Key Points
Question
Is there a difference in interferon-stimulated gene (ISG) expression in nonlesional skin and peripheral blood mononuclear cell (PBMC) levels in patients with lupus with and without a history of cutaneous lupus erythematosus (CLE)?
Findings
In this cross-sectional study including 101 participants, 36 patients with a history of CLE with or without SLE had significantly higher ISG expression in PBMCs compared with 10 patients with SLE without a history of CLE. Nonlesional cutaneous and PBMC ISG expression were strongly correlated.
Meaning
Nonlesional skin is an immune-activated site in CLE that relates to systemic interferon activation, whereas patients with SLE without a history of CLE showed less interferon activation.
This cross-sectional study compares nonlesional skin and peripheral blood mononuclear cell interferon-stimulated gene expression in patients with lupus with and without a history of cutaneous lupus.
Abstract
Importance
Interferons (IFNs) play a crucial role in systemic lupus erythematosus (SLE) pathophysiology and are increased in cutaneous lupus erythematosus (CLE) lesions and blood. Recently, IFN-stimulated genes (ISGs) have been shown to be expressed in nonlesional skin of patients with SLE, suggesting that the nonlesional skin functions as an immune-activated site. Whether this is the case in all patients with SLE remains to be understood.
Objective
To compare nonlesional skin and peripheral blood mononuclear cell (PBMC) ISG expression in patients with lupus with and without a history of cutaneous lupus erythematosus.
Design, Setting, and Participants
This cross-sectional study at a single-center at a tertiary referral center included patients with a history of cutaneous lupus without SLE (CLEwoSLE), patients with SLE with CLE (SLEwCLE), patients with SLE without CLE (SLEwoCLE), and healthy controls (HCs). All SLE patients met the European League Against Rheumatism (EULAR) and the American College of Rheumatology (ACR) classification criteria and cutaneous lupus was diagnosed by a dermatologist. Data analysis occurred from January 2024 to May 2025.
Main outcomes and measures
ISG expression in PBMCs and nonlesional skin was assessed via calculation of IFN score.
Results
Overall, 74 of 101 participants were female individuals (73%), and the median (IQR) age varied between 44 (41-50) and 64 (53-68) years between groups. IFN scores in PBMCs were higher in SLEwCLE compared with patients with SLEwoCLE. Similarly, SLEwCLE patients showed highest levels of IFN scores in nonlesional skin. IFN scores in PBMCs and nonlesional skin were strongly correlated (r = 0.83, P < .001).
Conclusion and relevance
This cross-sectional study found that ISGs, as represented by IFN scores, in nonlesional skin and PBMCs were elevated in patients with lupus with a history of CLE compared with patients without CLE, suggesting that patients with lupus with and without CLE comprise 2 endotypes, with stronger IFN dysregulation occurring in patients with CLE.
Introduction
Cutaneous lupus erythematosus (CLE) can present alone or as one of the most common manifestations of systemic lupus erythematosus (SLE).1 SLE is characterized by augmented production of interferon (IFN), which is highly expressed in CLE lesions and in circulating blood cells.1 Importantly, IFN expression is also increased in nonlesional, healthy-appearing skin of patients with SLE.2,3,4 Here, keratinocytes exhibit robust IFN-κ secretion resulting in education of myeloid cells and increased epidermal cell death, thereby contributing to photosensitivity.2,3,4 However, the relationship between the cutaneous IFN signature and systemic IFN activation remains unclear. Furthermore, it is unknown whether the nonlesional skin functions as an immune-activated site in patients with SLE with and without CLE because previous reports primarily examined patients with SLE with CLE. Therefore, we compared IFN-stimulated genes (ISGs) using a composite IFN score comparing the sum of normalized expression of individual ISGs in nonlesional skin and peripheral blood mononuclear cells (PBMCs) between patients with lupus with or without a history of CLE.
Methods
For this cross-sectional study, patients with cutaneous lupus without SLE (CLEwoSLE), patients with SLE with a history of CLE (SLEwCLE), patients with SLE without a history of CLE (SLEwoCLE), and healthy controls (HCs) were recruited at the University of Michigan between 2019 and 2024. HCs were self-identified as free of history of autoimmune or skin disease. All patients with SLE were assessed by a rheumatologist and met the 2019 European League Against Rheumatism and the American College of Rheumatology (EULAR/ACR) classification criteria,5 whereas CLE was confirmed by a dermatologist. The study was approved by the University of Michigan institutional review board (HUM00151834) and conducted in accordance with the Declaration of Helsinki. All participants provided written informed consent. Race and ethnicity were self-reported. Data analysis was completed from January 2024 to May 2025. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines.
