Exp Neurobiol 2019; 28(2): 172-182
Published online April 19, 2019
© The Korean Society for Brain and Neural Sciences
Seunghyuk Choi1, Ali Sadra1, Jieun Kang1, Jae Ryun Ryu2, June Hoan Kim2, Woong Sun2*, and Sung-Oh Huh1*
1Department of Pharmacology, College of Medicine, Institute of Natural Medicine, Hallym University, Chuncheon 24252, Korea.
2Department of Anatomy, Korea University College of Medicine, Brain Korea 21 Plus, Seoul 02841, Korea.
Correspondence to: *To whom correspondence should be addressed.
Sung-Oh Huh, TEL: 82-33-248-2615, FAX: 82-33-248-2612
Woong Sun, TEL: 82-2-2286-1404, FAX: 82-2-929-5696
Rheb (Ras homolog enriched in the brain) is a small GTPase protein that plays an important role in cell signaling for development of the neocortex through modulation of mTORC1 (mammalian-target-of-rapamycin-complex-1) activity. mTORC1 is known to control various biological processes including axonal growth in forming complexes at the lysosomal membrane compartment. As such, anchoring of Rheb on the lysosomal membrane via the farnesylation of Rheb at its cysteine residue (C180) is required for its promotion of mTOR activity. To test the significance of Rheb farnesylation, we overexpressed a farnesylation mutant form of Rheb, Rheb C180S, in primary rat hippocampal neurons and also in mouse embryonic neurons using
In the developing brain, axonal polarization and elongation are the earliest steps in neuronal differentiation  and the process of neuronal polarity establishment, having been studied in primary cultures of cortical and hippocampus neurons . The very first steps in the process of neuronal polarization have been defined as releasing of four to five lamellipodia and small protrusion veils for the developing neuron and one of neurites elongating rapidly to become the cell's axon while other minor neurites become dendrites [3,4]. In the process of axonal polarization/elongation, mTOR (mammalian target of rapamycin) has been reported as a key regulator of axogenesis. mTOR is a serine/threonine kinase and a central regulator of various processes in mammalian cells via its major two downstream targets of EBP1 (eIF-4E binding protein 1), which initiates eukaryotic translation, and S6K (70kDa ribosomal S6 protein kinase), which promotes cell growth. Rheb (Ras homolog enriched in brain) is a small GTPase Ras family member, and it plays an important role in the PI3K/AKT/TSC/mTORC1 signaling pathway to activate mTOR . In neuronal growth, mTOR signaling has been shown to promote cell growth and axon formation [6,7,8,9,10,11] and inhibition of mTORC1 by its specific inhibitor, rapamycin, blocks axon formation. Conversely, activation of Rheb activity enhances axonal growth and multiple axon formation . The key mechanism for mTORC1 apparently involves activation of local translation of proteins, required for axonal growth and axonal outgrowth in response to external cues such as Ephrins . mTOR-dependent local translation products related to axon specification, CRMP2 (collapsing response mediator protein 2) and tau, are also mediated by Rheb activation. Rheb-mTORC1 activation of S6K also increases in axon formation, and constitutively active S6K also leads to formation of multiple axons . Activation of Rheb also promotes intrinsic sensory axon regrowth in dorsal root ganglia (DRG) models of nerve injury .
Because mTOR signaling complex forms at the lysosomal membrane, it is believed that the anchoring of these complex components at the membrane is important. Accordingly, Rheb shares the ‘CAAX motif’ sequence identity with Ras, shown to be required for membrane localization of each GTPase via farnesylation. As such, this motif is a target for farnesyltransferase inhibitors that can block Rheb and Ras processing and function . With a farnesyl-transferase inhibition via farnesylthiosalicylic acid, Rheb was dispersed to the cytosol and its function was blocked. It also prevented S6K activation induced by a constitutively active mutant of mTOR, indicating that farnesylation of Rheb is critical for the mTOR activation .
