Cholecalciferol Plus Calcium Suppresses Abnormal PBMC Reactivity in Patients with Multiple Sclerosis

Abstract

Context:

The active metabolite of vitamin D, 1,25-dihydroxyvitamin D [1,25(OH)₂D], is a potent modulator of immune cells in vitro.

Objective:

Our objective was to determine whether the sun-dependent nutrient, cholecalciferol, can alter disease-associated cellular immune abnormalities in patients with multiple sclerosis (MS).

Design:

This was an open-label, 12-month, randomized controlled trial.

Setting:

Patients with MS were recruited from the MS Clinic at St. Michael's Hospital, Toronto.

Patients:

Forty-nine patients were matched (for age, sex, disease duration, disease-modifying drug, and disability) and enrolled (treated n = 25; control n = 24). Four patients were lost to follow-up (n = 2 from each group).

Intervention:

Treated patients received increasing doses of cholecalciferol (4,000–40,000 IU/d) plus calcium (1200 mg/d), followed by equilibration to a moderate, physiological intake (10,000 IU/d). Control patients did not receive supplements.

Main Outcome Measures:

At enrollment and at 12 months, peripheral blood mononuclear cell (PBMC) proliferative responses to disease-associated, MS-relevant, and control antigens were measured, along with selected serum biochemical markers.

Results:

At 12 months, mean serum 25-hydroxyvitamin D [25(OH)D] concentrations were 83 ± 35 nmol/liter and 179 ± 76 nmol/liter in control and treated participants, respectively (paired t, P < 0.001). Serum 1,25(OH)₂D did not differ between baseline and 1 yr. In treated patients, 12-month PBMC proliferative responses to neuron antigens myelin basic protein and exon-2 were suppressed (P = 0.002). In controls, there were no significant changes in disease-associated PBMC responsiveness. There were no significant differences between groups in levels of selected biomarkers.

Interpretation:

MS-associated, abnormal T cell reactivities were suppressed in vivo by cholecalciferol at serum 25(OH)D concentrations higher than 100 nmol/liter.

Multiple sclerosis (MS) is a demyelinating, neuroinflammatory, and neurodegenerative disease of the central nervous system (CNS). Although its etiology remains enigmatic, there is longstanding consensus that abnormal pools of disease-associated, T-lineage effector cells fundamentally contribute to the CNS tissue lesions that characterize MS pathology (1–3). Most of the current, partially effective, immune therapies for MS are thought to target these abnormal T cell pools (4).

Vitamin D₃ is a seco-steroid, readily metabolized by the liver to 25-hydroxyvitamin D [25(OH)D], an inactive metabolite which reflects vitamin D₃ nutritional status. UV exposure to skin can produce physiological serum 25(OH)D levels as high as 225 nmol/liter (5). Various cells metabolize 25(OH)D into 1,25-dihydroxyvitamin D [1,25(OH)₂D], the signaling molecule that interacts with the vitamin D receptor (VDR) in target tissues. Vitamin D₃ may have therapeutic potential in MS because preclinical in vitro and in vivo animal experiments show that 1,25(OH)₂D affects cell proliferation and apoptosis, differentiation of immune cells, and modulation of immune responses. At the cellular level, T cells and antigen-presenting cells express the vitamin D receptor (VDR) (6, 7) and possess the capability to produce 1,25(OH)₂D from 25(OH)D via 1α-hydroxylase expression (8, 9). Dendritic cells (DCs) respond to 1,25(OH)₂D with differentiation arrest and attenuated antigen presentation (10), and such tolerogenic dendritic cells can induce regulatory T (Treg) cells (11), and 1,25(OH)₂D promotes a shift toward T helper type 2 (Th2) predominance in proinflammatory tissue lesions (12). In the mouse, 1,25(OH)₂D pretreatment prevents the development of experimental allergic encephalomyelitis, and after onset of active disease, 1,25(OH)₂D ameliorates symptoms (13).

Much of the current vitamin D research has focused on the final metabolite, 1,25(OH)₂D, signaling through VDR, with little evidence that the nutrient compound, cholecalciferol (vitamin D₃), has effects in vivo in animals or humans. However, indirect human data also imply desirable effects of cholecalciferol in MS. The number of gadolinium-enhanced magnetic resonance imaging lesions are higher in winter than in summer (14, 15), and higher serum 25(OH)D concentrations predict lower relapse rates in both adults (16–18) and children (19). Small clinical trials report reduced development of new gadolinium-enhancing lesions (20) and reduction of circulating TGF-β1 after cholecalciferol supplementation in patients with MS (21).

