Proteomic analysis of the rat cerebellar flocculus during vestibular compensation

Transcription

1 Journal of Vestibular Research 19 (2009) DOI /VES IOS Press Proteomic analysis of the rat cerebellar flocculus during vestibular compensation Masahiko Fukasawa a,b, Kazuki Okamoto a,, Manabu Nakamura b, Koshi Mikami b, Sonoko Shimada b, Yasuhiko Tanaka b, Kouhei Nagai a, Mitsumi Arito a, Manae S. Kurokawa a, Kayo Masuko a, Naoya Suematsu a, Izumi Koizuka b and Tomohiro Kato a a Clinical Proteomics & Molecular Medicine, St. Marianna University Graduate School of Medicine, Kawasaki, Kanagawa , Japan b Department of Otolaryngology, St. Marianna University School of Medicine, Kawasaki, Kanagawa , Japan Received 16 October 2009 Accepted 4 February 2010 Abstract. Unilateral labyrinthectomy (UL) in rats is used as a human vertigo model. In this model, spontaneous nystagmus and dysequilibrium caused by UL are ameliorated within hours. The amelioration, termed vestibular compensation (VC), is long lasting. Although cerebellar flocculi have been reported to be involved in VC, the molecular mechanisms behind VC are unknown. In this study, we used 2D-DIGE to detect protein changes in flocculi during acute (48 hours) and chronic (1 week) stages of VC. We found 99 out of 967 protein spots that showed significant changes in their intensities. Of the 99 spots, 45 spots (ipsilateral side, 15; contralateral side, 30) changed unilaterally during the acute stage, whereas 46 spots (ipsilateral side, 21; contralateral side, 25) changed unilaterally during the chronic stage. Thus, the acute compensation mechanism is more complicated in the contralateral flocculus than in the ipsilateral flocculus. Using MALDI-TOF MS, we identified 10 proteins out of the 12 protein spots. Of these, 3 proteins involved in synaptic transmission, neuronal filament formation and vesicular transport, respectively, demonstrated altered expression only in the acute stage. Our results enhance the understanding of the role of the cerebellar flocculi in VC generation. Keywords: Unilateral labyrinthectomy, vestibular compensation, cerebellar flocculus, proteomics List of abbreviations Aco2: mitochondrial aconitate hydratase; Atp5b: mitochondrial ATP synthase subunit β; Atp5f1: mitochondrial ATP synthase subunit b; ER: endoplasmic reticulum; Hnrpk: heterogeneous nuclear ribonucleoprotein K; Inexa: alpha-internexin; Corresponding author: Kazuki Okamoto, Clinical Proteomics & Molecular Medicine, St Marianna University Graduate School of Medicine, Sugao, Miyamae-ku, Kawasaki, Kanagawa , Japan. Tel.: ; Fax: ; k2oka@marianna-u.ac.jp. Lmnb1: lamin-b1; MVN: medial vestibular nuclei; Nsf: N-ethylmaleimide-sensitive fusion protein; Pdia3: protein disulfide-isomerase A3; PTM: post-translational modification; Pygb: brain glycogen phosphorylase; RSI: relative spot intensity; UL: Unilateral labyrinthectomy; VC: vestibular compensation; VNC: vestibular nucleus complex; 2D-DIGE: 2-dimensional differential gel electrophoresis; 2-DE: 2-dimensional electrophoresis. ISSN /09/$ IOS Press and the authors. All rights reserved

2 84 M. Fukasawa et al. / Proteomic analysis of the rat cerebellar flocculus during vestibular compensation 1. Introduction Unilateral labyrinthectomy (UL) in rats is a model of human vertigo [6,16]. UL involves a static equilibrium disorder of the vestibuloocular and vestibulospinal systems, which presents clinically as spontaneous nystagmus and dysequilibrium [6,16]. These symptoms are ameliorated within hours [10], although the disordered inner ear is no longer able to send signals to the vestibular nucleus complex (VNC) [6,10,16,23]. This amelioration is called vestibular compensation (VC). The process of VC is usually divided into three stages: the non-compensation stage just after UL, the acute stage at 6 72 hours, and the chronic stage at 4 14 days after UL [16]. VC is accomplished via plasticity of the central nervous system (CNS) [6,10,16,23]. Recently, it was reported that the plasticity of the vestibulocerebellum also contributes to VC [14,18]. The vestibulocerebellum, which consists of the flocculus-paraflocculus, uvula, and nodulus, controls the vestibulo-ocular reflex. Excitatory signals delivered to the vestibulocerebellum pass through mossy fiber synapses and climbing fiber synapses via glutamate [12,14]. These excitatory signals stimulate Purkinje cells in the vestibulocerebellum, which suppress VNC neurons using GABA [11,14]. After UL, the inhibitory signals passing from the Purkinje cells to the VNC contribute to the recovery of the balance of spontaneous discharges between the bilateral VNC [5,15,16]. Thus, the contribution of the vestibulocerebellum is essential to VC. To achieve VC, the bilateral cerebellar flocculi are thought to integrate different neural signals delivered from the bilateral VNC. Indeed, cauterization of the vestibulocerebellum or disruption of the climbing fibers to the cerebellar flocculi results in a severe delay in VC [1,5]. Furthermore, even after VC is achieved, VC is broken down by flocculectomy [18]. Thus, the cerebellar flocculi play important roles in VC. Recently, Paterson et al. reported that 26 proteins showed a significantly altered expression in the medial vestibular nuclei in the VNC at 1 week after UL [21]. Similarly, Ban et al. reported the new expression of 8 proteins and the down-regulation of 5 proteins in the VNC following UL [2]. Kitahara et al. reported the down-regulation of the cerebellum-specific glutamate receptor δ2 and the up-regulation of the protein phosphatase 2A-β in the Purkinje cell layer of the vestibulocerebellum via a differential