Transgenic expression of the human Glucose Transporter1 (GLUT1) gene locus reduces disease burden in Glut1 deficiency syndrome model mice

Abstract

Proper brain function relies on an adequate supply of energy - mainly glucose - to power neuronal activity. Delivery of this nutrient to the neuropil is mediated by the Glucose Transporter1 (GLUT1) protein. Perturbing glucose supply to the brain is profoundly damaging and exemplified by the neurodevelopmental disorder, GLUT1 deficiency syndrome (GLUT1DS). Resulting from haploinsufficiency of the SLC2A1 (GLUT1) gene, GLUT1DS is characterized by intractable infantile-onset seizures and a disabling movement disorder. Ketogenic diets, which supply the brain with an alternate energy source, ketone bodies, are currently the preferred therapeutic option for Glut1DS patients but do not address the underlying cause - low brain glucose - of the disease. One intuitively appealing therapeutic strategy that does, involves restoring GLUT1 levels to the patient brain. Here, we demonstrate that transgenic expression of the human GLUT1 genomic locus in a mouse model of GLUT1DS raises brain GLUT1 levels and reduces disease burden. Augmenting GLUT1 levels in mutants correspondingly raised cerebrospinal fluid (CSF) glucose levels, improved motor performance and reduced the frequency of seizures characteristically observed in GLUT1DS. Interestingly, the increased GLUT1 in mutants harboring the human GLUT1 locus was at least partly the result of an increase in murine Slc2a1 (Glut1) activity, most likely the effect of a long non-coding RNA (lncRNA) embedded in the human transgene. Collectively, our work has not only shown that repleting human GLUT1 mitigates GLUT1DS but also has yielded transgenic mice that constitute a useful tool to test and optimize clinically promising agents designed to stimulate this gene for therapeutic purposes.

Keywords: GLUT1; Glut1 deficiency; Mouse models; Neurodevelopmental disease.

Fig. 1. –. Generation of transgenic mice harboring the genomic human GLUT1 gene locus.

(A) Schematic illustrating BAC RP11-125O1 and the Cla1 fragment containing human GLUT1 and the SLC2A1-DT lncRNA used to create transgenic mice. (B) Agarose gel showing BAC RP11-125O1 digested with Cla1 and an aliquot of the purified 79 kb transgene that was microinjected into fertilized oocytes to derive the transgenic mice. Note: Marker – Lambda DNA digested with HindIII. (C) Gel depicting the results of genotyping experiments to determine if the various lines that were produced harbored intact transgenes. Unlike lines 2829 and 2857, line 2833 lacks the 3′ end of the GLUT1 gene. Note: Marker – 1 kb Plus DNA ladder (Thermo Scientific Inc.).

Fig. 2. –. Mitigation of the GLUT1DS phenotype in mutant mice bearing the human GLUT1 gene locus.

(A) Quantified outcome of mouse motor performance on a rotarod. Note: *, **, ***, P < 0.05, P < 0.01 and P < 0.001, Kruskal-Wallis test, n = 5–21 subjects tested. (B) Quantification of CSF and blood glucose levels in controls and GLUT1DS mutants transgenic for the human GLUT1 gene locus. Note: *, ***, P < 0.05, P < 0.001; one-way ANOVA [F (3, 76) = 10.67] and Kruskal-Wallis tests to evaluate means for CSF and blood values respectively, n ≥ 16 mice of each cohort. (C) Graph depicting CSF:blood glucose ratios in controls and GLUT1DS mutants bearing the human GLUT1-containing transgene. Note: *, **, ***, P < 0.05, P < 0.01 and P < 0.001, [F (3, 70) = 96.82], one-way ANOVA, n ≥ 16 mice of each genotype. (D) Graphical representation of CSF lactate values in controls and GLUT1DS mutants transgenic for the human GLUT1 gene locus. Note: *, **, ***, P < 0.05, P < 0.01 and P < 0.001, [F (3, 18) = 8.27], one-way ANOVA, n ≥ 5 mice of each genotype.

