Absence of VGLUT3 Expression Leads to Impaired Fear Memory in Mice

Cue and spatial reference memory in VGLUT3^–/– mice. A–C, Watermaze experimental design: (A) Cuetask, (B) SRM task, and (C) SRM-reversal task. D–F, Mice were trained in 22°C water. VGLUT3^–/– mice showed no deficit of learning either in the cue (D), or the SRM and reversal (E), tasks. F, Memory assessment was performed 10 min (PT1 for SRM and SRM-R), 72 h (PT2 for SRM), and 48 h (PT2 for SRM-R) after training, and VGLUT3^–/– mice showed better performance at 72 h after training, and in the PT average. G–I, Mice were trained in water at 19°C. VGLUT3^–/– mice show no deficit of learning either in the Cue (G), or the SRM and reversal (H) tasks. I, No differences were observed in memory tests done at different times or on average. Data are mean ± SEM. Differences between genotypes: *p < 0.05. PT: probe test; R: reversal; NS: non significant. All corresponding statistics are presented in Extended Data Figure 1-1.

Object and spatial recognition in VGLUT3^–/– mice. A, Object recognition (OR): both groups show OR memory, with VGLUT3^–/– mice having higher scores. B, Spatial recognition (SR): both groups show comparable SR memory level. C, There is no correlation between learning sessions duration and VGLUT3^–/– mice performances in OR. Slopes are −0.01563 for WT and −0.001889 for VGLUT3^–/– mice. Data are mean ± SEM. Differences between genotypes: *p < 0.05. Differences to chance level: ##p < 0.01, ###p < 0.001. All corresponding statistics are presented in Extended Data Figure 2-1.

Contextual and cued fear memories of VGLUT3^–/– mice. A, Shock sensitivity assessment, i.e., the intensity for which the mice express a given behavior (movement, vocalization, running, or jump). B–F, Fear memories in VGLUT3^–/– mice. B, Freezing levels during fear conditioning consisting of two CS-US pairings. C, D, Contextual memory was tested 24 h after conditioning and revealed a more stable memory in VGLUT3^–/– mice. E, F, Cued memory test revealed high level of freezing to new context for VGLUT3^–/– mice. Data are mean ± SEM post hoc comparisons: *p < 0.05. All corresponding statistics are presented in Extended Data Figure 3-1.

Pattern separation of VGLUT3^–/– mice. A, Behavioral protocol. B, VGLUT3^+/+ mice performances. C, VGLUT3^–/– mice performances. D–G, Freezing levels on different days: (D) day 0, before conditioning; (E) day 1, VGLUT3^–/– mice already show a higher freezing level; (F) day 7, VGLUT3^+/+ mice start to discriminate the different contexts. G, On day 10, VGLUT3^–/– mice still do not discriminate the different contexts. Data are mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. All corresponding statistics are presented in Extended Data Figure 4-1.

Immediate shock in VGLUT3^–/– mice. Mice were subjected to a no shock (NS) or immediate shock (IS) protocol to assess their levels of spontaneous freezing after experiencing an electric footshock. They were either tested in the same context (IS-SC) or in a novel context (IS-NC). WT mice did not show any freezing after either procedure, whereas VGLUT3^–/– mice expressed a significant increase of freezing behavior only after the IS, in either context. Data are mean ± SEM post hoc comparisons: *p < 0.05; **p < 0.01; ***p < 0.001. All corresponding statistics are presented in Extended Data Figure 5-1.

Extinction fear in VGLUT3^–/– mice. A, Extinction learning over a 7-d period. Filled black and red circles represent the freezing levels of VGLUT3^+/+ mice and VGLUT3^–/– mice, respectively, before the presentation of the first CS. Open circles are used for the 10 subsequent CS. B–E, Fear memory in VGLUT3^–/– mice. B, C, The learning index (LI) was calculated to illustrate learning efficacy over time (B) and on average (C). D, E, Original memory was recalled on day 15. F, On day 18, freezing to CS was assessed in a new hexagonal context. Data are mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001. L: light (CS). All corresponding statistics are presented in Extended Data Figure 6-1.