DNA and memory: new theory proposes that genes play a role in declarative memory
JORGE LUIS BORGES’ FICTIONAL CHARACTER, FUNES, in the short story, “Funes, His Memory,” had the misfortune of remembering every experience of his life. The story mentions others in history who have shared the same fate: King Cyrus of Persia could call all of the soldiers in his armies by name; Mithridates Eupator, the king of Pontus, administered justice in the 22 languages of his empire; Greek poet Simonides of Ceos invented the mnemonic technique; and Greek philosopher Metrodorus professed the art of faithfully repeating what he had heard, even if it was only once. Although these are all cases of exceptional memory, most humans have the capacity to store a surprising amount of information throughout their lives, even though they rarely or never use it.
Sandra Peña, Ph.D., biologist at the University of Puerto Rico, Río Piedras Campus, has studied declarative memory for years at the behavioral level as well as the cellular and molecular levels. Declarative memory, shaped throughout life, is in charge of remembering data and events, and, along with the genome, determines human individuality. Peña and her team of researchers have formulated a new theory: a genomic mechanism that rearranges genes and may participate in the permanent storage of memories. Their work contributes to the understanding of diseases that affect memory.
Behavioral tests for memory and learning
In the laboratory, Peña and her team administer behavioral tests that help measure some types of learning and memory, such as spatial associative learning, which allows people to locate the necessary references to move through space. “For example, if a new student arrives at the university, she will learn to locate references so that she can move around campus. She can identify the bell tower as a reference point, which tells her how close she is to the theater. Rodents use the same cognitive mechanism. They use it to survive in their natural environment to find food and avoid predators,” explains Peña.
Spatial associative learning is dependent on the hippocampus. This brain structure is one of the areas most affected by Alzheimer’s disease. Peña tests hypotheses on transgenic mice, mice created in a laboratory by adding a gene that they did not previously possess. In this case, each experimental mouse is given a gene that causes Alzheimer’s disease. Next, the researchers develop experiments aimed at developing possible treatments to slow or alleviate the cognitive and neuropathological symptoms of Alzheimer’s disease.
“The test rodents undergo to measure spatial associative learning is known as the Spatial Discrimination Test, which can be useful for understanding other neuropsychiatric conditions. For example, patients with schizophrenia show disordered thoughts, give importance to irrelevant facts, and cannot correctly form associations,” says Peña. The researchers deactivate genes in the hippocampus that may be related to the development of schizophrenia.
Peña also constructs behavioral paradigms that simulate the development of conditioned aversions to tastes, smells, and visual stimulants in response to gastrointestinal problems. Using Conditioned Taste Aversion Tests, Peña’s team gives the rodents sugar water. As expected, the rodents drink it eagerly. Afterward, researchers add a substance to the sugar water, causing the rodents to become sick. The following day, when the researchers present the rodents with the option of drinking plain water or sugar water, they will drink the plain water.
“If the rodents hadn’t had the negative experience, they would have continued to drink the sugar water. Rodents use the same brain structures as humans to process and store this type of information. This behavior is analogous to when a person goes to a restaurant and eats a food that she has never tried before and it makes her sick. She is going to associate that taste with the gastrointestinal problem that she suffered and will avoid eating it in the future,” explains Peña. The results of these studies may one day help cancer patients to manage aversions they develop in response to chemotherapy.
Peña focuses the majority of her research on the study of cognitive emotional processes, such as the development of fear. Fear is a defense mechanism that can be innate or learned. For example, the ability to identify dangers in order to avoid them—the fear that rodents feel toward predators—is innate. But there are also learned fears that can get out of control, as in the case of phobias. In the laboratory, Peña and her team model situations that are similar to the way humans develop phobias by conditioning mice to fear a cage or a bell, in the hopes of understanding the neurobiological processes that are important for learning not to fear. This research may contribute to the development of medical alternatives for patients with post-traumatic stress disorder.
Learning and memory starts at the molecular and cellular levels
The hippocampus, the amygdala, and the prefrontal cortex are all associated with the behaviors that Peña studies in her laboratory. “The gustatory cortex and amygdala are important in taste conditioning. For example, when you enter a new place, the hippocampus activates to help you learn everything about it. Now, if you enter a place, and something bad happens, the amygdala activates because, while you are learning, an emotional factor is also present. The amygdala is the most important part of the brain related to emotions, both positive and negative,” explains Peña. The research team observes which genetic changes occur in the hippocampus, the amygdala, and the prefrontal cortex during and after the learning process.
