Vps27 in Endosomal Cholesterol Trafficking

0 downloads 0 Views 2MB Size Report
Jan 26, 2012 - and hVps20 has been reported to cause endosomal cholesterol accumulation (Peck et al., ... Hrs Knockdown Causes Endosomal Accumulation.
Cell Reports

Report An Essential Role of Hrs/Vps27 in Endosomal Cholesterol Trafficking Ximing Du,1 Abdulla S. Kazim,1 Andrew J. Brown,1 and Hongyuan Yang1,* 1School of Biotechnology and Biomolecular Sciences, University of New South Wales, Sydney, NSW, 2052, Australia *Correspondence: [email protected] DOI 10.1016/j.celrep.2011.10.004

SUMMARY

The endosomal sorting complex required for transport (ESCRT) plays a crucial role in the degradation of ubiquitinated endosomal membrane proteins. Here, we report that Hrs, a key protein of the ESCRT-0 complex, is required for the transport of low-density lipoprotein-derived cholesterol from endosomes to the endoplasmic reticulum. This function of Hrs in cholesterol transport is distinct from its previously defined role in lysosomal sorting and downregulation of membrane receptors via the ESCRT pathway. In line with this, knocking down other ESCRT proteins does not cause prominent endosomal cholesterol accumulation. Importantly, the localization and biochemical properties of key cholesterol-sorting proteins, NPC1 and NPC2, appear to be unchanged upon Hrs knockdown. Our data identify Hrs as a regulator of endosomal cholesterol trafficking and provide additional insights into the budding of intralumenal vesicles. INTRODUCTION Mammalian cells acquire exogenous cholesterol through receptor-mediated endocytosis of low-density lipoproteins (LDLs) (Brown and Goldstein, 1986). The endocytic pathway sorts and delivers LDL from early endosomes to late endosomes/lysosomes (LE/Ly) for the hydrolysis of cholesteryl esters, and the released free cholesterol exits LE/Ly efficiently to reach the plasma membrane and/or the endoplasmic reticulum (ER) for structural and regulatory functions (Chang et al., 2006; Ikonen, 2008; Kristiana et al., 2008; Mesmin and Maxfield, 2009). Niemann Pick type C (NPC) 1 and NPC2 are two key proteins that regulate the exit of LDL-derived cholesterol (LDL-C) from LE/Ly. Recent elegant studies suggest that NPC2 (a soluble, cholesterol binding protein that resides in the lysosomal lumen) accepts and delivers LDL-C to the N-terminal domain of NPC1 (an LE/Ly membrane protein with 13 transmembrane domains and three large lumenal loops), which then inserts LDL-C directly into the lysosomal membrane for export (Kwon et al., 2009). Putative proteins on the cytoplasmic side of the endosomal limiting membrane may be required to transport LDL-C to other membranes (Kwon et al., 2009; Yang, 2006). We have recently identified ORP5, an oxysterol binding protein (OSBP)-related

protein, as a possible key component in the post-LE/Ly transport of LDL-C (Du et al., 2011). It is likely that other yet-to-be identified proteins may also work with ORP5 to remove cholesterol from the endosomes. We have focused on components of the endosomal sorting complex required for transport (ESCRT) (Saksena et al., 2007), which mediates the degradation of ubiquitinated endosomal membrane proteins through the formation of intralumenal vesicles (ILVs) (Hurley, 2010; Hurley and Hanson, 2010; Raiborg and Stenmark, 2009; Williams and Urbe´, 2007). ESCRT consists of a highly conserved set of four hetero-oligomeric protein complexes (ESCRT-0, -I, -II, and -III) comprising more than a dozen subunits. ESCRT-0 is believed to sequester ubiquitinated cargo and recruit other ESCRT complexes. Recent elegant in vitro studies demonstrate that ESCRT-I and II can drive membrane deformation to form the ILVs whereas ESCRT-III is important for the abscission of the forming ILVs (Wollert and Hurley, 2010; Wollert et al., 2009). The AAA ATPase VPS4/SKD1 regulates the disassembly and recycling of ESCRT-III components, enabling further rounds of cargo sorting. There have been a few lines of evidence linking ESCRT function and cholesterol trafficking. We and others have shown that VPS4 can interact with yeast ORPs and regulate cholesterol sorting in yeast and mammalian cells (Bishop and Woodman, 2000; Wang et al., 2005; Yang, 2006). Overexpression of hSnf7 and hVps20 has been reported to cause endosomal cholesterol accumulation (Peck et al., 2004). A close functional relationship between ORPs and ESCRT function has been revealed in Caenorhabditis elegans (Kobuna et al., 2010). Interestingly, a recent study demonstrated that ORP5 specifically bound to the ubiquitin-interacting motif of Hrs (hepatocyte growth factor-regulated tyrosine kinase substrate, called VPS27 in yeast for vacuolar protein sorting 27) (Pridgeon et al., 2009). Hrs interacts with STAM (signal transducing adaptor molecule) to make up ESCRT-0, which plays a crucial role in initiating the ESCRT pathway (Bache et al., 2003b; Bilodeau et al., 2002). ESCRT-0 recruits ESCRT-I possibly through the direct interaction between Hrs and Tsg101(an ESCRT-I component that also binds ubiquitinated proteins), and hands the ubiquitinated cargo to ESCRT-I (Bache et al., 2003a; Bilodeau et al., 2003; Katzmann et al., 2003; Lu et al., 2003). Here, we report that Hrs, but not other ESCRT-0, -I, -II, or -III subunits, is required for the transport of LDL-C from endosomes to the ER. Our data indicate that Hrs regulates endosomal cholesterol transport independent of NPC1 and NPC2, and therefore identify Hrs as a regulator of cholesterol trafficking. Cell Reports 1, 29–35, January 26, 2012 ª2012 The Authors 29

Figure 1. Hrs Knockdown Causes LDLDerived Cholesterol Accumulation in HeLa Cells (A) HeLa cells were transfected with two siRNAs targeting different regions of Hrs (siHrs.1, siHrs.2), siNPC1 or a universal control siCtrl for 72 hr. Efficiency of knockdown was analyzed by immunoblotting using polyclonal anti-NPC1, monoclonal anti-Hrs, or anti-actin antibodies. (B) HeLa cells were transfected with indicated siRNAs for 72 hr. Cells were then fixed and stained by filipin for free cholesterol. Representative fluorescent images are shown. (C) Filipin intensities of cells from (B) were measured using ImageJ and the values were normalized to those of control images (mean ± SD, n > 30). (D) HeLa cells were transfected with siCtrl or siHrs for 54 hr, followed by incubation in medium A (FBS), medium B (LPDS), or medium B supplemented with 50 mg/ml of LDL (LPDS + LDL). Cells were then fixed and stained by filipin for free cholesterol. (E) Filipin intensities of cells from (D) were measured using ImageJ and the values were normalized to those of control cells grown in FBS medium (mean ± SD, n > 30). Bars = 10 mm.

