IGF-1 mediated Neurogenesis Involves a Novel RIT1/Akt/Sox2 Cascade

Insulin-like growth factor 1 (IGF-1) is known to have diverse effects on brain structure and function, including the promotion of stem cell proliferation and neurogenesis in the adult dentate gyrus. However, the intracellular pathways downstream of the IGF-1 receptor that contribute to these diverse physiological actions remain relatively uncharacterized. Here, we demonstrate that the Ras-related GTPase, RIT1, plays a critical role in IGF-1-dependent neurogenesis. Studies in hippocampal neuronal precursor cells (HNPCs) demonstrate that IGF-1 stimulates a RIT1-dependent increase in Sox2 levels, resulting in pro-neural gene expression and increased cellular proliferation. In this novel cascade, RIT1 stimulates Akt-dependent phosphorylation of Sox2 at T118, leading to its stabilization and transcriptional activation. When compared to wild-type HNPCs, RIT1 −/− HNPCs show deficient IGF-1-dependent Akt signaling and neuronal differentiation, and accordingly, Sox2-dependent hippocampal neurogenesis is significantly blunted following IGF-1 infusion in knockout (RIT1 −/−) mice. Consistent with a role for RIT1 function in the modulation of activity-dependent plasticity, exercise-mediated potentiation of hippocampal neurogenesis is also diminished in RIT1 −/− mice. Taken together, these data identify the previously uncharacterized IGF1-RIT1-Akt-Sox2 signaling pathway as a key component of neurogenic niche sensing, contributing to the regulation of neural stem cell homeostasis.


Results
Loss of RIT1 alters exercise-induced neurogenesis. We have shown that RIT1 deficiency alters hippocampal neurogenesis following traumatic brain injury by delaying the post-concussive recovery of immature neurons without affecting basal rates of neurogenesis 36 . However, whether RIT1 contributes to the regulation of adult neurogenesis more broadly in response to physiological stimuli is unknown. Importantly, RIT1 is known to be expressed in HNPCs, supporting a potential role for RIT1 in the regulation of neural progenitor function 39 . To assess the role of RIT1 in voluntary exercise-enhanced hippocampal neurogenesis 40 , RIT1 knockout (RIT1 −/− ) (n = 17) and wild-type (n = 23) mice were randomly assigned to either sedentary or running groups (running wheels were preinstalled in the housing cages permitting voluntary exercise), and the impact of RIT1 loss on exercise-enhanced neurogenesis assessed at either day 16 or 42 ( Fig. 1A). Both wild-type and RIT1 −/− mice in the runner groups ran an average distance of ~10 kilometers a day with no inherent difference arising from RIT1 deficiency (wild-type: 10.25 ± 1.14 km/d; RIT1 −/− : 10.10 ± 0.59 km/d, p = 0.86, n = 3). During the trial, mice received one daily intraperitoneal BrdU injection (50 mg/kg) for the first 2 weeks, to label proliferating neuroblasts (BrdU + /DCX + cells) (analysis day 16) (Fig. 1B) and maturing neurons (analysis day 42; 1 month post-BrdU chase) (BrdU + /NeuN + ) (Fig. 1C). While lineage tracing detected approximately equivalent numbers of BrdU labeled proliferating neuroblasts (p > 0.05) and mature neurons (p > 0.05) in sedentary housed mice (p > 0.05), RIT1 −/− mice displayed a significantly lower density of proliferating neuroblasts (*p < 0.01), and neurons that matured from these neuroblasts following running exercise than wild-type controls (*p = 0.01) (Fig. 1D,E). These data suggest that RIT1 signaling contributes to the proliferation and neuronal differentiation following voluntary exercise. RIT1 contributes to IGF-1 dependent neurogenesis. Running exercise increases the availability of several classes of growth factor, including BDNF and IGF-1, which have known roles in regulating adult neurogenesis 30 . While RIT1 plays a role downstream of diverse mitogen-activated receptors 34 , we have previously shown that BDNF signaling in primary hippocampal neuron cultures is not altered by RIT1 deficiency 36 . In agreement with earlier in vitro studies 41 , IGF-1 exposure (100 ng/ml, 15 min) led to robust ERK and Akt activation in wild-type hippocampal cultures ( Fig. 2A). Importantly the activation of both kinases was blunted (~55% of kinase phosphorylation of WT hippocampal neuronal cultures, n = 3, p < 0.05) in RIT1 −/− cultures as monitored by anti-phospho-specific immunoblotting ( Fig. 2A). Consistent with a role for RIT1 in IGF-1 signaling, wild-type primary hippocampal neural progenitor cells (HNPCs) (Fig. 2B) displayed increased proliferation (p < 0.01) following IGF-1 exposure (Fig. 2C,F), while RIT1 −/− HNPCs failed to respond (p > 0.05) (Fig. 2E,G), as assessed by confocal microscopy (co-stained Nestin + /Ki67 + cells). Expression of Myc-tagged RIT1 rescued IGF-1 dependent proliferation in RIT1 −/− HNPCs (p < 0.05) (Fig. 2D,E and G). These data suggest that RIT1 plays a key role in IGF-1 signaling and contributes to HNPC proliferation in vitro.
