QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

Vol. 18, Issue 12, 5113-5123, December 2007

This Article Abstract Full Text (PDF) Supplemental Materials All Versions of this Article: E07-04-0330v1 18/12/5113    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Koivusalo, M. Articles by Ikonen, E. Search for Related Content PubMed PubMed Citation Articles by Koivusalo, M. Articles by Ikonen, E.

Endocytic Trafficking of Sphingomyelin Depends on Its Acyl Chain Length

Mirkka Koivusalo^*,, Maurice Jansen^*, Pentti Somerharju, and Elina Ikonen^*

*Institute of Biomedicine/Anatomy and Institute of Biomedicine/Biochemistry, University of Helsinki, FIN-00014, Helsinki, Finland

Submitted April 12, 2007; Revised September 21, 2007; Accepted October 5, 2007
Monitoring Editor: Howard Riezman

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
To study the principles of endocytic lipid trafficking, we introduced pyrene sphingomyelins (PyrSMs) with varying acyl chain lengths and domain partitioning properties into human fibroblasts or HeLa cells. We found that a long-chain, ordered-domain preferring PyrSM was targeted Hrs and Tsg101 dependently to late endosomal compartments and recycled to the plasma membrane in an NPC1- and cholesterol-dependent manner. A short-chain, disordered domain preferring PyrSM recycled more effectively, by using Hrs-, Tsg101- and NPC1-independent routing that was insensitive to cholesterol loading. Similar chain length-dependent recycling was observed for unlabeled sphingomyelins (SMs). The findings 1) establish acyl chain length as an important determinant in the endocytic trafficking of SMs, 2) implicate ESCRT complex proteins and NPC1 in the endocytic recycling of ordered domain lipids to the plasma membrane, and 3) introduce long-chain PyrSM as the first fluorescent lipid tracing this pathway.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Sphingomyelin (SM) is the most abundant sphingolipid and one of the major phospholipids in the mammalian cell plasma membrane (PM). The acyl chain profile of SM differs from that of most other phospholipids, being enriched in saturated and either relatively short or very long chains. For example, in human primary fibroblasts SMs with a 16:0 or 24:0 acyl chain constitute 70% of the total (Blom et al., 2001). The physiological significance of such bipartite acyl chain length distribution is not clear. SM is enriched in the PM, and it has a relatively long half-life suggesting that upon endocytosis, it is efficiently recycled to the PM, with only a small fraction targeted to lysosomes for degradation (Koval and Pagano, 1990; van Meer and Holthuis, 2000).

SM has high affinity for cholesterol in model membranes, and it has been proposed to partition into liquid-ordered domains, termed lipid rafts in cells (Simons and Ikonen, 1997; Brown and London, 1998; Simons and Vaz, 2004). Endocytic organelles have been suggested to sort lipids based on their domain association. By using dialkylindocarbocyanine (DiI) probes, Mukherjee and Maxfield found that probes with a propensity to partition into domains of different fluidity were differentially sorted in endosomes. DiI analogs with short or unsaturated hydrocarbon chains that preferred disordered domains were targeted to the endocytic recycling compartment, whereas those with long and saturated chains preferring more ordered domains entered late endosomes (Mukherjee et al., 1999; Hao et al., 2004). Whether differential endocytic sorting based on domain partitioning applies to naturally occurring lipids has not been investigated. Moreover, the endosomal proteins involved have not been identified and the effect of such differential sorting on lipid degradation has not been studied.

There is limited information on the mechanisms of SM internalization from the PM and subsequent intracellular targeting. Most of the data are based on studies with fluorescently labeled SMs. N-(N-[6-[(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino]caproyl])-sphingosylphosphorylcholine (C₆-NBD-SM) incorporated into the PM was endocytosed and mainly sorted for recycling to the PM, whereas some was transported to lysosomes (Koval and Pagano, 1990). N-[5-(5,7-dimethyl Bodipy)-1-pentanoyl]-D-erythro-sphingosylphosphorylcholine (C₅-DMB-SM; BODIPY₅-SM), in contrast, predominantly labeled the Golgi apparatus (Puri et al., 2001). This Golgi labeling may reflect vesicular movement of BODIPY-SM from endosomes to the Golgi and/or hydrolysis of the probe to the corresponding ceramide, which has a high affinity for the Golgi (Pagano et al., 1999). However, neither C₆-NBD-SM nor BODIPY₅-SM partition into ordered domains in model membranes (Wang and Silvius, 2000), and the trafficking pattern of C₆-NBD-SM is very similar to that of the short chain DiI analogs (Hao et al., 2004).

In this work, we studied whether the acyl chain length and differential domain partitioning play a role in the endocytic trafficking of SMs. We used pyrene-labeled SMs (PyrSMs) and unlabeled SMs with acyl chain lengths similar to naturally occurring SMs. Pyrene is a polycyclic fluorescent moiety that is more hydrophobic than NBD or BODIPY. Therefore, pyrene does not distort the conformation of the labeled chain, i.e., the depth of the pyrene moiety in the bilayer is determined by the length of the labeled acyl chain (Eklund et al., 1992; Sassaroli et al., 1995). Pyrene lipids mimic the behavior of their natural counterparts also in several other aspects (Somerharju, 2002). We have recently shown that PyrSMs with long acyl chains partition preferentially into ordered domains in model membranes, whereas short-chain PyrSMs prefer disordered domains (Koivusalo et al., 2004). We therefore chose a short-chain Pyr₄SM and a long-chain Pyr₁₂SM (and Pyr₁₄SM) to study SM trafficking in living cells.

SMs were introduced into the PM of primary fibroblasts or HeLa cells and their distribution analyzed after various chase times. Two methods were used to quantify the fraction of the probes in the PM: fluorescence quenching and compositional analysis of PM-derived vesicles. We found a systematic difference in the PM recycling of both fluorescently labeled and unlabeled SMs, with the short-chain SMs becoming more efficiently recycled compared with the long-chain SMs. Furthermore, cholesterol loading or depletion of proteins involved in multivesicular body formation or late endosomal trafficking altered the recycling and/or lysosomal degradation of the probes in a chain length-dependent manner. We discuss the potential mode of action of these proteins and the relevance of the findings with respect to domain-based lipid sorting in endosomes.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Materials
Glycerophospholipids and commercially available sphingolipids were from Avanti Polar Lipids (Alabaster, AL). NBD₆-ceramide, BODIPY₁₂-SM, and pyrene-labeled fatty acids were from Invitrogen (Carlsbad, CA), unmodified fatty acids (15:0, 17:0, 21:0, 23:0, 25:0) were from Larodan (Malmö, Sweden), -cyclodextrin (-CD) was from Cyclodextrin Technologies Development (High Springs, FL), Lipofectamine 2000 and DiI-low-density lipoprotein (LDL) were from Invitrogen, and other chemicals were from Sigma-Aldrich (St. Louis, MO). PyrSMs and other SMs were synthesized from sphingosylphosphorylcholine and a fatty acid, and they were purified as described previously (Ahmad et al., 1985; Koivusalo et al., 2004). Trinitrophenyl-lysophosphatidylethanolamine (TNP-LPE 12:0) was synthesized from 1-lauroyl-2-hydroxy-sn-glycero-3-phosphatidylethanolamine and 2,4,6-trinitrobenzenesulfonic acid as described previously (Tanhuanpaa et al., 2000). All lipids were >98% pure, and they were stored in chloroform/methanol (9:1) (C/M 9:1, vol/vol) below –20°C.

