Transcription factor specificity protein 1 (Sp1) is the main regulator of nerve growth factor-induced sphingosine kinase 1 gene expression of the rat pheochromocytoma cell line, PC12

Animals, cell line and reagents

Wistar rats were used to prepare tissue RNA and protein samples. The rat pheochromocytoma cell line, PC12, was cultured in 10% fetal calf serum in Dulbecco's Modified Eagle's medium (Sigma, St. Louis, MO, USA) at 37°C in 5% CO₂ (Yoshimura et al. 1998). Rabbit anti-mouse SPHK1 antibody was prepared as described previously (Murate et al. 2001). NGF was purchased from Sigma, and K252a from Alomone Laboratories (Jerusalem, Israel). pBluescriptII KS(+) vector was purchased from Invitrogen (Carlsberg, CA, USA). pGL3 basic vector for luciferase assay was from Promega (Madison, WI, USA). Sp1 expression vector was kindly provided by Dr G. Suske (Institut für Molekularbiologie und Tumorforshung Phillipps-Universitat Marburg, Marburg, Germany). Anti-Sp1 (SC-59), anti-activator protein 2 (AP-2) antibodies (SC-184) and normal rabbit IgG were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA), and anti-β-actin antibody (#AANO2) from Cytoskeleton, Inc. (Denver, CO, USA).

Sphingosine kinase enzyme activity

The enzyme activity of SPHK was measured according to the procedure of Olivera et al. 1999 with minor modifications (Murate et al. 2001). Briefly, 50 µg protein was used for the enzyme assay. The reaction mixture (200 µL) consisted of 20 mm Tris-HCl, pH 7.4, 2.5 mm MgCl₂, 0.25 mm EDTA, 5% glycerol, 1.2 mm dithiothreitol, 1 mm Na₃VO₄, and 15 mm NaF, 20 µm sphingosine with 0.25% Triton X-100, and 50 µg enzyme. The reaction (30 min at 37°C) was started by the addition of 20 µL [γ-³²P]ATP (2 µCi) and 10 mm ATP complex, and terminated by the addition of 20 µL 1 N HCl, followed by 0.8 mL of chloroform/methanol/HCl (100:200:1). After vigorous vortexing, 240 µL chloroform and 240 µL 1 m KCl were added, and the suspension centrifuged at 2000 g for 10 min at room temperature. The organic phase containing the lipids was removed, concentrated, and applied to a silica gel 60 TLC plate, which was developed in butan-1-ol/acetic acid/water (3:1:1, v/v). The S1P spot was visualized by autoradiography, and quantified using a densitometer (Densitograph Atto, Tokyo, Japan).

Western blotting

Western blotting was performed using anti-SPHK1 antibody (Murate et al. 2001) and anti-Sp1 antibody, respectively. Five or 10 µg of protein were separated by 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis and electro-transferred onto Hybond-P membranes (Amersham Pharmacia Biotech, Little Chalfont, Buckinghamshire, UK). A positive signal was detected with an ECL Plus western blotting system (Amersham Pharmacia Biotech).

Semi-quantitative reverse transcription–polymerase chain reaction

For RNA preparation, cells were harvested, and total RNA was extracted using an RNA STAT-60 kit (Tel-Test, Inc., Friendswood, TX, USA). Semi-quantitative RT–PCR was performed according to the method described previously (Nakade et al. 2003). The first-strand cDNA was prepared with 5 µg of RNA using the Super Script First-Strand System (Invitrogen). In the preliminary experiments, the relative amounts of cDNA and the range of PCR cycles that permit the linear amplification of SPHK1 and β-actin were determined. The primer sequences were: SPHK1 exon 1a, sense 5′-ACCAAGGCATGTATTGCAGTGACGC-3′, antisense 5′-CAGTGTGCAGCTGCTTATCGGTGTT-3′; SPHK1 exon 1d, sense 5′-AGTTGGTGAGGATCGTGG-3′, antisense 5′-AGCTGCTTATCGGTGTTG-3′; and β-actin, sense 5′-TTCTACAATGAGCTGCGTGTGGC-3′, antisense 5′-CTCATAGCTCTTCTCCAGGGAGGA-3′. The PCR conditions for rat SPHK1 (rSPHK1) were 96°C for 15 s, 59°C for 30 s, followed by 72°C for 30 s. The number of cycles was 36, 38 and 40. The PCR conditions for β-actin were 96°C for 2 min followed by 19, 21 and 23 cycles at 96°C for 15 s, 58°C for 30 s, and 72°C for 30 s. Using the NIH image (version 6), each band was surveyed for its relative band intensity. Within the range of the linear PCR amplification, the relative expression of the SPHK1 message was evaluated by calculating the band intensity ratio of SPHK1/β-actin.

