Team:UCL/Background/Neuropathology

WHAT CAUSES ALZHEIMER'S?

Well, It's Debatable

There are many vying hypotheses that postulate how Alzheimer’s Disease (AD) may arise. Of these, the most well known is the ‘Amyloid Hypothesis’. The Amyloid Hypothesis is centred around the ‘senile plaques’ that form in AD brains, and suggests that their removal could be key in halting the progression of the disease (Hardy and Allsop 1991), though other hypotheses contradict this precept. So far, no single cause for the disease has been identified, though most AD research focuses around senile plaques. AD is generally accepted to cause three major histopathological changes in brain tissue.

Histopathology

Senile plaques are extracellular deposits of an abnormal form of the waste protein β-amyloid (Aβ), which tangle with cell matter in the brain (Selkoe 2001). These plaques are larger than cell bodies (15-25 um in diameter) and mature to become even denser. The tangling of the plaques, however, is not proportional to the amount of amyloid proteins in an area, and so the process by which the tangling and the plaques are created are still as yet unknown. Aβ is cleaved from a larger precursor protein - amyloid precursor protein (APP) (Nistor et al. 2007). The Aβ peptide is predominantly cleaved to be 40 amino acids in length, that is, Aβ1-40. However, Aβ1-42 and Aβ1-43 nucleate more rapidly into amyloid fibrils than Aβ1-40 does, and are neurotoxic via unknown mechanisms. Strittmatter and Salvensen theorised that other proteins complexed with Aβ to exacerbate the plaque, mainly ApoE, which is another cerebrospinal fluid protein that has a high affinity for Aβ (Srittmatter et al. 1993). The functional forms of ApoE aid protease-mediated degradation.

The second pathological feature is the collection of intraneuronal cytoskeletal filaments called neurofibrillary tangles, due to paired helical filaments (Selkoe 2001). These abnormal tangles are made up of poorly soluble, hyperphosphorylated isoforms of Tau, a microtubule-binding protein that is normally soluble. Functional Tau acts as a part of the cell’s ‘cytoskeleton’, which forms the structural support network of a cell. Its dysfunction disrupts the cytoskeleton, making it harder for the cell to carry out essential survival tasks, and engendering cell death. It is thought that tangle formation is aided and perhaps caused by the senile plaques (Hardy and Selkoe 2002). Plaques and tangles predominantly appear in brain areas involved in learning, memory and emotional behaviours.

The third sign is cell death. Cell death is most apparent in the neocortex, limbic structures, hippocampus, amygdala, and some of the brainstem nuclei. It is cell death that directly causes the symptoms of AD.

Genetics

Studying the genetics of AD has uncovered key genetic risk factors. While highly heritable, AD is genetically complicated, associated with multiple genetic factors, making genetic analysis difficult.

Down’s Syndrome is caused by having three copies of chromosome 21, causing an extra copy of the APP gene (Nistor et al. 2007). This may increase production of beta-amyloid, triggering the chain of biological events leading to AD. Early onset AD is a component of Down Syndrome, indicating that defects in chromosome 21 can lead to Alzheimer’s disease independently of Down’s syndrome. Heritable early-onset AD can also be caused by mutations in the genes presenilin 1 and presenilin 2, which modify how APP is processed. Late onset AD is not necessarily inherited, though relatives of those with AD are at greater risk. Again, why this should be is not fully understood.

The gene ApoE has three versions called ‘alleles’, e2, e3 and e4. The frequency of e4 is five times higher in AD patients than in the general population (Ertekin-Taner 2007). Those that have two copies (homozygous) for this allele have as much as 8 times higher chance for developing AD. However, inheriting e4 does not cause AD, only increases its likelihood.

Not all cases of AD are thought to be genetic (Ertekin-Taner 2007). Both genetic and non-genetic AD cases could arise from the activation of a type of brain cell receptor, p75NTR, by local sources of a protein called pro-NGF. This can initiate what is known as ‘tetraploidisation’ - genetic information doubles as it must for cell division, even though the cell does not divide. This generally occurs when pro-NGF outcompetes a survival factor called BDNF does, in older brains where the brain cells are more stressed. It has been suggested that this process can increase the dosage of genes responsible for the onset of AD, and so heralds in the disease-state (Frade & Lopez-Sanchez 2010).

Amyloid Hypothesis

It has been more than two decades since the first postulation that AD may be caused by deposition of Aβ in plaques in brain tissue (Hardy and Selkoe 2002). The ‘Amyloid Hypothesis’ rose as a seemingly strong idea that threaded together genetic and histological data. It posits that concomitant signs and symptoms of the disease arise directly and indirectly due to plaques that appear in quantity in older brains and the genetically susceptible, due to an imbalance in Aβ deposition and clearance (Hardy and Selkoe 2002). Our project is mainly, but not completely, built around assuming the Amyloid Hypothesis is correct.

