| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,,2
* Department of Medicine,
Department of Biochemistry and Molecular Biology, and
Ralph H. Johnson Veterans Administration Hospital at the Medical University of South Carolina, Charleston, South Carolina, USA
2Correspondence: 114 Doughty St., Rm. 646, STB, PO Box 250779, Charleston, SC 29425, USA. E-mail: maoc{at}musc.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: placenta • rate-limiting step • serum deprivation • S1P receptor • sphingolipid
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2
3)
. More recently S1P has been implicated in cardiovascular development (4)
, angiogenesis (5)
, immunity (6)
, and tumorigenesis (7
, 8)
. S1P mediates these biological responses through binding to the G-protein-coupled receptors (S1P1–5) in the endothelial differentiation gene (Edg) receptor superfamily (9)
.
Due to its diverse roles, S1P regulation is of great interest. S1P is generated from phosphorylation of sphingosine by the action of sphingosine kinases (10
, 11)
. S1P is then irreversibly cleaved by the S1P lyase (12
, 13)
. S1P is also attenuated by dephosphorylation via the S1P phosphatases (14
, 15)
. Many studies have focused on sphingosine kinases because they catalyze the immediate step for the formation of S1P. Two homologous sphingosine kinases, sphingosine kinase 1 (SK1) (10)
and sphingosine kinase 2 (SK2) (11)
, have been cloned from humans and mice. The deletion of both mouse SK1 (mSK1) and SK2 (mSK2) completely inhibits the formation of S1P in tissues (16)
, suggesting that the sphingosine kinases are essential for the generation of S1P. However, Jolly et al. (17)
demonstrated that overexpression of hSK1 causes a slight increase in S1P in RBL-2H3 mast cells due to limited availability of cellular sphingosine, the substrate of sphingosine kinases, suggesting that the generation of sphingosine can also be a rate-limiting step for the formation of S1P.
Sphingosine is essentially generated from ceramides through the action of ceramidases (18)
. Several mammalian ceramidases have been cloned and are classified as acid, neutral, or alkaline types according to their pH optima for in vitro activity (19)
. A recent study by Monick et al. (20)
showed that the human acid ceramidase, a lysosomal enzyme, is highly expressed in macrophages and that its high expression correlates with high levels of sphingosine, suggesting it may regulate the levels of sphingosine. Neutral ceramidases were cloned from various organisms, including mice (21)
, rats (22)
, and humans (23)
. A recent study by Tani et al. (24)
demonstrated that overexpression of the mouse neutral ceramidase in CHOP cells fails to increase sphingosine and S1P unless the cells are treated with a bacterial sphingomyelinase, which liberates ceramides from sphingomyelins in the plasma membrane, suggesting that the action of the neutral ceramidase per se is insufficient to increase the levels of sphingosine and S1P. Our laboratory cloned mammalian alkaline ceramidases, including the human alkaline phytoceramidase (haPHC) (25)
, the mouse alkaline ceramidase 1 (maCER1) (26)
and its human orthologue, haCER1 (unpublished data). We demonstrated that haPHC mainly utilizes phytoceramide, an analog of ceramide, as substrate (25)
. haPHC is highly expressed in most human tissues (25)
but its preferred substrate, phytoceramide, is present at very low levels (27)
, suggesting that haPHC may play a role in the clearance of phytoceramide in mammalian cells. Studies of its substrate specificity reveal that haPHC exhibits no activity on regular ceramides or dihydroceramides (unpublished data), suggesting a minimal role in regulating the generation of sphingosine or S1P. We showed that maCER1 catalyzes the hydrolysis of ceramide exclusively, and its expression causes an increase in both sphingosine and S1P in cells (26)
, suggesting that maCER1 can regulate S1P levels by controlling the generation of sphingosine. However, the role of maCER1 in controlling the generation of S1P is restricted to the skin, where it is predominantly expressed. We illustrated that haCER1 has a similar tissue-specific expression pattern as maCER1 and plays a similar role in regulating the levels of both sphingosine and S1P in human epidermal keratinocytes (unpublished data).
In this study, we report the cloning and functional analyses of a novel human ceramidase homologous to both haPHC and haCER1. Similar to haCER1 but different from haPHC, this novel ceramidase prefers ceramide over phytoceramide and dihydroceramide for substrate, thus it is termed the human alkaline ceramidase 2 (haCER2). We show that haCER2 regulates the levels of both sphingosine and S1P as well as cell proliferation and survival.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
in the Lipidomics Core at Medical University of South Carolina (MUSC).