MS patients, both adults and children, show abnormal peripheral blood mononuclear cells (PBMC) proliferative responses to a range of self and environmental antigens (2, 22). In vitro treatment of T cells from MS patients with 1,25(OH)₂D reduces their proliferative responsiveness to MS-relevant autoantigens (9, 23, 24). To our knowledge, no clinical intervention with a nutrient, vitamin D₃ (cholecalciferol), has been shown to moderate these responses. In the present clinical trial of high-dose cholecalciferol in MS patients, we have previously shown successful increase in serum 25(OH)D concentrations and observed positive impact on clinical MS parameters (25). Here we report that our trial protocol significantly reduced responses of disease-associated PBMC responses, providing a mechanistic correlate for the clinical impact of serum 25(OH)D concentrations within the physiological range in MS patients.

Materials and Methods

Vitamin D study design

Forty-nine participants with clinically definite MS, as defined by McDonald criteria (26), were enrolled in an open-label phase I/II safety, dose-escalation study of cholecalciferol (vitamin D₃) supplementation plus calcium. Briefly, patients with MS were matched (based on age, sex, Expanded Disability Status Scale scores, disease duration, and disease-modifying treatment) and each member of a pair randomly allocated to treatment (n = 25) or control (n = 24) groups. Treated participants received oral cholecalciferol in doses escalating from 4,000–40,000 IU/d, averaging 14,000 IU/d over 1 yr, with a steady supplement of tricalcium phosphate (1200 mg/d elemental calcium). Control participants were permitted up to 4000 IU/of vitamin D₃ as per the clinic's standard recommendations, but supplements were not provided. Details of the study design have been discussed in further detail elsewhere (25). A timeline of study visits, treatment regimen, and biomarker measurements are presented in Fig. 1.

Fig. 1.

Study timeline. Supplementation schedule presented was for the treated patients only; control group participants received no intervention. Measurements were conducted as indicated at time points marked by (blood drop, indicating blood collection).

Approval was obtained from the institutional review boards of St. Michael's Hospital, Hospital for Sick Children, University of Toronto, and McGill University and was registered with www.ClinicalTrials.gov (NCT00644904).

Proliferation and cytokine assays

Heparinized blood samples were collected from all patients at two time points, baseline and 1 yr later at end of study. PBMC were purified on Ficoll-Hypaque gradients, washed, and cultured (10⁵ PBMC plus 10 U IL-2 per well) for 1 wk in protein-free Hybrimax 2997 medium (Sigma Chemical Co., St. Louis, MO), when proliferative responses to an array of test and control antigens (0.01–10 μg/well) were measured by [³H]thymidine incorporation. Replicate responses ([counts per minute (cpm)] were averaged and normalized to generate a stimulation index (SI = cpm test antigen/cpm cells alone). Composite scores were derived by adding the number of positive responses (SI ≥ 1.5) (2) to test antigens within antigen subsets (dietary, islet, or neuron test antigens) and for the total proliferation score (1, 27).

Our T cell activation array has been described (1, 22) and validated for studies of disease-associated T cell pools in patients with MS (2) or type 1 diabetes (27, 28). The array included positive controls [tetanus toxin (TT) and phytohemagglutinin (PHA)], negative controls [cytochrome c (Cyt-c) and actin], dietary (milk) antigens [casein, β-lactoglobulin (BLG), BSA, and BSA epitopes BSAp193 and BSAp147(ABBOS)], human islet antigens [Tep69, glutamic acid decarboxylase (GAD), GADp555, proinsulin (PI)], and neuron antigens [myelin basic protein (MBP), exon-2 of MBP (Ex-2), glial fibrillary acidic protein, and a glial antigen, S100β].

Culture supernatants were cryopreserved. Supernatant pools were prepared for each subset of antigens based on proliferation responses, including positive (PHA plus TT), negative (Cyt-c plus actin), dietary (BSA plus BSAp193), islet (PI), and neuronal (MBP plus Ex-2) antigens. Cytokine concentrations were measured in stored supernatant pools for all treatment samples (n = 23) and a random selection of control samples (n = 13). Concentrations of IL-1β, -2, -4, -5, -6, -10, -12p40, and -13, interferon (IFN)-γ, and TNF-α were measured simultaneously in serum samples (below) and culture supernatants, using multiplex Luminex X100 bead immunoassays (Luminex, Austin, TX) and calibrated reagents (Bio-Plex Precision Pro, Bio-Rad Lab, Hercules, CA) (29).

