HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 276, Issue 28, 26577-26588, July 13, 2001

This Article Abstract Full Text (PDF) All Versions of this Article: 276/28/26577    most recent M102818200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Mao, C. Articles by Obeid, L. M. Search for Related Content PubMed PubMed Citation Articles by Mao, C. Articles by Obeid, L. M. Social Bookmarking                      What's this?

Cloning and Characterization of a Novel Human Alkaline Ceramidase

A MAMMALIAN ENZYME THAT HYDROLYZES PHYTOCERAMIDE^*

Cungui Mao^§, Ruijuan Xu, Zdzislaw M. Szulc^¶, Alicja Bielawska^¶, Sehamuddin H. Galadari, and Lina M. Obeid^**

From the Departments of  Medicine and ^¶ Biochemistry and the ^** Division of General Internal Medicine, Ralph H. Johnson Veterans Affairs Hospital, Medical University of South Carolina, Charleston, South Carolina 29425, and the  Department of Biochemistry, Faculty of Medicine and Health Sciences, United Arab Emirates University, Al Ain, Abu Dhabi, United Arab Emirates

Received for publication, March 29, 2001, and in revised form, May 16, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Ceramidases are enzymes involved in regulating cellular levels of ceramides, sphingoid bases, and their phosphates. Based on sequence homology to the yeast alkaline ceramidases YPC1p (Mao, C., Xu, R., Bielawska, A., and Obeid, L. M. (2000) J. Biol. Chem. 275, 6876-6884) and YDC1p (Mao, C., Xu, R., Bielawska, A., Szulc, Z. M., and Obeid, L. M. (2000) J. Biol Chem. 275, 31369-31378), we report the identification and cloning of a cDNA encoding for a novel human alkaline ceramidase (aPHC) that hydrolyzes phytoceramide selectively. Northern blot analysis showed that aPHC was ubiquitously expressed, with the highest expression in placenta. Green fluorescent protein tagging showed that it was localized in both the Golgi apparatus and endoplasmic reticulum. Overexpression of aPHC in mammalian cells elevated in vitro ceramidase activity toward N-4-nitrobenz-2-oxa-1,3-diazole-C₁₂-phytoceramide. Its expression in a yeast mutant strain devoid of any ceramidase activity restored the ceramidase activity and caused an increase in the hydrolysis of phytoceramide in yeast cells, thus leading to the decreased biosynthesis of sphingolipids. These data collectively suggest that, similar to the yeast phytoceramidase YPC1p, aPHC has phytoceramidase activity both in vitro and in cells; hence, it is a functional homolog of the yeast phytoceramidase YPC1p. However, in contrast to YPC1p, aPHC exhibited no reverse activity of ceramidase either in vitro or in cells. Biochemical characterization showed that aPHC had a pH optimum of 9.5, was activated by Ca²⁺, but was inhibited by Zn²⁺ and sphingosine. Substrate specificity showed that aPHC hydrolyzed phytoceramide preferentially. Together, these data demonstrate that aPHC is a novel human alkaline phytoceramidase, the first mammalian alkaline ceramidase to be identified as being specific for the hydrolysis of phytoceramide.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Ceramide and its intermediate breakdown product sphingosine have been shown to mediate many cellular events including growth arrest, stress responses, and apoptosis (for review, see Refs. 1-5). On the other hand, the subsequent product sphingosine-1-P, which is generated from sphingosine through phosphorylation by sphingosine kinases, promotes proliferation and migration of endothelial cells as a ligand for the Edg family of G protein-coupled receptors (6-8) or as an intracellular signaling molecule (6). In yeast cells phytosphingosine-1-P suppresses growth (9) and enhances heat stress tolerance (10). The diversity of cellular actions by these lipids implies that a balance among these different lipids may determine the physiological responses of cells. The balance involves many metabolic enzymes and their regulators, among which are ceramidases.

Ceramidases hydrolyze the amide linkage of ceramides to generate free fatty acids and sphingoid bases (11, 12). There are three types of ceramidases described to date (11). These are classified as acid, neutral, and alkaline ceramidases according to their pH optimum of enzymatic activity. The murine acid ceramidase was the first ceramidase to be cloned (13). It is localized in the lysosome and is mainly responsible for the catabolism of ceramide. Dysfunction of this enzyme because of a genetic defect leads to a sphingolipidosis disease called Farber disease (13). The neutral ceramidases have been purified from rat brain (14) and mouse liver (15), and recently they were cloned from Pseudomonas (16), mycobacterium (16), mouse (17), and human (18). These ceramidases share significant homology, and this homology extends to putative proteins deduced from expressed sequence tag (EST)¹ sequences of Dictyostelium discoideum and Arabidopsis thaliana (16, 18). These ceramidases have a broad pH optimum ranging from 5 to 9 for their activity (17, 18). They appear to hydrolyze unsaturated ceramide preferentially, saturated ceramide (dihydroceramide) slightly, and hardly hydrolyze phytoceramide (17). The Pseudomonas, mouse, and human neutral ceramidases have a reverse ceramidase activity of catalyzing the formation of ceramide from sphingosine and a fatty acid (16, 17, 19). El Bawab et al. (18) have shown previously that the human neutral ceramidase is localized in the mitochondria.

Recently, we cloned and characterized two Saccharomyces cerevisiae ceramidases YPC1p (20) and YDC1p (21). Both had an alkaline pH optimum of 9.5-10 and are classified as alkaline ceramidases. YPC1p showed a preference for phytoceramide as a substrate, whereas YDC1p showed a preference for dihydroceramide. Neither hydrolyzed the unsaturated mammalian type ceramide. In addition, YPC1p showed considerable reverse ceramidase activity, catalyzing the formation of phytoceramide and dihydroceramide from a free fatty acid and sphingoids, whereas YDC1p had very minor reverse activity (20, 21). We also demonstrated that these two ceramidases were involved in the regulation of sphingolipid metabolism in S. cerevisiae. Deletion or overexpression of these ceramidases significantly altered the turnover of many sphingolipids including complex sphingolipids, free sphingoid bases, and their phosphates (20, 21). Moreover, the critical role of these ceramidases in regulating sphingolipid metabolism is reflected by their localization to the endoplasmic reticulum (ER) where ceramides are generated and the synthetic enzymes of sphingolipids reside. Deletion of YDC1p increased cell tolerance to heat stress, which is in agreement with the established roles of sphingolipids in heat stress responses (10, 22-24).

