Calcium Signaling, PKC Gamma, IP₃R1 and CAR8 Link Spinocerebellar Ataxias and Purkinje Cell Dendritic Development

Abstract

Background

Spinocerebellar ataxias (SCAs) are a group of cerebellar diseases characterized by progressive ataxia and cerebellar atrophy. Several forms of SCAs are caused by missense mutations or deletions in genes related to calcium signaling in Purkinje cells. Among them, spinocerebellar ataxia type 14 (SCA14) is caused by missense mutations in PRKCG gene which encodes protein kinase C gamma (PKCγ). It is remarkable that in several cases in which SCA is caused by point mutations in an individual gene, the affected genes are involved in the PKCγ signaling pathway and calcium signaling which is not only crucial for proper Purkinje cell function but is also involved in the control of Purkinje cell dendritic development. In this review, we will focus on the PKCγ signaling related genes and calcium signaling related genes then discuss their role for both Purkinje cell dendritic development and cerebellar ataxia.

Methods

Research related to SCAs and Purkinje cell dendritic development is reviewed.

Results

PKCγ dysregulation causes abnormal Purkinje cell dendritic development and SCA14. Carbonic anhydrase related protein 8 (Car8) encoding CAR8 and Itpr1 encoding IP3R1were identified as upregulated genes in one of SCA14 mouse model. IP3R1, CAR8 and PKCγ proteins are strongly and specifically expressed in Purkinje cells. The common function among them is that they are involved in the regulation of calcium homeostasis in Purkinje cells and their dysfunction causes ataxia in mouse and human. Furthermore, disruption of intracellular calcium homeostasis caused by mutations in some calcium channels in Purkinje cells links to abnormal Purkinje cell dendritic development and the pathogenesis of several SCAs.

Conclusion

Once PKCγ signaling related genes and calcium signaling related genes are disturbed, the normal dendritic development of Purkinje cells is impaired as well as the integration of signals from other neurons, resulting in abnormal development, cerebellar dysfunction and eventually Purkinje cell loss.

1. Introduction

Cerebellar Purkinje cells are among the best known neurons in the brain due to their large and elaborated dendritic trees [1]. They are the principal output neurons of the cerebellar cortex, therefore Purkinje cell dysfunction results in multiple cerebellar diseases. Purkinje cell survival and function are compromised in spinocerebellar ataxias (SCAs).

This is a group of hereditary neurodegenerative diseases with impaired function of the spinocerebellum [2]. Although there are sporadic forms of SCAs, the term is most often used to refer to the hereditary forms [3, 4]. In hereditary SCA, cerebellar neurons or non-neuronal cells (e.g. Bergmann glial cells) can be affected by the gene defect [5, 6]. Currently, 42 SCA types are known (the number is continuously increasing). The clinical aspect of SCAs is heterogeneous with various symptoms, characterized by a slowly progressive incoordination of gait often associated with poor limb coordination, dysarthria and cerebellar oculomotor disorders [5]. The most common forms of SCAs are caused by expansion of CAG repeats in members of the ataxin gene family [7]. In contrast, several other forms of SCAs are caused by single missense mutations or deletions in genes which encoding proteins often related to the calcium signaling cascade in Purkinje cells. Examples for targets of such missense mutations are protein kinase C gamma (PKCγ) causing SCA14 [8] and the inositol 1, 4, 5-trisphosphate receptors type 1 (IP₃R1) causing SCA15/16 [9-11] and SCA29 [12]. Moreover, also some CAG mutations may affect calcium signaling in Purkinje cells [13, 14].

It is remarkable that in several cases in which SCA is caused by point mutations in an individual gene, the affected genes are involved in the PKCγ signaling pathway and calcium signaling which is not only crucial for proper Purkinje cell function but is also involved in the control of Purkinje cell dendritic development [15].

