Transcriptional Control of Amino Acid Homeostasis Is Disrupted in Huntington’S Disease

Transcriptional control of amino acid homeostasis is disrupted in Huntington’s disease

Juan I. Sbodioa,1, Solomon H. Snydera,b,c,2, and Bindu D. Paula,1,2

aThe Solomon H. Snyder Department of Neuroscience, The Johns Hopkins University School of Medicine, Baltimore, MD 21205; bDepartment of Pharmacology and Molecular Sciences, The Johns Hopkins University School of Medicine, Baltimore, MD 21205; and cDepartment of Psychiatry, The Johns Hopkins University School of Medicine, Baltimore, MD 21205

Contributed by Solomon H. Snyder, May 25, 2016 (sent for review April 18, 2016; reviewed by Rui Wang and X. William Yang) Disturbances in amino acid metabolism, which have been observed in ensure a constant supply of amino acids and maintain metabolic Huntington’s disease (HD), may account for the profound inanition of homeostasis (18, 20). In the present study we demonstrate deficits HD patients. HD is triggered by an expansion of polyglutamine repeats in the ATF4 response to cysteine deficiency, which can explain the in the protein huntingtin (Htt), impacting diverse cellular processes, defective amino acid disposition, overall metabolic dysfunction, ranging from transcriptional regulation to cognitive and motor func- and inanition of HD patients. tions. We show here that the master regulator of amino acid homeostasis, activating transcription factor 4 (ATF4), is dysfunctional in HD Results because of oxidative stress contributed by aberrant cysteine biosyn- Regulation of CSE and Its Transcription Factor ATF4 by Cysteine Levels. thesis and transport. Consistent with these observations, antioxidant Earlier we reported a deficit of CSE in HD, which could account supplementation reverses the disordered ATF4 response to nutrient for neuropathologic and clinical features of the disease (8). We stress. Our findings establish a molecular link between amino acid used the reverse transsulfuration pathway via which cysteine is disposition and oxidative stress leading to cytotoxicity. This signaling generated as a readout to study ATF4 function in HD (Fig. 1A). cascade may be relevant to other diseases involving redox imbalance Cysteine becomes essential when activity or expression of CSE is and deficits in amino acid metabolism. low. Thus, mice deleted for CSE lose weight on a cysteine-free diet (Fig. S1A) and die within 2 wk (21, 22) after undergoing

