- Email: [email protected]
Cell-type-specific alternative splicing in spinocerebellar ataxia type 6
Neuroscience Letters 447 (2008) 78–81
Contents lists available at ScienceDirect
Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet
Cell-type-specific alternative splicing in spinocerebellar ataxia type 6 Taiji Tsunemi a,b,∗ , Kinya Ishikawa a , Honglian Jin a , Hidehiro Mizusawa a a b
Department of Neurology and Neurological Science, Graduate School, Tokyo Medical and Dental University, Tokyo, Japan Department of Laboratory Medicine, University of Washington, Seattle, WA, United States
a r t i c l e
i n f o
Article history: Received 27 June 2008 Received in revised form 17 September 2008 Accepted 17 September 2008 Keywords: SCA6 Cav 2.1 gene Alternative splicing Purkinje cell Selective neurodegeneration
a b s t r a c t The ␣1A voltage-dependent calcium-channel (Cav 2.1) gene, the causative gene for spinocerebellar ataxia type 6 (SCA6), is transcribed into two major mRNA isoforms by alternative splicing at the intron 46–exon 47 boundary. One isoform has a stop codon upstream of the CAG repeat. The other “toxic isoform” has an alternatively spliced 5-nucleotide (GGCAG) insertion at the beginning of exon 47. This insertion leads to disruption of the following stop codon and transcription of a polyglutamine-encoding Cav 2.1 mRNA. The aim of our study is to investigate whether the expanded CAG repeat of exon 47 in Cav 2.1 gene increases the relative amount of the toxic isoform in Purkinje cells. Purkinje and granule cells were independently isolated in brain from subjects with SCA6 and quantified the amount of the toxic isoform mRNA by using real-time reverse transcription (RT)-PCR. We designed two sets of probe and primers: Set A for assessing total Cav 2.1 mRNA, and Set B for assessing the toxic isoform mRNA. The ratio of total Cav 2.1 mRNA to G3PDH mRNA was similar between Purkinje and granule cells in brain from both normal controls and patients with SCA6, and the ratio of toxic isoform mRNA to total Cav 2.1 mRNA did not differ between Purkinje and granule cells in control brains. However, this ratio was increased in Purkinje cells but not in granule cells in SCA6 brains. Our results suggest that toxic isoform mRNA is increased in a Purkinje cell-specific manner, which may result in SCA6-associated selective neurodegeneration. © 2008 Elsevier Ireland Ltd. All rights reserved.
Spinocerebellar ataxia type 6 (SCA6) is one of the nine inherited neurodegenerative disorders known as polyglutamine diseases, which include Huntington’s disease, Kennedy’s disease, dentatorubro-pallidoluysian atrophy, and six forms of spinocerebellar ataxia (SCA1, 2, 3, 6, 7, and 17) [8]. SCA6 is a disorder characterized by late-onset, pure cerebellar ataxia due to expansion of a CAG repeat in exon 47 of the Cav 2.1 gene [1,17]. Polyglutamine diseases, including SCA6, show selective neurodegeneration [8,16]. Although Cav 2.1 is widely distributed in various neurons throughout the central nervous system (CNS), SCA6-associated neuronal degeneration occurs predominantly in cerebellar Purkinje cells [2]. The reason for this selective neurodegeneration has not been elucidated. Alternative splicing at the intron 46–exon 47 boundary of the Cav 2.1 gene leads to two mRNA isoforms, only one of which is translated into polyglutamine-containing ␣1A calcium-channel protein [17]. The prototypal splice isoform has a stop codon at the beginning of exon 47; therefore, the CAG repeat residing downstream
∗ Corresponding author at: Department of Laboratory Medicine, HSB T163K UW Medical Center, 1959 Pacific Street, Box 357110, Seattle, WA 98195, United States. Tel.: +1 206 616 4564; fax: +1 206 616 4676. E-mail address: [email protected] (T. Tsunemi). 0304-3940/$ – see front matter © 2008 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2008.09.065
of this stop codon is not translated into polyglutamine. Because of alternative splicing, the other mRNA variant carries a 5-nucleotide (GGCAG) insertion just before the stop codon, thereby disrupting this stop codon and extending the open reading frame. Consequently, only this variant isoform (the “toxic isoform”) will contain the polyglutamine repeat, which can be toxic when the repeat expands beyond the normal range [17]. Our previous study using total cerebellar mRNA showed that patients with SCA6 had greater amounts of the toxic isoform than healthy subjects [2]. In mice, the amount of toxic isoform was greater in Purkinje cells than in granule cells [13]. From these previous observations, we hypothesized that transcription of the toxic isoform may predominate in the Purkinje cells of patients with SCA6, thereby contributing to the selective degeneration of these cells. To test this hypothesis, we used microdissection techniques and quantitative RT-PCR to compare the amount of toxic variant mRNA with the total quantity of Cav 2.1 mRNA in both Purkinje cells and granule cells. Fresh specimens of the cerebellar hemisphere from two patients with SCA6 and five neurologically normal patients were obtained at autopsy (Table 1). The two patients with SCA6 have been described previously [3]. Tissues were frozen immediately and stored at −80 ◦ C until use. The collection of tissues and their use for this study were approved by the Ethical Committee of Tokyo Medical and Dental University, Graduate School of Medicine.