PBMCs were isolated from whole blood by density centrifugation. Punch biopsies were collected from nonlesional, non–sun-exposed skin from the upper thigh. For RNA-bulk sequencing, RNA was isolated from PBMCs and skin via Zymo Direct-zol RNA isolation kits according to manufacturer protocols. Libraries were generated with poly-A kits and sequencing was performed using an Illumina-NovaSeq-X system through the University of Michigan Advanced Genomics Core. After quality control, trimming and read alignment, downstream analyses and normalization were performed with R statistical software (DESeq2 package; R Foundation).6 Age and sex were adjusted for. Combat-seq was used to remove batch effects.7 We detected 23 205 genes from PBMCs and 19 754 genes from nonlesional skin. Differentially expressed genes (DEGs) were defined as having an uncorrected P value of <.05 and an absolute log2-fold change of ≥0.5. Ingenuity pathway analysis from QIAGEN was performed on DEGs.8 Normalized expression of selected ISGs (IFN score) was summarized as the sum of Z scores (normalized for the HC group), using the following summation:
 |
based on 138 genes that were upregulated after stimulating keratinocytes with IFNα in vitro (eTable 1 in Supplement 1).9 These genes comprise commonly assessed ISGs. IFN scores are presented as means (SDs). We used ANOVA to compare IFN scores between groups and Tukey post hoc test for pairwise comparisons. Pearson coefficient was used to test for correlation. A 2-sided P value of <.05 was considered significant. Analyses were performed in R statistical software (version 4.4.0; R Core Team).
Results
We enrolled 7 patients with CLEwoSLE, 29 with SLEwCLE, 10 with SLEwoCLE, and 55 HCs. Demographics, clinical characteristics, and availability of nonlesional skin and PBMC RNA sequencing data are depicted in the Table. EULAR/ACR criteria per group are shown in eTable 2 in Supplement 1.
Table. Demographic and Clinical Characteristics.
| Characteristics | No. (%) | |||
|---|---|---|---|---|
| HC (n = 55) | CLEwoSLE (n = 7) | SLEwCLE (n = 29) | SLEwoCLE (n = 10) | |
| Available RNA-sequencing data | ||||
| PBMC | 55 (100) | 6 (86) | 21 (72) | 7 (70) |
| Nonlesional skin | 53 (96) | 6 (86) | 25 (86) | 8 (80) |
| PBMC and nonlesional skin | 53 (96) | 5 (71) | 14 (48) | 5 (50) |
| Demographics | ||||
| Age, median (IQR), y | 49 (29-60) | 64 (53-68) | 44 (41-50) | 50 (42-60) |
| Female | 35 (64) | 6 (86) | 25 (86) | 8 (80) |
| Male | 20 (36) | 1 (14) | 4 (14) | 2 (20) |
| Race | ||||
| Asian | 3 (6) | 0 | 1 (3) | 0 |
| Black | 4 (7) | 1 (14) | 5 (17) | 4 (40) |
| White | 42 (76) | 6 (86) | 19 (66) | 6 (60) |
| Othera | 1 (2) | 0 | 1 (3) | 0 |
| More than 1 | 5 (9) | 0 | 3 (10) | 0 |
| Disease characteristics, median (IQR) | ||||
| EULAR/ACR | NA | 4 (48) | 24 (1628) | 24 (2030) |
| SLEDAI | NA | 2 (12) | 4 (28) | 5 (46) |
| CLASI activity | NA | 6 (416) | 1 (07) | 0 |
| CLASI damage | NA | 0 (06) | 0 (06) | 0 |
| Cutaneous lupus subtype | ||||
| ACLE | NA | 0 | 17 (59) | 0 |
| SCLE | NA | 5 (71) | 8 (28) | 0 |
| DLE | NA | 3 (43) | 12 (41) | 0 |
| Chilblain lupus | NA | 0 | 1 (3.4) | 0 |
| Lupus tumidus | NA | 0 | 2 (6.9) | 0 |
| Panniculitis | NA | 1 (14) | 1 (3.4) | 0 |
| More than 1 subtype | NA | 2 (29) | 11 (38) | 0 |
| Antibodies | ||||
| Anti-SSA/Ro | NA | 4 (57) | 15 (52) | 3 (30) |
| Anti-SSB/La | NA | 0 | 5 (17) | 2 (20) |
