Possible Regulation of bFGF Expression by Mast Cells in Osteoarthritis Patients with Obesity: A Cross-Sectional Study.

PURPOSE: Obesity is associated with the risk of developing knee osteoarthritis (KOA). Furthermore, synovial basic fibroblast growth factor (bFGF) is linked to the severity of KOA. We previously demonstrated that bFGF and mast cell (MC) marker expression were elevated in the synovial tissues (ST) of KOA patients with obesity. However, it remains unclear whether MCs contribute to bFGF expression and regulation. PATIENTS AND METHODS: Radiographically diagnosed KOA patients (n=249) were assigned to groups based on the body mass index (BMI) classifications used by the World Health Organization: normal-weight (NW, BMI <25 kg/m2, n=95), overweight (OW, BMI ≥25 and <30, n=109) and obese (OB, ≥30 kg/m2, n=45). BFGF expression in the ST was examined using quantitative polymerase chain reaction and compared across the BMI groups. Additionally, BFGF and interleukin (IL) 13 expression were examined in freshly extracted MC-rich (THY-1-, CD3-, CD14-, and CD19-) and MC-poor (THY-1+, CD3+, CD14+, or CD19+) fractions from ST. Moreover, regulation of BFGF expression by IL-13 was studied in CD14-negative (fibroblast-rich) and CD14-positive (Mφ-rich) and cells in culture.
RESULTS: BFGF expression was significantly higher in OB than in NW patients. Furthermore, although IL13 was significantly higher in the MC-rich than the MC-poor fraction, BFGF expression was comparable. Recombinant human IL-13 stimulated expression of BFGF in synovial fibroblast cells.
CONCLUSION: BFGF expression is higher in the ST of KOA patients with obesity. Increased numbers of MCs may contribute to the elevated BFGF expression through IL-13 in KOA patients with obesity.

Introduction

Obesity has been shown to increase the prevalence and development of osteoarthritis (OA) in weight-bearing as well as non-weight-bearing joints.1,2 This suggests that other factors apart from mechanical loading may explain the association between obesity and OA. Such other factors may also contribute to OA pathology. However, these factors remain elusive.

Basic fibroblast growth factor (bFGF) regulates cell proliferation and contributes to the matrix metalloproteinase production and breakdown of the extracellular matrix in articular cartilage.3 Furthermore, bFGF contributes to the pathology of joint destruction,4 with bFGF levels in synovial fluid having been shown to be associated with radiographic severity of knee OA (KOA).5 In our previous report, we found that BFGF mRNA expression was increased in the synovial tissues (ST) of KOA patients with obesity.6 However, the mechanisms underlying the rise in bFGF in patients with obesity remain unclear.

Mast cells (MCs) are present in the ST, and are elevated in patients with rheumatoid arthritis (RA)7,8 and OA,9–11 where they are suggested to play a pivotal role in both acute and chronic inflammation. Recent evidence indicates a possible link between MCs and KOA severity.10 bFGF is among a range of growth factors generated by MCs under inflammatory conditions.12,13 We previously reported that BFGF expression was correlated with the MC marker expression.6 However, bFGF expression in MCs remains to be studied.

Interleukin (IL)-13 is an MC-secreted cytokine.14–16 Increased IL-13 levels are observed in the synovial fluid of patients with KOA and rheumatoid arthritis.17,18 IL-13 level in synovial fluid is correlated with OA pain and function.19 Interestingly, IL-13 has been shown to regulate bFGF in airway smooth muscle cells.20 We hypothesized that MCs express IL-13 and that IL-13 regulates bFGF in the ST of KOA patients.

We investigated BFGF and IL13 expressions in MCs, and the effect of IL-13 on BFGF expression in KOA patients.

Patients and Methods

Total knee arthroplasty (TKA) was performed on all participating subjects at our hospital. A total of 249 ST specimens were harvested from patients diagnosed with radiographic KOA during TKA. A small section of each specimen was immediately frozen in liquid nitrogen and placed at −80°C until RNA extraction. Specimens taken from 10 patients were subjected to experiments aimed at determining BFGF expression in MCs, and samples taken from 8 patients to experiments aimed at examining whether IL-13 regulates BFGF expression.

The study protocol received approval from Kitasato University Hospital institutional Ethics Review Board (approval number: KMEO B19–259). All participants provided written informed consent to participate and for removal of their synovial tissue the day prior to surgery. This study complied with the principles of the Declaration of Helsinki.

Patients were assigned to groups based on the body mass index (BMI) classifications used by the World Health Organization: normal-weight (BMI <25 kg/m2, n=95), overweight (BMI ≥25 and <30, n=109) and obese (≥30 kg/m2, n=45). The patients’ clinical characteristics are summarized in Table 1 by BMI group. MC marker (CD117, CD203c, TPSB2 (β-tryptase)) and BFGF expression in the ST were compared across the BMI groups using RT-PCR.