Serum samples

Sera from all participants were cryopreserved (−70 C) at baseline, twice at mid-study (at wk 7, corresponding to 2 wk of 4000 IU/d cholecalciferol, and at wk 29, corresponding to 6 wk of 40,000 IU/d cholecalciferol in the treatment group), and at 1 yr at protocol completion. Samples were analyzed together in the same run. Figure 1 depicts measurements with respect to the treatment group dosing schedule.

Other measurements

Replicate measurements of serum 1,25(OH)₂D concentrations employed an immunoassay-based kit (IDS Ltd., Tyne and Wear, UK). Serum 25(OH)D was measured by RIA according to the manufacturer's directions (DiaSorin, Stillwater, MN). High-sensitivity C-reactive protein and C telopeptide (CTx) were measured by an electrochemiluminescent immunoassay and an immunoturbidimetric assay, respectively, on a Roche Modular analyzer (Roche Diagnostic, Mannheim, Germany). Matrix metalloproteinase 9, tissue inhibitor of metalloproteinase 1, osteopontin, and bone-specific alkaline phosphatase (BAP) were measured by commercial ELISA kits (R&D Systems Inc., Minneapolis, MN), and kallikrein 6 was measured by an in-house ELISA developed in Dr. E. P. Diamandis's lab at Mount Sinai Hospital, Toronto, Canada (30).

Statistical analyses

Statistical analyses were performed with SPSS version 16.0 (SPSS Inc., Chicago, IL). For continuous variables, paired t testing was used to compare differences within groups (baseline vs. end of study) and changes in measures between groups (control vs. treated). Because results of proliferation assays and cytokine concentrations did not follow a normal distribution, the Wilcoxon signed ranks test was used for pair-based testing within- groups and for between-group comparisons. Bonferroni corrections were made for multiple comparisons. Mixed modeling procedures were used to compare markers that were measured at various time points throughout the study. Values are given in the text as mean ± sd.

Results

At baseline, the mean serum 25(OH)D concentration was 78 ± 27 nmol/liter, with no difference between groups randomized to receive cholecalciferol plus calcium or not. One year later, at study end, 25(OH)D concentrations were 179 ± 76 and 83 ± 27 nmol/liter for treated and control groups, respectively (paired t, P < 0.001). In treated patients, serum 25(OH)D concentrations were significantly higher with respect to baseline starting at wk 11 (after 6 wk at 10,000 IU/d of cholecalciferol) (P = 0.02) and at each time-consecutive measurement (paired t, P < 0.001). However, there were no significant changes in serum 25(OH)D concentrations in the treated group from wk 35 to wk 52, suggesting a steady-state vitamin D status when proliferation was evaluated. Serum 1,25(OH)₂D concentrations, measured at baseline and at 12 months, did not differ between baseline and end of study in treated (155 ± 59 and 184 ± 47 pmol/liter, respectively) or control participants (163 ± 45 and 165 ± 39 pmol/liter, respectively), nor did they differ between groups at either time point. No significant clinical or biochemical adverse events occurred. Clinical and biochemical responses have been presented in detail elsewhere (25).

We investigated the effects of the in vivo cholecalciferol supplementation on abnormal, disease-associated PBMC reactivities (2, 22) at baseline and 1 yr later (conclusion of this study), using a validated array of test antigens, including neuron, dietary, and islet antigens (1, 2, 22, 27). At baseline, proliferative responses did not differ between groups at baseline for any antigen, including the most MS-implicated responses to the BSA and MBP epitopes, which are also peptides that cause experimental allergic encephalomyelitis in mice (1, 22). During the ensuing year, no patient developed additional or enhanced responses to these antigens in either group, and PBMC responses to positive and negative controls remained similar across the trial. At the end of the trial, in the treated patients, several specific, previously MS disease-associated responses (38% of test antigens) were significantly reduced. These desirable changes were significant from comparison between baseline and 1 yr as well as from comparison of final data (i.e. difference over time) between treated and control groups (Table 1). As shown in Fig. 2, proliferative responses declined in treated patients at study completion for two of four neuron antigens (Ex-2 and MBP, P = 0.001), two of five milk antigens (BSA and BSAp193, P < 0.001), and one of four islet antigens (PI, P < 0.001). In contrast, the response to positive controls, polyclonal PHA and recall responses to tetanus toxoid, did not change.

Table 1.

