Research ArticleNeurodegeneration

Huntingtin promotes mTORC1 signaling in the pathogenesis of Huntington’s disease

See allHide authors and affiliations

Science Signaling  28 Oct 2014:
Vol. 7, Issue 349, pp. ra103
DOI: 10.1126/scisignal.2005633


In patients with Huntington’s disease (HD), the protein huntingtin (Htt) has an expanded polyglutamine (poly-Q) tract. HD results in early loss of medium spiny neurons in the striatum, which impairs motor and cognitive functions. Identifying the physiological role and molecular functions of Htt may yield insight into HD pathogenesis. We found that Htt promotes signaling by mTORC1 [mechanistic target of rapamycin (mTOR) complex 1] and that this signaling is potentiated by poly-Q–expanded Htt. Knocking out Htt in mouse embryonic stem cells or human embryonic kidney cells attenuated amino acid–induced mTORC1 activity, whereas overexpressing wild-type or poly-Q–expanded Htt in striatal neuronal cells increased basal mTOR activity. Striatal cells expressing endogenous poly-Q–expanded Htt showed an increase in the number and size of mTOR puncta on the perinuclear regions compared to cells expressing wild-type Htt. Pull-down experiments indicated that amino acids stimulated the interaction of Htt and the guanosine triphosphatase (GTPase) Rheb (a protein that stimulates mTOR activity), and that Htt forms a ternary complex with Rheb and mTOR. Pharmacologically inhibiting PI3K (phosphatidylinositol 3-kinase) or knocking down Rheb abrogated mTORC1 activity induced by expression of a poly-Q–expanded amino-terminal Htt fragment. Moreover, striatum-specific deletion of TSC1, encoding tuberous sclerosis 1, a negative regulator of mTORC1, accelerated the onset of motor coordination abnormalities and caused premature death in an HD mouse model. Together, our findings demonstrate that mutant Htt contributes to the pathogenesis of HD by enhancing mTORC1 activity.


Huntington’s disease (HD) is associated with an expansion of cytosine-adenine-guanine (CAG) trinucleotide repeats (>36) in exon 1 of the ITI5 gene, causing long N-terminal polyglutamine (poly-Q) tracts in the encoded protein huntingtin (Htt) (1). An N-terminal exon 1 fragment of Htt with expanded poly-Q elicits HD-related motor deficits and pathological phenotypes in mouse models (24). Except for prominent HEAT repeats, Htt has no homology with other proteins (5). Deletion of Htt causes embryonic lethality in mice at around embryonic day (E) 8.5 (6) and impairs vesicular transport and posttranscriptional RNA-mediated silencing (79). Previously, we demonstrated a role for the striatal-enriched guanosine triphosphatase (GTPase) Rhes, an activator of the Ser/Thr kinase mechanistic target of rapamycin (mTOR), in HD (10, 11). Rhes, which binds to an N-terminal fragment of poly-Q–expanded Htt, promotes striatal cell toxicity in multiple cell culture and mouse models of HD (10, 1217). Aside from protection offered by mTOR inhibitors against the motor abnormalities in mice and flies expressing the poly-Q Htt fragment (18, 19), it is of interest to delineate the exact role of mTOR signaling in the consequences of poly-Q–expanded Htt and, in turn, whether or how Htt modulates mTOR activity.

Amino acids activate mTOR and, in association with raptor and other proteins, form mTOR complex 1 (mTORC1), which phosphorylates targets S6 kinase (S6K) and eukaryotic translation initiation factor 4E (eIF4E)–binding protein 1 (4EBP1) at multiple sites (20). The Rag family of small GTPases is implicated in this process, in which they sense amino acids and translocate mTOR to perinuclear locations in close proximity with the small GTPase Rheb, a major activator of mTORC1 (21, 22). Growth factor–induced signaling, which requires amino acid–induced signaling for full activation of mTORC1 (23), also converges on Rheb, presumably through the phosphatidylinositol 3-kinase (PI3K)–Akt pathway, which inactivates tuberous sclerosis proteins 1 and 2 (TSC1/2), a Rheb GTPase–activating protein (24). Consistent with this, the PI3K-Akt pathway inhibitor wortmannin suppresses amino acid–mediated mTORC1 signaling in mammalian cells (25), implying that TSC1/2 may inactivate Rheb to reduce mTORC1 signaling. However, experiments in TSC1/2−/− mouse and human cells indicate that guanosine triphosphate (GTP)–loaded Rheb alone is not sufficient, and some other wortmannin-sensitive factors, acting independently of the classical PI3K-Akt pathway, may cooperate with Rheb to induce mTORC1 activity (2527). Here, we investigated how Htt regulates mTORC1 signaling and the relevance of this mechanism to behavioral deficits in HD.


Htt promotes amino acid–induced mTORC1 activity in the regulation of cell size

Htt, like mTOR, is a ubiquitously expressed protein. To determine the effects of Htt on mTORC1 activity, we used three lines of cultured cells: human embryonic kidney (HEK) 293 cells, mouse embryonic stem (ES) cells, and mouse striatal cells. Knocking down endogenous Htt in HEK293 cells with three different short hairpin RNAs (shRNAs) caused a marked loss of steady-state mTORC1 activity, as measured by phosphorylation of S6K at Thr389 (pS6K) (Fig. 1A). Because amino acids, such as leucine, robustly activate mTORC1, we tested whether Htt modulates amino acid–induced activation of mTORC1, using ES cells that either had or lacked endogenous Htt. Leucine had a lesser effect on mTORC1 activity in Htt knockout (Htt−/−) than in wild-type (Htt+/+) ES cells (Fig. 1B). On the basis of these data, we hypothesized that ectopic overexpression of Htt should potentiate mTORC1. As predicted, expression of full-length Htt containing 23 glutamines (FL-HTT-23Q) in striatal cells expressing endogenous wild-type Htt with 7 glutamines [STHdhQ7/Q7 (28)] markedly increased leucine-induced mTORC1 activity, as assessed by the abundance of pS6K and the phosphorylation of S6 at Ser235/236 (pS6), which was further increased by full-length Htt with an expanded poly-Q tract (FL-HTT-86Q) (Fig. 1C). This effect was reproducible in HEK293 cells (fig. S1A). Striatal cells expressing endogenous mutant Htt with 111 glutamines [STHdhQ111/Q111 (28)] also responded more robustly to leucine-induced mTORC1 compared to STHdhQ7/Q7 striatal cells (Fig. 1D).