Fig. 3. –. Seizures are suppressed in GLUT1DS mice expressing the human GLUT1 gene.

(A) Quantification of spike-wave discharges (SWDs) in controls and GLUT1DS mutants transgenic for the human GLUT1 gene locus. Note: *, **, ***, P < 0.05, P < 0.01 and P < 0.001, [F (3, 17) = 15.4], one-way ANOVA, n = 3–6 mice of each genotype. Representative EEG spectrograms (at 0-30 Hz) and corresponding traces below and to the right of them depicting (B) a relatively normal EEG wave pattern obtained from a 2829^Tg;Glut1^+/− subject and (C) one from a 2857^Tg;Glut1^+/− mouse exhibiting a SWD – highlighted by orange box – sometimes seen in this cohort of mice.

Fig. 4. –. Expression of GLUT1 and the SLC2A1-DT lncRNA from the 79 kb transgene.

(A) Agarose gel showing bands representing RT-PCR products – from brain tissue – originating from the human GLUT1 gene in the 79 kb transgene. Note that only mice from lines 2829 and 2857, carrying the intact transgene, express GLUT1. (B) Quantified levels of the human GLUT1 gene in 2829 and 2857 transgenic mice. Note: ***, P < 0.001, [F (2, 9) = 592.3], one-way ANOVA, n = 4 mice of each genotype. (C) Graph depicting levels of human GLUT1 expression in transgenic mice relative to those of its murine Slc2a1 (Glut1) counterpart in brain tissue of the mice. Note: Value depicted for WT mouse is that of murine Glut1 expression. Also note: ***, P < 0.001, one-way ANOVA, n = 3–8 mice of each genotype. (D) Agarose gel showing bands representing RT-PCR products – in brain tissue – from the human SLC2A1-DT lncRNA-expressing element in the 79 kb transgene. In this instance, all three transgenic lines expressed the lncRNA. (E) Quantified levels of the human lncRNA in the three lines of transgenic mice. Note: ***, P < 0.001, [F (2, 12) = 68.34], one-way ANOVA, n ≥ 4 mice of each genotype.

Fig. 5. –. Transgenic expression of the human GLUT1 locus raises murine Slc2a1 and total GLUT1 levels.

(A) Quantified levels of murine Slc2a1 (Glut1) transcripts in brain tissue of WT and transgenic mice. Note: ***, P < 0.001, [F (3, 14) = 131.3], one-way ANOVA, n ≥ 3 mice of each genotype. (B) Representative western blot of total GLUT1 protein in brain tissue of controls, and mutants expressing the human GLUT1 locus in line 2829. Note: 55 kDa band – endothelial GLUT1; 45 kDa band – astrocytic GLUT1. (C) Quantified levels of total GLUT1 protein in controls and mutants expressing the human GLUT1 locus in lines 2829 and 2857. Note: *, P < 0.05, [F (3, 19) = 4.55], one-way ANOVA, n ≥ 4 mice of each genotype. (D) Representative western blot of total GLUT1 protein in brain tissue of controls, and mutants expressing the human GLUT1 locus in line 2857. Note: 55 kDa band – endothelial GLUT1; 45 kDa band – astrocytic GLUT1.

Fig. 6. –. Transgenic expression of the human GLUT1 locus in lines 2829 and 2857 does not compensate for absence of murine Glut1.

(A) Graphical representation of observed and expected genotypes resulting from crosses between Glut1 haploinsufficient mice homozygous for the 2829 transgene. Mouse numbers are tabulated below the graph. (B) Graphical representation of observed and expected genotypes resulting from crosses between Glut1 haploinsufficient animals heterozygous for the 2857 transgene. Mouse numbers are tabulated below the graph. Note: The numbers in the two tables correspond to the genotypes on the x-axes of the graphs above them.