Systemic scleroderma (SS) is a systemic autoimmune disorder characterized by vasculopathy followed by progressive fibrosis of the skin and internal organs manifestations [1] and high mortality rate: 25% of patients die within the first 5 years after making the diagnosis, 37.5% die within the first 10 years [2–5]. The mean life expectancy of SS patients is 16–34 years less than the population average [4]. The disease prevalence varies between 38 and 341 per million population annually, and the disease incidence varies between 8 and 56 individuals per million per year based on the data from different countries [6–11]. The most common cause of death associated with SS in damage to internal organs: the lung, heart, gastrointestinal tract [4, 5]. In addition to these manifestations, in 35% of patients the disease course is complicated by digital ulcers, which, in turn, often leading to gangrene and fingers amputation, which causes severe functional impairment in such patients [12–13].

Although the disease is multisystemic, skin lesion is a distinctive feature that determines the clinical and prognostic stratification. SS is classified based on the extent of skin involvement into the diffuse (dSS) and limited (lSS) forms that have different progression rate, disease course features, and prognosis [1].

The molecular mechanisms that underlie the development of SS are poorly understood. This hinders the selection of targeted therapy selection. Currently, the range of potential treatment options is extremely limited, necessitating the search for new therapies. Although the T cell responses of types Th2 and Th17 play an important role in the SS pathogenesis [15, 16], it has been shown that activation of interferon pathways, especially type I, is more strongly associated with SS, than immune responses of other types [17, 18].

Recent studies have shown similar expression of the genes involved in the interferon cascade in patients with systemic lupus erythematosus and SS [19–21]. Upregulation of the IFN-associated genes is also typical for rheumatoid arthritis, Sjogren's syndrome, and polymyositis [23–29].

Activation of the IFN-I pathway and, as a result, the “interferon” gene signature are observed in blood and skin of a large number of patients with even early-stage SS and, according to some data, are associated with severity of lesions (including pulmonary and skin lesions) [30, 31]. The conclusion about the relationship between the expression of IFN-induced genes and the disease severity was disproven by subsequent research [28, 32].

The up-to-date EULAR guidelines (updated 2023) reflect a shift towards targeted approaches to SS, which increases the importance of the validated stratification biomarkers [33]. Such observations raise the question about the potential effectiveness of the interferon-mediated pathway blocking in some patients with SS.

The study aimed to assess the expression of interferondependent genes in peripheral blood cells and affected skin areas of SS patients in order to estimate the potential for patient stratification and the prognosis of a promising, but not yet approved therapy with antibodies against interferon receptors in SS [34]. The hypothesis was tested that the IFN-I-signature can be reflected in peripheral blood of patients with SS, the analysis of which can become an alternative to repeated skin biopsy when performing patient stratification and monitoring.

METHODS

Patients

The patients aged 21–77 with the limited or diffuse cutaneous systemic scleroderma form, who were admitted to the Rheumatology Department of the Moscow City Research Center Hospital No. 52 from April 2023 to February 2025 and met the ACR/EULAR2013 criteria, were included in the study [35].

Inclusion criteria: detection of the antinuclear factor in patient’s blood by the indirect immunufluorescence method. When detecting the antinuclear factor, the range of nuclear antibodies was tested by immunoblotting (tab. 1).

Exclusion criteria: another systemic autoimmune disorder (such as rheumatoid arthritis, idiopathic inflammatory myopathy, systemic lupus erythematosus); signs of infectious disease (upon physical examination).

The patients having no specific antibodies or negative immunoblot test results were not excluded from the study, since:

1. according to the ACR/EULAR criteria, the fact of having specific antibodies is not a mandatory criterion for establishing the diagnosis of SS [35];
2. antibodies against RNA polymerase III, which are not detected in the Russian Federation, are also typical for SS; furthermore, SS can be associated with the antibodies that are outside the spectrum assessed.

The screening tests for antinuclear antibodies were performed by enzyme-linked immunoassay (ELISA) with the Multiscan FC semi-automatic ELISA analyzer (Thermo Fisher Scientific Inc., USA) using the ANA-Screen ELISA IgG reagent kit (Euroimmun AG, Germany) in accordance with the manufacturer’s instructions. Confirmatory testing for specific antinuclear antibodies was performed by immunoblotting using the ANA profile 1 IgG reagent kit (Euroimmun AG, Germany) in accordance with the manufacturer’s instructions.

