Expression of calcitonin gene‐related peptide induces ligament degeneration through endochondral ossification in osteoarthritis

Osteoarthritis (OA) is a disease in which degeneration occurs in various tissues such as cartilage and subchondral bone. Degeneration of ligaments also plays an important role in OA progression, resulting in an increase in chondrocytes and ossification, but the factor that causes this is still unclear. It is reported that the expression of calcitonin gene‐related peptide (CGRP) increases OA progression, and CGRP might play a role in ligament degeneration because CGRP has a function in endochondral ossification. The purpose of this study is to analyze the mechanism of ligament degeneration and the function of CGRP.


| INTRODUC TI ON
Osteoarthritis (OA) is a progressive joint disorder that increases its prevalence annually.Although OA is a common disease encountered in daily clinical practice, the mechanisms of OA initiation and progression have not been completely elucidated.In the pathogenesis of OA, various tissues, including articular cartilage, subchondral bone, meniscus, joint capsule, and ligament, exhibit progressive degenerative changes, which is one of the reasons for the difficulty with OA treatment. 1Hence, elucidating the mechanism of degenerative changes in all joint structures is important.
However, few studies on the degeneration mechanism of ligaments have been conducted compared with those on the cartilage, bone, and meniscus, despite the importance of ligament degeneration in OA pathogenesis.Cushner et al. reported that 47% of patients with OA have ligament degeneration, which may cause joint laxity. 2 Levy et al. demonstrated that histological degenerative changes of the posterior cruciate ligament (PCL) were observed before articular cartilage degeneration. 3The anterior cruciate ligament (ACL) and PCL insufficiency frequently cause meniscal and cartilage injuries, which can accelerate the progression of OA due to lateral thrust. 4ucidation of the mechanism of ligament degeneration would enable the prevention of OA progression.Increasing chondrogenic potential in the ligament has been reported to be a key factor in the degeneration process. 5During the degenerative process of the ligament, chondrogenic cells increase, and ossification finally occurs in the fibroblast-based ligament. 6This suggests that some factors may induce chondrogenesis and osteogenesis in the ligament during OA pathogenesis.
Neuropeptides, such as calcitonin gene-related peptide (CGRP) and substance P, are well known to play an important role in OArelated pain in the joint. 7Neuropeptides have various functions, including bone metabolism and angiogenesis, to maintain homeostasis. 8In the pathogenesis of OA, excessive expression of neuropeptides in the subchondral bone causes sclerotic changes in the subchondral bone, and subsequently, cartilage degeneration progresses. 9,10CGRP is a neuropeptide belonging to the calcitonin superfamily; it consists of 37 amino acids and is released from sensory nerves. 11CGRP has various functions, including blood vessel dilation, accelerating inflammatory pain, and promotion of bone remodeling by differentiation of bone marrow mesenchymal cells into osteoblasts, stimulation of osteoblast proliferation, and acceleration of bone metabolism by inhibiting osteoclast differentiation. 124][15][16] CGRP plays a role in not only pain but also the pathophysiology of OA such as subchondral bone sclerosis. 9,17In addition, CGRP is recognized as an important factor in heterotopic ossification and fracture healing, which suggested that the expression of CGRP might be related to increasing the chondrogenic and osteogenic potential in the ligament. 18,19As CGRP receptor antagonists or antibodies have been developed for the treatment of migraine, there is the possibility that these drugs could be used to treat OA. 20 We focused on CGRP expression and its function in ligament degeneration during OA progression.We hypothesized that the higher expression of the CGRP in the ligament plays a role in increasing the chondrogenic potential during OA progression.The aim of this study was to analyze the expression pattern and function of CGRP in the ligament cells in OA pathogenesis.

| MATERIAL S AND ME THODS
Between June 2016 and October 2018, 30 patients with advanced knee OA underwent total knee arthroplasty were included; 11 men and 19 women, with a mean age of 73.5 years.Knee joints were classified as Kellgren-Lawrence grades 3 and 4. Patients with a history of ligament injury, malunited fractures, intra-articular steroid injection in the previous 6 months, joint infection, or previous knee surgery were excluded.During surgery, PCL was harvested for histological analysis according to a previous report. 21This study was approved by the institutional review board and ethics committee of our hospital and was conducted in accordance with the Helsinki Declaration.Informed consent was obtained from all patients.

