GTC
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE ADVANCED SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Genes to Cells (2007) 12, 889-901. doi:10.1111/j.1365-2443.2007.01099.x
© 2007 Blackwell Publishing or its licensors
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kobayashi, T.
Right arrow Articles by Kakizuka, A.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Kobayashi, T.
Right arrow Articles by Kakizuka, A.

Involvement of valosin-containing protein (VCP)/p97 in the formation and clearance of abnormal protein aggregates

 Taeko Kobayashi1,2, Atsushi Manno1,2 and Akira Kakizuka1,2,*

1 Laboratory of Functional Biology, Graduate School of Biostudies, Kyoto University, Kyoto 606-8501, Japan
2 SORST, Japan Science Technology Agency, Japan


   Abstract
Top
 Abstract
Introduction
 Results
 Discussion
 Experimental procedures
 References
 
Abnormal protein aggregates are commonly observed in affected neurons in many neurodegenerative disorders. We have reported that valosin-containing protein (VCP) co-localizes with protein aggregates in patients’ neurons and in cultured cells expressing diseased proteins. However, the significance of such co-localization remains elucidated. Here we report the involvement of VCP in the re-solubilization process of abnormal protein aggregates. VCP recognized and accumulated onto pre-formed protein aggregates created by proteasome inhibition. VCP knockdown or the expression of dominant-negative VCP both significantly delayed the elimination of ubiquitin-positive aggregates. VCP was involved in the clearance of pre-formed polyglutamine aggregates as well. Paradoxically, VCP knockdown also diminished polyglutamine aggregate formation. Furthermore, its ATPase activity was required for the re-solubilization and re-activation of heat-denatured proteins, such as luciferase, from insoluble aggregates. We thus propose that VCP functions as a mediator for both aggregate formation and clearance depending upon the concentration of soluble aggregate-prone proteins, indicating dual VCP functions as an aggregate formase and an unfoldase.


   Introduction
Top
 Abstract
 Introduction
Results
 Discussion
 Experimental procedures
 References
 
Many neurodegenerative disorders, such as polyglutamine diseases and Parkinson disease are characterized by the accumulation of abnormally folded diseased proteins in the neurons of the affected regions, resulting in the neuronal toxicity (Kakizuka 1998). The toxicity of diseased proteins likely depends on their potency to make aggregates, for example, nuclear inclusions in polyglutamine diseases and Lewy bodies in Parkinson disease. However, it is still unclear as to how protein aggregates and inclusion bodies contribute to the pathogenesis. Several possibilities have been proposed; that inclusion body formation might sequester cellular factors required for cells to survive, resulting in cellular toxicity; or oppositely, that inclusion body formation might sequester diseased proteins to decrease their toxicity (Shimohata et al. 2000; Arrasate et al. 2004). In either case, such inclusions are a kind of garbage, and thus cells are expected to have inherent systems for the elimination of them. Namely, the cellular protein quality control system built up by molecular chaperones and the ubiquitin-proteasome systems, monitors the amounts of mis-folded proteins in cells and alleviates the toxic property by refolding or degrading them. In addition, when a high amount of mis-folded proteins are produced by environmental stresses, proteasome dysfunction or genetic mutations as observed in human neurodegenerative disorders, the cellular quality control system may convert them to detergent-insoluble protein aggregates in cells. Indeed, cellular factors functioning in the protein quality control system such as Hsp70, Hsp40, ubiquitin and proteasome subunits, co-localize with these protein aggregates (Waelter et al. 2001).

Previously, we have purified valosin-containing protein (VCP)/p97 via affinity chromatography using Machado–Joseph disease (MJD) protein with expanded polyglutamine (ex-polyQ) (Kawaguchi et al. 1994; Hirabayashi et al. 2001). We have shown that, interestingly, VCP co-localizes in protein aggregates in neurons of patient brains suffering from not only MJD but also Huntington disease and Parkinson disease, and so on, as well as in cultured cells expressing ex-polyQ or cells treated with proteasome inhibitors (Hirabayashi et al. 2001; Mizuno et al. 2003; Ishigaki et al. 2004). In addition, we have identified ter94, Drosophila VCP, as a genetic modifier of ex-polyQ-induced eye degeneration (Higashiyama et al. 2002). Namely, in several ter94 mutants, especially in ter9426–8 background, ex-polyQ-induced eye degeneration was dramatically mitigated (Higashiyama et al. 2002). These results have emphasized the importance of VCP in the pathology of human neurodegenerative disorders (Hirabayashi et al. 2001; Higashiyama et al. 2002; Kimura & Kakizuka 2003; Kobayashi & Kakizuka 2003; Mizuno et al. 2003; Ishigaki et al. 2004).

Regarding the effects of VCP on aggregates, over-expression of a C. elegans VCP homologue has been shown to decrease ex-polyQ aggregates in C. elegans (Yamanaka et al. 2004). However, in mammalian cells, severe VCP knockdown or expression of a dominant-negative form of VCP has been shown to suppress proteasome inhibitor-induced aggresome formation (Wójcik et al. 2004; Kitami et al. 2006). These results have provided apparently opposite functions of VCP towards aggregates. Thus, the biological significance of VCP in protein aggregate formation and/or clearance is still controversial and remains to be elucidated.


   Results
Top
 Abstract
 Introduction
 Results
Discussion
 Experimental procedures
 References
 