Cell Culture and Transfections
Control human skin fibroblasts F92-99, Niemann-Pick type A (NPA) fibroblasts CC-95-62, and HeLa cells were cultured as described previously (Holtta-Vuori et al., 2000, 2005). The Rab5Q79L construct (a kind gift from Marino Zerial, Max-Planck Institute, Dresden, Germany) was transfected into HeLa cells for 24 h using Lipofectamine 2000. RNA interference (RNAi) oligonucleotides against hepatocyte growth factor receptor substrate (Hrs), tumor susceptibility gene 101 (Tsg101), Niemann-Pick C1 (NPC1), and the control RNAi (GL2) were as described previously (Holtta-Vuori et al., 2005; Ganley and Pfeffer, 2006; Razi and Futter, 2006), and they were transfected into HeLa cells with Lipofectamine 2000. At 48 or 72 h (Hrs + Tsg101 or NPC1, respectively) after transfection, the cells were retransfected for another 48 or 72 h, followed by labeling with PyrSM.

Labeling of Cells with PyrSM for Microscopy and Metabolic Studies
PyrSMs were introduced into cells by using their -CD complexes. Lipid in C/M 9:1 was first evaporated under nitrogen and further dried under vacuum. One hundred mM -CD in PBS was added, and the mixture was probe-sonicated (Soniprep 150; Sanyo Gallenkamp, Leicestershire, United Kingdom) for 3 x 2 min at room temperature. Then, 80–90% confluent cell monolayers on glass-bottomed dishes (MatTek, Ashland, MA) were labeled for 5 min at 37°C in serum-free medium. To obtain similar levels of cellular incorporation, the final lipid/-CD concentrations were 3.3 µM/3.3 mM for Pyr₄SM, 6.7 µM/6.7 mM for Pyr₁₂SM, and 13.4 µM/6.7 mM for Pyr₁₄SM. After the pulse, cells were washed with phosphate-buffered saline (PBS) and chased in serum-free medium at 37°C. Imaging was done within 2 h after labeling to ensure that at least 75–80% of the pyrene fluorescence derived from the parental PyrSM. Control experiments indicated that the differential -CD concentrations used did not account for the observed differential distributions of the Pyr_nSMs (data not shown).

Labeling of Cells with PyrSM Species for Electrospray Ionization-Mass Spectrometry (ESI-MS) Analysis
PyrSMs were introduced into cells from donor vesicles (PyrSM/1-palmitoyl-2-oleyl-sn-glycero-3-phosphocholine [POPC]/1-palmitoyl-2-oleyl-sn-glycero-3-phosphate [POPA]/di23:0-PC 40/40/5/2 nmol per Ø 6-cm dish) with the aid of -CD (Tanhuanpaa and Somerharju, 1999). The molar ratio of Pyr₄SM/Pyr₁₂SM/Pyr₁₄SM was 1:2:2 to obtain similar levels of incorporation of the probes. Vesicles were generated from dried lipids in PBS by probe sonication for 4 x 2 min at 50°C. Then, 80–90% confluent cell monolayers were incubated with donor vesicles (26.7 µM PyrSM) and -CD (6.7 mM) for 1 h at 37°C in serum-free medium, washed with PBS, and chased in serum-free medium. The amounts of PyrSMs incorporated into cells using this method were similar to those obtained with direct transfer from -CD complexes (see above). This protocol was used for ESI-MS analyses because it allowed the assessment of possible vesicle attachment to the cell surface or dish by determination of the nontransferable di23:0-PC marker. ESI-MS analysis showed that the amount of di23:0-PC was <0.5% of total cellular phosphatidylcholine (PC), indicating an insignificant amount of adhering vesicles.

Labeling of Cells with Unmodified SM Species
We introduced 15:0-SM and 21:0-SM into cells from donor vesicles (SM/cholesterol/POPA/di23:0-PC, 25:25:5:2 nmol per Ø 6-cm dish) by using methyl- β-cyclodextrin (mβCD) as a carrier. mβCD was used instead of -CD because the size of its hydrophobic cavity is more compatible with SMs containing unlabeled acyl chains. Cholesterol was included in the vesicles to maintain the cellular cholesterol level (loss <10%; data not shown). The labeling protocol was essentially the same as for PyrSMs. The molar ratio of 15:0-SM/21:0-SM was 1:2 to obtain a similar degree of incorporation of both lipids. Cells were incubated with donor vesicles (16.7 µM SM) and mβCD (10.7 mM) for 1 h at 37°C.

Acid SMase In Vitro Assay
The specificity of fibroblast acid SMase toward SM species was determined essentially as described previously (Lusa et al., 1996; Liu and Hannun, 1999). Fibroblast extracts was prepared freshly before use by sonication of cell pellet in lysis buffer (50 mM Na-acetate, 1 mM EDTA, 4 mM Na-taurocholate, pH 5.0, and a protease inhibitor cocktail of chymostatin, leupeptin, antipain, and pepstatin A at 25 µg/ml each) for 5 x 15 s on ice. Undispersed material was pelleted (10 000 x g at 4°C for 5 min), and the supernatant was used to assay SM hydrolysis. A dried lipid mixture (SM/POPC/POPA, 9:8:1, mol/mol/mol) was dispersed in assay buffer (250 mM Na-acetate 1 mM EDTA, and 4 mM Na taurocholate, pH 5.0) by sonication in water bath at 50°C for 15 min. This micellar substrate (9 nmol of total SM) was incubated with the cell-free fibroblast extract (15 µg of protein) at 37°C for 4 h, the reaction stopped by addition of C/M 1:1, the lipids were extracted and analyzed with ESI-MS.

Isolation of PM vesicles
PM vesicles were isolated after induction of cell surface blebbing with formaldehyde (Scott et al., 1979; Holowka and Baird, 1983). Fibroblast or HeLa cell monolayers were incubated for 20–30 min at 37°C in blebbing buffer (10 mM HEPES, 0.15 M NaCl, and 2 mM CaCl₂, pH 7.4) containing 25 mM formaldehyde and 2 mM dithiothreitol. The suspension of vesicles was collected and centrifuged at 130 x g for 5 min to remove detached cells. Lipids from vesicles and cells were extracted and analyzed with ESI-MS.

