Background
Upon CNS injury, myelin- and astrocyte-associated inhibitors limit re-growth of lesioned nerve fibers [
1]. Myelin-associated inhibitors such as Nogo signal through the Nogo receptor complex (NgR) or PIR-B (paired immunoglobulin-like receptor B) expressed on CNS neurons [
2]. Astrocyte-derived CSPGs activate the transmembrane protein tyrosine phosphatase receptor, RPTPσ, as well as NgRs to prevent axon growth [
2]. Once activated, the intracellular signaling pathways employed by these receptors are not well understood. So far, Rho-GTPase and Rho kinase (ROCK) as well as PKC signaling are known to be recruited by NgR receptors [
3]. Further downstream LIM kinase-cofilin signaling is connecting NgR with the actin cytoskeleton thereby contributing to axon stalling [
4].
Notably many other signaling pathways including cAMP/PKA, PTEN/AKT, mTOR, GSK3β and MAP kinase signaling have been implicated in axon regeneration [
3]. For instance, elevating neuronal cAMP levels emerges as potent mechanism to bypass axon injury [
5]. In addition, MAP kinases such as p38, ERK and JNK are involved in CNS axon regeneration [
6]-[
8]. ERK is involved in peripheral [
9]-[
11] and central [
12]-[
14] axon regeneration. However it has not been investigated in much detail whether these signaling cascades are activated upon NgR, RPTPσ or PIR-B engagement by their ligands. In this study we analyzed whether Rho-GTPase, cAMP/PKA and MAP kinase signaling are mediating signaling upon stimulation of primary CNS neurons with total CNS myelin, purified Nogo and CSPGs.
In addition, we investigated whether these axon regeneration inhibitors modulate gene expression. So far, activation of NgRs, PIR-B or RPTPs has not been connected to modulation of gene expression. We focused on SRF (serum response factor), a gene regulator mediating an immediate early gene (IEG) response of e.g. c-Fos [
15], a hallmark of neuronal activation as well as regulatory switch of neuronal apoptosis vs. survival [
16]. SRF cooperates with TCFs (ternary complex factors) such as Elk-1 to convey IEG induction. Besides IEGs, SRF regulates actin cytoskeletal gene abundance [
17]. SRF has not been studied in CNS axonal regeneration so far. In PNS axon regeneration, using facial nerve regeneration as model system, we demonstrated a stimulatory SRF function in motoneuron survival [
18] and axonal regeneration [
19] involving a cytoplasmic SRF localization [
19].
Here we show that stimulation of CNS neurons with total myelin, Nogo or CSPGs activated SRF-dependent c-Fos reportergene activity. Pharmacological inhibition of MAP kinase, and to la lesser extent Rho-GTPase/ROCK, but not cAMP/PKA signaling prevented SRF gene activity induced by all three inhibitors. MAP kinases (i.e. ERK) were activated upon incubation of neurons with myelin, Nogo or CSPGs. Further downstream of ERK we observed c-Fos induction by myelin, a process blocked by SRF ablation. Finally, we show that SRF is not only a signaling target of axon regeneration inhibitors. Employing constitutively-active SRF-VP16 circumvented neurite growth impaired by myelin, Nogo and CSPGs. This provides first in vitro data unraveling an SRF potential in CNS axon regeneration.
Discussion
In recent years many extrinsic factors posing regeneration obstacles to injured CNS axons were identified [
27]. However how these myelin- or astrocyte-associated molecules transmit their axon growth prohibiting activity within neurons is not understood in much molecular depth. So far Rho-GTPases have been the main signaling intermediate identified in neurons to connect these extrinsic factors with intrinsic signaling programs resulting in axonal stalling [
3].
