Vitamin D treatment modulates immune activation in cystic fibrosis

Abstract

Persistent inflammatory response in cystic fibrosis (CF) airways is believed to play a central role in the progression of lung damage. Anti-inflammatory treatment may slow lung disease progression, but adverse side effects have limited its use. Vitamin D has immunoregulatory properties. We randomized 16 CF patients to receive vitamin D2, vitamin D3 or to serve as controls, and investigated the effect of vitamin D supplementation on soluble immunological parameters, myeloid dendritic cells (mDCs) and T cell activation. Three months of vitamin D treatment were followed by two washout months. Vitamin D status at baseline was correlated negatively with haptoglobin, erythrocyte sedimentation rate and immunoglobulin A concentration. Total vitamin D dose per kg bodyweight correlated with the down-modulation of the co-stimulatory receptor CD86 on mDCs. Vitamin D treatment was associated with reduced CD279 (PD-1) expression on CD4+ and CD8+ T cells, as well as decreased frequency of CD8+ T cells co-expressing the activation markers CD38 and human leucocyte antigen D-related (HLA-DR) in a dose-dependent manner. There was a trend towards decreased mucosal-associated invariant T cells (MAIT) cell frequency in patients receiving vitamin D and free serum 25-hydroxyvitamin D (free-s25OHD) correlated positively with CD38 expression by these cells. At the end of intervention, the change in free-s25OHD was correlated negatively with the change in CD279 (PD-1) expression on MAIT cells. Collectively, these data indicate that vitamin D has robust pleiotropic immunomodulatory effects in CF. Larger studies are needed to explore the immunomodulatory treatment potential of vitamin D in CF in more detail.

Trial registration: ClinicalTrials.gov NCT01321905.

Keywords: T cells; cystic fibrosis; immunity; immunoglobulins; vitamin D.

© 2017 British Society for Immunology.

Figures

Figure 1

Haptoglobin correlates inversely with vitamin D status at baseline and change in vitamin D during treatment. Lipopolysaccharide (LPS) and soluble triggering receptor expressed on myeloid cells‐1 (sTREM‐1) correlate with change in vitamin D levels during treatment. Free serum 25‐hydroxyvitamin D (free‐s25OHD) was calculated and haptoglobin, LPS and sTREM‐1 analysed as described in Subjects and methods. Haptoglobin was correlated negatively with free‐s25OHD at baseline (a). Changes in haptoglobin concentration were correlated negatively with changes in free‐s25OHD at the end of intervention (b), and with changes in tot‐s25OHD at the end of the study (c). Change from baseline in tot‐s25OHD versus change from baseline in plasma LPS levels at the end of supplementation in all patients (d). Change from baseline in free‐s25OHD versus change from baseline in plasma sTREM‐1 levels at the end of supplementation in all patients (e). Change from baseline in LPS versus change from baseline in sTREM‐1 at the end of supplementation in all patients (f). Change from baseline in free‐s25OHD versus change from baseline in circulating monocyte count at the end of supplementation in all patients (g). P < 0·05 in all. Results are plotted as raw data from all subjects participating in the study irrespective of treatment arm (a–g). The number of patients presented in the individual graphs varies due to a few missing samples (see Table 2).

Figure 2

Vitamin D levels correlate with erythrocyte sedimentation rate (ESR). Free serum 25‐hydroxyvitamin D (free‐s25OHD) was calculated and ESR was analysed as described in Subjects and methods. At baseline, tot‐s25OHD (a) and free‐s25OHD (b) were correlated negatively with ESR. At the end of supplementation, the change in ESR was correlated inversely with the change in tot‐s25OHD from baseline (c). All Ps < 0·05. Results are plotted as raw data from all subjects participating in the study irrespective of treatment arm (a–c). The number of patients presented in the individual graphs varies due to a few missing samples (see Table 2).

Figure 3

Vitamin D levels correlate with immunoglobulin concentrations. Free serum 25‐hydroxyvitamin D (free‐s25OHD) was calculated and immunoglobulin (Ig)A, IgM, IgG and IgE were analysed as described in Subjects and methods. Sun exposure was quantified as described previously 26. Total‐s25OHD (a) and free‐s25OHD (b) were correlated negatively with serum total IgA levels at baseline. Changes in free‐s25OHD were correlated negatively with changes in serum total IgM at 1 week of supplementation (c). Changes in free‐s25OHD correlated with changes in plasma total IgE at 4 weeks of supplementation (d). At the end of supplementation, there was an inverse correlation between both sun exposure (e) and total vitamin D dose ingested (f) and plasma total IgE. P < 0·05 in all, unless stated otherwise. Results are plotted as raw data from all subjects participating in the study irrespective of treatment arm (a–f). The number of patients presented in the individual graphs varies due to a few missing samples (see Table 2).

