1. Neuroscience

Inhibition of PIP4Kγ ameliorates the pathological effects of mutant huntingtin protein

  1. Ismael Al-Ramahi
  2. Sai Srinivas Panapakkam Giridharan
  3. Yu-Chi Chen
  4. Samarjit Patnaik
  5. Nathaniel Safren
  6. Junya Hasegawa
  7. Maria de Haro
  8. Amanda K Wagner Gee
  9. Steven A Titus
  10. Hyunkyung Jeong
  11. Jonathan Clarke
  12. Dimitri Krainc
  13. Wei Zheng
  14. Robin F Irvine
  15. Sami Barmada
  16. Marc Ferrer
  17. Noel Southall
  18. Lois S Weisman
  19. Juan Botas  Is a corresponding author
  20. Juan Jose Marugan  Is a corresponding author
  1. Jan and Dan Duncan Neurological Research Institute, Texas Children's Hospital, United States
  2. Texas Medical Center, United States
  3. University of Michigan, United States
  4. National Center for Advancing Translational Sciences, United States
  5. Northwestern University, United States
  6. University of Cambridge, United Kingdom
https://doi.org/10.7554/eLife.29123
Share this article
Cite this article
  1. Ismael Al-Ramahi
  2. Sai Srinivas Panapakkam Giridharan
  3. Yu-Chi Chen
  4. Samarjit Patnaik
  5. Nathaniel Safren
  6. Junya Hasegawa
  7. Maria de Haro
  8. Amanda K Wagner Gee
  9. Steven A Titus
  10. Hyunkyung Jeong
  11. Jonathan Clarke
  12. Dimitri Krainc
  13. Wei Zheng
  14. Robin F Irvine
  15. Sami Barmada
  16. Marc Ferrer
  17. Noel Southall
  18. Lois S Weisman
  19. Juan Botas
  20. Juan Jose Marugan
(2017)
Inhibition of PIP4Kγ ameliorates the pathological effects of mutant huntingtin protein
eLife 6:e29123.
https://doi.org/10.7554/eLife.29123
  2. Download BibTeX
  3. Download .RIS

Abstract

The discovery of the causative gene for Huntington’s disease (HD) has promoted numerous efforts to uncover cellular pathways that lower levels of mutant huntingtin protein (mHtt) and potentially forestall the appearance of HD-related neurological defects. Using a cell-based model of pathogenic huntingtin expression, we identified a class of compounds that protect cells through selective inhibition of a lipid kinase, PIP4Kγ. Pharmacological inhibition or knock-down of PIP4Kγ modulates the equilibrium between phosphatidylinositide (PI) species within the cell and increases basal autophagy, reducing the total amount of mHtt protein in human patient fibroblasts and aggregates in neurons. In two Drosophila models of Huntington’s disease, genetic knockdown of PIP4K ameliorated neuronal dysfunction and degeneration as assessed using motor performance and retinal degeneration assays respectively. Together, these results suggest that PIP4Kγ is a druggable target whose inhibition enhances productive autophagy and mHtt proteolysis, revealing a useful pharmacological point of intervention for the treatment of Huntington’s disease, and potentially for other neurodegenerative disorders.

https://doi.org/10.7554/eLife.29123.001

Introduction

Huntington's disease (HD) is an autosomal dominant neurodegenerative disorder with no curative or preventative treatment options. The disease is caused by the expansion of a translated CAG trinucleotide repeat within exon 1 of the huntingtin gene (HTT), resulting in a mutant huntingtin (mHtt) protein with an abnormally long N-terminal tract of glutamine residues (Ross and Tabrizi, 2011). Individuals with more than 36 to 39 repeats develop the disorder, and the length of the repeat correlates with the age of disease onset (Walker, 2007). The poly-glutamine repeat expansion impacts the physical (Kazantsev et al., 1999) and physiological (Hipp et al., 2012; Verhoef et al., 2002; Fernandez-Estevez et al., 2014) properties of the huntingtin protein, producing aggregates in aged striatal neurons that eventually precipitate to form neuronal inclusion bodies (Miller et al., 2010a). Accumulation of mHtt triggers a variety of insults that lead to striatal degeneration, however, the nature of the specific mHtt species, soluble, oligomeric or aggregate, that triggers neurodegeneration remains unclear (Arrasate et al., 2004; Lajoie and Snapp, 2010). In the last decade, a number of potential therapeutic avenues have been proposed to prevent or attenuate the neurodegeneration induced by mHtt, including examining the effects of mHtt-induced oxidative stress (Wyttenbach et al., 2002; Giuliano et al., 2003; Lu et al., 2014), huntingtin posttranscriptional modifications (Steffan et al., 2004; Greiner and Yang, 2011; Bhat et al., 2014; Pavese et al., 2006), microglia activation (Gusella and MacDonald, 2009), a systematic exploration of coding (Gusella and MacDonald, 2009) and non-coding (Zhang and Friedlander, 2011) DNA, and autophagy (Sarkar et al., 2009; Williams et al., 2008). However, it has been difficult to identify druggable targets that reduce disease progression (Bard et al., 2014). In addition to targeting mHtt-induced downstream pathogenic events, an attractive alternative for developing HD therapies is reducing the levels of mHtt protein, thus addressing pathogenesis at its root. The therapeutic potential of this approach is supported by observations in animal and cellular models of HD (Sarkar et al., 2009; King et al., 2008; Singh et al., 2014; Giorgini, 2011; Yamamoto et al., 2000; Lin and Qin, 2013; Sarkar et al., 2007). Here we present PIP4Kγ as a novel therapeutic target for HD. PIP4Kγ [Phosphatidylinositol-5-phosphate 4-kinase, type II γ] is a lipid kinase expressed by the PIP4K2C gene. The protein is predominantly localized in several tissues, including the brain (Sasaki et al., 2009; Rameh et al., 1997; Clarke et al., 2008; Clarke et al., 2009). Enzymatically, PIP4Kγ phosphorylates phosphatidylinositol-5-phosphate [PI5P] to produce phosphatidylinositol 4,5-bisphosphate [PI(4,5)P2] (Lietha, 2011). The biological function of PIP4Kγ is not completely understood, although recent reports suggest a role in the modulation of vesicle trafficking (Clarke et al., 2008), and mTOR signaling (Mackey et al., 2014). Recently we presented the first selective inhibitor of PIP4Kγ (Clarke et al., 2015). Here we introduce an additional chemotype with striking cell-based activity which prompted us to explore the utility of inhibiting PIP4Kγ in the context of pathologic mHtt expression. We show that inhibiting PIP4Kγ activity modulates productive autophagy, reduces mHtt protein levels in patient fibroblasts, and clears mHtt aggregates in neuronal cell models. Moreover, we show that inhibition of PIP4Kγ rescues mHtt-induced neurodegeneration in two Drosophila HD models.

Results

Identification of novel PIP4Kγ inhibitors

NCT-504 (Figure 1A) is an analogue obtained upon medicinal chemistry optimization of a series of 5-phenylthieno[2,3-d]pyrimidine compounds identified in a high-throughput phenotypic screen (Titus, 2010). Expression of GFP-Htt(exon1)-Q103 in PC12 cells produces detergent-resistant GFP-labeled aggregates (Titus et al., 2012). NCT-504 caused a robust reduction of GFP-Htt(exon1)-Q103 levels, as measured by lowered GFP signal (Figure 1B and C). NCT-504 treatment also decreased huntingtin aggregates in HEK293T cells transiently transfected with GFP-Htt(exon1)-Q74 (Figure 1—figure supplement 1). As thienopyrimidines have been associated with kinase activity (Elrazaz et al., 2015) we profiled NCT-504 against a panel of 442 human kinases http://www.discoverx.com/technologies-platforms/competitive-binding-technology/kinomescan-technology-platform. Using a cutoff of >65% inhibition at 10 μM, NCT-504 was active against only a single kinase, PIP4Kγ (Table 1). Similarly, another analogue from the same thienopyrimidine series, ML168 (Titus, 2010), had activity against six kinases in the same panel, but was most potent against PIP4Kγ.

Figure 1 with 3 supplements see all
Download asset Open asset
Identification of NCT-504 and its inhibition of PIP4Kγ.

(A) Structure of NCT-504. (B) NCT-504 treatment reduces Htt(exon1)-Q103 in PC12 cells. Cells with stable expression of ecdysone-inducible GFP-Htt(exon1)-Q103 (green), induced for 24 hr, and treated with DMSO (top panels) or 23 μM NCT-504 (bottom). Cells stained with DAPI (blue). Scale Bar = 50 μm. (C) Concentration-response curve of NCT-504 inhibition of cellular accumulation of GFP-Htt(exon1)-Q103 in PC12 cells. (D) NCT-504 inhibition of PIP4Kγ binding to an immobilized proprietary active site ligand (DiscoverX KINOMEscan https://www.discoverx.com/services/drug-discovery-development-services/kinase-profiling/kinomescan). (E) NCT-504 exhibits dose-dependent inhibition of phosphorylation of PI4P by full length isolated PIP4Kγ. (F) The intrinsic ATPase specific activity of full length isolated PIP4Kγ in the absence of PI5P substrate as a function of NCT-504 concentration is the same in the presence (blue) or in the absence (purple) of NCT-504.

https://doi.org/10.7554/eLife.29123.002
Table 1
Kinase profiling results for NCT-504 and ML168.

Percent activity remaining at 10 μM exposure of NCT-504 and ML168 in KINOMEscan kinase panel/profiling http://www.discoverx.com/technologies-platforms/competitive-binding-technology/kinomescan-technology-platform. Top 3 NCT-504 inhibited kinases are reported as single replicate data. Full data set is provided in Table 1 – source data file. PIP4K2γ potencies were confirmed in triplicate concentration-response testing (Figure 1D).

https://doi.org/10.7554/eLife.29123.006
KinaseML168NCT-504
PIP4K2C234.9
RSK1(Kin.Dom.2-C-terminal)2040
GAK1042

  1. % Control Legend

    0% ≤ x < 10%

  2. 10% ≤ x < 35%

    35% ≤ x

To better characterize the biochemical action of NCT-504, we evaluated its inhibitory activity in several in vitro kinase assays. NCT-504 modulated the activity of PIP4Kγ in the DiscoverX binding assay (https://www.discoverx.com/services/drug-discovery-development-services/kinase-profiling/kinomescan) with a Kd = 354 nM (Figure 1D). Using a reconstituted assay of phosphorylation of the PI5P substrate by full length PIP4Kγ, NCT-504 inhibited enzyme activity with an IC50 of 15.8 μM (Figure 1E). Notably, in the absence of PI5P substrate, the compound did not impair the intrinsic ATP-hydrolytic activity of PIP4Kγ (Figure 1F), suggesting that NCT-504 is an allosteric inhibitor of this kinase. This may account for the differences in potency observed in the enzymatic assay vs the DiscoverX binding assay. Similar differences in potency between these two assays have also been observed for allosteric modulators of other kinases (Rudolf et al., 2014; Smyth and Collins, 2009). NCT-504 function as an allosteric inhibitor may also explain why NCT-504 is exquisitely selective in the kinase profiling assay. In isolated enzyme assays against other PIP4K isoforms, 50 μM NCT-504 did not inhibit PIP4Kbeta or PIP4Kalpha (IC50 between 50 μM and 100 μM) (Figure 1—figure supplement 2). We also characterized the compound using an alternate PIP4Kγ+ functional assay, which employs PIP4Kγ with a mutated G-loop and two additional mutations (described as PI5P4Kγ G3 + AB in [Clarke and Irvine, 2013]) to increase the low intrinsic ATP turnover of the kinase in the presence of PI5P (Clarke and Irvine, 2013). NCT-504 was largely inactive against PIP4Kγ+ with a potency >500 μM (Figure 1—figure supplement 3).

PIP4Kγ inhibition modulates cellular phosphatidylinositide levels in complex ways

Cellular inhibition of PIP4Ks should impair the production of PI(4,5)P2 from PI5P, resulting in an elevation of PI5P cellular levels as previously described in the Drosophila mutant (Gupta et al., 2013). Note that other PI levels were not tested in the dPI4PK Drosophila mutant. We hypothesized that elevation of PI5P might further impact the equilibrium between various PI species (Lietha, 2011; Emerling et al., 2014; Balla, 2013). To test this hypothesis, we exposed wild type mouse embryonic fibroblasts to nontoxic concentrations of NCT-504 (10 μM) for 12 hr, and then evaluated the levels of PI by HPLC (Figure 2; toxicity assay in Figure 2—figure supplement 1). As expected, exposure to NCT-504 elevated cellular levels of PI5P (Figure 2D). Surprisingly, NCT-504 also robustly increased PI(3,5)P2 levels, and to a lesser extent increased levels of PI3P (Figure 2B and E). We did not observe an effect on PI(4,5)P2 levels (Figure 2F), which is consistent with other reports indicating that the cellular levels of this lipid are mostly generated from PI4P via type I PI4P 5-kinases (Lietha, 2011). Kinetic measurement of PI levels showed that NCT-504 causes an increase in PI5P, PI(3,5)P2 and PI3P levels along with a decrease in PI4P, progressively over 12 hr (Figure 2—figure supplement 2). These statistically significant changes were not observed at 30 or 120 min suggesting that direct inhibition of PIP4Kγ eventually impacts other lipid kinases and phosphatases. Moreover treatment of unaffected human fibroblasts with NCT-504 elevated these three lipids in a dose dependent manner (Figure 2—figure supplement 3). Further evidence that the changes in PI levels are due to the specific inhibition of PIP4Kγ, is the finding that shRNA-mediated silencing of PIP4Kγ resulted in a similar PI profile to that observed with NCT-504 inhibition, namely an elevation of PI5P, PI(3,5)P2 and PI3P (Figure 2B,D and E). Note that during shRNA-mediated silencing of PIP4Kγ transcripts, PIP4Kγ protein was no longer detected (Figure 2G).

Figure 2 with 3 supplements see all
Download asset Open asset
Pharmacologic and genetic inhibition of PIP4Kγ elevates the levels of PI(3,5)P2, PI3P and PI5P in MEFs.