display-pcr method [16]. Despite the above reports, the molecular mechanisms behind VC that occur in the cerebellar flocculi are not fully understood. To promote understanding of these UL mechanisms, we investigated the changes that occur in proteins in the bilateral cerebellar flocculi at 48 hours (acute stage) and 1 week (chronic stage) after UL in rats, using 2-dimensional differential gel electrophoresis (2D-DIGE). Furthermore, using mass spectrometry, we identified 10 proteins that are involved in VC. 2. Material and methods 2.1. Animal treatment All animal experiments were approved by the Animal Care Committee of St. Marianna University School of Medicine. Twelve male Sprague-Dawley rats (7 weeks old; g) were used. They were divided into four groups (UL- and sham-operated rats at 48 hours or 1 week after surgery). All surgeries were performed between 10:00 and 15:00 hours. Rats were anesthetized with pentobarbital (50 mg/kg, i.p.) and underwent a right UL, using a retro-auricular approach. Local anesthesia with 1% lidocaine was used in the wound margin. After removal of the tympanic membrane, malleus, and incus, the vestibule just above the ampullae of the horizontal and anterior semicircular canals was drilled out. After aspiration of labyrinthine fluids and the membranous labyrinth from the drilled vestibule and the ventral portion of the oval window, the labyrinth was rinsed with 0.1 ml of absolute ethanol, and perfused through the ventral portion of the oval window and the drilled vestibule [12]. For the sham operation group, rats had their right retro auricular skin incised in the same way as for the UL with preservation of the tympanic membrane, malleus, and incus [12]. Histologically, we confirmed that the saccular maculae in the semicircular canals of the inner ear had been destroyed without regeneration of vestibular hair cells in the UL-rats, but that the vestibular ganglion remained intact (Fig. 1A). In the sham-operated rats, all the structures were intact (Fig. 1B). We observed spontaneous nystagmus and postural asymmetry such as ocular tilt and head deviation after recovery from anesthesia only in the UL-rats and not in the sham-operated rats (data not shown). At 48 hours after UL, we observed the disappearance of the spontaneous nystagmus and postural asymmetry. Since the disappearance of the symptoms was brought about by VC, we used only the rats which showed the disappearance at 48 hours after UL in following experiments.

3 M. Fukasawa et al. / Proteomic analysis of the rat cerebellar flocculus during vestibular compensation 85 (A) (B) Fig. 1. Histology of the rat vestibular endo-organs after unilateral labyrinthectomy (A) and sham operation (B). Histological examination after unilateral labyrinthectomy showed that surgical destruction of the membranous labyrinth (S) had been achieved, that no vestibular hair cells had regenerated, and that the vestibular ganglion (G) was intact. S: Saccular macula, G: Ganglion cells, Scale bar = 500 µm Sample preparation At 48 hours or 1 week after the surgeries, the whole brain was removed and the bilateral cerebellar flocculi were dissected. The dissected samples were immediately frozen in liquid nitrogen and stored at 80 C until use. The floccular samples were homogenized on ice in ice-cold phosphate buffered saline (PBS) using a glass- Teflon homogenizer. The lysate was diluted in 10 fold volumes of cell lysis buffer (30 mm Tris-HCl, ph 8.0, 7 M urea, 2 M thiourea and 4% 3-[(3-cholamidopropyl)dimmethylammonio]propanesulfonate (CHAPS)). Whole cell extract was obtained by centrifugation at 14,000 g for 30 min at 4 C as described previously [22] and stored at 80 C until use. Protein concentrations were determined by the Bradford method using bovine serum albumin as a standard D-DIGE analysis and protein identification The procedures for 2D-DIGE and protein identification were described previously [22]. Briefly, 2.5 µg of each floccular protein sample from rats with UL (ULrats) or sham-operated rats was labeled with Cy5 saturation dye. Similarly, 2.5 µg of a standard sample, a mixture consisting of the equal amounts of all the samples, was labeled with Cy3 saturation dye. Then, the Cy3-labeled internal standard sample and each of the individual Cy5-labeled protein samples were mixed and subjected to 2D-DIGE. The separated proteins were detected using an image analyzer (Typhoon 9400 Imager; GE Healthcare UK, Buckinghamshire, England). The acquired gel images were analyzed using the Progenesis software (PerkinElmer, Waltham, Massachusetts, USA), which can measure at least a 1.1-fold difference in the samples. For the identification of proteins, gel fragments that corresponded to protein spots of interest were recovered. Peptides, generated by in-gel digestion with trypsin, were extracted from the gel fragments and were subjected to MALDI-TOF/TOF mass spectrometry (Ultraflex; Bruker Daltonics, Bremen, Germany). The determined peptide masses were compiled to allow searches of the National Center for Biotechnology Information (NCBI) protein database using the Mascot software program (Matrix Science, London, UK) Statistical analysis Statistical significance of the normalized intensity of each sample was calculated by using the Student s t-test (comparing one sample with others in each group) and one-way ANOVA (comparing one sample with all other samples). A value of p<0.05 was considered to be statistically significant. To be assigned as a protein of interest, the proteins had to comply with two additional requirements: (1) average changes 1.3-fold and (2) an appearance on all of the gel images [3].

