Effects of carbonic anhydrase VIII deficiency on cerebellar gene expression profiles in the wdl mouse

Introduction

The waddles (wdl) mouse is an autosomal recessive mutant that exhibits ataxia and dystonia. Like several other murine models of ataxia and dystonia, the wdl mouse exhibits no gross morphological or histological abnormalities of the nervous system [10, 21, 26, 44], The wdl mouse harbors a loss-of-function mutation in the carbonic anhydrase VIII (Car8) gene (Car8). Car8 is an acatalytic member of carbonate dehydratase family. The sole known function of Car8 is to inhibit inositol 1, 4, 5-trisphosphate (IP3) binding to IP3 receptor 1 (Ip3r1) [17], which plays a critical role in the modulation of intracellular calcium (Ca²⁺) signaling [19]. It is, therefore, quite possible that altered Ca²⁺ signaling may, at least partially, underlie the motor phenotype observed in the wdl mice. Ca²⁺ signaling is involved in the regulation of a wide variety of cellular activities including synapse recycling, proliferation, fertilization, learning and memory, long term potentiation, long term depression, apoptosis, contraction, metabolism, and modulation of other signaling systems. To investigate the molecular abnormalities associated with the unique motor syndrome of wdl mice, we performed genome-wide cerebellar gene expression profiling using Affymetrix oligonucleotide microarrays. We found that many genes involved in synaptogenesis, synaptic vesicle formation and transport, cellular proliferation and differentiation, and signal transduction were dysregulated in wdl mice.

Materials and methods

Mice were housed in animal care facilities at the University of Tennessee Health Science Center under conditions of 14 hr light/10 hr darkness at an ambient temperature of 20 ± 2°C with relative humidity of 30–60%. Experimental animal procedures and mouse husbandry were performed in accordance with the National Institutes of Health’s Guide for the Care and Use of Laboratory Animals and approved by the UTHSC Institutional Animal Care and Use Committee. For the Affymetrix microarray experiments, individual cerebella from 2-week female homozygous wdl mice (N = 3) and age-matched, wide-type (+/+) littermates (N = 3) were employed for each array. For relative quantitative real-time RT-PCR (QRT-PCR) validation of a small subset of differentially-expressed transcripts, another panel of individual cerebella (N = 9) from either wild-type or homozygous wdl mice were pooled to yield 3 replicates for each group. Total RNA was extracted using Trizol reagent (Life Technologies). RNA quality was checked with an Eppendorf BioPhotometer. Microarray hybridization and data generation were performed according to Affymetrix guidelines using GeneChip® Mouse Genome 430 2.0 arrays, related reagents, and GCOS 1.4 software (www.affymetrix.com). Comparative analysis was carried out between each individual wild-type and mutant sample, which generated 9 comparisons. Genes showing a ≥1.5 fold change on ≥7/9 comparisons were chosen for further statistical analysis using SAM (Significance Analysis of Microarrays, http://www-stat.stanford.edu/~tibs/SAM/) with a false discovery rate and q value of < 0.01%. Data clustering were performed with Cluster and TreeView (Eisen Laboratory, http://rana.lbl.gov/EisenSoftware.htm). QRT-PCR was performed with SYBR Green (Applied Biosystems, ABI) and the ABI PRISM 7900 HT Sequence Detection System according to ABI protocol. To be brief, 100ηg of total RNA was reverse transcribed into cDNA. Two-step PCR cycling was carried out as follows: 10 min at 95°C × 1 cycle followed by 15 s at 95°C and 1 min at 60°C × 40 cycles. The housekeeping gene 18S rRNA was used as an endogenous control. The comparative C_T method was used to determine relative expression levels. QRT-PCR primer pairs and assay results are presented in Table 1. Gene ontology functional analysis of differentially expressed genes was performed using GenMAPP software (www.genmapp.org). Lastly, literature-based biological profiling of altered gene expression patterns was carried out with Ingenuity Pathways Analysis (www.ingenuity.com).

Results

In this study, 192 transcripts were found to be significantly dysregulated in wdl mice (upregulated 90, downregulated 102). Among this set, 79 (41%) were expressed sequence tags (ESTs) of unknown function (Supplementary Table 1). Hierarchical clustering analysis, a standard method of identifying coexpressed or coregulated genes, detected 6 major clusters of transcripts (Supplementary Table 2). Strikingly, many of the dysregulated transcripts with the largest fold differences were parceled into two major clusters: upregulated genes in Cluster 3 (Glp1r, H2-D1, Paip1 and Scn2b) and downregulated genes in Cluster 6 (Car8, IMAGE cDNA clone 6309338 and C030014A21Rik). Consistent with previous Northern blotting results [21], the expression level of Car8 transcript as determined with QRT-PCR was markedly depressed in wdl mice. Many transcripts in Cluster 6 are involved in synapse vesicle recycling, transcriptional regulation, and signal transduction regulation. In Cluster 3, upregulated genes are related to synapse vesicle transport, protein biosynthesis, immune response, and cell adhesion and migration. Of note, three genes known to be involved in cerebellar function and implicated in the pathophysiology of ataxia (Cbln3, Chn2, and Etv1) were found in Cluster 4 [2, 24, 32]. Microarray analysis and QRT-PCR yielded similar results for the four genes assessed with both techniques: Car8, the cerebellar-enriched Cbln3, and the inhibitory receptors Gabra2 and Gabra6 (Supplementary Table 2).

