ERK signaling mediates CaSR-promoted axon growth

Highlights • We have investigated how CaSR activation enhances sympathetic axon growth.• CaSR activation promotes phosphorylation of ERK1 and ERK2.• Inhibition of ERK1/ERK2 phosphorylation blocks CaSR-promoted axon growth.• CaSR-promoted axon growth requires a discrete region of the cytoplasmic domain.


Introduction
The CaSR plays a crucial role in monitoring and maintaining [Ca 2+ ] o within very narrow physiological limits and is conspicuously expressed in the tissues and organs involved in systemic calcium homeostasis [1]. The CaSR is also widely expressed in the peripheral and central nervous system, where it has been implicated in a diversity of functions [2]. These include regulating axon and dendrite growth [3], the migration and/or maintenance of hypothalamic GnRH neurons [4] and the regulation of neuronal excitability and synaptic transmission [5,6].
The molecular mechanisms by which CaSR exerts its effects on neurons are poorly understood. The CaSR is a member of the C family of G-protein coupled receptors that associates with three main heterotrimeric G protein complexes, G q/11 , G i/o and G 12/13, and thereby modulates the activity of a wide variety of downstream signaling networks, including PLC-mediated Ca 2+ mobilization, cAMP, Rho kinase and the MAP kinases ERK1/2, p38 and JNK [7]. The aim of this study was to ascertain how CaSR activation influences axon growth and branching using the well-characterized, experimentally tractable sympathetic neurons of the mouse superior cervical ganglion (SCG) [8]. Previous work has shown that expression of the CaSR peaks in these neurons in the immediate perinatal period and that activating the CaSR during this stage of development enhances NGF-promoted axon growth, and that this is important for the establishment of the appropriate level of sympathetic innervation in vivo [3]. Our  without the complication of concomitant NGF signaling. We show that CaSR-promoted ERK activation contributes to CaSR-promoted axon growth and identify the region of the CaSR C-terminal domain required for axon growth.

Neuron cultures
Dissociated cultures of SCG neurons from CD-1 mice were grown on poly-ornithine/laminin coated 35 mm tissue culture dishes (Greiner) in Hams F14 medium [9] with 0.25% Albumax I (Invitrogen). Survival was estimated as described [9]. The neurite arbors of non-transfected neurons were labelled with calcein-AM (1:1000, Invitrogen). Neurons transfected with plasmids encoding full-length CaSR or CaSR mutants were co-transfected with a YFP plasmid. Fast-Sholl analysis was carried out on imaged neurons [10].

Plasmids
The pFLCaSR plasmid was generated by cloning the open reading frame of human CaSR into pcDNA3.1. The pG907stopCaSR and pA877stopCaSR plasmids were generated by site directed mutagenesis. Transfection was carried out using the Neon Transfection system (Invitrogen).

CaSR-promoted neurite growth is NGF-dependent
Previous work has shown that activating the CaSR in cultured SCG neurons with elevated levels of [Ca 2+ ] o enhances NGF-promoted axon growth in the immediate perinatal period [3]. To ascertain whether or not CaSR activation is able to enhance neurite growth independently of NGF, we compared neurite growth from E18 SCG neurons cultured with and without NGF in media containing 2.3 mM (maximally-activating) and 0.7 mM (minimallyactivating) levels of [Ca 2+ ] o [3]. Because E18 SCG neurons are dependent on NGF for survival, we added a broad-spectrum caspase inhibitor (Boc-D-FMK) to the medium to prevent apoptosis.
In accordance with published observations [3], the neurite arbors of NGF-supplemented neurons grown in medium containing 2.3 mM [Ca 2+ ] o were much larger and more branched than those of neurons grown with 0.7 mM [Ca 2+ ] o . There were highly significant differences in neurite length (Fig. 1A) and branch point number (Fig. 1B), and the Sholl profiles displayed clear differences in NGF-supplemented cultures (Fig. 1C). In contrast, the size and complexity of the neurite arbors of neurons grown without NGF were not significantly different in media containing 0.7 mM and 2.3 mM [Ca 2+ ] o (Figs. 1A-C). These findings suggest that activation of the endogenous CaSR is insufficient to enhance the low level of neurite growth that occurs in the absence of NGF, but enhances the magnitude of NGF-promoted neurite growth.

