Disruption of the Zdhhc9 intellectual disability gene leads to behavioural abnormalities in a mouse model

Abstract

Protein S-acylation is a widespread post-translational modification that regulates the trafficking and function of a diverse array of proteins. This modification is catalysed by a family of twenty-three zDHHC enzymes that exhibit both specific and overlapping substrate interactions. Mutations in the gene encoding zDHHC9 cause mild-to-moderate intellectual disability, seizures, speech and language impairment, hypoplasia of the corpus callosum and reduced volume of sub-cortical structures. In this study, we have undertaken behavioural phenotyping, magnetic resonance imaging (MRI) and isolation of S-acylated proteins to investigate the effect of disruption of the Zdhhc9 gene in mice in a C57BL/6 genetic background. Zdhhc9 mutant male mice exhibit a range of abnormalities compared with their wild-type littermates: altered behaviour in the open-field test, elevated plus maze and acoustic startle test that is consistent with a reduced anxiety level; a reduced hang time in the hanging wire test that suggests underlying hypotonia but which may also be linked to reduced anxiety; deficits in the Morris water maze test of hippocampal-dependent spatial learning and memory; and a 36% reduction in corpus callosum volume revealed by MRI. Surprisingly, membrane association and S-acylation of H-Ras was not disrupted in either whole brain or hippocampus of Zdhhc9 mutant mice, suggesting that other substrates of this enzyme are linked to the observed changes. Overall, this study highlights a key role for zDHHC9 in brain development and behaviour, and supports the utility of the Zdhhc9 mutant mouse line to investigate molecular and cellular changes linked to intellectual disability and other deficits in the human population.

Keywords: H-Ras; Intellectual disability; Palmitoylation; S-acylation; zDHHC enzymes; zDHHC9.

Fig. 1

PCR and immunoblotting analysis of Zdhhc9 expression in wild-type and mutant mice. (A) Agarose gel electrophoresis of end-point PCR products from WT and mutant brain cDNA samples. mRNA confirmation assay 1 (MA1) primers were designed to amplify a Zdhhc9 mRNA region of 283 bp from the 5′UTR to Exon 2. mRNA confirmation assay 2 (MA2) primers were designed to amplify a Zdhhc9 mRNA region of 154 bp from Exon 2 to Exon 3. z9v1 primers were designed to amplify a mRNA region of 139 bp from exon 5 to exon 7, and z9v2 primers were designed to amplify a mRNA region of 106 bp from exon 5 to exon 7. HyperLadder 100 bp ladder (Bioline) was used as a marker of DNA size. (B) Comparison of the average ΔCt values of WT and mutant mouse brain samples (n = 3 WT, 3 mutant) for MA2 and z9v1 after normalisation against Tbp and Hprt1 reference genes. cDNA from WT and mutant mouse brain was amplified for 40 cycles using specific primers for the different targets (MA2, z9v1, Tbp and Hprt1) and SYBR Select Master Mix. Statistical analysis (unpaired t-test, Minitab version 17) revealed a significant effect of genotype for MA2-Tbp (p = .042), MA2-Hprt1 (p = .046) and z9v1-Hprt1 (p = .037). p value for z9v1-Tbp was 0.075. (C) PCR products amplified from cDNA derived from mRNA extracted from WT and Zdhhc9 mutant mouse brain. Primers were designed to anneal to a region in exon 1 (see MAO F in Table 1 for sequence of primer) and the 3′-UTR (3UTR R in Table 1 for sequence of primer). HyperLadder 1 kb ladder (Bioline) was included as marker of DNA size. (D) Lysates from HEK293T cells transfected with HA-tagged zDHHCs were probed with antibodies against zDHHC9 (top) and HA (bottom). (E) Brain lysates from WT and mutant mice were probed with antibodies against zDHHC9 (top) and beta actin (bottom). Position of molecular weight markers is shown on the left of all immunoblots.

Fig. 2

Comparison of wild-type and Zdhhc9 mutant mice in hanging wire and rotarod tests. (A) Average time spent by mice on the wire. Statistics were conducted using an unpaired t-test in Minitab version 17, t(31) = −3.3, p = .002 for effect of genotype on time spent on the wire. (B) Average time spent by mice on the rotarod. Statistics were conducted using an unpaired t-test in Minitab version 17, ns for genotype (n = 20 WT, 14 mutant).

Fig. 3

Comparison of wild-type and Zdhhc9 mutant mice in the open field test. Time spent by WT and mutant mice in the outer and inner sections of the test and their velocity during this time. (A–C) displays these parameters during the habituation phase of the test; (D–F) displays these parameters during the test time (n = 14 mutant, 20 WT). Statistics were conducted using general linear model, repeated measures in SPSS version 22 with time bin as the within-subjects factor and genotype as the between-subjects factor; F(1,32) = 4.646, p = .039 for effect of genotype in time spent in inner and outer zones and F(1,32) = 3.44, p = .073 for effect of genotype on velocity during habituation. Moreover, F(2,64) = 32.573, p < .001 for effect of time bin on time spent in inner and outer zones during habituation. There was no interaction between time bin and genotype for the time spent in inner and outer zones, F(2,64) = 0.482, p = .482. During test time, p = ns (non-significant) for effect of genotype on time spent in inner and outer zones and velocity. Each time bin represents a period of 5 min.