Fig. 1 Htt mediates amino acid–induced mTORC1 signaling.

(A) Western blotting of HEK293 cells transfected with one of three human HTT shRNA (H1 to H3) and cultured for 48 hours in full Dulbecco’s modified Eagle’s medium (DMEM). (B to F) Western blotting analysis of mTORC1 targets (pS6K-Thr389, pS6-Ser235/236, or p4EBP1-Ser65) and others as indicated in response to leucine in (B) serum-starved wild-type (Htt+/+) or Htt knockout (Htt−/−) mouse ES cells grown in F12 medium containing all amino acids (AA) (+) or F12 medium lacking l-leucine, l-lysine, and l-methionine (−), or F12 (−) stimulated with Leu (+Leu); (C) serum-starved STHdhQ7/Q7 cells transfected with normal (FL23Q) or poly-Q–expanded (FL86Q) myc-tagged full-length human HTT (1 μg), grown in F12 (−) or stimulated with Leu (+); (D) STHdhQ7/Q7 or STHdhQ111/Q111 cells treated as in (C); (E) STHdhQ7/Q7 cells transfected with human N171-82Q (1 μg), stimulated with Leu as in (C); or (F) HEK293 cells transfected with N171-82Q (0.6 μg) and stimulated with Leu as in (C). Data are means ± SEM from three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001 compared to control, Student’s t test. (G and H) Western blotting of HEK293 cells transfected with (G) N171-82Q (0.6 μg) or (H) poly-Q–expanded ataxin-1 (Ataxin-82Q, 1 μg) and stimulated with leucine (3 mM or as indicated). (I) Cell size analysis in HEK293 cells transfected with N171-82Q (0.6 μg), measured by forward scatter (FSC-H). Data are means ± SEM from three experiments. ***P < 0.001, Student’s t test. Blots are representative of three independent experiments.

The N-terminal portion of Htt with expanded poly-Q repeats can promote HD-related pathology. Therefore, we tested whether the N-terminal mutant Htt containing 82 glutamines (N171-82Q), a fragment that elicits HD pathology in mice (4), can promote mTORC1 activity. N171-82Q potentiated leucine-induced pS6K and pS6, as well as the phosphorylation of 4EBP1 at Ser65 (p4EBP1, another mTORC1 substrate), in both striatal cells and HEK293 cells (Fig. 1, E and F), in a manner possibly influenced by the concentration of both leucine (Fig. 1G) and N171-82Q (fig. S1B). This mutant Htt fragment–mediated effect on mTORC1 appeared to be more potent than that of full-length mutant Htt (fig. S1C). However, this could be due to different transfection efficiencies of the two constructs.

Next, we tested whether ataxin, another poly-Q–containing protein, promotes mTORC1 similar to poly-Q–expanded Htt. Compared to Htt, an overexpression of poly-Q–expanded ataxin (Ataxin-82Q) had a minimum effect on leucine-induced mTORC1 activity in HEK293 cells (Fig. 1H). Because mTORC1 activation is known to promote cell size, we investigated whether N171-82Q might affect the amino acid regulation of cell size (29). We found that the average size of HEK293 cells in cultures expressing N171-82Q was significantly larger than that of the control cells when cultured in serum-free medium containing the full complement of amino acids (Fig. 1I). This finding is consistent with the previous report demonstrating a larger soma size of striatal neuron–like cells expressing FL-HTT-140Q compared to cells expressing FL-HTT-7Q (30). Together, these data indicate that Htt mediates amino acid–induced mTORC1 signaling, and that the poly-Q–expanded fragment of Htt potentiates mTORC1 activity in an expected cellular outcome (cell size).

Htt promotes mTORC1 activity in a Rheb-dependent manner

To determine the potential mechanisms of how Htt may regulate mTORC1 activity, we used inhibitors of mitogen-activated protein kinase (MAPK) and PI3K, two upstream regulators of mTORC1 (31). Whereas the MAPK inhibitor U0126 was ineffective, the PI3K inhibitor wortmannin prevented N171-82Q–induced mTORC1 activity in HEK293 cells to a similar extent as did the mTOR inhibitor rapamycin (Fig. 2A). These effects were also observed in STHdhQ111/Q111 striatal cells (Fig. 2B). Like wortmannin, the Akt inhibitor MK-2206 also blocked mutant Htt–induced mTORC1 activity in response to amino acid stimulation in HEK293 cells (fig. S2). This suggests that the PI3K-Akt pathway, a well-established mTORC1 promoter (25, 32, 33), is crucial for Htt-mediated mTORC1 activity. A direct effect of Htt on the PI3K-Akt pathway is unlikely, because Htt neither appreciably altered the phosphorylation of mTOR at Ser2448, a target of PI3K-Akt signaling (34, 35), nor significantly increased the phosphorylation of Akt at Thr308, a PI3K target phosphosite (figs. S2 and S3). Due to reasons that are yet unclear, striatal cells expressing mutant Htt exhibited low Akt phosphorylation under amino acid–deprived conditions, compared with wild-type Htt expressing striatal cells (Fig. 2B).

Fig. 2 The Htt-mediated mTORC1 pathway is wortmannin-sensitive and Rheb-dependent.