Over-expression of CaSR promotes neurite growth in the absence of NGF
The requirement for NGF in CaSR-promoted neurite growth complicates investigation of the signaling pathways downstream of CaSR that mediate this effect. Because enhancement of neurite growth by CaSR activation is only observed in SCG neurons at the developmental peak of CaSR expression [3], we tested whether overexpression of CaSR would enhance neurite growth in the absence of NGF. Robust high-level CaSR expression was achieved by transfecting neurons with a pcDNA3.1 vector containing full-length CaSR (pFLCaSR). Quantification of CaSR immunofluorescence confirmed increased CaSR expression in neurons transfected with pFLCaSR ( Fig. 2A).
In the absence of NGF, the neurite arbors of E18 SCG neurons transfected with pFLCaSR were significantly larger than those of control transfected neurons in medium containing 2.

ERK1/ERK2 activation by CaSR over-expression contributes to neurite growth
To elucidate the molecular mechanism underlying the enhancement of neurite growth from SCG neurons by activated CaSR, we explored a common link in intracellular signaling between CaSR signaling and neurite growth. ERK1 and ERK2 are activated by the CaSR in parathyroid cells, fibroblasts and kidney cell lines [11][12][13][14][15] and NGF-promoted ERK1/ERK2 activation in PC12 cells and SCG neurons contributes to the neurite growth response [16][17][18][19][20].
To investigate whether ERK1/ERK2 signaling contributes to CaSR-promoted neurite growth from E18 SCG neurons, we used immunofluorescence to estimate the relative levels of phospho-ERK1/ERK2 in neurons. Phospho-ERK1/ERK2 immunofluorescence was clearly elevated in NGF-supplemented neurons grown in 2.   ( Fig. 3A). There were no differences in total ERK1/ERK2 immunofluorescence in neurons grown under these different conditions (not shown). This suggests that ERK1/ERK2 signaling is elevated in neurons grown with NGF and activating levels of [Ca 2+ ] o . We subsequently examined whether CaSR overexpression can activate ERK1/ERK2 in E18 SCG neurons grown in NGF-free medium. The level of phospho-ERK1/ ERK2 immunofluorescence was clearly elevated in neurons overexpressing CaSR compared with control transfected neurons (Fig. 3B). Phospho-ERK1/ERK2 immunofluorescence in neurons transfected with pFLCaSR after 24 h in medium containing 2.3 mM [Ca 2+ ] o without NGF revealed a significant elevation compared with control-transfected neurons (Fig. 3C), whereas immunofluorescence for total ERK1/ERK1 was not significantly different (Fig. 3D).
To test whether ERK1/ERK2 activation contributes to the enhanced neurite growth brought about by CaSR over-expression, we examined whether U0126, a selective MEK1/MEK2 inhibitor that interferes with activation of ERK1/ERK2 by MEK1/MEK2 [21], could prevent the increase in CaSR-promoted neurite growth. In these experiments, we plated pFLCaSR-transfected and control-transfected SCG neurons in NGF-free medium containing Boc-D-FMK and either U0126 or its inactive analog U0124. Quantifying neurite arbor size and complexity 24 h later revealed that U0126, but not U0124, completely prevented enhanced neurite growth accompanying CaSR overexpression ( Fig. 3E and F). Neither U0126 nor U0124 affected neuronal survival in these cultures (not shown). This suggests that MEK/ERK signaling contributes to CaSR-promoted neurite growth.