Huntington’s disease | cysteine | CSE | oxidative stress | ATF4 massive muscle wastage and associated motor deficits as assessed NEUROSCIENCE by the rotarod assay (Fig. S1B). These abnormalities, reminiscent untington’s disease (HD) is a lethal autosomal dominant neu- of several mouse models of HD, can be reversed by cysteine Hrodegenerative disease caused by an expansion of glutamine supplementation (Fig. S1B and Movie S1). In mouse embryonic repeats in the protein huntingtin (Htt) (1). Mutant huntingtin (mHtt) fibroblasts (MEFs) deprivation of cysteine for as little as 1 h leads elicits toxicity by impacting diverse cellular processes ranging from to increased CSE protein levels (Fig. 1B). To determine the cys- transcriptional regulation to cognitive and motor functions (2–4). teine concentration at which significant induction of CSE occurs Although several pathways affected by mHtt have been extensively without compromising cell viability, we cultured MEFs in media studied, exact mechanisms whereby mHtt disrupts cellular function in containing various cysteine levels (Fig. 1C and Fig. S1C). Re- HD have been unclear. ducing basal concentrations of cysteine from 0.2 mM (regular Abnormalities in amino acid metabolism have been frequently medium) to 0.05 mM doubles CSE levels, whereas in cysteine-free reported in HD, which may explain the weight loss and inanition medium CSE levels are about six times the value in regular observed during disease progression (5–7). We reported a substantial depletion of cystathionine γ-lyase (CSE), the primary biosynthetic Significance enzyme for the amino acid cysteine, in HD (8). Depletion of CSE occurs due to the sequestration of SP1, the basal transcription factor Aberrant amino acid metabolism has been reported in Huntington’s for CSE leading to lowered cysteine levels and elevated oxidative disease (HD), but its molecular origins are unknown. We show here stress (9–11). Increased levels of oxidative stress markers have been that the master regulator of amino acid homeostasis, activating linked to disease progression in HD, in which the antioxidant transcription factor 4 (ATF4), is dysfunctional in HD, reflecting oxi- defense pathways are compromised (12, 13). The CSE/cysteine dative stress generated by impaired cysteine biosynthesis and deficiency appeared to be pathogenic because treatment with transport. Accordingly, antioxidant supplementation reverses the cysteine or its precursor N-acetylcysteine was beneficial (8). diminished ATF4 response to nutrient stress. We identify a molec- Cysteine is a semiessential amino acid with multifaceted cellular ular link between amino acid disposition and oxidative stress that functions (14, 15). Cysteine is obtained from both, the diet and underlies multiple degenerative processes in HD. This disruption by endogenous production via the reverse transsulfuration may be relevant to cellular dysfunction in other neurodegenerative pathway (16, 17). In addition to serving as a building block for conditions involving oxidative stress. Agents that restore cysteine protein synthesis, cysteine is a component of the major antioxi- balance may provide therapeutic benefit. dant glutathione and a substrate for the biosynthesis of the gasotransmitter hydrogen sulfide, which itself has substantial Author contributions: J.I.S., S.H.S., and B.D.P. designed research; J.I.S. and B.D.P. performed antioxidant efficacy (16, 17). Cysteine is also the precursor research; S.H.S. and B.D.P. supervised research; J.I.S. and B.D.P. contributed new reagents/ analytic tools; J.I.S., S.H.S., and B.D.P. analyzed data; and J.I.S., S.H.S., and B.D.P. wrote of sulfur-containing molecules such as taurine and lanthio- the paper. nine, which have cytoprotective functions. In addition to SP1, Reviewers: R.W., Laurentian University; and X.W.Y., Semel Institute and David Geffen School expression of CSE can also be regulated by the transcription of Medicine at the University of California, Los Angeles. factor, activating transcription factor 4 (ATF4) under conditions The authors declare no conflict of interest. of stress (18, 19). Freely available online through the PNAS open access option. ATF4 is the master regulator of amino acid metabolism and is 1J.I.S. and B.D.P. contributed equally to this work. activated by amino acid deprivation (18). Levels of amino acids are 2To whom correspondence may be addressed. Email: [email protected] or bpaul8@jhmi. finely controlled in cells by a balance between their synthesis, up- edu. take, and utilization. ATF4 regulates the expression and activity of This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. several amino acid biosynthetic genes as well as their transporters to 1073/pnas.1608264113/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1608264113 PNAS Early Edition | 1of6 Downloaded by guest on September 26, 2021 Fig. 1. CSE is induced in response to cysteine deprivation and is regulated by ATF4. (A) Schematic representation of the reverse transsulfuration pathway

leading to cysteine and glutathione biosynthesis. CSE uses cystathionine to generate cysteine, which in turn is used to produce hydrogen sulfide (H2S). (B) Kinetics of CSE induction in MEFs. MEFs derived from wild-type (WT) and CSE−/− (CSE KO) mice were incubated in cysteine-free medium, and induction of CSE was monitored at various time points by Western blotting using actin as a loading control. (C) CSE levels are inversely proportional to the cysteine content in the growth medium. MEFs were incubated with media containing varying concentrations of cysteine for 24 h and analyzed by Western blotting. (D)CSE and ATF4 are induced in response to low cysteine. Wild-type and CSE KO MEFs were grown in low-cysteine (LC) medium for 24 h. The levels of CSE, ATF4, SP1, and Actin (loading control) were monitored by Western blotting. (E) Quantitation of D. n = 3 (means ± SEM). *P < 0.05; ***P < 0.001. (F) Schematic representation of SP1 sequestration by mHtt disrupting CSE expression in Huntington’s disease. (G) ATF4 is induced in response to cysteine depletion in Q7, but not in Q111, cells. Q7 and Q111 striatal cells were grown in medium containing different cysteine concentrations for 24 h, and ATF4 induction was monitored by Western blotting. (H) CSE is up-regulated under low-cysteine conditions in Q7, but not in Q111, striatal cells. Cell lysates used in G were used to monitor CSE induction. (I) Striatal Q7 and Q111 cells were grown in low cysteine (0.05 mM) for different time periods, and induction of CSE and ATF4 levels was monitored by Western blotting.