VCP recognizes and co-localizes with pre-formed protein aggregates

It has been shown that VCP co-localizes with ubiquitin-positive protein aggregates created by proteasome inhibition and ex-polyQ expression in cultured cells (Fig. 1A). These protein aggregates have been shown to passively include many cellular proteins, such as TATA-binding protein (TBP), CREB binding protein (CBP), and so on, resulting in sequestering of proteins important for cellular functions (Shimohata et al. 2000). VCP might be one of these proteins. Given that VCP belongs to the AAA (ATPase associated with various cellular activities) class of ATPase, which includes several proteins functioning in the quality control of cellular proteins, and that VCP itself has been shown to function in such processes, an alternative possibility that VCP has yet unknown biological functions onto these aggregates has arisen. In order to address this possibility, we performed the following experiments. First, protein aggregates were created by either the treatment of the cells with reversible proteasome inhibitors MG132 or PSI, or the expression of ex-polyQ using the tet-off inducible system in neuronally differentiated PC12 cells. Then the formation of protein aggregates was stopped by washing out the proteasome inhibitor or shutting off the expression of ex-polyQ by the addition of tetracycline. In the next step, GFP-tagged VCP (VCP-GFP) was expressed in these cells by transfection, and the localization of GFP signals was examined by fluorescent microscopy. Just after transfection, the culture medium was replaced with fresh medium without the proteasome inhibitor or with tetracycline. Remarkably, VCP-GFP co-localized with these protein aggregates even after the aggregate formation (Fig. 1B, upper panels and Supplementary Fig. S1). The aggregates gradually shrank with continued culture of these cells, but VCP remained localized with them until complete disappearance of these aggregates (Fig. 1B, middle panels). After this last moment, GFP signals were diffusely distributed throughout the cells (Fig. 1B, lower panels). These observations suggest the possibility that VCP is capable of recognizing aggregates, even when aggregates are pre-formed and keeps associated with them until their complete disappearance. By tracing GFP signals in a single cell with time-lapse video-microscopy (Fig. 1C and Supplementary Movie S1), we confirmed that this was indeed the case. Considering the potential protein quality control activity of VCP, these results strongly indicate that VCP plays an important role in the clearance of protein aggregates.


Figure 1
View larger version (45K):
[in this window]
[in a new window]

  		Figure 1  Localization of endogenous VCP in cells harboring protein aggregates. Protein aggregates were induced by the treatment of 0.5 µM MG132, 0.1 µM PSI or polyglutamine expression for 2 days in neuronally differentiated PC12 cells. (A) Cells were fixed in ice cold EtOH, and were then processed for immunofluorescence staining. VCP and ubiquitin were visualized via fluorescent microscopy with FITC (green signals)- and Texas Red (red signals)-conjugated second antibodies, respectively. White arrows indicate positions of aggregates. (B, C) VCP-GFP was expressed (time 0) in cells harboring pre-formed protein aggregates, and the cells were analyzed by fluorescent microscopy. Representative photos of cells 24, 48, 60 and 96 h after the expression of VCP-GFP are presented in (B). Alternatively, GFP signals were traced by time-lapse fluorescent microscopy for 3 days in a single cell (C). See also Supplementary Movie S1.

 
VCP in the clearance of proteasome inhibitor-induced protein aggregates

We have developed conditions for VCP knockdown (KD) in HeLa cells by the RNA interference (RNAi) technique (N. Kato & A. Kakizuka, manuscript in preparation). In the VCP KD condition used in this experiment, the VCP level was reduced to 30.6% (± 9.2%) of that in non-treated HeLa cells (Fig. 2A, middle panels). We then used this condition to examine the role of VCP in aggregate clearance. VCP KD cells were considered to be vulnerable to proteasome inhibition so we adopted a mild proteasome inhibition using a low concentration of MG132 to induce aggregate formations. As a result, protein aggregates, so called aggresomes, were successfully observed in VCP KD cells as well as in control cells with 18 h treatment of 0.5 µM MG132 (Fig. 2B), although another report has shown that VCP KD inhibits aggresome formation by the treatment of high concentrations of proteasome inhibitors (Wójcik et al. 2004). Aggregate clearance is theoretically composed of multiple steps. At first, the aggregated proteins should be unfolded, and then degraded probably by the proteasome. We then used these cells to examine the role of VCP in aggregate clearance. It has been reported that VCP is involved at least in the transport of ubiquitinated proteins to the proteasome (Dai & Li 2001).


Figure 2
View larger version (44K):
[in this window]
[in a new window]

  		Figure 2  Effects of VCP knockdown on the clearance of protein aggregates formed by proteasome inhibition. (A) Effects of VCP depletion by VCP siRNA. HeLa cells were transfected with control siRNA (lanes 1–4 and 9–12) and VCP siRNA (lanes 5–8 and 13–16). Cells were grown for 3 days after transfection, and then treated with 0.5 µM MG132 for 18 h (day –1). MG132 was washed out with medium 5 times (day 0). Cells were then cultured for an additional 1 or 2 day(s) after the wash (day 1 and day 2, respectively). Samples were dissolved in 1% Triton-containing buffer and fractionated by high-speed centrifugation. Ubiquitinated proteins (Ub) and VCP (VCP) and tubulin (tubulin) were visualized by Western blot analysis using the specific antibodies. (B) Formation of protein aggregates in mildly VCP-depleted (lower panels) or control cells (upper panels) treated as described in (A). Cells were fixed after 18 h treatment with 0.5 µM MG132 and visualized by immunofluorescence, using VCP (green signals) and ubiquitin (red signals)-specific antibodies. Black arrows denote protein aggregates.

 
After the induction of aggregate formation for 18 h, MG132 was washed out from the culture medium (day 0). Cells were harvested 1 day before (day –1), 1 day after (day 1) and 2 days after (day 2) the removal of MG132 as well as at day 0. Each cell lysate was separated into Triton X-100-soluble and -insoluble fractions, and both fractions were examined by Western blot with an anti-ubiquitin antibody; protein aggregates are known to be fractionated in the Triton X-100-insoluble fraction and are poly-ubiquitinated (Fig. 2A, upper panels). Poly-ubiquitinated proteins were apparently accumulated in control HeLa cells at day 0, rapidly disappearing as early as at day 1 in both soluble and insoluble fractions (Fig. 2A, lanes 2, 3, 10 and 11). In contrast, poly-ubiquitinated proteins stayed up to day 2 in both insoluble and soluble fractions in VCP KD cells (Fig. 2A, lanes 6–8, 14–16). These results indicate that levels of VCP are crucial for the clearance of aggregated proteins as well as detergent-soluble poly-ubiquitinated proteins.

Alternatively, these results may reflect that VCP knockdown simply induced ER stress (Kobayashi et al. 2002), which in turn induced accumulation of ubiquitinated proteins especially in a detergent-insoluble fraction. In order to rule out this possibility, we examined whether the treatment of cells by tunicamycin, a well-known ER stress inducer (Nishitoh et al. 2002), can affect the levels of ubiquitinated proteins (Fig. 3). Tunicamycin treatment did induce CHOP expression (Nishitoh et al. 2002), a well-known ER stress marker, but could not induce the accumulation of ubiquitinated proteins in both detergent-soluble and -insoluble fractions (Fig. 3).


