Background
Osteoarthritis (OA) is the most common joint disorder worldwide and one of the leading causes of disability in the elderly [
1]. Treatment for knee OA aims to relieve pain and improve function in order to mitigate reductions in physical activity. Intra-articular (IA) injection of hyaluronic acid (hyaluronan, HA), which is a linear glycosaminoglycan composed of repeating disaccharides of glucuronic acid and
N-acetylglucosamine, is a recognized treatment for pain associated with symptomatic knee OA [
2‐
4]. HA is distributed throughout the body, and especially in synovial fluid [
5,
6]. Balazs reported that the molecular weight of HA in the synovial fluid of OA patients was much lower than that of healthy humans [
6]. Dahl et al. reported that the HA concentration in synovial fluid and its molecular weight in patients with OA and rheumatoid arthritis (RA) are lower than those in healthy subjects [
7]. In the knee joint, HA plays a major role in maintaining the lubricant properties of synovial fluid under both dynamic and static conditions, and in exerting various physiological activities such as reduction of pro-inflammatory cytokine levels for mitigating cartilage degeneration, and reduction of COX-2 production for reducing pain sensation [
8‐
10]. Therefore, it was anticipated that intra-articular injection of HA would be beneficial for the management of knee OA. Notably, Asari et al. reported that positive staining of HA was distinctly reduced in synovial lining cells following anterior cruciate ligament (ACL) transection in dogs, but HA staining was maintained in the synovial lining cells of the HA-treated animals [
11]. Furthermore, Ikeya et al. reported that exogenous HA induced the production of endogenous HA in synoviocytes from OA and RA patients [
12]. The increase in endogenous HA is considered to normalize pathological synovial fluid and contributes to a long-term analgesic effect.
In pursuit of next generation OA therapeutics, we developed a novel conjugated compound, SI-613, which is a derivative of high-molecular-weight HA produced by fermentation (600,000 to 1200,000 Da) tethered to the non-steroidal anti-inflammatory drug diclofenac (DF) via a 2-aminoethanol linker extended from the glucuronic acid moieties [
13]. SI-613 releases DF locally in a sustained manner and remains in the joint for a long period, similar to the existing IA-HA injection formulation. We previously reported that intra-articularly administered SI-613 exerts much more robust anti-inflammatory and analgesic effects than native HA or orally-administered DF-Na in experimental animal models [
13]. SI-613 is thus a promising candidate in clinical development for symptomatic knee OA.
In the present study, we investigated the effects of SI-613 on the production of high molecular weight HA in synoviocytes from OA and RA patients and compared these effects with those of HA. Furthermore, we clarified the mechanism by which SI-613 induces the production of high molecular weight HA.
Methods
Materials
ARTZ Dispo® (1% HA) and hyaluronidase were purchased from Seikagaku Corporation (Tokyo, Japan). Diclofenac sodium was from Wako Pure Chemical Industries, Ltd. (Osaka, Japan). Water for injection (WFI) was from Otsuka Pharmaceutical Factory, Inc. (Tokushima, Japan). Human fibroblast-like synoviocytes from rheumatoid arthritis (HFLS-RA) patients and those from osteoarthritis (HFLS-OA) patients were purchased from Cell Applications, Inc. (San Diego, CA). Growth medium and a subculture reagent kit were from Cell Applications, Inc. α-MEM, α-MEM powder, penicillin/streptomycin and Dulbecco’s phosphate buffered saline (D-PBS) were from Life Technologies Corporation (Waltham, MA). FBS was from MP Biomedicals LLC. (Santa Ana, CA). IL-1β was from R&D Systems, Inc. (Minneapolis, MN). Bovine serum albumin (BSA) was from Sigma-Aldrich Co. LLC. (St. Louis, MO). Glucosamine hydrochloride, D-[6-3H (N)] ([3H]glucosamine) (37 MBq/mL) and Ultima-FloTMM (scintillation solution) were from PerkinElmer Co., Ltd. (Waltham, MA). OH pak SB-805 HQ and OH pak SB-807 HQ columns were from SHOWA DENKO K.K. (Tokyo, Japan). Select-HATM 500 k, Select-HATM 1000 k, and Select-HATM 2500 k were from Hyalose, LLC. (Oklahoma City, OK).