RESULTS AND DISCUSSION Hrs Knockdown Causes Endosomal Accumulation of LDL-Derived Cholesterol We have recently shown that ORP5 plays an important role in endosomal cholesterol transport (Du et al., 2011). The possible link between ORP5 and Hrs (Pridgeon et al., 2009) prompted us to investigate whether Hrs and other ESCRT components may regulate intracellular cholesterol transport and homeostasis. Strikingly, an NPC-like cholesterol accumulation phenotype was observed in HeLa cells when Hrs was depleted by either of the two distinct siRNA oligos (Figures 1A–1C). Both free and total cholesterol was increased upon Hrs knockdown (Figure S1A available online). The accumulation of cholesterol in Hrs knockdown cells is dependent on the presence of LDL (Figures 1D and 1E), indicating that Hrs may regulate the transport of LDL-C in HeLa cells. To test whether Hrs silencing causes the accumulation of LDL-C in LE/Ly, we investigated the colocalization between accumulated free cholesterol and the late endosomal marker, Lamp-1. Cholesterol mainly accumulated in Lamp-1-positive compartment in Hrs knockdown cells, similar to that in NPC1 knockdown cells (Figures S1B and S1C). Cholesterol in Hrs knockdown cells colocalized mainly with GFP-Rab7 and partially with GFP-Rab5 or GFP-EEA1 (Figures S1D and S1E), further indicating that Hrs depletion resulted in late endosomal accumulation of free cholesterol. Together, these data indicate that Hrs is specifically involved in endosomal transport of LDL-C in HeLa cells. Hrs Has a Distinct Role in Endosomal Cholesterol Transport To investigate whether other ESCRT components may also regulate cholesterol trafficking, HeLa cells were treated with 30 Cell Reports 1, 29–35, January 26, 2012 ª2012 The Authors

siRNAs targeting Hrs and selected ESCRT subunits: Hrs (ESCRT-0), Tsg101 (ESCRT-I), EAP20 (ESCRT-II), and CHMP6 (ESCRT-III) (Figure 2A). Among these ESCRT subunits, only Hrs appears to be involved in cholesterol transport, since the downregulation of other ESCRT subunits in HeLa cells did not cause any apparent cholesterol accumulation (Figure 2B). We carefully examined the effect of Tsg101 knockdown because Hrs and Tsg101 are closely related, both functionally and physically (Bache et al., 2003a; Lu et al., 2003). Tsg101 was efficiently depleted after siRNA transfection as indicated by immunoblotting (Figure 2C), and by the enlargement and/ or clustering of Lamp-1-positive compartments that indicates impaired Tsg101 function as previously described (Doyotte et al., 2005; Figure 2D). To further ensure that Tsg101 is effectively knocked down, we examined the trafficking of caveolin-1 (CAV1) tagged with mEGFP and mCherry at its C terminus (Hayer et al., 2010) in Hrs and Tsg101 single or double knockdown cells. Consistent with the previous report, most of LE/Ly compartments were red in control cells (Figure S2A), whereas both red and green LE/Ly compartments were observed in cells depleted of Hrs or Tsg101 due to defects in ILV targeting of CAV1 (Figures S2B and S2C). This phenotype became more severe in HeLa cells depleted of both Hrs and Tsg101 (Figure S2D, line profile). These results indicate that the function of Tsg101 is compromised by siRNA, but cholesterol transport is not, further supporting the distinct role of Hrs in cholesterol trafficking. A recent study revealed a close relationship between sorting nexin 3 (SNX3) and Hrs (Pons et al., 2008). Interestingly, silencing of SNX3 in HeLa cells appeared to have no effect on cholesterol transport (Figures S2E and S2F). We also knocked down VPS26 (Figure S2G), a component of the retromer complex that associates with endosomes and mediates

Figure 2. Hrs Is Specifically Involved in Intracellular Cholesterol Trafficking (A) HeLa cells were transfected with siRNAs targeting Hrs (ESCRT-0), Tsg101 (ESCRT-I), EAP20 (ESCRT-II), or CHMP6 (ESCRT-III) for 72 hr. Efficiency of knockdown was analyzed by qRT-PCR (means ± SD, n = 3 replicate cultures). (B) HeLa cells were transfected with the indicated siRNAs for 72 hr. Cells were then fixed and stained by filipin for free cholesterol. Representative fluorescent images are shown. (C) HeLa cells were transfected with siHrs, siTsg101, or a universal control siCtrl for 72 hr. Efficiency of knockdown was analyzed by immunoblotting using monoclonal anti-Hrs, antiTsg101, or anti-actin antibodies. (D) HeLa cells were transfected with the indicated siRNAs for 72 hr, followed by processing for immunofluorescence staining with the monoclonal antibody to Lamp-1 and filipin staining for free cholesterol. Representative confocal images are shown. Bars = 10 mm.

