Introduction

Neuritin, discovered by Nedivi and colleagues in 19931, is an activity-induced neurotrophin factor, named for its function of promoting neurite outgrowth. Neuritin mRNA is predominantly expressed in the nervous system, although minor expression can also be detected in other organs2. Although most studies regarding neuritin have focused primarily on its regulation and functions in neuroprotection and regeneration, other functions including roles in angiogenesis, tumorigenesis, and immunomodulation have also been reported. For example, new evidence has shown that neuritin expression can be down-regulated or up-regulated among different cancer types3,4,5. Another study found that recombinant neuritin protein increased human umbilical vein endothelial cell migration and overexpression of neuritin significantly promoted tumor angiogenesis6. In addition, Barbi et al found that neuritin could maintain and promote regulatory T cell (Treg) function in autoimmune diseases and cancer. This study further characterized neuritin as underappreciated immunoregulatory molecule and a potential target for therapies aimed at the fine-tuning of Treg function in cancer and inflammatory/autoimmune diseases7.

Although several reviews8,9 have discussed the physiological and pathological functions of neuritin in nerve and non-nervous tissue, the primary mechanisms of neuritin, including whether or not it acts through a receptor or which downstream signals might be activated following binding, are not fully understood. In this brief review, in addition to the biological characteristics of neuritin and its neuronal functions, we also discuss work published by this lab and others regarding advances in uncovering the mechanisms involved in neuritin-mediated signaling10,11,12.

The biological characteristics of neuritin

Using northern-blot analysis, scientists discovered that neuritin is expressed in many organs of the human body during puberty and adulthood. Among these organs, the highest expression levels are found in the brain, followed by liver, lung, heart, and skeletal muscles, respectively. Since its discovery nearly two decades ago, new techniques have aided in identifying the genes which encode for neuritin2,13.

The structure of the neuritin gene and its subsequent activation

The human neuritin gene (candidate plasticity gene 15, CPG15) is located on chromosome 6 (6p25.1); is 2072 bp long; and is highly conserved across species, with a homology of 97% between human and mouse (cpg15). The open reading frame is 426 bp long and contains two introns and three exons which encode the amino acid sequences 1–19, 20–67, 68–142, separately. The entire coding DNA sequence (CDS) of cpg15 encodes 142 amino acid residues; the first 27 of which act as a signal peptide; while residues 116–142 residues act as the glycosylphosphatidylinosotol (GPI) anchor. The remaining 88 amino acid residues are considered as the core of neuritin2.

The immediate-early gene, cpg15, can be induced by Ca2+ influx through N-methyl-D-aspartate (NMDA) receptor and L-type voltage-gated calcium channels (VGCC). Influx of Ca2+ can further activate the Ca2+/calmodulin-dependent protein kinase pathway and also the mitogen-activated protein kinase (MAPK) pathway, thus initiating transcription of the cpg15 gene2. Although the activity-dependent expression of neuritin does not require the activation of protein kinase A (PKA), Fujino and colleagues found that cpg15 could be induced by cyclic adenosine monophosphate (cAMP) in active neurons14. Furthermore, they proved that CREB can bind to the promotor region of cpg15 in vivo and partially regulate the activity-dependent expression of neuritin.

In addition to cpg15 and its encoded protein, neuritin, Fujino et al also identified the only paralogue of cpg15 in the mouse and human genome, cpg15-2 15. Specifically, they found that both cpg15 and cpg15-2 were regulated by neuronal activity; the proteins for which they encoded shared basic biochemical properties; and both proteins had similar functions in neurite growth and neuronal survival. However, cpg15 and cpg15-2 were found to be differentially expressed, with cpg15 mRNA most abundant in the brain and liver, while cpg15-2 mRNA was most abundant in the eye and brain. Within the central nervous system (CNS), cpg15 mRNA was most abundant in the cerebral cortex, followed by the hippocampus and thalamus, respectively; whereas cpg15-2 was expressed most abundantly in the retina, followed by the olfactory bulb and striatum, respectively. As a result of the discovery of cpg15-2, several researchers termed the protein encoded by this gene as neuritin-2. However, to date, additional studies regarding neuritin-2 have been lacking.