(A–F) Pharmacologic (NCT-504 10 μM, 12 hr) and genetic (shRNA) inhibition of PIP4Kγ leads to increased levels of PI5P (D), PI(3,5)P2 (E) and PI3P (B), with no significant change in the levels of phosphatidylinositol (A), PI4P (C) or PI(4,5)P2 (F). However, there was a modest reduction in PI4P. Note in Figure 2—figure supplement 2, this small change was statically significant. Measurements were performed in MEF cells incubated with 3H-inositol labeled media for 48 hr. Statistical significance was analyzed using paired one tailed student t-test (n = 3), *p<0.05, **p<0.01. (G) Anti-PIP4Kγ western blot showing the effective silencing of the enzyme using shRNA. (GAPDH used as loading control).

https://doi.org/10.7554/eLife.29123.007

PIP4Kγ inhibition stimulates productive autophagy

Numerous studies have shown that mHtt upregulates autophagy, but impairs incorporation of client proteins into autophagosomes (Cortes and La Spada, 2014; Ochaba et al., 2014; Martin et al., 2015; Martinez-Vicente et al., 2010; Tsvetkov et al., 2013). Importantly, a number of autophagy modulators have been described that reduce mHtt aggregates (Sarkar et al., 2009; Roscic et al., 2011; Zhang et al., 2007; Renna et al., 2010). That NCT-504 elevates the levels of three PI species implicated as positive regulators of autophagy suggests that the observed reduction in HTT-exon1-polyQ aggregates observed with NCT-504 treatment may occur due to upregulation of autophagy. Autophagy can be monitored by following the fate of microtubule-associated protein 1 light chain 3B (LC3-I). During autophagosome formation LC3-I gets conjugated to phosphatidylethanolamine to form LC3-II, which is degraded upon autophagosome-lysosome fusion (Tanida et al., 2008). We tested and found that a two hour incubation of HEK293T cells with 5 or 10 μM NCT-504 did not significantly increase LC3-II levels (Figure 3A and B). However, LC3-II levels depend on the rate of autophagosome formation, the rate of autophagosome-lysosome fusion, and on the rate of LC3-II degradation in mature autolysomes. Bafilomycin A1 inhibits the lysosomal v-ATPase, prevents autophagosome-lysosome fusion, and thus prevents autophagy-mediated degradation of LC3-II. Comparison of cells treated with and without bafilomycin A1 is a common method to monitor the rate of autophagosome formation within the cell independent of later steps (Barth et al., 2010). Bafilomycin A1 treatment for 2 or 6 hr elevated the total amount of LC3-II (Figure 3A and C). Importantly, treating cells with 10 μM NCT-504 and 100 nM bafilomycin A1 for two hours and six hours resulted in a 38% and 51% increase in LC3-II levels respectively compared with bafilomycin A1 treatment alone, which indicates that NCT-504 induces autophagosome formation. Similarly, treating cells with 5 μM NCT-504 and bafilomycin A1 for two and six hours resulted in a 30% and 46% increase in LC3-II levels respectively. Importantly, an elevation in LC3-II levels by NCT-504 in the presence of bafilomycin A1 but not in the absence of bafilomycin A1, suggests that NCT-504 elevates both the induction of autophagy as well as the rate of turnover of autophagic cargo (autophagy flux). To further evaluate the effects of NCT-504 on autophagosome formation and autophagy flux, we used a 293A cell line stably expressing a GFP-mCherry-LC3 reporter (Figure 3—figure supplement 1). This double tagged LC3 is commonly used to distinguish between autolysosomes and autophagosomes or phagophores (Hundeshagen et al., 2011; Kimura et al., 2007). Phagophore and autophagosome membranes conjugated with GFP-mCherry-LC3 are positive for both GFP- and mCherry-fluorescence. Upon generation of mature autolysosomes via fusion of autophagosomes with lysosomes, the GFP fluorescence from the internalized GFP-mCherry-LC3 is quenched in the acidic lysosomes; whereas mCherry fluorescence is insensitive to acidic pH and remains detectable. Thus, membrane structures positive for mCherry fluorescence, but not GFP fluorescence are autolysosomes. We determined the dose and time response of NCT-504 on autophagosomes and autolysosomes using GFP-mCherry-LC3; bafilomycin and torin-1 were used as controls (Figure 3—figure supplement 1). As previously reported, bafilomycin treatment resulted in an increase in autophagosomes because the subsequent formation of autolysosomes is blocked. In addition, as previously reported, torin treatment elevated both the number of autophagosomes and autolysosomes, because inhibition of mTORC1 causes an increase in the induction of autophagy as well as an increase in autophagic flux. In contrast, NCT-504 treatment caused a robust increase in the formation of autolysosomes with only a modest elevation in autophagosomes, which indicates that NCT-504 increases autophagic flux, with only a modest increase in autophagy initiation.

Figure 3 with 3 supplements see all
Download asset Open asset
Inhibition of PIP4Kγ increases autophagy flux.

(A) Representative Western blots showing the levels of LC3-I, LC3-II and Tubulin (loading control) in HEK293T cells treated with either 5 or 10 μM NCT-504 or DMSO (control) for two or six hours in the presence or absence of 100 nM bafilomycin. (B–C) Quantification of LC3-II levels detected by western blot normalized to α-tubulin (loading control). Changes in LC3-II with drug treatment alone is presented relative to levels in the DMSO control cell lysates (B) and changes in LC3-II with drug treatment plus bafilomycin is presented relative to DMSO plus bafilomycin (C). Statistical significance was quantified from three independent experiments using Dunnett's multiple comparisons test, *p<0.05, **p<0.01, ***p<0.005.

https://doi.org/10.7554/eLife.29123.011

While mechanisms of autophagy are highly similar in all cells, neurons exhibit some key differences. For example starvation does not upregulate autophagy (Mizushima et al., 2004). In addition, autophagy is spatially regulated (Maday and Holzbaur, 2014). Thus, we tested whether NCT-504 impacts autophagy in neurons. We tested several doses and time points (up to 72 hr after treatment) and measured autophagy flux in DIV4 rat primary cortical neurons transfected with Dendra2-LC3, a photoconvertable reporter (Figure 3—figure supplement 2). Dendra2 has excitation-emission maxima that are similar to GFP. However, exposure to intense blue light raise these maxima, and thus red light is emitted. Since the photoconversion reaction is irreversible, and LC3 is both a marker of autophagy as well as a substrate, the disappearance of red Dendra2-LC3 over time can be used to assess autophagy flux in a noninvasive manner (Gupta et al., 2017; Barmada et al., 2014). As a positive control for an increase in autophagic flux in neurons, we co-expressed Beclin-1, a positive regulator of autophagy that increases autophagy activity when overexpressed (Kang et al., 2011). Importantly, treatment of rat primary cortical neurons expressing Dendra2-LC3 with either 500 nM or 1 µM NCT-504 enhanced the rate of Dendra2-LC3 turnover. Thus, NCT-504 stimulates autophagy flux in primary rodent cortical neurons in a statistically significant manner up to 72 hr following treatment.

That NCT-504-induced changes in autophagic flux were dose dependent, led us to test whether the resultant increase in autophagy correlated with changes in Htt levels. We found that 293A cells display a high content of wt Htt, which enabled us to use an anti-Htt FRET assay (Cui et al., 2014). Using this assay, we found that NCT 504 treatment resulted in a dose dependent decrease of Htt protein levels at levels that did not impact cell viability (Figure 3—figure supplement 1C and D).

To further test whether NCT-504 reduces mHtt aggregates via increasing autophagic flux, we tested the ability of NCT-504 to lower GFP-Htt(exon1)-Q74 aggregates in a cells with a defect in macroautophagy. We found that while NCT-504 lowered the levels of GFP-Htt(exon1)-Q74 aggregates in Atg7+/+ MEF, it failed to lower aggregates in Atg7-/- MEF; Atg7 is essential for autophagosome formation and its loss inhibits the autophagy pathway (Figure 3—figure supplement 3).

That PI3P is a critical regulator of autophagy (Shibutani and Yoshimori, 2014), and that PI5P and PI(3,5)P2 have also been implicated in the autophagy process (Vicinanza et al., 2015; Hasegawa et al., 2017), suggests that upregulation of one or more of these lipids is the driver behind the increase in autophagic flux. Importantly, NCT-504 treatment contrasts with the action of other autophagy modulators such as mTORC1 inhibitors which produce stable increases in LC3-II (Boland et al., 2008), accelerating the initiation of autophagy but not necessarily later steps which require mTOR reactivation (Munson and Ganley, 2015).

Blocking PIP4Kγ activity reduces levels of full-length mutant huntingtin protein and levels of Htt(exon1)-polyQ aggregates

To test whether PIP4Kγ inhibition lowers full-length mHtt protein, we used immunoblots to determine the effect of NCT-504 on mHtt levels in patient fibroblasts and immortalized striatal neurons from a knock-in HD mouse model. Notably, treatment with 5 μM NCT-504 for 12 hr, conditions that did not affect cell viability (Figure 4—figure supplement 1), significantly reduced mHtt levels in fibroblasts from two different HD patients HD(Q68) or HD(Q45) (Figure 4A and C). To further test whether the reduction of mHtt levels was due to selective modulation of PIP4Kγ, we individually silenced PIP4K2A, PIP4K2B and PIP4K2C RNA in the HD(Q68) patient fibroblast cell line. Only silencing of PIP4K2C exhibited an appreciable and robust reduction of huntingtin protein levels (Figure 4B). Note that silencing of PIP4K2A, PIP4K2B and PIP4K2C was effective and specific for each isoform (Figure 4—figure supplement 2). We also tested the effect of NCT-504 on the levels of mutant full-length huntingtin protein in immortalized striatal neurons. We treated a striatal cell line from a knock-in HD mouse (STHdhQ111) (Trettel et al., 2000), with 5 μM NCT-504 for 12 hr and observed a 40% decrease in mHtt levels (Figure 4D).

Figure 4 with 5 supplements see all
Download asset Open asset
Chemical inhibition of PIP4Kγ or knock-down of the corresponding mRNA, PIP4K2C, lowers mHtt protein levels in cells from HD patients and HD knock-in mice.

(A) Reduction of mHtt protein levels in an HD patient fibroblast cell line (Q68) following exposure for 12 hr to NCT-504 (5 μM) (B) mHtt protein levels in patient fibroblast cell line (Q68) were analyzed following siRNA-mediated silencing of PIP4K2A, PIP4K2B and PIP4K2C genes. Note that only PIP4K2C knockdown lowers mHtt levels. Control experiments showing silencing specificity on PIP4K protein levels are in Figure 4—figure supplement 3. (C) Reduction of mHtt protein levels in an HD patient fibroblast cell line (Q45) following exposure to NCT-504 (5 μM). (D) Reduction of mHtt protein levels in immortalized striatal cells from knock-in HD mice (STHdhQ111) treated for 12 hr with NCT-504 (5 μM).

https://doi.org/10.7554/eLife.29123.015

To examine the impact of NCT-504 on the levels of huntingtin-related aggregates in neurons, we evaluated the effect of NCT-504 in wild-type mouse primary cortical neurons transfected with Htt(exon1)-Q74. We tested and found that concentrations of NCT-504 of 5 μM or lower did not impact the viability of cortical neurons (Figure 4—figure supplement 3). Importantly, 2.5 or 5 μM NCT-504 lowered the levels of Htt(exon1)-Q74 in primary cortical neurons (Figure 4—figure supplement 4A). Moreover, depletion of PIP4Kγ in cortical neurons via PIP4K2C-shRNA treatment also led to a decrease in Htt(exon1)-Q74 levels and Htt(exon1)-Q74 aggregates (Figure 4—figure supplement 4B). Furthermore, NCT-504 treatment and PIP4K2C silencing each reduced Htt(exon1)-polyQ aggregates in neuroblastoma N2a cells transfected with Htt(exon1)-polyQ mutants (Figure 4—figure supplement 5).

Collectively, these studies show that NCT-504, a PIP4Kγ kinase inhibitor, at non-toxic concentrations, reduced full length huntingtin protein in patient fibroblasts, in immortalized striatal neurons from STHdhQ111 mutant mice and in HEK293T cells. Moreover, NCT-504 reduced the levels of Htt(exon1)-polyQ aggregates in primary cultured neurons and several cell lines. Similarly, specific silencing of the PIP4K2C gene led to reduction in the levels of full-length huntingtin and HTT-exon1-polyQ protein and aggregates. The lowering of huntingtin and Htt(exon1)-polyQ by NCT-504 was concentration dependent. Moreover, the levels of NCT-504 that reduced these mutant proteins increased autophagic flux. Importantly, NCT-504 did not lower Htt(exon1)-polyQ protein in Atg7-/- MEF, but lowered Htt(exon1)-polyQ protein in the corresponding Atg7+/+ MEF. Together these studies indicate that inhibition of PIP4Kγ lowers mutant Htt, via an increase in autophagic flux.

Phenotypic effects of PIP4K modulation in Drosophila models of Huntington’s disease

Unlike mammals, which have three PIP type II enzymes (PIP4Kalpha, PIP4Kbeta and PIP4Kγ), there is only one type II PIP kinase homologue in Drosophila (dPIPK4 also called CG17471) (Mackey et al., 2014). We used a well-established HD Drosophila model (Kaltenbach et al., 2007; Branco et al., 2008; Miller et al., 2010b; Lu et al., 2013; Yao et al., 2015) to evaluate the impact of modulating the dPIP4K gene on the pathogenesis induced by mHtt expression. The GAL4/UAS system (Elliott and Brand, 2008) is used to drive expression of an N-terminal human 128Q mHtt (HttN231Q128) fragment to the cell type of choice. First, we assessed the Drosophila retina and its photoreceptor cells. Control HD model animals with wild-type activity of dPIP4K show prominent mHtt-induced photoreceptor degeneration. This phenotype is ameliorated by reducing dPIP4K activity with either one of two different shRNAs (Figure 5A). In a second set of experiments we tested the potential of dPIP4K to modulate mHtt pathogenesis using a behavioral readout. Neuronal-specific expression of HttN231Q128 leads to a late-onset motor impairment that can be quantified in a climbing assay. This phenotype was also ameliorated by reducing the activity of dPIP4K using a previously described (Gupta et al., 2013) classical loss-of-function mutant allele in heterozygosis and a kinase dead allele (Figure 5B). Additionally, we also evaluated these approaches (loss-of-function by a heterozygous mutant allele and kinase dead allele) in animals expressing full length Htt carrying a 200 polyQ expansion in exon1. Notably, we observed a mitigation of the motor performance decline in this full-length HD model. Decreasing the levels of PIP4K with the same alleles in the absence of mHtt did not affect motor performance when compared to controls (Figure 5—figure supplement 1). Thus, reducing the activity of dPIP4K using different genetic approaches mitigates mHtt pathogenesis in three different assays.

Figure 5 with 1 supplement see all
Download asset Open asset
Reduced dPIP4K gene activity ameliorates photoreceptor degeneration and behavioral impairments in a Drosophila HD model.