Preparation of SI-613
SI-613 was prepared by conjugating high-molecular-weight fermented HA (600,000 to 1200,000 Da) and DF via a 2-aminoethanol linker extended from the glucuronic acid moieties. The SI-613 active pharmaceutical ingredient was manufactured in accordance with good manufacturing practice (GMP) guidelines, and its solution was prepared in a laminar flow cabinet to maintain its sterility. SI-613 was dissolved in 5 mM phosphate-buffered saline (pH 6.0) at a concentration of 10 mg/mL and diluted appropriately with PBS before use. Although SI-613 has been used at 10 mg/mL in clinical trials, the maximum concentration of SI-613 was set at 1 mg/mL in this study because highly-concentrated SI-613 solution is very viscous and difficult to pipette.
Cells and cell culture
Three lots of human fibroblast-like synoviocytes from rheumatoid arthritis (HFLS-RA) patients and three from osteoarthritis (HFLS-OA) (Cell Applications, Inc.) patients were cultured separately in basal medium containing 10% growth supplement and 1% penicillin/streptomycin (Cell Applications, Inc.). Cells were incubated at 37 °C under 5% CO2. The medium was changed every 2 days. After confluence, the cells were seeded at a density of 3.0 × 105 cells/2 mL/well in α-MEM medium containing 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin in 6-well plates.
Challenge to test materials
One day after cell seeding, the culture medium was replaced with 2 mL of α-MEM medium containing 10% FBS, 10 ng/mL recombinant human IL-1β/IL-1F2 (IL-1β) (R&D Systems, Inc.), 1% penicillin/streptomycin, 370 kBq/mL glucosamine hydrochloride D-[6-3H (N)] ([3H]glucosamine) and each test material: HA (ARTZ Dispo®), SI-613 (Seikagaku Corporation), diclofenac sodium (DF-Na) (Wako Pure Chemical Industries, Ltd.), or a mixture of DF-Na and HA (DF-Na + HA). For mRNA expression measurements, [3H] glucosamine was excluded from the medium. The cells were incubated at 37 °C for the indicated periods under 5% CO2.
Collection of the culture supernatant and cell lysate
Culture supernatant fraction for the measurement of high molecular weight HA (HMW-HA, MW > 2400 kDa) production was collected at each incubation time. Cell lysate for the measurement of RNA was collected at 48 h and stored frozen at -80 °C.
Fractionation of the culture supernatant and measurement of the radioactivity levels
Radiolabeled HA in the culture supernatant was separated using a size exclusion HPLC column (an OH pak SB-805 HQ column or an OH pak SB-807 HQ column) in an HPLC system (JASCO Corporation). HPLC was performed at 0.5 mL/min of mobile phase (5 mmol/L phosphate buffer, 0.82% NaCl: acetonitrile = 2: 1) at 35 °C. The injection volume of culture supernatant was set at 10 μL. Eluent was collected continuously as 0.5 min fractions from 9 min (OH pak SB-805 HQ column) or 14 min (OH pak SB-807 HQ column) after starting the elution using a fraction collector (JASCO Corporation). Liquid scintillator (Ultima-FloTMM, PerkinElmer Co., Ltd.) was added to each HPLC fraction and mixed well. The disintegrations per minute (dpm) of each fraction were measured in a liquid scintillation counter (PerkinElmer Co., Ltd.). Each sample was analyzed in a single HPLC run.
Evaluation of the molecular weight of HA in each fraction using HA standard solutions
Select-HA™ 500 k, Select-HA™ 1000 k and Select-HA™ 2500 k, for which weight-average molecular weights were 528 kDa, 1076 kDa, and 2420 kDa respectively, were used for molecular weight calibration. Each standard was subjected to size exclusion HPLC analysis with online monitoring of UV absorption at 210 nm. The peak top fraction of each HA standard was calculated by considering the lag between the UV monitor and the fraction collector.
