Abstract

   Huntington’s disease (HD) is a devastating neurodegenerative disorder
   that manifests from an N-terminal polyQ-expansion (>35) in the
   Huntingtin (HTT) gene leading to axonal degeneration and significant
   neuronal death. Despite evidence for a scaffolding role for HTT in
   membrane-related processes such as endocytosis, vesicle transport, and
   vesicle fusion, it remains unclear how polyQ-expansion alters membrane
   binding during these processes. Using quantitative Mass
   Spectrometry-based proteomics on HTT-containing light vesicle membranes
   isolated from healthy and HD iPSC-derived neurons, we found significant
   changes in the proteome and kinome of signal transduction, neuronal
   translation, trafficking, and axon guidance-related processes. Through
   a combination of in vitro kinase assays, Drosophila genetics, and
   pharmacological inhibitors, we identified that GSK3β and ERK1
   phosphorylate HTT and that these events play distinct and opposing
   roles during HD with inhibition of GSK3β decreasing polyQ-mediated
   axonal transport defects and neuronal cell death, while inhibition of
   ERK enhancing these phenotypes. Together, this work proposes two novel
   pathways in which GSK3β phosphorylation events exacerbate and ERK
   phosphorylation events mitigate HD-dependent neuronal dysfunction
   highlighting a highly druggable pathway for targeted therapeutics using
   already available small molecules.

   Subject terms: Huntington's disease, Mechanisms of disease

Highlights

     * The proteome and kinome of vesicular/light membranes are
       dysregulated in HD.
     * HTT phosphorylation levels are elevated in HD.
     * GSK3β phosphorylates HTT, contributing to HTT and global axonal
       transport defects.
     * ERK1 phosphorylates HTT and is neuroprotective against HD-mediated
       neuronal death.
     * GSK3β phosphorylation events exacerbate while ERK phosphorylation
       events mitigate HD-dependent neuronal dysfunction

Introduction

   Huntington’s disease (HD) is a neurodegenerative disorder that
   manifests from an expansion of CAG repeats (>36) in the Huntingtin
   (HTT) gene. Although identified more than two decades ago, the normal
   function of HTT remains elusive despite being ubiquitously expressed
   [[36]1], enriched within neurons [[37]1, [38]2], and essential for
   embryonic development [[39]3]. One proposed function is a role as a
   scaffolding protein which is supported by the broad distribution of N-
   and C-terminal HEAT (Huntingtin, Elongator factor3, PR65/A regulatory
   subunit of PP2A, and Tor1) repeats thought to enhance protein-protein
   interactions [[40]4–[41]6]. Indeed, early yeast two-hybrid ([[42]7,
   [43]8]) or immunoprecipitation-mass spectrometry (MS; [[44]9, [45]10])
   studies identified >350 HTT-binding partners linked to gene expression,
   metabolism, proteostasis, endocytosis, and trafficking [[46]11].
   Although polyQ-expansion can dramatically alter the HTT interactome
   [[47]9, [48]12], there is no clear insight yet into the normal
   functions of HTT with these binding partners, thus hindering our
   current understanding of pathways involved in the induction of neuronal
   pathogenesis in HD.

   Despite lacking a transmembrane domain, HTT can localize to membranes
   [[49]13] and can associate with phospholipids [[50]14–[51]16], likely
   through a putative N-terminal amphipathic α-helical domain in the first
   17/18 amino acids of HTT [[52]14, [53]17–[54]20]. In fact, HTT
   localizes across a broad range of different membranous structures:
   vesicles [[55]13, [56]21], plasma membrane [[57]14], endoplasmic
   reticulum, Golgi, and endosomes [[58]17, [59]22–[60]24]. Although work
   has shown that HTT’s association with membranes is altered with
   polyQ-expansion [[61]12, [62]14, [63]15, [64]25, [65]26], little is
   known about the interacting partners at these membranes, hampering our
   ability to identify candidate targets for therapeutic interventions
   aimed at defective pathways such as neural adhesion [[66]27], vesicle
   fusion [[67]28], endocytosis [[68]29], and axonal transport of synaptic
   [[69]30, [70]31] and Rab-containing vesicles [[71]32–[72]34].
   Therefore, elucidating the proteomic landscape of HTT-associated
   membranes under normal and diseased states is critical in unraveling
   the underlying complicated mechanisms of HD pathogenesis.

   Here, we address these gaps in knowledge by testing the hypothesis that
   polyQ-expansion hinders HTT-mediated scaffolding at membranes by
   enriching HTT-containing light membranes (LMs) from healthy (Q17) and
   diseased (Q109) iPSC-derived neurons for quantitative MS analysis. We
   found a dramatic redistribution of kinases at HD LMs, which coincided
   with increased levels of phosphorylated HTT. Although prior MS-based
   studies identified dozens of putative phosphorylation consensus sites
   on HTT [[73]35, [74]36], the functional relevance of these is yet to be
   examined. In addition to AKT, which is well characterized in the
   context of HTT’s normal function and during HD [[75]37, [76]38], we
   identified that polyQ-expansion altered the distribution of several
   kinases: CDK1/5, ERK1/2, and GSK3α/β. Interestingly, the sole putative
   phosphorylation site for GSK3α/β overlaps with one of three putative
   sites for ERK1/2, suggesting a potential interplay between GSK3- and
   ERK-related pathways with normal HTT (Q17) and/or pathogenic (Q109) HTT
   and in HD. By combining in vitro kinase assays and pharmacological
   inhibitors, we identified that GSK3β and ERK1 can phosphorylate both
   normal and pathogenic HTT. In a humanized model of HD in Drosophila
   GSK3β inhibition rescued polyQ-mediated axonal transport defects,
   synaptic dysfunction, and neuronal cell death. In contrast, ERK
   inhibition enhanced neuronal cell death and axonal defects, which were
   rescued by adding excess ERK. Taken together, our observations identify
   previously unknown discrete and opposing roles for ERK and GSK3β
   kinases, with ERK playing an antagonist role in pathogenic HTT-mediated
   cell death and GSK3β exacerbating pathogenic HTT-mediated axonal
   transport defects. This work highlights a novel pathway for potential
   therapeutic interventions that can aid to mitigate early (transport
   deficits), and terminal (cell death) consequences seen in HD.

Results

Pathogenic HTT disrupts the kinome network in membranes

   To test the hypothesis that expansion of polyQ repeats alters the
   scaffolding role of HTT at membranes, we isolated LMs involved in
   trafficking, regulation of trafficking, and downstream events of
   trafficking by using terminally differentiated neurons from Q17 (WT) or
   Q109 (HD) patient-derived iPSCs (iNeurons) (Fig. [77]1A–C; Fig.
   [78]S1A; [[79]33, [80]34]). LMs were first isolated as we have done
   previously using a subcellular fractionation and membrane flotation on
   sucrose step gradient protocol where membrane-associated proteins are
   enriched in the 35/8 sucrose interface membrane fraction with
   syntaxin-1, synaptotagmin1 (SYT), synaptobrevin (SYB) and Rab proteins,
   while Golgi, ER, and mitochondrial proteins are enriched in the pellet
   [[81]34, [82]39–[83]42]. Previous proteomics analysis has shown that
   these membrane proteins are involved in signal transduction and
   membrane trafficking [[84]41].

Fig. 1. The proteomic network on HTT containing light membranes is
dramatically altered in HD iPSC-derived neurons.

   [85]Fig. 1
   [86]Open in a new tab

   A Representative images from healthy/normal (WT, Q17) or diseased (HD,
   Q109) human iPSCs stained with the pluripotent marker OCT-4, the
   neuronal precursor (NPC) marker Nestin and the mature neuronal markers
   MAP2 and βIII-Tubulin. Hoechst stains nuclei. Scale = 25 μm.
   Differentiated neurons show Synaptophysin (SYP) positive staining.
   Scale = 10 μm. B Electrophysiological analysis of WT and HD human
   neurons differentiated from iPSCs show action potentials, which are
   abolished in the presence of TTX or TEA. C Schematic diagram of human
   iNeuron lysate fractionation into perinuclear supernatant (PNS), light
   membrane (LM), soluble (SF), and heavy membrane (P1) fractions by
   ultra-centrifugation and sucrose gradient separation. D Workflow for
   quality control and quantification of unique peptides identified from
   LC-MS of HTT-IPs from WT or HD human iNeurons. E Hierarchical cluster
   heat map showing the avg. relative abundance (spectral count; SpC) of
   800 proteins (≥3 unique peptides/trial across ≥2 biological replicates)
   quantified across the WT and HD HTT-IPs with a normalized fold change
   (FC) threshold of ±2X and a significance threshold of p < 0.05
   determined by a Welch’s t test across three independent biological
   replicates. Increased in HD HTT-IP = red, decreased in HD
   HTT-IP = blue. In addition, proteins were identified in only WT HTT-IP
   (lost = green) or in only HD HTT-IP (gained = orange). F Volcano plot
   with the y axis depicting significance (−log[10][p value]) and the x
   axis depicting fold change of individual peptides between HD and WT
   HTT-IPs (log[2][FC]). Three independent biological replicates were
   performed for each genotype. A negative, no-antibody IP was performed
   to account for non-specific peptide association with magnetic beads. G
   Representative western blot of HTT-IP from WT or HD LMs, probed against
   HTT, KIF5A, KIF5B, KIF5C, DNCT, MAP1B, MAP2, RAB2, RAB5, RAB7, VPS35,
   or SUMO2. Except for KIF5A, all show presence in WT and HD HTT-IP. No
   bands are seen in the negative no antibody control (−Crtl). n = 3.
   Statistical analysis was conducted using the two-sample two-sided
   Student’s t test comparing signal/noise intensity between bands in WT
   and HD conditions normalized to WT. Data represented as mean ± SEM.
   ns = p > 0.05, *p < 0.05, **p < 0.005.

   To evaluate how pathogenic HTT changes the proteomic network of
   HTT-associated membranes, we immunoprecipitated (IP’d) HTT from WT or
   HD LMS isolated from iNeurons and then subjected these samples to MS
   analysis. Our quality control pipeline (Fig. [87]1D–F) required
   identified proteins to be detected in at least two biological
   replicates with the presence of at least three unique peptides
   (spectral count >3). The proteomic network of the HTT-LMs from HD
   iNeurons exhibited 894 gained/increased (orange/red) and 99
   lost/decreased (green/blue) proteins compared to WT iNeurons (Fig.
   [88]1D–F). In contrast, the proteomic network of the total membranes
   isolated from HD iNeurons exhibited 269 gained/increased (orange/red)
   and 1,186 lost/decreased (green/blue) proteins compared to WT iNeurons
   (Fig. [89]S2A–C). Interestingly, HTT showed a decreased presence across
   total LMs in HD iNeurons, suggesting that HTT’s intracellular
   localization and/or membrane association is likely altered with
   pathogenic polyQ-containing HTT. Notably, we isolated previously
   identified targets capable of interacting with HTT, such as
   HTT-associated protein 40 (HAP40; [[90]43]), HTT-interacting protein K
   (HYPK; [[91]44]), HTT-interacting protein 1 (HIP1; [[92]45, [93]46]),
   HTT-associated protein 1 (HAP1; [[94]47]), dynein [[95]48], and
   dynactin [[96]49].

   To better understand the biological relevance of the observed
   HD-mediated proteomic shift, we coupled Gene Ontology (GO) analysis
   with reactome pathway enrichment analysis using clusters of decreased,
   increased, lost, or gained proteins from total LMs (Fig. [97]S3A, C, E,
   G) or HTT LMs (Fig. [98]S3B, D, F, H). We found that pathogenic HTT
   largely affected protein clusters that show enrichment at synapses,
   axons/dendrites, the endomembrane system, and synaptic vesicles (Fig.
   [99]S3A, B, E, F) with involvement in pathways related to translation,
   trafficking, axon guidance, and intracellular signaling (Fig. [100]S3C,
   D, G, H). Having previously identified HTT-relevant roles in the
   trafficking of a sub-type of RAB-containing vesicles [[101]32, [102]33,
   [103]50], we next focused on the validation of our observations.
   Indeed, we observed RAB7 with non-pathogenic HTT LMs with a
   significantly increased association seen in pathogenic HTT LMs (Fig.
   [104]1G), suggesting a possible accumulation of this complex due to
   defective transport and/or endolysosomal degradation mechanisms. We
   further validated significant changes in molecular motor proteins
   kinesin-1 (KIF5C) and dynactin (DNCT), the microtubule-binding proteins
   MAP1B and MAP2, the RAB-GTPases RAB2 and RAB5, and other vesicle or
   trafficking-related proteins such as SYT1, tyrosine hydroxylase (TH),
   and VPS35 (Fig. [105]1G and [106]S2D, E) collectively supporting our
   hypothesis that pathogenic HTT disrupts its scaffolding function on
   membranes. This disruption consequently led to widespread alterations
   in proteomic networks involved in trafficking, and other pathways that
   will need to be further examined (translation, axon guidance, and
   intracellular signaling) in future studies.

