Inositol Hexakisphosphate Kinases Induce Cell Death in Huntington Disease

Inositol pyrophosphate diphosphoinositol pentakisphosphate is ubiquitously present in mammalian cells and contains highly energetic pyrophosphate bonds. We have previously reported that inositol hexakisphosphate kinase type 2 (InsP6K2), which converts inositol hexakisphosphate to inositol pyrophosphate diphosphoinositol pentakisphosphate, mediates apoptotic cell death via its translocation from the nucleus to the cytoplasm. Here, we report that InsP6K2 is localized mainly in the cytoplasm of lymphoblast cells from patients with Huntington disease (HD), whereas this enzyme is localized in the nucleus in control lymphoblast cells. The large number of autophagosomes detected in HD lymphoblast cells is consistent with the down-regulation of Akt in response to InsP6K2 activation. Consistent with these observations, the overexpression of InsP6Ks leads to the depletion of Akt phosphorylation and the induction of cell death. These results suggest that InsP6K2 activation is associated with the pathogenesis of HD.

Inositol phosphates occur in many animal and plant tissues. The most extensively characterized member of this molecular family is inositol 1,4,5-trisphosphate, which releases calcium from intracellular storage (1,2). Inositol hexakisphosphate (InsP 6 ) 2 is the most abundant inositol polyphosphate in mammalian cells. It is also a precursor of inositol pyrophosphate molecules, such as diphosphoinositol pentakisphosphate (InsP 7 ) (3,4). Three inositol hexakisphosphate kinases (InsP 6 Ks) are primarily responsible, at least in mammalian cells, for InsP 7 synthesis using InsP 6 and ATP as substrates.
Notably, under stress conditions, InsP 6 K2 (an isoform of InsP 6 Ks) strongly induces apoptotic cell death and is translocated from the nucleus to the cytoplasm (5). Inositol pyrophosphates also undergo rapid turnover in cells, suggesting a potential signaling role for their metabolism (6). Several studies have linked inositol pyrophosphates to disparate cellular functions, from vesicular trafficking to telomere maintenance (7)(8)(9). Their involvement in disease processes, such as cancer and diabetes, has also been suggested (10 -12).
Interestingly, InsP 6 Ks were originally purified from rat brain because neuronal tissues possess the highest level of InsP 6 K activity in mammalian tissues (13). However, the mechanism by which inositol pyrophosphates control brain functions and how alterations in InsP 7 signals affect neurological disorders remain unclear.
We have previously reported that InsP 6 Ks promote autophagy (14), a cellular clearance mechanism that occurs in patients with Huntington disease (HD). Therefore, in this study, we investigated whether InsP 6 K signaling is activated in HD lymphoblast cells. We show that inositol pyrophosphates contribute to the HD lymphoblast cell death process, possibly by promoting apoptotic cell death. We also suggest that InsP 6 Ks, and thus inositol pyrophosphates, are associated with the pathophysiology of neurodegenerative disorders.

Immunoelectron Microscopy in HD and Control Lymphoblast
Cells-Lymphoblast cells were prepared from the peripheral blood of an HD patient with an expanded CAG repeat of 82 triplets and a healthy control and were maintained as described previously (15,16). The lymphoblast cells were fixed with 4% paraformaldehyde for 2 h with 0.05% saponin in phosphate buffered saline (PBS) and treated with 2% normal goat serum. Fixed cells were incubated overnight at 4°C with anti-InsP 6 K2 antibody and anti-LC3 antibody, an autophagosome marker. The cells were than stained with 3,3Ј-diaminobenzidine (DAB) (Sigma-Aldrich). Samples were also embedded in Epon, and ultrathin sections were cut using an ultramicrotome (LKB Bromma 2088 UltroTome, Leica Instruments, Bannockburn, IL). The cells were then examined under an electron microscope (JEM-1200EX, Nippon Denshi) (17). The number of LC3-positive cells was counted for both the HD and control lymphoblast cells. The experiments were repeated three times, and the results are expressed as the mean Ϯ S.D. Statistical analysis was performed using analysis of variance. A value of p Ͻ 0.05 was considered statistically significant.
