GSK5182, 4-Hydroxytamoxifen Analog, a New Potential Therapeutic Drug for Osteoarthritis

Estrogen-related receptors (ERRs) are the first identified orphan nuclear receptors. The ERR family consists of ERRα, ERRβ, and ERRγ, regulating diverse isoform-specific functions. We have reported the importance of ERRγ in osteoarthritis (OA) pathogenesis. However, therapeutic approaches with ERRγ against OA associated with inflammatory mechanisms remain limited. Herein, we examined the therapeutic potential of a small-molecule ERRγ inverse agonist, GSK5182 (4-hydroxytamoxifen analog), in OA, to assess the relationship between ERRγ expression and pro-inflammatory cytokines in mouse articular chondrocyte cultures. ERRγ expression increased following chondrocyte exposure to various pro-inflammatory cytokines, including interleukin (IL)-1β, IL-6, and tumor necrosis factor (TNF)-α. Pro-inflammatory cytokines dose-dependently increased ERRγ protein levels. In mouse articular chondrocytes, adenovirus-mediated ERRγ overexpression upregulated matrix metalloproteinase (MMP)-3 and MMP-13, which participate in cartilage destruction during OA. Adenovirus-mediated ERRγ overexpression in mouse knee joints or ERRγ transgenic mice resulted in OA. In mouse joint tissues, genetic ablation of Esrrg obscured experimental OA. These results indicate that ERRγ is involved in OA pathogenesis. In mouse articular chondrocytes, GSK5182 inhibited pro-inflammatory cytokine-induced catabolic factors. Consistent with the in vitro results, GSK5182 significantly reduced cartilage degeneration in ERRγ-overexpressing mice administered intra-articular Ad-Esrrg. Overall, the ERRγ inverse agonist GSK5182 represents a promising therapeutic small molecule for OA.

The estrogen-related receptors (ERRs) consist of ERRα, -β, and -γ [21,22]. ERRs are orphan nuclear receptors that possess high similarity sequence DNA-binding domains of estrogen receptors (ERs) [22]. However, ERRs do not bind to 17β-estradiol as an estrogen ligand [23]. ERRs are involved in various metabolic processes, including alcohol, bone, cholesterol, glucose, iron, and lipid metabolism [24]. They are expressed in the liver, muscle, heart, and bone [23,24]. Our group has reported that ERRγ is a novel catabolic regulator of OA pathogenesis [7]. GSK5182 (a 4-hydroxy tamoxifen analog) is a selective inverse agonist of ERRγ [25] and inhibits Esrrg transcriptional activity by recruiting small heterodimer partner (SHP)-interacting leucine zipper protein (SMILE) [26]. The inhibitory effects of GSK5182 on pro-inflammatory cytokine-mediated OA pathogenesis are limited [24,27]. Therefore, the aim of this study was to elucidate whether the small molecule GSK5182 is a potential therapeutic molecule for OA pathogenesis.

The Ectopic Expression or Genetic Ablation of ERRγ in the Mice
To investigate the role of ERRγ in OA pathogenesis in vivo, we ectopically overexpressed ERRγ in the knee joint tissues of 12-week-old male mice. We employed an intra-articular (IA) injection to deliver an adenovirus expressing ERRγ (Ad-Esrrg). The adenovirus delivery system to joint tissues using IA injection has been well established [6,7,32]. ERRγ overexpression was induced by IA injection (three weekly IA injections) of Ad-Esrrg, which induced a loss of glycosaminoglycans in articular cartilage above the tidemark, as determined by safranin O staining ( Figure 3A, left panel). Cartilage degeneration was quantified using the Osteoarthritis Research Society International (OARSI) grade. ERRγ overexpression significantly increased the OARSI grade (p < 0.0001) ( Figure 3A, right panel).
To further examine ERRγ cartilage-specific functions in OA pathogenesis, we used cartilage-specific ERRγ Tg mice (Col2a1-Esrrg) [7]. Compared with wildtype (WT) littermates, destabilization of the medial meniscus (DMM)-operated Col2a1-Esrrg Tg mice exhibited dramatically more cartilage damage, as shown by safranin O staining and the OARSI grade (p < 0.0001) ( Figure 3B). Other symptoms of OA, including subchondral sclerosis and osteophyte formation, were also dramatically enhanced in Col2a1-Esrrg Tg mice when compared with WT littermates ( Figure 3B). Collectively, the results of our current experiments demonstrated that ERRγ is a key player in OA pathogenesis. Additionally, we used ERRγ-knockout (KO) mice as a reverse approach. ERRγ-null mice demonstrate embryonic lethality [33], therefore we used heterozygous mice (Esrrg +/− ) for the OA experiments. ERRγ-knockout (KO) mice revealed that DMM-induced cartilage erosion, osteophyte formation, and subchondral bone sclerosis were dramatically attenuated in Esrrg +/− mice (p < 0.0001) ( Figure 3C). This result supported the conclusion that ERRγ is an important catabolic regulator in OA pathogenesis.