Furthermore, PKCγ signaling pathway related proteins in Purkinje cells such as transient receptor potential cation channel type 3 (TRPC3) and metabotropic glutamate receptor type 1 (mGluR1) have influence on Purkinje dendritic development and mutations in these genes cause cerebellar ataxia in mice or humans [16, 17]. In this article, we will focus on the role of the affected calcium signaling molecules both for Purkinje cell dendritic development and for spinocerebellar ataxia.

2. CA2+ homeostasis in Purkinje cell dendritic development and ataxia

Disruption of intracellular calcium (Ca²⁺) homeostasis in Purkinje cells is thought to be a key mechanism in the pathogenesis of SCA and at the same time is known to control Purkinje cell dendritic development. Chronic activation of mGluR1 induces a pronounced reduction of the size and branching of the Purkinje cell dendritic tree [18]. This inhibition of Purkinje cell dendritic development after chronic activation of mGluR1 is partially rescued by inhibition of T-type and P/Q-type calcium channels [16], indicating that Ca²⁺ influx through T-type and P/Q-type channels is important for mGluR1 mediated dendritic growth inhibition.

SCA6 is caused by CAG expansions in the CACNA1A gene [19] which codes for the P/Q-type calcium channel Cav2.1. Although SCA6 is a polyglutamine disease, the polyglutamine stretch was shown to change the channel properties of Cav2.1 [20] causing a dysfunction of this channel [13, 14]. However, the pathogenic significance of this effect for the development of the SCA phenotype is still open [13]. The tottering mouse has a mutation in Cav2.1 [21], which results in reduced Ca²⁺ currents in cerebellar Purkinje cells. These mice have cerebellar ataxia and show intermittent absence seizures, which indicate the important role of Cav2.1 function in Purkinje cells [22]. In agreement with the important role of P/Q-type calcium channels, the dendritic arbor of the Purkinje cells in the tottering mouse is reduced in size and complexity [23].

The importance of the Ca²⁺ homeostasis for Purkinje cell dendritic development is further demonstrated by lurcher mutant mice which have increased calcium entry via a mutated GluR-delta2 channel resulting in a much reduced dendritic development which can be rescued by blocking Ca²⁺ influx through this channel [24]. Interference with Ca²⁺ clearance mechanisms also affects Purkinje cell dendritic development. Inhibition of the plasma membrane Ca²⁺-ATPase2 (PMCA2) activity by carboxyeosin resulted in a reduction of Purkinje cell dendritic growth [25]. Interestingly, it is known that PMCA2 does co-immunoprecipitate with mGluR1, Homer3 and IP₃R1, which suggests that the Ca²⁺ pump PMCA2, mGluR1, Homer3 and the IP₃R1 might be forming a complex and regulate each other [26].

Another mutation affecting the Ca²⁺ homeostasis in Purkinje cells is found in Moonwalker (Mwk) mice which have a point mutation in the TRPC3 leading to increased calcium influx. Mwk mice develop cerebellar ataxia [17] and also have abnormal dendritic arborization during cerebellar development [27]. Recently, mutations in the TRPC3 gene were linked to spinocerebellar ataxia in humans [28] and have been classified as SCA41. Interestingly, Trpc3 knockout mice showed normal dendritic development [16], indicating that an increased Ca²⁺ entry through the TRPC3 channel and not a loss of function did cause abnormal dendritic development and ataxia in the Mwk mice. Another report showed that CHO cells transfected with PKCγ carrying the G118D-PKCγ mutation showed increased Ca²⁺ entry through TRPC3 channels due to decreased phosphorylation of this channel by the mutant PKCγ [29]. This raises the possibility that PKCγ might be mediating Ca²⁺ entry through TRPC3 channels also in Purkinje cells. Dulneva et al. showed that CaMKIV is hyper-phosphorylated in the Mwk cerebellum and might be one candidate for the downstream signaling of the TRPC3 mediated Ca²⁺ overload [30]. One of the downstream targets of CaMKIV is retinoid-related orphan receptor α (RORα) which is a key factor for early dendritic development of Purkinje cells [30, 31].

