PDZD8 Disruption Causes Cognitive Impairment in Humans, Mice, and Fruit Flies

Abstract

Background: The discovery of coding variants in genes that confer risk of intellectual disability (ID) is an important step toward understanding the pathophysiology of this common developmental disability.

Methods: Homozygosity mapping, whole-exome sequencing, and cosegregation analyses were used to identify gene variants responsible for syndromic ID with autistic features in two independent consanguineous families from the Arabian Peninsula. For in vivo functional studies of the implicated gene's function in cognition, Drosophila melanogaster and mice with targeted interference of the orthologous gene were used. Behavioral, electrophysiological, and structural magnetic resonance imaging analyses were conducted for phenotypic testing.

Results: Homozygous premature termination codons in PDZD8, encoding an endoplasmic reticulum-anchored lipid transfer protein, showed cosegregation with syndromic ID in both families. Drosophila melanogaster with knockdown of the PDZD8 ortholog exhibited impaired long-term courtship-based memory. Mice homozygous for a premature termination codon in Pdzd8 exhibited brain structural, hippocampal spatial memory, and synaptic plasticity deficits.

Conclusions: These data demonstrate the involvement of homozygous loss-of-function mutations in PDZD8 in a neurodevelopmental cognitive disorder. Model organisms with manipulation of the orthologous gene replicate aspects of the human phenotype and suggest plausible pathophysiological mechanisms centered on disrupted brain development and synaptic function. These findings are thus consistent with accruing evidence that synaptic defects are a common denominator of ID and other neurodevelopmental conditions.

Keywords: Brain structure; Endoplasmic reticulum; Intellectual disability; Long-term memory; PDZD8; Synaptic plasticity.

Figure 1

Two families with PDZD8 mutations. (A) Pedigree of four-generation family A showing cosegregation of PDZD8 p.(S733∗) homozygosity with syndromic ID in 3 affected siblings (represented by filled symbols). (B) Pedigree of four-generation family B showing cosegregation of PDZD8 p.(Y298∗) homozygosity with syndromic ID in the affected individual (represented by filled symbol). Two progeny who died in utero are represented by small triangles. The numbers in generation III indicate brothers and sisters of the parents (B.III.1 and B.III.2). (C) Sanger sequence chromatograms showing the PDZD8 4-bp (AGTT) deletion (c.2197_2200del) identified in family A. (D) Sanger sequence chromatograms showing the PDZD8 nonsense mutation (c.894C>G) identified in family B. (E) Schematic diagram depicting domain structure and functions of PDZD8 in human (Q8NEN9; top), mouse (B9EJ80; middle), and Drosophila (Q9VYR9; bottom). The ER-TM domain (2–24 aa) and a region between the PDZ and phorbol-ester/diacylglycerol binding (C1) domains (466–797 aa) are required for interaction with protrudin (13,34,49). The SMP domain is required for the formation of PDZD8 dimers or oligomers (49). The SMP and PDZ domains are required for the extraction of lipids from the ER to late endosomes and lysosomes (13,49). The C1 domain is required for interaction with PS and PI4P associated with the late endosome/lysosome membrane (13,50). The CC domain is required for interaction with Rab-7a (34,49,51). Black horizontal lines indicate interactor binding sites; broken vertical red lines indicate the location of PTC (human: p.Y298∗ & p.S733∗; mouse: p.F333Nfs1∗). Numbering is from published sources (13,34,35). C, carboxyl-terminus; CC, coiled-coil; ER, endoplasmic reticulum; ID, intellectual disability; N, amino-terminus; PR, proline-rich; PS, phosphatidylserine; PI4P, phosphatidylinositol 4-phosphate; PDZ, PSD-95/DlgA/ZO-1-like; PTC, premature termination codon; SMP, synaptotagmin-like mitochondrial lipid-binding; TM, transmembrane; UAE, United Arab Emirates.

