B-vitamin deprivation induces hyperhomocysteinemia and brain S-adenosylhomocysteine, depletes brain S-adenosylmethionine, and enhances PS1 and BACE expression and amyloid-β deposition in mice

Abstract

Etiological and molecular studies on the sporadic form of Alzheimer's disease have yet to determine the underlying mechanisms of neurodegeneration. Hyperhomocysteinemia is associated with Alzheimer's disease, and has been hypothesized to promote neurodegeneration, by inhibiting brain methylation activity. The aim of this work was to determine whether a combined folate, B12 and B6 dietary deficiency, would induce amyloid-β overproduction, and to study the mechanisms linking vitamin deficiency, hyperhomocysteinemia and amyloidogenesis in TgCRND8 and 129Sv mice. We confirmed that B-vitamin deprivation induces hyperhomocysteinemia and imbalance of S-adenosylmethionine and S-adenosylhomocysteine. This effect was associated with PS1 and BACE up-regulation and amyloid-β deposition. Finally, we detected intraneuronal amyloid-β and a slight cognitive impairment in a water maze task at a pre-plaque age, supporting the hypothesis of early pathological function of intracellular amyloid. Collectively, these findings are consistent with the hypothesis that abnormal methylation in association with hyperhomocysteinemia may contribute to Alzheimer's disease.

Introduction

The sporadic form of Alzheimer Disease (AD) is a multi-factorial disease; studies on nutrition, metabolism and neurodegeneration support a strong link between dietary and genetic factors with ageing and AD (Grant et al., 2002, Mattson et al., 2002, Luchsinger et al., 2004, Dosunmu et al., 2007). Specifically, the involvement of homocysteine (HCY) and its dietary determinants (vitamins B6, B12 and folate) in dementia has been a topic of intense investigation (Kado et al., 2005); although it was first hypothesized in the 80s (Abalan, 1984), it has only recently become widely recognized as a risk factor for AD and dementia (Joosten et al., 1997, Clarke et al., 1998, McCaddon et al., 1998, Seshadri et al., 2002). HCY is a sulfur containing amino acid that doesn't participate in protein synthesis; its complex biochemical pathway is regulated by the presence of folate, vitamin B12 and B6 (among other metabolites) and leads to the production of methyl-donor molecule S-adenosylmethionine (SAM). Folate and B12 are cofactors in trans-methylation reactions transforming HCY to methionine, whereas B6 is the cofactor in trans-sulfuration reactions, responsible for HCY removal and glutathione synthesis. SAM can donate a methyl group to different substrates (lipids, proteins and DNA) being converted in S-adenosylhomocysteine (SAH), a strong competitive inhibitor of methyltransferases. The SAM/SAH ratio is an important indicator of cellular methylation potential (Chiang et al., 1996). Several studies have found positive correlations between plasma HCY concentrations and plasma amyloid β (Aβ) levels (Irizzarry et al., 2005) suggesting a possible involvement of HCY metabolites in amyloidogenesis. SAM, the main methyl donor for methyltransferase enzymes, appears to be altered in some neurological disorders, including AD (Bottiglieri and Hyland, 1994). Data suggest that HCY might exert its toxicity in vascular disease and AD by inhibiting methylation activity (Lee and Wang, 1999, Miller, 2003, Kennedy et al., 2004). Such hypothesis is consistent with our previous in vitro findings on the relation of Aβ processing and methylation activity, showing that the alteration of HCY metabolism is associated with PS1 (Presenilin 1) promoter methylation status and with Aβ production, (Scarpa et al., 2003, Fuso et al., 2005) and provide the basis for the theory that nutrition-related impairments of SAM dependent methylation activity might promote the pathogenesis of AD (Ulrey et al., 2005, Tchantchou et al., 2005).

