□ ORIGINAL ARTICLE □

Association of Soluble -like 2 with the Severity of Coronary Artery Disease

Jing Cheng 1, Yingying Chen 2, Banglong Xu 2, Jixiong Wu 2 and Fei He 2

Abstract

Objective The purpose of this study was to investigate the relationship between circulating soluble fibrinogen-like protein 2 (sFGL2) concentrations and the severity of coronary artery disease (CAD) in pa- tients who underwent first-time angiography for suspected CAD. Methods Serum sFGL2 concentrations were measured in 102 consecutive patients by an enzyme-linked im- munosorbent assay (ELISA). The number of circulating CD4+CD25+CD127low T regulatory cells (Tregs) was determined by flow cytometry and effecter cytokines, including transforming growth factor-β1and interleukin-10 (IL-10), were also evaluated by an ELISA. Associations between sFGL2 and Tregs with angi- ographic indexes of the severity of CAD (i.e., number of diseased vessels and the modified Gensini score) were estimated. Results The sFGL2 levels in patients with angiographically confirmed CAD were significantly lower than those in patients with normal coronary arteries (26.95±8.53 vs. 9.88±5.46 ng/mL, p<0.001). Significant corre- lations were observed between the serum sFGL2 level and number of diseased vessels (r=-0.860, p<0.001) and modified Gensini score (r=-0.833, p<0.001). Using a multivariate analysis, the serum sFGL2 level was independently associated with the presence and severity of CAD. Conclusion The serum sFGL2 levels are significantly lower in the presence of CAD and correlate with the severity of the disease. Further clinical studies are needed to confirm the use of sFGL2 as a biomarker for the detection and extent of CAD.

Key words: coronary artery disease, Gensini score, T regulatory cell, soluble fibrinogen-like protein 2

(Intern Med 55: 2343-2350, 2016) (DOI: 10.2169/internalmedicine.55.6149)

teins super family and has been demonstrated to be a novel Introduction effector mainly secreted by CD4+CD25+ Tregs (5). The FGL2 was originally cloned from cytotoxic T lympho- Inflammation is implicated in the development and rup- cyte (CTL), and the encoded protein shares 36% homology ture of atheromatous plaques, and there is considerable evi- with the fibrinogen β-andγ-chains and 40% homology with dence supporting the involvement of CD4+CD25+ T regula- the fibrinogen-related domain (FRED) of tenascin (6-8). To tory cells (Tregs) in this process (1, 2). A previous study date, two forms of the FGL2 protein have been character- demonstrated that naturally occurring CD4+CD25+ Treg ized, the membrane bound FGL2 (mFGL2) and the soluble numbers were reduced and functional properties were com- FGL2. The former form is a 70 kDa molecule that exerts a promised in patients with acute coronary syndrome procoagulative activity (9) and promotes (10), (ACS) (2). Animal experiments have also confirmed that generating thrombin directly and playing an important role upregulated CD4+CD25+ Tregs can attenuate both early and in innate immunity (6). sFGL2, which has a 50 kDa weight, advanced atherosclerotic lesions (3, 4). Soluble fibrinogen- exhibits immunoregulatory and contradictory properties dur- like protein 2 (sFGL2) belongs to the fibrinogen-related pro- ing tissue injury (11). A recent study has confirmed that

１School of Nursing, Anhui University of Traditional Chinese Medicine, China and ２Department of Cardiology, Second Affiliated Hospital of An- hui Medical University, China Received for publication July 10, 2015; Accepted for publication November 19, 2015 Correspondence to Dr. Fei He, [email protected]