endosomes to trans-Golgi trafficking (Bonifacino and Hurley, 2008). Consistent with a previous report (Arighi et al., 2004), VPS26 depletion caused a dispersed distribution of cationindependent mannose 6-phosphate receptor (CI-MPR) in HeLa cells (Figure S2H). However, no detectable defects in cholesterol distribution upon VPS26 knockdown were observed in HeLa cells (Figure S2H). Taken together, these data indicate that Hrs has a unique role in endosomal cholesterol trafficking. Hrs Is Required for LDL-C Transport to the ER To monitor the arrival of LDL-C at the ER, HeLa cells depleted for Hrs were incubated with LDL, followed by examining the conversion of cholesterol to cholesteryl ester that is catalyzed by the ER-localized enzyme: ACAT (acyl coA cholesterol acyl transferase). NPC1 knockdown in HeLa cells ablated LDL-induced cholesterol esterification (Figure 3A, lane 3). Similar to NPC1 knockdown, Hrs depletion also significantly decreased LDLinduced cholesterol esterification (Figure 3A, lanes 3 and 4 versus lane 2, Figure 3B), indicating a blockage of endosomal cholesterol transport to the ER. The lack of LDL-C in the ER as indicated by the cholesterol esterification assay was further confirmed by the sterol-regulatory element binding protein (SREBP)-2 processing assay. Cholesterol-loading in the ER

blocks the transport of SREBP-2 precursor from the ER to the Golgi complex for proteolytic cleavage, inhibiting the formation of nuclear form of SREBP-2 (nSREBP-2) and inactivating genes involved in cholesterol uptake and synthesis (Goldstein et al., 2006). SREBP-2 processing was revealed by immunoblotting using an antibody (IgG1D2) against both precursor and nSREBP-2 (Figure 3C). As expected, NPC1 silencing impeded LDL-C transport to the ER because the nuclear form of SREBP-2 was still produced in the presence of LDL (Figure 3C, lane 4 versus lane 2, and Figure 3D). Consistent with the results from cholesterol esterification assay (Figures 3A and 3B), Hrs knockdown also blocked LDL-C transport to the ER and abolished the inhibitory effect of LDL on SREBP-2 processing (Figure 3C, lane 6 versus lane 2, and Figure 3D). This effect was further tested using qRT-PCR to examine the expression of two SREBP-dependent genes: LDL receptor (LDLR) and 3-hydroxy-3-methyglutarylCoA reductase (HMGR). Under control conditions, LDL treatment almost completely shut down LDLR or HMGR expression (Figure 3E); however, when NPC1 or Hrs was depleted, the inhibitory effect of LDL on LDLR or HMGR expression was significantly reduced (Figure 3E). Collectively, these data demonstrate a deficiency of LDL-C transport to the ER when Hrs is depleted, further supporting that Hrs is required for this process. NPC1 and NPC2 Still Localize to LE/Ly upon Hrs Depletion NPC1 and NPC2 work in concert to deliver LDL-C out of LE/Ly membranes. Hrs knockdown may interfere with the level and localization of NPC1 or NPC2, thereby indirectly causing cholesterol accumulation in LE/Ly compartments. However, the levels of NPC1 and NPC2 were not compromised upon Hrs depletion Cell Reports 1, 29–35, January 26, 2012 ª2012 The Authors 31

Figure 3. Hrs Is Required for LDL-Derived Cholesterol Transport to the Endoplasmic Reticulum (A) HeLa cells grown in medium A were transfected with siNPC1, siHrs, or a universal control siCtrl for 48 hr, followed by incubation in medium B for 18 hr. Cells were then chased with LDL (50 mg/ml) and [14C]-palmitate in medium B for 7 hr. Lipids extracted from cell lysates were standardized for total cell proteins and separated by TLC. Cholesteryl [14C]-esters were revealed by phosphorimaging. A representative phosphorimage of three experiments with similar results is shown. (B) Quantification of cholesteryl [14C]-esters formed in (A) by densitometry. Values of siCtrl + LDL were arbitrarily set as 100, against which experimental data were normalized. The results are expressed as means ± SD from three independent experiments. (C) HeLa cells grown in medium A were transfected with siNPC1, siHrs, or a universal control siCtrl for 48 hr, followed by incubation in medium B for 18 hr. Cells were then treated with LDL (50 mg/ml) in medium B for 6 hr. Whole cell lysates were subjected to SDS-PAGE and immunoblotting with a monoclonal antibody (IgG-1D2) against both precursor (P) and nuclear forms (N) of SREBP-2. The membrane was stripped and reblotted with antibodies to NPC1, Hrs, and actin. (D) Relative intensity of the N band relative to the total (N+P bands) in (C) was quantified by densitometry and presented relative to none LDL-treated control in each siRNA treatment. The results are expressed as means ± SD from three independent experiments. (E) qRT-PCR analysis of cells with the same treatment in (C). mRNA levels for LDL-receptor (LDLR) or HMG-CoA reductase (HMGR) were measured and normalized to actin mRNA levels. Data are presented relative to no LDL-treated control in each siRNA treatment and are means ± SD (n = 3 replicate cultures).

(Figure S3). In fact, whereas NPC2 appeared to be unchanged (Figures S3A and S3B), NPC1 was upregulated in Hrs knockdown cells in a time-dependent manner (Figures S3C and S3D). NPC1 or NPC2 may be mislocalized upon Hrs silencing, which could lead to cholesterol accumulation in LE/Ly. To examine this possibility, NPC1 localization was investigated by immunofluorescence in cells treated with control, Hrs or NPC1 siRNAs. In control cells, NPC1 almost completely colocalized with the LE/Ly marker, Lamp-1 (Figure 4A). Depletion of either NPC1 or Hrs by siRNA led to the accumulation of free cholesterol in Lamp-1-positive compartments (Figures 4B and 4C). In Hrs knockdown cells, NPC1 clearly overlapped with Lamp-1-positive structures, which encircled the accumulated cholesterol (Figure 4C, inset). These observations demonstrate that NPC1 is not mislocalized upon Hrs depletion. To examine NPC2 localization, HeLa cells treated with control siRNA or siRNAs against NPC1, NPC2, or Hrs were immunolabeled for endogenous NPC1 and NPC2, and stained with filipin for free cholesterol. In control cells, NPC1 and NPC2 colocalized to the same compartments (Figure 4D), which were shown to be Lamp-1-positive structures (Figures 4A–4C). Cholesterol accumulation in LE/Ly caused by NPC1 depletion overlapped with NPC2, also confirming the cellular localization of NPC2 (Figure 4E). This was also the case for NPC1, which overlapped 32 Cell Reports 1, 29–35, January 26, 2012 ª2012 The Authors

with the accumulated cholesterol caused by NPC2 depletion (Figure 4F). Importantly, in Hrs-depleted cells, NPC1, NPC2 and also cholesterol clearly colocalized (Figure 4G, inset). These microscopic data reveal that Hrs knockdown causes cholesterol accumulation in LE/Ly without affecting the localization of NPC1 and NPC2. Hrs and Cholesterol Sorting Here, we identify Hrs, but not other ESCRT components, as a regulator of endosomal cholesterol trafficking. Compared with other ESCRT components, Hrs is unique in at least two aspects: (1) Hrs functions at the very beginning of the ESCRT pathway, and therefore may have a critical role in cargo sorting and the initiation of intralumenal vesicle (ILV) budding (Babst, 2011; Hurley et al., 2010). (2) Hrs has multiple interacting partners that include many non-ESCRT components and has been implicated in a number of trafficking/sorting events. Besides STAM, clathrin and Tsg101, Hrs has also been reported to interact with SNX1 and Vps35, components of the retromer complex (Popoff et al., 2009). Moreover, Hrs, but none of the other ESCRT components, is required for the recycling of plasma membrane G protein coupled receptors (Hanyaloglu et al., 2005). Together, it seems that the ability of Hrs to interact with multiple partners forms the basis for its multiple functions. Therefore, it is not entirely surprising

