Neuritin promotes neurite and spine growth in rat cerebellar granule cells via L‐type calcium channel‐mediated calcium influx

Abstract Neuritin is a neurotrophic factor that is activated by neural activity and neurotrophins. Its major function is to promote neurite growth and branching; however, the underlying mechanisms are not fully understood. To address this issue, this study investigated the effects of neuritin on neurite and spine growth and intracellular Ca2+ concentration in rat cerebellar granule neurons (CGNs). Incubation of CGNs for 24 h with neuritin increased neurite length and spine density; this effect was mimicked by insulin and abolished by inhibiting insulin receptor (IR) or mitogen‐activated protein kinase kinase/extracellular signal‐regulated kinase (ERK) activity. Calcium imaging and western blot analysis revealed that neuritin enhanced the increase in intracellular Ca2+ level induced by high K+, and stimulated the cell surface expression of CaV1.2 and CaV1.3 α subunits of the L‐type calcium channel, which was suppressed by inhibition of IR or mitogen‐activated protein kinase kinase/ERK. Treatment with inhibitors of L‐type calcium channels, calmodulin, and calcineurin (CaN) abrogated the effects of neuritin on neurite length and spine density. A similar result was obtained by silencing nuclear factor of activated T cells c4, which is known to be activated by neuritin in CGNs. These results indicate that IR and ERK signaling as well as the Ca2+/CaN/nuclear factor of activated T cells c4 axis mediate the effects of neuritin on neurite and spine growth in CGNs. Open Practices Open Science: This manuscript was awarded with the Open Materials Badge. For more information see: https://cos.io/our-services/open-science-badges/ Cover Image for this issue: doi: 10.1111/jnc.14195.

Neuritin is a neurotrophic factor activated by neural activity and neurotrophins (Nedivi et al. 1993;Naeve et al. 1997). Neuritin has multiple roles in neural development, including synaptic plasticity and maturation and neurite outgrowth (Naeve et al. 1997;Son et al. 2012). Neuritin is also involved in neuronal apoptosis and regeneration after ischemia and traumatic brain injury (Rickhag et al. 2007;He et al. 2013;Yan et al. 2016). Although it is known to exert non-cell-autonomous effects by binding to receptors (Nedivi et al. 1998;Cantallops et al. 2000), a specific neuritin receptor has yet to be identified, and the downstream effectors are unknown. We previously reported that neuritin stimulated the expression of transient potassium (K + ) outward current (I A ) channel K V 4.2 subunits in rat cerebellar granule neurons (CGNs) via insulin receptor (IR) activation and mitogen-activated protein kinase kinase (MEK)/extracellular signal-regulated kinase (ERK) and Akt/mammalian target of rapamycin signaling pathways (Yao et al. 2012), and that activation of the Ca 2+ /CaN/nuclear factor of activated T cells (NFAT)c4 axis was required for this effect (Yao et al. 2016). We also showed that activation of IR and MEK/ERK signaling by neuritin enhanced synaptic transmission in the medial prefrontal cortex (Lu et al. 2017).
Neuritin was first described as a factor that promotes neuritogenesis (Naeve et al. 1997). Treatment of primary rat hippocampal and cortical neuron cultures with recombinant neuritin for 4 days induced extensive neuritogenesis (Naeve et al. 1997). Similar effects have been reported in PC12 cells and motor neurons (Marron et al. 2005;Cappelletti et al. 2007) as well as in hippocampal neurons in depression models, in which neuritin over-expression in the hippocampus prevented the atrophy of dendrites and dendritic spines under conditions of chronic stress (Son et al. 2012). We recently showed that neuritin over-expression increased dendritic spine density in mouse hippocampal neurons and reversed deficits in novel object associative recognition memory in mice caused by exposure to extremely low frequency (50 Hz) electromagnetic fields (Zhao et al. 2015). However, the mechanistic basis for the effects of neuritin has yet to be established.
