Kartogenin mediates cartilage regeneration via stimulating the IL-6/Stat3-dependent proliferation of cartilage stem/progenitor cells

Background: Declination of endogenous stem cells in cartilage is regarded as the cause of cartilage degeneration. Kartogenin (KGN) is well known to play an important role in chondrogenesis of mesenchymal stem cells (MSC). Methods: Using MSCs isolated from rat cartilages, we analyzed the changing of transcriptomics after the treatment of KGN in vitro . DMM animal models were then applied to identified the effect of MSCs proliferation in vivo after KGN capsule injection. Furthermore, we explored the potential mechanisms how KGN mediates cartilage regeneration and proliferation of cartilage progenitor cells. Results: In this study, we demonstrate that KGN can promote the proliferation of MSC from cartilage, respectively. The percentage of G2-M phase cells in culture reached over 10%, nearly twice as the control group with KGN treatment. Transcriptomic profiling of rat cartilage stem/progenitor cells (CSPC) revealed that the expression of at least 20 cell cycle related genes was significantly changed in response to the KGN treatment. IL-6 and its co-receptor Gp130 gene expression level are much higher than the untreated control. The phosphorylation of the IL-6-downstreamt molecular Stat-3 was enhanced upon the KGN stimulation. The knee joint injury animal model further showed the increased articular cartilage thickness after KGN treatment. Interestingly, the IHC staining also demonstrated the up-regulated level of Stat-3 phosphorylation and enhanced distribution of CD44 + /CD105 + cells in the KGN-treated cartilage. Conclusion: Taken together, our data suggest that KGN promotes cartilage regeneration at least partially by stimulating the IL-6/Stat3-dependent proliferation of stem cells resident in the cartilage.

Introduction levels eventually lead to characterized loss of articular cartilage, osteophyte formation, thickening of subchondral bone and subchondral cyst formation. [3] Tissue resident stem cells/progenitor cells are essential in maintaining homeostasis of all tissue types and are susceptible to disease-causing factors such as aging and inflammatory factors. [11] Cartilage stem/progenitor cells (CSPCs) have been found within human articular cartilage [12][13][14] and proposed to maintain cartilage homeostasis, including the cell number and their anabolic metabolisms. These CSPCs can respond to injury and migrate to the injured tissue area to achieve minimal cartilage repair, and are characterized as stem/progenitor cells due to their self-renewal ability, multiple differentiation potential and stem cell markers. Some mesenchymal progenitor cell markers, including Notch-1, VCAM (CD166), STRO-1, were expressed in early OA, but less expressed in normal human articular cartilage, suggesting that OA may activate the endogenous CSPCs in the tissue. [15] At the middle to late phases of OA, however, the functionality and number of cartilage progenitors typically decline and may be attributed for the abnormality in pathologies of OA in human and in murine models. [15,16] Recently, the small molecule Kartogenin has been proposed to be able to promote the multipotent mesenchymal stem cells (MSC) into chondrocytes by regulating the CBFb-RUNX1 transcriptional program. [17] Since then, the established capability of KGN in mediating cartilage regeneration both in vitro and in vivo have been documented in a number of studies using different models. [18][19][20][21] During the process of mouse limb development, KGN stimulated cartilage nodule formation by up-regulating genes encoding hedgehog and TGFβ. [18] For the purpose of integration with regenerative medicine approaches, KGN has been applied in a number of studies involved in biomaterials and scaffolds for full-thickness cartilage repair.
[19] An intra-articular delivery system contained KGN nano-or micro-particles was applied in a surgically-induced OA model in rats, where KGN could induce the chondrogenic differentiation of loaded bone marrow MSC.
[20] A pilot study in a rodent model of osteoarthritis using micro-CT as a main evaluation mean showed that the KGN treatment not only reduced the conversion rate of cartilage and bone, but also prevented the change of subchondral bone. [21] For the action of KGN on the cartilage regeneration, current evidences mainly indicate the differentiation of endogenous MSC within the cartilage by regulating the CBFb-RUNX1 transcriptional program and up-regulating the TGFβ/Smad2/3 pathway. In other experimental sittings, however, KGN may also function in promoting differentiation of MSC towards intervertebral disc cell types or osteogenic lineages. [

