Dispensability and dynamics of caveolin-1 during liver regeneration and in isolated hepatic cells

Authors

  • Rafael Mayoral,

    1. Instituto de Investigaciones Biomédicas Alberto Sols, Consejo Superior de Investigaciones Científicas, Universidad Autónoma de Madrid, Madrid, Spain
    2. Centro Nacional de Investigaciones Cardiovasculares, Madrid, Spain
    Search for more papers by this author
  • Amalia Fernández-Martínez,

    1. Instituto de Investigaciones Biomédicas Alberto Sols, Consejo Superior de Investigaciones Científicas, Universidad Autónoma de Madrid, Madrid, Spain
    2. Centro Nacional de Investigaciones Cardiovasculares, Madrid, Spain
    Search for more papers by this author
  • Rosa Roy,

    1. Instituto de Investigaciones Biomédicas Alberto Sols, Consejo Superior de Investigaciones Científicas, Universidad Autónoma de Madrid, Madrid, Spain
    2. Centro Nacional de Investigaciones Cardiovasculares, Madrid, Spain
    Search for more papers by this author
  • Lisardo Boscá,

    Corresponding author
    1. Instituto de Investigaciones Biomédicas Alberto Sols, Consejo Superior de Investigaciones Científicas, Universidad Autónoma de Madrid, Madrid, Spain
    2. Centro Nacional de Investigaciones Cardiovasculares, Madrid, Spain
    • Instituto de Investigaciones Biomédicas Alberto Sols, Consejo Superior de Investigaciones Científicas, Arturo Duperier 4, 28029 Madrid, Spain
    Search for more papers by this author
    • fax: 34915854401

  • Paloma Martín-Sanz

    Corresponding author
    1. Instituto de Investigaciones Biomédicas Alberto Sols, Consejo Superior de Investigaciones Científicas, Universidad Autónoma de Madrid, Madrid, Spain
    2. Centro Nacional de Investigaciones Cardiovasculares, Madrid, Spain
    • Instituto de Investigaciones Biomédicas Alberto Sols, Consejo Superior de Investigaciones Científicas, Arturo Duperier 4, 28029 Madrid, Spain
    Search for more papers by this author
    • fax: (34) 915854401


  • Potential conflict of interest: Nothing to report.

Abstract

Caveolae participate in several cellular processes such as vesicular transport, cholesterol homeostasis, regulation of signal transduction, integrin signaling, and cell growth. The expression and functional role of caveolin (Cav), the most abundant protein of caveolae, has been reported in liver and in different hepatocyte cell lines, in human cirrhotic liver, and in hepatocellular carcinomas. The role of Cav-1 in liver regeneration after partial hepatectomy (PH) has been investigated as a model of liver proliferation in vivo. Our results show that Cav-1 increases in liver after PH with a redistribution of the protein from the caveola-enriched domain to the noncaveolar fraction. Moreover, the Cav-1 located in the noncaveolar fraction is phosphorylated in tyrosine 14, even though the Cav-1 gene is dispensable for liver regeneration after PH, as deduced from data obtained with commercially available animals lacking this gene. In addition to this, the proinflammatory stimulation of hepatocytes induces Cav-1 translocation to a noncaveolar fraction and tyrosine 14 phosphorylation mainly through the activation of tyrosine kinases such as Src. Conclusion: These results support a dynamic role for Cav-1 in liver proliferation both in vivo after PH and in vitro in cultured hepatic cell lines, but with minimal implications for the liver regeneration process. (HEPATOLOGY 2007.)

Caveolae were originally identified with an electron microscope as flask-shaped membrane invaginations on the surface of epithelial1 and endothelial cells.2 These plasmalemmal vesicles are enriched in sphingolipids, cholesterol, and caveolin (Cav), the major protein constituent of these structures.3 Three Cavs have been identified in differentiated cells with specific patterns of distribution. Cav-1 and Cav-2 are found in adipocytes and endothelial cells,4 whereas Cav-3 is selectively expressed in muscle cells.5

Caveolae participate in many cellular processes, including vesicular transport,6 cholesterol homeostasis,7 regulation of signal transduction,8 integrin signaling,9 and cell growth.10 Despite this, it is rather surprising that all Cav-null mice are viable, including the complete caveola-less mouse; however, the combined loss of Cav-1 and Cav-3 has profound effects on the cardiovascular function.11

Caveolae have now been demonstrated to concentrate a wide variety of signaling molecules, including Src family tyrosine kinases, H-Ras, heterotrimeric G protein α subunits, protein kinase C isoforms, and endothelial nitric oxide synthase (NOS; i.e., NOS-3).12 Many signaling molecules directly interact with Cav-1 through a defined modular protein domain, which is known as the Cav-scaffolding domain. The Cav-scaffolding domain has been shown to directly inhibit the activation of NOS-3,13 epidermal growth factor receptor,14 and c-neu15 among other molecules. Moreover, Cav-1 is a potent inhibitor of proliferative pathways such as the Ras-p42/p44 MAP (mitogen activated protein kinase) kinase cascades.16 Cav-1 expression negatively regulates cell cycle progression through a p53/p21-dependent pathway,17 and several cellular oncoproteins down-regulate Cav-1 expression.10

