Prokineticin 2 as a potential biomarker for the diagnosis of Kawasaki disease

Kawasaki disease is a pressing acute self-limiting inflammatory disorder disease which lack of specific biomarkers. Our research aims to investigate the serum expression of a novel immune regulator PK2 in children with Kawasaki disease and to evaluate the ability of PK2 to predict Kawasaki disease. A total of 70 children with Kawasaki disease in the Children’s Hospital of Chongqing Medical University who were first diagnosed, 20 children with common fever admitted to hospital due to bacterial infection during the same period, and 31 children underwent physical examination were included in this study. Venous blood was collected for complete blood count, CRP, ESR, PCT, and PK2 before clinical intervention. The predictive potential of PK2 as a biomarker for the diagnosis of Kawasaki disease was judged by correlation analysis, the receiver operating characteristic (ROC) and combined score. Compared with healthy children and children with common fever, children diagnosed with Kawasaki disease had significantly lower serum PK2 concentrations (median 28,503.7208 ng/ml, 26,242.5484 ng/ml, and 16,890.2452 ng/ml, respectively, Kruskal–Wallis test: p < 0.0001). Analysis of the existing indicators in other laboratories showed that WBC (Kruskal–Wallis test: p < 0.0001), PLT (Kruskal–Wallis test: p = 0.0018), CRP (Mann–Whitney U: p < 0.0001), ESR (Mann–Whitney U: p = 0.0092), NLR (Kruskal–Wallis test: p < 0.0001), and other indicators were significantly increased compared with healthy children and children with common fever, RBC (Kruskal–Wallis test: p < 0.0001), and Hg (Kruskal–Wallis test: p < 0.0001) were significantly decreased in children with Kawasaki disease conversely. In the analysis of the Spearman correlation, it was found that serum PK2 concentration and NLR ratio were significantly negatively correlated in children with Kawasaki disease (rs = -0.2613, p = 0.0301). In the analysis of the ROC curves, it was found that the area under the PK2 curve was 0.782 (95% confidence interval 0.683–0.862; p < 0.0001), the ESR was 0.697 (95% confidence interval 0.582–0.796; p = 0.0120), the CRP was 0.601 (95% confidence interval 0.683–0.862; p = 0.1805), and the NLR was 0.735 (95% confidence interval 0.631–0.823; p = 0.0026). PK2 can significantly predict Kawasaki disease independently of CRP and ESR (p < 0.0001). The combined score of PK2 and ESR can significantly improve the diagnostic performance of PK2 (AUC = 0.827, 95% CI 0.724–0.903, p < 0.0001). The sensitivity was 87.50%, the sensitivity was 75.81%, the positive likelihood ratio was 6.0648, and the Youden index was 0.6331. PK2 has the potential to be a biomarker for early diagnosis of Kawasaki disease, and the combined use of ESR can further improve its diagnostic performance. Our study identifies PK2 as an important biomarker for Kawasaki disease and provides a potential new diagnostic strategy for Kawasaki disease.


