Background/Purpose:

Rheumatoid arthritis (RA) disease activity can often quiesce during pregnancy. Nevertheless, most women will experience a disease flare postpartum (PP). We hypothesized that changes in blood immune cell gene expression patterns during healthy pregnancy, which promote fetal tolerance in order to protect against fetal loss, will also be relevant in regulating RA disease activity.  

Methods:

Whole-blood gene expression profiles (15,050 expressed; p<0.05) from 19 pregnant RA subjects and 13 pregnant age-matched healthy controls (CON) were assayed at first trimester of pregnancy (T1), T2, T3, and PP using Illumina HumanHT-12 v3 and v4 BeadChip arrays. At identical time points, disease activity in RA women was assessed using the DAS28-CRP. A Generalized Estimating Equations (GEE) approach was applied to model gene expression changes during pregnancy (T1-T3) and between pregnancy and PP, to take into account the correlation within samples. A version effect (v3 vs v4) was also included in the model.

Results:

Within the CON samples, 581 blood cell genes demonstrated either a >20% change in mRNA expression between T1—>T2 or T2—>T3 of pregnancy, or a >20% change in expression between the average of T1—T3 and the PP period (BH q<0.05). Many of these genes mapped to Ingenuity Pathways associated with (p<0.01) hematopoietic system development and function (n=149 molecules), cell death (n=185), and immunological diseases (n=127). These 581 "pregnancy-related" genes were then selected for further study in the pregnant RA cohort. Of the 581 genes, 74 were found to be significantly associated with the DAS28-CRP score after correcting for multiple hypothesis tests (Bonferroni correction, p<8.6x10^-5), with 11 of these demonstrating a >20% change in mRNA expression for every 1 unit change in DAS28-CRP score either during pregnancy or in the PP period. For the 50 pregnancy-related genes whose expression fell PP in association with rising DAS28-CRP, 7 genes (CD247, CD5, CD6, FLT3LG, IL7R, RBM38, and TUBB) were functionally associated with autoimmune diseases by Ingenuity Pathways Analysis; likewise, of the 24 genes whose expression directly correlated with DAS28-CRP PP, 5 genes (IL1B, MYD88, NCF4, PFKFB3, and SLC22A4) were associated with autoimmune diseases (p=0.0015).

Conclusion:

Our experiments identify 74 RA disease activity-correlated blood cell genes whose expression is modulated during normal pregnancy. Insofar as these genes reflect alterations in immune cell reactivity normally associated with pregnancy and promotion of fetal tolerance, they may represent a window into processes and pathways with a capacity to regulate disease activity in RA patients. Pregnancy-modulated genes that fluctuate with RA disease activity may also have utility as disease activity biomarkers.