Circulating miRNA-181b-5p, miRNA-223-3p, miRNA-210-3p, let 7i-5p, miRNA-21-5p and miRNA-29a-3p in patients with localized scleroderma as potential biomarkers

Localized scleroderma (LoSc) is a rare disease manifested by an inflammation and sclerosis of the skin. The latest studies focused on glycoprotein Krebs von den Lungen-6, surfactant protein-D, chemokine ligand 18 and dipeptidylpeptidase 4 as potential biomarkers of skin fibrosis in systemic scleroderma. Our study aimed to identify 6 miRNAs with elevated or decreased levels in 38 LoSc patients (31 females, 7 males) compared to healthy volunteers (HVs) and to correlate the selected miRNAs’ serum levels with the severity and the clinical symptoms of LoSc and some laboratory parameters with the selected miRNAs’ serum levels. The serum levels of miRNAs, i.e. miRNA-181b-5p, miRNA-223-3p, miRNA-21-5p, let 7i-5p, miRNA-29a-3p and miRNA-210-3p were significantly increased in the LoSc patients compared to the HVs. The level of let-7i increase in the female LoSc patients correlated negatively with BSA (r =  − 0.355, p = 0.049) and mLoSSI (r =  − 0.432, p = 0.015). Moreover, the female patients with inactive LoSc had significantly higher level of let-7i (2.68-fold on average) in comparison to those with active disease (p = 0.045). The exact role of those molecules has not been revealed in LoSc and a long-term longitudinal research is pivotal to confirm their prognostic value.


Scientific Reports
| (2020) 10:20218 | https://doi.org/10.1038/s41598-020-76995-2 www.nature.com/scientificreports/ did not correlate with age, gender, duration of the disease, the clinical subtype of LoSc, BSA, mLoSSI, LoSDI, disease activity, the presence of sclerotic areas, erythematous patches without sclerosis, patches with a lilac ring, hyperpigmented patches, pruritus, extracutaneous manifestations, distribution of lesions, positive ANA or RF as well as abnormal ESR and CRP levels (p > 0.05). However, in the female LoSc patients we have found some significant correlations between some miRNAs' serum levels and the severity, the clinical symptoms of LoSc and some laboratory parameters of the LoSc characteristics (Figs. 3, 4). The level of let-7i increase in the female LoSc patients correlated negatively with BSA Table 1. Summary of miRNA PCR array. MiRNAs selected for further studies by RT-qPCR are presented in bold. Position position on the plate for specific miRNA, Mature ID mature miRNA ID, Fold Change 2 −ΔΔCt , cel-miR-39-3p and cel-miR-39-3p miScript primer used for alternative data normalization using exogenously spiked Syn-cel-miR-39 miScript miRNA mimic, SNORD61, SNORD68, SNORD72, SNORD95, SNORD96A, RNU6-6P 6 miScript PCR controls used data normalization using the ΔΔCT method of relative quantification, miRTC reverse transcription control used for assessment of reverse transcription performance, PPC positive PCR control used for assessment of PCR performance.  Fig. 3b), i.e. the higher the severity of the disease measured by BSA and mLoSSI, the lower the increase level of let-7i. Moreover, the female patients with inactive LoSc had significantly higher level of let-7i (2.68-fold on average) in comparison to those with active disease (p = 0.045). In the female LoSc patients who had patches without lilac ring, significantly higher level of let-7i (2.11-fold on average) than in the female patients with lilac ring (a feature of disease activity) was observed (p = 0.022) (Fig. 4). These results suggest that the higher increase level of let-7i co-exist with less severe LoSc.    11 . Furthermore, our study revealed that higher level of miRNA-21 was observed in the female LoSc patients without erythematous patches without sclerosis (2.48-fold on average) than in those with erythematous patches without sclerosis (a marker of the inflammatory stage of the disease) (p = 0.042), indicating that miRNA-21 serum level is inversely correlated with inflammation ( Fig. 4). We also demonstrated that the level of miRNA-21 was higher in the female LoSc patients with positive RF (3.88-fold) than in those with negative RF (p = 0.045) (Fig. 4). Additionally, our study showed a higher level of miRNA-29a in the female LoSc patients without the presence of concomitant autoimmune diseases (1.44-fold on average) than in those with autoimmune diseases (p = 0.029), as well as a higher level of miRNA-210 in female LoSc patients with the normal