Biological Activities of Phosphocitrate: A Potential Meniscal Protective Agent

Phosphocitrate (PC) inhibited meniscal calcification and the development of calcium crystal-associated osteoarthritis (OA) in Hartley guinea pigs. However, the mechanisms remain elusive. This study sought to examine the biological activities of PC in the absence of calcium crystals and test the hypothesis that PC is potentially a meniscal protective agent. We found that PC downregulated the expression of many genes classified in cell proliferation, ossification, prostaglandin metabolic process, and wound healing, including bloom syndrome RecQ helicase-like, cell division cycle 7 homolog, cell division cycle 25 homolog C, ankylosis progressive homolog, prostaglandin-endoperoxide synthases-1/cyclooxygenase-1, and plasminogen activator urokinase receptor. In contrast, PC stimulated the expression of many genes classified in fibroblast growth factor receptor signaling pathway, collagen fibril organization, and extracellular structure organization, including fibroblast growth factor 7, collagen type I, alpha 1, and collagen type XI, alpha 1. Consistent with its effect on the expression of genes classified in cell proliferation, collagen fibril organization, and ossification, PC inhibited the proliferation of OA meniscal cells and meniscal cell-mediated calcification while stimulating the production of collagens. These findings indicate that PC is potentially a meniscal-protective agent and a disease-modifying drug for arthritis associated with severe meniscal degeneration.


Introduction
Osteoarthritis (OA) is one of the most prevalent causes of disability in the aging population and has enormous economic and social consequences. However, existing nonsurgical treatment options only provide symptomatic relief but have no effect on the progression of the underlying disease or cartilage degeneration. The lack of progress in the development of disease-modifying drugs for OA therapy is largely due to our limited understanding of the pathogenesis of OA and our insufficient knowledge about the molecular targets for OA therapy.
Knee OA is not merely an articular cartilage disease, but a disease of the whole joint. An important local factor is the structural integrity of the menisci. In recent years, there has been a dramatic advance in our understanding of the integral role of the menisci for knee function and the consequences of meniscal abnormality on the development of OA. Studies have found that meniscal degeneration is a general feature of knee OA and contributes to joint space narrowing [1,2]. Meniscal lesions at baseline were more common in knees that developed OA than in the knees that did not develop OA during a 30-month follow-up period [3]. OA meniscal cells displayed a distinct gene expression profile different from normal meniscal cells [4]. These observations indicate that the meniscus is not a passive bystander in the disease process of knee OA [5,6].
Basic calcium phosphate crystal and calcium pyrophosphate dihydrate crystal are the two most common articular calcium crystals. The presence of these crystals in OA articular cartilage and synovial fluid is well recognized. These crystals are also present in knee menisci of patients with endstage OA [7,8]. Studies found that these crystals stimulate cell mitogenesis, cell endocytotic activity, and the production of matrix metalloproteinases (MMPs) and inflammatory cytokines including interleukin-1 (IL-1) and prostaglandinendoperoxide synthase 2/cyclooxygenase-2 (PTGS2/Cox-2) [9][10][11]. However, there is still controversy as to whether these crystals are causative factors, factors that exacerbate OA, or simply bystanders.
Phosphocitrate (PC) is a naturally occurring compound originally identified in rat liver mitochondrial extracts and crab hepatopancreas [12]. Moro et al. and Romanello et al. suggested that PC could be formed in vivo through cytosolic phosphorylation of citric acid [13,14], which explains why it is nontoxic. Since its original identification, PC has been shown to be a powerful calcification inhibitor [15,16]. Tew et al. speculated that PC prevented CaPO 4 precipitation in cells or cellular compartments containing high concentration of Ca 2+ and PO 4 [12]. PC prevented soft tissue calcification in vivo and did not produce any significant toxic side effect in rats when administered through intraperitoneal injections in doses up to 150 mol/kg/day [17]. In cell cultures, PC inhibits calcium crystal-induced mitogenesis, expression of MMPs, and crystal-induced cell death [18][19][20]. In the Hartley guinea pig model of crystal-associated OA, PC inhibited meniscal calcification and reduced the severity of cartilage degeneration [21]. These observations provide support for the notion that crystals may play an important role in the development of OA and that calcification inhibitors are potentially disease-modifying drugs for crystal-associated OA therapy. However, two bisphosphonates, which are potent calcification inhibitors, failed to inhibit the development of OA in animal models of OA [22,23], raising doubts as to whether calcification inhibitors are disease-modifying drugs for crystalassociated OA as well as the exact role of calcium crystals in the development of OA. In this study, we sought to examine the biological activities of PC in the absence of calcium crystals and test the hypothesis that PC has unique crystalindependent biological activities which may be responsible, at least in part, for its disease-modifying activity on OA and that PC is potentially a meniscal protective agent.

