The relationship between TNF-like protein 1A and coronary artery aneurysms in children with Kawasaki disease

Kawasaki disease (KD) is an acute, systemic vasculitis of unknown etiology that occurs predominantly in infants and children, and the most crucial complication of KD is coronary artery aneurysm (CAA). Tumor necrosis factor (TNF)-like protein 1A (TL1A) is a member of the TNF superfamily, which possesses the ability of maintaining vascular homeostasis and regulating immune responses. This study aimed to examine serum TL1A levels in KD patients, and to investigate the relationship between TL1A and CAAs in children with KD. Blood samples were recruited from 119 KD patients, 35 febrile controls (FCs), and 37 healthy controls (HCs). The KD group was further divided into KD with CAAs (KD-CAAs) and KD non-CAAs (KD-NCAAs) groups. Serum TL1A levels were measured using enzyme-linked immunosorbent assays, and clinical parameters were collected from KD patients. Serum TL1A levels of KD patients in the acute phase of KD were significantly higher than in the FC and HC groups. In particular, serum TL1A levels were substantially increased in the KD-CAA group compared with the KD-NCAA group. Furthermore, TL1A levels in the KD group were positively correlated with the duration of fever and the time point of IVIG and WBC levels, but negatively correlated with levels of RBC, Hb and albumin. TL1A might be involved in KD-associated vasculitis and in the development of CAAs.


Introduction
Kawasaki disease (KD) has been defined as an acute, systemic vasculitis that mainly occurs in infants and children aged < 5 years [1]. It predominantly affects systemic medium-sized blood vessels. The most critical complication of KD is coronary artery lesions (CALs). Indeed, KD has replaced rheumatic fever as the leading cause of pediatric acquired cardiac disease in developed countries [1]. Although a single dose of 2 g/kg intravenous immunoglobulin (IVIG) combined with aspirin is now considered as the standard treatment strategy, approximately 15% of children with KD still develop CALs or coronary artery aneurysms (CAA), which may lead to coronary artery stenosis or thrombosis, ischemic heart disease or even myocardial infarction and sudden death [2,3]. The etiology and pathogenesis of KD are still unclear. However, in recent years, several studies have suggested that endothelial dysfunction and excessive inflammatory responses may play a key role in the pathogenesis of KD [4].
Tumor necrosis factor (TNF)-like protein 1A (TL1A), also known as vascular endothelial growth inhibitor, is a member of the TNF superfamily. It has a high degree of sequence homology with TNF-α [5]. TL1A is primarily produced by endothelial cells under stimulation of TNF-α and IL-1, and is also expressed by lymphocytes, macrophages, dendritic cells, chondrocytes and synovial cells [6]. TL1A stimulates T cell activation, proliferation, and the production of large amounts of essential cytokines via death receptor 3 (DR3), which is mainly expressed on lymphocytes [5]. TL1A plays an important role in apoptosis, modulation of 1 3 immune responses, vascular homeostasis, glucose and lipid metabolism [5,7,8]. McLaren et al. [9] have demonstrated that TL1A promotes the formation of foam cells in vitro by regulating the levels of low-density lipoprotein and cholesterol that are involved in atherosclerosis. Recent studies have also demonstrated a relationship between serum TL1A and the occurrence of coronary artery disease and acute coronary syndrome [10,11], and serum TL1A might be a specific biomarker for monitoring coronary artery disease. In addition, TL1A can promote the expression of MiR-29b by activating the JNK-GATA3 signaling pathway [12], and MiR-29b further mediates the down-regulation of extracellular matrix proteins, which play a key role in the development of aortic dilatation and aortic aneurysm [13]. However, there is still no report on the relationship between TL1A and CAAs in children with KD.

Patients and their general characteristics
This study enrolled 119 children with acute phase KD from the Children's Hospital of Chongqing Medical University between May 2020 and November 2020. All patients (80 males and 39 females, 2.23 ± 1.59 years old) met the guidelines proposed by the Japanese Kawasaki Disease Research Committee [14]. Subjects with a previous history of KD, metabolic diseases or any other associated immunological disease were excluded from the study. Thirty-five subjects (23 males and 12 females, 2.39 ± 2.29 years old) with acute febrile disease were recruited as a febrile control (FC) group, and 37 children (24 males and 13 females, 3.66 ± 0.55 years old) were included as a healthy control (HC) group. FC subjects included patients with bronchopneumonia, infectious diarrhea, urinary tract infection, sepsis, hand-foot-mouth disease (HFMD), infectious mononucleosis, scarlet fever, suppurative meningitis, and herpangina. This study was approved by the Ethics Committee of Children's Hospital, Chongqing Medical University and informed consent was obtained from the parents or appropriate guardians of all subjects.
Echocardiography was measured on the KD patients one day before administration of intravenous immunoglobulin (IVIG) or anticoagulants. Subjects with a z-value ≥ 2.5 were included in the KD with CAA group (KD-CAAs, n = 59), while those with z-value < 2.5 were included in the KD non-CAA group (KD-NCAAs, n = 60) [15].

