Caveolin- and -ß Perform Nonredundant Roles in Early Vertebrate Development

【摘要】  Caveolins are integral membrane proteins that localize predominantly to lipid rafts, where they oligomerize to form flask-shaped organelles termed caveolae and play important roles in membrane trafficking, signal transduction, and other cellular processes. To investigate potential roles for caveolin-1 (cav-1) in development, cav-1 and -1ß cDNAs were functionally characterized in the zebrafish. Cav-1 and -1ß mRNAs exhibited overlapping but distinct expression patterns throughout embryogenesis. Targeted depletion of either Cav-1 isoform, using antisense morpholino oligomers, resulted in a substantial loss of caveolae and dramatic neural, eye, and somite defects by 12 hours after fertilization, the time at which mRNA levels of both isoforms substantially increased in wild-type animals. Morphant phenotypes were rescued by injection of homotypic (cav-1/cav-1) but not heterotypic (cav-1/cav-1ß) zebrafish and human cav-1 cRNAs, revealing nonredundant and evolutionarily conserved functions for the individual Cav-1 isoforms. Mutation of a known Cav-1 phosphorylation site unique to Cav-1 (Y14F) resulted in a failure to rescue the cav-1 morphant phenotype, verifying an essential role for Cav-1 specifically and implicating this residue in early developmental functions. Cav-1 and -1ß morphants also exhibited disruption in the actin cytoskeleton. These results support important and previously unanticipated roles for the Caveolin-1 isoforms in vertebrate organogenesis.
--------------------------------------------------------------------------------
Lipid rafts are liquid-ordered membrane microdomains, enriched in cholesterol and glycosphingolipids, that have been implicated in endocytosis, transcytosis, cholesterol transport, and signal transduction.1-8 Caveolae ("little caves") are a form of lipid raft identifiable as 50- to 100-nm plasma membrane invaginations visible in electron micrographs.9 Caveolae are predominantly found in the plasma membrane but also exist as detached vesicles and/or racemose structures that penetrate deep into the cell, forming transcytotic channels.10,11 Caveolae are abundant in endothelia, adipocytes, and muscle cells and have been implicated in diverse cellular processes, including polarized sorting of membrane proteins and regulation of signal transduction.12-14
The invaginated morphology characteristic of caveolae is conferred by homo- or hetero-oligomerization of members of the caveolin family of structural proteins.15,16 Four caveolins have been identified in vertebrates: Cav-1 and -1ß, which arise from differential splicing, and Cav-2 and -3, encoded by separate genes. Cav-1, -1ß, and -2 are present in nonmuscle cells, whereas myocytes express all four caveolins. Caveolins interact with and functionally regulate a number of signaling proteins via direct protein/protein interactions and, consequently, are believed to mediate many processes attributed to caveolar rafts.8 Despite extensive study, the in vivo roles of the caveolin proteins are not well understood. Surprisingly, cav-1-null mice, which lack most caveolae, and cav-1/3-null mice, which have been reported to lack all caveolae, are viable and fertile.17,18 However, cav-null mice do exhibit a variety of moderate pathologies, including hyperplasia of the genitourinary tract, cardiomyopathy and pulmonary hypertension, vascular abnormalities, and reduced lifespan.7,19,20 Very recently, cav-1-null mouse mammary epithelium was found to exhibit defects in three-dimensional acinar architecture21 and, when deprived of exogenous extracellular matrix, to undergo spontaneous epithelial-mesenchymal transition and reorganization of the actin cytoskeleton.21 Targeted depletion of the muscle-specific Cav-3 was found to result in severe defects in muscle differentiation and function.22 The fact that caveolin genes exhibit significant evolutionary conservation suggests important biological functions for these proteins.23-27
To investigate possible roles for the two Cav-1 isoforms in vertebrate development, zebrafish cav-1 and -1ß cDNAs were isolated, the expression profiles of cav-1 mRNAs were defined, and Cav-1 proteins were functionally characterized using a targeted protein depletion approach. Our studies revealed unique developmental functions for zebrafish Cav-1 and -1ß isoforms, including roles in actin cytoskeletal organization, which were rescued by expression of human Cav-1 and -1ß isoforms. We also demonstrate, for the first time in an in vivo context, that the Cav-1 Y14 phosphorylation site, which is a substrate of Src in other vertebrates, is required for Cav-1 function in early zebrafish development.

【关键词】  caveolin- nonredundant vertebrate development

Materials and Methods

Zebrafish Husbandry

Wild-type and heterozygous fli-enhanced green fluorescent protein (EGFP) transgenic28 zebrafish (Danio rerio) were maintained in The Forsyth Institute Zebrafish Facility. Embryos were generated by natural pairwise crosses, raised at 28.5??C, and staged as described previously.29,30

Cloning of Zebrafish and Human cav-1 and -1ß cDNAs

Full-length zebrafish cav-1 and -1ß cDNAs were isolated using degenerate oligonucleotide primers, polymerase chain reaction, and rapid amplification of cDNA ends (DOP-PCR/RACE). In brief, degenerate cav-1 primers were chosen from conserved amino acid sequences of human, bovine, mouse, rat, and chick caveolin-1 genes. The cav-1 forward primer was 5'-ATHGAYYTIGTIAAYMGIGAYCC-3', and the cav-1 reverse primer was 5'-AIGGIAYIAIIRYCCADATRTG-3'. Total RNA was isolated from adult zebrafish using the TRIzol Reagent method (Invitrogen, Carlsbad, CA). Reverse transcription (RT) reactions were performed using the SuperScript First-Strand Synthesis System (Invitrogen) followed by PCR amplification. 5'- and 3'-RACE was performed using the GeneRacer kit (Invitrogen), and the resulting PCR products were cloned into the pCR4-TOPO TA cloning vector. Nucleotide sequence comparison with the previously identified zebrafish cav-1 genomic clone confirmed the identity of full-length zebrafish cav-1 and -1ß cDNAs, whose nucleotide sequences were submitted to GenBank (accession nos. AY124572 and AY124573).

Human cav-1 cDNAs were generated as follows. Wild-type human cav-1 was PCR amplified from a construct kindly provided by Dr. Michael Lu,31 adding BamHI and EcoRI sites to 5' and 3' ends, respectively (sense primer, 5'-CAGGATCCATGTCTGGGGGCAAATACGTG-3'; and antisense primer, 5'-CAGAATTCTTATATTTCTTTCTGCAAGTTGATGCG-3'), and cloned into pcDNA3. The human cav-1a-Y14F mutant cDNA was generated by introducing an A to T base change at position 41, using the Stratagene QuickChange Site-Directed Mutagenesis kit (Stratagene, La Jolla, CA) and the following primers: sense primer, 5'-CGGAGGGACATCTCTTCACCGTTCCCATC-3'; and antisense primer, 5'-GATGGGAACGGTGAAGAGATGTCCCTCCG-3'. Human cav-1ß cDNA was cloned using the sense primer 5'-CAGGATCCATGGCAGACGAGCTGAGCGAGAAG-3' and the above antisense primer used to clone human cav-1, using a cDNA library from human prostate cancer cell (PC3) as template. The PCR product was cloned into pcDNA3 at BamHI/EcoRI sites. The nucleotide sequence of generated human cav-1 isoform expression constructs was confirmed by DNA sequence analyses.

Expression of Zebrafish Cav-1 and -1ß in Mammalian Cell Lines

Full-length zebrafish cav-1 and -1ß coding regions were cloned into pcDNA3.1/Myc-HisB (Invitrogen), generating the eukaryotic expression vectors pcDNA3.1-zfcav-1 and pcDNA3.1-zfcav-1ß, respectively. LNCaP and human embryonic kidney (HEK) 293 cells (American Type Culture Collection, Rockville, MD) were cultured and transfected using standard procedures. Protein extracts were generated from transfected cell lines and subjected to immunoblotting with the anti-human caveolin-1 polyclonal antibody (1:1000 dilution; Transduction Labs, San Diego, CA).