4 86 M. Fukasawa et al. / Proteomic analysis of the rat cerebellar flocculus during vestibular compensation 3. Results 3.1. Proteome analysis of cerebellar flocculi after unilateral labyrinthectomy (UL) by 2D-DIGE To clarify the molecular alterations that occur in the cerebellar flocculi during vestibular compensation (VC) after UL, we compared the profiles of proteins extracted from cerebellar flocculi of the UL-rats in the acute (48 hours) and chronic (1 week) stages of VC with those from the sham-operated rats. Specifically, protein samples extracted from the ipsilateral and contralateral flocculus of the UL- and sham-operated rats at 48 hours or 1 week after the surgeries were labeled with Cy5 fluorescent dye (n = 3 for each group, n = 24 in total). As a standard sample, equal amounts of the 24 protein samples were mixed and labeled with Cy3. Then, each of the Cy5 labeled samples together with the Cy3 labeled standard sample for comparison was loaded and separated by 2D-DIGE (Fig. 2). A total of 967 protein spots were detected on the 2D-DIGE. Intensities of the detected protein spots were normalized to the intensities of the standard sample. We then compared the normalized intensities of the spots from the UL-rats to those of the corresponding spots from the sham-operated rats to calculate the relative spot intensity (RSI). For each of the spots in the UL-rats, the spot intensity of the identical side and the identical sampling time in the sham-operated rats was defined as 1.0. Thus, we selected protein spots with RSIs that were 1.3 or 1/1.3 in the UL-rats. This procedure highlighted 120 protein spots in total (increased, 68; decreased, 52) as shown in Table 1. After excluding 21 spots that overlapped between the bilateral flocculi or between the acute and chronic stages, we obtained 99 protein spots with RSIs that were 1.3 or 1/1.3 in the UL-rats compared to the sham-operated rats during VC. Next, we assessed the chronological changes in the intensity of the 99 protein spots during VC. As summarized in Table 2, we classified the spots initially according to the changes in the acute stage and then according to changes in the chronic stage. At 48 hours after UL, only 4 (4%) of the 99 protein spots showed bilateral RSI changes, 45 spots (45%) showed unilateral RSI changes, and the remaining 50 spots (51%) showed similar expression levels. Among the 45 spots that showed RSI changes in unilateral side at 48 hours, 30 spots showed contralateral RSI changes and the remaining 15 spots showed ipsilateral RSI changes. At 1 week, 24 of 30 spots showed contralateral RSI Table 1 Number of protein spots changed in rat cerebellar flocculus during VC 48 hours 1 week RSI ipsi contra ipsi contra 2.0 x x < x < /1.5 < x 1/ /2.0 < x 1/ x 1/ RSI: relative spot intensity. changes and 8 of 15 spots showed ipsilateral changes that returned to a similar level as that in the shamoperated rats. The temporal increase or decrease of the former 24 spots in the contralateral flocculus and the latter 8 spots in the ipsilateral flocculus may be involved in the development of the acute stage of VC. This indicates that the contralateral cerebellar flocculus, as well as the ipsilateral cerebellar flocculus may play important roles to achieve the acute stage of VC. On the other hand, of the 50 spots that showed RSI changes at 1 week and not at 48 hours, the changes of 46 spots were observed unilaterally. Of the 46 spots, the RSI changes of 25 and 21 spots were observed in the contralateral and ipsilateral flocculus, respectively. These results suggest that the protein expression that changed after UL are mostly different between the ipsilateral and contralateral flocculus, and a similar number of proteins changed expression in each side at the chronic stage Identification of the protein spots by MALDI-TOF MS Next, we investigated the identity of the proteins that corresponded to the changed spots (Fig. 3A). Proteins extracted from the UL-rats were labeled with Cy3 and separated by 2-dimensional electrophoresis (2-DE). The changed spots of interest were picked from the 2D gels. Then, peptides were generated by in-gel digestion with trypsin and were subjected to MALDI- TOF/TOF mass spectrometry (Fig. 3B, 3C). We identified 10 proteins from the 12 spots as shown in Table Alteration of the identified proteins during VC To investigate the contributions of the identified proteins to VC, we compared the changes in protein expression between the acute and chronic stages of VC and between the bilateral cerebellar flocculi. As shown in the upper panels of Fig. 4A, lamin-b1 (Lmnb1) and