(A) Sections through the Drosophila retina showing loss of photoreceptor cells and retinal tissue in animals expressing N-terminal mHtt (HTTNT231Q128) in the eye (compare no modifier with negative control panels). The photoreceptor and retinal loss phenotype is ameliorated in HttNT231Q128 animals that also express anyone of two shRNAs targeting dPIP4K. (B) Chart shows motor performance (%) as a function of age in negative controls (dPIP4K+/+, blue dotted line), Drosophila expressing N-terminal mHtt in the CNS (HTTNT231Q128/dPIP4K+/+, black line) or animals expressing N-terminal mHtt in the CNS together with a dPIP4K heterozygous loss of function (HTTNT231Q128/dPIP4K+/-, red continuous line) or a dPIP4K kinase dead isoform (HTTNT231Q128/dPIP4K+/DN, green continuous line). Notice the amelioration of mHtt-induced deficits upon decreasing the activity of dPIP4K. (C) Chart shows motor performance (%) and climbing speed as a function of age in negative controls (dPIP4K+/+, blue dotted line), Drosophila expressing full length mHtt in the CNS (HTT-FLQ200/dPIP4K+/+, black line) or animals expressing FL mHtt in the CNS together with a dPIP4K heterozygous loss of function (HTT-FL200/dPIP4K+/-, red continuous line) or a dPIP4K kinase dead isoform (HTT-FL200/dPIP4K+/DN, green continuous line). Note amelioration of neural HttNT231Q128-induced motor deficits by decreasing the activity of dPIP4K. Genotypes in A: Negative control: GMR-GAL4/+; dPIP4K+/+. No modifier: GMR-GAL4/+; UAS:HTTNT231Q128/+; dPIP4K+/+. PIP4K2 sh1/sh2: GMR-GAL4/+; UAS:HTTNT231Q128/UAS:dPIP4Ksh-1 or sh-2. Genotypes in B: Negative control: elavc155GAL4/+; dPIP4K+/+.HTT231Q128: elavc155GAL4/+; UAS:HttNT231Q128/+; dPIP4K+/+. HTT231Q128/PIP4K2LOF: elavc155GAL4/+; UAS:HttNT231Q128/+; dPIP4K29/+ and HTT231Q128/PIP4K2DN: elavc155GAL4/+; UAS:HttNT231Q128/UAS:dPIP4K29[D271K].. Genotypes in C: Negative control: elavc155GAL4/+; dPIP4K+/+. HTT-FLQ200: elavc155GAL4/+; UAS:HttFLQ200/+; dPIP4K+/+. HTT-FLQ200/PIP4K2LOF: elavc155GAL4/+; UAS:UAS:HttFLQ200/+; dPIP4K29/+ and HTT-FLQ200/PIP4K2DN: elavc155GAL4/+; UAS:UAS:HttFLQ200/UAS:dPIP4K29[D271K]. elavc155GAL4 drives expression of mHtt to all neurons but not other cell types. Means between points at each age were analyzed by ANOVA followed by Dunnet’s post hoc test. Error bars indicate the s.e.m. *p<0.05.

https://doi.org/10.7554/eLife.29123.021

Discussion

Our unbiased screen for compounds that protect cells against a pathogenic huntingtin fragment reveal PIP4Kγ as a potential target for Huntington disease. The compounds identified led to the development of NCT-504, a selective fully efficacious inhibitor of PIP4Kγ. NCT-504 treatment or knock-down of PIP4Kγ lowers huntingtin fragments Htt(exon1)-polyQ in multiple cell types including cortical and striatal neurons, and lowers full-length mutant huntingtin in patient fibroblasts and mouse striatal neurons. Moreover, genetic targeting of PIP4K in two Drosophila models of HD, mitigated associated HD phenotypes. Importantly, we observed two major changes in cells following PIP4Kγ inhibition, an increase in autophagic flux, and an increase in the levels of three phosphoinositide signaling lipids. It is tempting to speculate that the changes in PI upregulate autophagic flux, and thereby lower mHtt levels.

Little is currently known about cellular roles of PIP4Kγ. However, in line with our current findings, previous studies observed that knock-down of PIP4Kγ resulted in an increase in autophagy (Mackey et al., 2014; Vicinanza et al., 2015), and a reduction of EGFP-HttQ74 aggregates in MEFs that was dependent on the presence of the autophagy gene, ATG7 (Mackey et al., 2014; Vicinanza et al., 2015).

While increasing the proteolysis of pathogenic huntingtin protein via the upregulation of autophagy is an attractive therapeutic approach for HD (Lin and Qin, 2013; Sarkar et al., 2007), there are potential challenges. Mutant huntingtin itself may alter autophagy. Wild-type huntingtin may be an adaptor for selected autophagic cargoes including itself (Martinez-Vicente et al., 2010). Consistent with this hypothesis, mutant huntingtin impairs the loading of ubiquitinated-tagged proteins into autophagosomes (Martinez-Vicente et al., 2010), and circumvents its own clearance (Martin et al., 2015). Moreover, mutant huntingtin sequesters diverse proteins required for key cellular processes (Kim et al., 2016), including mTOR, which plays key roles in the regulation of autophagy (Tsvetkov et al., 2013; Petersén A et al., 2001; Ravikumar et al., 2004; Pryor et al., 2014; Ashkenazi et al., 2017; Caviston et al., 2007; Wong and Holzbaur, 2014). In addition, the PI binding autophagy adaptor protein ALFY which plays a fundamental role in degrading mutant huntingtin is down-regulated in HD (Martin et al., 2015; Filimonenko et al., 2010). In contrast with these findings, there are studies that indicate that autophagy is not impaired in HD, and have revealed an elevation of autophagy flux in HD cells (Petersén A et al., 2001; Kegel et al., 2000). Despite possible mutant huntingtin-dependent changes on autophagy, upregulation of autophagy remains a viable approach for lowering levels of mutant huntintin and aggregates (Lin and Qin, 2013; Sarkar et al., 2007). Note that caloric restriction also raises the basal level of autophagy, leading to improvements in HD models (Duan et al., 2003) and increasing axonal autophagosome transport (Ikenaka et al., 2013), although it is less clear how to translate this observation into clinical practice.

One common approach to induce autophagy is via inhibition of the major metabolic kinase mTORC1. Indeed, rapamycin, an inhibitor of mTORC1 also reduces mHtt protein levels (Sarkar et al., 2009). However, while PIP4Kγ likely impacts mTORC1 activity, it is not yet clear whether PIP4Kγ inhibition results in mTORC1 inhibition or activation. One study showed that knock-down of PIP4K2C inhibits mTORC1 (Mackey et al., 2014). However, in PIP4K2C homozygous knock-out mice, mTORC1 is elevated (Shim et al., 2016). Thus, the precise link between PIP4Kγ and mTORC1 is not clear. Importantly, our data suggest that PIP4Kγ upregulation of autophagy has some differences with upregulation of autophagy via mTORC1 inhibition. While inhibition of mTOR via torin treatment exhibited a large increase in both autophagosome formation and autophagy flux, inhibition of PIP4Kγ had only a modest impact on autophagosome formation, but had a large increase in autophagic flux (Figure 3—figure supplement 1). Thus, elucidation of the mechanism whereby PIP4Kγ inhibition increases autophagic flux remains to be fully determined.

It is likely that inhibition of PIP4Kγ increases autophagic flux at least in part via the resulting impact on the levels of selected phosphoinositide lipids. PIP4Kγ is predicted to convert PI5P to PI(4,5)P2. However, it was not known which cellular pools of PI5P are substrates for PIP4Kγ. Using NCT-504 we found no significant change in PI5P levels or other PI lipids following up to two hours of inhibition of PIP4Kγ. This contrasts with other lipid kinases such as PIKfyve, where direct inhibition results in an acute loss of PI(3,5)P2 which can be observed within 5 min (Zolov et al., 2012; McCartney et al., 2014). The long delay prior to changes in PI5P following PIP4Kγ inhibition suggests that PIP4Kγ is not in contact with most of the cellular PI5P, or is only active under specific conditions. However, after 12 hr treatment with NCT-504, there was a 1.6 fold elevation of PI5P. The fact that this change occurred well after 2 hr of inhibition, suggests that it might not be directly due to an accumulation of the PI5P substrate normally used by PIP4Kγ. Indeed, at 12 hr PI(3,5)P2 levels were also elevated at 2-fold, which was even higher than the fold elevation in PI5P. This raises the possibility that long-term inhibition of PIP4Kγ indirectly results in the activation of PIKfyve. This activation of PIKfyve may account for the increase in PI5P as well as PI(3,5)P2 (Zolov et al., 2012; McCartney et al., 2014; Sbrissa et al., 2012). In addition, at 12 hr PI3P levels increased 1.3 fold, suggesting that VPS34 may be indirectly activated as well.

The elevation of PI3P, PI(3,5)P2 and/or PI5P likely contribute to the elevation in autophagic flux. PI3P has well characterized roles in autophagy, and acts in initiation of phagophore formation (Shibutani and Yoshimori, 2014) as well as in later steps of autophagosome maturation (Carlsson and Simonsen, 2015), including autophagosome-lysosome fusion (Ikonomov et al., 2006) and lysosome reformation (Rong et al., 2012; Yu et al., 2010). In some conditions, PI5P can induce autophagy independent of PI3P (Vicinanza et al., 2015). In contrast, PI(3,5)P2 functions at a late step in autophagy (Ikonomov et al., 2006; de Lartigue et al., 2009; Jin et al., 2008; Jin et al., 2014; Ikonomov et al., 2001; Martin et al., 2013; Rusten et al., 2007; Sano et al., 2016). These late functions, may contribute to the observed increase in autophagic flux. In addition to these changes the statistically significant decrease in PI4P may also contribute to changes in autophagy. A reduction of PI4P has been postulated to be necessary for lysosome reformation (Rong et al., 2012; Yu et al., 2010).

The elevation of PI3P, PI(3,5)P2 and PI5P may also have a role in compensating potential mutant huntingtin-dependent changes in PI or masking of selected PI. Several studies have indicated that there are polyglutamine-dependent alterations in PI binding of huntingtin protein (Burke et al., 2013; Kegel et al., 2009a; Kegel et al., 2009b; Kegel et al., 2005). Moreover, wild-type huntingtin binds phosphoinositide lipids including PI5P and PI(3,5)P2 (Kegel et al., 2009b). Notably, when assessed using unilamellar vesicles, huntingtin with a polyglutamine expansion bound these lipids even more tightly than wild-type hungtingtin, potentially reducing the free total levels of these lipids and impacting their downstream dependent signaling (Kegel-Gleason, 2013). Masking of PI lipids could negatively and progressively impact the function of proteins involved in autophagosome cargo recognition and loading, especially those effector proteins dependent on low abundance PI, such as PI5P and PI(3,5)P2. Additional studies need to be carried out to determine the effectors proteins (Alfy and/or others) responsible for the action of PIP4Kγ modulation, the mechanism of action behind the high cellular alteration in PI(3,5)P2 as well as the modulation of other PI levels, and the impact of those changes on mTOR function and autophagy dynamics (Ikonomov et al., 2006; Jin et al., 2014; Ikonomov et al., 2001). It has not escaped our attention that mHtt-dependent effects seem to be triggered by aging, which is known to limit the clearance of misfolded proteins (Komatsu et al., 2007), and deregulate phosphatidylinositide lipid signaling (Igwe and Filla, 1995).

The data presented in this manuscript demonstrates that pharmacological inhibition, or knock-down of PIP4Kγ produce a similar reduction in huntingtin levels, and a concomitant elevation of PI5P and PI(3,5)P2 and PI3P. These findings open the door to a new disease-modifying approach for this disorder and validate PIP4Kγ as a druggable target. In a recent report, a homozygous mouse knockout displayed no growth or behavioral abnormalities (Shim et al., 2016). From the translational point of view, the development of selective PIP4Kγ inhibitors could be extraordinarily useful for other neurodegenerative diseases as well. Alzheimer’s disease and Parkinson’s disease in particular are also mediated by the accumulation of toxic protein aggregates, whose catabolism by autophagy might rescue stressed neurons. Starvation increases life spans across species (Speakman and Hambly, 2007) and there are numerous diseases where upregulation of basal autophagy is beneficial (Hara et al., 2006; Seino et al., 2013). Further, dose response studies are necessary to fully evaluate the therapeutic potential of PIP4Kγ inhibition.

Materials and methods

Synthesis of NCT-504

Request a detailed protocol

General Experimental Procedure: Unless otherwise stated, all reactions were carried out under an atmosphere of dry argon or nitrogen in dried glassware. Indicated reaction temperatures refer to those of the reaction bath, while room temperature is noted as ~25°C. All anhydrous solvents, commercially available starting materials, and reagents were purchased from Aldrich Chemical Co. and used as received. Chromatography on silica gel was performed using forced flow (liquid) of the indicated solvent system on Biotage KP-Sil pre-packed cartridges and using the Biotage SP-1 automated chromatography system.

1H spectra were recorded on a Varian Inova 400 MHz spectrometer. Chemical shifts are reported in ppm with the solvent resonance as the internal standard (DMSO-d6 2.50 ppm, for 1H). Data are reported as follows: chemical shift, multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, br s = broad singlet, m = multiplet), coupling constants, and number of protons.

Analytical purity analysis and retention times (RT) reported here were performed on an Agilent LC/MS (Agilent Technologies, Santa Clara, CA). A Phenomenex Luna C18 column (three micron, 3 × 75 mm) was used at a temperature of 50°C. The solvent gradients are mentioned for each compound and consist of a percentage of acetonitrile (containing 0.025% trifluoroacetic acid) in water (containing 0.05% trifluoroacetic acid). A 4.5 min run time at a flow rate of 1 mL/min was used.

Mass determination was performed using an Agilent 6130 mass spectrometer with electrospray ionization in the positive mode.

Synthetic scheme to prepare NCT-504:

Scheme 1
Download asset Open asset
Synthetic scheme to prepare NCT-504.
https://doi.org/10.7554/eLife.29123.023

Synthetic Procedures:

B: A (5-bromo-4-chlorothieno[2,3-d]pyrimidine) was prepared according to WO2012/44993 A1, 2012; Location in patent: Page/Page column 45). A solution of A (1.03 g, 4.13 mmol) in THF (15 ml) was treated at 0°C under nitrogen with dropwise addition of isopropylmagnesium chloride (2.48 ml, 4.95 mmol, 2M in THF). The mixture was stirred for 15 min and then a solution of iodine (1.05 g, 4.13 mmol) in THF (10 mL) was added dropwise under nitrogen. The mixture was stirred at 0°C for almost 2 hr, quenched with saturated aqueous NH4Cl and then EtOAc was added. The mixture was stirred, the organic layer was separated, washed with saturated aqueous Na2S2O3, dried with MgSO4, filtered, concentrated to obtain crude 4-chloro-5-iodothieno[2,3-d]pyrimidine (1.17 g, 3.95 mmol, 96% yield). This appeared to be contaminated with a small amount of B 4-chlorothieno[2,3-d]pyrimidine (approximately ~5–10% by LC/MS).

1H NMR (400 MHz, DMSO-d6) δ 8.96 (s, 1H), 8.46 (s, 1H).

LC/MS Gradient 4% to 100% Acetonitrile (0.05% TFA) over 3.0 min; RT 3.290 min, ESI (M + 1)+calculated 296.9, found 296.8.

