An Overview Of Alzheimers Disease Biology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Alzheimers Disease (AD) is the most commonly diagnosed senile dementia (Alzheimer's Association, 2011). Initially, AD is characterized by mild cognitive impairment (MCI) effecting declarative and spatial memory that progresses to severe cognitive dysfunction and neuronal atrophy over the course of several years. Initial memory problems are associated with cell lose in the temporal lobes, which contain the hippocampus which is crucial to declarative and spatial memory function. In particular, the hippocampus is the first brain area to show hallmark signs of AD pathophysiology, which include the deposition of senile plaques (SP). SP are protein complexes comprised of insoluble amyloid beta (Aβ) protein aggregates. The gradual deposition of SP in AD suggests a causal connection between SP deposition with neuronal atrophy and progressive cognitive impairments. The purpose of this project is to examine the role of oligomeric Aβ (Aβo) aggregates in route to SP formation in relation to MCI in early stages of AD.

Amyloid-beta Precursor Protein Processing

The amyloid precursor protein (APP) is a type 1 transmembrane protein up to 770 amino acids long and has ten known isoforms (De Strooper & Annaert, 2000). Post-translational modification of APP by a variety of secretase complexes produces Aβ monomers of variable length. In particular, secretases differentially attach to and degrade hydrogenated bonds between residues on APP. The process is either amyloidogenic or non-amyloidogenic, being respectively initiated by β-secretase and α-secretase.

β-secretases are amyloidogenic proteinases which cleave APP at the N-terminus of the Aβ domain (Hong et al., 2004). Consequently, ectodomain shedding results in release of a soluble N-terminal APPβ fragment and a membrane bound C-terminus protein containing the entire Aβ domain. Proteolysis of the remaining C-terminal fragment by γ-secretase results in the production of clinically relevant Aβ alloforms. Additionally, α-secretases are non-amyloidogenic proteinase which cleave within the Aβ domain of APP (Lammich et al., 1999). Consequently, ectodomain shedding results in release of a soluble N-terminus APPα fragment containing part of the Aβ domain and a membrane bound C-terminal fragment. The C-terminal fragment is then proteolysed by γ-secretase to produce clinically irrelevant and truncated Aβ alloforms.

Amyloid-beta and Plaque Formation

Normally, the Aβ(1-40) alloform is most commonly produced at a rate of 10:1 to the Aβ(1-42) alloform (Alzheimer's Association, 2011, need better reference). Research indicates that Aβ(1-42) is more amyloidogenic than Aβ(1-42) since it is prone to misfolding and polymerizing with other Aβ monomers (Baukmetner et al., 2006; Collins et al., 2004; Kirkitadze & Kowalska, 2005; Yang & Teplow, 2008). In general, the structure of Aβ alloforms are amorphous, though there are certain characteristics of Aβ(1-42) that are connected to amyloidosis (i.e., SP formation and deposition). In particular, the 41 and 42 residues support additional hydrogen bonds to the central hydrophobic cluster of the Aβ monomer. These additional bonds results in a metastable C-terminal β-hairpin structure and a decrease in the configurational entropy of the protein. Subsequently, the Aβ(1-42) C-terminal may function as a seeding point for polymerization with other Aβ monomers. Initiallyyy, seeding results in the formation of oligomeric Aβ (Aβo) composed of multiple Aβ monomers which further polymerize to form strands ofibrilar Aβ that ultimately constitute SP (Kirkitadze & Kowalska, 2005; Yang & Teplow, 2008; Walsh et al., 1997).

Genetic Link between Amyloidosis and Alzheimer's-like Memory Deficits

Substantial insights into the causes of AD were facilitated by the discovery that SP in Down's Syndrome (DS) are constituted by Aβ proteins homologous to those found in AD (Podlisny, Tolan & Selkoe, 1991). DS is caused by trisomy of chromosome 21 which encodes APP. DS causes the over-production APP which subsequently initiates amyloidosis through the overproduction of Aβ(1-42). Since amyloidosis occurs in both diseases and DS patients invariable develop progressive Alzheimer's-like memory impairments, it is logical to assume a common mechanism of neurodegeneration between the diseases is amyloidosis caused by deleterious alterations in APP processing. In support of this claim, researchers have found autosomal dominant APP and presenilin mutants in families with a history of early onset AD (reference).