Neuritin expression and regulation

Neuritin is concentrated in the neuronal soma and has been shown to be unevenly dispersed along neurite projections in the rat brain2. Sato et al showed that neuritin is expressed in puncta throughout cell bodies, dendrites, and axons; and its expression in axon terminals is particularly strong16. A recent study by Merianda et al showed that neuritin mRNA shifted from neuronal cell bodies to axons, predominantly after nerve crush injury. Interestingly, neuritin mRNA also showed different localization elements for CNS and PNS axons: the 3' untranslated region (UTR) drove localization in hippocampal neurons and the 5' UTR drove localization in sensory axons17.

As cpg15 is an activity-regulated gene, late neuritin expression is contemporaneous with critical periods for activity-dependent plasticity and requires action potential activity. However, neuritin is also expressed in an activity-independent manner during early brain development, prior to circuit formation and mutations18.

Added to the complicated regulation of neuritin expression, scientists found two forms of neuritin expressed in the CNS; one being membrane-bound and the other soluble. Membrane-bound neuritin is attached to the cell plasma membrane through a GPI linkage and has been shown to function non-cell autonomously to regulate growth of opposing dendritic and axonal arbors; and also promotes synaptic maturation19. Putz et al found that soluble neuritin expressed in the embryonic brain could regulate survival of cortical progenitors by preventing caspase-mediated apoptosis20. These observations suggest that: neuritin has different expression patterns at different development stages; and the two forms of neuritin (membrane-bound and soluble) have diverse functions.

The neuronal functions of neuritin

As a neurotrophic factor, neuritin has multiple effects in the nervous system and many studies related to its neuronal functions have been investigated primarily through studies involving the overexpression, gene silencing, and application of neuritin. To date, scientists have showed that neuritin is important in neuritogenesis, neurite arborization, and neurite extension9.

Neuritin promotes neuritogenesis and neuronal regeneration

Naeve et al found that treatment of primary cultured rat hippocampal and cortical neurons with recombinant neuritin over 4 days produced extensive neuritogenesis compared with control cultures2. Other groups have found similar neuritogenesis effects in PC12 cells and in motor neurons21,22,23. Moreover, Picard et al indicated that the absence of neuritin by global knock-out of cpg15 induced aberrant development and plasticity of excitatory visual cortical networks24. In addition, the effect of neuritin in promoting neuritogenesis also has been found in hippocampal neurons isolated from depression models; whereby viral-mediated expression of neuritin in the hippocampus prevented atrophy of dendrites and dendritic spines caused by chronic stress25. Taken together, these studies indicate that neuritin plays an important role in neurodevelopment and neural plasticity.

Neuritin is also a potential target in the study of neuronal regeneration after injury. Di Giovanni et al found that in neurons and axons in an experimental rat spinal cord injury (SCI) model, neuritin mRNA and protein levels were down-regulated 24 h post injury; followed by strong re-induction during the following 2 weeks, which paralleled functional recovery26. This finding was recently confirmed by Gao et al, where they suggested that neuritin created an environment that promoted nerve cell survival and neurite regeneration after SCI, thus contributing to nerve regeneration and recovery of motor function27. Besides spinal cord injury, results obtained by Zhao et al suggest that soluble neuritin secreted by astrocytes promoted the recovery of injured hippocampal neurons via adherance to the surface of neurons in ischemic brains28.

Similar neuroprotective and regenerative effects were observed in retina, optic nerve and sciatic nerve fiber after injury. Using an optic nerve crush mouse model, Xu et al found that neuritin mRNA and protein were transiently increased in retina and optic nerve after injury. They also found infection of rat retina with neuritin-expressing lentivirus significantly increased the number of retinal ganglion cells (RGCs) and optic nerve axons compared with controls; these effects were subsequently confirmed by functional tests29. Using a similar optic nerve crush model, Sharma et al evaluated the axogenic properties of neuritin on axotomized RGCs in vitro and on the optic nerve in vivo, and found that recombinant neuritin significantly increased RGCs survival and neurite outgrowth. Furthermore, significant increases in Gap43, a classical marker of axonal regeneration, were observed within the optic nerves of the adeno-associated virus (AAV)2–neuritin group, suggesting that neuritin preserved RGC function on axotomized RGCs in vitro and after axonal injury in vivo 30. Using a sciatic nerve injury model in rats, Wang et al reported that recombinant human neuritin stimulated nerve regeneration and restored motor function, as assessed by the walking-track test and nerve conduction velocity; and thus promoted the repair of injured sciatic nerves31. Although, neuritin can be both down-regulated and up-regulated at different stages in different injury models, the overall levels of neuritin increase after nerve injury; indicating that axonal neuritin protein synthesis may provide a spatially restricted mechanism for promoting axonal growth. Taken together, these data suggest that neuritin plays a critical role in nerve regeneration and is a potential new target for optic and sciatic nerve injury therapy.

