Authors:  Yunlong Xu, Chenyu Jiang, Junyan Wu, Peidong Liu, Xiaofei Deng, Yadong Zhang, Bo Peng, Yingjie Zhu
Journal: CNS Neuroscience & Therapeutics, Volume 28: Issue 4; Pages 580-592 (Published December 10th, 2021)
DOI: https://doi.org/10.1111/cns.13779

Alzheimer’s disease (AD) is a terrible disease resulting from the accumulation of extracellular amyloid-beta plaques. The build-up of plaques around the neurons causes the neurons to lose connectivity to other neurons, lose conductivity of the electrical signal, and eventually die. Our brain is made up of many neurons, some of which are connected in a particular pattern, and these connections facilitate thinking, coordination, and emotion (and much more). Neurons communicate with each other by sending chemical signals, but an electrical signal from the "sending" neuron is required to facilitate the release of these chemical messengers.

To briefly highlight the neuroanatomy involved, a neuron can be stimulated by a variety of changes, including pH, pressure, and temperature. An example of this would be the sensation of touch. A neuron, or touch receptor, that sits near the surface of the skin (they have specialized names but for the sake of simplicity) can be activated when pressure is applied to the skin. This pressure stimulates the neuron to send an action potential (electrical signal) through its axon and eventually into its axon terminal, a region of the neuron that contains machinery responsible for releasing neurotransmitters. This electrical signal causes the release of a neurotransmitter into the synaptic cleft, a tiny gap between neighboring neurons. This neurotransmitter will diffuse across the synaptic cleft and bind to its respected receptor. This binding can cause this postsynaptic neuron to become excited or inhibited, which will result in the neuron firing or not firing an action potential, respectively. There is a series of neurons and connections that allow the sensation of touch from all over the body, and neurons eventually fire at a part in the brain called the primary somatosensory cortex, where the electrical signals are converted into the perception of touch. All sensory information enters the somatosensory cortex and begins to be processed. This perception of touch happens in the brain, not the skin.

In the case of AD, plaques build up around neurons. These plaques can act like a net. Imagine you are trying to throw a ball to your partner but you are surrounded by a net. There is only one open portion of the net. You can throw the ball a hundred times and maybe you will make it to your partner. The same happens when your partner throws you the ball; it is harder to receive the catch. Another way to think about these plaques is trying to swim through a viscous liquid like honey. These plaques are not only influencing the intensity and frequency of the incoming signal but are influencing the ability to send signals as well.

These plaques also recruit immune cells, which can release inflammatory molecules and cause inflammation in the brain. As stated in earlier posts, acute inflammation against a pathogen is incredibly beneficial. However, chronic inflammation caused by the plaques leads to an increase in neuronal apoptosis, or programmed cell death, thus leading to a reduction in neuron number. Adding to the example earlier, you are trying to swim in honey, and now you must add in big rocks. These are the immune cells. Also, you must add a terrible migraine. So not only are you trying to swim back and forth in honey, but there are rocks in the honey, and you have a terrible migraine. This scenario keeps getting worse, and, of course, there are additional factors that must be considered when accurately representing AD. These include the age of onset, history of illnesses, diet, etc.

AD is characterized by cognitive decline and memory loss. The build-up of amyloid-beta plaques and the subsequent impact on neuronal connectivity and neuroinflammation is the most accepted mechanism of how AD influences cognition. The ketogenic diet has been found to reduce neuroinflammation in multiple animal models, suggesting that a KD could alleviate neuroinflammation in AD patients and improve cognition. The KD influences gut microbiota composition and a variety of metabolic parameters that have an impact on many neural circuits, including GABA, dopamine, and serotonin.

This journal article investigates the impact of dietary intervention on AD progression in mice. While this has not been investigated in humans to this extent, animal models provide valuable data and direction for future clinical trials. To properly test the influence of the ketogenic diet on AD, these mice must develop AD. The researchers used the standard mouse model of AD, the 5XFAD model. 5XFAD mice are genetically programmed to overexpress the human amyloid-beta precursor protein. These mice also have mutations in the familial Alzheimer’s disease (FAD) and human presenilin 1 (PS1) genes. These genetic modifications cause accelerated AD at around 2 months of age. The advantage that this journal article offers over other articles involving KD and AD is the investigation of cognition. In theory, if a KD reduces neuroinflammation and improves cognition in AD patients, then a KD may benefit anyone suffering from mild to severe neuroinflammation.

