Báo cáo khoa học: The effect of ST2 gene product on anchorage-independent growth of a glioblastoma cell line, T98G

doi:10.1046/j.1432-1033.2003.03377.x

Eur. J. Biochem. 270, 163–170 (2003) (cid:1) FEBS 2003

The effect of ST2 gene product on anchorage-independent growth of a glioblastoma cell line, T98G

Yasushi Haga1, Ken Yanagisawa2, Hiromi Ohto-Ozaki2, Shin-ichi Tominaga2, Toshio Masuzawa1 and Hiroyuki Iwahana2 1Department of Surgical Neurology and 2Department of Biochemistry, Jichi Medical School, Minamikawachi-machi, Kawachi-gun, Tochigi, Japan

the results suggest that

continuously produce and secrete the ST2 protein, in order to study the effect of the ST2 protein on malignancy. Although we could not detect a remarkable difference in proliferation between transformants and control cells in conventional tissue culture dishes, the efficiency of colony formation in soft agar was significantly decreased in the case of cells that continuously produce the ST2 protein. Furthermore, inhibition of colony formation in soft agar was observed in wild-type T98G cells when purified soluble ST2 protein was added to the culture, in a dose-dependent the manner. Taken together, expression of ST2 suppressed the anchorage-independent growth and malignancy.

Keywords: ST2; glioblastoma; anchorage-independent growth; IL-1 receptor family; malignancy. The ST2 gene, which is specifically induced by growth sti- mulation in fibroblasts, encodes interleukin-1 receptor-rela- ted proteins and is widely expressed in hematopoietic, helper T, and various cancer cells. However, the physiological as well as pathological functions of the ST2 gene products are not yet fully understood. In this study, we analyzed the expression of the ST2 gene in human glioma cell lines and human brain tumor samples with real-time polymerase chain reaction method, the results of which revealed that the expression level of the ST2 gene in glioma cell lines and glioblastoma samples is significantly lower than that in a fibroblastic cell line, TM12, and benign brain tumors, sug- gesting the reverse relationship between malignancy and ST2 expression. As we could not detect the soluble ST2 protein in the culture fluid of the T98G glioblastic cell line by ELISA, we established stable transformants of T98G that

which are far apart from each other (for example, they are 25.4 kb apart in the case of human genes) [9]. Differential usage of the two distinct promoters by cell type may be a special means of regulation [10]. However, although the structures of the ST2 gene products are very similar to that of IL-1 receptor (IL-1R), these products never bind to IL-1a, IL-1b, or receptor antagonist [11]. The ligands for the receptor molecule are still unknown, thus leaving it as an orphan receptor system.

Correspondence to S.-i. Tominaga, Department of Biochemistry, Jichi Medical School, 3311-1 Yakushiji, Minamikawachi-machi, Kawachi-gun, Tochigi 329-0498, Japan. Fax: + 81 285 44 2158, Tel.: + 81 285 58 7323, E-mail: shintomi@jichi.ac.jp Abbreviations: IL, interleukin; MEM, minimum essential medium; fetal bovine serum, fetal bovine serum; DMEM, Dulbecco’s modified Eagle’s medium; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; ELISA, enzyme-linked immunosorbent assay; HRP, horseradish peroxidase; FITC, fluorescein isothiocyanate; TNF, tumor necrosis factor. (Received 4 September 2002, revised 12 November 2002, accepted 20 November 2002)

The ST2 gene, also known as T1, Fit-1, and DER4, was originally found as a gene induced by growth stimulation (hence the name ST2) in a murine fibroblastic cell line, BALB/c-3T3 [1–6]. The subsequent structural analysis of the ST2 protein, deduced from ST2 cDNA, revealed that it was a soluble secreted protein very similar to the extracel- lular portion of the interleukin (IL)-1 receptor [2]. To date, we have identified at least three ST2 gene products, generated by alternative splicing mechanisms. These prod- ucts are ST2 (soluble secreted form), ST2L (transmembrane receptor form), and ST2V (variant form of ST2) [2,7,8]. The gene is also interesting in that it has two distinct noncoding exon 1 regions and consequently two distinct promoters,

A research breakthrough revealed that the ST2 gene products were specifically expressed in type 2 helper T (Th2) lymphocytes and not in Th1 cells [11–13]. The evidence of suppression in eosinophilia by administrating anti-ST2 Ig or modified soluble ST2 protein in asthma model mice [13] was followed by discovery of the fact that the ST2 concentration in serum of patients suffering from asthma attacks was significantly higher than that in controls [14,15]. Elevated serum ST2 was also detected in various autoimmune diseases, such as systemic lupus erythematosus [16], suggest- ing a significant relationship between ST2 and immunolo- gical reactions. However, there is accumulating evidence that the ST2 gene is expressed by various cancer cell lines, such as those of hematological neoplasms [11,17,18], breast cancer [19], and colon cancer (Tago, K. & Tominaga, S., unpub- lished results). Furthermore, elevated ST2 protein concen- tration in pleural effusions of lung cancer imply a relationship between cancer and immunological responses [20]. Therefore, the investigation of ST2 should be widely based on both immunology and growth regulation.