| Histological analysis of human ligament
Harvested PCL was fixed in 4% paraformaldehyde and embedded in paraffin.Four-micrometer-thick sections were prepared for histological analysis.Sections were stained with hematoxylin & eosin (H&E) and safranin-O/fast green, and histologically graded using a scoring system according to a previous report. 15The following categories were examined and scored for each ligament: (1) inflammation in the ligament substance, (2) mucoid degeneration, (3) chondroid metaplasia, (4) cystic changes, and (5) orientation of collagen fibers.Five fields of the PCL were randomly selected and each slide was evaluated at 200× magnification.The histological changes were scored and graded with reference to the previous report as follows 22 : 0, no changes; 0.5, minimal changes; 1, mild changes; 2, moderate changes; and 3, severe changes.The highest summed score of ligament degeneration (total score) was 15 if all five histological categories were scored as severe.Furthermore, the total score was classified into three groups: mild (0-5), moderate (6-10), and severe (11-15).

| Animals
All procedures were performed in accordance with the Guidelines for Animal Experimentation at our university, and with the approval of the Committee of Research Facilities for Laboratory Animal Sciences, Graduate School of Biomedical Sciences, Hiroshima University.
Senescence-accelerated mouse prone 8 (SAMP8) mice, which develop spontaneous joint OA signs that resemble human disease, were used at the ages of 4, 18, and 42 weeks (n = 9 at each time point) for the evaluation of the spontaneous OA model.Harvested knee joints fixed in 4% paraformaldehyde were decalcified for 2 weeks in 20% EDTA and embedded in paraffin.Four-micrometer-thick sagittal sections, where the whole length of the PCL in the knee joint was observed, were prepared for histological analysis.Sections were stained with H&E and safranin-O and graded histologically using Osteoarthritis Research Society International (OARSI) and ligament scores. 22,23

| Immunohistochemical analysis
For immunohistochemical analysis, each section was immunostained using anti-CGRP antibody (1:500 dilution; Abcam), anti-SOX9 (1:800 dilution; Abcam), anti-matrix metallopeptidase 13 (MMP13) (1:20 dilution; Neo Markers), anti-CRLR/CGRP1 polyclonal antibody (1:100 dilution; Bioss) and anti-type II collagen (1:150 dilution; DSHB).For double immunofluorescence staining of CGRP and SOX9 or type II collagen, and MMP13, the anti-CGRP antibody was labeled with fluorescein isothiocyanate using Dojindo Ab-10 Rapid Fluorescein Labeling Kit (Dojindo Laboratories), and anti-SOX9, anti-collagen type II, and anti-MMP13 antibodies were labeled using the Dojindo Ab-10 Rapid HiLyte Fluor 555 Labeling Kit (Dojindo Laboratories) according to the method described in a previous report. 24DAPI (Dojindo Laboratories) solution was used for nuclear staining.For immunohistochemical signals, five fields in the ligament were randomly selected among the areas where ligamentous structures were found; the total number of cells was counted, and the number of SOX9-, MMP13-, and CGRP-positive cells was counted at 200× magnification using ImageJ software (National Institutes of Health) according to the method of a previous report. 23,25The numbers of both CGRP-and type II collagen-positive cells or MMP13-positive cells were also counted and their percentage in CGRP-positive cells was calculated.

| Cell culture
Human ligament cells were obtained from the ACL of five patients who had undergone total knee arthroplasty.Ligament tissues were placed in 10-cm diameter Petri dishes under sterile conditions and washed five times with phosphate-buffered saline (PBS; WAKO) to remove red blood cells after removing synovial tissues, adipose tissue, and small blood vessels.The tissue was minced into 1-2 mm pieces, and 0.25% type 1 collagenase was added.Tissues were shaken in a water bath at 37°C for 60 min, and then an equal volume of high-glucose Dulbecco's modified Eagle's medium (DMEM; Life Technologies) containing 10% heat-inactivated fetal bovine serum (FBS; Sigma-Aldrich Corp.) and antibiotics (at a final concentration of 100 U/mL penicillin, 100 μg/mL streptomycin, and 0.25 μg/mL amphotericin B; Nacalai Tesque) was added to stop degradation.The tissue was then filtered through a 200-mesh nylon filter and centrifuged at 400 rpm for 10 min.The supernatant was discarded, and the human ligament cells were suspended in DMEM containing 10% FBS and 1% antibiotics.The harvested cells were homogeneous and were further identified as ligament cells by real-time polymerase chain reaction (PCR) analysis as they expressed SCX (Hs 03054634_g1).
When the cell coverage reached 80%-90%, the cells were passaged.
The cells from passages 2 and 3 were seeded onto 24-well plates (BD Falcon), and incubated with or without 10 and 100 nM CGRP receptor agonist (Peptide Research Institute, Inc.) or 10 and100 nM CGRP receptor antagonist (BIBN4096, TOCRIS, Bioscience) or the same amount of PBS as a control group added to each well twice a week in a humidified 5% CO 2 /95% air atmosphere at 37°C (n = 8 each group).
After 24 h, 48 h, and 21 days, RNA was extracted for PCR analysis.