Fig. 4

Performance of WT and Zdhhc9 mutant mice in the elevated plus maze. Distance moved (A), time spent in closed arms (B), time spent in open arms (C), and time spent in final third of open arms (D) are shown (n = 20 WT, 14 mutant). Statistics were conducted using an unpaired t-test in Minitab version 17; t(22) = 2.49, p = .021 for effect of genotype on time spent in the open arms, t(29) = −2.84, p = .008 for effect of genotype on time spent in closed arms and t(24) = 2.74, p = .011 for effect of genotype on time spent in the final thirds of open arms.

Fig. 5

Comparison of acoustic startle and PPI response of wild-type and Zdhhc9 mutant mice. (A) Startle reactivity of WT and Zdhhc9 mutant mice. The startle response to a sound at a range of levels (dB) was measured. Statistics were conducted using general linear model, repeated measures in SPSS version 22; F(1,44) = 13.622, p = .001 for effect of genotype (n = 24 WT, 22 mutant). (B) Mice received a pre-pulse at the levels (dB) shown. Following this, the startle response to a 120 dB sound was measured. Statistics were conducted using general linear model, repeated measures in SPSS version 22; p > .05 for effect of genotype (n = 24 WT, 22 mutant).

Fig. 6

Performance of Zdhhc9 mutant and WT mice in the Morris water maze. Distance moved (A), latency (B) and mean velocity (C) during the 5 acquisition days of the experiment are shown. Statistics were conducted using general linear model, repeated measures in SPSS version 22 with experiment day as the within-subjects factor and genotype as the between-subjects factor; F(1,44) = 8.935, p = .005 for effect of genotype on distance moved, F(4,176) = 27.647, p < .001 for effect of day on distance moved while no interaction was found between day and genotype with F(4,176) = 1.020 and p = .399. F(1,44) = 9.677, p = .003 for effect of genotype on latency, F(4,176) = 23.499 and p < .001 for effect of day on latency while no interaction was found between day and genotype with F(4,176) = 0.868 and p = .484 (n = 26 WT, 20 mutant). The effect of genotype on velocity was non-significant. (D) Time spent in quadrants during the first 30 s and last 30 s of the probe trial to assess reference memory are shown. Statistics were conducted using general linear model, repeated measures in SPSS version 22; F(1,44) = 49.573, p < .001 for effect of quadrant during the first 30 s and F(1,44) = 41.805, p < .001 for effect of quadrant during the last 30 s (n = 26 WT, 20 mutant). (E) Time spent to reach the visible platform for WT and Zdhhc9 mutant mice during the visual cue trial of Morris water maze. Statistics were conducted using unpaired t-test in Minitab version 17; t(42) = −0.92, p = ns for effect of genotype (n = 26 WT, 20 mutant).

Fig. 7

MRI analysis of brains from WT and Zdhhc9 mutant mice. (A) Coronal images from WT and mutant mouse brains after ex vivo MRI scan in a 9.4 Tesla magnet. (B and C) Box and whiskers graphs showing the various data points for the volume of corpus callosum and hippocampus. The whiskers indicate the maximum and minimum data points while the thick horizontal line indicates the median. (B) without normalisation; (C) with normalisation to total brain volume (n = 3 WT, 3 mutant).

Fig. 8

Analysis of membrane association and S-acylation of H-Ras in whole brain and hippocampus from WT and Zdhhc9 mutant mice. (A) Membrane and cytosol fractions from whole brain homogenates were resolved by SDS-PAGE and transferred to nitrocellulose, and subsequently probed with GAPDH, Syntaxin and H-Ras antibodies. The left panel shows representative immunoblots (position of molecular weight markers is shown on the left), whereas the right panel shows quantified data for % membrane association. Statistical analysis using an unpaired t-test indicated that there was no significant difference in H-Ras membrane association between WT and mutant samples; p > .05 (n = 3 WT, 3 mutant). M: membrane fraction, C: cytosolic fraction. (B) Acyl-RAC samples from whole brain or hippocampi were resolved by SDS-PAGE and transferred to nitrocellulose, and subsequently probed with H-Ras antibody. The upper panel shows representative immunoblots (position of molecular weight markers is shown on the left), whereas the lower panel shows quantified data from whole brain or hippocampi. Statistical analysis using an unpaired t-test indicated that there was no significant difference in H-Ras S-acylation between WT and mutant samples; p > .05 (n = 3 WT, 3 mutant). TI: total input, UT: Unbound Tris treated fraction, BT: Bound Tris treated fraction, UH: Unbound HA treated fraction, BH: Bound HA treated fraction.