   Dysfunctional kinase signaling pathways involving AKT [[107]37,
   [108]51–[109]53], YAP [[110]54], ERK (extracellular signal-related
   kinase; [[111]55, [112]56]), mTOR (mechanistic target of rapamycin;
   [[113]57]), GSK3 (Glycogen synthase kinase 3; [[114]58]), and CDK5
   (cyclin-dependent kinase 5; [[115]59, [116]60]) were proposed to
   contribute to HD pathogenesis. However, aside from AKT [[117]37],
   little is known about these kinase-mediated mechanisms. Our reactome
   pathway enrichment analysis of proteins identified with HTT at
   membranes showed proteins involved in signal transduction and signaling
   pathways were collectively among the top-most represented identifiers
   (Fig. [118]S3C, D, G, H). Therefore, we focused on the human kinome
   with the prediction that pathogenic HTT alters kinase associations on
   membranes. Indeed, we observed a dramatic redistribution of kinases
   associated with HTT on membranes in HD iNeurons (compared to
   normal/WT), with 56 gained/increased and 1 lost/decreased associations
   in HD (Fig. [119]2A–D, right). However, since pathogenic HTT can also
   exert secondary effects that alter the localization/recruitment of
   kinases across membranes that may not necessarily contain HTT, we
   identified 86 lost/decreased and 13 gained/increased kinases across
   total membranes in HD compared to normal (Fig. [120]2A–D, left).
   Furthermore, there were 64 lost/decreased kinases and 5
   gained/increased kinases on total membranes that exhibited no
   association with HTT (Table [121]S1). These findings suggest that
   pathogenic HTT causes a shift in the localization of these kinases
   across membranes. This proposal is further supported by 21 kinases
   showing gained/increased associations with HTT in HD iNeurons compared
   to normal, and a concomitant lost/decreased association across total
   membranes (Table [122]S2), implicating that these proteins are likely
   sequestered with pathogenic HTT.

Fig. 2. Pathogenic HTT triggers abnormal kinase associations with HTT and on
membranes.

   [123]Fig. 2
   [124]Open in a new tab

   A Workflow for quality control and quantification of unique peptides
   identified as kinases from LC-MS of total LMs (left) or HTT-IPs (right)
   from WT or HD iNeurons. B Hierarchical cluster heat map showing the
   avg. relative abundance (spectral count; SpC) of kinases (≥3 unique
   peptides/trial across ≥2 trials) quantified across total LMs (left) and
   HTT-IPs (right) from WT and HD iNeurons with a normalized fold change
   (FC) threshold of ±2X and a significance threshold of p < 0.05
   determined by a Welch’s t test across three independent biological
   replicates. Increased in HD = red, decreased in HD = blue. In addition,
   kinases were identified in only WT (lost = green) or in only HD
   (gained = orange). C Volcano plot with the y-axis depicting
   significance (−log10[p value]) and the x axis depicting fold change of
   individual peptides between LMs (left), or HTT-IPs (right) enriched
   from HD or WT iNeurons (log2[FC]). Three independent biological
   replicates were performed for genotype. A negative, no-antibody IP was
   performed to account for non-specific peptide association with magnetic
   beads. D Depiction of the number of kinases in LMs (left) or HTT-IPs
   (right) that show increased (Red), decreased (blue), lost (Green),
   gained (orange), or unchanged (grey) abundance between WT and HD
   iNeurons. E Representative western blot with quantification of LMs
   (left) or HTT-IPs (right) from WT or HD iNeuron samples, probed with
   AKT1, pAKT (pSer473), ERK1/2, GSK3α/β, or pGSK3α/β (pTyr279/pTyr216),
   which all show presence in WT and HD HTT-IP. No bands are seen in the
   negative no antibody control (−Crtl). n = 3. Statistical analysis was
   conducted using the two-sample two-sided Student’s t test comparing
   signal/noise intensity between bands in WT and HD conditions normalized
   to WT. Data represented as mean ± SEM. ns = p > 0.05, *p < 0.05,
   **p < 0.005.

   Since pathogenic HTT led to altered kinase associations across a broad
   range of kinome families (Fig. [125]S4A), we next performed reactome
   pathway enrichment analysis on the kinases exhibiting HTT-dependent
   associations (Fig. [126]S4B) to begin dissecting the biological
   relevance of these alterations. Notably, we found that pathogenic HTT
   affected kinases involved in PI3K, ERK, mTOR, MAPK, AKT, GSK3, and TP53
   signaling pathways (Fig. [127]S4B). Intriguingly, previous MS-based
   evidence suggested that several of these kinases can putatively
   phosphorylate HTT (Table [128]1, [[129]35]), including AKT1 (Ser421),
   ERK1/2 (Ser2076, Ser2653, and Ser2657), CDK1/5 (Ser1181 and Ser1201),
   or GSK3α/β (Ser2657). While only CDK5 [[130]59, [131]60] and AKT1
   [[132]37] have been validated to directly phosphorylate HTT, we predict
   that the phosphorylation state of HTT at membranes becomes dysregulated
   in HD. Indeed, in addition to the elevated levels of total GSK3α/β at
   both HTT-LMs and total LMs in HD iNeurons (Fig. [133]2C–E, Table
   [134]1), we observed a significant increase in the levels of the
   phospho-active forms of GSK3α/β (pTyr279/pTyr216; [[135]61, [136]62]),
   indicating enhanced kinase activity at these membranes (Fig. [137]2E).
   In contrast, the levels of total AKT1 and active AKT1 (pSer473;
   [[138]63]) were decreased at total membranes in HD iNeurons (Fig.
   [139]2C–E, Table [140]1). The fact that AKT1 can directly inhibit GSK3
   activity [[141]64] could explain our observation of increased active
   GSK3 levels concomitant with decreased active AKT1 levels. Furthermore,
   ERK1 was also decreased at total membranes in HD iNeurons (Fig.
   [142]2C–E, Table [143]1) and has been previously shown to antagonize
   GSK3-mediated signaling in cortical neurons [[144]65]. Although
   AKT1-mediated phosphorylation of HTT (Ser421) has been well
   characterized in the context of scaffolding roles for kinesin-1 and
   dynein during axonal transport [[145]38], it remains unknown whether
   ERK1 or GSK3α/β can phosphorylate HTT and how these phosphorylation
   events contribute to neuronal dysfunction in HD.

Table 1.

   Kinases with putative phosphorylation consensus sequences in HTT
   exhibit abnormal associations with HTT in HD.
   ProteinAC Kinase Description TOTAL LMs HTT-IP LMs
   [146]P31749 AKT1 AKT serine/threonine kinase 1 DECREASED GAINED
   [147]P49841 GSK3β glycogen synthase kinase 3 beta SIG. INCREASED GAINED
   [148]P49840 GSK3α glycogen synthase kinase 3 alpha SIG. INCREASED
   GAINED
   [149]P27361 ERK1/MAPK3 extracellular signal-regulated kinase 1
   DECREASED GAINED
   [150]P06493 CDK1 Cyclin-dependent kinase 1 UNCHANGED SIG. INCREASED
   [151]Q00535 CDK5 Cyclin-dependent kinase 5 UNCHANGED GAINED
   Score Motif Motif Group Site Sequence Gene Info Referenced
   0.367 GSK3 (motif: S*-X3-pS,+4 S) Acidophilic serine/threonine kinase
   group (Acid_ST_kin) S2657 APSSPPTsPVNSRKH GSK3β Schilling et al.
   [[152]35]
   0.405 GSK3 (motif: S*-X3-pS,+4 S) Acidophilic serine/threonine kinase
   group (Acid_ST_kin) S2657 APSSPPTsPVNSRKH GSK3α Schilling et al.
   [[153]35]
   0.431 ERK (motif: PX-S*-P) Proline-dependent serine/threonine kinase
   group (Pro_ST_kin) S2076 DSLSPpSPPVSSHPL ERK1/MAPK3 Schilling et al.
   [[154]35]
   0.541 ERK (motif: PX-S*-P) Proline-dependent serine/threonine kinase
   group (Pro_ST_kin) S2653 APAPSsPPTSPVNSRKH ERK1/MAPK3 Schilling et al.
   [[155]35]
   0.42 ERK (motif: PX-S*-P) Proline-dependent serine/threonine kinase
   group (Pro_ST_kin) S2657 APSSPPTsPVNSRKH ERK1/MAPK3 Schilling et al.
   [[156]35]
   0.374 AKT-1 (motif: XXRXRXX-S*-X) Basophilic serine/threonine kinase
   group (Baso_ST_kin) S421 GGRSRSGsIVEL AKT1 Schilling et al. [[157]35]
   0.448 AKT-1 (motif: XXRXRXX-S*-X) Basophilic serine/threonine kinase
   group (Baso_ST_kin) T2066 LDRFRLStMQDSLSP AKT1
   0.274 CDK1 (motif: X-S*-PXR/K) Proline-dependent serine/threonine
   kinase group (Pro_ST_kin) S1181 LTNPPSLsPIRRKGK CDK1 Schilling et al.
   [[158]35]
   0.334 CDK5 (motif: X-S*-PK/RXR/K) Proline-dependent serine/threonine
   kinase group (Pro_ST_kin) S1181 LTNPPSLsPIRRKGK CDK5 Schilling et al.
   [[159]35]
   0.266 CDK1 (motif: X-S*-PXR/K) Proline-dependent serine/threonine
   kinase group (Pro_ST_kin) S1201 EQASVPLsPKKGSEA CDK1 Schilling et al.
   [[160]35]
   0.372 CDK5 (motif: X-S*-PK/RXR/K) Proline-dependent serine/threonine
   kinase group (Pro_ST_kin) S1201 EQASVPLsPKKGSEA CDK5 Schilling et al.
   [[161]35]
   0.348 CDK5 (motif: X-S*-PK/RXR/K) Proline-dependent serine/threonine
   kinase group (Pro_ST_kin) T1857 WWAEVQQtPKRHSLS CDK5
   0.281 CDK1 (motif: X-S*-PXR/K) Proline-dependent serine/threonine
   kinase group (Pro_ST_kin) S2911 VDRVNVHsPHRAMAA CDK1
   0.327 CDK5 (motif: X-S*-PK/RXR/K) Proline-dependent serine/threonine
   kinase group (Pro_ST_kin) S3124 EVVAAPGsPYHRLLT CDK5
   graphic file with name 41419_2025_7524_Taba_HTML.gif
   [162]Open in a new tab

   List of kinase names and family groups that were present in both HTT-IP
   and LMs of human iNeurons with which the HTT sequence was previously
   proposed [[163]35] to contain putative phosphorylation consensus
   sequences. Note, ERK and GSK3 are both predicted to phosphorylate
   Ser2657 of HTT, while CDK1 and CDK5 are both predicted to phosphorylate
   either Ser1181 or Ser1201. A schematic of the HTT protein illustrates
   the relative locations of these putative phosphorylation sites.

GSK3β and ERK1 both phosphorylate normal and pathogenic HTT

   HD progression is often characterized by cytoplasmic/nuclear
   inclusions, correlated with loss of striatal neurons, and clinical
   decline [[164]66, [165]67]. Synaptic plasticity/transmission
   deficiencies [[166]68–[167]71] and defective axonal transport [[168]30,
   [169]72, [170]73] have been proposed to occur early in disease
   progression, with transport defects arising prior to behavioral defects
   in mice expressing full-length HTT with expanded polyQ repeats
   [[171]74]. Therefore, since pathogenic HTT has been shown to have
   elevated phosphorylation levels compared to non-pathogenic HTT
   [[172]26, [173]53, [174]75], changes in the phosphorylation state of
   HTT could contribute to neuronal dysfunction observed in HD. Indeed,
   significantly increased levels of phospho-isoforms of HTT were seen in
   HD iNeurons compared to normal/WT (Fig. [175]3A, B). Using replicate
   HTT IPs from the LMs from normal/WT or HD iNeurons that were subjected
   to LC-MS (Fig. [176]3C), we conducted in vitro kinase phosphorylation
   assays to test the prediction that GSK3β and/or ERK1 phosphorylate HTT.
   A ~350 kDa band was visible after incubating the HTT-IP with
   recombinant GST-tagged GSK3β and ^32P ATP, suggesting that both normal
   and pathogenic HTT is phosphorylated by GSK3β (Fig. [177]3D-lanes 1,7).
   The addition of a GSK3β specific inhibitor, CHIR99021 [[178]76], which
   we previously used in [[179]77], significantly reduced the levels of
   HTT phosphorylation (Fig. [180]3D, E-lanes 2,8). However, complete
   inhibition was not seen as likely due to the presence of endogenous
   GSK3β and other kinases in the HTT-IP (Figs. [181]3C and [182]2E).

Fig. 3. GSK3β and ERK phosphorylate non-pathogenic and pathogenic HTT in
vitro.

   [183]Fig. 3
   [184]Open in a new tab

   A HTT-IPs from WT and HD iNeurons were run on a Zn^2+-phos-tag™ AAL-107
   gel, transferred via western blot, and probed for HTT. Note the
   increased density of high-molecular weight bands in HD samples
   indicative of multiple phosphoprotein isoforms that were not present in
   WT samples, suggesting increased phosphorylation of HTT occurs in HD
   iNeurons. B Quantification of total HTT signal intensity ≥350 kDa,
   normalized to WT Q17 HTT-IPs. C HTT-IPs from WT and HD iNeurons show
   that increased levels of active GSK3α/β pull down with HTT from HD
   iNeurons compared to WT. D HTT-IPs from WT and HD iNeurons were used as
   a substrate incubated with GST-GSK3β (~74 kDa) for an in vitro kinase
   assay in the presence of 1 mCi/100 γ32P-ATP. To account for a baseline
   level of HTT-phosphorylation by endogenous kinases that IP with HTT,
   1 mCi/100 γ32P-ATP and HTT-IP substrate were incubated without
   GST-GSK3β. The autoradiograph shows a strong band at ~350 kDa,
   indicating the phosphorylation of HTT by GSK3β that is largely
   diminished in the presence of CHIR99021, a GSK3β inhibitor. The 350 kDa
   shows increased intensity with pathogenic HTT from HD iNeurons,
   indicative of increased GSK3β-mediated phosphorylation of HTT. Note
   that a strong band is also observed at ∼47 kDa supporting previous
   findings that GSK3β can auto-phosphorylate itself. n = 3. E
   Quantification of the band intensity at 350 kDa from WT or HD HTT-IPs
   in the presence/absence of CHIR99021. F HTT-IPs from WT and HD iNeurons
   were used as a substrate incubated with GST-ERK1 (~72 kDa) for an in
   vitro kinase assay in the presence of 1 mCi/100 γ32P-ATP. To account
   for a baseline level of HTT-phosphorylation by endogenous kinases that
   IP with HTT, 1 mCi/100 γ32P-ATP and HTT-IP substrate were incubated
   without GST-ERK1. The autoradiograph shows a strong band at ∼350 kDa,
   indicating the phosphorylation of HTT by ERK1 that is largely
   diminished in the presence of SCH772984, an ERK1/2 inhibitor. The
   350 kDa shows increased intensity with pathogenic HTT from HD iNeurons,
   indicative of increased ERK1-mediated phosphorylation of HTT. n = 3. G
   Quantification of the band intensity at 350 kDa from WT or HD HTT-IPs
   in the presence/absence of SCH772984. H HTT-IPs from WT iNeurons were
   used as a substrate incubated with GST-GSK3β or GST-ERK1 for an in
   vitro kinase assay in the presence of 1 mCi/100 γ32P-ATP. To account
   for a baseline level of HTT-phosphorylation by endogenous kinases that
   IP with HTT, 1 mCi/100 γ32P-ATP and HTT-IP substrate were incubated
   without GST-GSK3β or GST-ERK1. The autoradiograph shows a strong band
   at ∼350 kDa that is largely diminished in the presence of CHIR99021 or
   SCH772984. I Quantification of the band intensity at 350 kDa from WT
   HTT-IPs normalized to DMSO-only lanes (lane#1, #4, or #7). Statistical
   analysis was conducted using the two-sample two-sided Student’s t test
   comparing signal/noise intensity between bands in WT and HD conditions
   normalized to WT. Data represented as mean ± SEM. ns = p > 0.05,
   *p < 0.05, **p < 0.005.

