Personal view in Lancet Respiratory Medicine:
Vitamin D and COPD: hype or reality.
Wim Janssens*, Marc Decramer*, Chantal Mathieu£, Hannelie Korf£
*Laboratory of Pneumology, Department of Experimental and Clinical Medicine, KULeuven, Leuven, Belgium
£Laboratory of Clinical and Experimental Endocrinology, Department of Experimental and Clinical Medicine, KULeuven, Leuven, Belgium
WJ, CM and HK are supported by the Flemish Research Funds (FWO Vlaanderen)
Address of correspondence:
Wim Janssens, MD, PhD
Respiratory Medicine, UZ Leuven, Herestraat 49, 3000 Leuven, Belgium
Tel: +32 16 34 68 00
Fax: + 32 16 34 68 03
Email:
Keywords: vitamin D, COPD, VDR, supplementation, exacerbation
Search strategy and selection criteria:
PubMed was searched for English reports before March 1, 2013 with the terms ‘vitamin D’ in combination with ‘smoking’, ‘COPD’, ‘airways disease’, ‘lung function’ or ‘exacerbations’. References from identified papers were also searched for relevant articles.
Abstract
Abundant evidence from laboratory studies is supporting an important role for vitamin D in the innate and adaptive immune system. In humans, observational studies have associated vitamin D deficiency to an increased risk for different inflammatory, infectious and auto-immune diseases. With regard to chronic obstructive pulmonary disease, conflicting data have been reported. Most epidemiological studies have been limited by their design so that larger longitudinal studies of population based samples and of cohorts with COPD are warranted. An alternative explanation for the discordant observations in COPD may be found in the complexity of the intracellular vitamin signaling pathway, which is not reflected in systemic levels of precursor 25-hydroxyvitamin D. For COPD in particular, we speculate that local down-regulation of vitamin D signaling from and beyond the receptor may clarify why pro-inflammatory processes in the airways are not or insufficiently countered by vitamin D dependent control mechanisms. In a disease already characterized by glucocorticoid resistance, the potential (re)activation of an intrinsic comprehensive immune control system should attract more attention in order to design appropriate interventions with promising therapeutic potential.
1. Introduction
Increasing but conflicting evidence is currently arising on the role of the vitamin D pathway in chronic obstructive pulmonary disease (COPD). Although several studies have highlighted associations between vitamin D deficiency, pulmonary function and disease characteristics, others have failed to confirm these findings. A recent placebo controlled intervention trial did not support clinical benefits of vitamin D supplementation on exacerbations in a severe COPD population.1 Moreover, the claim that vitamin D deficiency would be rather consequence than cause of the disease, has dampened the enthusiasm for new clinical studies on vitamin D in COPD.2;3 This is surprising since most mechanistic studies in cell culture systems and animal models clearly point to important anti-inflammatory, anti-infectious and anti-proliferative effects of vitamin D.4 Similarly, large cross-sectional and prospective studies have shown associations between low levels of vitamin D or polymorphisms in the vitamin D receptor with increased risk for asthma and poor asthma control, suggesting that the vitamin D pathway has anti-inflammatory effects in the human respiratory system.5-7 Here, we aim to review the current literature on vitamin D and COPD with a focus on epidemiological findings, underlying mechanisms, missing links and future challenges. We are aware of the role vitamin D may have in common comorbidities of COPD such as osteoporosis, muscle dysfunction, cardiovascular disease, diabetes and cancer8, but the focus of this review will be strictly on the respiratory involvement.
2. Vitamin D deficiency and COPD: epidemiologic associations
The majority of vitamin D3 (cholecalciferol) is synthesized in the skin by solar UVB exposure but an alternative source is derived from dietary intake, mainly fish oils, dairy products, fortified grains or oral supplements. Cholecalciferol is hydroxylated in the liver by 25-hydroxylase (CYP2R1) into 25-hydroxyvitamin D (25(OH)D3) that serves as a substrate for generation of the bioactive ligand, 1,25-dihydroxyvitaminD3 (1,25(OH)2D3), which is synthesized in the kidney or locally in immune and epithelial cells by 1α-hydroxylase (CYP27B1). Circulating 25(OH)D3 is mainly bound to its carrier protein, vitamin D binding protein (DBP), and because of its relative stability and long half-life, is often used to determine vitamin D status.9 Based on skeletal benefits, a recent report of the Institute of Medicine (IOM) has defined the cut-off for deficiency at serum 25(OH)D3 levels < 20 ng/ml, as there was no conclusive evidence for extra-skeletal benefits nor for the need of higher serum levels in humans.10
Low 25(OH)D3 levels have been associated with a variety of diseases including infections, auto-immune disease, cardiovascular diseases, diabetes and cancer.9 When focusing on pulmonary function, Black and colleagues were the first to describe associations with 25(OH)D3 in 14,091 subjects from the Third National Health and Nutrition Survey (NHANESIII).11 After multiple adjustments for confounders including intake of supplements and outdoors activity, the mean FEV1 was 106 mL (SE, 24 mL), and the mean FVC was 142 mL (SE, 29 mL) greater for the highest quintile of serum 25(OH)D3 levels compared with the lowest quintile (p < 0·0001). However, the authors did not find any relationship with FEV1/FVC ratio, indicating that the cross-sectional observation could also be determined by impaired lung growth or reduced respiratory muscle force rather than lung function decline per se. Moreover, a more recent analysis in 2,997 adults of the Hertfordshire Cohort found positive associations between lung function variables (FEV1, FVC, FEV1/FVC ratio) and vitamin D intake but failed to confirm the findings of NHANESIII with serum 25(OH)D3 levels.3 It was therefore concluded that vitamin D status did not influence adult lung function but that other (dietary) factors closely linked to vitamin D, could presumably explain the positive associations.
