Abstract
Cell type-specific localization of sphingosine kinase 1a (SPHK1a) in tissues was analyzed with a rabbit polyclonal antibody against the 16 C-terminal amino acids derived from the recently reported mouse cDNA sequence of SPHK1a. This antibody (anti-SPHK1a antibody) can react specifically with SPHK1a of mouse, rat, and human tissues. Utilizing its crossreactivity to human SPHK1a, the cell-specific localization of SPHK1a in human tissues was histochemically examined. Strong positive staining for SPHK1a was observed in the white matter in the cerebrum and cerebellum, the red nucleus and cerebral peduncle in the midbrain, the uriniferous tubules in the kidney, the endothelial cells in vessels of various organs, and in megakaryocytes and platelets. The lining cells of sinusoids in the liver and splenic cords in the spleen showed moderate staining. Columnar epithelia in the intestine and Leydig's cells in the testis showed weak staining patterns. In addition, TPA-treated HEL cells, a human leukemia cell line, showed a megakaryocytic phenotype accompanied with increases in immunostaining of both SPHK1a and SPHK enzyme activity, suggesting that SPHK1a may be a novel marker of megakaryocytic differentiation and that this antibody is also useful for in vitro study of differentiation models.
Keywords
T
A method for measuring SPHK enzyme activity was reported by Olivera et al. (1994). Using this method, it is reported that SPHK enzyme activity is activated and increases the cellular level of Sph-1-P through stimuli by various substances such as platelet-derived growth factor (PDGF) (Olivera and Spiegel 1993), serum (Rius et al. 1997), nerve growth factor (Edsall et al. 1997), vitamin D3 (Kleuser et al. 1998), activators of protein kinase C or protein kinase A (Mazurek et al. 1994; Machwate et al. 1998), antigens (Melendez et al. 1998), and muscarinic agonists (Meyer zu Heringdorf et al. 1998).
The tissue-specific distribution of SPHK has mainly been analyzed by measuring enzyme activity (Gijsbers et al. 1999) or Sph-1-P content in tissue (Yatomi et al. 1997; Edsall and Spiegel 1999). However, these methods cannot provide pertinent information on the cell type-specific localization of SPHK in the tissue, which is indispensable for understanding the physiological and pathological significance of this enzyme.
Genomic information on SPHK has recently become available. Mouse SPHK1, which was cloned as the first mammalian form of SPHK, has an apparent molecular weight of 49 kD and has two very similar types, 1a and 1b (Kohama et al. 1998). The human homologue of SPHK1 was also reported (Melendez et al. 2000). In addition, other isoforms of SPHK have been also reported while the present study was in progress (Melendez et al. 2000; Liu et al. 2000).
Based on the mouse cDNA sequence of SPHK1a (Kohama et al. 1998), we prepared a rabbit polyclonal antibody against the C-terminal oligopeptide, which was cross-immunoreactive to human SPHK1a. Using this anti-SPHK1a antibody, the cell type-specific localization of SPHK1a in human tissues was analyzed immunohistochemically. The change in SPHK1a levels was also studied immunohistochemically during TPA-induced differentiation of HEL cells.
Materials and Methods
Antibody Preparation
The antibody against SPHK1a (anti-SPHK1a antibody) was generated in New Zealand White rabbits by injecting the synthetic oligopeptide corresponding to the 16 C-terminal residues (APSGRDSRRGPPPEEP) of the mouse SPHK1a sequence (Kohama et al. 1998) that was conjugated to keyhole limpet hemocyanin with glutaraldehyde. Antiserum was purified using a CNBr-activated agarose gel bound to the same synthetic peptide and eluted with 100 mM glycine, pH 2.5, and neutralized immediately with 1 M Tris.
Cell Line and Culture Conditions
A human erythroleukemia cell line, HEL, was cultured in 10% FCS in RPMI 1640 medium and treated with 10−7 M of phorbol ester, TPA (phorbol 12-myristate 13 acetate), as described previously (Murate et al. 1993). After 2 days, cell samples were collected to measure SPHK enzyme activity, and cytospin slides were prepared for immunostaining with anti-SPHK1a. CHO cells and COS 7 cells were used for the transfection experiments.