   A ~ 350 kDa band was also visible after incubating HTT IP from
   normal/WT or HD iNeurons with recombinant GST-tagged ERK1 and ^32P ATP,
   suggesting that both normal and pathogenic HTT is phosphorylated by
   ERK1 (Fig. [185]3F-lanes 1,8). Adding an ERK-specific inhibitor,
   SCH772984 [[186]78], significantly reduced the levels of HTT
   phosphorylation (Fig. [187]3F, G-lanes 2, 8), but not completely,
   perhaps due to the presence of endogenous ERK1 and other kinases in the
   HTT-IP (Fig. [188]2E). To access the coordination of both GSK3β and
   ERK1 phosphorylation events on HTT, we combined both CHIR99021 and
   SCH772984 inhibitors in the absence of GST-tagged GSK3β or GST-tagged
   ERK1 and failed to observe complete elimination of HTT phosphorylation
   (Fig. [189]3H, I-lane 10), suggesting that other kinases, perhaps AKT1,
   present in the HTT-IP (Fig. [190]2E and Tables [191]1, [192]S1 and
   [193]S2) phosphorylates HTT. The Coomassie-stained gels show equal
   protein levels (Fig. [194]S5A–C), suggesting that the observed
   difference in phosphorylation is likely not due to changes in the
   number of total proteins loaded. Taken together, our observations
   identify that both GSK3β and ERK1 can phosphorylate normal and
   pathogenic HTT. However, the biological relevance of GSK3β- or
   ERK1-mediated phosphorylation of HTT and how these phosphorylation
   events contribute to HD is unknown.

GSK3β-dependent events on HTT contribute to axonal transport and synaptic
morphological defect, and neuronal cell death seen in HD

   Given that GSK3β phosphorylates both normal and mutant HTT (Fig.
   [195]3), and previous MS-based findings identified a putative GSK3β
   phosphorylation site at Ser2657 (Table [196]1, [[197]35]), we next
   tested the hypothesis that active GSK3β contributes to HD pathogenesis.
   To do so, we employed a humanized Drosophila HD model to assess
   GSK3β-dependent effects in a whole organism. Similar to differentiated
   neurons from HD patient-derived iPSCs, larval segmental neurons from
   larvae expressing pathogenic HTT contained axonal blockages that
   stained with HTT and synaptic proteins [[198]33], show disrupted
   transport of RAB proteins [[199]32, [200]34], accumulation of HTT/polyQ
   aggregates and cell death [[201]30, [202]79, [203]80] indicating that
   both human HD iNeurons and Drosophila neurons expressing pathogenic HTT
   show similar dysfunctions. In Drosophila, a specialized neuronal
   network known as the central pattern generator coordinates the
   peristaltic movement of larval muscles, resulting in wave-like
   contractions along the posterior-to-anterior axis of the larvae to
   facilitate larval crawling [[204]81–[205]83]. We previously showed that
   larvae expressing pathogenic human HTT had severe locomotor defects
   together with axonal blockages within their axons and synaptic
   morphological defects at their neuromuscular junctions (NMJs) [[206]33,
   [207]34, [208]79]. To test how active GSK3β influences these
   phenotypes, we first generated larvae expressing either non-pathogenic
   normal HTT (HTTex1.Q25-eGFP) or disease-causing, pathogenic HTT
   (HTTex1.Q103-eGFP) using the pan-neuronal driver APPL-GAL4 and
   inhibited GSK3β by feeding larvae on food laced with either 10 µM
   CHIR99021, a potent inhibitor of GSK3β and an inducer of the
   Wnt/beta-catenin pathway [[209]76], or 0.1% DMSO for 24 h prior to the
   behavioral studies. CHIR99021 was previously characterized in
   Drosophila S2 (Kc167) cells showing activated Wnt signaling at 12.5 µM
   and 25 µM CHIR99021 after 24 h, but not at 6.3 µM [[210]84]. We
   previously showed that 10 µM CHIR99021 but not 5 µM was sufficient to
   produce an axonal blockage phenotype similar to larvae expressing
   dominant negative shaggy (sgg), the Drosophila homolog of GSK3β
   [[211]77, [212]85]. Higher concentrations of CHIR99021 led to embryonic
   lethality, comparable to homozygous loss of function sgg mutants
   [[213]86, [214]87]. Consistent with our previous observations [[215]33]
   larvae expressing pathogenic HTT (Q103) showed significant larval
   locomotor defects compared to non-pathogenic normal HTT (Q25; Fig.
   [216]4A, B). Surprisingly, larvae cultured on food laced with the GSK3β
   inhibitor (CHIR99021) showed significantly attenuated locomotion
   defects (Fig. [217]4A, B). However, larval growth was not affected by
   pathogenic HTT expression or feeding food laced with CHIR99021 (Fig.
   [218]4C).

Fig. 4. GSK3 inhibition mitigates larval locomotion defects, axonal transport
blockages, abnormal synaptic morphology, and elevated neuronal cell death
elicited by pathogenic HTT.

   [219]Fig. 4
   [220]Open in a new tab

   A HTT.Q103-eGFP expressing larva that was fed fly food laced with
   either 0.1% DMSO or 10 µM CHIR99021 for 24 h were subjected to larval
   crawling assays for 1.5 mins at 25 °C, 60% humidity and compared to
   DMSO-treated HTT.Q25-eGFP larvae. n = 20. Bar = 1 cm. B Quantification
   of the avg. larval crawling velocity (cm/min) comparing
   CHIR99021-feeding to DMSO-treatment of HTT.Q103-eGFP expressing larvae
   and DMSO-treated HTT.Q25-eGFP expressing larvae. C Quantification of
   avg. larval size (area; cm^2) normalized to HTT.Q25 expressing larvae
   (AU) comparing CHIR99021-feeding to DMSO-treatment of HTT.Q103-eGFP
   expressing larvae and DMSO-treated HTT.Q25-eGFP expressing larvae. D
   Schematic of Drosophila larval nervous system showing brain, segmental
   nerves, and neuromuscular junctions (NMJs). Red box = imaged area.
   Representative images of NMJs from muscle 6/7 segment A4–5 of
   HTT.Q25-eGFP or HTT.Q103-eGFP expressing larva that were fed fly food
   laced with either 0.1% DMSO or 10 µM CHIR99021 for 24 h prior to
   dissection and staining with HRP-TxRED. Top panels = HRP-TxRED
   staining. Bottom Panel = HTT-eGFP fluorescence. n = 6 larvae.
   Bar = 15μm. Note an even distribution of colocalization (yellow arrows)
   between HTT.Q25 (green) at boutons (red) is observed, which is lost in
   the HTT.Q103 expressing larva. E Quantification of the avg. synapse
   length (μm) per larvae comparing CHIR99021-feeding to DMSO-treatment of
   HTT.Q25-eGFP or HTT.Q103-eGFP expressing larvae. F Quantification of
   the avg. HTT-eGFP (green) signal intensity at boutons normalized to HRP
   (red) signal intensity (AU) per larvae comparing CHIR99021-feeding to
   DMSO-treatment of HTT.Q25-eGFP or HTT.Q103-eGFP expressing larvae. G
   Schematic of Drosophila larval nervous system showing brain and
   segmental nerves. Red box = imaged area. Representative larval
   segmental nerves stained with the synaptic vesicle marker Cysteine
   String Protein (CSP) from HTT.Q25-eGFP (non-pathogenic) or
   HTT.Q103-eGFP (pathogenic) expressing larva that were fed fly food
   laced with either 0.1% DMSO or 10 µM CHIR99021 for 24 h prior to
   dissections. n = 6 larvae. Bar = 10 µm. Note while DMSO-treated larvae
   expressing HTT.Q25-eGFP show smooth staining in larval nerves,
   HTT.Q103-eGFP expressing larva show axonal blockages (white arrows)
   that stain with CSP and contain HTT. H, I Quantification of the average
   number (#) of axonal HTT H accumulations or I CSP blockages within
   larval segmental nerves in CHIR99021-feed compared to DMSO-fed
   HTT.Q25-eGFP or HTT.Q103-eGFP expressing larvae. J Representative
   larval brains evaluated for cell death using TUNEL-labelling of
   HTT.Q25-eGFP or HTT.Q103-eGFP expressing larva that were fed fly food
   laced with either 0.1% DMSO or 10 µM CHIR99021 for 24 h. Red box in
   schematic = imaged are. n = 5 larvae. Bar = 10 µm. Larval brains
   expressing HTT.Q103-eGFP show TUNEL-positive cells, in contrast to
   HTT.Q25-eGFP expressing larvae. However, HTT.Q25 expressing larvae show
   robust TUNEL-positive cells in the context of 1 µg/µL DNAase I
   (positive-control). K Quantification of the average number (#) of
   TUNEL-positive cells normalized to imaged brain area comparing
   CHIR99021-feeding to DMSO-treatment of HTT.Q25-eGFP or HTT.Q103-eGFP
   expressing larvae. L Representative larval brains were evaluated for
   aggregations of HTT in HTT.Q25-eGFP or HTT.Q103-eGFP expressing larva
   that were fed fly food laced with either 0.1% DMSO or 10 µM CHIR99021
   for 24 h. n = 5 larvae. Bar = 10 µm. M Quantification of the average
   number (#) of HTT accumulates/aggregates per larval brain comparing
   CHIR99021-feeding to DMSO-treatment of HTT.Q25-eGFP or HTT.Q103-eGFP
   expressing larvae. Statistical significance was determined using the
   two-sample two-sided Student’s t test. Data represented as mean ± SEM.
   ns = p > 0.05, *p < 0.01, **p < 0.001, ***p < 0.0001, ****p < 0.00001.

   To examine how active GSK3β contributes to pathogenic HTT-mediated
   axonal transport defects and synaptic growth defects, we evaluated
   type-1 synaptic boutons in NMJs between muscle 6/7 at larval abdominal
   segments A4-A5 as we had previously done [[221]33, [222]34, [223]79,
   [224]88]. We found that CHIR99021 feeding alleviated the synaptic
   growth defects observed in larval NMJs expressing pathogenic HTT (Q103;
   Fig. [225]4D, E). Notably, concomitant increases in the levels of HTT
   at synapses were seen in these larvae (Fig. [226]4D, F) compared to
   DMSO-fed larvae, suggesting that the movement of HTT to NMJs is likely
   regulated by active GSK3β. Since larval locomotor defects correlate
   with the number of axonal blockages [[227]88] and axonal transport
   defects caused by loss of motor proteins or excess linker/receptor
   proteins result in synaptic defects [[228]79], we next tested the
   proposal that GSK3β plays a role in pathogenic HTT-mediated axonal
   transport defects. As expected, DMSO-fed larvae expressing pathogenic
   HTT (Q103) showed axonal blockages containing both CSP (cysteine string
   protein) and HTT, but CHIR99021-fed larvae showed significantly
   decreased numbers of CSP and HTT-containing axonal blockages (Fig.
   [229]4G–I). Unsurprisingly CHIR99021-fed larvae expressing
   normal/non-pathogenic HTT (Q25) also caused axonal blockages that
   contained both CSP and HTT. Although the extent of these blockages was
   not as comparable to those seen with pathogenic HTT expression, the
   amounts of blockages were significant compared to DMSO-fed controls.
   These observations are similar to what was observed for GSK3β loss of
   function mutants [[230]86] and confirm GSK3β’s role in the axonal
   transport of synaptic vesicles [[231]42, [232]87]. Therefore, while
   inhibition of GSK3β can cause axonal transport defects by decreasing
   the number of motors associated with synaptic membranes [[233]42],
   these findings implicate active GSK3β in escalating pathogenic
   HTT-mediated axonal transport defects, either by directly
   phosphorylating HTT (Fig. [234]4D, E) and/or by regulating the
   recruitment of motor proteins onto CSP- and/or HTT-containing membranes
   [[235]77, [236]85]. Indeed, this proposal is validated via larvae
   expressing excess sgg with pathogenic HTT, which still showed brain
   accumulations, axonal blockages, synaptic morphology defects, and
   decreased levels of HTT at NMJs (Fig. [237]S6)

   Previously, we showed that the extent of axonal blockages correlated
   with the extent of HTT aggregates and cell death [[238]30, [239]80].
   However, the mechanism by which pathogenic HTT initiates cell death
   remains obscure. To test how active GSK3β affects pathogenic
   HTT-mediated neuronal cell death, we examined larval brains expressing
   pathogenic HTT (Q103) and normal HTT (Q25) in the context of the GSK3β
   inhibitor CHIR99021. While expression of pathogenic HTT (Q103) caused
   significant amounts of cell death in larval brains, CHIR99021-fed
   larval brains showed significant reductions in cell death (Fig.
   [240]4J, K). Significantly decreased levels of HTT accumulations within
   larval brains (Fig. [241]4L, M) were also seen, perhaps due to
   stimulating lysosomal degradation of HTT aggregates as reported
   previously in mice brains [[242]58]. In contrast, larvae expressing
   excess sgg with pathogenic HTT did not change the levels of HTT
   accumulations within larval brains (Fig. [243]S6A). Therefore, excess
   active GSK3β likely causes neuronal cell death through multiple
   pathways, including those that are triggered by phosphorylation of HTT
   and/or a cascade of events activated by pathogenic HTT-mediated
   transport defects.