Introduction
Kawasaki disease (KD), also named as mucocutaneous lymph node syndrome (MLNS), frequently occurs in children younger than 5 years old [1].Kawasaki disease is a global disease [2], which has occurred in more than 60 countries up to now since it was first discovered and described by Dr. Tomisaku Kawasaki in Japan in 1967 [1].Its incidence varies by ethnicity and region and is increasing year by year, relatively high in Asia, especially in Japan.In 2019, Japan has reported the incidence of Kawasaki disease nationwide, from 218.6 per 100 thousand in 2008, up to 243.1 per 100 thousand in 2011, 330.2 per 100 thousand in 2015 [3,4].The most serious complications of Kawasaki diseases may cause coronary inflammation or aneurysms that may lead to myocardial ischemia, infarction, and sudden death, which bring about a cause of obtaining heart disease in developed countries [5], causing a heavy psychological and economic burden to the society and family.
Although a lot of basic researches and epidemiological investigations on the etiology and pathogenesis of Kawasaki disease have been conducted by many well-known experts and scholars at home and abroad, they are still unclear however.As for the cause, it relates to many causative factors such as immune inflammation [6], genetic susceptibility [7], and the environment [8] possibly.Its pathogenesis is complex, unclear yet.The abnormal activation of the immune system and the damage and dysfunction of vascular endothelial cells are the cause of important pathogenesis of Kawasaki disease [9].With reference to the new guideline [10], it is divided into complete Kawasaki disease and incomplete Kawasaki disease according to clinical symptoms, Kawasaki disease with coronary artery injury and Kawasaki disease without coronary artery injury to the presence or absence of coronary artery injury.At present, Kawasaki disease, lack of specific biomarkers, is mainly diagnosed by clinical manifestations [3], but easily confused with the symptoms of febrile diseases such as scarlet fever and measles, wherefore it is liable to cause misdiagnosis and missed diagnosis of Kawasaki disease.Intravenous immunoglobulin combined with aspirin is currently the first-line drug for clinical treatment of Kawasaki disease [11], which mainly reduces the inflammatory response and the incidence of coronary artery disease [12].Therefore, further exploring the pathogenesis of Kawasaki disease and finding potential laboratory diagnostic markers are of great value for the early diagnosis and treatment of Kawasaki disease.
Prokineticin 2 (PK2), also known as Bv8, a biologically active peptide composed of 81 amino acids with a molecular weight of 8KD, is provided with multiple physiological functions such as regulating neurogenesis [13], angiogenesis [14], and inflammatory responses [15].Prokineticin 2 can be highly expressed in lymphoid organs, peripheral blood cells, dendritic cells and other immune cells and even participate in the function regulation of various immune cells, whose genes are located in the regions of human chromosome 3p21 and mouse chromosome 6 respectively [15].As a novel immune regulator, its immunological function has attracted more and more scholars' attention.PK2 can participate in the function of regulating important immune cells [16].Martucci C et al. reported that PK2 can promote macrophage infiltration by way of the PK2-PKR1, induce inflammatory phenotype macrophages (M1) to secrete IL-1 and IL-12, suppress the expression of IL-10 [17], drive mobilization neutrophil to inflammation site [18], as well as inhibit the secretion of typical anti-inflammatory factors IL-10 and IL-4 in mouse splenocytes by PK2-PKR1 pathway [19].In addition, Silvia Franchi et al. proposed that PK2 has potential value as a regulator of lymphatic function even more.Many studies have been reported about the roles in various autoimmune disease such as EAE [20], rheumatoid arthritis [16], psoriasis [21], and metabolic diseases such as obesity [22] and the inflammatory diseases like sepsis [23].
Taking these factors above into consideration, in addition, related research about the immunomodulatory correlation between PK2 and Kawasaki disease has not yet been reported until now.Therefore, we detected the concentration of PK2 in peripheral blood of children with Kawasaki disease for the first time, compared it with the existing laboratory auxiliary metrics to explore the value as a potential marker for early clinical diagnosis of Kawasaki disease.

Study subjects' recruitment
According to the diagnostic criteria proposed by the Japanese Kawasaki Disease Research Committee, we recruited a total of 70 children first diagnosed with Kawasaki disease (sex: male/female: 38/32; age: 27.00 mouth (29.50)) in the Children's Hospital of Chongqing Medical University to participate in this study from October 2019 to June 2021.Simultaneously, 20 children with common fever (sex: male/ female:12/8; age: 27.50 mouth (33.25)) who were admitted to hospital owing to bacterial infection during the same time period were recruited as the fever control group (the fever) and 31 healthy children (sex: male/female: 18/13; age: 48.00 mouth (72.00)) who participated in the physical examination were recruited as the normal control group (the control).This study was approved by the Ethics Review Committee of Children's Hospital Affiliated to Chongqing Medical University.
According to the typical degree of clinical symptoms, we divided the recruited Kawasaki disease subjects into the incomplete Kawasaki disease group consisting of 32 children and the complete Kawasaki disease group consisting of 38 children.According to the presence or absence of coronary artery injury in accordance with the size of the Z value in the echocardiographic within two weeks of onset, we divided them into a Kawasaki disease with coronary artery injury group consisting of 34 children(the CAL, sex: male/ female: 16/18; age: 22.50 mouth (25.75)) and a Kawasaki disease without coronary artery injury sick group consisting of 36 children (the NCAL, sex: male/female:22/14; age: 28.50 mouth (27.75)).Within one week of the onset, collect 2-4 mL of venous blood from the child before the first intravenous injection of immune globulin.After centrifugation at 3000g × 6min, immediately transfer them to a − 80 °C refrigerator for storage until testing.Repeated freezing and thawing should be avoided.