Discussion
OA is the most well-known form of arthritis; its symptoms include cartilage destruction, synovial inflammation, osteophyte formation, and subchondral bone sclerosis [5,8,35]. Moreover, OA is an arthropathy and a leading cause of pain and disability with a sizable socioeconomic cost. However, no effective therapies for OA have been developed. Our group has reported that ERRγ acts as a catabolic regulator of cartilage degeneration and OA pathogenesis [7]. In addition, we have demonstrated that the inverse agonist of ERRγ, GSK5182, inhibits OA pathogenesis in a mouse model [7]. However, previous studies have reported limited information regarding the relationship between pro-inflammatory cytokines, ERRγ expression, and GSK5182. Therefore, we further investigated the relationship between pro-inflammatory cytokines and ERRγ expression, and GSK5182 function in the pro-inflammatory cytokine-mediated cartilage catabolism in the OA joint.
Previously, we failed to define whether pro-inflammatory cytokines induced ERRγ expression [7]. IL-1β, TNF-α, and IL-6 appear to be the central pro-inflammatory cytokines involved in OA pathophysiology [15,28,29]. It has been reported that IL-1β and TNF-α are elevated in the synovial fluid and synovial membrane, subchondral bone, and cartilage during OA [3,15,16]. These cytokines suppress type II collagen, proteoglycan, and aggrecan expression while stimulating MMP-1, MMP-3, and MMP-13 expression [17,[36][37][38][39][40]. Additionally, IL-6 levels are highly elevated in the synovial fluid and serum of patients with OA, with MMP functions [41]. The current study demonstrated that the mRNA and protein ERRγ levels were significantly increased in chondrocytes exposed to IL-1β, IL-6, and TNF-α (Figures 1 and 2). Notably, OA pathogenesis is mediated by an imbalance between anabolic and catabolic factors. OA-causing primary mechanisms include mechanical stresses (joint instability and injury), which induce the activation of biochemical pathways in chondrocytes, resulting in a loss of ECM by matrix metalloproteinases (MMPs) and aggrecanases (ADAMTSs). MMP-3, MMP-13, and ADAMTS5 are known to play crucial roles in OA cartilage destruction [12][13][14]. Our study showed that ERRγ overexpression via transduction with Ad-Esrrg dramatically elevated the mRNA or protein levels of MMP-3 and MMP-13 ( Figure 2D-F). These results suggest that in articular chondrocytes, pro-inflammatory cytokines, ERRγ, and MMPs are closely associated. Based on our previous report and other available evidence, ERRγ might directly regulate MMP transcription [7,42].
We further analyzed the genetic function of ERRγ in vivo in OA pathogenesis systems. ERRγ (NR3B3, Esrrg) is one of the ERR isoforms (ERRα; NR3B1, Esrra), (ERRβ; NR3B2, Esrrb), which was first identified as an orphan nuclear receptor [21]. ERRs are closely related to the ER without binding to the ER ligand but share high homology in their DNA-binding domain [21]. ERRα positively regulates osteoblast differentiation and bone formation [21], but ERRγ has demonstrated the opposite function [33,43]. ERRs have functions in chondrocytes and OA. For example, ERRα plays dual roles in OA chondrocytes; ERRα increases pro-chondrogenic factor (SOX9) and cartilage-degenerative factor (MMP-13) in response to pro-inflammatory factors [7,42,44]. Our animal studies demonstrated that ERRγ overexpression by either adenovirus delivery system (Ad-Esrrg) or cartilage-specific ERRγ Tg mice (Col2a1-Esrrg) enhanced cartilage degeneration, osteophyte formation, and subchondral bone sclerosis, which are hallmarks of OA pathogenesis ( Figure 3A,B). In contrast, an opposite phenomenon was observed in ERRγ-knockout (KO) mice ( Figure 3C). This result indicated that ERRγ is a crucial mediator in OA pathogenesis. Finally, we focused on whether the ERRγ inverse agonist GSK5182 blocks pro-inflammation mediated MMP-3 and MMP-13 expression as well as ERRγ expression. The ERRγ binding activity for synthetic ligands diethylstilbestrol (DES), tamoxifen (TAM), and 4-hydroxytamoxifen (4-OHT) was evaluated and 4-OHT was shown to be the most specific and had high binding affinity with ERRγ at the micro-molar level [45]. Synthetic GSK5182, a 4-OHT analog, was developed to exhibit higher affinity (IC50 = 79 nM) for ERRγ [46] and was shown to regulate the transcriptional activity of ERRγ [47,48]. As expected, treatment with GSK5182 dramatically inhibited IL-1β, IL-6-, or TNF-α induced MMP-3 and MMP-13 expression in primary cultured chondrocytes ( Figure 4A-C). Notably, GSK5182 unquestionably inhibited ERRγ overexpression-mediated MMP-3 and MMP-13 expression as well as ERRγ expression ( Figure 4D). It could be postulated that GSK5182 regulates transcription of ERRγ as well as post-transcriptional regulation [47][48][49]. Following GSK5182 delivery to the mouse knee joint, protection against ERRγ-induced cartilage degeneration was observed ( Figure 4E). Treatment with tamoxifen induced cancellous bone and longitudinal growth, while treatment with GSK5182 reduced DMM-induced bone remodeling including cartilage destruction, osteophyte development, and subchondral bone sclerosis [7,50]. Data from cellular and animal studies revealed that GSK5182 is a potential therapeutic drug for OA by blocking inflammatory pathways.