CGNs are the most abundant neurons in the brain (Herculano-Houzel 2010). Cultured CGNs typically have a round soma and extensive network of neurites (Powell et al. 1997), and have been widely used to study the mechanisms of neuronal survival, programmed cell death (Borodinsky et al. 2002;Zhou et al. 2012), and neurite and synapse development (Borodinsky et al. 2002;Sato et al. 2005). The latter was shown to be modulated by high K + (HK) depolarization-induced Ca 2+ signaling (Borodinsky et al. 2002;Sato et al. 2005), while voltage-gated calcium channel (VGCC) activation induced neurite growth, which was mediated by Ca 2+ /calmodulin (CaM)-dependent protein kinase II in immature CGNs (Borodinsky et al. 2002;Gaudilliere et al. 2004). We also observed that neuritin increased intracellular Ca 2+ concentration ([Ca 2+ ] i ) and HKinduced Ca 2+ influx in mouse CGNs and activated the Ca 2+ / CaN/NFATc4 axis, leading to up-regulation of K V 4.2 protein (Yao et al. 2016). Whether neuritin promotes neurite and spine growth in CGNs by increasing [Ca 2+ ] i and whether the CaN/NFATc4 axis is involved remain to be determined.
In this study, we used rat CGNs to evaluate the effects of neuritin on neurite and spine growth and [Ca 2+ ] i . We also examined whether IR and ERK signaling pathways and the Ca 2+ /CaN/NFATc4 axis mediate the effects of neuritin.

Cell culture
The procedure was approved by the Committee on the Ethics of Animal Experimentation at Fudan University (Permit No. 20090614-001). CGNs were obtained from the cerebella of 7-dayold Sprague-Dawley female rat pups (average weight 180 g, SLAC Laboratory Animal Co., Shanghai, China) as following steps. Briefly, the meninges and vessels of the cerebellum were removed in ice-cold Kreb-Ringer solution (120 mM NaCl, 25 mM NaHCO 3 , 14.3 mM glucose, 4.8 mM KCl, 2.5 mM MgSO 4 , 1.2 mM KH 2 PO 4 , and 0.3% w/v bovine serum albumin) after decapitation. The tissues were minced with scissors, resuspended in the same solution and centrifuged at 2000 rpm for 1 min. The supernatant was discarded and the pellet was resuspended in Kreb-Ringer solution containing 0.025% w/v trypsin (cat. no. T8003; Sigma-Aldrich) and digested at 25°C for 6 min. The digestion was terminated by adding 0.1% w/v trypsin inhibitor (cat. no. T9128; Sigma-Aldrich), 0.008% w/v DNAse I (cat. no. 10104159001; Roche, Milano, Italy, RRID:SCR_008987), and 3.2 mM MgSO 4 followed by centrifugation (900 g for 1 min). The sediment was triturated until there were no visible clusters of tissue. Kreb-Ringer solution containing 2.5 mM MgSO 4 and 0.1 mM CaCl 2 was then added to the cell suspension, followed by centrifugation at 400 g for 3 min. The sediment was resuspended in fresh culture medium consisting of Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 25 mM KCl, and 1% antibioticantimycotic solution and seeded in 35-mm Petri dishes (cat. no. 010035; Shenyou, Shanghai, China) coated with poly-L-lysine (10 g/mL) or on coverslips (cat. no. 12-545-100; Fisher brand, Leicestershire, UK) coated with poly-L-lysine (50 g/mL) at a density of 10 5 cells/cm 2 . Cells were cultured on coverslips dotted with paraffin at a low density (50-130 cells/mm 2 ) for morphological analysis (Kubota et al. 2013). At 6 h after plating, the coverslips were turned over onto cells cultured at a normal density (Kubota et al. 2013) followed by incubation at 37°C and 5% CO 2 in the culture medium. Cytosine 1-beta-D-arabinofuranoside (5 lM) was added to the culture medium 24 h after seeding to inhibit the proliferation of non-neuronal cells. Cells were used for experiments after culturing for 3-5 days unless otherwise indicated.