The clinical and immunological assessment of patients was conducted that included the following: assessment of disease activity based on EScSG (tab. 1) [36], evaluation of the capillaroscopic pattern at the time of examination, and laboratory testing for specific antibodies. All the patients were tested for the presence/absence of damage to possible target organs: skin (Rodnan skin score was assessed), joints (the number of painful and swollen joints was estimated), lung (all the patients underwent chest multislice computed tomography (MSCT)), heart and pulmonary artery (electrocardiography, echocardiography and gastrointestinal catheterization (e.g. barium swallow test in one patient) were performed).

 Biomaterial selection and RNA extraction

Sample collection was performed in the clinic during general patient examination at admission to the hospital. A total of 48 patients with SS were included in the study. In 25 patients, peripheral blood samples only were collected as biomaterial for testing. In another 10 patients, both the affected skin specimens and peripheral blood samples were collected; in 13 patients, the affected skin specimens only were collected. Skin biopsy specimens were collected from the forearm (area with the thickest skin) by incisional biopsy.

The healthy skin samples collected from three donors not diagnosed with SS were used as the reference samples, along with the peripheral venous blood samples of 31 healthy donors aged 20–54 years. A total of 20% of the cohort of healthy donors were males.

The affected skin samples 4 mm in diameter were placed in the MACS® Tissue Storage Solution (Miltenyi, USA) at +4 °С and transferred to the laboratory for RNA extraction. The resulting samples were ground in liquid nitrogen with RLT lysis buffer added simultaneously. RNA extraction was performed using the HiPure Total RNA Kit (Magen, China) in accordance with the manufacturer’s instructions. RNA concentration was measured using the Qubit 3.0 fluorometer and the reagent kit (Thermo Fisher Scientific, USA).

Blood samples (4 mL) were collected into the EDTA-coated tubes (final concentration of 2 mg/mL).

After collection, whole blood was stored at 4 °С until mononuclear cells were isolated. Mononuclear cells were isolated from peripheral blood by sedimentation (Ficoll-Paque density gradient centrifugation (density 1.077 g/cm³)) (PanEco, Russia). The resulting cell fraction was placed in the RLT lysis buffer (Qiagen, Germany) and stored at –80 °С until total RNA was extracted. The total RNA was extracted using the HiPure Total RNA Kit (Magen, China) in accordance with the manufacturer’s instructions.

The RNA concentration was measured using the Qubit 3.0 fluorometer (USA) and the reagent kit (Thermo Fisher Scientific, USA). The quality of the RNA sample extracted was evaluated by agarose gel electrophoresis. The extracted RNA was frozen and stored at the temperature of –80 °С until the reverse transcription and real-time PCR were launched.

RT-PCR and data analysis

The one-tube quantitative RT-PCR (reverse transcription PCR) was performed using the One-Tube RT-PCR TaqMan reagent kit (Evrogen, Russia) [37]. The kit contains a readyto-use master mix comprising a reaction buffer for RT and PCR, nucleotide triphosphates, and the hot start Taq DNA polymerase. In the first phase, during the first-strand cDNA synthesis, the Taq polymerase was inactivated by monoclonal antibodies; heating at 95 °C prior to PCR ensured the rapid hot start. The modified MMLV reverse transcriptase was provided to the reaction mixture separately. Primers for first-strand cDNA synthesis and subsequent PCR amplification were added to the reaction in a concentration of 0.4 µМ. Real-time PCR product accumulation imaging was performed using the TaqMan type fluorescent probes. The probes were added to the reaction at a concentration of 0.1 µМ. All the oligonucleotides specific for the test and reference genes were selected such that optimal performance was achieved under a single, versatile RT-PCR protocol.

Reverse transcription: 55 °C, 15 min, one cycle, without fluorescence acquisition. This was followed by a reverse transcriptase inactivation/polymerase activation step: 95 °C, 1 min, one cycle, without reading. Amplification was carried out for 40 cycles: denaturation — 95 °C, 15 s (without acquisition); annealing — 60 °C, 20 s (with fluorescence acquisition); elongation — 72 °C, 20 s (without acquisition).

To analyze the interferon signature, the expression of IFIT1, IFIT3, IFI27, IFI44, ISG15, XAF1 was assessed, that was previously validated in the peripheral blood samples of patients with systemic lupus erythemathosus (SLE), along with that of the TBP reference gene [29, 31].

The same method was used to assess the expression of the SFRP4 marker gene which is a recognized marker associated with fibrosis progression in scleroderma [22].

Assessment of the relative expression of the interferondependent genes involved normalization to the expression of the TBP housekeeping gene.

Relative expression was determined by the ΔΔCt method with normalization to the reference gene amplified in the same PCR. Relative expression of the test gene was determined based on the amplification effeciency and the difference in cycle thresholds (ΔCt) between the target and reference gene.

