Association between endothelin-1 and systemic lupus erythematosus: insights from a case–control study

Systemic lupus erythematosus (SLE) is a chronic rheumatic disorder. Endothelin-1, a vasoconstrictor, belongs to the endothelin family and is associated with vascular-related damages. To date, association between ET-1 and pathogenesis of SLE remains unclear. This case–control study was carried out by 314 SLE, 252 non-SLE diseases patients and 500 healthy controls. Serum ET-1, CCN3, IL-28B levels were detected by ELISA, and ET-1 gene polymorphisms (rs5369, rs5370, rs1476046, rs2070699, rs2071942, rs2071943, rs3087459, rs4145451, rs6458155, rs9369217) were genotyped with Kompetitive Allele-Specific PCR. SLE patients had high levels of ET-1, which were correlated with some clinical, laboratory features. Serum CCN3, IL-28B levels were higher in SLE patients, and ET-1 levels were positively correlated with the two cytokines. Rs5370, rs1476046, rs2070699, rs2071942, rs2071943, rs3087459, rs6458155 and rs2070699 were associated with SLE risk. Rs2070699 (T, TT) was related to SLE patients with alopecia. Rs5370 (T, TT, TG), rs1476046 (G,GA), rs2071942 (G,GA) and rs2071943 (G,GA) were associated with SLE patients with pericarditis, pyuria and fever manifestation, respectively. Rs3087459 (CC) and rs9369217 (TC) were related to SLE patients with positive anti-SSB antibody. Rs5369 (AA) was associated with IgG and CRP levels in SLE patients. In conclusion, elevated serum ET-1 in SLE patients may be a potential disease marker, and its gene polymorphisms were related to SLE susceptibility.

interferons, including IL-28A, IL-28B, IL-29, are functionally similar to type I interferon.IL-28B involves in antiviral and antitumor responses, and contributes to development of immune-related disorders 14 .For example, levels of IL-28B were increased in SLE patients and were related to disease activity 15 .However, the relationship between ET-1 and CCN3, as well as the relationship between ET-1 and IL-28B, in SLE remains obscure.
Therefore, role of ET-1 in SLE is not clear, and whether ET-1 regulates CCN3, IL-28B, and then contributes to SLE pathogenesis is not discussed.In this study, we conducted a case-control study with SLE patients and healthy controls to reveal serum levels of ET-1 in SLE, gene polymorphisms of ET-1 in SLE.Second, we examined serum levels of CCN3, IL-28B in SLE patients, and discussed association of ET-1 and CCN3, IL-28B.If the two inflammatory components (CCN3, IL-28B) were abnormally expressed in SLE patients and there is significant association of ET-1 and CCN3, ET-1 and IL-28B, in the future, we will conduct functional study to clarify how ET-1 contributes to SLE development, and whether ET-1 contributes to SLE development by regulating CCB3, IL-28B.

Elevated serum levels of ET-1 in SLE from validation cohort
To evaluate potential of serum ET-1 as a biomarker for SLE, a validation cohort including 102 SLE patients and 252 non-SLE diseases patients was discussed further (90 RA, 95 osteoarthritis (OA), 55 sjogren syndrome (SS), 38 anlylosing spondylitis (AS), 17 SSc).Information about age and sex of these diseases patients is shown in Table 1.Results indicated that serum ET-1 levels in SLE patients were significantly higher than that in other rheumatic diseases (all P < 0.001, Fig. 2A).AUC was 0.803 by ROC curve analysis when comparing serum levels of ET-1 between SLE with RA patients (Fig. 2B).Similarly, serum ET-1 in SLE patients compared to that in OA, SS, AS, SSc, showed AUC 0.979, 0.856, 0.892 and 0.903, respectively (Fig. 2C-F).We compared serum ET-1 levels between SLE patients and non-SLE patients, showing that there was a significant difference (P < 0.001, Fig. 2G) and AUC was 0.900 (95% CI = 0.869-0.932)(Fig. 2H).

