Two Very Long Chain Fatty Acid Acyl-CoA Synthetase Genes, acs-20 and acs-22, Have Roles in the Cuticle Surface Barrier in Caenorhabditis elegans

In multicellular organisms, the surface barrier is essential for maintaining the internal environment. In mammals, the barrier is the stratum corneum. Fatty acid transport protein 4 (FATP4) is a key factor involved in forming the stratum corneum barrier. Mice lacking Fatp4 display early neonatal lethality with features such as tight, thick, and shiny skin, and a defective skin barrier. These symptoms are strikingly similar to those of a human skin disease called restrictive dermopathy. FATP4 is a member of the FATP family that possesses acyl-CoA synthetase activity for very long chain fatty acids. How Fatp4 contributes to skin barrier function, however, remains to be elucidated. In the present study, we characterized two Caenorhabditis elegans genes, acs-20 and acs-22, that are homologous to mammalian FATPs. Animals with mutant acs-20 exhibited defects in the cuticle barrier, which normally prevents the penetration of small molecules. acs-20 mutant animals also exhibited abnormalities in the cuticle structure, but not in epidermal cell fate or cell integrity. The acs-22 mutants rarely showed a barrier defect, whereas acs-20;acs-22 double mutants had severely disrupted barrier function. Moreover, the barrier defects of acs-20 and acs-20;acs-22 mutants were rescued by acs-20, acs-22, or human Fatp4 transgenes. We further demonstrated that the incorporation of exogenous very long chain fatty acids into sphingomyelin was reduced in acs-20 and acs-22 mutants. These findings indicate that C. elegans Fatp4 homologue(s) have a crucial role in the surface barrier function and this model might be useful for studying the fundamental molecular mechanisms underlying human skin barrier and relevant diseases.


Introduction
The surface barrier is essential for maintaining the internal environment in multicellular organisms. The stratum corneum comprises this barrier in mammals [1,2]. Fatty acid transport protein 4 (FATP4) is one of the key factors involved in forming the stratum corneum barrier [3,4,5] and was initially identified as a transport protein that facilitates the uptake of long chain fatty acids [6]. FATP also possesses acyl-coenzyme A (CoA) synthetase activity, which esterifies fatty acids into metabolically active CoA thioesters [7]. The mammalian FATP family comprises six members that differ in their biochemical activities, tissue expression patterns, and physiologic roles [8]. FATP1, 2, and 4 possess acyl-CoA synthetase activity towards very long chain fatty acids (VLCFAs) [9]. FATP5 is involved in metabolizing bile acids [10]. FATP3 and FATP6 were recently cloned and characterized, but their physiologic roles remain largely unclear [11,12]. Several lines of evidence indicate the involvement of FATPs in some human diseases. FATP1 is the major FATP family member expressed in adipose tissue and in skeletal muscle [6] and FATP1 knockout (KO) mice are completely resistant to diet-induced obesity, insulin desensitization, and other parameters of metabolic syndrome [13]. FATP4, one of the most well studied FATP members, is predominantly expressed in intestinal epithelial cells, and widely expressed in many organs including the skin [8]. Among the six family members, only FATP4 plays a crucial role in skin development and function. A mouse strain, wrinkle-free (wrfr), containing a spontaneous mutation that abrogates Fatp4 expression, and Fatp4-KO mice display features of lethal dermatology, such as tight, thick, shiny skin and a defective skin barrier, and die shortly after birth, symptoms that are strikingly similar to those of a human skin disease called restrictive dermopathy [3,4,5]. How Fatp4 contributes to the skin barrier function, however, is unknown.
A database search for sequences similar to human FATP members revealed homologues in various animals, including the simple multicellular animal, Caenorhabditis elegans, suggesting that FATP plays a fundamental role in animals [14]. Here we characterized two C. elegans FATP genes, acs-20 and acs-22, and revealed that C. elegans FATPs are involved in forming the cuticle barrier and can be substituted for by human Fatp4.

