Distinct and common functions of mTORC1 and mTORC2 in Purkinje cells

In mammalian cells, the serine/threonine protein kinase mTOR (mammalian target of rapamycin) is present in two complexes, called mTORC1 and mTORC2. While several of the components are common to both complexes, raptor and rictor are only associated with mTORC1 or mTORC2, respectively. Due to differences in their molecular composition mTORC1 and mTORC2 possess distinct functions and properties (Laplante & Sabatini, 2012). For example, mTORC1 but not mTORC2 is sensitive to the immunosuppressive drug rapamycin. mTORC1 integrates various extracellular signals (e.g. growth factors, energy status or amino acid availability) to promote protein synthesis, to regulate lipogenesis and to inhibit autophagy (Shimobayashi & Hall, 2014). In line with these features, mTORC1 was found to be essential for cell growth and proliferation. In contrast, activation and function of mTORC2 is less well understood. It phosphorylates and activates members of the AGC kinase family, including Akt, SGK1 and PKC, suggesting a role in cell survival/metabolism and actin cytoskeleton organization.
In the brain, mTOR signalling has been implicated in several neurodevelopmental and neurodegenerative disorders like autism spectrum disorders (ASD) or Huntington’s disease. The availability of approved drugs, such as rapamycin and its analogs (called rapalogs), has made the mTOR signalling pathway an attractive target for the treatment of those diseases. Although rapamycin has been shown to preferentially target mTORC1, prolonged exposure also inhibits mTORC2 (Sarbassov et al., 2006). Thus, it is important to unravel the specific and the common functions of mTORC1 and mTORC2 in the central nervous system.
In this study, the roles of mTORC1 and mTORC2 were analysed in Purkinje cells by conditionally deleting floxed Rptor or Rictor genes, respectively, using an L7/Pcp-2-driven expression of the Cre recombinase. The resulting mouse lines are called RAPuKO or RIPuKO, which stands for raptor or rictor Purkinje knockout, respectively, and allowed to study the functions of mTORC1 and mTORC2 in developing and adult Purkinje cells and to investigate the effect on mouse behaviour.
We found that the phenotypes of RAPuKO and RIPuKO mice only sparsely overlapped but mostly differed, which assigns mTORC1 and mTORC2 distinct functions in these neurons. (I) mTORC1, but not mTORC2 abrogation in Purkinje cells reduced the social interest of mice. (II) Ablation of either mTORC1 or mTORC2 in Purkinje cells was sufficient to cause motor coordination deficits, yet, for RAPuKO mice the onset of these deficits was age-dependent while motor deficits of RIPuKO mice could be detected at any tested age. (III) The motor phenotype of RIPuKO mice was accompanied by developmental aberrations, such as impaired climbing fibre synapse elimination and hampered dendritic self-avoidance, while the age-dependent motor phenotype of RAPuKO mice seemed to be driven by Purkinje cell degeneration that finally led to apoptosis and a loss of these neurons. Vice versa, no signs for deficient climbing fibre elimination or Purkinje cell loss could be detected for RAPuKO or RIPuKO mice, respectively. (IV) mTORC1 and mTORC2 ablation in Purkinje cells both affected neuron morphology in a similar manner, which included multiple primary dendrites and a reduction of the neuron size, yet, last was more pronounced for raptor-deficient cells.
Altogether, both mTORC1 and mTORC2 ablation in Purkinje cells had a pronounced, yet distinct, effect on these neurons and the mouse behaviour, unlike in other tissues where inactivation of mTORC2 has been reported to result in a minor phenotype in comparison to mTORC1 ablation (Bentzinger et al., 2008; Godel et al., 2011).
While ablation of mTORC1 and mTORC2 in Purkinje cells resulted in mostly distinct phenotypes, we found that sustained mTORC1 activation in these neurons by a TSC1 knockout (TSCPuKO) caused a phenotype that was similar to the one of RAPuKO mice. In both RAPuKO and TSCPuKO mice an age-dependent loss of Purkinje cells due to apoptosis was observed, which was paralleled by reactive gliosis. Moreover, in both cases Purkinje cell apoptosis was preceded by signs of neurodegeneration in form of axonal swellings that accumulated neurofilaments. Also in terms of behaviour similar phenotypes were observed since both knockout mice showed reduced social interest (Tsai et al., 2012). These behavioural phenotypes support the growing notion that the cerebellum is important for non-motor related functions (Schmahmann et al., 2007; Wang et al., 2014) and that mTORC1 plays a role therein. TSC1 knockout in Purkinje cells has been reported to cause also repetitive behaviour in mice in addition to abnormal social behaviour and therefore it has been suggested that these mice show an autism-like phenotype (Tsai et al., 2012).
In summary, this study provides in vivo data for the importance of mTORC1 and mTORC2 in developing and adult Purkinje neurons. We find that both complexes are crucial for Purkinje cells, yet, in mostly distinct manners. This finding is in line with the model that mTORC1 and mTORC2 largely feed separate downstream effectors, although they share many molecular components. The knowledge of the function of mTORC1 and mTORC2 in adult neurons is important for the development of treatment options that target the mTOR pathway. This work clearly suggests that such drugs need to be highly selective for the different complexes. Moreover, this work highlights that a complete inhibition of mTORC1 may have detrimental effects on the survival of neurons and that this may also precipitate autism-like pathologies.