Immunoprecipitation and Western Blotting Analysis of Human Tissue
Outdated human platelets were supplied from Aichi Red Cross (Aichi, Japan). Human kidney tissue was obtained from a small part of total nephrectomized kidneys with renal tumors resected from patients who had given their informed consent. The sample collection was in accordance with the Helsinki Declaration of 1975. Human tissues or platelets were suspended in lysis buffer (50 mM Tris/HCl, 1% (v/v) Triton X-100, 1% sodium cholate, 0.1% sodium dodecyl sulfate (SDS), 10 mM EGTA, 150 mM NaCl, pH 8.8) containing 1 mM phenylmethylsulfonyl fluoride (PMSF) and 20 μg/ml leupeptin. The lysate was then sonicated and centrifuged at 100,000 × g for 60 min, and the resulting supernatant (500 μg of total protein) was incubated with 2 μg of anti-SPHK1a and 40 μl of a 1:1 slurry of protein A–Sepharose for 16 hr at 4C. The immunoprecipitates were collected by centrifugation at 5000 × g for 5 min, washed three times with lysis buffer, and suspended in SDS sample buffer.
The immunoprecipitated proteins were separated by SDS-PAGE on a 10% gel and transferred to an Immobilon (Millipore; Bedford, MA) membrane in solution containing 25 mM Tris base, 192 mM glycine, and 20% methanol. After blocking with 5% skim milk in TBS-T buffer (0.05% Tween-20, 137 mM NaCl, 20 mM Tris-HCl, pH 7.5), membranes were incubated with the anti-SPHK1a antibody (final concentration 500 ng/ml) at 4C for 15 hr and then washed with TBS-T, followed by incubation for 1 hr at room temperature with 125I labeled protein A. The blots were again washed with TBS-T and detected using a Hamamatsu DVS 3000 image analyzer system (Hamamatsu Photonics; Hamamatsu, Japan).
Western Blotting Analysis of Rat Tissue
Whole rat tissue lysate without immunoprecipitation treatment was directly subjected to Western blotting analysis. The blots were detected by HRP-conjugated anti-rabbit IgG (Amersham-Pharmacia; Poole, UK) and an ECL Western blotting system (Amersham-Pharmacia).
Preparation of Recombinant Murine SPHK1a cDNA
The recombinant murine SPHK1a (Kohama et al. 1998) was amplified from mRNA of NIH 3T3 cells by reverse transcriptase polymerase chain reaction using primers 5′-CTGGCTATGGAACCAGAATG-3′ and 5′-TTATGGTTCTTCTGGAGGTG-3′. To the amplified SPHK1a cDNA, the BamHI site and the EcoRI site were introduced just before the start codon and just after the stop codon of cDNA sequence, respectively, by polymerase chain reaction using primers 5′-CG
Expression and Purification of Maltose Binding Protein (MBP)-tagged Recombinant SPHK1a
The cDNA amplified as described above was subcloned into the BamHI site and EcoRI sites of a bacterial expression vector pMALc2 (New England Biolabs; Beverly, MA). Maltose binding protein (MBP)–SPHK1a was expressed in E. coli strain BL21 (DE3) by incubating with 0.3 mM isopropyl β-thiogalactopyranoside (IPTG) for 5 hr. The cells were pelleted by centrifugation, washed, and homogenized in 50 mM Tris-HCl, pH 7.5, 2 mM MgCl2, 2 mM EGTA, 1% Triton X-100, 2 mM dithiothreitol, 1 mM PMSF, and 20 μg/ml aprotinin. Purification of MBP-SPHK1a was performed according to the manufacturer's recommended procedure.
Expression of FLAG-tagged Recombinant SPHK1a
To express the recombinant SPHK1a in mammalian cells, the mammalian expression plasmid pcDNA3 (Invitrogen; San Diego, CA) was ligated with the oligonucleotides 5′-AGCTTGCCACCATGGATTACAAGGATGACGACGATAAGG-3′ and 5′-GATCCCTTATCGTCGTCATCCTTGTAATCCATGGTGGCA-3′, yielding pcDNA3-FLAG1. Then, PCR-amplified SPHK1a was subcloned into the BamHI site and the EcoRI site of pcDNA3-FLAG1, yielding pcDNA3-FLAG-SPHK1a. A FLAG-tagged SPHK1a was then expressed in CHO cells or COS 7 cells with either stable or transient transfection by LipofectAMINE PLUS (Life Technologies; Gaithersburg, MD) according to the manufacturer's procedure.
