Increased Ratio of CD14++CD80+ Cells/CD14++CD163+ Cells in the Infrapatellar Fat Pad of End-Stage Arthropathy Patients

Abstract

Objectives: This study sought to identify the ratio of M1/M2 cells in the infrapatellar fat pads (IFP) and subcutaneous fat tissues (SC) of osteoarthritis (OA) and rheumatoid arthritis (RA) patients. The clinical features of OA and RA patients treated with or without biological disease-modifying anti-rheumatic drugs (bDMARDs) were also assessed.

Methods: IFP and SC were collected from patients with OA and RA who are undergoing total knee arthroplasty (TKA). CD14-positive cells were then isolated from these samples. Flow cytometry was used to determine the number of CD14⁺⁺CD80⁺ cells and CD14⁺⁺CD163⁺ cells. The expression levels of lipid transcription factors, such as sterol regulatory element-binding protein 1 (SREBP1) and liver X receptor alpha (LXRA), and inflammatory cytokines were also evaluated.

Results: Twenty OA patients and 22 RA patients were enrolled in this study. Ten of the RA patients (45.4%) received bDAMRDs before TKA. On average, a fivefold increase in the number of CD14-positive cells and lower expression levels of SREBP1C and LXRA were observed in OA IFP relative to OA SC; however, these results were not obtained from the RA samples. The median ratio of CD14⁺⁺CD80⁺ cells/CD14⁺⁺CD163⁺ cells of OA IFP was 0.87 (0.76-1.09, interquartile range), which is higher to that of OA SC with a lower ratio (p = 0.05835).

Conclusions: The quantity and quality of CD14-positive cells differed between IFP and SC in arthropathy patients. To our knowledge, this is the first study to characterize the ratio of M1/M2 cells in the IFP and SC of end-stage OA and RA patients. The increased ratio of CD14⁺⁺CD80⁺ cells/CD14⁺⁺CD163⁺ cells in the IFP from patients with OA and RA treated with bDMARDs indicated that inflammation was localized in the IFP. As adipose tissue-derived innate immune cells were revealed as one of the targets for regulating inflammation, further analysis of these cells in the IFP may reveal new therapeutic strategies for inflammatory joint diseases.

Keywords: inflammation; infrapatellar fat pad (Hoffa’s); monocytes/macrophage; osteoarthritis; rheumatoid arthritis.

Figure 1

(A) The anatomical location of infrapatellar fat pads. (B) Flow chart of the experimental method.

Figure 2

The infrapatellar fat pads (IFP) of arthropathy patients has a higher proportion of CD14-positive cells than the subcutaneous fat tissues. CD14-positive cells per gram of SC or IFP in the osteoarthritis (OA) (A), rheumatoid arthritis (RA) without bDMARDs (B), and RA with bDMARDs (C). Samples were obtained during total knee arthroplasty, digested in collagenase medium to obtain the stromal vascular fraction, and stained with human CD14 antibody to separate adipose-tissue-resident CD14-positive cells. **p < 0.01 and ***p < 0.001, as determined by the paired-sample Wilcoxon signed-rank test to identify differences between SC and IFP OA. n = 20; RA without bDMARDs, n = 12; RA with bDMARDs, n = 10.

Figure 3

CD14⁺⁺CD80⁺ and CD14⁺⁺CD163⁺ cell ratio increases by 1.36-fold in osteoarthritis (OA) infrapatellar fat pads (IFP). (A–C) Ratio of CD14⁺⁺CD80⁺/CD14⁺⁺CD163⁺ cells in the infrapatellar fat pads (IFP) and subcutaneous fat tissues (SC) from the OA, rheumatoid arthritis (RA) without bDMARDs, and RA treated with bDMARDs groups. To calculate ratios and proportions, adipose-resident CD14-positive cells were stained with anti-human CD80, CD163, and CD14 antibodies and evaluated using flow cytometry. (D) Proportions of CD14⁺⁺CD80⁺ and CD14⁺⁺CD163⁺ cells in the SC and IFP tissues of OA and RA patients (SC OA, n = 4; SC RA, n = 6; IFP OA, n = 11; IFP RA, n = 13). The x-axis displays samples from IFP and SC grouped by OA and RA, and the y-axis displays the proportions of CD14⁺⁺CD80⁺ cells or CD14⁺⁺CD163⁺ cells in CD14-positive cells. Lines connecting the dots between SC and IFP indicate samples from the same patients. Gating strategies for obtaining CD14⁺⁺CD80⁺ cells and CD14⁺⁺CD163⁺ cells as well as fluorescence minus one are presented in Supplementary Figures S1 and S2 . Statistics were calculated by the paired-sample Wilcoxon signed-rank test between CD14⁺⁺CD80⁺ and CD14⁺⁺CD163⁺ cells from the same patients. The Mann–Whitney U-test was used for analysis between different groups.

