|Home | About | Journals | Submit | Contact Us | Français|
Vitamin D is being touted as an anti-infective agent and it has even been suggested that vitamin D supplementation could be effective against the H1N1 influenza virus. The claims are largely based on the ability of vitamin D to induce antibacterial peptides and evidence that the immune system produces active vitamin D (1,25(OH)2D3) in situ. While there are many examples of immune production of 1,25(OH)2D3 in vitro, there is little in vivo evidence. In addition, it is not clear what role immune production of 1,25(OH)2D3 has on the course of disease. Vitamin D and 1,25(OH)2D3 inhibit T helper type 1 (Th1)/Th17-mediated immune responses and autoimmune diseases by acting on the innate and acquired immune system to inhibit the function of Th1 and Th17 cells. Th1 and Th17 cells are important in host resistance to many infections including tuberculosis (TB) caused by Mycobacterium tuberculosis. Paradoxically the innate immune system is induced to produce antibacterial peptides that are effective against TB in vitro. Data from several models of infection have so far not supported a role for vitamin D in affecting the course of disease. There is also very little evidence that vitamin D affects the course of human TB infection. Experiments have not been done in cells, mice or humans to evaluate the effect of vitamin D on influenza virus. At this time it would be premature to claim that vitamin D has an effect on TB, influenza or any other infection.
Recently, there has been a great deal of interest in the role that vitamin D might play in host resistance to infection. The increase in attention came largely as a result of reports of two findings: (1) the immune system could produce the enzyme that converts circulating vitamin D to active vitamin D, and (2) active vitamin D produced in the immune system led to the induction of cathelicidin, which in turn inhibited replication of Mycobacterium tuberculosis in vitro.1 This finding has then led to a number of claims being reported in the scientific and non-scientific communities that claim vitamin D as a broadly anti-infective agent. Two recent letters to the editor suggest that vitamin D supplementation would be beneficial in people infected with influenza.2,3 In fact, the title of one of them is ‘Pandemic influenza A (H1N1): Mandatory vitamin D supplementation?’3 Here we will review the evidence behind the claims that vitamin D status could affect the course of infection. We will first look at the evidence that in vivo the immune system produces the vitamin D 1α-hydroxylase, then briefly discuss the described roles of vitamin D in the control of innate and adaptive T-cell immunity and then end with a look at the evidence that vitamin D affects in vitro and in vivo clearance of selected infections, including M. tuberculosis and influenza. We conclude that at present the evidence does not support a positive or negative role for vitamin D in host resistance to infection.
In 1970 Fraser and Kodicek4 first described the kidney as the source of the vitamin D 1α-hydroxylase enzyme (Cyp27B1 gene). The data demonstrated that intact chickens produced but nephrectomized chickens failed to produce what was determined to be 1,25(OH)2D3.4 Later it was shown by two different groups that patients with advanced renal failure or nephrectomy (prior to transplantation) had no detectable 1,25(OH)2D3 in the blood.5,6 Experiments in rats using radiolabeled 25(OH)D3 (to increase the sensitivity of the assay) showed that normal, otherwise healthy nephrectomized rats failed to produce 1,25(OH)2D3 in the blood, bone or intestine.7-9 The exception was found in nephrectomized pregnant female rats where extra-renal production of 1,25(OH)2D3 was detected in the blood.7 More recently, transgenic mice that express the bacterial LacZ reporter gene in the place of the vitamin D 1α-hydroxylase confirmed the renal-only expression.10 No reporter gene activity was detected in skin, lung, intestine, skeletal muscle, liver and ovary.10 Consistent with the earlier finding in rats, expression was found in the placenta of pregnant females.10 Conversely, measurements by several groups done by mRNA expression and polyclonal antibody staining of tissues have shown 1α-hydroxylase expression in healthy tissues including the skin, lymph node and colon of both humans and mice.11-14 Unlike conversion of 25(OH)D3 to 1,25(OH)2D3, these techniques fail to measure enzymatic activity, which is the most reliable measure of the vitamin D 1α-hydroxylase enzyme. Therefore, we conclude that under normal physiological conditions, animals and humans (with the exception of pregnant females) express the vitamin D 1α-hydroxylase only in the kidneys.