Figure 3
View larger version (84K):
[in this window]
[in a new window]

  		Figure 3  Absence of correlation between tunicamycin-induced ER stress and accumulation of ubiquitinated proteins. HeLa cells were treated with 0.5, 1, 2, 5 µg/mL tunicamycin (Tm) for 18 h, or with 0.5, 1, 2 µM MG132 (MG132) for 18 h. After 18 h treatment, cells were harvested, and triton-soluble (soluble) and insoluble (insoluble) fractions were separated as described in Experimental procedures. Ubiquitinated proteins (Ub), actin (actin) as a loading control, and CHOP (CHOP) as an ER stress marker were visualized by Western blot analysis using the specific antibodies.

 
In order to further confirm the involvement of VCP in aggregate clearance, we took advantage of VCP(K524A), a dominant-negative form of VCP, in which the 524th lysine is substituted by alanine, and examined the effect of this mutant on aggregate clearance; this mutant VCP has completely lost its ATPase activity, and the expression of this mutant is able to inhibit the ATPase activity of endogenous VCP in vivo (Kobayashi et al. 2002). We had previously made two PC12 cell lines (TV and TmV cells), in which C-terminally GFP-tagged wild-type VCP and VCP(K524A), respectively, were expressed under the control of the tet-off inducible promoter (Kobayashi et al. 2002). First, TV and TmV cells were neuronally differentiated by addition of NGF up to 2 days before the treatment of MG132 (day –2). Cells were treated in the presence (tet-on) or absence (tet-off) of tetracycline for 1 day until the treatment of MG132 (day –1), and then treated with 0.1 µM MG132 or DMSO for 18 h. Triton X-100-soluble and -insoluble proteins were examined by Western blot analysis (Fig. 4). Proteasome inhibition apparently led TV cells to accumulate more poly-ubiquitinated proteins in the tet-off condition than those in the tet-on condition (Fig. 4, lanes 2 and 6), but these poly-ubiquitinated proteins equally disappeared 1 day after MG132 removal (Fig. 4, lanes 3 and 7). On the other hand, TmV tet-off cells could not efficiently eliminate poly-ubiquitinated proteins; nevertheless, their accumulation in tet-off cells was lower than that in tet-on cells (Fig. 4, lanes 14, 15, 18 and 19). In soluble fractions, however, more poly-ubiquitinated proteins were accumulated in TmV tet-off cells than in TmV tet-on cells (Fig. 4, lanes 14 and 18). In the condition used above, most of the visible aggregates disappeared 1 day after the removal of MG132; confirmed via fluorescent microscopy (Supplementary Fig. S2). These results clearly show that the ATPase activity of VCP is required to eliminate protein aggregates formed by proteasome inhibition, and also suggest that VCP is involved in creating protein aggregates from soluble poly-ubiquitinated proteins.


Figure 4
View larger version (97K):
[in this window]
[in a new window]

  		Figure 4  Effects of functional depletion of VCP. TV and TmV cells were differentiated by NGF and grown in tet-on (on) or tet-off (off) medium for 1 day (from day –2 to day –1). A 0.1 µM MG132 or DMSO was added to the cells in tet-on medium (day –1) and washed out the next day (day 0) with fresh tet-on medium. Cells were harvested 1 day (day 1) and 2 days (day 2) later, and the triton-soluble (soluble) and insoluble (insoluble) fractions were separated and analyzed by Western blot, as described in Experimental procedures. A schematic drawing of the experimental schedule is shown at the bottom.

 
VCP in the clearance of ex-polyQ aggregates

We next investigated the function of VCP on other aggregates. As shown in Fig. 1A, VCP co-localizes with ex-polyQ aggregates (Hirabayashi et al. 2001). To examine whether expression levels of VCP affect the state of ex-polyQ aggregates, we prepared a HeLa cell line expressing Q81-GFP-Myc under the tet-on inducible promoter, and set up the following experimental condition. After the completion of ex-polyQ aggregate formation (Fig. 5A), VCP siRNA was introduced to induce an effective decrease in VCP levels (Fig. 5B, lanes 3 and 5). To quantify aggregated ex-polyQ, cell lysates were subjected to the filter retardation assay (Wanker et al. 1999). In this assay, ex-polyQ aggregates were retained on cellulose acetate filters, and the amount of aggregates was quantified by immunoblotting using an anti-Myc antibody. VCP KD was performed after ex-polyQ was expressed for 4 days. Just after VCP siRNA transfection, the ex-polyQ expression was stopped by replacing the medium with a tetracycline-free medium. Days following the VCP siRNA transfection, VCP levels were gradually decreased in these cells (Fig. 5B, lanes 1, 3 and 5). The total amounts of ex-polyQ were indistinguishable between control and VCP KD cells at both day 1 and day 2 (Fig. 5B, lanes 2–5). The amounts of ex-polyQ aggregates were also indistinguishable between control and VCP KD cells at day 1 (Fig. 5B, lane 7). However, at day 2 much more ex-polyQ aggregates remained in VCP KD cells than in control cells (Fig. 5B, lane 8). Consistently more aggregated or dotted GFP signals were observed via fluorescent microscopy at day 2 in VCP KD cells than in control cells (Fig. 5C). Although strong VCP KD inhibits cell growth (N. Kato & A. Kakizuka, manuscript in preparation), the growth rates of these VCP KD cells appeared unaffected in this experimental condition (Fig. 5D). These results clearly show that VCP is involved in the processes of clearing the pre-formed ex-polyQ aggregates.


