- Top of page
We investigated metabolite levels during the progression of pathology in McGill-R-Thy1-APP rats, a transgenic animal model of Alzheimer's disease, and in healthy age-matched controls. Rats were subjected to in vivo 1H magnetic resonance spectroscopy (MRS) of the dorsal hippocampus at age 3, 9 and 12 months and of frontal cortex at 9 and 12 months. At 3 months, a stage in which only Aβ oligomers are present, lower glutamate, myo-inositol and total choline content were apparent in McGill-R-Thy1-APP rats. At age 9 months, lower levels of glutamate, GABA, N-acetylaspartate and total choline and elevated myo-inositol and taurine were found in dorsal hippocampus, whereas lower levels of glutamate, GABA, glutamine and N-acetylaspartate were found in frontal cortex. At age 12 months, only the taurine level was significantly different in dorsal hippocampus, whereas taurine, myo-inositol, N-acetylaspartate and total creatine levels were significantly higher in frontal cortex. McGill-R-Thy1-APP rats did not show the same changes in metabolite levels with age as displayed in the controls, and overall, prominent and complex metabolite differences were evident in this transgenic rat model of Alzheimer's disease. The findings also demonstrate that in vivo 1H MRS is a powerful tool to investigate disease-related metabolite changes in the brain.
Alzheimer's disease (AD) is a progressive neurodegenerative disease and the most common cause of dementia in the elderly. It is characterized by accumulation of extracellular plaques containing aggregated amyloid β (Aβ) peptides and intracellular neurofibrillary tangles composed of hyperphosphorylated tau proteins. In addition, regional loss of neurons and synapses, progressive cognitive decline and regional hypometabolism occurs (Mosconi 2005; Serrano-Pozo et al. 2011). Emerging evidence also suggest that, intraneuronal Aβ oligomers may contribute substantially to AD disease progression (Haass and Selkoe 2007).
There is no definite biomarker for the diagnosis of AD, which motivates the search for neuroimaging markers that may facilitate early detection of the disease. Using 1H magnetic resonance spectroscopy (MRS), the regional concentration of low-molecular-weight metabolites can be measured non-invasively and provides insight into neurochemical processes of normal and pathological conditions in vivo. Performing 1H MRS of patients with AD has revealed a consistent pattern of decreased levels of N-acetylaspartate (NAA) or NAA/total creatine (tCr) and increased myo-inositol (mIns) or mIns/tCr (Kantarci et al. 2003; Shiino et al. 2012). NAA is synthesised in neurons (Wiame et al. 2010) and the level decreases with neuronal loss or reversible neuronal or mitochondrial dysfunction (Gasparovic et al. 2001; Narayanan et al. 2001). This has enabled the widespread use of NAA as a marker of neuronal density, health and function (Moffett et al. 2007). In contrast, mIns is commonly considered to be a glial marker, and its increased content in humans and animal models is associated with elevated immunoreactivity of GFAP (Bitsch et al. 1999; Chen et al. 2009; Yang et al. 2011), which is rapidly synthesised during astrogliosis (Eng et al. 2000). Furthermore, less consistent alterations of glutamate or glutamate/tCr (Fayed et al. 2011; Rupsingh et al. 2011), total choline (tCho) or tCho/tCr (Jessen et al. 2000; Kantarci et al. 2003) and tCr (Jessen et al. 2009) have been reported in AD.
The study of transgenic animals is instrumental to achieve a better understanding of early aspects of the disease. A range of transgenic mouse models of AD have been investigated with in vivo 1H MRS, but to our knowledge, no longitudinal monitoring of metabolite content has been performed in vivo in a transgenic rat model of AD. This study was thus carried out to non-invasively investigate cerebral metabolite levels in a transgenic rat model of AD prior to and after the appearance of Aβ plaques, and to longitudinally characterize metabolite concentrations as pathology progressed. We employed the McGill-R-Thy1-APP model, in which rats express the human amyloid precursor protein (APP) carrying the double Swedish mutation and the Indiana mutation driven by the thymocyte antigen promoter (Thy 1.2). The latter restricts the expression of the transgene to neurons. Homozygous rats develop intraneuronal Aβ oligomers within 1 week after birth, and display cognitive symptoms within 3 months. Extracellular amyloid plaques occur in the hippocampal formation at age 6 months, appear in cortical areas around age 13 months and spread to large parts of the cerebrum within 20 months (Leon et al. 2010). Here, we analysed a volume comprising the dorsal hippocampus and subiculum at ages 3, 9 and 12 months, whereas an additional volume from the frontal cortex was analysed at ages 9 and 12 months.
- Top of page
LCModel provided modelled spectra with an average S/N ranging from 9 to 12 after the application of the above mentioned exclusion criteria (for a typical spectrum and the corresponding LCModel fit, see Fig. 2). The average linewidth at half height ranged from 0.033 to 0.039 ppm, well below the 0.1 ppm linewidth or less considered essential for in vivo 1H MRS spectra (Forster et al. 2012). The following metabolites could be reliably quantified with LCModel: glutamate, GABA, glutamine, NAA, mIns, taurine, tCr (creatine + phosphocreatine), and tCho [mainly glycerophosphocholine (GPC) and phosphocholine (PCh) with minor contributions from free choline and acetylcholine (Klein 2000)]. The majority of the metabolites had average CRLBs lower than 10%, with the exception of GABA (CRLB 12–15%) and glutamine (CRLB 8–12%). The estimated concentrations in control rats were for the most part in good agreement with those reported for the same rat brain regions by others (Kim et al. 2011, 2012).
