Introduction

The S100 proteins are calcium-binding proinflammatory proteins with a wide spectrum of functions, including inflammation, cell proliferation, differentiation, migration, apoptosis, and metabolic functions [1]. The S100A12 protein, also known as calgranulin C, is a part of the S100 protein family predominantly secreted by neutrophil granulocytes and its extracellular secretion leads to activation of the inflammation cascade, cytokine production, and induction of oxidative stress. It has chemotactic activity and is linked with innate immunity [2, 3]. The S100A12 protein is significantly elevated in systemic juvenile arthritis (sJIA) compared to other autoinflammatory diseases [4]. Previous studies have suggested the use of S100A12 as a useful diagnostic tool in a number of inflammatory disorders [5,6,7,8].

Until recently, disorders of adaptive immunity were thought to play an exclusive and major role in the pathogenesis of childhood-onset systemic lupus erythematosus (SLE). However, novel research suggests that disorders of the immune system in this disease affect both innate and adaptive immunity [9]. Neutrophils are key cells of innate immunity and are involved in the pathogenesis of cSLE and sJIA, but in different ways [9]. Therefore, the two diseases, despite differences in clinical manifestations, share some common pathogenic features.

In this study we compared serum levels of S100A12 protein in children with newly diagnosed cSLE and sJIA, to examine the role of this molecule in innate immunity as a diagnostic and discriminative marker of inflammation and to indirectly point out the importance of neutrophils and innate immunity in the pathogenesis of these diseases. We also included a group of children with recurrent fevers in the absence of infection with high inflammatory markers (systemic undefined recurrent fevers [SURFS]). We chose these diseases for comparison because the inflammatory component is pronounced in all of them, at the beginning of each disease it is not easy to make a diagnosis, and there are recent studies that have described the important role of neutrophils [9], the main source of S100A12 [2], in these diseases.

Patients and methods

Patients

The study was designed as a cross-sectional study in which a total of 68 children were enrolled and divided into three groups: 19 children with newly diagnosed cSLE, 18 children with newly diagnosed sJIA, 7 children with SURFS, and the control group.

All cSLE patients included in this study fulfilled the American College of Rheumatology diagnostic criteria for SLE [10], while sJIA patients fulfilled the International League of Associations for Rheumatology (ILAR) diagnostic criteria [11]. Systemic Lupus Erythematosus Disease Activity Index 2000 (SLEDAI-2K) was used to assess disease activity in patients with cSLE [12].

For scoring disease activity in sJIA, we used the Systemic Juvenile Arthritis Disease Activity Score based on the count of any involved joint up to a maximum of ten joints, physician’s global assessment of disease activity, parent global assessment of wellbeing, the modified Systemic Manifestation Score, and erythrocyte sedimentation rate (sJADAS10) or C‑reactive protein (sJADAS10-CRP) [13]. Both sJADAS10 and sJADAS10-CRP scores range from 0 (no activity) to 50 (maximum activity).

Patients with three or more episodes of fever (body temperature ≥38 °C) more than 7 days apart in a period of at least 6 months in the absence of infection, accompanied by high inflammatory markers (either C‑reactive protein >50 mg/L, erythrocyte sedimentation rate >40 mm/h, leukocytes >20 × 109/L, or ferritin >500 μg/L) and without a defined diagnosis were included in the SURFS group. Patients with leukocytopenia were excluded.

The control group included 24 children admitted to hospital for reasons not related to autoimmunity, autoinflammation, or infectious diseases (e.g., tension or psychogenic headaches, functional abdominal pain). Children with leukocytopenia were excluded from the control group.

All the children were patients of the Division of Paediatric Immunology, Rheumatology and Allergology, University Hospital Centre Zagreb, University of Zagreb School of Medicine, Croatia. The study was approved by the Ethics Committee of the University of Zagreb School of Medicine. All procedures performed in the study were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent was obtained from all individual participants included in the study.

Procedures

For all patients, blood samples were collected at the time of diagnosis, prior to initiation of immunosuppressive treatment. All patients with SURFS had fever at the time of blood collection. The blood samples were taken between 9 and 10 am, and collected into vacutainer tubes with clot activator. The tubes were centrifuged at 3500 rpm for 5 min at room temperature and the collected serum was frozen at −80 °C until analysis.

Serum levels of S100A12 were determined with commercially available enzyme-linked immunosorbent assay (ELISA) kits (R&D Systems, USA) according to the manufacturer’s protocol. According to the literature, the normal range of S100A12 is 32–385 ng/mL [5].

The following routine laboratory tests were performed: C‑reactive protein, erythrocyte sedimentation rate, hemoglobin, leukocyte count, platelets, ferritin, and α2 globulin. For patients with cSLE, antinuclear antibodies (ANA; positive when ≥1:100 using HEp‑2 indirect immunofluorescence assay), anti-double stranded DNA antibodies (anti-dsDNA; positive when >120 IU/mL using multiplex, fluorescent bead-based immunoassay), and C3 (low when <0.9 g/L) and C4 (low when <0.1 g/L) complement components levels were also analyzed.