Figure 4. NPC1 and NPC2 Localize to Lamp-1 Positive Compartments in Hrs Knockdown Cells HeLa cells were transfected with siCtrl (A and D), siNPC1 (B and E), siHrs (C and G), or siNPC2 (F) for 72 hr, followed by processing for immunofluorescence staining with antibodies to Lamp-1 or NPC1 (A–C), NPC1 or NPC2 (D–G), and filipin staining (A–G) for free cholesterol. Representative confocal images are shown. Bars = 10 mm.

that Hrs, but not other ESCRT components, is required for the proper delivery of LDL-C. Because of its well-established role in endosomal protein sorting, one way Hrs may regulate cholesterol transport is through controlling the level/degradation and localization of two endosomal proteins: NPC1 and NPC2. However, NPC1 and NPC2 both appear to be correctly localized upon Hrs knockdown (Figure 4).

The level of NPC1 is upregulated by 50% upon Hrs depletion, whereas NPC2 is unchanged. These results indicate that Hrs may regulate the efflux of cholesterol from endosomes through other mechanisms. Hrs may recruit putative lipid carriers in the cytoplasm, and facilitate cholesterol removal from the limiting membrane. In support of such a hypothesis, a genome-wide screen of Hrs-interacting proteins identified ORP5 (Pridgeon et al., 2009), which has recently been shown to mediate endosomal cholesterol transport (Du et al., 2011). We are currently examining the detailed physical and functional relationship between ORP5, Hrs, and NPC1. Additionally, disturbed ILV formation upon Hrs knockdown may further interrupt the efflux of LDL cholesterol from LE/Ly as ILVs may play a role in the storage and trafficking of lumenal cholesterol. It is likely that the striking cholesterol accumulation phenotype upon Hrs depletion results from its combined effect on ILV formation, and on the recruitment of cytoplasmic sterol carriers. In summary, although further work is needed to elucidate the interactions between NPC1, ORP5, and Hrs, it is clear that Hrs can regulate endosomal cholesterol sorting. Hrs, Cholesterol, and Intralumenal Budding The ESCRT pathway is characterized by the formation of ILVs through a unique mechanism, i.e., membrane budding away from the cytoplasm. An important unanswered question is what triggers the initial membrane deformation in this budding event, although both lipids and ESCRT proteins have been implicated (Babst, 2011; Hurley et al., 2010). Given the fact that most coatless budding mechanisms rely on membrane microdomains, it has been hypothesized that the ESCRT-mediated budding could involve cholesterol-rich domains, which may initiate the budding (Hurley et al., 2010). However, little is known about how ESCRTs may regulate the formation of such microdomains, which could form spontaneously as recently proposed (Babst, 2011). Here, a function of Hrs in endosomal cholesterol transport is revealed. Hrs may initiate the ESCRT pathway by orchestrating the formation of putative cholesterol rich microdomains that Cell Reports 1, 29–35, January 26, 2012 ª2012 The Authors 33

help drive the ‘‘budding away’’ from cytosol. In one hypothetical scenario, Hrs may selectively facilitate the removal of cholesterol from specific regions of the limiting membrane, thereby generating a cholesterol rich neighboring region that forms the basis of membrane deformation, which initiates the ESCRT pathway. The potential role of Hrs in generating microdomains on the endosomes may have broader implications. For instance, the recycling of a prototypical sequence-dependent recycling receptor, the beta-2 adrenergic receptor, is an active process mediated by distinct endosomal subdomains that are stabilized by a highly localized but dynamic actin machinery (Puthenveedu et al., 2010). The fact that knocking down Hrs, but none of the other ESCRT components, can significantly impair this recycling pathway suggests that cholesterol-rich microdomains generated by Hrs may also be involved in the formation of this actinstabilized subdomain (Hanyaloglu et al., 2005). In summary, our results uncover a role of Hrs in cholesterol sorting and indicate that cholesterol efflux from endosomal compartments and the formation of ILVs may be coregulated by the multifunctional protein: Hrs.

SUPPLEMENTAL INFORMATION Supplemental Information includes Extended Experimental Procedures and three figures and can be found with this article online at doi:10.1016/j.celrep. 2011.10.004. LICENSING INFORMATION This is an open-access article distributed under the terms of the Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 Unported License (CC-BY-NC-ND; http://creativecommons.org/licenses/by-nc-nd/3.0/ legalcode). ACKNOWLEDGMENTS This work is jointly supported by research grants from the Ara Parseghian Medical Research Foundation and the National Health and Medical Research Council of Australia (#510271). H.Y. is a Future Fellow of the Australian Research Council. Received: August 11, 2011 Revised: September 27, 2011 Accepted: October 25, 2011 Published online: January 26, 2012

