[CANCER RESEARCH 48. 2998-3001, June, 1988] Variants of Phosphohexose Isomerase in Gastrointestinal and Mammary Carcinoma: Isoelectric Focusing Patterns of Normal and Tumor Tissues Derived from Surgical Specimens of the Same Patient1

Matthias Baumann, Damián Jezussek, Ralf-Torsten Richter, and Karl Brand2

Institute of Biochemistry, Medical Faculty, University of Erlangen-Nuremberg, Federal Republic of Germany

ABSTRACT H2PO4 H2O, 5; MgSO4 7H2O,0.77; Krebs saline buffer: H2O, 1:4; pH 7.5). The occurrence of charge variants of phosphohexose isómeras? was Then the homogenate was sonicated by 6 pulses of 10 s each at 50 elucidated in cancerous and, for comparison, in noncancerous tissues W with cooling intervals of 1 min to make sure that the temperature of the preparation remained below 10'C. Subsequently, the extract was obtained from the same organ. Surgical specimens from 35 patients with gastrointestinal and mammary carcinoma were studied by means of the rehomogenized and centrifuged at 12,000 x g for 10 min. The super isoelectric focusing (IEF) technique. Differences in the IEF pattern could natant was dialyzed against 1% glycine buffer, pH 7.5, overnight at 4'C. LKB Multiplier 2117, LKB Power Supply Unit 2103, and Servalyt be demonstrated in 90% of the patients with gastric cancer. Correspond ing values for the patients with colorectal and mammary carcinoma were Precotes, pH 3-10, were used for the isoelectric focusing experiments. 50 and 73%, respectively. Almost all normal specimens proved to be A Julabo Paratherm FT20p electronic thermostat served to cool the plate to 4'C during the run. monomorphic, revealing only the major basic band with a pi of 9.1. In most of the tumors, additional bands of pi 8.9 and 8.6 were found. This The dialyzed probes were diluted with 2% LKB ampholine, pH 3- "typical" IEF pattern in tumor tissues was observed in 60% of gastric 10, achieving a final PHI activity of 1000-3000 »¿unitsina sample and 73% of mammary tumors, but in only 36% of colorectal tumors. The volume of 20 ¡A.Aconstant power of 4.0 W was preset and the Precote results obtained suggest that the IEF pattern of phosphohexose isomerase was prefocused until the voltage reached 500 V. Then the probes were in tumor tissue might be a useful tool for tumor diagnosis. applied and a voltage limit of 2000 V was chosen. When this voltage was reached the run was carried out for a further 30 min, so that the whole procedure took 3-4 h. INTRODUCTION Subsequently the precote was specifically stained for PHI, according to the method described by Carter and Parr (25), using phenazine The analysis of electrophoretic variants of the glycolytic methosulfate and the tetrazolium salt 3-(4,5-dimethyl-2-thiazolyl)-2,5- enzyme PHI3 is a widespread method in human genetics and diphenyl-2H-tetrazolium bromide. For each run, the protein standard internal medicine (1-7). Furthermore, serum PHI activity is a Test Mix 9, supplied by Serva, was used to determine the isoelectric well established tumor marker in oncology (8-19). The molec points (pi) of the PHI bands. This standard was visualized by Coomassie ular and kinetic properties of human PHI were investigated by staining. PHI activity was assayed spectrophometrically according to using hemolysates and muscle extracts as enzyme source (20- Tilley et al. (2). 22). There are relatively few reports dealing with the quantita tive and, in particular, qualitative comparison of PHI in normal and tumor tissues derived from surgical specimens of the same RESULTS patient (23, 24). Thirty-five patients with malignant tumors were included in The purpose of the present study was to evaluate the occur rence of PHI variants, comparing authentic normal and tumor this comparative study, 10 with gastric, 14 with colorectal, and tissues derived from the same organ immediately after surgical 11 with mammary carcinoma. By IEF, three charge variants of removal. Tissue specimens from 35 patients with gastric, colo PHI could be demonstrated in the tissues studied; a major basic rectal, and mammary carcinoma were studied by means of the band A with an isoelectric point of 9.1, and two more acidic IEF technique. bands B and C, with a pi of 8.9 and 8.6, respectively. In almost all normal specimens only the major band A could be found. In most tumor tissues additional bands could be detected. On MATERIALS AND METHODS the basis of this observation a "typical" IEF pattern of PHI, which is illustrated in Fig. 1, was defined as follows: normal Normal (noncancerous) and tumor tissue probes from the same organ tissue containing only band A and tumor specimen revealing of the patient were obtained immediately after surgical removal and two additional bands B and C. histopathological examination of each specimen. Only solid, surgically The IEF patterns of PHI found in normal and tumor tissue and histologically homogeneous tumor tissue, and closest noninvolved, histologically normal tissue as control were used. Blood and coagula specimens derived from the same organ of the patients are given in Tables 1-3. Table 1 shows that for gastric cancer PHI band were removed very thoroughly. All further steps were carried out at 4'C. Normal and carcinoma specimens were prepared identically. An A could be detected in all normal and tumor probes, whereas aliquot of 400 mg of each specimen was minced and homogenized in a band B is found in all tumor samples but only in 20% of the 10-fold volume of hypotonie Krebs saline buffer (in HIMconcentrations: normal specimens. Band C appeared in 8 of 10 malignant NaCl, 136; KC1, 4.7; CaCl2 2H2O, 1.29; K2HPO4 3H2O, 1.18; Na- probes but only in 1 of 10 benign tissues. In Table 2 the results obtained from the patients with colo Received 6/16/87; revised 12/10/87; accepted 2/23/88. The costs of publication of this article were defrayed in part by the payment rectal carcinoma are listed. In all 14 cases the normal tissue of page charges. This article must therefore be hereby marked advertisement in contained only band A. With respect to the tumor specimens, accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1This work was supported by the Marohn-Stiftung Grant Nr. Br/85. additional bands B and C were found, however, in only SOand 2To whom requests for reprints should be addressed, at Institute of Biochem 36% of the patients. In one instance only band B and C, but istry, Medical Faculty, Fahrstraße17, D-8520 Erlangen, Federal Republic of not band A (pi 9.1) could be detected. Germany.. 3The abbreviations used are: PHI, phosphohexose isomerase; IEF, isoelectric Table 3 illustrates that all normal breast tissues revealed band focusing. A exclusively. In 73% of the malignant breast tissues band B 2998