Figure 4

Total vitamin D dose per kg bodyweight correlates with the down‐modulation of the co‐stimulatory molecule CD86 on myeloid dendritic cells (mDCs). The total vitamin D dose per kg bodyweight correlated with the down‐modulation of CD86 on mDCs at the end of the study. Results are plotted as raw data from all subjects participating in the study irrespective of treatment arm. The number of observations presented in the graph is smaller than the total number of patients due to a few missing samples (see Table 3).

Figure 5

Vitamin D treatment reduces CD279 [programmed death 1 (PD‐1)] expression on CD8+ T cells as well as frequency of CD8+ T cells co‐expressing CD38 and human leucocyte antigen D‐related (HLA‐DR). The combined group of patients receiving D2 or D3 reduced their HLA‐DR (a) and CD279 (PD‐1) (b) expression on CD8+ T cells at the end of supplementation. The frequency of CD8+ T cells co‐expressing CD38 and HLA‐DR was also decreased at the end of supplementation (c). HLA‐DR expression on CD8+ T cells correlated with the total vitamin D dose per kg bodyweight at the end of intervention (d), and the down‐regulation of the HLA‐DR expression on CD8+ T cells correlated with change in total‐serum 25‐hydroxyvitamin D (s25OHD) at the end of the study (e). Patients receiving D3 displayed decreased CD279 (PD‐1) expression on CD8+ T cells at the end of supplementation (f). P < 0·05 in all. Results are plotted as raw data from the subjects in the pooled vitamin D2 and vitamin D3 treatment group (a–c), raw data from all subjects participating in the study irrespective of treatment arm (d,e) and as raw data from subjects in the vitamin D3 treatment group (f). The number of patients presented in the individual graphs varies due to a few missing samples (see Table 3).

Figure 6

Vitamin D supplementation decreases the frequency of CD4+ T cells expressing CD279 [programmed death 1 (PD‐1)] in a dose‐dependent manner. At baseline, free‐serum 25‐hydroxyvitamin D (s25OHD) was correlated negatively with the expression of human leucocyte antigen D‐related (HLA‐DR) by CD4+ T cells (a) and with the frequency of CD4+ T cells co‐expressing CD38 and HLA‐DR (b). At the end of the study, the combined group of patients receiving vitamin D2 or D3 experienced a decreased frequency of CD4+ T cells expressing CD279 (PD‐1) (c). At this time‐point, total vitamin D dose received was correlated negatively with change in CD279 (PD‐1) expression by CD4+ T cells (d) and correlated positively with change in CD4/CD8 ratio (e). P < 0·05 in all. Results are plotted as raw data from all subjects participating in the study irrespective of treatment arm (a,b,d,e) and as raw data from the subjects in the pooled vitamin D2 and vitamin D3 treatment group (c). The number of patients presented in the individual graphs varies due to a few missing samples (see Table 3).

Figure 7

There is a trend towards decreased mucosal‐associated invariant T cell (MAIT) frequency in patients receiving vitamin D and free‐serum 25‐hydroxyvitamin D (s25OHD) correlates positively with CD38 expression by these cells. At baseline, free‐s25OHD was correlated positively with CD38 expression, P = 0·008, (a) and tended to correlate with human leucocyte antigen D‐related (HLA‐DR) expression, P = 0·055, (b) by MAIT cells. MAIT cells were identified by the combined expression of CD161 and Vα7.2 by T cells (c). There was a trend towards decreased MAIT cell frequency at the end of intervention compared with baseline in the combined group of patients receiving vitamin D2 or D3, P = 0·0506 (d). At the end of intervention, the change in free‐s25OHD was correlated negatively with the change in CD279 (programmed death 1) expression on MAIT cells (e). P < 0·05 in all, unless stated otherwise. Results are plotted as raw data from all subjects participating in the study irrespective of treatment arm (a,b,e) and as raw data from the subjects in the pooled vitamin D2 and vitamin D3 treatment group (d). The number of patients presented in the individual graphs varies due to a few missing samples (see Table 3).