CRP level (3.31-fold on average) than in those with abnormal CRP level (a marker of the inflammatory stage of the disease) (p = 0.027) (Fig. 4). Anti-inflammatory properties of miRNA-210 have been already reported by Qi et al. 38 . Computational prediction of miR-181b-5p, miR-223-3p, miR-210-3p, let 7i-5p, miR-21-5p, and miR-29a-3p, using the aforementioned algorithms, generates a long list of potential targets and pathways involved in the process of fibrosis (Table 2, Fig. 5) [39][40][41][42] . miRNA181-5p is predicted to target S1PR1 (sphigosine-1-phosphate receptor 1), KLF6 (Krueppel-like factor 6), BCL2L11 (Bcl-2-like protein 11), OSBPL3 (oxysterol binding protein-like 3), PTEN (phosphatase and tensin homolog), LOX (lysyl oxidase family of enzymes), IL-2 (interleukin 2), as well as NF-kappa B signaling pathway and PI3K-Akt signaling pathway. Sphingosine 1-phosphate (S1P) is a bioactive sphingolipid involved in the profibrotic inflammatory process through i.a. hematopoietic stem cells (HSCs) proliferation and differentiation to myofibroblasts 43 . S1P is synthesized by sphingosine kinases (SphKs) and acts through S1P specific cell   www.nature.com/scientificreports/ surface receptors (S1PR1-5) 43 . Overexpression of S1PR1 has already been reported in cardiac hypertrophy and fibrosis through angiotensin II and interleukin-6 44 . Moreover, González-Fernández et al. delineated that targeting SphKs/S1P/S1P receptors signalling pathway has a very promising therapeutic potential in the treatment of hepatic fibrosis 43 . KLF6 is the transcription factor that plays roles in differentiation, development, apoptosis, and angiogenesis 45 . It activates a number of genes crucial for the development of liver fibrosis, including collagen 1, TGF-β1 and TGF-β1 receptors types I and II 45 . Starkel et al. reported that oxidative stress, KLF6 and TGF-β1 upregulation discriminate non-alcoholic steatohepatitis advancing to fibrosis from uncomplicated steatosis in rats 46 . www.nature.com/scientificreports/ Furthermore, accumulating evidence demonstrates that KLF6 is involved in the occurrence of oral submucous fibrosis 47 . BCL2L11, also known as BIM, is a pro-apoptotic member of the B cell CLL/lymphoma 2 protein family and is an essential modulator of apoptosis 48 . In the course of fibrotic diseases, the excessive collagen deposition leads to increased ECM stiffness. Myofibroblasts activated by ECM stiffness are prepared for apoptosis by death signals such as the BCL2L11. Upregulation of BIM results in expression of anti-apoptotic proteins to provide myofibroblast survival 49 . OSBPL3 was delineated to increase in advanced stages of non-alcoholic fatty liver disease/non-alcoholic steatohepatitis leading to fibrosis 50 . PTEN is a lipid/protein phosphatase that negatively regulates proliferation by inhibiting the integrin-PI3K/Akt pathway 51 . Loss of PTEN expression has already been demonstrated in patients with dsSSc and IPF 51,52 . LOX is a family of ECM cross-linking enzymes that have been disclosed to play a crucial role in fibrogenesis 53 . Increased LOX gene expression has been described in IPF tissue compared with healthy controls in two independent studies 53,54 . miRNA-223-3p is anticipated to target ETC2 (epithelial cell transformation-2), SMAD1, FGFR2 (fibroblast growth factor 2 receptor), FGF2 (fibroblast growth factor 2), FOXO3 (Forkhead box O3), SPP1 (osteopontin), MMP16 (matrix metalloproteinase 16), PTEN, BCL2L11, and JAK-STAT signalling pathway. ETC2, a factor with established roles in cytokinesis and extracellular-signal-regulated kinase (ERK) signalling, was reported to be increased in IPF 55 . SMAD1 is a protein mediating TGF-b signalling, activated by α-SMA, that was delineated to be elevated in the skin of patients with SSc i liver fibrosis 56 . FGF2 and FGFR2 are profibrogenic factors that may support the proliferation and activation of kidney fibroblasts, which contribute to the development of renal fibrosis 57 . FOXO3 is an important integrator of profibrotic signaling in lung fibrosis and was found to be downregulated in IPF myofibroblasts 58 . SPP1, a protein with proinflammatory and profibrotic properties, has already been demonstrated to take part in cardiac and lung fibrosis 59,60 . MMP16 protein activates MMP2 protein which in turn degrades type III collagen and was demonstrated to be decreased in the myocardial fibrosis 61 .