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1988 American Association for Cancer Research. VARIANTS OF PHI IN CANCER

GASTRIC CARCINOMA COLORECTAL CARCINOMA MAMMARY CARCINOMA

••

NORMAL TUMOR NORMAL TUMOR NORMAL TUMOR Fig. 1. Isoelectric focusing patterns of phosphohexose isomerase in normal (noncancerous) and tumor tissues from patients with gastrointestinal and mammary carcinoma. Aliquots of extracts from normal and tumor tissues were subjected to isoelectric focusing on Servalyt Precotes (3-10), as described in "Materials and Methods." Band A, pi 9.1; Band B, pi 8.9; Band C, pi 8.6.

Table 1 Isoelectricfocusing patterns of phosphohexose isomerase in normal and Table 3 Isoelectricfocusing patterns of phosphohexose isomerase in normal and tumor tissues from patients with gastric carcinoma tumor tissues from patients with mammary carcinoma TumorBand B+*--A° Band B BandC BandA Band Patient1234S678910n%NormalBandB+»+A* Band C BandC+ A Band B Band Patient1234567891011n%Normal ++ + +— + —+ ++ + + ++ — — + ++ +— + + —+ — — + ++ +— + + ++ — — + —4. +— + + ++ - - + —+ —— + + —+ — — + —+ +— + + ++ — — + —+ —- + -t- ++ — — + —+10 +++ + —+ — — + +1 + ++ — — + +11 + 2100 810 10 10 8100 0 0 11 20TumorBand 100 100 80 " Band A, pi 9.1; Band B, pi 8.9, Band C, pi 8.6. 0 0 100 73BandC++—++—++—++873 * +, presence of band; —,absenceof band. " BandA, pi 9.1; Band B,pi 8.9; BandC, pi 8.6. +, presenceof band; -, absenceof band.

