Trehalose Alleviates Polyglutamine-Mediated Pathology in a Mouse Model of Huntington Disease

Motomasa Tanaka; Yoko Machida; Sanyong Niu; Tetsurou Ikeda; Nihar R Jana; Hiroshi Doi; Masaru Kurosawa; Munenori Nekooki; Nobuyuki Nukina

Disclosures

Nat Med. 2004;10(2) 

In This Article

Results

in Vitro

To screen for small molecules that inhibit the formation of polyglutamine aggregates, we used a mutant myoglobin containing a 35-glutamine repeat (Mb-Gln35) that readily formed aggregates upon incubation at 37 °C (ref. 23). We mixed Mb-Gln35 with potential inhibitors and monitored its aggregation as indicated by absorbance at 550 nm. Among a wide variety of compounds tested (more than 200), a disaccharide, trehalose, caused a statistically significant and dose-dependent reduction of aggregation (Figure 1a). N-acetylgalactosamine tetramer (GalNAcGT), an oligosaccharide, also had similar inhibitory effects. The effects of trehalose and GalNAcGT were comparable to that of the QBP1 peptide, which was previously reported to reduce polyglutamine aggregation in vitro.[25] Because trehalose inhibited polyglutamine aggregation, we investigated the ability of other disaccharides to prevent the formation of polyglutamine aggregates. We found that disaccharides have a general tendency to decrease the polyglutamine-mediated aggregation (Figure 1b).

Inhibition of aggregation and cell death by saccharides in vitro. Error bars, s.e.m. n > 3 for each data set. (a, b) Turbidity of Mb-Gln35 solution in the presence of inhibitors. (c-e) Number of tNhtt-60Q-EGFP cells with green foci (aggregates) in the presence of inhibitors. (f-h) Viability of tNhtt-150Q-EGFP cells in the presence of inhibitors. (a, c, f) (1) 50 µM PBS, (2) 50 µM glucose, (3) 50 µM trehalose, (4) 100 µM trehalose, (5) 1,000 µM trehalose, (6) 50 µM N-acetylgalactosamine tetramer, (7) 50 µM N-acetylneuraminic acid, (8) 25 µM QBP1. (b, d, g) A series of disaccharides (all at 50 µM): (1) PBS, (2) trehalose, (3) sucrose, (4) maltitol, (5) turanose, (6) cellobiose, (7) melibiose, (8) melezitose, (9) mannose. a-d, f, g, *P < 0.05 versus PBS (control). (e) Number of cells with green foci of aggregates in the tNhtt-150Q-EGFP cells in which the indicated protein was overexpressed. *P < 0.01 versus LacZ. (h) Viability of tNhtt-150Q-EGFP cells in which the indicated protein was overexpressed. *P < 0.01 versus mock treatment with ponasterone A and dbcAMP.

To investigate the inhibitory mechanism, we examined the stability of wild-type myoglobin, of Mb-Gln35 and of Mb-Gln12, which contains a shorter 12-glutamine repeat,[23] in the presence or absence of trehalose. To evaluate protein stability, we observed the guanidine hydrochloride-induced unfolding of cyanomyoglobins by monitoring the absorbance at 419 nm.[23] We calculated the concentration of guanidine hydrochloride at the midpoint of the unfolding transition (Cm), which is an index of protein stability[26] ( Table 1 ). The presence of trehalose did not affect the C m value of wild-type myoglobin but resulted in an increase in the C m of Mb-Gln35 as well as a smaller increase in the C m of Mb-Gln12. Thus, trehalose stabilized proteins containing an expanded polyglutamine.

