Cryogenic Electron Microscopy Helps Clarify Synapses Bear No Resemblance to Information Storage Structures

We can classify several different types of scientific truth claims, along with some tips on how to recognize the different types. 

Type of truth claim	How to recognize it
Citation of established fact	Typically occurs with a discussion of the observational facts that proved the claim.
Citation of a claim that is not yet established fact	Typically occurs with phrases such as “scientists believe” or “it is generally believed” or an appeal to a “scientific consensus.” The claim of a “scientific consensus” is often unfounded, and there may be many scientists who do not accept the claim.
Citation of a claim that has little basis in observations, and that there may be good reasons for doubting	Often occurs with a phrase such as “it is widely believed,” or maybe a more confident-sounding phrase like “it is becoming increasingly clear” or “there is growing evidence.”

Claims that memories are stored in synapses fall into the third of these categories.  Such claims often are made using the weak-sounding phrase "it is widely believed." To show that, I may cite some of the many times in which writers or scientists suggested that memories are stored in synapses, and merely used the weak phrase "it is widely believed" as their authority. 

• "It is widely believed that synaptic plasticity mediates learning and memory"  (link). 
• "It is widely believed that synapses in the forebrain undergo structural and functional changes, a phenomenon called synaptic plasticity, that underlies learning and memory processes" (link).
• "It is widely believed that synaptic modifications underlie learning and memory" (link).
• "As with other forms of synaptic plasticity, it is widely believed that it [spike-dependent synaptic plasticity] underlies learning and information storage in the brain" (link).
• "It is widely believed that memories are stored as changes in the number and strength of the connections between brain neurons, called synapses" (link).
• "It is widely believed that modifications to synaptic connections – synaptic plasticity – represent a fundamental mechanism for altering network function, giving rise to phenomena collectively referred to as learning and memory" (link).
• "It is widely believed that encoding and storing memories in the brain requires changes in the number, structure, or function of synapses"  (link).
• "It is widely believed that long-term changes in the strength of synaptic transmission underlie the formation of memories" (link).
• "It is widely believed that the brain's microcircuitry undergoes structural changes when a new behavior is learned" (link).
• "It is widely believed that long-lasting changes in synaptic function provide the cellular basis for learning and memory in both vertebrates and invertebrates (link).
• "It is widely believed that the brain stores memories as distributed changes in the strength of connections ('synaptic transmission') between neurons" (link).
• "It is widely believed that the long-lasting, activity-dependent changes in synaptic strength, including long-term potentiation and long-term depression, could be the molecular and cellular basis of experience-dependent plasticities, such as learning and memory" (link).
• "It is widely believed that a long-lasting change in synaptic function is the cellular basis of learning and memory" (link).
• "It is widely believed that the modification of these synaptic connections is what constitutes the physiological basis of learning" (link).
• "It is widely believed that memory traces can be stored through synaptic conductance modification" (link).
• "It is widely believed that memories are stored in the synaptic strengths and patterns between neurons" (link).
• "It is widely believed that long-term changes in the strength of synaptic connections underlie learning and memory" (link).
• "It is widely believed that long-term synaptic plasticity plays a critical role in the learning, memory and development of the nervous system" (link).
• "It is widely believed that learning is due, at least in part, to long-lasting modifications of the strengths of synapses in the brain" (link).
• "It is widely believed that long-term memories are stored as changes in the strengths of synaptic connections in the brain" (link). 
• "It is widely believed that activity-dependent modification of synapses is the brain's primary mechanism for learning and memory" (link).
• "It is widely believed that synaptic modifications are one of the factors underlying learning and memory" (link).
• "Learning, it is widely believed, is based on changes in the connections between nerve cells" (link).
• "It is widely believed that memories are stored as changes in the number and strength of the connections between brain cells (neurons)" (link).
• "It is widely believed that memories are stored as changes in the strength of synaptic connections between neurons" (link). 
• "It is widely believed that memory formation is based on changes in synapses" (link).

All such claims are weak claims, and are not even claims of a scientific consensus or a majority opinion. Something can be widely believed when much less than a majority of experts believe in it, as long as maybe 10% or 20% of the experts believe in it. There is no good evidence that any memories are stored in synapses or stored through a strengthening of synapses or stored by a modification of synapse weights, or stored anywhere in the human brain through any means.  We know that humans remember things, but do not know that brains store or retrieve memories. 