RT-PCR

Developmentally staged zebrafish were collected, frozen in liquid nitrogen, and stored at C80??C. Total RNA was isolated using TRIzol Reagent (Invitrogen), and 5 µg of each sample was used in RT reactions using the SuperScript III First-Strand Synthesis System for RT-PCR (Invitrogen). One seventh of each RT reaction was used as template in 50-µl PCRs, which were performed per the manufacturer??s specifications using the Platinum TaqDNA Polymerase High Fidelity (Invitrogen) and the following reaction conditions: denaturing at 95??C for 5 minutes; followed by 35 cycles at 95??C for 40 seconds, 60??C for 1 minute, and 72??C for 2 minutes; and final extension at 72??C for 7 minutes. Ten-microliter aliquots of each PCR reaction were resolved by agarose gel electrophoresis, stained with ethidium bromide, and digitally photographed. PCR primers were ß-actin upstream, 5'-GGAGAAGATCTGGCATCACACCTTCTAC-3'; ß-actin downstream, 5'-TGGTCTCGTGGATACCGCAAGATTCCAT-3'; cav-1 specific upstream, 5'-AATATCCCCACGCCTCCCACACTG-3'; cav-1ß specific upstream, 5'-CTCCCACTCCACGCTCATGAACTTC-3'; and cav-1 and -1ß common downstream, 5'-GCTGTCAGCAGCCTGTAGCACC-3'.

Lysate Preparation and Immunoblotting of Embryonic and Adult Zebrafish

Staged zebrafish embryos were collected, quick frozen in liquid nitrogen, and stored at C80??C. Frozen embryos were solubilized in immunoprecipitation assay buffer (50 mmol/L Tris-HCl, pH 7.4, 1% Nonidet P-40, 0.25% sodium deoxycholate, 150 mmol/L NaCl, 1 mmol/L NaF, and 1 mmol/L sodium orthovanadate) and COMPLETE protease inhibitor cocktail (Roche, Mannheim, Germany). After a 30-minute incubation on ice, insoluble material was removed by centrifugation at 18,000 x g for 15 minutes at 4??C, and the supernatants were collected as whole cell lysates. Adult zebrafish lysates were prepared from anesthetized zebrafish quick frozen in liquid nitrogen and ground with mortar and pestle. Protein concentrations were determined using the Bio-Rad Dc protein assay kit (Bio-Rad Laboratories, Hercules, CA). Equal amounts of total cell lysate were solubilized in 3x cell lysis buffer , boiled for 15 minutes, and fractionated by SDS-polyacrylamide gel electrophoresis using 20% gels. Ponceau S staining was used to confirm equal loading of protein extracts. Adult zebrafish lysates were diluted 1:20 relative to developmentally staged extracts, due to the very high Cav-1 expression in adults. Fractionated proteins were electrotransferred to a nitrocellulose membrane (Bio-Rad Laboratories), blocked with 5% milk, and immunoblotted with the anti-human Cav-1 polyclonal antibody (1:1000 dilution; Transduction Labs). Cav-1 proteins were visualized using horseradish peroxidase-conjugated secondary antibody and the Supersignal Chemiluminescence System (Pierce, Rockford, IL).

Whole-Mount in Situ Hybridization

Digoxigenin-labeled sense and antisense riboprobes (Roche Applied Science, Indianapolis, IN) were generated from unique 5'- and shared 3'-untranslated regions of cav-1 and -1ß. Whole-mount in situ hybridization (WISH) was performed as previously described.32,33

Targeted Depletion and Rescue of Cav-1 and -1ß

Antisense morpholino oligomers (MOs) targeted to the ATG start sites of cav-1 and -1ß mRNAs were designed and synthesized by Gene Tools, LLC (Eugene, OR). Antisense and 4-mismatch negative control MOs at 1, 2, and 4 mg/ml were injected at 1- to 3-picoliter volumes into single-cell staged embryos, which were then allowed to develop at 28.5??C with periodic examination using Leica MZ12 (Wetzlar, Germany) and Zeiss M2Bio (Thornwood, NY) dissecting microscopes. The lowest MO concentration resulting in a consistent phenotype, 1 mg/ml, was used in all subsequent experiments. The specificity of each cav-1 isoform morphant phenotype was determined using human cav-1 (hcav-1) isoform cRNAs that do not contain the antisense MO binding site. Translation-competent cav-1 and -1ß cRNAs were transcribed with mMESSAGE mMACHINE (Ambion, Austin, TX) and co-injected with antisense MOs, as detailed in Figure 6 . We injected 100 pg per embryo of cav-1 cRNA, 100 pg per embryo of cav-1 Y14F cRNA, and 200 pg per embryo of cav-1ß cRNA.

Figure 6. Rescue of morphant phenotypes by co-injection of isoform specific human cav-1 cRNAs. Human wild-type cav-1 isoform cRNAs and Y14F cav-1 mutant cRNAs were tested for their ability to rescue zebrafish cav-1 morphant embryo phenotypes. Representative rescue of cav-1ß morphant phenotype by homotypic and not heterotypic human cav-1ß cRNAs are shown (top), along with all of the experimental rescue results. Zebrafish cav-1 morphants were almost fully rescued by homotypic human cav-1 (98.6%) but not by hcav-1-Y14F or heterotypic hcav-1ß cRNAs. Likewise, zebrafish cav-1ß morphants were rescued by homotypic human cav-1ß cRNAs (96.6%) but not by heterotypic hcav-1 or hcav-1 Y14F cRNAs. Embryos injected with hcav-1, hcav-1-Y14F, or hcav-1ß cRNAs alone appeared normal.

Phalloidin Staining of Actin Microfilaments

At 12 hours after fertilization (hpf), embryos were fixed in 3.7% paraformaldehyde (PFA) in actin stabilizing buffer (ASB) (1 mmol/L phenylmethylsulfonyl fluoride, 10 mmol/L EGTA, 10 mmol/L PIPES, pH 7.3, 5 mmol/L MgCl2, and 900 mmol/L KCl), washed extensively in ASB, blocked with 1% fetal bovine serum (FBS)/3% bovine serum albumin in ASB overnight, and labeled with 66 nmol/L Alexa 546 phalloidin (Molecular Probes, Eugene, OR) for 1 hour. At 24 hpf, embryos were fixed in sucrose/PFA for 3 hours and washed in PBS-Tween (PBT) and then in distilled water. Washed embryos were placed in prechilled C20??C acetone for 5 minutes and washed in PBT again. Embryos were then blocked with either PBT-2% dimethylsulfoxide (PBTD) and 10% goat serum or 5% goat serum and 5 mg/ml bovine serum albumin in PBTD for 2 hours and labeled with 165 nmol/L Alexa 546 phalloidin in PBDT and 1% goat serum for 2 hours and washed in PBTD. Labeled embryos were analyzed using a Leica TCS SP2 AOBS Confocal Laser Scanning Microscope.