Materials and Methods
Dulbecco's Modified Eagle Medium, StemPro chondrogenesis differentiation medium, fetal bovine serum, Hank's balanced salt solution, and stock antibiotic and antimycotic mixture were products of Invitrogen (Carlsbad, CA, USA). PC was synthesized according to the procedure described [24]. All other chemicals are purchased from Sigma (St. Louis, MO, USA).

Cells.
OA meniscal cells were prepared from menisci derived from patients with end-stage OA. Briefly, the medial menisci derived from patients with end-stage OA were processed to remove fatty and synovial tissues, minced into small pieces, and cultured in 100 mm plates at 37 ∘ C in medium containing 0.5% antibiotic/antimycotic solution and 10% serum. Every three or four days, the culture medium was changed. When the cells reached 70% confluence, they were passaged and maintained in medium containing 10% serum. Human foreskin fibroblasts were obtained from American Type Culture Collection (CRL-2429, Manassas, VA, USA). OA meniscal cells prepared from three OA patients were used in this study. OA menisci were collected with the approval of the authors' Institutional Review Board from OA patients undergoing knee joint replacement surgery. The need for informed consent was waived because those menisci were surgical waste, and no private patient information was collected.

Cell
Culture and RNA Extraction. OA meniscal cells derived from three OA patients were harvested from cell culture plates and mixed and replated in four 100 mm cell culture plates at 90% confluence. On the second day, medium containing 1% serum was added. Twenty-four hours later, the medium in two plates was replaced with medium containing 1% serum and PC (1 mM), and the medium in the other two plates was replaced with medium containing 1% serum without PC. Twenty-four hours later, total RNA was extracted from these cells using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and purified using Oligotex kit (Qiagen, Valencia, CA, USA). We repeated the experiment twice. Microarray was performed using these RNA samples (total six RNA samples).

2.3.
Microarray. RNA samples extracted from three independent experiments were used for microarray analysis experiments. Briefly, double-stranded DNA was synthesized using SuperScript Double-Stranded cDNA Synthesis Kit (Invitrogen, San Diego, CA, USA). The DNA product was purified using GeneChip Sample Cleanup Module (Affymetrix, Santa Clara, CA, USA). cRNA was synthesized and biotin-labeled using BioArray high yield RNA transcript labeling kit (Enzo Life Sciences, Farmingdale, NY, USA). The cRNA product was purified using GeneChip Sample Cleanup Module and subsequently chemically fragmented. The fragmented and biotinylated cRNA was hybridized to HG-U133 Plus 2 GeneChip using Affymetrix Fluidics Station 400 (Affymetrix, Santa Clara, CA, USA). The fluorescent signals were quantified during two scans by Agilent Gene Array Scanner G2500A (Agilent Technologies, Palo Alto, CA, USA) and GeneChip operating Software (Affymetrix, Santa Clara, CA, USA). Genesifter (VizX Labs, Seattle, WA, USA) was used for the analysis of differential gene expression and gene ontology.

Real-Time RT-PCR.
After microarray analyses, we mixed the RNA samples extracted from PC-treated OA meniscal cells (PC-treated RNA sample) and the RNA samples extracted from untreated OA meniscal cells (untreated RNA sample) and performed RT-PCR experiments. Briefly, cDNA was synthesized using TaqMan Reverse Transcription Reagents (Applied Biosystems, University Park, IL, USA) using the RNA samples described. Quantification of relative transcript levels for selected genes and the housekeeping gene glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was performed using ABI7000 Real-Time PCR system (Applied Biosystems, University Park, IL, USA). TaqMan Gene Expression assay (Applied Biosystems, University Park, IL, USA) was used. CDNA samples were amplified with an initial Taq BioMed Research International 3 DNA polymerase activation step at 95 ∘ C for 10 minutes, followed by 40 cycles of denaturation at 95 ∘ C for 15 seconds and annealing at 60 ∘ C for one minute. Fold change was calculated, and the expression level of the genes to be examined was normalized to the expression level of GAPDH. RT-PCR experiment was performed in triplicates using the same RNA sample.