Anti-haCER2 antibody (Ab) production
A peptide (DAASEIPEQGPVIKFWPNEK) corresponding to the carboxyl terminus of the haCER2 protein was chemically synthesized and used to immunize rabbits in the Ab production facility at MUSC.
RT-polymerase chain reaction (RT-PCR) analysis
Total RNA was isolated from HeLa T-Rex cells using an EasyRNA kit (Qiagen, Chatsworth, CA, USA) according to the manufacturer’s instructions. One microgram of total RNA from each cell sample was reverse transcribed into cDNAs as described (26)
. One-tenth of the transcribed cDNAs was subjected to polymerase chain reaction (PCR) analysis for expression of haCER2 and S1P1 mRNA under the condition of 1 cycle of 94°C for 30 s, 30 cycles of 94°C for 20 s, 60°C for 30 s, 72°C for 50 s, and 1 cycle of 72°C for 5 min. PCR analysis for 28s rRNA or G3PDH was performed under the same conditions but with one-thirtieth of the transcribed cDNAs. The following primer sets were used: 5'-CTGGACTCTTTTGGTTGTAGTGGG-3'/5'-AGGCAGCAA GGCAGATGAGG-3'(haCER2); 5'-GACTCTGC-TGCTGGCAAATTCAAGCGAC-3'/5'-ACCCTTCCCAGTGCATTGTTCACAG-3'(S1P1); 5'-GAAAGATGGTGAA CTATGCC-3'/5'-TTACCAAAAGTGGCCCACTA-3'(28s rRNA); and 5'-TGAAGGTCGGAGTACACGGATTTGGT-3'/5'-CATGTGGG-CCATGAGGTCC ACCAC-3' (G3PDH). PCR products were verified by DNA sequencing.
qRT-PCR analysis
Real-time PCR was performed on an iCycler system (Bio-Rad, Hercules, CA, USA) using primers for haCER2 (F, 5'-AGTGTCCTGTCTGCGGTTACG-3', and R, 5'-TGTTGTTGATGGCAGGCTTGAC-3') and ß-actin (F, 5'-CAATGTTCGGTGCAATTCAGAG-3', and R, 5'-GGATCCCATTCCTACCACT-GTG-3'). Standard reaction volume was 25 µl, including 12.5 µl of iQTM SYBR Green Supermix (Bio-Rad), 10 µl of cDNA template, and 2.5 µl of a primer mixture. The initial PCR step was 3 min at 95°C, followed by 40 cycles of a 10 s melting at 95°C and a 45 s annealing/extension at 60°C. The final step was 1 min incubation at 60°C. All reactions were performed in triplicate. Real-time RT-PCR results were analyzed using Q-Gene software which expresses data as mean normalized expression (MNE) (29)
. MNE is directly proportional to the amount of mRNA of the target gene (haCER1) relative to the amount of mRNA of the reference gene (ß-actin).
Northern blot analysis
Northern blot analysis was performed as described (25)
.
Plasmid construction
The open reading frame (ORF) of haCER2 was amplified by PCR from a human liver cDNA library (BD Biosciences, Bedford, MA, USA) using primers (F, 5'-GTTGATCAATGGGCGCCCCGCACTGGTG-3', and R, 5'-CGGAATTCTCACG TGGTCTTGACTGATGATTTC-3'). The amplified ORF was cloned into a PCR cloning vector pCR2.1 (Invitrogen, Carlsbad, CA, USA) to generate the plasmid pCR2.1-haCER2, from which the haCER2 ORF was excised and cloned into a mammalian expression vector pcDNA4. The resulting expression construct pcDNA4-haCER2 was verified by DNA sequencing.
Stable cell line construction
Hela-T-Rex cells (Invitrogen) were transfected with the empty vector pcDNA4 or the expression construct pcDNA4-haCER2 using Lipofectamine 2000 (Invitrogen) as described (25)
. The transfected cells were selected in MEM medium containing zeocin (200 µg/ml) and blasticidin (5 µg/ml) (Invitrogen). The antibiotic resistant clones were screened for expression of haCER2 by Western blot analysis using the anti-haCER2 Ab as described (25)
. The clones expressing haCER2 in the presence of tetracycline (Tet), but not ethanol (Et), the vehicle control, were maintained in the same medium as above but supplemented with 50 µg/ml zeocin and 5 µg/ml blasticidin. A representative clone, termed haCER2-TET-ON, was chosen for further studies.
siRNA transfection
Control siRNA [UUCUCCGAACGUGUCACGUdTT (sense)/ACGUGACACGUUCGGAGAAdT (antisense)], siRNAs against haCER2 (GCCUGCCAUCAACAACAUCdTT/GAUGUUG-UUGAUGGCAGGCdTT) and S1P1 (CUGACCUCGGUGGUGUUCAdTT/UGAACACCACCGAGGUCAGdTT) were chemically synthesized in Qiagen The siRNAs were transfected into cells at a concentration of 5–10 nM using Oligofectamine (Invitrogen) according to the manufacturer’s instructions.