5 M. Fukasawa et al. / Proteomic analysis of the rat cerebellar flocculus during vestibular compensation 87 Fig. 2. 2D-DIGE of the floccular proteins extracted from the UL-rats (A, B, E, F) or the sham-operated rats (C, D, G, H). Proteins extracted from cerebellar flocculi were labeled with CyDye and separated by IEF (ph 3-11) and then by 12.5% SDS-PAGE. (A) ipsilateral/48 hours/ul-rat, (B) contralateral/48 hours/ul-rat, (C) ipsilateral/48 hours/sham-operated rat, (D) contralateral/48 hours/sham-operated rat, (E) ipsilateral/1 week/ul-rat, (F) contralateral/1 week/ul-rat, (G) ipsilateral/1 week /sham-operated rat, (H) contralateral/1 week /sham-operated rat. MW: molecular weight markers.

6 88 M. Fukasawa et al. / Proteomic analysis of the rat cerebellar flocculus during vestibular compensation Table 2 Chronological changes in the 99 protein spots 48 hours 1 week Change The number Ipsi Contra The number Ipsi Contra The number of spots of spots of spots Bilateral 4 H H 1 L L 1 L L 2 L S 1 S S 1 H L 1 L S 1 Unilateral 45 H S 7 H H 1 L L 1 L S 1 S H 1 S S 3 L S 8 H S 1 L S 3 S S 4 S H 16 H H 2 H S 2 L S 1 S H 1 S S 10 S L 14 L L 1 L H 1 H S 1 L S 1 S H 1 S S 9 No change 50 S S 50 H H 2 at 48 hours L L 1 L H 1 H S 16 L S 5 S H 14 S L 11 H: higher than the level of sham operated rats ( x 1.3). S: same as the level of sham operated rats (1/1.3 < x < 1.3). L: lower than the level of sham operated rats ( x 1/1.3). Table 3 Identified protein spots by MALDI-TOF MS Spot no. Protein MW(kDa)/pI MOWSE Hit/ Accession Protein (observed) (calculated) score submitted no. function 132 lamin-b1 70.6/ / /13 P heterogeneous nuclear ribonucleoprotain 66.6/ / /5 P61980 transcription-related K 383 heterogeneous nuclear ribonucleoprotain 66.6/ / /9 P61980 K 092 brain glycogen phosphorylase 80.3/ / /22 P mitochondrial aconitate hydratase 82.6/ / /15 Q9ER34 energy production-related 086 mitochondrial ATP synthase F1 57.2/ / /19 P10719 complex subunit β 827 mitochondrial ATP synthase F0 27.4/ / /5 P19511 complex subunit b 365 alpha-internexin 66.7/ / /24 P23565 neuronal filament formation 008 alpha-internexin 66.7/ / /25 P protein disulfide-isomerase A3 66.1/ / /12 P11598 vesicular transport 315 N-ethylmaleimide sensitive factor 74.2/ / /14 Q9QUL6 synaptic transmission 304 serum albumin 74.2/ / /8 P02770 blood protein

7 M. Fukasawa et al. / Proteomic analysis of the rat cerebellar flocculus during vestibular compensation 89 (A) (C) (B) Fig. 3. (A) Locations of picked and identified protein spots. Proteins extracted from ipsilateral cerebellar flocculus of UL-rats were labeled with CyDye and separated by 2-DE. Among the significantly changed spots, picked and identified spots are marked by a circle. The spot numbers shown in Table 3 are indicated nearby. Broken line square is magnified in Fig. 4B. (B) An example of an MS spectrogram of the trypsin digest of spot No The 25 peptide ion peaks were matched for peptide mass fingerprinting against the National Center for Biotechnology Information (NCBI) protein database, resulting in the identification of alpha-internexin with an MOWSE score of 520. (C) Representative MS/MS spectrum of the peptide ion peak with the m/z values of ( ), ( ), ( ). heterogeneous nuclear ribonucleoprotein K (Hnrpk), both of which are known to be directly involved in gene transcription, showed similar chronological changes. In the ipsilateral flocculus of the UL-rats, the expression levels of both proteins were decreased at 48 hours and then increased to a high level at 1 week. In contrast, in the contralateral flocculus, the expression levels of both proteins were higher only slightly at 48 hours and decreased to lower levels even though they showed 1/1.3 < RSI < 1.3. The inverse changes of Lmnb1 and Hnrpk between the two time points suggest different roles between the acute and chronic stages of VC. The changes in the expression of brain glycogen phosphorylase (Pygb), mitochondrial aconitate hydratase (Aco2), mitochondrial ATP synthase F1 complex subunit β (Atp5b), and mitochondrial ATP synthase F0 complex subunit B1 (Atp5f1) are shown in the middle panels of Fig. 4A. All are enzymes that are necessary for energy production. However, their changes in expression were not similar. For example, Atp5b was decreased in the chronic stage of VC, while that of Atp5f1 was increased. These enzymes may have other biological activities besides energy production. As shown in the lower panels of Fig. 4A, the expression of other proteins such as alpha-internexin (Inexa), N-ethylmaleimide-sensitive fusion protein (Nsf), and protein disulfide-isomerase A3 (Pdia3) changed in the acute stage of VC and returned to similar levels as in the sham-operated rats in the chronic stage of VC. Thus, these proteins may influence VC in the acute

