CC BY-NC-ND
0Cell Reports
Article
Sphingosine Kinase 1 Cooperates with Autophagy to Maintain Endocytic Membrane Trafficking
Graphical Abstract
Authors
Megan M. Young, Yoshinori Takahashi, Todd E. Fox, Jong K. Yun, Mark Kester, Hong-Gang Wang
Correspondence
hwang3@hmc.psu.edu
In Brief
Young et al. show that Sphkl facilitates endocytic trafficking. Inhibition of Sphkl with a lipid-based inhibitor or in the presence of sphingosine accumulates enlarged, dysfunctional late endosomes that impair autophagy. The autophagic machinery coordinates with lysosomal biogenesis to restore endocytic trafficking, thus highlighting crosstalk between autophagy, sphingolipids, and endocytosis.
Highlights
• Inhibition of Sphkl impairs endocytic trafficking to enlarge late endosomes
• Sphkl aids in endosomal membrane fusion independent of catalytic activity
• Autophagy is induced, but flux is impaired by inhibition of Sphkl with SK1-I
• Restored autophagic flux and lysosome biogenesis restore endosomal trafficking
Young et al., 2016, Cell Reports 17,1532-1545 ciossMark November 1, 2016 © 2016 The Author(s).
http://dx.d0i.0rg/l 0.1016/j.celrep.2016.10.019
CelPress
Cell Reports
Article
Sphingosine Kinase 1 Cooperates with Autophagy to Maintain Endocytic Membrane Trafficking
Megan M. Young,1'2 Yoshinori Takahashi,2 Todd E. Fox,3 Jong K. Yun,1 Mark Kester,3 and Hong-Gang Wang124 *
department of Pharmacology, Penn State University College of Medicine, Hershey, PA 17033, USA 2Department of Pediatrics, Penn State University College of Medicine, Hershey, PA 17033, USA 3Department of Pharmacology, University of Virginia, Charlottesville, VA 22908, USA 4Lead Contact
'Correspondence: hwang3@hmc.psu.edu http://dx.doi.org/10.10167j.celrep.2016.10.019
SUMMARY
Sphingosine kinase 1 (Sphkl) associates with early endocytic membranes during endocytosis; however, the role of sphingosine or sphingosine-1-phosphate as the critical metabolite in endocytic trafficking has not been established. Here, we demonstrate that the recruitment of Sphk1 to sphingosine-enriched endocytic vesicles and the generation of sphingosine-1-phosphate facilitate membrane trafficking along the endosomal pathway. Exogenous sphingosine and sphingosine-based Sphk1 inhibitors induce the Sphkl-dependent fusion of endosomal membranes to accumulate enlarged late endo-somes and amphisomes enriched in sphingolipids. Interestingly, Sphkl also appears to facilitate endosomal fusion independent of its catalytic activity, given that catalytically inactive Sphk1G82D is recruited to endocytic membranes by sphingosine or sphingosine-based Sphk1 inhibitor and promotes membrane fusion. Furthermore, we reveal that the clearance of enlarged endosomes is dependent on the activity of ceramide synthase, lysosomal biogenesis, and the restoration of autophagic flux. Collectively, these studies uncover intersecting roles for Sphk1, sphingosine, and autophagic machinery in endocytic membrane trafficking.
INTRODUCTION
Sphingolipids are a class of inter-convertible, bioactive lipids that have dynamic functions in cellular signaling and membrane composition. Sphingosine kinase (Sphk) catalyzes the conversion of pro-apoptotic sphingosine (Sph) to pro-survival sphingo-sine-1-phosphate (S1P) (Hannun and Obeid, 2008). In addition to cellular signaling, Sphk modulates membrane dynamics by converting the largely membrane-restricted substrate Sph to soluble S1P (Goni et al., 2014). Sphingoid base metabolism also has a conserved role in endocytosis, as it is required for yeast endocytosis (Hannich et al., 2011), synaptic vesicle recycling in D. melanogaster (Rohrbough et al., 2004) and C. elegans (Chan
and Sieburth, 2012; Chan et al., 2012), and endolysosomal trafficking in D. melanogaster photoreceptors (Acharya et al., 2003, 2004; Yonamine et al., 2011).
The regulation of membrane fusion events by Sph and S1P was initially characterized in neurons during synaptic vesicle exocytosis. The sole C. elegans isoform, sphk-1, localizes to presynaptic terminals for the generation of S1P and recruitment of the soluble W-ethylmaleimide-sensitive factor activating protein receptor (SNARE) Munc13 for synaptic vesicle release (Chan and Sieburth, 2012; Chan et al., 2012). Moreover, Sph alters the local membrane environment to activate synaptobrevin for SNARE complex assembly and membrane fusion (Darios et al., 2009). Local accumulation of Sph also induces a conformational change in the SNARE Syntaxin1 to enhance its interaction with the docking molecule Munc18 and sequester synaptic vesicles away from the active zone (Camoletto et al., 2009). As neuronal and endocytic SNARE proteins are highly conserved in structure and function (Antonin et al., 2000), we hypothesize that sphingoid base metabolism regulates membrane fusion events in the endocytic pathway. Indeed, knockdown of both mammalian Sphk isoforms impairs transferrin uptake and recycling to suggest a conserved function in endocytosis (Shen et al., 2014). While Sphk1 recruitment to early endocytic intermediates has recently been reported, the identification of Sph or S1P as the important metabolite in this process has not been firmly established.
The endocytic pathway converges with the highly conserved cellular recycling pathway of autophagy. During autophagy, damaged proteins and/or organelles are engulfed within double-membrane vesicles known as autophagosomes and delivered to late endosomes (LEs) and/or lysosomes (Lys) for degradation (Amaya et al., 2015). Autophagy is regulated by over 30 identified autophagy-related genes (Atgs) (Feng et al., 2014). Autophagic degradation requires functional LEs and Lys; thus, it is not surprising that several components of the autophagic machinery positively regulate the endocytic pathway (Lamb et al., 2013), including Beclin1 (Liang et al., 2008; Ruck et al., 2011), Atg14L (Kim et al., 2012), Atg5 (Peng et al., 2014), and Atg3 (Mur-row et al., 2015). Bidirectional crosstalk also occurs between autophagy and sphingolipids, because sphingolipids positively and negatively regulate autophagy (Young et al., 2013) and accumulate in autophagy-deficient cells (Alexaki et al., 2014).
Here, we utilized exogenous Sph, Sphk1-selective inhibitors, and genetic approaches to demonstrate a role for Sph and
1532 Cell Reports 17, 1532-1545, November 1, 2016 © 2016 The Author(s).
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
CrossMark
Sphkl in endocytic membrane trafficking. Moreover, as the autophagic machinery restores endocytic function in response to disruption by a Sphkl-selective inhibitor, our studies validate the importance of autophagy-related proteins in the maintenance of endocytic trafficking.
RESULTS
The Sphkl-Selective Inhibitor and Sph Mimetic, SK1-I, Impairs Endocytic Membrane Trafficking
Tools that allow for the study of dynamic changes in endogenous sphingolipids within localized cellular membranes during endo-cytosis remain to be established. Therefore, to investigate the role of Sphkl activity in endocytic trafficking, we utilized several pharmacological inhibitors. Unexpectedly, Sph-based Sphk inhibitors SK1-I, N,N-dimethyl-sphingosine (DMS), and FTY720 induce the rapid appearance of dilated cytosolic structures that we will call "vacuoles" (Figure 1A). Vacuolization is dose and time dependent and occurs in transformed mouse embryonic fibroblasts (MEFs) as well as multiple cancer cell lines, including HeLa cells (Figures 1A and S1A-S1C). In contrast to macropinocytosis, the large vacuoles (>5 mm in diameter) arise from the progressive fusion of small perinuclear vacuoles (Movies S1 and S2). Interestingly, the Sph-based inhibitors also generate small vacuoles in Sphk1-deficient MEFs; however, these structures do not fuse upon prolonged monitoring to suggest that vacuole enlargement is Sphk1 dependent (Figures 1A, S1B, S1D, and S1E; Movies S3 and S4).
