Altered emotionality and neuronal excitability in mice lacking KCTD12, an auxiliary subunit of GABAB receptors associated with mood disorders

Abstract

Gamma-aminobutyric acid (GABA), the major inhibitory neurotransmitter in the brain, is fundamental to brain function and implicated in the pathophysiology of several neuropsychiatric disorders. GABA activates G-protein-coupled GABAB receptors comprising principal GABAB1 and GABAB2 subunits as well as auxiliary KCTD8, 12, 12b and 16 subunits. The KCTD12 gene has been associated with bipolar disorder, major depressive disorder and schizophrenia. Here we compare Kctd12 null mutant (Kctd12(-/-)) and heterozygous (Kctd12(+/-)) with wild-type (WT) littermate mice to determine whether lack of or reduced KCTD12 expression leads to phenotypes that, extrapolating to human, could constitute endophenotypes for neuropsychiatric disorders with which KCTD12 is associated. Kctd12(-/-) mice exhibited increased fear learning but not increased memory of a discrete auditory-conditioned stimulus. Kctd12(+/-) mice showed increased activity during the inactive (light) phase of the circadian cycle relative to WT and Kctd12(-/-) mice. Electrophysiological recordings from hippocampal slices, a region of high Kctd12 expression, revealed an increased intrinsic excitability of pyramidal neurons in Kctd12(-/-) and Kctd12(+/-) mice. This is the first direct evidence for involvement of KCTD12 in determining phenotypes of emotionality, behavioral activity and neuronal excitability. This study provides empirical support for the polymorphism and expression evidence that KCTD12 confers risk for and is associated with neuropsychiatric disorders.

Figure 1

Generation and characterization of Kctd12-deficient mice. (a) Schematic representation of WT and mutated Kctd12 alleles. The coding sequence (CDS) of the Kctd12 gene was replaced with a loxP-flanked neomycin resistance cassette (PGKneo) by homologous recombination in embryonic stem (ES) cells. Correctly targeted ES cells (Kctd12 knockout+PGKneo allele) were injected into blastocysts. A founder mouse was crossed with a mouse constitutively expressing Cre-recombinase to excise the neomycin cassette, leaving one loxP site behind (Kctd12 knockout allele). PCR primers for genotyping are indicated (P1, P2 and P3). (b) PCR of genomic DNA from Kctd12 WT (+/+), heterozygous (+/−) and null mutant (−/−) mice. The sizes of the amplified PCR products are indicated in base pairs (bp). (c) Western blot showing the absence of KCTD12 protein in whole-brain lysates of Kctd12^−/− mice. Kctd12^+/− and Kctd12^−/− mice express normal levels of GABA_B1 (GB1a,b), GABA_B2 (GB2), KCTD12b, KCTD8 and KCTD16 proteins. Equal loading was controlled with anti-tubulin antibodies. (d) Quantitative analysis of KCTD12 protein expression in selected brain regions microdissected from adult brain and in peripheral blood mononuclear cells (PBMCs). Kctd12^+/− mice express 30–40% of WT KCTD12 protein. In PBMC samples KCTD12 runs as a double band, most likely reflecting differences in posttranslational modification. Asterisks indicate crossreacting bands. Bar graphs summarize the amount of KCTD12 protein normalized to tubulin on the same blot (in % of WT). Data are means±s.e.m., n=4 mice per genotype. GABA, gamma-aminobutyric acid; KCTD, K⁺-channel tetramerization domain.

Figure 2

Effects of Kctd12 genotype on fear-conditioned freezing. CS–US fear conditioning: (a–f) Twenty-four mice per genotype underwent CS–US fear conditioning in a specific context on day 2 (a and d) and fear expression was then tested in either the same (N=12, b and e) or a novel (N=12, c and f) context on day 3. Freezing during the CS: (a) CS–US conditioning, (b) CS expression test in the same context, (c) CS expression test in a novel context. Freezing during inter-trial intervals: (d) CS–US conditioning, (e) expression test in the same context, (f) expression test in a novel context. Context–US fear conditioning (g and h): Twelve mice for each of WT and Kctd12^−/− genotypes underwent (g) context fear conditioning on day 1 and (h) context expression test on day 2. Data are mean±s.e.m. GT, genotype. ^*P<0.05 for Kctd12^−/− versus WT, ***P<0.001 for Kctd12^−/− versus WT. ^##P<0.01 for Kctd12^−/− versus Kctd12^+/−, ^###P<0.001 for Kctd12^−/− versus Kctd12^+/−. CS, conditioned stimulus; KCTD, K⁺-channel tetramerization domain; US, unconditioned stimulus; WT, wild type.

Figure 3

Effects of Kctd12 genotype on activity and operant drinking behavior in IntelliCage during a 20-day period. Light period 1900–0700 h: (a) total visits to operant corners, (b) total nosepokes in operant corners, (c) total water licks in operant corners. Dark period 0700–1900 h: (d) total visits to operant corners, (e) total nosepokes in operant corners, (f) total water licks in operant corners. Values are overall mean±s.e.m. for 10 mice per genotype. KCTD, K⁺-channel tetramerization domain; WT, wild type.

Figure 4

Increased electrical excitability and reduced Ba²⁺-sensitive K⁺ currents in Kctd12^−/− CA1 pyramidal neurons. (a) Square pulse current injection to determine the rheobase in WT and Kctd12^−/− neurons. The current amplitude required to generate at least one action potential (AP) was significantly reduced in Kctd12^−/− neurons. Note that Kctd12^−/− neurons fired APs at more hyperpolarized membrane potentials and with a reduced delay compared with WT neurons (quantification in Table 1). Scale bars, 200 ms, 20 mV. (b) Resting conductance in WT and Kctd12^−/− neurons during wash-in of 200 μM Ba²⁺. The resting conductance was monitored once per second by a single voltage step. (c) Current traces produced to a series of voltage steps in WT and Kctd12^−/− neurons to accurately determine the resting conductance before (CTRL) and after wash-in of Ba²⁺ at the time points depicted in b. The resting conductance was determined as the slope of the voltage–current (V–I) plots constructed from the current traces. Scale bars, 200 ms, 20 pA. (d) Summary bar graph of the resting conductance determined as in c. Data are mean±s.d. **P<0.01, ***P<0.001. KCTD, K⁺-channel tetramerization domain; WT, wild type.