No one has any understanding or any credible coherent theory of how learned information or episodic memories could ever be stored using synapses or any other part of the brain. We know of the strongest reason for rejecting all of the claims in the bullet list above, which is that the average lifetime of the proteins in synapses is only about two weeks or less.  The proteins in synapses last an average of only about a thousandth of the longest length of time that humans can remember things (50 years or more). Moreover, humans can form permanent new memories instantly, which could never occur if forming such memories required synapse strengthening (something that would take quite a few minutes or hours, because it would require the synthesis of new proteins). 

An interesting question is: does the latest and greatest technology offer any support for claims that memories are stored in synapses? Cryo-electron microscopy is a rather recently developed technique involving examining tissue stored at super-cold temperatures. We read here that "Cryo-electron microscopy (cryo-EM) single particle analysis (SPA) is a technique for reconstructing the three-dimensional structure of a biomacromolecule using projected images acquired with an electron microscope and was the subject of the Nobel Prize for Chemistry in 2017." A 2020 article is entitled "Cryo–electron microscopy breaks the atomic resolution barrier at last." We read this:

" Now, for the first time, scientists have sharpened cryo-EM's resolution to the atomic level, allowing them to pinpoint the positions of individual atoms in a variety of proteins at a resolution that rivals x-ray crystallography's. 'This is just amazing,' says Melanie Ohi, a cryo-EM expert at the University of Michigan, Ann Arbor. 'To see this level of detail, it's just beautiful.' Because the heightened resolution reveals exactly how complex cellular machines carry out their jobs, improvements in cryo-EM should yield countless new insights into biology."

Neuroscientists have long claimed that memories are stored in synapses, even though there is no robust evidence to support such a claim. I suspect that claims that synapses store memories arose mainly because:

1. synapses are extremely numerous in the brain, with there being multiple synapses for every neuron (the page here refers to a ratio of about 4 to 1, although other sources suggest a much higher ratio);
2. synapses are much smaller and harder to observe than neurons. 

Item 2 is very important for anyone claiming that memories are stored in synapses, because the harder something is to observe, the more you can get away with speculating that such a thing contains something that no one one has observed in it or on it. But now it is getting easier and easier to precisely observe the contents of synapses, because of technology such as the cryo-electron microscopy discussed above.

Let us look at some of the cryo-electron microscopy studies that have been done on synapses, and examine whether they offer any support for speculations that memories are stored in synapses.  A 2018 paper is entitled "Differentiation and Characterization of Excitatory and Inhibitory Synapses by Cryo-electron Tomography and Correlative Microscopy." We have some very strong photos and diagrams that seem to clarify the exact contents of synapses. 

Figure 8 (photo A) shows an actual photographic closeup of a synaptic terminal, and it looks like the circled closeup in the diagram below, except for two differences:

1. Instead of seeing five microtubules, as in the circled part of the diagram, we seem to see only one.
2. Instead of seeing only 9 of the circles that are synaptic vesicles, as in the diagram below,  we see about 100 in the photo, and about 200 in a diagram clarifying the contents of the photo. In the diagram and in the photo, the position of these vesicles seems as random as the position of soap bubbles in a bubble bath. 

What are these vesicles that make up most of the end of the synapse? They are merely holes filled up with neurotransmitter chemicals. Scientists are convinced that these vesicles are very short-lived. A paper tells us, "In mature neurons, synaptic vesicles continuously recycle within the presynaptic nerve terminal." 

It is interesting that the 2018 paper discussing the closeup physical details of synapses makes no substantive reference to memory. The only references to memory are the two vague insubstantial "lip service" mentions below:

• "As key functional units in neural circuits, different types of neuronal synapses play distinct roles in brain information processing, learning, and memory."
• "The existence of multiple functional and plasticity states in excitatory synapses could be critical for optimal learning and memory storage in neuronal circuits, as suggested by theoretical studies (Fusi et al., 2005; Fusi and Abbott, 2007)."

None of the details of synapses mentioned in the paper do anything at all to suggest information storage. Similarly, the 2022 paper "Digitalizing neuronal synapses with cryo-electron tomography and correlative microscopy" deals with microscopic examination of synapses using the latest and greatest technology. But it makes no mention of memory other than to say, "Therefore, deciphering the molecular and structural basis of synaptic transmission and changes in its plasticity is a core focus of research attempting to understand learning, memory, and the pathogenesis of brain diseases." No discussion is made of anything found in synapses bearing any resemblance to information storage. 