Here, we employed CNS neurons to analyze which signaling pathways are activated when neurons encounter the regeneration inhibitors total myelin, Nogo or CSPGs. We further investigated whether these regeneration inhibitors modulate gene expression, an event largely neglected in CNS compared to PNS axon regeneration. Our data show that all three regeneration obstacles stimulated SRF-dependent gene expression (Figure
1). Interestingly, both, growth-promoting BDNF as well as growth-inhibiting regeneration inhibitors stimulated SRF, yet with different temporal patterns (Figure
1A). Axonal growth inhibitors activated SRF only at shortest but failed to activate SRF at longer time-points (8 h; Figure
1A). In contrast, BDNF typically provides sustained SRF activation ranging from short time-points [
28] to long-term stimulation (Figure
1A). Such prolonged BDNF-mediated SRF activity and thereby overall cellular response might eventually enhance neuronal survival. In opposite to this, axonal growth inhibitors stimulate a rapid but transient IEG response (Figure
1A). As IEGs are well-known molecules regulating a cellular switch between cell death and survival [
16], axonal growth inhibitors might recruit IEGs to modulate an initial cellular response mediating axonal regeneration. Notably, both BDNF and axonal growth inhibitors employ small Rho family GTPases as downstream signaling effectors [
1],[
29]. Rho-GTPases are known to activate F-actin polymerization which in turn will enhance SRF activity [
17],[
30]. Thus, despite different time-scale of operation, pro- and anti-growth signals might share Rho-GTPase-to-SRF signaling as common downstream effector.
Axonal growth inhibitors did not modulate SRF activity through alterations in either SRF’s nuclear abundance nor nuclear localization (data not shown). Instead, our data are congruent with models of SRF activation involving specific cofactor recruitment to stimulate SRF-dependent gene transcription [
15],[
17]. Using reportergene assays (Figure
1C), we demonstrate an interaction of SRF with TCF family transcription factors to convey an axon regeneration inhibitor mediated IEG response. TCF family members such as Elk-1 are activated through MAP kinase phosphorylation [
15],[
17]. In this study, MAP kinases were identified as critical downstream effectors activated by myelin, Nogo and CSPGs (Figures
1,
2 and
3). Thus, besides Rho-GTPases, MAP kinases emerge as a further signaling pathway activated by myelin- and astrocyte associated growth inhibitors. Our data suggest that such MAP kinase activation might eventually result in TCF phosphorylation and TCF-SRF mediated gene transcription.
In this study, we have not addressed which of the receptors (i.e. NgRs, PIR-B or RPTPs) are engaged by axon regeneration inhibitors. However, we noted that oligodendrocyte- (total myelin and Nogo) as well as astrocyte-derived (i.e. CSPGs) axon growth inhibitors all shared a similar signaling profile with regard to intermediates recruited (i.e. MAP kinase, Rho-GTPase, SRF) and temporal signaling sequence followed (i.e. shorter stimulation time-points were more effective than pro-longed stimulation). This result suggests that all regeneration inhibitors signal through the same receptor molecules, a finding supporting the current model that all regeneration inhibitor indeed share the same neuronal receptor [
2].
How do these in vitro findings relate to axon injury in vivo?
Induction of an IEG response is reported in various brain injuries including spinal cord injury [
31]. Rapid but transient induction of c-Fos is a key event in regulation of survival vs. elimination of injured neurons [
16]. Thus, our
in vitro results suggest that upon encounter of myelin- or astrocyte-associated regeneration inhibitors, propagation of such an IEG response in neurons might also require MAP kinases and SRF
in vivo (Figure
3). In such a scenario, SRF might be an effector fulfilling the detrimental impact of signaling initiated by axon growth inhibitors. However, our data also reveal that SRF might be employed to circumvent the growth inhibitory potential of these regeneration obstacles. For this constitutively-active SRF-VP16 was used which unlike wild-type SRF is not subject to neuron-endogenous regulatory mechanisms. SRF-VP16 rescued myelin, Nogo or CSPG evoked neurite growth inhibition in primary neurons (Figure
4). The latter is likely due to SRF’s potential to modulate neuronal actin cytoskeletal dynamics [
17],[
22]. For instance, SRF-VP16 enhances the cellular F-actin content [
23] and the activity of the actin severing factor cofilin [
32] and might thereby allow for neurite growth on inhibitory substrates. In line with a beneficial SRF role in axonal regeneration, we demonstrated before an enhanced motoneuron survival by SRF-VP16 and impaired axon regeneration in
Srf deficient mice in the peripheral nervous system [
18],[
19].
In sum our data show that extrinsic regeneration barriers activate a neuronal signaling pathway involving MAP kinases and Rho-GTPases. In addition we show for the first time that axon growth inhibitors such as Nogo elicit a neuronal gene expression program mediated by SRF. Due to its dual access to regulation of IEGs and actin cytoskeletal dynamics, SRF might be an interesting gene regulator to analyze in CNS axon regeneration in vivo.
Competing interests
The authors declare that they have no competing interests.