Because of the roles of these ceramidases in sphingolipid metabolism and their regulation of yeast heat stress, we elected to identify and characterize their mammalian counterparts. In this study, we report on the cloning and characterization of a novel human alkaline ceramidase based on sequence similarity. This new ceramidase shares many biochemical properties with the yeast ceramidases, although it lacks the reverse activity. It has an alkaline pH optimum and is activated by Ca²⁺. Interestingly, this enzyme hydrolyzes phytoceramide selectively. It shares neither sequence similarity nor substrate specificity with other mammalian ceramidases; therefore it represents a novel mammalian ceramidase, the first mammalian enzyme to be identified as being specific for hydrolysis of phytoceramide.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Lipid Preparations-- D-erythro-Sphingosine, D-erythro-dihydrosphingosine, and D-ribo-phytosphingosine were purchased from Avanti Polar Lipids, Inc. D-erythro-C₁₂-NBD-4,5-dihydroceramide, D-ribo-C₁₂-NBD-phytoceramide, and D-erythro-C₁₂-NBD-ceramide were synthesized in our laboratory following the methods described by Marchesini et al. (25).

Cell Lines, Media, and Tissue Culture Reagents-- HEK293 and COS-1 cell lines purchased from ATCC were respectively cultured in minimum essential medium and Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 100 units/ml penicillin G, and 100 µg/ml streptomycin at 37 °C in a humidified atmosphere of 5% CO₂ and 95% air. Minimum essential medium, Dulbecco's modified Eagle's medium, fetal bovine serum, trypsin-EDTA, PBS, and penicillin/streptomycin were purchased from Life Technologies, Inc.

Transfection of HEK293 and COS-1 Cells-- HEK293 and COS-1 cells were transfected by LipofectAMINE (Life Technologies, Inc.) according to the manufacturer's instructions.

Cloning of the cDNA-- Using the advanced BLAST program, the nonredundant data bases and the EST data bases in GenBank were searched for homologous sequences of the yeast alkaline ceramidases YPC1p and YDC1p. No significantly homologous sequences with known function were found. However, two mouse EST sequences (AA271224 and AI607051) and three human EST sequences (N57268, BF086933, and AI375670) were found to be highly homologous to the yeast YPC1p and YDC1p at the protein level. The three overlapping human EST sequences were assembled into a longer sequence using the software MacVector. The assembled sequence was used further to search the human EST data base, from which the longest sequence with a length of 1.8 kb was assembled. This 1.8-kb sequence contains a putative open reading frame encoding for a 200-amino acid polypeptide, which has 38% identity to the yeast ceramidases YPC1p and YDC1p. Several polyadenylation signals downstream of the stop codon and in-frame ATG codons were also found; however, there was no in-frame stop codon upstream of any of these ATG codons, suggesting that the translation initiation site ATG might not be included. To clone the full-length cDNA, 5'-RACE was performed using RACE ready marathon kidney cDNA (CLONTECH) as template, the gene-specific primer 5'-AAACTGGAGCGAAAGGAGGAGGGG-3' and the adaptor primer AP1, and Advantage-GC2 DNA polymerase (CLONTECH, Inc.) according to the manufacturer's instructions. The first round PCR products were diluted 50-fold and used as the templates for the second round PCR using the gene-specific primer 5'-GCCATGAGCTAAACACCCAGCAGGCCCTAC-3' and the adaptor primer AP2, thus resulting in a major 1.0-kb product that was cloned into a TA vector pCRII (Invitrogen, Inc.) and sequenced. Sequencing showed that 5'-RACE extended 87 nucleotides from the existing sequence. This extended sequence contains an in-frame ATG codon preceded by an in-frame stop codon. The flanking sequence of this ATG codon matches the Kozak consensus sequence, thus it is very likely to be the translation initiation codon. To obtain the 3'-end of the cDNA, 3'-RACE was performed using primer 5'-CAAGCACCTACCATAGACCTGGC-3' and the adaptor primer AP1, which resulted in a single 2.7-kb PCR product. Sequencing showed that the 3'-RACE product was extended from the previously assembled sequence. The sequences of 5'- and 3'-RACE products were assembled into the full-length cDNA whose coding region was identified by the software MacVector. The full-length cDNA was amplified from a kidney cDNA library with primers 5'-TCGCCAGCCTAACCCGGCAC-3' and 5'-TGGCTTCCCTTTCTGCTTCC-3' using the Advantage-2 DNA polymerase, cloned into the vector pCRII, and sequenced.

In Vitro Translation-- The human full-length cDNA was cloned downstream of the T7 promoter of the vector pCRII in the sense orientation. The resulting construct pCRII-aPHC was amplified in Escherichia coli and purified by a Midi-prep kit (Qiagen, Inc.). The trace RNase in the plasmid preparation was removed by phenol-chloroform extraction followed by chloroform. The in vitro translation was performed with the empty vector pCRII or the plasmid pCRII-aPHC (0.5 µg) as a template using a Single Tube Protein System 3 kit (Novagen, Inc.) in the presence of [³⁵S]methionine (PerkinElmer Life Sciences). Translation products were resolved by SDS-PAGE and detected by autoradiography according to the manufacturer's instructions. The molecular mass of the products was estimated according to prestained protein standards (Life Technologies, Inc.).

Northern Blot Analysis-- A multiple human tissue mRNA blot (CLONTECH, Inc.) was hybridized with DNA probes radiolabeled by [³²P]dCTP using a random priming kit (Amersham Pharmacia Biotech). To determine aPHC mRNA, the DNA probe spanning the entire coding region was generated by PCR. The actin mRNA probe was provided with the blot by the manufacturer and used for the normalization of cDNA from different tissues. Hybridization was performed according to the manufacturer's instructions.

