Fig 2: Caspase-2 deficiency increases spine density both in vitro and in vivo. a Caspase-2 inhibitor (Z-VDVAD) significantly increased spine density in cultured rat hippocampal neurons compared with vehicle DMSO. DMSO, n = 19 neurons; Z-VDVAD, n = 15 neurons; Scale bar, 10 μm. b Knockdown of caspase-2 with shRNA (C2-sh) increased spine density in cultured rat hippocampal neurons compared with control shRNA (Con-sh). Con-sh, n = 9 neurons; C2-sh, n = 10 neurons; Scale bar, 10 μm. c Increased spine density in CA1 neurons of Casp2 KO mice. WT and KO, n = 6 mice; Scale bar, 10 μm. d, e Live-cell imaging revealed that spine elimination, but not spine formation, was impaired in caspase-2 knockdown neurons. Con-sh, n = 14 neurons; C2-sh, n = 19 neurons; Scale bar, 5 μm. Neurons were from at least three independent rat hippocampal cultures for analysis of dendritic spines. Data are expressed as mean ± SEM. Two-tailed Student’s t test: *p < 0.05 and **p < 0.01. Source data are provided as a Source Data file

Fig 3: Caspase-2 regulates the Akt-GSK3β signaling cascade through cleavage of Rictor. a Immunoblotting analysis and quantification of total Akt, Ser473-Akt, total GSK3β, and Ser9-GSK3β in the hippocampus of WT and Casp2 KO mice. Tubulin was used as a loading control. n = 4 per genotype. b Levels of Rictor and mTOR in the hippocampus of WT and Casp2 KO mice. WT, n = 4; KO, n = 5. c Cleavage of Rictor in cultured mouse hippocampal neurons infected with lentivirus expressing caspase-2L. Note that full-length Rictor is 200 kDa, whereas the two cleavage products are above and below 100 kDa, respectively. d Immunoblotting analysis and quantification of total Akt, Ser473-Akt, total GSK3β, Ser9-GSK3β, mTOR, and pS6K in cultured mouse hippocampal neurons. Neurons on DIV6 were infected with lentivirus expressed GFP or caspase-2L and lysed 96 h later. n = 3 for each viral vector. e Immunoblotting analysis and quantification of Rictor, total Akt, Ser473-Akt, total GSK3β, and Ser9-GSK3β in cultured WT and Casp2 KO mouse hippocampal neurons treated with vehicle or NMDA. n = 3. f Knockdown of Rictor with shRNA (Rictor-sh) reduced spine density in cultured hippocampal neurons compared with control shRNA (Con-sh). Con-sh, n = 10 neurons; Rictor-sh, n = 14 neurons; Scale bar, 5 μm. Data are presented as mean ± SEM. Two-tailed Student’s t test: n.s. for p > 0.05, *p < 0.05, **p < 0.01, and ***p < 0.001. Source data are provided as a Source Data file

Fig 4: Casp2 KO mice display elevated anxiety, enhanced fear memory and impaired cognitive flexibility. a, b Time spent and distance traveled in the center zone in open field tests. n = 18 WT mice and 21 KO mice. c Transition numbers between two chambers in light–dark exploration tests. n = 15 WT mice and 22 KO mice. d, e Percentage of time spent in freezing and distance traveled in contextual fear conditioning (FC) tests. n = 18 WT mice and 22 KO mice. f Schedule for training sessions and probe trials in water maze tests. g Latency for mice to locate the hidden platform during reversal spatial learning. n = 15 WT mice and 21 KO mice. h Percentage of time spent in the target quadrant during the 2nd probe trial and the remote probe trial. n = 15 WT mice and 20 KO mice. i Number of time when mice cross the location of the hidden platform during the 2nd probe trial and the remote probe trial. n = 15 WT mice and 20 KO mice. j Performance at the probe trials. Duration spent in crossing the location of the hidden platform was significantly shorter on day 27 than on day 17 in WT mice but not KO mice. Two-tailed Student’s t test for a–c. Two-way ANOVA with Bonferroni post hoc test for d–j. *p < 0.05, **p < 0.01 and ***p < 0.001. Data are presented as mean ± SEM. Source data are provided as a Source Data file

Fig 5: Caspase-2 deficiency impairs NMDA-induced spine shrinkage, diminishes LTD expression and alters decay kinetics. a, b Effects of caspase-2 knockdown or inhibition on spine head size in cultured rat hippocampal neurons. Neurons were transfected with constructs expressing Con-shRNA (Con-sh) or Casp2 shRNA (C2-sh) on DIV14. Alternatively, the caspase-2 inhibitor Z-VDVAD-FMK (10 μM) or DMSO was added to neuronal cultures on DIV21. For visualization of spines, neurons were transfected with a construct expressing EGFP-actin on DIV14. On DIV28, neurons were treated with vehicle or 40 μM NMDA for 5 min and fixed 30 min after NMDA treatment for analysis of dendritic spines. Con-sh and Veh, 281 spines from 10 neurons; C2-sh and Veh, 292 spines from 10 neurons; Con-sh and NMDA, 178 spines from 6 neurons; C2-sh and NMDA, 356 spines from 12 neurons; DMSO and Veh, 369 spines from 13 neurons; Z-VDVAD and Veh, 308 spines from 11 neurons; DMSO and NMDA, 208 spines from 7 neurons; Z-VDVAD and NMDA, 299 spines from 10 neurons. c Normal LTP expression at Schaffer collateral-CA1 synapses of Casp2 KO mice, as revealed by measurement of field excitatory postsynaptic potentials (fEPSP). LTP was induced with high-frequency stimulation (HFS). WT, 10 slices from 6 mice; KO, 11 slices from 7 mice. Scale bars, 1 mV (vertical) and 20 ms (horizontal). d Impaired LTD expression at Schaffer collateral-CA1 synapses of Casp2 KO mice. LTD was induced with low-frequency stimulation (LFS). WT, 13 slices from 5 mice; KO, 14 slices from 8 mice. Scale bars, 1 mV (vertical) and 20 ms (horizontal). e Cumulative probability of mEPSC decay time in CA1 pyramidal neurons of WT and Casp2 KO mice. Mean decay time was shown in the bar graph. WT, 840 events from 17 neurons in 5 mice; KO, 991 events from 20 neurons in 6 mice. Scale bars, 20 pA (vertical) and 10 ms (horizontal). f Decay time of fEPSP at CA1 pyramidal neurons. WT, 22 slices from 9 mice; KO, 22 slices from 11 mice. Scale bars, 0.5 mV (vertical) and 10 ms (horizontal). Data are expressed as mean ± SEM. Data were analyzed using two-tailed Student’s t test, except e where Mann–Whitney test was used. n.s. no significance, *p < 0.05, **p < 0.01, and ***p < 0.001. Source data are provided as a Source Data file