164 Y. Haga et al. (Eur. J. Biochem. 270)

(cid:1) FEBS 2003

meningothelial meningiomas. The histological diagnosis was confirmed by pathologists using portions of the original tumor tissue.

RT-PCR

Although the brain has been considered to be an immunologically privileged site, there is accumulating evidence that the glia and brain tumors express several cytokines [21–28]. Among them, IL-1 has been shown to play an important role in the growth of glia [29,30]. The gene expression of IL-1b and IL-6 has been shown in some gliomas [24–28]. Up-regulation of IL-1 receptor expression has been observed in human glioblastoma cell lines after incubation with glucocorticoid [24]. Furthermore, the serum IL-1b levels were higher after radiation than those before treatment in pediatric patients with astrocytoma [25]. The expression of IL-6, which enhances the activity of natural killer cells and cytotoxic T lymphocytes, has also been shown to be induced in some glioblastoma cell lines after treatment with IL-1b [28]. Taken together, cytokines and their receptors are speculated to be regulating cell prolifer- ation in the brain, but the mechanisms of their action are still unclear. To synthesize the first-strand cDNA, 5 lg of total RNA extracted from cells using ISOGEN (Nippon Gene, Tokyo, Japan) was denatured with 4 lM of random hexamer (Takara, Tokyo, Japan) at 70 (cid:2)C for 10 min and immedi- ately chilled on ice. Next, the denatured RNA was reverse- transcribed with 10 lM each dNTPs, 200 U RNase inhibitor (Toyobo, Osaka, Japan), and 200 U M-MLV Reverse Transcriptase (Gibco BRL, Grand Island, NY, USA) in a buffer containing 50 mM Tris/HCl (pH 8.3), 75 mM KCl, 3 mM MgCl2, and 10 mM dithiothreitol in a total volume of 25 lL at 37 (cid:2)C for 60 min. Then, the reaction was terminated at 70 (cid:2)C for 10 min.

The expression of cytokines and their receptors in the brain prompted us to investigate ST2 expression and its possible implications in brain tumors. Here we report the mode of expression of the ST2 gene in malignant glioma cell lines and brain tumor samples. Stable transformants of the glioblastoma cell line T98G with cDNA for the ST2 protein revealed that ST2 suppresses anchorage-independent growth of the tumor cells in soft agar.

Materials and methods

Cell culture

PCR amplification was carried out using 1 lL of the first-strand cDNA as a template in a total volume of 20 lL containing 0.5 lM of each primer, 200 lM each of dNTPs, 2.5 mM MgCl2, and 0.1 lL AmpliTaq GoldTM DNA Polymerase (Roche, Branchburg, NJ, USA) in the buffer recommended by the manufacturer. The forward primer, hST2-582F; 5¢-GACGGCGACCAGGTCCTT-3¢, and the reverse primer, hST2-649R; 5¢-GGGCTCCG ATTACTGGAAACA-3¢, were both derived from the common region of human ST2 and ST2L [9,17]. After treatment at 94 (cid:2)C for 10 min, 30 cycles of 94 (cid:2)C for 1 min, 60 (cid:2)C for 1 min, and 72 (cid:2)C for 1 min were performed in the DNA Thermal Cycler 480 (Takara). The last polymerization step at 72 (cid:2)C was extended to 10 min. T98G, A172, U251, U373Mg and T430 cells were derived from human glioblastoma. Onda 11 cells were derived from human anaplastic astrocytoma.

Real-time PCR

T98G cells (from T. Kasahara, Kyoritsu College of Pharmacy, Tokyo, Japan) were cultured in minimum essential medium (MEM) with 10% fetal bovine serum and 1 mM sodium pyruvate. A172 and U251 cells (Japanese Cancer Research Resources Bank, Tokyo, Japan) were cultured in Dulbecco’s modified Eagle medium (DMEM) with 10% fetal bovine serum. Onda11 cells (from T. Kumanishi, Brain Research Institute, Niigata University, Niigata, Japan) and U373Mg and T430 cells (from T. Kasahara) were cultured in RPMI 1640 with 10% fetal bovine serum. TM12 cells (from S. Yonehara, Kyoto University, Japan) were cultured in DMEM with 10% fetal bovine serum. RT- and real-time PCR were carried out in an ABI Prism 7700 Sequence Detection System (Perkin-Elmer, Brauch- burg, NJ, USA) using TaqManTM Gold RT-PCR Kit with controls (Perkin-Elmer) according to the manufacturer’s protocol. The same primer set as described above, hST2- 528F and hST2-649R, and TaqMan probe, hST2C-TM1 [5¢-(Fam)-CGGTCAAGGATGAGCAAGGCTTTTCT- (Tamra)-3¢] was used for amplification. Glyceraldehyde- 3-phosphate dehydrogenase (GAPDH) was used as an internal control, and total RNA extracted from stimulated TM12 cells was used as a positive control.