Enzyme digestion
The culture supernatant of the 0.01% HA-treated group (90 μL) was mixed with 10 μL of 100 turbidity reducing units (TRU)/mL hyaluronidase or WFI and incubated overnight at 37 °C, except for the culture supernatant of the 0.01% SI-613-treated group, which was digested by hyaluronidase treatment (30 μL hyaluronidase solution at 60 °C for 3 h). This more aggressive treatment was required because HA in the culture supernatant of the 0.01% SI-613-treated group was more difficult to digest.
Cell counting
Immediately after collecting the culture supernatant, each well was washed with Dulbecco’s phosphate-buffered saline (D-PBS) (GIBCO, Life Technologies Corporation). Trypsin/EDTA solution (Cell Applications, Inc.) was added to the wells and the cells were incubated at 37 °C under 5% CO2. The cells were detached from the wells by tapping the plate, and 0.5 mL of trypsin-neutralizing solution (Cell Applications, Inc.) was added. The live cells were counted using a hematocytometer after staining with trypan blue.
RNA extraction and cDNA sample preparation
RNA was extracted from the cell lysates according to the instruction manual for the RNeasy plus mini kit (QIAGEN). RNA concentrations were measured by ultramicro spectrophotometry (Thermo Fisher Scientific, Inc.). The RNA was stored frozen at -80 °C. cDNA samples were prepared using the Super Script III First-Strand Synthesis System (Invitrogen, Life Technologies Corporation).
Real-time quantitative PCR
The mRNA expression levels of the target genes (HAS1, HAS2, HAS3, HYAL1, HYAL2, and HYAL3) and GAPDH were measured in duplicate using real-time quantitative polymerase chain reaction (PCR; RT-qPCR). RT-qPCR was performed using TaqMan Gene Expression Assay (TaqMan) (Applied Biosystems) and Premix Ex Taq (Perfect Real Time, Takara Pac LTD.) according to the instruction manuals. The following sets of TaqMan probes and primers were used to measure the expression levels of the target genes: HAS1 (ID: Hs00987418_m1), HAS2 (ID: Hs00193435_m1), HAS3 (ID: Hs00193436_m1), HYAL1 (ID: Hs00201046_m1), HYAL2 (ID: Hs01117343_g1), HYAL3 (ID: Hs00185910_m1), and GAPDH (ID: Hs03929097_g1). Reagents were aliquoted into each well at the following volumes: 1 μL of TaqMan, 12.5 μL of Premix Ex Taq, 2 μL of ROX II (× 40), and 7.5 μL distilled water. Next, 2 μL of the cDNA sample was added. The PCR conditions were as follows: first denaturation at 95 °C for 30 s., followed by a 40-cycle sequence of denaturation (95 °C, 5 s.), annealing (60 °C, 30 s.), and extension (72 °C, 20 s.). The cycle threshold (Ct) value was measured for each gene using a real-time PCR system (MX3000P; Stratagene Corporation). The Ct value was analyzed using the comparative Ct method (ΔΔCt method) to calculate the target mRNA expression level, followed by normalization to GAPDH level. The relative mRNA expression levels were presented as the times-fold increase compared to that of the control.
Discussion
We have investigated the novel HA derivative chemically linked with DF, SI-613, which is a potentially safer and more effective treatment for OA knee pain. In a previous study, we reported that a single intra-articular administration of SI-613 provided an analgesic effect via the sustained release of DF, and this pharmacological effect lasted at least 28 days [
13].
In the present study, we evaluated the effect of SI-613 on the production of HMW-HA in synoviocytes from OA patients. We found that SI-613 induced the production of HMW-HA of > 2400 kDa in synoviocytes from patients with OA, whereas HA or a mixture of HA and DF-NA induced the production of HMW-HA of 1000 kDa. Although HA also produced HMW-HA of > 2400 kDa in some tests (Fig.
1a and
4a), this effect was not reproducible (Fig.
2a and
3b), perhaps because the range of column fractionation was exceeded. Using an appropriate column, we rarely detected HMW-HA > 2400 kDa in HA-treated culture supernatant (Fig.
5). DF-Na had no effect beyond basal-level HMW-HA production. We could not accurately evaluate the molecular weight of HA produced by SI-613 because there is no molecular weight standard for HA above 2420 kDa. However, the molecular weight of HA produced by SI-613 is clearly higher than 2420 kDa, and is likely comparable with the average molecular weight of HA from healthy humans, which is 6000 kDa [
15]. Therefore, it appears that the stimulatory effect of SI-613 on HMW-HA production helps normalize the pathological synovial fluid of OA patients.