Rapid amplification of 5′ cDNA ends

RNA ligase-mediated rapid amplification of 5′-cDNA ends (5′-RACE) was performed according to the instruction manual with a GeneRacer™ kit (Invitrogen) to determine the transcription initiation point of rSPHK1 in PC12 cells. The reverse SPHK1 gene-specific primer and reverse SPHK1 gene-specific nested primer were 5′-GGCGCTGCTAGAGGTAACGG-3′, and 5′-TCGCCCTCTCCACCTCCCT-3′, respectively. Cloned PCR products were analyzed for their DNA sequences and the 5′-transcription initiation point was determined.

Transfection and luciferase assay

The vector containing rat SPHK1 gene and its 5′ promoter region was the kind gift of Dr K. Shiota (Tokyo University, Tokyo, Japan). The 877 bp (BglII and PstI digested) fragment derived from this original vector was introduced to the BamHI and PstI sites of pBluescriptII KS(+) vector using the Ligation high kit (Toyobo, Osaka, Japan). The introduced DNA fragment was digested by MluI (in rSPHK1 genome) and HindIII [in pBluescriptII KS(+) vector], and again introduced to the MluI and HindIII sites of pGL3 basic vector. This construct was tentatively named rSPHK exon 1d/−604 (+1 denotes the transcription initiation point of exon 1d). A further 5′ DNA fragment was added to this construct to form rSPHK exon 1d/−3942. The required DNA fragment of the 5′ promoter region was digested by KpnI and MluI, and ligated to the KpnI and MluI sites of rSPHK exon 1d/−604. This construct was named rSPHK exon 1d/−3942. The deletion mutants were prepared by the conventional enzyme-based digestion method. rSPHK exon 1d/−3942 was digested by either KpnI/StuI or KpnI/BstEII. Each fragment was blunted and self-ligated. The final constructs were designated as rSPHK exon 1d/−2008 and rSPHK exon 1d/−294, respectively. To circumvent several putative transcription factor-binding motifs within this region, further deletion mutants were constructed using PCR-based methods. According to the DNA sequence of the 5′ promoter region of rat SPHK1 exon 1d, the following primer sets were prepared. The upper primer for producing rSPHK exon 1d/−119 was 5′-GGGGGTACCGCATTTTCACGCGGAGTT-3′, and that for producing rSPHK exon 1d/−64 was 5′-GGGGGTACCATGCTAGCCTCTCCAATCCC-3′. A single underline denotes the KpnI enzyme site. The lower primer was GL primer 2 of the pGL3 basic vector (Promega). The PCR products were digested by KpnI and NcoI, purified, and ligated to the pGL3 basic vector to produce, respectively, the truncated promoters, rSPHK exon 1d/−119 (including AP-2 and 2 Sp1 binding motifs) and rSPHK exon 1d/−64 (deleted AP-2 and 2 Sp1 binding motifs). To introduce a mutation to the AP-2 and Sp1 binding motifs, the following primer sets were prepared, and PCR was performed using rSPHK exon 1d/−119 as the template. The upper primer for the mutated AP-2-binding motif was 5′-GGGGTACCGCATTTTCACGCGGAGTTCGCAATATTAGTGGGGAGCGGTG-3′, that for the mutated distal Sp1-binding motif was 5′-GGGGGTACCGCATTTTCACGCGGAGTTCGCGCCCGAATTAGTGAATTGCGGT-3′, and that for mutated proximal Sp1-binding motif was 5′-GGGGGTACCGCATTTTCACGCGGAGTTCGCCCGAAGGGGTGGGGAGCGGTGGGACTATTATTACATTGCTAGCCT-3′. A single underline denotes the KpnI enzyme site, and a double underline indicates the mutated sequence.