EXPERIMENTAL PROCEDURES REFERENCES Cell Culture and Transfection HeLa cells were purchased from the American Type Culture Collection (ATCC HTB-22; Rockville, MD). Monolayers of cells were maintained in specified medium supplemented with serum (medium A: 10% FBS; medium B: 5% LPDS), 100 units/ml penicillin, and 100 mg/ml streptomycin sulfate in 5% CO2 at 37 C. DNA transfection was performed using Lipofectamine LTX and Plus Reagent (Invitrogen) according to manufacturer’s instruction. For each transfection, 1–2 mg/well of plasmid cDNA were used in 6-well plates. siRNA transfection was carried out in cells grown in medium A at 20% confluence according to standard methods using Lipofectamine RNAiMAX transfection reagent (Invitrogen). Forty picomoles of duplexes of siRNA were used for transfection of one well of cells grown in 6-well plate. Filipin Staining and Fluorescence Microscopy Cells grown on coverslips were fixed with 4% paraformaldehyde for 30 min at room temperature. Cells were stained with freshly prepared 50 mg/ml of filipin in PBS for 1 hr at room temperature. Stained cells were imaged using a Leica CTR5500 microscope (Wetzlar, Germany) equipped with an EL6000 fluorescent lamp and a DFC300 FX digital camera. Quantification of intracellular free cholesterol was carried out as previously described (Du et al., 2011). Immunofluorescence and Confocal Microscopy All immunofluorescence steps were performed at room temperature and cells were extensively rinsed with 3% BSA/PBS after each step. Cells grown on glass coverslips were fixed with 4% paraformaldehyde for 15 min. For NPC2 staining, cells were fixed with Bouin’s solution (Sigma-Aldrich) for 1 hr at room temperature. Cell permeabilization was carried out using 0.1% saponin/ PBS for 30 min, followed by blocking with 3% BSA/0.05% saponin in PBS for 1 hr. Incubation with primary antibodies and appropriate conjugated secondary antibodies were performed at room temperature for 1 hr. Cells were mounted in ProLong Gold antifade reagent (Invitrogen). Confocal images were acquired on an Olympus FV1000 laser-scanning microscope. The manufacturer’s software and Adobe Photoshop CS4 were used for data acquisition. For colocalization analysis, Image-Pro Plus 6.0 software was used. Cholesterol Esterification, SREBP-2 Processing, and Immunoblotting Analysis Cholesterol esterification, SREBP-2 processing, and immunoblotting analysis were carried out as previously described (Du et al., 2006, 2011) and are detailed in Supplemental Experimental Procedures.

34 Cell Reports 1, 29–35, January 26, 2012 ª2012 The Authors

Arighi, C.N., Hartnell, L.M., Aguilar, R.C., Haft, C.R., and Bonifacino, J.S. (2004). Role of the mammalian retromer in sorting of the cation-independent mannose 6-phosphate receptor. J. Cell Biol. 165, 123–133. Babst, M. (2011). MVB vesicle formation: ESCRT-dependent, ESCRT-independent and everything in between. Curr. Opin. Cell Biol. 23, 452–457. Bache, K.G., Brech, A., Mehlum, A., and Stenmark, H. (2003a). Hrs regulates multivesicular body formation via ESCRT recruitment to endosomes. J. Cell Biol. 162, 435–442. Bache, K.G., Raiborg, C., Mehlum, A., and Stenmark, H. (2003b). STAM and Hrs are subunits of a multivalent ubiquitin-binding complex on early endosomes. J. Biol. Chem. 278, 12513–12521. Bilodeau, P.S., Urbanowski, J.L., Winistorfer, S.C., and Piper, R.C. (2002). The Vps27p Hse1p complex binds ubiquitin and mediates endosomal protein sorting. Nat. Cell Biol. 4, 534–539. Bilodeau, P.S., Winistorfer, S.C., Kearney, W.R., Robertson, A.D., and Piper, R.C. (2003). Vps27-Hse1 and ESCRT-I complexes cooperate to increase efficiency of sorting ubiquitinated proteins at the endosome. J. Cell Biol. 163, 237–243. Bishop, N., and Woodman, P. (2000). ATPase-defective mammalian VPS4 localizes to aberrant endosomes and impairs cholesterol trafficking. Mol. Biol. Cell 11, 227–239. Bonifacino, J.S., and Hurley, J.H. (2008). Retromer. Curr. Opin. Cell Biol. 20, 427–436. Brown, M.S., and Goldstein, J.L. (1986). A receptor-mediated pathway for cholesterol homeostasis. Science 232, 34–47. Chang, T.Y., Chang, C.C., Ohgami, N., and Yamauchi, Y. (2006). Cholesterol sensing, trafficking, and esterification. Annu. Rev. Cell Dev. Biol. 22, 129–157. Doyotte, A., Russell, M.R., Hopkins, C.R., and Woodman, P.G. (2005). Depletion of TSG101 forms a mammalian ‘‘Class E’’ compartment: a multicisternal early endosome with multiple sorting defects. J. Cell Sci. 118, 3003–3017. Du, X., Kristiana, I., Wong, J., and Brown, A.J. (2006). Involvement of Akt in ERto-Golgi transport of SCAP/SREBP: a link between a key cell proliferative pathway and membrane synthesis. Mol. Biol. Cell 17, 2735–2745. Du, X., Kumar, J., Ferguson, C., Schulz, T.A., Ong, Y.S., Hong, W., Prinz, W.A., Parton, R.G., Brown, A.J., and Yang, H. (2011). A role for oxysterol-binding protein-related protein 5 in endosomal cholesterol trafficking. J. Cell Biol. 192, 121–135.

Goldstein, J.L., DeBose-Boyd, R.A., and Brown, M.S. (2006). Protein sensors for membrane sterols. Cell 124, 35–46.

Mesmin, B., and Maxfield, F.R. (2009). Intracellular sterol dynamics. Biochim. Biophys. Acta 1791, 636–645.

Hanyaloglu, A.C., McCullagh, E., and von Zastrow, M. (2005). Essential role of Hrs in a recycling mechanism mediating functional resensitization of cell signaling. EMBO J. 24, 2265–2283.

Peck, J.W., Bowden, E.T., and Burbelo, P.D. (2004). Structure and function of human Vps20 and Snf7 proteins. Biochem. J. 377, 693–700.

