20 October 2006. If you are one of those people who can recall verbatim a conversation held a year ago, then you might want to thank your parents—not necessarily for sending you to that prestigious school, forcing you to memorize all of the Canterbury Tales, or spending endless hours playing The Memory Game, but for simply passing on a specific letter of the genetic code. A paper in today’s Science reports that a single allelic variation (T for C) in the KIBRA gene can increase memory performance by about 20 percent. Furthermore, non-carriers of the T allele show greater neuronal activation in the hippocampus and parahippocampal gyrus, brain centers that are intimately involved in learning and memory, suggesting that they have to work harder when remembering. The findings could lead to new strategies to develop memory-enhancing drugs that might help patients with mild cognitive impairment or even those with Alzheimer disease and other dementias.
The KIBRA gene codes for a protein that has binding motifs for dendrin and a protein kinase C isoform called PKMζ, both linked to long-term potentiation and synaptic plasticity (see Pastalkova et al. 2006, ARF related news story, and Kremerskothen et al., 2003)—processes that are essential for human learning and memory. But it was not these protein-protein interactions that led to the discovery of the memory allele. Instead, Andreas Papassotiropoulos and Dominique de Quervain at the University of Zurich, Switzerland, Dietrich Stephan at the Translational Genomics Research Institute, Phoenix, Arizona, and their colleagues found the KIBRA allele in an unbiased screen for genetic variations that impact memory.
The authors compared a genome-wide screen of more than half a million single nucleotide polymorphisms (SNPs) with performance in a verbal memory task. They recruited 351 young adults (mean age 22 years) from Switzerland to take part in the screen, testing their verbal episodic memory by having them remember 30 unrelated nouns. The authors found that two SNPs were associated with volunteers with the best memories: rs17070145, a T-to-C substitution in intron 9 of the KIBRA gene; and rs643988, another T-to-C substitution in the first intron of the gene encoding the synaptic protein calsyntenin 2 (CLSTN2). Volunteers with the T variant of the KIBRA gene had 24 percent better recall 5 minutes after being presented with the words and 19 percent better recall after 24 hours. These differences were highly significant. The differences between those with the two CLSTN alleles were not as high—16 and 15 percent for the short and long recall, respectively.
Genetic association studies in large diverse populations can be notoriously difficult to interpret because the cohort being studied may not necessarily be genetically similar. So the authors verified the findings in two ways. First, to eliminate the possibility of false positives, they checked for heterogeneity within the study sample, finding that the volunteers were genetically homogeneous apart from 10 outliers who were discounted from the study. Second, they performed the same analysis on an American cohort of 256 cognitively normal older people (mean age 55 years) using neuropsychological tests and DNA samples from a study under the direction of Eric Reiman at the Banner Alzheimer’s Institute in Phoenix. Again, KIBRA T allele carriers performed significantly better in two episodic memory tasks, the Buschke’s Selective Reminding Test and the Rey Auditory Verbal Learning Test. But the CLSTN2 SNP failed to show any association with memory, suggesting that it has no effect in the American population or that the effect may be age dependent. In a second Swiss cohort they also found that T allele carriers performed better in tests of visual episodic memory.
What role does KIBRA play in memory and how is that influenced by the intron allele? The authors found that non-T allele carriers exhibited significantly higher neuronal activation when trying to match previously memorized faces and professions. The greater neuronal activation was limited to the hippocampus and parahippocampal gyrus, areas of the brain that are integral to learning and memory, suggesting that the allele affects these areas of the brain in particular (see image below). It appears to have no effect on working memory, executive function, or attention because both carriers and non-carriers of the T allele performed equally well in the Wisconsin Card Sort Test, while in the verbal recall tests there were no differences in immediate recall ability.
KIBRA Allele Makes a Difference
During a face-profession association task, individuals lacking the KIBRA T allele show significantly greater activation (shown by color-coded t values) in the hippocampus (H) and parahippocampal gyrus (P). The coronal image is made by overlaying data from 15 T allele carriers and 15 non-carriers. [Image © Science]
"The exciting thing about KIBRA is that we completely abandoned the hypothesis-based approach and scanned the genome in an unbiased manner to identify novel genes and pathways related to human memory performance. KIBRA is the first memory-related gene identified by genome-wide scanning. Now that we know of its existence we can track the entire KIBRA-related pathway to understand human memory better and ultimately help people with memory problems,” said Papassotiropoulos.