Statistical data processing

The nonparametric methods (appropriate for small sample sizes) were used for statistical processing.

The Mann–Whitney test was used to compare gene expression in the donor samples of the experimental (SS patients) and control groups. When comparing gene expression values of the skin and blood samples, the significance of differences between samples was estimated using the paired Wilcoxon signed-rank test. All p-values were further adjusted using the Benjamini–Hochberg (False Discovery Rate (FDR)) procedure. Spearman’s rank correlation coefficient was used to assess correlations between variables.

The IFN-I signature value was calculated as the mean of the standardized scores (z-score) of relative expression of five genes (IFIT3, IFI27, IFI44, ISG15, XAF1). 

The standardized scores for the IFN-I signature (averaged between skin and blood) were calculated for the subgroup of patients from whom paired blood and skin samples were obtained (n = 10).

The standardized scores of the IFN-I signature values in SS patients were calculated relative to healthy donors, separately in blood and skin samples).

Mean values and the standard deviation were calculated for clinical data (tab. 2 and tab. 3). However, due to the indicators’ limited applicability to small samples (with the unproven hypothesis about the normally distributed values) the median and interquartile range were also calculated.

The standard deviation of binary types of data (such as clinical mainfestations) was calculated for the Bernoulli distribution.

Work with the tables, data correction, charting, and statistical analysis were accomplished using the R integrated features and supplementary libraries: tidyverse, ggplot2, and corrplot.

RESULTS

The study included 48 SS patients, mostly females (90%), a mean age 61 years, and mean disease duration 13.5 of years, primarily with the disease chronic course (84%) and limited cutaneous form (62.5%). The leading organ damage and manifestations: interstitial lung disease (56%, mainly NSIP); esophageal (71%) and joint (58%) lesions. were observed. The average skin activity was low (mRSS 7), the Raynaud’s phenomenon was diagnosed in almost all patients (98%), digital ulcers were present in 31%. The results of the clinical and immunological assessment of SS patients are presented in tab. 1. 

The integrated IFN-I signature assessment involved the use of the modified test system conprising five genes (IFI27, XAF1, IFI44, IFIT3, ISG15) using one reference gene instead of two, which had been previously tested in peripheral blood samples of patients with SLE [32, 38]. SFRP4 (secreted frizzled-related protein) was used as a marker gene of the disease, the expression levels of which are associated with skin fibrosis in SS [22].

Significant differences in expression levels of all five test genes between the groups of SS patients and healthy donors were observed in the RNA samples obtained from the skin; significant differences in expression of 4 test genes out of 5 (except IFI27) between the groups of SS patients and healthy donors were reported for blood samples (fig. 1A, B; tab. 2). The SFRP4 expression levels were higher in the skin samples of patients with SS compared to the skin samples of healthy donors (fig. 1А).

Comparison of expression levels between compartments performed for the paired blood/skin samples (fig. 2; tab. 3) showed that the XAF1, IFI44, IFIT3, ISG15 expression levels in patients’ blood were significantly higher than in skin samples. The IFI27 gene expression, in contrast, was more pronounced in the skin (fig. 2А, C). The direction of changes was consistent across both sample types, which supports the effectiveness of using both skin biopsy samples and blood samples for the diagnosis and dynamic monitoring (fig. 2B, D). The IFN-I integral index of patients with SS was significantly higher compared to the reference threshold interval calculated based on the healthy donors’ values in both peripheral blood samples and the affected skin biopsy samples (fig. 2E, F). A limitation of the skin IFN-I signature index comparison is associated with the small number of samples collected from healthy donors. The integral index based on blood samples showed that the values were above the reference interval in 62% (22 patients).

The correlation analysis of clinical parameters and gene expression (fig. 3; tab. 4) has shown that the IFN-I signature genes (IFIT3, IFI27, IFI44, ISG15, XAF1) form a tightly coexpressed module (Rs between 0.52 and 0.87; p < 0.05 for the skin) with the significant positive correlations between genes, while SFRP4 exhibited little association with this module reflecting a distinct fibrosis-associated component.

In blood samples, the IFIT3, IFI27, ISG15, XAF1 genes also form a correlation cluster with each other, but the correlation values are lower (Rs between 0.38 and 0.63). The IFIT27 gene is most closely correlated to ISG15 (Rs = 0.77), but it is outside the common correlation cluster (Rs with IFIT3, IFI27, ISG15, XAF1 < 0.15).