Detection of Expressed Murine SPHK1a cDNA
Purified MBP protein from bacterial lysate or crude cellular extracts of FLAG transfectants were separated by SDS-PAGE, transferred to an Immobilon membrane, and then probed with anti-SPHK1a antibody or anti-FLAG M2 antibody (Sigma; St Louis, MO). The signals were detected on an ECL+Plus Western blotting detection system (Amersham-Pharmacia).
Enzyme Activity of SPHK
The enzyme activity of SPHK was measured according to the procedure of Olivera and Spiegel (1993) with minor modifications. Briefly, cells were harvested in a kinase buffer containing 20 mM Tris-HCl, pH 7.4, 20% glycerol, 1 mM β-mercaptoethanol, 1 mM EDTA, 15 mM NaF, 20 mM Na3VO4, 1 mM MgCl2, 0.5 mM 4-deoxypyridoxine, 1 mM PMSF, 10 μg/ml leupeptin, and 10 μg/ml aprotinin. After sonication, the extracts were obtained by centrifugation for 10 min at 14,000 × g, and 50 μg protein was used for enzyme assay. The reaction was started by adding 0.05 μCi [3-3H]-sphingosine (American Radiolabeled Chemicals; St Louis, MO) and 1 mM ATP in a final volume of 50 μl. After incubation at 37C for 30 min, the reaction was terminated with 5 μl of 1 N HCl followed by 200 μl of chloroform:methanol:HCl (1000:200:1, v/v). After vortexing, 62.5 μl of 2 M KCl and 62.5 μl of chloroform were added for phase separation. The resulting chloroform phase was condensed by in vacuo evaporation, applied on a Silica Gel 60 thin-layer plate, and then developed with a solvent of butanol:ethanol:acetic acid:water (80:20:10:20, v/v). Cold authentic sphingosine and Sph-1-P were added to each sample just before chromatography and their position was identified with ninhidrin reagent. The spot corresponding to Sph-1-P was visualized by autoradiography, scraped off the plates, and its radioactivity was measured in a liquid scintillation counter.
Immunostaining
For the immunohistochemical identification of transfected FLAG-tagged SPHK1a, COS 7 cells were transfected transiently using LipofectAMINE PLUS as described above and grown on glass coverslips to 40% confluency. Coverslips were fixed in 3.7% formaldehyde in PBS at 37C for 30 min. Coverslips were rinsed in PBS before cells were permeabilized in PBS + 0.2% Triton X-100 for 5 min. After blocking for 10 min in 1% bovine serum albumin in PBS, coverslips were incubated with either 1 μg/ml of rabbit anti-mouse SPHK1a antibody or 25 μg/ml of mouse anti-FLAG M2 antibody (Sigma) in 1% bovine serum albumin in PBS. After washing coverslips three times with PBS, Alexa 488-conjugated anti-rabbit antibody (7 μg/ml) and Alexa 594-conjugated anti-mouse secondary antibody (7 μg/ml) (Molecular Probes; Eugene, OR) were used to visualize the antigens. Coverslips were washed in PBS three times and in deionized water twice. Then they were mounted with Mowiol 4–88 (Calbiochem; Darmstadt, Germany). For the fluorescent microscopy, a Carl Zeiss Axiophot 2 (Carl Zeiss; Mainz, Germany) equipped with a ×40 objective lens was used. The images were recorded in grayscale by a CCD camera, Mico-MAX-1300YHS (Princeton Instruments; Trenton, NJ) and MetaMorph software (Universal Imaging; West Chester, PA). Merged images were produced in pseudo-color (green and red) artificially with MetaMorph according to the manufacturer's instructions.
