Lack of a protective effect of the Tmem106b "protective SNP" in the Grn knockout mouse model for frontotemporal lobar degeneration

Abstract

Genetic variants in TMEM106B are a common risk factor for frontotemporal lobar degeneration and the most important modifier of disease risk in patients with progranulin (GRN) mutations (FTLD-GRN). TMEM106B is encoding a lysosomal transmembrane protein of unknown molecular function. How it mediates its disease-modifying function remains enigmatic. Several TMEM106B single nucleotide polymorphisms (SNPs) are significantly associated with disease risk in FTLD-GRN carriers, of which all except one are within intronic sequences of TMEM106B. Of note, the non-coding SNPs are in high linkage disequilibrium with the coding SNP rs3173615 located in exon six of TMEM106B, resulting in a threonine to serine change at amino acid 185 in the minor allele, which is protective in FTLD-GRN carriers. To investigate the functional consequences of this variant in vivo, we generated and characterized a knockin mouse model harboring the Tmem106b^T186S variant. We analyzed the effect of this protective variant on FTLD pathology by crossing Tmem106b^T186S mice with Grn^-/- knockout mice, a model for GRN-mediated FTLD. We did not observe the amelioration of any of the investigated Grn^-/- knockout phenotypes, including transcriptomic changes, lipid alterations, or microgliosis in Tmem106b^T186S/T186S × Grn^-/- mice, indicating that the Tmem106b^T186S variant is not protective in the Grn^-/- knockout mouse model. These data suggest that effects of the associated SNPs not directly linked to the amino acid exchange in TMEM106B are critical for the modifying effect.

Keywords: FTLD; Progranulin; Protective SNP; TMEM106B.

Fig. 1

A Alignment of the protein sequence of human and mouse TMEM106B around the critical threonine (red) that is coded by the rs3173615 SNP in humans. T185 in humans corresponds to T186 in mice. The codon coding for the threonine is indicated. B Sequencing chromatogram of genomic tail DNA of a wildtype and a homozygous Tmem106b^T186S/T186S mouse with the codons for threonine 186 in the wildtype mouse and serin 186 (red) in the knockin mouse after gene editing of the “ACT” codon to “AGC”. A third base in the following codon is edited in the wobble-position without affecting the coded amino acid asparagine. C Immunoblot of brain lysates from wildtype and Tmem106b^T186S/T186S mice (n = 5–7, nine months old) with antibodies against TMEM106B and GAPDH as a loading control. The quantification of the TMEM106B signal is depicted (average of wildtype set as “1”). D Breeding scheme depicting the crossing of Tmem106b^T186S/T186S and Grn^−/− mice to obtain Grn^−/− × Tmem106^T186S/T186S. E Representative images of immunofluorescence stainings of the cerebral cortex, thalamus, and the hippocampus (CA3 subfield) of a wildtype, Grn^−/−, and Grn^−/− × Tmem106b^T186S/T186S with an antibody against LAMP1 (green). The quantification of the LAMP1-positive area/section is provided (n = 5, age: 6 months). F Representative images of autofluorescence recorded with laser excitation at 594 nm of the cerebral cortex, thalamus, and the hippocampus (CA3 subfield) of a wildtype, Grn^−/−, and Grn^−/− × Tmem106b^T186S/T186S. The quantification of the autofluorescence-positive area/section is provided (n = 5. Age: 6 months)

Fig. 2

A Representative images of immunofluorescence stainings of the cerebral cortex, thalamus, and the hippocampus (CA3 subfield) of a wildtype, Grn^−/−, and Grn^−/− × Tmem106b^T186S/T186S with an antibody against CD68 (green). B The quantification of the CD68-positive area/section is indicated (n = 5; Age: 6 months). C Representative images of immunofluorescence stainings of the thalamus of a wildtype, Grn^−/−, and Grn^−/− × Tmem106b^T186S/T186S with an antibody against Iba1 (red). Age: 6 month. D Three-dimensional reconstructions from representative microglia cells of a wildtype, Grn^−/−, and Grn^−/− × ^T186S/T186S mouse stained for Iba1 (red) and CD68 (green) from 50 images in the Z-direction of the thalamus (age: 6 months). E Volumetric quantification of the microglia cell volume of individual cells depicted from 3D reconstructed images. F Volumetric quantification of the CD68-positive phagosome compartment in individual Iba1-positive microglia cells depicted from 3D reconstructed images. B, E, F *p < 0.05; **p < 0.01; ****p < 0.001; ns Not significant

Fig. 3

A Volcano plot presentation of the differently expressed transcripts between wildtype vs. Grn^−/− (left panel), Grn^−/− vs. Grn^−/− × Tmem106b^T186S/T186S and wildtype vs Grn^−/− × Tmem106b^T186S/T186S mice (n = 4, each genotype) in RNA isolated from 6-month-old mice. The thresholds for transcripts reaching statistically significant differences (fold change (fc) < − 1.5 / > 1.5; p < 0.05) are indicated (dotted lines), and the quadrants with transcripts reaching the threshold are colored in green. Grn as the top-down-regulated transcript in Grn^−/− and Grn^−/− × Tmem106b^T186S/T186S mice was excluded in the volcano-plot representation. B Principal component analysis (PCA) of the 12 total brain samples analyzed by Nanostring nCounter “Glia Profiling” panel. Every individual point represents one individual biological replicate: Wildtype = black, Grn^−/− = green Grn^−/− × Tmem106b^T186S/T186S = red. Note that the wildtype group can be clearly separated from the other two groups, while there is an overlap between Grn^−/− and Grn^−/− × Tmem106b^T186S/T186S. C Fold changes of the top 14 transcripts differentially expressed between wildtype and Grn^−/− mice determined by NanoString nCounter analysis. Statistics indicate significant differences between wildtype and Grn^−/− / wildtype and Grn^−/− × Tmem106b^T186S/T186S mice. For non of the selected genes, any significant differences were found between Grn^−/− and wildtype and Grn^−/− × Tmem106b^T186S/T186S. *p < 0.05; **p < 0.01; ***p < 0.001; ns = not significant; (n = 4, each genotype). D Heatmap of 86 gene transcripts analyzed by NanoString nCounter analysis in wildtype, Grn^−/−, Grn^−/− × Tmem106b^T186S/T186S mice (n = 4, each genotype) from total brain RNA isolated from 6-month-old mice. The expression-corrected and housekeeping gene-normalized RNA counts for each gene and sample were normalized to the mean value of the wildtype animals. Grn was excluded from this representation. E Pathway enrichment of differentially expressed genes (p < 0.01) between wildtype and Grn^−/− mice (upper panel) and wildtype and Grn^−/− × Tmem106b^T186S/T186S mice (lower panel) by the “Enrichr” tool. The top-enriched terms of the gene ontology term “Cellular Component 2021” are depicted

Fig. 4

A Specific enzymatic activity of the three lysosomal enzymes cathepsin D (CTSD), cathepsin L (CTSL), and β-Glucocerebrosidase (GCase) in brain lysates of 6-month-old wildtype, Grn^−/−, Grn^−/− × Tmem106b^T186S/T186S mice (n = 4, each genotype) (upper panel) and 10-month-old mice (lower panel). B The levels of the BMP species with different fatty acid compositions (di18:1, di22:6, and 20:4/22:6) were quantified in lipid extracts of the brain from 6-month-old wildtype, Grn^−/−, Grn^−/− × Tmem106b^T186S/T186S mice (n = 8, each genotype) by mass spectrometry