Strikingly, vacuoles are not observed under basal conditions in SphklMEFs or upon Sphk1 inhibition by the potent, small-molecule inhibitor PF-543 (Figure 1A). We hypothesize that treatment with the Sph mimetics SK1-I, DMS, and FTY720 overloads endocytic membranes with Sph-like lipids to promote Sphk1-dependent membrane fusion and vacuole formation. Consistently, phosphorylated FTY720 fails to induce vacuole formation, thus highlighting the importance of the Sph-like structure (Figure S1F). Moreover, vacuolization occurs independently of S1P receptor (S1PR) signaling, as pre-treatment with the S1PR agonist FTY720-phosphate fails to alter vacuole formation by FTY720 or SK1-I (Figure S1G).
Given that SK1-I is a Sphk1-selective inhibitor (Paugh et al., 2008) and that Sphk1 is implicated in endocytosis (Shen et al., 2014), we chose the isozyme-selective inhibitor for further study. Consistent with a role in endocytic trafficking, the limiting membranes of SK1-I-induced vacuoles contain the LE/Lys marker Lamp1-RFP (red fluorescent protein) (Figure 1B). In contrast, mitochondrial and endoplasmic reticulum structures remain relatively intact (Figures S2A and S2B). To visualize fluid-phase endocytosis, cells were co-treated with 10 kDa Alexa Fluor 488-dextran. Interestingly, basal fluid-phase endocytosis is not altered by the loss of Sphk1, as the internalization and delivery of fluorescent dextran to Lamp1-positive puncta is comparable in wild-type (WT) and Sphk1~'~ cells (Figure 1B). In contrast, inhibition of Sphk1 by SK1-I accumulates large Lamp1-positive vacuoles containing fluorescent dextran, while treatment of Sphklcells moderately accumulates fluorescent dextran puncta that are significantly smaller than the vacuoles of WT cells (Figure 1B).
To confirm that the enlargement of SKI-I-induced vacuoles is Sphkl dependent, Sphkl expression was rescued in Sphkl-deficient MEFs (Figures 1C and 1D). Indeed, transient expression of EGFP-tagged murine Sphkla significantly enlarges SKI-I-induced Lampl-positive vacuoles compared to EGFP control (Figures 1D and 1E). Unexpectedly, EGFP-tagged catalytically inactive Sphk1aG82D significantly enlarges SK1-I-induced vacuoles (Figures 1D and 1E). EGFP-Sphk1a and EGFP-Sphk1aG82D are both recruited to vacuole membranes during SK1-I treatment and are present during vacuole fusion (Figure 1D; Movie S5), leading us to hypothesize that Sphk1 can promote membrane fusion independent of its catalytic activity. Collectively, these studies reveal that inhibition of Sphk1 by SK1-I promotes LE/Lys fusion in a Sphk1-dependent manner.
Sphkl Activity Promotes Endocytic Membrane Trafficking
Next, we attempted to mimic the effect of SK1-I by overloading the endocytic pathway with a non-toxic dose of exogenous sphingosine (Sph) in the presence of PF-543. Exogenous Sph induces the dose-dependent formation of vacuoles in both WT and Sphk1-deficient MEFs (Figures 2A and S2C). However, in contrast to non-metabolizable SK1-I, Sph-induced vacuoles rapidly disappear by 4 hr in WT cells (Figure 2B). Interestingly, loss of Sphk1 or inhibition of Sphk1 by PF-543 significantly prolongs Sph-induced vacuoles to suggest a critical role for Sphk1 activity in restoring membrane trafficking in response to exogenous Sph (Figures 2B, S2D, and S2E). Similar to SK1-I, Sph-in-duced vacuoles are positive for Lamp1-RFP to indicate that an overload of Sph blocks endocytic trafficking at LE/Lys (Figure 2C).
To confirm the result using a genetic approach, EGFP-Sphk1a or EGFP- Sphk1aG82D was transiently expressed in Sphk1-defi-cient MEFs (Figures 1C and 2C). Indeed, EGFP- or EGFP-Sphk1aG82D-expressing cells sustain Sph-induced vacuoles compared to rescue cells expressing EGFP-Sphk1a (Figures 2C). Interestingly, the size of Sph-induced vacuoles is significantly increased by EGFP-Sphk1aG82D compared to EGFP control to once again suggest that recruitment of Sphk1 can promote the fusion of Sph-containing endocytic membranes independent of its catalytic activity (Figures 2C and 2D). Taken together, these results suggest a role for Sph and Sphk1 protein in endosomal maturation.
Inhibition of Sphkl by SK1-I Accumulates Enlarged Late Endosomes and Amphisomes Enriched in Sphingolipids
As exogenous Sph is a substrate of several sphingolipid-metab-olizing enzymes, we chose to utilize non-metabolizable SK1-I to further characterize the endocytic defect. As the disruption of lipid rafts at the plasma membrane recruits Sphk1 to early endocytic intermediates (Shen et al., 2014), cells were briefly treated with the cholesterol-depleting agent hydroxypropyl-b-cyclodex-trin (HPbCD) prior to SK1-I. Interestingly, cellular vacuolization and uptake of SK1-I were impaired by HPbCD to suggest that the lipid-based inhibitor enters the cell by cholesterol-dependent endocytosis (Figures S3A and S3B). Consistently, inhibition of endocytosis by pre-treatment with the vacuolar ATPase
Figure 1. The Sphkl-Selective Inhibitor and Sph Mimetic, SK1-I, Impair Endocytic Membrane Trafficking
(A) Representative phase-contrast images of wild-type (WT) and Sphk1^/^ MEFs treated with DMSO; 10 mM SK1-I, DMS, or FTY720; or 1 mM PF-543 for 6 hr.
(B) Live-cell fluorescence imaging of WTand Sphk1~'~ MEFs stably expressing Lamp1-RFP (L1-RFP) co-treated with 100 mg/mL Alexa Fluor 488-dextran (AF488) and DMSO or 7.5 mM SK1-I for 2 hr.
(C and D) Live-cell fluorescence images of Sphk1~'~ MEFs stably expressing Lamp1-RFP (L1-RFP) transiently transfected with EGFP, EGFP-Sphk1a (WT), or EGFP-Sphk1aG82D (G82D) and treated with (C) DMSO or (D) 7.5 mM SK1-I for 4 hr. DIC, differential interference contrast.
(E) Vacuole size of (D) is presented as the mean ± SEM from three independent experiments. n > 400 vacuoles per treatment. ****p < 0.0001, ANOVA. Scale bars represent 20 mm in (A), 10 mm in (B)-(D), and 5 mm in enlarged panels of (D). See also Figures S1 and S2 and Movies S1, S2, S3, S4, and S5.
C EGFP L1-RFP DIC merged enlarged D
Figure 2. Sphkl-Dependent Phosphorylation of Sphingosine Promotes Endocytic Membrane Trafficking
(A) Phase-contrast images of WT or Sphk1~'~ MEFs treated with DMSO or 10 mM sphingosine (Sph) for 1 hr.
(B) WT or Sphk1~'~ MEFs pre-treated with DMSO or 2.5 mM PF-543for 30 min followed by DMSO or 10 mM Sph for an additional 4 hr.
(C) Sphk1-/- MEFs stably expressing Lamp1-RFP (L1-RFP) transiently transfected with EGFP, EGFP-Sphk1G82D (G82D), or EGFP-Sphk1 (WT) plasmids and treated with 7.5 mM Sph for 4 hr.
(D) Vacuole size of (C) is presented as the mean ± SEM from two independent experiments. n > 800 vacuoles per treatment. ****p < 0.0001, ANOVA. Scale bars represent 20 mm in (A) and (B), 10 mm in (C) and enlarged panels of (B), and 5 mm in enlarged panels of (C). See also Figure S2.
inhibitor, Bafilomycin A1 (BafA1), also impairs SK1-I-induced vacuole formation (Figure S3A).