Figure 4

Expression levels of lipid transcription factors related to inflammation decrease in the infrapatellar fat pads (IFP). Expression levels of SREBP1A (A), SREBP1C (B), and LXRA (C) in the IFP and subcutaneous fat tissues (SC) from osteoarthritis (OA) and rheumatoid arthritis (RA) patients. The panels to the left contain results comparing the expression levels between SC and IFP without division by diseases. The panels to the right contain results comparing the expression levels of SC and IFP in patients with OA, RA treated with bDMARDs, and RA without bDMARDs treatment. Lines connecting the dots from SC to IFP indicate samples from the same patients. The samples were labeled and separated into CD14-positive cells from the stromal vascular fraction of the IFP and SC from each patient. The median (interquartile range, IQR) expression levels of SREBP1A, SREBP1C, and LXRA in the SC from all patients were 0.31 (0–5.25), 0.98 (0.22–2.98), and 0.69 (0.34–2.12); these expression levels changed to 1.1 (0.09–4.81), 0.30 (0.16–0.71), and 0.039 (0.0085–0.50) in the IFP. The median (IQR) expression levels of SREBP1C and LXRA in the SC from OA patients were 1.03 (0.03–3.69) and 0.69 (0.34–2.14); these expression levels changed to 0.31 (0.23–0.71) and 0.039 (0.018–0.50) in the IFP. The median (IQR) expression levels of LXRA in the SC and IFP from RA patients who were not treated with bDMARDs were 1.04 (0.065–20.89) and 0.017 (0.0067–0.81). Significant differences are displayed in each figure. OA, n = 15; RA without bDMARDs, n = 7; RA with bDMARDs, n = 9, respectively. The 2^−ΔΔCT method was used to quantify the qPCR data. GAPDH was used as the reference gene. The average GAPDH Ct values from all SC samples were used to normalize the Ct values. The Y-axes indicate the relative expression levels of mRNA, which were normalized by GAPDH. *p < 0.05, **p < 0.01, and ****p < 0.0001, as determined using the Mann–Whitney U-test for differences between the IFP and SC groups and the paired-sample Wilcoxon signed-rank test for differences between IFP and SC within the OA, RA without bDMARDs, and RA with bDMARDs groups.

Figure 5

Expression levels of inflammatory cytokines decrease in the IFP. Expression levels of CXCL10 (A), IL1B (B), and IL6 (C) in the infrapatellar fat pads (IFP) and subcutaneous fat tissues (SC) from osteoarthritis (OA) and rheumatoid arthritis (RA) patients. The panels to the left contain results comparing the expression levels between SC and IFP without division by diseases. The panels to the right contain results comparing the expression levels of SC and IFP in patients with OA, RA treated with bDMARDs, and RA who were not treated by bDMARDs. Lines connecting the dots from SC to IFP indicate samples from the same patients. The samples were labeled and separated into CD14-positive cells from the stromal vascular fraction of the IFP and SC of each patient. The median (interquartile range) expression levels of CXCL10 in the SC and IFP from patients were 1.13 (0.23–2.33) and 0.33 (0.10–0.55); the expression level in the SC was 0.98 (0.07–2.33) and changed to 0.17 (0.05–0.42) in the IFP of OA patients. Significant differences are displayed in each figure. OA, n = 15; RA without bDMARDs, n = 7; RA with bDMARDs, n = 9, respectively. The 2^−ΔΔCT method was used to quantify the qPCR data. GAPDH was used as the reference gene. The average GAPDH Ct values from all SC samples were used to normalize the Ct values. The Y-axes indicate the relative expression levels of mRNA, which were normalized by GAPDH. *p < 0.05, as determined using the Mann–Whitney U-tests for differences between the IFP and SC groups and the paired-sample Wilcoxon signed-rank test for differences between the IFP and SC within the OA, RA without bDMARDs, and RA with bDMARDs groups. Data for the sample whose CXCL10 2^−ΔΔCT value was 69.2-fold higher than the average were removed from the plot.