Cell
Proliferation. OA meniscal cells (2 × 10 4 ) were plated in six well cluster plates and cultured in medium containing 10% serum in the presence of increasing amounts of PC (triplicates). Medium was changed every three days until the OA meniscal cells in the wells without PC reached 85% confluence. Cells were then harvested and cell numbers were determined using a hemocytometer. This experiment was repeated 3 times using OA meniscal cells derived from different patients. This proliferation assay was also performed using human primary foreskin fibroblasts.
2.6. Micromass Culture and Histology. OA meniscal cells were harvested from several 100 mm culture plates and suspended in medium containing 10% serum. For preparing a micromass, a droplet of the cell suspension containing 6 × 10 6 cells was placed in a well of a 24-well plate. After placing all droplets, the plate was incubated for 4 hours at 37 ∘ C in a tissue culture incubator. These micromasses were then fed with StemPro chondrogenesis differentiation medium with PC (1 mM) or without PC every three days throughout the experiment for 14 days. Each well was then rinsed twice with 500 L of Hank's balanced salt solution, and two drops of eosin were added to these wells. Five minutes later, eosin was aspirated off and the micromasses were transferred to a strip of filter paper which sat on the top of an ethanol-soaked sponge within a plastic cassette. The cassettes sat in a 10% formalin solution for one hour. These micromasses then underwent routine paraffin embedding. Sections were cut at 5 m thick and stained with picrosirius red for collagens, alcian blue for proteoglycans, and alizarin red for calcium deposits.

Cell-Mediated Calcification in Monolayer
Culture. OA meniscal cells were plated in twenty-four well plates at 90% confluence. The next day, medium was replaced with StemPro chondrogenesis differentiation medium containing 1 mM adenosine-5 -triphosphate (ATP) in the absence or presence of PC (1 mM). The cells were cultured for 14 days and fed with StemPro chondrogenesis differentiation medium containing ATP every three or four days throughout the experiment. At the end of the experimental period, media were removed. Calcification was examined using alizarin red.

Statistical Analysis.
For cell proliferation, data are expressed as the mean ± SD, and the difference between two groups was analyzed using Student's t-test. For realtime RT-PCR, experiment was repeated in triplicates. The difference between two experimental groups was analyzed using Student's test. In all cases, P values less than 0.01 were considered significant. Statistical analysis was performed using the SAS software, version 9.3.

Effect of PC on Gene Expression.
We performed three microarray experiments (PC-treated RNA sample I and untreated RNA sample I; PC-treated RNA sample II and untreated RNA sample II; PC-treated RNA sample III and untreated RNA sample III) as described. The results of the three microarray experiments were quite similar. The results of the first microarray experiment showed that of the more than 50,000 transcripts, 2445 transcripts displayed significant differential expression (more than 1.6-fold changes) in the PC-treated OA meniscal cells compared with the untreated OA meniscal cells. A total of 1795 transcripts displayed decreased expression and 650 transcripts displayed increased expression. The genes that fell into specific biological processes previously implicated in OA, or suspected to have a role in OA are listed in Tables 1 and 2.
As shown in Table 1, the expression of numerous genes classified in cell proliferation was downregulated by PC. Of the 62 differentially expressed genes (more than 2-fold changes) classified in cell proliferation, the expression of 52 genes, including bloom syndrome recq helicase-like (BLM; −7.41-fold change), lymphoid-specific helicase (Hells; −4.46fold change), cell division cycle 7 homolog (CDC7; −4.11fold change), CDC 25 homolog C (CDC25C; −2.36-fold change), and cyclin E2 (CCNE2, −3.51-fold change), was downregulated by PC. This specific downregulatory effect of PC on the expression of genes classified in cell proliferation suggests that PC may inhibit the proliferation of OA meniscal cells.

Real-Time RT-PCR.
Real-time RT-PCR was used to confirm expression of selected genes. The results are listed in Table 3. As shown, the differential expression of the genes examined was confirmed by real-time RT-PCR ( < 0.01).