Transmission Electron Microscopy Analysis of Zebrafish Caveolae

Control and cav-1/-1ß morphant 24-hpf embryos were dechorionated and fixed for 3 hours at room temperature and then overnight at 4??C in 2% PFA, 2.5% glutaraldehyde, and 0.1 mol/L sodium cacodylate, pH 7.4. Fixed embryos were treated with osmium collidine for 1 hour at room temperature, washed in 0.05 mol/L sodium maleate, pH 5.2, and dehydrated in successive incubations of 50, 70, 85, 95, and 100% (2x) EtOH, for 15 minutes each. Dehydrated embryos were incubated in acetone for 3 minutes and infiltrated in a 1:1 araldite/acetone mixture overnight at 4??C, followed by overnight incubation in 100% araldite at 4??C. Embryos were then incubated in fresh 100% araldite for 2 hours at room temperature and in fresh araldite overnight at 68??C. Embedded embryos were thin sectioned for electron microscopy analysis. Transmission electron microscopy (TEM) was performed at the Massachusetts General Hospital East Core Facility (Charlestown, MA).

Results

Cloning and Expression of Zebrafish cav-1 and -1ß

Two full-length zebrafish cav-1 isoform cDNAs, cav-1 and -1ß, were cloned by DOP-PCR/RACE (GenBank accession nos. AY124572 and AY124573, respectively). Zebrafish cav-1 and -1ß encode proteins of 181 and 148 amino acids, with predicted molecular weights of 21 and 17 kd, respectively. Comparison of zebrafish cav-1 and -1ß nucleotide sequences to a zebrafish genomic cav-1 clone (GenBank clone 66I19, accession no. AC087105), revealed that zebrafish cav-1 and -1ß represent alternatively spliced transcripts, as is the case for mammalian cav-1 transcripts.25,34 The predicted amino acid sequence of zebrafish Cav-1 exhibits 72 and 71% identity to human and mouse Cav-1, respectively, whereas the predicted amino acid sequence of zebrafish Cav-1ß exhibits 77 and 76% identity to human and mouse Cav-1ß, respectively (Figure 1B) . Immunoblot analysis of the zebrafish proteins expressed in mammalian cells demonstrated that zebrafish Cav-1 and -1ß migrate as expected in SDS-polyacrylamide gel electrophoresis gels (Figure 1C) .

Figure 1. Cav-1 and -1ß genomic organization, predicted amino acid sequences, and expression in cultured cells. A: Alignment of zebrafish cav-1 and -1ß cDNAs with the zebrafish cav-1 genomic clone 66I19. Base pairs 244 to 1748 of the cav-1 cDNA, 76 to 1580 of the cav-1ß cDNA, and 31,441 to 31,612 and 41,369 to 42,703 of the zebrafish genomic cav-1 clone exhibit 100% nucleotide sequence identity (triple-line boxes). Base pairs 1 to 243 of cav-1 and 1 to 75 of cav-1ß are distinct from each other but are identical to cav-1 genomic clone bp 30,576 to 30,819 and 31,336 to 31,441, respectively. A single line indicates cav-1 intronic sequence; open boxes indicate unique sequences of individual cav-1 isoforms that are identical to the genomic cav-1 clone 66I19 nucleotide sequence; triple line boxed regions and arrows indicate identical sequences. The coding region of cav-1 consists of three exons, the boundaries of which lie between bp 243/244 and 412/413, whereas that of cav-1ß is composed of two exons, the boundary of which falls between bp 244/245. ATG indicates the predicted initiation codon, and TAA indicates the predicted termination codon. B: Comparison of the predicted amino acid sequence of zebrafish Cav-1 to mouse and human Cav-1. The cav-1 cDNA encodes a protein of 181 amino acids. The cav-1ß cDNA encodes a protein of 148 amino acids. Asterisks indicate amino acid identity, and hyphens indicate absence of amino acids. The arrowhead present at amino acid 34 indicates the amino terminus of the zebrafish Cav-1ß isoform. Zebrafish Cav-1 exhibits 72% amino acid identity to human Cav-1 and 71% identity to mouse Cav-1. Zebrafish Cav-1ß exhibits 77% amino acid identity to human Cav-1ß and 76% identity to mouse Cav-1ß. C: Expression of zebrafish Cav-1 and -1ß in human cells. Western analysis of LNCaP and HEK 293 cell lines transfected with vector alone (VO) or zebrafish cav-1 and -1ß expression constructs. Zebrafish Cav-1 and -1ß exhibited molecular masses of 20.6 and 16.8 kd, respectively, similar to mammalian Cav-1 isoforms.

Developmental Expression and Tissue Specificity of Cav-1 Isoform mRNAs

Isoform-specific primers were used to assess cav-1 and -1ß mRNA expression by RT-PCR at discrete stages of zebrafish development (Figure 2A) . Both genes were weakly expressed as maternal mRNAs, whereas zygotic cav-1 mRNA expression increased dramatically between 6 and 12 hpf. Caveolin-1 protein (Cav-1 and -1ß) expression in developing zebrafish was examined by Western blot analysis. Using an antibody that detects both mammalian Cav-1 and -1ß isoforms,35 zebrafish Cav-1 was first detectable as a faint band at 18 hpf (Figure 2B) , exhibited steady expression levels between 20 to 28 hpf, appeared up-regulated between 28 and 48 hpf, and was detected at even higher levels in adults. The discrepancy between the developmental times where cav-1 mRNAs (0.7 hpf) and Cav-1 proteins (18 to 24 hpf) are first detected is likely due to the relative difference in sensitivity of the two assays (RT-PCR versus Western blot). The developmental expression profile of zebrafish Cav-1ß was similar to that of Cav-1 but at reduced levels. The zebrafish Cav-1-to-Cav-1ß expression ratios are similar to those observed in mammalian cells.36

Figure 2. Developmental expression of cav-1 isoform mRNAs and gene products. A: Developmental expression of cav-1 isoform mRNAs. RT-PCR indicated that both cav-1 mRNAs were weakly expressed as maternal mRNAs at 0.7 hpf, exhibited an increase in zygotic expression between 6 and 12 hpf, and were maintained at these levels throughout adulthood. NTC, no cDNA template control. B: Developmental expression of zebrafish Cav-1 protein isoforms. Protein extracts from developmentally staged and adult zebrafish were used to examine expression of endogenous Cav-1 and -1ß isoforms by Western blot. Cav-1 was first detected as a faint band at 18 hpf and increased in intensity between 24 and 48 hpf. Cav-1ß was first detectable as a faint band at 20 hpf and increased dramatically between 28 and 48 hpf. Both isoforms were highly expressed in the adult (adult protein extracts were diluted 1:20, relative to other extracts). Extracts from wt embryos injected with zebrafish cav-1 and -1ß cRNAs served as positive controls. Ponceau S staining of the blot shows relative protein loading.

The tissue-specific expression of each cav-1 splice variant was examined by WISH using either a probe that recognizes both cav-1 and -ß isoform mRNAs or a probe specific for cav-1ß mRNA only (Figure 3) . We were unable to generate a cav-1-specific probe suitable for WISH, despite repeated attempts. The common probe detected cav-1 mRNAs in the notochord, somites, and skin of 24-hpf embryos (Figure 3, ACC) , whereas at 48 hpf, expression was also detected in cranial and trunk sensory neuromasts (Figure 3, E versus F) . At 90 hpf, the common probe detected cav-1 mRNAs in the pharyngeal vasculature, intestinal epithelium, and skin (Figure 3G) , whereas the cav-1ß-specific probe detected expression in all of these tissues except intestinal epithelium (Figure 3, H versus I) . Together, these data indicate that both cav-1 and -1ß isoform mRNAs are expressed as maternal transcripts (Figure 2 ; data not shown) and are subsequently expressed in heart, pharyngeal vasculature, notochord, somites, skin, and neuromast tissues. In contrast, only cav-1 mRNA was detectable in intestinal epithelium.