(A) Western blotting analysis of mTORC1 targets (pS6K-Thr389 or pS6-Ser235/236) and others as indicated in HEK293 cells transfected with myc or myc-tagged N171-82Q cultured in F12+ medium containing all amino acids (AA) and treated with vehicle (0.01% dimethyl sulfoxide), rapamycin (100 nM), U0126 (10 μM), or wortmannin (100 nM) for 2 hours. (B) Western blotting of serum-starved STHdhQ7/Q7 cells grown in F12 (−) medium lacking l-leucine, l-lysine, and l-methionine and treated with vehicle, or STHdhQ111/Q111 cells treated with vehicle, rapamycin, U0126, or wortmannin for 2 hours and then stimulated with leucine (3 mM, 10 min). Data are means ± SEM from three experiments. *P < 0.05, **P < 0.01 against vehicle-treated STHdhQ7/Q7 cells; #P < 0.05, ##P < 0.01 against vehicle-treated STHdhQ111/Q111 cells, Student’s t test. (C and D) Western blotting of (C) HEK293 or (D) STHdhQ7/Q7 cells transfected with myc or myc-tagged wild-type (WT) or mutant (D60K) Rheb (0.25 μg) in the presence or absence of N171-82Q (0.5 μg) in F12 (−) medium and stimulated with leucine (3 mM, 10 min). Data are means ± SEM from three experiments. *P < 0.05, **P < 0.01 against Leu-stimulated cells; #P < 0.05, ##P < 0.01 against starved cells. (E and F) Western blotting of HEK293 cells transfected as indicated for 48 hours, cultured in either F12+ or F12− medium for 1 hour, and then stimulated with leucine. In (F), HTT shRNA was the H1 construct. Blots are representative of at least three independent experiments.

Next, we wanted to determine the mechanism by which wortmannin and MK-2206 block Htt-mTORC1 signaling. One possibility is that these inhibitors may be relieving PI3K-Akt–mediated inhibitory restraints on TSC1/2, thus inactivating Rheb GTPase, a major promoter of amino acid–induced mTORC1 (27, 36, 37). Therefore, because Htt was unable to activate Akt but wortmannin blocked Htt-mediated mTORC1, we hypothesized that Htt might be acting downstream of PI3K—for example, in association with Rheb—to increase mTORC1 signaling. If this hypothesis is correct, we reasoned that Rheb and Htt must synergistically activate mTORC1. In support of this notion, we found a more potent activation of leucine-mediated mTORC1 in cells overexpressing both N171-82Q and Rheb. This enhancement is less effective with expression of Rheb D60K, a GTP binding–defective mutant (38), in both HEK293 cells and striatal cells (Fig. 2, C and D). These data suggest that the poly-Q–expanded Htt fragment cooperates with active Rheb to promote mTORC1 signaling. Consistent with this notion, N171-82Q was defective in activating mTORC1 in Rheb-depleted HEK293 cells (Fig. 2E), and Rheb was defective in activating mTORC1 in Htt shRNA–treated cells (Fig. 2F and fig. S4). Together, these data indicate that Htt can promote mTORC1 signaling in cooperation with Rheb.

Htt alters the intracellular localization of mTOR and enhances its interaction with Rheb in a ternary complex

To further dissect the mechanisms involved, we tested whether Htt and Rheb interact. Because the commercial antibodies we used were less optimal in our hands for endogenous coimmunoprecipitation, we used the glutathione S-transferase (GST) affinity pull-down method (10). In HEK293 cells, full-length wild-type Htt (FL-HTT-23Q) readily bound to GST-Rheb in the absence of amino acids, and this interaction increased about twofold in the presence of amino acids (Fig. 3A). Immunocytochemical analysis revealed that amino acids also stimulated the colocalization of Rheb and Htt in HEK293 cells (Fig. 3B). The poly-Q–expanded Htt (FL-HTT-86Q) bound more strongly to Rheb than did FL-HTT-23Q in the presence of amino acids (fig. S5). FL-HTT-86Q and N171-82Q strongly interacted with Rheb under amino acid–deprived conditions, and neither was further enhanced by the addition of amino acids (Fig. 3, C and D). Similarly, there appeared to be less induction of the interaction between Htt and Rheb after amino acid stimulation in mouse striatal cells expressing mutant Htt (STHdhQ111/Q111) than in those expressing wild-type Htt (STHdhQ7/Q7) (Fig. 3E and fig. S6). This indicates that whereas wild-type Htt binds to Rheb in an amino acid–dependent manner, poly-Q–expanded Htt or N-terminal poly-Q–expanded Htt fragments appear to have a strong affinity for Rheb regardless of amino acid conditions. Although the mechanisms are unclear, the poly-Q expansion of Htt is known to confer additional binding affinities, possibly because of altered conformations as demonstrated for other proteins (10, 39, 40).

Fig. 3 Htt binds and colocalizes to mTOR and Rheb in an amino acid–dependent manner.

(A) Pull-down analysis of the Rheb-HTT interaction in HEK293 cells transfected as indicated, deprived of amino acids in Krebs buffer for 1 hour (−), and then stimulated with 1× essential amino acids (AA) for 10 min (+). **P < 0.01, Student’s t test. (B) Colocalization of Rheb and Htt in HEK293 cells expressing hemagglutinin (HA)–Rheb or myc-FL-HTT-23Q, in (−) or (+) amino acids as described in (A). The Pearson’s r correlation was calculated from the average of 30 to 40 cells per group from three experiments. ***P < 0.001, Student’s t test. Scale bar, 20 μm. DAPI, 4′,6-diamidino-2-phenylindole. (C and D) Pull-down analysis of the interaction of Rheb with poly-Q–expanded full-length (C) or N-terminal fragment of (D) HTT in HEK293 cells transfected as indicated, in (−) or (+) amino acids as described in (A). ns, not significant. **P < 0.01, Student’s t test. (E) Western blotting of glutathione binding assay in striatal cells (STHdhQ7/Q7, STHdhQ111/Q111) expressing GST or GST-Rheb in (−) or (+) of amino acids as described in (A). (F) Immunofluorescence analysis of endogenous Htt and mTOR in HEK293 cells in (−) or (+) amino acids as described in (A). The Pearson’s r correlation was calculated from the average of 30 to 40 cells per group from three experiments. ***P < 0.001, Student’s t test. Scale bar, 20 μm. Blots are representative of three experiments. Data are means ± SD from three experiments.