Measurement of InsP 6 and InsP 7 in Both HD and Control Lymphoblast Cells-HD and control lymphoblast cells were incubated at a density of 10 6 cells in the presence of [ 3 H]inositol (final concentration, 100 Ci/ml) for 3 days. The cells were harvested and washed twice with ice-cold PBS. The cell pellets were lysed in 0.1 ml of ice-cold buffer containing 2 M perchloric acid, 0.2 mg/ml InsP 6 , and 2 mM EDTA. The lysates were neutralized with 0.1 ml of buffer containing 1 M K 2 CO 3 and 5 mM EDTA. Inositol phosphates were resolved using HPLC and a PartiSphere strong anion exchange column, and the different species were identified as described previously (18).
Overexpression of InsP 6 K and huntingtin in HEK293 Cells-The pEGFP-InsP 6 K1, pEGFP-InsP 6 K2, and pEGFP-InsP 6 K3 genes; the pEGFP-InsP 6 K2 K/A gene (which is a dominant-negative of the InsP 6 K2 gene); the pDsRed-normal huntingtin gene (17 CAG repeats,htt17); and the pDsRed-mutant huntingtin gene (68 CAG repeats, htt68) were transfected into HEK293 cells by electroporation (Amaxa Nucleofector kit, Lonza Group AG, Basel, Switzerland). The cells containing both GFP and DsRed tags were then sorted by fluorescence activated cell sorting (FACS) (FACSAria, BD Biosciences). We routinely determined the transfection efficiency for double-plasmid vectors to be ϳ30% for the entire cell population. We then performed the following experiments using double-transfected cells.
Overexpression of InP 6 Ks in Human Lymphoblast Cells-The pEGFP-InsP 6 K1, pEGFP-InsP 6 K2, and pEGFP-InsP 6 K3 genes were also transfected into lymphoblast cells from an HD patient and a healthy control. The cells with GFP tags were then sorted by FACS.
Cell Death Assays-Apoptosis was analyzed using propidium iodide and annexin V-biotin (Beckman Coulter, Fullerton, CA). Streptavidin-allophycocyanin (BD Pharmingen) was used to detect annexin V-biotin in the apoptosis analysis. Using FACS, both annexin-positive/propidium iodide-negative cells and annexin-positive/propidium iodide-positive cells were counted as dead cells at the time of sorting and after 9 h of incubation.
Immunoblotting-Cells that were cotransfected with huntingtin (the pDsRed-htt17 and pDsRed-htt68 genes) and InsP 6 Ks (the pEGFP-InsP 6 K1, pEGFP-InsP 6 K2, and pEGFP-InsP 6 K3 genes) were sorted by FACS as described above. The cells were used then for the following experiment. The cells were sonicated with cell lysis buffer (50 mM Tris-HCl (pH 7.4), 1% Triton X-100, 0.5 mM PMSF, 2 mM CaCl 2 , and proteinase mixture), and the protein concentrations were determined with a protein assay kit (Bio-Rad) using bovine serum albumin as a standard. The samples were then separated by gel electrophoresis with a 4 -12% gradient. After electrophoretic transfer to a polyvinylidene difluoride (PVDF) membrane (Immobilon-P, Millipore), the membranes were blocked with 4% bovine serum albumin in PBS. The membranes were then washed and incubated with the primary antibodies at 4°C overnight. After incubation with the pri- mary antibodies, the membranes were washed with PBS and 0.1% Tween 20 and incubated with appropriate horseradish peroxidase-conjugated secondary antibodies (Vector Laboratories) for 2 h at room temperature. The membranes were then examined using an ECL Western blot system (Amersham Biosciences). In this study, we used primary antibodies against Akt and phospho-Akt (Ser-473) (Cell Signaling Technology). Equal protein loading was confirmed using anti-␤-actin antibody (Sigma).

