

melting point of 47 oC and crystallization enthalpy of 1.5 J/g due to intra and inter hydrophobic aggregation between the hydrophobic blocks of the MC blocks. The DSC of the MC-C solution exhibited similar melting point of 48 oC and the increased crystallization enthalpy of 2.9 J/g, indicating

EP

the formation of an intra and inter hydrogen bonding between COOH groups on the MC blocks as well

AC C

as hydrophobic aggregation.

The phase transition and gelation time of MC and MC-C were examined by the visual inspection of gelation at different temperatures between 25 °C and 37 °C. All MC and MC-C solutions flowed at 25 °C, but MC and MC-C did not flow at 37 °C when tilted. The gelation time of MC and MC-C exhibited distinct sol-gel phase transitions of less than 20 s at 37 °C. The rheological measurements of the MC and MC-C copolymer solutions were obtained to examine phase transition behavior at temperatures between 10 °C and 60 °C (Fig. 2a). The MC and MC-C

ACCEPTED MANUSCRIPT diblock copolymer solutions at lower temperature exhibited a viscosity of approximately 1 cP, close to that of water, which indicated a homogeneous solution phase of diblock copolymers. At intermediate temperatures, the viscosities of MC and MC-C solutions increased at 36 and 31 °C, respectively, which

RI PT

was indicative of the onset temperature for gelation. The viscosities of the MC and MC-C solutions at 37 °C were 9 × 103 and 2.7 × 105 cP, respectively. This probably indicated the order of hydrogel strength. The maximum viscosities of MC and MC-C were 4.4 × 105 cP at 47 °C and 5.5 × 105 cP at

M AN U

behavior were affected by the COOH pendant group.

SC

44 °C, respectively. These results indicated that the gelation time, hydrogel strength and phase transition

3.3. Solution properties of drug-loaded MC and MC-C formulations First, zeta potential was measured to examine the electrostatic properties of MC and MC-C copolymer solutions prepared with different concentrations (Fig. 3). The zeta potential of 10, 50, 100 and 200 mg/mL concentrations of MC and MC-C copolymer solutions showed a little difference of zeta

TE D

potentials (Supplementary Information: Tab. S4). The zeta potential of the MC solution (200 mg/mL) was -6 mV, the MC-C solution exhibited a negative zeta potential of -18 mV. These result indicated that the electrostatic properties of the MC-C solution were affected by the presence of COOH pendant

EP

groups on the PLA segments. The zeta potentials of the Sul(-) and Min(+) also showed a little

AC C

concentration dependency (Supplementary Information: Tab. S5). The zeta potentials of the Sul(-) and Min(+) drugs were -9.6 mV and 5.8 mV, respectively. Each MC and MC-C solution was separately mixed with Sul(-) and Min(+). The electrostatic properties of the mixed Sul(-)-MC, Min(+)-MC, Sul(-)-MC-C and Min(+)-MC-C solutions were measured. The net zeta potentials of Sul(-)-MC were -13 mV, and slightly increased to more negative values as a result of the addition of Sul(-), which had a negative zeta potentials, and the original negative zeta potential of MC. Min(+)-MC showed a positive zeta potential of 2 mV owing to the neutralization of the zeta potential between cationic Min (+) and negative MC.

ACCEPTED MANUSCRIPT The net zeta potential of Sul(-)-MC-C was -26 mV and increased to more negative values after the addition of the negative zeta potentials of Sul(-) and the negative zeta potentials of MC-C. Meanwhile, Min(+)-MC-C showed a reduction of the negativity of the zeta potentials (-11 mV) as a result of the

RI PT

high negative potentials of MC-C. These results indicated that the arithmetic calculation of the zeta potentials of polyelectrolytes and electrolyte drugs were in good agreement with the expected electrostatic properties and was suggestive

SC

of the exact electrostatic attraction and repulsion that occurred between polyelectrolytes and electrolyte

M AN U

drugs.

3.4. Phase transition of drug-loaded MC and MC-C formulations

Drug-loaded MC and MC-C formulations was easily obtained by mixing Sul(-) and Min(+) with MC and MC-C solutions, respectively. The mixed suspensions showed good stability. The prepared Sul(-)MC, Min(+)-MC, Sul(-)-MC-C, and Min(+)-MC-C formulations were opaque and emulsions, which

TE D

indicated the complete suspension of the drugs (Supplementary Information: Fig. S2). The gelation time and phase transition of the drug-loaded MC and MC-C formulations were examined. Sul(-)-MC and Min(+)-MC formulations became hydrogels at 37 °C when tilted. MC

EP

solution only, Sul(-)-MC, and Min(+)-MC exhibited similar gelation times of approximately 20 s at

AC C

37 °C, which indicated that the electrolyte drugs induced little or no electrostatic interaction with MC in the gelation.

However, the gelation time of Sul(-)-MC-C and Min(+)-MC-C formulations were approximately 20 s and 5 s at 37 °C respectively, which implied that there was an electrostatic interaction between MC-C and Min(+). The phase transition behavior of the drug-loaded MC and MC-C formulations was measured in the temperature range from 10–60 °C. The drug-loaded MC and MC-C formulations showed similar phase transition behavior from solution at low temperature to hydrogel formation at intermediate temperatures.