Next, we investigated whether Htt, which normally resides at multiple intracellular locations (similar to mTOR) (41, 42), changes its localization when stimulated with amino acids. In HEK293 cells deprived of amino acids, we found that endogenous Htt and mTOR were dispersed throughout the cytoplasm as granular structures with sparse colocalization (Fig. 3F). When stimulated with amino acids, mTOR formed perinuclear punctate structures, consistent with previous reports (43, 44), whereas Htt also formed a rapid perinuclear accumulation with enhanced colocalization with mTOR (Fig. 3F). Although mTOR is localized to multiple compartments, it is evident that mTOR aggregates are enriched with lysosomal markers, such as LAMP1 or LAMP2, when stimulated with amino acids (22, 45). We tested whether Htt [which is also known to be associated with lysosomes (4649)] also displayed colocalization with LAMP1-positive vesicles. LAMP1 was present throughout the cells, consistent with previous reports (22, 50), and we found enhanced colocalization of Htt with LAMP1 in the perinuclear region in cells upon stimulation with amino acids (fig. S7). Because proper intracellular localization of mTOR is crucial for its activity (42), and Htt regulates intracellular protein trafficking (51, 52), we wondered whether Htt might alter the intracellular localization of mTOR upon amino acid stimulation. In concordance with previous reports (43, 44), mTOR was dispersed similarly throughout the cytoplasm in wild-type and Htt−/− ES cells (Fig. 4A). Upon stimulation of cells with amino acids, mTOR rapidly formed puncta in wild-type ES cells but displayed a markedly reduced tendency to form such puncta in Htt−/− ES cells (Fig. 4A). Because of technical limitations, we were unable to co-stain for LAMP1, but our data suggest that the amino acid–induced movement of mTOR is hindered in Htt-deficient cells. Because mutant Htt potentiates mTORC1 activity, we tested whether or how it modulates mTOR puncta formation in striatal cells. Both the number and the size of amino acid–induced mTOR puncta were significantly increased in mutant Htt–expressing striatal cells compared with those in wild-type striatal cells (Fig. 4B). Together, these data indicate that Htt alters the amino acid–induced intracellular movement of mTOR.

Fig. 4 Htt regulates the intracellular movement of mTOR and forms a ternary complex with Rheb and mTOR.

(A and B) Immunofluorescence analysis of mTOR in (A) WT (Htt+/+) or Htt knockout (Htt−/−) mouse ES cells (scale bar, 8 μm) or (B) STHdhQ7/Q7, STHdhQ111/Q111 mouse striatal cells (scale bar, 20 μm) cultured in Krebs medium (AA−), then stimulated with 1× essential amino acids (AA+) for 10 min. Data are mean numbers of mTOR puncta from 200 cells in each of two independent experiments (A) or from 20 to 30 cells in each of three independent experiments (B). *P < 0.05, **P < 0.01, ***P < 0.001, Student’s t test. (C and D) Pull-down assay (C) and two-step pull-down assay (D) in HEK293 cells transfected as indicated and cultured in amino acid–containing medium (F12+) for 1 hour to detect a ternary complex between Htt, Rheb, and mTOR. Blots are representative of three experiments.

Next, we wondered how this movement of mTOR by Htt contributes to mTORC1 activity. One mechanism would be that Htt might bring mTOR in close proximity to Rheb, analogous to the function of Rag GTPase (43). Previous work indicates that Rheb binds mTOR in an amino acid–dependent manner (53). We tested whether and how Htt influences the interaction between mTOR and Rheb. Using the GST pull-down assay, we found that overexpression of wild-type Htt (FL-HTT-23Q) markedly increased the binding of mTOR to Rheb (Fig. 4C), indicating that Htt promotes the mTOR-Rheb interaction. Consistent with this, a two-step coprecipitation assay revealed a ternary complex formation among transfected Htt, mTOR, and Rheb (Fig. 4D). Together, these data suggest that Htt (i) forms amino acid–dependent perinuclear accumulation, (ii) facilitates the intracellular movement of mTOR, (iii) binds to Rheb and enhances its association with mTOR, and (iv) forms a ternary complex with Rheb and mTOR.

Depletion of TSC1 in an HD mouse model increases behavioral abnormalities and causes premature death

Having established that a poly-Q–expanded full-length or N-terminal fragment of Htt potentiates mTORC1 activity, we aimed to determine if this pathway contributes to the in vivo progression of abnormalities in the N171-82Q transgenic mouse model of HD (herein called N171HD mice) (4). These mice progress through an asymptomatic phase to a symptomatic phase as they age, have a life span of 20 to 24 weeks, and exhibit abnormalities in the striatum that are representative of HD (4). Seven-week-old asymptomatic N171HD mice had no apparent changes in mTORC1 activity in the striatum (Fig. 5A), whereas mTORC1 activity was significantly increased in the striatum of 16-week-old symptomatic N171HD mice (Fig. 5B). To investigate whether or how mTORC1 promotes disease progression in this model, we increased mTORC1 signaling in the striatum of ~8-week-old asymptomatic N171HD mice. To do this, we first crossed wild-type or N171HD mice with mice expressing floxed alleles of TSC1, which encodes a protein that inhibits mTORC1 (54), generating TSC1flox/+/N171HD mice and TSC1flox/+/wild-type littermate controls. Then, using adeno-associated virus (AAV)–Cre injections that covered 40 to 60% of the striatum (fig. S8), we depleted TSC1 selectively in the striatum of these mice, which at 8 weeks old had no observable HD symptoms. As a control, we injected AAV–green fluorescent protein (GFP) into TSC1flox/+/N171HD mice and TSC1flox/+/wild-type littermate controls. In a separate cohort of mice, we confirmed that AAV-Cre injection consistently enhanced mTORC1 activity in the striatum of TSC1flox/+ mice (fig. S9). Two weeks after injection, we subjected the mice to a battery of behavioral tests repeated three times with 4 weeks in between testing (Fig. 5C). AAV-Cre–injected TSC1flox/+/N171HD mice had significant weight loss (fig. S10) and significantly impaired motor coordination and associated phenotypes, assessed by a rotarod test (Fig. 5D); a series of tests examining walking on a ledge, clasping, gait, kyphosis (spine curvature), and tremor (Fig. 5E); and an open field test (Fig. 5F), compared to AAV-GFP–injected TSC1flox/+/N171HD mice. By 4 months of age (~10 weeks after injection), 80% of the AAV-Cre–injected TSC1flox/+/N171HD mice had died from severe HD pathology (Fig. 5G), whereas the AAV-GFP–injected TSC1flox/+/N171HD mice lived despite severe motor defects. Together, these data indicate that activation of mTORC1 in the striatum expedites HD-associated motor phenotypes and death in N171HD mice.