T. Tsunemi et al. / Neuroscience Letters 447 (2008) 78–81
79
Table 1 Clinical and pathologic features of study subjects Subject
Age (y)/Gender
Diagnosis
CAG repeat
Duration of ataxia (y)
Purkinje cell loss
Patient 1 Patient 2 Control 1 Control 2 Control 3 Control 4 Control 5
68/female 79/female 79/female 56/female 94/male 63/male 66/male
SCA6 SCA6 Alzheimer ALS Stroke ALS ALS
22/13 22/13 13/13 13/13 13/13 13/13 13/13
18 14 Not applicable Not applicable Not applicable Not applicable Not applicable
Mild Severe None None None None None
ALS, amyotrophic lateral sclerosis. The two SCA6 patients have been described previously [5].
We purified genomic DNA from white blood cells from all seven subjects and analyzed the CAG repeat length in exon 47 of the Cav 2.1 gene. The number of CAG repeats was calculated as described [1]. PCR (Prism 7700 Sequence Detection System, Applied Biosystems, Foster City, CA) under the conditions previously described [1] was used to analyze the length of the CAG repeat in exon 47 of the Cav 2.1 gene. Portions of the cerebellar hemisphere were dissected without fixation, immediately embedded in OCT compound (Tissue-Tek, Miles Laboratories, Elkhart, IN), and frozen in liquid nitrogen. Frozen sections (thickness, 10 m) were processed, stained with 0.01% toluidine blue, washed in RNase-free water, and dried. Purkinje cells were isolated from sections by laser-capture microdissection (LCM, Leica LMD System, Nuhsbaum, McHenry, IL) according to the manufacturer’s protocol (Fig. 1A and B). To examine CAG repeat instability in Purkinje cells, 50 Purkinje cells were collected for DNA purification (DNeasy Micro kit, Qiagen, Hilden, Germany). To evaluate changes in the proportion of splice variants, Purkinje and granule cells (100 of each cell-type) were collected by LCM and subjected to RNA purification (RNeasy Micro kit, Qiagen) followed by oligo(dT)-primed cDNA synthesis by using SuperScript II (Gibco-BRL, Rockville, MD). The mRNA expression level of each individual brain was examined by TaqMan real-time quantitative PCR using probe–primer sets specific to target transcripts. PCR was performed (Prism 7700 Sequence Detection System, Applied Biosystems) according to the manufacturer’s recommendations, using cDNA derived from 100 Purkinje cells or granule cells of individual brains as a template. We selected glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as an endogenous control for normalization. We used calbindin-D28k (Calb1) and ˇ-2-chimaerin as markers for Purkinje cells and granule cells, respectively. In order to compare Cav 2.1 mRNA splice variants, two sets of specific primers and probes were designed. Set A reagents were designed to quantify total human Cav 2.1 mRNA (probe, 5 -6-carboxyfluoresin (FAM)-AGGTCTGTCCCCAAGCCTGTGTCCA6-carboxytetramethylrhodamine (TAMRA)-3 ; forward primer, 5 -CGGATTGGGTGGTCATGC-3 ; reverse primer, 5 -TGGGCCGCTACACCGAT-3 ). Set B was designed to quantify the toxic variant with the GGCAG insertion (probe, 5 -FAM-CACCGGCAGGGCAGTAGTTCCG-TAMRA-3 ; forward primer, 5 -GAGGGCCGAGAGCACATG-3 ; reverse primer, 5 -GGAGTGCTGGTACCAGATGTTG-3 ). We compared the amount of product generated with set A (total Cav 2.1 mRNA) to the amount of GAPDH mRNA and compared the amount generated with set B to that from set A to assess the ratio of toxic variant mRNA to total Cav 2.1 mRNA. Samples from each individual brain were analyzed three times independently, and the resulting values from five control brains or two SCA6 brains were averaged separately. First, we investigate the ratio of calbindin mRNA or ˇ-2chimaerin mRNA to G3PDH mRNA (endogenous controls) in Purkinje cells and granule cells. When the relative amount of calbindin mRNA in Purkinje cells was set at 1, that in granule cells was less than
0.01 (0.0078 ± 0.0023). When the relative amount of ˇ-2-chimaerin mRNA in granule cells was set at 1 that in Purkinje cells was around 0.02 (0.0204 ± 0.0044). These data suggest the specificity of the cell-type-based microdissection technique we used. The number of CAG repeat units in the Cav 2.1 gene was the same in Purkinje cells, granule cells and white blood cells from both normal subjects and those with SCA6 (data not shown). This finding suggests that the CAG repeat in the Cav 2.1 gene is stable and that cell-type-specific instability is not seen in SCA6. Next, we determined the ratio of total Cav 2.1 mRNA to G3PDH mRNA (endogenous controls) in Purkinje cells and granule cells. When the relative amount of total Cav 2.1 mRNA in Purkinje cells was set at 1, that in granule cells from normal brains was 0.12 ± 0.03 on average. In patients with SCA6, the average ratio of total Cav 2.1 mRNA to G3PDH mRNA was 0.97 ± 0.07 (Patient1; 0.91, Patient2; 1.02) in Purkinje cells and 0.12 ± 0.04 in granule cells (Patient1; 0.09, Patient2; 0.14) (Fig. 2A). The ratios of total Cav 2.1 mRNA in Purkinje cells and granule cells did not differ significantly between control and SCA6 brains (P = 0.24 for Purkinje cells; P = 1 for granule cells). Interestingly, the total Cav 2.1 mRNA expressed in Purkinje cells was about eight times that in granule cells in both normal controls and patients with SCA6. We then determined the ratio of toxic isoform mRNA to total Cav 2.1 mRNA in Purkinje cells and granule cells. The efficiency of PCR for the two mRNAs was comparable. In normal brains, the average ratio of