Lipid Analysis by ESI-MS
Lipids were extracted according to Folch et al. (1957), and the SM species were quantified after removal of glycerophospholipids with alkaline hydrolysis. For quantification of PC, SM, and ceramide (cer) lipid extracts were spiked with the following mixture of internal standards dissolved in C/M 1:2: di14:1-PC, di20:1-PC, di22:1-PC, 17:0-SM, 23:0-SM, 25:0-SM, 14:0-cer, 17:0-cer, and 20:0-cer. The concentrations of phospholipid and ceramide standards were determined as described previously (Bartlett and Lewis, 1970; Naoi et al., 1974). Extracts were evaporated under nitrogen and dissolved in C/M 1:2 containing 10 mM ammonium acetate. ESI-MS analysis was carried out with a Quattro Microtriple quadrupole mass spectrometer (Micromass, Manchester, United Kingdom). All the lipids were analyzed in the positive ion mode by scanning for the precursors of m/z 184 (PC and SM) or m/z 264 (ceramide) (Hermansson et al., 2005). Data analysis was performed with the LIMSA software (Haimi et al., 2006).

Live Cell Fluorescence Microscopy
Cells on glass-bottomed dishes were imaged at 37°C (unless otherwise stated) in serum-free CO₂-independent minimum essential medium (I-MEM; Invitrogen) containing oxygen depletion reagents to minimize pyrene photobleaching (Tanhuanpaa and Somerharju, 1999). Imaging was carried out with a TILLPhotonics imaging system (TILLPhotonics, Gräfelfing, Germany) equipped with the Polychrome IV light source, an Olympus IX70 cooled charge-coupled device camera (Olympus, Melville, NY), TILLVision 4.0 software (TILLPhotonics), and a Solent Scientific (Segensworth, United Kingdom) temperature-controlled chamber. Excitation light source was set at 345 nm for pyrene, 480 nm for NBD, and 550 nm for DiI-LDL. Emission filters were 405 nm (40-nm bandpass), 480 nm (60-nm bandpass), 525 nm (30-nm bandpass), and 630 nm (60-nm bandpass) for pyrene monomer, pyrene excimer, NBD and DiI-LDL, respectively. In all experiments PyrSMs were imaged at the monomer wavelength. We confirmed that imaging at the excimer wavelength gave identical results (data not shown). UApo 40x/numerical aperture (NA) 1.15 water or UApo/340 40x/NA 1.35 oil immersion objective was used. To selectively quench pyrene fluorescence at the plasma membrane, 1 ml of 30 µM TNP-LPE in PBS was added to the dish containing 2 ml of I-MEM (Tanhuanpaa and Somerharju, 1999). To calculate the fraction of fluorescence inside the cells and in the PM, the same cell was imaged before and after TNP-LPE addition and the latter image subtracted from the former. Both images were corrected for glass/medium background fluorescence before subtraction. Imaging of unlabeled cells showed that the contribution of cellular autofluorescence was negligible with the exposure times used. Image analysis was performed using Image-Pro Plus software (Media Cybernetics, Silver Spring, MD).

Other Methods
PyrSM degradation was analyzed by high-performance liquid chromatography (HPLC) according to (Silversand and Haux, 1997). Detergent-resistant membranes (DRMs) were isolated from cells by flotation at 4°C in an OptiPrep gradient in the presence of 1% Triton X-100 as described previously (Harder et al., 1998). To increase the resolution of the gradient in the low-density fractions, an otherwise similar gradient but with less OptiPrep was used (steps of 0, 12, 15, 18, 21, and 24% OptiPrep). To analyze DRM association, fluorescent lipids from OptiPrep fractions were analyzed by HPLC as described above. To analyze endogenous SM species from OptiPrep fractions by ESI-MS, Triton X-100 was first removed from lipid extracts with a reversed-phase HPLC cartridge (OASIS HLB, particle size 2.5 µm, 2.1 x 20 mm; Waters, Toronto, ON, Canada). Lipids were trapped to the column in methanol:H₂O 2:1 and eluted with C/M 1:1. To label endosomes with DiI-LDL, fibroblasts were starved overnight in medium containing 5% lipoprotein-deficient serum, followed by DiI-LDL (20–30 µg/ml) incubation at 37°C for 10–30 min. The quenching efficiency of PyrSM by TNP-LPE in liposomes was determined as described previously (Koivusalo et al., 2004). Cholesterol loading of cells was performed for 1 h at 37°C using cholesterol/mβCD complex as described in Blom et al. (2001). Cellular protein was determined as described previously (Lowry et al., 1951) and cholesterol concentration was determined with the Amplex Red kit (Invitrogen). Western blot analysis, and labeling of cellular unesterified cholesterol for microscopy were carried out as described previously (Holtta-Vuori et al., 2002, 2005). Statistical significance of differences was analyzed by Student's t test.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Introduction of SM Species into Cells
PyrSMs with different acyl chain lengths, i.e., Pyr₄SM, Pyr₁₂SM, or Pyr₁₄SM, were introduced into the PM of human fibroblasts by using -CD as a carrier at 37°C. Adequate labeling was not obtained at lower temperatures (data not shown). The length of the pyrene moiety is 5.5 in methylene units (Sassaroli et al., 1995). Thus, the length of Pyr₄-, Pyr₁₂-, or Pyr₁₄-fatty acid is roughly 10, 18, or 20 methylene units, respectively (Figure 1). PyrSM species with a labeled chain longer than 14 carbons could not be incorporated into cells in adequate amounts due to the limited ability of CDs to mediate the intermembrane translocation of very hydrophobic lipids (Tanhuanpaa and Somerharju, 1999). We also introduced short- and long-chain unlabeled SMs, i.e., 15:0-SM and 21:0-SM, into fibroblasts from vesicles by using mβCD as a carrier (Figure 1). These SM species are not naturally present in human fibroblasts, and they can thus be readily distinguished from the endogenous species by mass spectrometry. The labeling conditions were adjusted so that the amounts of short-chain and long-chain SM species introduced into the cells were similar. The amounts of PyrSMs and unlabeled SMs incorporated corresponded to 10–20% of total cellular SM (Figure 2A).

View larger version (25K): [in this window] [in a new window]   Figure 1. Structures of the SMs studied.

View larger version (28K): [in this window] [in a new window]   Figure 2. Incorporation of SMs into human fibroblasts. (A) ESI-MS analysis of exogenous SM species in cells (n = 6–10). (B) Association of SMs with DRMs as analyzed by flotation in an OptiPrep gradient in the presence of 1% Triton X-100. DRM association of only two major endogenous fibroblast SMs is shown but the result was similar for the other endogenous SM species (n = 1–4). (C) Fraction of unlabeled SMs in DRMs using a modified gradient with increased resolution. Values for exogenously introduced 21:0-SM were normalized to those of 15:0-SM, and values for endogenous 24:0-SM to those of 16:0-SM in the same cells (n = 3; *p < 0.05). Error bars are SEM.

