Vitamin D supplementation worsens Alzheimer's progression: Animal model and human cohort studies

Abstract

Vitamin D deficiency has been epidemiologically linked to Alzheimer's disease (AD) and other dementias, but no interventional studies have proved causality. Our previous work revealed that the genomic vitamin D receptor (VDR) is already converted into a non-genomic signaling pathway by forming a complex with p53 in the AD brain. Here, we extend our previous work to assess whether it is beneficial to supplement AD mice and humans with vitamin D. Intriguingly, we first observed that APP/PS1 mice fed a vitamin D-sufficient diet showed significantly lower levels of serum vitamin D, suggesting its deficiency may be a consequence not a cause of AD. Moreover, supplementation of vitamin D led to increased Aβ deposition and exacerbated AD. Mechanistically, vitamin D supplementation did not rescue the genomic VDR/RXR complex but instead enhanced the non-genomic VDR/p53 complex in AD brains. Consistently, our population-based longitudinal study also showed that dementia-free older adults (n = 14,648) taking vitamin D₃ supplements for over 146 days/year were 1.8 times more likely to develop dementia than those not taking the supplements. Among those with pre-existing dementia (n = 980), those taking vitamin D₃ supplements for over 146 days/year had 2.17 times the risk of mortality than those not taking the supplements. Collectively, these animal model and human cohort studies caution against prolonged use of vitamin D by AD patients.

Keywords: Alzheimer's disease; longitudinal study; non-genomic vitamin D signaling; p53; vitamin D; vitamin D receptor.

FIGURE 1

Dietary supplementation of vitamin D₃ aggravates AD pathology in APP/PS1 AD mice. (a) Serum 25(OH)D₃ levels in APP/PS1 (TG) and wild‐type (WT) mice. Mice were weaned at 4‐weeks of age (±3 days) and maintained on a vitamin D₃‐sufficient diet (600 IU/Kg of cholecalciferol). Serum vitamin D₃ levels were determined by 25(OH)D₃ enzyme‐linked immunosorbent assay (EMSA) at the indicated time points (n = 5). Results are shown as mean ± SD. *p < 0.05; **p < 0.01; ***p < 0.001 by unpaired t‐test. (b) WST‐1 cell viability assay. SH‐SY5Y cells were exposed to vitamin D₃ alone or Αβ (4 μM) plus vitamin D₃ (calcitriol or calcidiol) for 6 h prior to assays. Results are shown as mean ± SD. *p < 0.05 by One‐way ANOVA. (c) Western blot analysis of VDR, apoptotic and autophagic marker proteins in SH‐SY5Y cells exposed to Aβ42 or plus without or without Vitamin D₃. SH‐SY5Y cells were treated with 4 μM Aβ42 alone or Aβ42 plus 10, 30, or 100 nM calcitriol for 6 h before harvesting cell lysates for analysis. (d) Representative immunofluorescent micrographs of gliosis (anti‐GFAP, GA5) and amyloid aggregates (anti‐Αβ, D54D2) in hippocampal tissues of APP/PS1 mice. 4.5‐month‐old APP/PS1 mice were fed with vitamin D₃‐supplemented (8044 IU/Kg cholecalciferol/day, Vit. D) or vitamin D₃‐sufficient diets (600 IU/Kg cholecalciferol/day, Ctrl) for 3 months before harvesting brain tissues for analysis. Sections of cortex or hippocampus were stained with the indicated antibodies. The average percentage of surface area with Αβ plaques in five consecutive sections per animal (n = 4–7) was quantified by ImageJ in right panel. (e) Western blot analysis of Αβ production and β‐secretase 1 (BACE1) levels in hippocampal lysates of APP/PS1 mice supplemented with or without vitamin D_3. Densitometrical quantification of Aβ and BACE bands were normalized to GAPDH (right panel). (f) Cognitive performance for AD mice supplemented with vitamin D₃. 4.5‐month‐old APP/PS1 mice were fed with vitamin D₃–fortified (Vit. D) or vitamin D₃‐sufficient diets (Ctrl) for 7.5 months before Morris Water Maze. *p < 0.05 by One‐way ANOVA

