Antithetical roles of tau in hippocampal long-term synaptic depression in rats versus mice

Tau, a microtubule-associated protein crucial for the maintenance of neuronal cytoskeletal architecture, is disrupted in tauopathies such as Alzheimer’s disease (AD) [1]. Beyond its canonical pool, tau is also present at synapses [2] where it regulates key synaptic functions [1]. Since synaptic weakening and loss of synapses are early features of AD and other tauopathies [3], understanding the synaptic roles of tau is imperative for developing more efficacious therapeutics [4]. To date, the impact of tau deletion on synaptic function has primarily been investigated in microtubule-associated protein tau homozygous knockout (Mapt^−/−; Mapt KO) mice, where tau deletion results in deficits in long-term synaptic depression (LTD) in juveniles [2].

While studies have focused on the effects of tau deletion in mice, the Mapt gene exhibits considerable evolutionary divergence [5]. In the central nervous system (CNS) of humans, tau exists as six isoforms, each with distinct physiological functions and neuropathological profiles [6]. Unlike mice, rats express the same number of tau isoforms as humans, suggesting that rats may provide a better model for understanding human tau [7, 8]. Rats also more closely mimic aspects of human physiology [9] and offer a richer behavioural repertoire than mice [10], enabling more sensitive assessment of tau-dependent behaviours, while being less prone to environmental stressors [11] that drive tau phosphorylation and pathology [12]. Furthermore, there is evidence to suggest that, compared to mouse models of AD, tau pathology in rat AD models more closely resembles that of human AD patients [13], further supporting rats as a better translational model.

Synaptic weakening is relevant to neurodegenerative diseases involving tau [1, 3], and previous work has shown inhibition of hippocampal LTD in Mapt KO mice [2]. However, whether tau plays a comparable role in synaptic function in rats is unknown. To address this, we developed a constitutive rat Mapt KO line (SD-Mapt^em1Tcp) using CRISPR-Cas9 technology (Fig. 1A and Methods in Supplementary Materials). Sanger sequencing confirmed deletion of the fourth through fifth exon of the Mapt gene (Fig. S1A). Elimination of tau protein expression was confirmed by immunoblotting of whole-hippocampal homogenates from juvenile rats, using both an N-terminal monoclonal (Cat#: MABN2406, MilliporeSigma [14]) and a C-terminal polyclonal (Cat #: A1103, ABclonal [15]) antibody, directed against epitopes upstream and downstream, respectively, of the deleted region (Fig. 1B, C). Pups were born at expected Mendelian ratios from heterozygous breeding pairs (Fig. S1B), and Mapt KO pups exhibited no gross defects or alterations in body, brain and hippocampal mass compared to wild-type (Mapt^+/+; WT) littermates (Fig. S1C–E).

To assess synaptic weakening in this model, we examined low-frequency stimulation (LFS)-induced LTD in the CA1 region of acute ex vivo hippocampal slices prepared from juvenile male Mapt KO rats and their male WT littermates. We also examined LFS-LTD in slices from juvenile male Mapt KO mice and their male WT littermates in interleaved experiments. Surprisingly, we found that the impact of tau deletion on LTD depended on rodent species (two-way ANOVA: species × genotype interaction, F_(1,26) = 30.5, p < 0.0001). Specifically, LTD was significantly enhanced in male Mapt KO rats (40.3 ± 3.1%, n = 6) relative to WT littermate controls (22.0 ± 4.3%, n = 6; t₍₂₆₎ = 4.0, p = 0.002, Šídák’s post-hoc test, Fig. 1D, F). In contrast, LTD was reduced in male Mapt KO mice (14.9 ± 2.5%, n = 9) relative to WT littermates (29.4 ± 2.2%, n = 9; t₍₂₆₎ = 3.9, p = 0.003, Fig. 1E, F), consistent with previous findings [2], and differed significantly from the augmented LTD obtained in male Mapt KO rats (t₍₂₆₎ = 6.0, p < 0.0001). LTD levels between male WT mice and rats were similar (t₍₂₆₎ = 1.8, p = 0.3). Together, these findings indicate that tau has an opposite role in LTD dependent on rodent species, facilitating LTD in mice and constraining LTD in rats.

Our interleaved experiments using male rats and mice confirm that this is a bona fide species difference rather than a difference in experimental conditions. The reason for this species difference remains unknown and could be attributable to several factors. One plausible explanation is that it arises from the function of a tau isoform that rats possess but mice do not (e.g., 1N3R/2N3R) since isoforms are known to have unique roles [6]. Extrinsic factors, such as differences in how rats and mice handle or respond to stress [11] for which tau has been shown to play a role in mice [12], could also contribute. A third plausible explanation relates to intrinsic factors, such as different expression levels of synaptic plasticity genes in dendrites including calcium/calmodulin-dependent protein kinase II (CaMKIIβ), neurogranin, and several synaptotagmin isoforms [9]. Thus, the underlying mechanisms of induction and/or expression of LTD may differ between these rodent species.

Our interest in studying tau deletion in rats stems from the evolutionary divergence in tau isoform expression [5]. Specifically, like humans, juvenile rats express all six CNS tau isoforms, and this persists into adulthood [7, 8]. In contrast, juvenile mice predominantly express the 0N3R isoform, which wanes with age, as well as the 0N4R tau isoform, which increases alongside 1N4R and 2N4R into adulthood [8], while expression of 1N3R and 2N3R isoforms is lacking [6]. These differences may be critical since different tau isoforms stabilise microtubules to different extents, have unique interaction partners related to synaptic function, and have distinct sub-cellular localisation patterns [6]. As such, rats are increasingly being used to study tauopathies. For instance, recent findings in male rats demonstrate that acute intracerebroventricular injection of oligomer-enriched tau facilitates LTD while fibril-derived soluble tau aggregates inhibit LTD [16]. Therefore, the generation of our Mapt KO rat line (SD-Mapt^em1Tcp) offers a useful tool for studying the role of endogenous tau in AD and other neurodegenerative diseases.

Future experiments should consider (i) whether this species difference is also present in female rodents, given that sex differences in tau phosphorylation have been reported in rats [17], and (ii) how tau deletion impacts synaptic plasticity, learning and memory with age since the ratio of tau isoform expression [7, 8] and post-translational modifications [1] change across the lifespan. Indeed, tau deletion has been shown to not impact LTD induced by a stronger LFS protocol (2 Hz, 1800 pulses) in 6-month-old adult mice [18]. This may reflect the decline in LTD magnitude and a change of underling mechanism with age [19, 20]. Whether tau loss-of-function nonetheless reactivates LTD pathways in the context of tau pathology, and whether this differs between sexes or with age in rats, remains to be shown. In conclusion, the species difference we report here highlights the importance of selecting the most appropriate animal model for investigating the physiological and pathological functions of tau in relation to human health and disease.