8 90 M. Fukasawa et al. / Proteomic analysis of the rat cerebellar flocculus during vestibular compensation (A) (B) Fig. 4. (A) Chronological changes in RSI of the identified protein spots from the ipsilateral (open circles) and contralateral (closed circles) cerebellar flocculi are shown in logarithmic scale (X-axis). (B) Location (upper) and chronological (lower) changes in RSI of the post-translationally modified Nsf molecules from the ipsilateral (open circles) and contralateral (closed circles) cerebellar flocculi are shown in logarithmic scale (X-axis). Upper panel is the magnified area of the square framed by broken lines shown in Fig. 3A.

9 M. Fukasawa et al. / Proteomic analysis of the rat cerebellar flocculus during vestibular compensation 91 stage only. Among these 3 proteins, Nsf and Inexa are especially interesting, since they are directly involved in the morphogenesis of neurons and in the function of GABA receptors, as discussed below. From the 2-DE analysis (Fig. 3A), we found that the protein spot identified as Nsf (No. 315) was a member of a series of 4 protein spots with very similar molecular weights, but different pi values (Fig. 3A, broken line square). The upper panel of Fig. 4B shows a magnified view of the series of 4 protein spots. Since we confirmed that the far left spot (No. 308) was Nsf by mass spectrometry, these 4 spots were all considered to be Nsf. We hypothesized that the pi shifts of the protein spots without apparent changes in their molecular weights were caused by post-translational modification (PTM), although the type of PTM that occurred is unknown. Next, we analyzed the time course differences between these 4 spots. As shown in the lower panels of Fig. 4B, the left 3 Nsf spots (No. 308, 309, and 303) changed very little, while the intensity of the far right spot (No. 315) was significantly lower in the acute stage. This indicates that not only expression and/or degradation of proteins, but also PTM may be involved in the generation of VC. 4. Discussion Several studies have shown that cerebellar flocculi are involved in the recovery from static equilibrium disorder in the acute stage of VC [5,11,14] and in the maintenance of the compensation in the chronic stage of VC [16,17]. In this study, using 2D-DIGE proteomic analysis, we investigated alterations of protein expression in the contralateral and ipsilateral flocculi during VC. We found 99 protein spots that showed significant changes in RSI after UL (Table 1). In the acute stage of VC, 49 of the 99 protein spots showed changes in RSI. However, only 4 protein spots showed bilateral RSI changes. Of interest, 30 spots showed changes in RSI only in the contralateral flocculus, and the remaining 15 spots showed changes in RSI only in the ipsilateral flocculus (Table 2). These results indicate that distinct proteins contribute to the generation of the acute stage of VC between the ipsilateral and contralateral sides. Further, in the contralateral flocculus, the number of spots with changes in RSI was twice that on the ipsilateral side, which suggests more complicated regulatory mechanisms in the contralateral cerebellar flocculus than in the ipsilateral cerebellar flocculus to achieve the acute phase of VC. To achieve the chronic stage of VC, an additional 50 protein spots showed changes as summarized in Table 2. Similar to the acute phase, 45 of the 50 protein spots showed RSI changes unilaterally. These data demonstrate that the bilateral cerebellar flocculi maintain the chronic stage using distinct protein groups. In contrast with the acute stage, almost the same numbers of protein spots (21 vs. 25) displayed RSI changes in the bilateral cerebellar flocculi during the chronic stage. Therefore, the bilateral cerebellar flocculi may use equally complicated, but different processes to maintain the chronic compensation. Further, using 2-DE followed by MALDI-TOF MS analysis, we identified 10 proteins and analyzed the alterations of these protein spots during VC. Among the identified proteins, Lmnb1 and Hnrpk are known to have significant activity in transcription. Lmnb1 is a component of the nuclear lamina. Lmnb1 provides a framework for the nuclear envelope and also interacts with chromatin [19]. Thus, Lmnb1 is involved in chromatin remodeling and in the