RT-PCR Analysis-- Total RNA was isolated from HEK293 or COS-1 cells using the RNeasy mini kit (Qiagen, Inc.) and was subjected to RT-PCR analysis using a SuperScript One-Step RT-PCR system (Life Technologies, Inc.) according to the manufacturer's instructions. PCR was performed according to the manufacturer's instructions on human skin cDNA (Invitrogen, Inc) or a human multiple tissue cDNA panel (CLONTECH, Inc.) in which concentrations of cDNA reverse transcribed from mRNA of different tissues or organs were normalized to the mRNA expression levels of at least four different housekeeping genes. The PCR primers for determining aPHC mRNA were 5'-ATGTTCGGTGCAATTCAGAGT-3' (forward) and 5'-CCTTCTTTCGAAAGTTCCTCAGTG-3' (reverse), which generated a 428-base pair product. PCR products were verified by sequencing.

Expression of the Human Alkaline Ceramidase in Yeast-- To generate a construct (pYES2-FLAG-aPHC) expressing the FLAG-tagged aPHC under the control of the GAL1 promoter, the coding sequence of aPHC was amplified by PCR using the template pCRII-aPHC and the primers 5'-GAAGATCTATGGCTCCGGCCGCGGACCGAGAG-3' and 5'-CGGAATTCTCAATGCTTCCTGAGAGGCTCAAAC-3'. The PCR product was digested with the restriction enzymes BglII and EcoRI and cloned into the BamHI and EcoRI sites of pYES2-FLAG, which was constructed from pYES2 (Invitrogen, Inc.) by inserting the FLAG tag coding sequence (underlined) plus the Kozak sequence (ACC) and a start codon (ATG) 5'-ACCATGGACTACAAGGACGACGATGATAAG-3' into the HindIII and BamHI sites. The construct pYES2-FLAG-aPHC or the empty vector pYES2-FLAG was transformed into the yeast strain ypc1ydc1as described (21). Expression of the FLAG-tagged aPHC was induced by 2% galactose in synthetic complete medium (SC-ura) with the omission of uracil as described. Microsomes were prepared from yeast cells as described (21).

Expression of the FLAG- or GFP-tagged Human Alkaline Ceramidase in Mammalian Cells-- To generate a construct pcDNA3.1(+) -FLAG-aPHC expressing the FLAG-tagged aPHC in mammalian cells, the plasmid pYES2-FLAG-aPHC was digested with the restriction enzymes HindIII and EcoRI. The HindIII-EcoRI fragment containing the coding sequence for the FLAG-aPHC fusion protein was isolated and cloned into the HindIII and EcoRI sites of a mammalian expression vector pcDNA3.1(+). The aPHC coding sequence was also cloned into the BglII and EcoRI sites of pEGFP-C1 to create a construct pEGFP-aPHC expressing the GFP-tagged aPHC. Both constructs were sequenced and transfected into HEK293 or COS-1 cells.

Fluorescent and Confocal Microscopy-- HEK293 or COS-1 cells were transfected with pEGFP-aPHC. Thirty-six h after transfection, the cells were fixed by 3.7% paraformaldehyde in PBS for 10 min, washed twice by PBS, permeabilized for 10 min with 0.05% Triton X-100 in PBS, blocked for 20 min with 2% bovine serum albumin in PBS, and probed with a primary antibody anti-giantin (1:1000, Babco, Richmond, CA) followed by a secondary antibody donkey anti-rabbit IgG conjugated with rhodamine (1:200, Jackson ImmunoResearch Laboratories). The cells were examined under a fluorescent microscope (Zeiss) or a confocal laser scanning microscope (Olympus IX70) operated with an Ultraview software (PerkinElmer Life Sciences) that was set at the spinning disc mode.

Immunoprecipitation and Western Blot Analysis-- Cells (in a 10-cm culture dish) transfected with pcDNA 3.1(+) or pcDNA 3.1(+)-FLAG-aPHC were lysed in a lysis buffer (50 mM Tris-HCl, pH 7.4, containing 150 mM NaCl, 0.5% Nonidet P-40, 0.1 mM phenylmethylsulfonyl fluoride, and protease inhibitor mixture (1 tablet/50 ml, Roche)) after being washed twice with PBS. Cell lysates were centrifuged at 10,000 × g for 5 min, and the resulting supernatant (containing ~1 mg of protein) was incubated with 20 µg anti-FLAG antibody M2 (Sigma) for 1 h followed by 50 µl of protein A/G agarose beads (Sigma) for 2 additional h. The agarose beads were suspended in 50 µl of SDS-PAGE sample buffer, boiled for 4 min, and centrifuged for 5 min at 10,000 × g after being washed four times with the lysis buffer. The immunoprecipitated proteins were resolved by SDS-PAGE and transferred to a nitrocellulose membrane. The FLAG-tagged aPHC was detected by the anti-FLAG antibody M2 (Sigma) using an ECL Plus detection kit (Amersham Pharmacia Biotech) according to the manufacturer's instructions.

Assay of Ceramidase Activity and Its Reverse Activity-- Ceramidase activity was assayed by using fluorescent NBD-C₁₂-phytoceramide, ceramide, or dihydroceramide as substrate. NBD-ceramides (8 nmol) were dissolved in 20 µl of assay buffer A (25 mM glycine-NaOH, pH 9.4, containing 0.3% Nonidet P-40) and added to 20 µl of yeast microsomes (~50 µg of proteins) or mammalian postnuclear cell lysates (~15 µg of proteins) in a lysis buffer (25 mM Tris-HCl, pH 7.4, containing 5 mM CaCl₂ and the protease inhibitor mixture). For determining the pH optimum of enzymatic activity, the yeast microsomes were suspended in the buffer A but with 1 mM Tris-HCl instead. Enzymic reactions were incubated at 37 °C for 90 min and were stopped by boiling for 5 min. After the reactions were dried under a SpeedVac evaporator, lipids were dissolved in 35 µl of chloroform/methanol (2:1), and 25 µl was spotted onto a TLC plate. The product NBD-C₁₂-fatty acid was separated from the substrates by chloroform:methanol (90:30:0.5) and 25% ammonium hydroxide. The TLC plate was scanned by a PhosphorImager system (Storm 860) set at the fluorescence mode. NBD-fatty acid was identified by a standard and was analyzed by an Imagequant software. Both reaction time and protein concentrations were in the linear range of the assay. Reverse ceramidase activity was measured using [³H]palmitic acid and sphingoid bases as substrates as described (21).