Figure 5
View larger version (35K):
[in this window]
[in a new window]

  		Figure 5  Effect of VCP knockdown on ex-polyQ aggregate clearance. (A) A schematic drawing of the experimental schedule is shown. (B) Effect of VCP depletion on pre-formed ex-polyQ aggregates. Ex-polyQ was expressed in HeLa cells in the presence of tetracycline for 4 days to make aggregates. Then, the cells were transfected with control siRNA (lanes 2, 4 and "control" in lanes 7 and 8) and VCP siRNA (lanes 3, 5, and "KD" in lanes 7 and 8), and were cultured for another 1 or 2 days in the absence of tetracycline. Cells were harvested just before (day 0) (lanes 1 and 6), 1 day after (day 1) (lanes 2, 3 and 7), and 2 days after (day 2) (lanes 4, 5 and 8) the transfection. Cells were harvested and expression levels of VCP (VCP), actin (actin) as a loading control, and ex-polyQ (polyQ) were analyzed by Western blot (lanes 1–5). Ex-polyQ aggregates were visualized by the dot-blot assay (lanes 6–8) as described in Experimental procedures. (C) Ex-polyQ aggregates were visualized by immunostaining and counted in six different fields at each day as described in (B). Data are presented as the mean values. Bars, standard deviations. (D) Growth rates of VCP KD cells (white columns) and control cells (black columns) described in (B) were evaluated by measuring relative live cell numbers using the WST assay at 1 and 2 days after the siRNA transfection. Data are presented as the means of 12 samples. Bars, standard deviations.

 
VCP in the formation of ex-polyQ aggregates and aggresomes

We next examined the possible involvement of VCP in the formation of ex-polyQ aggregates by slightly modifying the assay described above. In this experiment, ex-polyQ expression and VCP KD were performed simultaneously (Fig. 6A). In this condition, less ex-polyQ aggregates accumulated in VCP KD cells than in control cells 4 days after induction of ex-polyQ, although the total amounts of ex-polyQ were equivalent in both cells (Fig. 6B, lanes 3, 4 and 6), indicating that VCP is also involved in the processes of making ex-polyQ aggregates.


Figure 6
View larger version (35K):
[in this window]
[in a new window]

  		Figure 6  Effect of VCP knockdown on ex-polyQ aggregate formation. (A) A schematic drawing of the experimental schedule is shown. (B) HeLa cells were transfected with control siRNA (lanes 1, 3 and "control" in lanes 5 and 6) and VCP siRNA (lanes 2, 4 and "KD" in lanes 5 and 6). Then the cells were grown for another 3 days in the absence (lanes 1 and 2) and presence (lanes 3 and 4) of tetracycline, in which the latter induces ex-polyQ expression. Cells were harvested and expression levels of VCP (VCP), actin (actin) as a loading control, and ex-polyQ (polyQ) were analyzed by Western blot (lanes 1–4). Ex-polyQ aggregates were visualized by the dot-blot assay (lanes 5 and 6) as described in Experimental procedures. In the absence of tetracycline, either ex-polyQ expression (lanes 1 and 2) or ex-polyQ aggregates (lane 5) were not observed.

 
We then examined the dose-dependency of VCP on aggresome formation in TV cells (see above), in which VCP-GFP was expressed under the control of the tet-off promoter. Indeed, the lower the tetracycline concentration, the more VCP-GFP was expressed (Fig. 7A). These cells were treated with different concentrations of MG132 and tetracycline for 18 h, and the amounts of ubiquitinated proteins were examined by Western blot analysis. The amounts of detergent-soluble ubiquitinated protein increased depending on the concentration of MG132 (Fig. 7B). In contrast, detergent-insoluble ubiquitinated proteins appeared not to increase up to 75 nM MG132 treatment even in the presence of exogenously introduced VCP-GFP (Fig. 7B,C). However, in cells treated with 100 nM MG132, more ubiquitinated proteins were accumulated in the detergent-insoluble fraction of the cells that expressed more VCP-GFP protein (Fig. 7B,C). These results suggest that aggregates are formed only in the presence of soluble ubiquitinated proteins above a certain (threshold) level and this threshold appears to be unaffected by the amount of exogenously introduced VCP. However, when soluble ubiquitinated proteins of more than the threshold level exist, aggregate formation is enhanced by exogenously introduced VCP in a dose-dependent manner.


Figure 7
View larger version (36K):
[in this window]
[in a new window]

  		Figure 7  Dose effects of VCP on proteasome inhibitor-induced protein aggregate formation. (A) TV cells were differentiated by NGF and incubated in media containing 50, 25, 16.7 and 12.5 µg/mL tetracycline (tet) for 2 days. Cells were harvested and expression levels of Ubiquitinated proteins (Ub), VCP-GFP (VCP-GFP), VCP (VCP), and actin (actin) as a loading control were analyzed by Western blot. The ratios of VCP-GFP expression levels to a loading control were quantified and are shown below the panels. (B) TV cells were differentiated by NGF and were grown in media containing 50, 25, 16.7 and 12.5 µg/mL tetracycline (tet) for 1 day, and then 50, 75 and 100 nM MG132 (50, 75, 100) or DMSO (0) was added to the medium. Eighteen hours after addition of MG132 or DMSO, cells were harvested, and Triton-insoluble (insoluble) and -soluble (soluble) fractions were separated as described in Experimental procedures and analyzed by Western blot. Ubiquitinated proteins (Ub), and actin (actin) as a loading control were visualized by Western blot analysis using the specific antibodies. (C) The ratios of Triton-insoluble ubiquitinated protein aggregates levels to a loading control, as shown in (B), are indicated. Each column number in (C) corresponds to each lane number in (B).

 
Chaperonic activity of VCP on heat-denatured luciferase

The above results suggest the possibility that VCP is able to function as an unfoldase toward aggregates. To investigate this possibility, we first analyzed whether VCP can possess chaperonic activities on heat-denatured proteins, namely cytoplasmic firefly luciferase. Cells expressing luciferase were heat-shocked at 45 °C for 15 min, and then incubated at 37 °C for several hours while inhibiting new protein synthesis by the addition of cycloheximide. Every hour, for up to 4 h after the heat shock, cells were harvested and their luciferase activities were measured. In this experiment, we compared the luciferase activities from four types of cells (Fig. 8A,B); control HeLa cells (referred to as control cells), VCP KD HeLa cells (referred to as KD cells), VCP KD HeLa cells re-introduced with wild-type VCP-GFP (referred to as KD + VCP cells) and VCP KD HeLa cells re-introduced with VCP(K251A)-GFP (referred to as KD + K251A cells). Since the target sequence of this VCP siRNA is located in the 5'-non-coding region of VCP mRNA and our VCP expression vectors did not contain the sequence, we could specifically reduce endogenous VCP but not exogenously introduced VCP levels (Fig. 8B). KD cells failed to reactivate luciferase after the heat shock, as compared with control cells (Fig. 8A). However, the re-introduction of VCP-GFP in KD cells led to a significant re-activation of luciferase activities, indistinguishable from those of control cells (Fig. 8A). To examine whether ATPase activity of VCP is required for this reactivation, VCP(K251A), harboring the Lys/Ala substitution at the 251st residue in the VCP D1 domain, was introduced into VCP KD cells. VCP(K251A) has been shown to have no measurable ATPase activity like VCP(K524A) but is less toxic than VCP(K524A) for unknown reasons. VCP(K251A) could not reactivate or even further reduced the luciferase activities (Fig. 8A). These results clearly indicate that VCP is involved in the re-folding and the re-activation of luciferase denatured by heat shock, and its ATPase activity is essential for this function.