We next investigated the effects of the small molecules selected by the in vitro screening in stable mouse neuroblastoma Neuro2a cells. In these cells, expression of a truncated N-terminal huntingtin (1-90 amino acids) containing 60 or 150 glutamines fused to an enhanced green fluorescence protein (tNhtt-60Q-EGFP, tNhtt-150Q-EGFP) can be induced by 1 µM ponasterone A, and the cells differentiate in response to treatment with 5 mM N6,2'-O-dibutyryladenosine-3',5'-cyclic monophosphate sodium salt (dbcAMP).[27] We found that trehalose decreased tNhtt-60Q-EGFP aggregates in a dose-dependent fashion without any cellular toxicity (Figure 1c). GalNAcGT also inhibited the formation of aggregates. We then tested whether other disaccharides could inhibit polyglutamine aggregation. Although the addition of these disaccharides did not alter the expression of tNhtt-EGFP upon treatment with ponasterone A (data not shown), most of the disaccharides mildly reduced aggregation in tNhtt-60Q-EGFP cells (Figure 1d). To confirm this effect, we transiently overexpressed the Escherichia coli proteins OtsA and OtsB, which produce trehalose intracellularly,[28] in tNhtt-150Q-EGFP cells and examined their inhibitory effects on polyglutamine-induced aggregation. OtsA and OtsB overexpression yielded trehalose in the tNhtt-150Q-EGFP cells (11.8 ± 1.5 nmol/µg protein) and significantly reduced the aggregation (Figure 1e). This decrease was comparable to that resulting from transient overexpression of a molecular chaperone, HDJ-1 (ref. 13; Figure 1e). We also verified, by western blotting, that the addition of trehalose and the overexpression of OtsA and OtsB did not induce HDJ-1, HDJ-2 or Hsp70 in the cellular model of Huntington disease (data not shown).

In addition, we investigated the viability of tNhtt-150Q-EGFP cells in the presence of inhibitors by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay, as more tNhtt-150Q-EGFP cells than tNhtt-60Q-EGFP cells die as a result of tNhtt-EGFP expression.[27] Trehalose produced a dose-dependent increase in cell viability, as compared with PBS control treatment (Figure 1f). Next, we explored cell survival in the presence of various disaccharides. The disaccharides tended to increase cell viability, though the effects were relatively small (Figure 1g). To confirm this protective effect of trehalose, we transiently overexpressed OtsA and OtsB in tNhtt-150Q-EGFP cells and found that this enhanced cell viability by more than 50% (Figure 1h), similar to the results of transient overexpression of HDJ-1 (ref. 13).

Trehalose was the most effective disaccharide among those screened, and its effect was confirmed by overexpressing OtsA and OtsB in the cellular model of Huntington disease (Figure 1e). Therefore, we tested the possible neuroprotective effects of trehalose in a mouse model of Huntington disease, R6/2 transgenic mice.[29] Trehalose (0.2%, 2% or 5%) was added to the drinking water of R6/2 transgenic and wild-type mice, which the mice spontaneously drank. Oral administration of trehalose reduced the weight loss of transgenic mice (Figure 2a), whereas it did not affect the body weight of wild-type mice (data not shown). Because this protective effect was most prominent in mice supplemented with 2% trehalose, we further evaluated the effects of this dosage of trehalose on the R6/2 transgenic mice. First, we examined the influence of trehalose on brain atrophy of the mice. Untreated R6/2 mice showed dilatation of lateral ventricles resulting from striatal atrophy (ratio of the ventricular area to total cerebrum area, 3.1 ± 0.3%), whereas administration of 2% trehalose reduced this ventricular dilatation (ratio 1.8 ± 0.3%, P = 0.008 by unpaired t-test) (Figure 2b). In addition, the average brain weight was also increased by the oral administration of trehalose (0.356 ± 0.006 g for untreated R6/2 mice (n = 11), 0.371 ± 0.006 g for R6/2 mice treated with 2% trehalose (n = 12)). We next investigated, by immunohistochemistry, the effects of trehalose on the formation of polyglutamine aggregates in motor cortex, striatum and liver.[29,30] Visualization of intranuclear inclusions with an antibody to ubiquitin[29] showed that ingestion of 2% trehalose resulted in a substantial decrease in the number of intranuclear aggregates (Figure 2c-e).

Effects of trehalose on polyglutamine-mediated pathology in vivo. (a) Reduction of body weight loss of R6/2 transgenic mice resulting from trehalose treatment. Concentrations of trehalose were 0% (), 0.2% (), 2% () and 5% (). Error bars, s.e.m. n > 8 for each group of mice. (b) Decrease in striatal atrophy of 12-week-old R6/2 transgenic mice resulting from trehalose treatment, as visible in Nissl-stained frozen sections of cerebrum. Shown are sections from R6/2 transgenic mice treated with 0% (left) or 2% trehalose (right). Scale bar, 400 µm. Values are mean ± s.e.m. (c-e) Inhibition of the formation of huntingtin aggregates by trehalose. Representative images of (c) motor cortex and (d) striatum in cerebrum of 8-week-old R6/2 transgenic mice and of (e) liver of 12-week-old R6/2 transgenic mice. Shown are images from R6/2 transgenic mice receiving 0% (left) or 2% trehalose (right). The huntingtin aggregates were visualized with an antibody to ubiquitin. Scale bar, 20 µm.