PBMC proliferative responses (reported as stimulation index) to MS-associated antigens in treated and control groups

Antigenic stimulus	Baseline		End of study		Within-group comparison		Control vs. treated, end of study
	Control	Treated	Control	Treated	Control	Treated
Milk
CS	1.19 ± 0.40	1.11 ± 030	1.12 ± 0.18	1.05 ± 0.22	NS	NS	NS
BLG	1.45 ± 0.87	1.18 ± 0.22	1.21 ± 0.27	1.11 ± 0.29	NS	NS	NS
BSA	2.36 ± 0.87	2.75 ± 0.97	2.09 ± 0.59	1.65 ± 0.36	NS	<0.001	<0.001
BSAp193	2.21 ± 0.88	2.44 ± 0.78	1.84 ± 0.27	1.54 ± 0.36	NS	<0.001	<0.001
ABBOS	1.10 ± 0.17	1.13 ± 0.17	1.03 ± 0.14	1.06 ± 0.09	NS	NS	NS
Islet
Tep	1.06 ± 0.15	1.07 ± 0.14	1.03 ± 0.14	1.07 ± 0.08	NS	NS	NS
Gad	1.08 ± 0.16	1.12 ± 0.27	1.04 ± 0.16	1.02 ± 0.11	NS	NS	NS
Gadp555	1.05 ± 0.16	1.04 ± 0.12	1.04 ± 0.08	1.35 ± 1.81	NS	NS	NS
PI	1.86 ± 0.76	1.95 ± 0.60	1.41 ± 0.40	1.36 ± 0.23	NS	<0.001	0.005
Neural
Ex-2	1.82 ± 0.68	1.95 ± 0.62	1.94 ± 0.44	1.61 ± 0.36	NS	0.001	0.001
MBP	1.79 ± 0.57	2.22 ± 0.91	1.89 ± 0.42	1.67 ± 0.46	NS	<0.001	0.001
GFAP	1.11 ± 0.17	1.14 ± 0.18	1.32 ± 1.13	1.03 ± 0.10	NS	NS	<0.001
S100	1.12 ± 0.15	1.10 ± 0.18	0.99 ± 0.16	1.00 ± 0.12	NS	NS	<0.001
Positive controls
PHA	30.1 ± 16.0	27.2 ± 6.57	26.9 ± 9.42	26.9 ± 6.06	NS	NS	NS
TT	16.1 ± 6.47	15.2 ± 4.39	14.3 ± 4.55	14.8 ± 3.07	NS	NS	NS
Negative controls
Cyt-c	1.05 ± 0.15	1.05 ± 1.15	1.02 ± 0.09	1.01 ± 0.11	NS	NS	NS
Actin	1.04 ± 0.18	1.07 ± 0.09	1.00 ± 0.10	0.95 ± 0.10	NS	NS	NS
n	24	25	22	23	22	23	22

ABBOS, BSA epitopes BSAp193 and BSAp147; BLG, β -lactoglobulin; CS, casein; GFAP, glial fibrillary acidic protein; NS, not significant. Comparisons were by Wilcoxon signed-ranks test; Bonferonni corrected P value for multiple comparisons, P = 0.0038; within-group comparison = baseline vs. 1 yr; between-group comparison = change in proliferative responses compared between treated and controls.

Fig. 2.

Proliferation responses to antigen challenge at baseline and end of study in control and treated groups. Thymidine incorporation was measured in response to antigen stimulation. Boxes represent the central 50% of values, whiskers the highest/lowest values, and the line represents the median value. a, Significant difference within group (baseline vs. end of study) (Wilcoxon, P < 0.01); b, difference between groups (control vs. treated) in change from baseline (Wilcoxon, P < 0.01).

Composite scores for antigen subsets are presented in Table 2. At study conclusion, proliferative scores for all antigen subsets were significantly lower in the treated group, whereas control group responses did not differ from baseline. Between group comparisons (of change in proliferation over time) revealed significant differences for the neuronal antigen subset (Wilcoxon, P = 0.03) and total proliferation scores (Wilcoxon, P = 0.05). We examined intra-individual changes (baseline vs. end of study) in total proliferation with respect to changes in serum 25(OH)D concentrations (Fig. 3). Spearman regression revealed that as 25(OH)D levels increased, there were corresponding decreases in total proliferative scores (P = 0.005), and proliferation in response to dietary (P = 0.005) and neuron (P = 0.03) antigen subsets. Therefore, in vivo treatment with cholecalciferol and the associated increased serum 25(OH)D concentrations selectively affected disease-associated PBMC in the circulation of treated MS patients.