ERK-dependent events on HTT are likely neuroprotective against pathogenic
HTT-mediated neuronal cell death

   A MS-based finding proposed three putative ERK phosphorylation sites on
   HTT (Table[244]1, [[245]35]). However, the functional significance of
   ERK-dependent events on HTT remains ambiguous. Consistent with our
   observations in HD iNeurons, cholinergic and somatostatinergic
   interneurons of the striatum in R6/2 mice showed higher levels of ERK
   compared to controls [[246]89]. Phosphorylated ERK levels also
   increased with age in these mice [[247]89]. However, work in a striatal
   cell line derived from a transgenic homozygous knock-in mouse with
   111CAG repeats (STHdhQ111/Q111) showed no change [[248]90]. In contrast
   to GSK3β [[249]58], ERK-mediated signaling was suggested to be
   neuroprotective in the context of pathogenic HTT expression [[250]55,
   [251]56, [252]91, [253]92], but the mechanisms are unknown. Unlike
   GSK3β [[254]77, [255]85, [256]86, [257]93], little is currently known
   about the role of ERK within the axonal transport pathway despite ERK
   being shown to phosphorylate dynein intermediate chain (DIC) at Ser80
   [[258]94]. Therefore, to first test whether ERK functions during axonal
   transport, we assessed larvae carrying a hypomorphic allele for rolled
   (rl), the Drosophila homolog of ERK, and larvae expressing excess
   wild-type rl or active rl [[259]95]. Previous work showed that rl
   mutant larvae exhibit aberrant nerve innervation patterns in abdominal
   body-wall muscles and premature defasciculation of axonal bundles
   [[260]96]. Furthermore, rl mutant flies have a rough eye phenotype due
   to missing photoreceptor cells [[261]97]. Surprisingly, neither
   condition (hypomorphic nor excess) showed axonal transport defects
   (Fig. [262]5A, B), which is in contrast to larvae expressing excess
   wildtype, excess active, or excess dominant negative sgg, which showed
   significant axonal blockages (Fig. [263]5A, B; [[264]85]). Therefore,
   unlike GSK3β, ERK likely has no major role during axonal transport.

Fig. 5. ERK does not play a major role in axonal transport regulation, unlike
GSK3β.

   [265]Fig. 5
   [266]Open in a new tab

   A Schematic of Drosophila larval nervous system showing brain and
   segmental nerves. Red box = imaged area. Representative larval
   segmental nerves from WT, sgg^OE, sgg^DN, sgg^active, rl^OE,rl^LoF, or
   rl^active larva stained with the synaptic vesicle marker CSP. n = 6
   larvae. Bar = 10 µm. OE=overexpression, LOF= loss of function, DN=
   dominant negative, Active=constitutively active. B Quantification of
   the average number (#) of axonal CSP blockages per larvae within
   segmental nerves. Statistical significance was determined using the
   two-sample two-sided Student’s t test. Data represented as mean ± sem.
   ns = p > 0.01, *p < 0.01, **p < 0.001, ***p < 0.0001, ****p < 0.00001.

   To test the prediction that ERK activity can specifically influence the
   movement of HTT within axons, we examined larvae expressing
   non-pathogenic (Q25) and pathogenic (Q103) HTT fed on food laced with
   10 µM SCH772984, which inhibits ERK1/2 [[267]78] and on 0.1% DMSO-laced
   food. We found that feeding 10 µM SCH772984 was not toxic to WT flies
   in contrast to feedings at 25 µM. In contrast to GSK3β inhibition,
   inhibition of ERK enhanced pathogenic HTT-mediated larval locomotor
   deficits (Fig. [268]6A–C), and increased axonal blockages containing
   HTT and CSP (Fig. [269]4G–I). However, no defects were observed at NMJs
   or synapses (Fig. [270]4D–F). Inhibition of ERK significantly increased
   neuronal cell death (Fig. [271]4J, K), concomitant with significant
   increases in HTT accumulations within pathogenic HTT-expressing larval
   brains (Fig. [272]4L, M). Strikingly, and in contrast to GSK3β
   inhibition, ERK inhibition did not induce CSP-containing axonal
   blockages but caused a significant number of HTT-containing blockages
   in larvae expressing non-pathogenic (Q25) HTT. Taken together, these
   observations suggest that while overall ERK likely exerts a
   neuroprotective role, ERK’s effect on axonal transport is restricted to
   HTT, in contrast to GSK3β, which has a broader role during axonal
   transport.

Fig. 6. ERK inhibition causes accumulation of non-pathogenic HTT in larval
nerves and exacerbates pathogenic HTT-induced neuronal cell death.

   [273]Fig. 6
   [274]Open in a new tab

   A HTT.Q103-eGFP expressing larva that were fed fly food laced with
   either 0.1% DMSO or 10 µM SCH772984 for 24 h were subjected to larval
   crawling assays for 1.5 mins at 25 °C, 60% humidity and compared to
   DMSO-treated HTT.Q25-eGFP larvae. n = 20. Bar = 1 cm. B Quantification
   of the avg. larval crawling velocity (cm/min) comparing
   SCH772984-feeding to DMSO-treatment of HTT.Q103-eGFP expressing larvae
   and DMSO-treated HTT.Q25-eGFP expressing larvae. C Quantification of
   avg. larval size (area; cm^2) normalized to HTT.Q25 expressing larvae
   (AU) comparing SCH772984-feed to DMSO-fed HTT.Q103-eGFP expressing
   larvae and DMSO-fed HTT.Q25-eGFP expressing larvae. D Schematic of
   Drosophila larval nervous system showing brain, segmental nerves, and
   neuromuscular junctions (NMJs). Red box = imaged area. Representative
   images of NMJs from muscle 6/7 segment A4–5 of HTT.Q25-eGFP or
   HTT.Q103-eGFP expressing larva that were fed fly food laced with either
   0.1% DMSO or 10 µM SCH772984 for 24 h prior to dissection and stained
   with HRP-TxRED. Top panels = HRP-TxRED staining. Bottom
   Panel = HTT-eGFP fluorescence. n = 6 larvae. Bar = 15μm. Note an even
   distribution of colocalization (yellow arrows) between HTT.Q25 (green)
   and HRP (red) at boutons is observed, which is lost in HTT.Q103
   expressing larva. E Quantification of the avg. synapse length (μm) per
   larvae comparing SCH772984-fed to DMSO-fed HTT.Q25-eGFP or
   HTT.Q103-eGFP expressing larvae. F Quantification of the avg. HTT-eGFP
   (green) signal intensity at boutons normalized to HRP (red) signal
   intensity (AU) per larvae comparing SCH772984-fed to DMSO-fed
   HTT.Q25-eGFP or HTT.Q103-eGFP expressing larvae. G Schematic of
   Drosophila larval nervous system showing brain and segmental nerves.
   Red box = imaged area. Representative larval segmental nerves stained
   with the synaptic vesicle marker CSP from HTT.Q25-eGFP (non-pathogenic)
   or HTT.Q103-eGFP (pathogenic) expressing larvae that were fed fly food
   laced with either 0.1% DMSO or 10 µM SCH772984 for 24 h prior to
   dissections. n = 6 larvae. Bar = 10 µm. Note while DMSO-fed larvae
   expressing HTT.Q25-eGFP show smooth staining in larval nerves,
   HTT.Q103-eGFP expressing larvae show axonal blockages (white arrows)
   that stain with CSP and contain HTT. H, I Quantification of the average
   number (#) of axonal H HTT accumulations or I CSP blockages within
   larval segmental nerves comparing SCH772984-fed to DMSO-fed
   HTT.Q25-eGFP or HTT.Q103-eGFP expressing larvae. J Red box = imaged
   are. Representative larval brains were evaluated for cell death using
   TUNEL-labelling of HTT.Q25-eGFP or HTT.Q103-eGFP expressing larvae that
   were fed fly food laced with either 0.1% DMSO or 10 µM SCH772984 for
   24 h. n = 5 larvae. Bar = 10 µm. Larval brains expressing HTT.Q103-eGFP
   show TUNEL-positive cells, in contrast to HTT.Q25-eGFP expressing
   larvae. However, HTT.Q25 expressing larvae show robust TUNEL-positive
   cells in the context of 1 µg/µL DNAase I (positive-control). K
   Quantification of the average number (#) of TUNEL-positive cells
   normalized to the imaged brain area comparing SCH772984-fed to DMSO-fed
   HTT.Q25-eGFP or HTT.Q103-eGFP expressing larvae. L Representative
   larval brains evaluated for aggregations of HTT in HTT.Q25-eGFP or
   HTT.Q103-eGFP expressing larva that were fed fly food laced with either
   0.1% DMSO or 10 µM SCH772984 for 24 h. n = 5 larvae. Bar = 10 µm. M
   Quantification of the average number (#) of HTT accumulates/aggregates
   per larval brain comparing SCH772984-fed to DMSO-fed HTT.Q25-eGFP or
   HTT.Q103-eGFP expressing larvae. Statistical significance was
   determined using the two-sample two-sided Student’s t test. Data
   represented as mean ± sem. ns = p > 0.01, *p < 0.01, **p < 0.001,
   ***p < 0.0001, ****p < 0.00001.

   To validate our prediction that ERK has a protective role in the
   context of pathogenic HTT-mediated neuronal dysfunction, we added
   wild-type ERK with pathogenic HTT. Co-expression of rl with HTT-Q103
   rescued pathogenic HTT axonal blockages within larval nerves (Fig.
   [275]7A, B) and HTT accumulations within larval brains (Fig. [276]7C,
   D). However, feeding these larvae food laced with the ERK inhibitor
   SCH772984 reverted these phenotypes (Fig. [277]7A–D), indicating that
   changing the level of ERK can directly contribute to HTT-mediated
   pathogenic phenotypes (Fig. [278]8A). Taken together, our observations
   support a role for ERK during HD neuronal dysfunction, and propose a
   mechanism in which ERK activity is likely neuroprotective against
   pathogenic HTT-mediated neuronal cell death via the depletion of HTT
   accumulations and rescue of cell death.

Fig. 7. Excess ERK rescues accumulations of pathogenic HTT in larval nerves
and brains.

   Fig. 7
   [279]Open in a new tab

   A Schematic of Drosophila larval nervous system showing brain and
   segmental nerves. Red box = imaged area. Representative larval
   segmental nerves from HTT.Q103-eGFP or HTT.Q103-eGFP; UAS-rl^OE
   expressing larva that were fed fly food laced with either 0.1% DMSO or
   10 µM SCH772984 for 24 h prior to dissections. n = 4 larvae.
   Bar = 10 µm. Note, HTT.Q103-eGFP expressing larva show axonal
   accumulations (arrows). B Quantification of the average number (#) of
   axonal HTT accumulations within larval segmental nerves comparing
   SCH772984-feeding to DMSO-treatment of HTT.Q103-eGFP or HTT.Q103-eGFP;
   UAS-rl^OE expressing larvae. C Schematic of Drosophila larval nervous
   system showing brain and segmental nerves. Red box = imaged area.
   Representative larval brain from HTT.Q103-eGFP or HTT.Q103-eGFP;
   UAS-rl^OE expressing larva that were fed fly food laced with either
   0.1% DMSO or 10 µM SCH772984 for 24 h prior to dissections. n = 4
   larvae. Bar = 10 µm. D Quantification of the average number (#) of HTT
   accumulations within larval brains comparing SCH772984-fed to DMSO-fed
   HTT.Q103-eGFP or HTT.Q103-eGFP; UAS-rl^OE expressing larvae.
   Statistical significance was determined using the two-sample two-sided
   Student’s t test. Data represented as mean ± sem. ns = p > 0.01,
   *p < 0.01, **p < 0.001, ***p < 0.0001. ****p < 0.00001.

Fig. 8. Schematic model for how ERK and GSK3β levels modulates HD
pathogenesis.

   [280]Fig. 8
   [281]Open in a new tab

   A Changing levels of ERK (left) or GSK3β (right) directly affects
   pathogenic HTT-mediated neuropathology. Expression of pathogenic HTT
   causes, HTT accumulations and cell death at cell bodies, HTT blockages
   that disrupt axonal transport and synaptic defects. Expression of
   pathogenic HTT with either decreased ERK or excess GSK3β enhances these
   phenotypes. However, expression of pathogenic HTT with either decreased
   GSK3β or adding excess ERK alleviates these phenotypes and restores
   axonal transport, synaptic morphology, and/or neuronal viability. B In
   HD, pathogenic HTT causes axonal transport defects and synaptic
   dysfunction that gives rise to locomotor deficits and neuronal cell
   death. While elevated GSK3β activity may exacerbate pathogenic
   HTT-mediated axonal transport defects via alternations to motor protein
   function and synaptic dysfunction via decreases in axonal transport of
   essential cargo, elevated ERK activity is likely neuroprotective and
   curtails neuronal cell death caused by aberrant structural and
   conformational changes to HTT.