Neuritin promotes synaptic maturation, stability and neurotransmitter release

The formation of CNS circuits is characterized by the development of neuronal structure and synaptic function. The effect of neuritin in promoting synaptic maturation was first reported by Cantallops and colleagues29. Using whole-cell recording from Xenopus optic tectal neurons, they found that neuritin expression promoted retinotectal synapse maturation by recruiting functional AMPA receptors to synapses. In addition, they found that when a truncated form of neuritin (lack of GPI anchor) was expressed, it did not promote axon arbor growth and actually blocked synaptic maturation19,32. These results not only suggest that neuritin can coordinately increase the growth of pre- and postsynaptic structures and also strengthen their synaptic contacts; but that different forms of neuritin coordinate different processes. To investigate the requirement for neuritin in maturation of synapses, Fujino et al generated a cpg15 knockout (cpg15 KO) mouse and found that in these mice, there was a developmental delay in axonal and dendritic arborization. Electrophysiological studies also showed that synaptic maturation was delayed, and electron microscopy confirmed that many dendritic spines initially lacked functional synaptic contacts33. These findings established a role for cpg15 in the maturation of synapses and efficient circuit formation.

The frequency of miniature excitatory post-synaptic currents (mEPSCs) in neuron recordings has been used for the evaluation of synaptic efficiency; including number of synapses and the probability of transmitter release. Picard et al showed that the density of spine protrusions was lower in neurons of cpg15 KO mice, which was consistent with the reduced mEPSC frequency measured at P28. These findings indicated that loss of neuritin likely resulted in the abnormal postnatal development of the excitatory network in the visual cortex and associated disruption in development of visual receptive field properties, suggesting loss of normal excitatory synapse maturation in neuritin deficient mice24. In accordance with the Picard et al study in the visual cortex, data from this lab demonstrated that the overexpression of hippocampal neuritin by an AAV vector significantly increased neuritin levels and also dendritic spine density of hippocampal neurons34. However, it is worth noting that a recent study from this lab involving the medial prefrontal cortex revealed that, although short-term incubation of cortical neurons with recombinant neuritin resulted in an increase in mEPSC frequency, this increase was not induced by increased dendritic spines density; but rather by the release glutamate. Moreover, up-regulation of the T-type VGCC α-subunit (CaV3.3) at the membrane surface was also involved. This work provides direct evidence for neuritin-mediated modification of synaptic plasticity and synaptic transmission in cortical neurons12.

The role of neuritin in Alzheimer's disease (AD) and psychiatric disorder-related cognitive function impairment

As synaptic plasticity has been associated with depressive disorders, schizophrenia, and cognitive performance, studies regarding the potential therapeutic use of neuritin in combating loss of cognitive function have been reported. Additionally, a possible role for neuritin gene variations in psychiatric disorder etiology have also been reported recently. Son et al found that in a chronic unpredictable stress-induced rat depression model, neuritin expression in the hippocampus was decreased. Furthermore, viral-mediated expression of neuritin in the hippocampus produced antidepressant effects and prevented the atrophy of dendrites and spines; as well as depressive and anxiety behaviors caused by chronic unpredictable stress25. These results demonstrate a unique action of neuritin in depressive disorders. A recent study from this lab also indicated that exposure of mice to extremely low-frequency (50 Hz) electromagnetic fields (ELF MFs) significantly decreased hippocampal dendritic spine density; and resulted in deficits in recognition memory, as detected by a novel object recognition test34. AAV-mediated overexpression of neuritin in the hippocampus was found to reverse both ELF MFs exposure-induced reduction in dendritic spine density and deficits in recognition and novel object associative recognition memory34.