Figure 3. Whole mount in situ hybridization analysis of zebrafish cav-1 and -1ß mRNAs in the embryo. At 24 hpf, a probe recognizing both cav-1 isoform mRNAs revealed ubiquitous cav-1 expression throughout the embryo, with distinct notochord and epidermal expression domains (ACC). At 48 hpf, cav-1 mRNAs were detected in skin (C) and in cranial and trunk neuromasts (E). At 90 hpf, cav-1 transcripts were detected in pharyngeal vasculature, intestinal epithelium, and skin (G and J, arrows). A probe specific for cav-1ß mRNAs shows expression of this isoform in pharyngeal vasculature, heart, and skin but not in intestinal epithelium (H). Sense controls were negative (D, F, and I). h, heart; ie, intestinal epithelium; nc, notochord; nm, neuromasts; pv, pharyngeal vasculature.

Zebrafish and Human Cav-1 Isoforms Exhibit Nonredundant and Conserved Functional Roles in Early Zebrafish Development

To identify putative developmental functions for individual Cav-1 isoforms, targeted protein depletion studies were performed using antisense MOs designed to hybridize to the unique translation start site of individual cav-1 isoform mRNAs and block their translation.37 Obvious defects in cav-1 morphant embryos were first evident at 12 hpf (not shown) and were quite distinct at 24 hpf, when neural, eye, and somite defects were observed, and morphant embryos appeared shortened along the anteroposterior (AP) axis relative to control embryos (Figure 4, E and I) . The specificity of targeted Cav-1 isoform protein depletions was confirmed by Western blot analysis of 28-hpf morphant embryos (Figure 4A , inset). At this time, Cav-1-depleted embryos exhibited obvious defects in retinal pigmented epithelium (RPE) differentiation, vacuolated neural ventricles, deformed notochord, and misshapen somites present in shortened tails composed of enlarged ventral tail tissue . Cav-1ß-depleted embryos exhibited flattened heads, neural ventricle defects, abnormal RPE, and irregularly shaped somites in otherwise fairly well-patterned tail tissue (Figure 4, ICL) . Analyses of thin-sectioned morphant embryo specimens confirmed abnormal cellular organization of eye, neural, RPE, and somite tissues (not shown). Molecular analyses revealed that 24-hpf cav-1 and -1ß morphant embryos exhibited reduced or lack of expression of neural markers pax2.1,38 engr3,39 fgf8,40 and otx2,41 examples of which are shown in Figure 5 , and aberrant expression of the muscle differentiation marker myoD42 in somite tissue, consistent with the observed neural and somite tissue defects. These data illustrate the dramatic effects on the embryo of the selective knockdown of either Cav-1 isoform.

Figure 4. Distinct phenotypes are exhibited by cav-1 and -1ß morphant embryos. ACD: Mismatch control MO-injected embryo; ECH: Cav-1 morphant embryo; ICL: Cav-1ß morphant embryo. A, inset: Western blot of 28-hpf Cav-1 and -1ß depleted and wt control embryos. C, G, K are dorsal views, and the remaining panels present lateral views, anterior to the left, dorsal up. Cav-1 morphant embryos exhibited curved bodies and kinked tails, lacked distinct mhb, and exhibited RPE defects (ECG) compared with control embryos (ACC). F: The heart was enlarged in cav-1 morphants, and neural tissues appeared vacuolated. G: Dorsal view shows disorganized RPE and abnormal neural folds. H: Lateral tail view reveals disorganized somites and shortened tails. I: Cav-1ß morphant embryos exhibited slightly curved bodies. J: Lateral head view reveals heart edema (arrow), eye pigmentation defects, and reduced neural tissues. K: Dorsal head view shows pigmentation defects and aberrant neural folding. L: Lateral tail view reveals notochord and somite defects, which appeared less severe than those of cav-1 morphants. e, eye; hb, hindbrain; mhb, midbrain hindbrain; s, somite; y, yolk. Scale bars = 200 µm.

Figure 5. Molecular characterization of 24-hpf cav-1 and -1ß morphant phenotypes. The neural markers otx2 and pax2.1 and the muscle marker myoD were used to characterize cav-1 and -1ß morphant phenotypes. The expression of neural markers was reduced and MyoD expression revealed aberrant somite patterning in Cav-1 and -1ß morphants.

Targeted Cav-1 isoform protein depletion studies were highly reproducible. At 24 hpf, 99.5% (426 of 428) of Cav-1-depleted embryos were phenotypic, compared with 2.3% (9 of 384) of cav-1 4-mismatch control MO-injected embryos. Likewise, 100% (178 of 178) of Cav-1ß-depleted embryos were phenotypic at 24 hpf, compared with 2.6% (5 of 192) of control MO-injected embryos. Specific depletion of only the targeted and not the alternate Cav-1 isoform was confirmed by Western blot analyses of 28-hpf morphant embryos (Figure 4A , inset). The specificity of our antisense MO-targeted protein depletion studies was confirmed by demonstrated rescue of morphant phenotypes with co-injected full-length cRNAs corresponding to human (Figure 6) and zebrafish (not shown) sequences. Criteria for rescue included proper neural fold formation, normal eye development including RPE differentiation, chevron-shaped somites, and proper vascular development. Human cav-1 and -1ß cDNAs were isolated, sequenced, cloned into expression vectors, and used to generate translation-competent cRNAs. Human cav-1 and -1ß cRNAs were tested for their ability to rescue isoform-specific and alternative isoform, morphant phenotypes as shown in Figure 6 . These experiments demonstrated that injection of human cav-1 cRNA reproducibly rescued the zebrafish cav-1 morphant phenotype (98.6%, 275 of 279), and co-injection of human cav-1ß cRNA reproducibly rescued the zebrafish cav-1ß morphant phenotype (96.6%, 86 of 89). In contrast, zebrafish morphant phenotypes were not rescued by co-injection of alternative human cav-1 isotype cRNAs. The rescue results were essentially identical using zebrafish instead of human cav-1 cRNAs (not shown). These results suggest functionally nonredundant roles for individual Cav-1 and -1ß isoforms in early zebrafish development. Importantly, they also demonstrate that human Cav-1 isoforms are capable of performing the functions of zebrafish Cav-1 isoforms within the early developmental time frame studied here.

The Cav-1 Y14 Phosphorylation Site Is Required for Early Zebrafish Development

An interesting difference between Cav-1 and -1ß is the presence of a Src kinase phosphorylation site at Y14 in the unique amino-terminal region of Cav-1. The Cav-1 Y14 Src phosphorylation site has been implicated in Cav-1 function using mammalian cell culture models.43 This site is conserved in the zebrafish Cav-1 isoform, suggesting that it performs a functional role. We took advantage of the zebrafish cav-1 morphant phenotypes to determine whether a conservative but nonphosphorylatable human Cav-1 mutant (Y14F) was capable of rescuing the cav1- morphant phenotype. We reasoned that if Y14 phosphorylation is not required for early developmental functions of Cav-1, then co-injection of cav-1 Y14F cRNA would rescue the cav-1 morphant phenotype. We also examined whether the human Y14F Cav-1 isoform could rescue the cav-1ß morphant phenotype, allowing for the possibility that the Y14F site is the primary functional difference between the two Cav isoforms and that cav-1 Y14F might functionally mimic Cav-1ß. If so, cav-1 Y14F cRNA would rescue the cav-1ß morphant phenotype. Our results, summarized in Figure 6 , demonstrated that cav-1 Y14F cRNAs could not rescue either the cav-1 or cav-1ß morphant phenotype. Together, these results suggest that, although the Y14 phosphorylation site is required for Cav-1 function in early zebrafish development, this difference in the two proteins does not by itself define the functional difference between the Cav-1 and -1ß isoforms.