InsP 6 K2 Exists in the Cytoplasm of HD Lymphoblast Cells-
We recently showed that InsP 6 Ks, particularly InsP 6 K2, regulate cell death and promote autophagy (14). During its activation process, InsP 6 K2 is translocated from the nucleus to the cytoplasm (5). Thus, we decided to investigate the localization of the InsP 6 K2 enzyme in HD lymphoblast cells. Immunoelectron microscopy against anti-InsP 6 K2 antibody revealed that InsP 6 K2 was present mainly in the cytoplasm of HD lympho-   1 and 3). In HD lymphoblast cells, a large number of autophagosomes were stained with anti-LC3 antibody (arrows).
blast cells, whereas InsP 6 K2 was localized in the nucleus in control lymphoblast cells (Fig. 1, A and C).
A confocal microscopic analysis of the control and HD lymphoblast cells revealed similar results regarding the distribution of InsP 6 K2, with prevalent nuclear localization in control lymphoblast cells and a dual cytosolic nuclear localization in HD lymphoblast cells (Fig. 1B).
These findings suggest that the translocation of InsP 6 K2 from the nucleus to the cytoplasm in HD lymphoblast cells likely results in enzyme activation, as we showed previously (5). To confirm this hypothesis, we measured the level of InsP 7 in our model system. The labeling of control and HD lymphoblast cells with [ 3 H]inositol and the subsequent analysis of the inositol polyphosphate profile performed using strong anion exchange HPLC revealed that the amount of InsP 7 was higher in HD lymphoblast cells than in control lymphoblast cells (Fig.  2). Usually, InsP 6 K2 is translocated from the nucleus to the cytoplasm under stress conditions. Consequently, our findings suggest that InsP 6 K2 is activated in HD lymphoblast cells, as we reported previously (5).
HD Promotes Autophagosomes-In this study, we investigated autophagy using lymphoblast cells from an HD patient. The LC3 molecule has two forms: LC3-I (with a molecular mass of 18 kDa), and LC3-II (with a molecular mass of 16 kDa). When an autophagosome is generated, LC3 changes from LC3-I to LC3-II, which exists on the membrane of the autophagosome.
Immunoblotting against anti-LC3 antibody showed that autophagosomes were generated in HD lymphoblast cells (Fig. 3A). Immunoelectron microscopy against anti-LC3 antibody revealed that the number of autophagosomes was larger in HD lymphoblast cells than in healthy control lymphoblast cells (Fig.  3C). Furthermore, the number of LC3-positive cells was significantly larger in HD lymphoblast cells compared with healthy control lymphoblast cells (Fig. 3B). These findings suggest that the autophagic pathway is activated in HD lymphoblast cells. Moreover, these findings are consistent with previously published reports demonstrating a relationship between autophagy and HD pathophysiology (20,21).
InsP 6 Ks Induce Cell Death in HD-To investigate the cell death mechanism caused by mutant huntingtins, we transfected HEK293 cell lines with huntingtins the wild-type, possessing 17 CAG repeats (htt17); and a mutant huntingtin, possessing 68 CAG repeats (htt68). The number of cell deaths in the htt68-transfected cells was larger than that in the htt17transfected cells. Cotransfection with InsP 6 K2 significantly increased the cell death ratio between the htt68-and htt17transfected cells, whereas transfection with the dominant-negative InsP 6 K2 gene did not affect the cell death ratio (Fig. 4A).
These data indicated that it is the synthesis of InsP 7 , and not the protein itself, that is responsible for the increase in the cell death ratio. To confirm that InsP 6 Ks induce apoptotic cell death, we performed a FACS apoptosis assay, sorting for double-positive cells (annexin V/propidium iodide); consequently, an increase in apoptotic cell death in htt68-and InsP 6 K2cotransfected cells was observed (Fig. 4B).
Conversely, the depletion of InsP 6 Ks using the siRNA method, especially the depletion of InsP 6 K2, tended to reduce the cell death ratio in huntingtin-cotransfected HEK293 cells, although the change was not significant (Fig. 4C).
