An understanding of how memory works has been the Holy Grail of psychology for the past century. Over the past fifty years, neuroscientists have joined that quest, searching for how and where the brain forms new memories and retrieves old ones. Over the past decade, research from both psychology and neuroscience has revised the fundamental notion of memory as a fixed image, word, or idea that we recollect. The modern scientific view of memory is much more fluid, as experiments have demonstrated that each time we recall a memory, we reshape and reconsolidate it.
Two new reports this week may be moving the study of memory into a brave new world, where we can not only monitor memory but manipulate it. Of course, we are manipulating and making memories every waking hour and perhaps with many of our sleeping hours as well, but the introduction of neurotechnologies that can target specific memories by tuning brain circuits takes us into a new era in the study of memory.
One report, in Nature, describes the use of optogenetics to manipulate memory in mice.1 Optogenetics uses a focused beam of light aimed directly into the brain to turn on or turn off specific circuits. The method requires genetic engineering of brain cells to render them light sensitive and the insertion of an optical probe to deliver the light, so this technique is not likely to serve as a treatment for human disorders. Previous studies in mice have used optogenetics to map the circuits for sleep, satiety, fear, and reward, among many other complex brain functions. The new work from Susumu Tonegawa and his colleagues at MIT uses optogenetics to turn on and turn off memory associations, changing negative associations to positive ones. While this may sound like psychotherapy (in mice), the memories in this case are for recent events and the associations can be tweaked in either a positive or negative direction.
A second report out today in Science uses a different technology to improve memory in humans.2 Joel Voss and colleagues at Northwestern University used repetitive transcranial magnetic stimulation (rTMS) to target a circuit connecting the hippocampus and the cortex which is thought to be essential for memory. After a few days of 20 minute-long magnetic stimulation treatments, participants' performance improved on a task requiring them to remember words associated with pictures of faces. One notable aspect of this study was the use of functional magnetic resonance imaging (fMRI) to pinpoint, for each participant, an exact site in the memory network to target for the association memory. By stimulating over that unique site with the rTMS magnet, Voss and colleagues administered a personalized intervention that optimized each participant's circuit connectivity. Not only did memory improve, but scan images confirmed that the stimulation brought activity into better synchrony, that is, improved connectivity within this circuit.
It's important to note that this intervention simply appears to improve the ability of one brain region to communicate with another region; it is not influencing the content of that communication—it may be improving the roads, as it were, not specifying what types of vehicles can travel on them. The tools for human neuroscience remain primitive—unlike optogenetics, rTMS stimulates broad regions of the brain and fMRI captures changes in blood flow over many seconds, not the activity of the brain at the millisecond speed of thought.
Nevertheless, these studies in mice and humans present us with an interesting new dilemma. On the one hand, manipulating memory can mean improving recall for people with dementia or relieving the symptoms of PTSD or depression. Indeed, if mental disorders can now be addressed as disorders of brain circuits, then treatments for mental disorders can be framed as interventions to tune circuits, as described in these two new reports. Psychotherapy can be considered the ultimate personalized and precise intervention to tune a brain circuit. On the other hand, manipulating memory with neurotechnologies raises thorny ethical issues. Aldous Huxley famously said, “Every man’s memory is his private literature.” Using fMRI to read the language of the brain has already raised questions of accuracy and privacy (thoughtfully reviewed in the recent book Brainwashed by Satel and Lilienfeld3), but writing that language raises questions of agency and even reality, as depicted in the film Total Recall. The technology is not there yet for humans, but turning on and off circuits and behaviors with optogenetics in mice is now routine. As we develop better tools for human neuroscience, it’s not too early to ask about unintended consequences.
The NIH BRAIN (Brain Research for Advancing Innovative Neurotechnologies) Initiative is about to announce its first awards for advancing human neuroscience. Neuroethics, the discipline that thinks about unintended consequences of new neurotechnologies, is part of the BRAIN Initiative with expertise on the Council that oversees the initiative and a full day workshop planned in November to struggle with these issues. How can we ensure that scientific progress will deliver better treatments without introducing tools for mind control? Part of the answer will be in careful planning and an honest appraisal of risks as we enter this complex new world.
In the future, these two new papers, in the most prestigious journals of science, may appear as milestones in the study of memory, moving the field from descriptive studies to more mechanistic ones. After a century of studying memory, the Holy Grail appears to be closer than ever. But as Indiana Jones learned in a similar quest, the closer we get the more complicated things become.
1 Redondo RL et al. Bidirectional switch of the valence associated with a hippocampal contextual memory engram . Nature. 2014 Aug 27. doi: 10.1038/nature13725. [Epub ahead of print]