Cav-1 and -1ß Are Required for Caveolae Formation, Actin Cytoskeletal Organization, and Vascular Development

Transmission electron microscopy (TEM) was used to determine the effect of Cav-1 depletion on the formation of caveolae in cav-1/1ß double morphant embryos (Figure 7) . At 24 hpf, numerous caveolae were clearly identifiable in notochord tissues of control MO-injected embryos, whereas few caveolae were evident in cav-1/-1ß double morphant notochord tissue (Figure 7, A versus B, arrows). Caveolae present in notochord tissue sections obtained from similar AP positions of 24-hpf cav-1/-1ß morphant and control embryos were quantified. Cav-1/-1ß morphant notochord tissue contained only 16% of the caveolae present in control embryos (0.6 versus 3.8 caveolae/µm). Similarly, although caveolae were found in neural and somite tissues of 24-hpf wt and control MO-injected embryos, none were found in cav-1/-1ß morphant neural tissues (not shown). Interestingly, caveolae appear to be transient structures in the developing zebrafish notochord, with numerous caveolae present at 24 hpf and few to none detectable in 4-day postfertilization wt embryos (not shown). Together, these results demonstrate that targeted depletion of Cav-1 and -1ß results in reduced numbers of caveolae. TEM analyses also revealed additional characteristics that distinguished wt and cav-1/-1ß morphant tissues. Control 24-hpf embryos exhibited abundant, polarized somitic myofibrils, which were surrounded by sarcoplasmic reticulum (SR) transverse systems (Figure 7C) . In contrast, 24-hpf cav-1/-1ß morphant somites contained only disorganized cytoplasmic structures, in which myofibrils were not readily identifiable, and organized SR was not evident (Figure 7D) .

Figure 7. TEM analysis of caveolae in 24-hpf Cav-1/-1ß-depleted embryos. Numerous caveolae were detected in notochord tissues of wt zebrafish (A, arrows), whereas very few were observed in cav-1/1ß morphant notochord (B, arrow). C: Wt somite tissues exhibit organized myofibrils surrounded by distinct SR (arrows). D: Cav-1/1ß morphant somite tissues exhibited few myofibrils, and those were not attended by SR. Statistical analyses, using one-tailed (P = 0.005) and two-tailed (P = 0.01) Student??s t-test, confirmed that the differences in the frequency of caveolae were significant. Cropped images were generated from original magnifications of 54,000 (A), 52,000 (B), 73,500 (C), and 73,500 (D).

We next examined actin cytoskeleton organization in cav-1 morphant embryos because of published reports in cell culture models implicating caveolins and/or caveolae in the maintenance of actin cytoskeletal organization44,45 and because of the apparent disruption in the myofibrillar cytoskeleton seen by TEM in the cav-1/-1ß morphants (Figure 7) . Actin-specific phalloidin staining46 of head and tail epidermis of 12-hpf Cav-1 and -1ß-depleted embryos revealed highly irregular cell borders and punctate fluorescence patterns, indicative of disrupted actin microfilaments in these tissues (Figure 8) . The actin microfilament networks in Cav-1ß-depleted somites appeared less disrupted than those of cav-1 morphant embryos (Figure 8 , 24-hpf tail panels), consistent with the relatively intact vasculature observed in Cav-1ß-depleted embryos in comparison with cav-1 morphant embryos (Figure 9) . At 24 hpf, somite defects were clearly evident in phalloidin-stained cav-1 morphant embryos (Figure 8 , bottom panels). Cav-1 morphants exhibited little to no organized somite tissues, whereas cav-1ß morphant somites lacked a midline myoseptum and the classic chevron shapes were shortened between each intersegmental boundary and exhibited thinner and disorganized myofibrils relative to that of control embryos. We also observed aberrant pigment granule movements in retinal pigment epithelium (RPE) and melanocytes of cav-1 and -1ß morphants (data not shown), suggestive of disrupted cytoskeletal organization and consistent with purported roles for actin microfilaments in intracellular pigment granule trafficking.47-49

Figure 8. Disrupted actin cytoskeletal organization in 12- and 24-hpf cav-1 and -1ß morphants. Confocal images of phalloidin-stained actin cytoskeletal filaments in head and tail epidermis of 12- and 24-hpf cav-1 and -1ß morphants and mismatch control-injected embryos. At 12 hpf, actin filaments of head and tail epidermis of cav-1 and -1ß morphants appeared disorganized, particularly in cav-1 morphants, in which extensive depolymerized and disrupted cytoskeletal architecture was evident by punctate staining at cell-cell junctions. Cell size varied extensively in cav-1 and -1ß morphants, consistent with cytokinesis defects. At 24 hpf, distinct defects were evident in actin cytoskeletal organization in somite myofibrils. cav-1 morphants exhibited little to no identifiable myofibrils, whereas myofibrils in cav-1ß morphant somite tissue appeared dysmorphic and disorganized, lacked midline myoseptum and chevron shape, and were shortened along the AP axis.

Figure 9. Cav-1 and -1ß are required for proper patterning of the vascular endothelium. Lateral views of control (A, A', B, and B'), cav-1 morphant (C, C', D, and D'), and cav-1ß morphant (E, E', F, and F') transgenic fli-EGFP embryos. Anterior is to the left, and dorsal is up. Control embryos exhibited green fluorescent protein expression in well-patterned head and tail vascular endothelial cells. In contrast, Cav-1-depleted head and tail vascular endothelium appeared mispatterned and did not form patent vessels (C, C', D, and D'). Head vasculature of Cav-1ß-depleted embryos appeared disorganized, whereas tail vasculature was less affected (E, E', F, and F').

Caveolae are abundant in endothelial cells where they have been implicated in vascular function.50 We took advantage of fli-EGFP transgenic zebrafish,28 which express green fluorescent protein in vascular endothelial cells, to examine blood vessel formation in Cav-1-depleted embryos. Confocal imaging of 24-hpf control fli-EGFP embryos revealed well-patterned blood vessels in both head and tail tissues (Figure 9, A, A', B, and B') . In contrast, Cav-1-depleted embryos exhibited substantial vascular patterning defects in both head and tail tissues (Figure 9, C, C', D, and D') , whereas Cav-1ß morphant embryos exhibited extensively disrupted head but fairly well-patterned tail tissue vasculature (Figure 9, E, E', F, and F') .

Discussion

This is the first report to demonstrate a requirement for both the Cav-1 and -ß isoforms during vertebrate development. We demonstrate here that targeted depletion of either isoform results in embryonic lethality in the zebrafish and that the heterotypic isoform is incapable of substituting for the alternative form of the protein. These results are in marked contrast to Cav-1 knockout phenotypes observed in mice, where loss of both Cav-1 isoforms results in animals that are viable and fertile, live well into adulthood, and exhibit rather modest functional defects. The results in the mouse are surprising because, in the Cav-1 knockout animals, a prominent and presumably important class of subcellular organelles (caveolae) is completely absent in essential nonmuscle tissues, such as the vasculature. In the zebrafish, targeted depletion of either Cav-1 isoform resulted in morphological defects evident at 12 hpf and distinct neural, somitic, and vascular tissue defects evident by 24 hpf. The inability of one isoform to substitute for the other is an unanticipated observation because the proteins are very similar: Cav-1ß lacks the first 33 amino acids of Cav-1 but is otherwise identical. In the present study, we demonstrate that early morphological defects observed in Cav-1 isoform-targeted morphant embryos coincided with a failure to form caveolae, a breakdown in actin cytoskeletal organization, and extensive disruption of vascular endothelial tissue formation, consistent with published functions for caveolins (and caveolae) in each of these processes. Embryonic lethality of Cav-1 depletion in the zebrafish embryo, without regard to isoform specificity, was recently reported by another group, and our conclusion about the necessity for Cav-1 functions in early zebrafish development is consistent with theirs.51