Results
C. elegans FATP Family Members, acs-20 and acs-22, Are Predominantly Expressed in the Hypodermis and Intestine, Respectively We identified and cloned two genes in the C. elegans genome, acs-20 (F28D1.9) and acs-22 (D1009.1a), both of which are highly homologous to the mammalian FATP family. Phylogenic tree analysis of worm and human FATP family proteins showed that worm ACS-20 and ACS-22 are closely related to human FATP1 and FATP4, and rather distantly related to the other FATP family members FATP2, FATP3, FATP5, and FATP6 (Fig. 1A).
To determine their expression patterns, we first performed expression analyses with transcriptional fusion genes. P acs-20 ::egfp was expressed predominantly in epithelial cells, such as Hyp7 and seam hypodermis (Fig.1D), while P acs-22 ::egfp was strongly expressed in the intestine, moderately expressed in the pharynx and in several sets of unidentified head neurons ( Fig.1B and D), and faintly expressed in the hypodermis (data not shown). Translational fusion P acs-20 ::acs-20::egfp was expressed in the vulval and anal muscles, in addition to Hyp7 and seam hypodermis (Fig. 1C). The P acs-20 ::acs-20::egfp transgene (Fig. 1D), which was expressed in a reticular pattern throughout the cytoplasm (Fig. 1D, ACS-20::EGFP), fully rescued the reduced body size of the acs-20;acs-22 double mutants (Fig. 1C), suggesting its functional expression in the endoplasmic reticulum (ER).

acs-20 Plays an Essential Role in Surface Barrier Function
First, to determine whether acs-20 and acs-22 mutants affect lipid metabolism, we examined fat storage using the lipophilic fluorescent dye, Nile Red. acs-20 mutant animals showed excessively high fluorescence intensity in the intestine ( Fig. 2A), while acs-22 mutant animals showed no obvious abnormal intensity (data not shown). Staining with another lipophilic fluorescent compound, C1-BODIPY-C12, produced similar results ( Fig. 2A). On the other hand, Sudan black staining, which is used for general lipid staining but requires the samples to be permeabilized, exhibited no significant difference between wildtype and acs-20 mutant animals ( Fig. 2A, right). Thus, we directly quantified the triacylglycerol and phospholipid contents. There was no significant difference between wild-type and acs-20 mutant animals (Fig. 2B, C). Based on these contradictory results, we considered the possibility that acs-20 mutant animals have a hyper-permeable surface barrier rather than increased lipid accumulation.
To examine the surface barrier function, living worms were incubated for 15 min in buffer containing a nuclei-staining dye Hoechst 33342, which does not usually permeate the worm cuticle barrier. The acs-20 mutant had increased Hoechst permeability compared to wild-type (Fig. 3A, B, and G). In contrast, the majority of acs-22 mutant animals did not have increased Hoechst permeability. The acs-20;acs-22 double mutant showed more intense staining than the acs-20 single mutant (Fig. 3B, C, and G), suggesting that acs-22 also contributes to the surface barrier function. The impaired surface barrier of acs-20 single mutants and acs-20;acs-22 double mutants was fully rescued by the P acs-20 ::acs-20::egfp transgene (Fig. 2D, F, and G). The expression of acs-20 or acs-22 with the dpy-7 promoter, which drives expression in hyp7 hypodermis restored the impaired barrier of acs-20 and acs-20;acs-22 mutants although not fully, indicating that acs-20 functions in hyp7 hypodermis for the surface barrier and acs-22 can substitute for acs-20 when acs-22 is expressed in hyp7 hypodermis with the dpy-7 promoter. These findings are similar to those in studies of Fatp4, which is expressed and functions in the skin and Fatp4-KO mice exhibit epidermal barrier defects [3,4,5].
To test the idea that acs-20 is a functional orthologue of mammalian Fatp4, we examined whether the expression of human Fatp4 cDNA rescues the defective surface barrier of acs-20 and acs-20;acs-22 mutants. The dpy-7 promoter-driven human FATP4 induced significant rescue activity ( Fig. 3E and G), and in rescued strains, the human FATP4::Venus fusion protein was expressed in a reticular pattern (Fig. 1D, middle lower panel).
To further analyze their barrier function, we established an assay to measure the resistance to hypotonicity. When living worms were placed in a drop of H 2 O and incubated for 20 min, wild-type animals actively moved throughout the assay (Fig. 3H upper panels and 3I). As expected, acs-20 animals were sensitive to hypotonicity: they gradually became rigid and some exhibited burst vulvae (Fig. 3H lower panels and 3I), probably due to the penetration of ambient H 2 O molecules into the worm coelom. The acs-20 transgene fully rescued the defect of acs-20 mutants (Fig. 3I). acs-20 mutants were also hypersensitive to levamisole, an antiparasitic agent (data not shown). These results demonstrate that acs-20 is required for barrier function, preventing the penetration of a variety of small molecules.