We next asked whether RIT1 signaling contributes to IGF-1-dependent in vivo stimulation of hippocampal neurogenesis 15 . In agreement with previous studies 15,16,18 , peripheral infusion of exogenous recombinant IGF-1 (500 ng/kg/day) (Fig. 3A) was found to induce neurogenesis in the mouse hippocampus (Fig. 3B). Using BrdU labeling, we found a significant increase in newborn BrdU + /DCX + immature neurons in the dentate granule cell layer of the hippocampus of WT mice after 7 d of peripheral IGF-1 administration, when compared to vehicle controls (Fig. 3B,C). While vehicle treated WT and RIT1 −/− mice displayed similar numbers of BrdU + /DCX + newborn immature neurons, IGF-1 dependent progenitor cell proliferation was significantly blunted in the dentate of RIT1 −/− mice (Fig. 3B,C) (p < 0.05). Taken together, both in vivo and in vitro data indicate that RIT1 plays a critical role in IGF-1 induced neurogenesis.
Adult neurogenesis involves the activation of a multistep transcriptional network 42 including the sequential expression of Ascl1 43 and NeuroD1 44, 45 transcription factors. To determine whether IGF-1 stimulation of HNPCs results in the activation of this transcriptional cascade, we used immunohistochemical analysis to determine the effect of IGF-1 on Ascl1 and NeuroD1 expression in HNPCs. As seen in Fig. 5, IGF-1 stimulation of HNPCs (50 ng/ml IGF-1, 24 h) resulted in a prominent increase in both Ascl1 ( Fig. 5A and C) and NeuroD1 ( Fig. 5B and D) expressing HNPCs. Consistent with a role for RIT1 in this process, IGF-1 stimulation failed to significantly increase levels of either Ascl1 or NeuroD1 in RIT1 −/− HNPCs ( Fig. 5E-H) (p > 0.05), but importantly, Myc-RIT1 re-expression was capable of restoring IGF-1-dependent increases in both transcription factors ( Fig. 5E-H) (p < 0.05). Consistent with these data, RT-PCR analysis demonstrates that IGF-1 stimulation leads to an increase in the expression of both NeuroD1 and Ascl1, in a manner that depends upon RIT1 (Fig. 5I). Furthermore, immunoblotting demonstrates a RIT1-dependent increase in Ascl1 protein levels following IGF-1 (20 ng/ml) stimulation in HNPCs (Fig. 5J). Thus, RIT1 deficiency blunts IGF-1-dependent regulation of pro-neurogenic gene expression and neural differentiation of HNPCs.

Discussion
Studies in the mammalian central nervous system support an essential role for insulin and insulin-like growth factors in stem cell self-renewal, neurogenesis, and cognitive function through distinct ligand-mediated receptor activation cascades 13 . Although IGF-1 has long been associated with the regulation of neural stem cell biology, and much is known about the diversity of IGF-1-dependent signaling cascades 19 , mechanisms by which IGF-1 governs neurogenesis remain incompletely characterized. Here, we provide evidence for the involvement of RIT1 as a critical downstream component of IGF-1-dependent, and exercise-mediated enhancement, of hippocampal neurogenesis. We demonstrate that, IGF-1 regulates Sox2 activation and neuronal differentiation in a RIT1-dependent fashion in HNPCs, with RIT1 directing IGF-1-mediated expression of the pro-neuronal Ascl1 and NeuroD1 genes, and Akt-dependent up-regulation of Sox2 transcriptional activity. Thus, we demonstrate that RIT1/Akt signaling plays a crucial role in IGF-1 dependent neural cell fate selection by controlling Sox2 function.