In order to show that the promoter region containing AP-2 and two Sp1 motif can confer NGF responsiveness to a heterologous promoter, a fragment of human acid sphingomyelinase (hASMase) promoter (Murate et al. 2002) was used as the partner. The luciferase vector containing the promoter region (−219 bp from the transcription start point to +112 bp) of hASMase was combined to the rSPHK1 promoter by PCR based method. For the convenience of ligation, the sequences of the upstream primers used for generating the hybrid promoter of hAMSase and rSPHK1 are: hASMase (−219/+112), 5′-GGGCTCGAGCGCAGCGTTGACAGCCGCCCGCCACC-3′; rSPHK exon 1d (−280/−33), 5′-GGGGGTACCCCTGGCGGCTTCTTTTTGTCC-3′. The downstream primers are: hASMase (−219/+112), 5′-GGGAAGCTTCCAGCGGCCAGGCCCATCCAAAGG-3′; rSPHK exon 1d (−280/−33), 5′-GGGCTCGAGGGGGCTCTCATCGGGATTGG-3′. A single, double and dotted underline denote the added KpnI enzyme site, the XhoI enzyme site, and the HindIII enzyme site, respectively. Each PCR product was subcloned into the pBluescriptII KS(+) vector, and constructs were digested by XhoI/HindIII [hASMase in pBluescriptII KS(+) vector] or KpnI/XhoI [rSPHK exon 1d in pBluescriptII KS(+) vector]. Both hASMase fragment and rSPHK exon 1d promoter fragment were ligated to the KpnI and HindIII sites of pGL3 basic vector to make the hybrid promoter, or the hASMase fragment was ligated to the XhoI and HindIII sites of pGL3 basic vector to make the control hASMase reporter.

The DNA transfection was performed by the calcium precipitation method as described by Wilson et al. 1993. Briefly, 5 µg luc vector and 2 µg β-galactosidase expression vector were cotransfected to PC12 cells. After 8 h, glycerol shock was induced. The cells were transferred to the complete medium with/without 100 ng/mL NGF for 20 h. Luciferase and β-galactosidase activities were measured, and the relative promoter activity was expressed as luc/β-galactosidase activity. When examining the effect of K252a, it was added 1 h before NGF stimulation.

Electrophoresis mobility shift assay

Nuclear extract was prepared from PC12 cells treated with/without 100 ng/mL NGF for the times indicated. Cells were subjected to a preparation of nuclear extract using NE-PER^® nuclear and cytoplasmic extraction reagents (Pierce, Rockford, IL, USA). For the preparation of Sp1 over-expressed nuclear extract, PC12 cells were transfected with 20 µg of Sp1 expression vector (as described above) by the calcium precipitation method. After 24 h, nuclear extracts were prepared similarly. Biotin-labeled probes containing the rat SPHK1 exon 1d type promoter fragment were synthesized and annealed. Electrophoresis mobility shift assay (EMSA) was performed with 2 µg nuclear extract and biotin-labeled double strand DNA probes (50 fmol) with a LightShift™ Chemiluminescent EMSA kit (Pierce) according to the manufacturer's protocol. Binding reactions were loaded onto a 6% polyacrylamide gel (8 × 8 × 0.1 cm) in 0.5 × TBE buffer (0.089 M Tris-borate, 0.089 M boric acid, 10 mM EDTA) and electrophoresed at 100 V for 3.5 h. Biotin-labeled double stranded DNAs were transferred to a positive-charge nylon membrane (Hybond-N, Amersham Pharmacia Biotech). Biotin-labeled DNA was then integrated with streptavidin–horseradish peroxidase conjugate. In some experiments, probes with mutations or deletions of the putative transcription factor binding motif were used. The sense oligomer of the AP-2 mutated probe was 5′-GTTCGCAATTATTAGTGGGGAGCGGTGGGACCACGCCCCCATGCT-3′, that of the AP-2 deleted probe was 5′- GCGGAGTTGTTGGGGTGGGGAGCGGTGGGACCACGCCCCCATGCT-3′, that of the distal Sp1 mutation was 5′-GTTCGCCCGAATTAGTGAATTGCGGTGGGACCACGCCCCCATGCT-3′, that of the distal Sp1 deletion was 5′-GCGGAGTTGTTCGCCCGAAGAGCGGTGGGACCACGCCCCCATGCT-3′, that of the proximal Sp1 mutation was 5′-GTTCGCCCGAAGGGGTGGGGAGCGGTGGGACTATTATTACATGCT-3′, and that of the proximal Sp1 deletion was 5′- GTTCGCCCGAAGGGGTGGGGAGCGGTGGGGCTAGCCTCTCCAATC-3′. A double underline denotes the mutated sequence. For the supershift experiment, anti-AP-2 or anti-Sp1 antibody was used. The antibody was mixed with nuclear extract for 15 min at room temperature before mixing with the labeled probe.

