Understanding the Immune-Endocrine Effects of Vitamin D in SARS-CoV-2 Infection: A Role in Protecting against Neurodamage

Vitamin D₃ is a fat-soluble steroid hormone with pleiotropic biologic effects [1]. It is derived from both food and from the physiological photoconversion of cutaneous 7-dehydrocholesterol into pre-vitamin D₃ and then into cholecalciferol, following exposure to UV-B solar rays. Cholecalciferol binds to circulating vitamin D-binding protein (VBP), and in the liver it is hydroxylated to calcifediol (25[OH]D₃) by 25-hydroxylase enzyme (cytochrome CYP2R1). Calcifediol is then further hydroxylated to calcitriol (1,25[OH]₂D₃) by 1α-hydroxylase enzyme (cytochrome CYP27B1). Calcitriol is the active hormonal final form of vitamin D₃ and can play both non-genomic and genomic effects, acting on vitamin D-receptor (VDR). Schematically, when hydroxylation in position 1 occurs in the kidney, calcitriol exerts endocrine rapid non-genomic effects on target cells (gut epithelial cells, osteoblasts, osteoclasts, parathyroid cells, tubular renal cells), regulating calcium-phosphorus homeostasis (shown in Fig. 1) [1]. On the other hand, when hydroxylation in position 1 occurs in immune cells, calcitriol exerts paracrine/autocrine slower genomic effects on immune cells themselves, downregulating autoimmune/inflammatory processes (shown in Fig. 1) [1]. Calcitriol is ultimately inactivated by 24-hydroxylase enzyme (cytochrome CYP24A1) into calcitroic acid, which is excreted in the bile and then eliminated in the faeces [1].

In addition to this canonical activation of pre-vitamin D₃, a non-canonical pathway has also been identified [2]. After prolonged sun exposure, pre-vitamin D₃ can be converted by CYP11A1 into two photoisomers, lumisterol (L₃) and tachysterol (T₃). L₃ and T₃ can be further hydroxylated to biologically active forms, such as 20(OH)L₃, 22(OH)L₃, 20,22(OH)₂L₃, 24(OH)L₃, 20(OH)T_3, 25(OH)T₃ by CYP27A1 [2]. These hydroxyderivatives interact not only with the VDR but also with human aryl hydrocarbon receptor, liver X receptor α and β, peroxisome proliferator-activated receptor γ, and retinoid-related orphan receptors α and γ [3]. Of note, they do not have calcemic properties, but they can affect immune function and pro-inflammatory pathways like calcitriol (Fig. 1) [3].

In daily clinical practice, serum 25(OH)D₃ concentrations are indicator of vitamin D₃ status of a person (calcifediol has a long half-life of about 3 weeks). The ranges of normality of serum 25(OH)D₃ concentrations have been established by the Endocrine Society in 2011: concentrations lower than 20 ng/mL are considered as “deficiency,” concentrations between 20 and 29 ng/mL are considered as “insufficiency,” while concentrations greater than 29 ng/mL are considered as “normality” [4]. Normal serum 25(OH)D₃ concentrations allow for adequate intestinal absorption of calcium and maintenance of normal serum parathormone values in most people, while a cut-off to ensure an immunomodulating effect has not yet been identified with absolute certainty [4].

The correlation between serum 25(OH)D₃ concentrations and paracrine/autocrine anti-inflammatory effects has been extensively investigated under autoimmune and inflammatory conditions, including infectious diseases [1]. Heliotherapy has been the only treatment of tuberculosis for centuries, until the discovery of antibiotics [1]. In recent decades, it has been clarified that the benefits of heliotherapy were due to the endogenous production of calcitriol, after the photoconversion of 7-dehydrocholesterol in cholecalciferol. In fact, calcitriol stimulates the synthesis and release of cathelicidin by innate immunity cells (monocytes and neutrophils) of tuberculosis patients [5]. LL-37 residue of cathelicidin is an antimicrobial peptide that damages lipoprotein membranes of Mycobacterium tuberculosis, hindering the formation of surface biofilms [5]. Moreover, LL-37 induces the production of interleukin (IL)-8 by monocytes/macrophages with chemotactic function for neutrophils [5].