Table 4 Comparison of isoelectricfocusing patterns of phosphohexose isomerase in normal and tumor tissues obtained by surgical removal from the same patient The typical IEF pattern of PHI is defined: in the noncancerous tissue only band A, but bands A, B, and C in the tumor tissue. Table 2 Isoelectricfocusing patterns of phosphohexose isomerase in normal and tumor tissues from patients with colorectal carcinoma of patients with differentIEF pat of patients with terns in normal typicalIEF pat TumorGastric and tumortissues90 tern60 Patient1234S6791011121314m%NormalBandB+»-1-A* Band C BandC+ A Band B Band +— •+ carcinoma —+ +— -1- + Colorectalcarcinoma 14 SO 367351 —+ —— + — MammarycarcinomaAlln10 1135% 7369% —++ —— + — —— + — —+* —- + — -! + - I+ -- + + and C could be detected in addition to band A. These data are -+ -—+ - summarized in Table 4. —+ —-+ + In 90% of the patients with gastric carcinoma a difference in —+ +—+ + the IEF pattern of PHI between normal and tumor tissue —+14 +- — + -0 + - derived from the same patient could be demonstrated. Corre sponding values for colorectal and mammary carcinoma were 0100 50 13 7 50 and 73%, respectively. The "typical" IEF pattern of PHI as

0TumorBand 93 50 36 defined above (Fig. 1) occurred in 60 and 73% of all cases with * Band A, pi 9.1; Band B, pi 8.9; Band C, pi 8.6. gastric and breast carcinoma but only in 36% of colorectal * +, presence of band; —,absenceof band. tumors. 2999

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1988 American Association for Cancer Research. VARIANTS OF PHI IN CANCER