toxic isoform mRNA to total Cav 2.1 mRNA was 0.51 ± 0.05 in Purkinje cells and 0.50 ± 0.04 in granule cells. These values suggest that the prototypal isoform lacks the polyglutamine tract when translated, and that the toxic isoform is transcribed in comparable amounts in both Purkinje cells and granule cells in normal brain. In granule cells from patients with SCA6, the relative amount of toxic isoform mRNA to total Cav 2.1 mRNA was 0.49 ± 0.05 on average (Patient1; 0.52, Patient2; 0.45). In contrast, the relative amount of toxic isoform mRNA to total Cav 2.1 mRNA in SCA6 Purkinje cells was increased to 0.98 ± 0.06 compared with that in granule cells (Patient1; 1.03, Patient2; 0.94) (Fig. 2B). Though we could not conduct statistical analysis from this small sample size, these data indicate that the toxic isoform is transcribed preferentially in SCA6 Purkinje cells. The mechanism of selective degeneration despite widespread distribution of mutant proteins remains unclear in neurodegenerative diseases including SCA6 [8]. To address this issue, we investigated postmortem SCA6 brains using LCM combined with real-time PCR methods. The CAG repeat length unstably expands in the striatum of patients’ brains with Huntington’s disease (HD) and also that of knock-in HD mouse models, suggesting that somatic repeat instability contributes to the tissue-specific patterns of HD pathogenesis [4]. However, CAG repeat instability was present but unrelated to selective cell vulnerability in the knock-in SCA1 mouse model [14]. Previously, our results showed that the CAG repeat numbers are same among different brain regions including in the deep white matter in SCA6 brains [3]. Our current study using LCM enabled us to show that CAG repeat length is stable among different
80
T. Tsunemi et al. / Neuroscience Letters 447 (2008) 78–81
Fig. 1. Microdissection of Purkinje cells in the cerebellar hemisphere. Sections were stained with 0.01% toluidine blue (A). Margins of Purkinje cells were dissected by laser (B).
types of cells, such as Purkinje cells, granule cells and white blood cells, suggesting that somatic repeat instability is not present in SCA6. It was hypothesized that abnormal expansion of the polyglutamine tract in Cav 2.1 channels causes Ca2+ channel dysfunctions, which eventually leads to Purkinje cell death. However, independent studies using different cultured cell models [6,9,10,12] and knock-in SCA6 mice [11] did not support the hypothesis that the channel dysfunction is critical step in SCA6 pathogenesis. Alternatively, the expanded polyglutamine tract which is located in C-terminus of Cav 2.1 may be toxic to Purkinje
cells. Calcium-channel protein aggregations, detected with the C-terminal antibody, are seen exclusively in the Purkinje cell cytoplasm of SCA6 brains [2]. In cell culture models, the C-terminus of Cav 2.1 was previously shown cleaved from full-length Cav 2.1, translocated to the nuclei based on the length of the polyglutamine tract, and exerted cell toxicity [5]. Taking these results together, it may be possible to speculate that polyglutamine-containing Cterminus of Cav 2.1 can form aggregations and contribute to cell toxicity in SCA6 Purkinje cells. Recent studies using LCM and real-time quantitative PCR have found that mis-splicing of Pre-mRNA causes neurodegeneration in increasing number of diseases [7]. While these studies analyzed the amount of mRNA between different cells, we investigated the expression level of splice variants between specific cells and showed toxic variants of Cav 2.1 channels are increased in a Purkinje cell-specific manner, which may facilitate subsequent degeneration. The underlying mechanism of this cell-specific alternative splicing in SCA6 is unclear but supposedly it may be caused by particular RNA binding proteins which mediate the splicing of Cav 2.1 gene. We must admit that the number of SCA6 brains (n = 2) utilized in this study is too small to draw a conclusion. This stems from the poverty of SCA6 frozen brain samples obtained by autopsy. Nevertheless, the present data clearly indicated a tendency that polyglutamine-coding Cav 2.1 mRNAs are preferentially expressed in SCA6 Purkinje cells. This implication appears important to clue the mechanism of selective neurodegeneration in SCA6, and also may provide a possible therapeutic target. Recently antisense oligonucleotides (AOs) succeeded to block the cryptic splice sites to restore normal gene expression in -thalassemia or HutchinsonGilford progeria syndrome in vitro [15]. Similarly, it may be possible to interfere toxicity of polyglutamine in SCA6 patients, if one can design and efficiently delivery AOs that can preferentially suppress the expression of Cav 2.1 toxic variants encoding polyglutamine. Acknowledgement We are grateful to Shiela Ganti for giving us critical suggestions about our manuscript. References
Fig. 2. (A) Quantitative analysis of total Cav 2.1 mRNA. About eight times more total Cav 2.1 mRNA is expressed in Purkinje cells that in granule cells from either normal controls or patients with SCA6. (B) Quantitative analysis of splice variant of Cav 2.1 mRNA. The ratio of GGCAG-inserted variant mRNA (toxic isoform) to total Cav 2.1 mRNA is almost same between Purkinje cells and granule cells from normal subjects but is increased in Purkinje cells compared with granule cells from patients with SCA6.
