Loss of TMEM106B leads to myelination deficits: implications for frontotemporal dementia treatment strategies

Abstract

Genetic variants that define two distinct haplotypes at the TMEM106B locus have been implicated in multiple neurodegenerative diseases and in healthy brain ageing. In frontotemporal dementia (FTD), the high expressing TMEM106B risk haplotype was shown to increase susceptibility for FTD with TDP-43 inclusions (FTD-TDP) and to modify disease penetrance in progranulin mutation carriers (FTD-GRN). To elucidate the biological function of TMEM106B and determine whether lowering TMEM106B may be a viable therapeutic strategy, we performed brain transcriptomic analyses in 8-month-old animals from our recently developed Tmem106b-/- mouse model. We included 10 Tmem106b+/+ (wild-type), 10 Tmem106b+/- and 10 Tmem106-/- mice. The most differentially expressed genes (153 downregulated and 60 upregulated) were identified between Tmem106b-/- and wild-type animals, with an enrichment for genes implicated in myelination-related cellular processes including axon ensheathment and oligodendrocyte differentiation. Co-expression analysis also revealed that the most downregulated group of correlated genes was enriched for myelination-related processes. We further detected a significant loss of OLIG2-positive cells in the corpus callosum of Tmem106b-/- mice, which was present already in young animals (21 days) and persisted until old age (23 months), without worsening. Quantitative polymerase chain reaction revealed a reduction of differentiated but not undifferentiated oligodendrocytes cellular markers. While no obvious changes in myelin were observed at the ultrastructure levels in unchallenged animals, treatment with cuprizone revealed that Tmem106b-/- mice are more susceptible to cuprizone-induced demyelination and have a reduced capacity to remyelinate, a finding which we were able to replicate in a newly generated Tmem106b CRISPR/cas9 knock-out mouse model. Finally, using a TMEM106B HeLa knock-out cell line and primary cultured oligodendrocytes, we determined that loss of TMEM106B leads to abnormalities in the distribution of lysosomes and PLP1. Together these findings reveal an important function for TMEM106B in myelination with possible consequences for therapeutic strategies aimed at lowering TMEM106B levels.

Keywords: TMEM106B; cuprizone; lysosome trafficking; myelin; oligodendrocytes.

Figure 1

Transcriptomics analysis reveals global changes in myelination pathways in TMEM106B deficient mice. (A) Diagram showing experimental procedures of transcriptomic analysis with Tmem106b^+/+ (wild-type, WT), Tmem106b^+/− and Tmem106b^−/− mouse brains. (B–D) Volcano plots show significant DEGs between Tmem106b^−/− and wild-type mice (B), Tmem106b^−/− and Tmem106b^+/− (C), and Tmem106b^+/− and wild-type mice (D). Cut-off step-up P < 0.05. Red and blue colours represent upregulated and downregulated genes in Tmem106b^−/− mice, respectively. (E) Hierarchical clustering of 213 DEGs between Tmem106b^−/− and wild-type mice, step-up P < 0.05. Red and blue colours represent upregulated and downregulated genes in Tmem106b^−/− mice, respectively. (F) Top 10 cellular processes identified by GO enrichment analysis from the DEGs between Tmem106b^−/− and wild-type mice. Myelination-related cellular processes are highlighted in orange. (G–K) Quantitative PCR validation of selected down- (G–J) and upregulated (K) DEGs using 8-month-old brains from wild-type and Tmem106b^−/− (KO) mice. Graphs represent the mean ± SEM. Data were analysed by Student’s t-test (n = 12 per group). ***P < 0.001, ****P < 0.0001.