Table 2.

Comparison of mean composite proliferation scores within and between groups for responsiveness to antigen subsets and total proliferation

	Control		Treated		Between-groups comparison
	Baseline	One year	Baseline	One year
Dietary	2.0 ± 1.2	1.8 ± 0.7	2.0 ± 0.6	1.2 ± 1.0b	0.007
Islet	0.7 ± 0.7	0.4 ± 0.5	0.9 ± 0.5	0.4 ± 0.5b	0.005
Neuronal	1.3 ± 0.8	1.6 ± 0.8	1.6 ± 0.9	1.1 ± 1.0a	0.023
Total	4.0 ± 1.9	3.9 ± 1.4	4.5 ± 1.6	2.7 ± 1.9b	0.001

Comparisons within groups (baseline vs. 1 yr) by Wilcoxon ranked-signs test.

^aP < 0.05.

^bP < 0.005.

Fig. 3.

Proliferation scores decrease with increasing serum 25(OH)D concentrations. Change in total proliferation scores is plotted against the change in serum 25(OH)D concentration between baseline and 1 yr. Regression analysis revealed a significant negative correlation (Spearman ρ = −0.419; P = 0.005).

As previously published, clinical efficacy was suggested with vitamin D₃ treatment, and compared with controls, treated patients tended to have fewer relapses, and a greater proportion had a stable or improved Expanded Disability Status Scale scores at end of study (25).

Cytokines can attenuate or enhance a proinflammatory tissue lesion, and we considered whether the vitamin D₃ treatment affected cytokine profiles in the context of a Th1:Th2 paradigm. However, levels of IL-1β, -4, -5, -6, -10, -12p70, and -13, IFN-γ, and TNF-α in pooled culture supernatants (Supplemental Table 1, published on The Endocrine Society's Journals Online web site at http://jcem.endojournals.org) and in sera (Supplemental Table 2) were mostly near or below detection sensitivities. Candidate markers associated with CNS inflammation (matrix metalloproteinase-9, tissue inhibitor of metalloproteinase 1, kallikrein-6, C-reactive protein, and osteopontin) revealed no differences between control and treated groups (Supplemental Table 3). Markers of bone formation, BAP, and degradation, CTx, did not differ over time or between groups (Supplemental Table 3). As expected, BAP and CTx were highly correlated (r = 0.77; P < 0.001). Correlations were also found between osteopontin and both bone markers, BAP (r = 0.51; P = 0.003) and CTx (r = 0.48; P = 0.009).

We conclude that cholecalciferol treatment resulted in stable increases in serum 25(OH)D levels that were associated with selective, significant attenuation of MS-associated T cell pools thought to drive progression of proinflammatory lesions in the CNS. The treatment did not measurably affect polyclonal or cognate recall T cell reactivities in these patients.

Discussion

Cross-sectional data suggest that in patients with MS, serum 25(OH)D concentrations are lower during relapse in comparison with remission (9, 18, 31) and correlate inversely with disease severity (17, 32–34). Improvement of MS clinical outcomes in small, uncontrolled clinical trials of vitamin D supplementation show lower rates of exacerbation (35) and reduced disease activity on magnetic resonance imaging (20). Furthermore, elevated steady-state 25(OH)D levels after oral administration of cholecalciferol had desirable clinical effects in the present cohort, as we previously reported (25). Biological plausibility for a therapeutic benefit is provided by in vitro and animal data that demonstrate immunomodulatory effects of 1,25(OH)₂D with a shift toward an antiinflammatory profile and an increase in Treg function (36). In support, MS patients supplemented with 20,000 IU/d vitamin D₃ found an increased proportion of IL-10-producing CD4⁺ T cells and decreased IFN-γ- and IL-4-producing CD4⁺ T cells (37). In addition, an improvement of Treg-suppressive function was suggested (37). Current therapies for MS are immunosuppressive and are thought to target immune cell function (e.g. IFN-β), or trafficking (e.g. natalizumab), to reduce immune activity in the CNS. However, these therapies are costly, can have severe side effects, and are only partially effective. Vitamin D supplementation offers a safe and inexpensive option that may be beneficial alone or as an add-on to therapy in patients with MS.