During both CNS development and adult neurogenesis, IGF-1/IGF-1R signaling has been found to regulate the proliferation and survival of neuronal progenitors as well as the generation, differentiation, and maturation of neurons 11,12,18,[49][50][51] . Circulating IGF-1 has been implicated in the beneficial effects of exercise on brain function, including increased hippocampal neurogenesis 15,52 . Previous findings have also identified roles for IGF-1 signaling in neuronal precursor proliferation and neuronal differentiation 24, 53-56 . Our results provide new insight into the molecular mechanisms underlying exercise-enhanced and IGF-1-driven hippocampal neurogenesis. The blunted neurogenic response of exercised RIT1 −/− mice suggests that RIT1 participates in the transduction of running stimuli to promote enhanced hippocampal neurogenesis, whether in response to IGF-1 or another exercise myokine 57,58 . Not surprisingly, RIT1 deficiency does not result in a complete block of exercise-enhanced neurogenesis ( Fig. 1), as neurogenesis is controlled by a broad and sophisticated regulatory network 6 . Indeed, other Ras-related G-proteins are likely involved, as Ras-GRF2, a calcium-regulated exchange factor for Ras and Rac GTPases, has been found to contribute neurogenesis in response to enriched environment 59 . However, the failure of IGF-1 infusion to stimulate neuroblast proliferation in RIT1 −/− mice suggests that RIT1 plays an essential role in IGF-1 dependent regulation of hippocampal neurogenesis, as other Ras family GTPases cannot compensate for RIT1 loss. While a number of studies have implicated elevated IGF-1 levels as a putative mediator of exercise induced neurogenesis 15,52 , and our data would support a role for IGF-1/RIT1 in the benefits of exercise on brain plasticity, a growing literature suggests that a variety of factors beyond IGF-1 likely contribute to this process. Indeed, recent studies have identified the exercise myokine, cathepsin B (CTSB) 58 , as an important mediator of the neurogenic benefits of exercise, and additional peripheral blood factors have been shown to improve brain plasticity in aged animals 57 . Therefore, further studies are needed to determine whether RIT1, or other Ras family GTPases, contribute to these signaling pathways.
A growing literature supports a central role for Akt is the regulation of neuronal stem cell proliferation 60, 61 , including the control of exercise-mediated hippocampal neurogenesis 28,62 and IGF-1 signaling in neuronal stem cells 22,41,63 . Our observation that IGF-1-mediated ERK and Akt activity is blunted in RIT1 −/− HNPC cultures (Fig. 2), but not following BDNF stimulation 36 , suggests that RIT1 deficiency might generate an IGF-1 selective, rather than global, defect in growth factor signaling within the HPNC niche. While these data implicate RIT1 as a key downstream regulator of neuronal IGF-1 signaling, the full contribution of RIT1 to IGF-1 signal transduction remains to be addressed, particularly details on how RIT1 deficiency impacts gene expression. In addition, it is important to identify the guanine nucleotide exchange factor(s) (GEF) involved in coupling IGF-1R activation to stimulation of neuronal RIT1 signaling. This is complicated by the fact that while activation of a SOS/Shc-Grb2 complex has been associated with in vitro RIT1 activation following either NGF or PCAP stimulation of pheochromocytoma cells, biochemical analysis has yet to identify a bona fide RIT1GEF. IGF-1 has been shown to be neuroprotective in models of traumatic brain injury 64, 65 , stroke and ischemic injury 66, 67 , preventing apoptotic death and promoting cell survival. As expression of active RIT1 has been shown to be neuroprotective in vitro 37 , studies are underway to determine activated RIT1 signaling is neuroprotective, capable of reducing behavioral deficits in the setting of brain trauma.