Immunocytochemistry
Cells were stained with the anti-haCER2 Ab (1:200), anti-GM130 Ab (1:200), or both. After washing, the above cells were stained with an anti-rabbit IgG Ab-FITC, antimouse IgG Ab-rhodamine (1:200), or both as described (25)
. Cellular localization was analyzed under a confocal laser scanning microscope (LSM 510 META; Zeiss, Thornwood, NY, USA).
Protein concentration determination
Protein concentrations were determined with BSA as a standard using a bicinchoninic acid (BCA) protein determination kit (Pierce, Rockford, IL, USA) according to the manufacturer’s instructions.
Ceramidase activity assay
Ceramidase activity was determined by the release of sphingoid bases from ceramides. Briefly, microsomes (
20 µg proteins) were incubated with ceramide substrate (75 µM) in 40 µl assay buffer (25 mM Tris-HCl, pH 6–9, containing 5 mM CaCl2 and MgCl2, 0.15% Nonidet P-40) at 37°C for 30 min. The reactions were stopped by extraction with chloroform and methanol. Sphingoid bases in the lipid extracts were determined by HPLC with D-e-C17-sphingosine as an internal standard as described (28)
.
Sphingolipid labeling
After 2 h pulse labeling with 2.5 µCi [3H]-sphingosine, 85% confluent haCER2-TET-ON cells in each 65 mm culture dish were incubated in the nonradioactive fresh medium containing ethanol or tetracycline (100 ng/ml µg) for an additional 24 h. Total sphingolipids were extracted and analyzed by TLC as described previously (26)
. The radiolabeled sphingolipid bands were detected by autoradiography and identified by their corresponding standards.
ESI/MS/MS lipid analysis
Electrospray ionization (ESI)/MS/MS analysis of sphingolipids was performed on a Thermo Finnigan TSQ 7000 triple quadrupole mass spectrometer, operating in a multiple reaction monitoring (MRM) positive ionization mode, according to the method (30)
modified from a previous study (31)
.
MTT assay
The number of viable cells was determined using an in vitro toxicology assay kit (MTT-based; Sigma, St. Louis, MO, USA) according to the manufacturer’s instructions.
FACS analysis
Cell cycle profiles were analyzed by fluorescence activated cell sorting (FACS) on a FACStarplus flow cytometer (BD Biosciences) as described (32)
.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A). This putative ORF was amplified by PCR from a human liver cDNA library (Clontech, Palo Alto, CA, uSA). DNA sequencing indicated that the cloned ORF has the predicted sequence. The pSORT II program predicted that haCER2 consists of 7 putative transmembrane domains (TM1–7) (Fig. 1A
). Protein sequence alignment revealed that haCER2 exhibits a 25%, 25%, 28%, 39%, and 39% identity in protein sequence to the yeast alkaline phytoceramidase (YPC1p) (33)
, the yeast dihydroceramidase (YDC1p) (34)
, haPHC (25)
, maCER1 (26)
, and haCER1, respectively (Fig. 1B
). It was also found that haCER2 contains multiple conserved domains (underlined) shared among these alkaline ceramidases (Fig. 1B
). These protein sequence features indicate that haCER2 may be a novel ceramidase. haCER2 was also found to have a protein sequence nearly identical to the product of the previously cloned mouse liver cancer-related gene, CRG-L1 (35)
, which we renamed maCER2.
|
haCER2 is highly expressed in placenta
To better understand the physiological roles of haCER2, we investigated the tissue-specific expression of haCER2 mRNA. Northern blot analysis showed that a 6.6-kb haCER2 mRNA is highly expressed in the placenta, but its expression is much lower in other organs, including brain, heart, skeletal muscle, thymus, spleen, kidney, liver, and lung (Fig. 2
A).
|
To quantitatively compare haCER2 mRNA in different tissues, we performed qRT-PCR analysis with cDNAs reverse transcribed from RNA in human tissues. The results showed that the haCER2 mRNA level in the placenta is 11-, 13-, 22-, 45-, 54-, 65-, 74-, and 112-fold higher than that in pancreas, heart, skeletal muscle, liver, colon, kidney, lung, and small intestine, respectively (Fig. 2B
).
