Page 61 - Read Online

Novati et al. Ageing Neur Dis 2022;2:17 https://dx.doi.org/10.20517/and.2022.19 Page 15 of 29

has been found to often not correlate [272-274] . And likewise, a proportion of unknown size of therapeutics fails
to enter the clinic, being not beneficial in animal models, though they might be effective in humans.

Despite the discouraging success rates in finding new therapies for NDs, rats have been essential for
discerning many aspects of neurological functions. However, with the more readily genetic manipulation of
mice and the discovery of many disease-causing genes for NDs, mice have outnumbered rats in studies
[275]
evaluating behavioral aspects of neuroscientific questions in the last two decades . Also, in studies
describing therapeutic approaches in AD, PD, and HD, this trend towards using mice is reflected by the
number of publications listed in PubMed [Figure 1].

Preclinical studies require a model to present a phenotype that is robust, fast developing, replicating key
aspects of the human disease, and compatible with the form of treatment investigated. Some aspects of
human disease are, however, only ever hardly modeled in animals. As one important example, cell loss is
often not found in genetic models of neurodegeneration or only towards the end of the lifespan.
Additionally, genetic rat models often display milder phenotypes than mouse models when based on the
same construct, and these phenotypes often take relatively long to develop . Therefore, we briefly describe
[276]
in this section models with induced cell loss - though fairly artificial - which have helped to model neuronal
demise and to evaluate therapies that can halt or even reverse this process. Commonly used models, with
such induced neurodegenerative phenotypes, are summarized in Table 2. Their fast-appearing nature and
cost-effectiveness, in comparison to generating new genetic rat lines, make them a resource to be relied
upon frequently.

Phenotypic rat models of Alzheimer’s disease
AD poses a challenge for finding appropriate models, because sporadic cases caused by mutations in AD-
[285]
risk genes outnumber familial cases . While rats are genetically closer to humans in terms of tau isoforms,
rats seem to be more resistant to developing characteristic neuropathological features of AD when
expressing human genes. They present fewer plaques and tau tangles are not present [67,276] . Injection of
neurotoxins or overexpressing constructs of Aβ into the brain are commonly used to induce local cell death
and to model the AD typical neuropathology. For this, the larger brain size of the rat offers advantages over
mice, as stereotactic injections can be performed more consistently and with larger volumes. Additionally,
these models have been mostly used in preclinical studies.

Rats with diminished cholinergic neuron populations or severed neuronal circuits show memory deficits
[278]
and impaired learning , thereby resembling the cognitive symptoms observed in patients, but not the
pathobiological origin of the disorder. Ibotenic and okadaic acid, amongst other cholinergic neuron
harming compounds, or surgically lesioned rats, have been used to study neuroprotective or even
regenerating therapies. Exemplary, with these phenotypic models, it was possible to demonstrate that
neuronal stem cells or mesenchymal stem cells can replace or protect cholinergic neurons and improve
spatial learning and memory [286-288] . Recapitulating early pathobiological events, Aβ-injected models have
also been used to investigate the beneficial effects of stem cell transplantation [289-291] . It should be noted in
this regard that the concentrations needed to induce the pathological phenotype by Aβ-injections exceed
any physiological concentrations, and the stereotactic injections always produce unwanted tissue damage at
the injection site. Genetic mouse models of AD have been used to elucidate the mechanisms underlying the
observed amelioration in the genetic context of AD [292,293] . A meta-analysis of preclinical studies on stem cell
therapy for AD found a large variation in the models used and origin of cells but concluded overall
beneficial effects on memory and learning. Approximately 60% of the analyzed studies were performed on
non-genetic rat models .
[294]