The question of whether extra-renal production of the 1α-hydroxylase occurs during disease, however, is still open. Several different reports suggest that during granulomatous diseases in humans, hypercalcemia can exist and may be a result of extra-renal production of 1,25(OH)2D3.11,15-17 The most convincing example of extra-renal production of the 1α-hydroxylase is a single anephric patient with sarcoidosis that had hypercalcemia.15 The source of the 1,25(OH)2D3 production was linked to lung macrophages isolated from several sarcoidosis patients with hypercalcemia.18,19 It was also shown that macrophages from other lung diseases did not produce 1α-hydroxylase activity.20 A number of research groups have reported 1α-hydroxylase activity in vitro from cultured cells. Data from some of these in vitro systems have shown that 25(OH)D3 is converted in vitro to 1,25(OH)2D3 by monocytes and dendritic cells (DCs).12,21 In particular, human macrophages have been shown to produce the 1α-hydroxylase when infected with M. tuberculosis and stimulated through toll-like receptors (TLRs1). However, there are no credible examples of 1α-hydroxylase production in vivo from experimental animals. There are reports in mice with experimental colitis that 1α-hydroxylase is induced, but the experiment was done using the same polyclonal antiserum that stained a variety of tissues from normal mice and humans described above.13 The lack of confirmatory data in vivo in experimental animals may be due to differences between humans and rodents, lack of vigorous investigation in animals, or perhaps a lack of understanding of the conditions under which the 1α-hydroxylase is induced in the immune system in vivo.
Reports of extra-renal production of the 1α-hydroxylase in the immune system have led investigators to propose a role for this enzyme in situ as a source of 1,25(OH)2D3 for use by the immune system.22 However, in sarcoidosis and other granulomatous diseases the more severe disease phenotypes were associated with increased levels of 1,25(OH)2D3 in the serum, and resolution of the disease following immunosuppressive therapy resulted in the correction of 1,25(OH)2D3 and calcium concentrations.16,17,23 The autocrine production of 1,25(OH)2D3 may not be beneficial; instead, it might be that immune production of 1,25(OH)2D3 is part of the pathogenic process in granulomatous diseases. More information is needed to determine under what conditions the 1α-hydroxylase enzyme is expressed in extra-renal tissues in vivo and whether or not expression of this enzyme is a positive or negative contributor to granulomatous diseases in humans.
T helper type 1 (Th1) immune responses are critical for the clearance of many bacterial, viral and parasitic pathogens and autoimmunity can occur when those responses go uncontrolled. Th1 immune responses are characterized by increased interferon (IFN)-γ and reduced interleukin (IL)-4. Vitamin D has been shown by several investigators using many different systems to suppress the generation of a Th1 response both in vitro and in vivo.24-28 1,25(OH)2D3 acts directly on T-cells to inhibit T-cell proliferation and IFN-γ production.29,30 In addition to the direct effects of 1,25(OH)2D3 on T-cells, indirect effects of 1,25(OH)2D3 acting on the antigen-presenting cells (APC) also results in decreased Th1 responses.29 1,25(OH)2D3 has been recognized as an immuno-suppressive agent that ameliorates the pathogenesis of several different experimental models of Th1 autoimmune diseases, including inflammatory bowel disease (IBD), diabetes, multiple sclerosis (MS), arthritis and several others (reviewed in ref.24). Furthermore, vitamin D deficiency and vitamin D receptor (VDR) deficiency in mice have been shown to exacerbate experimental IBD, MS, and diabetes (reviewed in ref.24)
More recently, in addition to Th1 cells, many of the auto-immune disease models have been shown to depend in part on Th17 cells, which make IL-17. There are reports that treatment of naïve CD4 T-cells during Th17 priming with 1,25(OH)2D3 inhibits IL-17 production.31 In addition, 1,25(OH)2D3 indirectly suppressed Th17 cell induction by inhibiting DC production of IL-6 and IL-23 that induce Th17 cells.32 In vivo, oral 1,25(OH)2D3 treatment reversed Th17-mediated experimental autoimmune uveitis in mice.31 The evidence from multiple investigators using several different in vitro and in vivo systems supports both direct and indirect effects of vitamin D that inhibit Th1 and Th17 responses.