Osteogenic differentiation medium (StemPro Osteogenic
Differentiation Kit, Life Technologies) and antibiotics with a final concentration of 100 U/mL penicillin, 100 μg/mL streptomycin, and 0.25 μg/mL amphotericin B (Nacalai Tesque) were used.A calcified medium containing 100 nM of CGRP receptor agonist, or the same amount of PBS was added and cultured for 3 weeks.Osteogenesis capacity was evaluated by PCR and staining with 0.5% alizarin red solution (Nacalai Tesque).To induce adipogenesis, the cells were cultured in adipogenic differentiation medium (StemPro adipogenesis differentiation kit, Life Technologies) and antibiotics, to which 100 nM of CGRP receptor agonist or the same amount of PBS was added.After 3 weeks, adipogenic capacity was assessed using PCR and staining with 0.3% oil red-O (Sigma O-0625; Sigma-Aldrich Corp.).To induce chondrogenesis, 5 × 10 5 cells were placed in 15-mL polypropylene tubes (BD Falcon) and pelleted by centrifugation at 500 × g for 5 min.The pellets were cultured in chondrogenesis medium (StemPro chondrogenesis differentiation kit, Life Technologies) and antibiotics, to which 100 nM of CGRP receptor agonist or the same amount of PBS was added for 3 weeks.Chondrogenic capacity was evaluated using PCR and histological analysis.For histological analysis, the pellets were embedded in paraffin, cut into 6 μm sections, and stained with safranin-O/fast green.

| Real-time PCR
RNA was isolated using TRIzol (Life Technologies) for real-time PCR analysis, and proteins were extracted for analysis of alkaline phosphatase activity.Complementary DNA was synthesized using 1 μg of total RNA using the Superscript VLIO kit (Invitrogen) according to the manufacturer's protocol: MMP13 (Hs 00233992_m1), RUNX2 (Hs 00231692_m1), VEGFA (Hs 0090055_m1), SOX9 (Hs 00165814_ m1), COL2A1(Hs 01064869_m1), COL10A1 (Hs 00166657_m1), PPARγ (Hs 01115513_m1), and GAPDH (Hs99999905_m1) using TaqMan gene expression assays probes (Life Technologies), and realtime PCR assays were performed.The expression level of each gene was evaluated relative to the expression level of GAPDH.The ΔΔCt method was used to analyze the real-time PCR data.The experiment was conducted at least three times with samples of five patients, and the averages of the data from the real-time PCR were compared.

| Statistical analysis
All results in this study are expressed as mean ± standard deviation (SD).Comparisons among the three or four groups were performed using the Tukey-Kramer post hoc test, and the Mann-Whitney U test was used to determine the differences between the two groups.
Statistical significance was set at p < .05.

