Background
Systemic sclerosis (SSc) or scleroderma is an intractable autoimmune disorder characterized by vasculopathy and fibrosis of the skin and visceral organs including the heart, kidney, and lung, which has high mortality and reduced survival than other autoimmune diseases [
1]. Although SSc often involves multiple organs, lung diseases, such as interstitial lung disease (ILD) and pulmonary arterial hypertension (PAH), are considered the main causes of mortality [
2]. To date, the pathogenesis and mechanism of SSc have been poorly understood. Progress in the development of effective therapies for SSc has been slow [
3]. Thus, besides novel biomarkers reflecting the progression of the disease, there is also a need to identify the potential targets for accurate therapy in SSc or SSc-ILD.
Fibrosis is the most remarkable clinical characteristics of SSc, which leads to impaired organ function and scarring. In epithelial organs, particularly the lung, skin, kidney, and liver, fibrogenesis is triggered by a wide range of initial injuries to the epithelium, such as trauma, toxins, inflammation, and infection [
4]. In addition to epithelial tissues, initial damages to endothelium and vascular dysfunction also play important roles in SSc-related vascular damage [
5]. However, regardless of the initial injuries, the accumulation of activated myofibroblasts eventually leads to excessive generation of extracellular matrix proteins and disease-related collagen in the connective tissue of multiple organs [
6]. Although the explicit origin of activated myofibroblasts in SSc remains varied and uncertain, profibrotic cytokine-mediated epithelial-to-mesenchymal transition (EMT) is considered a crucial mechanism for the production of activated myofibroblasts and the development of fibrosis in SSc [
7,
8]. Given the key role of EMT in SSc, we reasoned that relevant molecules that are involved in EMT would be attractive therapeutic targets for SSc.
We were interested in the association in the previous studies between various EMT-related and fibrotic diseases and the increased activity or expression of calpains—a family of calcium-dependent, non-lysosomal neutral cysteine endopeptidases [
9]. Humans can express 15 calpain genes including
CAPN1 to
CAPN16 except for
CAPN4, which can be categorized as conventional and unconventional subunits [
10]. Conventional subunits that comprised of
CAPN1 and
CAPN2 are the most studied members of the calpain family, and both are ubiquitously expressed in all tissues, which function as the catalytic subunits of calpain-1 and calpain-2, respectively [
11,
12]. Calpain small subunit 1 (
CAPNS1), also known as
CAPN4, is the common regulatory subunit of calpain-1 and calpains-2 and plays an important role in maintaining the activity and stability of calpain [
13]. Notably, calpain activity disorder contributes to the pathogenesis of many EMT-related diseases such as idiopathic pulmonary fibrosis, arterial sclerosis, cardiovascular disorders, and cancers [
10]. Several studies have reported that calpain-1 may lead to transforming growth factor-β1 induced EMT by mediating the phosphoinositide 3-kinase/Akt signaling pathway in human lung epithelial cells [
14,
15]. Peng et al. reported that selective deletion of endothelial cell calpain reduced cardiac fibrosis and cardiac endothelial cell dysfunction in the mouse model of diabetes [
16]. Although there is accumulating evidence that targeted inhibition of calpain serves as a potential therapeutic strategy for calpain-related diseases in animal models and clinical trials, few studies have investigated the roles of calpains in autoimmune diseases [
10,
17]. Given the detrimental role of calpains in EMT-related and fibrotic diseases, we advanced a hypothesis that calpains would be a potential therapeutic target for SSc. In particular, the roles of calpains in the pathogenesis of SSc have never been investigated.
High mobility group box 1 (HMGB1), an endogenous damage-associated molecular pattern (DAMP), is secreted when macrophages and monocytes are activated and is passively released as a consequence of cell damage or necrosis [
18]. HMGB1 is also a nuclear nonhistone chromatin-binding protein, which participates in the regulation of transcription and stabilization of nucleosome structure [
19]. Several studies have reported that serum HMGB1 level is increased and is associated with platelet-derived microparticles indicating that platelets are a source of DAMP HMGB1 in patients with SSc [
20‐
23]. Moreover, active calpain is also associated with platelet-derived microparticles and is known as a source of active calpain in the blood [
24]. Thus, we proposed a novel hypothesis that calpain activity in the serum and the levels of HMGB-1 may be directly correlative. So far, studies concerning the correlation between calpain activity and HMGB1 level in the serum from SSc or SSc-ILD patients are still missing.
In this study, we attempted to investigate clinical correlations of serum calpain activity and HMGB1 levels with immunological and clinical traits in 31 Chinese patients with SSc, 20 healthy control subjects (HC), and 10 patients with other connective tissue diseases (CTD). Furthermore, integrative microarray datasets of lung samples and peripheral blood mononuclear cell (PBMC) samples from patients with SSc or SSc-ILD and HC were utilized to explore the underlying mechanism by which calpain exerts its function in the pathogenesis of SSc through bioinformatic analysis. Our findings may point to serum calpain activity and HMBG1 levels could be promising treatment targets for SSc or SSc-ILD, and provide powerful evidence for better understanding the pathogenesis of SSc.