Consequently, great interest has aroused the correlation between serum 25(OH)D₃ concentrations and the course of other respiratory infections. A 2017 meta-analysis of 25 randomized clinical trials (11,321 participants) reported that vitamin D₃ supplementation was associated with a reduction of the risk of acute respiratory infections, with an adjusted odds ratio of 0.88 [6]. Protection seemed greater in those subjects with baseline serum 25(OH)D₃ concentrations <25 ng/mL [6]. Daily or at most weekly doses of vitamin D₃ were more effective than monthly doses in raising serum 25(OH)D₃ concentrations as monthly supplementation boluses activated 24-hydroxylase enzyme and therefore the catabolic pathway of vitamin D₃ [7]. In 2021, an update of the previous meta-analysis included more randomized clinical trials (n = 46) and participants (n = 75,541) and confirmed the protective effect of vitamin D₃ supplementation against respiratory infections, with an odds ratio of 0.92 [8].

After the spread of COVID-19 pandemic, the role of vitamin D₃ in SARS-CoV-2 infection has been the object of thousands of studies and reports [9]. Deficient/insufficient serum 25(OH)D₃ concentrations have been correlated with increased susceptibility to infection and more severe disease courses [10]. The biological basis of these observations will therefore be discussed below, first summarizing the pathophysiology of SARS-CoV-2 infection. Then, this narrative review will focus on neurological involvement of COVID-19, speculating on the protective role that vitamin D₃ may exert in neuroprotection.

SARS-CoV-2 is an RNA virus that is transmitted from human to human by airborne droplets [11]. Although there are some structural differences between the different viral variants, SARS-CoV-2 virion is usually formed by essential proteins, such as nucleocapside proteins (N), membrane proteins (M), and a glycoprotein envelope (E), from which two spike proteins (S₁ and S₂) protrude. S₁ and S₂ adhere to upper respiratory tract cells and nasal olfactory mucosa. S₁ binds to host receptor angiotensin-converting enzyme 2 (ACE-2), while S₂, cleaved and activated by host transmembrane protease serine-protease-2 (TMPRSS-2), fuses viral and host envelopes, integrating viral RNA within the human cells [11].

Subsequently, SARS-CoV-2 replicates and releases double-stranded RNA inside the cells, usually recognized by cytosolic pattern recognition receptors (PRRs), such as retinoic acid-inducible gene I (RIG-I) and melanoma differentiation-associated gene 5 (MDA5). These PRRs stimulate the production of type 1 (α and β) and type 3 (λ) interferons, which have a direct and indirect antiviral function, through the recruitment of the cells of the innate immunity. Neutrophils, monocytes, and dendritic cells in turn present surface PRRs, such as Toll-like receptor (TLR)-2 and TLR-4, which recognizes viral glycoproteic envelope. TLR-2 and TLR-4 stimulate nuclear factor kappa b (NF-kB)-signaling pathway, with activation of Nod-like receptor protein 3 (NLRP3) inflammasome. NLRP3 releases pro-inflammatory cytokines (IL-1β, IL-18) that drive pyroptosis, the inflammatory form of programmed cell death [11]. Innate immunity cells then activate the more specific adaptive immunity (T and B lymphocytes) and SARS-CoV-2 infection ends in most cases within a few days, with patients reporting only flu-like symptoms (fever, nasopharyngitis, arthralgia) [11].

However, in a minority of cases, depending on different risk factors, SARS-CoV-2 can evade defense mechanisms, downregulating interferons production and blocking autophagy [12, 13]. This allows SARS-CoV-2 to spread from the upper to lower respiratory tract, thus reaching the alveoli. The intense but ineffective inflammatory response that is activated in the lungs leads to a progressive worsening of respiratory function. Together with epithelial cell damage, inflammation disrupts integrity of the lung vascular endothelium, promoting a pro-coagulative phenotype [14].