2343 Intern Med 55: 2343-2350, 2016 DOI: 10.2169/internalmedicine.55.6149 sFGL2 is an important molecule of Tregs in its function and and PE labelled anti-CD127 (BD Biosciences-Pharmingen). development (12), similar to transforming growth factor beta Briefly, 5×105 fresh PBMCs were incubated with specific (TGF-β) and interleukin-10 (IL-10) (13). The administration antibodies conjugated with fluorochrome (0.42 μg/mL of of a neutralizing antibody of sFGL2 inhibited the Treg activ- CD25-APC, 0.33 μg/mL of CD4-FITC and 0.33 μg/mL of ity in vitro (14-16). Hence, sFGL2 might play a more criti- CD127-PE) for 30 minutes at 4°C. The cells were then cal role in the Treg function. Previous studies have demon- washed using phosphate-buffered saline (PBS), resuspended strated the role of TGF-β and IL-10 in atherosclerosis and in Pharmingen Stain Buffer (PSB) and analyzed using a coronary artery disease (CAD) (17-19). In the present study, flow cytometer (FACSCalibur, BD Biosciences, San Diego, therefore, we aimed to investigate the association between USA). serum sFGL2 levels and the severity of CAD as assessed by Serum levels of sFGL2 validated angiographic indexes. The serum sFGL2 concentrations were determined using a Materials and Methods commercially available enzyme-linked immunosorbent assay (ELISA) kit (Bio Legend, San Diego, USA) according to the manufacturer’s instructions. Briefly, serum samples (30 Participants μL) were added to each well of a plate precoated with anti- The study population consisted of 102 consenting patients human FGL2 antibodies. Following 2-hour incubation at who underwent first-time coronary angiography for sus- room temperature and four washes with buffer, 100 μL of pected CAD from November 2012 to February 2014 at No. human FGL2 detection antibody were added to each well 2 Hospital of Anhui Medical University. Patients with histo- for 1-hour incubation at room temperature. The plate was ries of malignant disease, recent myocardial infarctions, un- washed again and treated with secondary horseradish stable angina, major trauma or surgery, renal insufficiency, peroxidase-conjugated antibodies. After another five washes, acute or chronic infectious disease, any kind of immune- 100 μL of substrate solution were then added for a 20- mediated disease, or recent prescription of statins were ex- minute incubation in the dark. After the reaction was cluded from the study. The diagnosis of CAD was con- stopped, the absorbance was measured at 450 nm using an firmed by coronary angiography using a quantitative coro- ELISA plate reader. The concentration was determined using nary angiographic system. CAD was defined as significant a standard curve according to the kit’s instructions. coronary stenosis with at least one main coronary vessel Measurement of inflammatory cytokines with !50% luminal narrowing. Non-CAD patients were de- fined as those with no stenosis or <50% diameter narrowing The levels of cytokines including TGF-β1, IL-10, high according to the quantitative coronary analysis (QCA) meas- sensitivity C-reactive protein (hs-CRP) and IL-10 were ana- urement. The study protocol was approved by our ethics lyzed using an ELISA, according to the manufacturer’s pro- committee, and all patients provided their informed consent. tocol (Invitrogen, Rochester, NY, USA). Prior to analyzing the TGF-β1 levels, 100 μL of serum were activated by add- Blood samples preparation ing 4 μL of 1 N HCl and incubated at room temperature for Peripheral blood samples were taken from an antecubital 15 minutes and neutralized using 3 μL 1 N NaOH. The acti- vein after overnight fasting and before angiographic proce- vated serum was then diluted 4× with assay buffer and dis- dures. Blood samples for sFGL2 determination were drawn, pensed into 96-well plates which were pre-coated with TGF- centrifuged, and stored at -80°C for subsequent assays. Be- β1 antibody, which was followed by the addition of the de- fore coronary angiography, complete blood counts and se- tection antibody and streptavidin-horseradish peroxidase rum creatinine and serum lipid profiles were determined in (HRP) conjugate. Specific binding was visualized after add- all patients. All patients had normal serum creatinine levels ing the substrate solution. A standard curve (log-log curve and normal white blood cell counts. Total cholesterol, tri- fit) was generated using respective recombinant human cy- glycerides, and high-density lipoprotein cholesterol were tokine and the absolute concentration of the cytokine in the measured using enzymatic methods after overnight fasting. serum was determined from this curve. The detection of IL- The low-density lipoprotein cholesterol concentration was 10, hs-CRP and IL-10 levels were performed as described calculated using the Friedewald formula. above without the serum activation procedure. Determination of circulating CD4+CD25+CD127low Coronary angiograms and quantitative analysis Tregs Coronary angiograms were obtained according to standard Peripheral blood mononuclear cells (PBMCs) were iso- techniques, and the severity of stenosis was assessed us- lated from the peripheral blood using a Ficoll gradient as ing quantitative coronary angiography as previously de- described previously (20). The presence of Tregs was deter- scribed (21). The angiograms and quantitative coronary an- mined by measuring the following markers: PC5-labelled giographic analysis were evaluated by 2 experienced inter- anti-CD4 (BD Biosciences-Pharmingen, San Diego, USA), ventional cardiologists blinded to the clinical information FITC-labelled anti-CD25 (eBioscience, San Diego, USA) and were scored according to 2 scoring systems: (1) the

2344 Intern Med 55: 2343-2350, 2016 DOI: 10.2169/internalmedicine.55.6149

Table 1. Clinical and Angiographic Characteristics of Patients with and without CAD.

