A couple of things about memory

Children's Memory May Be More Reliable Than Adults' In Court Cases

                           Researchers Valerie Reyna, human development professor, and Chuck Brainerd, human development and law school professor--both from Cornell University--argue that like the two-headed Roman god Janus, memory is of two minds--that is, memories are captured and recorded separately and differently in two distinct parts of the mind.

They say children depend more heavily on a part of the mind that records, "what actually happened," while adults depend more on another part of the mind that records, "the meaning of what happened." As a result, they say, adults are more susceptible to false memories, which can be extremely problematic in court cases.

The implications of these results for legal testimony is not what I find especially interesting here. In fact, there are reasons why the testimony of children has sometimes been found to be less reliable than that of adults. Namely, in some cases, the techniques used to interview the children (before trial) have been improperly coercive or suggestive of particular interpretations.

What does seem interesting is the hypothesis that in adults memories of the same event tend to be stored in two distinct forms: literal details of "what happened", and interpretive judgments about the "meaning" of an event. But that in children it is primarily the actual details that are stored.

Reyna and Brainerd's Fuzzy Trace Theory hypothesizes that people store two types of experience records or memories: verbatim traces and gist traces.

Verbatim traces are memories of what actually happened. Gist traces are based on a person's understanding of what happened, or what the event meant to him or her. Gist traces stimulate false memories because they store impressions of what an event meant, which can be inconsistent with what actually happened.

The researchers have experimental evidence to support their conclusions. Some of this is noted in earlier accounts, such as this:

Children Less Prone To False Memories, Implications For Eyewitness Testimony, Study Shows

                                           In a study published in the May issue of Psychological Science, Brainerd and Reyna presented a list of words for groups of first, fifth and ninth graders. Many of the words from this "study list" were related to each other (by belonging to certain categories such as animals, furniture, men's names) while others were unrelated "filler" words.

After a short break, the students were presented with a new "test list" composed of study list words, new words belonging to the aforementioned categories (animals, furniture, etc.), and distracter words that were new and entirely unrelated to the categories or the study list. Their task was to identify whether they had previously heard a word or not.

As predicted, if the test list provided a new word with a closely related meaning (a "semantic relation") to a word from the study list, older children were more likely to assert that they had heard it before. Simply put, the older children had more false memories in this case than younger children.

One can speculate about what's going on here. As people mature through childhood, they are constantly learning about the interrelationship of isolated details and events. (For instance, "Dad acts more scary after he's been drinking beer.") In addition, the accumulation of details makes more literal forms of memory cumbersome (and liable to confusion), so people learn to make abstractions and interpretations that summarize details and make storage easier by associating similar details in more general categories. However, this kind of fuzzy storage (or "fuzzy traces" as Brainerd and Reyna call it) can misrepresent the facts. (For instance, "Dad was drunk when he hit me" – which might not actually be true.)

Second, and not directly related to this, there are two quite indepentdent studies that show something about the relationship between memory and experience of stress.

The first item concerns observation of squirrels:
Correct Levels Of Stress Hormones Boost Learning, Squirrel Study Suggests

                                            Tests on the influence that a stress-related hormone has on learning in ground squirrels could have an impact on understanding how it influences human learning, according to a University of Chicago researcher.

Jill Mateo, Assistant Professor in Comparative Human Development, has found that when they perform normal survival tasks, ground squirrels learn more quickly if they have a modest amount of cortisol, a hormone produced in response to stress, than those with either high or low levels of cortisol.

In humans, cortisol production is also related to stress and is known to have an impact on learning, but that impact is not well understood, Mateo said.

This should sound familiar to anyone who's been through even a few moderately difficult college courses. Namely, if the work in a particular course isn't difficult enough to cause at least a little stress, retention of the details may not be very complete. Without some stress, the material just doesn't seem "important" enough, even if it's new to the student, to compel the student's attention to the details and the complexity. But of course, if the material is difficult enough to cause excessive stress, anxiety can get in the way of successfully organizing the material in the student's mind.

