Intermittent fasting facilitates learning

Several very interesting papers on the physiological changes that may underlie the benefits of intermittent fasting have been published recently. In rodents, intermittent fasting (typically implemented as alternate day fasting) results in numerous physiological changes that are correlated with disease reduction, increased stress resistance, improved insulin sensitivity and upregulation of neuroprotective trophic factors (reviewed by Mattson, 2005; Varady & Hellerstein, 2007). While many IF studies have used relatively short dietary interventions, a recent study by Fontán-Lozano and colleagues (2007) highlights the intriguing possibility that long-term intermittent fasting can lead to improved cognitive function. In a nice series of experiments, these authors demonstrate that fasted mice exhibit improved learning and memory compared to controls (fed ad libitum), and they go on to provide evidence for the underlying mechanism for this cognitive improvement.

Fontán-Lozano et al. mice placed on an alternate day fasting (ADF) regimen for 6-8 months (starting post-natal week 9, the average lifespan of a mouse is about 2 years). Mice on an ADF diet will typically eat more food on their feed days to compensate for the fasting day, and after some adjustment period will eat a bit less than twice as much food on their feed day as a normal mouse would on an typical diet. So in addition to fasting every other day, there is usually some mild caloric restriction involved. It's well-documented that ADF mice live longer than controls.



Survival distributions (n=40) of male C57BL/6J mice fed ad libitum (left curve) or every other day (ADF, right shifted curve). Souce: Talan & Ingram, 1985




ADF mice will also typically weigh less than control mice (Anson et al., 2003), although the ADF mice in Fontán-Lozano et al.'s experiments weighed the same as control mice. This might be due to the longer time spent on the ADF diet, although there are mouse strain differences in addition to diet duration differences (Goodrick et al., 1990). Fontán-Lozano et al. used a simple battery of behavioral tests to assess the learning and memory capacities of ADF mice. They found that compared to control mice, ADF mice learned faster (in a motor learning task, operant food reward task, and Pavlovian eyeblink conditioning task) and were better able to discriminate novel from familiar objects from briefer presentations.

Subsequent physiological experiments revealed that ADF mice exhibited increased theta-band activity in the hippocampus. Explorations of synaptic plasticity revealed that ADF mice exhibited enhanced paired-pulse facilitation at the CA3-CA1 synapse. Moreover, LTP could be elicited at the Schaffer's collateral–CA1 synapse using fewer high-frequency stimulations. These results are interesting since it is a well-accepted hypothesis that learning and memory are based on modifications of synaptic strength among neurons, an idea that goes at least back to Donald Hebb (1949).

A possible mechanism for the observed potentiation of synaptic plasticity is suggested by the observation that ADF mice also show changes in NMDA receptor subunit composition; this glutamate receptor is critical for many types of synaptic plasticity, and various isoforms exist which are composed of different protein subunits. It's known that the relative expression of NR2B subunits change over the course of a rodent's life; the fraction of NR2B subunits declines (and NR2A increases) in adulthood (Moyner et al., 1994). Fontán-Lozano et al. noted increased expression of NR2B NDMAR subunits in the hippocampus and perirhinal cortex of ADF mice.

Photomicrographs and immunohistochemical analyses of the NR2B expression pattern in the hippocampus of IFD (ADF) and control (ad libitum, AL) mice. The graphs represent the densitometric analysis of NR2B expression in the different areas of hippocampus (n = 5 animals per group in all tests). s luc, Stratum lucidum; mol, molecular layer; l mol, lacunosum moleculare layer; Molec, molecular layer; Lac mol, lacunosum moleculare layer; O.D., optical density. ***p ≤ 0.001. Source: Fontán-Lozano et al., 2007

Remarkably, the behavioral improvements as well as the synaptic enhancements seen in ADF mice are returned to control levels when these mice are administered a NR2B antagonist, which strongly suggests that the relative increase in NR2B subunit expression is responsible for arresting the cognitive decline that accompanies natural aging. This is reminiscent of the genetically-engineered smart mice created by Joe Tsien's group (Tang et al., 1999).

The causal mechanisms underlying the changes in NR2B expression are still unknown. Interestingly, exercise increases NR2B subunit expression in the hippocampus (Farmer et al., 2004), and there is some evidence that ADF can increase basal levels of activity (Carlson & Hoelzel, 1946). Whether there is a link between fasting-induced increases in activity and NR2B subunit expression is an important question for future research.

One thing to keep in mind is that enhanced synaptic plasticity may not necessarily be a good thing. Presumably, there was some selective pressure that led to downregulating NR2B subunit expression with age. It's not clear what this might be, but it may turn out that increased NR2B subunit expression is actually maladaptive once we figure out what the selective pressure was (is, it may still be acting?).