Variables Total Non-CAD (n=50) CAD (n=52) t/x2 p Men/Women 51/51 27/23 24/28 0.428 0.553 Age(years) 60±11 59±12 60±10 -0.629 0.531 Body mass index(Kg/m2) 24.21±3.99 23.53±3.68 24.86±4.20 -1.698 0.093 Hypertension 52 27 25 0.358 0.560 Diabetes 48 22 26 0.368 0.559 Current Smoker 17 9 8 0.126 0.794 FPG (mmol/L) 6.24±0.92 6.22±0.88 6.27±0.97 -0.270 0.788 Triglycerides(mmol/L) 1.66±0.61 1.62±0.47 1.71±0.72 -0.756 0.451 Cholesterol(mmol/L) 5.13±0.79 5.04±0.80 5.22±0.79 -1.147 0.254 LDL-C(mmol/L) 3.41±0.77 3.26±0.85 3.56±0.66 -2.022 0.046 HDL-C(mmol/L) 0.96±0.21 1.05±0.20 0.88±0.18 4.294 0.000 White blood cell count (×109/L) 6.78±1.30 6.95±1.23 6.62±1.36 1.306 0.194 Lymphocytes count (×109/L) 1.73±0.41 1.80±0.43 1.66±0.38 1.727 0.087 Treg (%) 4.74±2.01 6.39±1.16 3.15±1.19 13.931 0.000 sFGL2(ng/mL) 18.25±11.13 26.95±8.53 9.88±5.46 11.978 0.000 No. of diseased vessels 0 50 1 29 2 12 3 11 Location of stenosis Left main coronary artery 8/122 Left anterior descending coronary artery 52/122 Left circumflex coronary artery 25/122 Right coronary artery 23/122 Gensini score 24±21 8±6 39±17 -12.295 0.000 Data are expressed as mean±SD, number (percentage), or median (interquartile range). HDL: high-density lipoprotein, LDL: low-density lipoprotein

possible scores of this index ranged from 0 to 3 diseased multivariate linear regression analysis to determine inde- vessels. The criterion for 1-, 2-, or 3-vessel disease was a pendent predictors of the serum sFGL2 level and a high 50% reduction in the internal diameter of the right, left ante- Gensini score. A 5% type I error level was used to infer sta- rior descending, or left circumflex coronary artery. A 50% tistical significance. A p level of less than 0.05 was consid- reduction in the internal diameter of the left main coronary ered to be statistically significant. artery was considered to indicate 2-vessel disease. (2) In the Gensini scoring system, narrowing of the lumen was graded Results as follows: 1 for 1-25% occlusion; 2 for 26-50% occlusion; 4 for 51-75% occlusion; 8 for 76-90% occlusion; 16 for 91- Clinical characteristics of CAD and non-CAD pa- 99% occlusion; and 32 for total occlusion. This score is tients multiplied by a factor accounting for the importance of the lesion position in the coronary arterial tree, such as 5 for The demographic and clinical characteristics of the pa- LM, 2.5 for proximal LAD, and 1 for proximal right coro- tients included in this study are tabulated in Table 1. The nary artery (RCA). The severity of the disease is expressed mean age of the CAD patients was 60±10 years. The mean as the sum of the scores for individual lesions (22). age of non-CAD patients was 59±12 years. The CAD group had higher low-density cholesterol (3.56±0.66 vs. 3.26±0.85 Statistics mmol/L, p=0.046) and lower high-density cholesterol (0.88± All statistical analyses were performed using the statistical 0.18 vs. 1.05±0.20 mmol/L, p<0.001) than the non-CAD software package SPSS 13.0 (SPSS Inc., Chicago, USA). group. There was no difference between the two groups in The independent t-test was used to compare the differences other traditional cardiovascular risks, such as age, body in the immune cells/cytokines levels between CAD and non- mass index, hypertension, diabetes, or triglyceride level. CAD patients. A one-way analysis of variance (ANOVA) Comparison of the number of circulating Tregs and was adopted to compare the differences in the immune cells/ serum sFGL2 levels between patients with CAD and cytokines levels between groups divided by the number of non-CAD patients diseased vessels. Spearman’s correlation test was used to as- sess the correlation between the sFGL2 level and Gensini A previous study demonstrated that the expression of hu- score, CD4+CD25+CD127low Tregs, and cardiovascular risk man Treg surface marker CD127 was downregulated on a factors. For the multivariate analysis, potential factors identi- subset of CD4+ T cells in the peripheral blood, which nega- fied by univariate analyses were further entered into the tively correlated with the suppression function of such