Databases predict that miRNA-210-3p targets FGFRL1 (fibroblast growth factor receptor-like 1), and WLS. FGFRL1, also known as FGFR5, is the atypical receptor that lacks an intracellular kinase domain and binds to FGF ligands (binds FGF3, FGF4, FGF8, FGF10, and FGF22). The FGF8/FGFRL1 complex regulates development of nephrons by controlling the mesenchymal-to-epithelial transition 62 . Modifications in the Wnt signaling pathway have been implicated in liver fibrosis 63 . Shutdown of Wnt secretion (WLS) from stellate cells is pivotal for development of liver fibrosis following hepatobiliary injury 63 .
Let-7i is foreseed to target collagens, NGF (nerve growth factor), HIF1AN (hypoxia-inducible factor 1-alpha inhibitor), TGFBR1 (transforming growth factor beta receptor 1), ADAMTS8 (ADAM metallopeptidase with thrombospondin type 1 motif 8), IL6R (interleukin 6 receptor), as well as PI3K signalling pathway, p53 signalling pathway, MAPK signalling pathway, and ECM-receptor interaction. NGF is a polypeptide which, in addition to its effect on nerve cells, is considered to play a function in inflammatory responses and tissue repair 64 . Fibroblasts express and release NGF protein, which may consequently mediate proliferation and hypertrophy. Overexpression of NGF has already been disclosed in SSc patients 65 . Chen et al. found miRNA-184 as a critical mediator to promote the renal fibrosis by targeting HIF1AN 66 . ADAMTS8 is a member of the ADAMTS family which is involved in a various of functions including migration, adhesion, proliferation, and growth factor signalling 67 . Increased ADAMTS8 protein levels were revealed in linear localized scleroderma lesions 67 .
Databases predict that miRNA-21-5p target collagens, ESM1 (endothelial cell-specific molecule 1), AGO2 (argonaute 2), TIMP3 (tissue inhibitor of metalloproteinases), BMPR2 (bone morphogenic protein receptor 2), and JAK-STAT signaling pathway. ESM1, also known as endocan, is reported to play a role in the pathogenesis of inflammation, endothelial dysfunction and vascular disorders 68 . Accumulating evidence suggest that elevated ESM1 expression promotes kidney fibrosis by inducing the endothelial to mesenchymal transition in a mouse model 68 . Moreover, Oak et al. revealed that AGO2, an essential component of the miRNA processing RISC (RNAinduced silencing complex) complex, was expressed at lower levels in rapidly progressive IPF biopsies compared with both normal and slowly progressive IPF biopsies 69 . According to Kassari et al. loss of tissue TIMP3, a key inhibitor of ECM remodeling, enhances renal fibrosis 70 . BMPR2 is a member of the TGF-β family that has a major role in suppressing TGF-β signaling 71 . Ning-Yuan et al. disclosed reduced BMPR2 expression in patients with IPF compared to healthy controls 71 .
miRNA-29a in anticipated to target collagens, CTNNBIP1 (catenin-beta interacting protein 1), LAMC1 (laminin gamma-1), AKT3 (Akt kinase isoform 3), CAV2 (caveolin 2), CTNND1 (catenin delta 1), WISP1 (WNT1-inducible-signaling pathway protein 1), PI3K signaling pathway, and ECM-receptor interaction. CTNN-BIP1 plays an important role in the integration of cell adhesion, motility, cell death, and suppresses Wnt/β-catenin signaling 72 . LAMC1 is a core structural protein present in the basement membrane of several organs and released by MMP-9 73 . Nielsen et al. hypothesized that dysregulation of LAMC1 remodeling could be associated with a higher risk of renal fibrosis 73 . AKT3 is a serine/threonine-protein kinases which regulate many processes including metabolism, proliferation, cell survival, growth and angiogenesis 74 . Wang et al. reported that knockdown of AKT3 inhibited α-SMA and collagen I, and enchanced apoptosis preventing liver fibrosis 74 . Latest evidence suggest that CAV2 overexpression strongly reduce the expression of fibrotic markers, such as connective tissue growth factor, α-SMA, fibronectin, SMAD2/3 75 . Yang et al. postulated that suppressing CAV2 expression lead to hepatic fibrosis by promoting TGF-β pathway 75 . Furthermore, Deng et al. reported that CTNND1 overexpression increased the protein expression of collagen I, and α-SMA. WISP1 is a matricellular protein encoded by a WNT target gene 76 . Increased expression of WISP1 in alveolar cells was revealed in patients with IPF 76 .