Consistent with a defect in endosomal trafficking, EEA1-pos-itive early endosomes are slightly enlarged by SK1-I (Figure 3A). While the large vacuoles contain the LE/Lys marker Lamp1-RFP, the vacuoles are devoid of cathepsin B activity and fail to accumulate the acidic probe LysoSensor to suggest impaired function (Figures 3B and 3C). Moreover, the enlarged vacuoles contain endogenous Lamp1 that colocalizes with LE marker Rab7 and autophagy marker GFP-LC3 (Figure 3D). Transmission electron microscopy (TEM) of SK1-I-treated cells reveals the
accumulation of massive endosomes containing intraluminal vesicles (ILVs) that are indicative of LEs (Figures 3Ei and 3Eii). Consistent with an accumulation of GFP-LC3, SKI-I-treated cells accumulate large amphisome-like structures containing ILVs in the presence of undigested cytoplasmic material to suggest impaired autophagy (Figures 3Eiii and 3Eiv).
Next, we sought to investigate the sphingolipid profile of the vacuoles. Consistent with previous reports (Kapitonov et al., 2009; Paugh et al., 2008), SK1-I alters global sphingolipid metabolism by decreasing cellular S1P and increasing sphingosine, ceramide, monohexosylceramide, and sphingomyelin
A DMSO SK1-I B AF488-Dex MR-CathB DIC merged
Figure 3. Inhibition of Sphk1 by SK1-I Accumulates Stalled LEs and Amphisomes
(A) Fluorescence images of WT MEFs stably expressing Lamp1-RFP treated with 10 mM SK1-I for 6 hr and Immunostalned with antl-EEA1.
(B) Live-cell fluorescence images of WT MEFs co-treated with 100 mg/mL Alexa Fluor 488-Dextran (AF488-Dex) and 7.5 mM SK1-I for 4 hr, followed by MagicRed cathepsin B substrate (MR-CathB).
(C) Live-cell fluorescence images of WT MEFs treated with 7.5 mM SK1-I for 4 hr, followed by labeling with LysoSensor DND-189.
(D) Confocal images of WT MEFs stably expressing GFP-LC3 treated with 7.5 mM SK1-I for 6 hr and immunostained using anti-Lamp1 (red) and anti-Rab7 (cyan).
(E) TEM images of WT MEFs treated with 10 mM SK1-I for 6 hr. LE, late endosome; A, amphisome. Dotted outline indicate cytoplasmic material; arrowheads indicate intraluminal vesicles.
Scale bars represent 20 mm in (A), 10 mm in (B)-(C), 15 mm in (D), 5 mm in (Ei), 1 mm in (Eii)—(Eiv), 7.5 mm in enlarged panels of (D), and 0.5 mm in enlarged panels of (Eiii) and (Eiv). See also Figure S3.
(Figure S4). However, localized changes of sphingolipids can dynamically alter the biophysical properties of membranes (e.g., membrane curvature, lipid raft assembly, and bilayer fusion) that are implicated in membrane trafficking (van Blitters-wijk et al., 2003). To examine sphingolipid changes within endo-cytic organelles, cell lysates were loaded onto a continuous OptiPrep gradient and fractionated by high-speed centrifugation
prior to immunoblotting and liquid chromatography-tandem mass spectrometry (LC-MS/MS) quantitative analysis of sphingolipids (Figure 4).
Consistent with the enlargement of LEs relatively devoid of contents (Figure 3E), Lamp1 and Rab7 are accumulated in lower density fractions 3 and 4 upon SK1-I treatment, compared to control (Figures 4A, S5A, and S5B). Moreover, a modest shift
Enlarged LE
Density:
Fraction #
Density:
Fraction #
Figure 4. SK1-I Accumulates Sphingolipids in Enlarged LEs
(A-H) WT (A-D) or Sphkl-'- (E-H) MEFs were treated with DMSO or 10 mM SK1-I for 4 hr, followed by subcellular fractionation. Protein expression was quantified by immunoblot and presented as percent enrichment in each fraction. Sphingolipids were quantified by LC-MS/MS and presented as pico-moles of lipid per fraction. (A and E) Lampl. (B and F) Ceramide. (C and G) Sphingomyelin. (D and H) Monohexosylceramide. See also Figures S4 and S5.
_ 2000
''S 1500
T3 ¡5
I s 1000-<s ^ " Ö £
^10000! c
= -j; 8000 « t> ä «
<3 _ m a)
C O. 4000
— DMSO
— SK1-I
^ 3000 c
i. o ® 'ti
^ a 2000
« 4= ° &
§ _ 1000 £ iE
12 3 4
^ 2000
<u 'Hi 1500 "D ¡5
<S Ô. « Ö E
9 10 11 12
^10000 c
= lg 8000 « t3
E i 6000
= a. 4000
5 6 7 8 9 10 11 12 Fraction #
^ 3000 c
1. O ® 'ti
^ ra 2000 <u <í=
§ _ 1000 £ i
9 10 11 12
5 6 7 Fraction #
9 10 11 12
— DMSO SM SM
12 3 4
5 6 7 8 9 10 11 12 Fraction #
12 3 4
in Rab5 to lower density fractions 7 and 8 by SK1-I is consistent with the slight enlargement of early endosomes (EEs) (Figures 3A and S5C). In contrast, the subcellular distribution of mitochondrial marker Hsp60 is not altered by SK1-I (Figure S5D). Interestingly, the enlarged LE fraction accumulates ceramide, monohexosylceramide, and sphingomyelin (fractions 3 and 4; Figures 4B-4D). Moreover, ceramide and sphingomyelin are slightly increased in the dilated EE fraction of SK1-I-treated cells (fractions 7 and 8; Figures 4B and 4C). Surprisingly, sphingosine does not accumulate in the enlarged LE fraction of SK1-I-treated cells but rather displays a sharp enrichment in fraction 9, which overlaps with endosomal and plasma membrane markers and is consistent with stalled endocytic trafficking (Figures S5E and S5F).
Despite comparable uptake of the inhibitor (Figure S5G), Sphk1-deficient MEFs do not exhibit a significant shift in LE markers to lower density fractions or accumulate sphingolipids after treatment with SK1-I (Figures 4E-4H and S5H-S5J). Inter-
estingly, the distribution of Lamp1 is slightly shifted toward lower density fractions in vehicle-treated Sphk1~'~ cells compared to WT (compare black lines in Figures 4A and 4E). Moreover, Sphk1-deficient cells appear to mildly accumulate sphingolipids under basal conditions (compare the black lines in Figures 4B-4D versus those in Figures 4F-4H). It is tempting to speculate that the loss of Sphk1 alters the lipid profile of LE/Lys. However, as Sphk1~'~ MEFs appear to maintain fluid-phase endocytosis (Figure 1B), and Sphk1-null mice are viable with no apparent abnormalities (Allende et al., 2004), the effect of Sphk1 deficiency on basal endosomal function requires further study. Collectively, these results demonstrate that SK1-I decreases S1P, accumulates sphingosine, and induces the Sphk1-dependent accumulation of sphingosine metabolites in enlarged LEs.