Figure 8
View larger version (51K):
[in this window]
[in a new window]

  		Figure 8  Involvement of VCP in the re-activation of heat-denatured luciferase. (A) Luciferase activities after heat shock. Cells described in the text and (B) were treated with 20 µg/mL cycloheximide for 30 min, heated at 45 °C for 15 min, and then cultured for 0 to 4 h at 37 °C to allow luciferase reactivation. Luciferase and ß-galactosidase activities at 0, 1, 2 and 4 h after the heat shock were measured, and luciferase activities normalized by ß-galactosidase activities were plotted as percentages of the activity before heat shock (100%). Data are presented as the mean values of three independent experiments performed in duplicates. Bars, standard deviations. (B) HeLa cells were transfected with pACT-ß-gal, pGL3 for firefly luciferase targeted to the cytoplasm and VCP siRNA (lanes 1–3) or control siRNA (lane 4), together with the plasmid encoding GFP (as a control; lanes 1 and 4), GFP-tagged wild-type VCP (lane 2) or GFP-tagged VCP(K251A) (lane 3). Cells were cultured for 3 days after transfection, and the total cell lysates were subjected to Western blot analyses to detect VCP, Hsp70, and actin, a loading control with the according specific antibodies. Note that expression levels of Hsp70 were not affected in VCP KD cells. (C) Western blot analysis of luciferase. Samples were taken at before (pre), just after (HS) and 4 h after (4 h) the heat shock. Total cell lysates (total), and Triton-soluble (sup) and -insoluble (ppt) fractions were prepared as described in Experimental procedures. Then, each fraction was analyzed by Western blot.

 
In order to obtain a clue as to how VCP reactivates heat-denatured luciferase, these cell lysates were fractionated by centrifugation to separate heat-aggregated proteins from soluble proteins, and the amounts of luciferase as well as ß-galactosidase, a control protein of the transfection were examined during the time course by Western blot. Luciferase as well as ß-galactosidase levels in the SDS-soluble total cell lysates kept constant during the time course, showing that luciferase and ß-galactosidase were not degraded in this time course (Fig. 8C, lanes 1–6). Luciferase levels in the aggregate fraction increased remarkably just after the heat shock in both KD cells and KD + VCP cells, but were found to become lower in KD + VCP cells than in KD cells 4 h after the heat shock (Fig. 8C, lanes 7–9). In KD cells, the amount of luciferase stayed constant in the pellet fraction (Fig. 8C, lanes 7–9). Consistently, soluble luciferase levels increased in KD + VCP cells, but not in KD cells (Fig. 8C, lanes 10–12). These results indicate that VCP plays an important role in reactivating luciferase through re-solubilizing it from aggregates. This is the first clear demonstration that VCP possesses an unfoldase-like activity in vivo.


   Discussion
Top
 Abstract
 Introduction
 Results
 Discussion
Experimental procedures
 References
 
In this study, we demonstrated lines of evidence showing that VCP is involved in aggregate clearance. VCP appeared to recognize, move towards and surround pre-formed protein aggregates. VCP kept localized on the aggregates even after large reductions of aggregate size. Similar co-localization was observed in ATPase activity–negative VCPs such as VCP(K524A) (Supplementary Fig. S1), but aggregates or insoluble poly-ubiquitinated proteins hardly disappeared in the VCP(K524A)-expressing cells, implicating that VCP plays an important role in the aggregate clearance with using its ATPase activity. Evidence for involvement of VCP in aggregate clearance was strengthened by the following results. In VCP KD cells, the elevation of basal and induced levels of poly-ubiquitinated proteins was observed in the absence and presence of proteasome inhibitors, respectively. Furthermore, a dramatic delay of poly-ubiquitinated protein degradation or clearance was evident after removal of proteasome inhibitors in both detergent-soluble and insoluble fractions, the latter containing protein aggregates. Over-expression of either one of two C. elegans VCP homologues has been shown to partially suppress the formation of ex-polyQ aggregates in a polyglutamine disease model in C. elegans (Yamanaka et al. 2004). Recently, it has been reported that VCP KD in cultured liver cells increases Mallory bodies, aggresome-like structures containing ubiquitin, cytokeratin, and so on (Nan et al. 2005). These observations are also consistent with the idea that VCP is involved in aggregate clearance.

In E. coli, mis-folded protein aggregates have been shown to be re-folded via an unfoldase activity of Hsp100 family, ClpA and ClpB, which are also AAA ATPases (Mogk et al. 1999; Wickner et al. 1999). However, it is not yet clear whether mammalian cells have an unfoldase similar to the Hsp100 family. From a structural point of view, mammalian cells do not have any ClpA homologue. As far as for heat-denatured luciferase, VCP was shown clearly to reactivate luciferase activities via re-solubilization of heat-denatured luciferase from insoluble fractions. Similar to as in aggregate clearance, ATPase activities of VCP were indispensable in this function. Our observations that VCP functions in aggregate clearance and re-solubilization of heat-denatured luciferase in mammalian cells are quite reminiscent of ClpA and ClpB (Hsp104 in yeast), and thus VCP may work as an unfoldase in mammalian cells.