RT-PCR with Southern blotting analysis for detection of promoter usage

TM12 human fibroblastic cells were stimulated to proliferate by changing the medium to DMEM with 10% fetal bovine serum, after the cells were incubated in DMEM with 0.5% fetal bovine serum for 48 h at 37 (cid:2)C in 5% CO2 in air, and total RNAs were extracted at 10 h after stimulation for RT-PCR analysis.

Specimens of brain tumors

The denatured RNA (5 lg) was reverse-transcribed with 10 lM each of dNTPs, 200 U RNase inhibitor (Toyobo), and 200 U SuperScriptTM II Reverse Transcriptase (Gibco BRL) in a buffer containing 50 mM Tris/HCl (pH 8.3), 75 mM KCl, 3 mM MgCl2, and 10 mM dithiothreitol in a total volume of 25 lL at 40 (cid:2)C for 60 min. The first-strand cDNA synthesized was precipitated by ethanol and dis- solved in 10 lL of distilled water.

Tumor specimens were obtained from eight patients suffering from meningioma and eight patients with gliob- lastoma. The study protocol was ethically approved by our Institutional Review Board for Human Studies, and informed consent was obtained from all subjects before enrollment. Histologically, all meningiomas consisted of PCR amplification was carried out using 1 lL of the first- strand cDNA as a template in a total volume of 20 lL containing 0.4 lM of each primer for ST2 (oBC001,

Suppression of anchorage-independent growth by ST2 (Eur. J. Biochem. 270) 165

(cid:1) FEBS 2003

absorbance of each well was determined using a micro- plate reader (ImmunoMini NJ2300, Inter Medical, Tokyo, Japan) at a wavelength of 450 nm against a reference wavelength of 620 nm.

Production and purification of recombinant human ST2 protein

5¢-GAGGAATTCGCTTTCTGAGTTGTGAAACTGT GGGC-3¢ or oBC009, 5¢-TCACTGACTCGAGGTTCAT CCCCTCTGTCTTTCAG-3¢, and oBC010, 5¢-CTCTTG GATCCACACTCCATTCTGCTTACACTTGC-3¢) or ST2L (oBC001 or oBC009, and oBC011, 5¢-TCAAA CTCGGATCCCTTTGCACATCACAGCAGGCA-3¢), 200 lM each of dNTPs, and 0.4 lL 50 · Advantage cDNA Polymerase Mix (Clontech, Palo Alto, CA, USA) in the buffer recommended by the manufacturer. At first, the denaturation was performed at 95 (cid:2)C for 1 min, and then 30 cycles of 94 (cid:2)C for 1 min and 70 (cid:2)C for 3 min were carried out in the DNA Thermal Cycler 480 (Takara). The last polymerization at 70 (cid:2)C was extended to 5 min.

The pEF-BOS-ST2H was transfected into HEK293 cells (from K. Tago, Jichi Medical School, Tochigi, Japan) as described previously [11]. Next, the recombinant hST2 protein was purified from cell culture supernatant of transfected HEK293 cells through a heparin-agarose col- umn [11]. The final preparation of purified hST2 protein showed a single band in silver staining after SDS/PAGE. The hST2 concentration was measured by a sandwich ELISA procedure [14].

Flow cytometry to assess binding of human ST2 protein to T98G cells

Southern blotting analysis was performed using DIG- High Prime DNA Labeling and Detection Starter Kit II (Roche, Mannheim, Germany) according to the manufac- turer’s protocol. Five microliters of the PCR products was separated by electrophoresis on a 1% (w/v) agarose gel and transferred onto a nylon membrane (Hybond N+, Amer- sham Pharmacia Biotech, Uppsala, Sweden). A 565-bp fragment, hST2EPv, which was excised from the human ST2 cDNA with EcoRI (Toyobo) and PvuII (Toyobo), was used as a probe. Human leukemic cell line, UT-7, was used as a positive control [10].

Isolation of stable transformants with human ST2

The purified hST2 protein was labeled with fluorescein isothiocyanate (FITC, Molecular Probes, Eugene, OR, USA) according to the manufacturer’s protocol. T98G cells were washed with NaCl/Pi containing 1% (w/v) BSA and resuspended in 50 lL of NaCl/Pi containing FITC-labeled hST2 protein. Then the cells were left in the dark for 30 min at room temperature. After being washed with NaCl/Pi containing 1% (w/v) BSA, the cells were analyzed by flow cytometry with a FACScan (Becton-Dickinson, Franklin Lakes, NJ, USA). Raji cells were used as a negative control, and RPMI 8226 cells were used as a positive control as described previously [11].

