Disruption of blood-brain barrier function by chronic intake of saturated fat and cholesterol : implications for Alzheimer’s disease risk

Abstract

It has been reported that lifestyle including diet is associated with Alzheimer’s disease (AD) risk and progression. Population studies indicate that the chronic consumption of diets enriched in saturated fats (SFA) and cholesterol significantly increase the risk of AD onset and progression. However, the mechanisms underlying the association of AD risk with dietary fat intake are presently unclearProteinaceous deposits enriched in amyloid-β (Aβ) within the cerebral parenchyma (amyloid plaque) and in the cerebrovasculature (cerebral amyloid angiopathy) are the hallmark pathological features of AD. Several animal and cell culture studies suggest that high-fat diets exacerbate amyloidosis by promoting Aβ secretion by neurons and increasing the propensity for oligomerization to occur. However, there is little evidence consistent with cerebral Aβ overproduction in AD. Rather, recent studies in animal models of AD suggest that efflux of Aβ relative to its delivery from the blood is pivotal to cerebral Aβ homeostasis.Two lines of evidence led me to develop the hypothesis that dietary fats may influence AD risk by modulating cerebrovascular exposure to circulating Aβ (Paper1 presented as Literature review of thesis). Firstly, Aβ has potent vasoactive properties and blood vessels treated with exogenous Aβ show substantial structural damage. Moreover, exaggerated plasma Aβ could occur because of chronic ingestion of diets enriched in fats. Dietary SFA were found to significantly increase Aβ abundance within the absorptive cells of small intestine (enterocytes) and thereafter, substantial plasma Aβ remains associated with triglyceride rich lipoproteins (TRLs). It is my contention that cerebrovascular integrity is compromised by the ingestion of fats which increases the plasma concentration of Aβ.An immunohistochemical approach was developed to explore the effects of dietary SFA and cholesterol on cerebrovascular integrity and Aβ kinetics at the bloodbrain barrier (BBB) (Paper 2 presented as Chapter 2 of thesis). Wild-type (WT) mice were used for the dietary intervention studies and appropriate comparisons were made with amyloid precursor protein/presenilin-1 (APP/PS1) amyloid transgenic mice, an established model of AD (Paper 3-5 presented as Chapters 3-5). Critical to the primary scientific objectives, three-dimensional colocalization analysis using double immunofluorescent microscopy was developed (Paper 2). This double immunofluorescent labelling technique enabled the simultaneous detection of two proteins utilizing polyclonal antibodies derived from the same species. Briefly, in order to avoid the cross-reactivity of two polyclonal antibodies that originate from the same species, the concentration of one of the primary antibodies was reduced, so that it was undetectable with conventional secondary antibody methodologies. Rather, avidinbiotin amplification that was specific for the diluted primary antibody was utilized to identify its specific immunoreactivity. The double labelling of proteins with certainty that cross-reactivity did not occur, enabled consideration of protein distribution and association within tissues and cells.In some cells Aβ is generated following processing of a precursor protein (APP) embedded within the plasma membrane. However, Chapter 5 of this thesis (Paper 5) shows that within enterocytes of the upper small intestine, Aβ genesis occurs within the Golgi apparatus and is likely secreted associated with nascent postprandial lipoproteins (chylomicrons). Apolipoprotein B immunoreactivity was used as a surrogate marker of enterocytic chylomicron distribution as it is an obligatory component of these macromolecules.Evidence that plasma derived apo B lipoproteins containing Aβ may contribute to the aetiology of AD is presented in Chapter 3 (Paper 3). Consistent with the hypothesis presented, APP/PS1 mice were previously reported to have significantly increased secretion of TRL-Aβ as a consequence of the genetically induced overexpression of Aβ. However, this study expanded on that finding and explored if there was evidence of blood-to-brain delivery of apo B and if this was positively associated with amyloid plaque distribution and abundance. In transgenic APP/PS1 mice, immunoreactive apo B was detected in the core and periphery of cerebral amyloid plaques with significant colocalization coefficience (Manders’ overlap coefficient = 0.85±0.004). The findings are consistent with the notion that cerebrovascular exposure to plasma TRL-Aβ is causally associated with cerebral