Although not yet fully elucidated, the mechanisms leading to endothelial damage are multiple. The binding between SARS-CoV-2 and ACE-2 receptor unbalances the renin-angiotensin-aldosterone (RAS) system [15]. Under physiological conditions, a reduction in systemic blood pressure stimulates renal renin production, which cuts circulating angiotensinogen produced by the liver into various fragments, including angiotensin I. Angiotensin I is converted into angiotensin II by ACE enzyme and in turn ACE-2 enzyme converts angiotensin II to angiotensin 1–7. Angiotensin 1–7 acts on angiotensin type 2 and Mas receptors, promoting the expression of endothelial nitric oxide synthase and reducing platelet aggregation, with an overall vasodilatory, anti-inflammatory, and antifibrotic effect (shown in Fig. 2) [15]. Binding between SARS-CoV-2 and ACE-2 dramatically reduces the production of angiotensin 1–7 and conversely leads to the accumulation of angiotensin II, which acts on angiotensin type 1 receptor downregulating the expression of endothelial nitric oxide synthase and promoting platelet aggregation, with an overall vasoconstrictor, pro-inflammatory and pro-fibrotic effect (shown in Fig. 2). At vascular level, there is therefore an oxidative stress with the release of oxygen-free radicals following episodes of hypoxia-ischemia which worsens endothelial damage [16].

Moreover, SARS-CoV-2 can induce a thrombotic endothelial damage, mediated by immune cells (immunothrombosis) (shown in Fig. 3). Following the activation of PRRs, monocytes express tissue factor on the surface, a protein that interacts with circulating coagulation factors to activate the extrinsic pathway of coagulation [11]. Furthermore, SARS-CoV-2 is recognized by the complement system via the mannose binding lectin, thus generating C5a fragment [17]. C5a is not only a chemoattractant for neutrophils but also stimulates neutrophils to express tissue factor on the surface [17]. Activated neutrophils then extrude nuclear material to trap and eliminate viral particles (NETosis) [18]. These traps can in turn activate the coagulation cascade by interacting with factor XII of the intrinsic pathway of coagulation [11]. Finally, SARS-CoV-2 can directly disrupt endothelial tight junctions, causing the exposure by the endothelium itself of the tissue factor [11]. The result of this redundant stimulation of the coagulation system is the formation of the fibrin clot, also favored by a deficit of the fibrinolysis pathway (shown in Fig. 3) [11]. Indeed, COVID-19 patients have high serum concentrations of plasminogen activator inhibitor 1, which inhibits the fibrinolytic activity of tissue plasminogen activator and urokinase [11].

A systemic endothelial damage has been demonstrated in the peripheral skin circulation by nailfold videocapillaroscopy, a non-invasive examination that allows to analyze the morphology and number of capillaries at the level of nailfold beds with a magnification of 40–200 times [19]. A videocapillaroscopic analysis performed on 61 subjects recovered from COVID-19 revealed a significant reduction in skin capillary density compared to healthy population and possibly involved in tissue and organs hypoxia in presence of long-COVID [19, 20]. In the most severe cases, the association between hyper-inflammatory cytokine storm and thrombotic events leads to systemic complications, among all acute respiratory distress syndrome, with multiorgan failure and patient’s death [11].

The link between serum 25(OH)D₃ concentrations and the course of SARS-CoV-2 infection has been extensively investigated and most studies agree that vitamin D₃ deficiency is related to a poorer prognosis of the disease [1, 9, 21]. Serum 25(OH)D₃ concentrations below 25 nmol/L have been associated with a higher risk of severe COVID-19 and systemic complications [22‒24]. Potential associations between VDR genetic polymorhisms, which can affect the expression and function of the protein, and the severity and/or mortality for COVID-19 have been also investigated. FokI (rs2228570), TaqI (rs731236), BsmI (rs1544410), and ApaI (rs7975232) are VDR single-nucleotide polymorphisms which have been variously associated with different aspects of COVID-19 (susceptibility, severity, mortality). However, the results of observational, retrospective, or case-control studies on this topic have been conflicting and do not allow to determine with certainty which polymorphisms contribute most to mitigating or aggravating the disease, also considering the different viral variants [25‒27].

