[CANCER RESEARCH 54, 2334-2336, May 1, 19941 The a Subunit of GTP-binding Protein G0 in Neuroblastoma: Correlation with Advanced Disease Stage

Yukio Ishiguro,' Kanefusa Kato, Tomiko Asano, Hiroshi Akatsuka, Hiroyuki Iwata, Fujio Ito, and Takahiro Ito

Department ofSurgery, Branch Hospital ofNagoya University School ofMedicine, Daiko-minam4 Higashi-ku, Nagoya 461 [V. L, H. A., H. L, F. I., T. I.J, and Department of Biochemistry, Institute for Developmental Research, Aichi Prefectural Colony, Kamiya-cyo, Kasugai [K K, T. A.J, Japan

ABSTRACT chemotherapy with cyclophosphamide and vincristine for 3 to 12 months. Stage III patients without recurrences (n = 5) had postoperative chemotherapy Tissue levels of the a subunit of G protein G0 (G0 a) were measured in for 2 years. Nine of 24 patients with distant metastases or recurrences had bone solid tumors from pediatric patients by immunoassay.G0 a concentra marrow transplants after complete clinical remissions. Survival time was tions were determined in the supernatant obtained by centrifugatlon of measured from the start of treatment. Diagnoses were confirmed histologically. tissue homogenates prepared in the presence (total G0 a) or absence of 2% Preparation of Tissue Extracts. Tissues were homogenized at 0°Cin 10 sodium cholate (soluble G0 a). Mean G0 a concentrations (total G0 a and volumes of 10 mM Tris-HCI (pH 7.5) containing 1 mM EDTA, using a soluble G0 a) in neuroblastomas (7 ganglioneuromas,13 ganglioneuro Polytron-type homogenizer. One-half of each homogenate was centrifuged at blastomas, and 50 neuroblastomas) were over 50-fold higher than those in 4°Cat 125,000 X g for 40 mm, and the supernatant was saved for G0 a other solidtumors from pediatric patients (n 13).Mean total G0a and analysis of the soluble fraction (soluble G0 a). Ten % sodium cholate was soluble G0 a concentrations were 207.0 ±166.0(SD)ng/mg of cholate added to the other one-half of the homogenate to a final concentration of 2%, extractable protein and 58.6 ±47.0 ng/mg ofsoluble protein, respectively, and the tissue was further homogenized. This homogenate was sonicated for 1 in the neuroblastoma group (n 70). Total G0 a concentration decreased mm and then centrifuged as described above. The supernatant fraction was with disease stage and was strongly correlated with outcome in patients then used for G0a analysis of the cholate-extractable fraction (total G0 a) (9). with neuroblastoma. The mean total G0 a concentration in tumors from G0a andAntibodyPreparations.G0awaspurifiedfrombovinebrain, younger patients (<1 year old) was 297.0 ±137.0 ng/mg of cholate human brain, and human neuroblastoma by the method of Sternweis and extractable protein, significantly higher than in twuors from older pa Robishaw (4) with modifications. Antibodies were raised in rabbits with dents (140.0 ±155.0ng/mg cholate-extractable protein, P < 0.0001). These bovine G0 a as an immunogen and purified to be monospecific to G0 a as results suggest that total C0 a levels in neuroblastoma may indicate the described previously (9). The purified antibodies cross-reacted with human G0 degree of malignancy. a, but not with G. a or fry subunits (9). Enzyme Immunoassay of G@a. G@,a was assayed by a sandwich-type INTRODUCTION immunoassay for bovine G@,aas described previously (9). This assay system consists of a solid phase (jolystyrene ball) with immobilized purified antibod G0 is a member of a family of heterotrimeric GTP-binding proteins ies to bovine G0 a and the same antibodies labeled with @-D-galactosidase (G proteins) which function as signal transducers (1, 2). Recently, from Escherichia coli. The reactivity of human brain and neuroblastoma 0,, a many investigators have reported that G0 interacts with several recep in the immunoassay was identical and was approximately 80% that of bovine tors and regulates Ca2@ and K@ channels as well as phospholipase