Neural Diseases in a Dish: Drug Testing on Multielectrode Arrays

Project Background Alzheimer's disease (AD) is characterised by progressive and irreversible neuronal cell death which discretely and predictably affects different brain regions, with the hippocampus and entorhinal cortex being primarily affected. These large-scale changes are paralleled and preceded by perturbations to functional network connectivity at the cellular level with the AD-pathogenic proteins Amyloid b (Ab) and tau possibly having distinct roles in this regard.

It is suggested that Ab may drive an initial hyperexcitability which may in turn instigate propagative neuropathology, whereas pathogenic tau may mediate later effects such as cell silencing and death (Busche et al., 2019). Indeed, tau has previously been shown to be a prerequisite for the decreases in long-term potentiation seen in mouse models of AD (Shipton et al., 2011). However, to date, no study has characterised the interplay of Ab and tau on network connectivity in models of AD progression. On the contrary, in vivo recordings from the brain and in vitro analyses of cell cultures or brain slices usually only provide a snapshot of this dynamic.

Multi-Electrode Arrays (MEAs) allow for spatial and temporal analyses of electrical signalling in in vitro networks. Unlike some other electrophysiological techniques, MEAs are non-invasive to cells and act as a culture platform which allows for regular, long-term analyses of integrated network activity. MEAs therefore provide an opportunity for the investigation of network disturbances during early AD progression. Furthermore, MEAs are highly amenable to drug application meaning that the opportunities for pharmacological intervention can be investigated.

Project Approach Initial work will be performed on primary hippocampal neurones derived from AppNL-G-F knock-in (KI) mice and wild-type (WT) mice which will act as a control. The AppNL-G-F model contains a KI of the human APP gene with three familial-AD-relevant mutations which cause the Swedish (KN670/671NL), Arctic (E693G) and Iberian (I716F) amino acid substitutions in the Ab precursor protein respectively (Saito et al., 2014).

Neurones will be cultured on MEAs to facilitate investigations of the networks formed. Measures of spontaneous and stimulated network activity will be collected and analysed using conventional burst analysis (Cotterill et al., 2016), graph theoretical measures (Schröter et al., 2015), and advanced multidimensional cross-correlational analysis. Additionally, primary neurones will be cultured on coverslips to facilitate two-photon Ca2+ imaging and whole-cell electrophysiological experiments, gaining insights into cell-type-specific and synaptic activities respectively. Immunohistochemical, Western blot and RT-PCR or RNA-Seq analyses are likely to be used to assess supplementary neuropathology such as Ab and tau burden, and changes in gene expression for the proteins implicated in AD and neuronal excitability.

Following this work, the humanised MAPT KI mouse line, which expresses both 3R and 4R tau isoforms at physiologically-relevant levels (Saito et al., 2019), will be cross-bred with the AppNL-G-F KI mice to produce AppNL-G-F/MAPT dKI mice and the aforementioned experiments repeated. Hippocampal neurones derived from the MAPT KI mouse line alone will be used as a control.

In all circumstances pharmaceuticals such as the GABAergic agonist benzodiazepine, used clinically in the treatment of anxiety and convulsive epileptic seizures, may be used to assess whether increased inhibitory signalling can influence the development of AD-like phenotypes in these hippocampal models.

Busche, M. A., et al. (2019) Nat Neurosci 22(1):57-64.
Cotterill, E., et al. (2016) J Neurophysiol 116(2):306-321.
Saito, T., et al. (2014) Nat Neurosci 17(5):661-663.
Saito, T., et al. (2019) J Biol Chem 294(34):12754-12765.
Schröter, M. S., et al. (2015) J Neurosci 35(14):5459-5470.
Shipton, O. A., et al. (2011) J Neurosci 31(5):1688-1692