A caveolin dominant negative mutant associates with lipid bodies and induces intracellular cholesterol imbalance

Abstract

Recent studies have indicated a role for caveolin in regulating cholesterol-dependent signaling events. In the present study we have analyzed the role of caveolins in intracellular cholesterol cycling using a dominant negative caveolin mutant. The mutant caveolin protein, cav-3(DGV), specifically associates with the membrane surrounding large lipid droplets. These structures contain neutral lipids, and are accessed by caveolin 1-3 upon overexpression. Fluorescence, electron, and video microscopy observations are consistent with formation of the membrane-enclosed lipid rich structures by maturation of subdomains of the ER. The caveolin mutant causes the intracellular accumulation of free cholesterol (FC) in late endosomes, a decrease in surface cholesterol and a decrease in cholesterol efflux and synthesis. The amphiphile U18666A acts synergistically with cav(DGV) to increase intracellular accumulation of FC. Incubation of cells with oleic acid induces a significant accumulation of full-length caveolins in the enlarged lipid droplets. We conclude that caveolin can associate with the membrane surrounding lipid droplets and is a key component involved in intracellular cholesterol balance and lipid transport in fibroblasts.

Figure 1

Cav^DGV is targeted to the membrane of intracellular vesicular structures. a, BHK cells were transfected with HA-tagged cav-3^DGV for 24 h and were then labeled with antibodies to HA tag, followed by specific Cy3-conjugated secondary antibodies. Cav^DGV mainly located in enlarged intracellular rings, and some protein was detected in the tubular elements of the ER (see Fig. 2) and in the perinuclear region (arrows and g and h). b, BHK cells were transfected with HA-tagged cav^DGV for 24 h and were then treated with 5 μg/ml of cycloheximide for an additional 24 h. After blocking of protein synthesis, the truncated protein was exclusively detected in very enlarged intracellular vesicles (arrows). c, BHK cells were transfected with GFP-tagged cav^DGV and frozen sections were processed for immunoelectron microscopy using anti-GFP antibodies, followed by protein A–gold. The mutant protein was detected enriched on the membrane of electron-lucent vesicles of 300–500-nm diam, often surrounded by intermediate filaments (arrows). d and e, Lowicryl sections (of freeze-substituted/low temperature embedded sections) of GFP-tagged cav^DGV-transfected cells labeled with antibodies to GFP followed by protein A–gold. Cav^DGV-containing vesicles (CDV) comprise a membrane bilayer (see arrows in d and e) surrounding a putative lipid-filled area. f, Epon sections showing a morphologically similar structure connected to tubular elements possibly corresponding to the ER (arrow and arrowhead), which were frequently observed in nontransfected cells. The typical morphology of lipid-filled body enclosed by a bilayer structure (see insert) is readily observed in epon sections. g, BHK cells were cotransfected with myc-tagged sialotransferase (a specific marker for trans Golgi) and HA-tagged cav^DGV and the distribution of the proteins studied by immunofluorescence by means of antitag specific antibodies. The distribution of sialotransferase (green) and cav^DGV (red) clearly overlapped in the Golgi area, however CDV rings did not contain the Golgi marker. h, HA-tagged cav^DGV-transfected cells were label with an mAb to cav-1 (which exclusively recognizes the Golgi complex conformation of cav-1) and a polyclonal antibody to HA-tag. When cav^DGV (red) accumulated in the Golgi region, the truncated protein clearly colocalized with the Golgi complex-associated pool of endogenous cav-1 (green). i and l, HA-tagged cav^DGV-transfected cells were labeled with polyclonal antibodies to the peroxisome marker SKL (l) or the late endosome marker, lysobisphosphatidic acid (h) and with an mAb to the HA-tag. The endosome and peroxisome markers (green) were excluded from the CDV compartment (red). j, Cav^DGV-transfected cells were incubated for 2 h at 37°C with 10 μg/ml of GM1 to allow insertion of the ganglioside into the plasma membrane and internalization. GM1, detected by using 1 μg/ml of Cholera toxin–FITC, was completely excluded from the CDV compartment (red). k, BHK cells were cotransfected with Niemann-Pick C1 protein and cav^DGV for 24 h and were then labeled with a polyclonal antibody to NPC1 and the mAb to the HA tag. Transfected NPC1 (green) does not colocalize with the CDV vesicles (red). Bars: (a, b, and g–l) 5 μm; (c–f) 100 nm.