Cav-1 was first identified as a major substrate for Src, which phosphorylates the protein on tyrosine 14 (Y14). Cav-1 can also be phosphorylated in response to epidermal growth factor (EGF) or insulin stimulation; however, the functional significance of Cav-1 tyrosine phosphorylation remains unclear. It has been proposed that Cav-1 phosphorylation on Y14 can serve as a docking site for Src homology 2 protein/PBD (phosphotyrosine binding domain) tyrosine binding domain–containing proteins such as Grb7 to activate downstream signaling cascades.18 In this sense, Cav-1 has peptide domains with opposing functions. In liver and in different hepatocyte cell lines, the expression of Cav has been demonstrated,19, 20 and the internalization of specific ligands through caveolae has been reported.21–23 Recent reports have identified changes in the expression of Cav-1 in liver regeneration after partial hepatectomy (PH), a process in which Cav-1 seems to be essential,24, 25 in human cirrhotic liver,26 and in hepatocellular carcinogenesis.27

In this study, we have investigated the role of Cav-1 in liver regeneration after PH and in liver cell lines. Our results show that the protein levels of Cav-1 increased in liver after PH with a redistribution of the protein from the caveola-enriched domain to the noncaveolar fraction, in which Cav-1 was found phosphorylated in Y14. Experiments with CHL liver cells or primary cultures of hepatocytes treated with tumor necrosis factor α (TNF-α)/interleukin-1β (IL-1β) or hepatocyte growth factor (HGF) showed a translocation of Cav-1 to a noncaveolar fraction and Y14 phosphorylation of Cav-1 mainly through the activation of Src. However, regarding the function of Cav-1 in liver, our data allow us to conclude that animals lacking Cav-1 regenerate to the same extent and kinetics as wild-type mice after PH.

Abbreviations

bp, base pair; Cav, caveolin; EGF, epidermal growth factor; ER, endoplasmic reticulum; HGF, hepatocyte growth factor; IL-1β, interleukin-1β; NOS, nitric oxide synthase; PBS, phosphate-buffered saline; PCNA, proliferating cell nuclear antigen; PH, partial hepatectomy; PMSF, phenylmethylsulfonyl fluoride; SD, standard deviation; TNF-α, tumor necrosis factor α; Y14, tyrosine 14.

Materials and Methods

Chemicals.

The antibodies were from Santa Cruz Laboratories (Santa Cruz, CA), Chemicon (Temecula, CA), BD Transduction Laboratories (San José, CA), and Cayman (Ann Arbor, MI). Lipopolysaccharide and other reagents were from Roche (Mannheim, Germany) or Sigma (St. Louis, MO). PP2 and pharmacological inhibitors were from Calbiochem (San Diego, CA). STI-571 (Gleevec) was from Novartis (Basel, Switzerland). Tissue culture dishes were from Falcon (Lincoln Park, NJ). Tissue culture media were from BioWhittaker (Walkersville, MD).

Animals.

BALB/c (Charles River) mice, 2 months old (20-25 g), were supplied with food and water ad libitum and exposed to a 12-hour light-dark cycle. For PH, mice were anesthetized with a 92:7 mg/kg mix of ketamine and xylacine and subjected to midventral laparotomy with 70% liver resection (left lower and upper and right upper lobes). Sham surgeries were performed after anesthesia and entailed midventral laparotomy. Untreated animals received 0.5 ml of NaCl 0.9%. Plasma was obtained from the aorta. In some experiments, Cav-1 KO mice, strain Cav-1tm1Mls/J, and their corresponding controls Cav-1+/+, obtained from the Jackson Laboratory (Bar Harbor, ME), were used. The animals were treated according to the institutional care instructions.

Cell Culture and Treatments.

Hepatocytes from mice were prepared by perfusion with collagenase (45 mg/100 m) according to the classic perfusion/recirculation protocol method.28 The human liver cell line CCL-13 (Chang liver, CHL), an immortalized nontumor cell line derived from normal liver that expresses hepatocyte markers, and the hepatoma cell line HepG2 were purchased from the American Type Culture Collection (Manassas, VA). HuH-7 cells were kindly provided by Dr. P. Schirmarcher (Institute of Pathology, University of Cologne, Germany) and are considered a hepatocellular carcinoma cell line. Cells were grown in DMEM supplemented with 10% fetal bovine serum and antibiotics. For transient transfection assays with vectors encoding wild-type Cav-1-GFP or Y14F-Cav-1-GFP (cloned into pEGF-N1), CHL cells at 60% confluence were exposed for 6 hours to Lipofectamine containing the Cav-1 expression vectors.

Flow Cytometry Analysis.

Liver sections (3 mm) were processed through a Dako Medimachine equipped with a 50-μm Medicon filter to digest the whole tissue and to obtain individual cells. The cells were fixed with 70% ethanol and stained with Nile red and Hoechst 33342 to evaluate the ploidy, the cell cycle, and the lipid content in a BD FACS Canto II cell cytometer.

