Mapping and Deciphering Neural Codes of NMDA Receptor-Dependent Fear Memory Engrams in the Hippocampus
(A) An example of various memory traces being retrieved during the 60-sec traced fear recall test in a control mouse (1-hr retention). A conditioned tone (2-sec) was presented without the reinforcing foot shock. The black bar on the top of spike raster indicates the non-freezing state, whereas the orange bar indicates the freezing state of the animal. Colored triangles or diamonds at the bottom indicate the various moments at which those memory traces were retrieved. (B) Examples of each type of memory traces retrieved during the traced recall. (C) Various memory traces being retrieved over all seven tone trials from a representative control mouse. Symbols: simple CS trace, blue triangle; simple US trace, red triangle; US-to-CS associative trace, red diamond; CS-to-US associative trace, blue diamond. The time window for recalling the anticipated foot shock memory is illustrated by dotted rectangle (18.5-24.5 seconds after the termination of the conditioned tone. Please note the simple or associative foot-shock traces at this time period. (D) Greatly reduced numbers of memory traces retrieved in the knockout mice during 1-hr traced retention test. Please note the lack of the simple US or associative US-to-CS traces at this trace-interval time period. (E) The average memory trace retrieval rates in the control and knockout group during 1-hr trace recall test (the central red lines are the medians, the edges of the box are the 25th and 75th percentiles, each row of ‘x’ markers indicate the result of seven trials from one animal, small random numbers were added to avoid overlap of the markers; student t-test, **p<0.01). (F) Memory trace-time interval analysis revealed that trace recall in the control group has the characteristics of exponential decay distribution, indicating it occurred in a ‘bursting’ manner. (G) Memory trace retrieved in the mutant mice did not show tight an exponential decay process. The inter-trace intervals of control and mutant mice formed different distribution (Two-sample Kolmogorov-Smirnov test, p = 5.3E-7). (H) Time distribution of US traces (include US simple trace and US-to-CS trace) in the control mice showed a distinct peak at about 22 second time-window at which a foot shock would be anticipated. This peak was absent in the mutant group. One way ANOVA test between two curve, p = 1.9E-6. (I) Time distribution of CS traces did not reveal any significant peaks in either control mice or knockout mice. (J) The occurrence of real-time shock memory traces around the 20-sec traced interval time window (18.5–24.5 sec after the offset of the recall tone) for all seven trials in five control mice. Red rectangles represent the occurrences of the correct foot-shock memory traces, whereas blue squares indicate the absence of foot-shock memory traces at the time point. (K) Lack of memory traces for time relationship in predicting foot shock event and its timing in five knockout mice. There is a significant difference in recalling memory of time relationship between CS and US between genotypes. Wilcoxon rank sum test, p<0.01.