acs-20 Mutants Have Impaired Cuticle Structures
To determine the surface barrier structure and function of acs-20 mutants, we first visualized the cuticle structure of the acs-20 mutant with a col-19::gfp marker. COL-19::GFP was localized along the alae (white arrows in Fig. 4A) and annular furrows corresponding to the regular stripe pattern in the wild-type background. In the acs-20 mutant background, although COL-19::GFP was localized correctly, the alae were obscured (white arrows in Fig. 4B).
To directly examine the cuticle structure of acs-20 mutants, ultrathin sections were observed under transmission electron microscopy. In acs-20 mutants, the alae were incompletely extended compared to those of wild-type animals ( Fig. 4C and D, black arrows), and unidentified matrix that was rarely observed in wild-type accumulated within the cuticle layers ( Fig. 4E and F, arrowheads). In addition, a collagen-like matrix was observed under the alae (Fig. 4D, bold white arrows). Because seam cells are responsible for producing alae [15], we examined the cell fates of hypodermal cells in acs-20 mutants using a seam cell marker (scm::gfp), which localized in the nuclei of seam cells. There were no obvious defects (data not shown). We also examined hypodermal cell integrity using an ajm-1::gfp reporter, which is expressed in the apical junctions between epithelial cells. No obvious defect was observed in Hyp7 and the seam cell junction (Fig. 4G, H). Thus, acs-20 mutants exhibited several abnormalities in the cuticle structure, but not in the epidermal cell fate or cell integrity. Finally, we analyzed the incorporation of radio-labeled fatty acids into the lipid fraction of worms. Because VLCFAs are not detected in glycerolipids, but in sphingolipids, in C. elegans, as shown in the study of Gerdt et al. [16] and Menuz et al. [17], we analyzed the incorporation of radio-labeled VLCFA into the sphingomyelin fraction of worms. In this system, free 14 C-fatty acids are converted to [ 14 C]acyl-CoAs by endogenous acyl-CoA synthetases and then incorporated into sphingomyelin. The incorporation of C26:0 fatty acids into sphingomyelin was reduced in acs-20 and acs-22 mutant animals (Fig. 5). On the other hand, the incorporation of C16:0 and C20:4 fatty acids was not affected in acs-20 and acs-22 mutants (Fig. 5). These data indicate that acs-20 and acs-22 are involved in incorporating VLCFA, but not other fatty acids, into the sphingomyelin in vivo.