ACCEPTED MANUSCRIPT The drug-loaded MC formulations showed almost similar phase transition behavior, except for small differences in maximum viscosities, which implied little or no electrostatic interaction between the electrolyte drugs and MC (Fig. 2b).

RI PT

For the drug-loaded MC-C formulations (Fig. 2c), the viscosity of the Min(+)-MC-C formulation increased to 5.5 × 105 cP at 37 °C and the maximum viscosity was 8.7 × 105 cP at 42 °C, compared with the maximum viscosity of the Min(+)-MC formulation of 2.7 × 105 cP at 37 °C. However, the viscosity

SC

of Sul(-)-MC-C formulation decreased to 8.1 × 104 cP at 37 °C, which indicated that the electrostatic attraction between Min(+) and MC-C increased the viscosity at 37 °C and the maximum viscosities;

M AN U

conversely, the electrostatic repulsion between Sul(-) and MC-C decreased the viscosity. Collectively, these results indicated that the gelation time and phase transition of drug-loaded MC and MC-C formulations were affected by the variation of the electrostatic interaction between the electrolyte drugs and MC or polyelectrolyte MC-C.

Next, the storage (G′), loss (G′’) modulus values and the phase angle of Sul(-)-MC, Min(+)-MC,

TE D

Sul(-)-MC-C, and Min(+)-MC-C were measured at 37 °C (Supplementary Information: Fig. S3). The difference between storage and loss values of Min(+)-MC-C were higher than those of other formulations at 37 °C, indicating high elastic and hydrogel-like characteristics of Min(+)-MC-C. The

EP

phase angle (tan delta) values of Sul(-)-MC, Min(+)-MC, Sul(-)-MC-C, and Min(+)-MC-C were

AC C

determined to compare their elasticity. The value of all formulation showed tan delta <1, indicating a predominant elastic behavior. Min(+)-MC-C exhibited a lower tan delta value than those of other formulations, indicating significantly more stiff, hydrogel strength features of Min(+)-MC-C. For the subsequent RA experiments, we used the MC-C formulations as the injectable electrostatically interacting drug depot formulation and the MC formulations as the control.

3.5. In vitro release of drug-loaded MC and MC-C depots

ACCEPTED MANUSCRIPT To investigate the drug release from MC and MC-C drug depots, the in vitro Sul(-) and Min(+) release behavior from Sul(-)-MC, Min(+)-MC, Sul(-)-MC-C, and Min(+)-MC-C was examined in PBS at 37 °C for 40 days (Fig. 4).

RI PT

For the Sul(-)-MC and Min(+)-MC formulations, the released amounts of Sul(-) and Min(+) were 64% and 57% at day 1 and then showed a gradual release, which reached approximately 90% at 5 days. The MC drug depots showed an initial burst release of Sul(-) and Min(+) at day 1. The slight difference

SC

in the release of Sul(-)-MC over 5 days was most likely attributable to the slightly higher hydrogel strength of Min(+)-MC in comparison with that of Sul(-)-MC, which indicated that the hydrogel elastic

M AN U

modulus affected the drug release. Additionally, the initial burst at day 1 is a result of little or no electrostatic interaction between Sul(-) or Min(+) and the MC drug depot. For the Sul(-)-MC-C and Min(+)-MC-C formulations, the Sul(-)-MC-C formulation showed 37% release of Sul(-) after 1 day, which represented a slight lower initial burst in comparison with the MC drug depot. Subsequently, the release of Sul(-) increased gradually to 53% at 2 days, 74% at 3 days, 83%

TE D

at 4 days, and then reached 90% at 24 days. Although MC-C had electrostatic repulsive interactions with Sul(-), the suppression of the initial burst was a result of the greater hydrogel strength of MC-C at 37 °C than that of MC (Fig. 2a).

EP

The Min(+)-MC-C formulation resulted in 16% release of Min(+) after 1 day and then a gradually

AC C

increased release of 29% at 2 days, 33% at 3 days, 41% at 5 days, 52% at 10 days, 74% at 15 days, 83% at 20 days, and 90% at 30 days (Fig. 4b). The Min(+)-MC-C formulation indicated the remarkably prolonged release of Min(+) over 40 days, although Min(+) release from Min(+)-MC-C depot showed sigmodal shape with inflection at 1 day probably due to the suppression of the initial burst by the greater hydrogel strength of MC-C and at 15 days due to the slightly dissipation of MC-C depot. It is likely that the sustained Min(+) release from the Min(+)-MC-C depot resulted from the attractive electrostatic interaction between Min(+) and MC-C and the greater hydrogel strength of MC-C in comparison to that of MC.

ACCEPTED MANUSCRIPT

3.6 In vitro cell viability effects of drug-loaded MC and MC-C depots Because RA of the joints is associated with the inflammation of synovial cells in the synovial

RI PT

membrane, we first examined the viabilities of normal synovial cells and RAW 264.7 cells exposed to drug alone, Sul(-) and Min(+) with different concentrations of 1, 5, and 10 mg/mL, and hydrogel depot alone (200 mg/mL) (Supplementary Information: Figs. S4 and S5). Synovial cells and RAW 264.7

SC

cells that were not treated with any drug or hydrogel depot were used as controls.