Our findings suggest that SI-613 exerts not only anti-inflammatory and analgesic effects due to diclofenac, but also a normalizing effect on pathological synovial fluid by producing endogenous HA. Additionally, this effect of SI-613 may contribute to a long-lasting analgesic effect.
Next, we assessed the mechanism of production of HMW-HA by SI-613. SI-613 or HA was added to the synoviocytes from OA patients to evaluate the mRNA expression levels of HA synthases (HAS1, 2 and 3) and hyaluronidases (HYAL1, 2 and 3). The results revealed that SI-613 significantly suppressed HYAL2 mRNA expression and significantly enhanced HAS2 mRNA expression. On the other hand, HA by itself did not significantly change the mRNA expression levels of HYAL1, HYAL2, HYAL3, HAS1, HAS2, and HAS3.
The abundance and molecular size of HA are thought to be controlled by HA synthases and hyaluronidases. Three isoforms (HAS1, HAS2 and HAS3) are known as human HA synthases. It has been reported that HAS1 and HAS3 are responsible for producing HA with molecular weights ranging from 200 kDa to 2000 kDa, and HAS2 synthesizes HA larger than 2000 kDa [
16]. It has also been reported that HAS2 synthesizes long-chain HA larger than 3900 kDa, whereas HAS3 synthesizes various sizes of HA ranging from 120 kDa to 1000 kDa, and HAS1 synthesizes short-chain HA with a molecular weight of 120 kDa [
17]. In contrast, HAYL1, 2 and 3 have been reported as human hyaluronidases. It has been reported that HYAL2 degrades HMW-HA into low-molecular (20 kDa) HA, HYAL1 degrades HA into disaccharides, and HYAL3 has no enzymatic activity and its function remains unknown [
18]. Taken together with the result of this study and information reported by others, we suggest that the mechanism by which SI-613 induces production of HMW-HA larger than 2400 kDa involves the enhancement of HAS2 mRNA expression. Decreased expression of HYAL2 mRNA is also likely involved in the retention of HMW-HA by suppressing breakdown of HMW-HA. In contrast, HA did not change the expression levels of genes responsible for the synthesis and breakdown of HA, and HA and DF-Na had no effect on gene expression levels, indicating that the effect of SI-613 was specific.
The effect of exogenous HA on high molecular weight sodium hyaluronate production in synoviocytes from OA and RA patients was previously reported by Ikeya et al. However, the molecular weight of newly synthesized HA was not determined. Smith and Ghosh also reported that high molecular weight HA (MW = 4700 kDa) stimulated the HA synthesis, but low molecular weight HA (MW < 500 kDa) showed little or no effect [
19]. In our study, HA with a molecular weight of 1000 kDa induced the production of only a small amount of HMW-HA (MW > 2400 kDa). In contrast, SI-613 composed of HA with a molecular weight of 600–1200 kDa induced the production of HMW-HA. This finding suggested that SI-613, i.e. DF-modified HA, probably acted in a manner similar to that by the high molecular weight HA (MW = 4700 kDa). Modification with DF may change the properties of HA (MW = 600–1200 kDa). David-Raoudi et al. have also reported that CS increases hyaluronan production in human synoviocytes through differential regulation of hyaluronan synthases via p38 and Akt. CS significantly suppressed hyaluronan synthase 3 (
HAS3) mRNA expression and significantly enhanced
HAS2 mRNA expression. It had no effect on the expression levels of
HAS1 [
14]. Our findings on the enhancement of
HAS2 mRNA expression in the present study appears to be in agreement with their findings, although the tested glycosaminoglycans are different.
In the present study, we confirmed that exogenous HA induced the production of endogenous HA with molecular weights above 1000 kDa but did not induce the production of HMW-HA > 2400 kDa or change the expression levels of HAS 1, 2, 3 and HYAL 1, 2, 3, in contrast to SI-613. Therefore, HA-induced newly synthesized HA might serve as a substitute substrate for hyaluronidase rather than as functional HA, thus preventing the digestion of high molecular weight endogenous HA.