3 July 2009. A report in today’s issue of Science describes an intriguing potential strategy for boosting memory. By overexpressing a protein that regulates G-protein signaling in specific neurons of the rat visual cortex, researchers led by Zafar Khan, University of Malaga, Spain, were able to convert short-term object recognition memory lasting under an hour to long-term memory persisting for many months. What’s more, selective elimination of the targeted cells wiped out the memory enhancement. These findings suggest that neurons in the V2 visual cortex are critical for object recognition memory, an activity long believed to occur within a separate brain area, the medial temporal lobe.
Unexpected circumstances led first author Manuel López-Aranda and colleagues to focus on area V2. They observed in their previous experiments with regulator of G-protein signaling-14 (RGS-14) that this protein modulated calcium signaling when transfected into cell lines. “Since that is one of the processes intimately related to memory processing, we thought, ‘why not use (RGS-14) to stimulate some part of the brain in vivo and see what happens?’” Khan said in an interview with ARF. “We wanted to target the hippocampus, knowing that this area of the brain has been implicated in memory function, and use other brain areas (including V2) as controls.” However, injecting RGS-14 into rat hippocampus or any other brain region produced no effects—except in area V2 where they saw a boost in memory.
In the monkey, V2 is a highly interconnected brain area that receives signals from the primary visual cortex (V1) and sends projections to other secondary visual cortices (Stepniewska and Kaas, 1996; Gattas et al., 1997). The rat brain is comparatively less complex. However, V2 is one area that is not only similarly developed in both rats and monkeys, but is also clearly separated from the medial temporal lobe in rats, Khan said.
V2 itself contains several levels of cortex, including layers 2 and 3 that handle information processing, and layer 6, which has a largely unknown role in visual memory. To explore this issue, Khan’s team injected rats in their layer 6 V2 region with lentivirus carrying the RGS-14 gene, and assessed their performance in a standard novel object recognition test. The test measures recognition memory by the amount of time the rat spends exploring a previously encountered object relative to one it has never seen, the idea being that less attention should go to the familiar object. Typically, rats will completely forget a novel object in less than an hour, and will explore the object as if it was a totally novel one.
However, RGS-14-injected rats could remember objects as much as 24 weeks later. Their ability to store long-term visual memories persisted even 14 months after RGS-14 injection. No memory effects appeared in rats injected with RGS-12 (a protein belonging to the same family as RGS-14), saline solution, or vehicle lentivirus; or when RGS-14 was injected into areas adjacent to V2 (e.g. layer 2/3 neurons, and CA1 and dentate gyrus of hippocampus).
The RGS-14-overexpressing animals were also better able to retain information on multiple objects. Judged by the same recognition memory test, the RGS-14 group could remember up to six objects presented 45 minutes earlier, whereas untreated rats had poor recall ability when just four objects were used. Immunocytochemical analyses done after the behavioral studies confirmed that most of the lentivirus RGS-14 protein went where it was supposed to go—layer 6 V2 neurons.
To further tease out the role of these neurons in object recognition memory, Khan’s team removed them by injecting a neuron-selective immunotoxin (Ox7-SAP) into layer 6 of area V2 in the rats. The neuronal damage mostly stayed within layer 6, and did not extend into other brain areas, e.g., hippocampus, as verified by cresyl violet staining. But the selective elimination of those layer 6 neurons had a huge effect. The Ox7-treated rats could not remember objects seen just 45 minutes earlier, a timepoint at which untreated animals still retained recognition memory. The Ox7-SAP treatment also blocked memory enhancement in RGS-14-treated animals. However, loss of the layer 6 neurons did not seem to affect the rats’ ability to recall objects encountered before the Ox7-SAP treatment, indicating that layer 6 neurons are important for formation but not storage of object recognition memory. The knockout studies “support the view that the entire stream of ventral visual-to-hippocampus, and not the medial temporal lobe alone, is important for memory processing,” the authors write.
Howard Eichenbaum, a memory expert at Boston University, Massachusetts, finds the RGS-14 data tantalizing from a pharmaceutical perspective. “The fact it enhances memory somewhere in the brain is an interesting finding,” he told ARF. “It gives us another potentially interesting candidate for improving memory.”
However, Eichenbaum and others expressed concern about the particular memory test used in the study. He said it is not so surprising that brain areas outside the medial temporal lobe played a role in the object recognition task. He described an analogous situation. “If someone reads a paragraph as fast as they can, and comes back to read it an hour later, they can read it faster that second time. But it’s not really a memory test,” Eichenbaum said. The problem with the object recognition test used on the rats “is that we don’t know entirely how the task is solved. We don’t know which brain structures are used to solve it.”
Eric Kandel, of Columbia University in New York, had similar reservations, contending that responses to novel and familiar objects may not boil down strictly to a memory task. He also felt that the choice of animal model in this study could compromise its validity. “We know a great deal about how visual images are represented in the brain, but we know that almost exclusively in the monkey. We know almost nothing in the rat,” Kandel told ARF. “Rats don't have a good visual system. They're practically blind.” He also felt that scientists should be cautious about extrapolating from the new data because “we don’t know how the drug (RGS-14) is working.”
Khan’s group is hoping to change that. RGS-14 is a microtubule-associated protein that is important for mitosis (Martin-McCaffrey et al., 2005). It has a GoLoco domain through which it controls how much G-protein signal reaches the cell. Analyzing monkeys and rats, Khan and colleagues have determined that RGS-14 is expressed throughout the brain at postsynaptic excitatory synapses (López-Aranda et al., 2006). Getting a handle on how it enhances object recognition memory mechanistically is the next challenge. “We know that we are expressing the protein in one area, but (those neurons) are sending the proteins to other areas,” Khan said. “We are tagging the protein and trying to track it down” to determine which brain areas may be involved in the visual memory processing.—Esther Landhuis.
López-Aranda MF, López-Téllez JF, Navarro-Lobato I, Masmudi-Martín M, Gutiérrez A, Khan ZU. Role of Layer 6 of V2 Visual Cortex in Object-Recognition Memory. Science. 2009 July 3;325:87-89. Abstract
Saksida LM. Remembering Outside the Box. Science. 2009 July 3;325:40-41. Abstract