Methods
Patients and controls
A total of 31 patients with a positive diagnosis as SSc according to ACR/EULAR 2013 classification criteria was recruited at Zhongshan Hospital (Fudan University, Shanghai, China) [
25]. Ten patients with other CTD were recruited as disease control subjects. Twenty people with no history of pulmonary, autoimmune, cardiovascular, or other diseases were recruited as HC. Detailed characteristics of SSc patients, HC, and patients with other CTD were shown (Table
1). The presence of ILD was defined by the identification of bibasilar fibrosis on chest high-resolution computed tomography or chest radiography [
25]. Patients presenting pulmonary vascular resistance ≥ 3 Wood units, mean arterial systolic pressure > 25 mmHg, and pulmonary capillary wedge pressure ≤ 15 mmHg were defined as PAH [
26]. The study was approved by the Zhongshan Hospital Research Ethics Committee. Written informed consent was acquired from all subjects.
Table 1Laboratory and clinical characteristics of the patients and control subjects included in the study
Age (mean ± SD) | 51 ± 13 | 49 ± 14 | 53 ± 12 | 47 ± 11 | 40 ± 11 |
Female,n(%) | 27 (87) | 10 (91) | 17 (85) | 8 (80) | 17 (85) |
Course of disease (mean ± SD) | 8 ± 7 | 8 ± 8 | 9 ± 8 | NA |
lSSc/dSSc % | 55/45 | 45/55 | 60/40 |
PAH,n(%) | 3 (10) | 0 (0) | 3 (15) |
Raynaud’s phenomenon,n(%) | 28 (90) | 9 (82) | 19 (45) |
Anti-Scl-70 antibody,n(%) | 13 (42) | 4 (36) | 9 (45) |
ANA,n(%) | 30 (97) | 10 (91) | 20 (100) |
Anti-RNP,n(%) | 8 (26) | 4 (36) | 4 (20) |
Anti-centromere,n(%) | 8 (26) | 2 (18) | 6 (30) |
mRSS (mean ± SD)n = 25 | 12 ± 10 | 21 ± 14 | 10 ± 7 |
Immunosuppressive therapy ever,n(%) | 11 (35) | 5 (45) | 6 (25) |
Cyclophosphamide,n(%) | 6 (19) | 3 (27) | 3 (15) |
Methotrexate,n(%) | 2 (6) | 0 (0) | 2 (5) |
Mycophenolate mofetil,n(%) | 3 (10) | 2 (18) | 1 (5) |
FVC % predicted,n = 17 (mean ± SD) | 74.5 ± 14.2 | 78 ± 8.5 | 75.6 ± 16.2 |
DLCO% predicted,n = 8 (mean ± SD) | 74.2 ± 22.3 | 70.9 ± 3.5 | 76.2 ± 27.8 |
FEV1/FVC%,n = 17 (mean ± SD) | 85.5 ± 11.6 | 76.2 ± 7.5 | 88.4 ± 11.1 |
ESR, mm/h (mean ± SD) | 17.10 ± 11 | 17.80 ± 14.36 | 16.76 ± 9.13 |
PLT, 109/L (mean ± SD) | 206 ± 67 | 209.40 ± 79.98 | 204.62 ± 61.60 |
MPV, fL (mean ± SD) | 11.95 ± 1.49 | 11.23 ± 1.95 | 10.81 ± 1.24 |
PDW, fL (mean ± SD) | 0.22 ± 0.07 | 0.23 ± 0.08 | 0.22 ± 0.07 |
PCT, % (mean ± SD) | 31.17 ± 11.13 | 34.69 ± 14.07 | 29.49 ± 9.36 |
P-LCR, % (mean ± SD) | 13.47 ± 3.30 | 14.12 ± 4.50 | 13.16 ± 2.63 |
RA,n(%) | NA | 1 (10) | NA |
ANCA-associated vasculitis,n(%) | 1 (10) |
SLE,n(%) | 2 (20) |
SS,n(%) | 1 (10) |
Dermatomyositis,n(%) | 2 (20) |
Polymyositis,n(%) | 3 (30) |
Blood sampling
Blood samples were collected in serum tubes with a gel separation plug (BD Biosciences, USA). All samples were gently mixed, and the serum tubes were placed at room temperature for coagulation for 30 min. Then, all samples were then centrifuged at 3000g for 20 min at 4 °C and the top volumes of the serum were collected in 1.5-mL centrifuge tubes (Axygen, USA). All samples were frozen within 30 min and preserved at − 80 °C before tests. Further detection was done within 1 month.