A discrete region of the CaSR cytoplasmic tail is required for enhanced neurite growth
Because many of the signaling functions of the CaSR are dependent on the C-terminal cytoplasmic domain [22,23], we transfected E18 SCG neurons with plasmids that express C-terminal truncation mutants and quantified neurite growth after 24 h culture in NGF-free medium containing 2.3 mM [Ca 2+ ] o . The G907stop CaSR mutant lacks the high-affinity binding site for filamin-A [15] and a PKC phosphorylation site. The A877stop CaSR mutant lacks additional PKC phosphorylation sites, a PKA phosphorylation site, arginine rich motifs and a low-affinity binding site for filamin-A [15,23].
Sholl analysis showed that the neurite arbors of E18 SCG neurons transfected with the pG907stopCaSR expression plasmid were of similar size to those of neurons transfected with pFLCaSR, whereas the neurite arbors of control-transfected neurons were very much smaller (Fig. 4A). There were no significant differences in neurite length and branch point number between pG907stopCaSR-transfected and pFLCaSR-transfected neurons (P = 0.9, ANOVA, data not shown), whereas controltransfected neurons were significantly shorter and less branched (P < 0.0001, Kruskal-Wallis test, data not shown). In contrast, the neurite arbors of pA877stopCaSR-transfected neurons were no larger than those of control-transfected neurons (Fig. 4B), and there were no significant differences in neurite length and branch point number between these groups (P = 0.8, ANOVA). This suggests that a region between residues alanine 877 and glycine 907 is required for CaSR-enhanced neurite growth.

Discussion
Previous work has shown that the CaSR regulates neurite growth and tissue innervation in the developing sympathetic nervous system [3]. In the current study, we have investigated how CaSR activation enhances neurite growth from SCG neurons. Our observation that activating levels of [Ca 2+ ] o enhance axon growth and branching from cultured SCG neurons overexpressing the CaSR in the absence of NGF has enabled us to investigate how CaSR signaling affects neurite growth without the confounding effect of NGF signaling.
We focused on a common link in intracellular signaling between CaSR signaling and the regulation of neurite growth. Several studies have shown that CaSR activation in a variety of cell types, including parathyroid cells, fibroblasts and kidney cell lines, leads to phosphorylation and activation of ERK1 and ERK2 [11][12][13][14][15] and ERK1/ERK2 activation in neurons and neural cell lines by NGF contributes to the neurite growth response [16][17][18][19][20].
Two observations suggest that ERK1/ERK2 activation is a key step by which CaSR activation enhances neurite growth. First, activating but not minimally-activating levels of [Ca 2+ ] o promote phosphorylation of ERK1 and ERK2 in SCG neurons overexpressing the CaSR. Second, pharmacological inhibition of the kinases that lie upstream of ERK1 and ERK2, MEK1 and MEK2, prevents activating levels of [Ca 2+ ] o enhancing axon growth. While these observations do not rule out the possibility that other signaling pathways activated by the CaSR may contribute to its influence on neurite growth, these findings suggest that ERK1/ERK2 activation is a necessary step in the neurite growth-promoting effects of the CaSR.
To determine the region of the CaSR required for enhancing neurite growth, we overexpressed CaSR deletion mutants to ascertain   [13,15,24], we used deletion mutants that lack one or both binding sites for filamin A. Overexpression of the G907stop mutant, which lacks the high-affinity filamin-A binding site [15], was just as effective as overexpressed full-length CaSR in promoting neurite growth in medium containing activating levels of [Ca 2+ ] o , indicating that the high-affinity filamin A binding site is not required for CaSR-promoted neurite growth. It is possible that this partially truncated mutant may enhance neurite growth by a mechanism that differs from that of full-length CaSR. A recent study in HEK-293 cells has also shown that the high-affinity filamin A binding site is also dispensable for ERK1/ERK2 activation by the CaSR [15]. Overexpression of the A877stop CaSR mutant failed to enhance neurite growth with activating levels of [Ca 2+ ] o , suggesting that the region between residues alanine 877 and glycine 907 is required for CaSR-enhanced neurite growth. This region not only possesses a low-affinity filamin A binding site [15], but also contains PKC phosphorylation sites [25][26][27][28] and residues necessary for PI-PLC activation [29]. These findings are consistent with the possibility that the low-affinity filamin A binding site is important, but raise the possibility that other sites within this short stretch of the C-terminal domain contribute to CaSR-promoted neurite growth. Further molecular dissection of this region will be required to address this possibility.