Addiction is tenacious: People who are in recovery remain vulnerable to relapse even after years of abstinence. Often, it is memory that reawakens the craving. The dormant desire can return with a vengeance when cued by people, places, or situations once associated with drug use. For this reason, memory has become an important area of study for addiction researchers.
Neuroscientists over the last two decades have made important advances in tracing the molecular events that occur in the brain as memories are formed and preserved. In recent experiments, NIDA-supported researchers Drs. Courtney A. Miller, J. David Sweatt, and colleagues at the University of Alabama at Birmingham and The Scripps Research Institute in Jupiter, Florida, delineated what appears to be a key pathway in this process: Reversible changes in DNA regulate the production of proteins that give memories their staying power. The new findings may spur breakthroughs in the study of drug relapse, as well as in a broad range of neurological problems, including age-related memory loss and neurodevelopmental disorders.
The Neural Basis of Memory Formation
Someone enters a restaurant and recalls the taste of its veal scallopini; a man arrives at a shore and thinks of his father, with whom he fished there as a child; a person in recovery passes a certain corner and recalls the high from the drugs he purchased there. This is remembering: Something in our present experience brings to mind something that was linked with a similar experience in our past. What brain activity makes this happen?
The emerging neuroscientific understanding of memory posits that the world's various features—for example, the interior of a restaurant or the taste of veal scallopini—each evokes particular patterns of neuron firing in the brain. The brain registers these patterns for future reference by strengthening the synapses between neurons that fire simultaneously. Stronger synapses make neurons more sensitive to each other, so that the subsequent firing of some readily stimulates the others to fire as well. Now, when the restaurant interior reactivates some of the same neurons that it did before, they in turn stimulate the neurons previously excited by the veal taste—even though no veal is actually present.
The brain is constantly registering, storing, and reaccessing neural patterns in this way. The resulting memory associations can powerfully drive emotions and behavior: The person passing the restaurant suddenly feels famished; the man at the shore feels an urge to call his parents; the person in recovery begins to crave the drug he used long ago.
| Methylation Suppressor Administered Before Foot Shocks | Methyl Groups Attach to Calcineurin Gene | Messenger RNA Produced | Calcineurin Produced | Fear Memory Formed |
|---|---|---|---|---|
| No | Yes | No | No | Yes |
| Yes | No | Yes | Yes | No |
Methylation and Memory Formation
Neuroscientists have found that the process of forging the new synaptic links between neurons involves a host of proteins. Drs. Miller and Sweatt reasoned that the process might be strongly influenced by epigenetic factors, which act upon genes to control the rate of production and availability of their proteins. Applying a hypothesis first suggested by Dr. Francis Crick in the 1980s, the researchers focused on one epigenetic process, DNA methylation, which prevents the first step in protein production (see box below).
In an initial experiment, Drs. Miller and Sweatt established that DNA methylation is essential for memory formation. They first placed rats in a special cage and administered mild electric shocks to the animals' feet. Normal rats remembered the unpleasant experience, and most froze with fear when put back into the cage 24 hours later. But rats that before being shocked were treated with a compound that prevents methylation did not appear to remember the shocks when they were returned to the special cage the next day. They behaved as if nothing untoward had happened to them there.
Drs. Miller and Sweatt next demonstrated that the changes in methylation that occur during memory formation favor synaptic strengthening between neurons in the hippocampus, the brain area known to initially register certain memory associations. Once again, the investigators administered mild shocks to rats, and the animals' behavior the next day showed that they remembered being shocked. This time, the researchers compared hippocampal tissue from these rats and from control groups that were not trained to associate the cage with shocks. The rats with the fear memories showed two changes related to strengthened memory.
- Increased methylation of the gene for protein phosphatase 1 (PP1), an enzyme that inhibits memory formation by interfering with synaptic strengthening;
- Decreased methylation of the gene for reelin, a protein that promotes memory formation by assisting in synaptic strengthening.
Drs. Miller and Sweatt confirmed that these changes resulted in less abundant PP1 and more abundant reelin. They concluded that methylation plays a dual role in memory formation, both increasing the availability of a memory-promoting protein and reducing the presence of the memory-suppressing protein.
Methylation Blocks Protein Production
Protein production begins with transcription, the transfer of information contained in a gene's DNA to a messenger RNA (mRNA) molecule. The mRNA molecule provides the template for assembling amino acids into protein.