regulation of transcription. Hnrpk, an intranuclear pre-mrna-binding protein, has many functions in gene transcription, processing, and translation [4]. Interestingly, our proteomic analysis revealed that these two proteins showed a similar time course of expression. In addition, both proteins showed inverse changes between the bilateral flocculi. The transcriptional activity in the ipsilateral flocculus may be decreased in the acute stage and increased in the chronic stage, compared to that in the contralateral flocculus. Several reports have indicated that inhibitory neural signals from cerebellar flocculi to the VNC are decreased on the ipsilateral side, but increased on the contralateral side during the acute stage of VC [10,11, 14,17]. Thus, our finding that transcription-related proteins were decreased in the ipsilateral cerebellar flocculus during the acute stage may explain the abrogation of the neural signals from the ipsilateral cerebellar flocculus. Other proteins examined, Pygb, Aco2, Atp5b and Atp5fl, are enzymes required for energy production. However, their expression patterns were not similar to each other. Atp5b was decreased in the chronic stage of VC, while that of Atp5f1 was increased. Neither showed any changes in the acute stage. In contrast, Pygb and Aco2 showed decreased expression during the acute stage, but both recovered their expression in the chronic stage. Paterson et al. reported that proteins associated with energy production and energy metabolism in mitochondria were up-regulated in the medial vestibular nuclei (MVN) during the chron-

10 92 M. Fukasawa et al. / Proteomic analysis of the rat cerebellar flocculus during vestibular compensation ic stage in UL-rats [21]. They speculated that the upregulation of mitochondrial energy production in the MVN resulted from the energy demands for VC processes such as gliosis, neuronal growth, and synaptic re-modeling. However, our results are not consistent with theirs. Only Atp5f1 expression was increased in the chronic stage, and the expression of the remaining three proteins, Pygb, Aco2, and Atp5b, was decreased or unchanged. Interestingly, Aco2 expression, which decreased in the contralateral flocculus during the acute stage of VC in our experiment, is known to be inactivated by nitric oxide (NO) [24]. NO produced in the unipolar brush cells is reported to be involved in VC [17]. Thus, the decrease in Aco2 expression detected in our experiment may have resulted from an increase in NO. The heterogeneous expression patterns of these energy production-related molecules suggest that they have roles related to balancing the bilateral cerebellar flocculi, in addition to those of energy production. Further studies are needed to clarify this point. It was reported that Inexa, Nsf and Pdia3 are involved in neuronal filament formation, synaptic transmission, and vesicular transport. Two spots (No. 008 and No. 365) were assigned as Inexa, as shown in the lower panels of Fig. 4A. In the acute stage of VC, the expression of Inexa increased contralaterally, but decreased ipsilaterally, and then returned to the level observed in the sham-operated rats in the chronic stage. Inexa, a neuronal intermediate filament protein, was reported to be involved in neuronal filament formation and to contribute to the maintenance of axon structure [26]. Reportedly, inhibitory neural signals from the cerebellar flocculi to the VNC are increased in the contralateral side and decreased in the ipsilateral side during the acute stage [6,12,15,16,18]. The signals returned to their normal level in the bilateral flocculi in the chronic stage. Taking these data together with ours, Inexa may be involved in the acute stage of VC via synaptogenesis, such as in the establishment of axonal outgrowth. The expression of Pdia3 (No. 390) was decreased bilaterally in the acute stage and returned to the same level as seen in the sham-operated rats during the chronic stage of VC (Fig. 4A). Pdia3, which is localized to the endoplasmic reticulum (ER), is involved in the vesicular transport of newly made proteins from the ER to the Golgi apparatus [8]. Pdia3 modulates the folding of newly synthesized glycoproteins by interacting with lectin chaperones (calreticulin and calnexin) and by promoting the formation of disulfide bonds in glycoprotein substrates [7]. Thus, protein secretion may be decreased in the bilateral cerebellar flocculi during the acute stage. N-ethylmaleimide-sensitive fusion protein (Nsf) is required for vesicle-mediated transport, and is thought to catalyze