Determination of Protein Concentrations-- Protein concentrations were determined using the BCA (Pierce) or Bradford (Bio-Rad) protein assay kit according to the manufacturer's instructions.

Labeling of Sphingolipids-- Yeast cells were labeled with [³H]palmitic acid or [³H]inositol, and lipids were extracted and analyzed by TLC as described (20).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cloning of a Human cDNA That Encodes a Protein Homologous to Yeast Alkaline Ceramidases-- A human kidney cDNA, homologous to the yeast genes YPC1 (20) and YDC1 (21) at the protein level, was cloned by a combination of 5'-RACE and 3'-RACE as described under "Experimental Procedures." The cDNA sequence was deposited in the GenBank with the accession number AF214454. The coding sequence of the cDNA was predicted using a MacVector program and is depicted in Fig. 1A. The cDNA encodes a putative protein of 253 amino acids with a calculated molecular mass of 31.6 kDa (Fig. 1A). The deduced amino acid sequence of this putative protein was aligned to the yeast ceramidases and to putative proteins of A. thaliana and Caenorhabditis elegans. It exhibited 38% identity to the yeast ceramidases YPC1p and YDC1p and 28% identity to the putative proteins of A. thaliana and C. elegans (Fig. 1B). These proteins share several conserved regions, suggesting that these regions may be important for catalysis. Hydropathy plot showed that, similar to the yeast ceramidases, the putative human protein is a very hydrophobic protein (Fig. 1C). Five putative transmembrane domains were predicted by the pSORTII program (Fig. 1A). Based on sequence similarity, this putative protein is postulated to be a ceramidase (aPHC).

View larger version (113K): [in this window] [in a new window]   Fig. 1.   Coding sequence and deduced protein sequence of the human aPHC. Panel A, the cDNA of the human aPHC was sequenced to completion on an Applied Biosystems sequencer. The coding region (plain letters) was predicted, and the encoded protein sequence (bold letters) was deduced by a MacVector software. Potential transmembrane domains (underlined) and a Golgi to ER retrieval-like sequence (double underlined) were predicted using the pSORTII program. Panel B, protein sequences were aligned using the ClustalW method. The alignment includes the deduced protein sequences of the yeast ceramidases YPC1p and YDC1p, the putative proteins of A. thaliana and C. elegans, and the human homolog aPHC. Identical amino acids are shaded, and conserved regions are boxed. Panel C, the hydropathy was plotted by the Kyte/Doolittle method.

To verify that the cDNA does encode the predicted protein, we performed an in vitro translation with the full-length cDNA as a template in the presence of [³⁵S]methionine. The translated product labeled by [³⁵S]methionine was resolved by SDS-PAGE and detected by autoradiography. A single polypeptide band with a molecular mass of ~30 kDa was translated from the cDNA-containing plasmid but not from the empty vector (Fig. 2), indicating that the cDNA isolated from human kidney indeed encodes the protein (aPHC) as predicted.

View larger version (57K): [in this window] [in a new window]   Fig. 2.   In vitro translation of aPHC. In vitro translation was performed with the empty vector pCRII (Vec) or the aPHC cDNA-containing plasmid pCRII-aPHC (aPHC) as a template as described under "Experimental Procedures." The translated proteins were resolved on 12% SDS-PAGE and detected by autoradiography. The in vitro translated aPHC is marked with the arrow. MM, molecular mass.

Human Ceramidase (aPHC) Transcripts Are Expressed Ubiquitously-- To determine the tissue distribution of the aPHC mRNA, we performed Northern blot analysis on mRNA from different tissues using the entire coding region as a probe. Fig. 3A shows that three transcripts with respective sizes of 3.4, 4.4, and 7.6 kb are expressed in most tissues or organs examined. All three transcripts were expressed most abundantly in placenta and least in skeletal muscle. The 7.6-kb transcript was also expressed considerably in heart, brain, lung, kidney, and pancreas. The 3.5- and 4.3-kb transcripts showed tissue distribution patterns similar to those of the 7.6-kb transcript but with a lesser abundance. The cloned cDNA is closest to the 3.4-kb transcript in size. Whether the other two transcripts are alternatively spliced products or alternative polyadenylation isoforms awaits further investigation. To investigate total mRNA levels of aPHC in the above organs, we performed PCR on a cDNA panel (CLONTECH, Inc.) reverse transcribed from the mRNA of different organs. A 428-base pair PCR product was generated by using the primers within the coding sequence. Fig. 3B shows that placenta had the highest expression, and skeletal muscle had the lowest. These data are consistent with those seen in the Northern blot analysis.

View larger version (67K): [in this window] [in a new window]   Fig. 3.   Tissue distributions of aPHC mRNA. Panel A, a multiple human tissue mRNA blot (CLONTECH, Inc.) was hybridized with a radioactive DNA probe for aPHC or actin mRNA as described under "Experimental Procedures." The blot was exposed to an x-ray film for 2 days (aPHC probe) or overnight (actin probe) at 70 °C. Upper lanes, aPHC mRNA; lower lanes, actin mRNA. The sizes of the three species of mRNA were predicted according to standards and are indicated on the right. Kb, kilobases. Panel B, RT-PCR for aPHC mRNA was performed on a multiple human tissue cDNA panel (left lanes) or on human skin cDNA (right lane) as described under "Experimental Procedures."

aPHC Is a Golgi/ER Protein-- To understand better the physiologic functions of aPHC, its cellular localization was studied. A Golgi to ER retrieval-like sequence (LRKH) was found at its the carboxyl terminus, suggesting that similar to its yeast homologs YPC1p and YDC1p, it is localized in the ER. To verify this, cellular localization of aPHC was investigated by GFP tagging. A construct pEGFP-aPHC expressing the GFP-aPHC fusion protein (with GFP tagged to the amino terminus of aPHC) was transfected into HEK293 or COS-1 cells as described under "Experimental Procedures." The GFP-aPHC fusion protein was detected in the 100,000 × g membrane pellet, whereas GFP was detected in the supernatant as revealed by Western blot analysis (Fig. 4A). The GFP-aPHC fusion protein was expressed at very low levels in HEK293 cells, which prevented us from observing fluorescence in the cells (data not shown). However, COS-1 cells expressed sufficient amounts of the GFP-aPHC fusion protein. The cells expressing GFP or the GFP-aPHC fusion protein were fixed, immunostained by the anti-giantin followed by the donkey anti-rabbit IgG conjugated with rhodamine, and scanned by a confocal laser scanning microscope. The GFP-aPHC fusion protein exhibited a perinuclear reticulum network or crescent-shaped fluorescence pattern (Fig. 4B). The crescent-shaped pattern was overlapped with that of giantin, a Golgi marker, as seen in the merged image. These results suggest that the GFP-aPHC has a Golgi/ER dual localization.