Blood sample collection
All blood samples were collected one day before treatment with IVIG or anticoagulants for subjects in the KD patient group. The blood samples were immediately centrifuged at 3000 rpm for 10 min and the resulting serum samples were immediately stored at − 80 °C. The same procedure was followed to obtain serum samples from subjects in the FC and HC groups.

Measurement of serum TL1A concentrations and clinical parameters
Serum TL1A levels were measured using an enzyme-linked immunosorbent assay (ELISA) according to the manufacturer's instructions (Cloud-Clone Corp, China).

Statistical analysis
All statistical analyses were performed using the SPSS25.0 software for Windows (SPSS, Chicago, IL). A two-sided P value < 0.05 was considered statistically significant. χ 2 tests were used to compare frequencies between groups. Differences between groups were assessed using unpaired twotailed t tests. Spearman rank correlation was used to analyze the association between serum TL1A levels and various clinical parameters. All values are quoted as mean ± SD or number and percent (n, %). The presence of KD development was expressed by area under curve (AUC) using receiver operating characteristic (ROC) analysis.
There were no significant differences in MPV, PDW, AST or ALT between the KD and FC groups (P > 0.05). However, compared to the FC group, the levels of RBC, albumin, Hb and L% were decreased, and the levels of WBC, PLT, N% and CRP were increased in the KD group (P < 0.05). (suppl 1).

TL1A levels and clinical parameters in the KD-CAA and KD-NCAA groups
There were no significant differences in the duration of fever, the time point of IVIG, WBC, RBC, PLT, albumin, MPV, PDW, CRP, ESR, AST, ALT, or CK-MB between the KD-CAA and KD-NCAA groups (P > 0.05). However, compared to the KD-NCAA group, the levels of Hb and N% were decreased and the levels of TL1A and L% were increased in the KD-CAA group (P < 0.05) ( Table 1).

Correlations between TL1A levels and clinical parameters in the KD Group
In the KD group, N%, L%, MPV, PDW, ALT, AST, CK-MB, CRP, ESR were not significantly correlated with serum TL1A levels. However, in the KD group, serum TL1A levels were positively correlated with the duration of fever, time point of IVIG, WBC and PLT levels (P < 0.05), and were negatively correlated with RBC, Hb and albumin levels (P < 0.05, Table 2).

Correlations between TL1A levels and clinical parameters in the KD-CAA and KD-NCAA groups
Serum TL1A levels were positively correlated with the duration of fever, WBC, PLT and CRP levels in the KD-NCAA group. (P < 0.05), but were not significantly correlated with the duration of fever, time point of IVIG, WBC, Hb, N%, L%, MPV, PDW, ALT, AST, CK-MB, CRP, and ESR. Furthermore, serum TL1A levels were positively correlated  with the PLT levels (P < 0.001), and were negatively correlated with RBC and albumin. (P < 0.05) in the KD-CAA group (Table 3).

ROC curves of TL1A used to predict Kawasaki disease (KD)
The AUC value for the KD group was 0.8678 (95% confidence interval 0.8175-0.9180, P < 0.0001), compared to both FC and HC groups. The cut-off value of KD was 123.4, with a sensitivity of 88.89% and specificity of 75.63% (Fig. 2).