C Go a (7). Purified human G0 a was used as the standard for immunoassay of @ (1, 2). G@was initially isolated from bovine brain (3, 4); G0 a2 G0 a, and results were expressed as ng of human G0 a per mg of soluble protein, per mg of cholate-extractable protein, or per g of wet tissue. mainly localized to nervous tissues and neuroendocrine cells (5). We Proteins were determined by the method of Schaffer and Weissmann (11) have reported that G0 a is present in tumors derived from neuroen with bovine serum albumin as a standard. docrine cells (6, 7) and that serum G0 a is a useful marker for Statistics. Mann-Whitney-Wilcoxon (2-sample) rank sum test, a nonpara neuroblastoma (8), a tumor in the amine precursor uptake and decar metric method, was used for the analysis of the separation between two groups. boxylation group. In the bovine brain, more than 98% of 0,, a is The event-time distributions were estimated by the product limit method of membrane bound (9), while in neuroblastomas a significant amount of Kaplan and Meier. Differences between event-time distributions were tested G0 a is found in the soluble fraction (6). In this study, we measured for statistical significance by using the generalized Wilcoxon t test. G0 a concentrations in the soluble and cholate-extractable fractions of neuroblastomas and evaluated the relationship between G0 a tissue RESULTS levels and disease stage and outcome in patients with neuroblastoma. G0 a Concentrations in Pediatric SolidTumors. Table 1 shows the mean G0 a concentrations in solid tumors from pediatric patients. MATERIALS AND METHODS Mean G0 a concentrations (total G0 a and soluble G@,a) in the Tissues. Tumor tissues obtained at operation from 83 patients included 50 neuroblastoma group (ganglioneuroma, ganglioneuroblastoma, and neuroblastomas, 13 ganglioneuroblastomas, 7 ganglioneuromas, 4 Wilms' neuroblastoma) were over 50-fold higher than those in other solid tumors, 4 rhabdomyosarcomas, 1 hepatoblastoma, 1 dysgerminoma, and 3 tumors from pediatric patients. The mean total G0 a and soluble G0 a adrenal cortical tumors. All samples were rapidly frozen and kept at —80°Cconcentrations were 207.0 ±166.0 (SD) ng/mg of cholate-extractable until analysis. Neuroblastoma patients were classified into five groups (I, 17; protein and 58.6 ±47.0 ng/mg of soluble protein, respectively, in the II, 11; III, 17; IV, 12; IVs, 6) according to the pretreatment staging procedures neuroblastoma group (n = 70). The mean total G0 a concentration in of Evans CtaL (10). Nine of the 17 stage III neuroblastoma patients and 9 of the neuroblastoma group was approximately one-tenth the value re the 12 stage IV neuroblastoma patients received intensive preoperative che ported for human and bovine cerebral cortex (9, 12). On the other motherapy with cyclophosphamide, vincristine, doxorubicin, and cisplatin; tumor tissues from these patients were obtained within 6 months after begin hand, the mean total and soluble G0 a concentrations in other solid ning of chemotherapy. After surgery, stage I, H, and IVs patients had adjuvant tumors from pediatric patients were 3.34 ±3.20 ng/mg cholate extractable protein (P < 0.0001) and 0.93 ±0.54 ng/mg of soluble Received 7/19/93; accepted 3/2/94. protein (P < 0.0001), respectively. The costs of publication of this article were defrayed in part by the payment of page G0 a Concentrations and Clinical Stage in Neuroblastomas. charges. This article must therefore be hereby marked advertisement in accordance with Table 2 shows the mean total G0 a and soluble G0 a concentrations 18 U.S.C. Section 1734 solely to indicate this fact. I To whom requests for reprints should be addressed. in different disease stages of neuroblastoma. More extensive disease 2The abbreviation used is: G0 a, the a subunit of GTP-binding protein G0. was associated with a decreased total G0 a concentration. The mean 2334

