The X-Linked Intellectual Disability Gene, ZDHHC9 , Is Important for Oligodendrocyte Subtype Determination and Myelination

Abstract

Two percent of patients with X-linked intellectual disability (XLID) exhibit loss-of-function mutations in the enzyme, ZDHHC9. One of the main anatomical deficits observed in these patients is a decrease in corpus callosum volume and a concurrent disruption in white matter integrity. In this study, we demonstrate that deletion of Zdhhc9 in mice disrupts the balance of mature oligodendrocyte subtypes within the corpus callosum. While overall mature oligodendrocyte numbers are unchanged, there is a marked increase in MOL5/6 cells that are enriched in genes associated with cell adhesion and synapses, and a concomitant decrease in MOL2/3 cells that are enriched in genes associated with myelination. In line with this, we observed a decrease in the density of myelinated axons and disruptions in myelin compaction in the corpus callosum of Zdhhc9 knockout mice. RNA sequencing and proteomic analysis further revealed a reduction in genes and proteins essential for lipid metabolism, cholesterol synthesis, gene expression, and myelin compaction, offering insights into the underlying mechanisms of the pathology. These findings reveal a previously underappreciated and fundamental role for ZDHHC9 and protein palmitoylation in regulating oligodendrocyte subtype determination and myelinogenesis, offering mechanistic insights into the deficits observed in white matter volume in patients with mutations in ZDHHC9.

Keywords: X‐linked intellectual disability; corpus callosum; myelination; oligodendrocytes; palmitoylation.

FIGURE 1

There is a change in mature oligodendrocyte gene expression and genes involved in lipid metabolism in the corpus callosum of Zdhhc9‐KO mice. (a) Illustration showing pipeline for RNAseq analysis in P60 mice. (b) Quantification of fold change in gene expression (n = 5 Zdhhc9‐KO mice, n = 6 control mice). Each point represents an individual gene; blue denotes genes that are downregulated, and red denotes genes that are upregulated in Zdhhc9‐KO mice relative to controls. The horizontal line denotes the statistical significance threshold required for two tailed t‐tests; p‐values were adjusted to account for multiple comparisons. Top 5 downregulated and upregulated genes are annotated, along with genes of interest. (c) Enrichment analysis for Biological Process GO terms for both up‐ and downregulated genes. The 10 GO terms with largest fold enrichment based on FDR value are shown. (d) Graph of RT‐qPCR data for a subset of differentially expressed genes with known roles in myelination (n = 4–6 Zdhhc9‐KO mice, n = 4–6 control mice). Normalized expression of each gene relative to control (dotted line). (e) Heatmap showing the expression of the 173 differentially expressed genes in Zdhhc9‐KO mice across various oligodendrocyte subtypes, as derived from single‐cell RNA sequencing data from the Allen Brain Cell (ABC) Atlas. Genes are listed (top to bottom) by decreasing log₂ Fold change, ranging from the most upregulated to the most downregulated. Dotted black boxes highlight predominant cell type enrichment of upregulated and downregulated genes. (f) Bisque bioinformatic estimation of oligodendrocyte cell type proportions in the corpus callosum, performed on bulk RNAseq data from this study and using Allen brain ABC atlas as a reference scRNAseq dataset.

FIGURE 2

There is a change in mature oligodendrocyte subtype densities, and an increase in the density of OPCs in the corpus callosum of Zdhhc9‐KO mice. (a) Representative images of fluorescent in situ hybridization for MOL2/3 and MOL5/6 (Hapln2—MOL2/3; C030029H02Rik—MOL5/6; Ptgds—MOL5/6) in the P60 corpus callosum (scale bar, 50 μm) (b) Representative image of corpus callosum cells colored by oligodendrocyte subtype classification. MOL2/3 cells are colored green; MOL5/6 cells are colored blue; unclassified cells are colored gray. c Graph of oligodendrocyte subtype proportions (n = 5 mice per genotype). (d) Representative images of fluorescent in situ hybridization for immature oligodendrocyte cell‐lineage‐subtype marker genes (Pdgfra‐OPCs, Itpr2‐NFOLs, Ctps‐MFOLs) in the corpus callosum (scale bar, 50 μm). (e) Representative image of corpus callosum cells colored by oligodendrocyte subtype classification. OPC cells are colored yellow; NFOL cells are colored cyan; MFOL cells are colored magenta; unclassified cells are colored gray. (f) Graph of OPC, NFOL and MFOL proportions (n = 4 Zdhhc9‐KO mice, n = 5 control mice). (g) Representative images of OLIG2 (pan oligodendrocyte marker) and CC1 (mature oligodendrocyte marker) immunolabeling in the P60 corpus callosum. h Graph of OLIG2+ and OLIG2+/CC1+ proportions (n = 5 mice per genotype). Data in (c, f, h) show mean ± SEM. (c, f, h) multiple t‐tests using Holm‐Sidak method, alpha = 0.05.

FIGURE 3

There are fewer myelinated axons and decreased myelin compaction in the corpus callosum of Zdhhc9‐KO mice. (a) Representative transmission electron microscopy images of the P60 corpus callosum (scale bar, 2 mm). (b) Quantification of axon density and (c) the percentage of axons that are myelinated (n = 3 mice per genotype). (d) The average diameter of unmyelinated axons (n > 1000 axons from 3 mice per genotype). (e) The average diameter of myelinated axons (n > 1000 axons from 3 mice per genotype). (f) Frequency distribution graph showing the relative frequency of myelinated axons with a given axon diameter (n > 1000 axons from three mice per genotype). (g) The average g‐ratio of myelinated axons (n > 1000 axons from 3 mice per genotype). (h) Scatter plot of g‐ratio versus axon diameter. i Representative images (left) and quantification (right) of axons with myelination decompaction phenotypes (myelin “loops”) (scale bar, 1 mm). (j) Representative images of a myelinated axon, showing major dense line (MDL) and intraperiod line (IL) within the myelin sheath (scale bar, 100 nm) (left) and quantification of the average distance between IL and MDL (n > 30 axons from 3 mice per genotype). Data in (b, c, d, e, f, g, i, j) show mean ± SEM. (b, c, d, e, i, j) unpaired two‐tailed Student's t‐tests; (d, g) Welch's t‐test; (h) Type II Sum of Squares Analysis of Covariance (ANCOVA).

FIGURE 4

Proteins associated with myelination and myelin compaction are decreased in Zdhhc9‐KO mice. (a) Illustration showing the pipeline for myelin proteomic analysis of P23 mice. (b) Quantification of fold change in protein expression (n = 4 mice per genotype). Each point represents an individual protein; blue denotes proteins that are downregulated, and red denotes proteins that are upregulated in Zdhhc9‐KO mice relative to controls. The horizontal line denotes the statistical significance threshold required for two tailed t‐tests. Labeled proteins denote those that are both differentially expressed and highly abundant (top 10% of control myelin proteome). (c) Graph of Max LFQ Intensity data for the 15 most abundant myelin proteins (n = 4 mice per genotype). Normalized expression of each protein relative to control (dotted line). Differentially expressed proteins are highlighted in green. (d) Graph of the biological process GO Term associated with each differentially expressed protein along with their log₂ Fold Change. (e) Graph of Max LFQ Intensity data for all palmitoylating enzymes identified in myelin (n = 4 mice per genotype). f Western blots used to validate proteomics results. (g) Quantification of changes in myelin protein levels as determined by western blot. (n = 3 control mice; n = 4 Zdhhc9‐KO mice). All statistical analyses were done using unpaired two‐tailed Student's t‐tests.