medium (Fig. 1C). We used a 0.05-mM cysteine (low cysteine, Fig. S1E). This decrease observed in the HD striatal cells re- LC) concentration to reflect cysteine deficiency without elic- semblesthatobservedinCSE-depletedMEFs(Fig. S1E). High iting major pathologic changes. ATF4 levels, as occur in amino acid deprivation, induce CSE Under basal conditions SP1 is the principal transcription factor expression. Accordingly, overexpressing ATF4 in Q111 cells regulating levels of CSE, whereas in response to stress, especially rescues CSE expression as well as growth in cysteine-free me- with lowered amino acid levels, ATF4 is a major determinant of CSE dium, consistent with a role for ATF4 in regulating CSE in expression. Induction of CSE and ATF4 occurs in response to cys- striatal cells (Fig. S1 F and G). ATF4 levels are significantly teine deprivation with particularly striking influences upon ATF4 augmented by cysteine depletion at 0.05 mM and progressively (Fig. 1D). For wild-type MEFs, levels of ATF4 are almost ninefold increase at lower cysteine concentrations in wild-type Q7 higher with low-cysteine medium than with regular medium, which is striatal cells, exhibiting a concentration–response relationship significantly greater than in CSE-deleted cells (Fig. 1 D and E). By (Fig. 1G). However, in Q111 cells, low-cysteine concentrations contrast, SP1 levels are essentially the same in wild-type and CSE- fail to induce ATF4. The deficient ATF4 induction in Q111 deleted MEFs with either normal or low-cysteine concentrations cells is paralleled by a similar lack of CSE inducibility (Fig. 1H). (Fig. 1E). ATF4 expression is dynamically and rapidly regulated in The failure of Q111 cells to respond to low-cysteine levels is not response to cysteine (Fig. S1D). Thus, elevated levels of ATF4 an artifact of altered temporal regulation, as at multiple intervals elicited by low cysteine are restored to baseline values within 1 h from 1 to 24 h after low-cysteine exposure, ATF4 and CSE levels of cysteine supplementation. in Q111 cells remain negligible (Fig. 1I). ATF4 expression steadily As HD is characterized by aberrant cysteine metabolism, with increases with time in Q7 cells in 0.05 mM cysteine, whereas in CSE as well as cysteine transporters altered alongside abnormalities Q111 cells, this increase is virtually absent (Fig. 1I). These alter- in amino acid levels (8, 23, 24), we explored the regulation of ATF4 ations of CSE and ATF4 are not attributable to toxic sequelae and its targets in a striatal cell culture model of HD harboring the of altered cysteine concentrations, as cell survival is not markedly expanded glutamine repeats characteristic of mHtt (25). In HD, different in Q7 and Q111 cells in regular medium and is only basal CSE expression is diminished due to sequestration of SP1, the modestly decreased in Q111 cells maintained on low-cysteine basal transcriptional activator for CSE, by mHtt (8) (Fig. 1F and medium (Fig. S1H).

2of6 | www.pnas.org/cgi/doi/10.1073/pnas.1608264113 Sbodio et al. Downloaded by guest on September 26, 2021 NEUROSCIENCE