Chemicals and Laboratory Ware
Unless specified otherwise, all chemicals and laboratory wares were purchased from Sigma Chemical Co., (St. Louis, MO, USA) and Falcon Labware (Becton-Dickinson, Franklin Lakes, NJ, USA), respectively. Dulbecco's modified Eagle's medium (DMEM) and fetal bovine serum (FBS) were purchased from Gibco Co., (Gibco BRL, New York, NY, USA).

Reverse Transcription-Polymerase Chain Reaction (RT-PCR)
Total RNA was extracted from primary cultured chondrocytes using TRI reagent (Molecular Research Center Inc., Cincinnati, OH, USA). The quality and concentration of RNA were evaluated using a NanoDrop™ 2000 Spectrophotometer (Thermo Scientific, Waltham, MA, USA). The RNA was reverse transcribed, and the resulting cDNA was amplified by PCR or CFX96™ Real-Time System (BIO-RAD, Hercules, CA, USA,in Bio-Health Materials Core-Facility, Jeju National University) using SYBR premix ExTaq reagents (TaKaRa Bio, Mountain View, CA, USA). The PCR primers and experimental conditions are summarized in Table 1. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as internal control.

Histology
Mouse knee joints presenting with experimental OA were fixed in 4% paraformaldehyde, decalcified in 0.5 M EDTA, and embedded in paraffin. The paraffin blocks were sectioned at a thickness of 5 µm, and sections were deparaffinized in xylene, hydrated with graded ethanol, and stained with safranin O. Cartilage destruction was scored by five observers under blinded conditions using the OARSI scoring system (grades 0-6) [7,18,20,54] The results of OARSI grade scoring represent the mean of the maximum score in each mouse, and the representative safranin O-stained image was selected from the most advanced lesion among serial sections.

Statistical Analysis
All statistical analyses were performed using IBM SPSS Statistics 21 (IBM Corp., Armonk, NY, USA). Data from the cell-based in vitro assays were evaluated using two-tailed Student's t-tests with unequal sample sizes and variances and two-way analysis of variance (ANOVA) with post-hoc tests (LSD) for pairwise comparisons and multi-comparisons, respectively. Data collected from the mouse experiments were analyzed using the non-parametric Mann-Whitney U test. Data distribution was evaluated for normalcy using the Shapiro-Wilk test. Herein, "n" indicates the number of independent experiments or mice. Significance was accepted at the 0.05 level of probability (p < 0.05).

Conflicts of Interest:
The authors declare no conflict of interest.