Generation of a new transgenic mouse model for assessment of tau gene silencing therapies

Mutations in tau cause familial frontotemporal dementia (FTD-tau) and aggregates of wild-type tau are a prevalent pathology of Alzheimer’s disease (AD) and other sporadic tauopathies [1]. Although tau has clearly been a target of interest in these diseases, there have been relatively few good leads for small molecule therapeutics to reduce tau pathology and slow disease progression. Dominantly inherited familial neurodegenerative diseases, such as FTD, are prime candidates for biologic therapies (biotherapies) in which expression of the mutant gene is directly targeted by gene silencing approaches. Multiple approaches to knocking down mutant gene expression have been described in the literature including viral vector delivery of shRNAi or microRNA mimics, delivery of naked RNAi and RNAi complexed with various reagents to facilitate uptake, and delivery of modified antisense DNA oligonucleotides (ASOs) [2–5]. Preclinical testing of such therapies has met with encouraging outcomes, with examples of good to moderate efficacy [6–9].

The accumulation of misfolded tau appears to correlate well with disease progression in human AD [10] and there are ample data to suggest that patients with high amyloid burden and mild memory impairment would be high risk for subsequent development of tauopathy and further degeneration [11, 12]. Second, the clinical course of AD is relatively slowly progressing, offering a larger window for therapeutic intervention. Third, familial cases of FTD-tau present a highly motivated patient population in which silencing mutant tau expression would appear to be a highly relevant target. Furthermore, FTD represents a particularly attractive target for gene knockdown as it may be possible to direct such strategies against a specific isoform of tau [four repeat (4R)] which is either elevated or preferentially aggregated in many cases of the familial disease [13]. Fourth, although multiple anti-amyloid therapies have reached the human clinic (with limited success so far), the pipeline for modulators of tau biology has been limited [12, 14, 15]. Thus, despite the hurdles created by the need to use invasive procedures to deliver these therapeutics and the difficulties in assessing target engagement in human trials, there is a strong rationale for pursuing gene-silencing therapies for tau in FTD and possibly AD.

In FTD cases caused by mutations in tau, it might be possible to use allele-specific targeting to diminish the levels of mutant tau. Such strategies might be very effective in cases involving mutations that alter tau splicing. In AD cases, however, there are no mutations or obvious splicing alterations and thus a gene silencing approach would target wild-type tau. A recent study to target tau expression in adult mice has demonstrated the potential efficacy of antisense-oligonucleotide gene-silencing approaches and demonstrated that reductions of tau in the adult nervous system may not produce adverse effects [16]. However, it should be noted that humans that are heterozygous for a microdeletion of chromosome 17, which includes the MAPT locus and at least two other genes, have multiple developmental defects including mental retardation [17]. Mice lacking Mapt have been reported to show defects in neuronal migration [18], and defects in long-term potentiation have been reported in both knockout mice and slice cultures from adult rats treated with tau shRNAi [19]. Although these data suggest a potential for tau-silencing therapies to produce adverse effects in humans, mice with heterozygous or homozygous deletions of tau are cognitively normal, and partial or complete elimination of tau in mice that overproduce human A? has been shown to ameliorate A?-mediated cognitive deficits [20]. Thus, a gene silencing approach to tau may well be a viable therapeutic strategy for multiple human tauopathies, including AD.

In the present study, we have endeavored to create a new transgenic mouse model expressing the human 4R tau that is designed for the sole purpose of optimizing gene silencing therapies for tau. Optimizing the type of therapeutic, the dose, and the route of delivery for each of these potential therapies in vivo typically involves cross-sectional studies that provide a snapshot of efficacy at whatever times are chosen to sacrifice animals and analyze tissues. In order to more effectively assess the efficacy of these types of therapeutics in vivo, we designed and generated transgenic animals that encode human 4R tau, containing the P301L mutation which is associated with FTD, that is fused in-frame to luciferase for the purpose of generating transgenic mice that could be used to report the efficacy of knockdown therapeutics in a mouse model of tauopathy. By placing the human tau transgene under the control of a doxycycline (Dox) suppressible promoter, we built in an internal positive control into this transgenic model. We demonstrate the utility of this model and the dynamic range of the luciferase reporter as an in vivo indicator of gene expression.