Gene ontology functional analysis showed a remarkable increase in the expression of genes involved in cellular structure and function of the Golgi apparatus and decreased expression of genes of associated with intracellular signaling, cell division, and metal ion (mainly calcium and zinc) binding in wdl mice (Table 2). Literature-based pathway analysis with Ingenuity Pathway Analysis provided additional insight into those biological processes altered by the absence of Car8. The most significantly affected categories were cellular maintenance, growth and proliferation, organismal survival, signaling, development, and morphology (Table 3).

Discussion

Although Car8 is known to be a negative regulator of intracellular Ca²⁺ signaling [17], its exact role in the IP3 signaling and cytoplasmic calcium homeostasis is unclear. Calcium plays an important role in the regulation of numerous neuronal processes including the formation and adaptive modification of neural circuits [27], neurotransmitter release, and gene transcription [5]. Calcium-binding proteins play critical roles in mediating many of the various activities of calcium ions [4]. In wdl mice, there were no significant changes in the expression of several major common calcium-binding protein genes including Calb1, Calb2, and Pvalb. However, wdl mice did show cerebellar upregulation of the Ca²⁺-binding protein encoding genes Hpcal1 and Slc8a1(Table 3), which are predominantly expressed in Purkinje cells [30, 33]. It is interesting to note that two other Ca²⁺- binding protein genes, Nptx1 and Pclo, whose protein products are enriched in presynaptic plasma membranes [9], were downregulated in the mutant mice (Table 2). Based on predominant expression of Car8 in Purkinje cells and the Car8 interaction with IP3R1, Car8 deficiency is predicted to exert most of its effects via increased concentration of free cytosolic in cerebellar Purkinje cells. However, due to the effects of Ca²⁺ on multifarious neuronal processes, it is difficult to determine if Car8 may have additional cellular roles independent of its effects on Ca²⁺.

Hippocalcin like 1 (Hpcal, Table 3), a member of the family of intracellular neuronal calcium sensors involved in the calcium-dependent regulation of signal transduction cascades, may be involved in the control of clathrin-coated vesicle trafficking [20] whereas solute carrier family 8 (sodium/calcium exchanger) member 1(Slc8a1, Table 3) is a primary regulator of intracellular Ca²⁺ concentration by pumping Ca²⁺ out of cells [22]. Neuronal pentraxin 1 (Nptx1) has been implicated in the modulation of hypoxic-ischemic injury through an interaction with glutamate receptors [18]. The neuronal pentraxins are characterized by calcium-dependent ligand binding. Piccolo or presynaptic cytomatrix protein (Pclo) is a novel shared component of glutaminergic and GABAergic synapses that functions as a presynaptic calcium sensor and may regulate neurotransmitter release [13]. Given the variety of cell types that contribute to olivocerebellar and cerebellar cortical local area networks, it is difficult to dissect direct versus compensatory effects of Car8 deficiency in Purkinje cells. For instance, neuronal pentraxin 1 is expressed in both granule and Purkinje cells [37].

An important finding in this study was the significant effects of Car8 deficiency on genes involved in the structure and function of the Golgi apparatus (GA), particularly vesicle assembly and transport. In neurons, GA is the place for the packing and transport of synaptic vesicles containing neurotransmitters. Fragmentation of the neuronal GA has been implicated in several neurodegenerative disorders including spinocerebelar ataxia type 2 [14]. Interestingly, we found increased expression of GA genes including Arcn1, Clasp2, Pik4cb, St6galnac5, and Tctex1 in wdl mice (Table 2). Archain vesicle transport protein 1 (Arcn1) is one of major components of synapse vesicle coat assembly [31]. CLIP-associated protein 2 (Clasp2) participates in the local regulation of distal microtubule dynamics [1], which is related to vesicle trafficking. Phosphatidyl-inositol 4-kinase (Pik4cb) interacts with frequenin (also know as neuronal calcium sensor 1) in the exocytosis of secretory vesicles [29]. Finally, t-complex testis-expressed 1 protein (Tctex1), a light-chain subunit of the dynein motor complex, interacts directly and selectively with N- and P/Q-type Ca²⁺ channels [23]. Taken together, these gene expression changes suggest that increased activity of the GA and downstream vesicle transport may contribute to the pathophysiology of ataxia and/or dystonia by virtue of increased receptor delivery to the neuronal cell surface. Since the GA has been confirmed as an important distinct IP3-sensitive Ca²⁺ store [35], our findings also suggest that Car8 deficiency may exert direct rather than compensatory effects on GA activity.