Colony formation assay in soft agar

The entire coding region of human ST2 (hST2) cDNA was subcloned into the pEF-BOS expression vector [17,18]. The pEF-BOS vector containing hST2 cDNA (pEF-BOS- ST2H) was introduced into T98G cells with pEF-BOS vector containing the blasticidin resistance gene [9] for stable transformation using Lipofectamine (Gibco BRL) accord- ing to the manufacturer’s protocol. The transfected cells were isolated in the presence of 3 lgÆmL)1 blasticidin S (Kaken Seiyaku, Tokyo, Japan). Among 91 blasticidin- resistant clones, we could obtain only four clones that expressed enough of the ST2 protein to be detectable by enzyme linked immunosorbent assay (ELISA). Genomic Southern blotting analysis was performed to confirm the individuality of each clone.

ELISA

First, 0.8 mL of MEM containing 0.5% Agar Purified (BD Diagnostic Systems, Sparks, MD, USA) and 10% (v/v) fetal bovine serum was poured into each well of 12-well plates. Then, the layer was covered with cell suspension (6 · 102 cells) in 1.2 mL of MEM containing 0.3% (w/v) agar and 10% (v/v) fetal bovine serum. Finally, the layer of MEM- 0.3% agar containing the cells was further covered with 1 mL of MEM containing 10% (v/v) fetal bovine serum. Medium was exchanged every 96 h. On day 7 after plating, the colonies were counted under a microscope.

Statistical analysis

Statistical evaluation of all data was by Student’s t-test. P < 0.05 was accepted as statistically significant.

Results

Expression of ST2 in brain tumors

We studied the ST2 gene expression in six cell lines derived from malignant glioma, A172, U251, U373Mg, T430, T98G, and Onda11, using RT-PCR (Fig. 1A). The ST2 gene was expressed in all malignant glioma cell lines examined. However, the expression levels of the ST2 gene in the tumor cell lines compared to that in stimulated TM12 The concentration of the hST2 protein in the cell culture supernatant was measured by the sandwich ELISA [14]. Microtiter plates containing 96 wells were coated with 1.75 lg per well of anti-human ST2 monoclonal Ig, FB9. One hundred microliters of the supernatant from each of the cell lines was added to the wells (run in duplicate), and the wells were kept at room temperature for 1 h. After washing with phosphate-buffered saline (NaCl/Pi) con- taining 0.05% (w/v) Tween 20, 100 lL of biotinylated 2A5 in NaCl/Pi containing 0.1% (w/v) BSA was added to each well, and the resulting mixtures were kept for 1 h at streptavidin–horse room temperature. After washing, radish peroxidase (HRP) containing solution was added to the wells, and the plates were kept at room temperature for 30 min. Finally, 140 lL of 10 mM o-phenylenediam- ine-0.01% (v/v) H2O2 in 100 mM sodium acetate buffer (pH 5.0) was added to each well. After 20 min, the

166 Y. Haga et al. (Eur. J. Biochem. 270)

(cid:1) FEBS 2003

Fig. 2. Real-time quantitative PCR analysis of ST2 gene expression in meningiomas and glioblastomas. Distribution of the relative quantity of ST2 mRNA of brain tumor samples of patients is presented. Total RNA was prepared from tumor specimens, then real-time PCR was performed using the TaqMan Gold RT-PCR Kit. The amount of ST2 mRNA was normalized to the level of GAPDH. The value of TM12 cells was taken as the standard, and final relative quantity of ST2 mRNA in meningioma (Mg) and glioblastoma (GB) was expressed relative to the standard. All experiments were performed in duplicate. Bars indicate the mean value in each group.

Fig. 1. Real-time quantitative PCR analysis of ST2 gene expression in lines. (A) Real-time-PCR was performed using glioblastoma cell primers corresponding to the common region of hST2 and hST2L, with GAPDH as an internal control. PCR products were separated by electrophoresis on 5% polyacrylamide gel. Lane 1, TM12; lane 2, A172; lane 3, U251; lane 4, U373Mg; lane 5, T430; lane 6, T98G; lane 7, Onda11. (B) Relative quantities of ST2 mRNA to TM12 are shown. All experiments were performed in duplicate and repeated three times. The amount of the ST2 was normalized to the level of GAPDH. A normalized ST2 value of TM12 was taken as the standard, and the final relative quantity of ST2 mRNA was expressed relative to the standard. Error bars designate the standard deviation.

Promoter usage in human glioma cell lines

fibroblastic cells used as a positive control [10], were very low in real-time PCR (Fig. 1B). The expression of the ST2 gene was lowest in T98G cells.

