Memory is mysterious. How and why we are able to maintain some memories while others are ephemeral is a worthy, albeit tricky scientific pursuit. It is also what Stowers Scientific Director Kausik Si, Ph.D., has devoted his career toward researching.

The human brain is composed of approximately 86 billion neurons that communicate with each other through synapses. Each neuron may be associated with hundreds or thousands of synapses: trillions of potential messages and exchanges.

Memory formation is a process where interactions between neurons via synapses in the brain interpret, store, and recall information from experiences. For example, a fruit fly must remember where to find food or else it will die. Human memory is much more complicated, yet likely shares fundamental molecular processes with the memories of animals elsewhere in the evolutionary tree of life.

The phenomenon of memory is extraordinarily complex and difficult to study. However, after many decades of research, scientists have developed ways to characterize different aspects of memory, for example, by duration (short or long), by system (conscious or unconscious), and by kind (rewarding or punishing).

Most memories begin from a sensory experience, or how our brain responds to sensations such as sight, sound, touch, smell, and taste. Whether these are retained very briefly like the bark of a dog or a whiff of perfume, or if they are converted, memories depend on various factors, many of which are still not understood.

Nearly all memory can be classified as explicit (conscious) or implicit (unconscious), irrespective of their temporal duration. These memories are stored in different areas of the brain yet share similar molecular mechanisms for memory storage.

Memorizing a piece of music, your phone number, or your first kiss are examples of explicit memory, the conscious recollection of facts or events. These are memories we deliberately took the time and effort to form. Learning how to play an instrument or how to ride a bike are examples of implicit memory: although deliberate, they are retained without conscious awareness. Think of the age-old adage, “It’s like riding a bike.”

It is also useful to classify memory by duration. Repeating a phone number several times in order to dial it, and then promptly forgetting it is an example of working memory. This is briefly retained information required to perform and complete tasks.

Imagine taking an evening stroll and greeting an adorable puppy along the way. This is an example of short-term memory, the recall of an experience for a limited time period (around a minute). Riding a bike, playing an instrument, or knowing that there are 52 cards in a deck are examples of long-term memory.

Following maturation, most nerve cells cannot proliferate, or divide, to create new cells. From a molecular perspective, short-term memory storage is a result in the modification of existing neuronal proteins and synaptic connections while long-term memory requires the synthesis of new proteins and connections.

Different types of memories are stored within different structures of the brain. The amygdala stores emotional responses like fear. The hippocampus is where explicit memories are located, while the striatum houses those implicit bike-riding skills. Both the temporal lobes and hippocampus play a variety of roles in forming and recalling memories.

There are distinct forms of memory loss, ranging from forgetting something temporarily, to amnesia, to the progressive loss of memory, or dementia.

Amnesia refers to the loss of memories—facts, information, and experiences—and typically results from a form of brain trauma, stroke, excessive alcohol consumption, and even PTSD. However, amnesia differs from dementia in that memory loss is not progressive.

There are many forms of dementia, yet many share the build-up of stable aggregations of proteins within the brain that are associated with toxicity and neuronal cell death; it is this slow accumulation that accounts for the progressive nature of dementia.

Different proteins throughout the body are required for normal cellular function. Proteins are continuously generated via translation of messenger RNA transcribed from DNA. Before a protein is functional, it is first synthesized one amino acid at a time prior to folding into a unique three-dimensional shape that governs its role.

However, proteins within cells can unfold and fold into different conformations, or shapes. Some of these configurations can induce surrounding proteins to adopt the same shape, stick together (aggregate), and form very stable, permanent structures called amyloids.

Amyloids have been investigated—and blamed—for hundreds of years as pathogenic, but more recently some have been identified as “functional amyloids,” which have normal, healthy, required roles within cells. The Si Lab investigates functional nervous system amyloids, and in 2020, published a study in Science resolving the structure of amyloid at the atomic level from the brains of fruit flies. These were directly implicated in the formation of long-term memory.

Essentially, very similar amyloid structures can form within synapses in the brain. Whether they serve as templates to hold memories or act to destroy neuronal cells is protein dependent. Thus, memory maintenance has a molecular basis.

Alzheimer’s disease is a debilitating decline in memory that can lead to emotional and behavioral problems, and eventually results in death. It is the most common type of dementia with approximately 50 million people worldwide currently living with the disease; that number is expected to triple by 2050.

The exact causes for Alzheimer’s are aggressively debated. There appears to be some component that is genetic—particularly for the early-onset form—yet environmental factors including lifestyle and education can raise or lower one’s risk of developing the disease.

Smell, or the olfactory system, is the focus of Stowers Investigator Ron Yu, Ph.D. Smell as a sensory memory, and the closely related sensation of taste seem more likely to become long-term as their endurance may be a mechanism of survival. Animals need food, at least until we figure out how plants do it. And, the inability to smell is one of the first symptoms in neurodegenerative diseases that result in memory loss.

Like all forms of dementia, Alzheimer’s is associated with the assembly of intricately ordered aggregates of a protein, in this case a fragment called Aβ42 from amyloid-beta precursor protein (APP) interacting with other related genes.

Other forms of dementia that do not seem to have a concrete correlation with genetics include Parkinson’s disease, frontal lobe dementia, and ALS. Also, professional football players are susceptible to chronic encephalopathy (CTE) caused by repeated concussions; symptoms are similar to Alzheimer’s, Parkinson’s and ALS, are progressive, and not treatable.

On the other hand, Huntington’s Disease has a strong genetic component. Within the huntingtin protein, the presence of a sequential strand of the amino acid glutamine above a threshold value (36) guarantees disease development.

Huntington’s and eight other neurodegenerative conditions are collectively called PolyQ—the Q is the symbol for glutamine. Research from Stowers Associate Investigator Randal Halfmann, Ph.D., is actively investigating the stochastic, or random, and energy-intensive first, rate-limiting step for amyloid called nucleation. In theory, the lab’s methodology can elucidate this ordered core structure for any amyloid.

Memory formation, and in particular long-term memories, in fruit flies and mice are supported by intricately ordered protein amyloids. Rather than causing disease, these three-dimensional structures are in fact required for maintaining a memory and thus serve as functional amyloids. Currently, the Si Lab team is examining whether the same memory-amyloid relationship occurs in the human brain.

If scientists can uncover a precise mechanism for how functional, memory-maintaining amyloids form, and since memories can fade, in theory it should be possible to elucidate how these amyloids can be taken apart. If the same principles hold for amyloid that leads to memory loss, it may be possible to figure out how to take pathogenic ones apart as well.

As we all know, memory can be fickle. While certain memories seem to be ingrained in our very sense of identity, other very memorable experiences frequently disappear. The Si Lab is investigating whether the departure of a memory is associated with the dissolution of the amyloid substrate securing it.

The seemingly random formation or dispersal of amyloid may yet reveal itself to be a tightly regulated process.