For the immunohistochemical examination of human tissues, paraffin blocks stored in the First Department of Pathology, Nagoya University School of Medicine, were used throughout the present study. After removal of paraffin, the section was dipped in 0.01 M EDTA in Tris buffer (pH 8.0) and was microwave-treated for 10 min. After washing three times with PBS, immunostaining was performed using the avidin–biotin–peroxidase complex (ABC) method as described previously (Kato et al. 2000). The final concentration of the primary antibody was 500 ng/ml, and the incubation with the primary antibody was performed overnight at room temperature. For absorption of the antibody with the C-terminal 16 amino acids, which had been used for rabbit immunization, the diluted antibody (500 ng/ml in PBS) was incubated with the synthetic 16 amino acids (1 μg for 1 ml of diluted antibody) for 60 min at room temperature. After brief centrifugation, the supernatant was used as the absorbed antibody. Dako's LSAB kit (Carpinteria, CA) was preferred to the ABC method for HEL cells. Bone marrow smears were obtained from patients with idiopathic thrombocytopenic purpura (ITP) or essential thrombocytopenia (ET). Informed consent was obtained from all patients.
Results
As the first step in the characterization of our polyclonal antibody against C-terminal amino acids (16 residues) of mouse SPHK1a, anti-SPHK1a antibody, homogenates of bacteria or cells transfected with tagged-mouse SPHK1a cDNA expression vectors were analyzed. Figure 1A shows the Western blotting analysis of purified MBP-tagged mouse SPHK1a protein expressed in E. coli BL21 (DE3) and the crude extract of CHO cells transfected with FLAG-tagged mouse SPHK1a cDNA. Bands consistent with the respective molecular weight (49 kD for FLAG-tagged SPHK1a and 83 kD for purified MBP-tagged SPHK1a) were clearly seen with anti-SPHK1a. The extract of mock transfectant (Figure 1A, Lane 4) did not produce any band with anti-SPHK 1a antibody. Coomassie Brilliant Blue staining of the same membrane (Figure 1B) identified this band of 83 kD as the purified MBP-tagged SPHK1a protein (Figure 1B, Lane 2). Figure 1C illustrates the staining pattern with anti-FLAG antibody. It is clear that Lane 3 only showed the band around 49 kD of SPHK1a protein.
(
The specificity of our anti-SPHK1a antibody was examined for tissues of the rat liver, spleen, kidney, and brain. All rat tissues examined by anti-SPHK1a showed a single band at 45–49 kD (Figure 1D). In contrast, human tissue showed a weaker band at around 45 kD (data not shown). Therefore, immunoprecipitation was performed before the Western blotting (Figure 1E). A reactive band was clearly visible at around 44–49 kD with human samples. The band with a molecular weight higher than 49 kD was immunoglobulin, and no other bands were observed. When the sample was incubated with anti-SPHK1a antibody adsorbed with the antigen peptide, the band of 49 kD disappeared (Figure 1E, Lane 5).
Figure 2 provides additional proof that our antibody indeed recognized SPHK1a protein in the immunohistochemical analysis. The FLAG-tagged mouse SPHK1a expression vector was transiently expressed in COS 7 cells. The localization of each antigen (SPHK 1a protein and FLAG tag) was detected with anti-SPHK1a antibody and anti-FLAG antibody, respectively (Figures 2A and 2B). As shown in the merged photos (Figure 2C), both anti-FLAG antibody and anti-SPHK1a antibody could visualize the co-localization of both FLAG tag and mouse SPHK1a proteins in the same cells and the same intracellular localization.
Figure 3A shows the immunostaining of a human leukemia cell line, HEL cells, by anti-SPHK1a antibody. TPA-treated HEL cells (open arrow in the middle of Figure 3A) were stained more strongly in the cytoplasm than were non-treated control cells. This positive staining was diminished by absorbing the antibody with the same peptide that was used for immunization (open arrow on the lower part of Figure 3A). Figures 3B and 3C show the simultaneous increase in SPHK enzyme activity in HEL cells treated with TPA compared with the activity in control cells.
We analyzed cell-specific localization of SPHK1a in human tissue using this antibody. As shown in Figure 4, anti-SPHK1a stained almost all human tissues including the brain, kidney, lung, liver, spleen, intestine, testis, and bone marrow. In bone marrow, megakaryocytes and platelets were strongly positive (Figures 4A and 4C). Neither myeloid nor erythroid cells in the bone marrow showed positive staining. Some megakaryocytes exhibited similar intense staining in the margin of the cytoplasm (data not shown), suggesting the release of platelets into the extracellular space or their preferential localization in the membrane area. This positive staining of megakaryocytes disappeared with the absorption of anti-SPHK1a antibody with the original 16 amino acids used for the immunization (arrow in Figure 4B).