In AD, there are multiple sites on APP and presenilins that are prone to non-synonymous mutations. APP and presenilin mutations alter the activity or affinity of secretase complexes for specific residues of APP (reference). For example, the iconic Swedish mutation has two non-synonymous substitutions on residue 670 and 671 which are adjacent to the β-secretase cleavage site. Consequently, Swedish mutants have a stronger affinity for β-secretase which increases the production of clinically relevant alloforms, including Aβ(1-42). In the Flemish mutation, there is a single substitution on residue 692 which is adjacent to the α-secretase cleavage site. Consequently, the Flemish mutant has a lower affinity for α-secretase which thereby increases the activity of β-secretase and the production of clinically relevant alloforms, including Aβ(1-42). Additionally, multiple mutations can be found adjacent to the γ-secretase cleavage site which decreases overall γ-secretase activity while increasing the specific production of Aβ(1-42). Furthermore, presenilin mutations are more frequent than APP mutations and account for approximately 50% of early onset AD cases. There are two presenilins in humans which constitute part of the γ-secretase complex. The presenilins are not identical to one another but are very similar and almost always have deleterious mutations on homologous protein exteins. In relation to AD etiology, presenilin mutations are similar to the Flemish mutation in that they result in decreased γ-secretase activity and increased Aβ(1-42) production.

The Plaque Theory of Alzheimer's Disease

Motivated by the genetic evidence linking Aβ(1-42) production with SP formation, the initial SP theory hypothesized that AD was caused by SP toxicity (Irie et al., 2005). In support of this claim, it has been found that Aβ toxicity is dependent upon aggregational state (reference). In particular, SP are said to attach to otherwise healthy neurons and initiate immune responses through agonism of the toll-like receptors (Jana et al., 2008; Takeda & Akira, 2004). In theory, as the amount of SP increases, the rate of cell lose is exacerbated by chronic and aberrant immune response.

While it is clear that SP contribute to cell death in AD, it is not clear how SP account for MCI in early stages of the disease. In particular, researchers observe that MCI in adults and transgenic animals occur before significant SP deposition in the temporal lobes (Catalano et al., 2006; Drake et al., 2003). Therefor, the connection between SP mediated immune responses and MCI is incomplete. Attempts to elucidate this relation have lead researchers to test immunization therapies against specific Aβ alloforms or SP (McLaurin et al., 2002; reference). In theory, if clinically relevant Aβ alloforms can be degraded before polymerization, then vaccinations against Aβ alloforms would directly inhibit amyloidosis and deter AD onset and or progression. Alternatively, if plaques could be degraded in a similar fashion, there would be an overall decrease in SP which would inhibit the induction of deleterious autoimmune responses. Researchers have found that MCI in transgenic animal models of AD are preventable by vaccinations against Aβ alloforms but not specifically against SP. If the plaque theory were true, improvement in cognitive functions should be the consequent of inhibiting the autoimmune response caused by SP. The opposite suggests that Aβo in route to SP formation may be the primary neurotoxins responsible for MCI in early stages of AD.

Glutaminergic Signaling and Synaptic Plasticity Mechanisms

Recent research has found that Aβo are potent neurotoxins in-vitro and in-vivo which cause synaptic disruptions and increase pro-death signaling (Catalano et al., 2006; Hu et al., 2008; Lecanu et al., 2006; Martins et al., 2008; O'Hare et al., 1999; Ramin et al., 2010; Renner et al., 2010; Tomic et al., 2009; Walsh et al., 2002). Considering that LTP in the hippocampus is activity dependent, recent research has focused upon studying the interaction between Aβo and glutaminergic signaling.

Glutamate is the primary excitatory neurotransmitter in the central nervous system (Anderson, 2007). Glutaminergic signaling is essential for certain types of synaptic plasticity such as long-term potentiation (LTP) and long term depression (LTD). LTP is a well known form of synaptic plasticity, though the exact contributions of LTP to memory is unclear. In particular, LTP causes physical changes at synapses which increases the number and strength of connection between neurons.

There are several types of glutamate receptors which are either ionotropic or metabotropic. Ionotropic glutamate receptors (iGluR) tend to directly induce LTP or LTD, while metabotropic glutamate receptors (mGluR) are coupled to second messengers known to regulate iGluR channel conductance, cytosolic calcium concentrations, transcriptions factors and cytoskeletal morphology.