SK1-I Induces Autophagy and Blocks Autophagic Flux
The endocytic pathway is critical for the maintenance of autophagic flux (Razi et al., 2009). Although the accumulation of dysfunctional LEs and amphisomes suggests impaired autopha-gic flux, autophagic activity was assessed by immunoblot for LC3-II, a widely used marker of autophagosomes, and the autophagic substrate p62/SQSTM1 (Klionsky et al., 2012). Autophagy induction is observed as the increase in LC3-II in the presence of lysosome inhibitors compared to Lys inhibitors alone. Indeed, SK1-I significantly increases LC3-II in the presence of the lysosomal protease inhibitors E64d, pepstatin A, and leupep-tin (E/P/L), compared to E/P/L alone, to indicate induced auto-phagosome formation (Figures 5A and 5B). Autophagic flux is indicated by an increase in LC3-II in the presence versus absence of Lys inhibitors. Indeed, LC3-II is further accumulated by the presence of Lys inhibitors in SK1-I-treated cells at 6 hr and 24 hrto indicate induced autophagic flux (Figures 5Aand 5C). In contrast, autophagic flux is impaired at 12 hr of SK1-I treatment (Figures 5A and 5C). The accumulation of p62 and a high-molecular-weight p62 oligomer (approximately 250 kDa) at 6 hr and 12 hr of SK1-I that is unaffected by E/P/L is consistent with
— DMSO mHexCer
5 6 7 8 9 10 11 12 Fraction #
A 6 h 12 h 24 h
SK1-I: - -+ +- -+ +--+ + E/P/L: - +- +- + - + - + - +
50 250 150 100 75
to —
OL LJÜ
LJÜ +
LC3-I LC3-II
p62 Short exposure p62 oligomer
p62 Long exposure p-actin
« 0.8H &
T3 0.4-| <u u
£ 0 2H
¿¡J2 4-
re > 3
merged
V. y • . t » t * ■f %
> V * > » V * k - ■ * -v-• ;
-î-f-î-
6 h 12 h 24 h
□ DMSO
□ E/P/L ■ SK1-I E3SK1-I + E/P/L
= 0.2-
S -a 0.1-
■a .c
= -0.1-<
n ■ * 1 s i * 1 ■f-
i j i L
12 h 24 h
merged
Figure 5. SK1-I Induces Autophagy and Blocks Autophagic Flux
(A-D) Atg5+/+ MEFs were treated with DMSO or 10 mM SK1-I for the indicated durations. (A) E64d (10 mM), pepstatin A (10 mg/mL~1), and leupeptin (1 mg/mL) (E/P/L) were added during the final 6 hr of treatment. Lysates were analyzed by immunoblotfor LC3, p62, and p-actin. (B) SK1-I-induced autophagy and (C) SK1-I-induced autophagic flux were calculated as described in the Experimental Procedures and are presented as mean ± SDfrom three independent experiments. (D) Quantification of p62 levels normalized to p-actin. Mean ± SD from three independent experiments.
(legend continued on next page)
impaired autophagic flux (Figure 5D). In contrast, the restoration of p62 to basal levels at 24 hr of SK1-I with accumulation by E/P/L indicates restored autophagic flux (Figure 5D). The discrepancy in autophagic flux at 6 hr, as determined by LC3-II and p62, may be due to autophagic flux in the absence of p62-associated cargo, autophagy-independent regulation of p62, or altered LC3-II delipidation.
Tandem-fluorescent LC3 (tf-LC3; RFP-EGFP-LC3) was used as an additional approach to monitor autophagic flux. Auto-phagosomes (RFP+EGFP+, or "yellow" puncta) and degra-dative amphisomes/autolysosomes (RFP+EGFP~, or "red" puncta) can be distinguished by the pH-sensitive quenching of EGFP. Consistent with immunoblotting, 6-hr treatment with SK1-I induces autophagosome formation, as RFP+EGFP+ LC3 puncta are accumulated by SK1-I in the presence of E/P/L compared to E/P/L alone (Figure 5E). As the phenotype is similar in cells treated with SK1-I in the presence or absence of E/P/L, autophagic flux appears to be minimal at 6 hr (Figure 5E). The accumulation of large RFP+EGFP~ LC3 structures by SK1-I is most likely a result of the fusion of previously functional LEs and amphisomes (Figure 5E). Consistently, prolonged treatment with SK1-I restores autophagic flux, as SK1-I-treated cells display an increase in RFP+EGFP~ LC3 puncta, indicative of degradative amphisomes/autolysosomes, that is impaired by lysosomal inhibition (Figure 5F). Collectively, these results indicate that SK1-I induces autophagosome biogenesis and dynamically regulates autophagic flux to impair and restore degradation.
Autophagy Machinery Coordinates with Lys Biogenesis to Clear Enlarged LEs in Response to Sphkl Inhibition
To determine the role of autophagy during Sphk1 inhibition, WT and autophagy-deficient Atg5~'~ or Atg3~'~ MEFs were treated with SK1-I. Atg5 and Atg3 are required for the lipidation of LC3-I to LC3-II during autophagosome biogenesis. Loss of autophagy does not alter vacuole formation by SK1-I. However, although WT cells clear SK1-I-induced vacuoles by 24 hr, vacuoles are significantly sustained in Atg5- or Atg3-deficient MEFs (Figure 6A). To verify this result, GFP-Atg5 or mutant Atg5K130R, which is unable to rescue LC3 lipidation, was stably expressed in Atg5~'~ MEFs. Although GFP-Atg5 rescues vacuole clearance, Atg5K130R fails to do so to indicate that LC3 lipidation is required for clearance (Figure 6B). To assess the role of autophagic flux in vacuole clearance, cells were treated with SK1-I for 6 hr, followed by the addition of lysosomal protease inhibitors. Similar to the loss of Atg5 or Atg3, Lys inhibition prolongs vacuoles in WT cells to suggest that autophagic flux is required for clearance (Figure 6B).
To examine the mechanism of vacuole clearance, MEFs stably expressing Lamp1-RFP were treated with SK1-I and monitored by time-lapse imaging. During vacuole clearance, small Lamp1-positive granule-like structures are recruited to the vacuole membrane prior to the gradual condensation of vacuoles
(Movie S6). This mechanism is distinct from autophagic Lys reformation in which proto-lysosomal tubules extend and bud from enlarged Lys (Yu et al., 2010). TEM imaging after vacuole clearance reveals the accumulation of significantly smaller vesicles containing multi-lamellar membranes (Figure 6C). As the enlarged LEs were initially devoid of membranes (Figure 3E), internalization of the limiting membrane appears to lead to vacuole shrinkage. Further study is required to uncover the mechanism of this process.
Sphingosine and ceramide have been implicated in the induction of autophagy and regulation of autophagic flux (Young et al., 2013). To determine whether sphingosine metabolism to ceramide is required for the autophagy-mediated clearance of SK1-I-induced vacuoles, ceramide synthase was inhibited by fumonisin B1 (FB1). Interestingly, FB1 significantly delays the clearance of SK1-I-induced vacuoles (Figure 6D). Of note, vacuole clearance is slightly accelerated in Sphk1+'+ MEFs (1220 hr; Figure 6D; data not shown) compared to Atg5+/+ MEFs (16-24 hr; Figures 6A-6C) to indicate that the precise timing of vacuole formation and clearance is likely dependent on genetic background. Importantly, LC3-II is increased upon co-treatment with SK1-I and FB1 compared to SK1-I alone (Figure 6E). Moreover, LC3-II accumulation appears to result from impaired flux, as LC3-II does not further accumulate with E/P/L. To assess the role of ceramide in the clearance of Sph-induced vacuoles in Sphk1~'~ MEFs, cells were treated with Sph in the presence of FB1. Consistent with SK1-I, inhibition of ceramide synthase by FB1 accumulates Sph-induced vacuoles in Sphk1~'~ MEFs (Figure 6F). A similar autophagic phenotype is observed where FB1 increases LC3-II in the presence of Sph compared to Sph alone (Figure 6G). Moreover, FB1 impairs Sph-induced autophagic flux as LC3-II accumulation is insensitive to E/P/L compared to Sph treatment alone (Figure 6G). More detailed studies are required to dissect the dual effects of Sph and Sph metabolites in autophagosome formation and degradation.
Although lysosomal protease activity is required for the efficient clearance of SK1-I-induced vacuoles (Figure 6B), cathepsin B activity is significantly suppressed by SK1-I after a few hours of treatment (Figure 3B). Therefore, we hypothesized that new Lys biogenesis may facilitate clearance. Transcription factor EB (TFEB) is a master regulator of autophagy and Lys-associated genes (Settembre and Medina, 2015). As lysosomal Sph can activate TFEB (Hoglinger et al., 2015), we examined TFEB localization during the clearance of SK1-I-induced vacuoles. Indeed, TFEB translocates to the nucleus prior to vacuole clearance (Figures 7A and 7B). Consistently, the expression of Lamp1 and Rab7, two TFEB target genes, are increased during vacuole clearance (Figure 7C).
Notably, the nuclear translocation of TFEB in response to SK1-I is markedly suppressed in Atg5-deficient cells, which have impaired vacuole clearance (Figures 7A and 7B). To monitor Lys in Atg5~'~ MEFs during SK1-I treatment,
(E and F) Confocal images of MEFs stably expressing RFP-EGFP-LC3 treated with DMSO or 7.5 mM SK1-I for (E) 6 hr or (F) 16 hr. E/P/L was added during the final 6 hr of treatment.