Figure 2

Associative learning and memory in CG10362 KD flies. (A) Expression of CG10362 in four pools of 8 to 10 whole adult CG10362-KD (KD: 0.4417 ± 0.09643), UAS-RNAi (UAS: 1.006 ± 0.04253), and Act5C-Gal4 (Gal4: 1.001 ± 0.04712) male flies (one-way analysis of variance: F_2,9 = 23.68, p < .0001; post hoc Bonferroni’s correction, Gal4 vs. UAS: p = 1.0, KD vs. UAS: p = .001, KD vs. Gal4: p = .001). (B) Aversive olfactory conditioning assay memory indices 30 seconds (learning) after training of KD (n = 8; 0.2393 ± 0.0442), UAS (n = 8; 0.35 ± 0.0634), and Gal4 (n = 7; 0.2024 ± 0.043) flies (one-way analysis of variance: F_2,20 = 2.189, p = .1382) and 30 minutes (short-term memory) after training of KD (n = 7; 0.2405 ± 0.0956), UAS (n = 7; 0.2969 ± 0.0394), and Gal4 (n = 7; 0.2161 ± 0.0512) flies (Kruskal-Wallis: χ²₂ = 1.2764, p = .5282). (C) Courtship conditioning assay memory indices immediately (0 hours) after training of KD (n = 17; 0.4327 ± 0.1782), UAS (n = 20; 0.3493 ± 0.0994), and Gal4 (n = 18; 0.3169 ± 0.0772) flies (Kruskal-Wallis: χ²₂ = 0.8324, p = .6595); 30 minutes after training of KD (n = 22; 0.4868 ± 0.1085), UAS (n = 24; 0.5326 ± 0.1542), and Gal4 (n = 19; 0.5144 ± 0.1067) flies (Kruskal-Wallis: χ²₂ = 0.8672, p = .6482); and 48 hours after training of KD (n = 33; 1.2102 ± 0.0902), UAS (n = 27; 0.7301 ± 0.0786), and Gal4 (n = 25; 0.8123 ± 0.0669) flies (one-way analysis of variance: F_2,82 = 10.52, p < .0001; post hoc Bonferroni’s correction, Gal4 vs. UAS: p = 1.0, KD vs. UAS: p = < .001, KD vs. Gal4: p = .003). Above dotted line (1.0) indicates no memory. Data are plotted as mean ± SEM. ∗∗p < .01 vs. controls; ^##p < .01 vs. Gal4; ∗∗∗p < .001 vs. UAS. KD, knockdown; UAS, upstream activation sequence.

Figure 3

Voxelwise volumetric differences in whole brain and specific brain regions in Pdzd8^tm1b mice determined by high-resolution structural magnetic resonance imaging. Significant differences in volume between Pdzd8^tm1b (n = 32 [10 males, 22 females]) and WT (n = 17 [7 males, 10 females]) are indicated by red (increased volume) and blue (reduced volume) contour shading on two-dimensional coronal slice images of the brain. (A) Absolute brain volume (mm³). (B) Cerebellum: relative volume (% total brain volume). (C) Olfactory bulb: relative volume (% total brain volume). (D) Hippocampus: relative volume (% total brain volume). (E) Retrosplenial cortex: relative volume (% total brain volume). (F) Thalamus: relative volume (% total brain volume). (G) Pallidum: relative volume (% total brain volume). (H) Superior colliculus: relative volume (% total brain volume). (I) Corpus callosum: relative volume (% total brain volume). ∗∗∗∗p < .0001 vs. WT. WT, wild-type.

Figure 4

Behavioral differences of Pdzd8^tm1b mice in OF and EPM. (A) Distance traveled (m) by Pdzd8^tm1b (n = 12) and WT (n = 14) mice in 10-minute intervals in OF, with lines of best fit shown (two-way repeated-measures analysis of variance, genotype: F_1,24 = 3.037, p = .094; time: F_2.44,58.65 = 23.17, p < .0001; interaction: F_2.44,58.65 = 3.795, p = .021). (B) Slope of habituation curve of Pdzd8^tm1b (−0.1625 ± 0.07680) and WT (−0.3727 ± 0.04413) mice (independent t test: t₂₄ = 2.458, p = .028). (C) Number of entries by Pdzd8^tm1b (92.67 ± 11.28) and WT (58.36 ± 8.195) mice to OF inner zone (independent t test: t₂₄ = −2.508, p = .019). (D) Time (s) spent by Pdzd8^tm1b (104.00 ± 21.44) and WT (51.69 ± 8.965) mice in OF inner zone (independent t test: t₂₄ = −2.710, p = .012). (E) Representative image of hindlimb jumping by Pdzd8^tm1b mouse in OF. (F) Number of jumps by Pdzd8^tm1b (n = 11; 165.9 ± 92.78) and WT (n = 12; 12.58 ± 6.289) mice in OF (Mann-Whitney U test: U = 97.00, p = .051). (G) Percentage of Pdzd8^tm1b and WT mice making more than 10 jumps per 10-minute interval in OF (Fisher’s exact test, 10–20 min: p = .037; 20–30 min: p = .037; 50–60 min: p = .037). (H) Number of entries to closed arms by Pdzd8^tm1b (17.0 ± 1.317) and WT (18.06 ± 1.184) mice in EPM (independent t test: t₃₂ = 0.597, p = .5541). (I) Number of entries to open arms by Pdzd8^tm1b (7.5 ± 1.258) and WT (4.056 ± 0.5686) mice in EPM (Welch’s t test: t₃₂ = −2.494, p = .021). (J) Number of head dips by Pdzd8^tm1b (21.81 ± 2.91) and WT (12.33 ± 1.06) mice in EPM (Welch’s t test: t₃₂ = −3.061, p = .006). (K) Total distance traveled (m) by Pdzd8^tm1b (n = 16; 12.69 ± 1.27) and WT (n = 18; 11.60 ± 0.79) mice in EPM (Mann-Whitney U test: U = 133, p = .720). Data are plotted as mean ± SEM. ∗p < .05; ∗∗p < .01 vs. WT. EPM, elevated plus maze; OF, open field; WT, wild-type.