Oberkochen, Germany) had no information about sample treatments. The images were analyzed by another investigator who was also blinded to the group assignment. A homogenous population of neurons was selected for neurite and spine measurements based on the following criteria: (i) the cell body and neurites were completely impregnated with dye; (ii) selected neurons were separate from surrounding neurons; and (iii) all of the neurites were visible within the plane of focus. Neurite lengths and branches of selected cells were reconstructed using NIH ImageJ software (RRID: SCR_003070) (Cho et al. 2017). Sholl analysis was carried out to quantify the number of branch intersections (Yao et al. 2016). Spine density is expressed as the average number of spines per 100 lm of neurite. Spines with a stubby or mushroom shape were regarded as mature, while those with a thin morphology were considered immature (Wei et al. 2011). A mushroom spine was defined as a neurite protrusion with a head size 50% larger than the neck (Wei et al. 2011).

Measurement of [Ca 2+ ] i levels in CGNs
CGNs grown on coverslips were loaded with Fura-2 acetoxymethyl (AM) ester (cat. no. F1201; Invitrogen) and 0.02% pluronic F127 (cat. no. P3000MP; Invitrogen) in Hanks' balanced salt solution (HBSS) composed of 126.0 mM NaCl, 2.5 mM KCl, 2.0 mM MgSO 4 , 2.0 mM CaCl 2 , 10.0 mM D-glucose, and 10.0 mM HEPES (pH 7.4) at 37°C for 45 min in the dark (Yao et al. 2016). The cells were rinsed three times in Fura-2 AM-free HBSS at 25°C. The coverslips were transferred to an open slide chamber containing 1 mL of Fura-2 AM-free HBSS that was placed on an inverted epifluorescence microscope (Eclipse Ti; Nikon, Tokyo, Japan). Excitation and emission wavelengths for Fura-2 AM were 340/380 and 505 nm, respectively. Baseline [Ca 2+ ] i was determined for 60 s prior to adding the HK solution (40 mM KCl) (Chen et al. 2005). Data were acquired at 4-s intervals throughout the experiment. Fluorescence intensity was quantified using MetaFluor software (RRID:SCR_014294). Calibration for calcium imaging was performed in vitro using a calcium calibration buffer kit (cat. no. F6774; Invitrogen) (Velazquez-Marrero et al. 2014

Patch clamp recording
Whole-cell currents of CGNs were recorded at 25°C using a standard patch-clamping technique . Prior to calcium current (I Ca ) recording, the culture medium was replaced with a solution composed of 147 mM tetraethylammonium chloride, 10 mM BaCl 2 , 10 mM HEPES (pH 7.4), 2 mM MgCl 2 , 1 lM tetrodotoxin, 2 mM 4-AP, and 10 mM glucose. Soft glass recording pipettes were filled with a solution composed of 145 mM CsCl, 10 mM EGTA, 10 mM HEPES (pH 7.3), 5 mM Na 2 -ATP, and 0.5 mM Na 2 -GTP. The resistance of the pipette filled with internal solution was 4-6 MΩ. Current responses were low-pass filtered at 1 kHz. The P/4 protocol was used to subtract the linear leak and capacitative currents, which were initiated from a holding potential of À80 mV.

Lentiviral transfection
A lentivirus encoding short interfering RNA targeting NFATc4 (siNFATc4) with the sequence 564-gggacggctctcctagagatt-584 (Benedito et al. 2005;Vashishta et al. 2009) was used to knock down NFATc4 expression. The lentivirus expressing scramble sequence 5'-ttctccgaacgtgtcacgt-3' was used as control. The lentivirus was generated by Genechem (Shanghai, China), with a titer of 10 9 /mL and multiplicity of infection of 100. CGNs were transfected with lentivirus for 3 days and then lysed for western blotting or stained by DiD solution.