Discussion
In the present study, we characterized two C. elegans FATP genes, revealing that acs-20 mutants are defective in the cuticle barrier that prevents penetration of small molecules. acs-22 mutants rarely exhibited a barrier defect, but the acs-20;acs-22 double mutant showed severely disrupted barrier function, suggesting a redundant function of the genes. We also showed that the barrier defects of acs-20 and acs-20;acs-22 mutants are rescued by acs-20, acs-22, and human Fatp4 transgenes. These data indicate that acs-20 and acs-22 are functional homologues of mammalian Fatp4.
In the rescued lines, we observed the reticular expression pattern of human FATP4::Venus protein throughout the cytoplasm, which is compatible with the finding that mammalian FATP4 is localized in the ER [18]. Although several initial studies suggested that FATP4 localizes to the plasma membrane of adipocytes and intestinal epithelial cells and facilitates fatty acid  uptake [19,20], other studies suggest that the protein localizes to the intracellular membrane and functions as an acyl-CoA synthetase [18,21]. In the present study, we observed the functional expression of C. elegans FATPs and human FATP4 in an ER-like pattern rather than a plasma membranous pattern, indicating that FATP localization in the ER is essential for surface barrier function. These findings further suggest that ACS-20, ACS-22, and FATP4 act as acyl-CoA synthetases, but not as fatty acid transporters, for surface barrier function.
The acs-20 mutants exhibited abnormalities in the cuticle structure. The C. elegans cuticle is a highly structured extracellular matrix (ECM) that is composed predominantly of cross-linked collagens, additional insoluble proteins termed cuticlins, and associated glycoproteins and lipids. ECM is synthesized and secreted by underlying epithelial cells, the hypodermis, and seam cells [22]. Although the mammalian skin appears to be largely different from the worm cuticle, some fundamental similarities exist, e.g., in the mammalian stratum corneum, the lipids discharged from the underlying epithelial cells, the keratinocytes, into the intercellular space and are cross-linked with a variety of proteins [23]. Given that Fatp4-deficient mice and acs-20 mutant worms have surprisingly similar defects in the skin and cuticle barrier, respectively, the defects are likely to be caused by common mechanisms between mammals and worms.
We propose two possible causes of the cuticle barrier defect: Given that acs-20 mutants displayed an unusual accumulation of collagen-like matrix under the alae, one possibility is an alteration of the secretion process of ECM from epithelial cells that results in the barrier defect. Because sphingomyelin is essential in lipid rafts of the plasma membrane [24] and our study showed reduced incorporation of C26:0, but not C16:0 or C20:4, into the sphingomyelin in acs-20 and acs-22 mutants, VLCFA-containing sphingomyelin in the epithelial cell membrane may be involved in regulating ECM secretion. Another possibility is that the altered lipid composition in the cuticle affects the cuticle barrier. Ceramide, which is synthesized from sphingomyelin and secreted from epithelial cells, might be involved in cuticle barrier function. Because VLCFA is highly hydrophobic, VLCFA-containing ceramide is likely to contribute to form a physicochemical barrier in the cuticle ECM. In this case, the unusual secretion of collagenlike matrix observed in acs-20 mutants may compensate for the impaired lipid barrier.
In addition to animals, several lines of evidence indicate that plant long-chain fatty acid acyl-CoA synthetases (LACSs), LACS1 and LACS2, are involved in normal cuticle development and barrier function [25,26]. The FATP family in animals and the LACS family in plants may have a common role in skin or cuticle development/barrier, although plant LACSs show very low overall similarity with mammal or C. elegans FATPs, except for a small stretch of amino acids, which is highly conserved AMPbinding motif in the ACS family. In addition, Tang et al. demonstrated that LACS2 mutations enhance susceptibility to Pseudomonas syringae but confer resistance to Botrytis cinerea [27], suggesting that FATP mutants might have an altered susceptibility to pathogens. Further studies are needed to elucidate the involvement of FATP in C. elegans defense mechanisms.
acs-20 mutants may be valuable for use in drug screening. Recently, C. elegans attracted the attention of drug screening and chemical genomics researchers [28,29]. One problem that remains to be resolved for this type of drug screening, however, is the low accessibility of compounds because the worm cuticle acts as a barrier to small molecule diffusion [30]. acs-20 mutant animals have an impaired cuticle barrier that results in an increase in small molecule accessibility. Moreover, basic physical parameters such  as viability and locomotion are rarely affected in acs-20 mutants, suggesting its utility in drug screens targeting neurologic disorders and many other diseases.
In summary, ACS-20 and ACS-22 are involved in the surface barrier in C. elegans. Our study is the first to show a biologic significance of FATPs in C. elegans and reveals that human FATP can substitute for worm FATP4. We propose that the C. elegans FATP mutant is a simple animal model for studying fundamental molecular mechanisms underlying the mammalian skin barrier. Future studies using the C. elegans model should facilitate further biologic and pharmacologic studies and may lead to therapeutic approaches to human skin diseases.

Measurement of Body Length
We used 1-day-old adult hermaphrodites (synchronized L4stage larvae were cultured for 24 h at 20uC). Animals were transferred onto a glass slide and paralyzed. Body lengths of animals were measured with a dissecting microscope and a micrometer.