Figure 5

Performance of Pdzd8^tm1b (n = 15) and WT (n = 10) mice in Barnes maze. (A) Latency (s) to first head entry to escape hole (Friedman’s analysis of variance, Pdzd8^tm1b: χ²₃ = 19.88, p < .0001; WT: χ²₃ = 21.30, p < .0001. Mann-Whitney U test, day 1: U = 79.50, p = .80; day 2: U = 85.50, p = .56; day 3: U = 78.00, p = .868; day 4: U = 98.00, p = .216). (B) Primary path length (m) (Friedman’s analysis of variance, Pdzd8^tm1b: χ²₃ = 11.32, p = .01; WT: χ²₃ = 23.88, p < .0001. Mann-Whitney U test, day 1: U = 60.00, p = .42; day 2: U = 76.00, p = 1.00; day 3: U = 83.00, p = .68; day 4: U = 92.00, p = .367). (C) Number of errors. (Friedman’s analysis of variance, Pdzd8^tm1b: χ²₃ = 10.50, p = .015; WT: χ²₃ = 9.39, p = .024. Mann-Whitney U test, day 1: U = 48.50, p = .144; day 2: U = 72.00, p = .892; day 3: U = 67.50, p = .683; day 4: U = 94.00, p = .311). (D) Time (s) spent by Pdzd8^tm1b (39.76 ± 4.983) and WT (53.45 ± 3.273) mice in target quadrant (Welch’s t test: t_22.24 = 2.296, p = .031. One-sample t test, Pdzd8^tm1b vs. chance [20]: t₁₄ = 3.965, p = .0014; WT vs. chance [20]: t₉ = 10.22, p < .0001). (E) Time (s) spent by Pdzd8^tm1b (11.39 ± 2.088) and WT (21.96 ± 2.579) mice in target sector (independent t test: t₂₃ = 3.223, p = .004. One-sample t test, Pdzd8^tm1b vs. chance [4]: t₁₄ = 3.540, p = .0033; WT vs. chance [4]: t₉ = 6.965, p < .0001). (F) Time (s) spent by Pdzd8^tm1b (6.587 ± 1.643) and WT (12.35 ± 2.3) mice within target hole annulus (Mann-Whitney U test: U = 36.00, p = .031). (G) Left, entry probability (%) of Pdzd8^tm1b (13.64 ± 2.359) and WT (27.8 ± 3.086) mice into the target hole annulus (independent t test: t₂₃ = 3.69, p = .001). Right, heat maps of mean entry probability (%) of Pdzd8^tm1b (right) and WT (left) mice. Data are plotted as mean ± SEM. ∗p < .05; ∗∗p < .01 vs. WT. T, target sector; WT, wild-type.

Figure 6

Analysis of hippocampal (long-term potentiation) in Pdzd8^tm1b mice. (A) Normalized change in fEPSP (% baseline) induced by 1× TBS in Pdzd8^tm1b (131 ± 4%; n = 11) and WT (136 ± 6%; n = 8) mice. (B) Normalized change in fEPSP (% baseline) induced by 3× TBS in Pdzd8^tm1b (131 ± 4%; n = 15) and WT (156 ± 7%; n = 16) mice. Scale bars = 0.3 mV and 10 ms in (A) and (B). (C) Normalized change in fEPSP (% baseline) induced by 1× HFS in Pdzd8^tm1b (145 ± 7%; n = 10) and WT (135 ± 6%; n = 12) mice. Insets: representative traces before (WT, light blue; Pdzd8^tm1b, pink) and after (WT, blue; Pdzd8^tm1b, red) the induction protocol. Scale bars = 0.2 mV and 10 ms. (D) Facilitation of fEPSP (% baseline) at 30 minutes after 1× TBS, 3× TBS, and 1× HFS. (E) Paired-pulse ratio of Pdzd8^tm1b (1.48 ± 0.03; n = 43) and WT (1.63 ± 0.04; n = 39) mice with 50-ms stimulus interval. Representative traces of WT (blue) and Pdzd8^tm1b (red) slices. Scale bars = 0.2 mV and 100 ms. Data are plotted as mean ± SEM. ∗p < .05 vs. WT. fEPSP, field excitatory postsynaptic potential; HFS, high-frequency stimulation; TBS, theta burst stimulation; WT, wild-type.