3. PKCΓ and SCA14

By now, almost 40 different mutations or deletions in the PRKCG gene which encodes PKCγ are known to cause SCA14, but it is still unclear how these mutations ultimately cause Purkinje cell dysfunction and death as seen in SCA14. Remarkably, PKCγ-deficient mice only show mild ataxia and no gross morphological abnormalities in the cerebellum [8, 32]. Furthermore, SCA14 is a dominantly inherited disease indicating that a toxic gain of function or a dominant negative function rather than a loss of function of PKCγ causes SCA14. There are several reports about the kinase activity of mutant PKCγ found in SCA14. An early report showed that two SCA14 missense mutations were functionally increasing PKCγ catalytic activity, linking Purkinje cell degeneration to a potential gain of function phenotype of PKCγ [33]. Another report showed that 19 out of the 20 tested PKCγ mutations showed an increased constitutive activity of PKCγ and increased Ca²⁺ levels in the cytoplasm in a cell based assay, which suggests that a gain of function of PKCγ could underlie the pathology of SCA14 [29]. On the other hand, there is evidence that in particular SCA14 mutations in C1 domain might be functionally defective due to decreased binding to Diacylglycerol (DAG) [34]. These controversial findings suggest that it may not be the increase or decrease of PKCγ activity as such but rather the loss of the ability of the Purkinje cell to rapidly adapt the activity of PKCγ to the changing requirements of different functional states which eventually will result in a long-lasting dysfunction and the development of the ataxic phenotype.

It was also reported that mutant PKCγ protein aggregates in the cytoplasm of cultured cells transfected with PKCγ mutations [35, 36]. When mutant PKCγ transfected cells were treated by rapamycin, an inducer of autophagy, cells demonstrated an accelerated clearance of aggregates, indicating that autophagy can contribute to the degradation of mutant PKCγ [37]. However, in primary cultures of Purkinje cells transfected with PKCγ mutations, abnormal dendritic development of Purkinje cells also occurred independent of aggregate formation [38] therefore the role of these aggregates in the pathogenesis of SCA14 is still not clear. There is another report that amyloid-like oligomers and fibril formation of mutant PKCγ may contribute to SCA14 pathogenesis [39]. In this report SCA14 related C1 domain mutations promoted the amyloid-like fibril formation of PKCγ both in cells and in vitro. Although endogenous PKCγ itself may form amyloid-like fibrils, SCA14 related mutations accelerated this process substantially. PKCγ protein amyloid-like fibril formation thus might also contribute to SCA14 pathogenesis. However, direct proof that amyloid or aggregate formation is involved in Purkinje cell death in SCA14 is still missing.

Several groups have been searching for proteins interacting with mutant PKCγ and being potentially involved in the pathogenesis of SCA14. However, it has been found that many known PKCγ substrate proteins are not directly associated with SCA14 or Purkinje cell degeneration, such as metabotropic glutamate receptor 5 (mGluR5), non-muscle myosin heavy chain II-B, myristoylated alanine-rich C-kinase substrate (MARCKS) and GAP43/B-50 [40]. Aprataxin (APTX), which is a DNA repair protein and associated with autosomal recessive ataxia with oculomotor apraxia type 1 (AOA1) [41], was shown to be a candidate substrate of SCA14 mutant PKCγ [40].