Nile Red and C1-BODIPY-C12 Staining
Staining experiments were performed as previously described [35]. Nile Red (Molecular Probes, Eugene, OR) was dissolved in acetone to produce a 0.5-mg/ml stock solution and stored at 4uC. C1-BODIPY-C12 (Molecular Probes) was dissolved in dimethylsulfoxide to produce a 5-mM stock solution and stored at 220uC. Stock solutions were freshly diluted in 1x PBS to 1 mg/ml (Nile Red) or 1 mM (C1-BODIPY-C12) and 150 ml of the diluted solution was poured onto Nematode Growth Media plates (6-cm diameter, 8 ml agar) seeded with E. coli OP50. Plates were allowed to air-dry before use. Starved worms were cultured on the plates for 3 days before observation.

Lipid Content Analyses
Phospholipid and triglyceride content: Lipids of synchronized young adult worms were extracted by the method of Bligh and Dyer [36]. The phosphorus content of the total lipids was determined by the method of Bartlett [37]. The triglyceride content was measured using an enzymatic kit (Wako Triglyceride E-test, Wako Pure Chemical Ltd., Osaka, Japan).

Hoechst Staining, Sensitivity to Hypo-Osmolality
Hoechst staining was performed as described previously [38] except that we used membrane-permeable and cuticle-impermeable Hoechst 33342 (formula weight, 561.93) instead of membrane-and cuticle-impermeable Hoechst 33258 (formula weight, 533.88) to evaluate the cuticle barrier rather than the plasma membrane barrier. Briefly, worms were incubated in M9 buffer containing 1 mg/ml Hoechst 33342 (Wako Pure Chemical Ltd., Osaka, Japan) at room temperature for 15 min, followed by washing several times with M9 buffer. Fluorescence images were captured with the identical parameters (lens and magnifier used, filters, exposure time, resolution) and with an 8-bit pixel depth (256 shades of grey). Fluorescence intensity was examined by processing fluorescence micrographs with ImageJ (Rasband, W.S., US National Institutes of Health, Bethesda, MD, http://rsb.info. nih.gov/ij/). To analyze the fluorescence of stained nuclei without including the auto-fluorescence of the intestine, the tail area was enclosed using a freehand selection tool, excluding the intestinal area. The average pixel intensity was determined. Sensitivity to hypo-osmolality was assayed by placing young adult worms into 20-ml drops of distilled H 2 O and measuring the time until they became rigid. All experiments were performed at 20uC.

Microscopy
Differential interference contrast and fluorescence images were obtained using a BX51 microscope equipped with a DP30BW CCD camera (Olympus Optical Co., Ltd, Tokyo, Japan).

Transmission Electron Microscopy
Transmission electron microscopy was performed by the Hanaichi Ultrastructure Research Institute Co. (Okazaki, Japan). Wild-type or mutant larvae were fixed with 4% paraformaldehyde and 1% glutaraldehyde, 1% paraformaldehyde in 100 mM cacodylate buffer for 1 h each at room temperature, followed by further fixation with 2% paraformaldehyde and 2% glutaralde-hyde in 100 mM cacodylate buffer for 1 day and an overnight wash in 100 mM cacodylate buffer at 4uC. They were then postfixed for 24 h with 2% osmium tetroxide in 100 mM cacodylate buffer, followed by dehydration and infiltration with a resin Q651 (Mitsui Chemicals, Tokyo, Japan). Ultrathin sections of the surface area were analyzed using an electron microscope (JOEL JEM-200EX, Akishima, Japan).

In Vivo Incorporation of Exogenous Fatty Acids
Synchronized first-stage larvae (approximately 6000 animals) were cultured with 1 mCi of 14 C-labeled palmitic acid, arachidonic acid, and cerotic acid (C16:0, C20:4, and C26:0, respectively) on C. elegans growth medium plates at 20uC until they reached the young adult stage. Lipids were extracted by the previously described methods [36] and separated by one-dimensional thin layer chromotography on silica gel 60 plates (Merck, NJ) in chloroform/methylacetate/1-propanol/methanol/0.25% KCl (25/25/25/10/9, v/v). Incorporation of 14 C-labeled fatty acids into sphingomyelin was calculated based on the percentages of the whole radioactivity incorporated into the total lipids.