Sox2 is involved in the maintenance and proliferation of NPCs in vivo, neurosphere formation in vitro 68,69 , and plays a key role in somatic cell reprogramming [70][71][72][73] . Furthermore, voluntary running is known to increase dividing Sox2 + HNPCs in the dentate gyrus leading to enhanced neurogenesis 27 . We find potentiation of Sox2 transcriptional activity, stabilization of Sox2 protein levels, and increased proliferation and neural differentiation of HNPCs following IGF-1 stimulation, suggesting that IGF-1/RIT1 regulation of Sox2 may represent a fundamental mechanism for controlling neurogenesis. IGF-1 has been shown to regulate Sox2 in a variety of systems, including human mesenchymal 74 and colonic stem cells 75 . Thus, it is possible that IGF-1/RIT1 signaling might regulate Sox2 activity in diverse cell populations. We present data indicating that Akt is required for IGF1/RIT1-dependent Sox2 activity. This extends recent studies showing that Akt signaling contributes to the regulation of embryonic stem cell fate by controlling Sox2 stabilization and transcriptional activity by a phosphorylation at threonine 118 (T118) 47 . Consistent with a conserved regulatory mechanism, we find that IGF-1 stimulates Sox2 T118 phosphorylation in HNPCs, resulting in higher levels of Sox2 T118 + nuclei (Figs 6 and 8). Using RIT1 −/− HNPCs, RNAi methods, and pharmacological Akt inhibition, we demonstrate that RIT1 and Akt are required for IGF-mediated Sox2 transcriptional activation and neuronal differentiation (Figs 6-8). Following traumatic brain injury we have shown that conditional overexpression of IGF-1 leads to increased hippocampal neurogenesis 64,76 , while post-TBI neurogenesis is delayed in RIT1 −/− mice 36 . Whether these changes rely upon changes in Sox2 function remains to be determined.
Interestingly, there is a growing literature suggesting that IGF-1 is critical for malignant transformation and metastasis of cancer cells [77][78][79] , and that Sox2 controls tumor initiation and cancer stem-cell function 80,81 . We have found that conditional RIT1 overexpression in the dentate gyrus leads to the robust generation of neuroblasts 39 . Recently, we also identified RIT1 as a novel driver oncogene in a subset of human lung adenocarcinomas, and our data suggest PI3K/Akt inhibition as a potential therapeutic strategy in RIT1-mutated tumors 38 . Studies are underway to determine whether oncogenic RIT1-dependent tumorigenesis involves activation of Akt-Sox2 signaling.
In summary, we demonstrate a key role for the Ras-related GTPase, RIT1, in both IGF-1 and exercise-induced neurogenesis. IGF-1 dependent NPC proliferation and neural differentiation are inhibited by genetic deletion of the RIT1 gene. We also present data that provides new insight into the mechanisms underlying IGF-1-directed neurogenesis. Our findings reveal that a previously uncharacterized IGF-1/RIT1/Akt/Sox2 signaling cascade is import for regulating the proliferation and differentiation of neural precursors within the dentate gyrus. Collectively, these data provide new insight into mechanisms by which IGF-1 signaling and Sox2 function in neural stem-cell maintenance, embryogenesis and neuronal development.

Materials and Methods
Mouse running and BrdU labeling. RIT1 −/− mice have been previously described 36 . 12-week-old male RIT1 −/− mice and their WT littermates were divided into 2 groups and randomly placed in standard-sized rat cages (6 mice per cage) with (running group) or without (sedentary group) 2 (6 inch) running wheels 82 . Wheel running activity for each genotype was monitored 24 h/day and 7 days/week using ClockLab software (Actimetrics, Inc). For lineage labeling experiments, each mouse independent of housing was intraperitoneally (i.p.) injected with BrdU (50 mg/kg, in 0.9% Saline) once daily for the first 14 days of the experiment. Newborn immature neuron assessment was performed 2 days after the last BrdU injection (day 16). Neural maturation was quantified 1 month after the last BrdU injection (day 42). Following the either length of housing (16 or 42 days) mice were anesthetized with sodium pentobarbital (65 mg/kg, i.p.), transcardially perfused with heparinized ) and immunostained for phospho-Sox2 T118 (green) and Nestin (red). Scale bar, 20 μm. (D,E) Quantification of Nestin + /Sox2 T118 + cells following