Discussion

   Previous studies have unmasked HTT’s role in many cellular processes
   such as intracellular trafficking, however how HTT affects
   membrane-associated functions likely vital for both normal neuronal
   physiology and HD pathogenesis have been overlooked. Our study bridges
   this gap in knowledge, revealing novel membrane-specific proteomic
   alterations induced by pathogenic HTT. Specifically, we found
   significant remodeling of the HTT-associated membrane proteome in HD
   patient-derived iNeurons (Fig. [282]1), with notable shifts in key
   kinases such as mTOR, AKT1, ERK1/2, CDK1/5, and GSK3α/β (Fig. [283]2).
   We further discovered that GSK3β and ERK kinases are capable of
   phosphorylating HTT under both normal and disease conditions (Fig.
   [284]3), a phenomenon previously unknown. Further, GSK3β and ERK likely
   have opposing roles during HD pathogenesis, with GSK3β inhibition
   rescuing pathogenic HTT-mediated axonal transport defects, synaptic
   dysfunction, and neuronal cell death (Fig. [285]4), and ERK inhibition
   enhancing these phenotypes (Fig. [286]6) and excess ERK rescuing
   HTT-containing accumulations caused by pathogenic HTT (Fig. [287]7).
   Moreover, unlike GSK3β, ERK had little-to-no effect on other synaptic
   vesicle transport (Fig. [288]5). Together our findings unveil a
   previously unknown role for GSK3β and ERK1 (Fig. [289]8A) as
   phospho-regulators of HTT at membranes (Fig. [290]8B), shedding new
   light on the pathogenic mechanisms underlying HD and offering a
   promising highly druggable pathway for therapeutic intervention that
   can be inhibited by already available small molecules.

GSK3β and ERK are likely phospho-regulators of HTT

   HTT exhibits a remarkable capacity for intramolecular interactions,
   giving rise to an array of three-dimensional conformations [[291]98].
   This versatility is evidenced by the observation of over one hundred
   structurally distinct conformations [[292]99], which likely aid in
   HTT’s ability to associate with hundreds of different proteins [[293]9,
   [294]10, [295]12]. Post-translational modifications, such as
   phosphorylation, could represent potential mechanisms to modulate these
   intramolecular interactions within HTT, thereby influencing its
   associations with proteins and membranes, leading to diverse cellular
   localizations. Indeed, normal HTT is phosphorylated, but pathogenic HTT
   exhibits hyperphosphorylation ([[296]26, [297]50, [298]75], Fig.
   [299]3A, B). While phosphorylation at Ser13/Ser16 by IKK increased HTT
   turnover [[300]100], phosphorylation at either Ser13 or Ser16 also
   promoted nuclear localization [[301]101, [302]102] and inhibited
   membrane binding [[303]103]. However, disruption of the Ser13 and Ser16
   sites in BACHD mice did not change HTT’s subcellular location,
   highlighting those other regions of HTT can play a role in its
   subcellular localization. Further, membrane interactions and
   intramolecular interactions within HTT have been proposed to be
   regulated by AKT (S421; [[304]37]) or CDK5 (S434, S1181, S1201;
   [[305]59]). While the mechanisms governing how these phosphorylation
   events impact HTT’s interaction with membranes are ambiguous, the
   functional significance of other potential HTT phosphorylation sites
   identified through MS remains unexplored [[306]35, [307]36].

   Here, we identified that GSK3β and ERK1 are both associated with and
   are capable of phosphorylating normal and pathogenic HTT (Fig. [308]3).
   Interestingly, while ERK has three putative phosphorylation sites
   (Ser2076, Ser2653, and Ser2657), GSK3β has only a sole putative
   phosphorylation site (Ser2657) at the C-terminus of HTT that overlaps
   with one of the three putative ERK1 sites [[309]35]. This suggests that
   there is likely crosstalk between these two kinases, and perhaps a
   regulatory mechanism exists. While a handful of studies have
   investigated GSK3β [[310]54, [311]104–[312]106] and ERK [[313]107] in
   the context of HD, little is known about functional interactions
   between these kinases and HTT. Our discovery that GSK3β and ERK1
   phosphorylates HTT at membranes proposes a novel mechanism for HD
   pathogenesis whereby these phosphorylation events become dysfunctional
   in HD, contributing to neuronal degeneration and cell death. One
   possible mechanism involves dysregulation of the regulatory interplay
   between pathways involving GSK3β and ERK1. Indeed, expression of excess
   or constitutively active GSK3β induces apoptosis in rat cortical
   neurons, which can be inhibited by expression of constitutively active
   ERK [[314]65]. However, unlike AKT1, which directly inhibits GSK3β via
   phosphorylation at Ser9 [[315]64], suppression of GSK3β-mediated
   apoptosis in rat cortical neurons by ERK was independent of GSK3β
   phosphorylation at Ser9 (inactivation) or at Tyr216 (activation),
   suggesting a novel regulatory mechanism between these two kinases. We
   propose that the overlapping phosphorylation site at Ser2657 can play a
   role in regulating GSK3β and ERK-mediated functions. Indeed, inhibition
   of GSK3β (Fig. [316]4) or excess ERK (Fig. [317]7) rescued HD
   phenotypes and decreased axonal transport defects and HTT aggregation,
   in contrast to inhibition of ERK (Fig. [318]6) or excess GSK3β (Fig.
   [319]S6). Previous work also supports this proposal since neuronal
   dysfunction in HD was ameliorated by either decreased GSK3β-mediated
   signaling [[320]58] or increased ERK-mediated signaling [[321]55].

ERK specifically regulates HTT functions, while GSK3β coordinates universal
axonal transport

   Axonal transport is a tightly regulated cellular process essential for
   neuron survival. Both HTT and GSK3β have independently emerged as
   coordinators of axonal transport in the context of motor proteins such
   as kinesin-1. While HTT acts as a scaffold protein and can interact
   with kinesin light chain (KLC) via huntingtin-associated protein 1
   (HAP1; [[322]108]), AKT1 phosphorylation of HTT at serine 421 is
   thought to enhance the recruitment of kinesin-1 to vesicles and
   microtubules, thereby promoting anterograde transport [[323]38]. In
   contrast, GSK3β phosphorylation can promote motor-membrane associations
   [[324]85], and GSK3β phosphorylation of kinesin heavy chain (KHC) at
   Ser314 stops motor activity [[325]77]. Therefore, GSK3β plays a key
   role in regulating the universal transport of all synaptic vesicles,
   including HTT-containing vesicles within axons. Our results also reveal
   that while HTT-containing membranes contain active GSK3β and motor
   proteins, consistent with our previous observations [[326]85, [327]87],
   increased levels of both active GSK3β and motor proteins were present
   on membranes enriched from HD patient-iNeurons (Figs. [328]1 and
   [329]2), likely contributing to the disruption of cargo motility.
   Furthermore, inhibition of GSK3β caused both CSP- and HTT-containing
   axonal blockages in larval nerves expressing non-pathogenic HTT (Fig.
   [330]4G), similar to the disruption of the motility of dense core
   vesicles [[331]42]. Moreover, loss or excess GSK3β also caused
   CSP-containing axonal blockages [[332]86, [333]87], indicating GSK3β is
   a universal regulator of axonal motility. Therefore, changes in GSK3β
   activity seen in HD could likely be responsible for defective axonal
   transport, which is proposed to be an early event in HD pathogenesis
   [[334]72].

   In contrast to GSK3β, surprisingly, we found that ERK has no direct
   role during axonal transport. Partial loss or excess of rl, the
   Drosophila homolog of ERK, did not cause axonal blockages (Fig.
   [335]5). Excess constitutively active ERK also failed to cause axonal
   transport defects (Fig. [336]5). Furthermore, inhibition of ERK in the
   context of larvae expressing non-pathogenic HTT only caused
   HTT-containing axonal blockages, but not CSP-containing blockages (Fig.
   [337]6). Taken together, ERK is likely only involved in regulating the
   transport of HTT and does not play a broader regulatory role in the
   motility of other axonal cargo. While ERK activity has been shown to be
   controlled by several different mechanisms, including scaffolding and
   anchoring proteins, dictating when and where ERK is activated and
   limiting ERK localization to specific subcellular compartments
   [[338]109], one prediction is that ERK phosphorylation of HTT could
   influence HTT’s ability to associate with membranes during initial
   assembly onto axonal cargoes, a process thought to occur at cell bodies
   and/or at terminals prior to long-distance transport. Here, we
   anticipate that ERK and HTT would only predominantly associate within
   cell bodies/terminals. Indeed, ERK localizes to synapses
   [[339]110–[340]113] and has been reported to modulate synaptic events
   such as functional synapse formation [[341]14], vesicle exocytosis
   [[342]114], synaptic vesicle mobilization and release [[343]115],
   glutamate neurotransmitter release [[344]116, [345]117] and synaptic
   vesicle trafficking [[346]118]. Intriguingly, in vitro work also
   demonstrated that ERK phosphorylates dynamin to inhibit
   dynamin-microtubule interactions [[347]118]. Perhaps there is a
   putative endocytic pathway at synapses involving crosstalk between
   ERK-mediated events and HTT, which are known to affect synaptic vesicle
   recycling [[348]29], likely through associations with HIP1 [[349]41],
   AP-2 [[350]119], and HIP1R [[351]120] to facilitate clathrin-dependent
   endocytosis. Notably, clathrin-dependent endocytosis can drive the
   internalization of neurotrophic receptors such as TrkB [[352]121] and
   EGFR [[353]122], which both lead to the activation of ERK1/2-mediated
   signaling [[354]122–[355]125]. It is also possible that ERK
   phosphorylation of HTT affects its association with receptors, synaptic
   proteins, and/or endocytosis-related proteins. Future work will be
   needed to test these predictions.

ERK dampens pathogenic HTT-mediated apoptotic pathways

   ERK signaling was shown to be deficient in striatal neurons of multiple
   transgenic HD mice [[356]126, [357]127]. Interestingly, either
   expression of constitutively active ERK [[358]126] or pharmacological
   stimulation of ERK activity via RB5 [[359]127] or Fisetin [[360]88]
   resulted in decreased cell death in HD mice. Similarly, we also found
   that changing the dose of ERK modulated HTT accumulations and cell
   death caused by pathogenic HTT. While pharmacological inhibition of ERK
   increased HTT accumulations and cell death in larval brains expressing
   pathogenic HTT (Fig. [361]6), adding ERK back with excess pathogenic
   HTT rescued these phenotypes (Fig. [362]7). Therefore, ERK signaling
   may play a compensatory role in preventing neuronal cell death elicited
   by neuronal dysfunction in HD [[363]89]. We propose that ERK’s
   protective effect is primarily a response to the cellular defects
   caused by pathogenic HTT as opposed to being directly involved in
   pathways contributing to HD pathogenesis. However, since we discovered
   that ERK1 is capable of associating with and phosphorylating normal and
   pathogenic HTT, it is possible that ERK also plays a more direct role
   in HD pathogenesis than previously thought.

   Neuronal survival hinges on the uptake of anti-apoptotic extrinsic
   factors known as neurotrophins, such as BDNF and NGF. Upon activation
   of their respective receptor tyrosine kinases, these factors undergo
   internalization into signaling endosomes, which are then transported to
   the cell body via dynein-mediated movement. Neurotrophin signaling
   fosters neuronal survival by initiating the ERK1/2 pathway [[364]128].
   Notably, active ERK1/2 can also facilitate the assembly of signaling
   endosomes by phosphorylating cytoplasmic dynein and promoting its
   association with RAB7-positive signaling endosomes [[365]94]. Indeed,
   while disrupted neurotrophic signaling is well-documented in HD models
   [[366]31, [367]129], we previously showed that HTT is present on
   retrogradely moving RAB7-containing signaling endosomes that likely
   contain BDNF and BMP [[368]34]. Perhaps ERK-mediated phosphorylation of
   HTT influences the assembly of signaling endosomes. Indeed, ERK1/2
   inhibition caused HTT cargo accumulations within axons but did not
   affect other synaptic cargo labeled by CSP (Fig. [369]6). Therefore,
   ERK1/2 may influence pathways involving only a subset of axonal cargoes
   that affect neuronal survival, such as HTT-Rab7-containing signaling
   endosomes. Alternatively, since ERK phosphorylates the intermediate
   chain of dynein [[370]94], which can bind HTT [[371]48, [372]49],
   perhaps ERK phosphorylation may also promote HTT-dynein associations,
   similar to how AKT1 phosphorylation promotes HTT associations with
   kinesin-1 [[373]38]. Accordingly, we predict that ERK may play a role
   in the assembly of signaling endosomes by phosphorylating HTT and
   dynein. While further study is needed to test the predictions of this
   proposal, our findings underscore the potential of targeting the ERK1/2
   pathway for therapeutic intervention in mitigating polyQ-mediated
   neuronal cell death.

   Alternatively, since the caspase cleaved 587-3144 C-terminal fragment
   containing the three putative ERK phosphorylation sites was found to be
   toxic [[374]130], with the C-terminus interacting with the N-terminal
   fragment begs the question whether ERK-mediated events on HTT are
   independent of polyQ-expansion. A shorter C-terminal HTT fragment
   (aa2568-3144) containing two ERK putative phosphorylation sites was
   also shown to be toxic [[375]131]. Therefore, C-terminus-dependent
   mechanisms that are mediated by ERK phosphorylation could contribute to
   HD pathogenesis. While the HTT C-terminal mechanisms are still
   ambiguous, several unanswered questions remain. Do ERK-mediated
   phosphorylation/dephosphorylation events cause conformational changes
   to full-length HTT or to the caspase-cleaved HTT C-terminal fragments
   for ON/OFF states? Does ERK phosphorylation at the C-terminus
   structurally alter HTT’s binding properties with membranes or dictate
   the recruitment of proteins that structurally re-organize and aid HTT’s
   scaffolding role at membranes? Further studies would be needed to test
   these predictions and to isolate the role of caspase-cleaved HTT
   fragments that are likely to be phosphorylated by GSK3β and ERK.