Because its mouse counterpart CRG-L1 was found to be up-regulated in cancerous liver (35)
, we investigated whether haCER2 is up-regulated in human tumors. By qRT-PCR, we found that haCER2 is up-regulated in liver and colon tumor tissues, compared to the corresponding normal tissues (Fig. 2C
). These results suggest that haCER2 mRNA is expressed highly in the placenta, moderately in the pancreas and heart, and slightly in other organs, and that haCER2 is up-regulated in some tumors.
haCER2 encodes ceramidase activity
To investigate whether haCER2 encodes ceramidase activity, we generated a stable HeLa-based cell line, haCER2-TET-ON, that expresses haCER2 in HeLa T-Rex cells under the control of a tetracycline-inducible promoter system (cytomegalovirus (CMV)-Tet-ON). This inducible expression system allowed us to turn on/off the expression of haCER2 by adding tetracycline (Tet) or vehicle control, ethanol (Et), respectively, to medium. Western blot analysis with an Ab against haCER2 showed that ectopic expression of haCER2 was induced with Tet (100 ng/ml), but not Et, in haCER2-TET-ON cells (Fig. 3
A).
|
Microsomes were then prepared from haCER2-TET-ON cells grown in the presence of Et or Tet (100 ng/ml) and were measured for ceramidase activity at pH 7.4. haCER2 expression caused a 20-fold increase in microsomal ceramidase activity with D-e-C24:1-ceramide as substrate (Fig. 3B
). However, when the same concentration of D-e-C24:1-dihydroceramide or D-ribo-C24:1-phytoceramide was used as substrate, haCER2 expression slightly increased ceramidase activity (Fig. 3B
). These results suggest that haCER2 encodes ceramidase activity with the sphingosine-containing ceramide as a preferred substrate.
To determine the pH optimum of haCER2 for its in vitro activity, the microsomes isolated from haCER2-TET-ON cells grown in the presence of Et or Tet were measured for ceramidase activity using D-e-C24:1-ceramide as substrate at different pHs. As shown in Fig. 3C
, the difference in ceramidase activity between the microsomes from haCER2-TET-ON cells grown in the presence of Tet and Et, respectively, was the highest at pH 7 to 9. These results suggest that haCER2 has a neutral-to-alkaline pH optimum.
haCER2 overexpression increases the generation of both sphingosine and S1P
Because haCER2 exhibited ceramidase activity in vitro, we determined the effects of haCER2 expression on the cellular levels of ceramides, sphingosine, and S1P. Because the Tet-induced ectopic expression of haCER2 reached the peak at 24 h, haCER2-TET-ON cells grown to a 80% confluence were treated for 24 h with Et or Tet (100 ng/ml) before they were harvested and subjected to lipid extraction. ESI/MS/MS analysis showed that the Tet-induced expression of haCER2 caused a decrease in the levels of D-e-C14-ceramide, D-e-C16-ceramide, D-e-C18:1-ceramide, D-e-C24:0-ceramide, and C24:1-ceramide (Fig. 4
A), with an increase in the levels of sphingosine and S1P (Fig. 4B
). These results suggest that haCER2 can regulate cellular ceramides, sphingosine, and S1P by controlling the hydrolysis of ceramides.
|
It has been shown that ceramide is generated through two major pathways: the de novo synthesis or sphingomyelin hydrolysis. To define the sources of ceramides that serve as the substrates of haCER2 in cells, we determined whether an inhibition of ceramide synthases inhibited the haCER2-mediated generation of sphingosine and S1P, and whether an increased hydrolysis of sphingomyelin had the opposite effect. ESI/MS/MS demonstrated that treatment of haCER2-TET-ON cells with fumonisin B1 (FB1), a ceramide synthase inhibitor, markedly attenuated the generation of both sphingosine and S1P in response to the Tet-induced haCER2 expression (Fig. 4B
). In contrast, treatment with bacterial sphingomyelinase (bSMase) robustly enhanced the generation of both sphingosine and S1P (Fig. 4B
). These results suggest that haCER2 catalyzes the hydrolysis of ceramides derived from both the de novo and sphingomyelin breakdown pathways.
