Impaired Glucose Homeostasis in a Tau Knock-In Mouse Model

Abstract

Alzheimer's disease (AD) is the leading cause of dementia. While impaired glucose homeostasis has been shown to increase AD risk and pathological loss of tau function, the latter has been suggested to contribute to the emergence of the glucose homeostasis alterations observed in AD patients. However, the links between tau impairments and glucose homeostasis, remain unclear. In this context, the present study aimed at investigating the metabolic phenotype of a new tau knock-in (KI) mouse model, expressing, at a physiological level, a human tau protein bearing the P301L mutation under the control of the endogenous mouse Mapt promoter. Metabolic investigations revealed that, while under chow diet tau KI mice do not exhibit significant metabolic impairments, male but not female tau KI animals under High-Fat Diet (HFD) exhibited higher insulinemia as well as glucose intolerance as compared to control littermates. Using immunofluorescence, tau protein was found colocalized with insulin in the β cells of pancreatic islets in both mouse (WT, KI) and human pancreas. Isolated islets from tau KI and tau knock-out mice exhibited impaired glucose-stimulated insulin secretion (GSIS), an effect recapitulated in the mouse pancreatic β-cell line (MIN6) following tau knock-down. Altogether, our data indicate that loss of tau function in tau KI mice and, particularly, dysfunction of pancreatic β cells might promote glucose homeostasis impairments and contribute to metabolic changes observed in AD.

Keywords: energy metabolism; glucose homeostasis; high-fat; mouse model; tau.

Figure 1

Generation and tau expression in the tau knock-in mouse model. (A) Generation strategy of tau knock-in (KI) mice. MAPT human transgene is composed of 10 exonic regions with P301L mutation on exon 10. A V5 epitope tag is inserted after exon 2. STOP codon is inserted followed by exogenous poly(A) Tail. The transgene is under the control of the Mapt mouse endogenous promoter. Insertion of the targeting vector was mediated by cre-loxP recombination (pLox sites shown as empty arrows). A floxed neomycin resistance (Neor) cassette used for positive selection was removed from the targeted allele by FRT (FRT sites shown as black rectangles) recombination sites. Positions and sizes of exons and introns are not to scale. (B) Representative expression of tau and V5 in the cortex and hippocampus of tau KI mice and WT littermates (two showed mice out of eight/genotype).

Figure 2

Metabolic phenotyping of tau KI mice under High Fat diet (HFD; given from 2 to 5 months of age). (A–L): Males. (A) Body weight gain of WT and tau KI mice under HFD from 2 to 5 months of age (Two-Way ANOVA; F_(14,210) = 1.807, *p < 0.05 vs. WT). (B–F) Metabolic cage evaluation of tau KI male mice: (B) 24 h-cumulative food intake (g) (Two-Way ANOVA; F_(23,207) = 2.401, *p < 0.05 vs. WT). (C) 24 h spontaneous locomotor activity (total beam breaks/h). (D) 24 h-respiratory exchange ratio (RER = VCO₂/VO₂). (E) 24 h-O₂ consumption. (F) Glycemia after 6 h of fasting before (2 m) and at the completion of HFD (5 m; NS, One-Way ANOVA). (G) Glycemia in fed condition (9 a.m) before (2 m) and at the completion of HFD (5 m; NS, One-Way ANOVA). (H) Insulinemia after 6 h of fasting before (2 m) and at the completion of HFD (5 m; One-Way ANOVA followed by Tukey’s post-hoc test; F_(3,27) = 13.34 p < 0.0001; *p < 0.05 vs. WT). (I) Glycemic variations during the 1st, 2nd and 4th h following re-feeding after 16 h of fasting at the completion of the HFD, i.e., 5 months of age (One-Way ANOVA followed by Tukey’s post-hoc test; F_(7,60) = 19.83 p < 0.0001; *p < 0.05 vs. WT *p < 0.05 vs. WT for glycemia after 16 h). (J) Intraperitoneal glucose tolerance test (IPGTT) at completion of the HFD, i.e., 5 months of age (*p < 0.05, Two-Way ANOVA). (K) Pyruvate tolerance test (PTT) at completion of the HFD, i.e., 5 months of age (NS, Two-Way ANOVA). (L) Rectal temperature at the completion of the HFD, i.e., 5 months of age (NS, Student’s t-test). (M–U): Females. (M) Body weight gain of WT and tau KI mice under HFD from 2 to 5 months of age (NS). (N) Glycemia after 6 h of fasting before (2 m) and at the completion of HFD (5 m; NS). (O) Glycemia in fed condition 9 a.m. before (2 m) and at the completion of HFD (5 m; NS). (P) Insulinemia after 6 h of fasting before (2 m) and at the completion of HFD (5 m; NS). (Q) Glycemic variations during the 1st, 2nd and 4th h following re-feeding after 16 h of fasting at the completion of the HFD, i.e., 5 months of age (NS). (R) Intraperitoneal glucose tolerance test (IPGTT) at the completion of the HFD, i.e., 5 months of age (NS). (T) Pyruvate tolerance test (PTT) at the completion of the HFD, i.e., 5 months of age (NS). (U) Rectal temperature at the completion of the HFD, i.e., 5 months of age (NS). Results are expressed as mean ± SEM. WT mice are indicated as white circles/bars, tau KI mice as black circles/bars.