Hayer, A., Stoeber, M., Ritz, D., Engel, S., Meyer, H.H., and Helenius, A. (2010). Caveolin-1 is ubiquitinated and targeted to intralumenal vesicles in endolysosomes for degradation. J. Cell Biol. 191, 615–629. Hurley, J.H. (2010). The ESCRT complexes. Crit. Rev. Biochem. Mol. Biol. 45, 463–487. Hurley, J.H., and Hanson, P.I. (2010). Membrane budding and scission by the ESCRT machinery: it’s all in the neck. Nat. Rev. Mol. Cell Biol. 11, 556–566. _ Hurley, J.H., Boura, E., Carlson, L.A., and Ro´zycki, B. (2010). Membrane budding. Cell 143, 875–887. Ikonen, E. (2008). Cellular cholesterol trafficking and compartmentalization. Nat. Rev. Mol. Cell Biol. 9, 125–138. Katzmann, D.J., Stefan, C.J., Babst, M., and Emr, S.D. (2003). Vps27 recruits ESCRT machinery to endosomes during MVB sorting. J. Cell Biol. 162, 413–423. Kobuna, H., Inoue, T., Shibata, M., Gengyo-Ando, K., Yamamoto, A., Mitani, S., and Arai, H. (2010). Multivesicular body formation requires OSBP-related proteins and cholesterol. PLoS Genet. 6, 6. Kristiana, I., Yang, H., and Brown, A.J. (2008). Different kinetics of cholesterol delivery to components of the cholesterol homeostatic machinery: implications for cholesterol trafficking to the endoplasmic reticulum. Biochim. Biophys. Acta 1781, 724–730. Kwon, H.J., Abi-Mosleh, L., Wang, M.L., Deisenhofer, J., Goldstein, J.L., Brown, M.S., and Infante, R.E. (2009). Structure of N-terminal domain of NPC1 reveals distinct subdomains for binding and transfer of cholesterol. Cell 137, 1213–1224. Lu, Q., Hope, L.W., Brasch, M., Reinhard, C., and Cohen, S.N. (2003). TSG101 interaction with HRS mediates endosomal trafficking and receptor down-regulation. Proc. Natl. Acad. Sci. USA 100, 7626–7631.

Pons, V., Luyet, P.P., Morel, E., Abrami, L., van der Goot, F.G., Parton, R.G., and Gruenberg, J. (2008). Hrs and SNX3 functions in sorting and membrane invagination within multivesicular bodies. PLoS Biol. 6, e214. Popoff, V., Mardones, G.A., Bai, S.K., Chambon, V., Tenza, D., Burgos, P.V., Shi, A., Benaroch, P., Urbe´, S., Lamaze, C., et al. (2009). Analysis of articulation between clathrin and retromer in retrograde sorting on early endosomes. Traffic 10, 1868–1880. Pridgeon, J.W., Webber, E.A., Sha, D., Li, L., and Chin, L.S. (2009). Proteomic analysis reveals Hrs ubiquitin-interacting motif-mediated ubiquitin signaling in multiple cellular processes. FEBS J. 276, 118–131. Puthenveedu, M.A., Lauffer, B., Temkin, P., Vistein, R., Carlton, P., Thorn, K., Taunton, J., Weiner, O.D., Parton, R.G., and von Zastrow, M. (2010). Sequence-dependent sorting of recycling proteins by actin-stabilized endosomal microdomains. Cell 143, 761–773. Raiborg, C., and Stenmark, H. (2009). The ESCRT machinery in endosomal sorting of ubiquitylated membrane proteins. Nature 458, 445–452. Saksena, S., Sun, J., Chu, T., and Emr, S.D. (2007). ESCRTing proteins in the endocytic pathway. Trends Biochem. Sci. 32, 561–573. Wang, P.Y., Weng, J., and Anderson, R.G. (2005). OSBP is a cholesterol-regulated scaffolding protein in control of ERK 1/2 activation. Science 307, 1472– 1476. Williams, R.L., and Urbe´, S. (2007). The emerging shape of the ESCRT machinery. Nat. Rev. Mol. Cell Biol. 8, 355–368. Wollert, T., and Hurley, J.H. (2010). Molecular mechanism of multivesicular body biogenesis by ESCRT complexes. Nature 464, 864–869. Wollert, T., Wunder, C., Lippincott-Schwartz, J., and Hurley, J.H. (2009). Membrane scission by the ESCRT-III complex. Nature 458, 172–177. Yang, H. (2006). Nonvesicular sterol transport: two protein families and a sterol sensor? Trends Cell Biol. 16, 427–432.