Recent research has shown that Rheb also activates pathways other than mTOR signaling. For instance, constitutively active mutant of Rheb (Rheb S16H) also induced GSK3β, which can contribute to axonal growth by induction of MAP1 expression . Thus, the importance of Rheb farnesylation in the neuronal growth is not yet completely understood. Here, we investigated the effect of mutation of the CAAX motif in Rheb as required for farnesylation of Rheb on axonal morphology changes in the primary hippocampal neuron cultures. The farnesylation-block of Rheb was achieved via C180S mutation of the molecule. We show that overexpression of Rheb C180S mutant also induces elongation of axonal length but without formation of multiple axons as seen with the overexpression of wild-type Rheb. Increased phosphorylation of 4EBP1 is also seen with overexpressed Rheb C180S and seen in an mTOR independent fashion. Therefore, farnesylation-defective Rheb points to an mTOR1 independent pathway in axogenesis and a non-canonical role for Rheb.
Experiments in C57BL/6J strain mice (Daehan Bio-Link, Eumseong-gun, South Korea) were conducted in compliance to animal handling guidelines by the Institutional Animal Care and Use Committee (IACUC) of Hallym University (ethical clearance number: Hallym2016-63, South Korea). Similarly, experiments with rats (Orient Bio, Sung-nam si, South Korea) were approved by the IACUC of Korea University (ethical clearance number: KOREA-2016-0195, South Korea). The animals were initially purchased and then housed in individual cages in 12-hour light/dark cycles at 22±2℃, 50±10% humidity with food and water available ad lib. Animal feed was procured commercially (Purina Inc., Sungnam si, South Korea).
Oligo primers via PCR were used to produce the mutations for various vectors (Tables 1 and 2). The base vectors that were used included pEGFP-C1 (expressing an N-terminal GFP fusion protein) (Sung Ho Ryu; Postech, South Korea) , pCAGIG (co-expressing a bicistronic EGFP) (Connie Cepko; plasmid 11159; Addgene, Watertown, MA, USA) and pmCherry-C1 (expressing an N-terminal mCherry, a mutant fluorescent protein) (Cat. No. 632524; Clontech, Mountain View, CA, USA). pEGFP-C1-Rheb WT and Rheb D60I constructs (GenBank sequence FQ219039.1) were kindly provided by Dr. Sung Ho Ryu from Postech, South Korea. pCAGGS-4EBP1 F113A construct was from Angelique Bordey (Addgene; plasmid # 81122). The 4EBP1 F113A insert was then subcloned into pCAGIG.
Tissue fragments were dissected out from embryonic day 18 (E18) rat hippocampus samples (Orient Bio, South Korea) and were digested with papain (Worthington Biochemical, Lakewood, NJ, USA) containing L-cysteine (15.76 mg/ml). Dissociated primary rat hippocampal neurons were then transfected by nucleofection (Mirus Bio, Madison, WI, USA) and cultured in Neurobasal medium (Invitrogen, Carlsbad, CA, USA) with B-27 supplement (ThermoFisher, Waltham, MA, USA). The dissected neurons were plated on coverslips coated with Poly-L-lysine (Sigma-Aldrich) at a concentration of 5x105 cells in a 24 well-plate and cultured at 37℃ in humidified, 5% CO2 incubator.
Primary neurons were fixed in 4% PFA in 30% sucrose for 15 min at room temperature. After a PBS wash, the cells were blocked for 1 h in 3% BSA in 0.02x TBST and incubated at 4℃ overnight with antibodies against GFP (Abcam; 1:1000), mCherry (Abcam; 1:500), GM130 (Abcam; 1:250), LAMP1 (Abcam; 1:1000), tau-1 (Millipore, Burlington, MA, USA; 1:200), MAP2 (Millipore; 1:2000), p4EBP1 (Thr37/46) (Cell Signaling, Danvers, MA, USA; 1:400) or pmTOR (Ser2448) (Cell Signaling; 1:100). After washing with 1X TBST, the samples were incubated with secondary antibodies (Alexa fluor conjugated; Life Technologies) for 2 h at room temperature and were mounted in mounting media after 3 times wash with 0.5X TBST. All images were observed and captured on a confocal microscope (Leica SP8; Leica, Wetzlar, Germany).