Association of ET-1 gene single nucleotide polymorphisms (SNPs) with SLE
To discuss whether polymorphisms of ET-1 gene correlate with SLE risk, a total of 314 lupus patients and 500 age, sex-matched healthy volunteers were selected.Hardy-Weinberg equilibrium (HWE) test showed no deviation in SLE patients and controls for all 10 polymorphisms (P > 0.05, Table 2).Information about alleles and genotypes of the SNPs was shown in Table 3. Frequency of rs5370 genotype TG was higher in SLE patients compared to healthy controls (TG vs. GG: OR = 1.443, 95% CI 1.072-1.943,P = 0.016).In the dominant model (TT + TG vs GG), frequency of genotypes TT + TG of rs5370 was increased in SLE patients (OR = 1.334, 95% CI 1.005-1.770,P = 0.046).For rs1476046, frequency of genotype GG in SLE patients was markedly lower than that in controls In the dominant model, higher frequencies of CA + CC genotypes of rs3087459 were found in patients with SLE (CA + CC vs AA: OR = 1.352, 95% CI 1.013-1.805,P = 0.040).In addition, SLE patients with rs6458155 had an enhanced frequency of TC genotype (TC vs CC: OR = 1.392, 95% CI 1.021-1.899,P = 0.0037).For rs5369, rs4145451, rs9369217 polymorphisms, we discussed the allele and genotypes distribution between SLE cases and healthy controls, showing no significant differences.

Relationship of ET-1 gene polymorphisms with clinical, laboratory features in SLE patients
Association of qualitative and quantitative clinical, laboratory features in SLE patients with ET-1 gene polymorphisms was shown in Supplementary table 7, 8, 9 and 10, respectively.Compared to patients without these clinical features, frequency of rs5070 genotypes TT + TG was elevated in patients with pericarditis and positive ANA (P = 0.034, P = 0.045), and was decreased in patients with fever (P = 0.005).Moreover, there was a higher frequency of allele T of rs5370 in patients with pericarditis, and a lower frequency of allele T of rs5370 in patients with pyuria compared with those in patients without the features (P = 0.013, P = 0.047).For rs1476046, decreased www.nature.com/scientificreports/G allele frequency in patients with pericarditis and increased G allele frequency in patients with pyuria was observed when compared to patients without these features.A lower frequency of genotype GA of rs1476046 in SLE patients with fever and pyuria was noted (P = 0.013, P = 0.028).Interestingly, there were significant differences for genotypes and allele frequencies of rs2070699 polymorphism between SLE patients with alopecia and SLE patients without alopecia (P = 0.013, P = 0.035).Distribution of GG, GA, AA genotypes of rs2071942 and rs2071943 was different between SLE patients with and without fever (P = 0.025, P = 0.010).Patients with pericarditis had a declined frequency of G allele of rs2071942 and rs2071943 compared to patients without pericarditis (P = 0.019, P = 0.025).Frequency of G allele of rs2071942 and rs2071943 in patients with pyuria was higher (P = 0.035, P = 0.028).With respect to rs3087459, decreased frequencies of C allele and CC genotype in patients with positive ANA and increased frequency of CA genotype in patients with positive anti-SSB were observed when compared to patients without these features (P = 0.045, P = 0.029, P = 0.027).In addition, distribution of TT, TC and CC genotypes of rs9369217 was different in SLE patients with and without positive anti-SSB    When discussing quantitative indicators of clinical, laboratory features, levels of IgG between SLE patients with GG + GA genotype and AA genotype for rs5369 was different (P = 0.025).SLE patients carrying rs5369 AA genotype had higher levels of CRP as compared to the patients carrying GG + GA genotype (P = 0.012) (Supplementary table 10).

ET-1 gene haplotypes and SLE risk
Considering genetic linkage disequilibrium, we analyzed correlation between ET-1 gene haplotypes and SLE risk.There were two blocks constructed, one consisting of rs6458155, rs4145451 and the other one including rs2071942, rs2071943, rs5370 (Fig. 4).Results revealed that frequency of haplotype CA was lower in SLE patients compared to healthy controls (P = 0.012).Reduced frequency of haplotype AGG was found in patients with SLE (P = 0.041).No significant differences were observed in other haplotypes (Table 4).