The comprehensive applicability of miRNAs may help early diagnosis of LoSc, preventing irreversible sequelae like e.g. contractures, limb asymmetry or scoliosis. Furthermore, miRNAs can have therapeutic significance and may possibly modulate skin fibrosis by enhancing the expression of anti-fibrotic miRNAs or decreasing the expression of profibrotic miRNAs.
Limitations of the study included a small cohort (N = 38) and insufficient diversity of LoSc clinical subtypes (lack of deep subtype).

Methods
The study groups. The  Medical history, including the duration of LoSc, possible triggering factors, and family history was taken. The presence of any extracutaneous manifestations (arthritis, joint pain, contractures, headache, symptoms of eye and nervous system involvement) as well as concurrent autoimmune diseases (Hashimoto's thyroiditis, vitiligo, lichen planus) were noted.
The exclusion criteria included coexistence of other autoimmune disorders manifested by fibrosis of different organs, i.e. systemic sclerosis, keloids, hypertrophic scars, cardiomyopathy, pulmonary fibrosis, liver fibrosis, and renal fibrosis.
The study protocol was approved by the Bioethics Committee of the Medical University of Lublin and the study was performed in accordance with the relevant guidelines and regulations. All subjects provided written informed consent prior to study enrolment. The presence of clinical features, i.e. sclerotic areas, erythematous patches without sclerosis, patches with lilac ring, hyperpigmented patches, atrophic patches, and pruritus, was assessed.
The active disease was defined as the appearance of a new erythematous lesion and/or sclerotic plaque or enlargement of the existing inflammatory lesions (erythema and/or lilac ring) during the preceding month. The non-active disease was defined as the presence of hyperpigmentations or discolourations and/or atrophic patches of the skin over at least 6 months. The site of skin lesions, as well as their distribution, were assessed.
Two indexes, i.e. Localized Scleroderma Skin Severity Index (LoSSI) for the assessment of disease activity/ severity and Localized Scleroderma Skin Damage Index (LoSDI) for the assessment of tissue damage, were calculated. An extent of the skin lesions was assessed with the use of BSA. The clinical characteristics of the LoSc patients are presented in Table 3.
Laboratory parameters in the studied localized scleroderma patients. In all the studied LoSc patients, routine laboratory tests, i.e., C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), antinuclear antibodies (ANA) and rheumatoid factor (RF), were performed. miRNA isolation. Whole blood (5-10 mL) was collected in BD Monovette plastic tubes (SARSTEDT, Germany). The samples were stored on ice and processed within 1 h of the draw. Serum was isolated by centrifugation at 3000×g for 10 min at 4 °C and stored at − 80 °C, in the absence of freeze-thaw cycles, until analysis. Before RNA extraction, we checked the serum samples to test for hemolysis by measuring the absorbance of free hemoglobin at 414 nm, and samples with OD 414 greater than 0.2 were excluded 77 , due to the potential of cellular miRNAs to confound the results. RNA was extracted from 200 µL serum using the Syngen miRNA Mini Kit (Syngen, Poland; SY391210) following the manufacturer's protocol. With this kit it is possible to isolate total RNA and miRNA, thus, we were able to evaluate the concentration of both fractions. Briefly, after mRL buffer (200 µL) was added and mixed, the samples were incubated at room temperature for 10 min, the addition of 450 µL phenol and 200 µL chloroform was useful to achieve aqueous and organic phase separation. In the next step, the aqueous phase was applied to an RNeasy spin column. The purified RNA was eluted in 50 µL mRE buffer and stored at − 80 °C.