Western Blot Analysis.

Equal amounts of protein (10-50 μg) from each fraction or 50 μl of the fractions obtained from the gradient was loaded into a 10%-12% SDS-PAGE. The amounts of cyclin E, D1, proliferating cell nuclear antigen (PCNA), p27kpi, total and phospho-Cav-1 Abs, and total and active c-Src Abs were determined with the commercial Abs given in parentheses. Proteins were size-fractionated, transferred to a Hybond P membrane (Amersham), and, after blocking with 5% nonfat dry milk, incubated with the corresponding Abs. Blots were normalized by the measurement of the amount of PI3K subunit p85 for cytosolic proteins, SP1 for nuclear proteins, 5′-nucleotidase as a marker for the particulate fraction, GM130 for the Golgi apparatus, calregulin for endoplasmic reticulum (ER), and annexin VI for recycling receptor endosomes. The blot was revealed, and different exposition times were performed for each blot with a charged coupling device camera in a luminescent image analyzer (LAS 3000, TDI, Madrid, Spain) to ensure the linearity of the band intensities.

Immunofluorescence Analysis.

Samples of liver were collected in a solution containing 30% sucrose in phosphate-buffered saline (PBS) and maintained for 24 hours at 4°C. After that, the tissue was frozen in methyl butane in contact with dry ice, and serial 10-μm-thick sections were cut with a cryostat (Leica Microsystems) in gelatinized glass. Liver sections or CHL cells (cultured on glass coverslips) were fixed for 15 minutes with 4% paraformaldehyde (pH 7.0), washed with PBS, and permeabilized with methanol for 15 minutes at room temperature. After blocking with 3% bovine serum albumin for 1 hour at room temperature, the sections were incubated overnight with the corresponding Abs diluted to 1:150 in 1% bovine serum albumin, washed several times, and incubated for 2 hours with fluorochrome conjugated Abs (Alexa 488 or Alexa 657, Invitrogen) raised against Fc of primary Abs. For the detection and quantification of apoptosis, a TUNEL commercial kit (Roche) was used according to the instructions of the manufacturer. Lipid bodies were stained with Nile red. The images were acquired with a Radiance 2100 confocal microscope (Zeiss). The quantification of colocalization analysis and image processing were performed with Laserpix software (Bio-Rad).

RNA Isolation and Reverse Transcriptase Polymerase Chain Reaction Analysis.

One microgram of total RNA, extracted with Trizol Reagent (Invitrogen) according to the manufacturer's instructions, was reverse-transcribed with 50 U of expand reverse transcriptase and pd(N)6 random hexamer as a primer according to the indications of the manufacturer (Amersham Biosciences). The resultant cDNAs were amplified with the following oligonucleotide sequences: Cav-1 [5′AGCCCAACAACAAGGCCAT3′ sense, 5′GCAATCACATCTTCAAAGTCAATCTT3′ antisense, 142 base pairs (bp)], NOS-2 (5′CTTTGACGCTCGGAACTGTAGC3′ sense, 5′GGCAGTGCATACCACTTCAACC3′ antisense, 550 bp), and 18S rRNA (5′GCAATTATTCCCCATGAACGA3′ sense, 5′CAAAGGGCAGGGACTTAATCAA3′ antisense, 100 bp).

Determination of Metabolites.

The protein levels were determined with Bradford reagent. Triacylglycerols were determined by enzymatic methods with specific kits from Biosystems (Barcelona) in liver homogenates.

Data Analysis.

The data are expressed as means ± the standard deviation (SD). The statistical significance was estimated with a Student t test for unpaired observation. The results were considered significant at P < 0.05. The data were analyzed by the SPSS for Windows statistical package, version 9.0.1.

Results

Increase in Cav-1 in Regenerating Liver After PH.

The total levels of Cav-1 and cyclin E, as markers of the regeneration process, were analyzed in liver samples from sham and PH animals. As Fig. 1A shows, Cav-1 increased 2.8-fold in total liver protein extracts and 3-fold in the corresponding microsomal fractions (Fig.1A, inset) at 12-24 hours after PH, returning to the control values at day 5 after PH. Cav-1 was not detected in the soluble fraction (not shown). Cyclin E reached a peak at 48 hours after PH in agreement with previous results,28 and these changes were absent in sham animals. As shown in Fig. 1B, RT-PCR analysis confirmed the increase in Cav-1 mRNA in the remnant liver. In addition to this, the NOS-2 mRNA levels, used as an additional control of the regeneration process, increased transiently after PH, exhibiting a peak at 24 hours.

Figure 1.

