Ketogenic diet improves motor performance but not cognition in two mouse models of Alzheimer's pathology

Abstract

Dietary manipulations are increasingly viewed as possible approaches to treating neurodegenerative diseases. Previous studies suggest that Alzheimer's disease (AD) patients present an energy imbalance with brain hypometabolism and mitochondrial deficits. Ketogenic diets (KDs), widely investigated in the treatment and prevention of seizures, have been suggested to bypass metabolic deficits present in AD brain by providing ketone bodies as an alternative fuel to neurons. We investigated the effects of a ketogenic diet in two transgenic mouse lines. Five months old APP/PS1 (a model of amyloid deposition) and Tg4510 (a model of tau deposition) mice were offered either a ketogenic or a control (NIH-31) diet for 3 months. Body weight and food intake were monitored throughout the experiment, and blood was collected at 4 weeks and 4 months for ketone and glucose assessments. Both lines of transgenic mice weighed less than nontransgenic mice, yet, surprisingly, had elevated food intake. The ketogenic diet did not affect these differences in body weight or food consumption. Behavioral testing during the last two weeks of treatment found that mice offered KD performed significantly better on the rotarod compared to mice on the control diet independent of genotype. In the open field test, both transgenic mouse lines presented increased locomotor activity compared to nontransgenic, age-matched controls, and this effect was not influenced by KD. The radial arm water maze identified learning deficits in both transgenic lines with no significant differences between diets. Tissue measures of amyloid, tau, astroglial and microglial markers in transgenic lines showed no differences between animals fed the control or the ketogenic diet. These data suggest that ketogenic diets may play an important role in enhancing motor performance in mice, but have minimal impact on the phenotype of murine models of amyloid or tau deposition.

Figure 1. Experimental design for the study of a ketogenic diet using two mouse models of Alzheimer’s pathology.

APP+PS1, Tg4510 and nontransgenic littermates received either a control (NIH-31) or a low carbohydrate, medium-chain triglyceride rich, ketogenic diet (KD) for 16 weeks. Blood was collected at 4 weeks and 16 weeks for measurement of circulating ketone and glucose levels. Near the end of the administration period, mice received a variety of behavioral tests. Tissue was collected after a 16 week diet administration.

Figure 2. KD increased ketosis and reduced glucose levels.

(A, C) KD (black bars) successfully increased peripheral β-hydroxybutyrate (BHB) levels after 4 weeks or 4 months, compared to a control diet (NIH-31, open bars). (B) Accordingly, circulating glucose levels were found to be decreased in KD-fed mice, in all genotypes. Glucose and BHB levels were measured utilizing a commercially available glucose/ketone monitoring system (Nova Max^© Plus). Data are presented as mean ± SEM (n=10). **p<0.0004, ***p<0.0001.

Figure 3. Changes in body weight and food intake throughout experiment.

(A) Assessments of body weight and (B) food intake in APP+PS1, Tg4510 and nontransgenic mice on either control diet (NIH-31, open symbols) or ketogenic diet (KD, solid symbols) for 4 months. Both transgenic mouse lines weighed significantly less than nontransgenic mice. (B) Smaller body weights did not result from reductions in food intake. The Tg4510 mice ate significantly more food than did nontransgenic mice. Data are presented as mean ± SEM. *p<0.02, ***p<0.0001.

Figure 4. AD transgenic mouse models had increased locomotor activity in the open field.

Both APP+PS1 and Tg4510 mice displayed greater total distance travelled in the open field, compared to nontransgenic control mice, regardless of the diet. Locomotor activity was assessed during a single 15min trial in the open field. Data are presented as mean ± SEM. **p<0.005 and ***p<0.0001.

Figure 5. KD enhanced motor performance in all genotypes.

Motor performance was assessed using two variations of the rotarod test. (A) Both genotype and diet effects were found in the standard accelerating rotarod test and (B) Average latency to fall in the accelerating rotarod was significantly greater in mice fed a KD. (C) Similarly, the ketogenic diet significantly enhanced motor performance in the non-accelerating variation of the test in all genotypes. Overall, latency to fall was greater in nontransgenic mice than in either APP+PS1 or Tg4510 mice. Data are presented as mean ± SEM. **p<0.007.

Figure 6. KD did not rescue spatial memory deficits.

Spatial memory was assessed by the 2-day radial arm water maze (RAWM). (A) Both APP+PS1 and Tg4510 (tau) mice made more errors on the 2-day RAWM compared to Ntg control mice, regardless of the diet. (B) APP+PS1 and Tg4510 mice consistently made more entries into the wrong arms, suggesting that neither were able to learn a new platform location on the reversal trial. Data are presented as mean ± SEM. **p<0.01 and ***p<0.0001.

Figure 7. KD did not rescue neuronal loss in Tg4510 mice.

Micrograph representation (2X) of neuronal staining (NeuN) in Ntg (A, D), APP+PS1 (B, E) and Tg4510 (C, F) mice kept on either NIH-31 (A, B, C) or KD diets (D, E, F). (G) Immunoreactivity for the neuronal marker NeuN was significantly reduced in Tg4510 mice compared to both nontransgenic and APP+PS1 mice. (H) Hippocampal volume (expressed in mm³) was calculated in Nissl stained sections. In agreement with the neuronal loss observed, hippocampal volume was significantly smaller in Tg4510 line, compared to the other genotypes. No diet effects were observed. Immunostaining was digitally quantified by Mirax image analysis. Data are presented as mean ± SEM. 2X Scale bar = 1000μm; 20X inset scale bar = 100 μm. **p<0.01 and ***p<0.0001.

Figure 8. Astrocytosis and microglial activation were observed in Tg4510 mice.

Micrograph representation (5X) of hippocampi stained for GFAP positive astrocytes (panels A-G) or Iba-1 positive microglia (panels H-N) in Ntg (A, D, H, K), APP+PS1 (B, E, I, L) and Tg4510 (C, F, J, M) mice kept on either NIH-31 (A, B, C, H, I, J) or KD diets (D, E, F, K, L, M). (G, N) Immunostaining was digitally quantified by Mirax image analysis. Data are presented as mean ± SEM. Scale bar = 500μm; inset scale bar = 100 μm ***p<0.0001.