Regarding the effect of VCP upon ex-polyQ aggregates, VCP behaved in a totally different way depending on the ex-polyQ-expressing phases, namely during or after the expression. VCP can enhance aggregate formation during the expression of ex-polyQ, and then change its function to eliminate the aggregates after the ex-polyQ expression is stopped. Both functions to form and to eliminate aggregates would require a chaperonic activity, as represented by Hsp104. Hsp104, another member of yeast AAA family protein, has been shown to have dual functions onto Sup35 yeast prion protein depending on the proportion of Hsp104 to its substrates. Hsp104 has been shown to catalyze the formation and elimination of Sup35 aggregates, when its substrate (soluble form) concentration relative to Hsp104 is high or low, respectively (Parsell et al. 1994; Shorter & Lindquist 2004). These results as a whole suggest the possibility that, like Hsp104, VCP catalyzes both aggregate formation and clearance depending upon the concentration of soluble aggregate-prone proteins rather than upon the concentration of already aggregated or insoluble proteins. Namely, during the expression of ex-polyQ, it is expected that the concentration of soluble aggregate-prone ex-polyQ is high, and after shutting off the ex-polyQ expression it becomes low. Indeed, both conditions induced aggregate formation and clearance, respectively, and both aggregate formation and clearance were inhibited by VCP knockdown.

Very recently, Boeddrich et al. have shown that VCP affects self-aggregate formation of full-length MJD1/ataxin-3 protein via the interaction of arginine/lysine-rich motif (Boeddrich et al. 2006). However, neither our ex-polyQ constructs nor luciferase contains such a motif. Rather, our results suggest that VCP functions as an unfoldase and an aggregate formase upon a broad range of protein aggregates and aggregate-prone proteins, respectively. When such VCP substrates are abundantly produced in soluble forms, for example in the cases of polyglutamine diseases and Parkinson disease, VCP would catalyze the formation of ex-polyQ aggregates and Lewy bodies, respectively, and thus resulting in co-localization of VCP in both ex-polyQ aggregates and Lewy bodies. Consistent with this, severe VCP KD (the absence of VCP) or expression of a dominant negative form of VCP has been reported to diminish aggresome formation (Wójcik et al. 2004; Kitami et al. 2006). We do not currently know any conditions or chemicals that can shift this VCP function toward either aggregate formation or clearance. Searching for such conditions and/or chemicals may provide a clue for further understanding as well as for novel treatments of neurodegenerative disorders such as polyglutamine diseases and Parkinson disease.


   Experimental procedures
Top
 Abstract
 Introduction
 Results
 Discussion
 Experimental procedures
References
 
Cell culture, cell lines, transfection and immunostaining

PC12 cells and HeLa cells were cultured as previously reported (Kobayashi et al. 2002). PC12 cell lines, which express GFP-fused wild-type VCP, VCP(K524A) mutant and ex-polyQ containing Q78 repeats with Flag-tag under the tet-off medium, were established previously (Hirabayashi et al. 2001; Kobayashi et al. 2002). HeLa cell line, stably expressing ex-polyQ only in the presence of 1 µg/mL tetracycline, was created by transfection of pT-REx-DEST30/polyQ81-GFP-myc plasmid into T-RExTM-HeLa cell line (InvitrogenTM) and drug-selected by G418 and blasticidin. For immunostaining, cells were fixed in ice cold EtOH, washed with PBS (phosphate-buffered saline), permeabilized with 0.1% Triton/PBS, blocked with 1% goat serum and visualized with the appropriate antibodies.

Antibodies, plasmids and siRNAs

The affinity-purified rabbit polyclonal anti-VCP antibody was described previously (Hirabayashi et al. 2001; Mizuno et al. 2003). The following antibodies were used in this study: monoclonal anti-ubiquitin (Chemicon); monoclonal anti-FLAG® M5 (SIGMA®); rabbit polyclonal anti-ß-tubulin (H235) (Santa Cruz Biotechnology, Inc.); monoclonal anti-actin (Chemicon); monoclonal anti-Hsp70 (Stressgen®); goat polyclonal anti-luciferase (Promega); monoclonal anti-polyglutamine (Chemicon); rabbit polyclonal anti-c-myc (A-14) (Santa Cruz Biotechnology, Inc.); FITC dye-conjugated goat polyclonal anti-rabbit IgG and Texas Red® dye-conjugated goat polyclonal anti-mouse IgG (Jackson ImmunoReserch Lab., Inc.); donkey polyclonal anti-mouse IgG-HRP and anti-rabbit IgG-HRP (Amersham Biosciences). pGL3-control vector (Promega) and pACT-ß-gal are expression vectors for cytoplasmic firefly luciferase and E. coli ß-galactosidase, respectively. Plasmids encoding GFP-fused wild-type VCP and VCP(K251A) mutant were constructed in pEGFP-N vector (Clontech). The polyQ81 insert in pT-REx-DEST30/polyQ81-GFP-myc was derived from MJD1 gene causative of MJD, and constructed by using Gateway technology (InvitrogenTM). VCP and control siRNAs were chemically synthesized, and the sequences were from position (–17) to (+2) relative to the start codon of VCP or its scrambled sequence (CGGACGCGUCAGGAGCCGGTT), respectively.

Quantification of aggregated proteins

Cells harboring protein aggregates formed by proteasome inhibitors or heat shock were lysed with lysis buffer (1% Triton X-100, 50 mM Tris–HCl (pH 8.0), 150 mM NaCl and 1 mM EDTA containing protease inhibitors). Supernatant and pellet fractions were separated by centrifugations at 12 000 g for 10 min, and analyzed by SDS-PAGE and Western blot analysis. For polyglutamine aggregates, we performed dot-blot filter retardation assay with minor modifications (Wanker et al. 1999). Cells were suspended with lysis buffer (0.5% NP-40, 50 mM Tris–HCl (pH 8.0), 100 mM NaCl, 5 mM MgCl2 and 1 mM EDTA containing protease inhibitors and 250 U/mL benzonase), followed by incubation on ice for 30 min. Samples were mixed in dot-blot buffer (20 mM EDTA, 2% SDS and 50 mM DTT), boiled at 98 °C for 5 min, and were filtered on a cellulose acetate membrane (Schleicher and Schuell, 0.2-µm pore size) pre-equilibrated with 2% SDS. A 30 µg cell extract was loaded on a well, washed 2 times with 0.1% SDS, blocked with 5% skim milk-containing TBST buffer and analyzed by Western blot using anti-myc antibody. All results of Western blot analyses in this study were visualized using the ECL detection kit (Amersham Biosciences) and a luminescence image analyzer (LAS-1000 PLUS, Fuji Film).