Data acquisition Origin 8.0 (RRID:SCR_014212) was used to analyze the data. The size of the samples was determined by power analysis. The results of Kolmogorov-Smirnov test indicated that all samples except for the neurite branching result were normally distributed. The Student's t-test was used to compare two samples and oneway analysis of variance (ANOVA) with Bonferroni's post hoc test was used for comparisons between multiple groups. Neurite branching was analyzed by Friedman test, with drug treatment and distance from the soma as two factors. Data out of the range of mean AE 3 SD were regarded as outliers and discarded. Data are presented as mean AE SEM, with n representing the number of neurons used for morphological analysis, calcium imaging, or electrophysiological recordings. Western blotting data are presented as individual data points in a scatter plot, with n representing the number of experiments. For electrophysiology, data were collected from at least four different batches of neurons prepared on different days to minimize bias resulting from culture conditions. p ≤ 0.05 was considered statistically significant in all statistical tests.

Neuritin increases neurite length and spine density by activating IR and MEK/ERK signaling pathways in CGNs
Neuritin has been shown to promote the growth of neurites in hippocampal and cortical neurons (Naeve et al. 1997). We therefore investigated whether neuritin promotes the growth of neurites in CGNs by evaluating neurite length and number of branches. Incubation of CGNs with 150 ng/mL neuritin increased neurite length by 32.2 AE 6.1% after 12 h (n = 43 and 48, p < 0.001) and by 46.6 AE 8.4% after 24 h (n = 41 and 52, p < 0.001) (. 1a and b). Similar effects were obtained by treatment with 150 ng/mL insulin, with increases in neurite length of 39.3 AE 6.0% and 53.6 AE 9.7% observed after 12 and 24 h, respectively (n = 35 and 33, p < 0.001) ( Fig. 1a and b). Sholl analysis and Friedman test revealed that neither neuritin nor insulin affected neurite branching after 12 or 24 h and the interactions between distance from the soma and the drug treatment were not significantly different (Fig. 1c).
As in the case of hippocampal and cortical neurons, the spines of CGN neurites can be classified as thin, stubby, and mushroom-shaped according to the ratio of spine head to spine neck width (Wei et al. 2011). We evaluated the effects of neuritin on the three types of spine and found that it had no effect on thin spines after 12 or 24 h (Fig. 2c). However, the density of stubby spines was increased by 77.8 AE 9.5 % (n = 33 and 39, p < 0.001) and 46.1 AE 8.2% (n = 37 and 40, p < 0.001) after 12 and 24 h, respectively (Fig. 2c). Although neuritin treatment did not significantly affect the density of mushroom spines after 12 h, an increase of 133.0 AE 18.1% was observed after 24 h (n = 36 and 39, p < 0.001) (Fig. 2c). Insulin treatment for 12 h increased the density of stubby spines by 55.6 AE 8.2% (n = 33 and 33, p < 0.001) (Fig. 2c); however, after 24 h, the density of stubby and mushroom spines was increased by 51.9 AE 7.7% (n = 37 and 41, p < 0.001) and 153.1 AE 15.5% (n = 36 and 40, p < 0.001), respectively (Fig. 2c).
We previously reported that neuritin-activated IR and MEK/ERK signaling leads to up-regulation of K V 4.2 subunits in CGNs and enhanced synaptic transmission in the medial prefrontal cortex (Yao et al. 2012;Lu et al. 2017). To determine whether neuritin-induced increases in neurite length and spine density in CGNs involve IR and MEK/ERK signaling pathways, we used HNMPA and U0126 to block IR and ERK activity, respectively. Treatment with 100 lM HNMPA (Yao et al. 2016), neuritin did not significantly increase neurite length (1.0 AE 5.7%, n = 41 and 43) or spine density (2.4 AE 4.6%, n = 31 and 26) relative to the control (p > 0.05) ( Fig. 3a and b). On the other hand, 10 lM U0126 (Lu et al. 2017) abrogated the neuritin-induced increases in neurite length and total spine density to 3.7 AE 6.2% (n = 38 and 43) and 1.5 AE 5.4% (n = 22 and 21) of the control values, respectively ( Fig. 3a and b).