[1] K. Ishikawa, H. Tanaka, M. Saito, N. Ohkoshi, T. Fujita, K. Yoshizawa, T. Ikeuchi, M. Watanabe, A. Hayashi, Y. Takiyama, M. Nishizawa, I. Nakano, K. Matsubayashi, M. Miwa, S. Shoji, I. Kanazawa, S. Tsuji, H. Mizusawa, Japanese families with autosomal dominant pure cerebellar ataxia map to chromosome 19p13.1-p13.2 and are strongly associated with mild CAG expansions in the spinocerebellar ataxia type 6 gene in chromosome 19p13.1, Am. J. Hum. Genet. 61 (1997) 336–346. [2] K. Ishikawa, H. Fujigasaki, H. Saegusa, K. Ohwada, T. Fujita, H. Iwamoto, Y. Komatsuzaki, S. Toru, H. Toriyama, M. Watanabe, N. Ohkoshi, S. Shoji, I. Kanazawa,
T. Tsunemi et al. / Neuroscience Letters 447 (2008) 78–81
[3]
[4]
[5]
[6]
[7] [8] [9]
T. Tanabe, H. Mizusawa, Abundant expression and cytoplasmic aggregations of [alpha]1A voltage-dependent calcium channel protein associated with neurodegeneration in spinocerebellar ataxia type 6, Hum. Mol. Genet. 8 (1999) 1185–1193. K. Ishikawa, M. Watanabe, K. Yoshizawa, T. Fujita, H. Iwamoto, T. Yoshizawa, K. Harada, K. Nakamagoe, Y. Komatsuzaki, A. Satoh, M. Doi, T. Ogata, I. Kanazawa, S. Shoji, H. Mizusawa, Clinical, neuropathological, and molecular study in two families with spinocerebellar ataxia type 6 (SCA6), J. Neurol. Neurosurg. Psychiatry 67 (1999) 86–89. L. Kennedy, E. Evans, C.M. Chen, L. Craven, P.J. Detloff, M. Ennis, P.F. Shelbourne, Dramatic tissue-specific mutation length increases are an early molecular event in Huntington disease pathogenesis, Hum. Mol. Genet. 12 (2003) 3359–3367. H.B. Kordasiewicz, R.M. Thompson, H.B. Clark, C.M. Gomez, C-termini of P/Q-type Ca2+ channel alpha1A subunits translocate to nuclei and promote polyglutamine-mediated toxicity, Hum. Mol. Genet. 15 (2006) 1587–1599. Z. Matsuyama, M. Wakamori, Y. Mori, H. Kawakami, S. Nakamura, K. Imoto, Direct alteration of the P/Q-type Ca2+ channel property by polyglutamine expansion in spinocerebellar ataxia 6, J. Neurosci. 19 (1999) RC14. P.T. Nelson, J.N. Keller, RNA in brain disease: no longer just “the messenger in the middle”, J. Neuropathol. Exp. Neurol. 66 (2007) 461–468. H.T. Orr, H.Y. Zoghbi, Trinucleotide repeat disorders, Annu. Rev. Neurosci. 30 (2007) 575–621. E.S. Piedras-Renteria, K. Watase, N. Harata, O. Zhuchenko, H.Y. Zoghbi, C.C. Lee, R.W. Tsien, Increased expression of alpha 1A Ca2+ channel currents arising from expanded trinucleotide repeats in spinocerebellar ataxia type 6, J. Neurosci. 21 (2001) 9185–9193.
81
[10] S. Restituito, R.M. Thompson, J. Eliet, R.S. Raike, M. Riedl, P. Charnet, C.M. Gomez, The polyglutamine expansion in spinocerebellar ataxia type 6 causes a beta subunit-specific enhanced activation of P/Q-type calcium channels in Xenopus oocytes, J. Neurosci. 20 (2000) 6394–6403. [11] H. Saegusa, M. Wakamori, Y. Matsuda, J. Wang, Y. Mori, S. Zong, T. Tanabe, Properties of human Cav2.1 channel with a spinocerebellar ataxia type 6 mutation expressed in Purkinje cells, Mol. Cell Neurosci. 34 (2007) 261–270. [12] S. Toru, T. Murakoshi, K. Ishikawa, H. Saegusa, H. Fujigasaki, T. Uchihara, S. Nagayama, M. Osanai, H. Mizusawa, T. Tanabe, Spinocerebellar ataxia type 6 mutation alters P-type calcium channel function, J. Biol. Chem. 275 (2000) 10893–10898. [13] T. Tsunemi, H. Saegusa, K. Ishikawa, S. Nagayama, T. Murakoshi, H. Mizusawa, T. Tanabe, Novel Cav 2.1 splice variants isolated from Purkinje cells do not generate P-type Ca2+ current, J. Biol. Chem. 277 (2002) 7214–7221. [14] K. Watase, K.J. Venken, Y. Sun, H.T. Orr, H.Y. Zoghbi, Regional differences of somatic CAG repeat instability do not account for selective neuronal vulnerability in a knock-in mouse model of SCA1, Hum. Mol. Genet. 12 (2003) 2789–2795. [15] M. Wood, H. Yin, G. McClorey, Modulating the expression of disease genes with RNA-based therapy, PLos Genet. 3 (2007) e109. [16] M. Yamada, M. Shimohata, T. Sato, S. Tsuji, H. Takahashi, Polyglutamine disease: recent advances in the neuropathology of dentatorubral-pallidoluysian atrophy, Neuropathology 26 (2006) 346–351. [17] O. Zhuchenko, J. Bailey, P. Bonnen, T. Ashizawa, D.W. Stockton, C. Amos, W.B. Dobyns, S.H. Subramony, H.Y. Zoghbi, C.C. Lee, Autosomal dominant cerebellar ataxia (SCA6) associated with small polyglutamine expansions in the alpha 1Avoltage-dependent calcium channel, Nat. Genet. 15 (1997) 62–69.