The analysis of human lymphoblast cells from control and HD patients revealed, as expected, that the ratio of cell death under normal incubation conditions was larger for the lymphoblast cells from HD patients (Fig. 5A). The transfection of HD lymphoblast cells with the InsP 6 Ks induced an increase in the cell death ratio, and this change was statistically significant for InsP 6 K2 compared with the cell death of control vector-transfected HD lymphoblast cells (Fig. 5A). On the other hand, the depletions of InsP 6 Ks significantly suppressed cell death in HD lymphoblast cells. In particular, the depletion of InsP 6 K2 dramatically decreased cell death (Fig. 5B). Therefore, in both HEK293 cells transfected with mutant huntingtin (htt68) and primary human lymphoblast cells from HD patients, the overexpression of InsP 6 K2 enhanced the cell death ratio. However, a robust reduction in cell death after the depletion of InsP 6 Ks using RNAi was observed only in lymphoblast cells from HD patients, suggesting a higher degree of sensitivity to inositol pyrophosphate signaling in primary human lymphoblast cells.
InsP 6 Ks Induce Depletion of Akt Phosphorylation-Akt possesses prosurvival activities, and its down-suppression might cause cell death. The protein level of Akt was not altered in cells transfected with InsP 6 Ks. However, the phosphorylation level of Akt, and thus its activation, was decreased in cells transfected

. InsP 6 K2 promotes apoptotic cell death in HD lymphoblast cells: overexpression and disruption of InsP 6 Ks in HD lymphoblast cells.
A, in HD human lymphoblast cells, the ratio of cell death was larger than that in control human lymphoblast cells under normal incubation conditions. Moreover, the cell death in InsP 6 K-transfected HD lymphoblast cells was augmented compared with that in control vector-transfected HD lymphoblast cells. Notably, the percentage of cell death in HD lymphoblast cells transfected with InsP 6 K2 was the largest. Control, control vector-transfected HD lymphoblast cells; K1, InsP 6 K1-transfected HD lymphoblast cells; K2, InsP 6 K2-transfected HD lymphoblast cells; K3, InsP 6 K3-transfected HD lymphoblast cells. B, on the other hand, the depletions of InsP 6 Ks significantly suppressed cell death in HD lymphoblast cells (Fig. 4B). Control, control human lymphoblast cells; siK1, disruption of InsP 6 K1 in HD lymphoblast cells; siK2, disruption of InsP 6 K2 in HD lymphoblast cells; siK3, disruption of InsP 6 K3 in HD lymphoblast cells.
with InsP 6 Ks (Fig. 6, A and B). Interestingly, in htt68 cells, the protein expression ratio of phospho-Akt and Akt was significantly lower than that in htt17 cells. Moreover, in InsP 6 K-transfected htt68 cells, the protein expression ratio of phospho-Akt and Akt was significantly lower than that in control vectortransfected htt68 cells (Fig. 6, A and B). These findings reveal that the higher concentration of InsP 7 , generated by InsP 6 Ks, suppressed the Akt signaling pathway by reducing the phospho-Akt level, ultimately inducing cell death.

DISCUSSION
This study provides evidence that InsP 6 K2, which generates InsP 7 , promotes cell death in HD, suggesting the primary role of inositol pyrophosphates in HD pathophysiology. We have shown that a large number of autophagosomes were generated in HD lymphoblast cells, whereas a small number of autophagosomes were present in control lymphoblast cells, similar to the results of our previous report (22). Activated InsP 6 K2 was localized in the cytoplasm of HD lymphoblast cells, controlling the phosphorylation of Akt in HD. Consequently, the Akt-activated phosphorylation of mTOR may be suppressed by elevated levels of InsP 7 , possibly increasing the generation of autophagosomes, whereas the depletion of InsP 6 Ks may suppress the generation of autophagosomes (23,24). The overexpression of InsP 6 K2 in HD lymphoblast cells induced a small increase in apoptotic cell death as determined by FACS analysis. There-fore, inositol pyrophosphates contribute to HD pathophysiology primarily by affecting the autophagic cell death pathway.