Cav-1 and -1ß isoforms, initially identified as alternative translation products from a single mRNA,36 were subsequently demonstrated to form by differential RNA splicing.25 Collectively, our studies indicate that the two Cav-1 isoforms perform nonoverlapping roles during zebrafish development. We arrived at this conclusion from several independent observations. 1) Defects arising from targeted depletion of one Cav-1 isoform were not rescued by the heterotypic isoform but were rescued by the homotypic isoform. The ability of the human cav-1 isoform cRNAs to rescue defects arising from depletion of the zebrafish proteins suggests strong functional conservation across a long evolutionary timeline. 2) Targeted depletion of Cav-1 resulted in severe neural and tail tissue defects, whereas targeted depletion of Cav-1ß more selectively affected neural tissues. Isoform-characteristic patterning defects of cav-1 and -1ß morphant tissues were also clearly evident 3) with phalloidin-staining of the actin cytoskeleton and 4) in the vasculature of fli-EGFP, cav-1 morphant zebrafish. 5) WISH analyses revealed that only cav-1 mRNA was detected in intestinal epithelium, consistent with a recent report of Cav-1/Annexin 2 heterocomplex formation in intestinal sterol transport.51 6) Mutation of the Y14 Src phosphorylation site to the nonphosphorylatable residue phenylalanine disrupted the ability of human cav-1 cRNA to rescue the zebrafish cav-1 morphant defect. This mutation also did not convert Cav-1 into a Cav-1ß mimic. These results suggest that this tyrosine residue is essential for Cav-1 function and therefore is an important molecular difference between the two proteins. 7) Finally, the genomic organization (including intron/exon boundaries) specifying the zebrafish and mammalian cav-1 isoforms are conserved, implying strong evolutionarily constraints on the expression of the Cav-1 splice variants. We submitted the zebrafish cav-1 and -1ß cDNA sequence data to GenBank in 2003 (accession nos. AY124572 and AY124573). After GenBank released this information to the public, the existence of two zebrafish cav-1 isoform mRNAs was reported by another group.22

To date, very little published evidence exists supporting discrete functions for individual Cav-1 isoform proteins. One recent study, reporting Cav-1 isoform-specific interactions with bone morphogenetic protein (BMP) receptors RIa and RII, suggests that Cav-1ß inhibits BMP signaling, whereas Cav-1 does not.52 Another study used immunofluorescence and Western blotting to show that, although both Cav-1 isoforms were initially co-distributed in cardiac tissue at birth, at later times, Cav-1 exhibited restricted expression in endothelial tissues.53

At the present time, it is not clear whether developmental defects observed in Cav-1-depleted embryos are a primary or secondary effect of cytoskeletal disruption. Actin microfilament defects preceded and mirrored later tissue specific defects, suggesting a functional link between these observations. This study is the first in vivo demonstration that down-regulation of Cav-1 isoforms results in disruption of the actin cytoskeleton, consistent with mammalian cell culture studies demonstrating that actin microfilament defects arise from loss of Cav-1.54 Izumi et al55 first demonstrated that caveolae present in endothelial membranes could be decorated with phalloidin, which binds to actin at the junction between subunits.56-58 More recently, filamin, an F-actin cross-linking protein and Cav-1 ligand, was shown to directly link Cav-1 to the actin cytoskeleton, and Cav-1-positive membranes were demonstrated to redistribute to filamin-decorated stress fibers in response to the low-molecular weight GTPase RhoB, known to promote cytoskeletal reorganization in vitro.44 That study also used a quantitative yeast two hybrid approach to demonstrate that Cav-1ß, and not Cav-1, might be the primary ligand for filamin. Further confirmation of this result would provide additional evidence for discrete biochemical functions for the two Cav-1 isoforms. Additional evidence for an actin-caveolin interaction includes the demonstration that Cav-1 resides in structures at the plasma membrane that appear to be assembled into the cortical actin cytoskeleton59,60 and that disruption of filamentous actin by cytochalasin D treatment results in enhanced mobility and clustering of cell surface caveolae.59 Finally, numerous reports have shown that a variety of proteins known to communicate with or participate in the formation of cytoskeletal structures localize to lipid rafts and/or caveolae.8,61-69

The biological function(s) of caveolae is incompletely understood, despite a substantial body of literature linking these specialized organelles to multiple signal transduction processes and membrane trafficking. Consistent with observations from the Cav mouse knockout studies, we have shown that simultaneous targeted depletion of Cav-1 and -1ß in zebrafish resulted in dramatically reduced numbers of caveolae, in accord with the view that caveolins are integral membrane proteins required for the distinctive caveolar architecture.7,44 The reason for the disparity in the requirement for Cav-1 during development between mouse and zebrafish is unknown. It is possible that Cav-1 serves specialized functions in zebrafish that can be compensated for by redundant system(s) in the mouse. Another possibility is that the difference may arise from features characteristic of placental versus nonplacental embryonic development. Further elucidation of the molecular mechanisms of mammalian and zebrafish Cav-1 signaling pathways will clarify this issue.

Conclusions

In this study, we present a functional characterization of Cav-1 and -1ß, key structural proteins of caveolar lipid rafts, in the developing zebrafish. Our primary conclusions are that 1) expression of both Cav-1 isoforms is required for embryonic development in the zebrafish, because targeted depletion of either Cav-1 isoform results in tissue patterning defects evident as early as 12 hpf; 2) morphological characterization of Cav 1 morphants, tissue distribution analysis, and rescue experiments indicate that Cav-1 and -1ß exhibit distinct and nonredundant roles in early zebrafish development; 3) human Cav-1 and -1ß rescued the early developmental functions of zebrafish Cav-1 isoforms; 4) the Cav-1 Y14 Src phosphorylation site is required for the developmental function(s) of Cav-1; and 5) targeted depletion of either Cav-1 or -1ß results in disrupted actin cytoskeletal organization, which likely contributes to later vascular and developmental defects. Together, these findings suggest important and previously unrecognized developmental roles for Cav-1 and, by implication, for caveolar lipid rafts in vertebrate organogenesis.

Acknowledgements

We thank members of the Yelick and Freeman laboratories and Dr. Christopher Pierson (Children??s Hospital Boston, Boston, MA) for helpful discussions; Loic Fabricant, Kevin Tong, and Seija Cope for expert zebrafish husbandry; Dan Brown (Beth Israel Deaconess Medical Center, Boston, MA) and Dr. Zie Skobe (The Forsyth Institute) for electron microscopy analysis and advice; Elka Pravda for confocal analyses; and Paul Guthrie for expert figure preparation.