DISCUSSION pattern of PHI found only in mice oocytes which undergo the postnatal growth phase might be a potential marker of oocyte Human PHI, derived from erythrocyte and muscle, revealed growth. a dimeric structure composed of probably identical subunits, On the posttranslational level, Cooper et al. (32-34) reported each with a molecular weight of about 60,000. The pi was the tyrosine-specific phosphorylation of proteins as a common reported to be rather basic in the region of 9.0 (2, 21, 22). phenomenon in virus-transformed cells. Several of the products Charge variants were found, detected by specific staining after of oncogenes were found to be tyrosine-specific protein kinases. electrophoresis or by using Chromatographie methods (1-5, 21, In this context it is of interest that glycolytic enzymes, e.g., 22). The majority of the authors explained those variants as the enolase and phosphoglycerate mutase, which are substrates of existence of a single autosomal gene locus and the dimeric these kinases, have a more acidic pi (32, 33). Unfortunately, structure of the enzyme, resulting in one electrophoretic band these authors were not able to study other glycolytic enzymes, in homozygous and three bands in heterozygous individuals (1, such as PHI, because of a lack of appropriate antisera. 3, 5, 22). Some investigators studied the activity and the elec With regard to point c, it is important to note that charge trophoretic behavior of PHI from different human organs and variants of PHI with different stability toward inactivating found identical electrophoretic patterns and comparable activ conditions are known from human genetics (35). In addition, ities (26, 27). With regard to the comparison of tumor and there are reports indicating that the levels of proteolytic en normal tissue, many reports in the literature deal with differ zymes in malignant transformed cells are different from those ences in enzyme activities, but to our knowledge, only Schwartz in the cognate cells (36, 37). In this sense, the charge variants and Bodansky (24) reported the occurrence of PHI variants by obtained would be "stability variants." means of starch gel electrophoresis. In a small percentage of Artifacts arising during the isolation procedure can be ruled cancer tissues they found at pH 8.6 a second band migrating out for the following reasons. Unspecific proteolysis or modi more slowly than the major basic band present in all normal fication of the enzyme such as oxidation of SH groups should and tumor probes. occur randomly in normal and tumor probes but this was not Using the high resolution capacity of the isoelectric focusing the case. Furthermore, oxidation of SH groups, by some au technique, we were able to demonstrate charge variants of PHI thors, suggested as a common cause of "pseudoisoenzyme" in a majority of carcinoma specimens compared with the cor formation (26, 38) could be excluded by control experiments in responding normal tissues from the same individual (Tables 1- which dithiothreitol was present during the whole preparation 4). and electrophoresis procedure. Benign tissues containing rap Noncancerous tissue probes of all three organs proved to be idly dividing cells such as intestinal mucosa (controls for colo very monomorphic. With the exception of two gastric speci rectal carcinoma; Table 2) showed only one band, suggesting mens, all probes showed the major band A exclusively. This that the variants of PHI observed in the tumors are not a feature finding is in good agreement with the results of other authors of proliferation per se but are related to malignancy. (26, 27), who found no organ specific PHI variants as is From the presence of PHI variants in most cases of tumor commonly known for pyruvate kinase or lactate dehydrogenase. specimens it therefore appears that this shift of IEF pattern The major band A is also found in all but one of the malignant might be a useful marker for malignant growth. Further inves probes from the three organs. Surprisingly, 90% of the gastric tigations dealing with the isolation and biochemical character and 73% of the mammary carcinoma samples revealed one or ization of those variants are in progress. two additional, more acidic PHI bands. These findings led to the definition of a typical IEF pattern of PHI in a tumor- bearing organ, as illustrated in Fig. 1 and Table 4. This typical ACKNOWLEDGMENTS pattern is observed in about two-thirds of the patients with gastric and mammary carcinoma but only in about one-third of The authors would like to thank Dr. F. P. Gall et al. from the Department of Surgery, University of Erlangen-Nuremberg for provid all cases of colorectal cancer. ing the tissue specimens. Thanks also to Dr. P. Hermanek and Dr. J. To explain this finding, three possibilities can be discussed. Giedl from the Department of Surgical Pathology, University of Erlan (a) The PHI variant found in cancer tissue is based on altera gen-Nuremberg, for the histopathological examination of the probes. tions of the genome of the tumor cell; (b) the PHI variant is Special thanks to Dr. Morton K. Schwartz from the Memorial Sloan- caused by posttranslational alterations of the enzyme in the Kettering Cancer Center, New York, for the fruitful discussions. The tumor cell; and (c) the different IEF patterns are due to marked excellent technical assistance of F. Wartenberg is gratefully acknowl differences in stability of the charge variants. edged. As mentioned above, in hemolysates of healthy individuals, normally only one band in homozygotes or a triplet pattern in REFERENCES hétérozygotesisfound. These data correspond to the observed typical IEF pattern in normal and tumor probes. This may lead 1. Mohrenweiser, H. W., Wade, P. T., and Wurzinger, K. H. Characterization to the presumption that a mutation of one alíele,occurring in of a seríesofelectrophoretic and enzyme activity variants of human glucose- the tumor cells results in a "heterozygous" PHI pattern in the phosphate isomerase. Hum. Genet., 75: 28-31, 1987. 2. Tilley, B. E., Gracy, R. W., and Welch, S. G. A point mutation increasing cancer tissue of the same homozygous individual. The fact that the stability of human phosphoglucose isomerase. J. Biol. Chem., 249:4571- mutations occur in the process of malignant transformation is 4579, 1974. 3. Detter, J. C., Ways, P. O., Giblett, E. R., Baughan, M. A., Hopkinson, D. well established, e.g., in the activation of protooncogenes in A, Povey, S., and Harris, H. Inherited variations in human phosphohexose virus-transformed cells (28, 29). Another explanation might be isomerase. Ann. Hum. Genet., 31: 329-338, 1968. the "switching on" of a second PHI gene which is not expressed 4. Satoh, <'..and Mohrenweiser, H. W. Genetic heterogeneity within an electro phoretic phenotype of phosphoglucose isomerase in a Japanese population. in normal cells but is activated in transformed cells. The exist Ann. Hum. Genet., 42: 283-292,1979. ence of a c/s-active regulator gene linked to the structural PHI 5. Fitch, L. I., Parr, C. W., and Welch, S. G. Phosphoglucose isomerase variation in man. Biochem. J., 110: 56, 1968. locus in the genome of mice is postulated by several authors 6. Baughan, M. A., Valentine, W. N., Paglia, D. E., Ways, P. O., Simons, E. (30, 31). They suggest that the shift of the electrophoretic R., and DeMarsh, Q. B. Hereditary hemolytic anemia associated with glucose 3000