Contents lists available at ScienceDirect
Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet
Cell-type-specific alternative splicing in spinocerebellar ataxia type 6 Taiji Tsunemi a,b,∗ , Kinya Ishikawa a , Honglian Jin a , Hidehiro Mizusawa a a b
Department of Neurology and Neurological Science, Graduate School, Tokyo Medical and Dental University, Tokyo, Japan Department of Laboratory Medicine, University of Washington, Seattle, WA, United States
a r t i c l e
i n f o
Article history: Received 27 June 2008 Received in revised form 17 September 2008 Accepted 17 September 2008 Keywords: SCA6 Cav 2.1 gene Alternative splicing Purkinje cell Selective neurodegeneration
a b s t r a c t The ␣1A voltage-dependent calcium-channel (Cav 2.1) gene, the causative gene for spinocerebellar ataxia type 6 (SCA6), is transcribed into two major mRNA isoforms by alternative splicing at the intron 46–exon 47 boundary. One isoform has a stop codon upstream of the CAG repeat. The other “toxic isoform” has an alternatively spliced 5-nucleotide (GGCAG) insertion at the beginning of exon 47. This insertion leads to disruption of the following stop codon and transcription of a polyglutamine-encoding Cav 2.1 mRNA. The aim of our study is to investigate whether the expanded CAG repeat of exon 47 in Cav 2.1 gene increases the relative amount of the toxic isoform in Purkinje cells. Purkinje and granule cells were independently isolated in brain from subjects with SCA6 and quantified the amount of the toxic isoform mRNA by using real-time reverse transcription (RT)-PCR. We designed two sets of probe and primers: Set A for assessing total Cav 2.1 mRNA, and Set B for assessing the toxic isoform mRNA. The ratio of total Cav 2.1 mRNA to G3PDH mRNA was similar between Purkinje and granule cells in brain from both normal controls and patients with SCA6, and the ratio of toxic isoform mRNA to total Cav 2.1 mRNA did not differ between Purkinje and granule cells in control brains. However, this ratio was increased in Purkinje cells but not in granule cells in SCA6 brains. Our results suggest that toxic isoform mRNA is increased in a Purkinje cell-specific manner, which may result in SCA6-associated selective neurodegeneration. © 2008 Elsevier Ireland Ltd. All rights reserved.
Spinocerebellar ataxia type 6 (SCA6) is one of the nine inherited neurodegenerative disorders known as polyglutamine diseases, which include Huntington’s disease, Kennedy’s disease, dentatorubro-pallidoluysian atrophy, and six forms of spinocerebellar ataxia (SCA1, 2, 3, 6, 7, and 17) [8]. SCA6 is a disorder characterized by late-onset, pure cerebellar ataxia due to expansion of a CAG repeat in exon 47 of the Cav 2.1 gene [1,17]. Polyglutamine diseases, including SCA6, show selective neurodegeneration [8,16]. Although Cav 2.1 is widely distributed in various neurons throughout the central nervous system (CNS), SCA6-associated neuronal degeneration occurs predominantly in cerebellar Purkinje cells [2]. The reason for this selective neurodegeneration has not been elucidated. Alternative splicing at the intron 46–exon 47 boundary of the Cav 2.1 gene leads to two mRNA isoforms, only one of which is translated into polyglutamine-containing ␣1A calcium-channel protein [17]. The prototypal splice isoform has a stop codon at the beginning of exon 47; therefore, the CAG repeat residing downstream
∗ Corresponding author at: Department of Laboratory Medicine, HSB T163K UW Medical Center, 1959 Pacific Street, Box 357110, Seattle, WA 98195, United States. Tel.: +1 206 616 4564; fax: +1 206 616 4676. E-mail address: [email protected] (T. Tsunemi). 0304-3940/$ – see front matter © 2008 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2008.09.065
of this stop codon is not translated into polyglutamine. Because of alternative splicing, the other mRNA variant carries a 5-nucleotide (GGCAG) insertion just before the stop codon, thereby disrupting this stop codon and extending the open reading frame. Consequently, only this variant isoform (the “toxic isoform”) will contain the polyglutamine repeat, which can be toxic when the repeat expands beyond the normal range [17]. Our previous study using total cerebellar mRNA showed that patients with SCA6 had greater amounts of the toxic isoform than healthy subjects [2]. In mice, the amount of toxic isoform was greater in Purkinje cells than in granule cells [13]. From these previous observations, we hypothesized that transcription of the toxic isoform may predominate in the Purkinje cells of patients with SCA6, thereby contributing to the selective degeneration of these cells. To test this hypothesis, we used microdissection techniques and quantitative RT-PCR to compare the amount of toxic variant mRNA with the total quantity of Cav 2.1 mRNA in both Purkinje cells and granule cells. Fresh specimens of the cerebellar hemisphere from two patients with SCA6 and five neurologically normal patients were obtained at autopsy (Table 1). The two patients with SCA6 have been described previously [3]. Tissues were frozen immediately and stored at −80 ◦ C until use. The collection of tissues and their use for this study were approved by the Ethical Committee of Tokyo Medical and Dental University, Graduate School of Medicine.