Fig. 2. Lack of induction of ATF4 in striatal cells expressing mutant huntingtin is specific to cysteine deprivation. (A) Transcriptional targets of ATF4 are not induced in response to cysteine deprivation in Q111 cells. Striatal Q7 and Q111 cells were grown in regular or low-cysteine media, and transcript levels of Cse, xCT, Cat1, and Lat1, which encode CSE, cystine transporter, cationic amino acid transporter 1, and large amino acid transporter 1, respectively, were monitored by real-time qPCR. n = 4 (means ± SEM). ***P < 0.001. (B) Inhibition of cystine uptake also fails to induce ATF4 in Q111 cells. Striatal Q7 and Q111 cells were treated with 0.5 μM erastin (ERA), an inhibitor of xCT, the cystine transporter, to induce cysteine deprivation for 24 h. ATF4 was monitored by Western blotting. (C) Transcriptional targets of ATF4 were not induced in response to erastin treatment in Q111 cells. Striatal Q7 and Q111 cells were grown in regular medium and treated with 0.5 μM ERA. Transcript levels of Cse, xCT, Cat1, and Lat1 were monitored by real-time qPCR. n = 3 (means ± SEM). ***P < 0.001. (D) ATF4 is induced in response to endoplasmic reticulum (ER) stress in both Q7 and Q111 cells. Striatal cells were incubated in regular or low-cysteine (LC) medium or in regular medium (Reg) containing DMSO (vehicle) or 1 μM thapsigargin (TG) for 24 h. ATF4 and SP1, the markers of ER stress (not induced in LC) phosphorylated PERK, the ER chaperone binding Ig protein (BiP), and actin (loading control) levels were monitored by Western blotting. (E) Immunofluorescence analysis depicting the up-regulation and nuclear localization of ATF4 in response to low cysteine and ER stress. Q7 and Q111 cells were incubated as in D and the cells were fixed, permeabilized, and stained for ATF4 (green) and DNA (DAPI). (Magnification, 60×.) (F) The up-regulation of ATF4 and CSE occurs at the transcriptional level. Striatal Q7 and Q111 cells were incubated as in D. Cells were scraped and the RNA was isolated and analyzed by real-time qPCR. n = 4 (means ± SEM). ***P < 0.001; ns: not significant. (G) ATF4 is induced in response to deprivation of other amino acids. Striatal cells were grown in media individually deprived of the amino acids arginine (−R), lysine (−K), glutamine (−Q), or leucine (−L) for 24 h. ATF4 induction was assessed by Western blotting.

Sbodio et al. PNAS Early Edition | 3of6 Downloaded by guest on September 26, 2021 Failure of ATF4 Induction Is Specific to Cysteine Deprivation. We We wondered whether the deficient ATF4 response in HD monitored transcriptional targets of ATF4, such as the genes for xCT, might stem from abnormalities in amino acid sensing, a process the transporter for cystine; cationic amino acid transporter 1 (CAT1), that is reflected by levels of phosphoGCN2 (20) that serve as a a transporter for positively charged amino acids; and LAT1, a marker for amino acid deprivation (Fig. S2A). Cysteine deprivation transporter for large amino acids. Similar to CSE, these genes exhibit elicits a similar augmentation of phosphoGCN2 in Q111 as in Q7 a diminished response to cysteine deprivation in Q111 cells (Fig. 2A). cells, indicating that amino acid