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1994 American Association for Cancer Research. G0 a IN NEUROBLASTOMA value in advanced (stage III and IV) neuroblastomas was 118.0 ± 1.0 136.0 ng/mg of cholate-extractable protein and was markedly lower than that in localized (stages I and II) neuroblastomas (296.0 ±131.0 .0 0.8 11@ ng/mg of cholate-extractable protein, P < 0.0001). However, the mean total G@a concentration, 346.0 ±185.0 ng/mg of cholate-extractable protein, in stage P/s neuroblastomas (in which metastases were lim .@ 0.6 ited to liver, skin, and bone marrow), was significantly higher than that of stage N neuroblastomas (62.1 ±61.8 ng/mg of cholate Cl) 0.4 extractable protein, P < 0.005), and was similar to that of the most localized disease, stage I neuroblastoma (300.0 ±117.0 ng/mg of :@ 0.2 cholate-extractable protein). Soluble G0 a concentration (ng/mg sol In uble protein) also decreased with disease stages of neuroblastoma, but U.', the changes were not so remarkable compared with those of total G0 0 5@ 1000 1500 2000 2500 a concentration. Therefore, the mean percentage of G0 a in the SurvivalTime(days) soluble fraction, 37.3 ±21.2%, in advanced neuroblastomas was Fig. 1. Correlation between total tissue G0 a concentration J: >200 ng/mg protein; •: significantly higher than in stage I (mean, 16.7%) or stage II (mean, <200 ng/g wet tissue) and overall survival rate in children with neuroblastoma as 19.4%) tumors (P < 0.0001). calculated by the Kaplan-Meier method. Since almost one-half of advanced neuroblastoma patients (stage III and IV) received preoperative chemotherapy, we examined the effects G0 a concentration of >200 was associated with a statistically better of chemotherapy on G0 a concentrations in neuroblastomas. Twenty survival rate (P < 0.01, generalized Wilcoxon t test) (Fig. 1). The nine patients with advanced disease were divided into two groups, soluble G0 a concentration did not correlate with patient survival; chemotherapy(—) and chemotherapy(+). In neuroblastomas from pa however, the percentage of G0 a in the soluble fraction did correlate tients receiving chemotherapy (n = 17), the mean total G0 a and with patient survival as well as with disease stage. soluble G0 a concentrations were 105.0 ±98.4 ng/mg of cholate Correlation between Total G0 a Concentration and Age of extractable protein and 48.4 ±49.8 ng/mg of soluble protein, respec Neuroblastoma Patients. The prognosis of neuroblastomapatients tively, which did not differ significantly from those (134.0 ±177.0 younger than 1 year is better than that of older patients. To examine ng/mg of cholate-extractable protein and 51.9 ± 48.8 ng/mg of the relationship between G0 a concentration and age, patients were soluble protein) in neuroblastomas from patients not receiving che divided into two groups, younger than 1 year and older. In 23 (70%) motherapy (n = 12). of 33 neuroblastomas from patients younger than 1 year, total G0 a Correlation between Total G0 a Concentration and Neuroblas was greater than 200 ng, ranged from 28.8 to 560.0 ng/mg of cholate toma Patient Survival. The correlationbetween the cumulativesur extractable protein, and in only 7 (23%) of 30 neuroblastomas from vival of neuroblastoma patients and total G0 a concentration was older patients, total G0 a was greater than 200 ng, and ranged from determined by the Kaplan-Meier method. Patients were divided into 6.0 to 629.0 ngJmg of cholate-extractable protein. The mean total G0 two groups, depending on whether their total G0 a concentration was a concentration in younger patients was 297.0 ±137.0 ng/mg of greater or lesser than 200 ng/mg of cholate-extractable protein. A total cholate-extractable protein, significantly higher than that in older patients (140.0 ± 155.0 ng/mg of cholate-extractable protein, P < 0.0001) (Fig. 2). Table1 G0a concentrationstumorsNo.Totalin solid pediatric G0 a― G0 a@' DISCUSSION SD)Neuroblastoma (mean ±SD)Soluble (mean ± group ±166.0 ±47.0 The present study demonstrates that tumor tissues in the neuroblas Ganglioneuroma 7 91.2 ±105.0 25.4 ±11.7 toma group contained much higher levels of G0 a than other solid Ganglioneuroblastoma 13 275.0 ±197.0 94.8 ±63.7 39.3OtherNeuroblastoma(70)c 50207.0 205.0 ±155.058.6 53.6 ± tumors from pediatric patients. G0 a is a protein which is mainly localized in nervous tissues and neuroendocrine cells (5); its level in solid pediatric ±3.200.93 ±0.54 the rat brain increases during ontogenic development (13). It has also tumors(13)3.34 been reported that the level of G0 a increases in a pheochromocyto a ng/mg of cholate-extractable protein. b ng/mg of soluble protein. ma-derived cell line (PC12) and a neuroblastoma X glioma hybrid C Numbers in parentheses, total. cell line (NG 108-15) during differentiation (14), suggesting that G0