We compared the number of polyglutamine aggregates in motor cortex, striatum and liver between R6/2 transgenic mice orally administered 2% trehalose and untreated R6/2 mice at 8 and 12 weeks of age. Trehalose administration unambiguously reduced the number of ubiquitin-positive aggregates in the brains of these mice (Figure 3a,b). We did not observe any aggregates or histological abnormalities in the brain of 8- and 12-week-old wild-type mice supplemented with 2% trehalose (data not shown). For liver, we estimated the number of polyglutamine aggregates in 12-week-old R6/2 transgenic mice, because aggregates were not clearly detectable in 8-week-old mice.[30] The number of polyglutamine aggregates in liver was also decreased by the ingestion of 2% trehalose (Figure 3c). We confirmed the presence of trehalose in the homogenates of brain (0.11 ± 0.02 nmol/µg protein) and liver (11.6 ± 1.1 nmol/µg protein) tissues only in the trehalose-supplemented mice. To gain more insights into the effects of trehalose on aggregation, we carried out western blotting analysis on R6/2 transgenic mice at age 8 weeks as well as at 3 weeks, when mice started to drink the trehalose-containing water. We detected aggregates in the stacking gel for untreated and 2% trehalose-supplemented 8-week-old R6/2 transgenic mice, whereas we did not observe any oligomeric states of huntingtin[31] in the high-molecular-weight region of the separation gel even after overexposure of the immunoblot (Figure 3d,e). The 1C2 antibody raised against expanded polyglutamine also did not detect any oligomeric huntingtin (data not shown). Oral administration of 2% trehalose resulted in a reduction in aggregates in the stacking gel (ratio 0.55 ± 0.04 for Figure 3d and 0.57 ± 0.05 for Figure 3e, by densitometric analysis), which was consistent with the decrease in aggregation seen by immunohistochemistry (Figure 3a,b). We observed small amounts of huntingtin aggregates in the stacking gel for R6/2 transgenic mice even at 3 weeks of age (Figure 3e). To confirm this, we carried out an immunohistochemical study of 3-week-old R6/2 transgenic mice. We did not detect ubiquitin-positive aggregates in motor cortex or striatum (data not shown), as reported previously.[3] In addition, 3-week-old R6/2 transgenic mice did not show any pathogenic symptoms. Next, we assessed neuron viability in the motor cortex and striatum of R6/2 transgenic and wild-type mice supplemented with 2% trehalose and in controls not given trehalose. We calculated the total area of neuronal nuclei in motor cortex and striatum using an antibody that recognizes the neuron-specific nuclear protein (NeuN).[32] The area stained by the antibody to NeuN did not differ statistically either between 8- and 12-week-old R6/2 trehalose-supplemented transgenic mice or between trehalose-supplemented and unsupplemented transgenic mice (Figure 3f,g). The NeuN-positive area in R6/2 transgenic mice was also similar to that in wild-type mice (Figure 3f,g), consistent with the observation that neuron loss was not detected throughout the central nervous system in R6/2 transgenic mice.[3,29]

Effects of trehalose on aggregation and neuron viability in vivo. Black and gray bars show the data from 8- and 12- week-old R6/2 transgenic mice, respectively. Error bars, s.e.m. *P < 0.01 versus 0% R6/2 transgenic mice. (a-c) Number of ubiquitin-positive aggregates (per mm2) in (a) motor cortex, (b) striatum and (c) liver. (d, e) Representative western blots for the brain homogenates of 3- and 8-week-old wild-type (WT) and R6/2 transgenic (Tg) mice. Signals were detected with antibodies to huntingtin or ß-tubulin (control) applied to SDS-PAGE gels of (d) total homogenate and (e) its pellet fraction. Molecular sizes are shown at right. (f, g) NeuN-positive area in (f) motor cortex and (g) striatum. n = 9 for mice receiving 0% and n = 8 for mice receiving 2% trehalose.

In addition, we assessed the effects of trehalose on the motor function of 7- to 11-week-old R6/2 transgenic mice by rotarod and footprinting tests. Oral administration of 2% trehalose improved the rotarod performance of R6/2 transgenic mice (Figure 4a) but did not affect that of wild-type mice (latency time >300 s). This evidence of a favorable effect of trehalose was reinforced by footprinting analysis. Average distance of stride and average width of the walking steps of R6/2 transgenic mice were increased and decreased by the ingestion of trehalose, respectively (Figure 4b,c), showing that trehalose improved the walking posture of the mice.