Cav-1 increases in regenerating liver after PH. Total tissue extracts, membrane fractions, and RNA were prepared from remnant or sham liver at different times. For the total extracts, cultured cells (2 × 106 to 3 × 106) or the remnant liver (50 mg) was homogenized in 1 ml (cells) or 5 volumes of an ice-cold buffer containing 10 mM Tris-HCl (pH 7.5), 1 mM ethylene glycol tetraacetic acid, 1 mM MgCl2, 5 mM β-mercaptoethanol, 10% glycerol, 0.1 mM phenylmethylsulfonyl fluoride (PMSF), 10 μg/ml aprotinin, 10 μg/ml leupeptin, and 0.5% 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate. The extracts were vortexed for 30 minutes at 4°C, and after centrifugation for 20 minutes at 13,000g, the supernatants were stored at −20°C. For microsomal fractions, cells or liver tissue was homogenized in 1 ml (cells) or 5 volumes of an extraction buffer (PBS, 1 mM EDTA, 2 mM PMSF, 10 μg/ml aprotinin, and 10 μg/ml leupeptin), and this was followed by 3 cycles of 15 seconds of sonication at 4°C and centrifugation at 6000g for 15 minutes. The resulting supernatants were centrifuged at 105,000g for 1 hour at 4°C, and the pellets [crude membrane fraction containing mainly the endoplasmic reticulum and plasma membrane (PM+ER)] and the supernatants (soluble fraction) were brought to equal volumes of the initial extraction buffer. (A) The protein levels of Cav-1 and cyclin E were determined by an immunoblot using total liver extracts. The inset shows the Cav-1 levels in the membrane fraction. Representative blots are shown. (B) The mRNA levels of Cav-1 and NOS-2 were determined by semiquantitative RT-PCR. The data are presented as the means ± SD of 5 independent experiments. *P < 0.01 versus the corresponding value at time 0. The blots were normalized with p85 for total extracts and with 5′-nucleotidase for the membrane fraction (A) and with 18S RNA for the PCR studies (B).

Cav-1 Translocates Outside the Caveolar Fraction During Liver Regeneration.

As shown in Fig. 2A, there was a significant redistribution of Cav-1 from the caveola-enriched plasma membrane fraction (CEF, fractions 5 and 6) in the control liver to fractions 9-12 corresponding to high-density membrane fractions and to the interface (fractions 7 and 8) at 12-48 hours after PH. At later stages of liver regeneration (120 hours), Cav-1 was relocated again in the CEF fraction, and this resembled the control situation. The quantification of the data by the measurement of the percentage of Cav in each fraction versus the total Cav-1 content is shown in Fig. 2B. The protein concentration in the sucrose gradient fractions is shown in Fig. 2C. To identify the subcellular localization in which Cav-1 accumulates after PH, we used different markers: GM130 and calregulin for Golgi and ER, respectively; annexin VI for recycling receptor endosomes, and 5′-nucleotidase for the CEF fraction. As shown in Fig. 2D, fractions 9-12, in which Cav-1 is located after 12 hours of PH, correspond mainly to the ER and Golgi apparatus, indicating an internalization of Cav-1. This dynamic behavior of Cav-1 is completed during later stages when Cav-1 relocates to the CEF fraction.

Figure 2.

Cav translocates outside the caveolar fraction after PH. Sucrose density gradients were performed to isolate caveola-enriched fractions from the remnant liver at different times after PH. To prepare Cav-enriched membrane domains, cultured cells (107) or liver tissue (50 mg) was homogenized with a glass-Teflon homogenizer in 2 ml (cells) or 5 volumes of an MBS buffer at 4°C: 25 mM 2-(N-morpholino)ethanesulfonic acid (pH 6.5), 150 mM NaCl, 1 mM phenylmethylsulfonyl fluoride, 10 μg/ml aprotinin, 10 μg/ml leupeptin, and 1% Triton X-100. Nuclei and unbroken cells were removed by centrifugation at 3000g for 5 minutes. The resulting supernatant was mixed with the same volume of 80% sucrose prepared in the MBS buffer and loaded at the bottom of a discontinuous sucrose gradient (40%-30%-5%). The tubes were centrifuged at 200,000g for 18 hours at 4°C in a TH641 rotor, and 12 fractions were unloaded from the bottom to the top of the gradient. Proteins were precipitated with cold acetone for 2 hours, centrifuged at 16,000g for 10 minutes, and dried. (A) The distribution of Cav-1 in the gradient was analyzed by a western blot. A representative sucrose gradient is shown. (B) Quantification of the distribution of Cav-1 protein levels. The results are expressed as the band intensity of Cav-1 in each lane with respect to the total Cav-1 levels. CEF is the Cav-enriched fraction; INF corresponds to the interface region; and NCEF is the noncaveolar fraction, which is mainly plasma membrane and ER. (C) Total protein concentration in the gradient fractions at 0 and 12 hours after PH. (D) Distribution of different subcellular markers in the sucrose gradient: GM130 (Golgi), calregulin (ER), annexin VI (recycling receptor endosomes), and 5′-nucleotidase (CEF fraction). A representative blot is shown. The results are the means ± SD of 4 independent experiments. *P < 0.01 versus the corresponding value at time 0.

Cav-1 Is Phosphorylated in Tyrosine 14 After PH.