neuroblastomasTotalTable 2 G0 a concentrations in 63

G0 aSoluble G0 a% of G0 a in soluble fraction 11AaBCD(ng/mg wetNo.extractable of cholate (ng/g of wet(ng/mg of soluble(ng/g of protein)tissue)protein)tissue)StageI17300.0

7.5II11292.0 ±117.013,700 ±4,43071.5 ±37.42,310 ±1,28016.7 ± 8.3III17150.0 ±153.013,100 ±5,54071.3 ±44.82,460 ±1,46019.4 ± 23.4IV1262.1 ±157.09,360 ±8,92060.5 ±52.72,440 ±1,72039.4 ± 18.4IVs6346.0 ±61.84,750 ±4,98035.0 ±39.31,230 ±1,34035.4 ± ±185.017,100 ±7,83081.5 ±71.62,890 ±2,29018.6 ±9.6 a A: I versus II = not significant (NS); I versus III = P < 0.005; I versus IV = P < 0.001; 1 versu.s IVs = NS; II versus III = P < 0.05; II versus IV = P < 0.001; II versus IVa

= NS; Ill versus N = NS; HI versus Na = P < 0.05; N versus Na = P < 0.005. C: I versus II = NS; I versus III = NS; I versus IV = P < 0.01; 1 versus IVa NS; II versus ifi = NS;II versusN P < 0.05;II versusIVs NS;Ill versusIV NS;IIIversusNa NS;IV versusIVa= NS.E: [SolubleG0a (ngof wettissue)-@-TotalG0a (ngof wet tissue)] X 100; I versus II = NS; I versus III = P < 0.0005; I versus IV = P < 0.01; I versus Na = NS; 11versus III = P < 0.05; II versus IV = P < 0.05; II versus IVs = NS; m versusN = NS;Ill versusNa NS;IV versusIVs NS.Mean±SD. 2335

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1994 American Association for Cancer Research. G0 a IN NEUROBLASTOMA

io3 while total G0 a concentrations in neuroblastoma tissues from younger patients were significantly higher than those from older patients. However, total G0 a concentrations varied in neuroblastoma 2 tissues from younger patients, and neuroblastomas with lower total G0 I.I 102I Ii a concentrations (<200 ng/mg protein) may be more malignant than others. . I In conclusion, the total G0 a in neuroblastoma seems to be an I U101UU100 indicator for the malignant grade as well as a prognostic factor. However, the biological function of soluble and membrane-bound G0 a in neuroblastoma has not been clarified as has that of brain G0 a.