Distribution of PyrSMs in Human Fibroblasts
The prominent cell surface staining immediately after the labeling indicated that PyrSMs were first incorporated into the PM (Figure 4A). Within a few minutes, PyrSMs were also found in punctate intracellular structures, some of which represent early endosomes as judged from partial colocalization with DiI-LDL, which had been internalized for 10 min (Figure 3A). This agrees with the partial overlap of BODIPY₅-SM and DiI-LDL stainings observed in previous studies (Puri et al., 2001). At later times, some of the PyrSMs were also found in late endosomal compartments (Figure 9A). BODIPY-labeled sphingolipids predominantly localize to the Golgi apparatus in human fibroblasts (Pagano et al., 1999; Puri et al., 1999). To investigate whether this is the case for PyrSMs, we colabeled fibroblasts with the Golgi marker NBD-ceramide (Pagano and Martin, 1994). At 1–1.5 h after labeling, PyrSMs did not stain the Golgi as prominently as NBD-ceramide (Figure 3B).

View larger version (115K): [in this window] [in a new window]   Figure 3. Partial colocalization of PyrSM with DiI-LDL and NBD-ceramide in human fibroblasts. (A) Localization of Pyr4SM, Pyr12 SM, or Pyr14SM, and DiI-LDL chased into early endosomes. Fibroblasts were labeled with PyrSM for 5 min at 37°C, chased for 50 min at 37°C, and labeled with DiI-LDL for 10 min at 37°C before imaging. Arrows indicate colocalization of DiI and pyrene fluorescence. (B) Localization of PyrSM and NBD-ceramide. Fibroblasts were labeled with PyrSM, chased for 30 min at 37°C, labeled with NBD-ceramide for 30 min on ice, and then chased for 30 min at 37°C. PyrSM was imaged at room temperature (RT) after TNP-LPE quenching of PM fluorescence. Arrows indicate the Golgi region. Bar, 10 µm.

View larger version (47K): [in this window] [in a new window]   Figure 4. Assessment of the PM versus intracellular pool of PyrSM fluorescence in human fibroblasts. (A) Images of Pyr4SM, Pyr12SM, and Pyr14SM fluorescence after a 5-min pulse and 3-min chase at 37°C before and after quenching of PM fluorescence with TNP-LPE. (B) Images of Pyr4SM, Pyr12SM, and Pyr14SM fluorescence after 5-min pulse and 2-h chase at 37°C before and after quenching with TNP-LPE. Images ± TNP-LPE within each lipid acquired with the same exposure time and displayed with the same intensity scale. The +TNP-LPE image also shown with intensity multiplied by the factor indicated. (C) Quantification of the fraction of PyrSM fluorescence inside cells at different chase times in wild-type (WT) or NPA fibroblasts (n = 30–56 cells). Error bars are SEM. **p < 0.025, ***p < 0.001. Bar, 10 µm.

PM vesicles were isolated from fibroblasts that had been labeled with Pyr₄SM, Pyr₁₂SM, and Pyr₁₄SM and chased for 2 h. The amount of PyrSMs in vesicles and remaining cellular material was determined by ESI-MS. We found that Pyr₁₂SM and Pyr₁₄SM were depleted in the PM vesicles compared with the short-chain Pyr₄SM (Figure 5, A and B). These results agree well with those of the PM fluorescence-quenching assay (Figure 4). We then investigated whether the exogenous unlabeled SMs, 15:0-SM and 21:0-SM, exhibit a similar chain-length dependent distribution in PM vesicles. After 2 h of chase, 21:0-SM showed a moderate yet significant depletion from the PM vesicles compared with 15:0-SM (Figure 5, C and D). The effect became more pronounced with increasing the chase time to 2–3 d (Figure 5, C and D). These results indicate that among the SMs introduced into cells, the short-chain SMs became enriched in the PM fraction relative to the long-chain SMs.

View larger version (42K): [in this window] [in a new window]   Figure 5. ESI-MS analysis of the distribution of PyrSMs and unlabeled SMs between PM vesicles and cells. Pyr4SM, Pyr12SM, and Pyr14SM association with PM vesicles isolated from fibroblasts after 1-h labeling and 2-h chase is expressed as percentage of total PyrSM in PM vesicles (A) and long-chain/short-chain PyrSM ratio in PM vesicles and cells (B) (n = 6). Distribution of unlabeled SM species between cells and vesicles isolated after 1 h labeling and 2-h or 2- to 3-d chase is expressed as percentage of SM in PM vesicles (C) and long-chain/short-chain SM ratio in PM vesicles and cells (D) (n = 9–22). Error bars are SEM. **p < 0.01, ***p < 0.001.

View larger version (35K): [in this window] [in a new window]   Figure 6. TNP-LPE quenching and ESI-MS analysis of SM distribution between PM and internal membranes in HeLa cells. (A) Images of Pyr4SM and Pyr12SM fluorescence after a 5-min pulse and 10-min chase before and after quenching of PM fluorescence with TNP-LPE. Images ± TNP-LPE with each lipid acquired with the same exposure time, +TNP-LPE images shown with intensities multiplied by a factor of 2.5. Bar, 10 µm. (B) Quantification of the fraction of Pyr4SM and Pyr12SM fluorescence inside the cells by TNP-LPE quenching after a 5-min pulse and 5- to 20-min chase (n = 24). (C) ESI-MS analysis of short-chain (Pyr4SM/15:0-SM) versus long-chain (Pyr12SM/21:0-SM) SM distribution between cells and PM vesicles. Vesicles from PyrSM-labeled cells were isolated after 30 min labeling + 30 min chase (n = 3–4) and from cells labeled with unmodified SMs after 30-min labeling and 30-min to 2 h chase (n = 6). Data normalized to the value of short-chain SM. Error bars are SEM. *p < 0.05, **p< 0.01.

View larger version (63K): [in this window] [in a new window]   Figure 7. PyrSM internalization in cells expressing Rab5Q79L-GFP. (A) Images of Pyr4SM and Pyr14SM fluorescence after 5-min pulse and 7-min chase in HeLa cells transfected with Rab5Q79L-GFP. Bar, 10 µm. (B) Quantification of the fraction of Pyr4SM and Pyr14SM fluorescence inside cells by TNP-LPE quenching after a 2-min pulse and 2- to 3-min chase (n = 21–37 cells). Error bars are SEM.