sensing is intact in HD (Fig. S2B). As another means of eliciting cysteine deficiency, we treated cells As cysteine is metabolized by CSE to H2S, the diminution of which with erastin, which blocks cystine import by inhibiting xCT (26). The in HD might be pathogenic, we considered whether a decrease induction of ATF4 by erastin in Q7 cells is lost in Q111 cells, in H2S might be responsible for the aberrant ATF4 induction. confirming that the deficient response in HD due to lack of cysteine However, the H2S donors, GYY4137 and NaHS, fail to rescue the islinkedtoregulatorymechanismsforaminoaciddeprivation(Fig. deficient ATF4 responses to low cysteine in Q111 cells (Fig. S2C). 2B). Moreover, erastin treatment fails to activate transcription of the ATF4 targets, Cth (Cse), xCT, Cat1,andLat1 significantly (Fig. 2C). Low Cysteine Promotes Oxidative Stress in HD. What distinguishes Is the deficient response of ATF4 in Q111 cells specific to cys- cysteine from other amino acids? Cysteine is unique among amino teine deprivation? We monitored the induction of ATF4 by other acids in displaying antioxidant properties so that its depletion stress stimuli in striatal cells. We evaluated thapsigargin, a known might elicit oxidative stress. MEFs from CSE knockout cells dis- inducer of ATF4, which elicits endoplasmic reticulum stress by de- play a doubling of basal protein carbonylation as well as enhanced pleting intracellular calcium stores. Thapsigargin markedly induces sensitivity to the oxidant H2O2 (Fig. 3A). Moreover, toxic effects ATF4 to a similar extent in Q7 and Q111 cells in addition to the ER of H2O2 and homocysteic acid, an oxidant and analog of gluta- stress markers, phosphorylated PERK, and binding Ig protein BiP mate, are increased in neurons of CSE-deleted mice (Fig. 3 B and (Fig. 2D). The subcellular localization of the transcription factor C). The depletion of CSE in HD cells appears to influence oxi- ATF4 was also monitored. Immunofluorescence analysis reveals dative stress responses, as oxidative stress is doubled in Q111 cells selective concentration of ATF4 in nuclei with no notable differ- in normal medium and increased 3.5-fold in cells on low cysteine ences in Q7 and Q111 cells in response to thapsigargin and in Q7 (Fig. 3D). Microscopic examination reveals augmented reactive cells in response to low cysteine. No induction of ATF4 is observed oxygen species (ROS) throughout the cells of Q111 preparations in Q111 cells under low cysteine (Fig. 2E). The deficient response of compared with Q7 cells, which is accentuated in response to cys- Q111 cells to low cysteine occurs at the transcriptional level as Atf4 teine deprivation. (Fig. 3E). We compared the extent of oxidative mRNA fails to increase in Q111 cells cultured in low cysteine (Fig. stress elicited by H2O2 in Q7 cells and Q111 cells cultured in low 2F). The disrupted ATF4 response is unique to cysteine deprivation cysteine (Fig. 3F). Stress levels in Q111 cells are almost doubled by as in Q111 cells induction of ATF4 by arginine, lysine, glutamine, low cysteine, reaching levels similar to that generated by 100 μM and leucine deprivation appears normal (Fig. 2G). H2O2 in Q7 cells.