Luciferase reactivation assay

Cells were transiently transfected with various combinations of siRNA and plasmids including pGL3, pACT-ß-gal and the pEGFP derivatives by Lipofectamine 2000 (InvitrogenTM). Twenty-four hours after transfection, media were replaced with the selection media containing 2 µg/mL G418 to select cells having the pEGFP derivative plasmid. Luciferase assay was performed at 3 days after the transfection. Cells were cultured for 30 min in the medium containing 20 µg/mL cycloheximide and 20 mM morpholinepropanesulfonic acid (MOPS; pH 7.0) at 37 °C, heat-shocked at 45 °C for 15 min in a water bath and grown at 37 °C (Nollen et al. 2000). Cell samples were moved on ice, washed with PBS for 3 times and lysed in Glo-lysis buffer (Promega). Insoluble proteins were removed by high-speed centrifugation. Supernatants were used for measurements of luciferase and ß-galactosidase activity. Data were taken from the ratio of luciferase to ß-galactosidase activity, because ß-galactosidase activity was not affected by the heat shock (Pinto et al. 1991).

Statistical analysis

Each experiment was conducted at least 3 times with consistent results. The gel or blot representative of each experiment is presented in this study. The statistical significance was analyzed using student's t test.


   Acknowledgements
 
We wish to thank K. Kuroiwa for technical assistance, our laboratory members for valuable discussions and Dr Y. Kimura for critical reading. This work was supported in part by research grants from the Ministry of Education, Culture, Sports, Science and Technology of Japan. A.M. is supported by the 21st Century COE Program of the Ministry of Education, Culture, Sports, Science and Technology of Japan to Graduate School of Biostudies and Institute for Virus Research, Kyoto University.


   Footnotes
 
Communicated by: Eisuke Nishida

aPresent address: Institute for Virus Research, Kyoto University, Kyoto 606–8507, Japan.

* Correspondence: E-mail: kakizuka{at}lif.kyoto-u.ac.jp


   References
Top
 Abstract
 Introduction
 Results
 Discussion
 Experimental procedures
 References
 
Arrasate, M., Mitra, S., Schweitzer, E.S., Segal, M.R. & Finkbeiner, S. (2004) Inclusion body formation reduces levels of mutant huntingtin and the risk of neuronal death. Nature 43, 805–810.

Boeddrich, A., Gaumer, S., Haacke, A. et al. (2006) An arginine/lysine-rich motif is crucial for VCP/p97-mediated modulation of ataxin-3 fibrillogenesis. EMBO J. 25, 1547–1558.[CrossRef][Medline]

Dai, R.M. & Li, C.C. (2001) Valosin-containing protein is a multi-ubiquitin chain-targeting factor required in ubiquitin-proteasome degradation. Nat. Cell Biol. 3, 740–744.[CrossRef][Medline]

Higashiyama, H., Hirose, F., Yamaguchi, M., Inoue, Y.H., Fujikake, N., Matsukage, A. & Kakizuka, A. (2002) Identification of ter94, Drosophila VCP, as a modulator of polyglutamine-induced neurodegeneration. Cell Death Differ. 9, 264–273.[CrossRef][Medline]

Hirabayashi, M., Inoue, K., Tanaka, K., Nakadate, K., Ohsawa, Y., Kamei, Y., Popiel, A.H., Sinohara, A., Iwamatsu, A., Kimura, Y., Uchiyama, Y., Hori, S. & Kakizuka, A. (2001) VCP/p97 in abnormal protein aggregates, cytoplasmic vacuoles, and cell death, phenotypes relevant to neurodegeneration. Cell Death Differ. 8, 977–984.[CrossRef][Medline]

Ishigaki, S., Hishikawa, N., Niwa, J., Iemura, S., Natsume, T., Hori, S., Kakizuka, A., Tanaka, K. & Sobue, G. (2004) Physical and functional interaction between dorfin and valosin-containing protein that are colocalized in ubiquitylated inclusions in neurodegenerative disorders. J. Biol. Chem. 279, 51376–51385.[Abstract/Free Full Text]

Kakizuka, A. (1998) Protein precipitation: a common etiology in neurodegenerative disorders? Trends Genet. 14, 396–402.[CrossRef][Medline]

Kawaguchi, Y., Okamoto, T., Taniwaki, M., Aizawa, M., Inoue, M., Katayama, S., Kawakami, H., Nakamura, S., Nishimura, M., Akiguchi, I., Kimura, J., Narumiya, S. & Kakizuka, A. (1994) CAG expansions in a novel gene for Machado–Joseph disease at chromosome 14q32.1. Nat. Genet. 8, 221–228.[CrossRef][Medline]

Kimura, Y. & Kakizuka, A. (2003) Polyglutamine diseases and molecular chaperones. IUBMB Life 55, 337–345.[Medline]

Kitami, M.I., Kitami, T., Nagahama, M., Tagaya, M., Hori, S., Kakizuka, A., Mizuno, Y. & Hattori, N. (2006) Dominant-negative effect of mutant valosin-containing protein in aggresome formation. FEBS Lett. 580, 474–478.[CrossRef][Medline]

Kobayashi, T. & Kakizuka, A. (2003) Molecular analyses of Machado–Joseph disease. Cytogenet. Genome Res. 100, 261–275.[CrossRef][Medline]

Kobayashi, T., Tanaka, K., Inoue, K. & Kakizuka, A. (2002) Functional ATPase activity of p97/VCP is required for the quality control of endoplasmic reticulum in neuronally differentiated mammalian PC12 cells. J. Biol. Chem. 277, 47358–47365.[Abstract/Free Full Text]

Mizuno, Y., Hori, S., Kakizuka, A. & Okamoto, K. (2003) Vacuole-creating protein in neurodegenerative diseases in humans. Neurosci. Lett. 343, 77–80.[CrossRef][Medline]

Mogk, A., Tomoyasu, T., Goloubinoff, P., Rüdiger, S., Röder, D., Langen, H. & Bukau, B. (1999) Identification of thermolabile Escherichia coli proteins: prevention and reversion of aggregation by DnaK and ClpB. EMBO J. 18, 6934–6949.[CrossRef][Medline]

