PDZD8-deficient mice manifest behavioral abnormalities related to emotion, cognition, and adaptation due to dyslipidemia in the brain

Abstract

Although dyslipidemia in the brain has been implicated in neurodegenerative disorders, the molecular mechanisms underlying its pathogenesis have been largely unclear. PDZD8 is a lipid transfer protein and mice deficient in PDZD8 (PDZD8-KO mice) manifest abnormal accumulation of cholesteryl esters (CEs) in the brain due to impaired lipophagy, the degradation system of lipid droplets. Here we show the detailed mechanism of PDZD8-dependent lipophagy. PDZD8 transports cholesterol to lipid droplets (LDs), and eventually promotes fusion of LDs and lysosomes. In addition, PDZD8-KO mice exhibit growth retardation, hyperactivity, reduced anxiety and fear, increased sensorimotor gating, and impaired cued fear conditioned memory and working memory. These results indicate that abnormal CE accumulation in the brain caused by PDZD8 deficiency affects emotion, cognition and adaptive behavior, and that PDZD8 plays an important role in the maintenance of brain function through lipid metabolism.

Keywords: Behavior; Cholesterol; Dyslipidemia; Knockout mouse; Lipophagy; PDZD8.

Fig. 1

PDZD8 promotes fusion of LDs with lysosomes mediated by cholesterol. a Confocal fluorescence images of PC12 cells transfected with siControl or siPDZD8 as well as with expression vectors for EGFP-PLIN2 (green) and then metabolically labeled with LysoTracker (red). The boxed regions in the left panels are shown at higher magnification in those to the right. b TEM images of PC12 cells transfected with siControl or siPDZD8. Arrowheads in the images at higher magnification (enlarged-1) indicate LDs. Arrows in the enlarged-1 indicate organelles during lipophagy, which are shown at much higher magnification in those to the right (enlarged-2). LD and lysosome (Ly) are also indicated in the enlarged-2. c Confocal fluorescence images of PC12 cells transfected with siControl or siPDZD8 as well as with expression vectors for mCherry-PLIN2 (red) and then metabolically labeled with filipin (blue). The boxed regions in the left panels are shown at higher magnification in those to the right. Filipin is shown in gray scale in the enlarged images. d and e Intensity of filipin per cell (n = 17 and 16 cells for siControl and siPDZD8, respectively) (d) or per LD (n = 772 and 719 LDs for siControl and siPDZD8, respectively) (e) in images similar to those in (c). Quantitative data are presented with SEM values. n.s., not significant; ***P < 0.001

Fig. 2

Gene targeting of the mouse Pdzd8 locus. a Schematic representation of the domain structure of PDZD8, the corresponding ORF, and the genomic structure of the Pdzd8 locus. Exons (Ex) and introns of the gene are shown as boxes and lines, respectively. b Schematic representation of the domain structure for the ORF products of wild-type (WT), Ex3d (from EUCOMM) [25], and Ex1d alleles of Pdzd8. The genotype, genotype abbreviation, as well as corresponding numbers of nucleotides and amino acids are also shown. c Schematic representation of the genomic structure around exon 1 of Pdzd8 showing the positions targeted by the two sgRNAs for generation of the Ex1d allele. The sequences of the WT and Ex1d alleles are also shown, with exon 1 and the adjacent upstream and intron sequences being highlighted in black and gray, respectively

Fig. 3

General health and neurological screen of PDZD8-KO mice. a, b Body weight (a) and body temperature (b) of male WT and PDZD8-KO mice at 10 to 20 weeks of age. c Grip strength. d Latency to fall in the wire-hang test. e Latency of the first paw response in the hot-plate test of pain sensitivity. f Latency to fall from the rotating rod in the rotarod test of motor function. All data are means ± s.e.m. (WT mice, n = 18; PDZD8-KO mice, n = 19). P values for differences between genotypes were determined with Student’s t test (a, c), the Mann–Whitney U test (b, d, e), or two-way repeated-measures ANOVA (f). ***P < 0.001; n.s., not significant. The results of all statistical analysis are provided in Additional file 2: Table S2

Fig. 4

Anxiety and fear-related behavior of PDZD8-KO mice. a–d Distance traveled in the light and dark chambers (a), time spent in the light and dark chambers (b), number of transitions between the light and dark chambers (c), and latency of entry into the light chamber (d) for the light–dark transition test. e–h, Distance traveled (e), vertical activity (f), center time (g), and stereotypic counts (h) for the open-field test. i–l, Distance traveled (i), total number of arm entries (j), percentage of entries into the open arms (k), and percent time spent in the open arms (l) for the elevated plus-maze test. All data are means ± s.e.m. (WT mice, n = 18; PDZD8-KO mice, n = 19 for a–d and n = 18 for e–l). P values for differences between genotypes were determined with Student’s t test (a–c, k), the Mann–Whitney U test (d, i, j, l), or two-way repeated-measures ANOVA (e–h). *P < 0.05, ***P < 0.001; n.s., not significant. The results of all statistical analysis are provided in Additional file 2: Table S2