Ca 2+ influx after depolarization occurs through VGCCs (Kirischuk et al. 1996;Toescu 1999). We recorded the VGCC current (I Ca ) by patch-clamping 5 DIC CGNs to determine whether VGCCs were activated by neuritin. Culture medium was replaced by recording bath solution with 150 ng/mL neuritin also for 20 min. I Ca was elicited by depolarization from a holding potential of À80 mV to 10 mV for 250 ms (Fig. 5a). Neuritin increased I Ca amplitude in CGNs in a time-dependent manner (Fig. 5b) by 39.7 AE 5.7% (n = 48 and 31, p < 0.001) and 43.5 AE 6.3% (n = 48 and 37, p < 0.001) after 3 and 5 DIC, respectively, but not after 7 DIC (6.3 AE 4.2%, n = 36 and 21, p > 0.05). The current-voltage (I-V) curves obtained by delivering 250ms depolarizing pulses from a holding potential of À80 mV to between À60 and 40 mV in 10-mV steps at 10-s intervals (Fig. 5c) revealed that I Ca was increased from a negative potential of À60 mV to a maximum value of 10 mV (Fig. 5d). We then evaluated steady-state I Ca activation by measuring membrane conductance of Ca 2+ (g Ca ) which was determined with the equation g Ca = I Ca /(V m À V rev ), where V m is the membrane potential and V rev is the reversal potential for Ca 2+ . We then plotted g Ca /g Ca-max against membrane potential and fitted the line with a sigmoidal function using the Boltzmann model-that is, g Ca/ g Ca-max = 1/ {1 + exp [(V m1/2 À V m )/k]}. The half-activation potential (the voltage at which the current amplitude is half-activated) in the absence or presence of neuritin obtained from the g-V curve was À7.8 AE 1.0 mV (n = 21) and À5.8 AE 0.7 mV (n = 28), respectively (p > 0.05, Fig. 5e). We also compared membrane capacitance between neuritin treatment (6.9 AE 0.3 pF, n = 48) and control (6.6 AE 0.3 pF, n = 37) groups and found no significant difference (p > 0.05, Fig. 5f). Taken together, these results indicate that neuritin increases I Ca amplitude without altering the voltage-gating properties of I Ca channels or membrane capacitance.
Our previous study showed that L-type Ca 2+ channels are the main VGCCs up-regulated by neurotrophic factors . To determine whether L-type Ca 2+ channels are responsible for the increases in Ca 2+ influx and I Ca amplitude induced by neuritin, CGNs were treated with the selective blocker nifedipine. Pre-incubation or perfusion of CGNs with nifedipine inhibited the neuritin-induced increases in HKevoked [Ca 2+ ] i and I Ca amplitude, respectively (Fig. 6a-c).
Consistent with the results from Ca 2+ imaging, nifedipine perfusion for 5 min reduced I Ca amplitude of the control and neuritin groups by 49.6 AE 6.0% and 55.2 AE 7.6% (Fig. 6c), respectively, suggesting that L-type Ca 2+ channels mediate the neuritin-induced increase in I Ca amplitude and [Ca 2+ ] i . Moreover, blocking IR and ERK with 100 lM HNMPA and 10 lM U0126, respectively, abrogated the increase in Ca 2+ influx and I Ca amplitude induced by neuritin ( Fig. 7a-d).
We examined whether the neuritin-mediated increase in I Ca is because of an increase in channel expression at the cell membrane. Ca V 1.2 and Ca V 1.3 are the major a subunits of Ltype Ca 2+ channels in the central nervous system (Forti and Pietrobon 1993;Koschak et al. 2007). We used biotin labeling to quantify the expression of Ca V 1.2 and Ca V 1.3 proteins on the membrane of CGNs following neuritin treatment. Ca V 1.2 and Ca V 1.3 levels were increased by 59.9 AE 12.4% (n = 6, p < 0.05) and 74.8 AE 11.8% (n = 6, p < 0.05), respectively, in the presence of neuritin, as determined by western blotting (Fig. 8a). Consistent with the observed effects of neuritin on HK depolarizationinduced Ca 2+ influx and I Ca amplitude, the cell membrane expression of Ca V 1.2 and Ca V 1.3 was blocked by HNMPA and U0126 (Fig. 8b).