We next analyzed the expression of the ST2 gene in the specimens of human brain tumors (Fig. 2) and found expression levels to be higher in meningiomas than in glioblastomas. Statistically, there was no significant dif- ference between the quantity of ST2 mRNA in TM12 cells and that in meningiomas. On the other hand, the expression level of the ST2 gene in glioblastomas was significantly lower than that in TM12 cells and in menin- giomas. Four primers, described previously [10], were used to detect transcripts of the human ST2 gene. The primer oBC001 corresponds to the distal exon 1a, and oBC009 corresponds to the proximal exon 1b. The reverse primers, oBC010 and oBC011, correspond to the unique 3¢-regions of cDNA for hST2 and hST2L, respectively. RT-PCR with Southern blotting analysis was carried out using 5 lg of total RNA extracted from each human glioma cell line. As shown in Fig. 3A, a DNA fragment of 1237 bp corresponding to the human ST2 mRNA containing proximal exon 1b was amplified with the primer pair oBC009 and oBC010 in five of six malignant glioma cell lines. On the other hand, no DNA fragment was amplified with the primer pairs oBC001 and oBC010, oBC001 and oBC011, or oBC009 and oBC011 (Fig. 3B–D). These results indicated that the main promoter for the expression of hST2 in malignant glioma cell lines resides in the proximal region and, further, that the transmembrane form of hST2L was not expressed in these cell lines.

Suppression of anchorage-independent growth by ST2 (Eur. J. Biochem. 270) 167

(cid:1) FEBS 2003

Fig. 4. Level of the hST2 protein in the supernatant of ST2 trans- formants measured by ELISA. The sandwich ELISA procedure was performed as described in Materials and methods. Cell supernatant was collected after every cell line had grown to confluency. ST2 protein was detected in the supernatants of C2.1, C8, C10, and C31; however, that of T98G, EV-4, EV-5, and EV-7 (data not shown) was under the detectable level. All experiments were performed in duplicate and repeated three times. Error bars represent the standard deviation.

Therefore, purified hST2 protein was labeled with FITC and used as a probe for flow cytometric analysis. As described previously [11], Raji cells hardly shift in the profile of flow cytometry; in contrast RPMI8226 cells showed a remarkable shift (Fig. 5A and B). As shown in Fig. 5C, a marked shift was also observed in T98G, suggesting that hST2 protein strongly bound to T98G cells.

Fig. 3. RT-PCR and Southern blotting analysis of the ST2 transcripts in human malignant glioma cell lines. RT-PCR was performed using primers oBC009 and oBC010, which anneal specifically to the exon 1b and the unique 3¢-regions of cDNA for ST2, respectively. The result of Southern blotting analysis using primer pair oBC009-oBC010 is shown (A). Using primer pairs of oBC001-oBC010 (B), oBC001-oBC011 (D), and oBC009-oBC011 (C), the amplified cDNA was not detected. Lane 1, A172; lane 2, U251; lane 3, U373Mg; lane 4, T430; lane 5, T98G; lane 6, Onda11; lane 7, UT-7 as a positive control [10]. Arrowheads and arrows indicate the position of the amplified cDNA for ST2 and ST2L, respectively.

Effects of hST2 on anchorage-independent growth of T98G cells

Introduction of hST2 cDNA into T98G cells

We examined growth properties of the stable transformants that could be affected by hST2 in an autocrine fashion. Direct cell counting and WST-1 colorimetric assay were carried out in conventional culture conditions. In these studies, we could not draw any significant conclusions about the effect of ST2 expression on cell proliferation of T98G (data not shown).

To examine the effects of ST2 on proliferation and malignancy of brain tumors, we introduced cDNA for hST2 into the human glioblastoma cell line T98G, which originally expressed very low levels of hST2. We isolated four clones (C2.1, C8, C10, and C31) that stably expressed the hST2 protein. The amount of the hST2 protein in the culture supernatant was measured by ELISA (Fig. 4). The hST2 protein was under the detectable level in the super- natant of wild-type T98G cells. The expression level of the hST2 protein was the highest in C2.1 and the lowest in C8. To confirm that each ST2 transfectant was an independ- ent clone, we performed genomic Southern blotting analy- sis. Exogenous ST2 was detected in all clones, and the lengths of the labeled fragments were different from one another, showing the different integration sites (data not shown).

To evaluate the effects of hST2 protein on anchorage- independent growth, we carried out colony formation assay using T98G, transfectants with the empty vector, and transfectants with hST2 cDNA, cultured in soft agar with MEM containing 10% fetal bovine serum and 1 mM pyruvate. On day 7 after seeding in soft agar, 151 ± 16 (mean ± SD), 148 ± 11, 165 ± 22, and 130 ± 6 colonies per well were observed in control cells, such as T98G, EV-4, EV-5, and EV-7, respectively. In contrast, 32 ± 11, 70 ± 12, and 56 ± 22 colonies per well were observed in C2.1, C8, and C10 clones, respectively. The experiment was repeated three times, and the results were reproducible. The numbers of colonies of ST2 transfectants were significantly lower than those of control cells (Fig. 6A). Binding of hST2 protein to T98G cells

The next question is whether this suppressive effect was caused by the hST2 protein secreted by the transformed cells to the culture environment or due to an internal In order to analyze the effect of hST2 on T98G cells, the first important question is whether hST2 binds to the cells.