Another potential effect of neuritin is against AD-related neural and cognitive defects35. Using the AD model Tg2576 mouse, An et al observed that exogenous application of recombinant neuritin fully restored dendritic complexity as well as spine density in hippocampal neurons prepared from Tg2576 mice; without affecting neurite branching of neurons from wild-type littermates. Chronically infused neuritin into Tg2576 mouse brains also aided in the restoration long-term potentiation in acute hippocampal slices35. Similarly, Choi et al indicated that viral-mediated overexpression of neuritin in the dentate gyrus of 13-month-old Tg2576 mice attenuated a deficit in learning and memory, as assessed by the Morris water maze test. This neuritin-mediated rescue effect on AD-induced cognitive impairments resulted in an approximate 3.6-fold increase in synaptophysin protein levels (a presynaptic marker). In addition, Choi et al also demonstrated that levels of neuritin mRNA were significantly decreased in the hippocampus and cerebral cortex of AD patients compared to age-matched control subjects36. Consistent with this study, Fei et al found that microRNA-574 (miR-574) was up-regulated in the hippocampi of 5-month-old APP/PS1 mice; while overexpression of miR-574 lowered neuritin and synaptic protein expression in Aβ25-35-treated primary hippocampal neurons37. These studies indicated that neuritin could used as a potential therapeutic approach for patients with AD.

Recent studies have demonstrated that there is an association between neuritin and psychiatric diseases, including schizophrenia, schizophrenia-spectrum disorders, and bipolar disorders. Chandler et al and Fatjo-Vilas et al analyzed common variants along the neuritin sequence and their corresponding association with intelligence and schizophrenia-spectrum disorders by testing neuritin gene sequence variations on susceptibility to schizophrenia and on general cognitive ability in patients and non-psychiatric control subjects. Results suggested that certain allelic variants in the neuritin gene modulated the risk for developing schizophrenia-spectrum disorders and also reduced general cognitive performance. Thus, the association between variations in neuritin and schizophrenia suggests a role for neuritin as a modifier of cognitive function38,39. A subsequent study by Prats et al found that multiple genes were involved in this modification process and gene–gene interactions between dysbindin-1 and neuritin increased the risk of schizophrenia-spectrum disorders. Interestingly, both of these genes are known to be involved in synaptic maturation, connectivity, and glutamate signaling40. These results suggest that the neuritin gene could be used as a potential marker for schizophrenia etiology.

Potential receptors of neuritin and related signaling pathways

Although previous studies implied that neuritin may function as a ligand15,41, few have investigated the receptor/receptors responsible for neuritin signal transduction. For several years, this lab has worked on neuritin related pathways and and the results may aide in the understanding of the role neuritin plays in the CNS.

Neuritin may be a potential ligand for the insulin receptor (IR)

We previously showed that incubation of cerebellar granule neurons (CGNs) with neuritin increased the density of transient outward potassium currents (I A) in a time- and concentration-dependent manner. Further studies revealed that these increases were due to the enhanced transcription and translation of Kv4.2, the main α-subunit of I A 10. Interestingly, as administration of insulin induces I A density increases as well as increases in Kv4.2 expression, we hypothesized that the enhanced expression of Kv4.2 caused by neuritin may be through the activation of the IR. Indeed, biochemical analyses showed neuritin could activate IR and that pharmacological blockade of IR completely offset the effect caused by neuritin. To date though, neuritin is still considered as an orphan ligand and its receptor remains unknown. Although there is a lack of evidence indicating that neuritin may directly bind to the IR, our data suggest neuritin can be a potential ligand of IR.