Most of these studies used in vitro assays based on cell lines like CHO or COS7 cells in order to search for pathogenic mechanisms of SCA14 and produced conflicting results (see above). In order to clarify the molecular mechanisms leading to SCA14 in Purkinje cells a detailed analysis of the effects of mutant PKCγ within Purkinje cells is required. Shuvaev et al. reported that after lentiviral transfection of the C1 domain mutant PKCγ, aggregated PKCγ formation and impaired LTD in transfected Purkinje cells were observed, but there was no signs of Purkinje cell degeneration within the three weeks of transfection [42]. There are also reports of transgenic mouse models of SCA14. In a mouse model with ubiquitous expression of a human mutant PKCγ carrying the C1 domain mutation H101Y, a loss of Purkinje cells at the age of four weeks and stereotypic clasping responses in the hind limbs were reported [43]. Previous work had suggested that this mutation had a dominant negative effect on endogenous wild type PKCγ enzyme activity leading to uncontrolled, open gap junctions in cells expressing the H101Y-PKCγ mutation [43, 44]. Since the original short report about this mouse model in 2009, no further studies have been published. Ji et al. introduced a transgenic mouse model in which PKCγ with the kinase domain mutation S361G was specifically expressed in Purkinje cells using the L7 promoter [45]. S361G-PKCγ transgenic mice showed a mild ataxic phenotype and abnormal Purkinje cell dendritic morphology in organotypic slice culture strongly indicative of a high constitutive PKC activity [45]. The S361G-PKCγ transgenic mice offered the possibility to search for genes specifically regulated in Purkinje cells with an increased PKCγ activity because of the presence transgene. Using this approach carbonic anhydrase related protein 8 (Car8) and Itpr1 were identified as upregulated genes in S361G transgenic mice [46].

4. IP₃R1 and SCA15/16 & SCA29

IP₃Rs are membrane glycoproteins activated by inositol trisphosphate (IP3) and IP₃Rs function as a ligand-gated ion channel that releases Ca²⁺ from intracellular stores [47]. There are three isoforms of IP₃Rs, called types 1 (IP₃R1), 2 (IP₃R2), and 3 (IP₃R3). In many mammalian cells more than one of these isoforms are expressed, often all three isoforms [48]. IP₃R1 is the major neuronal IP₃R isoform in the central nervous system and it is abundant in the cerebellum, predominantly in Purkinje cells [49]. It has been suggested to be one of the key proteins in the pathogenesis of SCAs [50]. Itpr1 knock-out mice have a severe ataxia [10, 51] and Purkinje cells from these mice show abnormal dendritic development in dissociated cultures [52, 53]. In humans, two SCA subtypes are associated with the ITPR1 gene: SCA15/16 is caused by deletions or missense mutations in the ITPR1 gene [9, 11, 54] and SCA29 is caused by missense mutations in ITPR1 [55]. SCA29 is distinguished from SCA15/16 by clinical characteristics like early onset and delayed motor development [12, 55] but SCA15/16 and SCA29 reflect the same genetic target. Both forms show a dominant inheritance and it is not yet clear whether the disease is caused by a lack of sufficient functional IP₃R1 protein or whether there also might be a toxic gain of function involved [56]. In heterozygous IP₃R1-deficient mice there was no major phenotype found besides a mild motor discoordination [57]. It has been reported that one SCA15 mutation still has a functional Ca²⁺ release channel [58] and no other neuropathological mechanisms in SCA15/16 have been proposed to date [50]. In SCA29, at least six different point mutations have been identified [55]. One of these mutations is located in the CAR8 binding region [59] which might change IP3 binding affinity to the IP₃R1. In a recent study, it was shown that “Gillespie syndrome”, a rare hereditary disease with iris malformation and cerebellar ataxia, is also caused by point mutations in the ITPR1 gene which are thought to have a dominant negative action [60, 61].

Interestingly, ataxin-2 and ataxin-3 which are the mutated proteins in SCA2 and SCA3 respectively have been shown to interact with the IP₃R1, probably by increasing Ca²⁺ release to the cytoplasm [62]. If IP₃R1-mediated Ca²⁺ signaling is blocked by overexpression of the IP3R1 suppressor inositol 1, 4, 5-phosphatase (Inpp5a) or by stabilizing intracellular calcium levels with the drug dantrolene [62], Purkinje cell death by dark cell degeneration in SCA2 and SCA3 mouse models could be reduced [63, 64]. These findings indicate that IP₃R1 mediated Ca²⁺ signaling is also a key mechanism in pathogenesis of SCA2 and SCA3 [50]. It is intriguing that while in SCA15 and SCA29 there is evidence that the cerebellar phenotype is caused by a loss of function of the IP₃R1 either through dominant negative effects or through haploinsufficiency, it seems to be converse in SCA2 and SCA3 where a gain of function of the IP₃R1 is assumed through binding of the mutated ataxin proteins [65]. These opposing results suggest that it may not be the reduced or increased activity of the IP₃R1 as such but rather the loss of a precise regulation of its activity that causes the disease. In such a scenario it does not matter to which side the functional IP₃R1 activity is shifted but it would be the loss of regulation that eventually compromises Purkinje cell function and survival [66].

