Deletion of the γ-secretase subunits Aph1B/C impairs memory and worsens the deficits of knock-in mice modeling the Alzheimer-like familial Danish dementia

Abstract

Mutations in BRI2/ITM2b genes cause Familial British and Danish Dementias (FBD and FDD), which are pathogenically similar to Familial Alzheimer Disease (FAD). BRI2 inhibits processing of Amyloid precursor protein (APP), a protein involved in FAD pathogenesis. Accumulation of a carboxyl-terminal APP metabolite -ß-CTF- causes memory deficits in a knock-in mouse model of FDD, called FDDKI.We have investigated further the pathogenic function of ß-CTF studying the effect of Aph1B/C deletion on FDDKI mice. This strategy is based on the evidence that deletion of Aph1B/C proteins, which are components of the γ-secretase that cleaves ß-CTF, results in stabilization of ß-CTF and a reduction of Aβ. We found that both the FDD mutation and the Aph1B/C deficiency mildly interfered with spatial long term memory, spatial working/short-term memory and long-term contextual fear memory. In addition, the Aph1BC deficiency induced deficits in long-term cued fear memory. Moreover, the two mutations have additive adverse effects as they compromise the accuracy of spatial long-term memory and induce spatial memory retention deficits in young mice. Overall, the data are consistent with a role for β-CTF in the genesis of memory deficits.

Keywords: APP; Alzheimer; Danish dementia; Gerotarget; Itm2b-bri2; gamma-secretase.

Figure 1. Elevated Zero Maze test on mice at 4 months of age

Data are expressed as means ± S.E.M. No significant effect of genotype was found.

Figure 2. Open Field test on mice at 4 months of age

Data are expressed as means (± S.E.M.) during the 10-min testing period over 3 days. A. Total distance traveled. B. Amount of time in which the animal ambulated at speed greater than 50 mm/s. FDD_KI mice were more active than WT mice, especially on Day 1. C. Amount of time the animal spent in the center of the arena (20 cm × 20 cm). D. Total number of entries into the arena center. FDD_KI mice entered and spent more time in the arena center on Days 1 and 2 than WT mice. * p < 0.05, ** p < 0.01, *** p < 0.001, WT vs. FDD_KI.

Figure 3. The Morris Water Maze task shows a mild deficit in accuracy of spatial long-term memory 4 month-old FDD_KI/Aph1BC^−/− mice

Data are expressed as means ± S.E.M. A.-B. Performance on the visible platform task. The visible platform task indicates that there were no significant differences among the genotypes in path length traveled (A) or swim speed (B). C. Acquisition of spatial reference memory in the hidden platform task. Mean path lengths across 6 daily trials are shown. Significant differences among the genotypes were found only on the first day: * p < 0.05, ** p < 0.01, *** p < 0.001. D.-F. Performance on the 60-s probe trial given 2 days after the last acquisition session. (D-F) Performance on the 60-s probe trial given 2 days after the last acquisition session. (D) Percentage of time spent in the four quadrants. ** p < 0.01, *** p < 0.001, **** p < 0.0001 (E) Number of counter crossings in the target quadrant (Quadrant 4). * p < 0.05 (F) A separate unpaired t-test showed a significant difference between WT and FDD_KI/Aph1BC^−/− mice (* p = 0.0177). G. Average proximity to the former platform location. H. Performance in the reversal learning task with a new platform location. Mean path length across 6 daily trials are shown. I.-K. Performance on the 60-s probe trail given 2 days after the last reversal learning session. (I) Percentage of time spent in the four quadrants. **** p < 0.0001 (J) Number of counter crossings in the target quadrant (Quadrant 1). (K) Average proximity to the former platform location.

Figure 4. Normal spatial memory in 7-8 month-old mice

Data are expressed as means ± S.E.M. A.-B. Performance on the visible platform task. There were no significant differences among the genotypes in path length traveled (A) or swim speed (B). C. Acquisition of spatial reference memory in the hidden platform task. Mean path lengths across 3 daily trials are shown. No significant differences were found among the genotypes. D.-F. Performance on the 60-s probe trial given 2 days after acquisition. (D) Percentage of time spent in the four quadrants. *** p < 0.001, **** p < 0.0001. (E) Number of counter crossings in the target quadrant (Quadrant 2). (F) Average proximity to the former platform location. No significant differences were found among the genotypes in any of the measures.

Figure 5. The 8-arm Radial Arm Water Maze task detects possible working memory impairment in FDD_KI/Aph1BC^−/− mice at 7-8 months of age

Data are expressed as means ± S.E.M. A. Number of errors on 3 daily acquisition trials averaged across the last 3 test days. The performance by WT, FDD_KI, and Aph1BC^−/− mice improved across trials to a greater degree, compared to FDD_KI/Aph1BC^−/− mice. *p < 0.05, **p < 0.01, ***p < 0.001, **** p < 0.0001. B. Number of errors on trials 2, 3 and retention (R) given 30 min after the third acquisition trial. No significant differences were found among the genotypes.