Chromatin immunoprecipitation assay

PC12 cells were crosslinked for 10 min by adding formaldehyde directly to the cell culture medium to a final concentration of 1%. Cross-linking was stopped to a final concentration of 0.125 m Glycine. Crosslinked cells were washed twice with phosphate-buffered saline, scraped and resuspended in the swelling buffer (10 mm Tris-HCl, PH 7.4, 3 mm CaCl₂, 0.1% Nonidet P-40 (v/v), and protease inhibitors). After a 10-min incubation on ice, cells were vortexed and nuclei were collected by centrifugation. Isolated nuclei were resuspended in lysis buffer (50 mm Tris-HCl, pH 8.0, 1% sodium dodecyl sulfate, 10 mm EDTA, and protease inhibitors) and sonicated using eight pulses of 30 s. After centrifugation, the supernatant was diluted 10-fold with dilution buffer (16.7 mm Tris-HCl, pH 8.0, 167 mm NaCl, 1.1% Triton X-100, and protease inhibitors) and precleared with protein A-Sepharose 4B (Upstate Biotechnology, Lake Placid, NY, USA) containing 0.5 mg/mL of salmon sperm DNA. After 10 µL of protein A-Sepharose (50% slurry containing 0.5 mg/mL salmon sperm DNA) was then added to the solution again, the supernatant was divided into three aliquots. Nothing was added to one aliquot, and to the others, normal rabbit IgG or Sp1-specific antibody (final concentration: 1 ng/µL) was added and incubated 4 h at 4°C on a rotating wheel. The beads were pelleted and washed sequentially in the following buffers: buffer A [50 mm HEPES, pH 7.9, 140 mm NaCl, 0.1% (w/v) deoxycholic acid (sodium salt), 1 mm EDTA, 1% (v/v) Triton X-100 and protease inhibitors]; buffer B [50 mm HEPES, pH 7.9, 500 mm NaCl, 0.1% (w/v) deoxycholic acid (sodium salt), 1 mm EDTA, 1% (v/v) Triton X-100, and protease inhibitors], and TE buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8.0) (twice). Immuno-complexes were extracted from the beads with elusion buffer [50 mm Tris-HCl, pH 8.0, 10 mm EDTA, 1% (w/v) sodium dodecyl sulfate]. Cross-linking was reversed by heating the elutes at 65°C overnight. The elutes were then digested with proteinase K at 50°C for 5 h. The solutions were extracted with phenol/chloroform/isoamyl alcohol. DNA was purified by ethanol precipitation and resuspended in 30 µL of TE buffer. The SPHK1 promoter region was amplified by PCR using primers 5′-CCTGGCGGCTTCTTTTTGTCC-3′ (forward) and 5′-GGGGCTCTCATCGGGATTGG-3′ (reverse).

Rat brain expresses sphingosine kinase 1 protein

The protein level of SPHK1 in rat brain was analyzed by western blotting (Fig. 1a). Rat cerebrum, midbrain and cerebellum showed higher levels of SPHK1 protein. Heart and kidney also expressed a significant but lower level of SPHK1 protein with the longer exposure of the same membrane. This result is slightly different from that based on the mouse data reported by Fukuda et al. (2003), which showed a modest SPHK1 expression in mouse heart and kidney as well as in brain.

Rat brain and PC12 cells utilizes exon 1d more than exon 1a

The rat SPHK1 gene structure was published recently (Imamura et al. 2001). RT–PCR was applied to total RNA from the cerebrum, midbrain and cerebellum to detect the transcript of SPHK1 first exon. For the analysis of gene transcription level, we used the semiquantitative RT–PCR method as reported previously (Nakade et al. 2003). In the preliminary experiments, the relative amount of mRNA was checked by serially amplifying the β-actin message, after which the template volume for the RT–PCR of SPHK1 was adjusted between samples. Figure 1(b) shows that rat cerebrum, midbrain and cerebellum utilized exon 1d more than exon 1a. The other exons (1c, 1b, 1e, and 1f) were not detected by the present RT–PCR experiments (data not shown). PC12 cells also utilized exon 1d as the major first exon as shown in Fig. 2(d), whereas the signal of exon 1a was not detected, even after NGF stimulation under our experimental conditions (Fig. 2d).

Nerve growth factor stimulates sphingosine kinase 1 gene expression in PC12 cells

NGF was reported to stimulate SPHK enzyme activity (Ruis et al. 1997). In the present study, NGF increased SPHK enzyme activity (Fig. 2a) as well as the SPHK1 protein level to about twice those in the control (Fig. 2b). It was shown that NGF induced an increase in SPHK1 gene expression (exon 1d type) in PC12 cells 3.6 times that in the control (Fig. 2d).

Determination of transcription initiation points of exon 1

Using the 5′-RACE method, we examined the transcription initiation points of exon 1d (Fig. 2c) of PC12 cells and found three points, i.e. −1759, −1726, and −1696 from the first ATG in the second exon, which were very close to the reported initiating points available online. Other initiation points around exon 1a located upstream of exon 1d were not found.