Fig. 3. Elevated ROS is the causative factor for diminished ATF4 response in Q111 striatal cells. (A)CSE−/− (CSE KO) MEFs exhibit elevated oxidative stress. Wild-

type (WT) and CSE KO MEFs were incubated in regular medium or in regular medium containing 100 μMH2O2 for 24 h. Oxidation was assessed by measuring protein carbonylation. n = 3 (means ± SEM). *P < 0.05; **P < 0.01. (B) Primary cortical neurons from CSE KO mice are more vulnerable to oxidative stress. Primary

cortical neurons from wild-type or CSE KO mice were incubated in medium with or without H2O2 for 18 h. Cell viability was determined by MTT assay. n = 3 (means ± SEM). **P < 0.01; ***P < 0.001. (C) Primary cortical neurons from CSE KO mice are more susceptible to the oxidant homocysteic acid (HCA). Primary cortical neurons from wild-type or CSE KO mice were treated with or without HCA for 18 h. Cell viability was assessed by MTT assay. n = 3 (means ± SEM). ***P < 0.001. (D) Striatal Q111 cells have elevated reactive oxygen species. Striatal Q7 and Q111 cells were grown in regular or low-cysteine medium for 24 h. Cells were treated with the fluorescent indicator CellROX green to assay levels of ROS. n = 3 (means ± SEM). ***P < 0.001. (E) Striatal Q7 and Q111 cells were grown in low- cysteine (LC) medium for 24 h and incubated with the redox-sensitive fluorescent indicator CM-H2DCFDA and analyzed by fluorescence microscopy. (Magni-

fication, 10×.) (F) Relative oxidative stress in striatal Q7 and Q111 cells. Q7 cells were incubated with different concentrations of H2O2 for 24 h and compared with Q111 cells incubated in regular or low-cysteine medium for 24 h. Oxidative stress was measured as in D. n = 3 (means ± SEM).

4of6 | www.pnas.org/cgi/doi/10.1073/pnas.1608264113 Sbodio et al. Downloaded by guest on September 26, 2021 NEUROSCIENCE

Fig. 4. Diminished ATF4 induction caused by oxidative stress is reversed by antioxidant supplementation. (A) Oxidative stress suppresses ATF4 response to

cysteine deprivation in Q7 cells. Q7 striatal cells were grown in regular and low-cysteine (LC) medium with or without 100 μMH2O2 for 24 h. ATF4 levels were monitored by Western blotting using actin as a control. (B)QuantitationofA. n = 4(means± SEM). *P < 0.05. (C) Q7 striatal cells were treated as in A and harvested, and Atf4 mRNA levels were analyzed by real-time qPCR. n = 3(means± SEM). ***P < 0.001. (D) Transcriptional targets of ATF4 are diminished in wild-