The synovial cells was almost similar viability with the control for up 5 mg/mL, but showed

M AN U

significantly reduced cell viabilities of approximately 60% at 7 and 10 mg/mL in comparison with control. The viabilities of RAW 264.7 cells were also examined at 24 h after treatment with drug alone. All concentrations of Sul(-) and Min(+) showed reduced RAW 264.7 cell viabilities of approximately 15–20% in comparison with control.

Meanwhile, hydrogel depot alone exhibited similar growth of synovial cells and RAW 264.7 cells in

TE D

comparison with control as a function of culture time, indicating no significant inhibitory effect on cell proliferation over an extended period. Collectively, the formulation with Sul(-) or Min(+) of 5 mg/mL concentration and MC or MC-C (200 mg/mL) was used in the following experiment.

EP

Next, we examined the viabilities of normal synovial cells exposed to Sul(-), Min(+), Sul(-)-MC,

AC C

Min(+)-MC, Sul(-)-MC-C, and Min(+)-MC-C for 21 days (Fig. 5a). Synovial cells that were not treated with any drug were used as controls. For all formulations, synovial cells exhibited cell growth as a function of culture time. The viability of synovial cells was not significantly different to the control for up to 10 days. After 10 days, the synovial cells exhibited slightly lower viabilities than the control cells. However, these differences were not significant, which implied that the synovial cells exhibited gradual growth after treatment with all drug formulations. To examine the anti-inflammatory effects of all formulations, the viabilities of RAW 264.7 cells were

ACCEPTED MANUSCRIPT examined after treatment with the test compounds for 21 days (Fig. 5b). The drug untreated RAW 264.7 cells exhibited cell growth for 21 days. On day 1, all drug formulations showed reduced cell viabilities of approximately 20–30% in comparison with the corresponding untreated cells. This indicated that the

RI PT

growth of RAW 264.7 cells was suppressed by the presence of Sul(-) and Min(+) in the formulation. Thereafter, other formulations, except Min(+)-MC-C, resulted in cell growth as a function of culture time.

SC

For the formulations of Sul(-)-MC, Min(+)-MC, and Sul(-)-MC-C, the growth of RAW 264.7 cells was suppressed for up to 4 days as a result of over 60% drug release (Fig. 4c). Thereafter, the viability

M AN U

of RAW 264.7 cells increased to approximately 30–50% at 10 days, and reached 70–80% at 21 days. This indicated that the drug release from the hydrogel depot was sustained for only 4–7 days and little release occurred after 10 days (Fig. 4b).

Meanwhile, the viability of RAW 264.7 cells after treatment with the Min(+)-MC-C formulation was 20–30% between 1–21 days in comparison with the untreated RAW 264.7 cells. This indicated that the

TE D

Min(+) release from the MC-C depot was sustained for 21 days as a result of the attractive electrostatic interaction between Min(+) and the MC-C depot.

EP

3.7. Evaluation of RA repair via imaging of the paws, ankle diameter, and articular index

AC C

The RA rats exhibited swollen paws, altered ankle diameter, and stiffness in their movement. To observe therapeutic RA repair, injectable formulations of Sul(-) alone, Min(+) alone, Sul(-)-MC, Min(+)-MC, Sul(-)-MC-C, and Min(+)-MC-C with equivalent drug concentrations were intra-articularly injected into the articular knee joints of RA rats. As shown in Fig. 6a, the untreated RA rats and the rats treated with Sul(-) alone, Min(+) alone, Sul()-MC, Min(+)-MC, and Sul(-)-MC-C showed severe edema and erythema for 6 weeks in comparison with untreated normal rats. However, the RA rats intra-articularly injected with Min(+)-MC-C showed some edema and erythema at 1–2 weeks, which had almost returned to at 4 and 6 weeks, which

ACCEPTED MANUSCRIPT indicated that Min(+)-MC-C produced significant ameliorative effects in animals with RA. Then, the hind paws of untreated RA rats, drug formulation-treated RA rats, and normal rats were examined to determine the AI score and ankle diameter (Fig. 6b and c). The mean AI score gradually

RI PT

decreased at 1 or 2 weeks and reached a plateau after 2 weeks in untreated RA rats, drug formulationtreated RA rats, and normal rats. However, Min(+)-MC-C-treated RA rats showed a gradual decreased in mean AI over 3 weeks. After 3 weeks, the mean AI drastically decreased and approached almost

SC

similar levels to that of normal rats at 6 weeks.

The ankle diameter of RA rats, untreated rats, and drug formulation-treated RA rats showed an

M AN U

irregular pattern. However, the Min(+)-MC-C-treated rats showed a gradual decrease in ankle diameter that reached a similar value to that of normal rats after 6 weeks. These results indicate that Min(+)-MCC resulted in the amelioration of RA and implied that it suppressed inflammation more effectively than the other formulations.