Ca 2+ /CaM/CaN/NFATc4 signaling mediates increases in neurite length and spine density induced by neuritin Ca 2+ and CaM play critical roles in regulating the growth and maturation of neurites and their spines (Borodinsky et al. 2002;Gaudilliere et al. 2004). We investigated whether Ca 2+ and CaM are involved in the neuritin-induced increases in neurite length and spine density. We found that these effects were abrogated by blocking L-type channels with nifedipine and by inhibiting CaM and CaN by treatment with CMZ and CsA, respectively ( Fig. 9a and b). Application of 10 lM nifedipine, 2 lM CMZ, or 5 lM CsA (Yao et al. 2016) could eliminate the neuritin-induced increases in neurite length and spine density ( Fig. 9a and b).
NFATc family transcription factors (NFATc1-4) are downstream effectors in the Ca 2+ /CaN pathway (Sato et al. 2005;Vashishta et al. 2009). Our previous work showed that neuritin can activate this pathway and induce the expression of K + channels, thereby increasing neuronal excitability (Yao et al. 2016). In this study, we used a lentivirus expressing siNFATc4 to silence NFATc4 expression and determine its role in the stimulation of neuritogenesis in CGNs. The high transfection efficiency of lentivirus expressing green fluorescent protein along with a scrambled sequence or siNFATc4 was confirmed by fluorescence microscopy (Fig. 10a). Treatment of CGNs with the siNFATc4 lentivirus for 3 days reduced NFATc4 expression by 71.4 AE 8.5% relative to cells infected with the lentivirus expressing the scrambled sequence (p < 0.001, n = 3) (Fig. 10b). siNFATc4 treatment alone did not affect neurite length, but suppressed the neuritin-induced increase in neurite length, which did not differ between the siNFATc4 only and siNFATc4 plus neuritin groups (n = 22 and 17, p > 0.05) but differed between the scrambled sequence plus neuritin and the siNFATc4 plus neuritin groups (n = 20 and 17, p < 0.001) (Fig. 10c). The neuritin-induced increase in total spine density was also suppressed by siNFATc4 treatment, with no difference observed between the siNFATc4 and siNFATc4 plus neuritin groups (n = 19 and 19, p > 0.05), although the scrambled sequence plus neuritin and siNFATc4 plus neuritin groups differed significantly (n = 19 and 19, p < 0.001) (Fig. 10d).

Discussion
We previously reported that neuritin increased I A in rat CGNs by increasing the expression of K V 4.2 and enhanced glutamate release and miniature excitatory post-synaptic current frequency in mouse cortical neurons via upregulation of the Ca V 3.3 subunit of T-type VGCCs at the membrane surface, with activation of IR and ERK signaling mediating these effects (Yao et al. 2012;Lu et al. 2017). In this study, we found that neuritin increased neurite length and spine density in CGNs via the same receptors and signaling pathways through up-regulation of the Ca V 1.2 and Ca V 1.3 subunits of L-type VGCCs at the membrane surface, which enhanced Ca 2+ /CaN/NFATc4 signaling.
Although previous studies have suggested that neuritin functions as a neurotrophic factor to stimulate neurite growth (Naeve et al. 1997;Marron et al. 2005;Cappelletti et al. 2007), there is little information on how it induces the formation of neurites and spines in CGNs. We confirmed that neuritin has this effect in CGNs, which is consistent with what is known of CGN development (Powell et al. 1997). We noted that neuritin exerted variable effects on spine density depending on spine morphology and primarily increased the number of mushroom spines, which are mature and mediate more robust synaptic transmission as the socalled memory spines (Bourne and Harris 2007). Thus, neuritin not only increases spine density but may also promote spine maturation in CGNs. It has been proposed that thin and mushroom spines are interconvertible and that longlasting enhancement of synaptic transmission enlarges the head size of thin spines, transforming them into stable mushroom (Matsuzaki et al. 2004). Given our observation that neuritin increased the total number of spines without reducing thin spine density, it is likely that it acts not by stimulating the interconversion of thin and mushroom spines, but by promoting mushroom spine formation, leading to enhanced synaptic plasticity.