| Expression pattern of the CGRP in human PCL
Of the 30 knees with human ligament degeneration, 7 were in the mild group, 12 were in the moderate group, and 11 were in the severe group.The mean histopathological score of degeneration of the mild group was 3.3 ± 0.3 points (range 2-4.5), the ligament was composed of parallel fiber arrangements, and the cells were mostly fibroblasts.
In the moderate group, the mean histopathological score of degeneration was 6.7 ± 1.5 points (range 5-9), the number of chondrocytelike cells with round nuclei increased, and the surrounding tissue was stained with safranin O.In addition, the ligament fibers became wavy, and the parallel fiber arrangement was disrupted.In advanced degeneration, the mean histopathological score of degeneration was 12.9 ± 1.5 points (range 10-14), the number of chondrocyte-like cells increased, and the area stained with safranin O stain also increased.
In addition, ossification areas were observed in this study.
The number of SOX9-positive cells was significantly higher in the moderate group than in the mild and advanced groups (p < .01).
The number of MMP13-positive cells was significantly higher in the moderate group than in the advanced group (p < .01).The number of CGRP-positive cells was lower in the mild group, significantly higher in the moderate group than in the other groups (p < .01),and decreased again in the severe group (Figure 1A,B).Immunofluorescent analyses revealed that CGRP and SOX9 were co-expressed in the chondrocyte-like cells in the ligament (Figure 1C).CGRP-positive cells was significantly higher at 18 weeks than at 4 and 42 weeks (Figure 2A,C).Immunohistochemical analysis of the CGRP receptor showed few positive cells at 4 weeks.However, the positive cells in both fibroblasts and chondrocyte-like cells increased at 18 weeks, and they decreased and were expressed only in chondrocyte-like cells at 42 weeks.Type II collagen-positive cells were increased, and the percentage of both CGRP-and type II collagen-positive cells was significantly increased as OA progressed (Figure 2D).Immunohistochemical analysis of CGRP and SOX9 showed that the co-expression of CGRP and SOX9 in the chondrocyte-like cells was highest at 18 weeks, and then decreased at 42 weeks (Figure 3).

| Functional analysis of CGRP on the ligament in vitro
To investigate the chondrogenic effects of CGRP on ligament cells, the expression of SOX9, RUNX2, COL10A1, VEGF, and MMP13 was evaluated using real-time PCR.SOX9 expression increased in a CGRP dose-dependent manner at all stages and was significantly suppressed by BIBN4096 treatment.Runx2 expression increased in a CGRP dose-dependent manner at all stages and was suppressed by BIBN4096 treatment.COL10a1 expression significantly increased in a CGRP dose-dependent manner at 3 weeks post-treatment and was significantly suppressed by BIBN4096 treatment (Figure 4).VEGF expression was significantly upregulated at 24 h post-treatment in the CGRP 100 nM group and at 3 weeks post-treatment in the CGRP 10 nM group.MMP13 expression increased in a dose-dependent manner at 3 weeks and was significantly suppressed by BIBN4096 treatment (Figure 5).Real-time PCR analyses revealed that CGRP has an effect on ligament cells to promote endochondral ossification, which is inhibited by BIBN4096.
The effect of CGRP on chondrogenesis, osteogenesis, and adipogenesis of ligament-derived cells was examined.In chondrogenesis, the diameter of the pellet with CGRP treatment was larger than that without CGRP treatment (Figure 6A).In addition, the expression of SOX9 and COL2a1 was significantly higher in the pellet with CGRP treatment than in the pellet without CGRP treatment (Figure 6A).In osteogenesis, alizarin red staining showed more staining in the ligament cells with CGRP treatment (Figure 6B).In addition, RUNX2 expression in the ligament cells with CGRP treatment was significantly higher than that without CGRP treatment (Figure 6B).In adipogenic differentiation, oil red staining showed no staining of the ligament after CGRP treatment (Figure 6C).The expression of PPARγ was significantly lower in the ligament cells treated with CGRP than in those without CGRP treatment (Figure 6C).