T. Tsunemi et al. / Neuroscience Letters 447 (2008) 78–81
79
Table 1 Clinical and pathologic features of study subjects Subject
Age (y)/Gender
Diagnosis
CAG repeat
Duration of ataxia (y)
Purkinje cell loss
Patient 1 Patient 2 Control 1 Control 2 Control 3 Control 4 Control 5
68/female 79/female 79/female 56/female 94/male 63/male 66/male
SCA6 SCA6 Alzheimer ALS Stroke ALS ALS
22/13 22/13 13/13 13/13 13/13 13/13 13/13
18 14 Not applicable Not applicable Not applicable Not applicable Not applicable
Mild Severe None None None None None
ALS, amyotrophic lateral sclerosis. The two SCA6 patients have been described previously [5].
We purified genomic DNA from white blood cells from all seven subjects and analyzed the CAG repeat length in exon 47 of the Cav 2.1 gene. The number of CAG repeats was calculated as described [1]. PCR (Prism 7700 Sequence Detection System, Applied Biosystems, Foster City, CA) under the conditions previously described [1] was used to analyze the length of the CAG repeat in exon 47 of the Cav 2.1 gene. Portions of the cerebellar hemisphere were dissected without fixation, immediately embedded in OCT compound (Tissue-Tek, Miles Laboratories, Elkhart, IN), and frozen in liquid nitrogen. Frozen sections (thickness, 10 m) were processed, stained with 0.01% toluidine blue, washed in RNase-free water, and dried. Purkinje cells were isolated from sections by laser-capture microdissection (LCM, Leica LMD System, Nuhsbaum, McHenry, IL) according to the manufacturer’s protocol (Fig. 1A and B). To examine CAG repeat instability in Purkinje cells, 50 Purkinje cells were collected for DNA purification (DNeasy Micro kit, Qiagen, Hilden, Germany). To evaluate changes in the proportion of splice variants, Purkinje and granule cells (100 of each cell-type) were collected by LCM and subjected to RNA purification (RNeasy Micro kit, Qiagen) followed by oligo(dT)-primed cDNA synthesis by using SuperScript II (Gibco-BRL, Rockville, MD). The mRNA expression level of each individual brain was examined by TaqMan real-time quantitative PCR using probe–primer sets specific to target transcripts. PCR was performed (Prism 7700 Sequence Detection System, Applied Biosystems) according to the manufacturer’s recommendations, using cDNA derived from 100 Purkinje cells or granule cells of individual brains as a template. We selected glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as an endogenous control for normalization. We used calbindin-D28k (Calb1) and ˇ-2-chimaerin as markers for Purkinje cells and granule cells, respectively. In order to compare Cav 2.1 mRNA splice variants, two sets of specific primers and probes were designed. Set A reagents were designed to quantify total human Cav 2.1 mRNA (probe, 5 -6-carboxyfluoresin (FAM)-AGGTCTGTCCCCAAGCCTGTGTCCA6-carboxytetramethylrhodamine (TAMRA)-3 ; forward primer, 5 -CGGATTGGGTGGTCATGC-3 ; reverse primer, 5 -TGGGCCGCTACACCGAT-3 ). Set B was designed to quantify the toxic variant with the GGCAG insertion (probe, 5 -FAM-CACCGGCAGGGCAGTAGTTCCG-TAMRA-3 ; forward primer, 5 -GAGGGCCGAGAGCACATG-3 ; reverse primer, 5 -GGAGTGCTGGTACCAGATGTTG-3 ). We compared the amount of product generated with set A (total Cav 2.1 mRNA) to the amount of GAPDH mRNA and compared the amount generated with set B to that from set A to assess the ratio of toxic variant mRNA to total Cav 2.1 mRNA. Samples from each individual brain were analyzed three times independently, and the resulting values from five control brains or two SCA6 brains were averaged separately. First, we investigate the ratio of calbindin mRNA or ˇ-2chimaerin mRNA to G3PDH mRNA (endogenous controls) in Purkinje cells and granule cells. When the relative amount of calbindin mRNA in Purkinje cells was set at 1, that in granule cells was less than
0.01 (0.0078 ± 0.0023). When the relative amount of ˇ-2-chimaerin mRNA in granule cells was set at 1 that in Purkinje cells was around 0.02 (0.0204 ± 0.0044). These data suggest the specificity of the cell-type-based microdissection technique we used. The number of CAG repeat units in the Cav 2.1 gene was the same in Purkinje cells, granule cells and white blood cells from both normal subjects and those with SCA6 (data not shown). This finding suggests that the CAG repeat in the Cav 2.1 gene is stable and that cell-type-specific instability is not seen in SCA6. Next, we determined the ratio of total Cav 2.1 mRNA to G3PDH mRNA (endogenous controls) in Purkinje cells and granule cells. When the relative amount of total Cav 2.1 mRNA in Purkinje cells was set at 1, that in granule cells from normal brains was 0.12 ± 0.03 on average. In patients with SCA6, the average ratio of total Cav 2.1 mRNA to G3PDH mRNA was 0.97 ± 0.07 (Patient1; 0.91, Patient2; 1.02) in Purkinje cells and 0.12 ± 0.04 in granule cells (Patient1; 0.09, Patient2; 0.14) (Fig. 2A). The ratios of total Cav 2.1 mRNA in Purkinje cells and granule cells did not differ significantly between control and SCA6 brains (P = 0.24 for Purkinje cells; P = 1 for granule cells). Interestingly, the total Cav 2.1 mRNA expressed in Purkinje cells was about eight times that in granule cells in both normal controls and patients with SCA6. We then determined the ratio of toxic isoform mRNA to total Cav 2.1 mRNA in Purkinje cells and granule cells. The efficiency of PCR for the two mRNAs was comparable. In normal brains, the average ratio of