View larger version (35K): [in this window] [in a new window]   Fig. 4.   Cellular localization of aPHC. aPHC was tagged with GFP as described under "Experimental Procedures." COS-1 cells expressing GFP or GFP-tagged aPHC were homogenized by passing through an insulin syringe several times. Cell lysates were centrifuged at 1,000 × g for 5 min to remove nuclei, and the resulting supernatants were centrifuged further at 100,000 × g for 45 min to sediment all membranes. 30 µg of proteins from the final supernatant (S) and membrane (P) fraction were subjected to Western blot analysis for the expression of GFP or the GFP-tagged aPHC (panel A). COS-1 cells expressing GFP or the GFP-tagged aPHC were fixed, immunostained by anti-giantin (a Golgi marker), and examined under a confocal laser scanning microscope as described under "Experimental Procedures" (panel B).

aPHC Encodes a Ceramidase Activity-- To study whether aPHC has ceramidase activity similar to that of YPC1p or YDC1p, the coding sequence was cloned in-frame with an epitope tag FLAG into the mammalian expression vector pcDNA3.1 to generate the construct pcDNA3.1-FLAG-aPHC, which expresses the FLAG-tagged aPHC. The construct and the vector were transfected into HEK293 cells as described under "Experimental Procedures." Expression of the FLAG-tagged aPHC was verified by immunoprecipitation with the anti-FLAG antibody followed by Western blot analysis (Fig. 5A). It should be noted that expression levels of the FLAG-tagged aPHC in HEK293 cells were extremely low such that 2 mg of total protein from the aPHC-transfected cells was required to detect the FLAG-tagged protein by immunoprecipitation. Postnuclear cell lysates were prepared from cells expressing the FLAG-aPHC or containing the empty vector and assayed for ceramidase activity using NBD-C₁₂-phytoceramide, ceramide, or dihydroceramide as substrates. Fig. 5B demonstrates that the cells transfected with pcDNA3.1-FLAG-aPHC had 2-, 1.3-, and 1.3-fold increases in ceramidase activity toward NBD-C₁₂-phytoceramide, ceramide, and dihydroceramide, respectively, compared with those transfected with the empty vector. These results suggest that the human homolog (aPHC) encodes a ceramidase activity and hydrolyzes phytoceramide most efficiently.

View larger version (50K): [in this window] [in a new window]   Fig. 5.   Elevation of ceramidase activity by overexpression of aPHC in HEK293 cells. The FLAG-tagged aPHC expressed in HEK293 cells was immunoprecipitated using anti-FLAG antibody and was detected by Western blot analysis (panel A). MM, molecular mass. Postnuclear lysate prepared from HEK293 cells transfected with an empty vector pcDNA3.1-FLAG (Vec) or pcDNA3.1-FLAG-aPHC (aPHC) was assayed for ceramidase activity on NBD-C12-ceramide (CER), dihydroceramide (DHC), or phytoceramide (PHC) (panel B). Data represent the mean ± S.D. of three independent experiments.

aPHC Is a Bona Fide Ceramidase-- Low expression of the FLAG-tagged aPHC and the high basal ceramidase activity in mammalian cells presented obstacles for further characterization of the enzyme. We therefore took advantage of a yeast mutant (ypc1ydc1) that had undetectable endogenous ceramidase activity because of the knockout of the two yeast ceramidases YPC1p and YDC1p (21). To express aPHC in the ypc1ydc1mutant, the construct pYES2-FLAG-aPHC, which expresses FLAG-aPHC under the control of the GAL1 promoter, or the vector pYES2-FLAG, was transformed into the ypc1ydc1mutant. The expression of the FLAG-tagged aPHC was induced in galactose-containing medium and was verified by Western blot analysis using anti-FLAG antibody (Fig. 6A). The FLAG-tagged aPHC was membrane-associated. Microsomes were prepared from yeast cells and were assayed for ceramidase activity with NBD-C₁₂-phytoceramide, dihydroceramide, and ceramide as substrates. Fig. 6B shows that the (ypc1ydc1) mutant cells transformed with the vector had negligible ceramidase activity toward the NBD-C₁₂-ceramides, whereas the cells expressing the FLAG-tagged aPHC exhibited a substantial activity toward phytoceramide, ceramide, and dihydroceramide.

View larger version (36K): [in this window] [in a new window]   Fig. 6.   Complementation of ceramidase activity but not reverse activity by aPHC in the yeast mutant ypc1ydc1. Microsomes were prepared from cells of the yeast mutant strain (ypc1ydc1) containing the empty vector pYES2-FLAG (Vec) or expressing the FLAG-tagged aPHC (FLAG-aPHC). An equal amount of proteins (20 µg) from each microsome preparation was subjected to SDS-PAGE, and the expression of the FLAG-tagged aPHC was verified by Western blot analysis (panel A). MM, molecular mass. The microsomes from the vector control or aPHC-expressing cells (aPHC) were assayed for ceramidase activity (panel B) or the reverse activity of ceramidase (panel C). Data represent the mean ± S.D. of three independent experiments. PHC, phytoceramide; CER, ceramide; DHC, dihydroceramide.

To investigate whether the human ceramidase catalyzes the reversal reaction of ceramidase, we measured the reverse ceramidase activity of aPHC microsomes by using [³H]palmitic acid plus phytosphingosine, dihydrosphingosine, or sphingosine as substrate. Fig. 6C shows that in contrast to the yeast ceramidase YPC1p, the human ceramidase aPHC had no detectable reverse activity.