TE D

3.8. Confirmation of RA amelioration via histology The therapeutic effects of intra-articularly injected RA rats were examined by H&E staining (Fig. 7), SO staining (Fig. 8), and TNF-α staining (Fig. 9) of the articular knee joint.

EP

In the images stained with H&E, normal rats showed clear chondrocytes within lacunae in the

AC C

cartilage. Sul(-) alone, Min(+) alone, Sul(-)-MC, Min(+)-MC, and Sul(-)-MC-C-treated RA rats and untreated RA rats showed few chondrocytes in their cartilage, even at 6 weeks. However, the Min(+)MC-C-treated RA rats exhibited a few chondrocytes at 1 week, but exhibited clear chondrocytes within lacunae at 4 and 6 weeks.

The cartilage thicknesses in sections of synovial tissue were determined from the H&E images of RA rats intra-articularly injected with all formulations and normal rats (Fig. 7b). The cartilage thickness of normal rats increased as a function of time, from 140 µm at 1 week, to 150 µm at 2 weeks, to 230 µm at 4 weeks, and to 430 µm at 6 weeks.

ACCEPTED MANUSCRIPT For Sul(-) alone, Min(+) alone, Sul(-)-MC, Min(+)-MC, and Sul(-)-MC-C-injected RA rats, the cartilage thicknesses showed little change in length as a function of time, increasing from 80–92 µm in 1 week to only 130–150 µm at 6 weeks. Meanwhile, the cartilage thickness of Min(+)-MC-C-treated RA

RI PT

rats increased as a function of time, from 99 µm in 1 week to 313 µm at 6 weeks; 80% cartilage thickness of the normal rats increased, which implied almost complete amelioration of RA.

The glycosaminoglycan (GAG) deposition (pink color) in the cartilage was observed through SO

SC

staining (Fig. 8). In SO staining for Sul(-) alone, Min(+) alone, Sul(-)-MC, Min(+)-MC, and Sul(-)-MCC injected RA rats, little or no positive staining with SO was observed at 1 week in sections of synovial

M AN U

tissue. However, the SO staining of Min(+)-MC-C-treated rats with RA exhibited some positively stained areas of GAGs at 2 and 4 weeks. At 6 weeks, the positive area of GAGs was increased. Additionally, mature chondrocytes with rounded shapes were observed within the cartilage, which indicated the formation of new cartilage-like tissue. These results showed that the Min(+)-MC-C depot could successfully support the regeneration of in vivo cartilage tissue via the sustained release of Min(+).

TE D

The immunohistochemical TNF-α staining of Sul(-) alone, Min(+) alone, Sul(-)-MC, Min(+)-MC, Sul(-)-MC-C, and Min(+)-MC-C injected rats with RA was performed to investigate TNF-α expression

synovial tissue.

EP

in synovial tissues of RA animals (Fig. 9). The normal rats showed almost no reddish staining in

AC C

However, untreated RA rats and Sul(-) alone and Min(+) alone-treated RA rats showed reddish staining in widespread areas of synovial tissue, which was indicative of abundant TNF-α expression. The reddish stained area was maintained without large difference in synovial tissue for 6 weeks. Sul(-)-MC, Min(+)-MC, and Sul(-)-MC-C-treated RA rats showed a decrease in the reddish stained area in sections of synovial tissue at 1 week in comparison with the untreated, Sul(-) alone, or Min(+) alone-treated RA rats. Over time, the reddish stained areas decreased in considerable synovial tissue area. Meanwhile, the reddish stained areas in Min(+)-MC-C-treated RA rats rapidly decreased and almost disappeared at 6 weeks, and similar levels to those seen in normal rats were achieved.

ACCEPTED MANUSCRIPT The quantification of the TNF-α expression in synovial tissues was obtained by the normalization of each mean value by the reddish stained areas in the total stained tissue area (Fig. 9b). Untreated RA rats and Sul(-) alone, Min(+) alone, Sul(-)-MC, Min(+)-MC, and Sul(-)-MC-C-treated RA rats maintained a

RI PT

high intensity of reddish stained areas, whereas the TNF-α staining in Min(+)-MC-C-treated RA rats gradually decreased and approached a similar expression level to that of normal rats. These results demonstrated that Min(+)-MC-C suppressed inflammation in the cartilage of RA rats in vivo, which led

SC

to the amelioration of RA.

M AN U

3.9. Evaluation of RA treatment via inflammatory TNF-α and IL-1β measurements To investigate the expression of the inflammatory proteins TNF-α and IL-1β in the articular knee joint in untreated RA rats, drug formulation-treated RA rats, and normal rats, we determined the amount of TNF-α and IL-1β expression in the articular knee joint (Fig. 10). Normal rats showed very low expression levels of TNF-α and IL-1β in the articular knee joint up to 6 weeks.