toxic isoform mRNA to total Cav 2.1 mRNA was 0.51 ± 0.05 in Purkinje cells and 0.50 ± 0.04 in granule cells. These values suggest that the prototypal isoform lacks the polyglutamine tract when translated, and that the toxic isoform is transcribed in comparable amounts in both Purkinje cells and granule cells in normal brain. In granule cells from patients with SCA6, the relative amount of toxic isoform mRNA to total Cav 2.1 mRNA was 0.49 ± 0.05 on average (Patient1; 0.52, Patient2; 0.45). In contrast, the relative amount of toxic isoform mRNA to total Cav 2.1 mRNA in SCA6 Purkinje cells was increased to 0.98 ± 0.06 compared with that in granule cells (Patient1; 1.03, Patient2; 0.94) (Fig. 2B). Though we could not conduct statistical analysis from this small sample size, these data indicate that the toxic isoform is transcribed preferentially in SCA6 Purkinje cells. The mechanism of selective degeneration despite widespread distribution of mutant proteins remains unclear in neurodegenerative diseases including SCA6 [8]. To address this issue, we investigated postmortem SCA6 brains using LCM combined with real-time PCR methods. The CAG repeat length unstably expands in the striatum of patients’ brains with Huntington’s disease (HD) and also that of knock-in HD mouse models, suggesting that somatic repeat instability contributes to the tissue-specific patterns of HD pathogenesis [4]. However, CAG repeat instability was present but unrelated to selective cell vulnerability in the knock-in SCA1 mouse model [14]. Previously, our results showed that the CAG repeat numbers are same among different brain regions including in the deep white matter in SCA6 brains [3]. Our current study using LCM enabled us to show that CAG repeat length is stable among different
80
T. Tsunemi et al. / Neuroscience Letters 447 (2008) 78–81
Fig. 1. Microdissection of Purkinje cells in the cerebellar hemisphere. Sections were stained with 0.01% toluidine blue (A). Margins of Purkinje cells were dissected by laser (B).
types of cells, such as Purkinje cells, granule cells and white blood cells, suggesting that somatic repeat instability is not present in SCA6. It was hypothesized that abnormal expansion of the polyglutamine tract in Cav 2.1 channels causes Ca2+ channel dysfunctions, which eventually leads to Purkinje cell death. However, independent studies using different cultured cell models [6,9,10,12] and knock-in SCA6 mice [11] did not support the hypothesis that the channel dysfunction is critical step in SCA6 pathogenesis. Alternatively, the expanded polyglutamine tract which is located in C-terminus of Cav 2.1 may be toxic to Purkinje
cells. Calcium-channel protein aggregations, detected with the C-terminal antibody, are seen exclusively in the Purkinje cell cytoplasm of SCA6 brains [2]. In cell culture models, the C-terminus of Cav 2.1 was previously shown cleaved from full-length Cav 2.1, translocated to the nuclei based on the length of the polyglutamine tract, and exerted cell toxicity [5]. Taking these results together, it may be possible to speculate that polyglutamine-containing Cterminus of Cav 2.1 can form aggregations and contribute to cell toxicity in SCA6 Purkinje cells. Recent studies using LCM and real-time quantitative PCR have found that mis-splicing of Pre-mRNA causes neurodegeneration in increasing number of diseases [7]. While these studies analyzed the amount of mRNA between different cells, we investigated the expression level of splice variants between specific cells and showed toxic variants of Cav 2.1 channels are increased in a Purkinje cell-specific manner, which may facilitate subsequent degeneration. The underlying mechanism of this cell-specific alternative splicing in SCA6 is unclear but supposedly it may be caused by particular RNA binding proteins which mediate the splicing of Cav 2.1 gene. We must admit that the number of SCA6 brains (n = 2) utilized in this study is too small to draw a conclusion. This stems from the poverty of SCA6 frozen brain samples obtained by autopsy. Nevertheless, the present data clearly indicated a tendency that polyglutamine-coding Cav 2.1 mRNAs are preferentially expressed in SCA6 Purkinje cells. This implication appears important to clue the mechanism of selective neurodegeneration in SCA6, and also may provide a possible therapeutic target. Recently antisense oligonucleotides (AOs) succeeded to block the cryptic splice sites to restore normal gene expression in -thalassemia or HutchinsonGilford progeria syndrome in vitro [15]. Similarly, it may be possible to interfere toxicity of polyglutamine in SCA6 patients, if one can design and efficiently delivery AOs that can preferentially suppress the expression of Cav 2.1 toxic variants encoding polyglutamine. Acknowledgement We are grateful to Shiela Ganti for giving us critical suggestions about our manuscript. References
Fig. 2. (A) Quantitative analysis of total Cav 2.1 mRNA. About eight times more total Cav 2.1 mRNA is expressed in Purkinje cells that in granule cells from either normal controls or patients with SCA6. (B) Quantitative analysis of splice variant of Cav 2.1 mRNA. The ratio of GGCAG-inserted variant mRNA (toxic isoform) to total Cav 2.1 mRNA is almost same between Purkinje cells and granule cells from normal subjects but is increased in Purkinje cells compared with granule cells from patients with SCA6.