We also evaluated if aPHC had reverse activity in cells by testing if its overexpression imparts to cells resistance to fumonisin B1 (FB1), an inhibitor of the CoA-dependent ceramide synthetase. FB1 inhibits growth of S. cerevisiae by blocking the synthesis of sphingolipids or accumulating sphingoid bases and their phosphates. We have demonstrated previously that overexpression of YPC1p rescued from FB1 induced growth inhibition by generating sphingolipids through its CoA-independent reverse activity (20). Fig. 7 shows that in contrast to that of yeast ceramidase, overexpression of aPHC resulted in an increased sensitivity to FB1, hence suggesting that the human ceramidase aPHC did not display reverse activity in cells. The increased sensitivity to FB1 could result from a synergistic effect of overexpression of aPHC and FB1 on the depletion of yeast inositol-phosphorylated sphingolipids, which are essential for yeast viability (26). Alternatively, the FB1 sensitivity could result from the accumulation of phytosphingosine in cells because the accumulation of phytosphingosine or its phosphate inhibits cell growth (9).

View larger version (32K): [in this window] [in a new window]   Fig. 7.   Elevation of a sensitivity to fumonisin B1 by overexpression of aPHC in yeast cells. The yeast cells described in Fig. 5 grown in SC-ura medium with 2% glucose were serially diluted to certain cell densities as indicated. The dilutions were spotted onto SC-ura plates (containing 2% galactose) with or without 450 µM FB1 and were incubated at 30 °C for 3 days. Plates were photographed under an imaging system (Alpha Innotech, Inc.).

To investigate whether overexpression of aPHC alters the metabolism of sphingolipid through its ceramidase activity in cells, the metabolism of sphingolipids was evaluated by labeling yeast cells with [³H]palmitic acid. Fig. 8 demonstrates that overexpression of the human ceramidase led to a decrease in radiolabeled phytoceramide and inositol-phosphorylated sphingolipids but to an increase in phytosphingosine. These results suggest that, similar to the yeast counterpart, aPHC acted as a ceramidase in cells to regulate metabolism of sphingolipids by hydrolyzing yeast ceramides. These results also explain why overexpression of aPHC enhanced the sensitivity of cells to FB1.

View larger version (42K): [in this window] [in a new window]   Fig. 8.   Alteration of metabolism of sphingolipids by overexpression of aPHC in yeast cells. Yeast mutant (ypc1ydc1) cells (6 × 107) containing an empty vector pYES2-FLAG or expressing the FLAG-tagged aPHC were labeled with 5 µCi of [3H]palmitic acid (Pal acid) for 90 min. Total lipids were extracted, hydrolyzed with a mild base dimethylamine, resolved by TLC, and detected by autoradiography as described (20). The labeled sphingolipids were identified according to known standards. DHS, dihydrosphingosine; PHS, phytosphingosine; IPC, inositolphosphoryl ceramide; MIPC, mannosylinositolphosphoryl-ceramide; M(IP)2C, mannosyldi-inositol-phosphorylceramide.

aPHC Is an Alkaline Ceramidase-- Three types of ceramidases, acid, neutral, and alkaline, have been described in mammalian cells according to their pH optima for activity (12). To evaluate the pH optimum of aPHC, its activity was measured using NBD-C₁₂-phytoceramide as substrate at different pH. Fig. 9 shows that aPHC had slight or negligible activity at pH 4.5-6, with the highest activity around pH 9.5, indicating that, similar to its yeast homolog YPC1p, aPHC is also an alkaline ceramidase.

View larger version (13K): [in this window] [in a new window]   Fig. 9.   pH optimum of aPHC. Microsomes were prepared from the yeast mutant ypc1ydc1 expressing FLAG-tagged aPHC. The microsomes were assayed for ceramidase activity using NBD-C12-phytoceramide as substrate at different pH values in the presence of 0.15% Nonidet P-40.

aPHC Is Activated by Ca²⁺ but Inhibited by Zn²⁺-- The Pseudomonas ceramidase requires Ca²⁺ for its activity (16); however, neither acid nor neutral ceramidases in mammalian cells require a cation for enzymatic activity. To investigate the cation dependence of aPHC, its activity was measured in the presence of different cations at a concentration of 5 mM. Fig. 10A shows that Ca²⁺ activated the activity 3-fold, whereas Zn²⁺ inhibited the activity by 60%. Mg²⁺, Mn²⁺, and Cu²⁺ had no effect on the activity at the same concentration. Ca²⁺ activation was concentration-dependent, with 50% activation seen at around 1 mM (Fig. 10B). Depletion of free calcium ion in the reaction mixture by a chelator EGTA (5 mM) inhibited aPHC activity by 50% (data not shown), further suggesting that the calcium ion is an activator of the enzyme.

View larger version (16K): [in this window] [in a new window]   Fig. 10.   Effects of cations on the activity of aPHC. The aPHC microsomes as described above were assayed for ceramidase activity using NBD-C12-phytoceramide as substrate at pH 9.4 in the presence of 5 mM cations (panel A) and different concentrations (0.1-5 mM) of Ca2+ (panel B). Data are expressed as a percent of the ceramidase activity (control) in the absence of cations.

Sphingosine Inhibits aPHC Activity-- We next examined the effects of various lipids on the activity of aPHC. Fig. 11A shows that sphingosine at a concentration of 100 µM inhibited the activity of ceramidase by 40%, whereas neither phytosphingosine nor dihydrosphingosine at the same concentration was inhibitory. The sphingosine inhibition was dose-dependent (Fig. 11B), suggesting that sphingosine is a competitive inhibitor of aPHC. We also found that none of the phospholipids (100 µM) including phosphatidylinositol, phosphatidylcholine, and phosphatidylserine, affected the activity (Fig. 11C).