Here we analyzed MS disease-associated, T cell responsiveness to an array of test antigens previously validated in blinded MS and diabetes studies (1, 22). The rise of mean serum 25(OH)D concentrations in the treated group, from 78 nmol/liter at baseline to 179 nmol/liter, significantly reduced abnormal PBMC responsiveness to disease-associated antigens, whereas polyclonal and recall T cell responses were unaffected. In the untreated control group, PBMC responsiveness to those same antigens remained unchanged over this period. Suppression of antigen responsivity was persistent for 7 d in the absence of cells seeing vitamin D metabolites, suggesting a shift in cell populations or alterations in cell behavior, perhaps via epigenetic changes induced by the signaling molecule 1,25(OH)₂D. Furthermore, reductions in proliferation as measured by composite scores indicate a dose-dependent relationship. To our knowledge, this is the first indication that a nontoxic intervention can selectively reduce disease-associated T cell pools in humans with tissue-selective autoimmune disease. Moreover, although it was a small trial, this is the first study to link these pools with clinical course.

Serum 25(OH)D concentrations correlate inversely with the percentage of Treg cells in patients with MS (38). Addition of 1,25(OH)₂D in vitro improves Treg-mediated suppression of CD4⁺ T cell proliferation (9, 23). However, it has remained uncertain whether provision of circulating 25(OH)D via supplementation with cholecalciferol will promote local production and use of the signaling metabolite, 1,25(OH)₂D, by cells of the immune system. In fact, the ultimate vitamin D effector mechanisms remain unknown. In support, circulating 1,25(OH)₂D concentrations were not affected by increased 25(OH)D levels, and this implies local use by immune cells.

Although we measured several serum markers related to compromised blood-brain barrier integrity in patients with MS and a general marker of inflammation, C-reactive protein, we did not detect any differences between treated and control groups. We attribute the lack of changes in these markers to high baseline vitamin D status. At baseline, serum 25(OH)D concentrations were already at what is widely regarded as optimal (>75 nmol/liter). The inverse relationship previously shown between 25(OH)D with C-reactive protein and cytokine concentrations have occurred at lower 25(OH)D baseline concentrations (21, 38–40). However, normal or optimal levels of vitamin D remain a moving target and ultimately will require formal (and larger) trials such as ours in relevant target populations of healthy as well as affected subjects from different environments. We had expected to see a treatment effect on markers of bone turnover, either an increase in formation or a decrease in resorption, but we did not detect any influence of increased serum 25(OH)D concentrations on either marker of bone metabolism. Bone density studies demonstrate an optimal 25(OH)D concentration of 75 nmol/liter, meaning that above this value, the influence of 25(OH)D on bone turnover is negligible (41). We also attribute this lack of change to high baseline serum 25(OH)D concentrations. Furthermore, this indicates that high serum 25(OH)D did not adversely affect bone deposition by causing excessive bone resorption, which some fear. Correlations found between osteopontin and bone turnover markers, BAP and CTx, support recent findings by Vogt et al. (42).

This was a randomized, interventional, clinical trial designed with objective outcomes that cannot be attributed to a placebo effect. The final serum 25(OH)D concentrations in the treated patients were well within the natural physiological range achievable through environmental exposure; in fact, no adverse events occurred. Because of the 2- to 3-month half-life of 25(OH)D (5), serum 25(OH)D concentrations were stable during the months leading up to the time of final sampling for the proliferation assays (25). The number of subjects in this randomized controlled trial was limited because of its exploratory character, yet we were surprised by the extent and the strength of statistical confidence obtained for the changes elicited for key, disease-associated T cell pools.

Our observations of 25(OH)D-induced amelioration of abnormal MS-associated PBMC reactivities provide a mechanistic explanation of recent evidence that higher serum 25(OH)D concentrations, or vitamin D₃ supplementation, are associated with lower rates of relapse (9, 16, 18, 43) and to fewer gadolinium-enhancing lesions (20). We attempted to characterize this response in the context of a Th1:Th2 paradigm by measuring phenotypic cytokine concentrations, but we did not characterize cell populations within the PBMC preparations. Since the design of this trial, the importance of a role for abnormal Treg function, pathogenic Th17 cells (3, 37, 44), and the role of antigen-presenting cells in MS pathology has been recognized (45); thus, the paradigm we were working with was likely too simplified. However, VDR and 1α-hydroxylase expression are up-regulated by T cell activation (46). In the absence of cytokine changes, conversion of 25(OH)D to the signaling molecule, 1,25(OH)₂D, would be the simplest explanation for reduced responsiveness of PBMC observed in the cholecalciferol-treated patients in the present study.