Immunohistochemical detection of expressed mouse SPHK1a protein and FLAG tag in transiently transfected COS 7 cells. COS cells were transiently transfected with FLAG-tagged SPHK1a expression vector as described in Materials and Methods. After fixation, cells on the coverslip were stained with either anti-SPHK1a antibody or anti-FLAG antibody followed by the respective second antibody as described in Materials and Methods. (
In the cerebral cortex, white matter was diffusely immunostained (Figure 4D). In the midbrain, only the red nucleus and cerebral peduncle were positive (Figure 4F). In the cerebellum, the white matter was positive, and the gray matter was negative (Figure 4H). Higher magnifications (Figures 4E, 4G, and 4I) show that the axon areas were strongly positive. In the kidney, uriniferous tubules were strongly positive, but Bowman's capsule was negative (Figure 4J). In the lung, endothelial cells of the vessels and submucosal bronchial glands were moderately positive, whereas alveoli were negative (Figure 4K). In the spleen, endothelial cells and splenic cords were positive but most lymphocytes in the follicle were negative (Figure 4L). In the liver, the lining cells of the sinusoids were positive, but hepatocytes were negative (Figure 4M). In the testis, endothelial cells and smooth muscle cells of the vessels were positive, and interstitial tissue, such as Leydig's cells, was weakly positive (Figure 4N). In the intestine, columnar epithelia were weakly positive (Figure 4O).
(
Discussion
Our polyclonal antibody against C-terminal amino acids (16 residues) of mouse SPHK1a, anti-SPHK1a antibody, was assessed to be reactive for the recombinant mouse SPHK1a. Using a tagged mouse SPHK1a cDNA expression vector, we tried to detect the MBP-tagged mouse SPHK1a protein of transfected E. coli BL21 and FLAG-tagged mouse SPHK 1a protein of transfected CHO cells in Western blotting. Anti-SPHK1a antibody could detect respective bands with the consistent molecular weight. In Figure 1A, a band with an approximate molecular weight of 200 kD observed in Lane 1 of MBP-tagged SPHK1a may be an aggregation of proteins, but this remains to be determined. Anti-FLAG antibody could detect the band around 49 kD (Figure 1C) as well as a band with the molecular weight of 32.5 kD. The latter band appears to be nonspecific because it was observed in both FLAG-tagged SPHK1a transfectant and FLAG-tagged mock transfectant.
We analyzed rat tissue homogenates to assess the reactivity of anti-SPHK1a antibody because the full cDNA sequences of rat and mouse are very similar (Kohama T, Sankyo Pharmaceutical Co. Tokyo, Japan, personal communication). Under our experimental conditions, a single band with the molecular weight of 45–49 kD was observed. However, it is reported that mouse SPHK1 has two isoforms, SPHK1a and 1b. The difference between them is only 7 amino acids located at its N-terminal (Kohama et al. 1998). It is not clear why we could not observe two bands in our rat tissue samples. The SPHK enzyme activity in rat tissue reported by Olivera et al. (1998) showed that the enzyme activity in rat kidney and spleen is higher than in rat liver and brain. Our Western blotting also shows that SPHK1a protein content is higher in rat kidney and spleen than that in liver. Of course, further quantitative analysis is necessary.
Human tissue homogenates were used to prove the crossreactivity of our anti-SPHK1a antibody. Immunoprecipitation followed by Western blotting was effective in detecting the band at around 44–49 kD with human samples. Prior incubation of anti-SPHK1a antibody with the antigen peptide erased the band at around 44–49 kD, supporting the identity of this band as SPHK1a. From these results, it was concluded that anti-SPHK1a specifically reacted with SPHK1a in mouse, rat, and human tissues.
During preparation of this manuscript, the cDNA sequence of human SPHK1a was reported (Melendez et al. 2000). According to that report, human SPHK1a cDNA has six completely matched amino acids (PPPEEP) in the mouse C-terminal peptide for production of anti-SPHK1a antibody. Our anti-SPHK1a is beleived to recognize this sequence. In checking the human protein database, we can not completely rule out the possibility that several proteins with these matching amino acids (Bat 2 protein; S36152,
It is reported that TPA-treated HEL cells are induced to a megakaryocytic lineage (Murate et al. 1993). HEL cells have also been reported to increase SPHK activity by TPA treatment (Buehrer et al. 1996). We previously reported that the appearance of large and multinucleated cells among TPA-treated MEG-01 cells, a human megakaryoblastic leukemia cell line, indicates the most advanced stage of this differentiation model (Murate et al. 1991). We observed similar large and multinucleated cells in TPA-treated HEL cells that were immunostained strongly with anti-SPHK1a antibody. These results suggest that SPHK1a is a novel differentiation marker for megakaryocytic lineage. Furthermore, anti-SPHK1a is useful for analysis of human SPHK1a in various differentiation models in vitro.