Cav-1 staining in control hepatocytes is located mainly in the plasma membrane (Fig. 3A); however, at 12-24 hours after PH, there is an increase in Cav-1 labeling concomitant with significant staining in intracellular domains. This distribution of the labeling of Cav-1 returned to the initial situation at day 5 after PH. Figure 3B shows the quantification of the fluorescence in the membrane and cytoplasm. Because Cav-1 is phosphorylated in response to oxidative stress and growth factors,29, 30 we analyzed whether the Cav-1 redistribution after PH is a result of Cav-1 phosphorylation, using an anti–phosphoY14-specific Ab recognizing phosphorylated Cav-1 on Y14. As Fig. 3A-C shows, a significant amount of Cav-1 is phosphorylated at 12-24 hours after PH, and this protein is located in the noncaveolar fraction from the sucrose gradient, as visualized and quantified by immunocytochemistry (Fig. 3A, right panel) and determined by sucrose gradient separation followed by a western blot (Fig. 3C). In agreement with the data of Figs. 2A and 3A, 5 days after PH, both P-Cav-1 and Cav-1 returned to the initial location in the sucrose gradient. The subcellular localization of Cav-1 and Y14-P-Cav-1 is shown with higher magnification in Fig. 3D: the protein was dephosphorylated on Y14 at 0 hours; however, at 24 hours, there was a significant increase in Y14-P-Cav-1 staining, mainly localized in cytoplasmic domains.

Figure 3.

Cav-1 is internalized and phosphorylated in regenerating liver. (A) Immunohistochemical analysis by confocal microscopy of Cav-1 and Y14-P-Cav-1 in liver sections at 0, 12, 24, and 120 hours after PH. Cav-1 staining is green, Y14-P-Cav-1 is red, and nuclear staining with Hoechst 33258 is blue. The average red fluorescence is given in the panels. (B) Quantification of the subcellular distribution of Cav-1 from panel A. The results are expressed as the percentage of Cav fluorescence in the cell membrane fraction versus the cytosol. (C) Sucrose density gradients were performed as described in Fig. 2. The levels of Cav-1 and Y14-P-Cav-1 were determined by a western blot. (D) Higher magnification of the images in panel A at 0 and 24 hours. The results are the means ± SD of 4 independent experiments. *P < 0.01 versus the corresponding value at time 0.

Cav-1 Is Dispensable for the Liver Regeneration Process.

Because the preceding data suggested a functional role for Cav-1 in the regeneration process, we investigated the effect of its absence, using Cav-1−/− mice. We performed PH in 2 groups of at least 25 Cav-1+/+ and 25 Cav-1−/− mice and followed the regeneration up to 7 days after PH. The mortality of both groups was very similar and occurred mainly during the initial 24 hours after PH (20% for both groups; Fig. 4A). When the liver regeneration index was determined, the Cav-1−/− mice exhibited an accelerated mass gain at 80 hours after PH versus the corresponding Cav-1+/+ counterparts (Fig. 4B,C). Figure 4D shows the mRNA and protein levels of Cav-1 in Cav-1+/+ and Cav-1−/− mice, respectively.

Figure 4.

Liver regeneration after PH in Cav-1−/− mice. (A) Survival rates of Cav-1+/+ and Cav-1−/− mice at 120 hours after PH. (B) Liver regeneration index at the indicated times after PH. The number of animals for each point is given in parentheses. (C) Representative picture comparing the liver recovery at 80 hours after PH. (D) Determination of the mRNA and protein levels of Cav-1 in Cav-1+/+ and Cav-1−/− mice. 18S RNA and the p85 subunit of PI3K were used for normalization. The results are the means ± SD of the indicated number of animals. *P < 0.01 versus the corresponding value of Cav-1+/+ mice.

To investigate the time course of the regeneration process, parameters related to the cell cycle, such as cyclins D1, E, PCNA, and p27, were determined. As Fig. 5A shows, cell cycle protein levels did not exhibit significant differences versus the corresponding Cav-1+/+ mice. Cyclin E and PCNA reached a maximum at 48 hours after PH, with an early and moderate increase at 12-24 hours in Cav-1−/− mice. When the ploidy of the cells of the remnant liver was analyzed, a significant increase in the 4N population was observed in Cav-1−/− mice, which might be related to the accelerated mass gain after PH (Fig. 5B). Figure 5C shows the absence of differences in size (forward scatter) and granularity/complexity (side scatter) between the liver cells isolated from Cav-1+/+ and Cav-1−/− mice at 80 hours after PH.

Figure 5.

Molecular and cellular markers of liver regeneration in Cav-1+/+ and Cav-1−/− mice after PH. (A) The protein levels were determined in total liver extracts from regenerating liver at the indicated times after PH by a western blot and normalized with SP1. (B) Distribution of cell ploidy of regenerating liver after cell isolation in a Medimachine device and staining with Hoechst 33342 followed by flow cytometry analysis. (C) Forward scatter (FSC), indicating the size, and side scatter (SSC), indicating the granularity and complexity of cells isolated from liver at 80 hours after PH. A representative scatter distribution is shown. The results are the means ± SD of the 5 animals. *P < 0.01 versus the corresponding value of Cav-1+/+ mice.