Nerve growth factor-induced sphingosine kinase 1 exon 1d promoter activity is blocked by TrkA inhibitor

The TrkA (high affinity-NGF receptor) specific inhibitor, K252a, suppressed SPHK1 message expression induced with NGF (Fig. 3a). Using the DNA fragment of the 5′ promoter region of rat SPHK1, we prepared luciferase vectors with inserts of various lengths (exon 1d/−3943, exon 1d/−2008, exon 1d/−604, exon 1d/−294, exon 1d/−119 and exon 1d/−64, in Fig. 4 as explained in the Materials and methods). Figure 3(b) shows that NGF increased SPHK1 exon 1d type promoter activity (exon 1d/−604 as shown in Fig. 4 left) more than three times as compared with non-NGF treated exon 1d/−604, which was consistent with the increase in the message level. It was shown that inhibition of the NGF receptor suppressed NGF-induced promoter activity, suggesting that the NGF-induced SPHK1 gene expression was located downstream of TrkA receptor signaling.

Determination of nerve growth factor-responsive promoter region

We tried to determine the 5′ promoter region of SPHK1 exon 1d responsible for NGF stimulation in PC12 cells. About a 4-kb length 5′ of the exon 1d was analyzed. As shown in Fig. 4(a), a 55-bp length (−119/−64) was sufficient (7.69-fold increase) for NGF-induced exon 1d transcription as compared to NGF-induced exon 1d/−64 luciferase vector. The relative ratio of luc activity (promoter activity) between NGF(+)1d/−119 luc/NGF(–)1d/−119 luc was 2.3 and was lower than that of 1d/−294 luc (8.6). However, the decrease of the ratio is due to the high luciferase activity of NGF(–) 1d/−119 luc, and the luciferase activity of NGF(+)1d/−119 itself was comparable to NGF(+)1d/−294 luc. So, we focused on this region. The decrease of promoter activity was observed in NGF-treated 1d/−604/luc vector. The presence of suppressor elements between −604 and −294 is the possible answer; however, the reason still remains to be determined. Available online data of the putative binding motif of transcription factor, TFSEARCH (http://www.cbrc.jp/research/db/TFSEARCH.html), demonstrated three putative boxes (one AP-2, and two Sp1 boxes) within this 55-bp length. By introducing a mutation into the respective transcription factor binding motifs, our analysis revealed that both the AP-2 and Sp1 motifs regulated the exon 1d expression, and that the proximal Sp1 binding motif was the prerequisite promoter motif (Fig. 5). Our experiment using the hybrid of hASMase promoter and rSPHK1 promoter (containing this 55-bp region) showed that the addition of the rat SPHK1 promoter region containing AP-2 motif and Sp1 motifs could also confer the NGF responsiveness to hASMase promoter (Fig. 4b).

Nerve growth factor increases specificity protein 1 (Sp1) protein binding to the proximal Sp1-like motif

Figure 6(a) illustrates the oligonucleotides used for EMSA experiments. Figure 6(b) shows that our EMSA probe could detect two bands (a and b), both of which were increased time-dependently after NGF treatment. Band a was stronger than band b. Cold competitor greatly reduced bands a and b, suggesting their specificity (Fig. 6c). Figure 7(a) showed the mutated or deleted oligonucleotides used for further analysis. Mutation or deletion of the proximal Sp1 motif but not the distal Sp1 or AP-2 motifs diminished bands a and b in NGF-stimulated nuclear extract (Fig. 7b). This suggested that both bands were derived from the proximal Sp1-like motif but not from the distal Sp1-like motif (which slightly overlapped with AP-2 motif) or the AP-2 motif in NGF-induced SPHK1 gene expression. Anti-AP-2 antibody did not change the band pattern, whereas the addition of anti-Sp1 antibody greatly reduced band a intensity and diminished band b (Fig. 8a). We also analyzed changes in Sp1 protein levels during NGF stimulation. It was observed that the Sp1 level increased during the observation period (Fig. 8b).

Sp1 over-expressed nuclear extract showed an increase in both bands a and b, thus confirming the involvement of Sp1 protein (Fig. 8a, lanes 6 and 7). CHIP assay shown in Fig. 8(c) demonstrated that anti-Sp1 antibody but not normal IgG could precipitate the complex between Sp1 protein and the rSHPK1 promoter, suggesting that Sp1 protein did bind with the region containing AP-2 and Sp1 in NGF-stimulated PC12 cells in vivo.