High ectopic expression of haCER2 causes fragmentation of the Golgi complex and growth arrest in HeLa cells
haCER2 expression alters the levels of ceramide, sphingosine, and S1P, all of which have been implicated in regulating cell growth and survival; thus, we investigated the cellular effects of haCER2 expression in HeLa cells. MTT assays demonstrated that the haCER2 expression induced by 100 ng/ml Tet markedly decreased cell viability (Fig. 5
A). Tet at this concentration caused a 5-fold increase in the levels of sphingosine, which we previously showed fragmented or disrupted the Golgi complex in HeLa cells (36)
. To investigate whether sphingosine generated through the action of haCER2 caused fragmentation of the Golgi complex, we labeled the Golgi complex with an Ab against GM130, a Golgi resident protein. Immunostaining revealed that the Golgi complex was markedly fragmented in haCER2-TET-ON cells grown in the presence of 100 ng/ml Tet whereas it was compact in haCER2-TET-ON cells grown in the presence of Et (Fig. 5A
). Tet at this concentration failed to cause fragmentation of the Golgi complex in wild-type (WT) HeLa T-Rex cells (data not shown). These results suggest that the haCER2 expression induced by 100 ng/ml Tet causes the fragmentation of the Golgi complex probably due to an accumulation of sphingosine in cells.
|
Tet at 100 ng caused a 20-fold increase in haCER2 activity (Fig. 5B
), which might be beyond a physiological level. Accordingly, we decreased the ectopic expression of haCER2 by lowering the concentration of Tet. Tet at 1 and 10 ng/ml caused a 4-fold and 13-fold increase in ceramidase activity with D-e-C24:1-ceramide as substrate, respectively (Fig. 5B
). Confocal microscopy showed that the Golgi complex was compact in haCER2-TET-ON cells grown in the presence of 1 ng/ml of Tet but was slightly fragmented in the presence of 10 ng/ml Tet (Fig. 5C
).
To investigate whether sphingosine, which caused the fragmentation of the Golgi complex, was generated in the same organelle, we determined whether haCER2 is colocalized with GM130 by staining haCER2-TET-ON cells with the anti-haCER2 Ab. Confocal microscopy revealed that haCER2 was colocalized with GM130 in haCER-TET-ON cells grown in the presence of different concentrations of Tet (Fig. 5C
), suggesting that haCER2 is localized to the Golgi complex independent of its expression level.
Low ectopic expression of haCER2 promotes serum-independent cell proliferation
Because Tet at 1 ng/ml caused a moderate increase in ceramidase activity in haCER2-TET-ON cells without fragmenting the Golgi complex, we investigated cellular responses to haCER2 expression induced by 1 ng/ml of Tet. At 1 ng/ml, Tet caused an 82.4% and 408.9% increase in the levels of sphingosine and S1P (Fig. 6
A), respectively, with a slight decrease in levels of D-e-C14-ceramide, D-e-C16-ceramide, D-e-C18:1-ceramide, D-e-C24:0-ceramide, and C24:1-ceramide in haCER2-TET-ON cells (Fig. 6B
). MTT assays showed no discernable difference in the viability of haCER2-TET-ON cells cultured in regular MEM containing 1 ng/ml Tet or Et (Fig. 6B
), suggesting that low ectopic expression of haCER2 exerts no effect on cell growth under a normal culture condition.
|
Low ectopic expression of haCER2 causes a substantial increase in the levels of S1P that has been shown to promote proliferation and survival of a variety of cell types in the absence of serum. Thus, we analyzed the effects of haCER2 expression on cell growth and survival in serum-free medium. After haCER2-TET-ON cells were grown to 50% confluence in serum-containing MEM, they were switched to serum-free MEM in the presence of Et or Tet. MTT assays revealed that 48 h after the medium switch, the number of haCER2-TET-ON cells grown in the presence of Tet (1 ng/ml) was significantly higher than the number of viable haCER2-TET-ON cells in the presence of Et (Fig. 6C
). FACS analysis revealed that the cell population in the S-phase of the cell cycle in haCER2-TET-ON cells grown in the presence of Tet was greater than in the presence of Et (Fig. 6D
). No sub-G1/G0 DNA content was detected in haCER2-TET-ON cells grown in the presence of either Et or Tet (Fig. 6D
). These results suggest that haCER2 expression promotes the serum-independent proliferation of cells.
haCER2 expression promotes cell proliferation by activating the S1P/S1P1
To confirm that the S1P increase was responsible for the proliferative effect of haCER2, the formation of S1P in haCER2-TET-ON cells was blocked by dimethylsphingosine (DMS), a competitive inhibitor of sphingosine kinases. ESI/MS/MS revealed that DMS at 2.5 µM markedly blocked the generation of S1P in response to haCER2 expression (Fig. 7
A). MTT assays demonstrated that treatment with DMS totally inhibited the haCER2-mediated proliferation of HeLa T-Rex cells in the serum-free medium, suggesting that an increased generation of S1P may be associated with the growth-promoting effect of haCER2 expression (Fig. 7B
). To further validate this hypothesis, we tested whether exogenous S1P promoted proliferation of HeLa T-Rex cells in serum-free medium. MTT assays revealed that S1P at 100 nM potently promoted proliferation of HeLa T-Rex cells grown in serum-free MEM (Fig. 7C
), further supporting the role of S1P in the haCER2-mediated cell proliferation.