However, in CGNs, insulin-like growth factor type 1 receptor (IGF-1R) is also present42. Although IR and IGF-1R differ in ligand binding kinetics and functions, they share a high degree of structural homology43. Therefore, it cannot be discounted that neuritin may also bind to IGF-1R. However, results indicated that pharmacological inhibition of IGF-1R affected neither neuritin induced upregulation of I A density nor neuritin induced increases in Kv4.2 expression10. Interestingly, Yan et al recently have shown that exogenous IGF-1 can rescue Schwann cells from hyperglycemia-induced apoptosis. In the same study, they found that neuritin mediated the rescue of IGF-1 and they considered neuritin as a downstream element of the IGF-1-PI3K signaling pathway44. Taken together, these studies indicate that neuritin may not directly activate the IGF-1R.

Neuritin activates the MEK-extracellular signal-regulated kinase (ERK) and PI3K-Akt-mTOR pathways

ERK is an important regulator of neuronal functions, including synaptic plasticity, and long term memory formation45,46. Previous studies indicated that the ERK pathway is a target for growth factors such as NGF and also IR signal transduction47,48. Studies from this lab have shown that: incubation of CGNs with neuritin for 30 min dramatically increased the phosphorylation of ERK1/2; and elevated phosphorylation levels of ERK1/2 returned to baseline levels after 60 min of incubation. Notably, co-incubation of neuritin and U0126, an inhibitor of the MEK-ERK pathway, significantly blocked both neuritin induced Kv4.2 expression and increases in I A density.

In addition to the MEK-ERK pathway, we also tested if neuritin activated the PI3K-Akt-mTOR pathway. This signaling pathway is essential for several forms of synaptic plasticity for learning and memory49,50. Similar to ERK, following incubation of CGNs with neuritin, the phosphorylation of both Akt and mTOR increased. However, unlike the quick response of ERK, the onset of Akt phosphorylation occurred 1 h after incubation and lasted up to 6 h. Interestingly, increased phosphorylation of mTOR was not found until 18 h after neuritin incubation, and lasted up 24 h. Similarly, co-incubation of neuritin with either the PI3K inhibitor, LY294002, or the mTOR blocker, rapamycin, reversed neuritin-mediated increases in Kv4.2 expression and elevated I A density.

Neuritin-mediated expression of Kv4.2 requires the activation of Ca2+-CaN-NFATc4 pathway

Most studies regarding neuritin have focused on determining how and when the neuritin-encoding gene, cpg15, is activated14,18,20. As neuritin is also a neurotrophic factor, it may modulate the expression of other genes; for example, Kv4.2. Recently, we found that the Ca2+-calcineurin (CaN)-nuclear factor of activated T-cells (NFAT) c4 axis is required for this modulation effect.

Using Ca2+-sensitive fluorescent dye, Fura-2, in vitro calcium imaging data showed that neuritin-incubation for 15 min increased intracellular Ca2+ ([Ca2+]i) levels in mouse CGNs at both basal and high K+ conditions. Potentiated [Ca2+]i at high K+ conditions could be a consequence of modulated T-type or L-type VGCCs51, while elevated [Ca2+]i at basal levels may depend on increased release of calcium from endoplasmic reticuli (ER) through inositol 1,4,5-triphosphate receptors and/or ryanodine receptors52,53,54. Additionally, increased [Ca2+]i could also be the result of inhibited up-take of Ca2+ into ER by IR pathway activation55. A recent study showed that neuritin up-regulated the surface expression of the CaV3.3 α-subunit9. Although the precise mechanisms of increased [Ca2+]i by neuritin need further exploration, there is little doubt that elevated [Ca2+]i activate related pathways. Indeed, when neuritin was co-incubated with calmidazolium chloride, a calmodulin (CaM) inhibitor, the effects of neuritin on Kv4.2 expression and I A density were blocked.

It is well established that activated CaM modulates CaN, a Ca2+ and CaM dependent serine/threonine protein phosphatase56. CaN may activate NFATc4, a transcription factor, via dephosphorylation. Activated NFATc is then translocated into the nucleus, where it regulates gene expression57. Using the CaN inhibitor, cyclosporine A, or ascomycin (FK520), we found that translocation of NFATc4 was blocked, resulting in the enhanced transcription of Kv4.2 and also increased I A density; both mediated by neuritin. More importantly, in nfatc4 −/− mice, neuritin no longer affected Kv4.2 expression or I A density. Taken together, these data suggest that neuritin elevates [Ca2+]i, which subsequently activates the CaN-NFATc4 pathway resulting in up-regulation of Kv4.2 and increased I A density.