5. CAR8 and ataxia

CAR8 is a member of the carbonic anhydrase family which lacks enzymatic activity. It is predominantly expressed in the central nervous system, in particular there is a strong expression in cerebellar Purkinje cells [59, 67, 68]. In zebrafish, Car8 knockdown resulted in abnormal cerebellar development and ataxia [69]. In human patients, mutations in the CA8 gene were identified which are associated with cerebellar ataxia and mild cognitive retardation [70, 71]. Autoantibodies directed against CAR8 have been identified as a cause of Purkinje cell degeneration and cerebellar ataxia in paraneoplastic syndromes [72]. Mice carrying a 19-base-pair deletion in the Car8 gene have no CAR8 protein expression and the loss of function of CAR8 in these waddles (wdl) mouse mutants [67] results in ataxia accompanied by synaptic changes in the cerebellum which are associated with alterations in calcium regulation [73]. Interestingly, this marked ataxic syndrome occurs in the absence of gross anatomical defects and loss of Purkinje cells. Only a certain developmental delay was observed during early cerebellar development in wdl mice [67, 74]. In contrast, the zonal architecture of the cerebellar cortex was changed and abnormal Purkinje cell firing in vivo was observed [74]. This phenotype has similarities to that of mice with a Purkinje cell specific inactivation of the Cav2.1 Ca²⁺ channels [75]. These findings indicate that Ca²⁺ dysregulation through CAR8 or Cav2.1 can cause ataxia in the absence of Purkinje cell loss. Both Car8 and Itpr1 mRNAs were found to be downregulated in the staggerer mouse and in a SCA1 mouse model confirming the association of both genes to ataxia [76]. Recently it was shown that CAR8 was upregulated in a SCA14 mouse model with increased PKCγ activity [46]. CAR8 overexpression in developing Purkinje cells in dissociated cultures impaired Purkinje cell dendritic development raising the possibility that CAR8 might also be a regulator of Purkinje cell dendritic development [46]. There is evidence that CAR8 is associated with the IP₃R1 and may modulate IP₃R1 function by interfering with binding of IP3 to the IP₃R1 [59] although a direct interaction of the two proteins in Purkinje cells has not yet been proven [46]. These findings suggest that the ataxic phenotype of the wdl mouse could be explained by a loss of the precise regulation of IP₃R1 activity and that the altered activity of the IP₃R1 could contribute to the abnormal dendritic development of Purkinje cells.