View larger version (90K): [in this window] [in a new window]   Figure 8. Effect of Hrs + Tsg101 depletion, NPC1 depletion or cholesterol loading on PyrSM distribution between the PM and internal membranes. (A) Images of Pyr4SM and Pyr12SM fluorescence after quenching of PM fluorescence with TNP-LPE in HeLa cells transfected with control (GL2), Hrs + Tsg101 siRNA, or NPC1 siRNA. Cells were imaged after a 5-min pulse and 5- to 30-min chase. Same display range in all images. Bar, 10 µm. (B) Quantification of the fraction of Pyr4SM and Pyr12SM fluorescence inside the cells by TNP-LPE quenching (n = 54–126 cells). Error bars are SEM. *p < 0.05, **p < 0.01, ***p < 0.001. (C) Effect of cholesterol loading on the fraction of Pyr4SM and Pyr12SM fluorescence inside fibroblasts as quantified by TNP-LPE quenching. Control (ctr) or cholesterol-loaded (chol-load) fibroblasts were labeled with PyrSM and chased for 3-min to 2 h (n = 30–39). Error bars are SEM. ***p < 0.001.

Because cholesterol loading has been shown to result in sphingolipid accumulation in late endocytic organelles (Puri et al., 1999), we analyzed whether cholesterol loading affects PyrSM sorting. Cholesterol was loaded to primary fibroblasts from a mβCD complex for 1 h, resulting in an approximately twofold increase in cellular cholesterol levels (Blom et al., 2001). Thereafter, the cells were labeled with PyrSMs as described above. Cholesterol loading enhanced the cellular incorporation of PyrSMs by 30% for both the short- and long-chain species. The PM fluorescence quenching assay revealed that cholesterol loading significantly increased the intracellular fraction of the long-chain PyrSM but not that of the short-chain PyrSM (Figure 8C).

Lysosomal Degradation of PyrSMs
The partial colocalization of PyrSMs with DiI-LDL chased for 2 h (Figure 9A) suggested that some of the probes reached late endocytic organelles. Given the preferential endomembrane localization of the exogenous long-chain SMs, we considered that they may be preferentially sorted for lysosomal degradation. SM is hydrolyzed by lysosomal acid sphingomyelinase (SMase) and in principle, degradation can be used to monitor the delivery of SMs to the hydrolytic compartments. However, the interpretation is complicated by the potential acyl chain selectivity of acid SMase and contribution of neutral SMases (Levade et al., 1999). To address the first issue, we determined the rate of hydrolysis of short- and long-chain PyrSMs by acid SMase in vitro. The results indicated that acid SMase preferred Pyr₁₂SM as substrate (Supplemental Figure 4). Instead, Pyr₁₄SM was degraded only slightly more efficiently than Pyr₄SM and was chosen as the long-chain PyrSM to study lysosomal targeting in cells.

View larger version (31K): [in this window] [in a new window]   Figure 9. Degradation of PyrSM in human fibroblasts and in Hrs + Tsg101- or NPC1-depleted HeLa cells. (A) Partial colocalization of Pyr4SM or Pyr14SM with DiI-LDL chased into late endosomes in human fibroblasts. Cells were first labeled with DiI-LDL for 30 min and then with PyrSM for 5 min and chased for 1 to 1.5 h at 37°C. PyrSM was imaged after TNP-LPE quenching of PM fluorescence at RT. Arrows indicate colocalization of DiI and pyrene fluorescence. (B) Pyr4SM and Pyr14SM hydrolysis in control and NPA fibroblasts after 1–2 h of chase (n = 10 for control and 5 h for NPA cells). (C) Hydrolysis of Pyr4SM and Pyr14SM after a 30-min labeling and 2-h chase in HeLa cells transfected with control (GL2), Hrs + Tsg101, or NPC1 RNAi (n = 3–7). Error bars are SEM. ***p < 0.001.

We then analyzed the effect of Hrs + Tsg101 depletion on the degradation of PyrSMs. The degradation of both the long- and short-chain PyrSM was decreased, but the effect was stronger for the long-chain species. The hydrolysis of Pyr₁₄SM was reduced approximately fourfold and that of Pyr₄SM only approximately twofold (Figure 9C). This agrees with the enhanced PM retrieval of the long-chain PyrSM in Hrs + Tsg101-depleted cells. Instead, NPC1 depletion did not alter PyrSM degradation despite the intracellular sequestration of the long-chain PyrSM (Figure 9C). The long-chain SM might be targeted from late endosomes to other endomembranes, such as the Golgi, although no evident redistribution of the long-chain PyrSM to the Golgi was observed in NPC1 RNAi cells (Figure 8A). Alternatively, the unaltered degradation of the long-chain SM may reflect increased shunting of the lipid to the degradative compartments in combination with posttranslational inhibition of acid SMase in NPC cells (Reagan et al., 2000).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
In this work, we introduced pyrene-labeled and unlabeled SMs into the PM, and we addressed the question whether the acyl chain length plays a role the endocytic routing of SM.

We found that the initial internalization of PyrSMs from the PM was rapid and occurred with closely similar kinetics for all PyrSM species. Soon after internalization, all PyrSMs entered early endosomes as judged by colocalization with DiI-LDL, which had been internalized for 10 min. Moreover, the Rab5Q79L-GFP–containing compartments were PyrSM positive immediately after labeling, indicating rapid uptake of both the short- and long-chain species into the enlarged early endosomes. In contrast, our results provide several lines of evidence to suggest that SMs differing only in their acyl chain length are differentially recycled to the PM.

We found that after a short (30 min–2 h) chase, a smaller fraction of the long-chain PyrSM was returned to the PM compared with the short-chain PyrSM. A similar difference was observed between unlabeled short- and long-chain SMs, indicating that the difference was not induced by the fluorescent modification. The involvement of specific proteins was then studied by using RNAi. Depletion of Hrs and Tsg101 selectively enhanced the PM recycling of the long-chain SM. In contrast, depletion of NPC1 inhibited the recycling of the long-chain SM, and a similar effect was observed upon cholesterol loading. A possible interpretation is that the long-chain SM was preferentially targeted Hrs and Tsg101 dependently toward late endosomal compartments and recycled to the plasma membrane in an NPC1- and cholesterol-dependent manner, whereas the short-chain SM recycled more effectively, by using Hrs-, Tsg101-, and NPC1-independent routing that was insensitive to cholesterol loading (see Figure 10 for a schematic model). This model is based on the available information on the localization and sites of action of Hrs, Tsg101, and NPC1. However, it should be pointed out that the assays used do not identify the precise sites or mechanisms of SM recycling. Therefore, other scenarios not necessarily involving differential recycling pathways for the short-and long-chain SMs could also be envisaged. Moreover, the routing of the short-and long-chain SMs probably overlaps in part, as suggested by the moderate differences in biochemical analyses and relatively similar subcellular distributions of the fluorescent SMs (also when comparing the extent of colocalization e.g., to early, recycling and late endosomal Rab GTPases; our unpublished data).

