Work

Alzheimer's disease mouse models based on presenilin-2 N141I: an in vivo study of spontaneous electrical activity by means of hippocampal extracellular recordings

Year: 2016
Type: dissertation
Abstract: Alzheimer's disease (AD) is a neurodegenerative pathology that affects an increasing number of elderly people. It is characterized by progressive impairment in cognition and memory and it is the most frequent cause of dementia, being responsible for 60 to 70 % of the cases over 65 years (World Health Organization, 2015). The major neuropathological hallmarks of the disease are the deposition of neurofibrillary tangles and senile plaques, primarily in hippocampus, entorhinal cortex and neocortex, and a widespread neuronal loss (Hardy and Selkoe 2002). AD is divided in sporadic and familial (FAD) forms. FAD is caused by highly penetrating mutations in three genes involved in amyloid-ß (Aß) metabolism: the Aß precursor protein (APP), the presenilin1 (PSEN1) and the presenilin2 (PSEN2) (Bertram and Tanzi 2011). The earliest pathological changes are believed to take place in the hippocampal formation and entorhinal cortex (Squire, Stark, Clark 2004). These regions are part of the medial temporal lobe and are fundamental for the encoding of new memories as well as for the fixation of recent ones. Moreover, they are among the first targets of the pathology in terms of tissue abnormalities and neuro-physiological alterations; these latter well correlate with memory deficits, being early symptoms of the disease (Sperling, Mormino, Johnson 2014; Squire, Stark, Clark 2004). The most prominent network activity patterns in the hippocampus are field potential oscillations in the theta (4 - 12 Hz) and gamma (30 - 100 Hz) frequency bands. Theta and gamma oscillations are considered to play a pivotal role in memory as their properties, namely the amplitude, the frequency and the degree of their coupling, change during memory processes; moreover, they are predictive of learning performance (Lisman and Jensen 2013). Several brain oscillations have been shown to interact, a phenomenon known as cross-frequency coupling (CFC) (Canolty and Knight 2010). One particular type of interaction consists in the phase of a slower oscillation modulating the amplitude of a faster one, hence the name phase-amplitude CFC (PAC). The PAC of theta on gamma oscillations in the hippocampus has drawn a growing interest since it has proven to be a physiological feature remarkably suited for predicting memory performance (Axmacher et al. 2010; Tort et al. 2009). Most notably, theta-gamma decoupling in the hippocampus impairs memory performance (Shirvalkar, Rapp, Shapiro 2010). Further, a characteristic of brain electrophysiological signals is the "1/f" behavior of the frequency-domain power spectrum, meaning that the amplitude of each frequency component decays as a function of the frequency. Along with other features, the steepness of the decay has recently been shown to be informative about network activity. At present, there are no therapies to reliably revert or stop AD. A biomarker (or a combination of biomarkers) for a non-invasive early diagnosis of the pathology "i.e. before the emergence of the cognitive deficits" would improve the efficacy of current and future strategies to contrast the progression of the disease (Sperling and Johnson 2013a). Notwithstanding, FAD mutations of presenilin 1 (PS1) and, in particular of presenilin 2 (PS2) were shown to determine calcium (Ca2+) homeostasis impairment, a common feature in AD (Zampese et al. 2011a). In particular, Ca2+ defects due to PS2-N141I appear to occur independently of Aß load (Kipanyula et al. 2012). However, at the brain network level, outcomes due to the mutant PS2 have not been explored. The aims of this work were (i) the assessment of the effects of the PS2-N141I FAD mutation on the above-mentioned aspects of the hippocampal network activity and (ii) the identification of novel potential electrophysiological markers of the preclinical phase of AD. To these ends, we employed two FAD mouse models, PS2.30H and B6.152H, which express the human PS2-N141I mutation respectively alone or in combination with the human APP Swedish mutation (Ozmen et al. 2009; Richards et al. 2003). We investigated the local field potential (LFP) signal in the dentate gyrus (DG) region of the hippocampus in the condition of urethane anesthesia. Through the exploitation of frequency and time-frequency methods, we extracted several signal features, including amplitude, spectral steepness and theta-higher frequencies PAC. The temporal evolution of these features was assessed by investigating three age points, namely 3, 6 and 12 months. Additionally, we characterized our transgenic lines for three known molecular and histological biomarker of the disease. We quantified the degree of Aß42 load at each time point while the presence of amyloid plaques and astrogliosis was addressed at 3 and 6 months of age. At 6 months of age, we report a significant power increase of slow-gamma (SG, 25 - 40 Hz) and high-gamma (HG, 40 - 90 Hz) oscillations in the PS2.30H line and of beta (10 - 25 Hz) and SG oscillations in the B6.152H line. Gamma oscillations and, more generally, a broadband power increase are generated by local network activity and are linked to active computation (Buzsaki and Wang 2012). The enhancement of beta and gamma power that we observe in our models likely reflects neuronal hyperactivity, consistently with similar observations in both sporadic and familial AD as well as in several AD mouse models (Stargardt, Swaab, Bossers 2015). In our mouse lines, this is the first time that the condition of neural network hyperactivity is described, although in vitro neuronal Ca2+ hyperexcitability was previously reported (Kipanyula et al. 2012). In B6.152H mice, hyperactivity was timed with the appearance of amyloid plaques and astrogliosis, whereas in PS2.30H mice no molecular or histological biomarker significantly differed from the wt (wild-type) line at any investigated age. The observed hyperactivity is possibly due to the expression of the mutant PS2, rather than to the Aß load which we probed to be profoundly different in the two AD mouse lines. Only in the B6.152H line we found a more complex pattern of alterations. In addition to the broadband power increase, we report an enhancement of theta-beta and -SG PAC in this line at 6 months. Moreover, B6.152H mice displayed a steeper power spectrum slope compared to wt mice at 3 and 6 months of age. Both the overcoupling and the steeper spectral slope are ascribable to a condition of neuronal hyper-synchronicity, though with notable differences (Voytek and Knight 2015). Remarkably, we found that the difference between B6.152H and wt mice in terms of steepness was strongest at 3 months of age and diminished with ageing. Therefore, although further investigation will be required to validate this finding, hippocampal network hyper-synchronicity appears to be a promising early marker of the disease, at least in the context of a high Aß load.
Author: Roberto Fontana
Institution: University of Padua
Language: English
Open Access status: green
OpenAlex ID: W2557394052
Cites: 254
Cited by:
Related to: 10