An elevation in [Ca 2+ ] i is critical for activation of Ca 2+ / CaM-dependent neurite growth and synapse maturation in hippocampal neurons and CGNs (West et al. 2002). [Ca 2+ ] i  is increased via Ca 2+ influx through VGCCs that are opened by HK stimulation (Kirischuk et al. 1996;Toescu 1999) (Borodinsky et al. 2002;Sato et al. 2005), and is consistent with our previous study showing that neurotropic factors or hormones such as growth/differentiation factor-15 and melatonin increased HK-induced Ca 2+ influx but not basal [Ca 2+ ] i in rat CGNs Lu et al. 2016). However, it also contradicts our previous finding in mouse CGNs, in which neuritin not only enhanced HK-induced increases in [Ca 2+ ] i but also caused a slight increase in basal [Ca 2+ ] i (Yao et al. 2016). Additional studies are needed to determine whether there are species-specific differences in the mechanisms of calcium release from intracellular stores in CGNs.
Various types of VGCC are expressed in immature and mature CGNs (Koschak et al. 2007). L-type VGCCs are responsible for CaN-mediated synaptogenesis (Sato et al. 2005), whereas N-type channels are critical for CGN migration (Dangelo et al. 1994). The results presented here demonstrate that [Ca 2+ ] i following membrane depolarization and Nif-sensitive I Ca were increased by neuritin, suggesting the activation of L-type VGCCs, although we cannot exclude the possibility that nifedipine-insensitive Ca 2+ channels or N-, P/Q-, R-type Ca 2+ channels are also involved. We found that I Ca amplitude and HK-evoked [Ca 2+ ] i in CGNs increased with developmental stage, which is consistent with observations in cerebellar slice cultures (Dangelo et al. 1994). That is, the neuritin-induced increase in I Ca amplitude and HK depolarization and resultant Ca 2+ influx only occurred in CGNs after 3 and 5 DIC. This may be explained by the fact that CGN Ca 2+ channels are more sensitive to neuritin at early developmental stages, and is consistent with the function of neuritin in promoting spine development and maturation in CGNs.
L-type Ca channels include Ca V 1.1-1.4 a subunits (Koschak et al. 2007). Ca V 1.1 and 1.4 are mainly expressed in skeletal muscle and retinal cells, whereas Ca V 1.2 and Ca V 1.3 are abundant in the brain where they regulate neuronal excitability, synaptic plasticity, and activity-dependent gene transcription (Forti and Pietrobon 1993;Koschak et al. 2007). Although Ca V 1.2 and Ca V 1.3 a subunits differ in their biophysical properties, brain distribution, and function (Tuckwell 2012;Berger and Bartsch 2014), our western blot analysis showed that Ca V 1.2 and Ca V 1.3 a subunits were similarly up-regulated by neuritin, which is consistent with a previous finding that functionally distinct L-type Ca 2+ channels coexist in rat CGNs (Forti and Pietrobon 1993). Increased surface expression of Ca V 1.2 and Ca V 1.3 a subunits can also explain why neuritin treatment increased I Ca amplitude but did not affect L-type VGCC gating properties. Similar to our earlier studies examining the effect of neuritin on K V 4.2 expression in rat CGNs and T-type Ca V 3.3 membrane expression in mouse cortical neurons (Yao et al. 2012;Lu et al. 2017), neuritin-induced effects on spine density, I Ca amplitude, and Ca 2+ influx elicited by HK were mimicked by insulin and abolished by inhibition of IR and ERK signaling, suggesting that these pathways mediate the activities of neuritin. We also found that neuritin-induced increases in Ca V 1.2 and Ca V 1.3 expression were dependent on IR and ERK signaling. ERK can directly phosphorylate ion channel subunits and may alter the gating properties of K + channels in the same manner that neurotransmitter regulates I A (Hu et al. 2007). Since neuritin did not alter the gating properties of Ca 2+ channels but instead increased Ca V 1.2 and Ca V 1.3 expression on the cell membrane, we propose that activation of ERK signaling by neuritin enhances Ca V 1.2 and Ca V 1.3 trafficking to the membrane in the same way that it increases cell surface T-type VGCC Ca V 3.3 expression via activation of ERK signaling in mouse cortical neurons (Lu et al. 2017). It is likely that neuritininduced activation of IR and ERK signaling caused the upregulation of the Ca V 1.2 and Ca V 1.3 subunits of L-type VGCCs on the cell membrane, increased Ca 2+ influx, and activated the Ca 2+ /CaN/NFATc4 pathway (Fig 11).