To address causality, Lange and colleagues performed a longitudinal study to investigate vitamin D status and lung function decline in 626 healthy community-based elderly men over 20 years. Multivariable-adjusted analysis revealed that vitamin D status alone was not associated with spirometric measurements but that a significant vitamin D status-by-smoking interaction for all measures of lung function was present. In addition, faster rates of lung function decline per pack-year of smoking were found in subjects with vitamin D deficiency compared with non-deficient subjects.12 The study indicated that in the general population smoking individuals are particularly at risk for developing COPD if being deficient for vitamin D. As a Belgian Cohort study showed that vitamin D deficiency was more prevalent in mild COPD (39%) compared to smoking controls (31%), these latter findings indirectly supported such a hypothesis.13 However, the study of Lange et al. did not adjust for physical activity nor for sun exposure, which are main determinants of vitamin D status. Physical inactivity is an independent risk for lung function decline and COPD onset14, and additional research is needed to unravel how these conditions interact with respect to COPD.
Several cross-sectional studies have described poor vitamin D status in patients with established COPD.13;15;16 For instance, in the Belgian cohort of 414 individuals not taking vitamin D supplements, mean 25(OH)D3 levels were significantly lower in the COPD group compared to the smoking controls (19·9 ± 8·2 vs. 24·6 ± 8·7 ng/ml) with respectively 39, 47, 60 and 77% of the GOLD stage I, II, III and IV exhibiting vitamin D deficiency.13 Multivariate analysis retained FEV1, obesity, smoking and winter season as main determinants of low vitamin D status. The Belgian study also confirmed previous reports that polymorphisms in genes encoding for DBP (GC) were independent determinants of 25(OH)D3 levels.17 Interestingly, genotypes that associated with lower 25(OH)D3 levels also depicted an increased risk of COPD in the Belgian study. In other studies the same genetic variants of GC also associated with COPD and diffuse pan-bronchiolitis.18;19 Although DBP has pleiotropic functions and may activate macrophages in COPD20, the association of the same risk variants with lower 25(OH)D3 levels may offer an alternative explanation for the underlying biological mechanism. Overall, these cross-sectional studies highlight that vitamin D deficiency is prevalent in COPD. Notwithstanding the fact that lower serum levels in certain genotypes may exist long before disease onset, impaired intake and reduced synthesis capacity because of low outdoors activities and accelerated skin ageing, are likely the main causes for such deficiency. As such, vitamin D deficiency can be considered as a direct consequence of COPD which does not exclude that deficiency once established, may accelerate disease onset or progression.8
Currently, only few longitudinal studies have explored relationships between 25(OH)D3 and disease outcomes in patients with COPD. One nested matched case-control study in a limited sample of 196 COPD patients of the Lung Health Study, did not find lower mean 25(OH)D3 levels in rapid decliners compared to non-decliners over a 6-year follow-up period.21 A larger secondary analysis of a randomized intervention trial with Azithromycin in 1,117 COPD patients at risk for exacerbations, found no relationship between baseline 25(OH)D3 levels and incident exacerbations over 1 year follow-up.2 Unfortunately, patients taking supplements were not excluded from this analysis. As supplementation in particular is often started in more severe COPD for reasons of osteoporosis, it may explain why the quintile with the highest 25(OH)D3 levels (>40ng/ml) presented with a higher number of exacerbations per patient year (1·79 ± 1·96 exacerbations per year). Notably, patients from the same trial with lowest 25(OH)D3 titers (<10ng/ml) exhibited the highest risk for relapse (2·2 ± 5·3 exacerbations per year), suggesting that severe vitamin D deficiency is a useful biomarker for future exacerbation risk.22 Indeed, a recent analysis in 402 cases of non-cystic fibrosis bronchiectasis demonstrated an inverse relationship of 25(OH)D3 with exacerbations23, although similar associations were not retrieved in a small London cohort of 97 COPD patients with 3 years follow-up24. Finally, Holmgaard and colleagues found no association between vitamin D status and all-cause mortality in 462 COPD patients followed for 10 years.25 Again, their design was limited by small sample size including patients on supplementation. Taken together, new longitudinal studies in large population-based samples or patient cohorts with COPD, are urgently needed to investigate if 25(OH)D3 levels determine risk of COPD, lung function decline, exacerbation frequency and even all-cause mortality.26