| Anti-dsDNA | NA | 0 | 16 (55) | 7 (70) |
| Anti-Sm | NA | 0 | 13 (45) | 2 (20) |
| Medication | ||||
| Hydroxychloroquine | NA | 3 (43) | 22 (76) | 7 (70) |
| Prednisone | NA | 1 (14) | 19 (66) | 9 (90) |
| Methotrexate | NA | 0 | 1 (3.4) | 0 |
| Mycophenolate mofetil | NA | 0 | 8 (28) | 4 (40) |
| Azathioprine | NA | 0 | 2 (6.9) | 1 (10) |
| Rituximab | NA | 1 (14) | 3 (10) | 2 (20) |
| Belimumab | NA | 0 | 4 (14) | 1 (10) |
| Dapsone | NA | 1 (14) | 1 (3.4) | 0 |
Abbreviations: ACLE, acute cutaneous lupus erythematosus; CLASI, cutaneous lupus disease area and severity index; CLE, cutaneous lupus erythematosus; CLEwoSLE, CLE without SLE; DLE, discoid lupus erythematosus; dsDNA, double stranded DNA; EULAR/ACR, European Alliance of Associations for Rheumatology/American College of Rheumatology; HC, healthy controls; NA, not applicable; PBMCs, peripheral blood mononuclear cells; SCLE, subacute cutaneous lupus erythematosus; SLEDAI, systemic lupus erythematosus disease activity index; SLE, systemic lupus erythematosus; SLEwCLE, SLE with history of CLE; SLEwoCLE, SLE without history of CLE; ; sm, smith.
Other was a self-reported category.
First, we evaluated IFN scores in PBMCs. Patients with CLEwoSLE (mean [SD], 148 [91.1]) and those with SLEwCLE (mean [SD], 188 [184]) exhibited increased IFN scores compared with HCs (mean [SD], 0 [51.8]; P < .01 and P < .001, respectively), whereas PBMCs from SLEwoCLE did not (mean [SD], 62.9 [122]) (Figure, B). We then examined IFN scores in nonlesional skin. Similarly, we found that SLEwCLE had elevated IFN scores compared with HCs (mean [SD] skin IFN score, 0 [56]; P < .001; Figure, C) and also demonstrated the highest mean (SD) IFN score (119 [136]) in nonlesional skin among all groups. We did not find significant differences in cutaneous IFN scores between other groups.
Figure. Peripheral Blood Mononuclear Cell (PBMC) and Nonlesional Interferon (IFN) Scores.
A, General study design. B and C, PBMC and cutaneous IFN scores depicted per group. D, Scatterplot illustrating correlation between cutaneous and PBMC IFN scores. PBMC and cutaneous IFN scores for SLEwCLE and CLEwoSLE stratified by CLE subtype (13/35 patients had more than one subtype of CLE) (E). An IFN score of 0 is equivalent to the mean score of the healthy control group. A negative score thus indicates a lower score than the mean of the HC group. ACLE indicates acute cutaneous lupus erythematosus; CLE, cutaneous lupus erythematosus; CLEwoSLE, CLE without SLE; DLE, discoid lupus erythematosus; HC, healthy controls; IFN, interferon; NS, not significant; PBMC, peripheral blood mononuclear cells; SCLE, subacute cutaneous lupus erythematosus; SLE, systemic lupus erythematosus; SLEwCLE, SLE with history of CLE; SLEwoCLE, SLE without history of CLE.
aP < .001.
bP = .009.
cP = .04.
The increased IFN scores in PBMCs and nonlesional skin in patients with lupus with a history of CLE did not seem to be associated with a specific CLE subtype (Figure, E) nor by medication use (eTable 3 in Supplement 1). Overall, more genes were differentially expressed in PBMCs than in nonlesional skin. Mainly IFN-related pathways (antiviral responses) were enriched in PBMCs from patients with SLEwCLE compared with those with SLEwoCLE (eFigure 1A, eFigure 2A in Supplement 1). As in PBMCs, predominantly IFN-related pathways were enriched in nonlesional skin (eFigure 1B, eFigure 2B in Supplement 1).