Fig. 5 Depletion of TSC1, an mTORC1 inhibitor, in the striatum of N171HD transgenic mice exacerbates HD-associated behavioral symptoms and causes premature death.

(A and B) Western blotting analysis of mTORC1 targets (pS6K-Thr389 or pS6-Ser235/236) and others as indicated in lysates from the striatum of (A) asymptomatic or (B) symptomatic N171HD mice. Data are means ± SEM from three mice in the nonsymptomatic group, six mice in the symptomatic group, and a corresponding number of age-matched controls. ns, not significant. **P < 0.01, Student’s t test. (C) Experimental timeline of striatal injections and behavioral tests. (D to F) Behavioral analysis of HD mice with striatal-specific knockout of TSC1. Data are means ± SEM of (D) rotarod performance in three trials per day for 4 days, (E) composite performance on various motor tasks (walking on a ledge, clasping, gait, kyphosis, and tremor), and (F) the total distance and velocity of movement in open field tests. *P < 0.05, **P < 0.01, Student’s t test. (G) Kaplan-Meier survival analysis and Wilcoxon rank test of Cre-injected TSC1flox/+/N171HD mice. ***P < 0.001. The numbers of mice assessed in (D) to (G) were as follows: TSC1flox/+/WT;AAV-Cre, n = 7; TSC1flox/+/WT;AAV-GFP, n = 5; TSC1flox/+ /N171HD;AAV-Cre, n = 7; and TSC1flox/+/N171HD;AAV-GFP, n = 6.


We demonstrate a novel functional connectivity between Htt and mTOR, two developmentally important genes that in adult animals promote cell proliferation and cell survival and can facilitate behavioral dysfunction. For instance, the embryos of both mTOR knockout mice and Htt knockout mice fail to survive beyond E7.5 to E8.5 because of impaired proliferation (6, 55). Whereas wild-type Htt promotes neuronal survival in mice (56), abnormal mTOR activity promotes neurodegeneration in diverse disorders (57). Similarly, whereas Htt promotes anxiety and depression-like behaviors in mice (58), mTOR inhibition with rapamycin blocks these behaviors (59). Our finding that Htt potentiates the mTORC1 pathway offers a new perspective on the biological relationship between Htt and mTOR in this evolutionarily conserved nutrient signaling pathway. Our data assemble a working model in which amino acids stimulate the perinuclear accumulation of Htt, its interaction with Rheb and mTOR, and its enhancement of mTORC1 activity (Fig. 6). mTOR is activated by multiple stimuli: growth factors (various), nutrients (amino acids, glucose, lipids), and the intracellular energy status of cells (60). Our data indicate that Htt responds to amino acid–induced mTORC1, but its role in other stimulus-induced mTORC1 signaling requires further investigation. However, amino acids play a crucial role in growth factor–mediated mTORC1 signaling (32, 61). Therefore, in physiological settings, where separation of amino acids and other mTORC1 signaling is difficult, a possibility of Htt orchestrating mTORC1 induced by other stimuli, including growth factors, cannot be ruled out.

Fig. 6 Model for Htt-mediated mTORC1 signaling.

Our model predicts that WT huntingtin (Htt) in the presence of amino acids rapidly accumulates in the perinuclear structures, presumably lysosomes, and facilitates mTOR movement. The majority of HD patients are heterozygous for CAG mutation in Htt, which means they harbor one WT copy of Htt and one copy of poly-Q–expanded Htt. In pathological conditions, the poly-Q–expanded Htt has a conformational change that might sustain Rheb and mTOR in a ternary complex that is further stabilized on perinuclear structures by amino acids signals, leading to sustained mTORC1 activity, through PI3K-Akt-TSC signaling–dependent mechanisms.

Mechanistically, our data also imply that Htt might be involved in the intracellular trafficking of mTOR, which is crucial for mTORC1 activation (42). How poly-Q–expanded Htt potentiates mTOR accumulation is as yet obscure. A growing body of evidence indicates that Htt plays a crucial role in intracellular vesicular trafficking (62). Previously, the Rag family of GTPases has been shown to regulate mTORC1 activation upon amino acid stimulation by altering the intracellular accumulation of mTOR in lysosomes (43, 44, 63). We speculate that Htt might associate with the Rag family of GTPases in regulating the intracellular trafficking of mTOR. This notion is supported by previous findings demonstrating that Htt regulates Rab GTPase activity and its post-Golgi trafficking to lysosomes (8, 48). How does poly-Q Htt maintain sustained mTORC1 activity despite amino acid deprivation? Our data suggest that this may be due to the enhanced affinity of poly-Q Htt for the Rheb/mTOR complex (Fig. 6). This is analogous to the findings by Sancak et al. (43) where Rag mutants that constitutively recruit mTOR to perinuclear locations resist amino acid deprivation–induced loss of mTORC1 activity. Because of conformational changes, poly-Q–expanded Htt binds certain proteins with greater affinity than does wild-type Htt (10, 39, 40). Such altered binding may have an important role in the poly-Q Htt–mediated striatal-specific damage in HD pathology. It remains to be determined what structural component in Htt, in addition to poly-Q, is required for mTORC1 activity in HD.

Our demonstration that striatum-specific deletion of TSC1, which encodes a protein that inhibits mTORC1, exacerbates behavioral deficits and accelerates death in HD mice raises two important questions: Does poly-Q Htt–mTORC1–mediated cellular dysfunction also occur in parts of the brain other than the striatum and in peripheral tissue? Or, if it is only restricted to the striatum, what mechanisms contribute to this striatal selectivity? HD is an age-dependent disorder. Patients born with the Htt mutation develop normally, but the first appearance of symptoms, which is directly proportional to the number of Htt poly-Q repeats, occurs between 30 and 50 years of age. The age-related factors that contribute to this delayed onset of the disease are unclear, but pharmacological studies show that mTORC1 signaling, a major regulator of mammalian life span (64), participates in many neurodegenerative diseases (65). Moreover, expression of hyperactive mTOR kinase in the forebrain can promote cortical neurodegeneration (66).