Table 1

Patients’ Clinical Characteristics

	Normal (n=95)	Overweight (n=109)	Obese (n=45)	P
Age (years)	75.8±7.5	73.2±6.9a	71.1±7.4a,b	<0.001
Male/female, n	17/78	32/77	7/38	0.099
KL (3/4), n	29/66	31/78	11/44	0.634
BMI (kg/m2)	22.2±1.9	27.2±1.4a	32.8±2.2a,b	<0.001

Notes: Data are mean±standard deviation unless otherwise indicated. aStatistical difference between normal and obese groups. bStatistical difference between overweight and obese groups.

Abbreviations: KL, Kellgren and Lawrence grade; BMI, body mass index.

qPCR

Details of RNA extraction, cDNA synthesis and qPCR using SYBR Green are published elsewhere.8,21 qPCR primer sequences are shown in Table 2. All gene expression was normalized to that of GAPDH.

Table 2

Primer Sequences

Gene	Direction	Primer sequence (5′–3′)	Product size (bp)
TPSB2	F	CGCAAAATACCACCTTGGCG	138
	R	GTGCCATTCACCTTGCACAC
IL13	F	CCTCATGGCGCTTTTGTTGAC	134
	R	TCTGGTTCTGGGTGATGTTGA
CD117	F	TGACTTACGACAGGCTCGTG	126
	R	CCACTGGCAGTACAGAAGCA
CD203c	F	CGACTGCACTATGCCAAGAA	164
	R	GGTCCATGTGCCAGAAAGAT
BFGF	F	AGAGCGACCCTCACATCAAG	80
	R	ACGGTTAGCACACACTCCTT
GAPDH	F	TGCCACTCAGAAGACTGTGG	129
	R	TTCAGCTCTGGGATGACCTT

Expression of BFGF and IL13 in MCs

Cells isolated from synovial tissue were separated into MC-rich and MC-poor fractions as described previously,21 and examined for BFGF and IL13 expression. To digest ST, samples were immediately immersed in collagenase solution at 37°C for 2 h. Extracted cells reacted with the following antibodies for 30 minutes at 4°C: biotin-conjugated anti-THY-1 (fibroblast marker), anti-CD3 (T cell marker), anti-CD14 (Mφ marker), and anti-CD19 (B cell marker). The cells were subsequently washed twice with Hanks’ balanced salt solution, and exposed to streptavidin-conjugated magnetic particles (BD Biosciences, CA, USA). A magnetic separation system (BD Biosciences) was used to separate the cells into negative (MC-rich; THY-1-, CD3-, CD14-, and CD19-) and positive (MC-poor; THY-1+, CD3+, CD14+, or CD19+) fractions. qPCR was used to determine the BFGF and IL13 expression in negative (MC-rich) and positive fractions.

Effect of IL-13 on Synovial Fibroblasts and Mφ

Synovial fibroblasts and Mφ were isolated from ST samples harvested from 8 KOA patients to study the role of IL-13 in the synovium. CD14-positive (Mφ-rich) and CD14-negative (fibroblast-rich) cells were separated using the magnetic separation technique described in a previous protocol.22 CD14-negative and CD14-positive cells were seeded in α-MEM in six-well plates and cultured for 7 days. Thereafter, the cells were exposed to vehicle (serum-free media) or 100 ng/mL of recombinant human IL-13 (rh-IL13; BioLegend) for 6 hours. The IL-13 concentration was chosen based on a protocol reported previously.20 Following stimulation, qPCR was conducted to measure BFGF expression.

Statistical Analyses

All analyses were conducted based on a pre-specified statistical analysis plan. All analyses were performed using SPSS 25.0. Continuous variables were compared between two groups using the Wilcoxon signed-rank test and among three groups using the Kruskal–Wallis test followed by the Mann–Whitney U-test using Bonferroni correction. Categorical variables were compared using the Fisher's exact test. P<0.05 was indicative of statistical significance.

Results

Clinical Characteristics by BMI

OB patients were significantly younger than NW and OW patients (Table 1). The proportion of male to female patients (P=0.099) and those with Kellgren and Lawrence grade 3 and 4 (P=0.634) was comparable across the BMI groups (Table 1).

Expression of Synovial MC Markers (CD117, CD203c, TPSB2) and BFGF by BMI

OB patients had significantly higher levels of CD117 than NW patients (P=0.011), while NW- and OW patients had comparable levels (P=0.535) (Figure 1A). OB patients also had significantly higher levels of CD203c than NW patients (P=0.027) but comparable levels to OW patients (P=0.161) (Figure 1B). Likewise, OB patients had significantly higher levels of TPSB2 than NW patients (P=0.002) but comparable levels to OW patients (P=0.247) (Figure 1C).