Essential for synaptic plasticity are the iGluR receptors α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) and n-methyl d-aspartate (NMDA) receptors. AMPAR is a ligand gated ion channel containing four extracellular binding domain for AMPA. Agonism of AMPAR causes a sodium ion pore to open in the cell membrane. Upon entering the cell, sodium ions depolarize the membrane and increase the probability of excitatory postsynaptic potentials (EPSP). NMDAR has three extracellular binding sites for NMDA, the co-agonist glycine and variety of allosteric modulators. The NMDAR ion channel is similar to the AMPAR channel except that it is not selective for particular types of ions, meaning that the NMDAR channels allow for the influx of any and all ions. Additionally, NMDAR is distinct from AMPAR in that its membrane domain contains residues that readily form ionic bonds with magnesium ions. At rest, the NMDAR ion channel is blocked by a magnesium ion which can only be removed by depolarization of the cell membrane sufficient to elicit EPSP. For this reason, NMDAR is both ligand and voltage gated. When glycine and glutamate bind to their respective domains, a conformational change occurs which creates a pore in the cell membrane. If the cell membrane is not sufficiently depolarized the magnesium ion continues to block the channel, thereby inhibiting unabashed influx of ions into the cell. It is important to note that NMDAR ion influx is graded. In particular, the channel conductance of NMDAR responds to gradual changes in membrane potential, meaning that when the magnesium block is in place and an ESPS is not evoked, certain ions, like calcium, diffuse across the channel but at variable rates relative to the voltage of the cell membrane.

In regards to synaptic plasticity, LTP is controlled by changes in cytosolic calcium concentrations. When the magnesium block is removed from NMDAR, calcium ions enter the cell at sufficient concentrations to activate calcium-calmodulin-dependent kinase II which induces LTP by phosphorylating specific residues of intracellular AMPAR. Phosphorylation of AMPAR subunits increases AMPAR channel conductance, and as a consequent, the AMPAR ion channel allows more ions into the cell making agonism of AMPAR more likely to elicit EPSP and remove the magnesium block on NMDAR. Additionally, calcium activates various protein kinases and transcription factors which upregulate the expression of AMPAR. Consequently, the relationship between AMPAR and NMDAR in LTP induction is reciprocal, such that AMPAR initially depolarizes the cell causing EPSP, which opens NMDAR and increases cytosolic calcium concentrations enough so to upregulate AMPAR.

LTD is the complement of LTP and decreases the strength of an association between cells having undergone LTP. LTD occurs at cytosolic calcium concentrations below the threshold needed to induce LTP. The general rule of thumb at excitatory synapses is that connections between neurons only strengthen during EPSP, otherwise the default condition is that unused connections between cells are gradually pruned through the induction of LTD. Calcium concentrations below the threshold for LTP induction activate phosphatases which induce the endocytosis of AMPAR. Activated phosphatases additionally dephosphorylate residues on the intracellular domain of AMPAR which decrease AMPAR channel conductance and make AMPAR agonism less likely to elicit EPSP. Therefore, LTD is best viewed as a negative feedback mechanism which counters LTP through the desensitization and downregulation of AMPAR.

Of particular interest to AD researchers are the group 1 mGluRs found primarily on post-synaptic densities (PSD). Group 1 mGluR are coupled with the Gq second messenger which activates phospholipase C (PLC). PLC cleave phospholipids such as phosphatidylinosital biphosphate into inositol triphosphate (IP3) and diacylgycerol (DAG). IP3 is a ligand for receptors located on the plasma membrane of the endoplasmic reticulum. When activated, IP3R transport calcium ions into the cytosol and contribute to calcium dependent signaling cascades such as LTP and LTD. Furthermore, DAG contributes to protein kinase C activity which is implicated in actin reorganization. Actins are proteins which polymerize and form scaffolds that integrate adapter proteins into the cell membrane. In particular, scaffolds provide binding sites for receptors at the cell membrane and therefor contribute to trafficking of glutamate receptors to and from the cell membrane. Agonism of group 1 mGluR5, especially metabotropic glutamate receptor subtype 5 (mGluR5), has been show to induce LTD and activate factors within the mitogen-activated protein kinases (MAPK) transcription pathway such c-Jun N-terminal protein kinase (JNK), which are strongly associated with capsase-8 dependent apoptosis.