Discussion
SLE is a rheumatic disease with heterogeneous clinical symptoms that may result from excessive endothelial cells activation 16 .When activated by inflammatory stimuli, endothelial cells exhibit increased levels of surface adhesion molecules, which promote the migration and accumulation of leukocytes to the endothelial cells.Then, vascular obstruction and tissue hypoxia were induced, leading to apoptosis and tissue fibrosis 17 .Therefore, measuring biomarkers related to endothelial cells activation is useful for diagnosis of SLE.Endothelium-derived ET-1 plays a pathogenic role in connective tissue diseases, pulmonary hypertension, and cancer by affecting angiogenesis, inflammation, and fibrosis 18 .In OA mice, ET-1 induces increased levels of IL-18 through ET-1/ ETAR axis and PI3K-dependent manner, promoting osteoblasts proliferation and exacerbating the disease 19 .As compared to control group, there were increased ET-1 levels in RA patients, especially in the patients with kidney and cardiovascular damages 20 .ET-1 levels were also elevated in other autoimmune diseases such as SSc, psoriasis and type 1 diabetes 21,22 .According to our study, serum ET-1 levels were significantly higher in SLE patients in training cohort when compared with healthy controls, which is consistent with previous findings 6,23,24 .Moreover, we found that serum ET-1 levels were associated with SLE patients complicated with cylindruria, alopecia and positive anti-Sm antibody.Similarly, Yoshio et al. showed that serum ET-1 in SLE patients correlated with IgM   23 .SLE patients with visceral manifestation showed higher ET-1 levels as compared to that in patients without the feature 23 .In our study (the training cohort), we also found a positive correlation between serum ET-1 levels and SLEDAI score, and SLE patients with active disease activity had higher ET-1 levels.Urinary ET-1 may be a useful measurement of renal inflammatory activity and may serve as a marker in lupus nephritis disease activity 25 .In the present study, we also explored potential of serum ET-1 as a biomarker for SLE.In training cohort, AUC of ET-1 was 0.912 (95% CI 0.866-0.959),indicating that serum ET-1 could distinguish SLE patients from healthy subjects.In the validation cohort, ET-1 levels were higher in SLE patients than in other rheumatic diseases, including RA, OA, SS, AS, SSc.Compared with non-SLE diseases patients, AUC of serum ET-1 in SLE was higher than 0.800.Therefore, serum ET-1 is a promising marker for SLE.How ET-1 contributes to SLE or what mechanisms of ET-1 may involve in development of the disease?ET-1 could induce levels of connective tissue growth factor (CTGF/CCN2) by activating the MEK/ERK kinase pathway after binding to ETA and ETB receptor 26 .Indeed, these signalling pathways are pro-inflammatory pathways in SLE pathogenesis.After treatment with ET-1 receptor antagonist (bosentan), anti-dsDNA levels were reduced, glomerulosclerosis and renal T cells infiltration were improved in lupus mice 27 .These findings suggest that ET-1 may promote SLE progression.IL-17 and Th17 cells are important cytokine and cells in SLE pathogenesis.It is known that ET-1/ETAR signaling pathway regulates production of IL-17 in Th17 cells by an autocrine or paracrine manner 28 .To date, no study has discussed effects of ET-1 on Th17 cells in SLE.Thus, in future studies, we will consider to explore whether highly expressed ET-1 in SLE will contribute to lupus development by regulating Th17 cells.In patients with lupus nephritis, IgM anti-endothelial cell antibody and immune complexes can bind to endothelial cells in glomerular capillary rings and stimulate production of ET-1, leading to vascular injury 24 .CCN3 involves in some autoimmune diseases pathogenesis 12 .Serum CCN3 levels were elevated in RA patients and positively correlated with levels of IL-6 29 .CCN3 levels were increased in SSc, multiple sclerosis patients 30,31 .For OA, CCN3 could inhibit PI3K/AKT/mTOR pathway by reducing HMGB1 levels and decrease extracellular matrix catabolism 32 .In our study, we found that serum CCN3 levels were significantly higher in SLE patients compared to normal subjects and were associated with thrombocytopenia.In addition, we observed a positive correlation between ET-1 levels and CCN3 levels in SLE patients.CCN3 inhibits levels of vascular adhesion molecules and