Figure 3

The cav^DGV phenotype is a feature of all caveolin family members. a–c, BHK cells were cotransfected with full-length VSV-G–tagged cav-1 and HA-tagged cav^DGV, and after 24 h were labeled with a rabbit antibody to VSV-G tag and a mouse antibody to HA tag. In a low but significant proportion of transfected cells (2–5% of transfectants) full-length cav-1 (green) accumulated in the same intracellular rings as cav^DGV (red). Although in those cells, cav-1 was detected on the PM (a and c, arrowheads) cav^DGV was totally excluded from the PM. Some ring-like structures were strongly labeled for full-length cav-1 but not for cav^DGV (b and c arrows). d and e, BHK cells were transfected with HA-tagged cav-3^DGV equivalent truncation mutants of cav-1 (cav-1^DGI, d) and cav-2 (cav-2^DKV, e) and the cells labeled with mouse anti-HA tag. Both equivalent mutants accumulated in morphologically identical rings to cav-3^DGV, demonstrating that the CDV compartment is a general characteristic of all the caveolin family members. f, BHK cells were transfected with HA-tagged cav-3^LLS (a cav-3 truncation lacking the entire NH₂-terminal domain up to the putative intramembrane region) and labeled with mouse anti-HA tag. Cav-3^LLS accumulated, as cav-3^DGV, in CDV-like structures, showing that the caveolin scaffolding domain is not involved in the targeting of the protein to this compartment. Bars, 5 μm.

Figure 2

CDV rings contain markers for the ER. a–d, BHK cells were transfected with cav^DGV for 24 h, and were then labeled with a polyclonal antibody to PDI (a marker for the ER) and an mAb to the HA tag. PDI (green) was located in the reticular network characteristic of the ER (b) and in some vesicular structures that colocalized to some extent with CDV rings (red) (b and c, arrows). The colocalization between both proteins was corroborated in frozen sections of cav^DGV-transfected cells (d) labeled with the same antibodies, followed by anti-rabbit antibodies conjugated to 15-nm gold (arrows) and anti-mouse antibodies conjugated to 5-nm gold (arrowheads). e–g, BHK cells were double transfected with CFP-tagged ER marker and HA-tagged cav^DGV. The cells were labeled with a polyclonal antibody to GFP (to increase the signal) and the mAb to HA tag (e). Clear colocalization could be observed between some of the CDV rings (red) and the ER marker (green) (f and g, arrows). Bars: (d) 100 nm; (a–e) 5 μm.

Figure 4

Real-time microscopy. CDV are immobile structures formed by maturation processes in some specific regions of the ER. a, Living GFP-tagged cav^DGV-transfected cells were incubated in a medium containing a marker for acidic compartments, lysoTracker. Cells expressing GFP-cav^DGV were selected under the microscope and the focus was kept constant. Images of both channels (green/red) with a delay of 10 s between channels were captured over 5 min. The figure shows a selected sequence of 6 consecutive frames (20-s interval between images) representing a total period of 100 s. LysoTracker-labeled vesicles showed a rapid bidirectional motility (arrows) but cav^DGV–GFP positive vesicles showed no significant motility in any axis (arrowheads). b and c, In an attempt to capture the formation of CDVs, cells were transfected with GFP-tagged cav^DGV for 18 h and were then observed by video microscopy. Cells were selected for the presence of forming rings (see arrows in the first panel) and the focus was kept constant during the entire experiment. Images were captured with a constant interval of 7 min. The figure shows a selected sequence of 10 consecutive frames representing a total period of 63 min. At time 0, cav^DGV–GFP was abundant in the tubular elements of the ER and in the Golgi area. Progressively the protein is less abundant in the ER and Golgi complex and accumulates in the forming rings (arrows) to shape well-defined CDV ring-like structures in ∼60 min. c shows a selected high magnification example of CDV formation. The protein can be detected moving in waves along the tubules of the ER (arrowheads) until the formation of the DGV-containing vacuolar structures (arrows). Bars: (a and b) 5 μm; (c) 1 μm. Supplemental videos are available at http://www.jcb.org/cgi/content/full/152/5/1057/DC1.

Figure 5

Biochemical and morphological characterization of the CDV compartment. a, HA-tagged cav^DGV expressing cells were incubated in a medium containing 5 μg/ml of BFA for 4 h and were then labeled with a polyclonal antibody to p23 (a cis Golgi marker) and an mAb to the HA tag, followed by secondary antibodies and finally incubated with DAPI to visualize the resistant cells. BFA treatment induced a dramatic tubulation of Golgi stacks (green), however CDV rings (red) were completely resistant to the treatment. b and c, HA-tagged cav^DGV-transfected cells were incubated before fixation in a tris-buffered medium containing 1% Triton X-100 at 4°C for 3 min and were then fixed and labeled with an mAb to β-tubulin and a polyclonal antibody to HA tag, followed for secondary antibodies and DAPI to visualize the resistant cells. The CDV compartment (green) and the microtubule network (red) were both resistant to the extraction with the detergent. When the cells were pretreated in a medium containing 1.6 μM nocodazole for 4 h and were then extracted with the detergent, the microtubule network appeared completely disorganized in a very diffuse cytoplasmic staining (compare c with b) but CDV rings remained resistant to the detergent. d and e, Epon sections of detergent extracted GFP-tagged cav^DGV-transfected cells were labeled with antibodies to GFP followed by protein A–gold before embedding. Cav^DGV was detected in electron-dense aggregates (arrowheads) on the membrane of CDV ring-shaped structures surrounded by a network of intermediate filaments (arrows in d). Other internal membranes like the nuclear envelope (arrows in e) or mitochondrial membranes were solubilized. The membranes of the rough ER did not contain cav^DGV (asterisk in d). f–h, Cav^DGV-transfected cells were extracted with 1% Triton X-100 at 4°C for 3 min and were then labeled with an mAb to HA tag and a polyclonal antibody to PDI. Some domains of the ER were resistant to the extraction with detergents (h, compare with Fig. 2 e) and clearly colocalized with cav^DGV. i, Cav^DGV-transfected cells were homogenized at 4°C in a Tris buffered solution or in Tris buffered solution containing pH 11 0.5 M NaCO₃ or 1% Triton X-100. Then, a postnuclear supernatant of each treatment was centrifuged at 75,000 rpm for 30 min to separate membrane associated proteins (p75) or soluble proteins (sol). 5 μg of protein of each sample were resolved by SDS PAGE electrophoresis, transferred to p-immobilon membranes and the presence of HA, cav-1, and PDI detected by Western blotting by means of specific antibodies. The three studied proteins were associated with the membrane fraction of the cells homogenized in the control buffer and in the buffer containing NaCO₃. Although cav-1 was resistant to the solubilization with 1% Triton, HA and PDI were solubilized. Bars: (d and e) 100 nm; (a–c and f) 5 μm.