The researchers fed the 5XFAD mice a standard KD consisting of a high amount of fat, low carbohydrates, and moderate protein. To be exact, the researchers used 76% fat, 16% protein, and 3% carbohydrate chow. The control diet (SD-standard diet) consisted of 12% fat, 23% protein, and 65% carbohydrate. Mice in the standard diet group were given chow ad libitum. To control for caloric restriction or surplus, the KD mice were fed an equal amount of ketogenic chow in terms of calories. This is extremely good control, thus allowing the researchers to compare the KD and SD at the same number of calories. An issue with the methodology comes when analyzing the data and setting up another control group. The researchers used three groups: 5XFAD mice fed a KD, 5XFAD mice fed a SD, and normal mice fed a SD. A group containing normal mice fed a KD would have allowed for another comparison between groups. With this being said, the goal of this study was not to study the impact of the KD in wild-type mice, but the data would have been useful for analysis.

The goal of this article was to investigate the impact of a dietary change (KD) on cognition in an AD model. To test for cognition, multiple field tests can be used. A field test is a constructed area, like a box, with objects or a maze. These field tests allow the researcher to observe the mice's ability to navigate a novel area, interact with their environment, and maintain memory. While field tests used to be purely qualitative data, software designated for these tests has been developed, thus allowing quantification of field test data. At 7 months of age, a group of 5XFAD mice switched from a SD to a KD, while a group of 5XFAD mice remained on a SD, and a group of wild-type (normal=no AD) mice remained on a SD.

The researchers then used a habituation technique to analyze memory recall and spatial learning between groups. This habituation technique involves allowing the mouse to be exposed to the Barns maze daily for 8 consecutive days. The maze is a cylinder consisting of holes in which one hole is the escape hole, and the time to find the escape hole is deemed the latency period. When compared to AD groups fed either a KD or a SD, the wild-type group fed a SD consistently had a lower latency period. While the wild-type mice performed significantly better than AD mice fed a KD, the AD mice fed a KD performed significantly better than AD mice fed a SD, suggesting that the KD helps restore memory loss in AD.

Another field test used was the T maze, which allows for the analysis of working memory. In brief, the T maze consists of two steps. In step one, a T-shaped maze is revealed, with one arm of the T containing food and the other arm blocked. There is only one way for the mouse to go. The second step is oftentimes referred to as the free-choice step because it removes the block. Instead of the food being placed in the same spot as it was in step one, the food is now in the arm that was blocked. If the mouse chose to enter the arm that previously held the food (step one), then the response was marked as incorrect. If the mouse chooses the previously blocked arm, which now houses the food pellet, then the response is marked as correct. There was no significant difference between wild-type mice and AD mice fed a KD, while there was a significant decrease in the correct response percentage in AD mice fed a SD. These results suggest that four months of KD in AD mice allows for full recovery of memory recall.

The last field test evaluated anxiety and locomotion. The open field test is an open area in which the mouse is allowed to freely explore. Data can be captured on a camera and put into a program where the total distance traveled and time in the center of the area can be calculated. The total distance traveled is a marker for locomotion, while time in the center is used to test anxiety. An anxious mouse (or person) will spend more time around the edges of a room when compared to a less anxious mouse. The open field test is not a totally reliable test for anxiety due to the introduction of an unfamiliar area and isolation from the group.

To evaluate structural differences between the treatment groups, immunohistochemical staining can be used. Immunohistochemical staining uses antibodies that bind to a specific antigen and contain a fluorophore. An antigen can be anything from genetic material to a protein on the cell’s surface. Companies can synthesize antibodies to attach to specific molecules in/on the cell. When you incubate the cells with these fluorescently-labeled antibodies, the antibody will bind to the target molecule. You can then place the cells under a fluorescent microscope and the antibodies’ fluorophore will emit light. The emitted light will allow us to highlight where the target molecule is present. To learn more about this, please refer to this YouTube video.

In this paper, the researchers wanted to visualize neuronal density in several regions of the brain. To do this, they synthesized antibodies that bind to synaptophysin, a protein found in presynaptic vesicles, and contain a fluorophore. There was a significant reduction in the measured fluorescence of the hippocampus, a region in the brain involved in memory, in the AD mice fed a SD, while there was no significant difference between wild-type mice fed a SD and AD mice fed a KD. This trend was also observed when the researchers stained PSD-95, a postsynaptic protein. Next, the researchers decided to stain for neuronal density in the lateral entorhinal cortex (LEC), a region in the brain that acts as a gatekeeper for the hippocampus, and the somatosensory cortex, a region in the brain involved in receiving and processing signals related to the senses. While neuronal density in the LEC was significantly lower in both AD mouse groups compared to wild-type mice, AD mice fed a KD had significantly higher neuronal density than AD mice fed a SD. Interestingly, the somatosensory cortex neuronal density was significantly reduced in AD mice fed a SD when compared to the wild-type group and AD mice fed a KD. There was no significant difference in somatosensory cortex neuronal density between the wild-type mice and AD mice fed a KD. While extensive staining was not performed on a variety of brain regions, the researchers demonstrated that the KD protects from the loss of neurons in the hippocampus and somatosensory cortex. Likewise, it also seems that in mice with AD, the KD results in significantly higher neuronal retention in the LEC compared to the SD.