Effects of trehalose on motor function and survival in vivo. (a-d) Data are from R6/2 transgenic mice orally administered 0% () or 2% trehalose (). Error bars, s.e.m. Shown are effects of trehalose on (a) rotarod performance, distance of (b) stride and (c) width in walking steps, and (d) lifespan of R6/2 transgenic mice. a-c, n = 10 for mice receiving 0% and n = 11 for mice receiving 2% trehalose. *P < 0.05 versus mice receiving 0% trehalose. d, n = 13 for mice receiving 0% and n = 15 for mice receiving 2% trehalose; P = 0.0015 by log-rank test.

Because oral administration of 2% trehalose ameliorated the motor dysfunction of R6/2 transgenic mice, we investigated its effect on survival. Oral administration of 2% trehalose extended the lifespan of R6/2 transgenic mice, and the effect was statistically significant (P = 0.0015 by log-rank test; 107.5 ± 2.3 d for 2% trehalose-supplemented R6/2 transgenic mice, 96.6 ± 2.4 d for unsupplemented R6/2 transgenic mice) (Figure 4d).

R6/2 transgenic mice develop diabetes, a feature that mimics the elevated diabetes rate in individuals with Huntington disease.[33,34] We examined the effects of trehalose administration on blood glucose in R6/2 transgenic and wild-type mice. We found that the fasting blood glucose level of 2% trehalose-supplemented R6/2 transgenic mice (157 ± 7.7 mg dl-1) was not statistically different from that of R6/2 mice not receiving the supplement (146 ± 6.5 mg dl-1). Wild-type mice ingesting 2% trehalose also showed a fasting blood glucose level (127 ± 4.5 mg dl-1) comparable to that of untreated wild-type mice (122 ± 6.3 mg dl.-1

Trehalose orally administered to mice is metabolized to glucose, which might have some effects on R6/2 transgenic mice. Therefore, we assessed the effect of oral administration of 2% glucose on the mice. First, we investigated the frequency of foot clasping, a cardinal phenotype of R6/2 transgenic mice,[29] in mice 5 to 7 weeks of age. Administration of 2% trehalose clearly delayed the onset and reduced the frequency of the foot-clasping posture, whereas administration of 2% glucose did not (Figure 5a,b). Next, we examined the amount of aggregation in brain and motor function in treated and untreated R6/2 transgenic mice. Mice administered 2% glucose showed comparable numbers of ubiquitin-positive aggregates in motor cortex and striatum and comparable rotarod performance to untreated mice (Figure 5c-e). In addition, the lifespan of mice supplemented with 2% glucose was similar to that of untreated mice (96.1 ± 1.5 d for mice receiving 2% glucose (n = 9), 96.6 ± 2.4 d for unsupplemented transgenic mice (n = 13)).

Effects of glucose on polyglutamine-mediated pathology in vivo. (a) A typical foot-clasping phenotype of R6/2 transgenic mice with or without 2% trehalose or 2% glucose supplementation. 7-week-old R6/2 mice supplemented with 0% (left) and 2% glucose (right) show the foot-clasping posture when suspended by the tail, whereas an R6/2 transgenic mouse supplemented with 2% trehalose (center) holds its hind limbs outward to steady itself. (b) Frequency of the foot-clasping behavior was investigated for 5- to 7-week-old R6/2 transgenic mice. The frequency of clasping was scored as follows: 3, >10 s; 2, 5-10 s; 1, 0-5 s; 0, 0 s. An average score was calculated in each group of mice; values are shown as mean ± s.e.m. Filled circle, no supplement; filled triangle, 2% trehalose; open triangle, 2% glucose. n = 7 for 0% mice, n = 9 for 2% trehalose-supplemented mice, n = 14 for 2% glucose-supplemented mice. (c, d) Number of ubiquitin-positive aggregates (per mm2) in (c) motor cortex and (d) striatum of 7-week-old R6/2 transgenic mice. n = 5 for mice receiving no supplement, n = 5 for mice receiving 2% trehalose, n = 6 for mice receiving 2% glucose. (e) Effects of glucose on rotarod performance (s) for 7-week-old R6/2 transgenic mice. n = 10 for mice receiving no supplement, n = 11 for mice receiving 2% trehalose, n = 11 for mice receiving 2% glucose. *P < 0.05 versus 0% R6/2 transgenic mice.

processing....