|
S1P has been shown to promote cell proliferation by binding to the G-protein-coupled receptors, S1P1–2 (37)
. To determine whether the S1P receptors played a role in the cell proliferation induced by haCER2 expression, we inhibited their expression by RNAi. RT-PCR analysis showed that S1P1 is the major S1P receptor in haCER2-TET-ON cells (Fig. 7D
); thus, RNAi was performed to down-regulate its expression. RT-PCR analysis showed that expression of S1P1 was significantly inhibited by a S1P1-specific siRNA compared to a control siRNA (Fig. 7E
). We then determined the effects of the down-regulation of S1P1 on the cell proliferation induced by haCER2 expression. Twenty-four hours after siRNA transfection, haCER2-TET-ON cells were grown for an additional 48 h in serum-free MEM in the presence of Et and Tet (1 ng/ml), respectively. MTT assays showed that transfection with the S1P1 siRNA, but not the control siRNA, inhibited the proliferation of HeLa T-Rex cells in response to the Tet-induced expression of haCER2 (Fig. 7F
), suggesting that S1P1 plays a role in the cell proliferation induced by haCER2 expression.
RNAi-mediated down-regulation of haCER2 enhances cell growth arrest and apoptosis in response to serum depletion
The above results showed that haCER2 up-regulation caused cell proliferation; thus, we investigated whether haCER2 down-regulation would have the opposite effect. Expression of haCER2 in HeLa T-Rex cells was down-regulated by RNAi. qRT-PCR analysis demonstrated that transfection of HeLa T-Rex cells with a siRNA specifically against haCER2 (the haCER2 siRNA) caused a significant decrease in the levels of haCER2 mRNA compared with the control siRNA (Fig. 8
A). In vitro activity assays demonstrated that ceramidase activity on D-e-C24:1-ceramide was decreased in HeLa T-Rex cells transfected with the haCER2 siRNA compared to the control siRNA (Fig. 8B
). These results suggest that both haCER2 mRNA and protein were knocked down by the haCER2 siRNA.
|
We then determined the effects of haCER2 knockdown on the levels of sphingolipids. ESI/MS/MS revealed that the RNAi-mediated knockdown of haCER2 markedly decreased the levels of S1P and sphingosine with a slight increase in the levels of ceramides (Fig. 8C, D
).
MTT assays demonstrated that haCER2 knockdown caused a slight decrease in the number of viable HeLa T-Rex cells grown in the presence of 10% FBS (Fig. 8E
). However, haCER2 knockdown markedly decreased the number of HeLa cells grown in the serum-free medium (Fig. 8F
). Consistently, FACS analysis revealed that haCER2 knockdown significantly decreased the percentage of cells in the S-phase of the cell cycle in the serum-free medium, with a substantial increase in the sub-G1 cell population (Fig. 8G
), suggesting that haCER2 knockdown enhances both growth arrest and apoptosis of HeLa cells in serum-free medium. To confirm that the above cellular effects were due to a block in the formation of S1P, we tested whether addition of the exogenous S1P could rescue cells from the inhibitory effects of the haCER2 siRNA. Addition of S1P to the medium significantly suppressed both the cell cycle arrest and apoptosis induced by haCER2 knockdown (Fig. 8G
). Taken together, these results suggest that haCER2 knockdown potentiate the serum deprivation-induced growth arrest and apoptosis of cells by inhibiting the formation of S1P.