C: A microwave vial filled was charged with 4-chloro-5-iodothieno[2,3-d]pyrimidine B (0.48 g, 1.62 mmol), (3-(methylsulfonyl)phenyl)boronic acid (0.389 g, 1.94 mmol), Pd(PPh3)4 (0.094 g, 0.081 mmol), sodium carbonate (1.42 ml, 2.83 mmol) followed by dimethoxyethane (8 mL) and water (1 mL). The mixture was heated in the microwave under ‘high’ settings at 120°C for 20 min in the microwave. The mixture was then cooled; celite was added, and concentrated. The adsorbed material was purified by flash silica gel chromatography with a gradient of 0% to 30% EtOAc in DCM that separated unreacted starting iodide (~20% recovery) from the required product 4-chloro-5-(3-(methylsulfonyl)phenyl)thieno[2,3-d]pyrimidine C (0.19 g, 0.59 mmol, 36% yield).

1H NMR (400 MHz, DMSO-d6) δ 9.02 (d, J = 0.4 Hz, 1H), 8.19 (d, J = 0.4 Hz, 1H), 8.08 (td, J = 1.8, 0.5 Hz, 1H), 8.02 (ddd, J = 7.8, 1.9, 1.1 Hz, 1H), 7.90 (ddd, J = 7.7, 1.7, 1.2 Hz, 1H), 7.77 (td, J = 7.7, 0.5 Hz, 1H), 3.29 (s, 3H). LC/MS Gradient 4% to 100% Acetonitrile (0.05% TFA) over 3.0 min, RT 3.014 min, ESI (M + 1)+ calculated 325.0, found 324.9.

NCT-504 4-Chloro-5-(3-(methylsulfonyl)phenyl)thieno[2,3-d]pyrimidine C (0.18 g, 0.55 mmol) with DME (10 mL) was treated with 1-methyl-1H-tetrazole-5-thiol (0.084 g, 0.720 mmol) and Hunig's Base (0.194 mL, 1.11 mmol), and heated at 120°C for 30 min in a sealed tube. The mixture was cooled, concentrated, re-dissolved in minimal DCM and the purified by silica gel column chromatography (5% to 60% EtOAc/DCM) to provide NCT-504 4-((1-methyl-1H-tetrazol-5-yl)thio)−5-(3-(methylsulfonyl)phenyl)thieno[2,3-d]pyrimidine (190 mg, 0.470 mmol, 85% yield).

1H NMR (400 MHz, DMSO-d6) δ 8.76 (d, J = 0.4 Hz, 1H), 8.23 (td, J = 1.8, 0.5 Hz, 1H), 8.15 (d, J = 0.4 Hz, 1H), 8.10 (ddd, J = 7.8, 1.9, 1.1 Hz, 1H), 8.02 (ddd, J = 7.7, 1.7, 1.1 Hz, 1H), 7.86 (td, J = 7.8, 0.6 Hz, 1H), 3.96 (s, 3H), 3.34 (s, 3H).

LC/MS Gradient 4% to 100% Acetonitrile (0.05% TFA) over 3.0 min, RT 3.024 min, ESI (M + 1)+ calculated 405.0, found 405.0.

Enzyme preparation and biochemical assays

Request a detailed protocol

Protein from human PIP4K2A (UniGene 138363), PIP4K2B (Unigene 269308) and PIP4K2C (UniGene 6280511) was expressed in pGEX6P (GE Healthcare) and purified from E. coli BL21(DE3). GST fusion proteins from cell lysates were bound to glutathione sepharose beads (GE Healthcare) and cleaved in situ with 50U of PreScission protease (GE Healthcare) for 4 hr at 4°C.

Lipid kinase assays were performed essentially as described previously (Wang et al., 2010; Clarke et al., 2001). In brief, dried substrate lipid (6 μM PI5P final reaction concentration) was resuspended in kinase buffer (50 mM Tris pH 7.4, 10 mM MgCl2, 80 mM KCl, and 2 mM EGTA) and micelles were formed by sonication for 2 min. Recombinant lipid kinase, preincubated with inhibitor for 10 min on ice (where required), was added to the micelles and the reaction started by the addition of 10 μCi [32P]ATP (200 µl final volume), and incubated at 30°C for 10–60 min (dependent on isoform). Lipids were extracted using an acidic phase-separation (Bligh and Dyer, 1959) and separated by one-dimensional thin layer chromatography (2.8:4:1:0.6 chloroform:methanol:water:ammonia). Radiolabelled PI(4,5)P2 product was detected by autoradiography, extracted from the plate and Cerenkov radiation was counted in the presence of Ultima Gold XR scintillant (Packard) on a LS6500 scintillation counter (Beckman Coulter). Specific enzyme activities, under these assay conditions, were calculated as nmoles of PI5P converted into PI(4,5)P2 per minute per mg of purified recombinant enzyme.

Intrinsic ATPase activities of the enzymes were determined using the Transcreener ADP2 fluorescence polarization assay (BellBrook Labs). PIP4Kγ (1 μM, [Clarke and Irvine, 2013]) was pre-incubated (10 min on ice) at range of inhibitor concentrations and assayed with ATP substrate (100 µM ATP, 60 min incubation at 22°C) in the absence of lipid substrate. Polarization units (mP) were read using a PHERAstar Plus microplate reader (BMG Labtech). Experimental values were interpolated from an ADP/ATP utilization standard curve and plotted using nonlinear regression analysis with Prism 5 (GraphPad).

Measurement of phosphorylated phosphoinositide (PI) levels by HPLC

Request a detailed protocol

PI measurements were performed as previously described (Zolov et al., 2012). Briefly, mouse primary fibroblasts were generated from P1 pups (129P2/OlaHsd × C57BL/6) and were cultured in DMEM supplemented with 15% FBS and 1X Pen-Strep-Glutamine and human patient fibroblast were cultured in MEM supplemented with 15% FBS, 1x Pen-Strep and 1x Glutamax in 100 mm dishes to 60–70% confluence. MEF cells and patient fibroblasts were tested using MycoFluor Mycoplasma Detection Kit (Thermo Scientific Fisher) and were negative for mycoplasma contamination. Cells were washed with PBS and incubated with inositol labeling medium (containing custom-made inositol-free DMEM (11964092; Life Technologies), 10 μCi/mL of myo-3H-inositol (GE Healthcare), 10% dialyzed FBS (26400; Life Technologies), 20 mM Hepes, pH 7.2–7.4, 5 μg/mL transferrin (0030124SA; Invitrogen), and 5 μg/mL insulin (12585–014; Invitrogen) for 48 hr. For experiments with NCT-504 treatments, cells were treated with indicated concentrations of NCT-504 or DMSO for indicated duration before the end of the labeling. Extraction and HPLC measurements were performed as described (Zolov et al., 2012).

Silencing of PIP4Kγ

Request a detailed protocol

Primary mouse embryonic fibroblast cells generated from P1 pups (129P2/OlaHsd × C57BL/6) were infected with MISSION shRNA lentiviral plasmid pLKO.1-puro with shRNA target sequence CTCCAAGATCAAGGTCAACAA (TRCN0000024702; Sigma) containing 237–257 nucleotides of mouse PIP4Kγ cDNA; MISSION nontarget shRNA lentiviral control vector SHC002 (Sigma) was used as control. Transduction-ready viral particles were produced by the Vector Core (University of Michigan, Ann Arbor, MI) with a concentration of 107 transduction units per ml. Mouse primary fibroblast grown on two 35 mm dishes were treated at an MOI of 5. After overnight incubation, cells were treated with 2 μg/ml puromycin. After two days of infection, cells from two 35 mm dishes were transferred to a 100 mm dish and maintained in puromycin containing media for another three days. Cells were either analyzed by western blot or incubated with inositol labeling medium for 48 hr for PI measurements. Immunoblots were performed with antibodies against PIP4K2C (17077–1-AP RRID: AB_2715526, ProteinTech; 1:5000) and GADPH (AM4300 RRID: AB_437392, Thermo Scientific Fisher; 1:50000).

LC3 measurements in HEK cells

Request a detailed protocol

HEK 293T cells grown on 35 mm Dishes till 60–70% confluency were either untreated or treated with DMSO or NCT-504 with or without 100 nM Bafilomycin for two hours. Cells were lysed and immunoblotted with antibodies against LC3A/B (12741 RRID:AB_2617131; Cell Signaling) and α-tubulin (A-11126 A11126 RRID:AB_221538; Life Technologies). Blots were analyzed using Adobe Photoshop. HEK293T cells were purchased from ATCC (RRID:CVCL_0063) and were certified authentic and mycoplasma free.

Htt HTRF assay

Antibodies

The monoclonal antibodies used in the HTRF assay were 2B7 (gift from collaborator) which binds to the first 17 amino acids of normal and mutant Htt, and MAB2166 (EMD Milipore #MAB2166), which binds to an Htt epitope (amino acid 181 to 810), and recognizes both normal and mtHtt. The antibody 2B7 was conjugated to Tb as a donor, and 2166 was conjugated to d2 as an acceptor (both were custom labeled by Cisbio). The labeled antibody pairs were diluted in the 1X HTRF assay buffer: 50 mM NaH2PO4, 400 mM NaF, 0.1%BSA, 0.05% Tween 20. The HTRF assay were performed at 1536-well plate. For the experiments, cells were seeding (6 μL/well) 24 hr in advanced and culture at 37°C 5%CO2 followed by compound addition (23 nL). After incubating with compounds for 24 hr, cells were lysed by adding 2 μL of 4x lysis buffer (Cisbio Lysis buffer #2), incubated at room temperature for 2 hr then add labeled antibodies. The labeled antibody pairs were diluted in the 1X HTRF assay buffer: 50 mM NaH2PO, 400 mM NaF, 0.1%BSA, 0.05% Tween 20. The final reaction is 8 μL/well. The signal ratio between 665 nm and 615 nm have been calculated as the raw HTRF ratio. The cell viability was measured by using CellTiter-Glo Luminescent Cell Viability Assay (Promega). The 293A cells were purchased from ThermoFisher Scientific, Cat#R70507, Lot # 1657360. They tested negative for mycoplasma. They were not sent out for STR since it was first use right after purchase from company.

Htt exon1 aggregation assay

Request a detailed protocol

GFP-Htt-exon1-Q23 (Plasmid #40261)1 and GFP-Htt-exon1-Q74 (Plasmid #40262)1 were purchased from Addgene (Cambridge, MA) (Narain et al., 1999). Immortalized Atg7+/+ and Atg7-/- MEF cells were generously provided by Dr. Masaaki Komastu (School of medicine, Niigata University) (Komatsu et al., 2005). Atg7+/+ and Atg7-/- MEF cells were tested using MycoFluor Mycoplasma Detection Kit (Thermo Scientific Fisher) and are negative for mycoplasma contamination. They were validated for the presence and absence, respectively, of Atg7 by the western blot shown in Figure 3C. These cell lines are not included in the list of commonly misidentified cell lines maintained by International Cell Line Authentication Committee were not authenticated further. HEK293T, Atg7+/+ and Atg7-/- cells grown on coverslips were transfected with either GFP-Htt-exon1-Q23 or GFP-Htt-exon1-Q74 using Lipofectamine 2000 (Invitrogen). After 2 hr of transfection, cells were incubated with DMSO or 2 μM of NCT-504 for 48 hr and fixed. Transfected cells with mHtt aggregates were quantified (Narain et al., 1999; Komatsu et al., 2005).

HTT quantification in fibroblasts and StHdh cells

Request a detailed protocol

Immortalized StHdhQ111 (Coriell-CH00095, RRID:CVCL_M591) cells (Trettel et al., 2000), immortalized wild type (Coriell-GM02153) and HDQ45 (Coriell-GM03868, RRID:CVCL_1H73) HD fibroblasts ((using SV40 large T antigen) (Lu et al., 2013) and non-immortalized HDQ68 (Coriell-GM21757, RRID:CVCL_1J85) were grown in 15%FBS DMEM with GlutaMax (Life Technologies). For drug treatments, cells were plated overnight until they reached 70% confluence in 12-well plates, drug was added at the desired concentration for 48 hr. For siRNA treatment cells were nucleofected using Amaxa at a final concentration of 30 nM and grown in 6-well plates for 72 hr. StHdh cells were grown in DMEM (Life Technologies) 10% FBS and drug treatment was carried out as described above. Cell identity was confirmed using STR profiling (GenePrint 10 System from Promega Corp.) and tested mycoplasma negative (Hoechst staining).

Cells were collected using trypsin, homogenized in RIPA buffer, sonicated and incubated in ice for 30 min. Supernatant was collected after a 10 min centrifugation and protein concentration was measured. For western blot analysis 15 μg of each protein sample was loaded in a 4–12% Bis-tris gel, transferred into a nitrocellulose membrane, blocked with 5% milk and incubated overnight with anti-Huntingtin antibody MAB5492 (Millipore) for fibroblasts or MAB2166 (Millipore) for StHdhQ111 cells.

Ethical treatment of animals

Request a detailed protocol

All vertebrate animal work was approved by the Institutional Animal Use and Care Committee at the University of Michigan (PRO00007096). Experiments were carefully planned to minimize the number of animals needed. Pregnant female wild-type, non-transgenic Long Evans rats (Rattus norvegicus) were housed singly in chambers equipped with environmental enrichment. They were fed ad libitum a full diet (30% protein, 13% fat, 57% carbohydrate; full information available at www.labdiet.com), and cared for by the Unit for Laboratory Animal Medicine (ULAM) at the University of Michigan. Veterinary specialists and technicians in ULAM are trained and approved in the care and long-term maintenance of rodent colonies, in accordance with the NIH-supported Guide for the Care and Use of Laboratory Animals. All rats were kept in routine housing for as little time as possible prior to euthanasia and dissection, minimizing any pain and/or discomfort. Pregnant dams were euthanized by CO2 inhalation at gestation day 20. For each animal, euthanasia was confirmed by bilateral pneumothorax. Euthanasia was fully consistent with the recommendations of the Guidelines on Euthanasia of the American Veterinary Medical Association and the University of Michigan Methods of Euthanasia by Species Guidelines. Following euthanasia, the fetuses were removed in a sterile manner from the uterus and decapitated. Primary cells from these fetuses were dissected and cultured immediately afterwards.