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1988 American Association for Cancer Research. VARIANTS OF PHI IN CANCER isomerase (GPI) deficiency—a new enzyme defect of human erythrocytes. 21. I'sulini. K. K., Fukunaga, K., and Chervenka, C. H. Phosphoglucose isomer Blood, 32: 236-249, 1968. ase from human erythrocyte. Preparation and properties. J. Biol. Chem., 7. Arnold, H. Inherited glucosephosphate isomerase deficiency. Blut, 39: 405- 246:7586-7594,1971. 417, 1979. 22. Carter, N. D., and Yoshida, A. Purification and characterization of human 8. Bodansky, ().. and Schwartz, M. K Phosphohexose isomerase: clinical phosphoglucose ¡somerase.Biochim. Biophys. Acta, 181: 12-19, 1969. aspects. In: W. A. Wood, (ed.), Methods in Enzymology, Vol. 9, pp. 568- 23. Balinsky, D., Cayanis, E., Geddes, E. W., and Behrsohn, I. Activities and 575. New York: Academic Press, Inc., 1966. isoenzyme patterns of some enzymes of glucose metabolism in human 9. Schwartz, M. A., West, M., Walsh, W. S., and Zimmermann, H. J. Serum primary hepatoma. Cancer Res., 33: 249-255, 1973. enzymes in disease. VIII. Glycolytic and oxidative enzymes and transami- 24. Schwartz, M. K., and Bodansky, O. Relationship of the electrophoretic nases in patients with gastrointestinal carcinoma. Cancer (Phila.), 15: 346- patterns of phosphohexose isomerase and glutamic oxaloacetic transaminase 353, 1962. in human tissues to patterns in serum of patients with neoplastic disease. 10. Schwartz, M. K., Greenberg, E., and Bodansky, O. Comparative values of Am. J. Med., 40: 231-241, 1966. phosphatases and other serum enzymes in following patients with prostatic 25. Carter, N. D., and Parr, C. W. Isoenzymes of phosphoglucose isomerase in carcinoma. Consideration of phosphohexose isomerase, glutamic oxalacetic mice. Nature (Lond.), 216: 511, 1967. transaminase, isocitric dehydrogenase, and acid and alkaline phophatases. 26. Arnold, H., Hoffmann, K. G., Blume, R. E., and Lohr. G. W. Zur Frage Cancer (Phila.), 16: 583-594, 1963. gewebespezifischer Isoenzyme der Glucosephosphatisomerase des Menschen. 11. Griffith, M. M., and Beck, J. C. The value of serum phosphohexose isomerase Klin. Wochenschr., 52: 149-151, 1974. as an index of metastatic breast carcinoma activity. Cancer (Phila.), 16: 27. Payne, M. D., Porter, D. W., and Gracy, R. W. Evidence against the 1032-1041, 1963. occurrence of tissue-specific variants and isoenzymes of phosphoglucose 12. .lupini. G. F., and Jegatheesan, K. A. Serum glycolytic enzymes and acid isomerase. Arch. Biochem. Biophys., 122:122-127, 1972. phosphatases in mammary carcinomatosis. Br. Med. J., 5281: 827-831, 28. Alitalo, K., Koskinen, P., Mäkelä,T.P., Saksela, K., Sistonen, I... and 1962. Winqvist, R. myc oncogenes: activation and amplification. Biochim. Biophys. 13. Rose, A., West, M., and Zimmermann, H. J. Serum enzymes in disease. V. Acta, 907: 1-32, 1987. Isocritric dehydrogenase, malic dehydrogenase, and glycolytic enzymes in 29. Weinberg, R. A. Molekulare Grundlagen von Krebs. Spektrum Wiss., /: 58- patients with carcinoma of the breast. Cancer (Phila.), 14: 726-733, 1961. 71, 1984. 14. Caffier. H., and Brandau, H. Serum tumor markers in metastatic breast 30. Peterson, A. C., and Wong, G. G. Genetic regulation of glucosephosphate cancer and course of disease. Cancer Detect. Prev., 6:451-457, 1983. isomerase in mouse oocytes. Nature (Lond.), 276: 267-269, 1978. 