Materials and methods

Neuronal culture, characterizations, and biochemical assays

WT and HD iPSC cultures and neuronal differentiation

   The following cell lines were obtained from the NIGMS Human Genetic
   Cell Repository at the Coriell Institute for Medical Research: iPSCs
   from WT (ND38555-polyQ = 17, 48 y, female) and HD (ND42222-polyQ = 109,
   9 y, female) patients were purchased from the NINDS Repository. iPSCs
   were grown and expanded on corning matrigel (Fisher) using E8 iPSC
   media (Invitrogen). Pluripotency was analyzed using an antibody against
   OCT-3/4 (Santa Cruz, 1:200), and Hoechst was used as a nuclear
   staining, as detailed below. After 4 passages iPSCs were differentiated
   into neuronal precursors (NPCs) using PSC neural induction media
   (Invitrogen) and published protocols (publication #MAN0008031). NPCs
   were identified using an antibody against Nestin (Santa Cruz, 1:200)
   and then differentiated into mature iNeurons using neurobasal media
   supplemented with 1× B27 and 2 mM glutamine (Invitrogen, ThermoFisher).
   Differentiated neurons, identified using antibodies against MAP2 (BD
   Biosciences, 1:100), βIII-tubulin (Biolegend, 1:100) and Synaptophysin
   (Phosphosolutions, 1:100) exhibited an extensive neurite network after
   21 days at which time they were used for electrophysiology experiments,
   biochemical experiments, immunofluorescence, and transfections.

Electrophysiology in iNeurons

   To study the functionality of the differentiated neurons, cells were
   selected for whole-cell patch-clamp electrophysiology as previously
   done [[376]33]. In all experiments, cells were bathed in artificial
   cerebrospinal fluid containing (in mM): 124 NaCl, 2.5 KCl, 2.5 CaCl[2],
   2 MgSO[4], 1.25 KH[2]PO[4], 26 NaHCO[3], 10 glucose, 4 sucrose, pH 7.2
   (NaOH), 324 mOsm. ACSF was prepared fresh daily.

Voltage-clamp experiments

   To reveal currents through voltage-gated Na+ channels, borosilicate
   glass pipettes (8—12 MΩ) were filled with intracellular solution
   containing (in mM): 117 CsMeSO4, 5 tetraethylammonium (TEA), 2 MgATP,
   0.3 Na2GTP, 20 HEPES, 0.4 EGTA, 2.8 NaCl, pH 7.2 (CsOH), 263 mOsm. To
   reveal currents through voltage-gated K+ channels, pipettes contained
   (in mM): 120 K-gluconate, 20 KCl, 10 HEPES, 0.2 EGTA, 2 MgCl2, 2 MgATP,
   0.3 Na2GTP. To verify the identity of the currents, suspect Na+ channel
   currents were tested for inhibition due to 1 µM tetrodotoxin (TTX) bath
   applied while suspect K+ channel currents were tested for inhibition
   due to 50 mM TEA bath applied. Currents elicited in the presence of an
   inhibitor were normalized to the maximum current elicited in the
   absence of an inhibitor. Currents were low-pass filtered (2 kHz, 4-pole
   Bessel, Axon Instruments 200B) and sampled (10 kHz, Molecular Devices,
   Digidata 1440A).

Current-clamp experiments

   To elicit action potential waveforms, borosilicate glass pipettes
   (8—12 MΩ) contained the K + -based intracellular solution described
   above. Upon break-in, cells were held at −70 mV. Action potentials were
   elicited upon increasing the current injection in 10 pA increments.
   Membrane potentials were low-pass filtered (2 kHz, 4-pole Bessel,
   Multiclamp 700B) and sampled (10 kHz, Molecular Devices Digidata
   1550B). All files were analyzed using Clampfit 10.5 (Molecular
   Devices).

Immunocytochemistry

   human iPSCs/NPCs/iNeurons: WT and HD iPSC/NPC colonies or 21-day-old,
   differentiated neurons were washed in PBS pH 7.2 and fixed in 4%
   paraformaldehyde. A blocking solution (1× PBST with 5% BSA) was added
   to cells for 60 min prior to antibody incubation. Cells were incubated
   in primary antibodies, OCT-4 (Santa Cruz 1:100), Nestin (Santa Cruz
   1:100), MAP2 (BD Biosciences 1:100) βIII-tubulin (Biolegend 1:100), or
   Synaptophysin (Phosphosolutions 1:100) for 16 h at 4 °C and appropriate
   secondary antibodies (AlexaFluor® 488 or 568, ThermoFisher) for 1 h at
   25 °C. DAPI was used to stain nuclei. Cells were then imaged at 20×–40×
   (for iPSC and NPC) or 60×–100× (iNeurons). iNeurons were imaged on
   glass slide-bottom dishes (In Vitro Scientific China, D29–14-1-N).
   Fixed images were taken at 20x-100x using DAPI, FITC, TxRed, and Cy5
   filters using a NikonTE-2000E inverted fluorescence microscope with a
   beam splitter containing narrow single-band GFP/DsRED filters and a
   Cool Snap HQ cooled CCD camera which were merged into a single RGB
   image to analyze colocalization noted as yellow or white puncta.

Preparation of protein extracts from human iNeurons

   WT and HD iNeurons were manually removed from 6-well or 12-well plates
   using ice-cold homogenization buffer (10 mM HEPES, pH 7.4, 100 mM K
   acetate, 150 mM sucrose, 5 mM EGTA, 3 mM Mg acetate, 1 mM DTT)
   containing a cocktail of protease inhibitors (Roche) and 5 mM EDTA,
   blended for 30 s on ice using a motorized pestle, and then quickly
   snap-frozen in liquid nitrogen. Neuronal extracts were centrifuged at
   1000 × g for 15 mins, and the supernatant was analyzed using western
   blot.

Sucrose gradient fractionations

   PNS samples human iNeuron extracts were further fractionated into
   soluble fractions (SF), heavy membrane pellet (P1), and LM fractions by
   sucrose gradient ultra-centrifugations using lysis buffer (4 mM HEPES,
   320 mM sucrose pH 7.4) containing a phosphatase and protease inhibitor
   cocktail (Pierce) as previously done [[377]33, [378]34, [379]85,
   [380]87]. Briefly, 300 µl of PNS was combined with 300 µl 62% sucrose
   and layered onto a sucrose gradient (35% and 8% sucrose), and
   centrifuged at ~100,000 × g for 90 mins. The LM fraction (35/8 layer),
   the SF, and the heavy membrane pellet (P1) were removed and used in
   western blot analysis. 100 µl of lysis buffer was used to dissolve the
   heavy membrane pellet (P1).

Co-immunoprecipitation analysis

   HTT-containing vesicles were isolated as described previously [[381]33,
   [382]34]. In brief, 10ug of mouse monoclonal antibody to HTT (EMD
   Millipore, 1:250) were crosslinked with magnetic beads (Pierce Protein
   A/G Magnetic Beads) with rotation for 30 min using Pierce™ Crosslink
   Magnetic IP/Co-IP Kit with 0.25 mM DSS. 1000 μg of total protein from
   the human iNeuron LMs were incubated with the HTT-antibody crosslinked
   Magnetic Beads for 16 h at 4 °C with rotation and eluted with Pierce
   elution buffer (0.1 M Glycine, pH 2.8). Eluents were then either
   separated by SDS-PAGE and analyzed via western blot analysis as
   detailed below and/or stored in liquid nitrogen for Trypsin-digestion
   and LC-MS analysis as detailed below. Isolation of HTT vesicles using
   crosslinked magnetic beads occurred in the absence of detergents to
   preserve vesicular membranes.

SDS-PAGE and western blot analysis

   Human iNeuron fractions in NuPage LDS sample buffer with 4 mM
   β-mercaptoethanol were run on 4–12% Bis-Tris gels (Invitrogen) and
   transferred to nitrocellulose membranes. Alternatively, human iNeurons
   HTT-IP samples were run on a Zn^2+-phos-tag™ AAL-107 gel as per
   manufacturer’s recommended protocol (FujiFilm-Wako) using 50 µM AAL-107
   solution. For the enhanced resolution of full-length HTT (350 kDa), a
   control gel with a ladder ran simultaneously beyond the 62 kDa standard
   “running-off” the gel. Similarly, phos-tag gels were transferred to
   nitrocellulose membranes. Blots were blocked using TBST with 5% BSA for
   60 mins at 25 °C and incubated with primary antibodies (SYT1
   (Thermofisher 1:1000), Rab4 (Abcam 1:1000), Rab5 (Abcam 1:1000), Rab2
   (SCBT 1:500), Rab7 (SCBT 1:500), VPS35 (SCBT 1:500), SUMO2/3
   (Cytoskeleton 1:500), KIF5A (Goldstein 1:250), KIF5B (Goldstein 1:250),
   KIF5C (Goldstein 1:250), DIC (Abcam 1:1000), DNCT (Abcam 1:1000), Actin
   (ThermoFisher 1:1000), Tubulin (Abcam 1:2000), HTT rabbit polyclonal
   (Abcam 1:1000), HTT mouse monoclonal (EMD Millipore 1:1000), Golgi
   (Millipore Sigma 1:1000), Cytochrome C (Santa Cruz 1:1000), TOM20
   (CellSignaling Technology 1:500), MAP1B (SCBT 1:1000), MAP2 (BD
   Pharmigen 1:1000), Total AKT1 (CellSignaling Technology 1:1000), pAKT1
   (Ser473, CellSignaling Technology 1:1000), Total GSK3α/β (CellSignaling
   Technology 1:1000), pGSK3α/β (pY279/pY216; Abcam 1:1000), or ERK
   (pan-ERK; BD Transduction Laboratories 1:1000) for 16 h at 4 °C. Blots
   were then incubated with anti-mouse or anti-rabbit secondary
   HRP-conjugated antibodies (ThermoFisher 1:1000) and imaged using a
   BioRad Chemi-doc system with Pierce ECL or diluted Femto substrate (1:5
   in TBS). Images from three blots were quantified using ImageLab.

In vitro phosphorylation kinase assay

   Recombinant GST-GSK3β or GST-ERK1 (SignalChem) were used for in vitro
   kinase assays. Potential substrates (HTT-IP) were incubated with 50 ng
   GSK3β or 50 ng ERK1 and 1 mCi/100 γ32P-ATP for 30 min at 37°C. The
   reaction was terminated using 4× sample buffer. Control reactions
   containing 30 µM of GSK3β inhibitor CHIR99021 (Selleck Chemical) and/or
   30 µM ERK1 inhibitor SCH772984 (Selleck Chemical) were done to evaluate
   the specificity of the GSK3β or ERK1 phosphorylation assay. Proteins
   were separated by SDS-PAGE, and the gel was dried and sealed in saran
   wrap and exposed to X-ray film overnight. After exposure, gels were
   stained with Coomassie brilliant blue to visualize proteins.

Statistical analysis of immunoblotting assays

   The statistical analysis used for each experiment is indicated in each
   figure legend. First, power and sample size (n) calculations were
   performed for each experimental paradigm: comparing two means from two
   samples, with two-sided equality to identify the sample size that
   corresponds to α = 0.05. For each experiment, a significance threshold
   of p < 0.05 (95% confidence) was used. Based on the power analysis for
   western blot assays, three separate membranes from three independent
   experiments were analyzed using ImageJ software. Individual data points
   for each analysis were averaged for each n and compared. Unless
   otherwise specified, the data compared was found to be normally
   distributed. Statistical significance of normal distributions was
   calculated by a two-sample two-tailed Student’s t test and/or ANOVA
   test in Excel, Minitab18 and/or by using the Proc GLM procedure
   followed by the pdiff mean comparison option in SAS Studio 3.81.
   Overlaid dot plots were constructed for all figures using
   OriginLab/OriginPro to represent mean ± SEM.

Proteomics assays and analyses

Protein preparation and digestion

   HTT-IP or total LM samples were dissolved in a detergent-containing
   buffer (50 mM Tris-formic acid (FA), 150 mM NaCl, 1% sodium
   deoxycholate, 2% sodium dodecyl sulfate (SDS), 2% IGEPAL® CA-630, pH
   8.0) plus the protease and phosphatase inhibitor tablets (Roche Applied
   Science, Indianapolis, IN), and were sonicated for 30 s
   (non-continuously, 5 s as a burst) using a high-energy sonicator
   (Qsonica, Newtown, CT). The solution was then centrifuged at
   20,000 × g, 4 °C for 30 min, and the supernatant portion was carefully
   transferred to Eppendorf tubes. Protein concentration was measured by
   bicinchoninic acid assay (BCA) kit (Pierce Biotechnology, Inc.,
   Rockford, IL). One hundred μg of extracted proteins from every sample
   were utilized for LC-MS analysis. Reduction and alkylation of proteins
   were achieved by 30 min incubation with 3 mM tris (2-caboxyl) phosphine
   (TCEP) and 30 min incubation with 20 mM iodoacetamide (IAM),
   respectively. Both steps were conducted under 37 °C in darkness with
   constant vortexing in an Eppendorf Thermomixer® (Eppendorf, Hauppauge,
   NY). The proteins were then subjected to a unique surfactant-aided
   precipitation/on-pellet digestion (SOD) procedure. Precipitation of
   proteins was performed by stepwise addition of 9 volumes of cold
   acetone with continuous vortexing and incubation at −20 °C overnight.
   After centrifugation at 20,000 × g, 4 °C for 30 min, the supernatant
   (containing undesirable constituents which may severely impair the
   overall quality of MS analysis in the protein mixture, e.g. detergents,
   components of a cellular matrix) was removed, and the pellets were
   rinsed with 800 μl of cold acetone/water mixture (85/15, v/v %) and
   air-dried. Two phases of enzyme addition were employed for the
   on-pellet-digestion. In phase 1 (pellet-dissolving phase), 50 µL of
   Tris buffer (50 mM, pH 8.5) containing trypsin at an enzyme/substrate
   ratio of 1:40 (w/w) was added to the pellets and incubated at 37 °C for
   6 h in an Eppendorf Thermomixer®; in phase 2 (complete-cleavage phase),
   another 50 µL of trypsin solution was added at an enzyme/substrate
   ratio of 1:40 (w/w). Then, the mixture was incubated at 37 °C overnight
   (12–16 h) to achieve complete digestion, and digestion was terminated
   by the addition of 1 μl formic acid. Supernatant for individual samples
   containing tryptic peptides derived from 6 μg of proteins was used for
   LC-MS analysis.