There are three categories that some matter might have regarding information:

Category 1: The matter might have no sign of storing any information.

Category 2: The matter might seem to have some storage of information, but information that cannot be currently deciphered by someone examining it.

Category 3: The matter might have information that a person can read and understand. 

The average rock is an example of Category 1. The paragraphs of this blog or the genes in DNA are examples of Category 3. You can read particular paragraphs of this blog and understand what they signify. A scientist can read some gene within DNA, and understand the chemical information that is being specified, such as which amino acids make up a particular protein. 

An example of Category 2 is shown below. You see what clearly looks like some type of encoded information, but you cannot understand what the information is. The visual below shows two of the hallmarks of stored information: (1) the repetition of a limited number of tokens or symbols; (2) an arrangement of such tokens in a linear sequence. There are only three tokens used: a dot, a dash, and a slash. A person looking at the visual below will get the impression that particular combinations of these tokens seem to stand for particular things, but a person will be unable to understand how the coding system works. If all of these dots, dashes and slashes were instead to exist in a bucket, then we would no longer have a sequential arrangement of tokens, and it would be much less likely that any real information storage was going on.  The text below is one of the paragraphs of this post, written in Morse Code. 

Neuroscientists have found no Category 3 information in brains corresponding to memory information. For example, it is impossible to examine the brain tissue of someone who recently died, and read some information that the person learned during his life. Another important fact is that scientists have not even found any Category 2 information in the brain that could be memory information. Scientists cannot even find anything in synapses that looks like encoded information using some code that scientists have not been able to crack. 

Synapses look no more like something storing encoded information than clouds look like something storing encoded information. Just as clouds don't seem to be storing any symbols and lack any structure that might allow any linear sequence of symbols to be stored,  synapses don't seem to be storing any symbols, and lack any structure that might allow any linear sequence of symbols to be stored.  Clouds are unstable things that don't persist for more than a few hours or days. Synapses are a little more stable, but there is no reason to believe that a synapse can last for years. Synapses are built of proteins with lifetimes of less than two weeks. Synapses are connected to units call dendritic spines, which typically have lifetimes of only weeks or months, and do not last for years. 

Here is a quote from the recent paper "Synaptic plasticity in human cortical circuits: cellular mechanisms of learning and memory in the human brain?" : "Direct evidence that synaptic plasticity is the actual cellular mechanism for human learning and memory is lacking." The paper tells us that "only a very modest number of studies of long-term plasticity of human synapses exist to date."

Since mere strengthening is not any such thing as a storage of complex information, the idea that memory is stored through synapse strengthening never made any sense. The idea started out with a speculative slogan, the claim that "neurons that fire together wire together."  That isn't a theory of memory; it's merely a rhyme. A real theory of neural memory would require explaining things such as instant memory recall, instant memory formation and the 50-year persistence of memories in old people, none of which today's neuroscientists can explain in any credible manner. The difference between the mere slogan of "neurons that fire together wire together" and a full detailed theory of memory occurring physically in the brain would be like the difference between a single couplet and all the poetry of Dante's Divine Comedy. And when scientists make the vacuous claim that memories are stored by synapses strengthening, they are not advancing a workable theory of memory, but are merely parroting a little sound bite.  Nothing in synapses looks anything like stored encoded information. 

When stored information is encoded in the body, and read, that leaves quite a big "footprint." Chemical information is read from DNA, information about which amino acids make up a protein. The machinery of such information reading and decoding leaves quite a big "footprint" that scientists have been able to detect. Many molecules serving such a task have been identified. Were there to exist some apparatus for reading memory information from synapses and storing human learned information in synapses, it would have quite a big "footprint" that would have allowed scientists to discover it long ago. No sign of any such memory-reading mechanism or memory encoding mechanism has been found in the body. No such thing has been found in synapses or anywhere else. The job of translating all of the many types of things humans can learn into synapses states or neural states (and also decoding such encoded information) would be the most gigantic job requiring thousands of dedicated proteins, but no such proteins can be found in the human body. We know from analysis of old DNA that humans have no proteins they did not have thousands of years ago. So could there ever be a protein handling the job of translating into neural or synapse states remembered words written in an alphabet that is only about two thousand years old?  It seems not.  This point is explained more fully in my post here.