Rodent primary neuron isolation and culturing

Request a detailed protocol

Primary mixed cortical neurons were dissected from these embryos as described previously (Saudou et al., 1998), and plated in a poly-l-lysine/laminin coated 96 well plate at a density of 5 × 105 cells/ml. On day four in vitro, cells were transfected with Dendra2-LC3 with or without GFP-Beclin using Lipofectamine 2000 (Invitrogen). Thirty minutes post-transfection, neurons were treated with NCT-504 or DMSO. Optical pulse labeling experiments were performed as previously described (Tsvetkov et al., 2013; Gupta et al., 2017; Barmada et al., 2014). Briefly, Dendra2-LC3 was photoconverted 24 hr post-transfection by illuminating each imaging field with a 250 ms pulse of 405 nm light. Following photoconversion, neurons were longitudinally imaged using a custom-built automated fluorescence microscopy platform (Arrasate et al., 2004; Barmada et al., 2014; Barmada et al., 2015). A Nikon Eclipse Ti inverted microscope equipped with a high-NA 20X objective lens, a PerfectFocus3 system, and an Andor iXon3 897 EMCCD camera were used for image acquisition. GFP and TRITC images were taken immediately after photoconversion and four more times within the following 48 hr. Single-cell TRITC intensity values were fitted to a first-order exponential decay curve, generating a half-life value for each individual neuron. Neuronal survival analysis was assessed using original software written in Python. Only cells that lived the duration of imaging were included in the Dendra2-LC3 half-life analysis. Half-life was determined by fitting the TRITC intensity values at each time point to a first-order exponential function using scripts written in R. Comparisons between groups to determine statistical significance were accomplished using one-way ANOVA with Dunnet’s post hoc test and the Kruskal-Wallis test.

Drosophila experiments

Request a detailed protocol

Two different Drosophila HTT-expressing strains were used for this study, and N-terminal model expressing the first 336 amino acids of human HTT with a 128Q expansion (Branco et al., 2008) and a full length model expressing human HTT with a Q200 expansion (El-Daher et al., 2015). For retinal expression, we used the GMR-GAL4 driver at 25C and for panneuronal expression, we used the elav-GAL4 driver. These two drivers as well as the siRNAs targeting dPIP4K were obtained from the Bloomington Drosophila stock center. The dPIP4K-29 loss of function allele and the K271D kinase dead (PIP4K-DN) allele were previously described and kindly provided by Dr. Padinjat Raghu (Gupta et al., 2013).

For the retinal degeneration assay, animals were fixed with 4% formaldehyde in PBS. Heads were dehydrated in increasing concentrations of ethanol and embedded in paraffin. Ten μm serial sections were obtained and re-hydrated to PBS. Sections were stained with hematoxylin (SIGMA). Images were captured using an AxioCam MRc camera (ZEISS) attached to a MICROPHOT-FXA microscope (Nikon).

Motor performance of animals was assessed as a function of age. For the N-terminal model 15 age-matched virgin females per replica were used. Animals are taped to the bottom of a plastic vial and the number of animals reaching a height of 9 cm in 15 s is assessed using infrared sensors. Ten trials are carried out for each day represented. The plotted data corresponds to the average percentage of animals reaching 9 cm. Data was analyzed by ANOVA followed by Dunnet’s post hoc test. For the FL-HTTQ200 a similar procedure was used, the animals were video recorded and data was processed using a custom designed analysis software (Source code 1), which allowed for calculating speed.

References

    1. Emerling B
    2. Sasaki A
    3. Cantley LC
    4. Hurov J
    (2014)
    Modulation of Phosphatidylinositol-5-Phosphate-4-Kinase Activity
    Modulation of Phosphatidylinositol-5-Phosphate-4-Kinase Activity.
    1. Kegel KB
    2. Kim M
    3. Sapp E
    4. McIntyre C
    5. Castaño JG
    6. Aronin N
    7. DiFiglia M
    (2000)
    Huntingtin expression stimulates endosomal-lysosomal activity, endosome tubulation, and autophagy
    Journal of Neuroscience 20:7268–7278.
    1. Lietha D
    (2011)
    eLS
    Phosphoinositides – The Seven Species: Conversion and Cellular Roles, eLS.
    1. Speakman JR
    2. Hambly C
    (2007)
    Starving for life: what animal studies can and cannot tell us about the use of caloric restriction to prolong human lifespan
    The Journal of Nutrition 137:1078–1086.
    1. Titus S
    (2010)
    Probe Reports From the NIH Molecular Libraries Program
    Identification of compounds which inhibit cytotoxicity associated with mutant Huntingtin protein expression, Probe Reports From the NIH Molecular Libraries Program, Bethesda, National Center for Biotechnology Information..
    1. Titus SA
    2. Southall N
    3. Marugan J
    4. Austin CP
    5. Zheng W
    (2012)
    High-throughput multiplexed quantitation of protein aggregation and cytotoxicity in a Huntington's disease model
    Current Chemical Genomics 6:79–86.

Decision letter

  1. Harry T Orr
    Reviewing Editor; University of Minnesota, United States

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

[Editors’ note: a previous version of this study was rejected after peer review, but the authors submitted for reconsideration. The first decision letter after peer review is shown below.]

Thank you for submitting your work entitled "Inhibition of PIP4Kγ ameliorates the pathological effects of mutant huntingtin protein" for consideration by eLife. Your article has been reviewed by three peer reviewers, one of whom, Harry T Orr (Reviewer #1), is a member of our Board of Reviewing Editors, and the evaluation has been overseen by a Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Leslie Thompson (Reviewer #3).

Our decision has been reached after consultation between the reviewers. Based on these discussions and the individual reviews below, we regret to inform you that your work will not be considered further for publication in eLife.

Note that while reviews found considerable merit in your study, there are major limitations that need to be addressed in order for eLife to consider your paper in the future. First, it is important that these studies be performed using a full-length Htt fly model. Second, it is also critical that your analysis include a dosing regime of the PIP4K compound. Should you choose to perform these studies and submit the new manuscript in the future, we will attempt to get some of the same reviewers.

Reviewer #1:

In this study the investigators present a rather extensive body of work (and concise) in support of the concept that inhibition of the lipid kinase PIP4Kγ reduces levels of huntingtin and mitigates phenotypic effects of mutant Htt in a fly model, presumably via induction of autophagy. Building on their previous work, a novel PIP4K inhibitor and its ability to enhance autophagy in cells and reduce Htt levels in neurons are presented. Importantly they go on to show that in a fly model of mutant Htt pathology reduction of PIP4Kγ mitigates two disease-related phenotypes; retinal degeneration and a motor performance deficit. All in all the data are solid and present a strong basis for exploring further the therapeutic potential of targeting PIPK4γ.

Reviewer #2:

This manuscript is describing the inhibition of PIP4Kγ as a potential therapeutic target in Huntington's disease by the triggering of autophagy to clear huntingtin protein, in what they refer to as a druggable target.

This small molecule lead, NCT-504, is important because it appears to trigger the degradation of huntingtin, and thus falls into the hypotheses of therapies for HD designed to _specifically_ reduce or eliminate huntingtin protein in Huntington's disease, which include AAV siRNAs, and anti-sense oligonucleotides, which are either pending or are in clinical trials currently. The concept behind this study was driven by mouse exon1 model work that implies that increasing the rate of autophagy may be beneficial for neurodegeneration in HD.

The manuscript has some solid data, but is confusing to follow in that they go back and forth from small fragment over expression models, in which 3% of the mutant huntingtin ORF with synthetic allele is over expressed at massive levels to induce the formation of aggregates, to data from human HD fibroblasts, which is arguably the most exciting data in this work. The only assay of pathological effects comes from an outdated fly fragment model.

The overall picture seems exciting but the details in the data are not as clear as the writing would suggest. In many of the figures and assays, they only look at mutant HD cells, and not fibroblasts from unaffected individuals, and most of the readouts are only looking at one protein, while the proposed mechanism is to switch on autophagy in general. Autophagy, and its sub-categories of mitophagy, micro-autophagy, macro-autophagy, pexophagy are all dynamic system in a high rate of constant flux in the cell. If this is truly a drug lead for Huntington's disease, or any disease, they need to show some selectivity in the effected state and in the target. It seems that even in their loading controls, there is a massive global effect of shutting down PIP4Kγ either by SiRNA or global inhibitor. In other words, autophagy in all its forms is in dynamic flux for a reason, and just pushing it in one direction may be as problematic and inhibiting it.

I have an conceptual concern that this data suggest that 97% of the huntingtin ORF is irrelevant for regulating the turnover of this protein by autophagic mechanisms, and it appears from this data that the molar stoichiometry is irrelevant, as whether they look at endogenous huntingtin proteins, or massively over-expressed small molecule fragments at orders of magnitude above physiological molarity, the compound works just as well to clear 'huntingtin', and likely everything else.

My major concern is that I believe they have proven that PIP4K γ inhibition triggers a massive level of autophagy, but they have no data here that show any selectivity or health effects on the fibroblasts. Is this "curing dandruff by decapitation"? They reference a manuscript on a PIP4Kγ knockout mouse stating the investigators observed "no growth or behavioral abnormalities", except the manuscript actually describes a massive global hyperactivation of the immune systems in those knockouts, and the true validation of a drug target would be a conditional knockout. For what it's worth, Broad's Exac exomic sequencing database shows no homozygous null alleles of PIP4Kγ yet in humans. Thus, I have concerns that PIP4K γ inhibition could really be a drug target.

They do not state exactly how their HD fibroblasts are immortalized, but is seems that there is no difference in effect between HEK293, STHdHQ111, HD primary cells, or HD immortalized cells. This is rather surprising since the only known inducer of huntingtin expression is P53, and some of those lines have P53, and others do not. This has been a concern that has been ignored in most HD cell lines studies, despite over 80 manuscripts published consistently defining disrupted P53 biology in HD, from mouse to humans. So whether they have p53 or not, whether small fragments of full protein, everything gets cleared, in all lines.

Major concerns

Figure 1

From the high throughput screen, they found NCT-504 and by the structure guessed that it was a kinase inhibitor, and tested this by kinomic screening, and used a cutoff of >65% inhibition to claim specificity. This is a very high cutoff, and the entire kinome screening data should be included a supplemental, not just 3/442 kinases. From my experience with pharma, even 20% inhibition is cause for concern on these screens. The IC50 is about 16μM for NCT-504. This is very high, and by their own data, the IC50 exceeds the toxicity threshold concentration in neurons.

The concentrations of NCT-504 used for effects vary hugely across the figures. In Figure 1, and from text, IC50 is quoted at 15.8μM. In Figure 1—figure supplement 1, 2μM is used to reduce aggregation by 50% effect, so whatever reason they see effect, it seems to be unlikely through PIP4Kγ inhibition.

Figure 2: concentration of NCT-504 is now 10μM, less than the IC50, and only one dose tested.

The loading controls seem to switch between α-tubulin, actin and GAPDH. For these studies looking at autophagic flux, as in Figure 4 they need better controls, not an insoluble polymer of actin. I'd like to see GAPDH and at least two inner mitochondrial proteins to show me the cells are not just completely autophagic or apoptotic. Images are essential.

Again, for Figure, 4, now 2μM of NCT-504 is used again in Q68 cells, but then 10μM in Q45 cells. This is plus or minus 500%.

Figure 4: they need to show the whole blot in a supplemental figure. I see a severe hit on actin levels in Figure 4B with PIP4K2C_si, and in Figure 4D with 10μM NCT-504, which concerns me that these cells might be heavily auto-degrading or just dead.

In Figure 4—figure supplement 4, they show a very unconventional assay for neuronal toxicity, in that they are no showing any images of neurons or other measure of health, or even life, but their own data shows a pretty severe "toxicity" at 10μM NCT-504, the exact concentration used in HD fibroblasts, and still below their calculated IC50.

Figure 4—figure supplement 4B, that western blot is not appropriate data to conclude anything.

Figure 5. They now switched back to a severe allele Q128 (extremely rare) and only small fragment 1-231. Why did they not use full huntingtin?

Reviewer #3:

This is a very well written manuscript that describes strong data showing that PIP4K γ inhibition, either pharmacologically or through KD, causes reduction of levels of soluble HTT. In cases where aggregation or formation of HMW species are observed, PIP4K reduction also reduces these aggregated species as well. The authors used several different cell lines ranging from immortalized lines to primary neurons or human fibroblasts (N2a, stHDHq111, human fibroblasts, primary neurons). Experiments are elegant, well described and well controlled, with helpful supplemental data, such as the KD efficiency of the RNAis and kinetics of the phosphatidylinositide changes with treatment. There are a few minor concerns:

1) How was NCT-504 selected for these studies from published the high-throughput screen? Was this best hit? Because of the complexity of these pathways, it would be helpful to have a schematic early on to show the various components of the pathway, how each of the PIP derivatives are generated and where this kinase fits into general scheme.

2) Figure 4A - Why not show more complete set of westerns as in C and D? Need quantification for Figure 4.

3) Do the fibroblasts have HMW species? This is not mentioned and if they are present, why they are not included should be addressed.

4) A potentially important experiment would be using a pulse chase or other such assay to measure whether this truly increases degradation of the protein.

5) Please describe length of HTT fragment in the 128Q HTT transgene in flies.

6) A citation should be included for the Neuron paper from the Davidson lab when discussing Rhes and Rheb.

7) It is a stretch to say that compound modulates HTT proteolysis as there is a single Western blot with one panel.

[Editors’ note: what now follows is the decision letter after the authors submitted for further consideration.]

Thank you for submitting your article "Inhibition of PIP4Kγ ameliorates the pathological effects of mutant huntingtin protein" for consideration by eLife. Your article has been reviewed by three peer reviewers, one of whom is a member of our Board of Reviewing Editors, and the evaluation has been overseen by a Senior Editor. The reviewers have opted to remain anonymous.

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary

In this manuscript the ability of a small molecule to inhibit PIP4Kγ and impact HD-like phenotypes in cellular and animal models expressing mutant Htt (full-length and fragment) are presented. While one reviewer is concerned with the extent to which these findings might apply to a mammalian model of HD, it was noted by other reviewers that their use of primary cortical neurons from mice provides strong evidence of the biological relevance of this compound and pathway beyond its heavy use of fly HD models.

Essential revisions

It is critical that the authors pay close attention to and respond accordingly to reviewers 1 and 2 concerns that the strength of this manuscript is decreased considerably by the style/manner in which it is written. Extensive editing to the text and figures are needed so that the manuscript flows smoother and rationale for the studies are clearer.

Reviewer #1:

In this new submission of a previously submitted manuscript the ability of a small molecule to inhibit PIP4Kγ and impact HD-like phenotypes in cellular and animal models expressing mutant Htt (full-length and fragment) is presented. Overall while this work addresses a very interesting issue and additional data using a FL-Htt fly model have been added, it remains difficult to read/follow in several key places.

Specific points:

Overall the manuscript remains hard to read, needing English editing in many places.

In several places statements are not referenced when they should be. For example, second paragraph of Results – "Similar discrepancies in potency […]" needs to be referenced.

Clarity of Figure 3 would be improved by including concentrations of compounds used within body of the figure and not just in the legend.

Reviewer #2:

This is a revised manuscript and the authors were very responsive to the reviewers.