haCER2 is up-regulated by serum deprivation
The ectopic expression of haCER2 promotes cell proliferation and survival against serum deprivation. This prompted us to investigate whether expression of the endogenous haCER2 was up-regulated upon serum deprivation. Quantitative RT-PCR analysis (qRT-PCR) demonstrated that haCER2 mRNA was significantly up-regulated in response to serum deprivation in a time-dependent manner, reaching a peak at 8 h (Fig. 9
A). In vitro activity assays showed that serum deprivation markedly increased ceramidase activity on D-e-C24:1-ceramide at 12 h after its onset (Fig. 9B
). These results suggest that serum deprivation up-regulates both haCER2 mRNA and activity.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
haCER2 is a member of the multiple-transmembrane domain (MTD) ceramidase superfamily
Protein sequence alignment revealed that haCER2 shares a high degree of sequence similarity to the alkaline ceramidases we previously cloned from yeast (YPC1p and YDC1p) (33
, 34)
, mice (maCER1) (26)
, and humans [haPHC (25)
and haCER1 (26)
]. haPHC, haCER1, and haCER2 are encoded by distinct genes on chromosomes 11, 19, and 9, respectively. Unlike other alkaline ceramidases, haCER2 has a much broader pH optimum, with significant activity at pHs between 6.5 and 7.4. This is consistent with its cellular localization to the Golgi complex where pH is below 7. According to its pH optimum, haCER2 should be considered a neutral/alkaline ceramidase. Because haCER2 and its homologous ceramidases contain putative multiple transmembrane domains, we reclassify these enzymes as the multiple transmembrane domain (MTD) ceramidases. BLAST search also revealed that haCER2 has an 88% identity in protein sequence to the product of a previously cloned mouse liver cancer-related gene, CRG-L1, which is significantly up-regulated in liver tumors compared to normal liver (35)
. Due to its high sequence similarity to haCER2, the protein product of CRG-L1 is highly likely to be the mouse counterpart of haCER2; thus, we renamed it maCER2 (the GenBank accession number, AY312515). We also found that haCER2 is significantly homologous to a Drosophila protein encoded by the brain washing gene (BWA) (the GenBank accession number: AF323976) whose mutation alters the structure of mushroom bodies (38)
. Protein sequence alignment revealed that this protein contains highly conserved protein domains shared among all known alkaline ceramidases, suggesting it may be a ceramidase as well. We renamed this protein the Drosophila alkaline ceramidase 2, daCER2. These observations suggest that haCER2 is an evolutionarily conserved enzyme that belongs to a big protein family, the MTD ceramidase family.
haCER2 regulates the levels of S1P by controlling the generation of sphingosine
S1P is abundant in platelets and serum. Recent studies indicated that S1P can be generated in a variety of cell types (39)
, although regulation of its generation is unclear. Jolly et al. showed that due to limited availability of free sphingosine in cells, overexpression of hSK1 in macrophages only causes a slight increase in the levels of S1P (17)
. In agreement with this finding, Maceyka et al. reported that overexpression of either hSK1 or hSK2 in HEK293 cells fails to increase the levels of S1P unless exogenous sphingosine is added to cells (40)
. By ESI/MS/MS analysis, we showed that the levels of both sphingosine and S1P are at least 20-fold lower than those of its precursor, ceramides, in Hela T-Rex cells. These observations suggest that the conversion of ceramide to sphingosine is a rate-limiting step for the formation of S1P. In agreement with this notion, we revealed that haCER2 expression caused an increase in the levels of S1P in HeLa-T-Rex cells by enhancing the generation of sphingosine. These results collectively support that the hydrolysis of ceramides is the rate-limiting step for the formation of S1P. We demonstrated that an aberrant accumulation of sphinogsine in cells causes a fragmentation or total disruption of the Golgi complex (36)
. In agreement, we found that high ectopic expression of haCER2 caused an accumulation of sphingosine, leading to a fragmentation of the Golgi complex. These observations suggest that sphingosine is highly cytotoxic. This may explain why the generation of sphingosine must be tightly regulated in cells.
The human acid ceramidase is capable of controlling the generation of sphingosine as well, but sphingosine generated by the action of the acid ceramidase is not efficiently converted to S1P (20)
. Overexpression of the neutral ceramidase fails to increase the levels of sphingosine or S1P unless ceramides are liberated from sphingomyelin on the plasma membrane (24)
. In contrast to the acid and neutral ceramidases, haCER2 is highly effective in controlling the generation of both sphingosine and S1P. This may be associated with its cellular localization to the Golgi complex. After being synthesized in the ER through the de novo pathway, ceramides are transported to the Golgi complex (41
, 42)
. Ceramide transport is mediated by a ceramide transfer protein, CERT (43)
. We showed that treatment of haCER2-TET-ON cells with FB1, which specifically blocked the biosynthesis of dihydroceramides and ceramides, markedly inhibited the haCER2-mediated generation of both sphingosine and S1P. This suggests that haCER2 catalyzes the hydrolysis of ceramides derived from the de novo pathway. Ceramide is also derived from the hydrolysis of sphingomyelin through action of the acid or neutral sphingomyelinase. The neutral sphingomyelinase is localized to the Golgi complex (44)
, suggesting that ceramide is also generated in this organelle. Upon stimulation, the acid sphingomyelinase has been shown to translocate to the plasma membrane to hydrolyze sphingomyelin to liberate ceramide (45)
. We demonstrated that treatment of haCER2-TET-ON cells with bSMase liberated ceramide from the plasma membrane and enhanced the haCER2-mediated generation of both sphingosine and S1P, suggesting that the ceramide generated on the plasma membrane translocates to the Golgi complex where it is hydrolyzed by the action of haCER2. These observations suggest that ceramides generated from different pathways converge in the Golgi complex and serve as the substrates of haCER2. This allows haCER2 to efficiently catalyze the generation of sphingosine due to unlimited substrate availability. We demonstrated that sphingosine generated in the Golgi complex could be rapidly phosphorylated to S1P, suggesting that the Golgi sphingosine is highly accessible by sphingosine kinases. A recent study by Delon et al. (46)
showed that upon the up-regulation of PLD or treatment with 12-myristate 13-acetate (PMA), hSK1 is also translocated to a perinuclear Golgi-like membrane structure, suggesting that sphingosine generated by the haCER2 action can be locally phosphorylated. The coupling of the generation of sphingosine to its phosphorylation in the same cellular compartment further enhances the ability of haCER2 to control the formation of S1P. These observations suggest that its Golgi localization makes haCER2 a key ceramidase in controlling the generation of S1P.