amyloidosis.Chapter 4 (Paper 4) investigates the differential effects of dietary fatty acids on BBB integrity. WT mice were fed either low-fat (LF) control chow, or physiologically relevant diets enriched in SFA, monounsaturated fatty acid (MUFA) or polyunsaturated fatty acid (PUFA) for either 3 or 6 months. Blood-to-brain delivery of apo B was found in SFA fed mice and exaggerated with a longer duration of feeding. The distribution of cerebral apo B in SFA fed mice, closely paralleled with the distribution of Aβ, consistent with blood-to-brain delivery as a lipoprotein complex. The cerebral extravasation of apo B was more evident in the cortex region (CTX) than in hippocampal formation (HPF). Mice fed the LF control, MUFA or PUFA diets for 3 or 6 months showed no evidence of apo B/Aβ parenchymal extravasation. The cerebral distribution of immunoglobulin G (IgG) was used as a surrogate marker of non-specific plasma protein leakage into the brain.In SFA fed mice alone, significant peri-vascular leakage of IgG was observed, suggesting that the endothelial dysfunction induced by SFA feeding was a non-specific or leakage phenomenon. Consistent with the latter, plasma S100B, a marker of brain-to-blood protein kinetics, was significantly increased in SFA fed mice but not in LF control, MUFA or PUFA supplemented mice. SFA group mice also had significantly attenuated occludin-1 expression, the primary BBB endothelial tight junction protein. Apo B and IgG extravasation greater in CTX than in HPF, increased plasma S100B, and decreased occludin-1 abundance were also observed in APP/PS1 amyloid transgenic mice. Hence the findings in SFA fed mice are consistent with a causal role in cerebral amyloidosis.The data presented in this thesis and consideration of other studies reported particularly since commencement of my candidacy are then presented in Chapter 5, which was published as a review in Progress in Lipid Research (Paper 5). Briefly, several studies suggest that significant plasma Aβ is associated with TRLs secreted by the small intestine as chylomicrons and from the liver as very low density lipoproteins. Evidence presented in this thesis of apo B colocalization within amyloid plaques is consistent with the concept that plasma derived lipoprotein-Aβ may be causally associated with cerebral amyloidosis. However, for delivery of lipoprotein-Aβ from blood to brain to occur, the breakdown of BBB function would be required as the cerebrovasculature architecture normally prevents transport of large macromolecules such as lipoproteins.In this study, chronic consumption by WT mice of food enriched in SFA resulted in non-specific leakage of plasma proteins within the brain parenchyma, analogous to that observed in an established transgenic murine model of AD (APP/PS1 mice). Within the review, dietary cholesterol is shown to elicit the same response. The mechanisms by which SFA and cholesterol cause the BBB dysfunction are presently unclear. Increased BBB exposure to circulating TRL-Aβ is one possibility, however, there was no significant difference in fasting plasma Aβ in mice maintained on SFA or cholesterol supplemented diets compared to LF control. Postprandial transient increases in plasma TRL-Aβ, not necessarily detected in fasting blood, may have been sufficient to cause the BBB dysfunction, but this was not specifically investigated. Alternatively SFA and cholesterol may have compromised cerebrovascular integrity via Aβ independent mechanisms, many of which are considered in depth in the review article. The review manuscript also discusses the potential mechanisms that could contribute to the extracellular retention of apo B lipoproteins enriched in Aβ and the inflammatory sequelae that may ensue thereafter. There is a positive association of apo B/Aβ retention with the abundance of the heparin-sulphate proteoglycans; perlecan, biglycan and decorin. A number of studies show that apo B lipoproteins are avidly metabolized by inflammatory cells and indeed under certain conditions may trigger the inflammatory cascade. In the review article, this paradigm is discussed in the context of apo B lipoproteins containing Aβ and putative interaction with activated glial cells.Collectively, the results presented in this thesis suggest that dietary SFA and cholesterol may increase the risk of AD and/or accelerate the progression of disease by compromising cerebrovascular integrity and promoting the cerebral delivery of Aβ from the blood. The findings support the contention that diet is an important consideration in the context of disease prevention, but also raise the intriguing notion that nutritional intervention approaches could be developed to treat AD.