Fig. 5

Contextual and cued fear memory of PDZD8-KO mice. a–j, Percent freezing time (a–e) and distance traveled (f–j) during conditioning (a, f), in a context test performed 1 day (b, g) or 28 days (d, i) after conditioning, and in a cued test with altered context performed 1 day (c, h) or 28 days (e, j) after conditioning in the contextual and cued fear conditioning test. Mice were presented three times with white noise as a conditioned stimulus (CS) for 30 s (horizontal black bars) followed by foot shock as an unconditioned stimulus (US) for the last 2 s of the conditioned stimulus (vertical arrows) during the conditioning session. k Shock sensitivity assessed on the basis of the distance traveled during and after exposure to the unconditioned stimulus. All data are means ± s.e.m. (n = 17 for WT and PDZD8-KO mice). Statistical analysis of differences between genotypes was performed with two-way repeated-measures ANOVA (a–k). For the cued test (c, e, h, j), separate P values are shown for the first and second 3-min periods as well as for the time point of 6 min (vertical arrows). *P < 0.05, **P < 0.01; n.s., not significant. The results of all statistical analysis are provided in Additional file 2: Table S2

Fig. 6

Acoustic startle response and prepulse inhibition test. Startle amplitude (A.U., arbitrary units) (a) and prepulse inhibition (b) were determined in the acoustic startle response and prepulse inhibition test. All data are means + s.e.m. (WT mice, n = 18; PDZD8-KO mice, n = 19). Statistical analysis of differences between genotypes was performed with Student’s t test (a, for 120 dB; b, for 78/110, 74/120, and 78/120 dB) or the Mann–Whitney U test (a, for 110 dB; b, for 74/110 dB). *P < 0.05, **P < 0.01; n.s., not significant. The results of all statistical analysis are provided in Additional file 2: Table S2

Fig. 7

Social behavior of PDZD8-deficient mice. a–e Duration of contacts (a), number of contacts (b), duration of active contacts (c), mean duration per contact (d), and distance traveled (e) in the social interaction test of social behavior in a novel environment. Data are means + s.e.m. (n = 9 for WT and PDZD8-KO mice). f–i, Mean number of particles calculated for each hour over 7 days (f) and averaged over the last 3 days (g) as well as mean activity level (A.U., arbitrary units) for each hour over 7 days (h) and averaged over the last 3 days (i) in the home-cage social interaction test of social behavior in a familiar environment. Time is indicated in Zeitgeber time (ZT). Data are means ± s.e.m. (n = 8 for WT and PDZD8-KO mice). j–p, The time spent in the chamber containing an empty cage, in the center chamber, and in the chamber with a cage containing an unfamiliar mouse (stranger 1) (j); the time spent around each cage (l); the number of approaches to each cage (n); and the distance traveled (p) were determined in the three-chamber sociability test (1st trial). The time spent in the chamber containing caged stranger 1, the center chamber, and the chamber with a cage now containing a novel unfamiliar mouse (stranger 2) (k); the time spent around each cage (m); the number of approaches to each cage (o); and the distance traveled (p) were also determined in the three-chamber social novelty preference test (2nd trial). Data are means + s.e.m. (WT mice, n = 18; PDZD8-KO mice, n = 19). P values for differences between genotypes were determined with Student’s t test (a, b, d, e, j, l–p), the Mann–Whitney U test (c, j–l, o), two-way repeated-measures ANOVA (f–i), or Welch’s t test (k, m).*P < 0.05, **P < 0.01, ***P < 0.001; n.s., not significant. The results of all statistical analysis are provided in Additional file 2: Table S2

Fig. 8

Absence of depression-related behavior in PDZD8-deficient mice. a–d Percentage immobility time (a, b) and distance traveled (c, d) on day 1 (a, c) and day 2 (b, d) in the Porsolt forced swim test. e Percentage immobility time in the tail-suspension test. All data are means ± s.e.m. (WT mice, n = 18 in a–d and n = 17 in e; PDZD8-KO mice, n = 19 in a–d and n = 17 in (e). P values for differences between genotypes were determined by two-way repeated-measures ANOVA. *P < 0.05; n.s., not significant. The results of all statistical analysis are provided in Additional file 2: Table S2

Fig. 9

Learning and memory of PDZD8-KO mice. a–d Latency to reach the target hole (a) as well as the number of errors (b), distance traveled (c), and number of omission errors (d) before reaching the target hole during training sessions in the Barnes maze test of spatial-learning memory and remote memory. e, f Time spent around each hole in probe trials conducted 1 day (e) and 30 days (f) after the last training session of the Barnes maze test. All data for a–f are means ± s.e.m. (WT mice, n = 17; PDZD8-KO mice, n = 16). g Percent correct responses in the T-maze test of working memory. Data are means + s.e.m. (WT mice, n = 18; PDZD8-KO mice, n = 19). P values for differences between genotypes were determined by two-way repeated-measures ANOVA (a–d, g) or the Mann–Whitney U test (e, f). **P < 0.01; n.s., not significant. The results of all statistical analysis are provided in Additional file 2: Table S2