reduces monocytes adhesion.CCN3 negatively regulates activation of NF-κB pathway, affecting inflammatory cytokines production of endothelial cells and cardiovascular homeostasis 33 .Type III interferons (IFNs) and type I IFNs may promote THP-1 cells differentiation, thereby contributing to the activation of follicular B cells and participating in development of autoimmune diseases 34 .IL-28B regulates innate and adaptive immune responses.SSc patients with pulmonary fibrosis had higher serum levels of IL-28B and IL-28B gene polymorphism (rs12979860) is associated with risk of pulmonary fibrosis in Caucasian SSc patients 35 .With respect to SLE, IL-28B gene SNPs (rs8099917, rs12979860) are risk factors for lupus nephritis in Taiwanese.Moreover, serum levels of IL-28B were elevated in SLE patients compared to healthy controls, which were related to complement levels and SLE disease activity 36 .IL-28B levels are associated with SLE disease activity 15 .According to our study, IL-28B levels in SLE patients were positively associated with disease activity and ESR levels.SLE patients with hematuria, proteinuria, cylindruria showed higher levels of IL-28B, which are typical clinical manifestations of SLE.We analyzed correlation between ET-1 and IL-28B levels and observed that serum levels of ET-1 were positively correlated with IL-28B levels.Therefore, in the future, we will conduct functional study (for example, knock out ET-1 gene in animal models) to discuss whether ET-1 regulates CCN3 or IL-28B, and then contributes to SLE development.
It is now accepted that SNPs are new genetic markers that may be used for discovery of high risk patients.ET-1 gene polymorphisms are widely discussed in vascular-related diseases and cancer, such as hypertension, coronary atherosclerosis, and papillary thyroid cancer 37,38 .However, ET-1 gene polymorphisms are less studied in autoimmune diseases.In this study, we explored relationship between ET-1 gene polymorphisms and SLE risk by a case-control study from Chinese Han population.We found that genotypes of rs5370 (TG,TT + TG), rs1476046 (GG), rs2070699 (TT,TG), rs2071942 (GA), rs2071943 (GG), rs3087459 (CA + AA), rs6458155 (TC) and allele of rs2070699 (T) were associated with SLE susceptibility.For rs5370, TG, TT + TG genotype frequencies were higher in SLE patients, suggesting that rs5370 polymorphism may increase the risk of SLE.Frequency of rs2071942 GG genotype was increased in SLE patients compared to healthy subjects, which was positively associated with SLE risk.Mantaka et al. discussed association of ET-1 gene rs2071942 and rs5370 polymorphisms with primary biliary cirrhosis (PBC), showing that distribution of genotypes and alleles of both loci were not significantly different between controls and PBC patients, whereas rs2071942 allele A and rs5370 allele T were associated with stage of disease progression 39 .These inconsistency may be due to differences in sample size, ethnicity, disease type.Similarly, frequencies of genotypes and alleles of rs5370 and rs18000541 were not significantly different between RA patients and healthy controls, but TT genotype of rs5370, rs18000541 was related to RA patients complicated with hypertension, indicating that rs5370 and rs18000541 loci were associated with cardiovascular risk in RA 40 .In our study, rs5370, rs1476046, rs2070699, rs2071942, rs2071943 rs5369, rs3087459, and rs9369217 polymorphisms correlated with some clinical, laboratory features in SLE patients.For example, SLE patients carrying rs2070699 T allele and TT genotype were more likely to develop symptoms of alopecia.Regarding rs2071942 and rs2071943, allele G was associated with pericarditis and pyuria, and genotype GA was related to fever.Moreover, significant correlation between rs5370 T allele, rs1476046 G allele and pericarditis, pyuria in SLE cases was observed.Rs3087459 (CC) and rs9369217 (TC) were associated with SLE patients with positive anti-SSB antibody.Rs5369 AA genotype correlated with IgG and CRP levels, suggesting that mutations at the rs5369 locus may affect levels of these disease markers in SLE patients.Indeed, CRP is an acute phase protein produced by hepatocytes in response to inflammation.CPR has important roles in active lupus nephritis.In SLE patients or mice models, IgG immune complexes are deposited in the spleen, causing damage to immune barrier