Figure 3

Analysis for both α and β cell fraction. (A) Representative double immunofluorescence staining for insulin (red) and glucagon (green) in pancreas sections from adult WT and tau KI mice. DAPI nuclear counterstaining was used (blue). Magnification for pancreatic sections = x40, scale bars = 20 μm. (B) Automated analysis of α and β cell fraction in pancreatic sections from tau KI and WT male mice under HFD (N = 1,521–1,883 islets from three mice/group; Kruskal Wallis test, p < 0.001; **p = 0.0096 using Dunn’s multiple comparisons test). Results are expressed as mean ± SEM. WT mice are indicated as white bars, tau KI mice as black bars.

Figure 4

Tau mRNA expression in mouse pancreas. (A) Tau mRNA expression in pancreatic islets and cortex of adult wild type mice (5 months). Tau mRNA expression in endocrine pancreatic islets represents 10% of its expression in the hippocampus. Results are expressed as mean ± SEM. (B) t-Distributed Stochastic Neighbor Embedding (t-SNE) visualization of all mouse pancreatic cells collected by fluorescence-activated cell sorting (FACS). (C) Single-cell transcriptomic data of Mapt gene from adult mouse pancreatic tissue. (D) Differential expression of Mapt gene between different mouse pancreatic islets cells. For images (B–D) data were extracted from the Tabula Moris single-cell RNA-sequencing database.

Figure 5

Tau expression in mouse and human pancreatic islets and colocalization with insulin and glucagon. Panel 1 (9F6 antibody). Tau WT mice. (A–E) Double immunofluorescence staining for insulin (green) and tau (red) in pancreatic islets from adult WT mice. Blue: DAPI nuclear counterstaining. (F–J) Double immunofluorescence staining for glucagon (green) and tau (red) in pancreatic islets from adult WT mice. Blue: DAPI nuclear counterstaining. Tau KO mice (A’–E’) Double immunofluorescence staining for insulin (green) and tau (red) in pancreatic islets from adult tau KO mice. Blue: DAPI nuclear counterstaining. (F’–J’) Double immunofluorescence staining for glucagon (green) and tau (red) in pancreatic islets from adult tau KO mice. Blue: DAPI nuclear counterstaining. (K’–O’). Tau KI mice (A”–E”) Double immunofluorescence staining for insulin (green) and tau (red) in pancreatic islets from adult KI mice. Blue: DAPI nuclear counterstaining. (F”–J”) Double immunofluorescence staining for glucagon (green) and tau (red) in pancreatic islets from adult tau KI mice. Blue: DAPI nuclear counterstaining. These observations were reproduced in at least three independent experiments and samples. Panel 2 tau expression using htauE1 antibody against human tau. Tau WT mice. (A–E) Double immunofluorescence staining for insulin (green) and tau (red) in pancreatic islets from adult WT mice. Blue: DAPI nuclear counterstaining. (F–J) Double immunofluorescence staining for glucagon (green) and tau (red) in pancreatic islets from adult WT mice. Blue: DAPI nuclear counterstaining. Tau KO mice (A’–E’) Double immunofluorescence staining for insulin (green) and tau (red) in pancreatic islets from adult tau KO mice. Blue: DAPI nuclear counterstaining. (F’–J’) Double immunofluorescence staining for glucagon (green) and tau (red) in pancreatic islets from adult tau KO mice. Blue: DAPI nuclear counterstaining. Tau KI mice (A”–E”) Double immunofluorescence staining for insulin (green) and tau (red) in pancreatic islets from adult KI mice. Blue: DAPI nuclear counterstaining. (F”–J”) Double immunofluorescence staining for glucagon (green) and tau (red) in pancreatic islets from adult tau KI mice. Blue: DAPI nuclear counterstaining. Mice were 5 month-old. Scale: 20 μm. These observations were reproduced in at least three independent experiments and samples. Panel 3 (tau 993S5 antibody). Human islets. (A–D) Double immunofluorescence staining for tau (green) and insulin (red). Blue: DAPI nuclear counterstaining. (E–H) Double immunofluorescence staining for tau (green) and glucagon (red). Blue: DAPI nuclear counterstaining. Scale: 50 μm.

Figure 6

Glucose-stimulated insulin secretion (GSIS) on islets isolated from tau knock-in and knock-out mice. (A) Insulin content (μg/L) from isolated islets (N = 4–6 mice/group; NS using Kruskal Wallis test). (B) Insulin secretion from control, tau KI and tau KO islets stimulated with low (2.8 mM) and high (16.7 mM) glucose for 1 h (N = 4–6 mice/group; Two-Way ANOVA; F_(5,26) = 7.544 p < 0.001; Tukey’s post hoc test **p < 0.01, ^#p < 0.05 vs. 2.8 mM). WT, tau KI, and tau KO are indicated as open bars, black bars and gray bars, respectively. Mice were 5-month-old. ns, non significant.