Calcitriol and hydroxyderivatives of L₃ and T₃ could reduce SARS-CoV-2 invasion and replication, inflammation, and endothelial damage [3]. In course of COVID-19, inadequate serum 25(OH)D₃ concentrations correlate with reduced ACE-2 levels and ACE-2 mRNA expression, and calcitriol supplementation seems to restore ACE-2 levels, re-establishing a physiological ratio of angiotensin 1–7/angiotensin II concentrations (shown in Fig. 2) [28]. Calcitriol and hydroxyderivatives of L₃ and T₃ can also bind to SARS-CoV-2 receptor-binding domain of ACE-2, hindering the interaction between the virus and the receptor [29]. Moreover, they can cause a conformational change in TMPRSS-2 structure, further reducing the probability of virus entry into the host cell [29]. Furthermore, in vitro experiments have demonstrated that hydroxyderivatives of L₃ and T₃ could block some of the proteases used by SARS-CoV-2 to replicate (3CL-chymotripsin or main protease, RNA-dependent RNA polymerase) and calcitriol stimulates monocyte production of β-defensin 2 and cathelicidin, further reducing viral replication [30, 31]. Active vitamin D₃ also promotes the elimination of damaged (infected) cells by autophagy, upregulating the expression of Beclin 1 (activating factor of autophagy) and downregulating mTOR pathway (inhibitor pathway of autophagy) [32].

Regarding immune effects, calcitriol and hydroxyderivatives of L₃ and T₃ can mitigate inflammation, including the pro-inflammatory cytokine storm that can develop into the most severe cases of the disease. In an in vitro study, it has been demonstrated that calcitriol can downregulate NF-kB, a pivotal transcription factor for the activation of pro-inflammatory genes, in particular, IL-1, IL-6, IL-8, IL12, IL-17, IL-23, and tumor necrosis factor (TNF) α [33, 34]. L₃ and T₃-hydroxyderivatives of provitamin D₃ can downregulate IL-17 production antagonizing not only NF-kB but also retinoid-related orphan receptors α and γ and aryl hydrocarbon receptor [35]. They also upregulate the expression of Nrf2, a transcription factor for several proteins with antioxidant and anti-inflammatory effects (glutamate-cysteine ligase catalytic subunit, glutathione S-transferase, NAD [P]H quinone oxidoreductase 1, heme oxygenase-1) [36, 37].

Moreover, calcitriol can reduce neutrophil extracellular traps release in vitro, mitigating both inflammatory and endothelial damage (shown in Fig. 3) [38]. Calcitriol therefore stimulates the shift from T helper 1 lymphocytes to Th2 lymphocytes (IL-10 production) with an anti-inflammatory effect, through an autocrine signaling induced by C3b fragment of complement [39].

At last, healthy subjects with serum 25(OH)D₃ concentrations below 26 ng/mL show upregulation of the pro-coagulative platelet-monocyte and monocyte-endothelium interactions [40]. In vitro, calcitriol can also upregulate monocyte expression of thrombomodulin, a protein that reduces the activation of circulating factor VIII (intrinsic pathway of coagulation) and factor V (common pathway of coagulation) and which inhibits plasminogen activator inhibitor 1, with a final fibrinolytic effect and a protective role on the endothelium (shown in Fig. 3) [41].

Self-reported and/or objectively detectable neurological symptoms are described in more than 80% of hospitalized COVID-19 patients [42]. The most frequent subjective symptom is headache, followed by changes in smell (anosmia) and taste (ageusia), depression, and dizziness [42, 43]. Objective neurological manifestations have also been reported such as ischemic stroke, meningoencephalitis, myelitis, seizures, Guillain-Barré syndrome, demyelinating diseases, and others [43].