FIGURE 2

Vitamin D supplementation enhances VDR/p53 but not VDR/RXR complex in worsening brain pathology in APP/PS1 AD mice. (a) Mammalian two‐hybrid assays for studies of the interaction of VDR with RXR in neuronal cells exposed to Αβ plus with vitamin D₃. SH‐SY5Y cells were treated Aβ42 for 6 h and then co‐treated with 10 nM calcitriol for additional 6 h prior to harvesting for mammalian two‐hybrid luciferase assays. (b,c) Western blot analysis of co‐immunoprecipitation of VDR/p53 complex in SH‐SY5Y cells and hippocampal tissues of APP/PS1 mice. (d) Western blot analysis of VDR, p53, and MDM2 in the hippocampal lysates of APP/PS1 mice treated with or without p53 inhibitor. 4.5‐month‐old APP/PS1 mice raised on vitamin D₃‐sufficient diets were intraperitoneally injected weekly with 3 mg/kg of p53 inhibitor pifithrin‐α (PFTα) for 7.5 months before harvesting hippocampal tissues for analysis. Densitometrical quantification of VDR, p53, and MDM2 bands were normalized to GAPDH (lower panel). *p < 0.05; **p < 0.01; ***p < 0.001 by unpaired t‐test. (e) Western blot analysis of autophagic markers LC3, p62, and ser349 phosphorylated p62 (p62‐S349) in the hippocampal lysates of APP/PS1 mice injected with or without PFTα. Densitometrical quantification of LC3, p62‐S349, and p62 bands were normalized to GAPDH (right panel). (f) Western blot analysis of Αβ and BACE levels in the hippocampal lysates of APP/PS1 mice injected with or without PFTα. Densitometrical quantification of Aβ and BACE bands were normalized to GAPDH (right panel). (g,h) p53 inhibitor amelioration of vitamin D₃‐aggravated Aβ aggregation and apoptosis. The Aβ, GFAP, and TUNEL‐positive signals in five consecutive sections per animal (n = 5) was quantified by ImageJ and presented as the mean ± SD. Scale bars, 50 μm. (i) Cognitive performance assays for the AD mice treated with p53 inhibitor. APP/PS1 mice were given with or without weekly injections of PTFα (n = 6 mice) starting at the age of 4.5‐month. APP/PS1 mice at 12‐month of age were used for the Morris Water Maze test

FIGURE 3

Population‐based cohort study of the associations between the incidence of dementia and calcitriol supplementation. (a) Flow chart of cohort formation for epidemiological study of association between incident dementia and calcitriol supplementation. NHIRD, National Health Insurance Research Database; LHID2000, Longitudinal Health Insurance Database 2000; Comorbidities including calcium prescription, chronic renal disease, osteoporosis, thyroid, diabetes, hyperlipidemia, hypertension, and Charlson score. (b) The adjusted curves of dementia development in study subjects aged over 65 years with different average cumulative dosages of calcitriol with a follow‐up of up to 10 years (n = 14,648). The definition of the ‘dosage/year’ is the assumed average maintenance dose (mcg) per year for calcitriol used in the whole follow‐up. ***p < 0.001 by Maximum Likelihood‐ratio test

FIGURE 4

Population‐based cohort study of the associations between risk of mortality and calcitriol use in dementia. (a) Flow chart of cohort formation for epidemiological study of association between survival and calcitriol supplementation in patients with pre‐existing dementia. HV, Registry for catastrophic illness patients; ICD9, International Statistical Classification of Diseases and Related Health Problems (ICD) 9; NHIRD, National Health Insurance Research Database; ID, Registry for beneficiaries; RCIP, Registry of Catastrophic Illness Patients; Comorbidities including calcium prescription, chronic renal disease, osteoporosis, thyroid, parathyroid disorders, hyperlipidemia, hypertension, and Charlson score. (b) The adjusted survival curves among dementia patients with different average dosages of calcitriol. The relationship between mortality and calcitriol use was determined by using the Kaplan–Meier survival curves and log‐rank tests with a follow‐up of up to 10 years (n = 980). The model was adjusted for age, sex/gender, calcium prescription, chronic renal disease, osteoporosis, thyroid, parathyroid disorders, hyperlipidemia, and hypertension. ***p < 0.001 by Maximum Likelihood‐ratio test (Ctrl vs. high cumulative doses with >36.5 mcg/year)