Figure 2

Assessment of the myelin in Tmem106b^−/− mice. (A) Representative images of OLIG2 immunostaining in corpus callosum of 8-month-old Tmem106b^+/+ (wild-type, WT), Tmem106b^+/− (Het), and Tmem106b^−/− knockout (KO) mouse brains. (B–F) Quantification of OLIG2-positive cells in the corpus callosum of the wild-type, Het, and KO mouse brains at indicated ages (n = 5–14 per group). (G) Western blots show protein levels of CNP, MBP, and GAPDH in Tmem106b^+/+ (WT) and Tmem106b^−/− (KO) adult mice. (H) Quantification of the blots in g (n = 5 per group). (I) Representative EM images of myelinated fibres in corpus callosum from wild-type and KO adult mice. (J) Measurement of the G-ratios in myelinated fibres in G (n = 3 per group). (K) Representative toluidine blue-stained sections of sciatic nerves from wild-type and KO adult mice. (L) Measurement of the G-ratios in myelinated fibres in g (n = 3–4 per group). Graphs represent the mean ± SEM. Data were analysed by Student’s t-test. NS = not significant, *P < 0.05, **P < 0.01.

Figure 3

Increased demyelination and reduced remyelination in Tmem106b^−/− mice in a cuprizone-induced de- and remyelination mouse model. (A) Diagram shows experimental groups and procedures in the cuprizone-induced demyelination mouse model. (B) Representative images of Luxol fast blue (LFB) staining of corpus callosum from wild-type (WT) and Tmem106b^−/− (KO) mice treated with indicated diets: Cup 3W = 3 weeks of cuprizone diet; Cup 6W = 6 weeks of cuprizone diet; ND = normal diet; Rec 6W = 6 weeks of cuprizone diet plus 6 weeks of normal diet. (C) Quantification of Luxol fast blue stained in B. (D) Representative images of MBP immunofluorescence staining of corpus callosum from the mice in B. (E) Representative EM images of corpus callosum from wild-type and KO mice treated as described in B. Graph represents the mean ± SEM. Data were analysed by Student’s t-test (n = 4–7 per group). NS = not significant, ***P < 0.001.

Figure 4

Myelination-related protein levels in cuprizone diet-treated wild-type and TMEM106B-deficient mice. (A, C, E and G) Western blots show protein levels of CNP, MBP, and GAPDH in Tmem106b^+/+ (wild-type, WT), Tmem106b^+/− (heterozygous, Het), and Tmem106b^−/− (KO) mice treated with normal diet (ND, A), 3 weeks of cuprizone diet (Cup 3W, C), 6 weeks of cuprizone diet (Cup 6W, E), and 6 weeks of cuprizone diet plus 6 weeks of normal diet (Rec 6W, G), respectively. (B, D, F and H) Quantification of the blots in A, C, E and G, respectively. Graphs represent the mean ± SEM. Data were analysed by Student’s t-test (n = 4). NS = not significant, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 5

Reduction in differentiated oligodendrocytes and impaired PLP distribution in TMEM106B-deficient cells. (A) Diagram shows cellular markers of undifferentiated and differentiated oligodendrocytes. (B) Quantitative PCR shows the relative mRNA levels of immature and mature oligodendrocyte cellular markers in 3-month-old Tmem106b^−/− (KO) mice compared to Tmem106b^+/+ (wild-type, WT) (n = 6 per group). (C) Representative immunofluorescent images of LAMP1 and PLP1 from wild-type (WT) HeLa and TMEM106B CRISPR KO (KO) HeLa cells 24 h after PLP1 transfection. (D and E) Quantification of lysosomal (D) and plasma membrane (E) localized PLP1 in C (n = 3). (F) Representative immunofluorescent images of PSAP, PLP1, and DAPI from wild-type and Tmem106b^−/− (KO) mouse primary neuron-glial mixed cultures. (G and H) Quantification of perinuclear-distributed PSAP+ vesicles (G) and lysosomal localized PLP1 (H) in F (n = 12). Graphs in B, D, E, G and H represent the mean ± SEM. Data were analysed by Student’s t-test. NS = not significant, *P < 0.05, **P < 0.01, ***P < 0.001.