IGF-1 stimulation of WT HNPCs with or without Akt inhibition (1-4 uM Trib.) (*p < 0.05, one-way ANOVA) or IGF-1 stimulation of RIT1 −/− HNPCs with or without RIT1 complementation (*p < 0.05, nonparametric one tailed t-test) (≥200 cells counted from 5-6 fields). (F) Representative confocal images of HNPCs co-immunostained for Tuj1 + (green) and DAPI (nuclei, blue) following IGF-1 (50 ng/ml) stimulation with or without Akt inhibition (1-4 uM Trib.). (G) Quantification of Tuj1 + neurons in WT HNPCs following IGF-1 stimulation (50 ng/ml) ± Triciribine (3 μM) (≥500 cells from 5-6 fields) (*p < 0.05, nonparametric t-test). Data are presented as mean ± SEM calculated from three separate experiments. (H) Schematic diagram of the putative IGF-1/RIT1/Akt/Sox2 signal transduction cascade. The sites of pharmacological inhibition (⊥) and the target of shRNA silencing reagents (•) are indicated. saline followed by 10% buffered formalin, and decapitated. After 24 h of post-fixation in 10% buffered formalin, brains were removed from the skull, post-fixed for an additional 24 h, cryoprotected in 30% sucrose, and quickly frozen in isopentane. Serial coronal 40 µm sections were cut using a freezing sliding microtome (Dolby-Jamison). Every tenth section (400 µm intervals between sections) was selected as a set for further analysis. All experimental procedures were approved by the University of Kentucky Institutional Animal Care and Use Committee in accordance with guidelines established by the National Institutes of Health in the Guide for the Care and Use of Laboratory Animals. Animals were housed at up to 5 mice per cage in the University of Kentucky Medical Center vivarium with a 14:10-hour light/dark photoperiod and were provided food and water ad libitum.
Hippocampal neuronal stem cell (HNPC) cultures. Primary HNPC cultures were prepared as describe 39 . HNPCs were isolated from wild-type and RIT1 −/− mice as described 83 . Briefly, mice were euthanized, the brain dissected and placed in immersion buffer (HBSS (1x) with no Ca 2+ or Mg 2+ containing 1x antibiotic solution (Gibco)). Using a stereomicroscope, dentate gyrus (DG) from hippocampi were dissected and placed in ice-cold immersion buffer. DG (4-5/genotype) were washed with HBSS (1x) containing antibiotic, incubated at 37 °C for 30-45 min with frequent shaking in enzymatic digestion solution (0.25% trypsin in 1 × HBSS with activated papain), trypsin activity was quenched by repeated washing with DMEM (5-10 ml), and placed in 37 °C culture medium (containing DMEM/F12 (1:1), supplemented with 0.3% B27 without insulin, 20 ng/ml of EGF and 10 ng/ml bFGF, and antibiotics). HNPCs were released by trituration (3-4 times) into single cells using fire polished Pasteur pipettes. Approximately, 50 × 10 4 cells were plated in 12 well plates for suspension culture. Neurospheres were evident by day 3. For passage, neurospheres were pooled and mechanically dissociated into single cells and seeded into suspension in growth media in presence of EGF and bFGF (see above). For immunocytochemistry and immunoblotting, single cell suspensions derived from neurospheres were plated on poly D-Lysine coated coverslips or 6 well plates. HNPCs used in this study were Nestin + (HNPC lineage) and >2 passages which promotes homogeneity in the cell population.
RNAi-mediated silencing in HNPCs. Lentiviral vector pZIP-mCMV containing the RIT1 pri-shRNA sequence ( TG CT GTTG A CA GTGAGCGACACGAAGTTCGGGAGTTTAAATAGTGAAGCCACAGATGTA TTTAAACTCCCGAACTTCGTGGTGCCTACTGCCTCGGA) was purchased from transOMIC Technologies (Huntsville, AL). Lentivirus was generated in 293LTV cells using the packaging vectors PsPAX2 and pMD2.G (Univ. Kentucky Genetic Technology Core). The efficiency of RNAi silencing in HNPCs was determined to be >70% using RT-PCR and confocal microscopy. Briefly, HNPCs after passage were allowed to assume normal morphology for 24-48 h before RNAi silencing. Growth medium was removed and stored as per our previously described method 84 , 1 µl of polybrene (Santa Cruz Biotechnology) was added in 1x HBSS for 10 min at 37 °C followed by repeated washes in 1x DPBS. Growth medium was premixed with 3 MOIs of either non-targeting control or RIT1-RNAi expressing lentivirus and placed on cells for 4-12 h at 37 °C. Cells were allowed to recover in fresh growth medium for 48-72 h to permit RIT1 silencing.