【参考文献】
  Kurzchalia TV, Dupree P, Parton RG, Kellner R, Virta H, Lehnert M, Simons K: VIP21, a 21-kD membrane protein is an integral component of trans-Golgi-network-derived transport vesicles. J Cell Biol 1992, 118:1003-1014

Lisanti MP, Scherer PE, Vidugiriene J, Tang Z, Hermanowski-Vosatka A, Tu YH, Cook RF, Sargiacomo M: Characterization of caveolin-rich membrane domains isolated from an endothelial-rich source: implications for human disease. J Cell Biol 1994, 126:111-126

Uittenbogaard A, Ying Y, Smart EJ: Characterization of a cytosolic heat-shock protein-caveolin chaperone complex: involvement in cholesterol trafficking. J Biol Chem 1998, 273:6525-6532

Galbiati F, Volonte D, Brown AM, Weinstein DE, Ben-Ze??ev A, Pestell RG, Lisanti MP: Caveolin-1 expression inhibits Wnt/beta-catenin/Lef-1 signaling by recruiting beta-catenin to caveolae membrane domains. J Biol Chem 2000, 275:23368-23377

Harris J, Werling D, Hope JC, Taylor G, Howard CJ: Caveolae and caveolin in immune cells: distribution and functions. Trends Immunol 2002, 23:158-164

Razani B, Woodman SE, Lisanti MP: Caveolae: from cell biology to animal physiology. Pharmacol Rev 2002, 54:431-467

Hnasko R, Lisanti MP: The biology of caveolae: lessons from caveolin knockout mice and implications for human disease. Mol Interv 2003, 3:445-464

van Deurs B, Roepstorff K, Hommelgaard AM, Sandvig K: Caveolae: anchored, multifunctional platforms in the lipid ocean. Trends Cell Biol 2003, 13:92-100

Smart EJ, Graf GA, McNiven MA, Sessa WC, Engelman JA, Scherer PE, Okamoto T, Lisanti MP: Caveolins, liquid-ordered domains, and signal transduction. Mol Cell Biol 1999, 19:7289-7304

Pelkmans L, Kartenbeck J, Helenius A: Caveolar endocytosis of simian virus 40 reveals a new two-step vesicular-transport pathway to the ER. Nat Cell Biol 2001, 3:473-483

Dvorak AM, Feng D: The vesiculo-vacuolar organelle (VVO): a new endothelial cell permeability organelle. J Histochem Cytochem 2001, 49:419-432

Simons K, Idonen E: Functional rafts in cell membranes. Nature 1997, 387:569-572

Harder T, Simons K: Caveolae, DIGS and the dynamics of sphingolipid-cholesterol microdomains. Curr Opin Cell Biol 1997, 9:534-542

Kurzchalia TV, Parton RG: Membrane microdomains and caveolae. Curr Opin Cell Biol 1999, 11:424-431

Song KS, Tang Z, Li S, Lisanti MP: Mutational analysis of the properties of caveolin-1: a novel role for the C-terminal domain in mediating homo-typic caveolin-caveolin interactions. J Biol Chem 1997, 272:4398-4403

Li S, Galbiati F, Volonte D, Sargiacomo M, Engelman JA, Das K, Scherer PE, Lisanti MP: Mutational analysis of caveolin-induced vesicle formation: expression of caveolin-1 recruits caveolin-2 to caveolae membranes. FEBS Lett 1998, 434:127-134

Park DS, Woodman SE, Schubert W, Cohen AW, Frank PG, Chandra M, Shirani J, Razani B, Tang B, Jelicks LA, Factor SM, Weiss LM, Tanowitz HB, Lisanti MP: Caveolin-1/3 double-knockout mice are viable, but lack both muscle and non-muscle caveolae, and develop a severe cardiomyopathic phenotype. Am J Pathol 2002, 160:2207-2217

Park DS, Cohen AW, Frank PG, Razani B, Lee H, Williams TM, Chandra M, Shirani J, De Souza AP, Tang B, Jelicks LA, Factor SM, Weiss LM, Tanowitz HB, Lisanti MP: Caveolin-1 null (C/C) mice show dramatic reductions in life span. Biochemistry 2003, 42:15124-15131

Razani B, Engelman JA, Wang XB, Schubert W, Zhang XL, Marks CB, Macaluso F, Russell RG, Li M, Pestell RG, Di Vizio D, Hou H, Jr, Kneitz B, Lagaud G, Christ GJ, Edelmann W, Lisanti MP: Caveolin-1 null mice are viable but show evidence of hyperproliferative and vascular abnormalities. J Biol Chem 2001, 276:38121-38138

Zhao YY, Liu Y, Stan RV, Fan L, Gu Y, Dalton N, Chu PH, Peterson K, Ross J, Jr, Chien KR: Defects in caveolin-1 cause dilated cardiomyopathy and pulmonary hypertension in knockout mice. Proc Natl Acad Sci USA 2002, 99:11375-11380

Sotgia F, Williams TM, Schubert W, Medina F, Minetti C, Pestell RG, Lisanti MP: Caveolin-1 deficiency (C/C) conveys premalignant alterations in mammary epithelia, with abnormal lumen formation, growth factor independence, and cell invasiveness. Am J Pathol 2006, 168:292-309

Nixon SJ, Wegner J, Ferguson C, Mery P-F, Hancock JF, Currie PD, Key B, Westerfield M, Parton RG: Zebrafish as a model for caveolin-associated muscle disease: caveolin-3 is required for myofibril organization and muscle cell patterning. Hum Mol Genet 2005, 14:1727-1743

Glenney Jr, JR, Soppet D: Sequence and expression of caveolin, a protein component of caveolae plasma membrane domains phosphorylated on tyrosine in Rous sarcoma virus-transformed fibroblasts. Proc Natl Acad Sci USA 1992, 89:10517-10521

Tang Z, Okamoto T, Boontrakulpoontawee P, Katada T, Otsuka AJ, Lisanti MP: Identification, sequence, and expression of an invertebrate caveolin gene family from the nematode Caenorhabditis elegans: implications for the molecular evolution of mammalian caveolin genes. J Biol Chem 1997, 272:2437-2445

Kogo H, Fujimoto T: Caveolin-1 isoforms are encoded by distinct mRNAs: identification of mouse caveolin-1 mRNA variants caused by alternative transcription initiation and splicing. FEBS Lett 2000, 465:119-123

Chen D, Zangl AL, Zhao Q, Markley JL, Zheng J, Bird IM, Magness RR: Ovine caveolin-1: cDNA cloning, E. coli expression, and association with endothelial nitric oxide synthase. Mol Cell Endocrinol 2001, 175:41-56

Razani B, Park DS, Miyanaga Y, Ghatpande A, Cohen J, Wang XB, Scherer PE, Evans T, Lisanti MP: Molecular cloning and developmental expression of the caveolin gene family in the amphibian Xenopus laevis. Biochemistry 2002, 41:7914-7924

Lawson ND, Weinstein BM: In vivo imaging of embryonic vascular development using transgenic zebrafish. Dev Biol 2002, 248:307-318

Westerfield M: The Zebrafish Book. 1995 University of Oregon Press, Eugene, OR

Kimmel CB, Ballard WW, Kimmel SR, Ullmannm B, Schilling TF: Stages of embryonic development of the zebrafish. Dev Dyn 1995, 203:253-310

Lu ML, Schneider MC, Zheng Y, Zhang X, Richie JP: Caveolin-1 interacts with androgen receptor: a positive modulator of androgen receptor mediated transactivation. J Biol Chem 2001, 276:13442-13451

Thisse C, Thisse B, Schilling TF, Postlethwait JH: Structure of the zebrafish snail1 gene and its expression in wild-type, spadetail and no tail mutant embryos. Development 1993, 119:1203-1215

Payne-Ferreira TL, Yelick PC: Alk8 is required for neural crest cell formation and development of pharyngeal arch cartilages. Dev Dyn 2003, 228:683-696

Kogo H, Aiba T, Fujimoto T: Cell type-specific occurrence of caveolin-1 alpha and -1 beta in the lung caused by expression of distinct mRNAs. J Biol Chem 2004, 279:25574-25581

Fujimoto T, Kogo H, Nomura R, Une T: Isoforms of caveolin-1 and caveolar structure. J Cell Sci 2000, 113:3509-3517

Scherer PE, Tang Z, Chun M, Sargiacomo M, Lodish HF, Lisanti MP: Caveolin isoforms differ in their N-terminal protein sequence and subcellular distribution: identification and epitope mapping of an isoform-specific monoclonal antibody probe. J Biol Chem 1995, 270:16395-16401