The work described identifies PIP4Kγ as a therapeutic target for HD and shows that pharmacological inhibition of the enzyme promotes mutant HTT degradation and autophagy. They provide in vivo evidence that lower the levels of PIP4Kγ is beneficial in a fly model of HD-retinal assays and motor performance. This is a promising direction to pursue as a possible therapy for HD and is in line with current treatments directed at lowering mHTT protein levels. Generally the quality of the data is strong. Minor points are that there are a number of formatting typos. Also the figures are not very consistent. Formatting is varied-different fonts, different graphing and repetitive information in legend and graphs.

Reviewer #3:

This manuscript describes the effect of inhibition of the PIP4Kγ in models of Huntington's disease. There is extensive description of the effect of the drug on lipid metabolism and autophagy. There are also data from cell models including HD fibroblasts. The reduction of Htt levels is of interest. However effects on Htt aggregates are difficult to interpret since aggregates may be toxic or protective. There is also an interesting effect on an HD Drosophila model. It would be beneficial if the investigators had some more substantial data on models more directly relevant to HD pathogenesis. For instance instead of using HD fibroblasts it would be beneficial to use HD iPS cell models. In addition, there are no data from HD mouse models. The processes of autophagy are cell type specific, and may be different in Drosophila and mammalian systems, so in vivo data from mice rather than flies would be advantageous.

https://doi.org/10.7554/eLife.29123.027

Author response

[Editors’ note: the author responses to the first round of peer review follow.]

[…] Note that while reviews found considerable merit in your study, there are major limitations that need to be addressed in order for eLife to consider your paper in the future. First, it is important that these studies be performed using a full-length Htt fly model.

We have included now results evaluating the effect of PIP4K loss of function in a Drosophila model expressing full length Htt (new panels in Figure 5C). As shown decreasing dPIP4K(CG17471) by either a heterozygous mutant (PIP4K-dCG17471LOF) or by expression of a dominant negative allele (PIP4K-dCG17471DN) leads to a suppression of the HTT-FLQ200-induced motor impairments. We used two different parameters to measure motor performance: 1- a climbing test and 2- average climbing speed.

Second, it is also critical that your analysis include a dosing regime of the PIP4K compound.

In our original manuscript we included several concentration-response experiments showing the effect of our PIP4Kγ inhibitor at different dose ranges that now we have reinforced with additional experiments:

· The kinase inhibitor IC50 of NCT-504 was reported using two different methods: binding assay (DiscoverX KINOMEscan, Figure 1D) and lipid phosphorylation using full length isolated PIP4Kγ (Figure 1E).

· We also reported the full concentration-response curve of NCT-504 inhibiting mHtt cellular accumulation in stably transfected PC12 cells with a Q103 mHtt fragment, fused to GFP (Figure 1C).

· In addition, we have also included a dose response for autophagosome formation and autophagy flux using a stable reporter HEK293T cell line transfected with LC3-GFP-mCherry.

· The same cell line HEK293T was used with 2-different concentrations of NCT-504 in the presence and absence of bafilomycin A1 to evaluate LC3 levels.

· We have also evaluated the effect of NCT-504 at different doses on autophagy flux in rat primary neurons carrying LC3-Dendra2, a photoswitchable protein able to report LC3 turnover in individual neurons. With this technique previous authors have shown that the proteostasis of Htt-poly-Q varies among neurons and predicts neurodegeneration.

· We have also performed titration curves for NCT-504 in the Q45 patient fibroblasts. As shown in Figure 4—figure supplement 1, NCT-504 is not toxic until we reach the 10μM concentration. The dose response experiments on mHTT protein levels revealed that there was a statistically significant decrease of mHTT levels starting at 5μM (new representative western blot and quantification now included in Figure 4C).

· The dose-dependent effect of NCT-504 in phosphatidylinositols levels has been also included using human fibroblast.

Reviewer #2:

This manuscript is describing the inhibition of PIP4Kγ as a potential therapeutic target in Huntington's disease by the triggering of autophagy to clear huntingtin protein, in what they refer to as a druggable target.

This small molecule lead, NCT-504, is important because it appears to trigger the degradation of huntingtin, and thus falls into the hypotheses of therapies for HD designed to _specifically_ reduce or eliminate huntingtin protein in Huntington's disease, which include AAV siRNAs, and anti-sense oligonucleotides, which are either pending or are in clinical trials currently. The concept behind this study was driven by mouse exon1 model work that implies that increasing the rate of autophagy may be beneficial for neurodegeneration in HD.

We have included additional references to provide more context around this line of research.

· Sarkar S1, Perlstein EO, Imarisio S, Pineau S, Cordenier A, Maglathlin RL, Webster JA, Lewis TA, O'Kane CJ, Schreiber SL, Rubinsztein DC., Small molecules enhance autophagy and reduce toxicity in Huntington's disease models, Nat Chem Biol. 2007 Jun;3(6):331-8. Epub 2007 May 7.

· Lin F1, Qin ZH1, Degradation of misfolded proteins by autophagy: is it a strategy for Huntington's disease treatment?, J Huntingtons Dis. 2013;2(2):149-57. doi: 10.3233/JHD-130052.

The manuscript has some solid data, but is confusing to follow in that they go back and forth from small fragment over expression models, in which 3% of the mutant huntingtin ORF with synthetic allele is over expressed at massive levels to induce the formation of aggregates, to data from human HD fibroblasts, which is arguably the most exciting data in this work. The only assay of pathological effects comes from an outdated fly fragment model.

For historical and practical reasons, our primary screen was carried out in a cell-based aggregation readout screen. This model allowed us to quickly screen libraries for compounds that could modulate the levels of NT-mHtt fragment. A similar approach would have been impractical in other cell models including the human HD fibroblasts. Once we identified NCT-504 as a potential target in the primary screen, we proceeded to the validation of the compound as well as its putative target (PIP4Kγ) in more physiologically relevant models, because we completely agree with the reviewer that these models are better suited for understanding mHtt biology. Given the consistency we observe between the fragment models and the full length human patient fibroblasts or mouse STHdhQ111 we think that the screening strategy is validated. Although currently not a feasible option, in future studies we are aiming at performing these protein levels screens in the HD patient fibroblasts.

With regard to the Drosophila model, we have followed the reviewers’ advice by evaluating the effect of PIP4K loss of function in a Drosophila model expressing full length Htt with 200 glutamines (new panels in Figure 5C). As shown decreasing dPIP4K(CG17471) by either a heterozygous mutant (PIP4K-dCG17471LOF) or by expression of a dominant negative allele (PIP4K-dCG17471DN) leads to a suppression of the Htt-FLQ200-induced motor impairments. We used two different parameters to measure motor performance: 1- a climbing test and 2- average climbing speed.

Our previous submission presented data using an N-terminal exon 1 fragment Htt-poly-Q model. We now include data using full length Htt-poly-Q, obtaining similar results and indicating that silencing of the gene or elimination of a single allele ameliorates the neurodegeneration and loss in motor skills. There are numerous published studies by us and others using Drosophila models (PMID: 18184562, 18184562) and mouse models (PMID: 24052178) of Huntington disease expressing full length huntingtin poly-Q or just the N-terminal exon 1 fragment. Comparing the phenotype of the disease in both models, exon 1 Htt-poly-Q models tend to be more severe with an earlier onset and therefore more difficult to recover from.

The overall picture seems exciting, but the details in the data are not as clear as the writing would suggest. In many of the figures and assays, they only look at mutant HD cells, and not fibroblasts from unaffected individuals,

It is difficult to quantify levels of Htt protein in wildtype fibroblasts because they express very low levels of protein and in general in this cell line the mutant protein does not accumulate much. Recently we found that HEK293 cells have quite high levels of wildtype Htt protein. We have produced a stable HEK293 cell line transfected with LC3-GFP-mCherry. We now have included in our manuscript results from this cell line of NCT-504’s effect on autophagosome formation, autophagy flux and Htt level in dose response and in a time dependent manner. Indeed modulation of autophagy and reduction of wildtype Htt levels can be observed in HEK293, indicating a biological effect upon inhibiting PIP4Kγ in a context outside of HD. Other potential effects of inhibiting PIP4Kγ on unaffected individuals are commented on by Lewis Cantley in their PIP4Kγ knockout mouse (and referenced in our paper) where the only observable effect is the activation of the immune system. Additionally we have included in Figure 4—figure supplement 1 data showing that decreasing PIP4K activity using a heterozygous loss of function (lof) or a neuronal expressed dominant negative (DN) does not cause neuronal dysfunction in Drosophila measured as motor performance.

and most of the readouts are only looking at one protein, while the proposed mechanism is to switch on autophagy in general. Autophagy, and its sub-categories of mitophagy, micro-autophagy, macro-autophagy, pexophagy are all dynamic system in a high rate of constant flux in the cell. If this is truly a drug lead for Huntington's disease, or any disease, they need to show some selectivity in the effected state and in the target. It seems that even in their loading controls, there is a massive global effect of shutting down PIP4Kγ either by SiRNA or global inhibitor. In other words, autophagy in all its forms is in dynamic flux for a reason, and just pushing it in one direction may be as problematic and inhibiting it.

As we note in the manuscript, numerous authors have described mutant Htt-dependent autophagy dysregulation in Huntington’s disease (HD). Especially important are several studies describing severe deficits in productive autophagy in HD cells and an inability to load mHtt and aggregates into autophagosome vesicles. Our data show that inhibition of PIP4Kγ increases productive autophagy of mHtt and aggregates. In the discussion we hypothesize that this is due to the observed shift in PI’s, which seem to promote the loading of mHtt and aggregates. In addition, this shift also upregulates autophagy flux, which might potentially increase the proteolytic rate of other ubiquitinated proteins tagged for elimination by autophagy, as well as damaged mitochondria (mitophagy is impaired in HD) and other cell damaged organelles. The impact of shifting autophagy dynamics can be insult-dependent and has to be evaluated in relevant disease models. Indeed upregulation of autophagy can prevent or promote apoptosis, depending of the cell model and its capacity to respond to stress. In previous KO mouse studies by Lewis Cantley, PIP4Kγ-dependent upregulation does not seem to have a major biological impact in viability or behavioral dysregulation, with no observable phenotype other than hyperactivation of some T cell populations, which it seem to be controllable by rapamycin. In the context of HD, levels of mHtt are elevated not only in striatal neurons, but also many other cell types, and numerous authors have described non-neuronal Huntington impairments, and therefore a global reduction of mHtt is perceived to be beneficial, as has been demonstrated with other autophagy modulators. Indeed our in vivo models indicate that this is the case, with an amelioration of the disease progression. Moreover this benefit of inhibiting PIP4Kγ is titrated by elimination of a single allele. Further dose dependent efficacy studies with a fully optimized PIP4Kγ inhibitor will be necessary to determine the adequate dose, schedule and potential side effects of this novel mechanism.

We have performed titration curves for NCT-504 in the Q45 patient fibroblasts. As shown in Figure 4—figure supplement 1, NCT-504 is not toxic until we reach the 10μM concentration, at which it does cause cell loss. We do see a significant effect on mHtt levels at 5μM. Thus we are able to reduce mHtt levels in non-toxic conditions.

We have also assessed the effect of the heterozygous PIP4KLOF as well as the new PIP4KDN allele on motor performance on their own in the absence of mHtt. As shown in the new Figure 5—figure supplement 1 neither allele causes significant behavioral improvement or worsening compared to the negative control as measured by climbing ability or speed

I have an conceptual concern that this data suggest that 97% of the huntingtin ORF is irrelevant for regulating the turnover of this protein by autophagic mechanisms, and it appears from this data that the molar stoichiometry is irrelevant, as whether they look at endogenous huntingtin proteins, or massively over-expressed small molecule fragments at orders of magnitude above physiological molarity, the compound works just as well to clear 'huntingtin', and likely everything else.

Our initial data focused in demonstrating that inhibition of PIP4Kγ reduces levels of mHtt and N-terminal-exon 1 aggregates in several cell models, and that the mechanism has a therapeutic benefit in a relevant N-terminal Htt-poly-Q Drosophila model of HD. As discussed above, now we have added new data using a FLQ200 Drosophila model showing similar results, with amelioration of the decline in motor performance. Our studies in the mechanism of action indicate an increase in autophagy flux, which is impaired both by expression of NT Htt-poly-Q and by expression of FL Htt-poly-Q. Our control proteins indicate that this mechanism does not clear everything else, but eliminate specific insults in Huntington cells.

My major concern is that I believe they have proven that PIP4K γ inhibition triggers a massive level of autophagy, but they have no data here that show any selectivity or health effects on the fibroblasts. Is this "curing dandruff by decapitation"?

We respectfully disagree with the interpretation that PIP4K γ inhibition triggers a massive level of autophagy. Instead PIP4K γ inhibition seems to elevate productive autophagy of mHtt and fragments, ameliorating disease progression.

Autophagy is a continuous and dynamic process carried out by cells to eliminate properly tagged proteins and organelles, by polyubiquitination (macrophagy, mitophagy) or KFERQ-like sequences (chaperone-mediated autophagy). The basal level of autophagy depends of stimulation by environmental stressors but also by intracellular factors such as genetic stressors, lysosomal homeostasis and level of energy production. Elevating the basal level of productive autophagy does not have to be necessarily detrimental for the cell in general and even less in HD, as productive autophagy is downregulated by the expression of mHtt and its fragments. Indeed, our siRNA experiments across cell lines show that elimination of the PIP4Kγ kinase activity does not induce cell death. This is further corroborated by the viability of full knockout animals with no growth or behavioral abnormalities. Furthermore any theoretical risk could be further modulated using the right dose and schedule of a PIP4Kγ small molecule inhibitor.

It is not clear for us how to assess the potential health benefit on fibroblasts, besides analyzing the reduction of the accumulated insult and the effect on cell viability. Therapeutically, any intervention is a balance between potential benefits and adverse effects which need to be carefully evaluated in relevant disease models (cell and animals if available) and eventually in human clinical trials. We believe that our data demonstrate the therapeutic benefit of this mechanism in relevant models of HD.

As detailed above, analysis of toxicity by dose titration in cells indicates that we are causing mHtt decrease at a non-toxic concentration. Furthermore decreasing PIP4K levels or activity in Drosophila does not cause any visible toxic effect in our assays. This is consistent with results in mouse models. Overall we believe the PIP4Kγ is a very safe target, and any toxicity that maybe observed with the NCT-504 is likely due to side effects not related to PIP4Kγ inhibition.

They reference a manuscript on a PIP4Kγ knockout mouse stating the investigators observed "no growth or behavioral abnormalities", except the manuscript actually describes a massive global hyperactivation of the immune systems in those knockouts, and the true validation of a drug target would be a conditional knockout. For what it's worth, Broad's Exac exomic sequencing database shows no homozygous null alleles of PIP4Kγ yet in humans. Thus, I have concerns that PIP4Kγ inhibition could really be a drug target.