3. Potential mechanisms
It should be emphasized that epidemiological associations between vitamin D status and clinical outcomes are currently based on 25(OH)D3 serum levels. These measures are reliable markers for vitamin D synthesis or intake. The cutoffs for deficiency determine hazards for disease, particularly in the bone but possibly also in other organs.10 However, 25(OH)D3 serum levels conceptually reduce the entire signaling pathway to a single substrate which is not the active compound and which does not reflect local activation or activity. Therefore, lack of associations between substrate and clinical outcomes does not exclude vitamin D signaling being involved in the underlying pathophysiological process. One could even speculate that in COPD local or general alterations in the signaling cascade beyond 25(OH)D3 are critically determining vitamin D mediated actions. We will first review the main mechanisms of vitamin D relevant to the lung and then discuss how some of these may be downregulated.
3.1. Epigenetic regulation of vitamin D pathway
Upon binding of 1,25(OH)2D3 to the VDR, a heterodimer is formed with the retinoid X receptor (RXR) and this VDR/RXR complex further binds to specific genomic sequences in the promoter region of target genes (vitamin D response elements) thereby affecting gene transcription. To regulate transcription, the VDR/RXR dimer interacts with histone acetyltransferases (HATs) which are known as transcriptional activators. Binding of the VDR/RXR complex to negative VDREs with recruitment of histone deacetylases (HDACs), reverses HAT activity by making chromatin more condensed thereby promoting gene repression and transcriptional inactivation. 27 Histone modification enzymes may act alone or in concert to facilitate the activation or repression of chromatin-mediated gene expression in various inflammatory mediators, genes for cell cycle arrest, apoptosis, senescence, antioxidants and growth factors - notably, all disease processes involved in COPD.28;29 It is therefore plausible that epigenetic chromatin remodeling events important in COPD might be co-regulated by VDR-dependent signaling events (Figure 1). In fact, abnormalities in acetylation and methylation patterns on histones resulting from imbalance of HAT/HDAC and HMT/demethylases, are associated with alteration in gene expression and/or disease severity in COPD.30 Ligand activation of VDR may restore such imbalance by recruitment of HDAC2/SIRT1 deacetylases31;32, with the down-regulation of pro-inflammatory factors including NF-κB and IL-17 as well as the upregulation of the anti-inflammatory cytokine IL-10.33;34
3.2 Anti- inflammatory effects
In COPD, lung destruction is mediated in part through inflammation, oxidative stress and increased release of proteases and many of these processes are modulated by vitamin D.4;35 An overview of pathological processes and the interaction with VDR-dependent pathways, is depicted in Figure 2. Briefly, direct injury of airway epithelial cells by toxic and noxious gases activates pattern-recognition receptors such as Toll-like receptors on epithelial cells and alveolar macrophages, inducing NF- dependent inflammatory responses with subsequent recruitment of neutrophils, monocytes and dendritic cells to orchestrate the immune response.36 Interestingly, epithelial cells and macrophages express VDR at high levels and possess the enzymatic machinery to produce 1,25(OH)2D3 locally in the lung (CYP27B1).37;38 Macrophages in particular, respond to 1,25(OH)2D3 by preventing excessive expression of inflammatory cytokines and chemokines. VDR signaling may engage counter-regulatory mechanisms including the anti-inflammatory cytokine IL-10, the transrepression of NF--mediated responses or the targeting of MAPK phosphatase.32;39;40 In addition, dendritic cells in COPD portray elevated expression of co-stimulatory molecules to promote CD4+ cell differentiation and CD8+ cytotoxicity against antimicrobial or self-antigens. It is well established that 1,25(OH)2D3 inhibits dendritic cell maturation by lowering expression of MHC class II molecules, co-stimulatory molecules and pro-inflammatory cytokines/chemokines.41 With respect to T cells, 1,25(OH)2D3 can either directly or indirectly influence T cell reactivity by suppressing cytokines such as IFNand IL-17 while enhancing the regulatory markers FOXP3, CTLA4 and IL-10.33;42;43 As the efficacy of 1,25(OH)2D3 in vivo models of autoimmunity and asthma has been demonstrated repeatedly44;45, tapering down Th1/Th17 reactivity while promoting regulatory T cell responses may also protect against uncontrolled inflammation in COPD.36 Moreover, 1,25(OH)2D3 can attenuate MMP9 release thereby limiting tissue destruction46. Finally, either directly or indirectly via inflammatory cells, vitamin D may also modulate inflammation, muscle contraction and remodeling in airway smooth muscle cells. 47