View larger version (37K): [in this window] [in a new window]   Figure 10. Schematic model of SM recycling. Both long- and short-chain SMs internalized from the PM enter early endosomes (EE). Short-chain SMs are recycled from the limiting EE membrane to the PM via tubular intermediates. This route may involve the endocytic recycling compartment (not shown). Long-chain SMs are sorted to the inward invaginating membranes of EE that form intraendosomal vesicles of MVBs, from where they enter late endosomes (LE). Long-chain SMs may be recycled from LE to the PM or be targeted to lysosomes (LYS) for degradation by acid SMase. Hrs and Tsg101 are involved in the incorporation of the long-chain SMs into MVBs, whereas NPC1 may be involved in their recycling from late endosomes to the PM. All membranes consist of mixtures of ordered and disordered domains (not shown), but membranes of the EE recycling tubules are more disordered than those of the inward invaginating membranes of EE and internal vesicles of MVB and LE. Both differ from intralysosomal, hydrolytic membranes. Sorting efficiency is not absolute, i.e., some long-chain SM is recycled from EE, whereas some short-chain SM enters LE (not shown).

The mechanisms mediating inward budding of the endosome membrane and assigning particular lipids into the endovesicles of MVBs are not well understood. Previous studies have implicated bis(monacylglycero)phosphate (BMP, also known as lysobisphosphatidic acid) and the Alix protein in the invagination process (Matsuo et al., 2004). We envisage that components of the ESCRT complexes, possibly together with BMP and Alix, may increase the curvature of the limiting membrane locally and thus drive lipid sorting. Long-chain SMs could enrich in the highly curved invaginating domains because in the presence of cholesterol, their physical properties are more compatible with a high membrane curvature. Notably, studies in model membranes indicate that membranes consisting of cholesterol and 24:0-SM have a lower bending modulus (i.e., bend more easily) than those consisting of cholesterol and 16:0-SM (Li et al., 2003).

Inclusion into the intralumenal vesicles of MVBs is a mechanism to ferry protein and lipid cargo to lysosomes for degradation as MVBs fuse with the late endosomal hydrolytic organelles (Hurley and Emr, 2006). Accordingly, in Hrs + Tsg101-depleted cells we observed a striking inhibition in the degradation of the long-chain PyrSM. To our knowledge, these data provide the first evidence that ESCRT proteins are involved in the delivery of lipids for degradation. Earlier studies indicate that sphingolipid degradation takes place in intraendosomal/lysosomal membranes that contain high amounts of BMP and low amounts of cholesterol (Kolter and Sandhoff, 2005). This membrane composition favoring hydrolysis may, in part, be generated by the removal of cholesterol from the late endocytic organelles by NPC1 (see below). Cholesterol loading should thus slow down sphingolipid degradation by increasing the cholesterol content in the endosomal membranes. This may explain why SM degradation was not enhanced in NPC1-depleted cells.

Our findings also provide insights regarding the function of NPC1. The NPC defect is associated with mistrafficking of a variety of membrane-associated molecules to the endo/lysosomal storage organelles, but its relationship to the loss of NPC1 function is not clear. BODIPY-lactosylceramide and BODIPY-SM that are normally targeted to the Golgi, accumulated in the endo/lysosomal compartment in NPC cells (Puri et al., 1999; Puri et al., 2001). The NPC defect also affects early and recycling compartments, as indicated by cholesterol loading of early endosomes (Choudhury et al., 2004) and defective transferrin receptor recycling (Pipalia et al., 2007) in NPC cells. Moreover, DiIC₁₂ and DiIC₁₆ dyes, which in normal cells were differentially routed to the endocytic recycling compartment and late endosomes, respectively (Mukherjee et al., 1999), were both directed to late endocytic compartments in NPC cells (Pipalia et al., 2007). In contrast to the DiI probes, we found that the PM recycling of the short-chain SM was not affected in NPC1 depleted cells, whereas that of the long-chain species was impaired. Thus, the present data provide the first indication for a selective, chain length dependent sphingolipid-recycling defect in cells lacking NPC1 function.

We have previously shown that NPC1 is involved in recycling cholesterol to the PM and in reducing DRMs in late endosomal organelles (Lusa et al., 2001). This agrees with the enrichment of cholesterol in the internal membranes of early and late endosomes and depletion in the internal membranes of lysosomes (Mobius et al., 2003). It is also compatible with the present data, because NPC1 might regulate the recycling of cholesterol–sphingolipid-rich intraluminal contents from late endosomes to the PM (Figure 10). NPC1 could be involved, e.g., in the generation of tubules emanating from the limiting membrane or at an earlier step, in the back-fusion of intraluminal vesicles with the limiting late endosomal membrane (van der Goot and Gruenberg, 2006). Cholesterol removal from the late endocytic organelles should serve to generate a cholesterol-poor intralysosomal milieu favoring sphingolipid digestion (Kolter and Sandhoff, 2005). How individual sphingolipid molecules are destined for recycling versus degradation remains to be addressed.

In summary, this work provides evidence for differential endocytic trafficking of SMs depending on their acyl chain length: short-chain SMs recycle to the PM more effectively and independently of NPC1, whereas long-chain SMs are preferentially routed to the late endocytic pathway and recycle in an NPC1- and cholesterol-dependent manner. Different domain preferences of the short- and long-chain SMs are considered to contribute to their differential sorting. The affinity of the long-chain PyrSMs for DRMs, their slow degradation compared with other fluorescent lipids, e.g., BODIPY-SM, and potential for probing sphingolipid recycling from late endocytic compartments to the PM, make long-chain PyrSMs attractive tools for further studies on sphingolipid trafficking and the role of sterol-sphingolipid domains in cells.

	ACKNOWLEDGMENTS

 
We thank Harald Stenmark, Joost Holthuis, and Martin Hermansson for critical reading of the manuscript, Birgitta Rantala and Tarja Grundström for technical assistance, Prabuddha Sengupta for advice on PM vesicle isolation, Andreas Upphof and Martin Hermansson for help with MS analyses, and Kimmo Tanhuanpää for imaging support. This study was supported by grants from the Academy of Finland (to M.K., E.I., and P.S.), the Sigrid Juselius Foundation (to P.S. and E.I.), and the Finnish Cultural Foundation and Biocentrum Helsinki (to E.I.).

	Footnotes

 
This article was published online ahead of print in MBC in Press (http://www.molbiolcell.org/cgi/doi/10.1091/mbc.E07-04-0330) on October 17, 2007.

The online version of this article contains supplemental material at MBC Online (http://www.molbiolcell.org).

Present address: Programme in Cell Biology, Hospital for Sick Children, 555 University Ave., Toronto, ON, M5G 1X8, Canada.