Lewis Cantley et al. in their description of the phenotype of PIP4Kγ knockout mouse also mentioned a population of human subjects carrying polymorphisms at PIP4K2C locus linked with familial autoimmunity. Additionally these authors described the hyper-activation of the immune system in the null PIP4Kγ mouse, which they are able to correct using the mTORC1 inhibitor rapamycin, a well-accepted upregulator of autophagy initiation which is also able to ameliorate in vivo the HD phenotype, demonstrating that the autophagy-modulating effect of inhibiting PIP4Kγ is distinct from inhibition of mTOR. Importantly in our HD Drosophila models, we demonstrated a potential therapeutic benefit upon decreasing dPIP4K activity, indicating that a partial modulation of PIP4Kγ activity might be enough for observing a beneficial effect. Dose dependent efficacy studies with the HD animal models and eventually in HD patients with a fully optimized PIP4Kγ inhibitor will be necessary for determining the percentage of PIP4Kγ-inhibition and the schedule needed to minimize any potential immune side effects while maximizing the therapeutic benefit from mHtt elimination. Furthermore, the potential secondary effects of PIP4Kγ inhibitors such as a risk of hyper-activation of the immune system could be different in mHtt-carrying HD-patients (migration of primary microglia seems to be impaired by mHtt expression PMID: 27615381) that in wt HD-patients or in population carrying specific SNP’s at the PIP4K2C locus.

They do not state exactly how their HD fibroblasts are immortalized, but is seems that there is no difference in effect between HEK293, STHdHQ111, HD primary cells, or HD immortalized cells. This is rather surprising since the only known inducer of huntingtin expression is P53, and some of those lines have P53, and others do not. This has been a concern that has been ignored in most HD cell lines studies, despite over 80 manuscripts published consistently defining disrupted P53 biology in HD, from mouse to humans. So whether they have p53 or not, whether small fragments of full protein, everything gets cleared, in all lines.

To our knowledge all the cell lines used in our experiments, PC12, HEK293, MEF, non-immortalize and immortalize HD fibroblast, STHdHQ111, and mouse primary cortical neurons, are wt p53. For example:

HEK293: (https://www.ncbi.nlm.nih.gov/pubmed/0008504475).

PC12 cells: (https://www.ncbi.nlm.nih.gov/pubmed/16817227).

We have now indicated in the Materials and methods section that the fibroblasts were immortalized using SV40 large T antigen.

Major concerns

Figure 1

From the high throughput screen, they found NCT-504 and by the structure guessed that it was a kinase inhibitor, and tested this by kinomic screening, and used a cutoff of >65% inhibition to claim specificity. This is a very high cutoff, and the entire kinome screening data should be included a supplemental, not just 3/442 kinases. From my experience with pharma, even 20% inhibition is cause for concern on these screens. The IC50 is about 16μM for NCT-504. This is very high, and by their own data, the IC50 exceeds the toxicity threshold concentration in neurons.

We did include the entire kinome screening data as Table 1 – Source data 1, and apologize if it was not available by the reviewer somehow. It can be observed that at 10 μM NCT-504 only PIP4Kγ is robustly inhibited, with little or no effect in the rest of the 442 kinases in the panel, as might be expected for a non-competitive allosteric inhibitor.

Regarding the IC50 of NCT-504, we explained in the manuscript the differences in IC50’s obtained using several in vitro kinase assays. While the IC50 using isolated PIP4Kγ was 15.8 μM, the IC50 using DiscoverX binding assay was 350 nM. As a screening center, we are very familiar with these kind of discrepancies between biochemical assays, and in the manuscript refer to other authors describing similar problems with allosteric modulators. The main reason behind these discrepancies has to do with the assay conditions used for evaluating inhibition of isolated kinases, usually requiring extraordinary high levels of ATP in order to promote autophosphorylaton and activation and the need for using micelles or detergents to stabilize the active conformation of the kinase, something especially important for lipid kinases such as PIP4Kγ. In general, for these kind of kinases, the DiscoverX binding assay using E. coli or mammalian cell-expressed kinases labeled with DNA tag for qPCR readout, seems to more faithfully reproduce the activity of the kinase in its natural environment. Importantly, in order to properly evaluate the real PIP4Kγ inhibitory capacity of NCT-504 in a cell environment we measure its dose dependent modulation PI5P levels (and other PI’s), using siRNA as a control of the maximal elevation obtained upon elimination of the activity of PIP4Kγ. These measurements in cells clearly indicate that the concentration used in the cell assays to evaluate effects on mHtt and aggregates robustly inhibits PIP4Kγ and that blocking the activity of this kinase does not reduce cell viability. All the data reported in the original manuscript was carried out at non-toxic concentration of NCT-504 in the same cells. Now we have incorporated additional dose-response and viability in different cell lines and end points (PI’s levels, autophagy, mHtt levels, viability, etc.)

The concentrations of NCT-504 used for effects vary hugely across the figures. In Figure 1, and from text, IC50 is quoted at 15.8μM. In Figure 1—figure supplement 1, 2μM is used to reduce aggregation by 50% effect, so whatever reason they see effect, it seems to be unlikely through PIP4Kγ inhibition.

The basal level of autophagy is dynamic process and varies in an insult dependent manner (some cell lines expressing poly-Q’s accumulates more mHtt or aggregates than others), in a cell type dependent manner (genetic background affects the basal level of autophagy), upon assay conditions (confluency, passage, nutrient level), time (we evaluate modulations of PI’s upon exposure to NCT-504 in a time dependency manner) and of course in a concentration dependent manner. We previously indicated that the concentration of NCT-504 able to induce a robust effect is cell type dependent, as would be expected for a modulator of autophagy. Now we present the data using the same concentration across several cell lines (Figure 4) or dose-dependent in a single cell line (Figure 3—figure supplement 2).

Figure 2: concentration of NCT-504 is now 10μM, less than the IC50, and only one dose tested.

We include now PI-dose titration (Figure 2—figure supplement 2) and cell viability (Figure 2—figure supplement 1) in MEF with NCT-504. We use in these cell 10μm as we want to observed a robust effect on PI’s modulation at a non-toxic dose for comparison with the shRNA data. As we mention before the extent of the PIP4Kγ inhibitory effect is cell dependent, as different cells have level of basal autophagy and impediments (or not) in autophagy flux. Upon evaluation of a specific cell line, we select the adequate dose depending, among other things, of the impact of NCT-504 on cell viability, being sure that we report the effect a non-toxic doses.

The loading controls seem to switch between α-tubulin, actin and GAPDH. For these studies looking at autophagic flux, as in Figure 4 they need better controls, not an insoluble polymer of actin. I'd like to see GAPDH and at least two inner mitochondrial proteins to show me the cells are not just completely autophagic or apoptotic. Images are essential.

Again, for Figure, 4, now 2μM of NCT-504 is used again in Q68 cells, but then 10μM in Q45 cells. This is plus or minus 500%.

Figure 4: they need to show the whole blot in a supplemental figure. I see a severe hit on actin levels in Figure 4B with PIP4K2C_si, and in Figure 4D with 10μM NCT-504, which concerns me that these cells might be heavily auto-degrading or just dead.

As suggested by the reviewer, we have included GAPDH in our dose-dependent experiments using patient fibroblast. We also have included now the effect of NCT-504 treatment on cell viability of numerous cell lines. Impaired mitophagy has been reported to contribute to HD (PMID: 26268247) and therefore it is possible that the increment in productive autophagy by NCT-504 might also affects the levels of mitochondrial proteins, within being necessary a bad outcome. For example we do not see an effect on cell viability upon silencing PIP4Kγ, and definitively we observe a beneficial therapeutic effect in Drosophila. Additional studies to better characterize these effect in cells and the effector proteins are necessary but we believe that are out of the scope of these paper.

In Figure 4—figure supplement 4, they show a very unconventional assay for neuronal toxicity, in that they are no showing any images of neurons or other measure of health, or even life, but their own data shows a pretty severe "toxicity" at 10μM NCT-504, the exact concentration used in HD fibroblasts, and still below their calculated IC50.

The effect of NCT-504 in Htt aggregates and reactive species evaluated in mouse primary cortical neurons transduced with 72Q shown in Figure 4—figure supplement 3 is a 5μM and 2.5μM, The same concentrations were used in HD fibroblasts and striatal cells from knock-in HD mice (Figure 4) without major impact in viability.

Figure 4—figure supplement 4B, that western blot is not appropriate data to conclude anything.

We have eliminated Figure 4—figure supplement 4B, leaving only 4A and 4C.

Figure 5. They now switched back to a severe allele Q128 (extremely rare) and only small fragment 1-231. Why did they not use full huntingtin?

Please see answers regarding additional Drosophila data above.

Reviewer #3:

1) How was NCT-504 selected for these studies from published the high-throughput screen? Was this best hit?

NCT-504 was produced upon medicinal chemistry optimization toward potency and metabolic stability of the screening hit ML168 (reference 33). The manuscript has been corrected to clarify this question. Details of the structure-activity relationship will be published in a following manuscript and we consider that topic out of the scope for this paper which focuses on the validation of PIP4Kgamma as a druggable target for Huntington Disease.

Because of the complexity of these pathways, it would be helpful to have a schematic early on to show the various components of the pathway, how each of the PIP derivatives are generated and where this kinase fits into general scheme.

We already have a large amount of figures in this manuscript and we ask the editor for guidance in addressing this question. We can include such as scheme, similar to that found in references 32 and 42.

2) Figure 4A - Why not show more complete set of westerns as in C and D? Need quantification for Figure 4.

To do quantification in patient fibroblast in dose response is quite challenging because the overall levels of Htt protein are low and NCT-504 seems to impact cell viability of these cell lines at double digit μM concentrations. Alternatively we have observed high levels of Htt in HEK293 cells and in these cells NCT-504 des not impact viability at all the tested doses (Figure 3—figure supplement 2) that allow us to use a FRET assay to quantify Htt levels in dose response.

3) Do the fibroblasts have HMW species? This is not mentioned and if they are present, why they are not included should be addressed.

As far as we know there are no HMW species in the fibroblast lines. We do not see any HTT left in the stacking portion when running WB gels. In addition, when we have used oligomer specific TR-FRET analysis we have not detected any signal in these cells.

4) A potentially important experiment would be using a pulse chase or other such assay to measure whether this truly increases degradation of the protein.

We have included now the effect of NCT-504 in neurons on autophagic flux determined after 24 hours using a Dendra2-LC3 photoconverted reporter (Figure 3—figure supplement 3) and we have titrated Htt levels along autophagosome formation and autophagy flux in HEK293 cells (Figure 3—figure supplement 2).

5) Please describe length of HTT fragment in the 128Q HTT transgene in flies.

This has been added to the Materials and methods.

6) A citation should be included for the Neuron paper from the Davidson lab when discussing Rhes and Rheb.

We have included the requested citation.

7) It is a stretch to say that compound modulates HTT proteolysis as there is a single Western blot with one panel.

We have include now other methods of quantification of Htt levels using a FRET assay. The western bolt analysis was carried out by triplicated in each cell line and we used several types of cell lines.

[Editors' note: the author responses to the re-review follow.]

Essential revisions

It is critical that the authors pay close attention to and respond accordingly to reviewers 1 and 2 concerns that the strength of this manuscript is decreased considerably by the style/manner in which it is written. Extensive editing to the text and figures are needed so that the manuscript flows smoother and rationale for the studies are clearer.

We read the decision letter carefully and have addressed every concern. We have extensively edited the writing style/manner without changing the content or essence of the paper. The paper was reviewed and modified by several authors and colleagues who speak English as their native language. Almost all figures have been edited to have consistent font and better quality. We have also included concentrations of compounds in all figures.

Reviewer #1:

In this new submission of a previously submitted manuscript the ability of a small molecule to inhibit PIP4Kγ and impact HD-like phenotypes in cellular and animal models expressing mutant Htt (full-length and fragment) is presented. Overall while this work addresses a very interesting issue and additional data using a FL-Htt fly model have been added, it remains difficult to read/follow in several key places.

We have taken this criticism about the writing style seriously and changed the text in most places. While the basic content and essence of the paper remains unchanged the writing has been modified extensively. We believe the narration and discussion is much improved. Please let us know if you have any additional questions

Specific points:

Overall the manuscript remains hard to read, needing English editing in many places. In several places statements are referenced when they should be. For example, second paragraph of Results – "Similar discrepancies in potency […]" needs to be referenced.

This has been addressed by the addition of two new references:

Rudolf, A.F., et al., A comparison of protein kinases inhibitor screening methods using both enzymatic activity and binding affinity determination. PLoS One, 2014. 9(6): p. e98800.

Smyth, L.A. and I. Collins, Measuring and interpreting the selectivity of protein kinase inhibitors. J Chem Biol, 2009. 2(3): p. 131-51

Clarity of Figure 3 would be improved by including concentrations of compounds used within body of the figure and not just in the legend.

We have addressed this. We have included concentrations and added clear labels with a consistent Arial font throughout the paper.

Reviewer #2:

This is a revised manuscript and the authors were very responsive to the reviewers.

The work described identifies PIP4Kγ as a therapeutic target for HD and shows that pharmacological inhibition of the enzyme promotes mutant HTT degradation and autophagy. They provide in vivo evidence that lower the levels of PIP4Kγ is beneficial in a fly model of HD-retinal assays and motor performance. This is a promising direction to pursue as a possible therapy for HD and is in line with current treatments directed at lowering mHTT protein levels. Generally the quality of the data is strong. Minor points are that there are a number of formatting typos. Also the figures are not very consistent. Formatting is varied-different fonts, different graphing and repetitive information in legend and graphs.

We appreciate the reviewer’s support.

Our work was a collaboration between multiple institutes and we are proud that we were able to work together and bring different expertise to the table to discover to validate a new target for Huntington’s disease. However, as the work was done in different sites the data for the figures was compiled with various format/fonts/styles. We understand that the figures needed to be consistent and have addressed this by changing, wherever possible, the figures’ text to a consistent Arial font throughout the manuscript.

Reviewer #3:

This manuscript describes the effect of inhibition of the PIP4Kγ in models of Huntington's disease. There is extensive description of the effect of the drug on lipid metabolism and autophagy. There are also data from cell models including HD fibroblasts. The reduction of Htt levels is of interest. However effects on Htt aggregates are difficult to interpret since aggregates may be toxic or protective. There is also an interesting effect on an HD Drosophila model. It would be beneficial if the investigators had some more substantial data on models more directly relevant to HD pathogenesis. For instance instead of using HD fibroblasts it would be beneficial to use HD iPS cell models. In addition, there are no data from HD mouse models. The processes of autophagy are cell type specific, and may be different in Drosophila and mammalian systems, so in vivo data from mice rather than flies would be advantageous.