Isolation and In Vitro expansion of CSPCs
Fibronectin (FN, 10 mg/ml) (Gene Operation, 10-50-1101) in PBS was used to coat culture plate overnight at 4 degree for cell transplant. Rats were killed by overdose of chloral hydrate, and separated the joints and intercepted the joint part (keep the upper and lower femur and tibia about 1-2 cm) into the 50 ml Falcon tube containing PBS. The cartilage tissue scissors are about the size of 1-3mm fragments as shown in Figure 1A, transferred to a 15 ml centrifuge tube, adding 3 times volume of the amount of 0.25% trypsin, digest it at 37 degree for 30 min. Following complete medium termination of digestion, adding 300unit/ml type II collagenase (Worthington, 42B13253) dissolved in supernatant for overnight. The supernatant was removed and the complete medium containing 1% gentamicin was suspended and counted. The cell number was adjusted to 1×10 4 /ml, and then transplanted into the FN covered culture plate and then separated cells can be observed on the surface of culture dish as shown in left panel of Figure 1B. After 2-3 days, replace fresh culture DF12 medium. On the day 7, cell colonies can be identified as shown in right panel of Figure 1B.

Growth and proliferation assay
After 72 hours of cell culture, cells were counted and collected with cold ethanol solution at 4 degree for overnight. Cells were finally suspended with PBST (containing a TritonX-100 of 0.2% PBS) by adding a 50 μg/ml RNA enzyme (RNase A) at 37 degree for 30-minute incubation. 50 μg/ml PI was added to cell suspension and incubated for 15 mins. After centrifuge above cells at 800 rpm for 5min, and then discard supernatant. Re-centrifugate and remove excess supernatant, afterwards using PBST to suspense cells and transfer to the flow tube before loading to BD FACS Calibur.

Flow cytometry
To characterize the CSPC identity, positive (CD73, CD105 and CD90) and negative (CD45 and CD34) cell-surface markers were detected by flow cytometry. All antibodies are listed

RNA-sequencing and bioinformatics analysis
One million of CSPCs were seeded in T175 bottle and treated with 10 μM KGN for 3 days.
Then, total RNA from CSPCs was extracted using TRIzol reagent (Thermo, 15596-026), according to the manufacturer's instructions. RNA-sequencing and bioinformatics analyses were performed as previously described. [11] IL-6 ELISA assay of rat CSPC in culture IL-6 secretion in culture supernatants were detected with ELISA kit (eBioscience, San Diego, CA) according to the manufacturer's directions. Briefly, ELISA plates were coated with the capture antibody and incubated overnight at 4°C. Then, washed wells thoroughly and blocked for 1 h. Wells were incubated for 2 hrs with human IL-6 standards and supernatants from CSPC with or without KGN medium, which were diluted 100 times. After extensive wash steps, wells were incubated with detection antibody for 2 hrs at 4°C and then washed several times. Wells were incubated with Avidin-HRP for 30 min, washed thoroughly, incubated with substrate solution for 15 min, and then added stop solution.
Finally, A450 values were measured using a microplate reader.

RT-qPCR
5% Agarose gel with identical wells were casted and rat-CSPC were seeded at a density of 1×10 6 /10 μL. Cells were cultured in chondrogenic medium with either 10µM KGN or its solvent, DMSO as control. Medium were changed every other day for 28 days before harvested. Total RNA was extracted with TRIzol reagent (Life Technologies) following manufacture's instruction. 500ng RNA was reverse-transcribed to cDNA using PrimeScript TM RT reagent Kit (TaKaRa, Dalian, China). Specific genes were amplified with One Step SYBR PrimeScript RT-PCR Kit II (TaKaRa) on CFX96 real-time PCR detection system (BioRad, Hercules, CA, USA). The relative gene expression was normalized to GAPDH transcript level and calculated using the 2-ΔΔCt method.

STAT-3 siRNA knockdown assay
Rat-CSPCs were seeded in 6-well plate at 1×10 5 per well. Serum-free medium was given 2 hours pre-transfection. We designed a STAT-3 siRNA oligonucleotide

BrdU administration
The animals were exposed to 1 mg/mL BrdU (Sigma-Aldrich, Steinheim, Germany) for 15 days, which is effective and safe for label-retaining cell studies. [26] Oral BrdU administration was combined with exercise acclimatization, and rats were killed with excessive sodium pentobarbital (Apoteket Produktion & Laboratorier AB (APL), Sweden) at 30, 60 and 90 days after the start point. Two animals, not exposed to BrdU, were killed at day 30 and 90 as negative controls for BrdU technique.