The most commonly accepted mechanism for the onset of symptoms associated with AD is amyloid-beta plaque buildup. The researchers wanted to explore if this plaque was reduced in mice fed a KD. To do this, similar immunohistochemical techniques were used as before, but this time the target was the plaque itself. The staining revealed that AD mice given a KD had significantly less plaque than AD mice given a SD, but significantly more plaque than the wild-type group. This is to be expected because dietary intervention is not going to reverse disease onset, especially in a disease like AD. There is essentially no plaque build-up in wild-type mice, but in AD mice, switching to a KD has been shown to reduce plaque build-up.

The last piece of information investigated by the researchers was inflammation-related. An enzyme-linked immunosorbent assay, or ELISA, is a common technique used to quantify the amount of a specific molecule in a solution. There are several types of ELISAs, but the easiest to describe is the sandwich ELISA method. Firstly, the well is coated with a primary antibody. This primary antibody is specific to the target molecule we are looking to detect. When you add your sample, the target molecule will bind and stick to the antibody. Thus, when you wash out your solution, you are left with your target molecule bound to the primary antibody. After removing your solution, a secondary antibody solution is added. This secondary antibody binds to the target molecule as well. However, this secondary antibody contains a conjugated enzyme. When you add in a solution containing a substrate, like horseradish, if there are a lot of bound secondary antibodies, there will be a drastic color change. The more target molecule that is present in the solution, the more bound secondary antibody, which leads to a more intense color change.

This journal article did ELISAs using several inflammatory markers, including IL-1𝛽 and TNF-𝜶. These molecules are released by activated M1 macrophages, an immune cell, and promote inflammation. A hallmark of the KD in previous studies is the decrease in C-reactive protein, which is indicative of reduced inflammation. To further test this previous knowledge in this newer model, the researchers used ELISAs to compare the concentrations of the two inflammatory molecules between treatment groups. In the hippocampus, the AD mice fed a SD had significantly elevated IL-1𝛽 and TNF-𝜶. When compared to each other, AD mice fed a KD and wild-type mice fed a SD had no significantly different levels of either molecule. The results in the cortex of the brain tell a similar story: AD mice fed a SD had significantly higher levels of IL-1 and TNF- when compared to both wild-type mice fed a SD and AD mice fed a KD. In summary, the progression and onset of symptoms associated with AD are correlated with increased neuroinflammation. Mice with AD who were fed a SD had significantly higher levels of inflammation, which could lead to more plaque buildup and worsening cognition. On the other hand, mice with AD but fed a KD displayed moderate plaque build-up but not significantly increased levels of inflammation when compared to wild-type mice, so these mice scored higher on multiple cognition tests.

Alzheimer’s disease is a tremendously excruciating disease associated with loss of cognition and memory. The mechanism of why a build-up of amyloid-beta plaque results in these symptoms is two-fold. First, a buildup of plaque directly impacts neuronal signaling and connectivity. Secondly, amyloid-beta plaques trigger the immune system, thus mounting an attack against the plaque. The problem is that this overactivation of the immune system causes the release of cytotoxic molecules which damage the surrounding neurons, thereby affecting cognition and memory. Previous studies have demonstrated that dietary interventions such as reducing caloric intake and reducing carbohydrate load improve inflammatory status. This foundational research has shown scientists that a large hole exists which investigates the effect of diet on disease states.

According to the researchers, a KD decreased amyloid-beta plaque build-up, neuroinflammation, and neuron death in mice with AD. The researchers did not include a wild-type group fed a KD, so they could not compare wild-type mice fed a SD with wild-type mice fed a KD. Although this information would have been valuable for the field, it is outside the scope of their aim. Because the researchers did not include this control group, a verifiable conclusion that the KD may reduce inflammation and neuron death in healthy mice cannot be reached.

Disclaimer: 
This blog post was written before the amyloid beta hypothesis data has been fully investigated. Recently, insight about a fundamental paper written in 2006 has been found to contain tampered data. In light of this, the amyloid beta hypothesis for treating Alzheimer's disease needs to be further investigated. While there are multiple mechanisms for Alzheimer’s disease and the onset of symptoms associated the information in this journal review still contains vital information for treating the disease.