TE D

The amount of TNF-α expression in untreated RA rats (control) was 103–117 pg/mL at 1–6 weeks. After the treatments of Sul(-) alone, Min(+) alone, Sul(-)-MC, Min(+)-MC, and Sul(-)-MC-C, the TNFα expression in RA rats at 1-6 weeks was 80–100 pg/mL. The expression of Min(+)-MC-C-treated RA

EP

rats was 69 pg/mL at 1 week, which rapidly decreased to 31 pg/mL at 2 weeks, 20 pg/mL at 4 weeks,

AC C

and 11 pg/mL at 6 weeks, thereby reaching similar expression levels to that of normal rat (3 pg/mL at 6 weeks). This result agreed with the TNF-α staining examined in Fig. 9. The expression of IL-1β was 60–80 pg/mL for rats treated with Sul(-) alone, Min(+) alone, Sul(-)-MC, Min(+)-MC, and Sul(-)-MC-C. Min(+)-MC-C-treated RA rats also showed a similar IL-1β expression of 68 pg/mL at 1 week. However, the amount of IL-1β in Min(+)-MC-C-treated RA rats gradually decreased to 53 pg/mL at 2 weeks, 42 pg/mL at 4 weeks, and 31 pg/mL at 6 weeks, which approached a similar expression level to that of normal rat (2.5 pg/mL at 6 weeks). These results indicated that Min(+)-MC-C suppressed the expression of the inflammatory proteins

ACCEPTED MANUSCRIPT TNF-α and IL-1β in the articular knee joint, which resulted in superior RA treatment in comparison with the other formulations.

RI PT

Discussion On the previous works, we identified MC diblock copolymer solutions as an injectable hydrogel depot via the hydrophobic aggregation among hydrophobic polyester segments. Usually, an attractive or

SC

repulsive electrostatic interaction is reproducibly induced between positive and/or negative electrolytes. It is likely that RA drugs were homogenously entrapped inside MC diblock copolymer solutions through

M AN U

hydrophobic and electrostatic interactions and formed hydrogels as electrostatically interacting drug depot at body temperature. We anticipated that the anionic pendant group on MC diblock copolymers stabilize or destabilize the hydrophobic aggregation between the hydrophobic MC segments via electrostatic interaction between anionic or cationic electrolytes in the aqueous solutions. In this study, we sought to develop an injectable electrostatically interacting drug depot for the

TE D

sustained release of an ionic drug for RA treatment. In previous works, we reported the preparation of polyelectrolyte MC-C copolymer and examined the possibility of their use as an injectable hydrogel depot [3,4].

EP

Thus, we aimed to utilize these attractive or repulsive electrostatic interactions between the ionic

AC C

drug and the polyelectrolyte depot. The extent of electrostatic interactions of ionic drugs inside the polyelectrolyte MC-C hydrogels resulted in substantial changes in the zeta potentials, gelation time, phase transition, and hydrogel strength, which may have been a result of the magnitude of anionic and cationic electrolytes. Collectively, we successfully prepared the electrostatically interacting drug depot formulations. We hypothesized that the release of drugs with opposite charges to the polyelectrolyte depot can be sustained by attractive electrostatic interaction, whereas the release of drug with the same charges as polyelectrolyte depot generate repulsion, which results in fast drug release from depot. In the present

ACCEPTED MANUSCRIPT study, the in vitro release time of the anionic RA drug inside the polyelectrolyte MC-C depot and the anionic and cationic RA drugs inside MC without electrolytes was shorter than that of the cationic RA drug inside MC-C, which indicated that the attractive electrostatic interaction led to the sustained RA

RI PT

drug release from the electrostatically interacting depot. The cationic RA drug inside MC-C exhibited high viscosity than those of the anionic RA drug inside the MC-C depot, and the anionic and cationic RA drugs inside MC without electrolytes. Since high

SC

viscosity of cationic RA drug inside MC-C suggest high hydrogel strength due to high hydrophobic aggregation among hydrophobic polyester segments, the cationic RA drug inside MC-C exhibited a lag

M AN U

time and sustained release for the release of RA drug for an extended period in vitro. Collectively, we found that electrostatically interacting formulations successfully formed, providing support for our hypothesis that the anionic pendant group on MC diblock copolymers stabilize or destabilize the hydrophobic aggregation between the hydrophobic MC segments via electrostatic interaction between anionic or cationic electrolytes in the aqueous solutions.

TE D

Therefore, we evaluated whether the injectable, electrostatically interacting depot formulations prepared in this work could be injected into the articular joint to produce drug depots and determined whether the electrostatically interacting depot produced a restorative effect on RA joints through an

EP

extended period of drug release.

AC C

All the injectable electrostatically interacting depot formulations exhibited no significant cytotoxicities in synovial cells, which alleviated concerns about the in vivo safety of the RA treatment. The formulations that contained the drug alone [Sul(-) and Min(+)], weak electrostatic interactions [Sul(-)-MC and Min(+)-MC], and repulsive electrostatic interaction [Sul(-)-MC-C] slightly suppressed the growth of RAW 264.7 cells. The electrostatically interacting drug depot [Min(+)-MC-C] formulation suppressed almost 70% of RAW 264.7 cells for 21 days, which indicated the sustained Min(+) release from the depot for an extended period. The injectable electrostatically interacting depot formulations were prepared by simple mixing of the

ACCEPTED MANUSCRIPT hydrogel and the RA drug, which could be easily loaded into the syringe; this is another major advantage of hydrogels, as the formulations can flow in the syringe. The individual solutions of the injectable, electrostatically interacting hydrogel formulations, as well as the drug alone, were intra-articularly

RI PT

injected into rats by using a 26-G needle. Even though toughness of hydrogel depots is not compare with commercial available hydrogel such as Hylan GF 20 (Biomatrix, USA) [29], it is reasonable to conclude that the injectable, electrostatically interacting hydrogel designed herein can act as a drug depot

SC

formulation with clinical injectability and applicability.