15. Paulick, R., Caffier, H., and Paulick, M. Serum CEA and PHI as tumor 31. Buehr, M., and McLaren, A. Expression of glucose-phosphate isomerase in markers in the follow up of breast cancer patients. In: H. Peelers (ed.), relation to growth of the mouse oocyte in vivo and in vitro. Gamete Res., //. Protides of the Biological Fluids. Proceedings of the Thirty-second Collo 271-281, 1985. quium 1984, Vol. 32, Section B5, pp 707-710. Elmsford: Pergamon Press, 32. Cooper, J. A., and Hunter, T. Regulation of cell growth and transformation 1984. by tyrosine-specific protein kinases: the search for intracellular substrate 16. Damle, S. R., Talavdekar, R. V., and Panse, T. B. Studies on glycolytic proteins. Curr. Top. Microbiol. Immunol., 707:125-161, 1983. enzymes in relation to cancer. Part III. Clinical significance of serum phos 33. Cooper, J. A., Reiss, N. A., Schwartz, R. J., and Hunter, T. Three glycolytic phohexose isomerase and aldolase in cancer of the prostate. A comparative enzymes are phosphorylated at tyrosine in cells transformed by Rous sarcoma study with serum acid phosphatase and serum alkaline phosphatase. Indian virus. Nature (Lond.), 302: 218-223, 1983. J. Cancer, 8: 176-184, 1971. 34. Hunter, T. Die Proteine der Onkogene. Spektrum Wiss., IO: 106-119,1984. 17. Schwartz, M. K. Laboratory aids to diagnosis—enzymes. Cancer (Phila.), 35. Bonné-Tamir,B., Papiha, S. S., Ashbel, S., Brok-Simoni, F., Kende, G., and 3 7: 542-548, 1976. Ramot, B. PGI*3 (Israel), a new, unstable alíelein the phosphoglucose 18. Munjal, I >..Chawla, P. I... and Zamcheck, N. Carcinoembryonic antigen and isomerase system. Hum. Genet., 77: 76-79, 1987. phosphohexose isomerase, gamma glutamyl transpeptidase and lactate de 36. Bosmann, II B., and Hall, T. C. Enzyme activity in invasive tumors of hydrogenase levels in patients with and without liver métastases.Cancer human breast and colon. Proc. Nati. Acad. Sci. USA, 71: 1833-1837, 1974. (Phila.), 37: 1800-1807, 1976. 37. Bosmann, H. B. Elevated glycosidases and proteolytic enzymes in cells 19. Bowen, M., de Mello, J., and Cooper, E. H. The clinical significance of transformed by RNA tumor virus. Biochim. Biophys. Acta, 264: 339-343, phosphohexose isomerase levels in gastrointestinal and breast cancer. 1972. TumorDiagnostik, 2: 22-26, 1981. 38. Blackburn, M. N., Chirgwin, J. M., James, G. T., Kempe, T. D., Parsons, T. 20. Kahana, S. E., Lowry, O. H., Schulz. D. W., Passoneau, J. V., and Crawford, F., Register, A. D., Schnackertz, K. D., and Noltmann, E. A. Pseudoisoen- E. J. The kinetics of phosphoglucoisomerase. J. Biol. Chem., 235: 2178- zymes of rabbit muscle phosphoglucose isomerase. J. Biol. Chem., 247: 2184, 1960. 1170-1179, 1972.

3001

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1988 American Association for Cancer Research. Variants of Phosphohexose Isomerase in Gastrointestinal and Mammary Carcinoma: Isoelectric Focusing Patterns of Normal and Tumor Tissues Derived from Surgical Specimens of the Same Patient

Matthias Baumann, Damian Jezussek, Ralf-Torsten Richter, et al.

Cancer Res 1988;48:2998-3001.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/48/11/2998

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/48/11/2998. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1988 American Association for Cancer Research.