Seropositivity-Dependent Association between LINE-1 Methylation and Response to Methotrexate Therapy in Early Rheumatoid Arthritis Patients

Abstract

Background: Methotrexate (MTX) is considered the first choice among disease-modifying anti-rheumatic drugs (DMARDs) for rheumatoid arthritis (RA) treatment. However, response to it varies as approximately 40% of the patients do not respond and would lose the most effective period of treatment time. Therefore, having a predictive biomarker before starting MTX treatment is of utmost importance. Methylation of long interspersed nucleotide element-1 (LINE-1) is generally considered a surrogate marker for global genomic methylation, which has been reported to associate with disease activity after MTX therapy.

Methods: We performed a prospective study on 273 naïve early RA (ERA) patients who were treated with MTX, followed up to 12 months, and classified according to their therapy response. The baseline LINE-1 methylation levels in peripheral blood mononuclear cells (PBMC) of cases were assessed by bisulfite pyrosequencing.

Results: Baseline LINE-1 methylation level per se turned out not to predict the response to the therapy, nor did age, sex, body mass index, or smoking status. However, if cases were stratified according to positivity to rheumatoid factor (RF) and anti-citrullinated protein antibody (ACPA) or seronegativity, we observed an opposite association between baseline LINE-1 methylation levels and optimal response to MTX therapy among responders. The best response to MTX therapy was associated with hypermethylated LINE-1 among double-positive ERA cases (p-value: 0.002) and with hypomethylated LINE-1 in seronegative ERA patients (p-value: 0.01).

Conclusion: The LINE-1 methylation level in PBMCs of naïve ERA cases associates with the degree of response to MTX therapy in an opposite way depending on the presence of RF and ACPA antibodies. Our results suggest LINE-1 methylation level as a new epigenetic biomarker for predicting the degree of response to MTX in both double-positive and seronegative ERA patients.

Keywords: DNA methylation; LINE-1; anti-citrullinated protein antibody; biomarkers; methotrexate; pharmacoepigenetics; rheumatoid arthritis; rheumatoid factor.

Figure 1

Average LINE-1 methylation level in relation to the response to MTX therapy. Data are presented as the average methylation percentage of five CpG sites. NR, No Response; MR, Moderate Response; GR, Good Response.

Figure 2

Association of LINE-1 methylation level with response to the therapy stratified by RF positivity. Data are presented as the average methylation percentage of five CpG sites. NR, No Response; MR, Moderate Response; GR, Good Response; * significant nominal p-value.

Figure 3

Association of LINE-1 methylation level with response to the therapy stratified by ACPA positivity. Data are presented as the average methylation percentage of five CpG sites. NR, No Response; MR, Moderate Response; GR, Good Response; * significant nominal p-value.

Figure 4

Effect of seropositivity on the association between LINE-1 methylation level and the magnitude of response to MTX therapy. Data are presented as the average methylation percentage of five CpG sites. FR− & ACPA−: seronegative; FR− or ACPA−: single positive; FR+ & ACPA+: double-positive; MR, Moderate Response; GR, Good Response; * significant nominal p-value.

Figure 5

CpG island region of the human LINE-1 transposon (Locus X58075). The five CpG sites highlighted in yellow were investigated in the present study. The twelve CpG sites shown in green were studied by Gosselt et al. [31].