PC Inhibited Proliferation of OA Meniscal Cells.
The specific downregulatory effect of PC on the expression of genes classified in cell proliferation suggests that PC may inhibit the proliferation of OA meniscal cells. To test this, we cultured OA meniscal cells in the absence and presence of PC for 9 days and then determined the cell number using a hemocytometer. Indeed, we found that PC inhibited the proliferation of OA meniscal cells (Figure 1(a), right bar  group). There were 55% fewer OA meniscal cells in the PC-treated wells than that in the untreated wells ( < 0.01). In contrast, PC had no effect on the proliferation of human foreskin fibroblasts (Figure 1(a), left bar group). For comparison, we also examined the effect of disodium salt of ethane-1-hydroxy-1, 1-bisphosphonic acid (EHDP), which is a bisphosphonate, together with PC on the proliferation of OA meniscal cells. As shown in Figure 1(b), both PC and EHDP inhibited the proliferation of OA meniscal cells in a dose-dependent manner. In addition, the morphology of PCtreated OA meniscal cells and untreated OA meniscal cells was similar (not shown), indicating that the reduction in cell number was not due to cellular toxicity of PC.

PC Stimulated Production of Collagens.
The upregulatory effect of PC on the expression of genes classified in collagen fibril organization, including COL1A1 (1.62-fold change), COL11A1 (1.82-fold change), COL14A1 (2.06-fold change), COL5A1 (1.52-fold change), and COL5A2 (1.44-fold change), suggests that PC may stimulate the production of collagens by the OA meniscal cells. To examine this, we prepared micromasses of OA meniscal cells (triplicates) and examined the production of collagens using picrosirius red staining. The results confirmed that PC stimulated the production of collagens by OA meniscal cells. Representative images of picrosirius red staining are shown in Figure 2. As shown, the picrosirius red staining in the PC-treated micromasses ( Figure 2(b)) was stronger than that in the untreated micromasses (Figure 2(a)). These micromasses were also stained with alcian blue. Results indicated that PC had no detectable effect of the production of proteoglycans (not shown).

PC Inhibited Meniscal Cell-Mediated Calcification.
These micromasses were also stained with alizarin red for calcium deposits. Representative images of alizarin red staining are shown in Figure 3. As shown, many calcium deposits were detected in the micromasses of OA meniscal cells cultured in the absence of PC (Figure 3(a)), but no calcium deposits were detected in the micromasses cultured in the presence of PC (Figure 3(b)). PC treatment (1 mM) abolished OA meniscal cell-mediated calcium deposition in micromass culture completely.
The inhibitory effect of PC on OA meniscal cell-mediated calcification was also examined using monolayer culture. As shown in Figure 4, calcium deposits were detected in the monolayer culture of OA meniscal cells (Figure 4(a), but no calcium deposits were detected in the monolayer culture of OA meniscal cells cultured in the presence of PC (Figure 4(b)).