[1] K. Ishikawa, H. Tanaka, M. Saito, N. Ohkoshi, T. Fujita, K. Yoshizawa, T. Ikeuchi, M. Watanabe, A. Hayashi, Y. Takiyama, M. Nishizawa, I. Nakano, K. Matsubayashi, M. Miwa, S. Shoji, I. Kanazawa, S. Tsuji, H. Mizusawa, Japanese families with autosomal dominant pure cerebellar ataxia map to chromosome 19p13.1-p13.2 and are strongly associated with mild CAG expansions in the spinocerebellar ataxia type 6 gene in chromosome 19p13.1, Am. J. Hum. Genet. 61 (1997) 336–346. [2] K. Ishikawa, H. Fujigasaki, H. Saegusa, K. Ohwada, T. Fujita, H. Iwamoto, Y. Komatsuzaki, S. Toru, H. Toriyama, M. Watanabe, N. Ohkoshi, S. Shoji, I. Kanazawa,
T. Tsunemi et al. / Neuroscience Letters 447 (2008) 78–81
[3]
[4]
[5]
[6]
[7] [8] [9]
T. Tanabe, H. Mizusawa, Abundant expression and cytoplasmic aggregations of [alpha]1A voltage-dependent calcium channel protein associated with neurodegeneration in spinocerebellar ataxia type 6, Hum. Mol. Genet. 8 (1999) 1185–1193. K. Ishikawa, M. Watanabe, K. Yoshizawa, T. Fujita, H. Iwamoto, T. Yoshizawa, K. Harada, K. Nakamagoe, Y. Komatsuzaki, A. Satoh, M. Doi, T. Ogata, I. Kanazawa, S. Shoji, H. Mizusawa, Clinical, neuropathological, and molecular study in two families with spinocerebellar ataxia type 6 (SCA6), J. Neurol. Neurosurg. Psychiatry 67 (1999) 86–89. L. Kennedy, E. Evans, C.M. Chen, L. Craven, P.J. Detloff, M. Ennis, P.F. Shelbourne, Dramatic tissue-specific mutation length increases are an early molecular event in Huntington disease pathogenesis, Hum. Mol. Genet. 12 (2003) 3359–3367. H.B. Kordasiewicz, R.M. Thompson, H.B. Clark, C.M. Gomez, C-termini of P/Q-type Ca2+ channel alpha1A subunits translocate to nuclei and promote polyglutamine-mediated toxicity, Hum. Mol. Genet. 15 (2006) 1587–1599. Z. Matsuyama, M. Wakamori, Y. Mori, H. Kawakami, S. Nakamura, K. Imoto, Direct alteration of the P/Q-type Ca2+ channel property by polyglutamine expansion in spinocerebellar ataxia 6, J. Neurosci. 19 (1999) RC14. P.T. Nelson, J.N. Keller, RNA in brain disease: no longer just “the messenger in the middle”, J. Neuropathol. Exp. Neurol. 66 (2007) 461–468. H.T. Orr, H.Y. Zoghbi, Trinucleotide repeat disorders, Annu. Rev. Neurosci. 30 (2007) 575–621. E.S. Piedras-Renteria, K. Watase, N. Harata, O. Zhuchenko, H.Y. Zoghbi, C.C. Lee, R.W. Tsien, Increased expression of alpha 1A Ca2+ channel currents arising from expanded trinucleotide repeats in spinocerebellar ataxia type 6, J. Neurosci. 21 (2001) 9185–9193.
81
[10] S. Restituito, R.M. Thompson, J. Eliet, R.S. Raike, M. Riedl, P. Charnet, C.M. Gomez, The polyglutamine expansion in spinocerebellar ataxia type 6 causes a beta subunit-specific enhanced activation of P/Q-type calcium channels in Xenopus oocytes, J. Neurosci. 20 (2000) 6394–6403. [11] H. Saegusa, M. Wakamori, Y. Matsuda, J. Wang, Y. Mori, S. Zong, T. Tanabe, Properties of human Cav2.1 channel with a spinocerebellar ataxia type 6 mutation expressed in Purkinje cells, Mol. Cell Neurosci. 34 (2007) 261–270. [12] S. Toru, T. Murakoshi, K. Ishikawa, H. Saegusa, H. Fujigasaki, T. Uchihara, S. Nagayama, M. Osanai, H. Mizusawa, T. Tanabe, Spinocerebellar ataxia type 6 mutation alters P-type calcium channel function, J. Biol. Chem. 275 (2000) 10893–10898. [13] T. Tsunemi, H. Saegusa, K. Ishikawa, S. Nagayama, T. Murakoshi, H. Mizusawa, T. Tanabe, Novel Cav 2.1 splice variants isolated from Purkinje cells do not generate P-type Ca2+ current, J. Biol. Chem. 277 (2002) 7214–7221. [14] K. Watase, K.J. Venken, Y. Sun, H.T. Orr, H.Y. Zoghbi, Regional differences of somatic CAG repeat instability do not account for selective neuronal vulnerability in a knock-in mouse model of SCA1, Hum. Mol. Genet. 12 (2003) 2789–2795. [15] M. Wood, H. Yin, G. McClorey, Modulating the expression of disease genes with RNA-based therapy, PLos Genet. 3 (2007) e109. [16] M. Yamada, M. Shimohata, T. Sato, S. Tsuji, H. Takahashi, Polyglutamine disease: recent advances in the neuropathology of dentatorubral-pallidoluysian atrophy, Neuropathology 26 (2006) 346–351. [17] O. Zhuchenko, J. Bailey, P. Bonnen, T. Ashizawa, D.W. Stockton, C. Amos, W.B. Dobyns, S.H. Subramony, H.Y. Zoghbi, C.C. Lee, Autosomal dominant cerebellar ataxia (SCA6) associated with small polyglutamine expansions in the alpha 1Avoltage-dependent calcium channel, Nat. Genet. 15 (1997) 62–69.