SPHK activity is reportedly high in the testis, intestine, kidney (Gijsbers et al. 1999), and platelets (Banno et al. 1998), and the content of Sph-1-P is high in the brain, spleen, testis, and intestine (Yatomi et al. 1997). The partial discrepancy observed between these reports may be due to the different methods employed (Yatomi et al. 1997; Edsall and Spiegel 1999). Therefore, we have examined the cell-specific expression of SPHK in human tissues using an immunohistochemical method. To our knowledge, this is the first report that describes the cell type-specific distribution of SPHK1a in various human tissues. According to Northern blotting data (Melendez et al. 2000), the brain, heart, spleen, kidney, and lung show intense SPHK1a gene expression.
The most remarkable finding in the present study is the localized distribution of SPHK1a in the central nervous system. The diffuse and dense distribution of SPHK1a in the axons (white matter of the cerebrum and cerebellum as well as the red nucleus and cerebral peduncle in the midbrain) is of particular interest. It suggests that SPHK1a is concentrated in axons but not in neurons overall and that the possible increase of Sph-1-P (the product of SPHK) in these axons might play roles either in signal transduction of the nervous system or in preventing apoptosis of the axon. It is also interesting that hepatocytes in the liver and lymphocytes in the spleen were not immunostained by anti-SPHK1a, although SPHK activity has been reported in their whole homogenates (Yatomi et al. 1997). This implies that immunohistochemical analysis is essential for defining the cell type-specific distribution of SPHK1a.
Very recently, several isoforms of SPHK have been reported (Liu et al. 2000; Melendez et al. 2000). In human platelets exhibiting high SPHK activity, the protein immunoprecipitated by anti-SPHK1a showed a similar molecular weight to that in the human kidney (Figure 1C), but considerable SPHK activity was found in the non-precipitated soluble fraction (data not shown). This result may indicate that a different type(s) of SPHK is present in human platelets. Hepatocytes and lymphocytes may also have other SPHK isoforms than SPHK1a. At present, the significance of the presence of SPHK isoforms is not known.
The physiological significance of SPHK1a will be explained finally by future knockout mouse experiments. However, it is noteworthy that endothelial cells (and/or smooth muscle cells of the vessels) were immunostained with anti-SPHK1a in almost all human tissues examined. FCS or PDGF is reported to increase Sph-1-P by increasing SPHK activity (Olivera and Spiegel 1993; Olivera et al. 1999a). Increased SPHK promotes cell growth and survival (Olivera et al. 1999b), and endothelial cells are located very close to secreted PDGF from activated platelets. Therefore, it is likely that Sph-1-P produced by SPHK1a in endothelial cells plays a role in tissue remodeling and, in some cases, in pathological conditions such as atherosclerosis. Similarly, brain and megakaryocytes/platelets, which differ in function and in their differentiation mechanism, showed similar strong immunoreactivity against anti-SPHK1a, suggesting diverse functions of SPHK1a.
Immunological detection of SPHK1a in various human tissues. Using anti-SPHK1a antibody, immunohistochemistry was performed as described in Materials and Methods. (
Taken together, the findings of present study provide basic necessary information for understanding the roles of SPHK1a in cell responses. This approach is also promising as a potentially useful method to elucidate the pathogenesis of various diseases.
Footnotes
Acknowledgements
We gratefully thank Drs C. Inoue and T. Kamiya (Aichi Red Cross Center) for providing the outdated platelets. We also thank Dr T. Kinoshita (First Department of Internal Medicine, Nagoya University School of Medicine), Dr M. Suzuki (Disease Mechanism and Control, Nagoya University School of Medicine), Mr T. Yamada, and Mr Y. Nakade (Nagoya University School of Health Science) for helpful discussion and technical advice.