IFN scores from PBMCs and nonlesional skin from the same patients were strongly correlated (r = 0.83; 95% CI, 0.67-0.92; P < .001). PBMC and cutaneous IFN scores were both associated with systemic lupus erythematosus disease activity index (SLEDAI) scores (r = 0.37; 95% CI, 0.04-0.63; P = .03; and r = 0.43; 95% CI, 0.13-0.65; P = .007; eFigure 3 in Supplement 1). SLEDAI scores were similar between SLEwCLE and SLEwoCLE (Table). We did not find a correlation between CLE activity, assessed with the cutaneous lupus activity and severity index (CLASI) and IFN scores in PBMCs or nonlesional skin (eFigure 3 in Supplement 1).
Discussion
In this cross-sectional study, we found that a history of CLE, in both patients with or without SLE was associated with higher IFN scores in blood and nonlesional skin than in patients without a history of CLE. Furthermore, we found a strong correlation between IFN scores in nonlesional skin and PBMCs.
These findings support the hypothesis that nonlesional skin in SLE is an immune-activated site with increased ISG expression that relates to systemic IFN production.2,3 Importantly, patients with SLEwoCLE had lower levels of IFN scores in both blood and nonlesional skin compared with those with SLEwCLE, whereas patients with CLEwoSLE showed rather similar levels to those with SLEwCLE. These findings suggest that patients with lupus with and without CLE comprise 2 endotypes with stronger IFN dysregulation occurring in patients with CLE and a lower IFN signature in those without CLE. This may also explain why anti-IFN treatment is especially effective in patients with SLE with CLE10,11 and why patients with lower IFN signatures may not respond as well.12
Similar to earlier studies,13 SLEDAI score was correlated with PBMC and skin IFN scores in our data. However, we did not find a relationship between CLASI and IFN scores in either PBMCs or nonlesional skin.14,15 It should be noted, however, that only a few patients had high CLASI activity in our cohort, which may have skewed this analysis. Also, our results did not identify a preferential increase in the nonlesional skin vs PBMC IFN scores.4 This may be related to differences in UV exposure of nonlesional skin prior to sampling because UV exposure has been linked to higher IFN signatures,1,2,16 and our samples were from UV-unexposed skin.
The strong correlation between PBMC and nonlesional skin IFN scores found in this study, together with other reports,2,9,16 suggests that nonlesional skin is involved in local and systemic IFN response. Mechanisms of IFN production in nonlesional skin in lupus, and how this is connected to systemic IFN production is a relevant topic that needs to be further explored. Furthermore, given that type-1 IFNs are less prominent in patients with SLEwoCLE, it is critical to identify other molecular pathways that drive disease in these patients. Eventually, this will lead to personalized treatment approaches based on SLE endotypes.
Limitations
This cross-sectional study had some limitations, including the relatively small sample size and cross-sectional design. This may have particularly affected the data for those with CLEwoSLE because they only had a nonsignificant trend toward increased IFN scores in nonlesional skin compared directly with HCs. Also, given our relatively small sample size, we did not perform subanalyses based on cutaneous lupus manifestations for this study. Hence, findings should be replicated in other cohorts. In addition, we only focused on type-1 IFN related genes in this study; further analysis of other pathways may identify additional differences.
Conclusions
In this cross-sectional study, IFN scores in nonlesional skin and PBMCs were higher in patients with lupus with a history of CLE compared with patients without skin involvement, suggesting different immune activation between patients with and without CLE. A strong correlation between nonlesional skin and PBMC IFN scores supports that the nonlesional skin is an immune-activated site that has strong interplay with the systemic immune system in SLE.