Figure 7

Full-length caveolins and cav^DGV accumulate on lipid droplets formed in response to oleic acid treatment. a–c, BHK cells preincubated in a medium containing oleic acid for 16 h, were transfected with HA-tagged cav^DGV, incubated for an additional 24 h in a medium containing oleic acid, labeled with antibodies to HA tag and finally mounted in mowiol containing Nile red (see Materials and Methods for details). Cav^DGV (b) was exclusively detected surrounding very enlarged Nile red-positive lipid droplets (c). d–i, Untreated BHK (d, e, and f) or cells preincubated in a medium containing oleic acid for 16 h (g, h, and i) were transfected with YFP-tagged cav-3 (d and g), HA-tagged cav-3 (e and h), or myc-tagged cav-2 (f and i) and incubated for an additional 24 h in medium with or without oleic acid. Cells transfected with HA or myc tag were labeled with specific antibodies and cells transfected with the YFP-tagged constructs were mounted without any additional manipulation after fixation. 65% of cells transfected with cav-3-YFP showed, in addition to PM and Golgi complex usual labeling, some intracellular rings (d). In contrast, very few rings were observed in those cells transfected with HA-cav-3 (e). After oleic acid treatment, in 98% of the cells transfected with YFP-tagged (g) or 52% of the cells transfected with HA-tagged cav-3 (h) the protein, in addition to the PM labeling, was observed on lipid droplets (Nile red staining not shown). Myc-tagged cav-2 showed a dramatic redistribution from the Golgi area in control cells (f) to the lipid droplets after oleic acid treatment (i). Bars, 5 μm.

Figure 6

CDV is a lipid-enriched compartment and cav^DGV expression induces lipid accumulation. a–c, Cav^DGV HA-tagged transfected cells were labeled with mAbs to HA, followed by Cy3-conjugated secondary antibodies. The samples were mounted in mowiol, containing a solution of the lipid probe, Nile red. The cells were selected for the presence of rings in the red channel and the corresponding image in the green channel was captured. Then the samples were exposed for 30 s in the red channel before the corresponding image was captured. Nile red staining was within the center of CDVs, whereas cav^DGV formed the characteristic ring around the vesicle (arrowheads and merge panel). Cells expressing cav^DGV showed a marked accumulation of lipids, in CDVs, and in other vesicular structures (arrows in c), compared with nontransfected cells. d–f, Cav^DGV HA-tagged transfected cells were labeled with filipin (to visualize cellular FC, see Materials and Methods for details) and with an mAb to the HA tag. Cav^DGV expression induced a marked redistribution of filipin: 65% of cav^DGV expressing cells did not contain filipin on the PM (arrows in d and f) and in 72% of the transfected cells filipin was detected in very enlarged cytoplasmic vesicles (arrowheads and insert f). g–j, Cav^DGV HA-tagged transfected cells were labeled with filipin (blue) and with antibodies to the HA tag (red) and to LBPA (green). Cav^DGV expression induced redistribution of FC to very enlarged late endosomes (arrows in i and j). k–m, BHK were transfected with cav^DGV in a normal medium (k) or in a medium containing 4.5 μg/ml of U18666A (l and m) for different times and cholesterol distribution was monitored by means of filipin. In cells expressing cav^DGV, a complete accumulation of lipids in late endosomes was evident after 24–32 h (k). However, in those cells expressing cav^DGV in the presence of U18666A the complete accumulation of FC in late endosomes was evident after 16 h (l), demonstrating that the combination of the drug and the cav^DGV results in an earlier lysosomal storage disorder. Finally, after 24 h few differences could be observed between nonexpressing and cav^DGV expressing cells, but some very enlarged intracellular vesicles were present in transfected cells (arrows in m). Bar, 5 μm.