How might poly-Q Htt–mTORC1 circuitry contribute to the extensive loss of medium spiny neurons in HD (67)? We speculate that poly-Q–expanded Htt, like Rheb and mTOR, is ubiquitously present and may regulate mTORC1 both in the brain and in other peripheral tissues in HD. This notion is consistent with the high mTORC1 activity seen both in the cortex and in the atrophied skeletal muscle of an HD mouse model expressing poly-Q–expanded Htt (19, 68). Yet, poly-Q–expanded Htt may also exert a tissue-specific increase in mTORC1 signaling, for example, by interacting with tissue-specific regulators of mTOR. Previously, we showed that poly-Q Htt interacts with the striatal-enriched SUMO-E3-GTPase Rhes, which SUMOylates poly-Q–expanded Htt and increases its toxicity (10, 17, 6971). Because Rhes also activates mTORC1 (11), we hypothesize that the interaction of poly-Q–expanded Htt with Rhes may lead to abnormally high activation of mTORC1 in the striatum that may cause early and prominent striatal dysfunction in HD (72). Sustained mTORC1 activity may contribute to disease progression through its roles in protein translation, autophagy, or de novo pyrimidine synthesis (73, 74). We surmise that poly-Q–expanded Htt–mediated enhancement of mTORC1 might compromise autophagy, whose dysregulation is implicated in neurodegeneration (75). This notion is also supported by our previous findings showing that poly-Q–expanded Htt blocks Rhes-induced autophagy (70). However, whether this blockade results from enhanced poly-Q–expanded Htt/Rhes–mediated mTORC1 activation or from mutant Htt–mediated inhibition of Rhes-induced Beclin 1– and Bcl-2–dependent (mTOR-independent) autophagy (70) is currently unclear.

Overall, using a new genetic mouse model of HD, TSC1flox/+/N171HD, we demonstrate for the first time that enhanced activation of mTORC1 signaling selectively in the striatum before the onset of the disease leads to severe HD phenotype and premature death. Thus, interfering with the mTORC1 pathway early in the disease process may have therapeutic potential. Together, our data indicate that Htt promotes amino acid–mediated mTORC1 signaling and that the poly-Q–expanded Htt–mTORC1 circuitry may play an important role in the progression of HD.


Reagents, plasmids, and antibodies

Unless otherwise noted, reagents were obtained from Sigma. Myc-tagged full-length human (FL) HTT-23Q and FL-HTT-103Q were from CHDI (Cure Huntington’s Disease Initiative) Foundation Inc. (Biobank at Coriell Institute for Medical Research). FL-HTT-86Q and N171-82Q were subcloned into pCMV-myc vector from FL-HTT 103Q and AAV-GFP-171-82Q backbone, respectively, using a protocol described before (10). Myc-tagged Rheb wild type and Rheb D60K were a gift from K.-L. Guan (University of California, San Diego, San Diego, CA). Ataxin-82Q GFP-tagged construct was from H. Zoghbi (Baylor College of Medicine, Houston, TX). Htt shRNA sequences encoded by lentiviral vectors (NM_002111) were as follows: H1, GCTGCTGACTTGTTTACGAAAC; H2, CCGTGCAGATAAGAATGCTATC; H3, GCACTCAAGAAGGACACAATAC. The scrambled or lentiviral control vectors were from Addgene. Antibodies for Rheb, GST, and myc were obtained from Santa Cruz Biotechnology (130398, SC138, and SC40, respectively). Antibodies against Htt (5656S), mTOR (2972S), pS6K Thr389 (9234S), pS6 Ser235/236 (4858S), p4EBP Ser65 (9451S), pAkt Thr308 (2965P), and p44/42 extracellular signal–regulated kinase 1/2 (4695P) were from Cell Signaling Technology Inc. Glutathione-Sepharose beads were from Amersham Biosciences, and protein G/protein A agarose suspension was obtained from Calbiochem. Rapamycin, U0126, and wortmannin were from Selleckchem.

Cell culture, transfections, and amino acid treatments

HEK293 cells were grown in DMEM (Gibco 11965-092) with 10% fetal bovine serum (FBS), 1% penicillin/streptomycin (pen/strep), and 5 mM glutamine. Briefly, cells were seeded in 3.5- or 6-cm plates. After 24 hours, the cells were transfected with complementary DNA (cDNA) constructs, using PolyFect (Qiagen) as per the manufacturer’s instructions. For the amino acid starvation/stimulation protocol, after 48 hours, the growth medium was replaced with either DMEM/F12 Ham with all amino acids and without FBS (F12+; Sigma, D2906) or, for essential amino acid starvation, DMEM/F12 Ham devoid of l-leucine, l-lysine, and l-methionine without FBS (F12−; Sigma, D9785). Cells were kept in these media for 1 hour and then either lysed or stimulated for 10 min with 3 mM l-leucine (+Leu) unless otherwise noted. Wild-type and Htt knockout Hdhex4/5/ex4/5 mouse ES cells were described previously (76, 77) and cultured in KnockOut DMEM (Invitrogen) containing 15% FBS, pen/strep (50 IU/ml, 50 mg/ml; Invitrogen), GlutaMax (0.2 mM, Invitrogen), MEM nonessential amino acids (0.1 mM, Invitrogen), 2-mercaptoethanol (0.1 mM, Sigma), and leukemia inhibitory factor (1000 U/ml, Millipore) at 37°C in 5% CO2 either on feeder layers of mitotically inactive, γ-irradiated mouse embryonic fibroblasts (Global Stem Sciences) or on 1% gelatin (Millipore). Before Western blotting, the medium was replaced with F12− medium for 1 hour, stimulated with leucine, then lysed. Striatal cells (STHdh) expressing knock-in wild-type Htt with 7 Glu (STHdhQ7/Q7) or expressing knock-in mutant Htt with 111 Glu (STHdhQ111/Q111) (28) were cultured in DMEM as previously described (10). After 1 or 2 days in culture, the medium was replaced with F12− medium (lacking amino acids) for 1 hour, stimulated with leucine, and lysed.