Other possible signaling pathways

Although few studies have focused on the mechanism or signaling pathways involved in neuritin-related functions, several groups have done interesting work related to potential signaling pathways58,59. Using sensory neurons in an experimental diabetic neuropathy, Karamoysoyli et al found in rat L4/5 small-diameter dorsal root ganglia neurons, neuritin was colocalized with the tyrosine kinase A (TrkA) receptor58; however, there was no direct evidence indicating that neuritin activated the TrKA receptor. Recently, Shimada et al reported that neuritin and fibroblast growth factor (FGF) cooperated in inducing mossy fiber sprouting through FGF signaling pathways59. In addition, they found that neuritin activated the MAPK, ERK, PI3K-Akt, and phospholipase C-Ca2+ pathways; which is consistent with our results. Interestingly, they also found that activation of these pathways could be blocked by inhibiting the FGF receptor. This latter finding suggests that interaction partners of neuritin may vary by cellular location. Recently, Zhang et al reported a mechanism connecting neuritin with Notch signaling. Using yeast two-hybrid and co-immunoprecipitation assays, they discovered an interaction between neuritin and Neuralized E3 Ubiquitin Protein Ligase 1 (NEURL1), an important regulator of Notch. Notably, recombinant neuritin restored the retraction of neurites caused by activation of Notch, and neurite growth stimulated by neuritin was partially blocked by NEURL1; suggesting that neuritin is an upstream and negative regulator of NEURL1 that inhibits Notch signaling to promote neurite growth23. This mechanism provides a valuable foundation for further investigation into the role of neuritin in neurodevelopment and neural plasticity.

It is well established that neuritin is a GPI-linked protein2; and that GPI-linked proteins can act as heterophilic or homophilic adhesion molecules, which act as co-receptors for slit FGF and laminin60. Although neuritin lacks the common immunoglobulin G domain of GPI-linked adhesion molecules, it does have some structural similarity to ephrins; a family of GPI-linked ligands that can activate their receptors by cell-cell contact61. This indicates that GPI-linked neuritin could possibly be activated by the interaction of neuritin from adjacent neurons or between neuron and the extracellular matrix.

Conclusion and perspectives

Neuritin was discoverd more than two decades ago and a large body of research exists regarding its functions, yet little is known in regards to possible receptors of this protein or what signaling pathways may be activated. As mentioned in this review, neuritin-mediated activation of IR triggered the MEK-ERK, PI3K-Akt-mTOR, and Ca2+-CaM-CaN-NFATc4 pathways resulting in the expression of Kv4.2 α-subunit in mouse CGNs: A proposed model is shown Figure 1. We also confirmed neuritin enhanced synaptic transmission in mouse mPFC neurons, a function that requires the activation of both IR and the MEK-ERK pathways. It remains to be elucidated whether there are other receptors and related signaling pathways for neuritin in different tissues and cells.

Figure 1
figure 1

Schematic illustration depicting signaling pathways involved in neuritin-induced Kv4.2 expression and the subsequent increase in I A channel density.

Full size image

Although neuritin is anchored to the cell membrane, it has been shown that it can cleaved by phospholipase C, producing a soluble form of neuritin19. Most studies associated with neuritin have used expression plasmids which produce GPI-linked neuritin, while several labs including our own, used plasmids which produced the soluble form of neuritin. Although studies have proven that both GPI-linked and soluble neuritin are functional19,59, it is not known whether these two different forms use the same receptor or similar activate signaling pathways. Future research is needed to clarify these questions.

Fujino et al discovered cpg15-2 in both humans and mice15. Interestingly, besides having a similar gene sequences, cpg15-2 and cpg15 also have similar temporal and spatial expression patterns. Furthermore, the two isoforms (netrin 1 and 2) encoded by these genes can form heterodimers. Future studies are warranted to investigate whether the function and regulation of neuritin-2 is different from neuritin-1. In addition, more work is needed in order to fully understand neuritin and its possible therapeutic role in neurodegenerative and psychiatric diseases.