Long gradient nano-reverse-phase liquid chromatography/mass spectrometry
(LC-MS)

   An optimized gradient was utilized to resolve the complex peptide
   mixture, encompassing the following steps: 3 to 8% B over 15 min; 8 to
   24% B over 215 min; 24 to 38% B over 115 min; 38–63% B over 55 min; 63
   to 97% B in 5 min, and finally isocratic at 97% B for 15 min. Under
   such chromatographic settings, a peptide elution window of >345 min was
   achieved, with an average peak of <30 s and a peak capacity of >580.

   An LTQ/Orbitrap-ETD hybrid mass spectrometer (Thermo Fisher Scientific,
   San Jose, CA) was employed to analyze the identity of peptides in the
   mixture. An “overfilling” approach, which allowed the reinforcement of
   MS sensitivity while simultaneously guaranteeing accuracy and
   resolution, was applied for peptide detection. The spray tip was rinsed
   by dripping 50% methanol after every three runs to keep ionization
   efficiency stabilized. The instrument was operated in the
   data-dependent product ion mode. One scan cycle included one MS1 scan
   (m/z 310–2000) in the profile mode at a resolution of 60,000 followed
   by seven MS2 scans in collision-induced dissociation (CID) activation
   mode to fragment the seven most abundant precursor ions identified in
   the MS1 spectrum. The target value for MS1 by Orbitrap was 8 × 106,
   under which the Orbitrap was meticulously calibrated for mass accuracy
   and Fourier transform (FT) transmission. The use of a high target valve
   on the Orbitrap enabled ultra-sensitive detection with no compromise to
   the mass accuracy and resolution. The dynamic exclusion was enabled
   with the following settings: repeat count = 1; repeat duration = 30 s;
   exclusion list size = 500; and exclusion duration = 40 s. The
   activation time was 30 ms, the isolation width was 3 Da for the linear
   ion trap (LTQ), the normalized activation energy was 35%, and the
   activation q was 0.25. Three biological replicates from each biological
   group (db/db versus db/+) were analyzed in a random manner.

Protein identification and quantification

   Raw files were imported into PD, and DTA files were generated from MS2
   spectra. The search parameters used were as follows: 25-ppm tolerance
   for precursor ion mass and 1.0 Da for fragment ion mass. Two missed
   cleavages were permitted for tryptic peptides. Carbamidomethylation of
   cysteines and oxidation of methionine were set as fixed and variable
   modification, respectively. The false discovery rate (FDR) was detected
   by the usage of a target-decoy search strategy, in which the sequence
   database contains each sequence in both forward and reversed
   orientations and enables the estimation of FDR. Scaffold software
   (v4.3.2, Proteome Software, Portland, OR) was used to validate
   MS/MS-based peptide and protein identification based on
   cross-correlation (Xcorr) and Delta Cn values. The peptide filtering
   criteria included Delta Cn scores >0.1 and Xcorr scores >1.1, 1.4, 1.7,
   and 2.5 for singly, doubly, triply, and quadruply charged peptides.
   Stringent cutoffs for the Delta Cn and Xcorrs scores, plus the
   additional requirement that at least two distinct peptide sequences are
   needed for the identification of a protein, resulting in a considerably
   low FDR (0.19% at the peptide level). Shared peptides are retained on
   the identification level but are further evaluated on the
   quantification level for the congruity of including these peptides for
   quantification.

   SIEVE® software (v2.1.377, Thermo Scientific, San Jose, CA) was used to
   perform chromatographic alignment and global intensity-based MS1
   feature detection/extraction, consisting of: 1) Global chromatographic
   alignment of LC-MS runs via the application of ChromAlign algorithm.
   The alignment scores given by SIEVE, as well as the intensities of
   base-peak-ion current, were monitored and benchmarked for quality
   control; 2) Determination of quantitative “frames” based on
   mass-to-charge (m/z) and retention time in the aligned dataset. Only
   frames with high-quality area under the curve (AUC) with
   signal-to-noise ratio >10 were picked so that the quantitative
   reliability was assured; 3) Calculation of ion intensities among all
   “frames”. The output files were then merged with the spectrum report
   file exported from Scaffold to link the MS2 fragmentation scans with
   each “frame” using an in-house developed R package, IonStarStat
   (available at [383]https://github.com/shxm725/IonStarstat). The
   normalization of ion current intensities, the rejection of outlier
   peptides with aberrant intensities, and the aggregation of sum ion
   intensities from “frame” level to the protein level were also achieved
   by IonStarStat. The expression ratio for each protein was calculated
   based on the ion current peak areas of three replicates in control/WT
   and HD groups.

Exclusion criteria for raw LC-MS data from WT and HD LMs

   From 3 independent biological replicates, proteins were identified
   using UNIPROT accession IDs from the Homo sapiens proteome FASTA
   database. 3931 unique proteins from WT LMs and 3074 unique proteins
   from HD LMs were identified from LC-MS across 3 independent biological
   replicates. After excluding proteins that were present in less than 2
   independent trials (< 2) and excluding proteins with less than three
   unique peptides (spectral count (SpC) < 3) per trial, 3046 proteins
   remained in the WT LM data, and 2127 proteins remained in the HD LM
   data. Of these, 1998 proteins were present in both WT and HD LM
   samples, 1048 proteins were only in WT LMs, and 129 proteins were only
   in HD LMs.

Exclusion criteria for raw LC-MS data from WT and HD HTT-IPs

   From 3 independent biological replicates, proteins were identified
   using UNIPROT protein accession IDs from the Homo sapiens proteome
   FASTA database. 2332 unique proteins from WT HTT-IPs and 2308 unique
   proteins from HD HTT-IPs were identified from LC-MS across 3
   independent biological replicates. In addition to excluding proteins
   that were present in less than 2 independent trials (<2) and excluding
   proteins with less than three unique peptides (spectral count
   (SpC) < 3) per trial, the SpC of a protein in WT or HD HTT-IP was
   deducted from a negative control-IP in which WT or HD LMs were
   incubated with magnetic beads without antibody. Following these
   criteria, 1062 proteins remained in the WT HTT-IP data, and 1639
   proteins remained in the HD HTT-IP data. Of these, 800 proteins were
   present in both WT and HD HTT-IPs samples, 87 proteins were only in WT
   HTT-IPs, and 714 proteins were only in HD HTT-IPs.

Statistical analysis of HTT-IP and LM LC-MS data

   First power and sample size (n) calculations were performed on
   Minitab18 for each experimental paradigm: comparing 2 means from 2
   samples, with two-sided equality to identify the sample size that
   corresponds to α = 0.05. Statistical analysis used for each experiment
   is indicated in each figure legend. SpC’s of identified proteins from
   total LMs and HTT-IPs were analyzed using three biological replicates.
   SpC’s of identified proteins from HTT-IPs were first normalized to the
   HTT SpC within each trial to account for any difference in the amount
   of HTT IP’d between replicates and then analyzed using three biological
   replicates. Statistical significance was calculated by one-way
   ANOVA/post-hoc analysis to reduce Type I error, followed by Welch’s t
   tests to test to compare individual groups in Excel and Minitab18.
   Statistical analysis reported in figures report results from Welch’s t
   tests, as results from ANOVA/post-hoc and Welch’s t tests were
   consistent. Fold change (FC) between WT and HD samples was calculated
   by dividing the HD SpC of an individual protein by its respective WT
   SpC and normalizing to the total SpC from each sample. To reduce the
   false discovery of identified hits, an effect size cut-off for FC was
   set to log2FC > |1| (±2X) as in [[384]132] and as done previously in
   [[385]133]. This approach has also been applied in previous comparative
   proteomics studies focused on HTT [[386]12, [387]134].

Gene ontology (GO) analysis

   Functional enrichment analysis for biological processes, molecular
   functions, and cellular components of lists of proteins that were
   increased, decreased, gained, or lost with either HTT-IP or LMs between
   WT and HD iNeurons were analyzed using the g:Profiler resource
   ([388]http://biit.cs.ut.ee/gprofiler/gost) and cross-referenced with
   the Database for Annotation, Visualization and Integrated Discovery
   (DAVID) Bioinformatics Resource v6.7
   ([389]http://david.abcc.ncifcrf.gov). Cellular components assigned by
   g:Profiler were manually examined, ranked by p-value as well as
   coverage (%gene intersection), and then distributed into selected
   categories (Fig. 3.1, 3.2). Furthermore, reactome pathways were
   analyzed using the g:Profiler resource
   ([390]http://biit.cs.ut.ee/gprofiler/gost) and cross-referenced with
   GeneCodis4.0 [[391]135]. Enriched reactome pathways assigned by
   g:Profiler were manually examined, ranked by p value and coverage (%
   intersection).

Kinome enrichment analysis

   Details about kinase enrichment and signaling pathways were analyzed
   from lists of proteins that were increased, decreased, gained, or lost
   with either HTT-IP or LMs between WT and HD iNeurons using the Kinase
   Enrichment Analysis v.3 (KEA3; [[392]136]) from the Icahn School of
   Medicine at Mount Sinai ([393]https://maayanlab.cloud/kea3/) and/or the
   Kinome Mapping Hub Resource through Cell Signaling Technology, Inc
   ([394]http://www.kinhub.org/kinmap/).

Data visualization software

   lists of proteins that were increased, decreased, gained, or lost with
   either HTT-IP or LMs between WT and HD iNeurons were used to generate
   heatmaps through the Morpheus software by the Broad Institute
   ([395]https://software.broadinstitute.org/morpheus/) and volcano plots
   were generated through OriginLab/OriginPro and/or the VolcaNoseR data
   visualization software ([396]https://huygens.science.uva.nl/) through
   Rstudio.

Drosophila studies

Drosophila genetics

   Pan-neuronal driver APPL-GAL4 was used for neuronal expression of
   transgenic lines [[397]137–[398]139]. Unless stated, fly stock were
   obtained from BDSC. In brief, males from UAS-HTTex1.Q25-eGFP (gift from
   Norbert Perrimon; [[399]139]), UAS-HTTex1.Q103-eGFP (gift from Norbert
   Perrimon; [[400]139]), UAS-sgg.B (sgg^OE), UAS-sgg.S9A (sgg^active),
   UAS-sgg.A81T (sgg^DN), UAS-rl.K (rl^OE), UAS-rl^SEM (rl^active), or
   rl^1 (rl^LoF) were crossed to APPL-GAL4 virgins. UAS-HTTex1.Q103-eGFP
   or UAS-HTTex1.Q25-eGFP males were crossed to
   APPL-GAL4;T(2:3),CyO,TM6B,Tb/Pin88K virgin females. The chromosome
   carrying T(2:3),CyO,TM6B,Tb is referred to as B3 and carries the
   dominant markers, Hu, Tb and CyO. The larval Tb (Tubby) marker is used
   to select larvae of interest. APPL-GAL4/Y;UAS- HTTex1.Q103-eGFP or
   UAS-HTTex1.Q25-eGFP/Cyo,TM6B males were then crossed with UAS-rl.K
   (rl^OE) virgin females. In all cases, female 3rd instar larvae were
   selected. Crosses were maintained at 29 °C for protein overexpression
   and controlled for humidity (60%).

Larval feeding and incubations

   For chemical feeding experiments, 3rd instar larvae expressing
   UAS-HTTex1.Q25-eGFP, UAS-HTTex1.Q103-eGFP, or
   UAS-HTTex1.Q103-eGFP;UAS-r.K larvae were grown in fly flood containing
   buffer (0.1%DMSO), 10 µM CHIR99021 (Caymen Chemical) dissolved in
   buffer/0.1% DMSO [[401]77, [402]85], or 10 µM SCH772984 (Caymen
   Chemical) dissolved in buffer/0.1%DMSO [[403]78]. After 24 h, female
   larvae were selected for behavioral studies and/or dissections prior to
   fixation and imaging.

