To the Editor: Dermatomyositis (DM) is an autoimmune, multiorgan disease. The type I interferon signature is characteristic in DM, and viral infection may be a contributing factor.1 Retrotransposons are divided into 2 groups: non-long terminal repeat retrotransposons include long interspersed nuclear element-1 (LINE-1), Alu, and short interspersed nuclear elements (SINE)-variable number of tandem repeats (VNTR)-Alu (SVA), whereas long terminal repeat retrotransposons include endogenous retroviruses.2 LINE-1 is a representative retrotransposon that regulates the expression of type I interferon through the MDA5 pathway.3 The expression of LINE-1 is observed in autoimmune diseases, such as systemic lupus erythematosus and Sjögren syndrome4; however, its expression pattern in DM is unknown. We investigated viral infections and the expression of retrotransposons.
We enrolled 7 pre- and 28 posttreatment patients with DM and 10 healthy individuals recruited from Gunma University. The autoantibody and treatment profiles and detected viruses by polymerase chain reaction are shown in Table I . Only a few DNA viruses, which are known to chronically infect humans, were detected among peripheral white blood cells after treatment. Next, quantitative polymerase chain reaction was performed to compare the expression of messenger RNA (mRNA) from peripheral white blood cells in posttreatment patients with DM with that in healthy control individuals. Two-group comparisons were assessed using the Mann-Whitney U test. The expression levels of LINE-1 and other retrotransposon mRNAs were significantly upregulated in patients with DM (LINE-1, 2.70 ± 3.19 vs 0.30 ± 0.19; HERVK14C, 3.54 ± 6.24 vs 0.27 ± 0.19; and SVA, 2.71 ± 2.74 vs 0.44 ± 0.20; all P < .0001). This upregulation was seen in not only anti-MDA5 antibody–positive patients but also in other autoantibody-positive patients as well as autoantibody-negative patients. To clarify whether altered DNA methylation contributes to the inappropriate expression of retrotransposons, expression levels of the methylation enzymes DNA methyltransferase (DNMT) 3A and DNMT3B were detected and showed significant down- and upregulation, respectively (0.62 ± 0.44 vs 1.15 ± 0.59 and 2.00 ± 3.00 vs 0.30 ± 0.19, respectively; P < .05). The expression levels of IFN-α2 and IFN-β1 were significantly increased (2.59 ± 4.59 vs 0.27 ± 0.18 and 4.08 ± 5.11 vs 0.31 ± 0.24, respectively; P < .0001). The LINE-1 promoter methylation was assessed by pyrosequencing, showing significant hypomethylation of LINE-1 promoter in patients with DM (78.12% ± 1.21% vs 80.24% ± 1.25%; P < .005, Mann-Whitney U test). Furthermore, the significant positive correlation was found between retrotransposon (LINE-1, HERVK14C, and SVA) and IFN-β1 expression (IFN-β1: r = 0.953, r = 0.941, and r = 0.948, respectively; all P < .0001; nonparametric Spearman test) (LINE-1 vs IFN-β1 in Fig 1 ) as well as DNMT3B and retrotransposons (DNMT3B: r = 0.906, r = 0.717, and r = 0.833, respectively).
Table I.
Clinical characteristics of DNA and RNA virus detection in patients with posttreatment dermatomyositis
| Patient | Age, y | Sex | Antibody | Therapy (per day) | DNA and RNA virus detection |
|||
|---|---|---|---|---|---|---|---|---|
| After treatment |
Before treatment |
|||||||
| Serum | Extracted RNA | Serum | Extracted RNA | |||||
| 1 | 55 | F | ARS | PSL 10 mg, CyA 50 mg | ND | - | ND | ND |
| 2 | 55 | F | MDA5 | PSL 30 mg, Tac 3 mg, IVCY | ND | CMV | ND | ND |
| 3 | 41 | F | TIF1γ | PSL 8 mg | - | - | ND | ND |
| 4 | 55 | F | Mi2 | PSL 40 mg | - | CMV | ND | ND |
| 5 | 51 | F | MDA5 | Unknown | - | - | ND | ND |
| 6 | 50 | M | MDA5 | PSL 10 mg, Aza 100 mg | - | - | ND | ND |
| 7 | 75 | F | negative | PSL 5 mg | - | - | ND | ND |
| 8 | 43 | M | negative | PSL 5 mg | - | - | ND | ND |
| 9 | 72 | F | negative | PSL 5 mg | - | EBV | ND | ND |
| 10 | 54 | F | negative | PSL 5 mg | - | - | ND | ND |
| 11 | 45 | F | negative | PSL 4 mg | - | - | ND | ND |
| 12 | 40 | F | ARS | PSL 10 mg | - | - | ND | ND |