Cell Reports 1, 29–35, January 26, 2012 ª2012 The Authors 35

Supplemental Information EXTENDED EXPERIMENTAL PROCEDURES Materials Dulbecco’s modified eagle’s medium (DMEM), newborn calf serum (NCS), fetal bovine serum (FBS), penicillin-streptomycin, and Dulbecco’s phosphate-buffered saline (PBS) were obtained from Invitrogen (Carlsbad, CA). [1-14C]-Palmitate (specific activity: 51 mCi/mmol) was purchased from GE Healthcare (Chalfont St. Giles, UK). Lipoprotein-deficient serum (LPDS) was isolated from NCS by ultracentrifugation as described elsewhere (Kristiana et al., 2008). LDL was subfractionated by density gradient ultracentrifugation from the plasma of healthy male volunteers. Compactin (mevastatin), mevalonate, filipin, saponin, and protease inhibitor cocktail were obtained from Sigma-Aldrich (St. Louis, MO). Antibodies Antibodies used were rabbit polyclonal to the C-terminal region of human NPC1 (Novus Biologicals), Tsg101(Abcam), and SNX3 (a gift from Dr. Wanjin Hong, Institute of Molecular and Cell Biology, Singapore); mouse monoclonal to Hrs (Alexis Biochemicals), Lamp-1 (Santa Cruz Biotechnology), NPC2 (Santa Cruz Biotechnology), actin (Abcam), and SREBP-2 IgG-1D2 (a gift from Dr. Baoliang Song, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai, China). For immunoblotting, we obtained horseradish peroxidase-conjugated secondary antibodies from Jackson ImmunoResearch. For immunofluorescence, we obtained Alexa-Fluor conjugated secondary antibodies from Molecular Probes/Invitrogen. cDNA Constructs pEGFP-EEA1 was a gift from Dr. Rohan Teasdale (Institute of Molecular Bioscience, University of Queensland). pEGFP-Rab5 and pEGFP-Rab7 were gifts from Dr. Jennifer Lippincott-Schwartz (National Institutes of Health, Bethesda, MD). pCAV1-mGFP-mCherry was a gift from Dr. Ari Helenius, Institute of Biochemistry, ETH Zurich, 8093 Zurich, Switzerland (Hayer et al., 2010). siRNAs Targeting sequences of siRNAs against Hrs (Sigma-Aldrich) were siHrs.1: 50 -CGTCTTTCCAGAATTCAAA-30 and siHrs.2: 50 -CAGA TATTCTGTGGAAAGT-30 (unless otherwise stated, siHrs.1 was used in all experiments) Tsg101 (QIAGEN) was 50 -CAGTTTATCATT CAAGTGTAA-30 , EAP20 (Sigma-Aldrich) was 50 -CGATCCAGATTGTATTAGA-30 , CHMP6 (Sigma-Aldrich) was 50 -AGATCGAAAT GAAAGTGAT-30 , NPC1 (QIAGEN) was 50 - ACCAATTGTGATAGCAATATT-30 , NPC2 (QIAGEN) was 50 -CTGATGGTTGTAAGAGTG GAA-30 , SNX3 (Sigma-Aldrich) was 50 - AAGGGCTGGAGCAGTTTATAA-30 , and VPS26 (Sigma-Aldrich) was 50 - AACTCCTG TAACCCTTGAG-30 . An RNAi universal negative control (QIAGEN) was used as a negative control in all RNAi transfections. Quantitative Real-Time Reverse Transcription-PCR Total RNA was extracted using Trizol (Invitrogen). cDNA was generated from total RNA using a SuperScript VILO cDNA Synthesis Kit (Invitrogen). PCR was performed using Rotor-Gene RG-3000A (QIAGEN). Threshold cycle value for each gene was acquired at the log phase and transcript expression levels were normalized to b-actin using DCT values. We used the following primers purchased from Geneworks, Australia or Sigma-Genosys (listed 50 to 30 in the order of forward primer, reverse primer): b-actin, AGCGAG CATCCCCCAAAGTT, GGGCACGAAGGCTCATCATT; Hrs, CGCAGGCTGTACTATGAGGG, CTTCTGCCGCATTATCTCCAG; Tsg101, GAGAGCCAGCTCAAGAAAATGG, GGGATTGTTCCAGTGAGGTTC; EAP20, TCCCACCCTTCTTTACGTTACA, CTGAGC TTCCATCACCGTCAT; CHMP6, TTGAGTTCACCCAGATCGAAATG, TGGCAGCTCTATTTGTTCCTG; LDLR, AGTTGGCTGCGTT AATGTGAC, TGATGGGTTCATCTGACCAGT; HMGR, GGACCCCTTTGCTTAGATGAAA, CCACCAAGACCTATTGCTCTG; Vps26, GAAACAATCGCCAAATATGAA, TTTCCTCAGTTTTTCAGGAGC. Cholesterol Measurement HeLa cells grown in 6-well plates were transfected with indicated siRNA for 72 hr in medium A. Cellular cholesterol concentrations were measured using a Molecular Probes Amplex Red cholesterol assay kit (Invitrogen) according to manufacturer’s instruction. In brief, cells were lysed and incubated with cholesterol oxidase, horseradish peroxidase, and Amplex Red in the absence or presence of cholesterol esterase. The amount of cholesterol was determined indirectly by measuring resorufin absorbance at 560 nm. The values were normalized to the total cellular protein levels, which were determined by the BCA protein assay kit (Pierce). Cholesterol Esterification HeLa cells transfected with siRNA in medium A for 54 hr were incubated in medium B for 18 hr in a 6-well plate. The cells were then treated with or without 50 mg/ml LDL in medium B for 5 hr and metabolically labeled with [1-14C]-palmitate for an additional 2 hr. Lipid extraction was carried out as previously described (Goldstein et al., 1983). Briefly, the cells were washed twice with Buffer A (0.15 M NaCl, 0.05 M Tris-HCl, 2 mg/ml BSA [pH 7.4]) and once with Buffer B (0.15 M NaCl, 0.05 M Tris-HCl [pH 7.4]). The cells were then incubated with 1 ml of hexane-isopropanol (3:2) at room temperature for 30 min. The organic solvent was collected and the cells were incubated with another 1 ml of the same solvent for 15 min at room temperature. The two organic solvent extracts were combined in a 2-ml glass vial. The glass vials were kept in the fume hood until the solvent was evaporated to dryness. The cells remained in each well were harvested using 1 ml of 0.1 M NaOH and aliquots were removed for protein determination by the Cell Reports 1, 29–35, January 26, 2012 ª2012 The Authors S1

bicinchoninic acid (BCA) protein assay. The lipids in each vial were resuspended in 60 ml of hexane and normalized to the total amount of proteins. The lipids were spotted on a Silica Gel 60 F254 thin layer chromatography (TLC) plate. The plate was developed in heptane-diethyl ether-acetic acid (90:30:1) and then exposed to a BAS-MS imaging plate (Fujifilm, Tokyo, Japan) for 72–96 hr. The imaging plate was visualized using the FLA-5100 phosphorimager (Fujifilm). The relative intensities of bands corresponding to cholesteryl ester were quantified using Sciencelab ImageGauge 4.0 Software (Fujifilm) or ImageJ (NIH). Analysis of SREBP-2 Processing by Immunoblotting Cells treated with siRNAs for 54 hr were incubated in medium B containing 10 mM compactin and 50 mM mevalonate for 18 hr and then treated with or without LDL in the same medium for 6 hr. After incubation, the cells were washed once with cold PBS, resuspended in lysis buffer (10 mM Tris-HCl, pH 7.6, 100 mM NaCl, 2.5% [wt/vol] SDS) containing Protease Inhibitor Cocktail (1:50), and then passed through a 22-gauge needle 15 times. The resultant cell lysates were then shaken for 20 min at room temperature. The protein concentration of each whole cell lysate was measured (BCA kit; Pierce, Rockford, IL) after which an aliquot (50 mg) was analyzed by SDS-polyacrylamide gel electrophoresis (PAGE; 7.5% wt/vol) and immunoblotting using the monoclonal antibody to SREBP-2 (IgG-1D2). The relative intensities of bands were quantified using Sciencelab ImageGauge 4.0 Software (Fujifilm). Immunoblot Analysis Samples were mixed with 0.25 volume of gel loading buffer (250 mM Tris-HCl, pH 6.8, 10% [wt/vol] SDS, 25% (vol/vol) glycerol, 0.2% (wt/vol) bromphenol blue, and 5% (vol/vol) b-mercaptoethanol), boiled for 5 min at 95 C or incubated for 15 min at 37 C, and then subjected to 7.5% or 10% SDS-PAGE. After electrophoresis, the proteins were transferred to Hybond-C nitrocellulose filters (GE Healthcare). Incubations with primary antibodies were performed at 4 C overnight or at room temperature for 1-2h using the following antibodies: 1:1000 dilution of monoclonal Hrs, 1:1000 dilution of polyclonal NPC1, 1:500 dilution of polyclonal Tsg101, 1:2000 dilution of polyclonal SNX3, and 1:2500 dilution of monoclonal actin. Secondary antibodies were peroxidase-conjugated AffiniPure donkey anti-rabbit or donkey anti-mouse IgG (H+L; Jackson ImmunoResearch Laboratories) used at a 1:5000 dilution. The bound antibodies were visualized by ECL Western Blotting Detection Reagent (GE Healthcare). The filters were exposed to Hyperfilm ECL (GE Healthcare) for periods of 2 s to 3 min or analyzed by the Fujifilm Luminescent Image Analyzer LAS-3000 with Fujifilm ImageReader LAS3000 v1.1 software. The relative intensities of bands were quantified using Sciencelab ImageGauge 4.0 Software (Fujifilm, Tokyo, Japan). SUPPLEMENTAL REFERENCES Goldstein, J.L., Basu, S.K., and Brown, M.S. (1983). Receptor-mediated endocytosis of low-density lipoprotein in cultured cells. Methods Enzymol. 98, 241–260. Hayer, A., Stoeber, M., Ritz, D., Engel, S., Meyer, H.H., and Helenius, A. (2010). Caveolin-1 is ubiquitinated and targeted to intralumenal vesicles in endolysosomes for degradation. J. Cell Biol. 191, 615–629. Kristiana, I., Yang, H., and Brown, A.J. (2008). Different kinetics of cholesterol delivery to components of the cholesterol homeostatic machinery: implications for cholesterol trafficking to the endoplasmic reticulum. Biochim. Biophys. Acta 1781, 724–730.