Address correspondence to: Elina Ikonen (elina.ikonen{at}helsinki.fi).

Abbreviations used: BMP, bis(monoacylglycero)phosphate; CD, cyclodextrin; DRM, detergent-resistant membrane; ESCRT, endosomal sorting complex required for transport; ESI-MS, electrospray ionization mass spectrometry; Hrs, hepatocyte growth factor receptor substrate; MVB, multivesicular body; NPA, Niemann-Pick type A; NPC, Niemann-Pick type C; PM, plasma membrane; PyrSM, pyrenylacylsphingomyelin; SM, sphingomyelin; SMase, sphingomyelinase; TNP-LPE, trinitrophenyl-lysophosphatidylethanolamine; Tsg, tumor susceptibility gene.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Ahmad, T. Y., Sparrow, J. T., and Morrisett, J. D. (1985). Fluorine-, pyrene-, and nitroxide-labeled sphingomyelin: semi-synthesis and thermotropic properties. J. Lipid Res 26, 1160–1165.[Abstract]

Bartlett, E. M., and Lewis, D. H. (1970). Spectrophotometric determination of phosphate esters in the presence and absence of orthophosphate. Anal. Biochem 36, 159–167.[CrossRef][Medline]

Blom, T. S., Koivusalo, M., Kuismanen, E., Kostiainen, R., Somerharju, P., and Ikonen, E. (2001). Mass spectrometric analysis reveals an increase in plasma membrane polyunsaturated phospholipid species upon cellular cholesterol loading. Biochemistry 40, 14635–14644.[CrossRef][Medline]

Brown, D. A., and London, E. (1998). Functions of lipid rafts in biological membranes. Annu. Rev. Cell Dev. Biol 14, 111–136.[CrossRef][Medline]

Choudhury, A., Sharma, D. K., Marks, D. L., and Pagano, R. E. (2004). Elevated endosomal cholesterol levels in Niemann-Pick cells inhibit rab4 and perturb membrane recycling. Mol. Biol. Cell 15, 4500–4511.[Abstract/Free Full Text]

Eklund, K. K., Virtanen, J. A., Kinnunen, P. K., Kasurinen, J., and Somerharju, P. J. (1992). Conformation of phosphatidylcholine in neat and cholesterol-containing liquid-crystalline bilayers. Application of a novel method. Biochemistry 31, 8560–8565.[CrossRef][Medline]

Folch, J., Lees, M., and Sloane Stanley, G. H. (1957). A simple method for the isolation and purification of total lipides from animal tissues. J. Biol. Chem 226, 497–509.[Free Full Text]

Fridriksson, E. K., Shipkova, P. A., Sheets, E. D., Holowka, D., Baird, B., and McLafferty, F. W. (1999). Quantitative analysis of phospholipids in functionally important membrane domains from RBL-2H3 mast cells using tandem high-resolution mass spectrometry. Biochemistry 38, 8056–8063.[CrossRef][Medline]

Ganley, I. G., and Pfeffer, S. R. (2006). Cholesterol accumulation sequesters Rab9 and disrupts late endosome function in NPC1-deficient cells. J. Biol. Chem 281, 17890–17899.[Abstract/Free Full Text]

Haimi, P., Uphoff, A., Hermansson, M., and Somerharju, P. (2006). Software tools for analysis of mass spectrometric lipidome data. Anal. Chem 78, 8324–8331.[Medline]

Hao, M., Mukherjee, S., Sun, Y., and Maxfield, F. R. (2004). Effects of cholesterol depletion and increased lipid unsaturation on the properties of endocytic membranes. J. Biol. Chem 279, 14171–14178.[Abstract/Free Full Text]

Harder, T., Scheiffele, P., Verkade, P., and Simons, K. (1998). Lipid domain structure of the plasma membrane revealed by patching of membrane components. J. Cell Biol 141, 929–942.[Abstract/Free Full Text]

Hermansson, M., Kakela, R., Berghall, M., Lehesjoki, A. E., Somerharju, P., and Lahtinen, U. (2005). Mass spectrometric analysis reveals changes in phospholipid, neutral sphingolipid and sulfatide molecular species in progressive epilepsy with mental retardation, EPMR, brain: a case study. J. Neurochem 95, 609–617.[CrossRef][Medline]

Holowka, D., and Baird, B. (1983). Structural studies on the membrane-bound immunoglobulin E-receptor complex. 1. Characterization of large plasma membrane vesicles from rat basophilic leukemia cells and insertion of amphipathic fluorescent probes. Biochemistry 22, 3466–3474.[CrossRef][Medline]

Holtta-Vuori, M., Alpy, F., Tanhuanpaa, K., Jokitalo, E., Mutka, A. L., and Ikonen, E. (2005). MLN64 is involved in actin-mediated dynamics of late endocytic organelles. Mol. Biol. Cell 16, 3873–3886.[Abstract/Free Full Text]

Holtta-Vuori, M., Maatta, J., Ullrich, O., Kuismanen, E., and Ikonen, E. (2000). Mobilization of late-endosomal cholesterol is inhibited by Rab guanine nucleotide dissociation inhibitor. Curr. Biol 10, 95–98.[CrossRef][Medline]

Holtta-Vuori, M., Tanhuanpaa, K., Mobius, W., Somerharju, P., and Ikonen, E. (2002). Modulation of cellular cholesterol transport and homeostasis by Rab11. Mol. Biol. Cell 13, 3107–3122.[Abstract/Free Full Text]

Hurley, J. H., and Emr, S. D. (2006). The ESCRT complexes: structure and mechanism of a membrane-trafficking network. Annu. Rev. Biophys. Biomol. Struct 35, 277–298.[CrossRef][Medline]

Ikonen, E., and Holtta-Vuori, M. (2004). Cellular pathology of Niemann-Pick type C disease. Semin. Cell Dev. Biol 15, 445–454.[CrossRef][Medline]

Koivusalo, M., Alvesalo, J., Virtanen, J. A., and Somerharju, P. (2004). Partitioning of pyrene-labeled phospho- and sphingolipids between ordered and disordered bilayer domains. Biophys. J 86, 923–935.[Abstract/Free Full Text]

Kolter, T., and Sandhoff, K. (2005). Principles of lysosomal membrane digestion: stimulation of sphingolipid degradation by sphingolipid activator proteins and anionic lysosomal lipids. Annu. Rev. Cell Dev. Biol 21, 81–103.[CrossRef][Medline]

Koval, M., and Pagano, R. E. (1990). Sorting of an internalized plasma membrane lipid between recycling and degradative pathways in normal and Niemann-Pick, type A fibroblasts. J. Cell Biol 111, 429–442.[Abstract/Free Full Text]