The second study looked at the actual neurobiology of learning under conditions of acute stress:

Short-term Stress Can Affect Learning And Memory

                            Short-term stress lasting as little as a few hours can impair brain-cell communication in areas associated with learning and memory, University of California, Irvine researchers have found.

It has been known that severe stress lasting weeks or months can impair cell communication in the brain's learning and memory region, but this study provides the first evidence that short-term stress has the same effect.

As it turns out, another stress-related hormone besides cortisol is involved, corticotropin releasing hormone (CRH), and the latter is more significant under conditions of acute stress:

In their study, Baram and her UC Irvine colleagues identified a novel process by which stress caused these effects. They found that rather than involving the widely known stress hormone cortisol, which circulates throughout the body, acute stress activated selective molecules called corticotropin releasing hormones, which disrupted the process by which the brain collects and stores memories.

Learning and memory take place at synapses, which are junctions through which brain cells communicate. These synapses reside on specialized branchlike protrusions on neurons called dendritic spines.

In rat and mouse studies, Baram's group saw that the release of CRH in the hippocampus, the brain's primary learning and memory center, led to the rapid disintegration of these dendritic spines, which in turn limited the ability of synapses to collect and store memories.

The researchers discovered that blocking the CRH molecules' interaction with their receptor molecules eliminated stress damage to dendritic spines in the hippocampal cells involved with learning and memory.

The role of cortisol, in learning under conditions of moderate stress, remains somewhat less clear. In addition to the squirrel study, anecdotal experience with so-called "flashbulb memories" supports the idea that some degree of stress can assist the formation of memories. The Wikipedia article states, without references, "Some biologists believe that the hormone cortisol, which is released in response to stressful incidents, cooperate with epinephrine (adrenaline) to cause the formation of flashbulb memories by the brain, functioning to help remembering things to avoid in the future." The squirrel study suggests cortisol actually has some role in memory formation, rather than being just a coincidental byproduct of stress. (See also the article on Emotion and memory.

Memory and BDNF

Following up on part of this note, where I discussed relationships between memory and stress, it turns out that there are some interesting things known, related to this, which involve a "neurotrophic factor" called BDNF.

In fact, there's quite a lot to say. Let's begin with an explanation of some terms, a little about BDNF, and a look at some research from the past several years on the relationship between BDNF and memory. Later we'll take up more on how stress and depression enter the picture.

A neurotrophin is a type of protein that promotes the survival of neurons – which is in general a pretty good thing. (We'll get to examples in a moment.) One type of neurotrophin, known as a "neutotropic factor", is a growth factor that affects neurons in particular.

More generally, a growth factor is a proteins that signals certain types of cells to survive, differentiate, or grow. A growth factor that helps a cell survive does so by inhibiting programmed cell death. Other growth factors promote cell division, which results in growth of the tissue that contains the affected cells. Yet other growth factors may induce cells to differentiate into cells of a more specialized type.

An important example of a general growth factor is IGF-1, also known as "insulin-like growth factor 1", which we'll be looking at more extensively in upcoming posts.

In this post we're going to consider the specific neurotrophic factor known as BDNF, the brain-derived neurotrophic factor.

Research has shown that BDNF plays a role in memory formation and in the connection between stress and depression. For example, in rats the stress hormone corticosterone seems to decrease the expression of BDNF, and if stress is persistent, this eventually leads to the atrophy of the hippocampus. Since the hippocampus plays an important role in long term memory, this is one way in which stress can negatively impact memory.

Atrophy of the hippocampus has also been found in humans suffering from chronic depression. There is evidence that suggests a deficiency of BDNF may be at least in part implicated in such depression. For example, various factors (such as the neurotransmitter glutamate, exercise, calorie restriction, and antidepressant drugs) are known to stimulate expression of BDNF – and often ameliorate depression as well.