3.3 Anti-infectious effects
Viral and bacterial infections are the main cause of acute COPD exacerbations which accelerate disease progression and increase the risk of death.48 Chronic infection or colonization of the lower airways may also amplify and perpetuate airway inflammation.49 Interestingly, genes encoding for antimicrobial polypeptides are driven by VDRE-containing promoters (Figure 2).50 TLR activation of monocytes and macrophages results in the upregulation of VDR and other VDR-target genes thereby inducing cathelicidin antimicrobial peptide (CAMP) with subsequent intracellular eradication of Mycobacterium tuberculosis.51 CAMP may also kill a number of antibiotic-resistant strains such as Pseudomonas aeruginosa and Staphylococcus aureus, different viruses, and chlamydia.51;52 Besides CAMP, the gene encoding for the antimicrobial defensin-β2, is also direct and indirect target for 1,25(OH)2D3.50 Exposure to 1,25(OH)2D3 results in a strong induction of these peptides with enhanced antimicrobial activity in various cell types, including myeloid cells, neutrophils, and bronchial epithelial cells. Other signaling pathways have been proposed to participate in the anti-infectious activities of 1,25(OH)2D3. For example, phosphatidylinositol 3-kinase was found to regulate the anti-mycobacterial activity of 1,25(OH)2D3 via the enhanced generation of reactive oxygen species (ROS) in monocytes and macrophages.53 1,25(OH)2D3 was shown to induce an NF- inhibitor in airway epithelium with subsequent dampening of chemokine and IFNβ release upon viral infection.54 1,25(OH)2D3 also induces autophagy and mediates co-localization of M. tuberculosis with antimicrobial peptides facilitating the destruction of these bacteria.55 Finally, 1,25(OH)2D3 enhances chemotactic and phagocytic capacity of macrophages which is severely impaired in patients with COPD.56;57 To summarize, it is clear that in addition to antibiotic treatment for acute bacterial infections, appropriate ligand activation of VDR may offer a powerful tool for boosting-up host innate immune defenses against low grade, smoldering bacterial and viral infections in COPD.
3.4 Impaired signaling in the vitamin D pathway
Impaired signaling in the vitamin D pathway is not only dependent on levels of 25(OH)D3. Deficient signaling may also comprise of reduced 25(OH)D3 activation (1α-hydroxylase, CYP27B1), increased catabolism of the active 1,25(OH)2D3 (24-hydroxylase, CYP24A) or impaired functioning of the receptor (VDR) with its complex regulation. The expression and functionality of these critical enzymes and receptor may depend on factors other than vitamin D intake or synthesis, in particular on smoking, chronic inflammation and (epi)genetic determinants which may even be organ or cell-specific.28 First, smoking can trigger increased expression of CYP24A in macrophages, resulting in increased catabolism and reduced bioavailability of the active compound.58 In dendritic cells for instance, the local activation of 25(OH)D3 is crucial for mediating the anti-inflammatory effects of vitamin D 43 and not only determined by absolute 25(OH)D3 levels but also by the concentration and genetic variant of its carrier protein DBP.59 Secondly, smoking may inhibit VDR translocation from the nucleus to the cell membrane.60 In mice, absence of VDR results in an abnormal lung phenotype with characteristics of COPD, including airspace enlargement, decline in lung function, increased lung inflammatory cellular influx and immune-lymphoid aggregates formation.61 Similar mechanisms may occur in patients as VDR expression in alveolar macrophages and airway epithelial cells was shown to be critically down-regulated by Aspergillus Fumigatus toxins in cystic fibrosis, a process which could be reversed by antifungal treatment.62 Finally, multifunctional enhancers including 1,25(OH)2D3 increase VDR gene transcription with local accumulation of VDR and increased signaling.27 Together, these data suggest that a down-regulation of critical intracellular enzymes and receptors in the vitamin D/VDR signaling cascade may occur in patients with COPD thereby creating a pro-inflammatory environment. If so, an important question arises whether environmental risk factors (including long-term vitamin D deficiency) may dampen or inactivate local VDR signaling and whether oral supplementation of the precursor 25(OH)D3, at low or high dose, will be able to overcome this.