| DISCUSS ION
This study showed that CGRP is expressed in chondrocyte-like cells in the ligament as degeneration develops, and CGRP may be involved in chondrogenesis and osteogenesis in the ligament during OA progression.As ligament dysfunction due to degeneration is an important factor in the pathogenesis of OA, elucidating the factors inducing ligament degeneration will lead to the development of new therapeutic strategies.Neuropeptides, including CGRP, have been recognized as therapeutic targets for various diseases, and receptor antagonists or antibodies have been developed as drugs targeting neuropeptides. 26,27In addition, CGRP is a potential target to prevent OA progression by administering a CGRP receptor antagonist through inhibition of subchondral bone sclerosis. 9Our findings that CGRP is involved in ligament degeneration in OA progression would contribute to the development of a therapeutic strategy for OA.Ligament degeneration is histologically characterized by the disorganization of collagen fiber arrangement and chondroid metaplasia. 22,28,29Degenerating ligaments exhibit fibrocartilaginous regions with collagen type II and III and increasing distribution of chondrocytes in addition to disorganization of fiber arrangements. 2,30,31perimental tendon degeneration showed that approximately 5% of the degenerating area has calcific foci with chondrocytes expressing SOX9. 32Kumagai et al. demonstrated that Sclelaxis-positive cells decreased and SOX9-positive cells increased as ligament degeneration progressed. 5Hence, chondrogenic differentiation plays a crucial role in ligament degeneration.However, the factors that induce or promote chondrogenic differentiation of the ligament during OA progression have not been elucidated.We focused on CGRP regarding chondrogenic differentiation in ligament degeneration.CGRP has various functions, including pain perception, osteogenesis, and angiogenesis. 33The sensory nerve around the joint has been reported to elongate into the joint during OA progression, and the neuropeptides released from the sensory nerve then cause pain, inflammation, and subchondral bone sclerosis. 34CGRP is distributed in the synovium, subchondral bone, cartilage, and meniscus in the joints with OA. 35 In particular, CGRP expression in the subchondral bone increases in DMM (destabilization of the medial meniscus) mice and induces subchondral bone sclerosis. 9According to this evidence, CGRP may also induce ossification of the ligament during the degeneration process.However, there are few studies regarding the expression of CGRP in the ligament.In the normal collateral ligament of the rat, CGRP-positive nerve fiber was observed but its expression in the tendons was more abundant than in the ligaments. 36In another report using rabbits, it is reported that CGRP-positive nerve fibers increased 6 weeks after medial collateral ligament injury and decreased at 14 weeks. 37In the report using human OA PCLs which were harvested during total knee arthroplasty, there was no significant difference between OA and non-OA PCLs. 38Our results that CGRP expression increased with the progression of the ligament degeneration and decreased in the end-stage of OA were supported by these previous reports.Moreover, CGRP has been reported to be involved in heterotopic ossification, 39 and the administration of CGRP to the tibialis anterior muscle of mice promotes heterotopic ossification. 18 examined the effects of CGRP on the differentiation of ligament cells.Furumatsu et al. demonstrated that ligament cells exhibit chondrogenic properties with SOX9 expression in a chondrogenic induction medium. 40Takimoto et al. showed the direct conversion of tenocytes to chondrocytes by SOX9 overexpression. 41In our study, CGRP upregulated SOX9 expression in pellet cultures during chondrogenesis, and the size of the pellet was enlarged.Furthermore,  42 VEGF, an important factor in endochondral ossification, is also upregulated by CGRP.Hypervascularization in various tissues in the joint, including the ligament, has been reported in the OA joint. 43Our study revealed that CGRP might play a role in inducing endochondral ossification and angiogenesis in the ligament through VEGF expression.In our study, CGRP expression increased as the degeneration of the ligament progressed, but severe degeneration of ligaments did not exhibit a decrease in CGRP expression accompanied by MMP13 expression.It is suspected that CGRP expression in the ligament increases by the sensory nerve that elongates from the surrounding joint. 13From the previous reports and our in vitro study, CGRP may be expressed in the ligament cells, which have the chondrogenic potential and promote chondrogenesis from ligament cells. 40,41,44In the ligament showing severe degeneration, fibroblastic cells were sparsely present, and CGRP was no longer required.CGRP induces VEGF at 24 h after CGRP addition to the ligament cells.CGRP has been reported to induce angiogenesis. 