Previous reports described the absence in Cav-1−/− mice of the characteristic steatosis that occurs in regenerating liver after PH.25 Because this fact has been associated with an impaired regenerative response after PH in these mice, we investigated the time course of the lipid droplet accumulation in sections of the remnant liver after Nile red staining. As Fig 6A -C shows, both Cav-1+/+ and Cav-1−/− mice exhibited an important accumulation of lipids at 12-48 hours after PH. When the time course of the accumulation of these apolar lipids was quantified in isolated cells, a sharp peak at 24 hours was observed in both mice, and these data are in agreement with the staining in fixed sections of liver (Fig. 6A); however, at 48 hours after PH, Cav-1+/+ mice exhibited a 3-fold accumulation of lipids with respect to the Cav-1−/− mice (Fig. 6B). The same conclusion was obtained when the lipid content was referred to nuclear staining with Hoechst (Fig. 6C). These results are in agreement with the triacylglycerol levels determined in liver homogenates (Fig. 6D). Moreover, when the localization of Cav-1 and Nile red staining were combined in the tissue sections, Cav-1 failed to colocalize with the lipid droplets at 24 hours after PH in Cav-1+/+ mice (Fig. 6E), showing colocalization coefficients below 0.006.

Figure 6.

Time course of steatosis in Cav-1+/+ and Cav-1−/− mice regenerating liver after PH. (A) Cav-1+/+ and Cav-1−/− mice were hepatectomized, and Nile red staining was performed in at least 7 fixed sections of the remnant liver isolated at the indicated times. (B) The fluorescence of Nile red was quantified after cell isolation in a Medimachine and was plotted as a function of the time of regeneration. (C) The data of panel B were plotted after the normalization of the nuclear staining with Hoechst. (D) The levels of triacylglycerol were determined in the liver extracts with a commercial kit. (E) Costaining with Nile red (red) and Cav-1 (green) was analyzed by confocal microscopy, and the pixel distribution was digitally evaluated to determine the extent of pixels within both fluorescences (colocalization). (F) Liver sections were stained with TUNEL to determine the percentage of apoptotic cells in the sections. The results are the means ± SD of 4-7 samples for each time. *P < 0.01 versus the corresponding value of Cav-1+/+ mice.

In addition to this, the presence of apoptotic cells in the remnant liver was evaluated by TUNEL, and less than 2% of positive nuclei were observed at 24-80 hours after PH, regardless of the expression of Cav-1 (Fig. 6F). Together, these data clearly indicate that Cav-1 is dispensable for the regeneration process after PH with the genetic background analyzed in these experiments.

Cytokines and Growth Factors Induce the Phosphorylation of Cav-1 in Liver Cells.

To evaluate the physiological relevance of Y14 phosphorylation of Cav-1, we used primary cultures of hepatocytes and hepatoma cell lines. Figure 7A shows the constitutive levels of Y14-P-Cav-1 in cultured hepatic cells. Primary hepatocytes (PCH) and CHL cells contained a significant amount of Cav-1 and a slight band of Y14-P-Cav-1, whereas the protein was absent in hepatocellular carcinoma cells. Next, we analyzed the contribution of proinflammatory stimuli and growth factors to the phosphorylation of Cav-1 in hepatic cells. As shown in Fig. 7B,C, treatment for 1 hour with TNF-α/IL-1β and HGF stimulated Y14-Cav-1 phosphorylation without changes in the levels of Cav-1. Furthermore, to obtain a readout of different signaling pathways activated in the course of regeneration, cells were stimulated with TNF-α/IL-1β, and a panel of selective kinase inhibitors was used. As shown in Fig. 7D, the pretreatment of CHL liver cells with serine/threonine kinase inhibitors did not influence Y14-P-Cav-1 levels induced by TNF-α/IL-1β; however, pretreatment with PP2 or STI-571 significantly inhibited Y14-Cav-1 phosphorylation. Regarding the subcellular localization of Cav-1 after the TNF-α/IL-1β treatment, redistribution to the noncaveola fractions of the gradient was observed (Fig. 8A,B). However, when cells were treated with PP2, Cav-1 was located again in CEF fractions. We performed transient transfection experiments in CHL liver cells with an expression vector for Cav-1-GFP and mutated Y14F-Cav-1-GFP, and as shown in Fig 8C, in control cells, both endogenous Cav-1 (red) and exogenous Cav-1 (green) were located in the plasma membrane. After the TNF-α/IL-1β treatment, the staining corresponding to endogenous and transfected Cav-1-GFP was associated with intracellular compartments. However, when cells were treated with PP2, both exogenous Cav-1 and endogenous Cav-1 were retained in the plasma membrane. Interestingly, when cells were transfected with Y14F-Cav-1-GFP, the redistribution induced by TNF-α/IL-1β was lost (Fig. 8D). These results indicate that Cav-1 phosphorylation and redistribution are involved in the signaling network that occurs after the proinflammatory stimulation of the cells.

