Hippocampal and cortical primary cilia are required for aversive memory in mice

Abstract

It has been known for decades that neurons throughout the brain possess solitary, immotile, microtubule based appendages called primary cilia. Only recently have studies tried to address the functions of these cilia and our current understanding remains poor. To determine if neuronal cilia have a role in behavior we specifically disrupted ciliogenesis in the cortex and hippocampus of mice through conditional deletion of the Intraflagellar Transport 88 (Ift88) gene. The effects on learning and memory were analyzed using both Morris Water Maze and fear conditioning paradigms. In comparison to wild type controls, cilia mutants displayed deficits in aversive learning and memory and novel object recognition. Furthermore, hippocampal neurons from mutants displayed an altered paired-pulse response, suggesting that loss of IFT88 can alter synaptic properties. A variety of other behavioral tests showed no significant differences between conditional cilia mutants and controls. This type of conditional allele approach could be used to distinguish which behavioral features of ciliopathies arise due to defects in neural development and which result from altered cell physiology. Ultimately, this could lead to an improved understanding of the basis for the cognitive deficits associated with human cilia disorders such as Bardet-Biedl syndrome, and possibly more common ailments including depression and schizophrenia.

Figure 1. Conditional loss of IFT88 in the adult hippocampus and cortex.

(A) PCR genotyping from whole genomic DNA prepped from the following brain regions: olfactory bulb (Ob), hippocampus (Hip), cortex (Ctx), hypothalamus (Hyp), and cerebellum (Cer). Bands from top to bottom are as follows: Emx1-cre transgene (cre), Ift88 floxed conditional allele (F), wildtype allele (W), mutant allele (Δ). Note the induction of the mutant band (Δ) when Emx1-cre (cre) occurs only in Ob, Hip, and Ctx and not in Hyp and Cer. Control samples from genomic DNA isolated from ear punches are as follows: Ift88^flox/flox; Emx1-cre (F/F), Ift88^flox/Δ (F/Δ), Ift88^{flox/wildtype}; Emx1-cre (F/W). (B) Immunoblotting from whole proteins isolated from Ctx, Hip, Ob, and Hyp probed for Ift88 and a loading control Gapdh. Control samples from Ift88^flox/flox (F) are next to experimental Ift88^Δ/Δ; Emx1-cre (Δ). Note the diminished levels of Ift88 in Δ samples only occurs in ctx and hip. (C) Cre reporter lacZ staining of whole mount brains shows the regions of cre activity. Top row shows a sagittal half and the bottom row shows transverse sections with most rostral to caudal from left to right.

Figure 2. Conditional loss of Cilia in the adult hippocampus and cortex.

(A) Immunofluorescence for the neuronal cilia marker adenylate cyclase III (red) in brain sections from the following regions of the brain: Ob, Ctx, and regions of the hippocampus the pyramidal layers CA1, CA3, and granule layer of the dentate gyrus (DG). Ift88^flox/flox control samples (top row) are compared to Ift88^Δ/Δ; Emx1-cre (bottom row). Hoechst nuclear stain is in blue. Scale bar is 50 µm. (B) Graph indicating the amount of cilia counted in several regions of the brain in control (Ift88^flox/flox, N = 3) and mutants (Ift88^Δ/Δ N = 4). Statistically significant differences are indicated: Student's t-test, * p<0.05.

Figure 3. Fear conditioning and novel object recognition in mice lacking hippocampal and cortical primary cilia.

(A) Percent of time spent freezing during training on day 1, bars and diamonds represent tones and shocks respectively (B) Day 2: Percent time spent freezing during the contextual test in same arena. (C) Day 3: Percent time freezing during the cued test in new arena, bar represents when tone was played. (D) Comparison of freezing time between cilia mutant and wild type mice. Mutant mice spent less time freezing. (E) Novel object recognition index as the time spent investigating the novel object relative to the total object investigation and (F) discrimination index as the exploration time devoted to the novel object minus the time devoted to the familiar object. Statistically significant differences are indicated: Student's t-test, * p<0.05, ** p<0.01.

Figure 4. Behavior testing in mice lacking primary cilia in hippocampus and cortex.

(A) Morris Water Maze displayed as latency to platform in seconds (s) during the experimental day. (B) RotaRod test results displays as time on rod in seconds (s) at specific speeds in rotations per minute (RPM). (C) Elevated plus maze displayed in events per minute for explorations and entrances (D) Hotplate test displaying time in seconds till reaction (s) (E and F) Open field test result displayed in distance in centimeters (cm) from the chamber periphery (E) and center (F). Control (Ift88^flox/flox) and mutant (Ift88^Δ/Δ) mice are indicated throughout figure. Animal numbers are indicated in parenthesis next to group. No statistically significant differences in any behavioral test were determined using Student's t-test.

Figure 5. Altered paired-pulse facilitation in mice lacking hippocampal and cortical primary cilia.

Field recordings from CA1 hippocampus after schaffer collateral stimulation. (A) example traces of paired fEPSPs recorded in wild type (white) and mutant (black) hippocampal slices. (B) Cilia mutant neurons show an increase in paired pulse facilitation. (C) Long term potentiation recordings after 4 100 hz trains. Statistically significant differences are indicated: Student's t-test, * p<0.05.