type striatal Q7 cells under oxidative stress. Striatal Q7 cells were treated with 100 μMH2O2 to induce oxidative stress and grown in low-cysteine medium, and targets of ATF4, Cse, xCT, Cat1,andLat1 were analyzed at the transcript level. n = 3(means± SEM). ***P < 0.001. The induction of these transcripts was decreased under oxidative stress in Q7 cells. (E) Reducing oxidative stress by antioxidant supplementation rescues ATF4 response to cysteine deprivation. Striatal Q7 and Q111 cells were incubated in regular or low-cysteine medium for 24 h. Q111 cells were also grown in low-cysteine medium containing different concentrations of ascorbate for 24 h, and ATF4 levels were monitored by Western blotting. (F) Model for aberrant ATF4 response in HD. Schematic representation of the model for lack of ATF4 response in HD. Early in the disease, levels of CSE, the biosynthetic enzyme for cysteine, an amino acid with major antioxidant properties,arenormal. With disease progression, CSE levels decrease due to sequestration of SP1, the basal transcription factor for CSE, by mHtt. Low levels of CSE and low levels of cysteine transporters observed in HD result in decreased cysteine levels in cells and consequent oxidative stress. The progressive increase in oxidative stress elicited by cysteine deficit would normally be ameliorated by induction and activity of the transcription factor ATF4. Under these stress conditions, ATF4 regulates CSE and other target genes. However, chronic oxidative imbalance, such as that in HD, disrupts the up-regulation of ATF4 in response to stress. Poor inductionofATF4thus results in further elevation of oxidative stress, which leads to a further inhibition of ATF4 up-regulation and disruption of amino acid homeostasis.Thisvicious cycle culminates in reduced viability of cells. (G) Schematic representation of diminished ATF4 response as a function of oxidative stress caused by cysteine deprivation. When cysteine is depleted, ATF4 is induced (depicted by green line). In addition, cysteine deprivation (red line) also results in increased oxidative stress (blue line). The induction of ATF4, which is optimal during low-grade oxidative stress, starts declining with greater accumulation of reactive oxygen species, suggesting a “threshold” beyond which ATF4 induction is compromised.

Sbodio et al. PNAS Early Edition | 5of6 Downloaded by guest on September 26, 2021 Antioxidant Supplementation Restores the ATF4 Response to Cysteine is a component of the major antioxidant glutathione, depletion Deprivation. If oxidative stress underlies the abnormal response of cysteine augments oxidative stress. Low levels of oxidative stress to low cysteine in HD, then these aberrations might be mimicked physiologically enhance ATF4 and CSE expression, responses that by treating normal cells with oxidants. We incubated Q7 cells are disrupted during the chronic oxidative stress of HD. Lack of μ with 100 MH2O2 to induce oxidative stress equivalent to that CSE induction leads to further oxidative stress that, in turn, fur- observed in Q111 cells under low cysteine. Oxidative stress appears ther exacerbates redox imbalance. This cycle thereby maintains to blunt induction of ATF4 by low cysteine. In Q7 cells low cysteine diminished CSE along with heightened oxidative stress and elicits increased ATF4 levels, whereas this increase is diminished in blunted ATF4 response (Fig. 4F). Thus, oxidative stress within cells exposed to low cysteine in the presence of H2O2 (Fig. 4 A and B Atf4 a physiological range can be mitigated by the ATF4 pathway. ). Responses of mRNA are similar to the increases elicited in However, when the oxidative stress persists or exceeds a certain protein levels with low cysteine quadrupling Atf4 mRNA levels in Q7 C level, restorative responses of ATF4 decline. Thus, our results cells, whereas H2O2 substantially diminishes this response (Fig. 4 ). suggest a threshold for oxidative stress beyond which induction Similarly, target genes for ATF4, such as Cse, xCT, Cat1,andLat1, of ATF4 is diminished (Fig. 4G). are up-regulated by low cysteine, an effect prevented by H O 2 2 ATF4 is a prominent regulator for diverse aspects of amino acid treatment (Fig. 4D). The importance of redox balance in regu- disposition, which include metabolism of cysteine and branched lating the ATF4 inducibility during low-cysteine conditions is evident in experiments using the antioxidant ascorbate (Fig. 4E). chain amino acids (18). Depletion of branched chain amino acids In a concentration-dependent fashion, ascorbate increases the has been implicated in muscle shrinkage observed in HD (5). Similarly, a pathogenic role for cysteine deficiency in HD is sup- diminished ATF4 levels of Q111 cells. Thus, elevated oxidative N stress in Q111 cells appears to be responsible for the lack of ported by the beneficial effects of -acetylcysteine (8, 27) and induction of ATF4 in HD (Fig. 4E). couples with evidence for cysteine disturbances in other conditions of oxidative stress including aging, neurodegeneration, AIDS, and Discussion insulin resistance syndrome (28–32). Disturbances of cysteine The principal finding of our study is that disordered amino acid disposition in multiple diseases may reflect a form of “cysteine homeostasis in HD reflects dysfunctions of the transcription stress” with widespread impact. Thus, stimulating the reverse factor ATF4, a master regulator of amino acid disposition. The transsulfuration pathway offers therapeutic avenues for mitigat- perturbation stems from the oxidative stress generated by cys- ing symptoms associated with diseases involving oxidative stress. teine deficits associated with the disease. In addition to depletion of CSE, the biosynthetic enzyme for cysteine, levels of two cys- Materials and Methods teine/cystine transporters, the excitatory amino acid transporter Materials and reagents used as well as details of experimental procedures are 3 (EAAT3/EAAC1), and the solute carrier family 7, member 11 available in SI Materials and Methods and Table S1. All animals were treated in (SLC7A11/xCT), are also diminished in HD (23, 24). ATF4 compliance with the recommendations of the National Institutes of Health regulates transcription of SLC7A11/xCT as well as CSE. These and approved by The Johns Hopkins University Committee on Animal Care. findings may reflect a vicious cycle in HD, which commences ACKNOWLEDGMENTS. We thank Adele Snowman, Lynda Hester, and Lauren with the sequestration of SP1 by mHtt. This leads to decreased Albacarys for technical assistance. This work was supported by US Public Health CSE expression, depressed levels of cysteine, and diminished Service Grant MH18501 (to S.H.S.) and by a grant from the CHDI Foundation antioxidant defenses. As cysteine has antioxidant properties and (to S.H.S. and B.D.P.).