Levade, T., Andrieu-Abadie, N., Segui, B., Auge, N., Chatelut, M., Jaffrezou, J. P., and Salvayre, R. (1999). Sphingomyelin-degrading pathways in human cells role in cell signalling. Chem. Phys. Lipids 102, 167–178.[CrossRef][Medline]

Levade, T., Gatt, S., Maret, A., and Salvayre, R. (1991a). Different pathways of uptake and degradation of sphingomyelin by lymphoblastoid cells and the potential participation of the neutral sphingomyelinase. J. Biol. Chem 266, 13519–13529.[Abstract/Free Full Text]

Levade, T., Gatt, S., and Salvayre, R. (1991b). Uptake and degradation of several pyrenesphingomyelins by skin fibroblasts from control subjects and patients with Niemann-Pick disease. Effect of the structure of the fluorescent fatty acyl residue. Biochem. J 275, 211–217.[Medline]

Li, X. M., Momsen, M. M., Brockman, H. L., and Brown, R. E. (2003). Sterol structure and sphingomyelin acyl chain length modulate lateral packing elasticity and detergent solubility in model membranes. Biophys. J 85, 3788–3801.[Abstract/Free Full Text]

Liscum, L. (2000). Niemann-Pick type C mutations cause lipid traffic jam. Traffic 1, 218–225.[CrossRef][Medline]

Liu, B., and Hannun, A. Y. (1999). Sphingomyelinase assay using radiolabeled substrate. Methods Enzymol 311, 164–167.

Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951). Protein measurement with the Folin phenol reagent. J. Biol. Chem 193, 265–275.[Free Full Text]

Lusa, S., Blom, T. S., Eskelinen, E. L., Kuismanen, E., Mansson, J. E., Simons, K., and Ikonen, E. (2001). Depletion of rafts in late endocytic membranes is controlled by NPC1-dependent recycling of cholesterol to the plasma membrane. J. Cell Sci 114, 1893–1900.[Abstract]

Lusa, S., Myllarniemi, M., Volmonen, K., Vauhkonen, M., and Somerharju, P. (1996). Degradation of pyrene-labelled phospholipids by lysosomal phospholipases in vitro. Dependence of degradation on the length and position of the labelled and unlabelled acyl chains. Biochem. J 315, 947–952.[Medline]

Matsuo, H. et al. (2004). Role of LBPA and Alix in multivesicular liposome formation and endosome organization. Science 303, 531–534.[Abstract/Free Full Text]

Mobius, W., van Donselaar, E., Ohno-Iwashita, Y., Shimada, Y., Heijnen, H. F., Slot, J. W., and Geuze, H. J. (2003). Recycling compartments and the internal vesicles of multivesicular bodies harbor most of the cholesterol found in the endocytic pathway. Traffic 4, 222–231.[Medline]

Mukherjee, S., Soe, T. T., and Maxfield, F. R. (1999). Endocytic sorting of lipid analogues differing solely in the chemistry of their hydrophobic tails. J. Cell Biol 144, 1271–1284.[Abstract/Free Full Text]

Naoi, M., Lee, Y. C., and Roseman, S. (1974). Rapid and sensitive determination of sphingosine bases and sphingolipids with fluorescamine. Anal. Biochem 58, 571–577.[CrossRef][Medline]

Pagano, R. E., and Martin, C. O. (1994). Use of fluorescent analogs of ceramide to study the golgi apparatus of animal cells. In: Cell Biology: A Laboratory Handbook, San Diego, CA: Academic Press, 387–393.

Pagano, R. E., Watanabe, R., Wheatley, C., and Chen, C. S. (1999). Use of N-[5-(5,7-dimethyl boron dipyrromethene difluoride-sphingomyelin to study membrane traffic along the endocytic pathway. Chem. Phys. Lipids 102, 55–63.[CrossRef][Medline]

Pipalia, N. H., Hao, M., Mukherjee, S., and Maxfield, F. R. (2007). Sterol, protein and lipid trafficking in chinese hamster ovary cells with Niemann-Pick type c1 defect. Traffic 8, 130–141.[Medline]

Puri, V., Watanabe, R., Dominguez, M., Sun, X., Wheatley, C. L., Marks, D. L., and Pagano, R. E. (1999). Cholesterol modulates membrane traffic along the endocytic pathway in sphingolipid-storage diseases. Nat. Cell Biol 1, 386–388.[CrossRef][Medline]

Puri, V., Watanabe, R., Singh, R. D., Dominguez, M., Brown, J. C., Wheatley, C. L., Marks, D. L., and Pagano, R. E. (2001). Clathrin-dependent and -independent internalization of plasma membrane sphingolipids initiates two Golgi targeting pathways. J. Cell Biol 154, 535–547.[Abstract/Free Full Text]

Razi, M., and Futter, C. E. (2006). Distinct roles for Tsg101 and Hrs in multivesicular body formation and inward vesiculation. Mol. Biol. Cell 17, 3469–3483.[Abstract/Free Full Text]

Reagan, J. W., Jr, Hubbert, M. L., and Shelness, G. S. (2000). Posttranslational regulation of acid sphingomyelinase in Niemann-Pick type C1 fibroblasts and free cholesterol-enriched chinese hamster ovary cells. J. Biol. Chem 275, 38104–38110.[Abstract/Free Full Text]

Roux, A., Cuvelier, D., Nassoy, P., Prost, J., Bassereau, P., and Goud, B. (2005). Role of curvature and phase transition in lipid sorting and fission of membrane tubules. EMBO J 24, 1537–1545.[CrossRef][Medline]

Sassaroli, M., Ruonala, M., Virtanen, J., Vauhkonen, M., and Somerharju, P. (1995). Transversal distribution of acyl-linked pyrene moieties in liquid-crystalline phosphatidylcholine bilayers. A fluorescence quenching study. Biochemistry 34, 8843–8851.[CrossRef][Medline]

Scott, R. E., Perkins, R. G., Zschunke, M. A., Hoerl, B. J., and Maercklein, P. B. (1979). Plasma membrane vesiculation in 3T3 and SV3T3 cells. I. Morphological and biochemical characterization. J. Cell Sci 35, 229–243.[Abstract]

Shaw, J. E., Epand, R. F., Epand, R. M., Li, Z., Bittman, R., and Yip, C. M. (2006). Correlated fluorescence-atomic force microscopy of membrane domains: structure of fluorescence probes determines lipid localization. Biophys. J 90, 2170–2178.[Abstract/Free Full Text]

Silversand, C., and Haux, C. (1997). Improved high-performance liquid chromatographic method for the separation and quantification of lipid classes: application to fish lipids. J. Chromatogr. B Biomed. Sci. Appl 703, 7–14.[CrossRef][Medline]

Simons, K., and Ikonen, E. (1997). Functional rafts in cell membranes. Nature 387</