From the in vivo animal experiments, Sul(-) alone or Min(+) alone intra-articularly injected RA rats

M AN U

exhibited little or no amelioration of RA as a result of the rapid clearance of the RA drug from the injected synovial location. Some research groups, including our group, have already reported this limitation and attempted to increase the efficacy through extended exposure of the proliferating cells to the drug in the joint by using a drug depot [22-27].

In present results, the intra-articular injection of formulations with weak electrostatic interaction

TE D

[Sul(-)-MC, Min(+)-MC] and repulsive electrostatic interaction [Sul(-)-MC-C] exhibited some level of RA treatment in comparison with the drug alone. However, the extent of RA treatment was quite low,

location.

EP

owing to the comparably rapid clearance of the RA drug from the drug depot at the injected synovial

AC C

However, RA rats injected intra-articularly with the electrostatically interacting drug depot [Min(+)MC-C] exhibited almost complete RA amelioration and very low inflammatory TNF-α and IL-1β expression in the articular knee joint. The extent of RA repair resulted in histologies and expression levels that were almost similar to normal rats. Even though the injectable, electrostatically interacting hydrogel designed in this work is not utilize directly for the typical human systemic RA disease, we believe that the local drug delivery system using electrostatically interacting Min(+)-MC-C depot formulation can provide a rationale for the treatment of RA through intra-articular injection.

ACCEPTED MANUSCRIPT

Conclusions Collectively, we successfully prepared attractive and repulsive electrostatically interacting drug depot

RI PT

via electrostatic interactions between polyelectrolytes depots and electrolyte drugs. The attractive or repulsive electrostatic interaction affected the gelation time, phase transition of the depot, and the drug release from the depot. An injectable electrostatically interacting depot formulation, Min(+)-MC-C,

SC

administered through intra-articular injection, successfully provided almost complete amelioration of

M AN U

RA.

Acknowledgments

This study was supported by a grant from a Basic Science Research Program (2016R1A2B3007448) and Priority Research Centers Program (2010-0028294) through the National Research Foundation of

TE D

Korea (NRF) funded by the Ministry of Education.

AC C

EP

Competing financial interests: The authors declare no competing financial interests.

ACCEPTED MANUSCRIPT References [1]

L. Yu, J. Ding, Injectable hydrogels as unique biomedical materials. Chem. Soc. Rev. 37 (2008) 1473-1481.

[2]

D.Y. Kim, D.Y. Kwon, J.S. Kwon, J.H. Kim, B.H. Min, M.S. Kim, Stimuli-Responsive

[3]

RI PT

Injectable in situ-forming hydrogels for regenerative medicines. Polym. Rev. 55 (2015) 407-452. J.I. Kim, D.Y. Kim, D.Y. Kwon, H.J. Kang, J.H. Kim, B.H. Min, M.S. Kim, An Injectable Biodegradable Temperature-Responsive Gel with an Adjustable Persistence Window.

[4]

SC

Biomaterials 33 (2012) 2823-2834.

B. K. Lee, J.H. Park, S.H. Park, J.H. Kim, S.H. Oh, S.J. Lee, B.Y. Lee , M.S. Kim, Preparation of Pendant Group-Functionalized Diblock Copolymers with Adjustable Thermogelling Behavior,

[5]

M AN U

Polymers 9 (2017) 239.

T. Kuroiwa, I. Kobayashi, A.M. Chuah, M. Nakajima, S. Ichikawa, Formulation and stabilization of nano-/microdispersion systems using naturally occurring edible polyelectrolytes by electrostatic deposition and complexation. Adv. Colloid Interface Sci. 226 (2015) 86-100.

[6]

V. Bourganis, T. Karamanidou, O. Kammona, C. Kiparissides, Polyelectrolyte complexes as

(2017) 44-60. [7]

TE D

prospective carriers for the oral delivery of protein therapeutics. Eur. J. Pharm. Biopharm. 111

S. Whitebread, B. Dumotier, D. Armstrong, A. Fekete, S. Chen, A. Hartmann, P.Y. Muller, L.

EP

Urban, Secondary pharmacology: screening and interpretation of off-target activities – focus on translation. Drug Discov. Today 21 (2016) 1232-1242. S. Murdan, Electro-responsive drug delivery from hydrogels. J. Control. Release 92 (2003) 1-17.

[9]

M. Ferstl, A. Strasser, H.J. Wittmann, M. Drechsler, M. Rischer, J. Engel, A. Goepferich,

AC C

[8]

Nanofibers resulting from cooperative electrostatic and hydrophobic interactions between peptides and polyelectrolytes of opposite charge. Langmuir 27 (2011) 14450-14459. [10]

U. Han, Y. Seo, J. Hong, Effect of pH on the structure and drug release profiles of layer-by-layer assembled films containing polyelectrolyte, micelles, and graphene oxide. Sci. Rep. 6 (2016) 24158.