6. IP3R1, CAR8 and PKCΓ are active in a common signaling pathway

IP₃R1, CAR8 and PKCγ proteins are all known to be strongly and specifically expressed in Purkinje cells [46, 77-80]. The common function among them is that they are involved in Ca²⁺ influx to the cytoplasm and in the regulation of Ca²⁺ homeostasis in Purkinje cells [29, 59, 81]. As mentioned, Ca²⁺ signaling is a key factor for both, Purkinje cell dendritic development and the pathogenesis of SCA. As depicted in Fig. (1A), Ca²⁺ signaling is initiated through Ca²⁺ channels and external stimuli acting upon membrane receptors (i.e. G-protein-coupled receptors or receptor tyrosine kinases), followed by phospholipase C (PLC)-mediated hydrolysis of phosphatidylinositol 4, 5-bisphosphate (PIP2) into the second messengers IP3 and DAG [82]. This pathway is dynamically controlled by feedback loops which control the IP3 level and metabolism as well as IP₃R1 function [83-84]. At basal levels of IP3, cytoplasmic Ca²⁺ levels are maintained by influx and efflux from the endoplasmic reticulum (ER) and the plasma membrane. Once PLC-mediated IP3 and DAG production has occurred, IP3 binds to the IP₃R1 and increases the Ca²⁺ influx to the cytoplasm from the ER. The simultaneous elevation of the Ca²⁺ concentration and DAG in the cytoplasm activates PKCγ which in turn phosphorylates its substrates [85]. However, the majority of the well characterized substrate proteins do not seem to be directly associated with SCA14 pathogenesis [40]. It is known that the IP₃R1 is a target of PKC phosphorylation and that phosphorylation promotes the IP3-induced calcium release [86]. This suggests that PKC-mediated phosphorylation of the IP₃R1 may form part of a positive feedback loop. Activated PKCγ phosphorylates TRPC3-channels which reduces Ca²⁺ influx from the plasma membrane through these channels and limits the Ca²⁺ concentration in the cytoplasm [29, 87]. As discussed above, one mutation related to SCA14 was recently shown to result in a constitutive active form of PKCγ causing abnormal dendritic development of Purkinje cells linking an increased PKC activity to the pathogenesis of ataxia [45]. Moreover, the IP₃R1 and CAR8 are upregulated in this SCA14 mouse model carrying the mutant PKCγ protein with higher kinase activity [46]. The upregulation of CAR8 may appear somewhat surprising because CAR8 is thought to be a negative regulator of IP₃R1 [59] and would be antagonizing the increased IP₃R1 expression. However, the increased expression of CAR8 could reflect a negative feedback loop for controlling Ca²⁺ homeostasis, and thus would be the logical consequence of the increased IP₃R1 expression.

Fig. (1A).

Summary of the PKCγ calcium-signaling pathway. Ligand binding to the mGluR1 receptor is followed by PLC-mediated hydrolysis of PIP2 into the second messenger IP3 and DAG. Binding of IP3 to the IP₃R1 and induces Ca²⁺ release from the endoplasmic reticulum to the cytoplasm. This step is controlled by CAR8 which can inhibit binding of IP3 to the IP₃R1. Simultaneous binding of Ca²⁺ to the C2 domain and of DAG to the C1 domain induce translocation of PKCγ to the plasma membrane. There, the pseudosubstrate is released from the kinase domain, allowing phosphorylation of downstream target proteins. PKC exerts an inhibitory effect on TRPC3 channel activity directly through phosphorylation or indirectly. A rise of the intracellular calcium concentration can also be mediated by the voltage-gated Ca²⁺-channel Cav2.1 or TRPC3. Conversely, PMCA2 can efficiently remove Ca²⁺ from the cytoplasm.

Conclusion

Overall, the IP₃R1, CAR8 and PKCγ related ataxic phenotypes could in part be explained by changes in IP₃R1 activity leading to poorly controlled Ca²⁺ influx/efflux and poor control of the Ca²⁺ concentration in the cytoplasm. This concept is supported by the finding that mutations affecting the function of various molecules involved in this signaling pathway will cause several subtypes of SCA or other forms of cerebellar ataxia (Fig. 1B). Once PKCγ- IP₃R1 related Ca²⁺ signaling is disturbed the normal dendritic development of Purkinje cells is impaired as well as the integration of signals from other neurons, resulting in abnormal development, cerebellar dysfunction and eventually Purkinje cell loss.

Fig. (1B).

Many proteins of the PKCγ calcium signaling pathway are targets of mutations causing ataxia. Mutations causing cerebellar ataxia often have proteins of the PKCγ calcium signaling pathway as targets (shown in red). This highlights the importance of this pathway for the appropriate function of cerebellar Purkinje cells.

Consent for Publication

Not applicable.

CONFLICT OF INTEREST

The authors declare no conflict of interest, financial or otherwise.