Nan, L., Wu, Y., Bardag-Gorce, F., Li, J., French, B.A., Wilson, L.T., Nguyen, S.K. & French, S.W. (2005) RNA interference of VCP/p97 increases Mallory body formation. Exp. Mol. Pathol. 78, 1–9.[CrossRef][Medline]

Nishitoh, H., Matsuzawa, A., Tobiume, K., Saegusa, K., Takeda, K., Inoue, K., Hori, S., Kakizuka, A. & Ichijo, H. (2002) ASK1 is essential for endoplasmic reticulum stress-induced neuronal cell death triggered by expanded polyglutamine repeats. Genes Dev. 16, 1345–1355.[Abstract/Free Full Text]

Nollen, E.A.A., Brunsting, J.F., Song, J., Kampinga, H.H. & Morimoto, R.I. (2000) Bag1 functions in vivo as a negative regulator of Hsp70 chaperone activity. Mol. Cell. Biol. 20, 1083–1088.[Abstract/Free Full Text]

Parsell, D.A., Kowal, A. S., Singer, M.A. & Lindquist, S. (1994) Protein disaggregation mediated by heat-shock protein Hsp104. Nature 372, 475–478.[CrossRef][Medline]

Pinto, M., Morange, M. & Bensaude, O. (1991) Denaturation of proteins during heat shock. J. Biol. Chem. 266, 13941–13946.[Abstract/Free Full Text]

Shimohata, T., Nakajima, T., Yamada, M. et al. (2000) Expanded polyglutamine stretches interact with TAFII130, interfering with CREB-dependent transcription. Nat. Genet. 26, 29–36.[CrossRef][Medline]

Shorter, J. & Lindquist, S. (2004) Hsp104 catalyzes formation and elimination of self-replicating Sup35 prion conformers. Science 304, 1793–1797.[Abstract/Free Full Text]

Waelter, S., Boeddrich, A., Lurz, R., Scherzinger, E., Lueder, G., Lehrach, H. & Wanker, E.E. (2001) Accumulation of mutant huntingtin fragments in aggresome-like inclusion bodies as a result of insufficient protein degradation. Mol. Biol. Cell 12, 1393–1407.[Abstract/Free Full Text]

Wanker, E.E., Scherzinger, E., Heiser, V., Sittler, A., Eickhoff, H. & Lehrach, H. (1999) Membrane filter assay for detection of amyloid-like polyglutamine-containing protein aggregates. Methods Enzymol. 309, 375–386.[Medline]

Wickner, S., Maurizi, M.R. & Gottesman, S. (1999) Posttranslational quality control: folding, refolding, and degrading proteins. Science 286, 1888–1893.[Abstract/Free Full Text]

Wójcik, C., Yano, M. & DeMartino, G.N. (2004) RNA interference of valosin-containing protein (VCP/p97) reveals multiple cellular roles linked to ubiquitin/proteasome-dependent proteolysis. J. Cell Sci. 117, 281–292.[Abstract/Free Full Text]

Yamanaka, K., Okubo, Y., Suzaki, T. & Ogura, T. (2004) Analysis of the two p97/VCP/Cdc48p proteins of Caenorhabditis elegans and their suppression of polyglutamine-induced protein aggregation. J. Struct. Biol. 146, 242–250.[CrossRef][Medline]

Accepted: 25 April 2007




This article has been cited by other articles:


Home page
 JCBHome page
J.-S. Ju, R. A. Fuentealba, S. E. Miller, E. Jackson, D. Piwnica-Worms, R. H. Baloh, and C. C. Weihl
Valosin-containing protein (VCP) is required for autophagy and is disrupted in VCP disease
J. Cell Biol., December 14, 2009; 187(6): 875 - 888.
[Abstract] [Full Text] [PDF]

Home page
 GENES CELLSHome page
C. Mori-Konya, N. Kato, R. Maeda, K. Yasuda, N. Higashimae, M. Noguchi, M. Koike, Y. Kimura, H. Ohizumi, S. Hori, et al.
p97/valosin-containing protein (VCP) is highly modulated by phosphorylation and acetylation.
Genes Cells, April 1, 2009; 14(4): 483 - 497.
[Abstract] [Full Text] [PDF]

Home page
 JCBHome page
L. Conforti, A. Wilbrey, G. Morreale, L. Janeckova, B. Beirowski, R. Adalbert, F. Mazzola, M. Di Stefano, R. Hartley, E. Babetto, et al.
WldS protein requires Nmnat activity and a short N-terminal sequence to protect axons in mice
J. Cell Biol., February 23, 2009; 184(4): 491 - 500.
[Abstract] [Full Text] [PDF]

Home page
 FASEB J.Home page
Y. Wang, A. B. Meriin, N. Zaarur, N. V. Romanova, Y. O. Chernoff, C. E. Costello, and M. Y. Sherman
Abnormal proteins can form aggresome in yeast: aggresome-targeting signals and components of the machinery
FASEB J, February 1, 2009; 23(2): 451 - 463.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
J.-S. Ju, S. E. Miller, P. I. Hanson, and C. C. Weihl
Impaired Protein Aggregate Handling and Clearance Underlie the Pathogenesis of p97/VCP-associated Disease
J. Biol. Chem., October 31, 2008; 283(44): 30289 - 30299.
[Abstract] [Full Text] [PDF]

Home page
 GENES CELLSHome page
S. Nishikori, K. Yamanaka, T. Sakurai, M. Esaki, and T. Ogura
p97 homologs from Caenorhabditis elegans, CDC-48.1 and CDC-48.2, suppress the aggregate formation of huntingtin exon1 containing expanded polyQ repeat.
Genes Cells, August 1, 2008; 13(8): 827 - 838.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
J. Ren, N. Pashkova, S. Winistorfer, and R. C. Piper
DOA1/UFD3 Plays a Role in Sorting Ubiquitinated Membrane Proteins into Multivesicular Bodies
J. Biol. Chem., August 1, 2008; 283(31): 21599 - 21611.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kobayashi, T.
Right arrow Articles by Kakizuka, A.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Kobayashi, T.
Right arrow Articles by Kakizuka, A.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE ADVANCED SEARCH TABLE OF CONTENTS