Data are presented as the mean ± SD from three independent experiments. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001, ANOVA. ns, not significant. Scale bars represent 15 mm. See also Figure S6.
Vehicle FBI
^ ß-actin
LC3-II/ß-actin: 1.0 1.6 3.3 4.4 1.5 2.2 3.9 2.4
Figure 6. The LC3 Conjugation Machinery, Lysosomal Proteases, and Ceramide Synthase Are Required for the Clearance of Enlarged LEs Induced by SK1-I
(A) Phase-contrast images of WT, Atg5~'~, and Atg3~'~ MEFs treated with 10 mM SK1-I for the indicated time course.
(B) Phase-contrast images of WT, Atg5~'~, Atg5~'~ GFP-Atg5, or Atg5~'~ Atg5K130R-HA (hemagglutinin) MEFs treated with 10 mM SK1-I for 6 hr, followed by the addition of E64d, pepstatin A, and leupeptin (E/P/L) for an additional 18 hr.
(C) TEM images of WT MEFs treated with 10 mM SK1-I for 16 hr.
(D) Phase-contrast images of Sphk1+/+ MEFs pre-treated with 50 mM fumonisin B1 (FB1) for 30 min prior to the addition of 10 mM SK1-I for 4 hr or 12 hr.
(E) Immunoblot of Sphk1+/+ MEFs pre-treated with 50 mM fumonisin B1 (FB1)for30 min prior to the addition of 10 mM SK1-I for 12 hr. Where indicated, E/P/L was added during the last 6 hr of treatment.
(F) Phase-contrast images of Sphk1~'~ MEFs pre-treated with 50 mM FB1 for 30 min prior to the addition of 10 mM Sph for 4 hr or 12 hr.
(G) Immunoblot of Sphk1~'~ MEFs pre-treated with 50 mM FB1 for 30 min prior to the addition of 10 mM Sph for 12 hr. Where indicated, E/P/L was added during the last 6 hr of treatment.
Scale bars represent 20 mm in (A), (B), (D), and (F); 5 mm in (Ci); 1 mm in (Cii) and (Ciii); and 0.5 mm in the enlarged panel of (Ciii). See also Movie S6.
A GFP-TFEB DIC merged B WT Atg5-I-
Figure 7. The Clearance of SKI-I-Induced Vacuoles Occurs Concomitantly with the Activation of TFEB-Mediated Lys Biogenesis and Is Dependent on Atg5
(A) Live-cell fluorescence Images of WT and Atg5~'~ MEFs stably expressing GFP-TFEB treated with 7.5 mM SK1-I for 16 hr. White and yellow dotted lines represent outlines of nuclei lacking or containing GFP-TFEB, respectively.
(B) WTor Atg5~'~ MEFs were treated with 10 mM SK1-I for 24 hr and fractionated to obtain cytoplasmic (Cyt), nuclear binding proteins (NBP), or insoluble nuclear proteins (INB). Immunoblot forTFEB, a-tubulin, and poly(ADP-ribose) polymerase (PARP).
(C) WT MEFs were treated with DMSO or 10 mM SK1-I for the indicated time points. Cell lysates were analyzed by immunoblot for Lampl, Rab7, and p-actin.
(D) WT and Atg5~'~ MEFs were pulsed with 100 mg/mL AF488-Dex for 1 hr followed by a 2.5 hr chase in dextran-free DMEM to label Lys. Live-cell fluorescent images of cells treated with DMSO or 10 mM SK1-I for 4 hr followed by MagicRed cathepsin B substrate (MR-CathB).
Scale bars represent 20 mm in (A) and 10 mm in (D).
lysosomal compartments were pulse-chase labeled with 10 kDa Alexa Fluor 488-dextran. Unexpectedly, while the SK1-I-induced vacuoles of WT cells were labeled by fluorescent dextran, fluorescent dextran was maintained in punctate structures and not enlarged vacuoles in SK1-I-treated Atg5~'~ MEFs (Figure 7D). Loss of Atg5 has been reported to impair the retromer-mediated retrieval of mannose-6-phosphate receptorto result in hybrid LE/ Lys (Peng et al., 2014). Although beyond the scope of this paper, we propose that the hybrid LE/Lys lacks the machinery required for association with vacuoles for clearance. Given that Lys are not consumed in Atg5-deficient cells by autophagy, TFEB is not activated for Lys biogenesis. These results emphasize the complex crosstalk between the endocytic pathway and autophagy machinery to suggest a potential mechanism for the clearance of dysfunctional LEs.
DISCUSSION
The function of Sphkl in membrane trafficking appears to be highly conserved. Here, we provide evidence for the role of Sph and Sphkl in endocytic membrane trafficking. Our results suggest that SK1-I enters the cell via lipid-raft-mediated endocy-tosis to imply that, like Sph, the inhibitor is largely restricted to membranes. By overloading endocytic membranes with exogenous Sph or Sph mimetic while inhibiting Sphkl activity (e.g., by Sphk1aG82D, SK1-I, or PF-543), Sphkl is stabilized at LE membranes to promote membrane fusion and reveal a role for Sph and Sphkl in endosomal fusion. We believe that small-molecule inhibitors like PF-543 fail to selectively accumulate Sph and Sphk1 at endosomal membranes and, thus, do not induce vacuoles. While we recognize that our current system is inherently
artificial, we hope that future technologies will aid in monitoring sphingosine metabolism during dynamic processes such as endosomal trafficking.
Sph isafusogeniclipid that can promote bilayer fusion (Castro et al., 2014; Goni et al., 2014). Moreover, perturbation of the local membrane environment by Sph can enhance SNARE-mediated membrane fusion by the direct activation of SNARE proteins such as synaptobrevin (Darios et al., 2009). Sph also regulates the electrostatic interaction of vesicles and SNARE vesicle-associated membrane protein 2 (VAMP2) by neutralization of negatively charged membranes (Williams et al., 2009). As the core machinery for membrane fusion in neuronal and endosomal systems are highly conserved (Antonin et al., 2000), the identification of endosomal SNARE proteins as Sph interactors is an exciting future direction.
Unexpectedly, catalytically inactive Sphk1aG82D enhances the fusion of Lamp1-positive vacuoles induced by SK1-I or Sph to suggest that Sphk1 may promote membrane fusion independent of catalytic activity. A possible explanation for this result is that Sphk1aG82D binds endogenous Sph at endocytic vesicles to limit its metabolism and accumulate the fusogenic lipid. Alternatively, Sphk1 may facilitate membrane fusion independent of kinase activity. Sphk1 assembles dimers that contain a concave, positively charged groove separating two lipid-binding loop-1 (LBL-1) motifs (Adams et al., 2016). Docking of Sphk1 dimers to convex, negatively charged membranes orients the LBL-1 motifs to the membrane for substrate extraction with projection of the C terminus away from the surface (Adams et al., 2016). Notably, LBL-1 is implicated in the curvature-sensitive recruitment of Sphk1 to negatively charged membranes during endocytosis (Shen et al., 2014)to suggest that dimerization may regulate Sphk1 recruitment to endo-cytic membranes. We propose that the Sph-mediated recruitment ofSphk1 dimers to endosomal membranes primes membranes for fusion by acting as a membrane tether and/or recruiting endocytic factors via its C terminus. As Sph-induced SNARE complex assembly was determined using purified synaptic vesicles, the presence of Sphk1 cannot be excluded (Darios et al., 2009). Further investigation is required to test this hypothesis.
Strikingly, vacuoles do not accumulate under basal conditions in Sphk1~'~ MEFs. We propose that endocytic sphingolipid metabolism is tightly regulated to limit Sph accumulation and protonation during endosomal maturation, leading to membrane destabilization (Goni et al., 2014). We propose that Sphk2 compensates for Sphk1 to prevent the accumulation of Sph and LE fusion. While Sphk1-deficient mice lack any apparent abnormalities (Mizugishi et al., 2005), Sphk1/2 double-knockout mice are embryonic lethal with defective vascularization and highly vacuolated endothelial cells (Mizugishi et al., 2005). Whether impaired endocytic trafficking and/or Sph accumulation contribute to this phenotype is an interesting avenue for further investigation.