It has been demonstrated that SARS-CoV-2 can infect peripheral nervous system, interacting with ACE-2 and TMPRSS-2 expressed by olfactory epithelial cells [44]. A review of 24 autoptic studies of 149 brains of unvaccinated patients who died from COVID-19 revealed that viral RNA was detectable in the brain or olfactory nerve at low levels by targeted quantitative reverse transcriptase polymerase chain reaction [45]. SARS-CoV-2 was also identified by immunohistochemistry in the glossopharyngeal and vagal nerves of another cohort of 43 unvaccinated patients (53% of cases) who died from COVID-19 [46].

SARS-CoV-2 can also invade the central nervous system (CNS), using as receptors not only ACE-2 and TMPRSS-2 but also neuropilin-1, highly expressed by pericytes and astrocytes of the blood-brain barrier (BBB) [47]. In fact, a digital polymerase chain reaction investigation detected SARS-CoV-2 nucleocapsid gene expression in multiple areas of the CNS (cervical spinal cord, olfactory nerve, basal ganglia, cerebral cortex, brainstem, cerebellum, thalamus, hypothalamus, corpus callosum, and dura mater) of 44 unvaccinated patients who died from COVID-19 (100% of the study population) [48].

SARS-CoV-2 thus alters the permeability of the BBB, increasing the expression of matrix metalloproteinase-9 that destroys the basement membrane through the degradation of collagen IV and activates RhoA, a small G-protein, which promotes the disassembly of tight junctions through modifications of the cytoskeleton [49]. Moreover, the integrity of the BBB can be disrupted by peripheral inflammation [50]. Pro-inflammatory cytokines (i.e., IL-1, IL-6, IL-17) upregulate the expression of adhesion molecules on BBB endothelial cells (E-selectin, VCAM-1, ICAM-1) and enter the CNS [50]. Then, they polarize resting microglial immune cells toward an M1 phenotype, which promotes neurotoxicity via the release of further IL-1β, IL-6, TNF-α, reactive oxygen species [51]. M1 microglia induces also a neuroinflammatory reactive astrocyte phenotype, stimulating astrocytes to secrete pro-inflammatory cytokines and vascular endothelial growth factor, further weakening the BBB [50, 52]. So, peripheral lymphocytes/cytokines infiltration cause neuroinflammation that is detrimental to neurons, neurotransmission, and neural circuit functions [53].

Neuroinflammation has been confirmed by a single-cell transcriptomic study of the brains of 8 unvaccinated patients who died of COVID-19 [54]. In the choroid plexuses, there was an upregulation of genes (i.e., NQO1 and ZFP36), which caused a pro-inflammatory activation of microglia, through CCL and CXCL chemokine pathways [54]. Immunohistochemistry also showed a significant over-expression of CD68 (marker of macrophage activation) in the choroid plexuses of COVID-19 patients compared to controls [54]. In another cohort, 41 brains of unvaccinated patients who died from COVID-19 were autopsied, and microglial activation (positivity for CD68 at immunohistochemistry) was detected in 81% of cases, with inflammatory infiltrate of T lymphocytes (positivity for CD3) in 93% of cases [55].

Moreover, the viral and inflammatory damage of the BBB promotes the development of immunothrombosis at the level of the brain vessels [56]. In a 2021 meta-analysis, which considered 108,571 patients with COVID-19, acute cerebral vascular events were reported in 1.4% of cases, with cerebral ischemia as the main cause of stroke (87.4% of cases) [57]. Ischemic stroke was predominant in the large vessels with a multi-infarct distribution, supporting a thrombotic pathogenesis of the disease [57].

Of note, several studies have demonstrated anti-inflammatory and neuroprotective effects of vitamin D₃. Calcitriol acts at multiple levels, first by reducing the expression of adhesion molecules on BBB, thus limiting the entry of inflammatory cells into the brain [58]. In mouse models of vascular diseases (arterial hypertension, ischemic stroke), microglial cells are polarized toward an M1 phenotype (pro-inflammatory). However, they express VDR receptors on their surface and calcitriol promotes the shift from M1 to M2 (anti-inflammatory) phenotype. In fact, vitamin D₃ modulates NF-kB pathway, upregulating M2 microglial expression of IL-10 and downregulating production of reactive oxygen species, interferon γ, and TNF-α [59‒61]. Similarly, in mice in which cerebral oxidative stress has been induced to mimic memory impairment of Alzheimer’s disease, calcitriol downregulate NF-kB pathway and upregulate NRF-2 and HO-1 genes in the brain with an antioxidant effect [62].