eTable 1. Interferon stimulated genes
eTable 2. EULAR/ACR classification criteria items per group
eFigure 1. Volcano plots comparing sequencing data between SLEwCLE vs. SLEwoCLE
eFigure 2. Ingenuity pathway analysis (QIAGEN), graphical summary
eFigure 3. Scatterplots of CLASI activity, SLEDAI and IFN scores
Data Sharing Statement
References
- 1.Maz MP, Martens JWS, Hannoudi A, Reddy AL, Hile GA, Kahlenberg JM. Recent advances in cutaneous lupus. J Autoimmun. 2022;132:102865. doi: 10.1016/j.jaut.2022.102865 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Sarkar MK, Hile GA, Tsoi LC, et al. Photosensitivity and type I IFN responses in cutaneous lupus are driven by epidermal-derived interferon kappa. Ann Rheum Dis. 2018;77(11):1653-1664. doi: 10.1136/annrheumdis-2018-213197 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Billi AC, Ma F, Plazyo O, et al. Nonlesional lupus skin contributes to inflammatory education of myeloid cells and primes for cutaneous inflammation. Sci Transl Med. 2022;14(642):eabn2263. doi: 10.1126/scitranslmed.abn2263 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Psarras A, Alase A, Antanaviciute A, et al. Functionally impaired plasmacytoid dendritic cells and non-haematopoietic sources of type I interferon characterize human autoimmunity. Nat Commun. 2020;11(1):6149. doi: 10.1038/s41467-020-19918-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Aringer M, Costenbader K, Daikh D, et al. 2019 European League Against Rheumatism/American College of Rheumatology Classification Criteria for Systemic Lupus Erythematosus. Arthritis Rheumatol. 2019;71(9):1400-1412. doi: 10.1002/art.40930 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):550. doi: 10.1186/s13059-014-0550-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Zhang Y, Parmigiani G, Johnson WE. ComBat-seq: batch effect adjustment for RNA-seq count data. NAR Genom Bioinform. 2020;2(3):lqaa078. doi: 10.1093/nargab/lqaa078 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Krämer A, Green J, Pollard J Jr, Tugendreich S. Causal analysis approaches in ingenuity pathway analysis. Bioinformatics. 2014;30(4):523-530. doi: 10.1093/bioinformatics/btt703 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Tsoi LC, Hile GA, Berthier CC, et al. Hypersensitive IFN responses in lupus keratinocytes reveal key mechanistic determinants in cutaneous lupus. J Immunol. 2019;202(7):2121-2130. doi: 10.4049/jimmunol.1800650 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Werth VP, Furie RA, Romero-Diaz J, et al. ; LILAC Trial Investigators . Trial of anti-BDCA2 antibody litifilimab for cutaneous lupus erythematosus. N Engl J Med. 2022;387(4):321-331. doi: 10.1056/NEJMoa2118024 [DOI] [PubMed] [Google Scholar]
- 11.Morand EF, Furie R, Tanaka Y, et al. ; TULIP-2 Trial Investigators . Trial of anifrolumab in active systemic lupus erythematosus. N Engl J Med. 2020;382(3):211-221. doi: 10.1056/NEJMoa1912196 [DOI] [PubMed] [Google Scholar]
- 12.Jayne D, Rovin B, Mysler EF, et al. Phase II randomised trial of type I interferon inhibitor anifrolumab in patients with active lupus nephritis. Ann Rheum Dis. 2022;81(4):496-506. doi: 10.1136/annrheumdis-2021-221478 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Psarras A, Wittmann M, Vital EM. Emerging concepts of type I interferons in SLE pathogenesis and therapy. Nat Rev Rheumatol. 2022;18(10):575-590. doi: 10.1038/s41584-022-00826-z [DOI] [PubMed] [Google Scholar]
- 14.Gómez-Bañuelos E, Goldman DW, Andrade V, Darrah E, Petri M, Andrade F. Uncoupling interferons and the interferon signature explain clinical and transcriptional subsets in SLE. medRxiv. 2023. doi: 10.1101/2023.08.28.23294734 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Braunstein I, Klein R, Okawa J, Werth VP. The interferon-regulated gene signature is elevated in subacute cutaneous lupus erythematosus and discoid lupus erythematosus and correlates with the cutaneous lupus area and severity index score. Br J Dermatol. 2012;166(5):971-975. doi: 10.1111/j.1365-2133.2012.10825.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Skopelja-Gardner S, An J, Tai J, et al. The early local and systemic type I interferon responses to ultraviolet B light exposure are cGAS dependent. Sci Rep. 2020;10(1):7908. doi: 10.1038/s41598-020-64865-w [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. Interferon stimulated genes
eTable 2. EULAR/ACR classification criteria items per group
eFigure 1. Volcano plots comparing sequencing data between SLEwCLE vs. SLEwoCLE
eFigure 2. Ingenuity pathway analysis (QIAGEN), graphical summary
eFigure 3. Scatterplots of CLASI activity, SLEDAI and IFN scores
Data Sharing Statement