The authors have examined expression levels of the protein in postmortem human brain and detected the truncated version of the protein (lacking the first 223 amino acids) in memory-related areas, including the hippocampus and temporal lobe. The truncated protein does contain the motifs that interact with dendrin and PKMζ. It will be interesting to see how the SNP might alter expression of the protein or otherwise influence recall.—Tom Fagan.
Q&A with Eric Reiman, Andreas Papassotiropoulos, Dominique de Quervain. and Dietrich Stephan. Questions by Tom Fagan.
Q: How does this data fit with current theories on the effects of genetic variation on memory?
A: Twin studies suggest that about half of the individual differences in normal episodic memory are related to inherited genes. Our studies suggest that the KIBRA gene accounts for about a 20 percent difference between people with better or worse memory performance, remaining significant after correction for more than 500,000 comparisons, and independently confirmed in several independent studies. Each of the studies looked at a slightly different version of episodic memory performance, suggesting that it's related to the ability to recall words or pictures, 5 minutes, 30 minutes, or 24 hours later.
Q: Might these alleles impact memory losses later in life?
A: We do not yet know, but are very interested in determining the extent to which the effects of KIBRA on memory are modified in response to normal aging. While our studies found a difference between KIBRA T carriers and non-carriers in both young and older adults, we would need the same memory test in both groups to determine if KIBRA preferentially affects the non-disabling episodic memory declines that accompany normal aging.
Q: How might these allelic variations relate to mild cognitive impairment, AD, or other dementias? Would you expect those with the T allele to maintain cognition a bit better than those without, for example?
A: Based on our preliminary analyses, the KIBRA gene does not appear to distinguish AD cases from controls—but you ask a very interesting question! If, indeed, KIBRA T non-carriers have less cognitive reserve (at least in the domain of episodic memory), one might predict that they would have an earlier onset of MCI and perhaps dementia in response to the same severity of underlying disease. This is something that could be tested.
Q: How might the KIBRA protein influence human memory?
A: KIBRA's structure suggests at least three possible roles in memory: 1) it may be a modulator of synaptic plasticity, 2) it may interact with the isoform of protein kinase C (PKMζ) that has been implicated in memory formation and long-term potentiation, and 3) it may serve as a calcium sensor in vesicle exocytosis, bolstering the communication between brain cells. In other words, its structure suggests that it could strengthen the connections between brain cells involved in episodic memory by building new connections between these brain cells, by bolstering the electrical response of brain cells to chemical messengers, or by bolstering the transmission of the chemical message itself. Yet it might turn out that KIBRA possesses unique properties unrelated to the above-mentioned mechanisms. It will be very interesting to find out.
Q: Memory-linked genetic variations have been found in genes for BDNF, HTRA2, and now KIBRA. Is there some common denominator or pathway that is emerging?
A: Episodic memory is a highly polygenic trait. While it is possible to identify statistically genetic clusters (de Quervain et al., 2006 and ARF related news story) related to human memory, it is highly premature to interpret these relations in a mechanistical way. The common denominator will be synaptic plasticity after all, which represents an extremely complicated cascade of molecular events (Kandel, 2001).
Q: Does the magnitude of the effect of the BDNF, HRTA2, and KIBRA genetic variations tell us anything about the relative importance of these three proteins in learning and memory?
A: No. The effect of the genetic variations tells us that these genes are important for learning and memory; however, it is not possible to extrapolate a relative contribution. It’s like Alzheimer disease: genetic variations within APP and the presenilins have (if any) a barely measurable effect on AD risk. However, the importance of these genes for the AD-related pathophysiology is huge.
Q: Might the discovery of the KIBRA alleles open any therapeutic avenues for dealing with memory loss or cognitive decline?
A: We're excited about the chance to use this information to identify promising memory-enhancing compounds, several of which we've begun to test in aged rats. The KIBRA protein and its interactions with other proteins permit us to postulate a pathway that may participate in episodic memory. We can consider which compounds affect the actions of these proteins and then test the effects of these potential memory-enhancing compounds on memory performance in certain laboratory animals. The ideal memory-enhancing compound would be effective, extremely safe, and extremely well tolerated. And it could potentially enhance memory performance (even without slowing Alzheimer disease progression) in patients with dementia and MCI, aged persons, and even the public at large, depending on its efficacy, safety, and tolerability.