These comments, while very relevant, are outside the scope of the paper focused in the validation of PIP4kγ as a target for Huntington’s disease. We show in multiple cell-based assays, including patient fibroblast and stratial neurons, that inhibition of this lipid kinase via si or sh RNA leads to a reduction of mutant or wild-type Htt protein. In addition, we disclose a tool compound that acts as an inhibitor of PIP4kγ and is able to correct the disease phenotype in multiple cell lines.

Our team does not have access to iPS cells at the present moment. Moreover striatal neurons differentiated from Htt iPSC’s do not present aggregates, and therefore we would be able only to use them for measuring total levels of mHtt and autophagy dynamics. We already have measured these events with well accepted mammalian cellular models. Thus, using compound treatment we show the following in multiple different cell lines (included in bold are cells of neuronal origin):

1) Reduction of GFP-mHttQ103 in PC12 cells, a neuroblastic cell line of rat adrenal phaeochromocytoma origin (Figure 1B, C).

2) Reduction of wt Htt from 293A cells which is derived from primary embryonal human kidney cells (Figure 3—figure supplement 1)

3) Reduction of mHtt protein levels in an HD patient fibroblast cell line (Q68), Figure 4A

4) Reduction of mHtt protein levels in an HD patient fibroblast cell line (Q45), Figure 4C

5) Reduction of mHtt protein levels in immortalized striatal cells from knock-in HD mice (STHdhQ111), Figure 4D.

6) Reduction of accumulated HTT-exon1 aggregates in HEK293T cells transfected with GFP-HTT(exon1)-Q23 or GFP-HTT(exon1)-Q74, Figure 1—figure supplement 1.

7) Reduction of wt Htt in 293A cells (derived from HEK cells), Figure 3—figure supplement 1.

8) Reduction of GFP-Htt(exon1)-Q74 aggregates in Atg7+/+ MEF cells (Figure 3—figure supplement 3)

9) Reduction of mt Htt in mouse primary cortical neurons transduced with Htt(exon1)-Q72 lentivirus, Figure 4—figure supplement 4

The above data strongly indicates our approach of inhibiting PIP4Kγ to upregulate autophagy and reduce Htt will work across multiple cell line including those of neuronal origin, and therefore is not cell specific.

NCT-504 is a tool compound and needs to be optimized further towards experiments in animal models. We have evaluated the pharmacokinetics of the compound and while it has decent plasma exposure it fails to cross the blood brain barrier. We intend to improve the ADME (Absorption, Distribution, Metabolism, Elimination) properties of the compound via the generation of new analogues.

https://doi.org/10.7554/eLife.29123.028

Article and author information

Author details

  1. Ismael Al-Ramahi

    1. Jan and Dan Duncan Neurological Research Institute, Texas Children's Hospital, Houston, United States
    2. Baylor College of Medicine, Texas Medical Center, Houston, United States
    Contribution
    Conceptualization, Data curation, Formal analysis, Supervision, Visualization, Methodology, Writing—original draft, Writing—review and editing
    Competing interests
    No competing interests declared

  2. Sai Srinivas Panapakkam Giridharan

    Department of Cell and Developmental Biology, Life Sciences Institute, University of Michigan, Ann Arbor, United States
    Contribution
    Conceptualization, Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing—original draft, Writing—review and editing
    Competing interests
    No competing interests declared

  3. Yu-Chi Chen

    Division of Preclinical Innovation, National Center for Advancing Translational Sciences, Rockville, United States
    Contribution
    Investigation, Visualization, Methodology
    Competing interests
    No competing interests declared

  4. Samarjit Patnaik

    Division of Preclinical Innovation, National Center for Advancing Translational Sciences, Rockville, United States
    Contribution
    Data curation, Formal analysis, Investigation, Methodology, Writing—original draft, Project administration, Writing—review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-4265-7620

  5. Nathaniel Safren

    Department of Neurology, University of Michigan, Ann Arbor, United States
    Contribution
    Investigation, Methodology, Writing—original draft
    Competing interests
    No competing interests declared

  6. Junya Hasegawa

    Department of Cell and Developmental Biology, Life Sciences Institute, University of Michigan, Ann Arbor, United States
    Contribution
    Conceptualization, Data curation, Formal analysis, Investigation, Visualization, Methodology, Writing—review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-7041-890X

  7. Maria de Haro

    1. Jan and Dan Duncan Neurological Research Institute, Texas Children's Hospital, Houston, United States
    2. Baylor College of Medicine, Texas Medical Center, Houston, United States
    Contribution
    Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology
    Competing interests
    No competing interests declared

  8. Amanda K Wagner Gee

    Division of Preclinical Innovation, National Center for Advancing Translational Sciences, Rockville, United States
    Contribution
    Investigation, Visualization, Methodology
    Competing interests
    No competing interests declared

  9. Steven A Titus

    Division of Preclinical Innovation, National Center for Advancing Translational Sciences, Rockville, United States
    Contribution
    Validation, Investigation, Visualization, Methodology
    Competing interests
    No competing interests declared

  10. Hyunkyung Jeong

    The Ken and Ruth Davee Department of Neurology, Feinberg School of Medicine, Northwestern University, Chicago, United States
    Contribution
    Formal analysis, Investigation, Visualization, Methodology
    Competing interests
    No competing interests declared

  11. Jonathan Clarke

    Department of Pharmacology, University of Cambridge, Cambridge, United Kingdom
    Contribution
    Investigation, Methodology, Writing—review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-4079-5333

  12. Dimitri Krainc

    The Ken and Ruth Davee Department of Neurology, Feinberg School of Medicine, Northwestern University, Chicago, United States
    Contribution
    Conceptualization, Resources, Formal analysis, Supervision, Funding acquisition, Methodology
    Competing interests
    No competing interests declared

  13. Wei Zheng

    Division of Preclinical Innovation, National Center for Advancing Translational Sciences, Rockville, United States
    Contribution
    Conceptualization, Resources, Supervision, Funding acquisition, Methodology, Writing—review and editing
    Competing interests
    No competing interests declared

  14. Robin F Irvine

    Department of Pharmacology, University of Cambridge, Cambridge, United Kingdom
    Contribution
    Conceptualization, Resources, Methodology, Writing—original draft
    Competing interests
    No competing interests declared

  15. Sami Barmada

    Department of Neurology, University of Michigan, Ann Arbor, United States
    Contribution
    Resources, Supervision, Funding acquisition, Investigation, Methodology, Writing—review and editing
    Competing interests
    No competing interests declared

  16. Marc Ferrer

    Division of Preclinical Innovation, National Center for Advancing Translational Sciences, Rockville, United States
    Contribution
    Conceptualization, Resources, Formal analysis, Supervision, Funding acquisition, Methodology
    Competing interests
    No competing interests declared

  17. Noel Southall

    Division of Preclinical Innovation, National Center for Advancing Translational Sciences, Rockville, United States
    Contribution
    Data curation, Formal analysis, Writing—original draft, Writing—review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-4500-880X

  18. Lois S Weisman

    Department of Cell and Developmental Biology, Life Sciences Institute, University of Michigan, Ann Arbor, United States
    Contribution
    Conceptualization, Resources, Data curation, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing—original draft, Project administration, Writing—review and editing
    Contributed equally with
    Juan Botas
    Competing interests
    No competing interests declared

  19. Juan Botas

    1. Jan and Dan Duncan Neurological Research Institute, Texas Children's Hospital, Houston, United States
    2. Baylor College of Medicine, Texas Medical Center, Houston, United States
    Contribution
    Conceptualization, Resources, Data curation, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing—original draft, Project administration
    Contributed equally with
    Lois S Weisman
    For correspondence
    jbotas@bcm.edu
    Competing interests
    No competing interests declared

  20. Juan Jose Marugan

    Division of Preclinical Innovation, National Center for Advancing Translational Sciences, Rockville, United States
    Contribution
    Conceptualization, Resources, Data curation, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing—original draft, Project administration, Writing—review and editing
    For correspondence
    maruganj@mail.nih.gov
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-3951-7061

Funding

Foundation for the National Institutes of Health (R21NS096395)

  • Ismael Al-Ramahi

Darrel K Royal Research Fund for Alzheimers Disease

  • Ismael Al-Ramahi

American Heart Association (Post Doctoral Fellowship (14POST20480137))

  • Sai Srinivas Panapakkam Giridharan

National Institutes of Health (P30-AG053760)

  • Sami Barmada

National Institutes of Health (R01-NS097542)

  • Sami Barmada

University of Michigan (Protein Folding Diseases FastForward Initiative)

  • Lois S Weisman
  • Sami Barmada

National Institutes of Health (R01-NS064015)

  • Lois S Weisman

National Institutes of Health (R01-NS099340)

  • Lois S Weisman

CHDI Foundation

  • Juan Botas

Robert A. and Renee E. Belfer Family Foundation

  • Juan Botas

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

Immortalized MEF wild-type and MEF Atg7 knock-out cells were a gift from Dr. Masaki Komatsu (Niigata University, Japan). None of the cell lines used in this study were included in the list of commonly misidentified cell lines maintained by International Cell Line Authentication Committee. This work was supported in part by National Institutes of Health (NIH) grants R01-NS064015 and R01-NS099340 to LSW, R01-NS097542 and P30-AG053760 to SJB, and the Protein Folding Diseases Fast Forward Initiative, University of Michigan to LSW and SJB. SSPG was supported in part by AHA Postdoctoral Fellowship, 14POST20480137. IA was supported by R21NS096395 grant from the NIH and by the Darrell K Royal Research Fund for Alzheimer’s Disease. JB was supported by grants from the CHDI and the Robert A. and Renée E. Belfer Family Foundation.

Ethics

Animal experimentation: All vertebrate animal work was approved by the Institutional Animal Use & Care Committee at the University of Michigan (PRO00007096).

Reviewing Editor

  1. Harry T Orr, University of Minnesota, United States

Version history

  1. Received: May 30, 2017
  2. Accepted: November 13, 2017
  3. Accepted Manuscript published: December 19, 2017 (version 1)
  4. Version of Record published: December 26, 2017 (version 2)

Copyright

This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.

Metrics

  • 4,303
    Page views
  • 677
    Downloads
  • 41
    Citations

Article citation count generated by polling the highest count across the following sources: Scopus, Crossref, PubMed Central.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Ismael Al-Ramahi
  2. Sai Srinivas Panapakkam Giridharan
  3. Yu-Chi Chen
  4. Samarjit Patnaik
  5. Nathaniel Safren
  6. Junya Hasegawa
  7. Maria de Haro
  8. Amanda K Wagner Gee
  9. Steven A Titus
  10. Hyunkyung Jeong
  11. Jonathan Clarke
  12. Dimitri Krainc
  13. Wei Zheng
  14. Robin F Irvine
  15. Sami Barmada
  16. Marc Ferrer
  17. Noel Southall
  18. Lois S Weisman
  19. Juan Botas
  20. Juan Jose Marugan
(2017)
Inhibition of PIP4Kγ ameliorates the pathological effects of mutant huntingtin protein
eLife 6:e29123.
https://doi.org/10.7554/eLife.29123

Categories and tags

Research organisms

Further reading

    1. Computational and Systems Biology
    2. Neuroscience

    Dynamic organization of cerebellar climbing fiber response and synchrony in multiple functional components reduces dimensions for reinforcement learning

    Huu Hoang, Shinichiro Tsutsumi ... Keisuke Toyama
    Research Article Updated

    Cerebellar climbing fibers convey diverse signals, but how they are organized in the compartmental structure of the cerebellar cortex during learning remains largely unclear. We analyzed a large amount of coordinate-localized two-photon imaging data from cerebellar Crus II in mice undergoing ‘Go/No-go’ reinforcement learning. Tensor component analysis revealed that a majority of climbing fiber inputs to Purkinje cells were reduced to only four functional components, corresponding to accurate timing control of motor initiation related to a Go cue, cognitive error-based learning, reward processing, and inhibition of erroneous behaviors after a No-go cue. Changes in neural activities during learning of the first two components were correlated with corresponding changes in timing control and error learning across animals, indirectly suggesting causal relationships. Spatial distribution of these components coincided well with boundaries of Aldolase-C/zebrin II expression in Purkinje cells, whereas several components are mixed in single neurons. Synchronization within individual components was bidirectionally regulated according to specific task contexts and learning stages. These findings suggest that, in close collaborations with other brain regions including the inferior olive nucleus, the cerebellum, based on anatomical compartments, reduces dimensions of the learning space by dynamically organizing multiple functional components, a feature that may inspire new-generation AI designs.

    1. Neuroscience

    Rodent ultrasonic vocal interaction resolved with millimeter precision using hybrid beamforming

    Max L Sterling, Ruben Teunisse, Bernhard Englitz
    Tools and Resources Updated

    Ultrasonic vocalizations (USVs) fulfill an important role in communication and navigation in many species. Because of their social and affective significance, rodent USVs are increasingly used as a behavioral measure in neurodevelopmental and neurolinguistic research. Reliably attributing USVs to their emitter during close interactions has emerged as a difficult, key challenge. If addressed, all subsequent analyses gain substantial confidence. We present a hybrid ultrasonic tracking system, Hybrid Vocalization Localizer (HyVL), that synergistically integrates a high-resolution acoustic camera with high-quality ultrasonic microphones. HyVL is the first to achieve millimeter precision (~3.4–4.8 mm, 91% assigned) in localizing USVs, ~3× better than other systems, approaching the physical limits (mouse snout ~10 mm). We analyze mouse courtship interactions and demonstrate that males and females vocalize in starkly different relative spatial positions, and that the fraction of female vocalizations has likely been overestimated previously due to imprecise localization. Further, we find that when two male mice interact with one female, one of the males takes a dominant role in the interaction both in terms of the vocalization rate and the location relative to the female. HyVL substantially improves the precision with which social communication between rodents can be studied. It is also affordable, open-source, easy to set up, can be integrated with existing setups, and reduces the required number of experiments and animals.

    1. Neuroscience

    Different rules for binocular combination of luminance flicker in cortical and subcortical pathways

    Federico G Segala, Aurelio Bruno ... Daniel H Baker
    Research Article

    How does the human brain combine information across the eyes? It has been known for many years that cortical normalization mechanisms implement ‘ocularity invariance’: equalizing neural responses to spatial patterns presented either monocularly or binocularly. Here, we used a novel combination of electrophysiology, psychophysics, pupillometry, and computational modeling to ask whether this invariance also holds for flickering luminance stimuli with no spatial contrast. We find dramatic violations of ocularity invariance for these stimuli, both in the cortex and also in the subcortical pathways that govern pupil diameter. Specifically, we find substantial binocular facilitation in both pathways with the effect being strongest in the cortex. Near-linear binocular additivity (instead of ocularity invariance) was also found using a perceptual luminance matching task. Ocularity invariance is, therefore, not a ubiquitous feature of visual processing, and the brain appears to repurpose a generic normalization algorithm for different visual functions by adjusting the amount of interocular suppression.