REFERENCES ly (AGE) 1. Simon, M. I., Strathmann, M. P., and Gautam, N. Diversity of G proteins in signal transduction. Science (Washington DC), 252: 802—808, 1991. Fig. 2. Total G@,a concentrations in neuroblaatoma tissues versus patient age. Sixty 2. Hepler, J. R., and Gilman, A. G. G proteins. Trends Biochem. Sd., 17: 383—387, three neuroblastoma patients were divided with respect to age at diagnosis into two 1992. groups; 32 (51%) patients were diagnosed before 1 year of age. 3. Neer, E. J., Lok, J. M., and Wolf, L. G. Purifications and properties of the inhibitory guanine nucleotide regulatory unit of brain adenylate cyclase. J. Biol. Chem., 259: 14222—14229, 1984. 4. Sternweis, P. C., and Robiahaw, J. D. Isolation of two proteins with high affinity for is important in the differentiated cells. Among the neuroblastoma guanine nucleotides from membranes of bovine brain. J. Biol. Chem., 259: 13806— 13813, 1984. group, the G0 a levels were higher in malignant neuroblastomas 5. Asano,T., Semba,R.,Kamiya,N., Ogasawara,N.,andKato, K. G0,a GTP-binding (neuroblastoma and ganglioneuroblastoma) than in benign tumors protein: immunochemical and immunohistochemical localization in rat. J. Neuro (ganglioneuroma).Inmalignantneuroblastomas,meantotalG0a chem., 50: 1164—1169,1988. 6. Kato, K., Asano, T., Kamiya, N., Haimoto, H., Hosoda, S., Nagasaka, A, Ariyoshi, concentrations were similar in ganglioneuroblastomas and neuroblas Y., and Ishiguro, Y. Production ofthe a subunit of guanine nucleotide-binding protein tomas, as were soluble G0 a concentrations. However, these results G0by neuroendocrinetumors.CancerRca.,47: 5800—5805,1987. do not mean that there is no relationship between tumor cell differ 7. Asano, T., Morishita, R., and Kato, K. Identification of a guanine nucleotide-binding protein G0 in human neuroblastoma. Cancer Rca., 48: 2756—2759, 1988. entiation and the level of G0 a, since ganglioneuromas are mostly 8. lahiguro, Y., Kato, K., Asano, T., Akatsuka, H., Kishida, Y., and Ito, T. A new blood composed of neurofibrous tissue, with a few ganglion-like tumor cells marker for neuroblastoma: the a subunit of GTP-binding protein G0. J. Tumor Marker Oncol., 5: 39—44,1990. containing G0 a. In addition, the levels of total G0 a in malignant 9. Asano, T., Semba, R., Ogasawara, 0., and Kato, K. Highly sensitive immunoassay for neuroblastomas decreased with advanced tumor stage. These ad the a subunit of the GTP-binding protein G0, and its regional distribution in bovine vanced tumors (stage III and IV) consisted of 9 ganglioneuro brain. J. Neurochem., 48: 1617—1623,1987. 10. Evans, D., D'Angio, J., and Randolph, G. A proposed staging for children with blastomas and 20 neuroblastomas. Therefore, these results are con neuroblastoma. Cancer (Phila.), 27: 374—378,1971. sistent with stable expression of G0 a in the face of decreased tumor 11. Schaffner, W., and weissmann, C. A rapid, sensitive, and specific method for the differentiation. determination of protein in dilute solution. Anal. Biochem., 56: 502—514,1973. 12. Kato, K., Kurobe, N., Suzuki, F., Morishita, R., Asano, T., Sato, T., and Inagaki, T. In the brain, most G0 a is membrane bound and less than 2% of G0 Concentrations of several proteins characteristic of nervous tissue in cerebral cortex a is detectable in the soluble fraction (9). The subunits of G0, G.1, and of patients with Alzheimer's disease. J. Mol. Neurosci., 3: 95—99,1991. 13. Asano, T., Kamiya, N., Semba, R., and Kato, K. Ontogeny of the GTP-binding G12 remain bound to the plasma membrane even after activation by protein G0 in rat brain and heart. J. Neurochem., 51: 1711—1716,1988. GTPyS in bovine brain and human neutrophil membranes (15). In 14. Aaano, T., Morishita, R., Sano, M., and Kato, K. The GTP-binding proteins, G0 and contrast, G0 a is released from the plasma membranes of NG1O8-15 G42, of neural cloned celia and their changes during differentiation. J. Neurochem., 53:1195—1198,1989. cells (16), while the a subunit of G.2 (G12 a) is released from 