Figure 6. Mild deficit in accuracy of spatial long-term memory in 15 month-old FDD_KI, Aph1BC^−/− and FDD_KI/Aph1BC^−/− mice

Data are expressed as means ± S.E.M. A.-B. Performance on the visible platform task. There were no significant differences among the genotypes in path length traveled (A) or swim speed (B). C. Acquisition of spatial reference memory in the hidden platform task. Mean path lengths across 2 daily trials are shown. No significant differences were found among the genotypes. D.-F. Performance on the 30-s probe trial given 2 days after acquisition. D. Percentage of time spent in the four quadrants. *** p < 0.001, **** p < 0.0001. E. Number of counter crossings in the target quadrant (Quadrant 1). WT mice crossed the target counter significantly more than FDD_KI or Aph1BC^−/− mice. *** p < 0.01, WT vs. FDD_KI, ** p < 0.01, WT vs. Aph1BC^−/−; * p < 0.05, WT vs. FDD_KI/Aph1BC^−/−. F. Average proximity to the former platform location.

Figure 7. Mild deficit in working memory in FDD_KI, Aph1BC^−/− and FDD_KI/Aph1BC^−/− mice at 15 months of age

Data are expressed as means ± S.E.M. A.-B. Performance in the visible platform task. There were no significant differences among the genotypes in path length traveled (A) or swim speed (B). C. Path length traveled on the 3 daily trials of the working memory task averaged across the last 3 test days. WT mice significantly reduced the path length traveled to reach the platform between trial 1 and 3 (* p < 0.05), while mice of the other 3 genotypes did not.

Figure 8. The two-trial Y-maze test showed mild deficit of short-term spatial recognition memory in FDD_KI, Aph1BC^−/− and FDD_KI/Aph1BC^−/− mice at 18-19 months of age

Data are expressed as means ± S.E.M. A. Total number of arm entries. FDD_KI mice made significantly more arm entries than WT or FDD_KI/Aph1BC^−/− mice (*p < 0.05). B. Percentage of entries into the novel (N) and known (K) arms. WT and, to a smaller degree, FDD_KI mice entered the novel arm significantly more than the known arm, while FDD_KI/Aph1BC^−/− or Aph1BC^−/− mice did not (**p < 0.01; **** p < 0.0001). C. Time spent in the novel and known arms. WT mice and, to a smaller degree, FDD_KI mice spent significantly more time in the novel arm than in the known arm, while FDD_KI/Aph1BC^−/− or Aph1BC^−/− mice did not (*p < 0.05; **** p < 0.0001). In addition, there was a significant difference in time spent in the novel arm between WT and FDD_KI/Aph1BC^−/− (*p < 0.05).

Figure 9. Open Field test on mice at 18-19 months of age

Data are expressed as means (± S.E.M.) during the 10-min testing period over 3 days. A. Total distance traveled. B. Amount of time in which the animal ambulated at speed greater than 50 mm/s. C. Amount of time the animal spent in the center of the arena (20 cm × 20 cm). D. Total number of entries into the arena center. No significant differences were found among the genotypes in any of the measures.

Figure 10. Y-maze spontaneous alternation test at 18-19 months of age

Data are expressed as means ± S.E.M. A. Total number of arm entries. B. Percentage of alternations. Six mice with less than three arm entries are not included in (B). No significant differences were found among the genotypes in either measure.

Figure 11. Elevated Zero maze test on mice at 18/-9 months of age

Mean (± S.E.M.) percentage of time spent in the open areas of the elevated zero maze. There were no significant differences among the genotypes.

Figure 12. Mild spatial working memory deficits in 18-19 month-old FDD_KI, Aph1BC^−/− and FDD_KI/Aph1BC^−/− mice

Data are expressed as means ± S.E.M. (A) Number of errors on 3 daily acquisition trials averaged across days 6-9 of testing. (B) Number of errors on trials 2, 3 and retention (R) given 30 min after the third acquisition trial. The visible platform task showed no significant differences among the genotypes in path length traveled (C) or swim speed (D). *p < 0.05, **p < 0.01, ***p < 0.001, **** p < 0.0001.

Figure 13. Deficit in long-term contextual fear memory in FDD_KI, Aph1BC^−/− and FDD_KI/Aph1BC^−/− mice at 18-19 months of age

Data are expressed as means ± S.E.M. A. Percentage of freezing during the last 3-m period of the contextual test. WT mice froze significantly more than mice of the other genotypes. *p < 0.05, WT vs. FDD_KI/Aph1BC^−/−; ** p < 0.01 WT vs. FDD_KI and WT vs. Aph1BC^−/−. B. Time course of freezing behavior during the contextual test in 1-m time bins. There was a significant genotype main effect (p < 0.01), with WT freezing more than the other genotypes. *p < 0.05, ** p < 0.01, WT vs. FDD_KI; £ p < 0.05, ££ p < 0.01, WT vs. Aph1BC^−/−; # p < 0.05, WT vs. FDD_KI/Aph1BC^−/−. C. Percentage of freezing during the 3-m tone presentation in the altered context in the cued test. WT mice froze more than mice of the other genotypes, although not significantly. D. Time course of freezing behavior during tone presentation in the cued test. Only WT mice remained frozen for the entire period. E. Sensitivity to varying intensity levels of foot shock. Mice of all the genotypes reacted to shock at the four intensity levels in a similar manner.