Discussion
Increased number and size of cell clusters are the hallmark histological feature of OA articular cartilage [25,26]. Cell clusters are activated cells and represent an important source of pathological mediator. The number of proliferating chondrocytes increases during OA progression [26,27]. Cell clusters are also detected in OA menisci [28]. These observations indicate that abnormal cell activation and proliferation may play a role in the development and/or progression of OA. In this study, we demonstrated that PC downregulated the expression of numerous genes associated with cell proliferation. Consistent with this regulatory effect, PC inhibited the proliferation of OA meniscal cells. Although the exact implication of this effect on the development of OA is unclear at present, these findings suggest that PC may inhibit the activation of OA meniscal cells or inhibit the formation of cell clusters.
Studies consistently reported apoptotic cells in cell clusters [29,30]. One of the mechanisms of apoptosis is abnormal hypertrophic differentiation of chondrocytes and subsequent pathological calcification. There is abundant evidence indicating the expression of hypertrophic makers, including collagen X, osteocalcin, and MMP13, within the cell clusters [31][32][33]. In addition, staining of OA cartilage sections showed colocalization of calcium deposits and clusters containing apoptotic cells adjacent to the calcium deposits [34]. Mineralization of extracellular matrix surrounding cell clusters was also detected in OA menisci [28]. These observations indicate that abnormal cell activation, terminal differentiation, pathological calcification, and apoptosis may be mutually linked disease processes in OA. In this study, we demonstrated that PC inhibited the expression of numerous genes associated with cell proliferation, differentiation, and pathological calcification, including CDC7 (−4.11-fold change), CDC25C (−2.36-fold change), COL13A1 (−3.39fold change), SATB2 (−3.13-fold-change), SMAD1 (−2.19-fold change), ENPP1 (−2.02-fold change), and ANKH (−1.60-fold change). Consistent with this regulatory activity, PC inhibited the proliferation of OA meniscal cells and abolished meniscal cells-mediated calcification. These findings suggest that PC may exert its disease-modifying activity in part by targeting the genes associated with cell proliferation, differentiation, and pathological calcification. Several previous observations provide support for this potential mechanism. For example, studies found that COL13 was increased in OA articular cartilage and that transgenic mice overexpressing COL13 had abnormally high bone mineral density [35,36]. The expression of ENPP1 and ANKH was elevated in OA meniscal cells, menisci, or articular cartilage [4,34,37]. SMAD1 induced terminal differentiation of chondrocytes and promoted calcification [38]. SATB2 stimulated osteogenic differentiation of adult stem cells [39]. These previous findings demonstrated that elevated expressions of COL13, ENPP1, ANKH, SMAD1, and SATB2 in OA cartilage or OA menisci were associated with terminal differentiation and biomineralization. The downregulatory effect of PC on these genes indicates that PC may reverse the processes of terminal differentiation and biomineralization.
Early studies demonstrated that MGP was a calcification inhibitor and that inhibition of IGFBP5 breakdown reduced articular cartilage loss in an experimental OA [40,41]. In this study, we demonstrated that PC upregulated the expression of MGP (1.61-fold change) and IGFBP5 (1.78-fold change). This upregulatory effect of PC may lead to elevated production of MGP and IGFBP5.
FGF signaling pathway plays an important role in the regulation of osteogenesis and chondrogenesis. Studies found that FGF18 induced chondrocyte hypertrophy and mineralization [42,43], while FGF9 inhibited terminal differentiation of calvaria-derived cells and mineralization [44,45]. FGF7 was a potent inhibitor of phosphate transport [46]. In this study, we found that PC downregulated the expression of FGF18 (−1.73-fold change), while it upregulated the expression of FGF9 (2.38-fold change) and FGF7 (1.83-fold change). These regulatory effects indicate that PC may inhibit OA meniscal cell differentiation and pathological calcification in part by modulating the FGF signaling pathway. Consistently, studies have demonstrated that constitutive expression of thrombospondin 1 (THBS1), a protein involved in FGF signaling pathway, inhibited biomineralization and that THBS1 gene therapy suppressed the progression of arthritis in a rat model of OA [47,48]. Another study demonstrated that the absence of signaling through FGF receptor 3 (FGFR3) leads to premature cartilage degeneration and early arthritis [49]. These previous findings indicate a key role of FGF signaling pathway in the development of healthy articular cartilage. The specific upregulatory activity of PC on THBS1 (1.68-fold change) and FGF9 (a preferred ligand for FGFR3) indicates that PC may exert its disease-modifying activity on OA in part by modulating the FGF signaling pathway.
Severe loss of collagen occurs in OA menisci [50]. Disease-modifying drugs targeting meniscal degeneration for OA therapy should inhibit meniscal collagen loss. In this study, we demonstrated that PC stimulated the production of collagen by OA meniscal cells. This finding suggests that PC is not only potentially a disease-modifying drug for OA therapy but also potentially a meniscal protective agent.
Interestingly, many genes classified in the inflammatory response were upregulated by PC, including adenosine A1 receptor ADORA1 (2.48-fold change), CD24 (2.09-fold change), and BCL6 (1.98-fold change). It is worth noting that previous studies demonstrated that the activation of adenosine receptors reduced inflammation and joint destruction in a rat adjuvant-induced arthritis, CD24 repressed tissue damage-induced immune response, and that BCL6 inhibited the expression of chemokines and attenuated allergic airway inflammation [54][55][56]. These previous findings indicate that ADORA1, CD24, and BCL6 are potentially antiinflammation molecules. The upregulation of these genes by PC is consistent with the notion that PC is potentially an antiinflammation agent [57].

Conclusions
These specific biological activities of PC are intrinsic properties of PC and are not dependent on the presence of calcium crystals. PC is potentially an anti-inflammation and meniscal protective agent. PC is not only potentially a diseasemodifying drug for crystal-associated OA but also potentially a disease-modifying drug for arthritis associated with severe meniscal degeneration. The findings presented in this study provide further support for the development of PC, and/or its analogues, as disease-modifying drugs for OA therapy.