Calpain activity measurement
Calpain activity kit (Raybiotech, USA) was utilized to measure calpain activities in serum or plasma. Eighty-five microliters of serum was diluted in 10 μL of 10X calpain reaction buffer and 5 μL of calpain substrate Ac-LLY-AFC with or without 100 μM calpeptin (Abmole, USA). Free AFC was quantified using a fluorometer (excitation λ 400 nm, emission λ 505 nm) after incubating at 37 °C for 1 h in the dark. The difference of calpain activity was determined by comparing the relative fluorescent unit (RFU) of samples with and without calpeptin. The calpain activity was expressed as RFU per microliter serum of each sample.
Measurement of HMGB-1 concentrations in serum
The measurement of the serum HMGB-1 level was performed by enzyme-linked immunosorbent assays (IBL-International, Hamburg, Germany) according to the manufacturer’s instructions. The detection limit of this assay was 0.313 ng/mL. Each sample was tested in duplicate.
Microarray datasets and high-throughput sequencing datasets from NCBI Gene Expression Omnibus (GEO) (
https://www.ncbi.nlm.nih.gov/geo/) were thoroughly searched for available datasets involving SSc. Included datasets should meet the following criteria: (a) datasets with SSc or SSc-ILD lung tissue, skin, or blood samples; (b) datasets with platform information; and (c) datasets with healthy people as control. According to these criteria, six microarray datasets (GSE40839, GSE48149, GSE76808, GSE81292, GSE33463, and GSE58095) were obtained from the GEO database. Details of each microarray study, including sample descriptions and platform information, are shown in Table S
1.
Data processing
For datasets of lung tissue samples (GSE40839, GSE48149, GSE76808, and GSE81292), 50 SSc-ILD patients and 28 HC were included for further analysis. For datasets of PBMC samples (GSE33463), 69 SSc-ILD patients and 41 HC were included. For datasets of skin biopsy samples, 59 SSc patients and 43 HC were included. First, the raw data of each dataset was preprocessed by the R packages affy (under the R environment, version 3.6.1) and annotate methods to make normalized expression profiles with official gene names. Since datasets of lung samples were from different studies and based on different platforms, all lung samples of five datasets (GSE40839, GSE48149, GSE76808, and GSE81292) were integrated by batch normalization using sva package in R software to reduce batch effects and heterogeneity among different samples to significantly improve sample size (50 SSc-ILD vs 28 HC). Next, the differential expression analysis (Log2FC > |1|, p value < 0.05) of calpain-related genes was performed by comparing SSc or SSc-ILD samples to HC samples using the limma package. The boxplot was also utilized to visualize the expression of calpain-related genes.
To explore the function of calpain-related genes in SSc patients, we removed HC lung samples (
n = 20) and clustered SSc-ILD patients (
n = 50) into two clusters based on the expression status of calpain-related genes using “ConsensusClusterPlus” package in R [
27]. Then, the differential expression analysis was performed using the limma package in two clusters. The cutoff value was log
2FC > |1|,
p value < 0.05. Next, Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) analysis of differently expressed genes (DEGs) in two clusters were performed using GOplot package in R. To build the protein-protein interaction (PPI) network, we imported the genes into STRING database (
http://string-db.org) and visualized these genes by Cytoscape.
Statistical analysis
We utilized the Student t test, the Mann-Whitney U test, or the chi-squared test, as appropriate, to test comparisons of each group for statistical significance, and used univariate logistic regression (ULR) and multivariate logistic regression (MLR) to determine risk factors by SPSS 22.0. We also utilized the receiver operating characteristic curve (ROC) to calculate the area under the curve (AUC) by SPSS 22.0 and utilized Spearman’s rank correlation to evaluate the correlation between two continuous parameters by GraphPad Prism 6.0. A p value of < 0.05 was considered statistically significant.
Discussion
The present study is the first to reveal that calpain activity was increased in serum samples from SSc patients in comparison to HC and patients with other CTD. We also demonstrated that calpain activity in serum of patients with SSc was positively correlated with HMGB1 levels, mRSS, MPV, and PCT, but correlated inversely with ESR. Moreover, the elevation of serum calpain activity was accompanied by the presence of increased mRSS, PCT, and P-LCR; negative anti-Scl-70 antibody; and decreased ESR, indicating that serum calpain activity was associated with platelet dysfunction and skin thickness. Furthermore, serum-elevated calpain activity and HMGB1 levels were also independent risk factors for SSc-ILD. Finally, we explored the potential mechanisms and provided new insights into the pathogenesis of SSc or SSc-ILD. Collectively, our research suggested that serum calpain activity and HMGB1 levels were a new potential serological biomarker for the diagnosis of SSc-ILD. In addition, calpain and HMGB1 could be therapeutic targets for accurate therapy of SSc or SSc-ILD in the future.