Analysis of miRNA quality and integrity. miRNAs' concentration, A260/230, and A260/280 ratios were evaluated by NanoDrop UV/Vis spectrophotometer (2000, ThermoFisher SCIENTIFIC, Waltham, MA, USA). Later, two different chips: Agilent RNA 6000 Nano Kit for total RNA and Agilent Small RNA kit for low molecular weight RNA were used to check the quality and the integrity of total and small RNA. It was measured by capillary electrophoresis with 2100 Agilent Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). Electropherograms were visualized with the Agilent 2100 Expert software.
In the next step, the serum levels of a panel of 84 miRNAs were assessed according to the manufacturer's protocol, along with three control sets present in this panel; the first contained six miRNAs (snoRNA/snRNA) whose average of readings enables normalization of the array data using the relative quantification method, the www.nature.com/scientificreports/ second was used for assessing the performance of reverse transcription reaction, and the third was used for assuring PCR performance. qPCR analysis was performed in 96-well plates using the CFX96 real-time PCR detection system (BIO-RAD, Hercules, CA, USA) in conditions as follows: initial activation step at 95 °C for 15 min, 50 cycles of denaturation at 94 °C for 15 s followed by an annealing step at 55 °C for 30 s, then extension step at 70 °C for 30 s. The relative miRNAs' serum levels were calculated using the 2 −ΔΔCt method. The data were presented as the fold change in gene expression normalized to an endogenous reference gene (RNU48) and relative to the control (the HVs) 78 .
Target gene analysis. In our study, we attempted to identify the miRNAs' levels in LoSc patients compared to healthy volunteers (HVs) using Pathway-focused miScript miRNA PCR Arrays analysis (Table 1). Having done the bioinformatics analysis, we have selected 6 miRNAs. Therefore, we validated the chosen miRNAs in the group of LoSc patients using RT-qPCR assay. We used target gene prediction of the selected miRNAs and tried to check their potential role in LoSc pathogenesis. Relevant target genes were identified by using databases Target Scan v7.1 (https ://www.targe tscan .org) 79 , which predicts miRNAs biological targets, miRDB v4.0 (https ://mirdb .org/miRDB /) 80 , and miRSearch v3.0 (https ://www.exiqo n.com/miRSe arch). Moreover, we performed pathway analysis by DNA Intelligent Analysis (DIANA)-miRPath v5.0 software, based on the data from Ensembl v69 and miRBase v18 (https ://diana .imis. athen a-innov ation .gr) 81 , which uses miRNA targets based on DIANA-microT-CDS, and predicts miRNA-gene interaction. All the PCR reactions were run in 3 replicates. The Ct values were normalized to RNU48 for miRNAs. Gene expression was analyzed using the difference in cycle threshold (ΔCt) method using the manufacturer's software (BIO-RAD). The relative miRNAs' serum levels were calculated using the 2 −ΔΔCt method and presented as the fold change of the control (HVs). Statistical analysis. The data were statistically analyzed using STATISTICA 13 software. Minimum and maximum values, mean (M) and standard deviation (SD) or median and interquartile range (IQR) were estimated for continuous variables, as well as absolute numbers (n) and percentages (%) of the occurrence of items for categorical variables. All the analyses were conducted firstly for all the patients (both female and male), and secondly only for the female patients.
We used statistical tests as follows: Mann-Whitney's U test (to compare age as well as the selected miRNAs' levels in the serum between LoSc patients and HVs), χ 2 test (to compare gender between the LoSc patients and the HVs), Mann-Whitney's U test (to compare the selected miRNAs' serum levels between two groups of LoSc dichotomous characteristics), Kruskal-Wallis' H test (to compare the selected miRNAs' levels in the serum between three or more groups of LoSc polytomous characteristics), Spearman's correlation coefficient r (to correlate the selected miRNAs' serum levels with LoSc continuous characteristics).
Since miRNA levels for both groups of patients were assessed relative to the control, in case of significant differences in miRNA levels between the two groups of LoSc dichotomous characteristics, we calculated the relative fold change between two groups. We divided the relative fold change for the first group by the relative fold change for the second group where the former was the group with a higher relative fold change relative to the control.
The significance level was assumed at 0.05. www.nature.com/scientificreports/ Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Scientific Reports
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.