Venous blood sample
Peripheral blood concentrations of prokineticin 2 (Mybiosource, MBS2022546, California, USA) was assayed using an enzyme-linked immunosorbent assay (ELISA) according to the manufacturer's instructions.

Statistical analysis
The Shapiro-Wilk test was used to test the normality of measurement data before statistical analysis, and the standard deviation and mean were used to represent the normally distributed data; the interquartile range and the median were used to represent the non-normally distributed data.The non-parametric test (Mann-Whitney U test) was used to compare the data between the two groups of non-normally distributed data.The Kruskal-Wallis test was used to compare data between more than 2 groups of non-normally distributed data.Correlation analysis of non-normally distributed data was performed using Spearman's test.All above statistical analyses were performed using SPSS software (SPSS, version 21.0, Chicago, Illinois, USA) and were twosided with a significance level of P<0.05.
The receiver operating characteristic (ROC) was used to evaluate the diagnostic performance of different laboratory indicators, and the area under the curve (AUC), specificity, sensitivity, positive likelihood ratio (LR+), negative likelihood ratio (LR-), and Youden index were calculated.Comparisons between ROC curves were performed using the log-rank test.The scoring system for combination of biomarkers was applied.The above statistical analysis was performed using MedCalc software, and all were two-sided tests with a significance level of P< 0.05.

The serum concentration of PK2 in children with Kawasaki disease was significantly decreased
In order to explore the expression of PK2 in children with Kawasaki disease, ELISA was used to detect the serum concentration of PK2 from 31 healthy children, 20 children with common fever, and 70 children with Kawasaki disease.The results are shown in Figure 1.Compared with normal children and children with common fever, the level of PK2 with Kawasaki disease was significantly lower, and the difference was statistically significant.

Other laboratory auxiliary indicators in children with Kawasaki disease
At present, there is no specific diagnostic indicator for Kawasaki disease.Clinicians mainly make comprehensive judgments based on clinical symptoms combined with laboratory indicators.Therefore, we collected and sorted out the common laboratory auxiliary indicators of children with Kawasaki disease.Through the analysis of Kruskal-Wallis test and Mann-Whitney U test, results are shown in Figure 2 and Table 1, compared with healthy children and children with common fever, and the levels of WBC, PLT, NLR, and CRP in children diagnosed as Kawasaki disease were significantly increased, RBC and Hg on the contrary yet, which was consistent with the conclusions reported in the previous literature [24,25].Compared with children with common fever, Kawasaki disease children had obviously higher ESR, and all the differences were statistically significant.What is different from previous literature reports in our experiment were the days of fever and a common indicator reflecting bacterial infection-PCT, although increased in children with  Kawasaki disease, which were not significantly statistical difference compared with the results of children with ordinary fever.
Table 1 shows clinical characteristics of study population.Data were presented as median and interquartile range (IQR) with parentheses; P values from Kruskal-Wallis test or Mann-Whitney U test were appropriate.

Correlations between PK2 and patient characteristics
In the above results, we found that the serum PK2 was significantly decreased in children with Kawasaki disease, suggesting that PK2 may have the potential to be a biomarker in Kawasaki disease.Thus, so as to make a thorough inquiry the function of PK2 in Kawasaki disease, we first conducted a correlation analysis with other existing auxiliary indicators in laboratory.The results are shown in Fig. 3.After Fig. 3 Correlations between PK2 and patient characteristics.The correlation coefficient and P value (Spearman's test) were shown analyzing the correlation coefficient and P value, we concluded that PK2 was negatively associated with NLR in Kawasaki disease.