Clinical indicators of disease severity (activity based on EScSG, progression, diffuse form) significantly correlate with each other.

The IFN-I signature genes demonstrate a non-significant correlation with the diffuse disease form: in skin samples, Rs is between 0.23 and 0.34, skin samples show a slight correlation for ISG15 and IFIT27 only (0.21 and 0.27, respectively). Age shows a weak negative association with the expression of the IFN-I-induced genes (Rs between –0.32 and –0.36) in skin samples; the correlations for blood are insignificant (0 to –0.13).

DISCUSSION

Recently, determination of the expression of the IFN-I-stimulated genes by PCR is increasingly used in clinical trials. Various diagnostic test systems for assessment of the IFN signature expression have been developed. Expression levels of the genes IFI44L, IFI44, MX1, MX2, OAS1, OAS2, OAS3, SIGLEC1, IFI35 are most often determined when determining the IFN signature [39]. Russian scientists have also proposed solutions in this area [40, 41]. In particular, one system involves the analysis of three genes (RIG-1, IFIT-1, IFIH-1). However, the expression normalization approach used in these panels seems to be suboptimal: HPRT is used as a reference gene in the first one, but standardization by ΔCt is not described, it is proposed to use GAPDH as a reference gene in the second one, which can result in artifacts due to the presence of numerous pseudogenes in the human genome. In this study, we used integrated assessment of the IFN-I signature using the test system of five genes: IFI27, XAF1, IFI44, IFIT3, ISG15 with normalization to one reference gene. Previously, we tested these genes individually and as a test system in peripheral blood samples from patients with SLE [32, 38].

According to the findings, the SS patients showed the increased expression of the IFN-induced genes compared to healthy donors in both skin and peripheral blood samples. These data confirming involvement of type I interferons in the disease pathogenesis are consistent with the previously reported results for SS [17, 21, 30]. Based on the results of our analysis the expression of IFN-I-induced genes shows a trend toward correlation in blood and affected skin samples. Such a trend has been earlier demonstrated in the large cohort of SS patients using the transcriptome profiling (microarray). It has been shown that the expression of the IFN-associated genes in the skin is consistent with similar changes in peripheral blood [42]. These and other observations suggest the informativeness of assessing IFN-I signature in blood samples of patients with SS [18].

Clinical indicators of the disease severity (activity based on EScSG, progression, diffuse form) significantly correlate with each other, but demonstrate only weak and heterogeneous correlations with the IFN-I signature. Thus, the applicability of the developed test system for patient stratification and predicting the success of therapy with the IFN-I receptor inhibitors should be assessed independently from the disease severity and form.

Since the levels of the assessed IFN-I signature in some SS patients are within the range close to that of controls (fig. 2), and there is no correlation between gene expression and the disease severity (fig. 3), it can be assumed that SS patients have various immunological patterns, in addition to the type I interferon activation patterns (by analogy to the patterns reported for SLE [26]). The fact that SS patients have both significantly higher and lower (at the level of the control group) expression levels of the IFN-induced genes suggests the need for patient stratification to predict the response to therapy) with the interferon receptor antibodies with the interferon receptor antibodies.

Given the current clinical trials of the IFN-I receptor inhibitor in SS (DAISY; NCT05631227) [37], the elevated IFN-I signature value can potentially be a criterion for patient selection and the tool to monitor the effectiveness of the response to therapy with the interferon receptor blocker (anifrolumab). Similarity of the interferon-associated pathways in SS and SLE [14, 23, 42] together with the available data on the anifrolumab efficacy in SLE [43] suggest that the use of this drug in SS can be highly effective in patients having high interferon signature values.

CONCLUSIONS

SS patients show the increased expression of interferoninduced genes in both skin and peripheral blood samples compared to healthy donors, which suggests the involvement of type I interferons in the SS pathogenesis. The expression levels of the IFN-I-induced genes show a trend toward correlation in blood samples and affected skin areas. The RT-qPCR panel developed that comprises the IFN-I-induced genes (IFI27, IFI44, IFIT3, ISG15, XAF1) has a potential for stratification of SS patients, as well as for assessment of the efficacy of target therapy with the interferon receptor blockers (in case such therapy is approved for SS). To verify this conclusion it is required to conduct further research focused on the correlation between the decrease in the interferon signature levels and the condition improvement in SS patients treated with the antibodies against the interferon receptor. We recommend using peripheral blood from SS patients for IFN-I-signature analysis by RT-PCR with the proposed gene panel, as this biomaterial is more accessible, less invasive and more reproducible; it also shows the informativeness potentially comparable with that of skin samples.