2345 Intern Med 55: 2343-2350, 2016 DOI: 10.2169/internalmedicine.55.6149

Figure 1. Comparison of circulating subsets between Non-CAD and CAD patients. A) The percentage of CD4+CD25+CD127low Tregs in Non-CAD patients analyzed by flow cytometry. B) The percentage of CD4+CD25+CD127low Tregs in CAD patients analyzed by flow cytometry. C) The num- ber of Tregs was significantly decreased in CAD patients. D) Percentages of Tregs in CAD patients categorized according to the number of diseased vessels. ★: Statistically significant difference in the percentage of Tregs compared with the no stenosis group.

T cells (23). Thus, the combination of CD4, CD25, and (defined by stenosis which resulted in more than 50% re- CD127 resulted in a highly purified population of Tregs ac- duction in the internal diameter of a coronary artery) on counting for significantly more cells that previously identi- coronary angiography, 50 patients were categorized as non- fied according to other cell surface markers. In the current CAD patients, while 52 patients were categorized as having study, the percentage of CD4+CD25+CD127low Tregs was sig- 1-, 2-, or 3-vessel disease (Table 1). We observed that nificantly decreased in CAD patients compared to non-CAD among patients categorized according to the number of dis- patients (3.15±1.19% vs. 6.39±1.16%, p<0.001, Fig. 1A-C eased vessels, there was a significant difference in the per- and Table 1). Furthermore, the serum sFGL2 levels in pa- centage of circulating CD4+CD25+CD127low Tregs (Fig. 1D tients with CAD were significantly higher than those in non- and Table 3), and an association between the percentage of CAD patients (26.95±8.53 vs. 9.88±5.46 ng/mL, p<0.001, circulating Tregs and the number of diseased vessels was Fig. 2). In addition, as shown in Table 2, the serum TGF-β1 evident (r=-0.829, p<0.001). Post hoc tests suggested that and IL-10 levels in patients with CAD were significantly Tregs of 3- (2.09±0.73%), 2- (3.01±0.56%), and 1-diseased lower than those in non-CAD patients while hs-CRP and IL- vessel patients (3.61±1.26%) were all significantly decreased 10 levels in patients with CAD were significantly higher compared with that of patients without CAD (6.39±1.16%, p than those in non-CAD patients. <0.001 for all). In addition, among patients categorized ac- cording to the number of diseased vessels, there was a sig- Comparison of the number of circulating Tregs and nificant difference in the serum sFGL2 levels (Fig. 2 and serum sFGL2 levels among patients categorized ac- Table 3), and an association between the number of diseased cording to the number of diseased vessels vessels and the serum sFGL2 levels was also evident (r=- According to the presence of significant luminal stenosis 0.860, p<0.001). Post hoc tests suggested that the serum lev-

2346 Intern Med 55: 2343-2350, 2016 DOI: 10.2169/internalmedicine.55.6149

Figure 2. Comparison of the serum sFGL2 levels between non-CAD and CAD patients. A) The se- rum sFGL2 levels were significantly decreased in CAD patients. B) The serum sFGL2 levels in CAD patients categorized according to the number of diseased vessels. ★: Statistically significant differ- ence in the serum sFGL2 levels compared with the no stenosis group.

Table 2. Comparison of Inflammatory Cytokines be- severity of CAD. The dependent variable was sFGL2 and tween Patients with and without CAD. the regression equation included HDL-c, inflammatory cy-