Vitamin D and 1,25(OH)2D3 are required for the optimal development and function of several regulatory T-cells. Regulatory CD4/CD25 + FoxP3 + T (reg) cells are responsible for suppressing immune responses and limiting tissue damage and inflammation. Several groups have shown that while expression of the VDR is not required for development of T regs, 1,25(OH)2D3 increases both number and function of T reg cells.33,34 1,25(OH)2D3 has direct effects on the T reg cells and indirect effects via the induction of tolerogenic DCs that promote T reg function to alleviate autoimmune disease.33 In the gut specialized regulatory T-cells that express the CD8αα homodimer protect the gastrointestinal tract from bacterial microflora and other antigens found there. VDR knockout (KO) mice have reduced numbers of CD8αα T-cells and are more susceptible to several different models of experimental IBD.34 Vitamin D is required for the development of CD8αα T-cells, and in the gut reduced IL-10 production by the CD8αα T-cells is associated with increased experimental IBD in the VDR KO mice.34 Invariant natural killer T (iNKT) cells are T-cells that can act as regulatory cells and bridge innate and adaptive immunity. Induction of iNKT cells has been shown to be protective in several autoimmune diseases including experimental IBD and MS. Vitamin D regulates iNKT cell development and function and 1,25(OH)2D3 treatment induces cytokine production in iNKT cells.35 Vitamin D is a positive regulator of several regulatory T-cells and the induction of regulatory T-cells is associated with the beneficial effects of 1,25(OH)2D3 as an inhibitor of Th1/Th17-mediated autoimmunity.
Th2-mediated immune responses are also regulated by vitamin D. Two studies showed that 1,25(OH)2D3 increased the production of IL-4 and IL-10 by CD4+ T-cells under Th2 cell culture conditions in vitro.36,37 Conversely, it has also been reported that addition of 1,25(OH)2D3 inhibits the production of IL-4 in Th2 cells and does not affect the expression of genes important in Th2 differentiation.38 In vivo experiments in the Th2-mediated disease experimental allergic asthma also provide conflicting results. Using the ovalbumin/alum model of experimental allergic asthma, 1,25(OH)2D3 has been shown to inhibit, induce or to have no effect on symptoms of experimental asthma.39-41 VDR KO mice failed to develop allergic asthma, while vitamin D deficiency had no effect in the same model.40 Th2 cells and IL-4 production are inhibited by T reg and Th1 cells, so perhaps the disparate results reflect indirect regulation of Th2 cells via effects on other cell types. The conflicting results on the effects of vitamin D on Th2 cells in vitro and in experimental asthma suggest a complicated and as yet not well understood effect of vitamin D on the Th2 cell responses.
The innate immune system plays an important role in early defense and antigen presentation. Regulation of macrophage and DC function by 1,25(OH)2D3 is important for the inhibition of experimental autoimmunity. 1,25(OH)2D3 treatment of DCs in vitro inhibited differentiation and maturation of DC and resulted in DCs that when transferred, induced in vivo suppression of alloreactive T-cells.42 1,25(OH)2D3-treated DCs produced lower levels of IL-12, and expressed less co-stimulatory and MHC class II molecules than control-treated DCs.33 Conversely, 1,25(OH)2D3- and lipopolysacharide (LPS)-treated DCs secreted increased amounts of IL-10, and were capable of inducing T reg cell development.33 The dextran sodium sulfate (DSS) model of acute colitis does not require T-cells and is mediated by the innate immune system. 1,25(OH)2D3 treatment suppressed secretion of several macrophage and DC products in vitro (tumor necrosis factor [TNF]-α, IL-1β and IL-6) and experimental DSS induced colitis in vivo.24,32,33,43 VDR KO mice overproduced TNF-α, IL-1β and IL-12 and as a result were extremely susceptible to DSS colitis.43,44 Vitamin D inhibits DC and macrophage functions such that IL-12-mediated responses are reduced and IL-10-mediated responses are bolstered. In addition, 1,25(OH)2D3 treatment of APC results in the reduced induction of Th1-mediated immune responses while maintaining the ability to induce T regs.
