Alterations of NK Cell Phenotype During Pregnancy in Multiple Sclerosis

Abstract

In multiple sclerosis (MS), relapse rate is decreased by 70-80% in the third trimester of pregnancy. However, the underlying mechanisms driving this effect are poorly understood. Evidence suggests that CD56^bright NK cell frequencies increase during pregnancy. Here, we analyze pregnancy-related NK cell shifts in a large longitudinal cohort of pregnant women with and without MS, and provide in-depth phenotyping of NK cells. In healthy pregnancy and pregnancy in MS, peripheral blood NK cells showed significant frequency shifts, notably an increase of CD56^bright NK cells and a decrease of CD56^dim NK cells toward the third trimester, indicating a general rather than an MS-specific phenomenon of pregnancy. Additional follow-ups in women with MS showed a reversal of NK cell changes postpartum. Moreover, high-dimensional profiling revealed a specific CD56^bright subset with receptor expression related to cytotoxicity and cell activity (e.g., CD16⁺ NKp46^high NKG2D^high NKG2A^high phenotype) that may drive the expansion of CD56^bright NK cells during pregnancy in MS. Our data confirm that pregnancy promotes pronounced shifts of NK cells toward the regulatory CD56^bright population. Although exploratory results on in-depth CD56^bright phenotype need to be confirmed in larger studies, our findings suggest an increased regulatory NK activity, thereby potentially contributing to disease amelioration of MS during pregnancy.

Keywords: FlowSOM clustering; immune profiling; immune tolerance; multiple sclerosis; natural killer cells; pregnancy.

Figure 1

CD56bright NK cells increase throughout pregnancy. Frequency of CD56^bright NK cells (A) and CD56^dim NK cells (B) in healthy pregnant women (HC, gray) and pregnant MS patients (MS, blue, Hamburg cohort) shown for each trimester of gestation. Boxplots depict median and inter-quartile range, overlaid with datapoints of individuals. Statistical analysis performed by Wilcoxon paired test between trimesters 1 and 3.

Figure 2

Pregnancy-induced shifts of NK cells throughout and after gestation in Berlin MS cohort. Frequency of CD56^bright NK cells (A) and CD56^dim NK cells (B) in pregnant MS patients (MS, Berlin cohort) shown for each trimester of gestation and 1 year postpartum. Boxplots depict median and inter-quartile range, overlaid with datapoints of individuals. Gray boxes represent pregnancy. Statistics performed by Wilcoxon paired test between trimesters 1 and 3, as well as trimester 3 and 3 months postpartum.

Figure 3

Clusters within CD56^bright and CD56^dim NK cell populations display dynamic shifts in MS pregnancy. In a detailed phenotypic analysis during and after pregnancy in MS patients (Berlin cohort), FlowSOM identified 20 distinct clusters among pre-gated Lin^- CD56⁺ NK cells. Clusters are numbered and indicated by shades of grey and differently coloured backgrounds (for details of FlowSOM see Supplementary Figure 2 and Methods) (A). Surface marker expression of each cluster is shown in the heatmap. For all markers except CD56, dark blue indicates negative expression and dark red indicates highly positive expression. For CD56, dark blue to green indicates dim expression (as all analyzed cells are CD56⁺) and dark red indicates bright expression (B). Clusters with frequency changes throughout pregnancy and the postpartum period (six clusters with smallest p-values shown). Cluster frequency referred to as percentage of Lin^- CD56⁺ NK cells (C). Cluster 3 and 7 contain CD56^bright and CD56^dim NK cells as indicated by a medium CD56 expression in the heatmap (yellow to orange). Mixed composition of cluster 3 and 7 has been additionally confirmed by manual gating. Boxplots depict median and inter-quartile range, overlaid with datapoints of individuals. The gray box represents pregnancy. Statistical analysis performed by Wilcoxon paired test between trimesters 1 and 3, as well as trimester 3 and 3 months postpartum.