the fusion of transport vesicles to the plasma membrane using ATP during exocytosis [25,27]. Nsf is reported to be involved in both inhibitory GABAergic and excitatory glutamatergic synaptic transmission [25, 27]. In addition, Nsf is reported to interact with GluR2, a subunit of the AMPA receptor, by binding with Ca 2+, which increases AMPAR receptor trafficking during synaptic plasticity [25,27]. As mentioned in Fig. 4B, we found that Nsf undergoes PTM and that the PTM of Nsf changes during the acute stage of VC. PTM is deeply involved in the regulation of protein functions [9]. Evidence [13,20,27] indicates that phosphorylation of Nsf is associated with depolarizationdependent neurotransmitter release from synaptosomes and that S-nitrosylation of Nsf mediates the surface expression of AMPA receptors. In this experiment, although the type of PTM and the effects of PTM on the activity of Nsf are unknown, PTM of Nsf plays an important role during the acute stage of VC. Further experiments are needed to elucidate the significance of PTM during VC. Combining the results for Inexa, Pdia3 and Nsf in the acute stage of VC, we suggest that synaptogenesis in the cerebellar flocculi increases contralaterally and decreases ipsilaterally, but vesicular transport and synaptic transmission decrease bilaterally. Of note, the altered expression levels of these proteins returned to the same levels as seen in the sham-operated rats during the chronic stage. We did not obtain direct evidence that the protein RSI changes are causally related to the induction nor maintenance of VC, due to the lack of pharmacological confirmation. However, these proteins are good candidates as the pharmacological targets. Further studies on these proteins may provide important information on the molecular mechanisms responsible for VC. 5. Conclusions Using 2D-DIGE, we showed chronological changes in protein expression in the cerebellar flocculi that may contribute to acute and chronic stages of VC. We found that 99 out of 967 protein spots showed significant changes in their intensities during these stages. Similar numbers, but different protein spots bilaterally showed changes in intensity in the chron-

11 M. Fukasawa et al. / Proteomic analysis of the rat cerebellar flocculus during vestibular compensation 93 ic stage. In contrast, two thirds of the protein spots that showed changes in intensity during the acute stage were detected in contralateral flocculus. Using MALDI-TOF MS, we successfully identified 10 proteins. Of these, 2 transcription-related proteins showed synchronous changes. In addition, 4 energy productionrelated proteins showed different patterns between the bilateral flocculi and between the acute and chronic stages. Three proteins involved in synaptic transmission, neuronal filament formation, and vesicular transport, respectively, demonstrated altered expression only in the acute stage. Furthermore, alterations in posttranslational modifications during VC were observed for N-ethylmaleimide-sensitive fusion protein. Our data are important in the understanding of the cerebellar floccular molecules participating in VC generation. Acknowledgements The authors thank Drs. Arata Horii, Tadashi Kitahara and Chisako Masumura, Department of Otolaryngology, Osaka University Graduate School of Medicine for technical advice and Ms Mie Kanke, Hiroko Murakami, Michiyo Yokoyama and Mayumi Tamaki, Clinical Proteomics & Molecular Medicine, St. Marianna University Graduate School of Medicine for their technical assistance. References [1] G.B. Azzena, O. Mameli and E. Tolu, Cerebellar contribution in compensating the vestibular function, Prog Brain Res 50 (1979), [2] J.H. Ban, J. Chang, G.J. Im and H.H. Jung, Proteomic analysis of the rat vestibular nucleus complex following unilateral labyrinthectomy, Acta Otolaryngol 21 (2008), 1 9. [3] C.F. Bisgaard, M.N. Jayatissa, J.J. Enghild, C. Sanchez, R. Artemychyn and O. Wiborg, Proteomic investigation of the ventral rat hippocampus links DRP-2 to escitalopram treatment resistance and SNAP to stress resilience in the chronic mild stress model of depression, J Mol Neurosci 32 (2007), [4] K. Bomsztyk, O. Denisenko and J. Ostrowski, hnrnp K: One protein multiple processes, Bioessays 26 (2004), [5] J.H. Courjon, J.M. Flandrin, M. Jeannerod and R. Schmid, The role of the flocculus in vestibular compensation after hemilabyrinthectomy, Brain Res 239 (1982), [6] C.L. Darlington and P.F. Smith, Molecular mechanisms of recovery from vestibular damage in mammals: recent advances, Prog Neurobiol 62 (2000), [7] B. Di Jeso, Y.N. Park, L. Ulianich, A.S. Treglia, M.L. Urbanas, S. High and P. Arvan, Mixed-disulfide folding intermediates between thyroglobulin and endoplasmic reticulum resident oxidoreductases ERp57 and protein disulfide isomerase, Mol Cell Biol 25 (2005), [8] R.B. Freedman, P. Klappa and L.W. Ruddock, Protein disulfide isomerases exploit synergy between catalytic and specific binding domains, EMBO Rep 3 (2002), [9] H. Fujisawa, R. Ohtani-Kaneko, M. Naiki, T. Okada, K. Masuko, K. Yudoh, N. Suematsu, K. Okamoto, K. Nishioka and T. Kato, Involvement of post-translational modification of neuronal plasticity-related proteins in hyperalgesia revealed by a proteomic analysis, Proteomics 8 (2008), [10] K.F. Hamann and J. Lannou, Dynamic characteristics of vestibular nuclear neurons responses to vestibular and optokinetic stimulation during vestibular compensation in the rat, Acta Otolaryngol Suppl 445 (1987), [11] A. Horii, T. Kitahara, P.F. Smith, C.L. Darlington, C. Masumura and T. Kubo, Effects of unilateral labyrinthectomy on GAD, GAT1 and GABA receptor gene expression in the rat vestibular nucleus, Neuroreport 14 (2003), [12] A. Horii, P.F. Smith and C.L. Darlington, Quantitative changes in gene expression of glutamate receptor subunits/subtypes in the vestibular nucleus, inferior olive and flocculus before and following unilateral labyrinthectomy in the rat: real-time quantitative PCR method, Exp Brain Res 139 (2001), [13] Y. Huang, H.Y. Man, Y. Sekine-Aizawa, Y. Han, K. Juluri, H. Luo, J. Cheah, C. Lowenstein, R.L. Huganir and S.H. Snyder, S-nitrosylation of N-ethylmaleimide sensitive factor mediates surface expression of AMPA receptors, Neuron 46 (2005), [14] M. Ito, Cerebellar long-term depression: characterization, signal transduction, and functional roles, Physiol Rev 81 (2001), [15] A.R. Johnston, J.R. Seckl and M.B. Dutia, Role of the flocculus in mediating vestibular nucleus neuron plasticity during vestibular compensation in the rat, J Physiol 545 (2002), [16] T. Kitahara, N. Takeda, H. Kiyama and T. Kubo, Molecular mechanisms of vestibular compensation in the central vestibular system review, Acta Otolaryngol Suppl 539 (1998), [17] T. Kitahara, N. Takeda, T. Kubo and H. Kiyama, Nitric oxide in the flocculus works the inhibitory neural circuits after unilateral labyrinthectomy, Brain Res 815 (1999), [18] T. Kitahara, N. Takeda, T. Saika, T. Kubo and H. Kiyama, Role of the flocculus in the development of vestibular compensation: immunohistochemical studies with retrograde tracing and flocculectomy using Fos expression as a marker in the rat brainstem, Neuroscience 76 (1997), [19] A. Malhas, C.F. Lee, R. Sanders, N.J. Saunders and D.J. Vaux, Defects in lamin B1 expression or processing affect interphase chromosome position and gene expression, J Cell Biol 176 (2007), [20] E.A. Matveeva, S.W. Whiteheart, T.C. Vanaman and J.T. Slevin, Phosphorylation of the N-ethylmaleimide-sensitive factor is associated with depolarization-dependent neurotransmitter release from synaptosomes, J Biol Chem 276 (2001), [21] J.M. Paterson, D. Short, P.W. Flatman, J.R. Seckl, A. Aitken and M.B. Dutia, Changes in protein expression in the rat medial vestibular nuclei during vestibular compensation, J Physiol 575 (2006), [22] S. Shimada, M. Nakamura, Y. Tanaka, K. Tsutsumi, M. Katano, K. Masuko, K. Yudoh, I. Koizuka and T. Kato, Crosslinking of the CD69 molecule enhances S100A9 production in activated neutrophils, Microbiol Immunol 51 (2007),

12 94 M. Fukasawa et al. / Proteomic analysis of the rat cerebellar flocculus during vestibular compensation [23] H. Straka, N. Vibert, P.P. Vidal, L.E. Moore and M.B. Dutia, Intrinsic membrane properties of vertebrate vestibular neurons: function, development and plasticity, Prog Neurobiol 76 (2005), [24] V. Tórtora, C. Quijano, B. Freeman, R. Radi and L. Castro, Mitochondrial aconitase reaction with nitric oxide, S- nitrosoglutathione, and peroxynitrite: mechanisms and relative contributions to aconitase inactivation, Free Radic Biol Med 42 (2007), [25] T.Y. Yoon and Y.K. Shin, Progress in understanding the neuronal SNARE function and its regulation, Cell Mol Life Sci 66 (2009), [26] A. Yuan, M.V. Rao, T. Sasaki, Y. Chen, A. Kumar, Veeranna, R.K. Liem, J. Eyer, A.C. Peterson, J.P. Julien and R.A. Nixon, Alpha-internexin is structurally and functionally associated with the neurofilament triplet proteins in the mature CNS, J Neurosci 26 (2006), [27] C. Zhao, J.T. Slevin and S.W. Whiteheart, Cellular functions of NSF: not just SNAPs and SNAREs, FEBS Lett 581 (2007),