haCER2 regulates cellular responses through S1P and its receptor S1P1
We demonstrated that haCER2 expression promoted proliferation of HeLa cells in the absence of serum, where the RNAi-mediated inhibition of haCER2 enhanced cell growth arrest and apoptosis in response to serum deprivation. The pharmacological inhibition of sphingosine kinases that catalyze conversion of sphingosine to S1P inhibited the proliferative effect of haCER2 expression. The RNAi-mediated inhibition of S1P1, the major receptor of S1P in HeLa cells, significantly suppressed the proliferative effect of haCER2 expression. The exogenous S1P could rescue cells from the inhibitory effects of the down-regulation of haCER2. These results strongly suggest that an inhibition of S1P formation is responsible for the effects of haCER2 down-regulation.
Neither haCER2 up-regulation nor down-regulation affected cell proliferation or survival in the presence of 10% FBS. FBS contains a high content of S1P, which may mask the effects of haCER2, or growth factors in FBS might have a redundant role in cell proliferation and survival. We demonstrated that haCER1 expression in HeLa cells was significantly up-regulated by serum deprivation. These results strongly suggest that haCER2 plays a critical role in serum-independent cell proliferation and survival by controlling the formation of S1P.
With the discovery of the G-protein-coupled receptors (S1P1–5) specific for S1P, a greater appreciation of the role of S1P as an extracellular signaling molecule in regulating various biological responses has emerged (47)
. The results presented in this study strongly support the extracellular action of S1P, because down-regulation of the major receptor S1P1 in HeLa T-Rex cells substantially inhibited the proliferative effect of haCER2 expression. Taken together, these results suggest that haCER2 mediates cell proliferation through the S1P/S1P1 pathway.
haCER2 may play a role in the development of the placenta
By both Northern blot analysis and qRT-PCR, we demonstrated that haCER2 is expressed in the placenta at a much higher level than in other organs. Our in vitrostudies strongly suggest that haCER2 is an important regulator of the generation of S1P, which has been shown to play a critical role in vascular development (4
, 48
, 16)
. Because blood vessels are enriched in the placenta, we speculate that haCER2 may play an important part in the development of the placenta by regulating the formation of blood vessels.
haCER2 up-regulation may have a role in tumorigenesis
As mentioned earlier, haCER2 is the orthologue of the protein product of the mouse liver cancer-related gene (CRG-L1/maCER2). The CRG-L1/maCER2 gene was identified as a gene that is significantly up-regulated in the mouse cancerous liver compared to the normal liver (35)
. The same study further showed that expression of the human orthologue, namely haCER2, is also up-regulated in all hepatocellular carcinomas although its biochemical function has been unknown. In agreement, we found that haCER2 is up-regulated in liver and colon tumor tissues. We have now revealed that haCER2 is a ceramidase whose up-regulation promotes proliferation and survival of cervical tumor cells under serum deprivation conditions by elevating the levels of S1P. Based on these findings, we speculate that haCER2 up-regulation may contribute to tumorigenesis by enhancing tumor cell proliferation and survival under the stress of limited growth factors. If this hypothesis is correct, haCER2 may serve as an important target for chemotherapeutic intervention of cancers.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication January 11, 2006. Accepted for publication April 10, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  X. Ye Lysophospholipid signaling in the function and pathology of the reproductive system Hum. Reprod. Update, September 1, 2008; 14(5): 519 - 536. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