15. Eide, B., Gierachik, P., Milligan, G., Mullaney, I., Unson, C., Goldsmith, P., and membranes of rat glioma C6BU1 cells (17) by incubation with Spiegel, A. GTP-binding proteins in brain and neutrophil are tethered to the plasma Gpp(NH)p. In mastocytoma P-815 cells, more than 50% of total G,2 membrane via their amino termini. Biochem. Biophys. Res. Commun., 148: 1398— 1405, 1988. a is found in the cytosol fraction (18). In addition, in lymphoma and 16. McArdle, H., Mullaney, I., Magee, A., Unson, C., and Milligan, G. GTP analogues mastocytoma cells, the a subunit of G@(G@a) is translocated from cause release of the alpha subunit of the GTP binding protein, G0, from the plasma membrane of NG1O8-15 cells. Biochem. Biophys. Rex. Commun., 152: 243—251, membranes to the soluble fraction when the cells are stimulated by @3 1988. adrenergic agonist and prostacyclin receptor ligands (19—21).In neu 17. Milligan, G., Mullaney, I., Unson, C. G., Marshall, L., Spiegel, A. M., and McArdle, roblastoma tissues, significant amounts of G0 a are found in the H. GTP analogues promote release of the a subunit of the guanine nucleotide binding protein, G.2, from membranes ofrat glioma C6BU1 cells. Biochem. J., 254: 391—396, soluble fraction (6), and the present study showed that the percentage 1988. of G0 a in the soluble fraction (as opposed to the membrane-bound 18. Takahashi, S., Hashida, IC, Yatsunami, K., Fukui, T., Negishi, M., Katada, T., Ui, M., fraction) increased with advanced tumor stage. The amino-terminal Kanaho, Y., Asano, T., and Ichikawa, A. Characterization of cytosolic pertussis toxin-sensitive GTP-binding protein in mastocytoma P-815 cells. Biochim. Biophys. glycine of the a subunits of G0 and G, is myristoylated (1, 2), while Acta, 1093: 207—215,1991. the cysteine residue near the amino termini of the a subunits of G0, G., 19. Rananas, L. A., Svoboda, P., Jasper, J. R., and Inset, P. A. Stimulation of @-adrenergic receptors of 549 lymphoma cells redistributes the a subunit of the stimulatory G and G, is palmitoylated (22, 23). These lipid modifications serve to protein between cytosol and membranes. Proc. Natl. Acad. Sd. USA, 86: 7900—7903, anchor the a subunits to the membrane. Because the release of the a 1989. subunit of G proteins from membranes was mostly observed in tumor 20. Levis, M. J., and Bourne, H. R. Activation of the a subunit of G, in intact cells alters its abundance, rate of degradation, and membrane avidity. J. Cell Biol., 119: 1297— cells, there may be an acyltransferase defect in advanced tumors. 1307,1992. However, mean soluble G0 a concentration (ng/mg soluble protein) 21. Negiahi, M., Hashimoto, H., and Ichikawa, A. Translocation of a subunits of the also decreased with disease stages as total G0 a concentration but stimulatory guanine nucleotide-binding proteins through stimulation of the proatacy dine receptor in mouse mastocytoma cells. J. Biol. Chem., 267: 2364—2369, 1992. soluble G0 a concentrations had no correlation with patient outcome. 22. Linder, M. E., Middleton, P., Hepler, J. R., Taissig, R., Gilman, A. G., and Mumby, Our study also showed that total G0 a concentration correlated with S. M. Lipid modifications of G proteins: a subunits are palmitoylated. Proc. Natl. Acad. Sci. USA, 90: 3675—3679,1993. patient age, one of the important prognostic factors in neuroblastoma. 23. Parenti, M., Vigano, M. A., Newman, C. M. H., and Magee, A. I. A novel N-terminal Patients younger than 1 year had a better prognosis than older patients, motif for palmitoylation of G-protein a subunits. Biochem. J., 291: 349—353, 1993.

2336

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1994 American Association for Cancer Research. The α Subunit of GTP-binding Protein Go in Neuroblastoma: Correlation with Advanced Disease Stage

Yukio Ishiguro, Kanefusa Kato, Tomiko Asano, et al.

Cancer Res 1994;54:2334-2336.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/54/9/2334

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/54/9/2334. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1994 American Association for Cancer Research.