PK2 has better diagnostic performance in Kawasaki disease
To probe the ability of PK2 and other laboratory parameters to distinguish children with Kawasaki disease from children with common fever, ROC curves were used to evaluate their diagnostic performance.The results are shown in Fig. 4A, indicating that PK2 performed the best in distinguishing common fever from Kawasaki disease, with an AUC of 0.782.To further demonstrate its predictive power, we plotted Table 2, showing that PK2 was higher than other biomarkers in specificity, positive likelihood ratio, and Youden index.

Predictive value of multiple infection biomarker combination scores for Kawasaki disease
According to the analysis results of the ROC curve in Fig. 4A, we tried to further combine the commonly used clinical laboratory biomarkers (WBC, PLT, CRP, ESR, and NLR) with PK2.We tested combinations including two and three biomarkers, and the results were shown in Fig. 4 B and Table 2.The results showed that combining PK2 and WBC increased its AUC value from 0.782 to 0.802, and its specificity increased from 80 to 95%; combining PK2 and ESR increased its AUC value from 0.782 increased to 0.827, and its specificity increased from 80 to 87.50%; when WBC, ESR, and PK2 were combined, the AUC value was adjusted from 0.827 to 0.826, and the specificity did not change significantly.Therefore, combining PK2 with ESR has better early predictive value for Kawasaki disease.

Serum expression of PK2 differed in different degrees of coronary artery injury
Coronary artery injury is the most common and serious complication of Kawasaki disease.Thence, to quest whether the expression level of PK2 in the clinical manifestations of Kawasaki disease is complete and whether there is coronary artery injury, we further divided the Kawasaki disease group into incomplete Kawasaki disease group (incomplete KD) and complete Kawasaki disease group (complete KD) according to the latest guideline, as well as Kawasaki disease group without coronary artery injury (the NCAL) and with coronary artery injury (the CAL).Analyzing the ELISA test results, we discovered the results shown in Fig. 5. Interestingly, as shown in Fig. 5A, compared with the incomplete KD, the serum PK2 in the complete KD was lower; compared with the NCAL, PK2 was higher in the CAL shown in Fig. 5B.According to the Z value, further analysis that PK2 was different in different severity of coronary injury, showing a positive correlation trend, as shown in Fig. 5C.