1. The Huntington’s Disease Collaborative Research Group (1993) A novel gene containing a 18. Harding HP, et al. (2003) An integrated stress response regulates amino acid metabolism trinucleotide repeat that is expanded and unstable on Huntington’s disease chromosomes. and resistance to oxidative stress. Mol Cell 11(3):619–633. Cell 72(6):971–983. 19. Dickhout JG, et al. (2012) Integrated stress response modulates cellular redox state via 2. Davies SW, et al. (1997) Formation of neuronal intranuclear inclusions underlies the neuro- induction of cystathionine γ-lyase: Cross-talk between integrated stress response and logical dysfunction in mice transgenic for the HD mutation. Cell 90(3):537–548. thiol metabolism. J Biol Chem 287(10):7603–7614. 3. Sugars KL, Rubinsztein DC (2003) Transcriptional abnormalities in Huntington disease. 20. Kilberg MS, Shan J, Su N (2009) ATF4-dependent transcription mediates signaling of Trends Genet 19(5):233–238. amino acid limitation. Trends Endocrinol Metab 20(9):436–443. 4. Zhai W, Jeong H, Cui L, Krainc D, Tjian R (2005) In vitro analysis of huntingtin-mediated 21. Mani S, Yang G, Wang R (2011) A critical life-supporting role for cystathionine γ-lyase transcriptional repression reveals multiple transcription factor targets. Cell 123(7):1241–1253. in the absence of dietary cysteine supply. Free Radic Biol Med 50(10):1280–1287. 5. Mochel F, Benaich S, Rabier D, Durr A (2011) Validation of plasma branched chain amino 22. Ishii I, et al. (2010) Cystathionine gamma-lyase-deficient mice require dietary cysteine to pro- acids as biomarkers in Huntington disease. Arch Neurol 68(2):265–267. tect against acute lethal myopathy and oxidative injury. J Biol Chem 285(34):26358–26368. 6. Perry TL, Diamond S, Hansen S, Stedman D (1969) Plasma-aminoacid levels in Huntington’s 23. Li X, et al. (2010) Aberrant Rab11-dependent trafficking of the neuronal glutamate chorea. Lancet 1(7599):806–808. transporter EAAC1 causes oxidative stress and cell death in Huntington’s disease. 7. Phillipson OT, Bird ED (1977) Plasma glucose, non-esterified fatty acids and amino J Neurosci 30(13):4552–4561. acids in Huntington’s chorea. Clin Sci Mol Med 52(3):311–318. 24. Frederick NM, et al. (2014) Dysregulation of system xc(-) expression induced by mutant 8. Paul BD, et al. (2014) Cystathionine γ-lyase deficiency mediates neurodegeneration in huntingtin in a striatal neuronal cell line and in R6/2 mice. Neurochem Int 76:59–69. Huntington’s disease. Nature 509(7498):96–100. 25. Trettel F, et al. (2000) Dominant phenotypes produced by the HD mutation in 9. Dunah AW, et al. (2002) Sp1 and TAFII130 transcriptional activity disrupted in early STHdh(Q111) striatal cells. Hum Mol Genet 9(19):2799–2809. Huntington’s disease. Science 296(5576):2238–2243. 26. Dixon SJ, et al. (2012) Ferroptosis: An iron-dependent form of nonapoptotic cell death. Cell 10. Ishii I, et al. (2004) Murine cystathionine gamma-lyase: Complete cDNA and genomic 149(5):1060–1072. sequences, promoter activity, tissue distribution and developmental expression. Biochem J 27. Wright DJ, et al. (2015) N-acetylcysteine improves mitochondrial function and ameliorates 381(Pt 1):113–123. behavioral deficits in the R6/1 mouse model of Huntington’sdisease.Transl Psychiatry 5:e492. 11. Yang G, Pei Y, Teng H, Cao Q, Wang R (2011) Specificity protein-1 as a critical regulator of 28. Hildebrandt W, Kinscherf R, Hauer K, Holm E, Dröge W (2002) Plasma cystine concentration human cystathionine gamma-lyase in smooth muscle cells. JBiolChem286(30):26450–26460. and redox state in aging and physical exercise. Mech Ageing Dev 123(9):1269–1281. 12. Browne SE, Beal MF (2006) Oxidative damage in Huntington’s disease pathogenesis. Antioxid 29. Hack V, et al. (1997) Cystine levels, cystine flux, and protein catabolism in cancer Redox Signal 8(11-12):2061–2073. cachexia, HIV/SIV infection, and senescence. FASEB J 11(1):84–92. 13. Ayala-Peña S (2013) Role of oxidative DNA damage in mitochondrial dysfunction and 30. Dröge W (1993) Cysteine and glutathione deficiency in AIDS patients: A rationale for Huntington’s disease pathogenesis. Free Radic Biol Med 62:102–110. the treatment with N-acetyl-cysteine. Pharmacology 46(2):61–65. 14. Pohlandt F (1974) Cystine: A semi-essential amino acid in the newborn infant. Acta 31. Carter RN, Morton NM (2016) Cysteine and hydrogen sulfide in the regulation of Paediatr Scand 63(6):801–804. metabolism: Insights from genetics and pharmacology. J Pathol 238(2):321–332. 15. Sturman JA, Gaull G, Raiha NC (1970) Absence of cystathionase in human fetal liver: Is 32. Han Y, et al. (2015) Abnormal transsulfuration metabolism and reduced antioxidant capacity cystine essential? Science 169(3940):74–76. in Chinese children with autism spectrum disorders. IntJDevNeurosci46:27–32.

16. Paul BD, Snyder SH (2012) H2S signalling through protein sulfhydration and beyond. 33. Yang G, et al. (2008) H2S as a physiologic vasorelaxant: Hypertension in mice with Nat Rev Mol Cell Biol 13(8):499–507. deletion of cystathionine gamma-lyase. Science 322(5901):587–590.

17. Paul BD, Snyder SH (2015) H2S: A Novel gasotransmitter that signals by sulfhydration. 34. Sbodio JI, Paul BD, Machamer CE, Snyder SH (2013) Golgi protein ACBD3 mediates Trends Biochem Sci 40(11):687–700. neurotoxicity associated with Huntington’s disease. Cell Reports 4(5):890–897.

6of6 | www.pnas.org/cgi/doi/10.1073/pnas.1608264113 Sbodio et al. Downloaded by guest on September 26, 2021