Non-CAD CAD patients tokines, BMI, Treg, and Gensini score. As a result, the Variables t p (n=50) (n=52) sFGL2 level was significantly and positively associated with TGF-ȕ 15.39±1.76 14.76±1.97 1.684 0.095 the number of circulating Tregs and negatively associated IL- 17.46±5.01 7.42±3.81 11.436 0.000 with the Gensini score (Table 5). Hs-CRP 2.40±0.94 5.01±1.48 -10.591 0.000 TNF-Į 120.39±12.09 145.66±15.86 -9.024 0.000 Discussion els of sFGL2 in 3- (4.25±1.51 ng/mL), 2- (6.65±2.30 ng/ The present study demonstrated that patients with angiog- mL), and 1-diseased vessel patients (13.36±4.68 ng/mL) raphically confirmed CAD had significantly decreased levels were all significantly decreased compared with that of pa- of sFGL2 compared to patients without CAD. Another find- tients without CAD (26.95±8.53 ng/mL, p<0.001). ing of our study was that the serum sFGL2 concentration significantly and positively correlated with the number of Correlation between the serum sFGL2 level with circulating CD4+CD25+CD127low Tregs. Furthermore, a sig- Tregs, cytokines, Gensini score, and cardiovascular nificant and negative correlation was observed between the risk factors serum sFGL2 concentration and the modified Gensini index The modified Gensini score in our study population was in patients who underwent first-time coronary angiography significantly higher in CAD patients than non-CAD patients for suspected CAD. (39±17 vs. 8±6, p<0.001). Post hoc tests suggested that the Atherosclerosis is the main cause of CAD. It is currently modified Gensini score of 3- (63.73±7.94), 2- (44.42±9.29), appreciated that chronic inflammatory cell-mediated immune and 1-diseased vessel patients (27.90±10.78) were all sig- responses are involved in the pathogenesis of atherosclero- nificantly higher than that of non-CAD patients (7.57±6.39, sis (24). Tregs are an anti-inflammatory subset of CD4+ T p<0.001). cells and a previous study has demonstrated significantly de- We then analyzed the relationship among the sFGL2 creased numbers of Tregs in patients with ACS compared level, Tregs, inflammatory cytokines, the Gensini score and with that of patients with stable angina and controls (25). cardiovascular risk factors. As shown in Table 4, we found Moreover, another study reported that the ability of Tregs to that the serum sFGL2 level positively correlated with the suppress responder CD4+CD25- T-cell proliferation in ACS circulating Treg numbers (r=0.827, p<0.001, Fig. 3A), IL-10 patients was significantly attenuated (26). In the current (r=0.615, p<0.001), TGF-β1 (r=0.226, p=0.022), and high- study, we observed that the circulating Treg number, charac- density lipoprotein (HDL)-c (r=0.349, p<0.001), while nega- terized by CD4, CD25, and CD127, was dramatically de- tively correlating with the Gensini score (r=-0.836, p<0.001, creased in CAD patients compared with controls, and an as- Fig. 3B), tumor necrosis factor (TNF)-α (r=-0.522, p< sociation between the number of circulating Tregs and the 0.001), hs-C-reactive protein (CRP) (r=-0.470, p<0.001), and number of diseased vessels was revealed. Consistent with BMI (r=-0.226, p=0.022). In addition, a multivariate linear our results, another study showed decreased numbers of pe- regression analysis using the stepwise method was per- ripheral blood CD4+CD25+CD127low Tregs in non-ST eleva- formed to further estimate the effects of the serum sFGL2 tion ACS patients compared with chronic stable angina and level with the variables mentioned above in the presence and chest pain syndrome patients (27). This is consistent with

2347 Intern Med 55: 2343-2350, 2016 DOI: 10.2169/internalmedicine.55.6149

Table 3. Comparison of Circulating Treg, Serum SFGL2 and Gensini Score in CAD Pa- tients Categorized according to Diseased Vessels.

Non-CAD 1-vessel 2-vessels 3-vessels Variables F p (n=50) (n=29) (n=12) (n=11) CD4+CD25+CD127low Treg 6.39±1.16 3.61±1.26 3.01±0.56 2.09±0.73 78.504 0.000 sFGL2(ng/mL) 26.95±8.53 13.36±4.68 6.65±2.30 4.25±1.51 63.183 0.000 Gensini Score 7.57±6.39 27.90±10.78 44.42±9.29 63.73±7.94 173.426 0.000

Table 4. Correlation of SFGL2 with and attenuating the function of CD8+ T cells. As result of Clinical Factors, Treg, Cytokines, and decreased numbers and the impaired function of Tregs in the Gensini Score. setting of ACS, this inhibitory effect might occur in patients