Discussion
Kawasaki disease is a pressing acute self-limiting inflammatory disorder disease which main pathological manifestation is inflammatory lesions of small and medium blood vessels in whole body.The pathogenesis of Kawasaki disease is complex and uncertain.Inflammatory cytokines such as TNF-α, IL-6, and IL-1 are drastically released, leading to the production of inflammatory response, the loss and dysfunction of vascular endothelial cells simultaneously leads to vasculitis stimulated by various pathological factors [9].The existing laboratory indicators such as CRP, ESR, and WBC lack specificity and sensitivity [11].Acute or persistent fever, oral mucositis, conjunctival congestion, and lymphadenopathy are typical clinical symptoms of Kawasaki disease.Incomplete Kawasaki disease is not typical because of the symptoms and easily missed.Once the timely treatment is missing, the inflammatory lesions of coronary artery and the coronary artery even are more likely to occur.Although the clinical treatment of Kawasaki disease is diverse, these approaches have limitations in IVIG-resistant children with Kawasaki disease, whose efficacy is frequently minimal [10,26].PK2 just right plays an important regulatory role in many autoimmune system diseases.PK2 in experimental autoimmune encephalomyelitis can adjust the TH1 and TH17 reactions to present inflammation by promoting the promotion of inflammation.PK2 can up-regulate typical pro-inflammatory factor IFN-gamma and IL-17A secretion and inhibit IL-10 secretion [20].In the mouse arthritis induced by collagen in vivo, the PK2 antagonist-PKRA7 has the ability to mitigate the inflammation severity in mouse joints via inhibiting the release of cytokines like IL-6 and TNF-α, thereby exerting its regulatory effect on inflammation [27].A large number of research data indicate that cytokines such as IL-6, TNF-α, IFN-gamma, IL-17A, and IL-10 occupy a pivotable position in the progression of Kawasaki disease, and this is the reason to explore the relationship between Kawasaki disease and PK2, which is necessary and meaningful meanwhile [24,25].
From our research, we collected and detected the serum concentration of PK2 in children with Kawasaki disease by ELISA.It is found that compared with healthy children and children with common fever, the expression of PK2 in children with Kawasaki disease was amazingly decreased as shown in Fig. 1, suggesting that PK2 may be involved in the process of Kawasaki disease.NLR is a potential biomarker that can reflect the imbalance of immune balance and systemic inflammatory response in the body and is closely related to many autoimmune system diseases, allergic disease [28], and infectious diseases.For example, NLR is an excellent biomarker for predicting the severity of sepsis [29] and is also associated with prognosis in acute myocardial infarction [30].Through further correlation analysis, it was found that PK2 and NLR had a strongly negative correlation in the diagnosis of Kawasaki disease as shown in Fig. 3.At the same time, through the ROC curve, we found that PK2 had the best diagnostic performance in Kawasaki disease as shown in Fig. 4A and Table 2, indicating that PK2 has the ability to reflect the immune balance of Kawasaki disease in vivo, which means PK2 has the good potential to be a biomarker of state and inflammatory responses in Kawasaki disease.Although the specificity of PK2 was not superior, the area under the curve and specificity were both improved by combining with ESR as shown in Fig. 4B and Table 2, whose AUC value from 0.782 increased to 0.827, and its specificity increased from 80 to 87.50%.Thence, combining PK2 with ESR has better early predictive value for Kawasaki disease.
In this process, we found an interesting discovery, as shown in Fig. 5B and C, compared with children with Kawasaki disease without coronary artery injury, the concentrations of PK2 in children with coronary artery injury were higher, and the expression level of PK2 was correlated with coronary artery injury, which is related to the severity.As reported by Farbod Shojaei [14], PK2 has the function of promoting angiogenesis.In vitro studies have shown that PK2 can induce cardiac endothelial cell angiogenesis by binding to PKR1 receptor, and PK2 concentration is positively correlated with the degree of angiogenesis [31].Therefore, children with Kawasaki disease complicated with coronary artery injury showed increased feedback of PK2 in serum, which played a regulatory function of angiogenesis, which was also consistent with the results in Fig. 5C.It may be one of the reasons for its higher expression level.In order to verify our conjecture, we need to further study its specific mechanism in the future.
There are still some limitations in our study.First of all, we are a single central study, and samples in this research are mainly collected from Children's Hospital affiliated to Chongqing Medical University, and lack of Kawasaki disease data in other hospitals or regions.Secondly, we only collected and measured the Kawasaki disease children's specimens from October 2019 to June 2021, the collection time is too short, and the sample size is not enough; as a result, the overall situation cannot be fully and comprehensively reflected.Third, this study mainly focused on clinical studies, just only the human serum PK2 concentrations were detected.There are no internationally recognized methods for establishing the animal model of Kawasaki disease, and we cannot verify the expression of PK2 in animal models of Kawasaki disease.At the same time, this study did not explore the possible mechanism of PK2 in the development of Kawasaki disease.
In conclusion, PK2 has the potential to be a biomarker for early diagnosis of Kawasaki disease, and the combined use of ESR can further improve its diagnostic performance.Our study identifies PK2 as an important biomarker for Kawasaki disease and provides a potential new diagnostic strategy for Kawasaki disease.

Fig. 1 Fig. 2
Fig.1The serum concentration of PK2 in children with Kawasaki disease was significantly decreased.The control: the healthy control; the fever: the fever control group; The KD: the Kawasaki disease group.Kruskal-Wallis test was used to compare the differences between three groups of data.ns: p > 0.05.***: p < 0.0001

Fig. 4
Fig. 4 Predict ability of biomarkers to distinguish Kawasaki disease from common fever

Fig. 5
Fig. 5 Serum expression of PK2 differed in different degrees of coronary artery injury.Mann-Whitney U test was used in A and B. Kruskal-Wallis test was used in C. P values are shown in Figures.