Implications for Glutaminergic Signaling in Alzheimer's Disease

In relation to AD, researchers have suggested that Aβo alter the trafficking of mGluR5 at synapses (Renner et al., 2010). Aβo have been found to interact with adapter proteins (e.g., 95kDa post synaptic density adapter protein or PSD-95) which express mGluR5 at synapses. Normally, proteins diffuse randomly across the cell membrane until they interact with adapter proteins that bind and localize proteins to a specific area of the cell. Synaptic transmission is dependent upon localization of receptor proteins, since only certain parts of the cell facilitate synaptic transmission between cells (i.e., synapses and junctions). Researchers have found that Aβo initially diffuse randomly across cell membranes until reaching synapses where they cross-link and cluster mGluR5. Consequently, Aβo toxicity is possibly the consequent of aberrant stabilization and over-activity of mGluR5 at synapses enough so to induce LTD and capsase-8 dependent apoptosis.

Therefore, MCI in early stages of AD can be accounted for by exacerbated IP3R and JNK signaling. In particular, researchers have found that even a single infusion of Aβo is accompanied by increased JNK activation and reversible spatial memory deficits attenuated using a JNK inhibitor (Ramin et al., 2010). Additionally, research shows that infusions of Aβo almost invariable result in delayed, as opposed to spontaneous, spatial memory deficits suggesting gradual cell-mediated death. The remaining question to answer is the degree to which the induction of JNK signaling is mediated through over-activity of mGluR5 caused by Aβo crosslinking. In particular, JNK is known to associate with a variety of adapter proteins raising the possibility that Aβo may activate JNK signaling in a manner that is ancillary to mGluR5 activity (Fanger et al., 1997).

In order to test the theory that Aβo exerts neurotoxicity by overactivity of mGluR5, the following experiment will test the efficacy of a selective mGluR5 antagonists in attenuating spatial memory deficits in rats acutely exposed to Aβo. Research suggests that allosteric mGluR5 antagonists may inhibit mGluR5 and Aβo cross-linking altogether. Additionally, decreasing the intrinsic activity of mGluR5 should result in decreased calcium and JNK signaling thereby ameliorating Aβo dependent memory deficits in early stages of AD.



21 male Long-Evan's rats are housed three per cage under a 12h light/dark cycle with food and water available ad lib. At 4-6 weeks of age, rats are randomly divided equally into three groups and given ear tags for identification. All behavioral procedures are derived from Ramin et al. (2010) and are performed in accordance with IACUC regulations.

Preparation of Amyloid-beta Oligomers

1 mg of Aβ(1-42) monomers will be dissolved in phosphate buffered solution (PBS) at 200 ng/ul and incubated at 37C for 5 days. Following incubation, the working concentration is titrated to 10ng/ul by the adding PBS. Aβo are then stored at -20C until use.


Rats are anesthetized with ???? and placed in a stereotaxic frame. The site of surgical incision is shaved, disinfected and anesthetized with lidocaine. An incision is made down the midline from anterior to posterior. The skull is cleaned, bleached and incision points are marked in accordance with coordinates from the atlas of Paxino and Watson (?) for subregion CA1 of the hippocampus. Three holes are then drilled in the skull and stainless steel cannula are placed bilaterally into region CA1 of the hippocampus. Cannula are secured using a jewelers screw and dental cement. Finally, the wound is clipped and rats are closely monitored until awake. Upon having regained ventral recumbency, rats are returned to the housing room and are checked daily for health for three days post-surgery. Post-operative analgesia will be provided by free access to a water bottle containing acetaminophen (dosage ?).