ACCEPTED MANUSCRIPT [11]

C. Xie, P. Li, L. Han, Z. Wang, T. Zhou, W. Deng, K. Wang, X. Lu, Electroresponsive and cellaffinitive polydopamine/polypyrrole composite microcapsules with a dual-function of ondemand drug delivery and cell stimulation for electrical therapy. NPG Asia Mater. 9 (2017) e358. A.C. Lima, C. Alvarez-Lorenzo, J.F. Mano, Design Advances in Particulate Systems for

RI PT

[12]

Biomedical Applications. Adv. Healthc. Mater. 5 (2016) 1687–1723. [13]

D. Uppalapati, B.J. Boyd, S. Garg, J. Travas-Sejdic, D. Svirskis, Conducting polymers with defined micro- or nanostructures for drug delivery. Biomaterials 111 (2016) 149-162.

B. Philipp, H. Dautzenberg, K.J. Linow, J. Kötz, W. Dawydoff, Polyelectrolyte complexes —

SC

[14]

recent developments and open problems. Prog. Polym. Sci. 14 (1989) 91-172. M.R. Park, C. Chun, C.S. Cho, S.C. Song, Enhancement of sustained and controlled protein

M AN U

[15]

release using polyelectrolyte complex-loaded injectable and thermosensitive hydrogel. Eur. J. Pharm. Biopharm. 76 (2010) 179-188. [16]

Y. Okada, T. Kishikawa, S. Sakaue, J. Hirata, Future Directions of Genomics Research in Rheumatic Diseases. Rheum. Dis. Clin. North Am. 43 (2017) 481-487. G. Boleto, M. Dramé, I. Lambrecht, J.P. Eschard, J.H. Salmon, Disease-modifying anti-

TE D

[17]

rheumatic drug effect of denosumab on radiographic progression in rheumatoid arthritis: a systematic review of the literature. Clin. Rheumatol. 36 (2017) 1699-1706. F. Hayat, S.H. Lee, E.J. Lee, S.J. Kim, K.A. Jung, S.K. Lee, J. Youn, J.K. Min, M.L. Cho, D.Y.

EP

[18]

Shin, STX0119 Ameliorates Arthritis in SKG Mice via Inhibiting T Helper 17. Tissue Eng. Reg.

[19]

AC C

Med. 17 (2016) 91-99.

S. Ramiro, C. Gaujoux-Viala, J.L. Nam, J.S. Smolen, M. Buch, L. Gossec, D. van der Heijde, K. Winthrop, R. Landewé, Safety of synthetic and biological DMARDs: a systematic literature review informing the 2013 update of the EULAR recommendations for management of rheumatoid arthritis. Ann. Rheum. Dis. 73 (2014) 529-535.

[20]

S. Tarp, D.E. Furst, M. Boers, G. Luta, H. Bliddal, U. Tarp, K.H. Asmussen, B. Brock, A. Dossing, T.S. Jørgensen, S. Thirstrup, R. Christensen, Risk of serious adverse effects of biological and targeted drugs in patients with rheumatoid arthritis: a systematic review metaanalysis. Rheumatology (Oxford) 56 (2017) 417-425.

ACCEPTED MANUSCRIPT [21]

H. O'Mary, I. Del Rincόn, Z. Cui, Nanomedicine for Intra-Articular Drug Delivery in Rheumatoid Arthritis. Curr. Med. Chem. 23 (2016) 2490-2506.

[22]

M. Rahman, G. Sharma, K. Thakur, F. Anwar, O.P. Katare, V.G. Goni, V. Kumar, M.A. Zamzami, S. Akhter, Emerging Advances in Nanomedicine as a Nanoscale Pharmacotherapy in Rheumatoid

[23]

RI PT

Arthritis: State of the Art. Curr. Top. Med. Chem. 17 (2017) 162-173. G. Gevers, J. Dequeker, M. van Holsbeeck, E. van Vliet-Daskalopoulou, A high dose (up to 200 mg) tolerance and efficacy study of intra-articular rimexolone (Org 6216) in rheumatoid synovitis of the knee. Clin. Rheumatol. 13 (1994) 103-109.

C. Larsen, J. Ostergaard, S.W. Larsen, H. Jensen, S. Jacobsen, C. Lindegaard, P.H. Andersen,

SC

[24]

Intra-articular depot formulation principles: role in the management of postoperative pain and

[25]

M AN U

arthritic disorders. J. Pharm. Sci. 97 (2008) 4622-4654.

X.M. Liu, L.D. Quan, J. Tian, F.C. Laquer, P. Ciborowski, D. Wang, Syntheses of Click PEG−Dexamethasone

Conjugates

for

the

Treatment

of

Rheumatoid

Arthritis.