Histology and immunohistochemistry
The knee joint samples were cut vertically at the cartilage defect sites with a thickness of immunochemistry. The images were captured with a microscope (Olympus BX51, Japan). The destabilization of medial meniscus (DMM) model in rats was created according to the previously reported protocol. [26] Briefly, male rats weighted at 100 ± 10 g were anesthetized with 2% isoflurane in air using an anesthesia machine (RWD Life Science Co., Ltd., Shenzhen, China). After opening the capsule of knee joint, the tibial ligament of medial meniscus was cut off to destabilize the joint. Then, the knee joint capsule was closed immediately. Thirty days after surgery, 50μL KGN dissolved in saline at the concentration of 10 mM was injected into knee capsule every 7 days for seven weeks. Two weeks later, rats were executed followed by knee joint isolation for staining as shown in Figure 5A.

Statistical analysis
All data were expressed as the means ± SD. Differences between groups were evaluated by independent-sample t test or one-way analysis of variance (ANOVA) followed by Tukey's Multiple Comparison Test in SPSS 22.0 software. A p-value less than 0.05 was considered to be statistically significant.

Identification of primary CSPC original from rat cartilage tissues
To confirm stem cell characterization of CSPC isolated from cartilage tissue, we checked the stem cell markers including three positive markers CD105/CD90/CD73 and two negative markers CD34/CD45 of these cells. Our results showed that the proportion of CD105/CD90/CD73 positive cells are, respectively, 99.96%, 99.93% and 99.95%; the proportion of CD34/CD45 positive cells are 0.19% and 0.10% suggesting that our CSPC from cartilage tissue are mesenchymal stem cells as shown in Figure 1C.

KGN promotes rat cartilage derived from CSPCs proliferation In Vitro
In our study, 10 μM KGN containing media treated CSPC isolated from knee joint of rats aged at 4 weeks led to a significant cell number increase. Cell number counts were performed at day 3, day 5 and day 7 time points with constant KGN concentration. A sustained increase of the total cell number was achieved, as shown in Figure 2A. We applied the flow cytometry to check the percentage of G2-M stage cells corresponding to the increasing cell number. We found the percentage of G2-M stage cells were 8.6%, 9.8% and 15.6% respectively at the end of day3, day5 and day7 as shown in Figure 2B, C.
Compared to control groups, both the cell number and the percentage of G2-M of KGN treated group had a significant differences as shown in Figure 2B. At the end of cell culture (day 7), we exanimated not only the percentage of G2-M for both control and KGN treated group but also the percentage of apoptotic cells induced by IL-1β. It was found that the KGN treatment results in the percentage of cell apoptosis is 15.17% which is lower than control treated with IL-1β groups shown in Figure 2D. This means KGN is able to protect cells from the IL-1β cytotoxicity. Taken together, these results shown that the KGN treatment resulted in an increasing of cell growth and a lower percentage of apoptosis.

IL-6 was identified as a key effector of KGN by total RNA sequencing
With the development of NGS, RNA-seq is becoming a powerful tool and has allowed great advances in the characterization and quantification of transcriptomes. With an attempt to reveal the possible mechanism underlying KGN effects on cell proliferation, the comprehensive transcriptomics studies were then carried out between the KGN-treated and the untreated CSPC derived from rat. As shown in Figure 3A, B, 738 genes were identified to be altered for their RNA expression profiling by using 1.5 as a threshold value in duplicate samples. A functional enrichment analysis and a protein-protein interaction network analysis were performed with bioinformatics, and IL-6 was identified as a key effector following the KGN treatment in vitro, as shown in Figure 3C. The value of log expression multiples of IL-6 between the KGN treatment and control groups reached 13, the highest hit in all genes as shown in Figure 3D. To further verify the elevation of the IL-6 expression by the KGN treatment, qRT-PCR was performed and the results matched the RNA-seq data, i.e. the mRNA level of IL-6 in CSPC was significantly higher than that of control cells (data not shown). IL-6 is one of secreted cytokines by MSC. [27] We first detected the level of IL-6 in the supernatant of KGN treated CSPC using an ELESA kit. The results show that the KGN-treated group contained significantly higher IL-6 level in the supernatant than that of control group, as shown in Figure 4A.
The KGN treatment results in the activation of IL-6/STAT3 pathway Activation of the TGFβ pathway by KGN has been reported by several groups during introduction of chondrogenic differentiation of MSC. [15,16] We therefore wondered whether the KGN effects on CSPC proliferation were TGFβ-signaling dependent. To address this question, SB431542, a reported dual inhibitor of activin/Nodal/TGFβ and BMP signaling pathways [28] was added to the experimental system. Interestingly, SB431542 was not able to block the IL-6 excessive secretion ( Figure 4A). At the time points of both day 3 and day 5, the KGN treatment resulted in an increasing of IL-6 secretion, which could only be partially inhibited by SB431542. Further, we were able to observe KGN effects on directly phosphating Smad2/3 without the external TGFβ addition, as shown in the column b in Figure 4B. It is interesting that the elevated pSmad2/3 level could be detected by adding KGN even in the presence of the TGFβ inhibitor SB431542, as shown in the column d, meaning that there should other pathways involved in the CSPC regulation by KGN. To investigate this possibility, we detected the Stat3 activation using P-Stat3 antibody and we found that the KGN treatment was able to increase the phosphorylation of Stat3, as shown in Figure 4C. Stat3 is well known to be activated by IL-6 via the IL-6/JAK/STAT3 pathway. [29] Interestingly, we found that KGN could also activate Stat3 ( Figure 4E).
Phosphorylation of Stat3 was significantly deduced when the IL-6 receptor gp130 was blocked with a gp130 antibody, as compared with either KGN or IL-6 addition alone ( Figure   4D, E). Taken all above together, KGN exhibited its role in activating the phosphorylation of Stat3 by increasing the secretion of IL-6.