Discussion
It has been reported that TL1A can regulate metabolism [8,16] in physiological status and that it may play an important role in the modulation of vascular homeostasis and autoimmune diseases [5,7]. The pathogenesis of KD is well-known as being a result of inflammation and an imbalance of immunology resulting in vascular lesions. The purpose of this study was to ascertain the relationship between TL1A and CAA in children with KD. Our results showed that serum TL1A levels were significantly higher in KD patients compared with either healthy control patients (HC group) or patients with fevers from other causes (FC group). Serum TL1A levels were much higher in the KD-CAA group than in the KD-NCAA group. Furthermore, TL1A levels were positively correlated with the   duration of fever, time point of IVIG, WBC and PLT levels, but negatively correlated with RBC, Hb and albumin in the KD group. TL1A, an important molecule related to inflammation, is characterized by a metalloprotease cleavage site and a C-terminal TNF homology domain [17]. It has previously been shown that levels of TL1A are significantly increased in patients with acute coronary syndrome, and are associated with both endothelial dysfunction and coronary atherosclerosis [11]. Recent studies have shown that TL1A promotes foam cell formation in vitro by regulating the levels of low-density lipoprotein and cholesterol [9], and it switches vascular smooth muscle cell phenotypes to participate in the development of atherosclerosis in mouse models of apolipoprotein E-deficient atherosclerosis [18]. Here, we found that the levels of TL1A in the acute phase of KD patients were significantly higher than those in FC and HC groups, indicating that TL1A might be acting as a proinflammatory factor in the vasculitis of KD. Prior research has demonstrated that as a T cell co-stimulator, TL1A plays an important role in the immune response [17]. Furthermore, studies have shown that increased levels of TNF-α, IL-1β, IL-2R IL-6and IFN-γ are involved in the inflammation associated with the development of CALs, and TL1A is known to promote secretion of these cytokines [5,[19][20][21][22]. This information caused us to speculate that TL1A might be a key factor in the excessive inflammatory responses associated with KD. In addition, it is known that TL1A induces apoptosis of endothelial cells in an autocrine manner and activates a series of signaling pathways, such as the SAPK/ JNK, p38 MAPK and NF-κB signaling pathways [23]. Furthermore, TL1A induces dysfunction and injury of coronary endothelial cells in the acute stage of KD, which leads to the dilatation of coronary arteries, then subsequently results in the gradual development of inflammation-mediated lesions in the whole arterial layer, especially smooth muscle layer, leading to destruction of the vascular wall's supporting system and eventually to formation of CAA [24]. Studies have also found that TL1A promotes the expression of MiR-29b via JNK-GATA3 [12]. MiR-29b has been shown to destroy the integrity of the vascular wall by targeting the mRNA that codes for extracellular matrix proteins such as collagens, fibrillin and elastin. Moreover, TL1A is considered to promote smooth muscle cell apoptosis [13]. Taken together, all of this information indicates a potential role for TL1A in the inflammatory process and development of CAA in patients with KD. When compared to both FC and HC Groups, the AUC value for the KD group was 0.8678 (95% confidence interval 0.8175-0.9180, P < 0.0001). The cut-off value of serum TL1A levels was 123.4 pg/mL and the sensitivity and specificity of predicting were 88.89% and 75.63%, respectively. This provides further evidence that TL1A also may play an important role in the development of KD.
Our correlation analysis showed that TL1A levels in KD patients were negatively correlated with levels of albumin and hemoglobin. Anemia and hypoproteinemia are common clinical features in children with KD [25,26]. In this study, the levels of albumin and hemoglobin in the KD-CAA group were lower than those in the KD-NCAA group. Studies have shown that TL1A can increase reactive oxygen to accelerate the inflammatory response via the TL1A/TNFR2 pathway [27]. Furthermore, Straface et al. [28] demonstrated that systemic oxidative stress and premature senescence of RBCs may play a key role in anemia and hypoproteinemia in children with KD, so TL1A may promote the observed changes of RBC and Hb in KD patients. On the other hand, albumin loss might result from the increased vascular permeability accompanied with the endothelium damage associated with KD [26]. TL1A can induce endothelial cell apoptosis and excessive inflammation leading to the destruction of the vascular barrier [5]. In this study, we observed significantly higher TL1A serum levels in the KD-CAA group than that in the KD-NCAA group, while the serum albumin levels were negatively associated with TL1A. These results suggests that hypoalbuminemia could reflect the extent of the vascular damage, and that high serum concentration of TL1A might be one of the causes of hypoalbuminemia in KD.
Furthermore, we found that the levels of PLT were higher in the KD-CAA group than in the KD-NCAA group, although this difference was not statistically significant. Serum TL1A levels were positively correlated with PLT in the KD group, especially in the KD-CAA group. Previous study has reported that TL1A was able to enhance PDGF-BB [20] and PDGF-BB is produced mainly by platelets. It seemly hinted that TL1A may be involved in the formation of thrombus in children with KD-CAA.
However, there are limitations in our study. First, it was a single-center study and the number of participants was relatively small. Second, due to lack of data on serum TL1A levels at different time points of patients with KD, we were unable to reveal the trend of TL1A changes or analyze the recovery of KD-associated vasculitis. To provide a more comprehensive understanding of TL1A in KD, further studies are required to enroll subjects from multicenter and explore the role of TL1A in the development of KD and the formation of CAAs.

Conclusion
In summary, our study was the first to demonstrate that the serum TL1A was higher in KD patients than in patients in the FC and HC groups. We found that serum TL1A levels were substantially increased in the KD-CAA group when compared to the KD-NCAA group. Serum TL1A levels were correlated with RBC and albumin levels in the KD group.