NFATc1-4 are key downstream effectors of CaN signaling (Sato et al. 2005;Vashishta et al. 2009). NFATc4 plays an important role in neural development, axon elongation, and neuronal survival (Benedito et al. 2005;Sato et al. 2005). We previously showed that neuritin along with Ca 2+ and CaN activated NFATc4 to increase K V 4.2 promoter activity in CGNs, which enhanced I A and neuronal excitability (Yao (Scramble, n = 20; Scramble + Nrn, n = 20; siNFATc4, n = 22; siNFATc4 + Nrn, n = 17) and total spine density (d) (Scramble, n = 22; Scramble + Nrn, n = 19; siNFATc4, n = 19; siNFATc4 + Nrn, n = 19) in CGNs cultured in the absence or presence of siNFATc4. Results represent mean AE SEM, n indicates the number of experiments in western blot or number of cells in morphological analysis. ***p < 0.001 between two groups connected with a straight line (oneway ANOVA followed by Bonferroni post hoc test). et al. 2016). Here we found that blocking I Ca channels, abolishing CaM or CaN function, or silencing NFATc4 abolished the effects of neuritin on neurite length and spine density, suggesting that these are mediated by Ca 2+ , CaN, and NFATc4. However, since neurite growth and spine maturation induced by Ca 2+ signaling involve a variety of mechanisms such as regulation of cargo loading and stabilization of the cytoskeleton and post-synaptic densities (Ichinose et al. 2015;Ji et al. 2017;Yadav et al. 2017), additional studies are required to identify the genes related to neurite and spine growth that are up-regulated by neuritin/ Ca 2+ /CaN /NFATc4. I A is the predominant K + current in many mature neurons (Kanold and Manis 1999;Tkatch et al. 2000;Plant et al. 2006) that is initially activated at the sub-threshold range of membrane potential and was previously implicated in the control of neuronal excitability (Carrasquillo et al. 2012;Yang et al. 2014). Interestingly, the expression of K V 4.2 increased with the duration of the micro-explant and CGN culture period (Shibata et al. 1999;Yao et al. 2012), and I A was developmentally regulated and associated with the maturation and migration of CGNs (Shibata et al. 2000;Yao et al. 2013). Notably, neuritin induced increases in I A and K V 4.2 expression from day 3-5 of culture (Yao et al. 2012), which corresponded to the neuritin-induced increase in neurite and spine growth. We speculate that the increased expression of K V 4.2 induced by neuritin may not be directly related to its effect on spine growth. Alternately, increased expression of K V 4.2 and spine growth may both require neuritin-induced increases in calcium influx and Ca 2+ /CaN/NFATc4 pathway activation, which may occur simultaneously during enhancement of CGN development and maturation induced by neuritin (Fig. 11).
In conclusion, we demonstrated that neuritin promotes neurite and spine growth via Ca 2+ -mediated activation of CaN and NFATc4. However, while depolarization did not affect the development of immature dendrites, it was shown to prevent terminal maturation of CGNs through activation of CaN in cultured cerebellar slices (Okazawa et al. 2009). Moreover, dendrite maturation was blocked by depolarization and NFATc4 activation in cultured CGNs (Ding et al. 2013). It is possible that the inconsistency between these results and ours is because of the fact that we used 25 mM rather than 5 mM K + medium for the cultured rat CGNs. Further study is needed to clarify the role of K + concentration in the neuritin-induced effects on neurite and spine development in cultured CGNs. study in the manuscript. The complete Open Science Disclosure form for this article can be found at the end of the article. More information about the Open Practices badges can be found at https://cos.io/our-services/openscience-badges/.