Vitamin D has been shown to regulate TLR-mediated events in multiple cell types. TLR-mediated production of IL-12 and TNF-α was inhibited by 1,25(OH)2D3.45,46 In neutrophils, 1,25(OH)2D3 suppressed the ability of LPS to induce IL-1β expression as well as inhibited some antimicrobial genes.47 More recently, 1,25(OH)2D3 has been shown to induce the expression of antimicrobial peptides (cathelicidin and β-defensin) in innate immune cells stimulated through TLR receptors, including monocytes, neutrophils and keratinocytes.48 Furthermore, 1,25(OH)2D3 treatment enhanced the phagocytosis of monocytes and induced autophagy in macrophages.48 Stimulation through TLRs in the presence of vitamin D inhibits inflammatory cytokine production (IL-12, TNF-α and IL-1β) in cells of the innate immune system while enhancing several antimicrobial pathways.
Our current understanding of the effects of vitamin D on immune function can be used to predict the effects of vitamin D on host immunity to infectious organisms. The same Th1/Th17 immune responses that are pathogenic in autoimmunity are protective during infections. In addition, regulatory T-cells not only suppress autoimmune diseases, but also suppress clearance of infectious organisms. Based on the ability of vitamin D to suppress innate and acquired immune responses that result in the suppression of Th1- and Th17-mediated immune responses, we would predict that vitamin D would impair the ability of the host to clear infections dependent on these cell types. Furthermore, vitamin D would be predicted to increase T reg cells that would suppress responses to infectious organisms. However, the ability of vitamin D to induce antibacterial peptides suggests a less straightforward outcome of changes in vitamin D in vivo and resultant effects on host resistance to infection.
The experimental data that has looked at the relationship between vitamin D and infections has been summarized in Table 1. Listeria monocytogenes is a food-borne pathogen that has commonly been used to study the Th1-mediated immune response to intracellular bacterial infections. Helming et al.49 showed that treating IFN-γ-activated macrophages with 1,25(OH)2D3 inhibits listeriacidal activity and suppressed oxidative burst. In vitro, VDR KO macrophages were as good as wild-type (WT) for killing Listeria.49 In vivo, VDR KO mice produced more IFN-γ but showed slightly delayed kinetics in Listeria clearance compared with WT.50 However, VDR KO mice were able to clear Listeria infections.50 1,25(OH)2D3 treatment of WT mice had no effect on the rate of Listeria clearance in vivo (unpublished data). Clearance of Leishmania major is also dependent on Th1-mediated responses and 1,25(OH)2D3-treated macrophages produced less nitric oxide and killed fewer parasites.51 Infection of VDR KO mice with L. major showed normal Th1 responses and clearance of the organisms.51 Vitamin D-deficient mice were more susceptible to Mycobacterium bovis infection than vitamin D-sufficient mice.52 The increased susceptibility of vitamin D-deficient mice to M. bovis infection was linked to an effect on nitric oxide production.52 1,25(OH)2D3 treatment did not alter the ability of mice to clear two infections that require Th1/Th17-mediated immunity (Candida albicans or herpes simplex virus-153). Host resistance to Shistosoma mansoni requires a strong Th2-mediated response. VDR KO mice had larger liver granulomas (Th2-mediated), but were no different from wild-type mice in their ability to clear a S. mansoni infection.54 In addition, 1,25(OH)2D3 treatment had no effect on the S. mansoni infection (unpublished data). Host immunity to Bordetella pertusis infection requires Th1 and Th2 responses and VDR KO mice were no different than WT mice in their ability to clear B. pertusis infection (unpublished data). The evidence in mice does not support either a beneficial or harmful effect of vitamin D on host immunity to infections that require either Th1/Th17- or Th2-mediated immune responses for clearance.