Figure 7.

Cytokines and growth factors induce the phosphorylation of Cav-1 in liver cells. (A,B) Total cell extracts from hepatocytes in primary culture (PCH), CHL, HuH7, and HepG2 cell lines were prepared, and the levels of Cav-1 and Y14-P-Cav-1 were determined by a western blot. Blots were normalized with αP85. (B) PCH was maintained with 2% fetal bovine serum 18 hours before the experiment and stimulated with 20 ng/ml TNF-α plus 20 ng/ml IL-1β for 1 hour. (C) CHL cells were treated with 0.5 μg/ml lipopolysaccharide, 20 ng/ml TNF-α plus 20 ng/ml IL-1β, 20 ng/ml HGF, and 20 ng/ml EGF for 1 hour. Total cellular extracts were prepared, and the levels of Cav-1 and P-Cav-1 were measured by a western blot. (D) CHL liver cells were treated for 1 hour with the following inhibitors (targets in parentheses): 20 μM SB202190 (p38 mitogen–activated protein kinase), 20 μM LY204002 (phosphatidylinositol-3-kinase), 10 μM PD98059 (mitogen-activated protein kinase/extracellular signal-regulated kinase), 0.4 μM CMI (protein kinase A), 1 μM bisindolylmaleimide and 0.1 μM Gö 6976 (protein kinase C), 50 μM PP2 (Src kinase family), and 20 μM STI-571 (c-Abl). After that, they were stimulated with a combination of 20 ng/ml TNF-α and 20 ng/ml IL-1β for 1 hour. Total cellular extracts were prepared, and the amounts of Cav-1 and Y14-P-Cav-1 were analyzed by a western blot. The data are the means ± SD of 5 independent experiments. *P < 0.01 versus the corresponding control value without the treatment.

Figure 8.

Cav translocates outside the caveolar fraction in liver cells after cytokine stimulation. (A) CHL cells were treated with 50 μM PP2, and this was followed by stimulation with a combination of 20 ng/ml TNF-α and 20 ng/ml IL-1β for 1 hour. Sucrose density gradients were performed to isolate caveola-enriched fractions, and the distribution of Cav-1 was analyzed by a western blot. (B) Quantification of the densitometric analysis of the bands is shown. (C,D) transient transfection assays with constructs encoding wild-type Cav-1-GFP or Y14F-Cav-1-GFP. The expression and distribution of Cav-1-GFP (C) or Y14F-Cav-1-GFP (D) were analyzed by immunofluorescence. Endogenous Cav-1 is in red; Cav-1-GFP staining is in green, and nuclear staining with Hoechst 33258 is in blue. Merged images are shown. The results show a representative density gradient, and the data are expressed as the level of Cav-1 in each line with respect to the total intensity and are the means ± SD of 4 independent experiments. *P < 0.01 versus the corresponding control value without the treatment.

c-Src Kinase Is Involved in the Proinflammatory Cytokine-Induced Y14 Phosphorylation of Cav-1 in Liver Cells.

To analyze which tyrosine kinase was implicated in the phosphorylation of Cav-1 in liver samples after PH and in liver cells by the effect of TNF-α/IL-1β, the protein levels of active and total c-Src were analyzed by a western blot. As shown in Fig 9A, the levels of active Src increased in liver at 12 hours after PH. Figure 9B shows that the treatment of CHL cells with proinflammatory cytokines induced a significant increase in the active Src kinase, an effect that was significantly inhibited after the treatment with PP2.

Figure 9.

c-Src kinase is involved in cytokine-induced Y14 phosphorylation of Cav-1 in liver cells. (A) Total liver extracts from PH animals and (B) CHL extracts were prepared from control and stimulated cells, and the total and active c-Src levels were determined in the same extract in the presence or absence of 50 μM PP2. The results show a representative western blot, and data are the means ± SD of 4 independent experiments. *P < 0.01 versus the corresponding control value at time 0 or without the treatment.

Discussion

In this work, we have investigated the dynamics and potential function of Cav-1 during liver regeneration after PH. Our data show that Cav-1 increases and translocates to noncaveolar fractions and is phosphorylated in Y14 in the course of liver regeneration. However, it is clear that in this mouse strain, and in contrast to previous work using a different genetic background,25 Cav-1 is dispensable for liver regeneration because Cav-1 KO animals survived and fully regenerated liver function and size after PH.