Statistical analysis

Numerical variables are descriptively presented with medians and interquartile ranges, while categorical data are presented as absolute and relative frequencies. Differences between the groups of patients and the control group were analyzed by Kruskal–Wallis test followed by a post-hoc test. A p-value <0.05 was considered as statistically significant. Correlations between variables were analyzed with the Spearman’s correlation test.

Results

Median time from symptom onset to blood collection was 2 months (range 0–4 months). Baseline features of patients in the cohort are summarized in Table 1. Patients with cSLE were statistically significantly older than other patients and controls. They had lower leukocyte counts in comparison to the patients with sJIA and SURFS. Five cSLE patients presented with renal involvement: four had biopsy-proven lupus nephritis class IV according to the histological classification of the International Society of Nephrology/Renal Pathology Society and one had lupus nephritis class V. Fourteen cSLE patients had arthritis.

Table 1 Clinical and laboratory findings of patients with childhood-onset systemic lupus erythematosus (cSLE), juvenile dermatomyositis (jDM), systemic juvenile arthritis (sJIA), systemic undefined recurrent fevers (SURFS), and children in the control group (CTRL)
Full size table

Patients with sJIA and SURFS had significantly higher C‑reactive protein levels in comparison to other patients and controls. When comparing inflammatory markers and generally used laboratory markers, including C‑reactive protein, erythrocyte sedimentation rate, hemoglobin, leukocyte count, platelets, ferritin, and α2 globulin, patients with sJIA and SURFS differed only in ferritin concentrations. Fourteen sJIA patients had arthritis at disease presentation, eleven had hepatomegaly, three splenomegaly, while no one had serositis.

The median (range) serum S100A12 levels were 469 (390–653) ng/mL in the cSLE group, 6103 (4087–7969) ng/mL in the sJIA group, 480 (453–763) ng/mL in the SURFS group, and 44 (40–72) ng/mL in the control group (Fig. 1). Serum S100A12 levels were significantly higher in patients with sJIA, cSLE, and SURFS compared to the control group (p < 0.0001). Patients with sJIA had significantly higher serum S100A12 levels compared to the patients with cSLE and SURFS (p < 0.0001). At the same time, there was no statistical difference in S100A12 serum levels between cSLE and SURFS patients (Table 2). S100A12 levels did not differ between patients with cSLE and nephritis and those without nephritis, nor between patients with cSLE and arthritis and those without arthritis. In patients with sJIA there was no difference in the level S100A12 regarding the presence of arthritis (at disease onset), hepatomegaly, or splenomegaly.

Fig. 1
figure 1

Boxplot of serum levels of S100A12 protein in the groups with sJIA, cSLE, SURFS, and CTRL. Serum S100A12 levels are plotted on the ordinate using a logarithmic scale. sJIA systemic juvenile arthritis, cSLE childhood-onset systemic lupus erythematosus, SURFS systemic undefined recurrent fevers, CTRL control group

Full size image
Table 2 S100A12 protein serum levels in patients with childhood-onset systemic lupus erythematosus (cSLE), systemic juvenile arthritis (sJIA), systemic undefined recurrent fevers (SURFS) and children in control group (CTRL)
Full size table

Among 19 patients with cSLE, 12 had a positive anti-dsDNA at the time of diagnosis. Patients who initially had negative anti-dsDNA during the follow-up period did not develop positive anti-dsDNA. The correlation between anti-dsDNA and S100A12 levels at the time of diagnosis was analyzed. We found a high positive correlation between serum S100A12 and anti-dsDNA levels in patients with cSLE (rs = 0.84, p = 0.002). S100A12 levels in serum furthermore strongly correlate with sJADAS10 (rs = 0.75, p = 0.003) and sJADAS10-CRP score (rs = 0.82, p = 0.0007) in sJIA patients. There was no correlation between S100A12 levels and ANA, C3 and C4 complement components, and SLEDAI-2K in patients with cSLE. Correlation was also absent between S100A12 and erythrocyte sedimentation rate, C‑reactive protein, hemoglobin, leukocyte count, platelets, ferritin, and α2 globulin in patients with cSLE, sJIA, and SURFS.

Receiver operating characteristic (ROC) curve analysis was performed to evaluate the sensitivity and specificity for various cut-off serum S100A12 levels. The cut-off serum S100A12 levels >763 ng/mL distinguished sJIA patients from patients with cSLE and different fever syndromes in the absence of infection with sensitivity of 100% and specificity of 100%.