Most autosomal recessive ataxias are multisystem disorders due to loss-of-function mutations. DNA repair and maintenance are critical to the molecular pathophysiology of seom types of ataxic disorders such as ataxia telangiectasia, AT-like disease, ataxia with oculomotor apraxia 1 and 2, and spinocerebellar ataxia with axonal neuropathy [40]. In our study, three genes (Cbln3, Chn2, and Etv1) that are either exclusively or predominantly expressed in the cerebellum were significantly downregulated in wdl mice (Supplementary Table 2). The peptide encoded by Cbln3 may play an important role in the synaptic integrity and plasticity of Purkinje cells via interaction with cerebellin (Cbln1) [16, 32]. Chn2 encodes a GTPase-activating protein (GAP) for p21 ras-related Rac [24]. For review, Rac, a primary member of Rho GTPase family, may be a key regulator of synaptogenesis and dendritogenesis by coordinating assembly of the actin and microtubule cytoskeleton [15, 43]. Lastly, Etv1 encodes a proprioceptive marker which may control the expression of other proprioceptive neuron-specific genes through collaboration with Runx3 for their expression [25]. Interestingly, Etv1 mutant mice exhibit a severe limb ataxia due to faulty assemblage of sensorimotor circuitry within the spinal cord [2]. These results suggest that Car8 deficiency could result in impaired neuronal connectivity through reduced synapse formation and dendritogenesis. By extrapolation, similar development defects may underlie some recessive ataxias in humans.

Although zinc plays an important role in cell surface signaling and acts as an enzymatic cofactor, little is known about cellular mechanisms mediating zinc homeostasis. The concentration of zinc has been shown to be altered in a number of disorders of the central nervous system including Friedreich’s ataxia [12]. Furthermore, mutations of several zinc ion-binding protein genes have been implicated in several ataxic disorders, such as Aptx in autosomal recessive ataxia [11], Egr3 in sensory ataxia [42], and Zic1 in knockout mice with severe ataxia [3]. In our study, transcripts for several zinc ion-binding proteins (Pip5k3, Chn2, Zfp52, Ttc3, Rph3a, 9830124H08Rik, and BC021442) were downregulated in wdl mice (Table 2). Among these, Rph3a is implicated in synaptic transmission [38], while the granule cell-specific and ataxia-related Chn2 has been confirmed to be involved in regulation of cell proliferation [48] and diacylglycerol signaling [8].

Purkinje cells have already finished cell division at birth whereas granule cells proliferate vigorously during the first three postnatal weeks [47] (http://www.cdtdb.brain.riken.jp/CDT/Top.jsp). Therefore, the down-regulation of several genes involved in cell division, including Sept3, Sept4, Pafah1b1, and Ccne2 (Table 2), may reflect decreased granule cell proliferation and/or neuritogenesis in the mutants. Septin 3 (Sept3) is a brain-specific member of the septin family of GTPase enzymes that assemble as intracellular filamentous scaffolds and are involved in cytokinesis or exocytosis. Most members of the septin family appear to interact with each other in order to assemble as filaments, such as the interaction of Septin 4 (Sept4) and Septin 8 (Sept8) in vesicle trafficking [6]. Since Car8 expression in cerebellar cortex is limited to Purkinje cells, any down-regulation of Sept3 and Sept4 transcripts that is occurring within granule cells must be some type of feedback response to abnormal Purkinje cell signaling.

Previous work has shown that the expression patterns of the α2 (Gabra2) and α6 (Gabra6) subunits of gamma-aminobutyric acid type A (GABAA) receptors are temporally associated with the early (α2) and late (α6) postnatal development of cerebellar granule cells [28, 39, 45]. Both granule and Purkinje cells express the α2 subunit whereas only granule cells express the α6 subunit [36]. Thus, the down-regulation of Gabra6 transcript in wdl mice must be an indirect effect of Purkinje cell Car8 deficiency. It is interesting to note that the α6 subunit is also down-regulated stargazer mice, a motoric mutant that exhibits dystonia and ataxia [41]. In the context of cerebellar development it should be noted that the α6 subunit may mediate cell surface anchoring [34]. The α6 subunit of GABAA receptors expressed by granule cell may regulate several aspects of normal cerebellar function [6, 7, 46].

In summary, our gene expression profiling analysis suggests that calcium homeostasis, synaptogenesis and dendritogenesis, vesicle formation and transport, and, possibly, granule cell proliferation and/or neuritogenesis, are dysregulated in wdl mice. Furthermore, cell-surface signaling pathways in wdl mice may be distorted through changes in intracellular Ca²⁺, alterations in the expression of zinc-binding proteins, and increased activity of the GA. These findings suggest that ultrastructural abnormalities of specific neuronal elements may be present in wdl cerebellar cortex. In addition, the substantial number of dysregulated genes in wdl mice indicates that Car8 plays an important physiological role in cerebellar cortex. Overall, the scope of our results is not surprising given that significant changes in gene expression profiles have been described in a phenotypically-normal mouse model of Friedrich’s ataxia (10).

Supplementary Material