Our results indicate that autophagy, ceramide synthase activity, and Lys biogenesis aid in the clearance of dysfunctional LEs. First, SK1-I induces autophagy while dynamically modulating autophagic flux. SK1-I was observed to suppress mammalian target of rapamycin complex 1 (mTORC1), a well-established negative regulator of autophagy, thus providing a potential mechanism for autophagy induction (Figure S6A). Intriguingly, mTOR associates with sphingolipid-enriched vacuoles to suggest that sphingolipids
may regulate mTOR localization (Figure S6). Moreover, sphingolipids are established regulators of autophagy (Young et al., 2013), and the recruitment of Sphk1 to endosomes, autophagosomes, and amphisomes has recently been linked to autophagosome biogenesis and degradation (Moruno Manchon et al., 2015, 2016). Additionally, ceramide directly interacts with LC3-II to mediate cargo recruitment to autophagosomes for efficient degradation (Sentelle et al., 2012; Dany et al., 2016). We found that inhibition of ceramide synthase by FB1 impairs autophagic flux to sustain SK1-I- and Sph-induced vacuoles (Figures 6D-6G). Further dissection of sphingolipid metabolites and Sphk1 in the regulation of autophagy is of great interest.
Finally, we demonstrate that Lys biogenesis and restored autophagic flux coordinate for the clearance of the enlarged LEs induced by SK1-I. As LysoSensor-positive puncta persist in the presence of SK1-I-induced vacuoles after a few hours of treatment (Figure 3C), autophagic flux may temporarily be maintained (6 hr; Figure 5C). However, we hypothesize that the continued induction of autophagy (Figure 5B) and the sequestration of mTOR at vacuoles (Figure S6B) will deplete the cell of functional Lys by impairment of mTORC1-dependent autophagic Lys reformation. Consequently, the activation of TFEB for Lys biogenesis restores autophagic flux and allows for vacuole clearance by an undetermined mechanism.
Our phenotype is remarkably similar to that observed in the lysosomal storage disorder Niemann-Pick type C disease (NP-C). NP-C is a sphingosine storage disease where the lysosomal accumulation of Sph impairs calcium storage, leading to secondary lipid accumulation (Lloyd-Evans et al., 2008). Similar to our phenotype, NP-C fibroblasts accumulate LEs enriched in sphingolipids and cholesterol (Sarkaret al., 2014). As Sphk activity is suppressed in NP-C (Lee et al., 2014), we propose that a failure to recruit Sphk to endosomal membranes may drive Sph accumulation during the pathogenesis of NP-C. Despite functional Lys, NP-C cells have impaired autophagic flux due to the failure to recruit autophagic SNARE machinery to LEs (Sarkar et al., 2013). Autophagy induction restores autophagic flux through the direct fusion of autophagosomes with Lys and has been proposed as a therapeutic strategy (Sarkar et al., 2013). Collectively, our studies implicate Sph and Sphk1 in endocytic membrane trafficking and may contribute to the development of new therapeutic strategies to restore endocytic membrane trafficking in sphingolipid storage disorders.
EXPERIMENTAL PROCEDURES Isolation and Immortalization of MEFs
Sphk1~'~ mice were kindly provided by Dr. Richard Proia (National Institute of Diabetes and Digestive and Kidney Diseases [NIDDK], NI H) (Allende et al., 2004). Sphk1~'~ mice were backcrossed with C57BL/6J mice, and heterozygous mice were intercrossed for MEF isolation from littermate day-12.5 embryos. MEFs were genotyped by PCR amplification of genomic DNA isolated from the embryo heads, as described previously (Allende et al., 2004). Passage-two primary MEFs were immortalized by transduction with recombinant retrovirus encoding simian virus 40 (SV40) large T-antigen and selected by continuous passage.
Live-Cell Imaging
Cells were seeded in NUNC Lab-Tek II chamber slides (Thermo Scientific) and imaged in a humidified live-cell chamber equilibrated to 5% CO2 and 37°C
using an Olympus IX81 deconvolution microscope equipped with SlideBook Software (Version 5.0.0.14, Intelligent Imaging Innovations). To monitor fluid-phase endocytosis, cells were treated with 0.1 mg/mL Alexa Fluor 488-dextran (10 kDa; Life Technologies). To label acidic organelles, cells were incubated with LysoSensor DND-189 (1 mM; Life Technologies) according to the manufacturer's protocol. For detection of cathepsin B activity, cells were incubated with MagicRed Cathepsin B(ImmunoChemistry Technologies) per the manufacturer's protocol.
Immunofluorescence Imaging
Cells were seeded in NUNC Lab-Tek II chamber slides (Thermo Scientific), treated as indicated, and fixed with 2% paraformaldehyde in PBS. For immu-nostaining, cells were permeabilized with 100 mg/mL digitonin in PBS, washed in PBS, and blocked with 5% normal goat serum prior to primary antibody incubation overnight followed by secondary antibody for at room temperature for 1 hr and mounting with ProLong Gold mounting medium (Invi-trogen). Antibody information is provided in the Supplemental Information. Confocal images were obtained using a Leica TCS SP8 confocal microscope (Leica Microsystems) and analyzed using Imaris 3D/4D Image Analysis software (Bitplane). All other imaging was performed using an Olympus IX81 de-convolution microscope. Vacuole size (a.u.; pixels per vacuole) was manually quantified from differential interference contrast (DIC) images using SlideBook software. Data are presented as the mean ± SEM.
Electron Microscopy
Cells were seeded on Thermanox coverslips (Electron Microscopy Sciences), followed by fixation in Karnovsky's fixative (1.5% paraformaldehyde, 3% glu-teraldehyde in 0.1 M phosphate buffer, pH 7.3) containing 4.2 mM ethylenegly-col-bis(b-aminoethyl)-N,N,N',N'-tetracetic acid for 30 min at room temperature followed by 1 hr at 4°C. Cells were washed in 0.1 M phosphate buffer and 0.1 M sodium cacodylate buffer (pH 7.36) and post-fixed in 1% osmium tetroxide, 1.5% potassium ferrocyanide in 0.1 M sodium cacodylate buffer for 30 min at 4°C. After thoroughly washing with water, cells were incubated with 1% uranium acetate at room temperature for 1.5 hr, followed by dehydration in a graded series of ethanol and propylene oxide. Coverslips were embedded in Embed 812 resin (Electron Microscopy Sciences), sectioned at a thickness of 70-90 nm, and mounted on mesh copper grids followed by staining with uranyl acetate and lead citrate. Images were obtained using a JEM 1400 digital capture transmission electron microscope (JEOL).
Subcellular Fractionation and Mass Spectrometry
MEFs were treated with DMSO or 10 mM SK1-I for 4 hr, washed twice with ice-cold PBS, scraped in ice-cold PBS, and pelleted by centrifugation (250 x g, 4°C). Cell pellets were resuspended in 0.9 mL of homogenization buffer (0.25 M sucrose, 140 mM NaCl, 1 mM EDTA, 20 mM Tris-HCl [pH 8.0]) containing protease and phosphatase inhibitors, passed through a 27-gauge syringe, and centrifuged (500 x g, 4°C) to obtain post-nuclear supernatant (PNS). Protein concentration was determined by bicinchoninic acid (BCA) assay, and equal amounts of protein were loaded onto 7.2 mL of a continuous 5%-20% OptiPrep (Sigma-Aldirch D1556) gradient prepared in homogenization buffer. Samples were centrifuged at 90,000 x g for 20 hr at 4°C using a SW-41 swing rotor. Twelve fractions were collected from the top of the gradient by aspiration of the meniscus. An aliquot of each fraction was removed for immunoblot analysis, and the resultant fractions were immediately mixed with ice-cold high performance liquid chromatography (HPLC)-grade methanol (1:1) and submitted to the Virginia Commonwealth University Lipidomics Core Facility for LC-MS/MS. Data are presented as the total lipid species (picomoles) per fraction. Equal volumes of each fraction were analyzed by immunoblot. Expression was quantified using Image Studio Lite (Version 5.0.21) and presented as the percentage of protein enrichment in each fraction.