We demonstrated previously that multiple cell stressors in diverse cell types all augment InsP 7 formation (5). The deletion of InsP 6 K2, but not the other forms of InsP 6 Ks, diminished cell death. Cell stress led to an 8-fold increase in InsP 7 synthesis, which we detected in both intact cells and cell lysates, with no associated increase in protein levels for any of the three InsP 6 K isoforms. Cytotoxicity was also associated with translocation of InsP 6 K2 from the nucleus to the cytoplasm, whereas the intracellular localization of the other isoforms of the enzyme did not change. Alternatively, the translocation of InsP 6 K2 from the nucleus to the cytoplasm may lead to more detectable enzyme activity (5). Notably, InsP 6 K2 existed in the cytoplasm of lymphoblast cells from an HD patient. These observations suggest that InsP 6 K2 is activated in the lymphoblast cells of HD patients, as demonstrated by an increase in InsP 7 levels. In HD, Akt phosphorylates polyglutamine-expanded huntingtin, and this phosphorylation abrogates its toxicity (25). Therefore, studying this intracellular signaling pathway could help to elucidate the cellular mechanisms that control toxicity in HD and other polyglutamine disorders (26). We have also shown previously that InsP 7 competes with phosphatidylinositol 3,4,5-trisphosphate (PIP 3 ) for binding to several mammalian pleckstrin homology domain-containing proteins, including Akt, PIKE, FIGURE 6. Depletion of Akt in mutant htt cells. A, InsP 6 K-transfected htt68 and htt17 cells were subjected to immunoblot analysis with anti-Akt and anti-phospho-Akt (p-Akt; Ser-473) antibodies. The level of Akt expression in InsP 6 K-transfected htt68 cells was higher than that in InsP 6 K-transfected htt17 cells. On the other hand, the level of phospho-Akt expression in InsP 6 K-transfected htt68 cells tended to be lower than that in InsP 6 K-transfected htt17 cells. C, control vector-transfected HD lymphoblast cells; K1, InsP 6 K1-transfected HD lymphoblast cells; K2, InsP 6 K2-transfected HD lymphoblast cells; K3, InsP 6 K3-transfected HD lymphoblast cells. B, the ratio of phospho-Akt and Akt expression levels in htt68 cells was significantly lower than that in htt17 cells. Moreover, the ratio of phospho-Akt and Akt expression levels in InsP 6 K-transfected htt68 cells was also significantly lower than that in control vector-transfected htt68 cells. The experiments were repeated three times, and the results are expressed as the mean Ϯ S.D. Statistical analysis was performed using an analysis of variance. A value of p Ͻ 0.05 was considered statistically significant. K1, InsP 6 K1; K2, InsP 6 K2; K3, InsP 6 K3. and Tiam (27). Because InsP 7 can compete with Akt for pleckstrin homology domain binding, a high concentration of InsP 7 will occupy more of the pleckstrin homology domain of Akt, thereby inhibiting the phosphorylation of Akt. Consequently, the activation of mTOR, which is downstream of the Akt/phosphoinositide 3-kinase pathway, will be suppressed (28).
In this manner, the generation of autophagosomes is increased. Interestingly, the suppression of mTOR was recently reported to mediate the impairment of synaptic plasticity in Alzheimer disease (29).
On the other hand, the inhibition of mTOR signaling by rapamycin improved cognitive defects in a mouse model of Alzheimer disease (30,31). Rapamycin also improved the transgenic mouse phenotype by reducing the accumulation of amyloid-␤ aggregates through autophagy induction (30,31).
We proposed previously that InsP 6 Ks might act through the synthesis of InsP 7 to induce caspase-independent cell death (14). We also observed an increase in the number of autophagosomes in InsP 6 K-activated cells. These findings reveal that InsP6Ks, especially InsP 6 K2, might at least induce autophagy and cell death concomitantly. The concept of "autophagic cell death" actually remains unsettled (32). Autophagy accompanying apoptosis might represent inefficient clearance of autophagosomes (33).
Our working hypothesis about the pathogenesis of neurodegenerative diseases is that the accumulation of disease-causing toxic proteins or certain external factors causes the activities of InsP 6 Ks to increase, thus inducing the generation of numerous autophagosomes that might facilitate clearance of protein aggregates.
These considerations suggest that mTOR or its upstream signal components, such as Akt and/or InsP 6 Ks, may be promising candidates as therapeutic targets. Further studies using human samples, such as lymphoblast cells obtained from HD patients, are very important because the availability of human samples is limited. The present findings suggest that InsP 6 Ks, especially InsP 6 K2, may be associated with autophagy via the Akt/phosphatidylinositol 1,4,5-trisphosphate pathway.