KGN increases rat cartilage thickness by promoting the stem cells proliferation In Vivo
To investigate whether KGN affect CSPCs, we used BrdU labelling [26] to investigate proliferation and label-retaining cells (i.e. slow cycling cells) in vivo by ether feeding the rat BrdU containing water or injecting BrdU into the joint space, in combination with immunostaining of the multi-potential stem cell marker nucleostemin. The results showed that articular cartilages were thickened by intra-articular injection of KGN ( Figure 5B). The KGN injection group appeared to have in a thicker cartilage layer, implicating a potential pharmaceutical value. Interestingly, the BrdU-labeling experiments showed that slow cycling cells were often assembled as columnar arrangements, like "the string of pearlslike" array. The software ImageJ V1.45 (http://imagej.en.softonic.com/) was applied to quantify the BrdU positive cells, showing a significant increase of the CSPC (the lower right panel in Figure 5B). Nucleostemin, a nucleolar protein, was first identified in neural stem cells and a number of cancer cell lines. [30] The molecule is mainly expressed in primitive cells and down-regulated before terminal differentiation. It is considered as a stem cell marker of chondrogenic cells. The results here showed that BrdU positive cells were well overlapped with nucleostemin expression ( Figure 5C). Furthermore, it was found that the KGN injection could result in the up-regulation of P-Stat3 and P-Smad2/3 positive cells as shown with IHC ( Figure 5D), well matching the in vitro findings.

KGN shows a good ability for cartilage regeneration In Vivo in DMM models
The destabilization of the medial meniscus (DMM) surgical model is a commonly accepted model of OA. [31] Utilizing the model, this study finally tried to assess the therapeutic capacity of the KGN injection. Fifty microliters of KGN at a concentration of 10mM were injected into the knee joints of DMM animal models, twice a week for 8-10 weeks, as schemed in Figure 6A. The DMM surgery could induce obvious degeneration and defects in articular cartilage. The intra-articular injection of KGN mediated a repair of damaged cartilage as shown in Figure 6B. Around the adjacent non-damaged areas, a thicker cartilage layer with more toluidine blue positive cells were observed, presenting an increased number of chondrocytes as shown in left panel of Figure 6B. Both CD44 and CD105 are commonly accepted MSC biomarkers and have been used to identify cartilageresident stem cells. [12,13] This inspired us to examine whether cells expressing the two markers could be enhanced with the KGN stimulation. Indeed, around the repairing area within cartilage of the KGN injection group animals, CD44 + CD105 + double positive cells were found to be clearly more than other groups, even the intact group animals.
Therefore, this study further implicated the role of KGN in mediating cartilage repair via enhancing CSPC proliferation ( Figure 6C).

