Transcriptome Analysis in Women with Rheumatoid Arthritis Who Improve or Worsen during Pregnancy

Abstract

Background/Purpose: Gene expression changes induced by pregnancy in women with rheumatoid arthritis (RA) and healthy women have not been examined. The few studies previously conducted did not have pre-pregnancy samples available as baseline. We have established a cohort of RA and healthy women followed prospectively from pre-pregnancy. In this study, we aimed to identify pregnancy-induced changes in gene expression among women with RA and healthy women, and to assess how those changes may differ between RA women who improve or worsen during pregnancy. Methods: Clinical data and samples collected from a subset of 11 women with RA and 5 healthy women from our cohort before pregnancy (T0) and at the third trimester (T3) were analyzed. Disease activity scores were used to determine whether the RA women improved or worsened during pregnancy. Global gene expression profiles were generated by RNA sequencing (RNA-seq). The raw RNA-seq reads were pseudo-aligned to the reference transcriptome and expression levels were estimated with kallisto. Differential expression analysis of normalized expression levels was performed using edgeR to identify genes differentially expressed within each group of women (T3 vs T0), using a foldchange cut-off of 2 and a significance threshold of q<0.05 (FDR-adjusted). Functional enrichment analysis was performed using WebGestalt. Results: Of the 11 women with RA, 8 showed an improvement in disease activity by T3 (RA_(improved)), while 3 worsened (RA_(worsened)). In the RA_(improved) group, a total of 161 genes were differentially expressed (DE) between T3 and T0. These included several genes whose expression have previously been associated with RA (e.g. S100A12, SLC14A1) as well as genes involved in the innate immune system (e.g. type I interferon-inducible genes). The majority of these genes (108 of 161) were also DE among healthy women. Of interest, most genes (30 of 31) that were significantly DE in both of the RA groups were also DE among healthy women (e.g. α-defensin genes). There were also differences between the RA_(improved) and RA_(worsened) groups. A set of IFN-inducible genes was over-expressed at T3 (vs T0) in the RA_(improved) but not the RA_(worsened) women. Additionally, some interesting candidate genes whose expression have previously been associated with RA (e.g. MMP9, PADI4 and PGLYRP1) were over-expressed at T3 (vs. T0) among RA_(worsened) but not among RA_(improved) women. Conclusion: Pregnancy-induced gene expression changes common between RA women who improved and those who worsened appeared to be normal pregnancy-related changes that were also observed among healthy women. Other genes that demonstrated different patterns of expression between the two RA groups are potential candidates that could be involved in the natural pregnancy-induced amelioration of RA.

Additional details