33CGRP may also induce angiogenesis in the early phase of OA, which may lead to further increased expression of CGRP. 15Our in vivo study revealed that the number of CGRP expression cells was reduced in severe degeneration but the percentage of CGRP expression cells with co-expressing type II collagen was increased.It is suggested that most CGRP-positive cells are involved in type II collagen production in the ligament during OA progression.This study had several limitations.First, the natural course of ligament degeneration during OA progression could not be observed in human samples.This is an unsolvable problem for histological evaluation, but we compensated for it by stratifying the degree of ligament degeneration.Moreover, SAMP8, which has been established as a spontaneous OA model, was used to evaluate the time course of ligament degeneration in OA progression. 45In the report detailing OA changes in SAMP8, 4 weeks of age showed no OA change in the knee joint, but a roughened articular cartilage surface, fibrillation, and proteoglycan loss in the articular cartilage started at 11 weeks of age and the severity of OA increased over time. 46As the natural course of OA progression in SAMP8 mice is quite similar to that in humans, SAMP8 mice are useful for investigating the pathogenesis of primary OA.Second, human ACL-derived cells were used in in vitro studies although in vivo studies focused on the PCL.ACL has less synovium than PCL, making it easier to isolate ligament cells without contamination of synovial cells.The histological findings of PCL are correlated with those of ACL in OA, 3 so using ACL-derived cells may yield the same results as using PCL-derived cells.In addition, there was a possibility that chondrocyte-like cells might be included in the isolated cells because the ACL was harvested from OA patients.It is desirable to use normal ACL without OA to evaluate the effect of CGRP on pure ligament cells.Third, the factors that induce the expression of CGRP in the ligament have not been elucidated, although this study explored the important role of CGRP in ligament degeneration.However, the factors that induce OA have not been clarified, despite many studies on OA pathogenesis.Finally, the gain or loss of function of CGRP was not examined in vivo in this study.To evaluate the role of CGRP on OA progression in vivo, analyses using CGRP transgenic or knockout mice are useful.However, our previous study revealed that administration of the CGRP receptor antagonist into the DMM mice reduced the OARSI score until 8 weeks, which suggested that inhibition of the CGRP expression may lead to preventing OA progression. 9As CGRP plays a role in increasing the chondrogenic potential in the ligament, administration of the CGRP receptor antagonist may ameliorate the ligament degeneration, subsequently preventing OA progression.Further studies to elucidate the OA progression through the CGRP and develop the therapeutic strategy for targeting the CGRP using several mouse models will be needed in the future.
In conclusion, the present study revealed that CGRP plays a role in increasing the chondrogenic potential in the ligament, and subsequently ligament degeneration.There is the possibility that CGRP would be the therapeutic target to prevent ligament degeneration in OA.
To investigate longitudinal ligament degeneration, PCL in a spontaneous OA mouse model (SAMP8 mice) were histologically analyzed.As the number of weeks increased, the OARSI and ligament scores significantly increased with articular cartilage degeneration (Figure 2A,B).H&E and safranin-O/fast green staining revealed parallel fiber arrangement and spindle-shaped nuclei.Ligaments were rarely stained with safranin-O at 4 weeks.At 18 weeks, the increased area in the ligament stained with safranin-O stain and the parallel fiber arrangement decreased, and chondrocyte-like cells were observed.At 42 weeks, ligaments stained almost exclusively with safranin-O stain and the parallel fiber structure disappeared, and the number of chondrocyte-like cells increased; its degeneration was similar to that of human ligament degeneration.The number of F I G U R E 1 Expression pattern of the calcitonin gene-related peptide (CGRP) in human posterior cruciate ligament.(A) Expression pattern of SOX9, MMP13, and CGRP with mild, moderate, and severe degeneration in safranin O staining and immunohistochemistry. (B) Ligament scores and percentage of SOX9-, MMP13-, and CGRP-positive cells.*p < .05,**p < .01. (C) Safranin O staining and immunohistochemistry of CGRP and SOX9.Bar 50 μm.

F I G U R E 4
Gene expression analyses of SOX9, RUNX2 and COL10a1 using polymerase chain reaction.*p < .05,**p < .01.F I G U R E 3 Double staining of the calcitonin gene-related peptide and SOX9 in the posterior cruciate ligament of senescence-accelerated mouse prone 8 mice.Bar 100 μm.

F I G U R E 5
Gene expression analyses of VEGF and MMP13 using polymerase chain reaction.*p < .05,**p < .01.CGRP promoted osteogenesis in the ligament cells while inhibiting adipogenesis.With the addition of CGRP to the ligament cells, the expression of SOX9 and RUNX2 was increased for up to 3 weeks, suggesting that CGRP promotes chondrogenesis and osteogenesis.At 3 weeks of culture, the expression of COL10a1 and MMP13, which are important factors in endochondral ossification, increased in a dose-dependent manner, indicating CGRP-induced endochondral ossification.

F I G U R E 6
Differentiation analyses of the human ligament cells by the calcitonin gene-related peptide.(A) Chondrogenesis capacity using pellet culture.(B) Osteogenesis was evaluated using alizarin red staining and real-time PCR of RUNX2.(C) Adipogenesis was evaluated using oil red O staining and real-time PCR of PPARγ.