Biomacromolecules 11 (2010) 2621–2628. [26]

K. Kim, J.H. Park, S.H. Park, H.Y. Lee, J.H. Kim, M.S. Kim, An Injectable, Click-Cross-Linked

TE D

Small Intestinal Submucosa Drug Depot for the Treatment of Rheumatoid Arthritis. Adv. Healthc. Mater. 5 (2016) 3105-3117. [27]

A.R. Son, D.Y. Kim, S.H. Park, J.Y. Jang, K. Kim, B.J. Kim, X.Y. Yin, J.H. Kim, B.H. Min, D.K. Han, M.S. Kim, Direct chemotherapeutic dual drug delivery through intra-articular injection for

[28]

EP

synergistic enhancement of rheumatoid arthritis treatment. Sci. Rep. 5 (2015) 14713. M. S. Kim, H. H., K. S. Seo, Y. H. Cho, G. Khang, H. B. Lee, Preparation and characterization of

AC C

MPEG-PCL diblock copolymers with thermo-responsive sol-gel-sol behavior, Journal of Polymer Science Part A: Polymer Chemistry, 2006, 44, 5413-5423. [29]

R. Barbucci, S. Lamponi, A. Borzacchiello, L. Ambrosio, M. Fini, P. Torricelli, R. Giardino, Hyaluronic acid hydrogel in the treatment of osteoarthritis, Biomaterials 23 (2002) 4503–4513.

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

EP

Figure 1. Schematic images for (a) MC-C copolymer, electrostatically attracted minocycline (Min(+)) and electrostatically repulsive sulfasalazine (Sul(-)) as an injectable, electrostatically interacting drug

AC C

depot and (b) the treatment of rheumatoid arthritis by intra-articular injection of electrostatically interacting Min(-) depot hydrogel. (The image (b) was drawn by J.H.P and M.S.K. using Adobe Photoshop 7.0 software).

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

EP

Figure 10. The expression of inflammatory (a) TNF-α and (b) IL-1β in rats with rheumatoid arthritis that received no treatment (control),

AC C

and received intra-articular injection of sulfasalazine (Sul(-)), minocycline (Min(+)), Sul(-)-MC, Min(+)-MC, Sul(-)-MC-C, and Min(+)MC-C, and of normal rats at 1–6 weeks (*p < 0.001, **p > 0.05).

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

MC-C, Sul(-)-MC-C, and Min(+)-MC-C.

TE D

Figure 2. Viscosity versus temperature curves for diblock copolymers: (a) MC and MC-C; (b) MC, Sul(-)-MC, and Min(+)-MC and (c)

SC

RI PT

ACCEPTED MANUSCRIPT

M AN U

Figure 3. Zeta potentials for (a) MC and (b) MC-C copolymer solutions without and with sulfasalazine

AC C

EP

TE D

and minocycline (sulfasalazine (Sul(-)) and minocycline (Min(+)).

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Figure 4. (a) Schematic images for MC-C copolymer, Sul(-) and Min(+) drugs, electrostatically attracted minocycline and electrostatically repulsive sulfasalazine, (b) cumulative drug release for 40 days, and (c) for magnified 5 days from Sul(-)-MC, Min(+)-MC, Sul(-)-MC-C, and Min(+)-MC-C.

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Figure 5. Viabilities of (a) synovial cells and (b) RAW 264.7 cells treated with control (no drug), sulfasalazine (Sul(-)), minocycline

AC C

EP

(Min(+)), Sul(-)-MC, Min(+)-MC, Sul(-)-MC-C, and Min(+)-MC-C at 1, 4, 7, 10, 14, 18, and 21 days (*p > 0.05, **p < 0.001).

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

Figure 6. (a) Paw photographs, (b) articular index (AI) score, and (c) ankle joint diameter of rats with rheumatoid arthritis that received no treatment (control), and received intra-articular injection of sulfasalazine (Sul(-)), minocycline (Min(+)), Sul(-)-MC, Min(+)-MC, Sul(-)-MC-C, and Min(+)-MC-C, and of normal rats at 1–6 weeks (scale bars = 10 mm).

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

TE D

Figure 7. (a) H&E staining and (b) the thickness of the articular knee joint cartilage of rats with rheumatoid arthritis that received no treatment (control), and received intra-articular injection of sulfasalazine (Sul(-)), minocycline (Min(+)), MC-Sul(-), MC-Min(+), MC-C-

AC C

EP

Sul(-), and MC-C-Min (+), and of normal rats at 1–6 weeks (scale bars = 100 µm; *p < 0.05, **p > 0.05).

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Figure 8. Safranin-O staining of the articular knee joints of rats with rheumatoid arthritis that received no treatment (control), and received

AC C

intra-articular injection of sulfasalazine (Sul(-)), minocycline (Min(+)), MC-Sul(-), MC-Min(+), MC-C-Sul(-), MC-C-Min(+), and of normal rats at 1–6 weeks (scale bars = 100 µm).

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Figure 9. (a) TNF-α staining of the articular knee joints of rats with rheumatoid arthritis that received no treatment (control), and received

EP

intra-articular injection of sulfasalazine (Sul(-)), minocycline (Min(+)), MC-Sul(-), MC-Min(+), MC-C-Sul(-), MC-C-Min(+), and of

0.001).

AC C

normal rats at 1–6 weeks, and (b) the TNF-α positive area calculated as a percentage of total cartilage area (scale bars = 100 µm; *p <