| 13 | 66 | M | TIF1γ | - | - | - | ND | ND |
| 14 | 52 | M | Mi2 | PSL 3 mg | - | - | ND | ND |
| 15 | 76 | F | TIF1γ | PSL 5 mg | - | - | ND | ND |
| 16 | 46 | F | ARS | PSL 5 mg | - | - | ND | ND |
| 17 | 68 | F | TIF1γ | PSL 6 mg | - | - | ND | ND |
| 18 | 68 | F | negative | PSL 1 mg | - | - | ND | ND |
| 19 | 63 | F | negative | PSL 1 mg | - | HHV7 | ND | ND |
| 20 | 38 | M | negative | PSL 4 mg | - | - | ND | ND |
| 21 | 51 | M | TIF1γ | PSL 5 mg | - | - | ND | ND |
| 22 | 53 | M | Mi2 | PSL 6 mg | - | - | ND | ND |
| 23 | 63 | F | negative | PSL 5 mg | - | - | ND | ND |
| 24 | 60 | F | negative | PSL 4 mg | - | - | ND | ND |
| 25 | 43 | M | negative | PSL 5 mg | - | - | ND | ND |
| 26 | 59 | F | MDA5 | Unknown | - | CMV, EBV | ND | ND |
| 27 | 54 | F | MDA5 | Unknown | - | - | ND | ND |
| 28 | 52 | F | MDA5 | Unknown | ND | EBV | ND | ND |
| 29 | 58 | M | negative | - | ND | ND | - | - |
| 30 | 60 | M | unknown | - | ND | ND | - | - |
| 31 | 71 | M | unknown | - | ND | ND | - | - |
| 32 | 44 | F | MDA5 | - | ND | ND | - | ND |
| 33 | 10 | F | MDA5 | - | ND | ND | - | ND |
| 34 | 45 | M | MDA5 | - | ND | ND | - | ND |
| 35 | 57 | F | MDA5 | - | ND | ND | - | ND |
The detected viruses were rhino/enterovirus, enterovirus 68, parechovirus, coronavirus, HSV, CMV, parvovirus B19, VZV, HHV6, HHV7, HHV8, and EBV.
Aza, Azathioprine; CMV, cytomegalovirus; CyA, cyclosporine A; EBV, Epstein-Barr virus; F, female; HHV, human herpesvirus; HSV, herpes simplex virus; IVCY, intravenous cyclophosphamide; M, male; ND, not described; PSL, prednisolone; Tac, tacrolimus; VZV, varicella zoster virus; -, negative.
Fig 1.
The correlation of IFN-β1 and LINE-1 mRNA in the blood cells of patients with DM. The correlation of the relative messenger RNA level of IFN-β1 with LINE-1. The results were tested for statistical significance by the nonparametric Spearman test. DM, Dermatomyositis; mRNA, messenger RNA.
In the present study, we detected elevated type I interferon expression of mRNA, even in posttreatment samples from patients with DM. We confirmed a strong positive correlation between the expression of retrotransposons and IFN-β1. In addition, the dysregulation of methylation enzymes was observed in our study. Although LINE-1–induced type I interferon regulates the activity and propagation of LINE-1,5 the precise mechanisms underlying the mutual regulation between LINE-1 and type I interferon need further investigation.
Footnotes
Funding sources: Supported by Japan society for the promotion of the science (JSPS) Grants-in-Aid for Scientific Research (KAKENHI), Japan (grant no. JP19K08766).
Conflicts of interest: None disclosed.
IRB approval status: The ethics commission of Gunma University (reference number: 1349).
Reprints not available from the authors.
References
- 1.Wong D., Kea B., Pesich R. Interferon and biologic signatures in dermatomyositis skin: specificity and heterogeneity across diseases. PLoS One. 2012;7(1):e29161. doi: 10.1371/journal.pone.0029161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ray D.A., Batzer M.A. Reading TE leaves: new approaches to the identification of transposable element insertions. Genome Res. 2011;21:813–820. doi: 10.1101/gr.110528.110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Zhao K., Du J., Peng Y. LINE1 contributes to autoimmunity through both RIG-I- and MDA5-mediated RNA sensing pathways. J Autoimmun. 2018;90:105–115. doi: 10.1016/j.jaut.2018.02.007. [DOI] [PubMed] [Google Scholar]
- 4.Mavragani C.P., Sagalovskiy I., Guo Q. Expression of long interspersed nuclear element 1 retroelements and induction of type i interferon in patients with systemic autoimmune disease. Arthritis Rheumatol. 2016;68:2686–2696. doi: 10.1002/art.39795. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Yu Q., Carbone C.J., Katlinskaya Y.V. Type I interferon controls propagation of long interspersed element-1. J Biol Chem. 2015;290:10191–10199. doi: 10.1074/jbc.M114.612374. [DOI] [PMC free article] [PubMed] [Google Scholar]