Calpain has been widely studied to play a critical role in vascular remodeling and collagen synthesis [
28,
29]. Peng and colleagues reported that calpain induces myocardial hypertrophy and fibrosis [
8]. Several studies have reported that inhibition of calpain attenuates bleomycin-induced pulmonary fibrosis in mice [
30‐
33]. Furthermore, a recent study has demonstrated that calpain activation by the renin-angiotensin system induces collagen-I synthesis and pleural fibrosis [
34]. However, studies involving the calpain activation and skin fibrosis or SSc-related pulmonary fibrosis have never been reported. It is worth mentioning that our study is the first clinical research that provides insightful viewpoints in the relationship between calpain activation and SSc or SSc-ILD and has demonstrated that increased serum calpain activity is associated with skin thickness and could predict the development of ILD in SSc patients. Moreover, several studies have demonstrated that microparticles are higher in the blood of patients with SSc compared with HC and that most of them are derived from platelets, indicating that platelet could be a critical factor in the activation of innate immunity [
35,
36]. Manfredi and colleagues have revealed that platelet-derived microparticles express DAMP HMGB1 and lead to initiate fibrosis and endothelial damage when injected into mice, indicating that platelet-associated HMGB1 may be a potential indicator of SSc [
22]. Consistent with previous findings, our results also showed that serum calpain activity was correlated with serum HMGB1 levels and platelet-related parameters (MPV, PCT, and P-LCR), which suggest that activated platelet-derived calpain and HMGB1 may be responsible for the vascular remodeling and endothelial damage in patients with SSc.
HMGB1 has been widely reported as a central role in the pathogenesis of many CTD, such as SLE, polymyositis or dermatomyositis with ILD, and RA [
37‐
41]. Our results clearly demonstrate that serum HMGB1 levels from SSc patients are significantly higher compared with those from HC. A consistent finding has also been reported in a cohort of 70 patients with SSc and 25 HC, which indicates that increased HMGB-1 levels have more frequent involvement of multiple organs and abnormalities of immunological parameters and correlate positively with mRSS [
20]. Several mechanisms have heen reported for HMGB1 to contribute to the development of SSc. A recent study reported that activated platelet-derived microparticles expressed DAMP HMGB1 and induced neutrophil activation in SSc patients, contributing to neutrophil extracellular trap production and autophagy [
22]. Moreover, DAMP HMGB-1 induces proinflammatory effects partly through interaction with toll-like receptor 4 (TLR4), resulting in the activation of nuclear factor-kappa B (NF-κB) pathway and several inflammatory genes [
42‐
45]. Notably, according to experimental findings of a myocardial infarction mouse model with cardiomyocyte-specific deletion of
Capn4,
Capn4 knock-out correlated with restoration of IκB protein and inhibition of NF-κB activation, resulting in the inhibition of proinflammatory cytokine expression and inflammatory cell infiltration in the
Capn4 knock-out heart after myocardial infarction [
46]. Consistent with previous results, our finding suggests that the toll-like receptor signaling pathway could be the potential mechanism that calpain and HMGB1 commonly function through bioinformatic analysis of integrative microarray datasets.
Most patients with SSc have circulating autoantibodies, such as anticentromere, anti-Scl-70, and anti-RNA polymerase III. Clinicians commonly combine the measurement of circulating autoantibodies into the diagnostic evaluation of SSc. The presence of anti-Scl-70 antibodies in patients with SSc was discussed previously as a risk factor associated with progressive ILD [
47]. However, none of these autoantibodies is absolute. Although the specificity of anti-Scl-70 antibodies is high, the sensitivity in detecting SSc is merely 34%, and this increases slightly to 40% in detecting dSSc [
48]. Inversely, in our study, we found that the elevation of serum calpain activity has a negative correlation with anti-Scl-70 antibodies, which suggested that due to the limitation of sensitivity and sample size, these autoantibodies should be interpreted by clinicians together with other clinical indicators in the evaluation of SSc or SSc-ILD. Also, the activity of serum calpain should be used as a guide for clinicians, but not as a definitive indicator for SSc or SSc-ILD. Furthermore, future research based on a larger sample size is needed to confirm these findings.
A limitation of this study is that our limited samples from largely middle-aged female patients cannot represent the general population. Furthermore, another key limitation is the lack of ethnic or geographic diversity in our samples, which may potentially affect the reliability of the results. Therefore, our findings may not generalize total SSc or SSc-ILD patients. Further studies with larger sample sizes, more diverse ethnicities, and less heterogeneity are needed for the generalization of our findings. Despite the limitation, our study firstly identified the serum calpain activity was elevated in patients with SSc and highlighted the potential mechanisms for further research.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.