Methylation regulates protein production by blocking transcription. Methylation occurs when a methyl group—a cluster of a carbon and three hydrogen atoms—attaches to DNA. Like other epigenetic processes, methylation is reversible—methyl groups attach, detach, and reattach to DNA. By doing so, they vary the rate of protein production to suit the brain's needs in changing situations.
The methylation associated with the shock memories proved to be transient, disappearing within an additional 24 hours. The researchers noted that this fits with a well-established role of the hippocampus: As the locus of initial memory formation, it provides short-term storage only. If a memory is to last, it must move to another part of the brain. Where this might be, and how the memory might be preserved, were the questions the researchers tackled next.
From a Week Ago to Yesteryear
Previous research had indicated that within a week of memory formation, the chemical and metabolic activities that support memory storage shift from the hippocampus to the dorsomedial prefrontal cortex (dmPFC). Drs. Miller and Sweatt and collaborators focused their attention on this region.
On the basis of their work on memory formation, Drs. Miller and Sweatt posited that methylation might also be key to the consolidation and preservation of long-term memory. However, they suspected that methylation of the gene for the enzyme calcineurin, rather than PP1 and reelin, might be important in establishing and sustaining memories in the dmPFC.
"There is a lot of published work showing that calcineurin suppresses memory and that memories are more easily formed and last longer if you block calcineurin," Dr. Miller says. Methylation of the calcineurin gene, by reducing the amount of calcineurin produced, might remove a barrier to the transfer into the dmPFC of memories first formed in the hippocampus.
The researchers returned to the animal lab and repeated the foot-shock procedure with a new set of rats. After they confirmed fear memory, the team assayed dmPFC brain tissue from groups of animals at different times after the procedure: 1 hour, 1 day, 7 days, and 30 days. The results bore out their hypothesis. The calcineurin gene did not exhibit increased methylation after 1 hour, but it did after 1 day and remained highly methylated 7 and 30 days later.
To solidify the link between memory and methylation of the calcineurin gene, the researchers infused the hippocampus of rats with a compound that prevents fear learning. When the animals were then given the foot-shock procedure, as expected, they did not demonstrate the fear response—and 7 days later there was no extra methylation of the calcineurin gene in the prefrontal cortex.
One crucial question remained: Is methylation necessary to maintain a memory after it has been consolidated in the prefrontal cortex? To test this, the researchers infused methylation-blocking molecules into the prefrontal cortex of rats 30 days after the foot-shock procedure. With methylation suppressed, the animals ceased to display fear when reexposed to the shock-associated cage. Without ongoing methylation, the researchers speculate, a memory effectively disappears.
Wider Application of Methylation
Although the latter study focused on the calcineurin gene, the complexities of memory formation and maintenance in the prefrontal cortex surely also entail DNA methylation in other genes, Dr. Miller says. Her team plans to identify these genes in future research.
Reducing methylation might, in theory, provide a means to weaken the memories that drive relapse to addiction, Dr. Miller says. But the difficulty of targeting specific memories without inducing a more general amnesia makes this approach a distant prospect at best.
Understanding the molecular basis for the persistence of memory, however, could have more immediate relevance to other problems, such as age-related memory loss. Dr. Miller says, "If we can boost the mechanisms involved in maintaining memory, we might help individuals suffering from cognitive decline."
What makes the recent findings particularly interesting, says Dr. John S. Satterlee of NIDA's Division of Basic Neuroscience and Behavioral Research, is the demonstration that methylation activity follows the path of memory from the hippocampus to the prefrontal cortex. "The molecules themselves aren't moving, of course, but something happens that shifts the process to the right place," he notes.
The duration of these processes is also intriguing, Dr. Satterlee says. "I wouldn't go so far as to conclude that DNA methylation is a way of encoding memory, but the fact that molecular changes in this brain region can last as long as 30 days is very intriguing," he comments.
While Dr. Satterlee agrees that issues of targeting and potential toxicity pose significant obstacles to methylation-based drug abuse interventions, he suggests that a deeper understanding of the process could have general implications for addiction research. "We know that epigenetic changes occur when the brain responds to drugs of abuse, but scientists haven't yet looked deeply at methylation in this regard," he says.
Dr. Satterlee notes that defects in a protein that can bind to methylated DNA have been identified in Rett syndrome. He suggests that studies like Dr. Miller's could help to advance our understanding of neurodevelopmental disorders.
Source
Miller, C.A., et al. Cortical DNA methylation maintains remote memory.Nature Neuroscience 13(6):664-666, 2010. (Full Text (PDF, 524KB))