Vitamin D and 1,25(OH)2D3 induce antibacterial peptides in vitro that effectively inhibit tuberculosis (TB). Early studies in 1985 showed that 1,25(OH)2D3 treatment of murine and human macrophages could potentiate the effects of IFN-γ to inhibit TB in vitro.55,56 Since 1,25(OH)2D3 treatment suppressed IFN-γ production, it remains unclear as to what the effect of 1,25(OH)2D3 would be when additional IFN-γ was not added. In mixed cultures of cells from TB patients, 1,25(OH)2D3 reduced the production of IFN-γ and TNF-α.57 Inhibition of IFN-γ, IL-12 and TNF-α by a variety of means are associated with an increased risk to TB infection.58-62 Unfortunately, there are no experiments that have looked at both the IL-12/IFN-γ-inhibiting and cathelicidin-inducing effects of 1,25(OH)2D3 on macrophages. Without a better understanding of the relationship between the antibacterial and IL-12 suppressive responses, it is difficult to predict what the net effect of vitamin D would be on host resistance to TB.
Epidemiological data suggest that low vitamin D status is associated with TB severity or susceptibility.63 Often mentioned is the anecdotal association that patients placed in the sun would see improvement in TB symptoms. A meta-analysis showed a positive association between VDR polymorphisms and host susceptibility to TB.63 Recently, a double-blind, randomized and placebo-controlled trial used three high dose (100,000 IU) vitamin D supplements in TB patients.64 The study showed no beneficial effect in clinical outcome or mortality in TB.64 Another recent report in dialysis patients showed no correlation between vitamin D supplementation and decreased risk of TB infection.65 Thus far, the evidence in humans with TB does not support a beneficial or harmful role of vitamin D supplementation.
It is estimated that at least one upper respiratory tract infection (URI) afflicts 72% of adults each year.66 The effect of vitamin D on URI is based on epidemiological data and therefore associations. 25(OH)D3 levels have been inversely associated with URI incidence.66 In a randomized, double-blind trial of vitamin D supplementation, vitamin D was shown to have no effect on the clinical course of URIs.67 In 2006 Cannell et al.68 suggested that children who received vitamin D supplements had a decreased incidence of respiratory infections and attributed this observation to vitamin D regulation of the antimicrobial peptides cathelicidin and defensin β2. Vitamin D has been shown to increase the expression of antibacterial peptides; however, the effect of vitamin D on these antibacterial peptides in vitro or in vivo against influenza has not been tested. Leikina et al.69 showed that retrocyclin (not shown to be a vitamin D target), a theta-defensin (that is not expressed by humans), can inhibit influenza virus in a canine-derived cell line. Any effect of either vitamin D or 1,25(OH)2D3 on influenza replication in vitro and/or in vivo has so far not been tested.
At present, there is not adequate information available to claim vitamin D as an anti-infective agent. The data supporting vitamin D as a factor that could improve resistance to infection are based on in vitro experiments that demonstrate immune cells make the vitamin D 1α-hydroxylase, and that 1,25(OH)2D3 induces antibacterial peptides that kill TB. However, contradictory data exist that host immune responses important in the control of TB are inhibited by vitamin D and 1,25(OH)2D3. Furthermore, it is unclear as to whether immune-mediated production of 1,25(OH)2D3 is protective or pathogenic. In vivo experiments to address the signals and role of immune produced 1,25(OH)2D3 have not been done. In experimental animals the data neither support nor refute an effect of vitamin D on several infections. Data from humans show associations between vitamin D status or genetic polymorphisms and TB. There are no data to support any relationship between vitamin D and host resistance to influenza. At this time it would be premature to suggest that vitamin D might be useful to improve host resistance to TB, influenza or any other infectious organism.
This work was supported by National Institutes of Health/National Institute of Diabetes and Digestive and Kidney Diseases DK070781 and National Center for Complementary and Alternative Medicine and the Office of Dietary Supplements AT005378 to MTC.
Author contributions: All authors participated in the design, interpretation and writing of the review. We thank Dr Avery August and the members of the Center for Immunology and Infectious Diseases for lively discussions.