Variables r p value with CAD due to the reduction in the sFGL2 level. + Treg 0.827 0.000 In addition to regulating the CD8 T cell function, sFGL2 Gensini score -0.836 0.000 might be involved in the pathogenesis and progression of IL-10 0.615 0.000 CAD through several other mechanisms. Recent evidence TGF- 0.117 0.834 ȕ demonstrated that sFGL2 regulates many cellular processes TNF-Į -0.522 0.000 Hs-CRP -0.470 0.000 that could play an important role in plaque progression, rup- FPG -0.127 0.202 ture, or thrombosis. sFGL2 remarkably induced apoptosis of Triglycerides -0.077 0.441 tubular epithelial cells and hepatocytes through upregulating TC -0.066 0.511 pro-apoptotic , including caspase-3, indicating the po- LDL-c -0.135 0.175 HDL-c 0.349 0.000 tential functions of sFGL2 in tissue injury and remodel- WBC 0.169 0.089 ing (35). Mitogen-activated protein kinases (MAPKs) are Lympho 0.175 0.078 key signaling molecules that influence a broad range of cel- Age 0.138 0.166 lular processes, such as proliferation, differentiation, migra- BMI -0.226 0.022 tion and apoptosis (36). In addition, a recent study showed that downregulation of sFGL2 decreases phosphorylation of what has been described in other conditions with low-grade extracellular signal-regulated kinase (ERK) and c-Jun N- inflammation, such as obesity, where decreased Treg levels terminal kinase (JNK) (37). This suggests that the MAPK in individuals exhibiting elevated markers of systemic pathway may be involved in the downstream signaling of inflammation or impaired glucose tolerance were ob- sFGL2. Furthermore, nuclear factor-κB(NF-κB) plays an served (28). These results suggest that the determination of important role in various biological processes. Many inflam- the number of circulating Tregs using flow cytometry might matory genes relevant to the pathogenesis of atherosclerosis be a useful indicator of the severity of atherosclerosis. are regulated by NF-κB, the activated form of which is pre- Previous data have suggested that sFGL2 is an important sent in atherosclerotic plaques (38). NF-κB translocation molecule of the function and development of Tregs, acting DCs was significantly reduced in the presence of sFGL2, in- similarly to TGF- β and IL-10 (29, 30). Thus, it is not sur- dicating that sFGL2 influenced the DC-dependent T cell dif- prising that sFGL2 may be involved in the pathogenesis of ferentiation through a NF-κB signaling pathway (6). Taken CAD. In the present study, the sFGL2 levels were signifi- together, in addition to Treg per se, its novel effector mole- cantly decreased in patients with CAD. We also demon- cule sFGL2 is also a promising therapeutic target against the strated that the serum sFGL2 levels positively correlated inflammatory response during atherosclerosis. Understanding with the circulating Treg number and negatively correlated of the link between sFGL2 and the severity of coronary with the severity of coronary stenosis in these patients. atherosclerosis may help in the treatment of patients with However, the mechanism could be complicated and has not CAD. Therefore, further research on the mechanism of been completely elucidated. Previous evidence showed that sFGL2 in the development of CAD is warranted. the proportions of CD8+ T cells were significantly higher in Limitations both ACS and stable angina patients compared with controls (31) and highly activated in atherosclerotic There are some limitations associated with our study. plaques (32). It was also reported that sFGL2 can inhibit First, we conducted a retrospective analysis of a single cen- CD8+ T cell proliferation in a dose-dependent manner and ter cardiac catheterization databank, and thus we were lim- this effect could be blocked by neutralizing antibod- ited by a small study sample size. In addition, because we ies (33, 34). Furthermore, studies have revealed that sFGL2 used angiographic determination to estimate the CAD pro- prevented the maturation of bone marrow-derived dendritic gression, some of the patients in the non-progression group cells (DCs) and markedly reduced the capacity to stimulate may actually have some atherosclerotic progression that was T cell proliferation, rather than induce T cell apoptosis (33). not recognized. Thus, sFGL2 might affect CAD by inhibiting proliferation

2348 Intern Med 55: 2343-2350, 2016 DOI: 10.2169/internalmedicine.55.6149

Figure 3. Correlation between the serum sFGL2 level and the number of circulating Tregs and Gensini score.