Furthermore, VDR and CYP27B1 are also expressed by astrocytes and oligodendrocytes. Through an autocrine loop, calcitriol can reduce the release of IL-1, IL-6, and TNF-α from reactive astrocytes [63]. Vitamin D₃ also promotes oligodendrogenesis, and therefore the production of myelin, inducing oligodendrocyte precursor cells differentiation [64]. It is well known that deficient serum 25(OH)D₃ concentrations are a risk factor for the development of demyelinating lesions in course of multiple sclerosis [65]. Vitamin D₃ therefore promotes the release of neurotrophic cytokines, including nerve growth factor and brain-derived neurotrophic factor, supporting neuronal differentiation, growth, and development [66].

The effects of L₃ and T₃-hydroxyderivatives on neuroimmunological mechanisms have not yet been elucidated, but they can interact with neuroinflammation, stimulating the activation of the hypothalamic-pituitary adrenal axis and therefore the release of glucocorticoids with immunosuppressive function [67]. At last, involvement of the CNS is part of the post-COVID-19 syndrome, known as long-COVID, which is defined as “the condition that occurs in individuals with a history of probable or confirmed SARS-CoV-2 infection, usually 3 months from the onset of COVID-19, with symptoms that last for at least 2 months and cannot be explained by an alternative diagnosis” [68]. A recent meta-analysis reported fatigue as the most common symptom of long-COVID, followed by memory problems (“brain fog”) [68]. Other commonly reported neurological symptoms are persistent changes in taste and smell, anxiety, depression, and sleep disorders [69]. The pathophysiology of long-COVID is still partly unknown, but it is reasonable that the mechanisms of neurological damage are superimposable to those described in the acute phase, i.e., the passage of inflammatory cells through a damaged BBB in association with micro-thrombotic vascular disease which maintains chronic hypoxia and brain damage [69]. A very recent investigation reported that low serum 25(OH)D₃ concentrations at baseline of SARS-CoV-2 infection are correlated to the development of long-COVID symptoms, including neurocognitive ones, with an odds ratio of 1.09 after multiple-regression analyzes [70].

Although nowadays the danger of acute SARS-CoV-2 infection has been mitigated by less aggressive viral variants and by mass vaccinations, adequate serum 25(OH)D₃ concentrations in COVID-19 patients could be protective against systemic complications, including acute and chronic neurological manifestations (long-COVID) [71, 72]. Calcitriol and hydroxyderivatives of L₃ and T₃ show interesting neuroimmunoendocrine effects in course of SARS-CoV-2 infection and can play an adjuvant role in neuroprotection, reducing BBB endothelial damage, antagonizing vascular immunothrombosis, and downregulating neuroinflammation. Specific studies in humans are desirable to confirm the evidence collected to date.

We thank the European Alliance of Associations for Rheumatology (EULAR) Study Group on Neuro Endocrine Immunology of the Rheumatic Diseases (NEIRD) for the continuous cultural support. All figures were created by Dr. Stefano Soldano with www.biorender.com.

The authors have no conflicts of interest to declare. V.S. is a senior clinical investigator of the Research Foundation Flanders (Belgium) (FWO) (1.8.029.20N). The FWO was not involved in study design, collection, analysis, and interpretation of data, writing of the report, or in the decision to submit the article for publication.

Authors received no funding for the manuscript.

E.G. and M.C. conceptualized the review, collecting data, and writing the manuscript. S.S. created all the figures. S.S., A. Casabella, E.H., A. Cere, C.P., S.P., A.S., and V.S. revised the manuscript for important intellectual content. All authors agreed to the content of the review and are accountable for all aspects of accuracy and integrity. All authors read and agreed to the current version of the manuscript.