Mass Spectrometry of Cellular Lipids
Cells were seeded in a six-well plate and harvested in 500 mL of ice-cold PBS containing phosphatase and protease inhibitors. A portion of the cell suspension was mixed with ice-cold HPLC-grade methanol (3:10) and submitted to the Virginia Commonwealth University Lipidomics Core Facility for LC-MS/ MS. The remaining cell suspension was sonicated and centrifuged prior to
protein quantification by BCA assay. Data were generated from three replicates and are presented as the mean picomoles per milligram of protein ± SD.
Western Blotting
Cells were washed with ice-cold PBS and scraped in 2x RIPA buffer containing protease and phosphatase inhibitor cocktails. Nuclear protein isolation was performed using the QIAGEN Qproteome Nuclear Protein Kit according to the manufacturer's protocol, with the addition of phosphatase inhibitors to all buffers. Proteins were separated by SDS-PAGE and transferred to poly-vinylidene fluoride (PVDF), as described previously (Young et al., 2012). Detailed antibody information is provided in the Supplemental Information. Blots were developed by chemiluminescent substrate (Thermo Scientific) using the LI-COR C-DiGit Blot Scanner or infrared imaging using the LI-COR Odyssey CLx. Immunoblots was quantified using Image Studio Lite (version 5.0.21). LC3-II and p62 were normalized to ß-actin, and normalized LC3-II was used for autophagic calculations, as described previously (Tooze et al., 2015). In brief, basal autophagic flux was calculated as [(DMSO + E/P/L) -(DMSO)]; induced autophagy was calculated as [(SK1-I + E/P/L) - (DMSO + E/P/L)]; and induced autophagic flux was calculated by [(SK1-I + E/P/L) — (SK1-I)].
Statistical Analysis
Data are representative of at least two independent experiments. Statistical significance was evaluated by two-tailed Student's t test or ANOVA with Tukey's post hoc test for multiple comparisons using GraphPad Prism software. Thresholds of significance are indicated in the figure legends.
SUPPLEMENTAL INFORMATION
Supplemental Information includes Supplemental Experimental Procedures, six figures, and six movies and can be found with this article online at http:// dx.doi.org/10.1016/j.celrep.2016.10.019.
AUTHOR CONTRIBUTIONS
Conceptualization, M.M.Y. and H.-G.W.; Methodology, M.M.Y., Y.T., and H.-G.W.; Investigation, M.M.Y., T.E.F., and Y.T.; Writing - Original Draft, M.M.Y. and H.-G.W.; Writing - Review & Editing, M.M.Y., J.K.Y., T.E.F., M.K., and H.-G.W.; Supervision, M.K. and H.-G.W.; Funding Acquisition, H.G.W.
ACKNOWLEDGMENTS
This work was funded in part by an NIH Program project grant (P01CA171983) and the Lois High Berstler Fund and Four Diamonds Fund of Penn State College of Medicine. Confocal images were generated using the Leica SP8 microscope (NIH Shared Instrumentation grant S10OD010756-01A1) located in the Penn State Microscopy Imaging Core Facility. LC-MS/MS data were generated by the Virginia Commonwealth University Lipidomics and Metabolomics Core (NIH-NCI Cancer Center Support grant P30CA016059 to the VCU Mas-sey Cancer Center and NIH Shared Instrumentation grant S10RR031535).
Received: April 6, 2016 Revised: September 15, 2016 Accepted: October 5, 2016 Published: November 1, 2016
REFERENCES
Acharya, U., Patel, S., Koundakjian, E., Nagashima, K., Han, X., and Acharya, J.K. (2003). Modulating sphingolipid biosynthetic pathway rescues photoreceptor degeneration. Science 299, 1740-1743.
Acharya, U., Mowen, M.B., Nagashima, K., and Acharya, J.K. (2004). Ceram-idase expression facilitates membrane turnover and endocytosis of rhodopsin in photoreceptors. Proc. Natl. Acad. Sci. USA 101, 1922-1926.
Adams, D.R., Pyne, S., and Pyne, N.J. (2016). Sphingosine kinases: emerging structure-function insights. Trends Biochem. Sci. 41, 395-409. Alexaki, A., Gupta, S.D., Majumder, S., Kono, M., Tuymetova, G., Harmon, J.M., Dunn, T.M., and Proia, R.L. (2014). Autophagy regulates sphingolipid levels in the liver. J. Lipid Res. 55, 2521-2531.
Allende, M.L., Sasaki, T., Kawai, H., Olivera, A., Mi, Y., van Echten-Deckert, G., Hajdu, R., Rosenbach, M., Keohane, C.A., Mandala, S., et al. (2004). Mice deficient in sphingosine kinase 1 are rendered lymphopenic by FTY720. J. Biol. Chem. 279, 52487-52492.
Amaya, C., Fader, C.M., and Colombo, M.I. (2015). Autophagy and proteins
involved in vesicular trafficking. FEBS Lett. 589, 3343-3353.
Antonin, W., Holroyd, C., Fasshauer, D., Pabst, S., Von Mollard, G.F., and
Jahn, R. (2000). A SNARE complex mediating fusion of late endosomes
defines conserved properties of SNARE structure and function. EMBO J. 19,
6453-6464.
Camoletto, P.G., Vara, H., Morando, L., Connell, E., Marletto, F.P., Giustetto, M., Sassoe-Pognetto, M., Van Veldhoven, P.P., and Ledesma, M.D. (2009). Synaptic vesicle docking: sphingosine regulates syntaxin1 interaction with Munc18. PLoS ONE 4, e5310.
Castro, B.M., Prieto, M., and Silva, L.C. (2014). Ceramide: a simple sphingolipid with unique biophysical properties. Prog. Lipid Res. 54, 53-67. Chan, J.P., and Sieburth, D. (2012). Localized sphingolipid signaling at presynaptic terminals is regulated by calcium influx and promotes recruitment of priming factors. J. Neurosci. 32, 17909-17920.
Chan, J.P., Hu, Z., and Sieburth, D. (2012). Recruitment of sphingosine kinase to presynaptic terminals by a conserved muscarinic signaling pathway promotes neurotransmitter release. Genes Dev. 26, 1070-1085. Dany, M., Gencer, S., Nganga, R., Thomas, R.J., Oleinik, N., Baron, K.D., Szulc, Z.M., Ruvolo, P., Kornblau, S., Andreeff, M., and Ogretmen, B. (2016). Targeting FLT3-ITD signaling mediates ceramide-dependent mitophagy and attenuates drug resistance in AML. Blood, Published online August 18, 2016. http://dx.doi.org/10.1182/blood-2016-04-708750. Darios, F., Wasser, C., Shakirzyanova, A., Giniatullin, A., Goodman, K., Munoz-Bravo, J.L., Raingo, J., Jorgacevski, J., Kreft, M., Zorec, R., et al. (2009). Sphingosine facilitates SNARE complex assembly and activates synaptic vesicle exocytosis. Neuron 62, 683-694.
Feng, Y., He, D., Yao, Z., and Klionsky, D.J. (2014). The machinery of macro-autophagy. Cell Res. 24, 24-41.
Goni, F.M., Sot, J., and Alonso, A. (2014). Biophysical properties of sphingosine, ceramides and other simple sphingolipids. Biochem. Soc. Trans. 42, 1401-1408.
Hannich, J.T., Umebayashi, K., and Riezman, H. (2011). Distribution and functions of sterols and sphingolipids. Cold Spring Harb. Perspect. Biol. 3, a004762.
Hannun, Y.A., and Obeid, L.M. (2008). Principles of bioactive lipid signalling: lessons from sphingolipids. Nat. Rev. Mol. Cell Biol. 9, 139-150. Hoglinger, D., Haberkant, P., Aguilera-Romero, A., Riezman, H., Porter, F.D., Platt, F.M., Galione, A., and Schultz, C. (2015). Intracellular sphingosine releases calcium from lysosomes. eLife 4, e10616.