Immunoblotting, pull-down assay, and in vitro binding

For Western blotting, HEK293 cells were directly lysed in 2× NuPAGE LDS sample buffer (Invitrogen) and sonicated. Brain tissue was snap-frozen and then homogenized in radioimmunoprecipitation assay buffer [150 mM NaCl, 1.0% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM tris (pH 8.0)] with 1× complete protease inhibitor cocktail (Roche). For pull-down assays, cells were grown to 70 to 80% confluency in 6- or 10-cm dishes and transfected with the indicated plasmids (1 μg each), and after 48 hours, they were pelleted and lysed in immunoprecipitation (IP) buffer [50 mM tris (pH 7.6), 1% CHAPS, 10% glycerol, 0.5 mM MgCl2, and 0.5 mM CaCl2]. The lysates were run several times through a 26-gauge needle in IP buffer and preincubated with glutathione beads for 1 hour to minimize nonspecific binding. GST-tagged proteins were pulled down with glutathione beads (60 μl/ml slurry) in IP buffer containing 0.5% CHAPS with protease inhibitor cocktail. After 12 hours, the beads were washed in IP buffer containing 0.5% CHAPS and 150 mM NaCl. Protein concentration was measured with BCA (bicinchoninic acid) protein assay reagent (Pierce). For ternary complex detection, a two-step pull-down assay was performed. First, GST-Rheb was precipitated with glutathione beads. Then, 10% of the precipitates was probed for Htt and mTOR, and the remaining 90% was subjected to precipitation with control rabbit immunoglobulin G or Htt antibody to detect mTOR and GST-Rheb. Protein lysates were loaded and separated by SDS–polyacrylamide gel electrophoresis on NuPAGE 4 to 12% bis-tris gels (Invitrogen), transferred onto polyvinylidene difluoride membranes, and probed with previously mentioned antibodies. Secondary antibodies were horseradish peroxidase (HRP)–conjugated (Jackson ImmunoResearch Inc.). Chemiluminescence was detected using WesternBright Quantum chemiluminescent HRP substrate (Advansta).

Cell size measurements

After 24 hours of plating, HEK293 cells were transfected with wild-type myc-Rheb (used as positive control), myc, or N171-82Q Htt. After 48 hours, the medium was replaced with FBS-free medium containing amino acids (F12+) for 12 hours. Cells were trypsinized, resuspended in F12+ medium, and filtered through a 40-μm cell strainer to a final density of 1 × 106 cells/ml. Cell size was analyzed using a flow cytometer (LSR II, BD Biosciences) measuring the forward scatter (FSC-H) of 20,000 events for each sample. Events were gated according to forward scatter and side scatter to exclude debris and aggregates. All scatter size gating criteria were used across all samples. The mean FSC-H of the cell population was calculated and displayed. Data were analyzed using FlowJo software (Tree Star). Each experiment was performed in triplicate.


HEK293 cells were grown on polylysine (0.1 mg/ml)–coated glass coverslips in culture dishes containing full DMEM. After 48 hours of transfection, the medium was changed to Krebs buffer medium [20 mM Hepes, glucose (4.5 g/liter), 118 mM NaCl, 4.6 mM KCl, 1 mM MgCl2, 12 mM NaHCO3, 0.5 mM CaCl2, 0.2% (w/v) bovine serum albumin (BSA)] devoid of serum and amino acids for 1 hour to simulate full starvation conditions. For the stimulation conditions, cells were stimulated for 10 min with 1× amino acid cocktail (Gibco 11130). Cells were washed with cold phosphate-buffered saline (PBS), fixed with 4% paraformaldehyde (PFA; 20 min), treated with 0.1 M glycine, and permeabilized with 0.1% (v/v) Triton X-100 (5 min). After being incubated with blocking buffer [1% normal donkey serum, 1% (w/v) BSA, and 0.1% (v/v) Tween 20 in PBS] for 1 hour at room temperature, cells were stained overnight at 4°C with antibodies against HA (for Rheb) (1:500, 631207, Clontech), myc (for Htt) (1:500, sc-40, Santa Cruz Biotechnology), Htt (1:75, MAB2166, Millipore), and mTOR (1:200, 2983S, Cell Signaling Technology). Alexa Fluor 488– or Alexa Fluor 588–conjugated secondary antibodies (Molecular Probes) were incubated together with the nuclear stain DAPI for 1 hour at room temperature. Glass coverslips were mounted with Vectashield mounting medium (Vector Laboratories). Images were taken with a Leica TCS SP8 confocal microscope.

ES cells were fixed in 4% PFA (Tousimis Research) for 10 min, followed by two brief washes in PBS. Cells were then exposed to 0.1 M glycine in PBS for 5 min and permeabilized with 0.1% Triton X-100 in PBS for an additional 5 min. After three washes with PBS and incubation in blocking solution (0.5% BSA, 1% normal goat serum, and 0.1% Triton X-100 in PBS) for 15 min, cells were incubated overnight at 4°C with mTOR antibody (1:150 in blocking solution, 2983S). After being washed in PBS, cells were labeled with Alexa Fluor 488–conjugated secondary antibodies (Invitrogen) for 60 min and DAPI (1 g/ml) for 5 min and mounted as described for HEK293 cells. Images of 200 cells in seven different visual fields were taken with a Leica SP5 confocal microscope (for ES cell experiment). Images for analysis were selected randomly, and two investigators blinded to the ESC genotypes manually quantified mTOR puncta on the basis of signal intensity and size. The data are presented as the number of puncta per nucleus.

Striatal cells were processed for immunochemistry as described for HEK293 cells after 10 to 15 min in medium with 1× amino acids. Images were analyzed for the number and size of mTOR puncta using Fiji software, as described previously (78). Where indicated, Pearson’s r colocalization threshold was calculated using Fiji software.