To reconcile this discrepancy, the existence of important genetic differences between the Cav-1 KO animals used by Fernandez et al.25, 31, 32 and those used in this work should be mentioned. First, the embryonic stem cells used by Fernandez et al.25 were fully derived from a 129Sv background, and the animals exhibited problems of fertility, whereas those available at Jackson were WW6, corresponding to a 75% 129Sv, 20% C57BL/6J, and 5% SJL mixture, in addition to the removal of exon 3 in the first case and exons 1 and 2 in our case. Second, death occurred between 48 and 72 hours, a time during which DNA synthesis started (the S-phase in the mice occurred at 36-40 hours after PH) and most of the increase in the liver mass occurred (3 days). However, initial events progressed normally: the 2 main signaling pathways that initiated recovery were activated, and the levels of fatty acids increased in the serum and inside cells, but regeneration collapsed abruptly at 72 hours, in the absence of apoptotic events.25 In contrast to this behavior, the animals used in this work exhibited a modest mortality, but always before the initial 24 hours after PH, similarly to Wt and Cav-1−/− mice, and a complete regeneration process was evident after 5 days. Third, we did not find a decrease in lipid droplet accumulation in the KO mice as reported previously,25 and we failed to observe a colocalization of Cav-1 with Nile red in the Wt mice after PH (>60 liver sections); these data fully agree with previous observations in liver cell lines.33 Interestingly, we observed prolonged lipid droplet accumulation in Wt animals, in comparison with the KO counterparts, that might explain a significantly different pattern of lipid metabolism between the two types of animals, although the pathophysiological relevance of these differences remains to be established. Finally, we should mention the ability of glucose supplementation to rescue the liver regeneration index and cell cycle progression in the Cav-1−/−-PH–sensitive mice. The comparison of the behaviors of the 2 models raises the question of the existence of a metabolic link connecting cell cycle progression, and not apoptosis, and lipid/carbohydrate metabolism as a key control point for restoring liver mass and function, the nature of which remains unidentified and deserves further work.

Several groups have described the inhibitory action of Cav-1 on proliferative/antiapoptotic pathways, and these data suggest that Cav-1 might have a role in liver regeneration. Pol et al.24 reported an important increase and redistribution of Cav-1 from the cell surface to the newly formed lipid bodies in liver regeneration after PH. These authors suggested that this association of Cav-1 with lipid bodies is due to the lipid mobilization process together with the arrest of endocytosis and membrane trafficking that accounts for regenerating liver. Our results demonstrate that Cav-1 is up-regulated and translocated to the ER and plasma membrane compartments in liver regeneration after PH. These results were confirmed by the immunohistochemistry of liver samples exhibiting an important intracellular staining of Cav-1 12-24 hours after PH. Although an important accumulation of lipid bodies was detected in mouse liver 24-48 hours after PH, we did not detect Cav-1 protein in the very low density fraction of the gradient or colocalization of Cav-1 with Nile red.

Recent data show that Y14 from Cav-1 can be rapidly phosphorylated in response to oxidative stress29, 34 and growth factors.30, 35 Although the functional consequences of this phosphorylation are unknown, it has been suggested that P-Y14-Cav-1 may function as a growth factor receptor that recruits Src homology 2 protein domain–containing proteins to the plasma membrane, thus activating cellular growth.10 It is known that, in addition to the plasma membrane, Cavs are present in the trans-Golgi network, in caveosomes, as secreted proteins when it is phosphorylated in S80 in response to different stresses.36 However, the precise localization of P-Cav-1 is not clear, and it has been detected in focal adhesions, colocalizing with paxillin, in caveola-derived fusion vesicles,37 or internalized in cytoplasmic vesicles after oxidative stress.38 Our results show that liver regeneration after PH causes some Cav-1 phosphorylation, and the phosphorylated protein is located in cytosolic domains.

In addition to liver regeneration after PH, recent reports26 have established a relationship between Cav-1 and hepatocellular carcinogenesis. A microarray analysis of macroregenerative and dysplastic nodules, a hepatocellular carcinoma precursor lesion, identified aberrant expression of Cav-1 and thrombospondin-1. Our results with hepatoma cell lines clearly demonstrate that nontumor cells and primary cultures of hepatocytes express a significant amount of Cav-1, whereas Cav-1 was undetectable in the HepG2 and HuH7 hepatoma cell lines. We demonstrated that nontumor liver cells stimulated with proinflammatory stimuli induce Cav-1 phosphorylation and translocation of the phosphorylated protein to the noncaveolar fraction. Moreover, the tyrosine kinase c-Src is implicated in Cav-1 phosphorylation on Y14, thus confirming the results obtained in the in vivo model of liver regeneration after PH. Besides the EGF and insulin receptors, the regulation by Cav-1 of different receptors signaling pathways, such as tumor necrosis factor receptor 1 and transforming growth factor β, some of them implicated in the process of liver regeneration, has been described recently.39, 40

In summary, we have studied the role of Cav-1 in liver regeneration after PH and in liver cell lines under proinflammatory stimulation. Our results show an increase in Cav-1 in the early phase of liver regeneration with a translocation of the protein to the noncaveolar fraction; however, Cav-1 is dispensable for liver regeneration, and its role in the process deserves further work.

Acknowledgements

We thank Dr. M. A. del Pozo for the vectors encoding WT and mutated Cav-1-GFP and for discussion, advice, and a critical reading of the manuscript. The technical support of Dr. A. Álvarez from the Department of Cytometry and Confocal Microscopy is greatly acknowledged.

Ancillary