Larval preparations, immunohistochemistry, and quantifications

   Third instar Drosophila larvae were dissected and fixed in 8%
   paraformaldehyde. DCSP-3 (DSHB, 1:10) antibody was used in conjunction
   with secondary antibodies anti-mouse or anti-rabbit AlexaFlour®488,
   AlexFlour®568, or Alexaflour®647 (ThermoFisher, 1:100). Images of
   segmental nerves were collected using a Nikon Eclipse TE 2000U
   microscope using the 40× or 60× objective (Nikon, Melville, NY, USA)
   alongside the FITC (488 nm), TxRED (568 nm), both FITC/TxRED
   (568/488 nm using a dual-view beam splitter attachment) filters. TxRED
   and FITC images were merged into a single RGB image to analyze
   colocalization noted as yellow puncta. Axonal blockages were quantified
   as previously done [[404]30]. For each genotype, a minimum of 6
   confocal optical images across six larvae were imaged. Subpixel imaging
   refers to the subpixel detection accuracy of fluorescent puncta, which
   was previously confirmed by directly comparing Gaussian fitting of
   conventional microscopy data analyzed to high-resolution imaging
   [[405]109]. For NMJ analysis, HRP-FITC or HRP-TxRED (1:50, Jackson
   ImmunoResearch Labs). Non-tubby, female larvae were dissected, fixed,
   and stained with HRP. Quantification of NMJ morphology and HTT-eGFP
   intensity was performed as previously done [[406]33, [407]79]. We
   examined type-1 synaptic boutons between muscles 6/7 at larval
   abdominal segments A4-A5 of third instar larval. Images of NMJs were
   collected using a Nikon Eclipse TE 2000 U microscope at ×60 (Nikon,
   Melville, NY, USA). For each genotype, a minimum of 12 optical images
   across 4–6 larvae were imaged. Total NMJ length (μm) was measured using
   NIH ImageJ software as done previously [[408]33, [409]34].

TUNEL assay

   Third instar larvae were dissected and fixed as described above prior
   to being permeabilized with 5% saponin for 20 min at 25 °C. TUNEL assay
   was performed on permeabilized larval brains using the In Situ Cell
   Death Detection Kit (Roche) per manufacturer’s instructions. Incubation
   in DNAse I was used as a positive control. The number of puncta in each
   brain was quantified in ImageJ (NIH) using the Threshold tool and
   Analyze Particles tool. At least four adult brains were imaged and
   quantified from each genotype.

Quantification of Drosophila larval velocities

   Larval velocity was performed by visualizing third instar larval
   crawling patterns on 1% agarose gel, dyed blue for added contrast, that
   was embedded with a 0.25 cm × 0.25 cm grid. Once placed at the center
   of the agarose gel, a 2-min interval recording began. Controlling for
   temperature (~25 °C) and humidity (~60%), twenty larvae were tested per
   condition across two independent trials. Quantifications of the larval
   velocity were measured using NIH ImageJ software and Track-mate. A
   total of 1.5 min of the 2-min recording was utilized for quantification
   purposes allotting the initial 15 s for larva self-adjustment after
   being placed in the center of the agarose gel.

Statistical analysis of Drosophila assays

   The statistical analysis used for each experiment is indicated in each
   figure legend. First power and sample size (n) calculations were
   performed for each experimental paradigm: comparing 2 means from two
   samples, with two-sided equality to identify the sample size that
   corresponds to α = 0.05. For each experiment, a significance threshold
   of p < 0.05 (95% confidence) was used. Based on the power analysis,
   quantifications were performed across 4–6 larvae for Drosophila assays.
   The n-value refers to the number of larvae. Individual data points for
   each analysis were averaged for each n and compared. Unless otherwise
   specified, the data compared was found to be normally distributed.
   Statistical significance of normal distributions was calculated by a
   two-sample two-tailed Student’s t test and/or ANOVA test in Excel,
   Minitab18 and/or by using the Proc GLM procedure followed by the pdiff
   mean comparison option in SAS Studio 3.81. Overlaid dot plots were
   constructed for all figures using OriginLab/OriginPro to represent
   mean ± SEM.
   RESOURCE SOURCE IDENTIFIER
   Antibodies and dyes
   Mouse anti-Nestin Santa Cruz

   Cat# sc-23927

   RRID: AB_627994
   Mouse anti-OCT-3/4 (C-10) Santa Cruz

   Cat# sc-5279

   RRID: AB_628051
   Mouse anti-MAP2 BD Biosciences

   Cat# 801212

   RRID: AB_2721321
   Mouse anti-βIII-tubulin (TUBB3) Biolegend

   Cat# ab152

   RRID: AB_390204
   Mouse anti-SYP (SY38) Phosphosolutions

   Cat# MAB5258

   RRID: AB_2313839
   Rabbit anti-SYT1 Thermofisher

   Cat# 1975-STG

   RRID: AB_2492251
   Mouse anti-HTT (1HU-4C8) EMD Millipore

   Cat# MAB2166

   RRID: AB_2123255
   Rabbit anti-HTT (EP867Y) Abcam

   Cat# ab45169

   RRID: AB_733062
   Rabbit anti-RAB4 (monoclonal) Abcam

   Cat# ab109009

   RRID: AB_10887396
   Rabbit anti-RAB4 (polyclonal) Abcam

   Cat# ab13252

   RRID: AB_2269374
   Rabbit anti-RAB5 Abcam

   Cat# ab18211

   RRID: AB_470264
   Mouse anti-RAB7 Santa Cruz Biotechnology

   Cat# sc-13156

   RRID: AB_627385
   Mouse anti-RAB2 Santa Cruz Biotechnology

   Cat# sc-133081

   RRID: AB_2176892
   Mouse anti-VPS35 Santa Cruz Biotechnology

   Cat# sc-374372

   RRID: AB_10988942
   Mouse anti-SUMO2/3 Cytoskeleton, Inc.

   Cat# ASM24

   RRID: AB_2884969
   Mouse anti-KIF5A Laboratory of Lawrence Goldstein [[410]140]
   Mouse anti-KIF5B Laboratory of Lawrence Goldstein [[411]140]
   Mouse anti-KIF5C Laboratory of Lawrence Goldstein [[412]140]
   Rabbit anti-DCTN1 (dynactin 1) Abcam

   Cat# ab96004

   RRID: AB_10677601
   Mouse anti-DYNC1I1 (74.1) Abcam

   Cat# ab23905

   RRID: AB_2096669
   Rabbit anti-ACTA1/Actin ThermoFisher Scientific

   Cat# MA5-32479

   RRID: AB_2809756
   Mouse anti- TUBA4A (tubulin, alpha 4A) Abcam

   Cat# ab7291

   RRID: AB_2241126
   Mouse anti-Golgi (7H6D7C2) Millipore Sigma

   Cat# 345867

   RRID: AB_564660
   Mouse anti-Cyt-c-p (A-8) Santa Cruz Biotechnology

   Cat# sc-13156

   RRID: AB_627385
   Rabbit anti-TOM20 (D8T4N) Cell Signaling Technology

   Cat# 42406S

   RRID: AB_2687663
   Mouse anti-MAP1B Santa Cruz Biotechnology

   Cat# sc-365668

   RRID: AB_10847224
   Mouse anti-MAP2 BD Pharmigen

   Cat# 556320

   RRID: AB_396359
   Rabbit anti-AKT1 Cell Signaling Technology

   Cat# 75692

   RRID:
   Rabbit anti-pAKT1 (Ser473) Cell Signaling Technology

   Cat# 9018

   RRID:
   Mouse anti-tyrosine hydroxylase EMD Millipore

   Cat# ab152

   RRID: AB_390204
   Rabbit anti-LAMP1 Abcam

   Cat# ab30687

   RRID: AB_775973
   Rabbit anti-GSK3α/β Cell Signaling Technology

   Cat# 5676

   RRID: AB_10547140
   Mouse anti-pGSK3α/β (pY279/pY216) Abcam

   Cat# ab68476

   RRID: AB_10013745
   Mouse anti-ERK (pan ERK) BD Transduction Laboratories

   Cat# 610124

   RRID: AB_397529
   Hoechst Thermofisher

   Cat# H3570

   RRID: AB_10626776
   Mouse anti-DCSP-3 (1G12) Developmental Studies Hybridoma Bank Cat#
   DCSP-3 (1G12) RRID: AB_528184
   Anti-Mouse Alexa Fluor® 488 Invitrogen

   Cat# A11001

   RRID: AB_2534069
   Anti-Mouse Alexa Fluor® 568 Invitrogen

   Cat# A11004

   RRID: AB_2534072
   Anti-Mouse Alexa Fluor® 647 Invitrogen

   Cat# A21235

   RRID: AB_2535804
   Anti-Rabbit Alexa Fluor® 488 Invitrogen

   Cat# A11008

   RRID: AB_143165
   Anti-Rabbit Alexa Fluor® 568 Invitrogen

   Cat# A11011

   RRID: AB_143157
   Anti-Rabbit Alexa Fluor® 647 Invitrogen

   Cat# A21244

   RRID: AB_141663
   Anti-mouse secondary antibody, HRP Invitrogen

   Cat# 32430

   RRID: AB_1185566
   Anti-Rabbit secondary antibody, HRP Invitrogen

   Cat# 32460

   RRID: AB_1185567
   Alexa Fluor® 594 Goat Anti-Horseradish Peroxidase Jackson Immuno
   Research Labs

   Cat# 123-585-021

   RRID: AB_2338966
   Fluorescein (FITC) Goat Anti-Horseradish Peroxidase Jackson Immuno
   Research Labs

   Cat# 123-095-021

   RRID: AB_2314647
   Experimental models: human cell lines
   ND42222, polyQ=109, 9 y, female NINDS Repository (Coriell Institute for
   Medical Research -Camden, NJ)

   Cat# ND42222

   RRID: CVCL_Y844
   ND38555, polyQ=17, 48 y, female NINDS Repository (Coriell Institute for
   Medical Research -Camden, NJ)

   Cat# ND3855

   RRID: CVCL_Y822
   Chemicals, peptides, recombinant proteins, and consumables
   GSK3β inhibitor CHIR99021 (CT99021) Selleck Chem

   Cat# S1263

   PubChem: 9956119
   ERK1/2 inhibitor SCH772984 Selleck Chem

   Cat# S7101

   PubChem: 24866313
   GST-GSK3β Signal Chem

   Cat# G09-10G

   RRID: N/A
   GST-ERK1 Signal Chem

   Cat# M29-10G

   RRID: N/A
   Corning Matrigel ThermoFisher Scientific

   Cat# CB40230A

   RRID: N/A
   Advanced DMEM/F12 Invitrogen

   Cat# 12634028

   RRID: N/A
   Essential 8 media Invitrogen

   Cat# A1517001

   RRID: N/A
   Neurobasal media Invitrogen

   Cat# 21103049

   RRID: N/A
   PSC neural induction media Invitrogen

   Cat# A1647801

   RRID: N/A
   B27 supplement media Invitrogen

   Cat# 17504-044

   RRID: N/A
   Protease inhibitor cocktail Pierce

   Cat# [413]PIA32965

   RRID: N/A
   Phosphatase inhibitor Pierce

   Cat# PI88667

   RRID: N/A
   Protein A/G Magnetic Beads Pierce

   Cat# PI88802

   RRID: N/A
   Vecta Shield Mounting Medium Vector Laboratories

   Cat# NC9265087

   RRID: N/A
   Phos-tag™ Acrylamide AAL-107 FujiFilm

   Cat# 300-93523

   RRID: N/A
   In Situ Cell Death Detection Kit Roche

   Cat# 11684795910

   RRID: N/A
   DNase I Roche

   Cat# 11284932001

   RRID: N/A
   Experimental Models: D. melanogaster organisms/strains
   P{Appl-GAL4.G1a}1, y^1 w* Bloomington Drosophila Stock Center

   BDSC: 32040;

   FlyBase: FBst0032040
   Appl-Gal4; T(2,3), CyO, TM6B,Ttb^1/Pin^88k Laboratory of Lawrence
   Goldstein [[414]137]
   pUAST-HTTex1-25Q-eGFP Laboratory of Norbert Perrimon [[415]139]
   pUAST-HTTex1-103Q-eGFP Laboratory of Norbert Perrimon [[416]139]
   w1118; P{UAS-sgg.B}MB5 (sgg^OE) Bloomington Drosophila Stock Center

   BDSC: 5361

   FlyBase: FBst0005361
   w1118; P{UAS-sgg.S9A}MB7/TM6C (sgg^active) Bloomington Drosophila Stock
   Center

   BDSC: 5362

   FlyBase: FBst0005362
   w1118; P{UAS-sgg.A81T}MB2 (sgg^LoF) Bloomington Drosophila Stock Center

   BDSC: 5359

   FlyBase: FBst0005359
   w1118; P{UAS-rl.K}2A (rl^OE) Bloomington Drosophila Stock Center

   BDSC: 36270

   FlyBase: FBst0036270
   y1 w*; P{UAS-rl^Sem.S}2 (rl^active) Bloomington Drosophila Stock Center

   BDSC: 59006

   FlyBase: FBst0059006
   rl^1 (rl^LoF) Bloomington Drosophila Stock Center

   BDSC: 386

   FlyBase: FBst0000386
   Software/algorithms
   MATLAB-based particle tracking program Laboratory of Danuser [[417]141,
   [418]142]
   ImageJ Schneider et. al., 2012 [419]https://imagej.net/ RRID:
   SCR_003070
   Metamorph/Metavue Imaging Software Molecular Devices, Sunnyvale, CA,
   USA RRID: SCR_002368
   Minitab18 [420]https://www.minitab.com/en-us/ RRID: SCR_014483
   Microsoft Excel [421]https://www.microsoft.com/en-gb/ RRID: SCR_016137
   RStudio [422]http://www.rstudio.com/ RRID:SCR_000432
   OriginLab/OriginPro [423]https://www.originlab.com/ RRID: SCR_014212
   [424]Open in a new tab

Supplementary information

   [425]Figure S1^ (1.2MB, png)
   [426]Figure S2^ (1.8MB, png)
   [427]Figure S3^ (1.6MB, png)
   [428]Figure S4^ (817.2KB, png)
   [429]Figure S5^ (1.5MB, png)
   [430]Figure S6^ (2.6MB, png)
   [431]Figure S7^ (2.2MB, png)
   [432]Table S1^ (27.7KB, docx)
   [433]Table S2^ (24.2KB, docx)

Acknowledgements