168 Y. Haga et al. (Eur. J. Biochem. 270)

(cid:1) FEBS 2003

Fig. 5. Flow cytometric analysis of T98G cells with FITC-hST2. Flow cytometric analysis was performed on Raji (A) as a negative control, RPMI8226 (B) as a positive control, and T98G (C). The filled area corresponds to the cells treated with FITC-hST2, and the lucent area corresponds to the cells without FITC-hST2 treatment.

ST2 is against malignancy. In fact, there are several reports suggesting an antiproliferative action of cytokines. Todo et al. have reported that the addition of recombin- ant human IL-6 to meningioma cell cultures caused a dose-dependent inhibition of basal DNA synthesis, and the secretion of IL-6 by meningioma cells was powerfully induced by both TNF-a and IL-1b [23]. The addition of IL-1b has been reported to down-regulate IL-1R expres- sion in a glioblastoma cell line [24]. Furthermore, the level of IL-1b in the sera of pediatric patients with glioma has been shown to increase after radiotherapy [25]. It should be noted that tumor samples from patients are certainly nonhomogeneous, containing other cell species. Conse- quently, the expression of the ST2 gene in tumor samples itself does not imply functional relevance of ST2 in glioma. Therefore, the experiments using cell lines are of importance.

effect in the cell as a result of the integration of the hST2 gene. Wild-type T98G cells were suspended in 1.2 mL of MEM containing 0.3% agar and 10% fetal bovine serum, and the cell suspension in MEM/0.3% agar was overlaid above the layer of MEM/0.5% agar prepared in each well beforehand. Then, 1 mL MEM containing various amounts of hST2 protein was added. As shown in Fig. 6B, on day 7 after plating, the inhibition of colony formation in soft agar was observed to be dose- dependent. The numbers of colonies of T98G cultured in the medium containing 10 ngÆmL)1 and 50 ngÆmL)1 of hST2 protein were significantly lower than those of T98G cultured in the medium without hST2. The assay was performed three times with reproducible results. The experiments suggest a significant possibility that the hST2 protein suppressed anchorage-independent growth of T98G glioblastoma cells. Considered together with the result of the binding experiment described above, the effect was judged to be conveyed from outside of the cells.

Discussion

The ST2 gene was revealed to be expressed in malignant glioma cell lines as in the case of fibroblast (Fig. 1), and usage of the proximal promoter to express the ST2 gene was also common among these cell species (Fig. 3A) [10]. Because of the accumulated knowledge about the expres- sion of the ST2 gene in fibroblasts, we carried out our study using a TM12 human fibroblastic cell line as a control.

To investigate the effect of ST2 on cell growth, we established stable transformants that constitutively express and secrete the human ST2 protein. Direct cell counting and WST-1 colorimetric assay resulted in no detection of a significant difference between the growth of the ST2-transformants and control cells under conventional culture conditions. However, a remarkable difference was reproducibly observed in the colony formation in soft agar plates. The ST2 transformants showed significantly lower numbers in colony formation compared to wild- type T98G cells as well as transformants with an empty vector (Fig. 6A). The intensity of inhibition correlated well with the concentration of ST2 protein in the culture supernatant of each cell line, thus corresponding to the efficiencies of the production of the ST2 protein (Figs 4 and 6A). Higher expression of ST2 mRNA in benign tumors compared to the scarce expression in malignant tumors led us to construct a working hypothesis that induction of

Suppression of anchorage-independent growth by ST2 (Eur. J. Biochem. 270) 169

(cid:1) FEBS 2003

exogenous purified recombinant human ST2 protein to wild-type T98G cells. The inhibitory effect was dose dependent and reproducible (Fig. 6B). It should be noted that the concentration of added ST2 was from 2.5 to 50 ngÆmL)1, which is the pathophysiological range of the ST2 concentration actually detected in sera of patients [14– 16]. In this possible mechanism, soluble ST2 is considered to function as a ligand attaching to an unknown counter- receptor on the cell surface. In fact, a recent report showing an inhibitory effect of ST2 on Toll-like receptor 4 expression suggests the same mechanism [31].

In conclusion, the ST2 protein suppresses anchorage- independent growth of T98G glioblastoma, suggesting the protein’s negative effect on malignancy. Further studies are necessary to reveal the mechanisms of action of the ST2 protein as well as the target molecule of the ST2 protein on the cell surface to convey the negative signal.

Acknowledgements

We are grateful to Dr H. Higashi of the Institute for Genetic Medicine, Hokkaido University, for his valuable advice and discussion. We also thank Mrs R. Izawa and Miss Y. Komine for excellent technical assistance. This work was supported in part by a grant for the High- Tech Research Center from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

References

1. Tominaga, S. (1988) Murine mRNA for the b-subunit of integrin is increased in BALB/c-3T3 cells entering the G1 phase from the Go state. FEBS Lett. 238, 315–319.