Table 5. Multiple Linear Regression Analysis with protein 2/fibroleukin exhibits immunosuppressive properties: sup- SFGL2 as the Dependent Variable. pressing T cell proliferation and inhibiting maturation of bone marrow-derived dendritic cells. J Immunol 170: 4036-4044, 2003. Variables exp(ȕ 95% CI p value 7. Doolittle RF, McNamara K, Lin K. Correlating structure and func- Treg 2.395 1.453 ~ 3.337 0.000 tion during the evolution of fibrinogen-related domains. Protein Gensini score -0.260 -0.351 ~ -0.169 0.000 Sci 21: 1808-1823, 2012. IL-10 0.047 0.070 ~ 0.528 0.491 8. Clark DA, Ding JW, Yu G, Levy GA, Gorczynski RM. Fgl2 TGF-ȕ 0.055 0.111 ~ 0.961 0.272 prothrombinase expression in mouse trophoblast and decidua trig- TNF-Į 0.105 0.159 ~ 0.551 0.114 gers abortion but may be countered by OX-2. Mol Hum Reprod 7: Hs-CRP -0.047 -0.083 ~ 0.750 0.412 185-194, 2001. HDL-c 0.012 0.023 ~ 0.841 0.822 9. Yuwaraj S, Ding J, Liu M, Marsden PA, Levy GA. Genomic char- BMI -0.093 -0.187 ~ 0.965 0.062 acterization, localization, and functional expression of FGL2, the Multiple linear regression analysis was performed. The dependent variable human gene encoding fibroleukin: a novel human procoagulant. was sFGL2. The regression equation included body mass index, HDL-C, Treg, Genomics 71: 330-338, 2001. IL-10, TGF-ȕ, TNF-Į, Hs-CRP, and Gensini Score. 10. Ning Q, Sun Y, Han M, et al. Role of fibrinogen-like protein 2 prothrombinase/fibroleukin in experimental and human allograft rejection. J Immunol 174: 7403-7411, 2005. The authors state that they have no Conflict of Interest (COI). 11. Zhao Z, Wang L, Yang C, et al. Soluble FGL2 induced by tumor necrosis factor-alpha and interferon-gamma in CD4+ T cells Acknowledgement through MAPK pathway in human renal allograft acute rejection. J The authors thank Dr. Jun Li in the Department of Orthope- Surg Res 184: 1114-1122, 2013. dics, Second Hospital of Anhui Medical University, for his assis- 12. Herman AE, Freeman GJ, Mathis D, Benoist C. CD4+CD25+ T regulatory cells dependent on ICOS promote regulation of effector tance with the image acquisition. cells in the prediabetic lesion. J Exp Med 199: 1479-1489, 2004. 13. Tang Q, Boden EK, Henriksen KJ, Bour-Jordan H, Bi M, Jing Cheng and Yingying Chen contributed equally to this Bluestone JA. Distinct roles of CTLA-4 and TGF-beta in CD4+ work. CD25+ function. Eur J Immunol 34: 2996-3005, 2004. 14. Shalev I, Liu H, Koscik C, et al. Targeted deletion of fgl2 leads to References impaired regulatory T cell activity and development of autoim- mune glomerulonephritis. J Immunol 180: 249-260, 2008. 1. Cheng X, Yu X, Ding YJ, et al. The Th17/Treg imbalance in pa- 15. Sarangi PP, Sehrawat S, Suvas S, Rouse BT. IL-10 and natural tients with acute coronary syndrome. Clin Immunol 127: 89-97, regulatory T cells: two independent anti-inflammatory mechanisms 2008. in herpes simplex virus-induced ocular immunopathology. J Immu- 2. Mor A, Luboshits G, Planer D, Keren G, George J. Altered status nol 180: 6297-6306, 2008. of CD4(+)CD25(+) regulatory T cells in patients with acute coro- 16. Miyara M, Sakaguchi S. Natural regulatory T cells: mechanisms nary syndromes. Eur Heart J 27: 2530-2537, 2006. of suppression. Trends Mol Med 13: 108-116, 2007. 3. Ait-Oufella H, Salomon BL, Potteaux S, et al. Natural regulatory 17. Wang XL, Liu SX, Wilcken DE. Circulating transforming growth T cells control the development of atherosclerosis in mice. Nat factor beta 1 and coronary artery disease. Cardiovasc Res 34: 404- Med 12: 178-180, 2006. 410, 1997. 4. van Puijvelde GH, Hauer AD, de Vos P, et al. Induction of oral 18. Heeschen C, Dimmeler S, Hamm CW, et al. Serum level of the tolerance to oxidized low-density lipoprotein ameliorates athero- antiinflammatory cytokine interleukin-10 is an important prognos- sclerosis. Circulation 114: 1968-1976, 2006. tic determinant in patients with acute coronary syndromes. Circu- 5. Li XL, Menoret S, Bezie S, et al. Mechanism and localization of lation 107: 2109-2114, 2003. CD8 regulatory T cells in a heart transplant model of tolerance. J 19. Smith DA, Irving SD, Sheldon J, Cole D, Kaski JC. Serum levels Immunol 185: 823-833, 2010. of the antiinflammatory cytokine interleukin-10 are decreased in 6. Chan CW, Kay LS, Khadaroo RG, et al. Soluble fibrinogen-like patients with unstable angina. Circulation 104: 746-749, 2001.