View larger version (44K): [in this window] [in a new window]   Fig. 11.   Effects of lipids on the activity of aPHC. The aPHC microsomes were assayed for ceramidase activity using NBD-C12-phytoceramide as substrate at pH 9.4 in the presence of 100 µM sphingoid bases (panel A), different concentrations of sphingosine (panel B), or phospholipids (panel C). Data are expressed as a percent of the control in the absence of the lipids. SPH, sphingosine; DHS, dihydrosphingosine; PHS, phytosphingosine; PI, phosphatidylinositol; PS, phosphatidylserine; and PC, phosphatidylcholine.

aPHC Shows Preference toward Phytoceramide as a Substrate-- To evaluate the substrate specificity of aPHC, the Michaelis-Menten constants (K_m) for NBD-C₁₂-phytoceramide, ceramide, and dihydroceramide were determined at pH 9.4 and in the presence of 0.15% Nonidet P-40. The apparent K_m value for NBD-C₁₂-phytoceramide was ~6.8 mol %, three times smaller than that of NBD-ceramide (22.8 mol %) and dihydroceramide (23.3 mol %) (Fig. 12). These data suggest that aPHC prefers NBD-phytoceramide to the other two ceramides as substrate.

View larger version (15K): [in this window] [in a new window]   Fig. 12.   Substrate specificity of aPHC. The aPHC microsomes as described under "Results" were assayed for ceramidase activity using NBD-C12-phytoceramide (), dihydroceramide (), or ceramide () as substrates at different concentrations in the presence of 0.15% (2.5 mM) Nonidet P-40. A double reciprocal plot was analyzed using CA-cricket graph program. Data represent the mean of three independent experiments.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Based on sequence similarity, we have identified and cloned the cDNA of a human homolog of the yeast alkaline ceramidases YPC1p and YDC1p. Three lines of evidence demonstrate that the cDNA encodes a novel human ceramidase (aPHC). First, transient expression of the cDNA in HEK293 cells elevates ceramidase activity. Second, expression of the aPHC cDNA restored ceramidase activity in a yeast mutant devoid of any ceramidase activity. Finally, overexpression of the cDNA resulted in a decrease in synthesis of complex sphingolipids in yeast cells by metabolizing phytoceramide (or dihydroceramide). This new ceramidase exhibited the highest activity at alkaline pH (9.5), thereby defining it as an alkaline ceramidase. Its overexpression led to growth suppression (or death) of yeast cells, suggesting that it may play a role in controlling the cell's well-being. Most interestingly, this mammalian ceramidase prefers phytoceramide as a substrate. Thus it is named as alkaline phytoceramidase.

aPHC does not share any sequence similarity with two other types of mammalian ceramidases, acid (13) and neutral (17), although these three types of ceramidase catalyze a similar reaction, cleaving the amide bond of ceramides. This sequence difference may account for their different substrate specificity, pH optimum, and cellular localization. aPHC uses phytoceramide as its best substrate, whereas the latter two enzymes use mammalian unsaturated ceramide best. The pH optimum for aPHC activity is considerably higher than both the acid and neutral ceramidases. As for cellular localization, aPHC is localized in both the Golgi apparatus and ER, whereas the acid and neutral ceramidases are localized in lysosomes (13) and mitochondria (18), respectively. A more striking difference between aPHC and the neutral ceramidases is the reverse activity of ceramidase. The neutral ceramidases in Pseudomonas (16), mouse (17), and human (27) have a considerable reverse activity, whereas aPHC does not have.

The unique sequences and different cellular localizations of these three types of ceramidase may also account for their different physiological roles. The role of acid ceramidase is better understood than the other two types of ceramidases. Its lysosomal localization and pathogenic phenotype caused by its mutation strongly suggest that it is mainly responsible for clearance of ceramide. As for the neutral and alkaline ceramidases, it is not completely known what functions they have in cells. However, it has been shown that activities of the neutral and alkaline ceramidases in mammalian cells are regulated by growth factors (platelet-derived growth factor) (28) and cytokines (interleukin-1 and tumor necrosis factor-) (28, 29), indicating that they may be important mediators of signaling events in cells.

Based on the structural differences among the sphingoid base moieties, three major types of ceramide have been found in eukaryotic cells: unsaturated ceramide with a double bond on the sphingoid base, dihydroceramide without the double bond, and phytoceramide with an additional hydroxyl group but no double bond on the sphingoid base (30). Each ceramide appears to have its own metabolizing enzyme, suggesting that (a) metabolism of each ceramide may be regulated independently; (b) each ceramide may have an unique pool in a different cellular compartment, and/or (c) each ceramide may have different physiological roles. In fact, there is ample evidence that ceramide, but not dihydroceramide, is a bioactive molecule in many cellular events.

Phytoceramide is a constituent of major complex sphingolipids in lower eukaryotes such as S. cerevisiae (23). However, the existence of phytoceramide in mammals has been mostly neglected because of its low content in many mammalian cell lines or tissues. Recently, phytoceramide has been found in a variety of mammalian tissues and organs including skin (31), cervical epithelium (32), uterine endometrium (32, 33), liver (34), and kidney (35). We found that the mRNA for aPHC is expressed in many human tissues and organs (Fig. 3). Whether distribution of aPHC mRNA correlates with the enzymatic activity and the content of phytoceramide in these tissues awaits further study. Skin has a significant proportion of phytoceramide as a building block of complex sphingolipids, and RT-PCR analysis of aPHC mRNA indicated that aPHC was indeed highly expressed in skin (Fig. 3B). The very recent study by Ardail et al. (34) showed that a free ceramide with t21:1 phytosphingosine as a sphingoid base was the highest in rat liver mitochondria. Our Northern blot and RT-PCR analysis did show that human liver also expressed high levels of aPHC mRNA. Moreover, we demonstrated that both human kidney epithelial cells and monkey fibroblast cells had both high levels of the aPHC mRNA (data not shown) and considerable amounts of ceramidase activity toward phytoceramide. These results indicate that levels of the aPHC mRNA may reflect the content of phytoceramide and its hydrolysis activity. The cloning of aPHC will provide an important tool to address a fundamental issue: what physiological significance phytoceramide and its breakdown products have in mammalian cells.