Neuro2A (N2A) cells were procured from JCRB Cell Bank (Osaka, Japan) and cultured at 37℃ in humidified, 5% CO2 incubator. Western blotting involved lysing N2A cells in RIPA buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM EGTA, 1 mM EDTA, 1% Triton X-100, 1 mM Na3VO4, 5 mM NaF and a protease inhibitor cocktail). Equal amounts of protein were then separated via SDS-PAGE and transferred onto polyvinylidene difluoride (PVDF) membranes (Millipore) and blocked with 5% skimmed milk in 1X TBST. Membranes were probed in primary antibody for overnight at 4℃ with antibodies (all from Cell Signaling) against p4EBP1(Ser65) (1:1000), p4EBP1 (Thr37/46) (1:1000), total 4EBP1 (1:1000), pS6K1 (1:1000) or total S6K1 (1:1000). After washing, the blots were incubated with the appropriate secondary antibody and the protein bands were resolved with Western HRP substrate (Luminata Forte; Millipore).
Statistical analysis was performed by using GraphPad Prism software (GraphPad, San Diego, CA, USA). The p-values were from one-way ANOVA analysis with a post-hoc Turkey test and data were expressed as mean and SEM.
A previous report has indicated that overexpression of wild-type Rheb is associated with an elongated axon . Here, we examined whether a farnesylation-deficient version of Rheb (Rheb C180S) would have any effect on neuronal polarity and axonal length in transfected cells. We cultured primary embryonic day 18 (E18) rat hippocampal neurons after transfection with EGFP co-expressing Rheb WT, Rheb C180S and control vectors and measured length and number of axons of the transfected cells after 3 days
To explore the activity of Rheb C180S in relation to the mTORC1 pathway, we transfected mouse neuroblastoma Neuro2A cells with EGFP co-expressing Rheb WT, Rheb C180S or control vectors (Fig. 2A and 2B). The transfection generated differential phosphorylation of mTORC1 effectors (phospho-S6K, pS6K Ser240/244; phospho-EBP1, p4EBP1 Ser65 and p4EBP1 Thr37/46). We observed an increase in pS6K levels with Rheb WT (124±6.7% of control) transfection compared with the control vector transfected cells; in contrast, there was a decrease in the levels of p70S6K with Rheb C180S transfection (70±6.1% of control) (Fig. 2A and 2B), suggesting a partial dominant-negative effect on the activity of p70S6K. Surprisingly, Rheb C180S transfected cells had an increase in their p4EBP1 levels compared with the control-vector transfectants. This was seen with both p4EBP1 Ser65 (127±9% of control) and p4EBP1 Thr37/46 (137±7.3% of control) antibody staining (Fig. 2A and 2B). This led us to conclude that altering the Rheb farnesylation status of Rheb has a differential effect on the activity of the two mTORC1 pathway substrates, namely S6K and 4EBP1. 4EBP1 is a known regulator of cap-dependent translation and is also reported to be downstream of Rheb-mTOR pathway with phosphorylation of 4EBP1 being specifically localized to the axon [15,16,17]. To further explore the axonal elongation link between Rheb C180S and 4EBP1, we examined levels of p4EBP1 Thr37/46 in Rheb WT, Rheb C180S and control transfected E18 rat hippocampal neurons. At the 3
Considering that Rheb WT and C180S induced differential mTOR downstream phosphorylation (Fig. 2A and 2B), we wondered whether increases in p4EBP1 levels due to Rheb C180S were indeed mTORC1 dependent. To verify this in our test cells, we investigated whether Rheb C180S also promoted the phosphorylation of mTOR (pmTOR Ser2448), which is a direct indicator of mTORC1 activity . After transfection in E18 rat primary hippocampal neurons, induction of pmTOR Ser2448 intensity was detected in Rheb WT transfectants more than the control vector transfected cells. By contrast, Rheb C180S transfected cells had decreased levels of pmTOR Ser2448 in both the axon and the cell body when compared with the control-vector cells (Fig. 3A and 3B), suggesting that Rheb C180S did not enhance mTORC1 activity. Accordingly, axon formation in Rheb WT and control vector transfected cells was completely abolished with rapamycin (200nM), an inhibitor for mTORC1, treated cells. But interestingly, Rheb C180S-transfected neurons maintained their axons at lengths similar for rapamycin-untreated cells (Fig. 3C and 3D). We next examined whether Rheb C180S leading to axon elongation depended on Rheb GTPase activity. We cultured primary embryonic day 18 (E18) rat hippocampal neurons after transfection with EGFP co-expressing GTPase defective Rheb (Rheb D60I), Rheb C180S, Rheb D60I C180S (double mutant) and control vectors. The axon lengths were measured after 3
Our results indicate that cytosolic localization of a farnesylation-defective version of Rheb (Rheb C180S) can induce axonal elongation, similar to Rheb WT. This was also independent of its GTPase activity, but still dependent on the phosphorylation of 4EBP1. We found that the wild-type and farnesylation-defective versions of Rheb exert their promotional roles in axonal growth using different downstream cascades as there were differences in the phenotype of the transfected neurons between the two groups. Rheb WT transfection induced multiple axons per cell whereas Rheb C180S only maintained axonal elongation to a specific single axon (Fig. 1). This Rheb mutant also induced callosal axonal elongation in the developing neocortex of mouse embryos (Fig. 1). Previous reports have shown a Rheb WT promoting axonal elongation and disrupting radial migration. This abnormal activation of mTORC1 impairs Reelin-Dab1 signaling, contributing to neuronal migration defects [42,43]. However, Rheb C180S only promoted callosal axon extension, but radial migration was only marginal (Fig. 1). Collectively, these results indicate that Rheb WT and Rheb C180S share a common activity in promoting axonal elongation, but not for the other phenotypes in the two different model systems.