Discussion

Currently, 25 members of the S100 protein family are known [1]. The most studied members of the S100 protein family are S100A4, S100A8, S100A9, and S100A12, and, among them, S100A8, S100A9, and S100A12 play the main role as mediators of the inflammatory response. S100A4 differs from other members of the S100 family in that it is involved in the pathogenesis of fibrotic diseases [8]. S100A12, on the other hand, stands out for its effect on neutrophils [3]. Namely, it is constitutively expressed in neutrophils and its role in promotion of neutrophil infiltration is significant. This study aimed to compare the serum levels of S100A12 protein in children with cSLE, sJIA, and SURFS as well as their correlation with clinical and laboratory findings at the onset of each disease in order to explore S100A12 as a potential diagnostic marker of inflammation and indirect marker of the role of neutrophils and innate immunity in the pathogenesis of these diseases.

We hypothesized that serum S100A12 protein levels would be elevated in cSLE and sJIA based on previous research on the importance of neutrophils, which secrete it, in the pathogenesis of all three diseases [9]. In children with sJIA, at initial presentation of the illness, it cannot be distinguished from systemic infection or other systemic diseases and different autoinflammatory diseases because of the absence of arthritis [14, 15]. cSLE and sJIA often present with constitutional symptoms [15, 16]. In our cohort of patients, we have shown statistically significant differences in serum S100A12 protein levels between children with sJIA and children with cSLE and other fever syndromes, which may indicate that serum S100A12 could be a discriminatory marker for distinguishing between sJIA and these diseases.

Evidence for neutrophil activation in SLE is contradictory and their exact role in pathogenesis has not yet been elucidated [9]. On the one hand there are studies on decreased neutrophil activity and impaired adhesion and phagocytosis in SLE [17], and on the other, there are studies on increased activity, especially in the vessel wall [18]. In patients with cSLE, neutrophils release neutrophil extracellular traps (NETs) [19, 20]. NETs can stimulate plasmacytoid dendritic cells to produce type I interferon, can cause endothelial cell damage, and serve as an antigen reservoir [21].

Recent data demonstrate use of S100 proteins as potential biomarkers of SLE and lupus nephritis [6,7,8]. In this study we found a positive correlation of serum S100A12 and anti-dsDNA levels in patients with cSLE, but there was no correlation between disease activity (SLEDAI-2K) and serum S100A12 levels. Although increases in anti-dsDNA titers are associated with disease activity, this does not apply to all patients [22]. Since it is known that the increase in anti-dsDNA levels could precede relapse of the disease [23, 24], our finding of a strong association between S100A12 and anti-dsDNA may indicate the use of this S100 protein in predicting relapse in patients with cSLE. We did not observe a difference in serum S100A12 levels between cSLE patients with and without nephritis, which is consistent with other research [7]. Considering the association of serum S100A12 levels in SLE with disease activity, there are conflicting results in the literature. Some studies showed that elevated serum levels of S100A12 were associated with disease activity in SLE patients [6, 8]. Nonetheless, the results of other research similar to ours have not confirmed this association [7]. Instead, it was found that higher urine S100 levels are associated with increased nephritis activity in cSLE [7].

In sJIA, neutrophils are generally activated and neutrophilia can sometimes be so severe that a leukemoid reaction occurs [9]. Neutrophils may be primary effectors and their aberrant activation, including other phagocytes, results in secretion of proinflammatory cytokines and the proteins S100A8, S100A9, and S100A12 [25]. In patients with sJIA, multiple elevated S100A12 values have been reported [4]. We found strong correlation between serum S100A12 and both sJADAS10 and sJADAS10-CRP score in sJIA patients, confirming the utility of S10012 in monitoring disease course [5].

In general, the small number of patients in all groups is a limitation of this study.

Since we detected significantly elevated serum S100A12 levels in children with newly diagnosed sJIA, cSLE, and different fever syndromes, but not in the control group, this may point to intense neutrophil activation, which may play an important role in chronic inflammation in these diseases. Considering that neutrophils are key cells in innate immunity, this finding is an indirect indicator that innate immune disorders are involved in the pathogenesis not only of sJIA, but also of cSLE. Nevertheless, because neutrophils secrete cytokines that can activate the cells of the adaptive immune system, they also form an important link between innate and adaptive immunity. In this way, the paradigm of rheumatic diseases being divided into autoimmune and autoinflammatory changes, and these diseases should be viewed as part of a continuum of disorders of immune regulation, where one or the other pattern overlap to varying degrees. In other words, cSLE can be seen as a disorder of immune regulation in which the autoimmune component is predominantly present, as opposed to sJIA in which the autoinflammatory component is predominantly present.

In summary, this research demonstrated elevated serum levels of S100A12 protein in children with newly diagnosed sJIA, cSLE, and different fever syndromes in the absence of infection with high inflammatory markers, but not in controls. Therefore, this protein may be used as a marker of inflammation in these diseases as well as a discriminative marker and indirect indicator of neutrophil activation and the role of an innate immunity disorder in their pathogenesis.

The contribution of neutrophils to the pathogenesis of rheumatic diseases is still insufficiently understood, but it arouses growing interest and further research is needed to elucidate their involvement with the aim of better diagnosis and development of new therapeutic modalities.