Treatment Groups

Controls (Aβo-) receive an injection of 10% Tween 80 administered intraperitoneally concurrently with anesthesia for stereotaxic surgery and a bilateral intra-CA1 injection of PBS (3ul/side) following surgery. The second group (Aβo+) receive an injection of 10% Tween 80 administered intraperitoneally concurrently with anesthesia for stereotaxic surgery and a bilateral intra-CA1 injection of Aβ (30ng/3ul PBS/side) following surgery. The final group (Aβo+MTEP) receive an injection of 3-((2-methyl-4-thiazolyl)ethynyl)pyridine ([email protected]/kg) dissolved in 10% Tween 80 administered intraperitoneally concurrently with anesthesia for stereotaxic surgery and a bilateral intra-CA1 injection of Aβ (30ug/3ul PBS/side) following surgery

Morris Water Maze Procedure

Nineteen days following the last injection, rats are placed in the Morris Water Maze (MWM) and allowed to free swim for a total of 60s to habituate to training conditions. On day 20, each rat will begin behavioral testing on MWM. MWM consists of a fairly large tub of water and is has been used successfully to assess spatial memory deficits in transgenic and non-transgenic animal models of AD (Vorhees & Williams, 2006). A small platform is located in the maze and is occluded from the view of subjects (Morris, 1984). In each trial, rats swim for no more than 90 seconds to find the escape platform. If a rat fails to find the platform, they are placed on or guided to the platform. Once upon the platform, rats are allowed to remain for 20 seconds. The rat is then removed from the platform and placed in its home cage for 30 seconds until the start of the next trial. Over a total of eight trials, rats start in each quadrant of maze twice determined in a quasi-random fashion. Twenty four hours following the last training session, the rats are subjected to a probe trial where the hidden platform is removed. All subjects start in the same quadrant and are allowed to free swim for a total of 90 seconds. All behavior will be tracked using automated software and a camera located above the apparatus.

Statistical Analysis and Expected Results

Eight training sessions will be divided and averaged into two blocks of trials, i.e., early and late training (Ramin et al., 2010). A repeated measures analysis of variance (RMANOVA) will be used to test average performance across blocks by treatment group. Performance will be measured in a variety of ways including escape latency, path length, swim speed, time spent in target quadrant, number of inner zone crossings, and trends in directionality. Escape latency is the amount of time it takes the animal to find the hidden platform (Morris, 1984; Vorhees & Williams, 2006). Time in target quadrant indicates the amount of time spent in the quadrant containing the platform. Number of inner zone crossing refers to an area surrounding the platform. Finally, directionality refers to a variety of search behaviors that the animal can use to learn the location of the platform. For example, at the start of trials rats tend to exhibit thigomotaxis. Thigomotaxis is roughly defined as swimming in a circle around the outside portion of the maze. As the animal learns the task, the amount of time in thigomotaxis deceases as new search behaviors are adopted to find the platform more effectively. Following a significant effect, post-hoc comparisons will be performed to determine specific differences between groups.

For probe trials, a one-way analysis of variance (ANOVA) will be used to determine main effects and interactions between treatment groups. Post-hoc comparisons will determine statistical differences on a variety of measures including time in quadrant, time in zone, number of platform crossings, average proximity to platform, and a novel entropy-based measure. Time in quadrant is as previously described. Time in zone refers to the amount of time spent in the previous platform area and the inner area previously described. Number of platform crossings refers to the number of times the animal returns to the area where the platform was previously located. Proximity is the average distance the animal is from the previous location of the platform. Entropy is defined as the pooled variance of proximity to the platform location and error in the swim path across two dimensions regardless of the platforms location

(Maei et al., 2009)

If this project is successful, results should indicate that animals in the Aβo+ group will show worse spatial memory performance as compared to both Aβo- and Aβo+MTEP groups. In particular, Aβo- and Aβo+MTEP group will have statistically significantly differences between early and late testing blocks, whereas Aβo+ will not. On probe trials, Aβo+ performance will perform significantly worse on all measures as compared to Aβo- and Aβo+MTEP groups. Results would indicate that comparisons between Aβo- and Aβo+MTEP groups are not significantly different by block or by the magnitude of the difference scores across blocks, whereas the differences between Aβo- or Aβo+MTEP should be significantly different from Aβo+.


If the experiment goes as planned, it will offer additional insight into a receptor-based explanation of MCI in early stages of AD. Such a finding would be beneficial to the development of novel therapeutics designed to inhibit the interaction between Aβo and glutaminergic signaling pathways. The approach has shown particular promise in the treatment of other neurodegenerative diseases which are similar in pathophysiology to AD (Lea & Faden, 2003). Alternatively, if the desired results are not forthcoming, the experiment will indicate that a specific receptor based theory is insufficient explaining MCI in early stages of AD. Either way, the following experiment will extend a line of current AD research in a way that has yet to be been done.