Skip to main content
. 2020 Jun 17;106(6):977–991.e4. doi: 10.1016/j.neuron.2020.03.013

Figure 1.

Figure 1

Spaced Training Induces LTM Comprised of Complementary CS+ and CS− Components

(A) Spaced training (six trials of CS+/CS− training with 15 min ITIs) generates 24 h LTM measurable when testing CS+ versus CS− odors. An aversive 24 h memory was measured when testing CS+ versus a novel odor, and this memory was reduced in comparison to the CS+ versus CS− 24 h performance. An appetitive 24 h memory was measured when CS− was tested against novel odor.

(B and C) If CS+ and CS− order was reversed during training (B) or intervals were omitted between training trials, massed training (C), the LTM (CS+ versus CS−) was not different from the aversive CS+ memory, and no approach was observed to CS−.

(D) A fasting LTM protocol that lacks repetition did not generate CS− approach memory, and CS+ memory was not different from that seen after training on CS+ versus CS−.

(E) Timeline of CS+ versus CS− memory performance after spaced training. Performance decays quickly for the first 3 h and stabilizes from 14 to 24 h. 96 h performance was reduced in comparison with 24 h performance.

(F) Timelines of CS+ versus novel odor (red) and CS− versus novel odor (blue) memory after spaced training. Significant CS− memory was only observed from 14 h and persisted for at least 96 h. The CS− memory was not significantly different from zero before 14 h. CS+ memory decayed between 3 and 14 h and remained constant between 14 and 24 h, and no 96 h performance was observed.

(G) rad mutant flies (hashed bars), in comparison to wild-type (WT) (Canton-S) flies (gray bars), had impaired LTM performance and lacked CS+ avoidance memory, but displayed normal CS− approach memory.

(H) CXM feeding impaired LTM performance after spaced training, causing a specific defect of CS− but not CS+ memory. WT flies fed 5% glucose (gray bars) or glucose laced with 35 mM CXM (stippled bars) for 12–16 h overnight before spaced training. Asterisks denote significant differences. Data are represented as means ± standard error of the mean (SEM). Individual data points are displayed as dots. See also Figure S1 and Table S1 for statistics.