Nasevicius A, Ekker SC: Effective targeted gene ??knockdown?? in zebrafish. Nat Genet 2000, 26:216-220

Picker A, Scholpp S, Bohli H, Takeda H, Brand M: A novel positive transcriptional feedback loop in midbrain-hindbrain boundary development is revealed through analysis of the zebrafish pax2.1 promoter in transgenic lines. Development 2002, 129:3227-3239

Scholpp S, Brand M: Morpholino-induced knockdown of zebrafish engrailed genes eng2 and eng3 reveals redundant and unique functions in midbrain-hindbrain boundary development. Genesis 2001, 30:129-133

Reifers F, Bohli H, Walsh EC, Crossley PH, Stainier DY, Brand M: Fgf8 is mutated in zebrafish acerebellar (ace) mutants and is required for maintenance of midbrain-hindbrain boundary development and somitogenesis. Development 1998, 125:2381-2395

Mori H, Miyazaki Y, Morita T, Nitta H, Mishina M: Different spatio-temporal expressions of three otx homeoprotein transcripts during zebrafish embryogenesis. Brain Res Mol Brain Res 1994, 27:221-231

Weinberg ES, Allende ML, Kelly CS, Abdelhamid A, Murakami T, Andermann P, Doerre OG, Grunwald DJ, Riggleman B: Developmental regulation of zebrafish MyoD in wild-type, no tail and spadetail embryos. Development 1996, 122:271-280

Li S, Seitz R, Lisanti MP: Phosphorylation of caveolin by src tyrosine kinases: the alpha-isoform of caveolin is selectively phosphorylated by v-Src in vivo. J Biol Chem 1996, 271:3863-3868

Stahlhut M, van Deurs B: Identification of filamin as a novel ligand for caveolin-1: evidence for the organization of caveolin-1-associated membrane domains by the actin cytoskeleton. Mol Biol Cell 2000, 11:325-337

Mundy DI, Machleidt T, Ying YS, Anderson RG, Bloom GS: Dual control of caveolar membrane traffic by microtubules and the actin cytoskeleton. J Cell Sci 2002, 115:4327-4339

Becker KA, Hart NH: Reorganization of filamentous actin and myosin-II in zebrafish eggs correlates temporally and spatially with cortical granule exocytosis. J Cell Sci 1999, 112:97-110

Burnside B, King-Smith CL: Actin dependent pigment granule transport in retinal pigment epithelial cells. Biol Bull 1997, 192:181-182

Lambert J, Vancoillie G, Naeyaert JM: Molecular motors and their role in pigmentation. Cell Mol Biol 1999, 45:905-918

McNeil EL, Taceloski D, Basciano P, Biallas B, Williams R, Damiani P, Deacon S, Fox C, Stewart B, Petruzzi N, Osborn C, Klinger K, Sellers JR, Smith CK: Actin-dependent motility of melanosomes from fish retinal pigment epithelial (RPE) cells investigated using in vitro motility assays. Cell Motil Cytoskeleton 2004, 58:71-82

Drab M, Verkade P, Elger M, Kasper M, Lohn M, Lauterbach B, Menne J, Lindschau C, Mende F, Luft FC, Schedl A, Haller H, Kurzchalia TV: Loss of caveolae, vascular dysfunction, and pulmonary defects in caveolin-1 gene-disrupted mice. Science 2001, 293:2449-2452

Smart EJ, De Rose RA, Farber SA: Annexin 2-caveolin 1 complex is a target of ezetimibe and regulates intestinal cholesterol transport. Proc Natl Acad Sci USA 2004, 101:3450-3455

Nohe A, Keating E, Underhill TM, Knaus P, Petersen NO: Dynamics and interaction of caveolin-1 isoforms with BMP-receptors. J Cell Sci 2005, 118:643-650

Ratajczak P, Oliviero P, Marotte F, Kolar F, Ostadal B, Samuel JL: Expression and localization of caveolins during postnatal development in rat heart: implication of thyroid hormone. J Appl Physiol 2005, 99:244-251

Navarro A, Anand-Apte B, Parat MO: A role for caveolae in cell migration. FASEB J 2004, 18:1801-1811

Izumi T, Shibata Y, Yamamoto T: Striped structures on the cytoplasmic surface membranes of the endothelial vesicles of the rat aorta revealed by quick-freeze, deep etching replicas. Anat Rec 1988, 220:225-232

Barden JA, Miki M, Hambly BD, DosRemedios CG: Localization of the phalloidin and nucleotide-binding sites on actin. Eur J Biochem 1987, 162:583-588

Faulstich H, Zobeley S, Heintz D, Drewes G: Probing the phalloidin binding site of actin. FEBS Lett 1993, 318:218-222

Steinmetz MO, Stoffler D, Muller SA, Jahn W, Wolpensinger B, Goldie KN, Engel A, Faulstich H, Aebi U: Evaluating atomic models of F-actin with an undecagold-tagged phalloidin derivative. J Mol Biol 1998, 276:1-6

Thomsen P, Roepstorff K, Stahlhut M, van Deurs B: Caveolae are highly immobile plasma membrane microdomains, which are not involved in constitutive endocytic trafficking. Mol Biol Cell 2002, 13:238-250

Pelkmans L, Zerial M: Kinase-regulated quantal assemblies and kiss-and-run recycling of caveolae. Nature 2005, 436:128-133

Leitinger B, Hogg N: The involvement of lipid rafts in the regulation of integrin function. J Cell Sci 2002, 115:963-972

Skubitz KM, Campbell KD, Skubitz AP: CD63 associates with CD11/CD18 in large detergent-resistant complexes after translocation to the cell surface in human neutrophils. FEBS Lett 2000, 469:52-56

Baron W, Decker L, Colognato H, Ffrench-Constant C: Regulation of integrin growth factor interactions in oligodendrocytes by lipid raft microdomains. Curr Biol 2003, 13:151-155

Thorne RF, Marshall JF, Shafren DR, Gibson PG, Hart IR, Burns GF: The integrins alpha3beta1 and alpha6beta1 physically and functionally associate with CD36 in human melanoma cells: requirement for the extracellular domain OF CD36. J Biol Chem 2000, 275:35264-35275

Green JM, Zhelesnyak A, Chung J, Lindberg FP, Sarfati M, Frazier WA, Brown EJ: Role of cholesterol in formation and function of a signaling complex involving avß3, integrin-associated protein (CD47), and heterotrimeric G proteins. J Cell Biol 1999, 146:673-682

Hocking DC, Kowalski K: A cryptic fragment from fibronectin??s III1 module localizes to lipid rafts and stimulates cell growth and contractility. J Cell Biol 2002, 158:175-184

Michaely PA, Mineo C, Ying YS, Anderson RG: Polarized distribution of endogenous Rac1 and RhoA at the cell surface. J Biol Chem 1999, 274:21430-21436

del Pozo MA, Alderson NB, Kiosses WB, Chiang HH, Anderson RG, Schwartz AM: Integrins regulate Rac targeting by internalization of membrane domains. Science 2004, 303:839-842

MacLellan DL, Steen H, Adam RM, Garlick M, Zurakowski D, Gygi SP, Freeman MR, Solomon KR: A quantitative proteomic analysis of growth factor-induced compositional changes in lipid rafts of human smooth muscle cells. Proteomics 2005, 5:4733-4742

作者单位：From the Urological Diseases Research Center,* and the Department of Orthopaedic Surgery, Children??s Hospital Boston, Harvard Medical School; Massachusetts General Hospital; and the Department of Cytokine Biology, The Forsyth Institute, and Department of Developmental Biology, Harvard School of Den