This imbalance leads to plaque-formation and then a cascade of macromolecular and cellular events which eventually culminate in dementia. In the typical AD brain, there is 7 years worth of un-cleared amyloid production from a healthy person deposited around the brain. Aβ is neurotoxic, and it has been suggested that this toxicity is due to diffusible forms of the protein permeating the areas around plaques, which are somewhere between the insoluble form and the non-pathological soluble form. They have subtle effects upon the chemical connections (synapses) between communicative brain cells called neurons, which impairs information flow (Hardy and Selkoe 2002). Neurons are further harmed by oxidative stress (Su et al. 2008) and neuroinflammation. This is caused by plaques, because they produce oxygen free radicals and other stress agents. Neuroinflammation occurs because the resident immune cells of the brain, mircoglial cells, are attracted to plaques. It seems that early on in the disease they help clear amyloid and reduce symptoms, but at later stages the inflammation caused by their activation is far more harmful.

The formation of intracellular protein (Tau) tangles is thought to occur downstream of plaque formation in the ‘amyloid cascade’ (Hardy and Selkoe 2002), though the two do not correlate in a linear fashion. Their ratios vary patient to patient. Tau tangles are also capable causing cell death on their own, as in frontotemporal lobe dementia, in which there are no protein deposits. In AD, deposits may help form tangles by altering neurons’ kinases and phosphatases (proteins that modify the functions of other proteins by taking away or adding a small ‘phosphate’ group) activities, making Tau hyperphosphorylated.

This hypothesis suggests that clearing amyloid, reducing its production or stopping its aggregation ought to halt the progression of AD. However, several therapeutics purported to deplete Aβ production/aggregation have failed in Phase III clinical testing (Mullane and Williams 2013). Some treatments trialled in animal models appear to remove vast quantities of plaques without effect. This may be because AD is caused at a certain amyloid threshold and that excess plaques have no further effect (Hardy and Selkoe 2002). Alternatively, it may be due to once the plaques have set AD in motion, it continues, and so plaques need to be removed at an earlier stage in the disease.

Other Hypotheses

There are many other theories behind the causation of AD. They are not generally mutually exclusive and it may be that most are correct. The key problem is that it is hard to distinguish the primary cause from secondary effects. Here we give a brief overview of some of these theories:

Neurotrophins - Neurotrophins are chemicals that affect neurons. An imbalance in their activity could trigger AD. They activate two classes of receptors - tyrosine receptor kinases (Trks), whose activation supports the neuron, and p75NTR, which may activate an apoptotic pathway. Over-expression of p75NTR and pro-NGF may be caused by oxidative stress (Su et al. 2008) in later life, possibly due to plaques. This leads to an increase in AD susceptibility gene dosage by initiating cell cycle re-entry without division. This generally occurs when pro-NGF signals to neurons more than another neurotrophin called BDNF does, in older brains where the brain cells are more stressed. It has been suggested that this process can increase the dosage of genes responsible for the onset of AD, and so heralds in the disease-state (Frade & Lopez-Sanchez 2010). BDNF acts to prevent apoptotic death, but this only means that these cells are fated to a slow death process and steady degeneration.

Neurotransmitters and signalling - AD patients have lower levels of various neurotransmitters that are believed to influence intellectual functioning and behavior, such as acetylcholine (Francis et al. 1999). The cause of this reduced production or something blocking their action may underlie a part of the AD pathology, for example, chemical imbalances or the great toxicity from heavy metals and homocysteine. AD may also be caused by dysfunction in ion channels proteins, which underlie the generation of suppression of electrical transmission in neurons (Francis et al. 1999).

Blood supply - A poor blood supply in older brains could damage neurons and impair their functioning, leading to the formation of plaques and tangles.

Viral infections - Viruses have been known to cause brain disorders that symptomatically resemble AD. It is conceivable that an infection could cause the onset of AD by initiating an neuroinflammatory response that is detrimental for neurons, predisposing the brain to the disease-state (Itzhaki and Wozniak 2008).

Neuroinflammation - It is thought that inflammatory processes could increase the production of waste products such as amyloid in cells, leading to the formation of plaques, then indirectly causing tangle formation (Streit 2006). Anti-inflammatory medication given to patients with conditions other than AD has resulted in a lower contraction of AD amongst these patients. Inflammation also damages cells directly. The microglial cells recruited to the plaques in AD may actually cause cell death by becoming activated and inflaming an area (Hensley 2010), while the plaques themselves have a comparatively minor toxic effect. Producing ‘vasoactive intestinal peptide’ in our circuit will act to de-active microglia around the plaque, subsequently reducing neuroinflammation.

Tau - Tau tangles could be the direct cause of cell death (Mudher and Lovestone 2002). Plaque build up could be the secondary event, with a comparatively minor toxic effect.

More Soluble Amyloid - More soluble, smaller abnormal pieces of amyloid could cause AD, while the plaques themselves may actually serve a protective role, by bundling excess amyloid together in discrete locations (Castellani et al. 2010) where, in early stages of the disease they can be removed by microglia. The plaques may even have an antioxidant, neuroprotective role. They could be constructs typical of end-stage downstream processes triggered by oxidative stress, cell cycle re-entry, inflammation, etc. MMP-9, the protease in our circuit, degrades soluble and insoluble amyloid.

A complete understanding of the underlying AD molecular mechanisms is vital for the creation of novel treatments able to modify the disease-state biology and efficiently combat the increase of AD with age in global society’s increasing life expectancy. Therefore, a much more effective system than the potential one we propose under the 'circuit overview' section of this website could be achieved with an entirely new suite of genes, as more light is shone on the pathogenesis of AD.