We reasoned that these were caused by different downstream signaling cascades. A previous report has shown that Rheb WT induces the formation of multiple axons dependent on the activation of the Rheb-mTORC1-S6K axis . However, Rheb C180S overexpression increased neither phosphorylation of S6K nor mTORC1 activation. On the other hand, Rheb C180S strongly promoted 4EBP1 phosphorylation. With respect to 4EBP1, its expression is important in promoting axon specification. Hyperphosphorylation of 4EBP1 is required to disrupt its interaction with the translation initiation factor eIF4E, leading to activation of cap-dependent translation. Previously, two reports have shown that mTOR-insensitive unphosphorylated-4EBP1 and 4EBP1 Δ24, both capable of constitutively binding to eIF4E, also prevented axon specification [15,17]. We conjecture that Rheb C180S induced p4EBP1 levels, but not through an activation of mTORC1. Rheb C180S transfected neurons had decreased pmTORC1 levels in both the cell body and the axons (Fig. 3). In neurons, rapamycin is known to block both pS6K and p4EBP1 levels and axon specification from suppression of mTORC1 . This was also seen in our experiments for both control and Rheb WT transfectants (Fig. 3D). Rheb C180S-transfected cells maintained their axons even in presence of rapamycin, suggesting that increases in p4EBP1 levels in transfected cells were not dependent on increased mTORC1 activity. This phosphorylation of 4EBP1 may be occurring non-canonically and occurring via an arm of the Rheb pathway that is mTORC1 independent.
We have currently no clear answer of what could be the mechanism for Rheb C180S in promoting axonal elongation in an mTORC1-independent manner, but we may speculate several possible mechanisms. As opposed to mTORC1, farnesylation-block of Rheb may be able to upregulate another member of mTOR, mTORC2. In this pathway, farnesylation-block of Rheb may promote cytoskeletal protein formation downstream of mTORC2 via PKC (protein kinase C), Rho, Rac, and thus promoting axon elongation [29,30,31,32]. It has been reported that Rheb regulates dendritic spine morphogenesis in mTORC1-independent ways via an interaction of Rheb with syntenin [40,41]. Thus, other partner kinases participating in phosphorylation of 4EBP1 may mediate cytosolic Rheb actions. For example, Cdc2 can phosphorylate 4EBP1 . Cdc2 activation is also positively linked to axon regeneration . Another candidate is PDK1 which can modulate the phosphorylation of 4EBP1 . As PDK1 contains a PH (pleckstrin homology) domain, disrupting its PH domain also decreased axon length . Src (p60src) also leads to 4EBP1 phosphorylation in a rapamycin-resistant manner  and Src kinases have been implicated in guidance and growth of axons as Src promotes stability of the growth cones . In addition, PAK2 is a newly described effector of TSC1/TSC2 complex, upstream of Rheb-mTORC1, and it affects 4EBP1 phosphorylation in an mTORC1-independent fashion .
In summary, farnesylation-defective Rheb can promote phosphorylation of 4EBP1 and axonal elongation in an mTORC1 independent manner, suggesting that Rheb may regulate axonal growth while dissociated from the lysosomal membrane during the axonal elongation in developing neurons.