Figure 1

CD117 (A), CD203c (B), TPSB2 (C), and BFGF (D) mRNA expression in the synovium of normal, overweight, and obese patients with knee osteoarthritis. *P<0.05. P values are indicated.

Similarly, OB patients had significantly higher levels of BFGF expression than NW patients (P=0.020) but comparable levels to OW patients (P=0.283) (Figure 1D).

Correlation Between MC Markers and BFGF Expression

Since our results indicated that KOA patients with obesity had higher expression levels of MC markers (CD117, CD203c, TPSB2) and BFGF, we next investigated the correlation between MC marker and BFGF expression. BFGF expression was significantly correlated with CD117 (ρ=0.273, P<0.001; Figure 2A) CD203c (ρ=0.432, P<0.001; Figure 2B) and TPSB2 (ρ=0.164, P<0.001; Figure 2C) levels.

Figure 2

Correlation between expression of BFGF and MC markers. Correlation between expression of BFGF and CD117 (A), CD203c (B), and TPSB2 (C).

MC Markers, BFGF and IL13 in MCs

Next, we measured the expression of MC markers (CD117, CD203c, TPSB2), BFGF and IL13 in MCs in MC-rich and MC-poor fractions isolated from ST specimens of KOA patients. The MC-rich fraction expressed significantly higher levels of MC markers (P=0.005; Figure 3A–C) and IL13 (P=0.009; Figure 3D) than the MC-poor fraction. Meanwhile, MC-rich and MC-poor fractions expressed comparable levels of BFGF (P=0.285; Figure 3E).

Figure 3

Expression of MC markers, BFGF and IL13 in mast cells. CD117 (A), CD203c (B), TPSB2 (C), IL13 (D), and BFGF (E) expression levels in mast cell (MC)-rich (THY-1-CD3-CD14-CD19-) and MC-poor (THY-1+, CD3+, CD14+, or CD19+) fractions. *P<0.05. P values are indicated.

Effect of IL-13 on BFGF Expression in Synovial Fibroblasts and Mφ

Given that BFGF expression was comparable between MC-poor and MC-rich fractions, we next studied the effect of rh-IL13 on BFGF expression in fibroblast- and Mφ-rich fractions isolated from the ST of KOA patients. Exposure to rh-IL13 induced BFGF expression in synovial fibroblast-rich fractions (P=0.043; Figure 4A) but not Mφ-rich fractions (P=0.093; Figure 4B) at 6 h following stimulation.

Figure 4

Effect of IL-13 on BFGF expression in synovial fibroblasts and Mφ. BFGF expression in CD14-negative (fibroblast-rich) (A) and CD14-positive (Mφ-rich) (B) fractions at 6 h after stimulation with vehicle or IL-13. *P<0.05. P values are indicated.

Discussion

Obesity alters local and systemic bFGF levels. Furthermore, high levels of plasma FGF2 have been shown to be correlated with fat mass in Han Chinese populations.23 In our previous study, we reported that KOA patients with obesity had higher expression of synovial bFGF and MC markers than normal-weight patients with OA.6 Consistent with our previous results, we found that KOA patients with obesity had elevated bFGF and MC marker expression in the present study. Accumulating evidence indicates that bFGF contributes to OA pathology and cartilage destruction. bFGF stimulates the expression of MMP-13 expression in human articular chondrocytes.24 Thus, elevation of bFGF may contribute to OA pathology in KOA patients with obesity.

Immunohistochemical studies have reported increased bFGF staining in hyperplastic lining synoviocytes, and vascular cells in ST derived from rheumatoid arthritis patients.7 To investigate the potential role of MCs in bFGF expression, we measured bFGF expression in MCs isolated from osteoarthritic ST. However, we found no differences between MC and non-MC populations. Thus, although MCs express bFGF, whether MCs contribute to the elevation of bFGF in the osteoarthritic ST remains unclear.

Previous studies have reported that IL-13 is released by several immune cells including T cells,25 MCs,15,26 eosinophils,27 and basophils.28 We found that MCs expressed higher IL-13 levels than non-MC populations. In addition, IL-13 stimulated BFGF expression in synovial fibroblasts. Thus, increased MC numbers may induce BFGF expression in OA pathology by activating synovial fibroblasts in KOA patients with obesity.

There were main two limitations in the present study. First, we lack data on non-OA patients. Second, we did not examine protein levels.

In conclusion, KOA patients with obesity showed higher BFGF expression than normal- and overweight patients. Increased MC numbers may contribute to the elevation of BFGF expression through IL-13 in KOA patients with obesity.