Discussion
Well-coordinated sequential proliferation and differentiation of CSPCs are the key cellular process for the cartilage regeneration and defect repair. This study provides evidences indicating that, in addition to the well-documented role in promoting chondrogenic differentiation, KGN also effects on CSPC proliferation as the essential step for cartilage regeneration. Interestingly, we also demonstrate that IL-6, a pro-inflammatory factor wellknown in mediating the OA pathology, may actually play a role in stimulating cartilage regeneration mediated with KGN.
A number of previous studies have shown that KGN can promote chondrogenic differentiation of MSC. [18][19][20][21] In line with these studies, this study also demonstrated that KGN activated Smad2/3 of in vitro cultured CSPC, implying that the TGFb signaling pathway, one of the most powerful inducers for chondrogenic differentiation, are activated. More importantly, this study demonstrates that KGN could promote CSPC into mitosis and lead to increased percentage of G2-M phase cells and cell counts. Here, we need to clarify the reason why we chose 72h as the time point to carry out RNA-Seq is that 72 hour after KGN treatment is the starting point of the cell proliferation phenotype displayed. Such results seemed in a way controversial because the basic concept would be that a single cell has to exit cell cycle for differentiation and it would be hard to interpret why a single stimuli could exert effects in both cell proliferation and differentiation.
However, it must be noted that MSC from any sources are highly heterogeneous and cells at different cell cycle phases or differentiation stages, and even different functional subgroups co-exist. As a result, the seemingly same or similar types of progenitor cells in vivo may contain different cells and respond differently to the same stimuli. Indeed, KGN has been proposed to exert effects on the proliferation of cells other than the cartilage origin. [25,32] Interestingly, TGFβ signaling inhibition was not able to inhibit the cell proliferation phenotype, further highlighting the involvement of mechanisms other than the TGFβ signaling pathway. These evidences may imply that the mixed cell subgroups co- be essential for tissue regeneration in several organs, such as hematopoiesis, liver cells, gastrointestinal tract mucosa, and muscle cells. Impaired IL-6 function in some of these organs is associated with reduced regeneration following injury, whereas IL-6 upregulation during chronic inflammation may in contrast contribute to the dysfunction of the same organs. Apparently, there should be certain regulatory mechanisms to balance the detrimental and anabolic effects of the IL-6 signaling, even for tissues such as liver, where both the effects of IL-6 have been fairly observed. One of the possible mechanisms could be that the cell status or even cell types in these tissues under different pathophysiological conditions responding to the IL-6 stimulation vary. In cartilage, it may be that the mature chondrocytes and stem/progenitor cells respond to IL-6 in different ways, consequently leading to its detrimental or anabolic effects, respectively. The local concentration of inflammatory factors may be another parameter, as shown for many other pro-inflammatory factors, including IL-1a, IL-6 and IFNg, which could also mediate regeneration. [41] In this study, IL-6 production by CSPC following the KGN stimulation was clearly demonstrated and the absolute concentration is not clear but might not be too.
Therefore, the anabolic effects may be more protruding in our study but not like most previous studies where mainly the role of IL-6 in mature and degenerative cartilage tissues were involved and the pro-inflammatory catabolic effects might be dominating.
However, this notion remains to be further confirmed by more detailed analysis in different experimental settings, such as the possible role of KGN in IL-6 deficient animal models.
Discovering new roles of KGN on CSPC or MSC and deciphering the underlying signaling mechanisms would be interesting also for several other aspects. As apoptosis and senescence of either differentiated chondrocytes or progenitor cells have been implicated in the OA pathology, [7][8][9][10] it might be worth of testing whether KGN could have an effect in delaying or reverse these cellular phenotypes. Considering stem cell therapy for OA and other joint diseases, pre-treatment or chondrogenic priming with KGN or an IL-6/Stat3 analogue, prior to the in vivo transplantation or loading onto a biomaterial scaffold, may be able to enhance the therapeutic efficacy of CSPC or MSC. [41,42] Furthermore, this study demonstrated that KGN injection could also mediate cartilage regeneration, suggesting the potential of KGN to be developed 8as clinically useful tablets. In line of this and even more importantly, the biosafety and pharmacological kinetics of the systemically and locally administrated KGN may have become the next issue to be addressed in further investigation in order to translate the small chemical to clinical application.

Conclusions
The proliferation of cartilage progenitor cells is crucial for regeneration and repair of cartilage degenerative defects. KGN mediates cartilage regeneration not only by stimulating CSPC differentiation but also by enhancing their proliferation via the IL-6/Stat3 signaling pathway. This study further implicate that KGN may be further translated into a reagent useful both locally and systemically. The involvement of IL-6/Stat3 in cartilage regeneration further implicates an novel insight into the association of cartilage regeneration with inflammation response of the local tissue environmental factors, with particular emphasis on their activating effects on tissue-resident stem/progenitor cells.
New therapies for OA may take advantage of the IL-6/Stat3 signaling rather than just inhibiting the inflammatory effects of IL-6, probably with necessary balancing of the IL-6 level and the stage of the OA pathology.

Availability of data and materials
Not applicable.

Consent for publication
Not applicable.