There's a lot of science behind all this. Let's just look at a few research announcements from the past several years to get a feel for the interactions of BDNF and memory.
Key Pathway In Synaptic Plasticity Discovered

                                                The researchers studied a major developmental event in newborn rodents. A rapid increases in synapse strength and visual circuit refinement occurs quickly after the animal's eyes first open. It was already known that the PSD-95 protein rushes to visual system synapses soon after eye opening. PSD-95 is a scaffold protein that anchors several types of receptors. Some of these receptors are for the neurotransmitter glutamate, and there is also the TrkB receptor for BDNF (and other neurotrophins).

A positive feedback loop is initiated, in which the NMDA glutamate receptor activates BDNF. BDNF then triggers a signaling pathway involving the kinases PI3 and Akt. This pathway leads to more PSD-95 production, completing the loop. The net result is to make synapses more responsive to BDNF, followed by production of additional PSD-95. Once this loop is started at just a few of a neuron's synapses, the rush of PSD-95 to other excitory synapses of the neuron is on. In this way a few very active synapses can prime larger regions of a neuron for long-term synaptic strengthening in response to subsequent stimulation in the newborn animal.
Proteins Necessary For Brain Development Found To Be Critical For Long-term Memory

                                                  This research indicates that BDNF, which is crucial for the growth of brain cells during development, is also equally important for the formation of long-term memories. The study was performed on the common marine snail Aplysia. When the snails are electrically shocked, the neurotransmitter serotonin is released and promotes the formation of long-term memories associated with the shocks. But when the researchers blocked interaction between BDNF and its TrkB receptor, long-term memories did not form, even though serotonin was still released at synapses. This indicates that serotonin alone was not sufficient for long-term memory formation. Short-term memory formation was not affected. Further investigation showed that interfering with the BDNF receptors blocked long-term enhancement of the connections between the brain cells in the reflex circuit normally induced by the shock treatment.

Drug Triggers Body's Mechanism To Reverse Aging Effect On Memory Process

                                                       A class of drugs known as "ampakines" (so-called because they target AMPA receptors) has been under study and development since the early 1990s to deal with neurological conditions, such as schizophrenia, problems of attention span and alertness, and memory impairment associated with dementia and Alzheimer's disease. The research reported here was conducted by a team that included Gary Lynch, who has long been associated with investigation of the biological bases of learning and memory. (See here for more about Lynch and long-term memory.)

In this study, rats were treated for four days with an ampakine drug. Of particular interest was the effect of the drug on the hippocampus of the brain, because of its known importance in the formation of long-term memories. In the hippocampus areas of rats treated with the drug, it was found that (compared to controls) there was a significant increase both of levels of BDNF and of long-term potentiation (LTP) of synapses (an indicator of memory formation). Further, even though the drug had a known half-life of only 15 minutes, elevated levels of BDNF and LTP were observed as long as 18 hours after drug administration was stopped.
Tiny RNA Molecules Fine-tune The Brain's Synapses

                               Synapses between two neurons are formed between locations at the tip of an axon of one neuron (the "presynaptic" neuron) and a dendrite on the body of another neuron (the "postsynaptic" neuron). In order to form a complete synapse, it is necessary for there to be protrusions called "dendritic spines" on dendrites of the postsynaptic neuron. In the process of synaptic signaling, it is these spines that absorb neurotransmitter molecules released by the axon of the presynaptic neuron. Consequently, any mechanism that affects the density of spines on dendrites will affect the total number of synapses that can form between neurons.

It had previously been established that BDNF activates a protein kinase called Limk1, which in turn promotes the growth of dendritic spines and hence the ability of synapses to form. This research on rats studied the effect of the microRNA miR-134 on growth of dendritic spines of hippocampal neurons. It was found that when neurons were exposed to miR-134, spine volume significantly decreased, and synapses weakened. Conversely, when miR-134 was inhibited, spines increased in size, strengthening synapses. However, increased levels of BDNF negated the effects of miR-134, indicating that miR-134 achieved its effect by suppressing Limk1.