Generation of TSC1flox/+/N171HD mice

Mouse protocols were carried out under the guidelines approved by the Institutional Animal Care and Use Committee. Transgenic mice expressing an N-terminally truncated human Htt cDNA that encodes 82 glutamines and encompasses the first 171 amino acids (N171-82Q) [B6C3-Tg(HD82Gln)81Dbo/J mice, herein called N171HD] were obtained from The Jackson Laboratory. Mutant mice containing targeted floxed alleles of TSC1 (STOCK Tsc1tm1Djk/J mice) were also from The Jackson Laboratory. Male N171HD mice were bred with homozygous TSC1-floxed females to derive offspring that were heterozygous for TSC1-floxed allele and N171-82Q Htt (TSC1flox/+/N171HD) or heterozygous for the TSC1-floxed allele with normal Htt (TSC1flox/+/wild type). The genotypes were confirmed by Transnetyx Inc.

Stereotaxic surgeries

For all surgical procedures, 8-week-old mice were anesthetized with constant delivery of isoflurane while mounted in a stereotaxic frame (David Kopf Instruments). Microinjections of AAV-Cre (AAV1.hSynap.HI.eGFP-Cre.WPRE.SV40) and AAV-GFP (AAV1.hSynap.eGFP.WPRE.bGH) (Vector Core, University of Pennsylvania) were injected bilaterally into the striatum according to the following coordinates: medial-lateral (ML) = ±1.50, anterior-posterior (AP) = +1.2, dorsal-ventral (DV) = −3.25/3.75 and ML = ±2.25, AP = 0, DV = −3.25/3.75 from bregma. Virus was injected in 0.5-μl volumes [5.9 × 1012 vg (viral genomes)/ml] per injection site in each animal (4 μl total volume). The animals were allowed to recover for 2 weeks before behavioral testing. The efficacy of the viral injections was determined by GFP expression in the striatum.

Behavioral analysis

Behavioral testing was performed at 10 weeks of age (Fig. 4C), and the investigator was blinded to the animal’s genotype. All behavioral testing was performed during the light phase of the light-dark cycle between 8:00 a.m. and 12:00 p.m. For each week of behavioral testing, rotarod performance was assessed on day 1, open field on day 2, and the battery of behavioral tests on day 3. Rotarod testing was performed using a linear accelerating rotation paradigm (Med Associates Inc.) in three trials separated by 20 min for four consecutive days each month. The mice were placed on the apparatus at 4 rpm and were subjected to increasing rpm, accelerating to 40 rpm over the course of a maximum of 5 min. The overall latency to fall for each mouse was calculated as the average of the three trials across 4 days. Open field activity was assessed in a single 30-min session in which a mouse was placed in the center of a square enclosure, and total distance moved and velocity were quantified by EthoVision XT software (Noldus Information Technology). Behavioral battery testing was adapted from a previous report (79). The battery of tests, each performed in triplicate, measured ledge walking, clasping, gait, kyphosis, and tremor. Individual measures were scored on a scale of 0 (absence of relevant phenotype) to 3 (most severe manifestation), as described before (79) with the addition of testing for tremor. To determine tremor, mice were placed in a clean cage and observed for 30 s. Each mouse was scored as follows: 0, no signs of tremor; 1, present but mild tremor; 2, severe intervals of tremor or constant moderate tremor; 3, outrageous chronic tremor. The composite score was generated as the mean of all five behavioral battery tests. The Scripps Research Institute Florida Institutional Animal Care and Use Committee approved all protocols.

Statistical analysis

Data are presented as means ± SD or SEM where indicated. All experiments were performed at least in triplicate and repeated twice at minimum. Pearson’s r was calculated for immunocytochemistry colocalization. Kaplan-Meier survival plot and Wilcoxon rank test were used for survival analysis. For most other data, statistical analysis was performed using Student’s t test (MS Excel).


Fig. S1. Effect of normal and expanded poly-Q Htt on amino acid–induced mTORC1 activation.

Fig. S2. Expanded poly-Q Htt fragment–mediated mTORC1 is abrogated by an Akt inhibitor.

Fig. S3. Expanded poly-Q Htt fragment does not potentiate phosphorylation of mTOR at Ser2448.

Fig. S4. Htt depletion inhibits Rheb-mediated mTORC1 activation.

Fig. S5. More FL-HTT-86Q than FL-HTT-23Q binds Rheb.

Fig. S6. Rheb binds Htt in striatal cells.

Fig. S7. Htt colocalizes with LAMP1.

Fig. S8. AAV-GFP expression in the striatum of a TSC1flox/+/HD mouse.

Fig. S9. Cre injection in TSC1flox/+/wild-type mice elicits mTORC1 activity.

Fig. S10. TSC1flox/+/HD mice injected with AAV-Cre show severe weight loss.


Acknowledgments: We thank M. Benilous, N. Norton, and T. Miles for administrative support. We are grateful to the people in The Scripps Research Institute, Florida, Jupiter, especially in the Department of Neuroscience, for their continuous support in setting up the laboratory and providing technical support whenever needed. We thank M. Bolton of The Max Planck Florida Institute, Jupiter, for the continued interest and support in this project. Funding: This work was supported by Scripps startup funds (to S. Subramaniam) and O’Keeffe Neuroscience Scholar Award (to W.M.P.). Author contributions: W.M.P. designed the study and carried out most of the Western blotting and mouse behavioral work. N.S. performed immunofluorescence staining and analysis, and S. Swarnkar maintained cells, performed Western blotting, and maintained the mouse colony. W.-C.H. and D.T.P. prepared the striatal injections images and provided technical support for the analysis of behavioral data. M.B. and M.E.M. performed the ES cell work and its analysis. W.M.P. and S. Subramaniam performed binding experiments and contributed to data analysis. S. Subramaniam provided conceptual input and wrote the paper with input from all co-authors. Competing interests: The authors declare that they have no competing interests.
View Abstract

Stay Connected to Science Signaling

Navigate This Article