2. Tominaga, S. (1989) A putative protein of a growth specific cDNA from BALB/c-3T3 cells is highly similar to the extracellular por- tion of mouse interleukin 1 receptor. FEBS Lett. 258, 301–304. 3. Werenskiold, A.K., Hoffman, S. & Klemenz, R. (1989) Induction of a mitogen-responsive gene after the expression of the Ha-ras oncogene in NIH3T3 fibroblasts. Mol. Cell Biol. 9, 5207–5214. 4. Klemenz, R., Hoffmann, S. & Werenskiold, A.K. (1989) Serum and oncoprotein-mediated induction of a gene with sequence similarity to the gene encoding carcinoembryonic antigen. Proc. Natl Acad. Sci. USA 86, 5708–5712.

5. Bergers, G., Reikerstorfer, A., Braselmann, S., Graninger, P. & Busslinger, M. (1994) Alternative promoter usage of the Fos- responsive gene Fit-1 generates mRNA isoforms coding for either secreted or membrane-bound proteins related to the IL-1 receptor. EMBO J. 13, 1176–1188.

6. Lanahan, A., Williams, J.B., Sanders, L.K. & Nathans, D. (1992) Growth factor-induced delayed early response gene. Mol. Cell. Biol. 12, 3919–3929.

Fig. 6. Effects of ST2 on colony formation in soft agar. (A) T98G, EV-4, EV-5, EV-7, C2.1, C8, and C10 cells were suspended in soft agar, after which the layer of soft agar was covered with MEM containing 10% fetal bovine serum. (B) T98G cells were suspended in soft agar and subsequently poured into each well. The layer of soft agar was further covered with MEM containing 10% fetal bovine serum and indicated amounts of ST2 protein. On day 7 after plating, colonies were counted under a microscope. Measurements were made from nine different fields in each well (the area of one field was 4 mm2; magnification, 40 ·). All experiments were performed in triplicate. The data represent the mean of total colonies per well. Error bars represent the standard deviation. *P < 0.02 vs. control; **P < 0.005 vs. T98G; ***P < 0.001 vs. T98G.

7. Yanagisawa, K., Takagi, T., Tsukamoto, T., Tetsuka, T. & Tominaga, S. (1993) Presence of a novel primary response gene ST2L, encoding a product highly similar to the interleukin 1 receptor type 1. FEBS Lett. 318, 83–87.

8. Tominaga, S., Kuroiwa, K., Tago, K., Iwahana, H., Yanagisawa, K. & Komatsu, N. (1999) Presence and expression of a novel variant form of ST2 gene product in human leukemic cell line UT-7/GM. Biochem. Biophys. Res. Commun. 264, 14–18.

9. Li, H., Tago, K., Io, K., Kuroiwa, K., Arai, T., Iwahana, H., Tominaga, S. & Yanagisawa, K. (2000) The cloning and nucleo- tide sequence of human ST2L cDNA. Genomics 67, 284–290. 10. Iwahana, H., Yanagisawa, K., Ito-Kosaka, A., Kuroiwa, K., Tago, K., Komatsu, N., Katashima, R., Itakura, M. & Tominaga, S. (1999) Different promoter usage and multiple transcription

One of the possible explanations was that the suppression was caused by the secreted ST2 protein, which may be at a relatively high concentration in the microenvironment, via cell surface attachment, in other words in an autocrine fashion. By flow cytometric analysis, we confirmed the binding of exogenous ST2 protein to T98G cells, supporting the possibility of an autocrine mechanism for the suppres- sion of colony formation (Fig. 5).

Finally, we confirmed the inhibitory effect of the ST2 protein on colony formation in soft agar by adding

170 Y. Haga et al. (Eur. J. Biochem. 270)

(cid:1) FEBS 2003

malignant pleural effusions. Am. J. Respir. Crit. Care Med. 165, 1005–1009.

21. Boyle-Walsh, E., Hashim, I.A., Speirs, V., Fraser, W.D. & White, M.C. (1994) Interleukin-6 (IL-6) production and cell growth of cultured human meningiomas: interactions with interleukin-1b (IL-1b) and interleukin-4 (IL-4) in vitro. Neurosci. Lett. 170, 129– 132.

initiation sites of the interleukin-1 receptor-related human ST2 gene in UT-7 and TM12 cells. Eur. J. Biochem. 264, 397–406. 11. Yanagisawa, K., Naito, Y., Kuroiwa, K., Arai, T., Furukawa, Y., Tomizuka, H., Miura, Y., Kasahara, T., Tetsuka, T. & Tominaga, S. (1997) The expression of ST2 gene in helper T cells and the binding of ST2 protein to myeloma-derived RPMI8226 cells. J. Biochem. 121, 95–103.

22. Levy, E.I., Paino, J.E., Sarin, P.S., Goldstein, A.L., Caputy, A.J., Wright, D.C. & Sekhar, L.N. (1996) Enzyme-linked immuno- sorbent assay quantification of cytokine concentration in human meningiomas. Neurosurgery 39, 823–829.

