Fig. 4.

(A) The probe trial showed a preference of exercised rats (Ex/Con) for the quadrant in which they had to swim for 60 s in the pool in which they received their training but with the escape platform removed. Blocking BDNF action prevented the exercise-induced preference for the target quadrant (Ex/IgG). Blocking the action of BDNF was selective for exercise as it did not affect the learning acquisition or recall abilities of sedentary animals (Sed/IgG). Each value represents the mean ± SEM (anova, *P < 0.05). (B) We performed a correlation analysis to evaluate a possible association between energy molecules and memory retention. We found a positive correlation between IGF-I mRNA levels and the memory retention ability (r = 0.68, P = 0.01) in the sedentary control (Sed/Con) and Ex/Con rats for the total pool of exercise and sedentary rats. The association between retention and levels of mRNAs was disrupted by the injection of the BDNF blocker TrkB-IgG (data not shown).