Cell transfection and neuronal differentiation. For HNPC transfection, fresh neurospheres were dissociated and approximately 10 5 cells were plated on poly-D-lysine coated coverslips/6 well plates. Cells were transfected 72 h after plating, using Jetprime transection reagent 85 according to the manufacturer's protocol, and allowed to recover cells 48 h before analysis. For IGF-1 stimulation of HNPCs, DMEM was supplemented with F12 (1:1) and FGF2 (10 ng/ml). For HNPC neuronal differentiation analysis, cells were placed in Neurobasal medium containing 1% FBS and RA (1 µM) for 4-6 days.
Image acquisition and analysis. Analysis was performed as previously described 86 . Briefly, isotype control slides were mounted to nullify nonspecific signals from either the Nikon C2 confocal or Nikon A2 confocal prior to mounting non-primary antibody treated slides and channels were adjusted to the individual secondary fluorophore. The numerical aperture and intensity values were optimized to minimize oversaturation and the channel data, pin hole details were recorded, and images recorded at 1024 pixels and 16x speed (Nikon C2 Confocal). Groups of images (3-6 images per specimen) were analyzed using ImageJ. For example, alternate color masks were applied to determine double positive (Nestin + /Ki67 + ) or single positive (Nestin + only) for ≥200 nuclei from 5-6 random fields (for normalization), and data converted to relative fold changes. A similar approach was used to determine neuronal differentiation and the number Sox2 + and Sox T118 + HNPCs. For Akt activation following IGF-1 stimulation, the intensity profile of ≥50 cells from 15 random fields were normalized to DAPI using ImageJ. The mean was determined from 5-8 images per group. Quantification of immature (DCX + /BrdU + ) and mature neuron analysis (NeuN + /BrdU + ) was performed as previously described 36 . Briefly, to quantify the fluorescently labeled cells in the dentate gyrus, three sections (pre-epicenter, epicenter, post-epicenter, 400 μm intervals) were counted using a Nikon Eclipse E600 fluorescence microscope (40× objective). The focal plane was moved throughout the z-axis to capture each positive cell. To estimate the volume of the dentate gyrus, images were collected using a Zeiss Axiovert 200 M fluorescence microscope (10× objective). The volume of the dentate gyrus was then calculated by multiplying the area of the dentate gyrus measured using ImageJ (NIH) by the thickness of each section (40 μm). Cell density was obtained by dividing total cell counts by the total volume of the dentate gyrus for the three sections that were counted. Western blotting. Tissues/cells were homogenized using a Next Advance Bullet Blender at 4 °C in lysis buffer (20 mM Tris·HCl pH 7.5, 250 mM NaCl, 10 mM MgCl 2 , 1% Triton X-100, 1 mM Na 3 VO 4 , 50 mM β-glycerophosphate, 1x protease inhibitor cocktail). Whole cell lysates were resolved on SDS-PAGE gels, transferred to nitrocellulose membranes (12 h, 0.08 mA), and protein abundance and phosphorylation determined by immunoblotting with the appropriate phospho-specific antibodies and band intensity quantified using a ChemiDoc MP with Image Lab software (Bio-Rad) 35, 36 . Luciferase gene reporter assay. Cignal Sox2 luciferase reporter was obtained from Qiagen (CCS-0038L).
HNPCs were transfected with reporter and transduced with RNAi (Control RNAi or RIT1 RNAi), while RIT1 −/− HNPCs were co-transfected with the reporter and Myc-RIT1 (as indicated) or empty vector, and stimulated with/ without IGF-1 (50 ng/ml, 24 h) in presence of FGF2 (10 ng/mL). Cells were washed with PBS and luciferase activity determined using a firefly luciferase assay kit (Promega) after passive lysis as described 39 .
Statistical analysis. The data is presented as Mean ± SEM. Statistical analysis was carried out by either one way/two way ANNOVA combined with post hoc analysis using Tukey Kramer multiple comparisons or nonparametric unpaired one tailed t-test. Significance reported in this manuscript is p < 0.05. Any changes with p > 0.05 were considered to be insignificant.