Amyloidogenesis is a central feature of neurodegeneration in AD. Although the causes of abnormal Aβ overproduction and aggregation in AD are far from clear, the roles of γ- and β-secretase mechanisms in producing toxic Aβ have been deeply studied (Edbauer et al., 2003, Takasugi et al., 2003, Iwatsubo, 2004, Capell et al., 2005) and the physiological functions of Aβ are increasingly understood (Tanzi, 2005, Mattson, 2004). In this work, we test in vivo the possibility that HCY metabolism could be associated with the regulation of γ- and β-secretases and with Aβ production.

We previously developed an in vitro model of neuroblastoma cells deficient of vitamin B (folate, B12 and B6), which we used to study the relation of altered DNA methylation to the expression of genes that are involved in amyloidogenesis. Folate and B12 deprivation inhibits remethylation of HCY to methionine, whereas B6 deprivation inhibits HCY clearance through the trans-sulfuration pathway (Hajjar, 2001, Medina et al., 2001). In cells cultured with vitamin B deficient medium, we found lower SAM concentrations and concomitant PS1 and BACE (Beta-site APP Cleaving Enzyme) up-regulation, with consequent overproduction of Aβ (Fuso et al., 2007). The observed demethylation of a specific CpG site in the gene promoter indicated that PS1 up-regulation could be due to the metabolic impairment of methylation reactions, whereas in contrast to PS1, we did not detect abnormal methylation of the BACE gene. Unfortunately, in these in vitro experiments, we couldn't measure HCY levels and SAH concentration; for this reason, the present in vivo work also gives a more complete view of the HCY metabolism under B-vitamin deficiency.

In the present study we used an in vivo approach to evaluate the potential pathogenic role of hyperhomocysteinemia induced by dietary deprivation of vitamin B (folate, B12 and B6), allowing us to test the hypothesis that vitamin-deficiency induced imbalances in homocysteine and brain methylation metabolism in vivo (indicated by plasma HCY and brain SAM and SAH concentrations) would be associated with changes in the regulation of key amyloidogenic genes, as observed in vitro. We chose to use the transgenic TgCRND8 mouse strain (Janus et al., 2000), which overexpress an APP (Amyloid Precursor Protein) gene containing the Swedish and the Indiana FAD mutation and show an early and age-related Aβ production and plaque deposition in the neocortical and hippocampal regions by 12 weeks of age (Chishti et al., 2001). In these mice, cognitive impairment (in terms of water maze deficits) parallels the onset of beta amyloid plaque deposition (Hyde et al., 2005). Moreover, it was useful for our purposes that these mice did not carry mutated or overexpressed PS1 or BACE, since, on the basis of the previous in vitro experiments, these genes were the main target of our study and we wanted to study their regulation. We chose not to use genetic models of hyperhomocysteinemia since they are not known to express detectable levels of Aβ and because dietary B-vitamin deficiency is a more common cause of human hyperhomocysteinemia than those genetic defects that are expressed in currently available mouse models.

TgCRND8 mice and wild type littermates were fed either on a vitamin B (folate, B12 and B6) deficient diet or a control diet starting from the third postnatal week. Neurochemical effects of vitamin B deprivation were evaluated after 45 and 60 days on: i) HCY, SAM and SAH in plasma and brain and cytosine and methylcytosine in brain DNA; ii) RNA and protein levels for APP, PS1, BACE and Notch1; iii) assembly of γ-secretase complex and levels of Aβ 1–40 and 1–42; and iv) intra- and extracellular Aβ deposition by immunohistochemistry. Finally, the behavioral effects of a shorter (18 days) vitamin B deprivation were evaluated in the pre-symptomatic phase (8 weeks) in another group of mice using a modified version (1-day protocol) of the water maze spatial learning task. The aim of this second experiment was two-fold: (i) to assess the effect of the deprived diet per se on spatial acquisition performances, and (ii) to verify if the deprived diet could accelerate the appearance of the learning impairment in mice carrying the APP mutation. Diet exposure in this group of animals was limited to 18 days to avoid that the decrease in body weight gain induced by vitamin B deprivation could in turn affect the motor abilities required for performance of the learning task.