Data from clinical, epidemiological, and preclinical studies have suggested that individuals with dysfunction in energy balance and a neuroinflammation state bear an increased risk of AD. In fact, it is well known that compromised mitochondrial bioenergetics lead to over-production and mitochondrial accumulation of Ab, which is coupled to oxidative stress. The prevailing ”amyloid cascade”hypothesis emphasizes the neurotoxic characteristics of Ab as the main contributor to disease progression. However, rather than being the cause of the disease, Ab may be considered as a reactive product that arise from increased vulnerability to genetic and environmental risk factors as a function of aging. Moreover, candidate drugs that directly target amyloid pathways,, failed to achieve efficacy in recent clinical trials, indicating the limited therapeutic value of amyloid-specific strategies. Increasing evidence suggests that Alzheimer’s disease is a multifaceted disease that could at least in part be attributed to a decline in mitochondrial function and altered brain metabolic activity. In this paper we report that a single icv injection of CNF1 improved 20S-Notoginsenoside-R2 memory processes in 12-month old apoE4 mice compared with apoE3 mice. To perform studies on memory, animals were behaviorally tested first in the Morris water maze and subsequently in the PA test. This temporal order took into account the level of stress associated with each procedure. Using MWM test we show that, during the acquisition phase, the escape latencies to find the hidden platform were shorter in saline-treated apoE3 mice compared to saline-treated apoE4 mice, indicating a clear impairment in learning skills. Anyway, apoE4 exposed to CNF1 showed an improvement during the acquisition phase and the escape latencies were comparable with those of the no-pathological control strain apoE3. Furthermore, while our behavioral results revealed that sal-apoE4 had a clear impairment in memory retention, CNF1 treatment improved apoE4 mice retention. It is noteworthy that CNF1 did not alter performances in apoE3 animals. The PA test is an amygdale-dependent test which evaluates the ability of mice to learn and to retain an associative rule. CNF1 apoE4-treated mice showed definite longer latencies to enter the dark compartment during the test session of the passive avoidance test. As a result, these animals were able to learn or to retain the rule: black compartment is equal to electrical shock. The results from the passive avoidance test suggest that at 12 months of age apoE4 mice have not impaired passive avoidance memory retention compare to apoE3. These results are in line with previous experimental evidence where it has been shown that there was no genotype on any classical measure of passive avoidance response, such as latency to enter the dark compartment during acquisition and retention sessions. On the contrary, CNF1-apoE4 retained contextual fear indicating that amygdala-dependent long-term memory is affected by CNF1. Taken toghether these data could indicate that CNF1 ameliorate no-spatial memory performances only in pathological condition. Since clinical and experimental evidences show that the presence of apoE4 correlated positively with an Alprostadil anxious state, we studied the effects of CNF1 on anxiety-like behavior tracts. These studies were performed prior WM and PA taking into account the level of stress associated with each procedure. Our data show that sal-apoE4 did not increase measures of anxiety when compared to sal-apoE3 mice and the icv treatment with CNF1 did not induce any changes in EPM and OF performances. Furthermore, there were no group differences in total locomotor activity measurements both in MWM, EPM and OF.