S2 Cell Reports 1, 29–35, January 26, 2012 ª2012 The Authors

Figure S1. Hrs Knockdown Causes Cholesterol Accumulation in Late Endosomes/ Lysosomes (A) HeLa cells were transfected with control siRNA (siCtrl), siNPC1 or siHrs for 72 hr, followed by AmplexRed cholesterol assay to measure unesterified free cholesterol and total cholesterol. Data shown (means + SEM) are from two separate experiments. There is no error bar in siCtrl because this condition was set at 100 for both separate experiments. *p < 0.05, **p < 0.01, unpaired, two-tailed t test. (B) HeLa cells were transfected with siNPC1 or siHrs for 72 hr, followed by processing for immunofluorescence staining with a monoclonal antibody to Lamp-1 and filipin staining for free cholesterol. Representative confocal images are shown. Bars = 10 mm. (C) Colocalization for filipin (blue) and Lamp-1 (green) in (B). The Pearson’s correlation coefficients (means ± SD) are shown to reveal a strong colocalization for cholesterol and Lamp-1 in either NPC1 or Hrs knockdown cells. (D) HeLa cells were transfected with siHrs for 48 hr, followed by transfection with cDNAs encoding GFP-tagged Rab7, EEA1, or Rab5 for 24 hr. Cells were then stained by filipin staining for free cholesterol. Representative confocal images are shown. Bars = 10 mm. (E) Colocalization for filipin (blue) and GFP (green) in (D). The Pearson’s correlation coefficients (means ± SD) are shown to reveal a strong colocalization for cholesterol and Rab7 in Hrs knockdown cells. *p < 0.05, **p < 0.01, unpaired, two-tailed t test.

Cell Reports 1, 29–35, January 26, 2012 ª2012 The Authors S3

Figure S2. Hrs Is Specifically Involved in Intracellular Cholesterol Trafficking (A–D) HeLa cells were transfected with control siRNA or siRNAs targeting Hrs or Tsg101 for 48 hr, followed by transfection with CAV1 with dual tags of EGFP and mCherry for 24 hr. Cells were then fixed for filipin staining, followed by confocal microscopy. Representative confocal images are shown. Fluorescence profiles are shown in the right panel. Note that most of endosomes are red in control cells (A), whereas endosomes become both red and green in cells depleted with Hrs (B) or Tsg101 (C) due to defects in ILV targeting of CAV1; this effect is even stronger in Hrs and Tsg101 double knockdown cells (D). Different from Hrs knockdown (B), Tsg101 knockdown did not cause cholesterol accumulation (C). Bars = 10 mm. (E–H) HeLa cells grown in medium A were transfected with siHrs, siTsg101, siSNX3, siVPS26 or a universal control siCtrl for 72 h: (E) Efficiency of knockdown was analyzed by immunoblotting using antibodies to Hrs, Tsg101, SNX3 or actin; (F) Filipin staining of cells treated with indicated siRNAs. Note that SNX3 knockdown has no effect on cholesterol accumulation; (G) qRT-PCR analysis of VPS26 knockdown in HeLa cells (means ± SD, n = 3); (H) Cells treated with control siRNA or siVPS26 were processed for immunofluorescence staining with a monoclonal antibody to CI-MPR and filipin staining for free cholesterol. Representative confocal images are shown. Bars = 10 mm. Note that VPS26 silencing causes dispersed CI-MPR distribution but has no effect on cholesterol accumulation.

S4 Cell Reports 1, 29–35, January 26, 2012 ª2012 The Authors

Figure S3. Expression of NPC1, but Not NPC2, Is Upregulated upon Hrs Depletion (A) HeLa cells were transfected with siHrs or a universal control siCtrl for 72 hr, followed by immunoblotting using antibodies to NPC1, Hrs, actin or NPC2. (B) Relative intensities of protein bands (triplicate) in (A) was quantified by densitometry and normalized to the values of actin. Data (mean ± SD) are representative of three independent experiments with similar results. (C) HeLa cells were transfected with indicated siRNAs for 0–72 hr, followed by immunoblotting using antibodies to NPC1, Hrs, or actin. (D) Relative intensities of protein bands in (C) was quantified by densitometry and normalized to the values of actin. Data (mean ± SD) are representative of two independent experiments with similar results.

Cell Reports 1, 29–35, January 26, 2012 ª2012 The Authors S5