12. Xu, D., Chan, W.L., Leung, B.P., Huang, F.P., Wheeler, R., Pierdrafita, D., Robinson, J.H. & Liew, F.Y. (1998) Selective expression of a stable cell surface molecule on type 2 but not type 1 helper T cells. J. Exp. Med. 187, 787–795.

23. Todo, T., Adams, E.F., Rafferty, B., Fahlbusch, R., Dingermann, T. & Werner, H. (1994) Secretion of interleukin-6 by human meningioma cells: possible autocrine inhibitory regulation of neoplastic cell growth. J. Neurosurg. 81, 394–401.

13. Lo¨ hning, M., Stroemann, A., Coyle, A.J., Grogan, J.L., Lin, S., Gutierrez-Ramos, J.-C., Levinson, D., Radbruch, A. & Kamradt, T. (1998) T1/ST2 is preferentially expressed on murine Th2 cells, independent of interleukin 4, interleukin 5, and interleukin 10, and important for Th2 effector function. Proc. Natl Acad. Sci. USA 95, 6930–6935.

24. Gottschall, P.E., Koves, K., Mizuno, K., Tatsuno, I. & Arimura, A. (1991) Glucocorticoid upregulation of interleukin 1 recep- tor expression in a glioblastoma cell line. Am. J. Physiol. 261, 362– 368.

14. Kuroiwa, K., Li, H., Tago, K., Iwahana, H., Yanagisawa, K., Komatsu, N., Oshikawa, K., Sugiyama, Y., Arai, T. & Tominaga, S. (2000) Construction of ELISA system to quantify human ST2 protein in sera of patients. Hybridoma 19, 151–159.

25. Gridley, D.S., Loredo, L.N., Slater, J.D., Archambeau, J.O., Bedros, A.A., Andres, M.L. & Slater, J.M. (1998) Pilot evaluation of cytokine levels in patients undergoing radiotherapy for brain tumor. Cancer Detect. Prev. 22, 20–29.

26. Jachimczak, P., Schwulera, U. & Bogdahn, U. (1994) In vitro studies of cytokine–mediated interactions between malignant glioma and autologous peripheral blood mononuclear cells. J. Neurosurg. 81, 579–586.

15. Oshikawa, K., Kuroiwa, K., Tago, K., Iwahana, H., Yanagisawa, K., Ohno, S., Tominaga, S. & Sugiyama, Y. (2001) Elevated soluble ST2 protein levels in sera of patients with asthma with an acute exacerbation. Am. J. Respir. Crit. Care Med. 164, 277–281. 16. Kuroiwa, K., Arai, T., Okazaki, H., Minota, S. & Tominaga, S. (2001) Identification of human ST2 protein in the sera of patients with autoimmune diseases. Biochem. Biophys. Res. Commun. 284, 1104–1108.

27. Lichtor, T. & Libermann, T.A. (1994) Coexpression of inter- leukin-1b and interleukin-6 in human brain tumor. Neurosurgery 34, 669–673.

17. Tominaga, S., Yokota, T., Yanagisawa, K., Tsukamoto, T., Takagi, T. & Tetsuka, T. (1992) Nucleotide sequence of a com- plementary DNA for human ST2. Biochim. Biophys. Acta 1171, 215–218.

28. Van Meir, E., Sawamura, Y., Diserens, A.C., Hamou, M.F. & de Tribolet, N. (1990) Human glioblastoma cells release interleukin 6 in vitro and in vivo. Cancer Res. 50, 6683–6688.

29. Giulian, D. & Lachman, L.B. (1985) Interleukin-1 stimulation of astroglial proliferation after brain injury. Science (Wash. DC) 228, 497–499.

18. Yoshida, K., Arai, T., Yokota, T., Komatsu, N., Miura, Y., Yanagisawa, K., Tetsuka, T. & Tominaga, S. (1995) Studies on natural ST2 gene products in the human leukemic cell line UT-7 using monoclonal antihuman ST2 antibodies. Hybridoma 14, 419– 427.

30. Giulian, D., Young, D.G., Woodward, J., Brown, D.C. & Lach- man, L.B. (1988) Interleukin-1 is an astroglial growth factor in the developing brain. J. Neurosci. 8, 709–714.

19. Prechtel, D., Harbeck, N., Berger, U., Ho¨ fler, H. & Werenskiold, A.K. (2001) Clinical relevance of T1-S, an oncogene-inducible, in secreted glycoprotein of the immunoglobulin superfamily, node-negative breast cancer. Laboratory Invest. 81, 159–165. 20. Oshikawa, K., Yanagisawa, K., Ohno, S., Tominaga, S. & Sugiyama, Y. (2002) Expression of ST2 in helper T lymphocytes of

31. Sweet, M.J., Leung, B.P., Kang, D., Sogaard, M., Schulz, K., Trajkovic, V., Campbell, C.C., Xu, D. & Liew, F.Y. (2001) A novel pathway regulating lipopolysaccharide-induced shock by ST2/T1 via inhibition of Toll-like receptor 4 expression. J. Immunol. 166, 6633–6639.