Figure 1 .
Figure 1.Comparison of ET-1 levels between SLE patients and healthy controls in the training cohort.(A) Serum ET-1 levels in 53 SLE patients and 80 healthy individuals were examined by ELISA.Each symbol stands for an independent sample.(B-D) ET-1 levels in SLE patients distributed according to alopecia, proteinuria and anti-Sm.(E) Difference of serum levels of ET-1 in SLE patients with less active disease and active disease.(F-G) Relationship between ET-1 levels and SLEDAI, ESR levels.(H) Receiver operating characteristic curve analysis of serum ET-1 as a biomarker for SLE.

Figure 2 .
Figure 2. ET-1 levels in SLE patients from validation cohort.(A) Comparison of serum ET-1 levels in SLE patients (N = 102) and diseases controls (N = 252, including 90 RA, 95 OA, 55 SS, 38 AS, 17 SSc) by ELISA.(B-F) Receiver operating characteristic curve analysis was used to assess potential of serum ET-1 in differentiating SLE from RA, OA, SS, AS, and SSc.(G) Analysis of difference in serum ET-1 between SLE and non-SLE patients.(H) Receiver operating characteristic curve analysis of serum ET-1 between SLE and non-SLE patients.

Figure 3 .
Figure 3. CCN3 and IL-28B levels in SLE patients and healthy controls.(A) Serum CCN3 levels were tested by ELISA in 53 SLE patients and 80 healthy controls.(B) Serum ET-1 in SLE patients distributed in accordance with thrombocytopenia.(C) Correlation between ET-1 and CCN3 levels in SLE patients.(D) Serum IL-28B levels were detected by ELISA in 53 SLE patients and 80 healthy controls.(E-G) Distribution of serum IL-28B in SLE patients with hematuria, proteinuria and cylindruria.(H) Difference of serum IL-28B in SLE patients with active disease and less active disease.(I-K) Relationship between IL-28B levels and SLEDAI score, ESR levels.(H) Correlation between ET-1 and IL-28B levels in SLE patients.

Figure 4 .
Figure 4. Linkage disequilibrium (LD) analysis for ten SNPs in ET-1 gene.The color and numerical value (D') of each box represent the Te intensity of LD.Red and pink indicate significant linkage, light blue and white indicate no linkage.The value of D' varies from 0 to 1, by which value of 1 represents the maximum link.Block 1 consists of rs6458155 and rs4145451.Block 2 consists of rs2071942, rs2071943 and rs5370. https://doi.org/10.1038/s41598-023-43350-0

Table 2 .
The Hardy-Weinberg's expectation test in SLE patients and healthy controls for SNPs.SNP, single nucleotide polymorphisms.SLE, systemic lupus erythematosustis; HC, healthy controls.

Table 3 .
Frequencies of alleles and genotypes for ET-1 gene polymorphisms in SLE patients and healthy controls.SLE, systemic lupus erythematosus; OR, odd ratio; 95% CI, 95% confidence interval.

Table 4 .
Haplotype analysis of ET-1 gene polymorphisms between SLE and healthy controls.antibody levels.A study from Poland SLE patients reported that serum ET-1 levels were higher in active SLE patients than in inactive SLE patients and healthy controls *Block 1 consists of rs6458155 and rs4145451.Block 2 consists of rs2071942, rs2071943 and rs5370.SLE, systemic lupus erythematosus.SLE patient versus controls by 2 × 2 chi-square test.