Figure 2. A typical spectrum from the dorsal hippocampus of a 3 months old control rat shown with (a) the LCModel fit; (b) Peak assignments. Abbreviations: mIns: myo-inositol, tCr: total creatine, tCho: total choline, Gln: glutamine, Glu: glutamate, NAA: N-acetylaspartate, MM: macromolecules. Localization sequence; PRESS, TE = 12 ms, TR = 5 s, NS = 512, VOI = 24 μL (line broadening 3 Hz).
Download figure to PowerPoint
Differences in metabolite levels between McGill-R-Thy1-APP rats and controls were found in dorsal hippocampus and frontal cortex at all ages investigated. At the pre-plaque stage at age 3 months, lower concentrations of glutamate (−14%, p < 0.001), mIns (−10%, p = 0.037) and tCho (−17%, p = 0.002) were evident in dorsal hippocampus compared with controls (Fig. 3). More widespread differences in metabolite levels were found in the same brain area at age 9 months, and several metabolite levels changed between age 3 and 9 months. In McGill-R-Thy1-APP rats, there was an increase in the level of glutamate (9%, p = 0.028), mIns (16%, p = 0.002), tCr (10%, p = 0.010) and tCho (20%, p = 0.001). In controls, the levels of GABA, NAA and tCho increased (16, 7 and 14% with p values 0.016, 0.027 and 0.001 respectively). Despite the rise in glutamate content, the level was still significantly lower than in controls (−10%, p = 0.005). GABA, NAA and tCho levels were also lower than in controls (−11, −8 and −12%, respectively, with the p-values 0.049, 0.035 and 0.021), whereas taurine and mIns levels were higher in McGill-R-Thy1-APP rats at this age (21 and 14% with p-values 0.003 and 0.015 respectively). Between ages 9 and 12, metabolite levels in McGill-R-Thy1-APP rats remained stable, whereas there was a decrease in the levels of glutamate (−8%, p = 0.035) and tCho (−6%, p = 0.044) in controls. At 12 months of age, the taurine concentration was higher in McGill-R-Thy1-APP rats (16%, p = 0.030) compared with controls, but no other significant differences were found between the groups.
Figure 3. Concentrations (mM) of metabolites in the dorsal hippocampus of control (white circles) and McGill-R-Thy1-APP (AD, black triangles) rats at age 3, 9 and 12 months obtained with in vivo 1H MRS (for details see Methods section). Values are averages ± SEMs. Statistical analysis was performed using mixed linear model and post hoc linear contrasts. Significant differences between controls and McGill-R-Thy1-APP rats at each age are indicated with *p < 0.05 and **p < 0.01, whereas dotted lines indicate significant (p < 0.05) change with age within the group. The only significant difference between 3 and 12 months old rats was found for total choline in McGill-R-Thy1-APP rats (p = 0.002), which was higher at 12 months. In addition, an increase in mIns was close to significance in McGill-R-Thy1-APP rats (p = 0.057).
Download figure to PowerPoint
In the frontal cortex at age 9 months, lower glutamate (−14%, p ≤ 0.001), glutamine (−13%, p = 0.013), GABA (−10%, p = 0.046) and NAA (−7%, p = 0.018) were apparent in the McGill-R-Thy1-APP group compared with controls (Fig. 4). From age 9 to 12 months, the concentrations of glutamine and tCr increased significantly in McGill-R-Thy1-APP rats (9 and 4% with p-values 0.007 and 0.017 respectively), while glutamate, glutamine, taurine, NAA and tCr levels decreased significantly in the control group (−12, −14, −21, −10 and −8% with p-values 0.001, 0.025, < 0.001, < 0.001 and < 0.001 respectively). As a consequence, increased levels of taurine (25%, p < 0.001), mIns (18%, p = 0.014), NAA (7%, p = 0.016) and tCr (10%, p = 0.003) levels were apparent in McGill-R-Thy1-APP rats at age 12 months compared with controls.
Figure 4. Concentrations (mM) of metabolites in the frontal cortex of control (white circles) and McGill-R-Thy1-APP (AD, black triangles) rats at age 9 and 12 months obtained with in vivo 1H MRS (for details see Methods section). Values are averages ± SEMs. Statistical analysis was performed using mixed linear model and post hoc linear contrasts. Significant differences between controls and McGill-R-Thy1-APP rats at each age are indicated with *p < 0.05 and **p < 0.01, whereas dotted lines indicate significant (p < 0.05) change with age within the group.
Download figure to PowerPoint
In general, metabolite levels in the McGill-R-Thy1-APP rats did not follow the variations observed in controls in this study. No decreases in metabolite levels occurred with age in McGill-R-Thy1-APP rats.