Kapitonov, D., Allegood, J.C., Mitchell, C., Hait, N.C., Almenara, J.A., Adams, J.K., Zipkin, R.E., Dent, P., Kordula, T., Milstien, S., and Spiegel, S. (2009). Targeting sphingosine kinase 1 inhibits Akt signaling, induces apoptosis, and suppresses growth of human glioblastoma cells and xenografts. Cancer Res. 69, 6915-6923.
Kim, H.J., Zhong, Q., Sheng, Z.-H., Yoshimori, T., Liang, C., and Jung, J.U. (2012). Beclin-1-interacting autophagy protein Atg14L targets the SNARE-associated protein Snapin to coordinate endocytic trafficking. J. Cell Sci. 125, 4740-4750.
Klionsky, D.J., Abdalla, F.C., Abeliovich, H., Abraham, R.T., Acevedo-Arozena, A., Adeli, K., Agholme, L., Agnello, M., Agostinis, P., Aguirre-Ghiso, J.A., et al. (2012). Guidelines for the use and interpretation of assays for monitoring autophagy. Autophagy 8, 445-544.
Lamb, C.A., Dooley, H.C., and Tooze, S.A. (2013). Endocytosis and autophagy: Shared machinery for degradation. BioEssays 35, 34-45. Lee, H., Lee, J.K., Park, M.H., Hong, Y.R., Marti, H.H., Kim, H., Okada, Y., Otsu, M., Seo, E.-J., Park, J.-H., et al. (2014). Pathological roles of the VEGF/SphK pathway in Niemann-Picktype C neurons. Nat. Commun. 5,5514. Liang, C., Lee, J.S., Inn, K.-S., Gack, M.U., Li, Q., Roberts, E.A., Vergne, I., De-retic, V., Feng, P., Akazawa, C., and Jung, J.U. (2008). Beclinl-binding UVRAG targets the class C Vps complex to coordinate autophagosome maturation and endocytic trafficking. Nat. Cell Biol. 10, 776-787.
Lloyd-Evans, E., Morgan, A.J., He, X., Smith, D.A., Elliot-Smith, E., Sillence, D.J., Churchill, G.C., Schuchman, E.H., Galione, A., and Platt, F.M. (2008). Niemann-Pick disease type C1 is a sphingosine storage disease that causes deregulation of lysosomal calcium. Nat. Med. 14, 1247-1255. Mizugishi, K., Yamashita, T., Olivera, A., Miller, G.F., Spiegel, S., and Proia, R.L. (2005). Essential role for sphingosine kinases in neural and vascular development. Mol. Cell. Biol. 25, 11113-11121.
Moruno Manchon, J.F., Uzor, N.-E., Dabaghian, Y., Furr-Stimming, E.E., Fink-beiner, S., and Tsvetkov, A.S. (2015). Cytoplasmic sphingosine-1-phosphate pathway modulates neuronal autophagy. Sci. Rep. 5, 15213. Moruno Manchon, J.F., Uzor, N.-E., Finkbeiner, S., and Tsvetkov, A.S. (2016). SPHK1/sphingosine kinase 1-mediated autophagy differs between neurons and SH-SY5Y neuroblastoma cells. Autophagy 12, 1418-1424. Murrow, L., Malhotra, R., and Debnath, J. (2015). ATG12-ATG3 interacts with Alix to promote basal autophagic flux and late endosome function. Nat. Cell Biol. 17, 300-310.
Paugh, S.W., Paugh, B.S., Rahmani, M., Kapitonov, D., Almenara, J.A., Kordula, T., Milstien, S., Adams, J.K., Zipkin, R.E., Grant, S., and Spiegel, S. (2008). A selective sphingosine kinase 1 inhibitor integrates multiple molecular therapeutic targets in human leukemia. Blood 112, 1382-1391. Peng, J., Zhang, R., Cui, Y., Liu, H., Zhao, X., Huang, L., Hu, M., Yuan, X., Ma,
B., Ma, X., et al. (2014). Atg5 regulates late endosome and lysosome biogenesis. Sci. China Life Sci. 57, 59-68.
Razi, M., Chan, E.Y.W., and Tooze, S.A. (2009). Early endosomes and endosomal coatomer are required for autophagy. J. Cell Biol. 185, 305-321. Rohrbough, J., Rushton, E., Palanker, L., Woodruff, E., Matthies, H.J.G., Acharya, U., Acharya, J.K., and Broadie, K. (2004). Ceramidase regulates syn-aptic vesicle exocytosis and trafficking. J. Neurosci. 24, 7789-7803. Ruck, A., Attonito, J., Garces, K.T., Nunez, L., Palmisano, N.J., Rubel, Z., Bai, Z., Nguyen, K.C.Q., Sun, L., Grant, B.D., et al. (2011). The Atg6/Vps30/Beclin 1 ortholog BEC-1 mediates endocytic retrograde transport in addition to autophagy in C. elegans. Autophagy 7, 386-400.
Sarkar, S., Carroll, B., Buganim, Y., Maetzel, D., Ng, A.H.M., Cassady, J.P., Cohen, M.A., Chakraborty, S., Wang, H., Spooner, E., et al. (2013). Impaired autophagy in the lipid-storage disorder Niemann-Pick type C1 disease. Cell Rep. 5, 1302-1315.
Sarkar, S., Maetzel, D., Korolchuk, V.l., and Jaenisch, R. (2014). Restarting stalled autophagy a potential therapeutic approach for the lipid storage disorder, Niemann-Pick type C1 disease. Autophagy 10, 1137-1140. Sentelle, R.D., Senkal, C.E., Jiang, W., Ponnusamy, S., Gencer, S., Selvam, S.P., Ramshesh, V.K., Peterson, Y.K., Lemasters, J.J., Szulc, Z.M., et al. (2012). Ceramide targets autophagosomes to mitochondria and induces lethal mitophagy. Nat. Chem. Biol. 8, 831-838.
Settembre, C., and Medina, D.L. (2015). TFEB and the CLEAR network. Methods Cell Biol. 126, 45-62.
Shen, H., Giordano, F., Wu, Y., Chan, J., Zhu, C., Milosevic, I., Wu, X., Yao, K., Chen, B., Baumgart, T., et al. (2014). Coupling between endocytosis and sphingosine kinase 1 recruitment. Nat. Cell Biol. 16, 652-662. Tooze, S.A., Dooley, H.C., Jefferies, H.B.J., Joachim, J., Judith, D., Lamb,
C.A., Razi, M., and Wirth, M. (2015). Assessing mammalian autophagy. Methods Mol. Biol. 1270, 155-165.
van Blitterswijk, W.J., van der Luit, A.H., Veldman, R.J., Verheij, M., and Borst, J. (2003). Ceramide: second messenger or modulator of membrane structure and dynamics? Biochem. J. 369, 199-211.
Williams, D., Vicôgne, J., Zaitseva, I., McLaughlin, S., and Pessln, J.E. (2009). Evidence that electrostatic interactions between vesicle-associated membrane protein 2 and acidic phospholipids may modulate the fusion of transport vesicles with the plasma membrane. Mol. Biol. Cell 20, 4910-4919. Yonamine, I., Bamba, T., Nirala, N.K., Jesmin, N., Kosakowska-Cholody, T., Nagashima, K., Fukusaki, E., Acharya, J.K., and Acharya, U. (2011). Sphingo-sine kinases and their metabolites modulate endolysosomal trafficking in photoreceptors. J. Cell Biol. 192, 557-567.
Young, M.M., Takahashi, Y., Khan, O., Park, S., Hori, T., Yun, J., Sharma, A.K., Amin, S., Hu, C.-D., Zhang, J., et al. (2012). Autophagosomal membrane
serves as platform for intracellular death-inducing signaling complex (iDISC)-mediated caspase-8 activation and apoptosis. J. Biol. Chem. 287, 1245512468.
Young, M.M., Kester, M., and Wang, H.-G. (2013). Sphingolipids: regulators of crosstalk between apoptosis and autophagy. J. Lipid Res. 54, 5-19.
Yu, L., McPhee, C.K., Zheng, L., Mardones, G.A., Rong, Y., Peng, J., Mi, N., Zhao, Y., Liu, Z., Wan, F., et al. (2010). Termination of autophagy and reformation of lysosomes regulated by mTOR. Nature 465, 942-946.