2349 Intern Med 55: 2343-2350, 2016 DOI: 10.2169/internalmedicine.55.6149

20. La Manna G, Ghinatti G, Tazzari PL, et al. Neutrophil gelatinase- uble fibrinogen-like protein 2 induces tubular epithelial cell apop- associated lipocalin increases HLA-G(+)/FoxP3(+) T-regulatory tosis in renal transplantation. Exp Biol Med (Maywood) 239: 193- cell population in an in vitro model of PBMC. PLoS One 9: 201, 2014. e89497, 2014. 31. Bergstrom I, Backteman K, Lundberg A, Ernerudh J, Jonasson L. 21. Katritsis DG, Pantos I, Efstathopoulos EP, et al. Three-dimensional Persistent accumulation of interferon-gamma-producing CD8+ analysis of the left anterior descending coronary artery: compari- CD56+ T cells in blood from patients with coronary artery dis- son with conventional coronary angiograms. Coron Artery Dis 19: ease. Atherosclerosis 224: 515-520, 2012. 265-270, 2008. 32. Grivel JC, Ivanova O, Pinegina N, et al. Activation of T lympho- 22. Liang KW, Sheu WH, Lee WL, et al. Decreased circulating pro- cytes in atherosclerotic plaques. Arterioscler Thromb Vasc Biol 31: tective adiponectin level is associated with angiographic coronary 2929-2937, 2011. disease progression in patients with angina pectoris. Int J Cardiol 33. Chan CWY, Kay LS, Khadaroo RG, et al. Soluble fibrinogen-like 129: 76-80, 2008. protein 2/fibroleukin exhibits immunosuppressive properties: Sup- 23. Liu W, Putnam AL, Xu-Yu Z, et al. CD127 expression inversely pressing T cell proliferation and inhibiting maturation of bone correlates with FoxP3 and suppressive function of human CD4+ T marrow-derived dendritic cells. J Immunol 170: 4036-4044, 2003. reg cells. J Exp Med 203: 1701-1711, 2006. 34. Sun Y, Xi D, Ding W, Wang F, Zhou H, Ning Q. Soluble FGL2, a 24. Hansson GK, Robertson AK, Soderberg-Naucler C. Inflammation novel effector molecule of activated hepatic stellate cells, regulates and atherosclerosis. Annu Rev Pathol 1: 297-329, 2006. T-cell function in cirrhotic patients with hepatocellular carcinoma. 25. Cheng X, Yu X, Ding YJ, et al. The Th17/Treg imbalance in pa- Hepatol Int 8: 567-575, 2014. tients with acute coronary syndrome. Clin Immunol 127: 89-97, 35. Selzner N, Liu H, Boehnert MU, et al. FGL2/fibroleukin mediates 2008. hepatic reperfusion injury by induction of sinusoidal endothelial 26. Mor A, Luboshits G, Planer D, Keren G, George J. Altered status cell and hepatocyte apoptosis in mice. J Hepatol 56: 153-159, of CD4(+)CD25(+) regulatory T cells in patients with acute coro- 2012. nary syndromes. Eur Heart J 27: 2530-2537, 2006. 36. Osaki LH, Gama P. MAPKs and signal transduction in the control 27. Zhang WC, Wang J, Shu YW, et al. Impaired thymic export and of gastrointestinal epithelial cell proliferation and differentiation. increased apoptosis account for regulatory T cell defects in pa- IntJMolSci14: 10143-10161, 2013. tients with non-ST segment elevation acute coronary syndrome. J 37. Liu Y, Xu L, Zeng Q, et al. Downregulation of FGL2/prothrom- Bio Chem 287: 34157-34166, 2012. binase delays HCCLM6 xenograft tumour growth and decreases 28. Wagner NM, Brandhorst G, Czepluch F, et al. Circulating regula- tumour angiogenesis. Liver Int 32: 1585-1595, 2012. tory T cells are reduced in obesity and may identify subjects at in- 38. Monaco C, Paleolog E. Nuclear factor kappaB: a potential thera- creased metabolic and cardiovascular risk. Obesity (Silver Spring) peutic target in atherosclerosis and thrombosis. Cardiovasc Res 61: 21: 461-468, 2013. 671-682, 2004. 29. Zhao Z, Yang C, Li L, et al. Increased peripheral and local soluble FGL2 in the recovery of renal ischemia reperfusion injury in a The Internal Medicine is an Open Access article distributed under the Creative porcine kidney auto-transplantation model. J Transl Med 12: 53, Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. To 2014. view the details of this license, please visit (https://creativecommons.org/licenses/ 30. Zhao Z, Yang C, Wang L, et al. The regulatory T cell effector sol- by-nc-nd/4.0/).

Ⓒ 2016 The Japanese Society of Internal Medicine http://www.naika.or.jp/imonline/index.html

2350