It has been shown that both ceramide and its intermediate breakdown product sphingosine induce apoptosis or growth suppression of many cancerous cell lines (36-38), whereas its subsequent product sphingosine-1-P has a proliferative effect on fibroblast or endothelial cells (39). Because of the presence of only trace amounts of phytoceramide and its breakdown product phytosphingosine in mammalian cells, their physiological functions have been largely unknown. However, a few studies have begun to suggest that phytosphingosine may play important roles in cell growth or death in eukaryotic cells. Tamiya-Koizumi et al. (40) demonstrated that phytosphingosine suppressed the growth of HL60 cells by targeting DNA primase. We and others demonstrated that phytosphingosine or phytosphingosine-1-P also suppressed the growth of yeast cells (9, 20), and phytosphingosine-1-P elevated heat tolerance (10). In this study, we found that overexpression of human aPHC substantially suppressed yeast growth (data not shown). These data collectively suggest that phytosphingosine and phytopshingosine-1-P are also important bioactive molecules in mammalian cell death or growth suppression in yeast.

Because aPHC is responsible for the regulation of the bioactive lipids, phytoceramide, phytosphingosine, and phytosphingosine-1-phosphate, it is conceivable that the activity of aPHC itself is regulated in cells. In this study, we have shown that aPHC was activated by Ca²⁺ in vitro. Ca²⁺ is a well known second messenger, which plays a critical role in signal transduction cascades. Flux of Ca²⁺ in and out of the ER through calcium channels occurs in response to many signaling events (41). The in vivo activity of aPHC, which is localized in the ER and Golgi apparatus, may be regulated by this Ca²⁺ flux. Moreover, a PROSITE motif search revealed that aPHC contains three putative protein kinase C phosphorylation sites, two casein kinase II phosphorylation sites, one tyrosine phosphorylation site, one N-glycosylation site, and two N-myristoylation sites, suggesting that the activity of this enzyme in cells may also be regulated by phosphorylation and other post-translational modifications. Coroneos et al. (28) demonstrated that pretreating cells with an inhibitor of protein tyrosine phosphatase (sodium vanadate) augmented the stimulation of alkaline ceramidase activity by platelet-derived growth factor in mesangial cells, suggesting that phosphorylation may be involved in the activation of alkaline ceramidase(s).

In yeast, in addition to being a bioactive molecule of cell regulatory events, phytoceramide and dihydroceramide have been shown to be indispensable for the transport of glycosylphosphatidylinositol (GPI)-anchored proteins out of the ER (42). Failure to transport GPI-anchored proteins out of the ER leads to slow growth of yeast cells (42, 43). However, mature GPI-anchored proteins do not contain the ceramide moiety (43), suggesting that ceramides may interact with GPI-anchored proteins noncovalently in assisting the transport, or ceramides are modified or simply removed from GPI-anchored proteins after these proteins exit the ER. A Golgi to ER retrieval sequence is found at the carboxyl termini of aPHC. Our cellular localization study showed that aPHC was localized in both the ER and Golgi apparatus (Fig. 4). Activity of ceramide synthase was shown to be localized in the ER as well (44). Colocalization of the phytoceramidase aPHC with ceramide synthase suggests that these two enzymes may also work as a pair in remodeling GPI anchors in the ER. The aPHC that resides in the Golgi apparatus may be responsible for the remodeling of GPI anchors of proteins that exit the ER with the phytoceramide moiety attached. Whether aPHC is involved in remodeling GPI anchors needs to be investigated thoroughly.

Thanks to the completion of the draft of human genome sequence, two additional aPHC homologs were identified. These homologs also share significant sequence similarity to the yeast alkaline ceramidases YPC1p and YDC1p (20, 21). It is most likely that these homologs are also alkaline ceramidases, but they may hydrolyze either ceramide or dihydroceramide. These data suggest that the alkaline ceramidase family consists of several members, and each member may have a distinct role. Characterization of these two homologs is in progress.

In conclusion, for the first time, this study demonstrates that an alkaline phytoceramidase activity is highly expressed in mammalian cells and is encoded by aPHC, a homolog of the yeast alkaline ceramidases, suggesting that the hydrolysis of phytoceramide is well conserved from the yeast S. cerevisiae to human. aPHC is distinguished from other cloned mammalian ceramidases by its very basic working pH, lack of reverse activity, localization to the ER and Golgi apparatus, and particularly the substrate specificity. aPHC is highly expressed in many human organs and tissues, which raises an intriguing issue: what physiological functions does the hydrolysis of phytoceramide serve? The cloning and characterization of the human aPHC will facilitate our understanding of roles that phytoceramide and its breakdown products have in higher eukaryotes.

	ACKNOWLEDGEMENTS

We thank Kevin P. Becker for excellent technical assistance with confocal laser scanning microscopy and Dr. Yusuf A. Hannun for critically reading the manuscript.

	FOOTNOTES

^* This work was supported in part by a Veterans Affairs merit award (to L. O.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in this paper has been submitted to the GenBank^TM/EMBL Data Bank with accession number(s) AF214454.

^§ To whom correspondence should be addressed: Dept of Medicine, 114 Doughty St., Rm. 605 STB, P. O. Box 250779, Charleston, SC 29425. Tel.: 843-876-5191; Fax: 843-876-5172; E-mail: maoc@musc.edu.

Published, JBC Papers in Press, May 16, 2001, DOI 10.1074/jbc.M102818200

	ABBREVIATIONS

The abbreviations used are: EST, expressed sequence tag; NBD, N-4-nitrobenz-2-oxa-1,3-diazole; kb, kilobase(s); RACE, rapid amplification of cDNA ends; PCR, polymerase chain reaction; aPHC, alkaline phytoceramidase; PAGE, polyacrylamide gel electrophoresis; RT-PCR, reverse transcription-polymerase chain reaction; GFP, green fluorescent protein; PBS, phosphate-buffered saline; FB1, fumonisin B1.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				B. X. Wu, C. F. Snook, M. Tani, E. E. Bullesbach, and Y. A. Hannun Large-scale purification and characterization of recombinant Pseudomonas ceramidase: regulation by calcium J. Lipid Res., March 1, 2007; 48(3): 600 - 608. [Abstract] [Full Text] [PDF]

				E. Houben, W. M. Holleran, T. Yaginuma, C. Mao, L. M. Obeid, V. Rogiers, Y. Takagi, P. M. Elias, and Y. Uchida Differentiation-associated expression of ceramidase isoforms in cultured keratinocytes and epidermis J. Lipid Res., May 1, 2006; 47(5): 1063 - 1070. [Abstract] [Full Text] [PDF]