The research team’s experiments confirm that protein synthesis (translation) and the synthesis of the RNA messenger (transcription) are required for long-term memory, while short-term memory does not require the synthesis of new proteins. The processes of translation and transcription are necessary for gene expression. “Genes are codified in the nucleus—in the DNA—but for a gene to express itself, it must first be transcribed to RNA and later translated into a protein that can express the DNA’s function. That is to say, genes are present in all cells of the body, but they need to be transcribed and translated in order to be expressed,” explains Peña.
“We have identified a group of genes, called candidate genes, which we believe are necessary for the processes of learning and memory. We use molecular and biochemical genetic techniques to study the changes in levels of RNA messengers and proteins for a particular candidate gene.”
Peña and her team test their hypothesis on transgenic mice. The added gene inhibits the function of a transcription activator in the processes of learning and memory. Peña has created animals with different genetic mutations to observe the effect these mutations have on various cognitive processes. For example, before exposing a rodent to the unpleasant experience of conditioned taste aversion, it is injected with a protein inhibitor that causes the animal to have only short-term memory.
Another approach allows Pena’s team to study hundreds or even thousands of genes simultaneously by utilizing DNA matrices. Complementary DNA microarrays contain representative sequences of thousands of genes on one glass slide. “For our experiments, it has been possible to synthesize the sequences that are important directly on the slide. We can study changes in the expression of 22,000 genes simultaneously,” says Peña.
Structural vs. genomic hypothesis
Currently, Peña is developing a new theory that questions the accepted hypothesis of how humans store memories and construct what is known as declarative memory.
The current models of declarative memory, based on the structural hypothesis, propose that memories are stored via structural modifications in the synaptic connections between neurons. According to the structural hypothesis, memories are formed by networks of interconnected neurons that contain our experiences in their connection structures. This hypothesis also states that proteins synthesized during memory consolidation contribute to changes in synaptic efficiency, including the establishment of new connections. It is presumed that the structural modifications in the synaptic connections cause a cascade of biochemical reactions to occur, activating many proteins. These proteins include kinases, which are transported to the nucleus from the cell where they activate transcription factors that regulate gene expression. It is believed that the final steps of this cascade activate the genes responsible for protein synthesis implicated in structural changes to synaptic connections. Numerous pharmacological experiments and experiments that deactivate genes have demonstrated the importance of this cascade in learning and memory formation.
“Changes in the levels of messenger RNA and proteins are short term. This is why it is difficult for us to understand how this type of change can be used to explain the formation of memory that lasts for years and, sometimes, for an entire lifetime.
“The structural hypothesis tries to construct stable memory using an unstable base. Permanent declarative memory cannot be based on the foundation of unstable synaptic plasticity. Although synaptic plasticity can be used to explain learning and storing short- and medium-term memory, it does not explain permanent memory,” says Peña.
Permanent declarative memory should be “carved in stone,” and the only stone-like structures in living cells are DNA molecules. This is one reason for suggesting that the function of DNA is not limited to storing hereditary information. The new hypothesis proposed by Peña and her team is known as the genomic hypothesis of declarative memory. They propose that a DNA recombination mechanism may participate in and support permanent memory storage.
Sandra Peña and graduate student Xiomara Ramos discuss the results of experiments relating the gene recombination activity in the hippocampus with learning in rats.
“In addition to transcription, a genomic rearranging mechanism, better known as DNA recombination, also exists. We propose that DNA recombination is used to produce permanent memory in the brain. Now other researchers are starting to agree. Our studies represent research that is going to change the way neurobiologists think in terms of how information is stored in the brain at the molecular level,” says Peña.
Peña and her researchers postulate that throughout a lifetime, recombined genes provide their protein products, which are used in the structural and functional changes necessary for storing and maintaining permanent memories. The number of facts, data, and events learned during an individual’s lifetime is enormous; in fact, the number of memories exceeds the number of genes. The team of researchers are studying the mechanisms that cause diversity in genetic modifications, which produce permanent changes in neurons. These changes facilitate long-term memory and help explain how a large number of memories can be stored in a finite number of genes.
Peña's team includes posdoctorate associates Jianpeng Wang, Adrinel Vázquez, Xiaobei Zeng and Michelle M. Martínez; and graduate students Wanda I. Colón Cesario, Melissa Colón Cesario, Itzamarie Chévere, Lorena Saavedra, Lixmar Pereira, Iván Santos, Xiomara Ramos and Sandra V. Rivera Beltrán.
“At this time, our biggest challenge is identifying the genes that are subject to recombination during the formation of memory. We hope that our work will help develop treatments for people who suffer from neurodegenerative and mental health conditions. We also hope that our results can be extrapolated to serve in the development of new and better life strategies that stimulate the proper development of behavior and cognitive aptitudes,” says Peña.
