| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Biochemistry |
Central Laboratories for Key Technology, KIRIN Brewery Co., Ltd., Kanazawa-ku, Yokohama 236-0004 [S. T., S. O., M. T. M., A. Y., K. N., M. T.], and Tokyo Metropolitan Institute of Gerontology, Itabashi-ku, Tokyo 173-0015 [A. K.], Japan
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-mannosidase II in normal human placentae and three human choriocarcinoma cell lines were investigated. GnT-IV activities in choriocarcinoma cell lines were increased from 16- to 66-fold and GnT-III activity was increased from 15- to 25-fold as compared with those in human placentae, whereas other enzyme activities were not increased significantly. The mRNA expression levels generally correlated with their enzyme activities. Among the two GnT-IV genes found in human tissues only GnT-IVa gene was strongly expressed in the cancer cells: from three to seven times as much as in the normal tissue, whereas that of GnT-IVb remained constant. On the basis of these results, we proposed that ectopic expression of GnT-IVa gene should occur along with the malignancy of trophoblastic tissues, and that the increased GnT-IV activity should be the main cause of the formation of abnormal biantennary sugar chains in choriocarcinoma. A possible enzymatic basis of the biosynthesis of abnormal biantennary sugar chains is discussed. |
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
- and ß-subunits, similar to the three mammalian pituitary glycohormones: lutenizing hormone, follicle-stimulating hormone, and thyroid-stimulating hormone (8)
. Each subunit of hCG contains two asparagine-linked sugar chains, and the ß-subunit also contains four serine-linked sugar chains. These sugar chains play important roles in expressing the biological activity of hCG (9
, 10)
.
Comparative studies of the oligosaccharides that are released by hydrazinolysis from hCG samples purified from the urine of pregnant women and patients with trophoblastic diseases revealed that extensive structural alteration exists in the asparagine-linked sugar chains of tumor hCGs (11)
. hCGs that are obtained from pregnant women and patients with hydatidiform mole contain the sialylated forms of oligosaccharides A (without the Fuc1–6 residue) and B in Fig. 1
. The hCGs from patients with invasive mole contain the sialylated forms of oligosaccharides A, B, and D. The hCGs from patients with choriocarcinoma contain either sialylated or nonsialylated forms of all of the eight oligosaccharides shown in Fig. 1
. These results indicated that an abnormal expression of GnT -IV is the key to alter the glycosylation of hCG in the malignant trophoblastic diseases. Because oligosaccharides C and D were not detected in the hCGs from pregnant women and patients with hydatidiform mole, GnT-IV, which catalyzes the formation of the GlcNAcß1–4 Man1–3 group, should not be expressed in their hCG producing cells. Presence of oligosaccharides D in the hCGs from invasive mole patients indicated that the enzyme is abnormally expressed in this disease. Expression of oligosaccharides C together with oligosaccharides D in the hCGs from choriocarcinoma indicated that the abnormally expressed GnT-IV in choriocarcinoma can act on monoantennary sugar chains as well as on biantennary sugar chains. It was reported by Gleeson and Schachter (12)
that GnT-IV solubilized from the Golgi membrane can use monoantennary sugar chains as acceptors. However, oligosaccharide C has not been detected in the glycoproteins produced by various normal cells. Hence, we called them "abnormal biantennary sugar chains," expecting them to become important tumor markers in the future. Actually, the abnormal biantennary sugar chains were later found in the 
-glutamyltransferase purified from human hepatoma (13)
and in the carcinoembryonic antigen obtained from colon cancer (14)
. Therefore, it is important to investigate the control mechanism that prevents formation of the abnormal biantennary sugar chains in normal cells.
|
Recently, we have established a sensitive assay method for GnT-IV (16) . By using this method, we successfully purified GnT-IVa from bovine small intestine (17) , and cloned its cDNA (18) . We also succeeded in cloning the cDNAs of human GnT-IV (19 , 20) . Unlike other GnTs, mammalian GnT-IV gene constructs an active gene family consisting of GnT-IVa and GnT-IVb genes (19 , 20) . Although precise enzymatic differences between the two gene products are under investigation, it is quite interesting to know which gene product should be the cause to form the abnormal biantennary sugar chains. In this study, we tried to ascertain the enzymatic and genomic background of the formation of abnormal biantennary sugar chains by analyzing and comparing related enzyme activities and their mRNA expression levels between choriocarcinoma cell lines and a normal placenta.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Determination of Activities of GnTs, GalT, and -Man’ase II.
The activities of GnT-I and GnT-II were measured by the method of Schachter et al. (21)
except that Man1–6(Man1–3)Manß1–4GlcNAcß1–4GlcNAc-PA(core-PA) and Man1–6(GlcNAcß1–2 Man1–3)Manß1–4GlcNAcß1–4GlcNAc-PA[Gn (2)core-PA] were used as substrates for GnT-I and GnT-II, respectively, at a concentration of 0.8 mM, and the incubation time was 1 h. The assays of GnT-III, -IV, and -V were performed according to the method of Oguri et al. (17)
, which was a modification of the method of Nishikawa et al. (22)
. The substrate Gn2(2',2)core-PA [GlcNAcß1–2 Man1–6(GlcNAcß1–2 Man1–3)Manß1–4GlcNAc ß1–4GlcNAc-PA] was prepared as reported previously (16)
. GalT activity was assayed as follows. Enzyme solution (3 µl) was incubated at 37°C for 30 min with 10 mM HEPES buffer (pH 7.2), containing 0.8 mM Gn2(2',2)core-PA, 10 mM UDP-Gal, 33 mM NaCl, 3 mM KCl, 1.5% Triton X-100, 5.5 mM D-Galactono-1,4-lactone, and 10 mM MnCl2 in a total volume of 20 µl. -Man’ase II activity was measured according to the method of Chui et al. (23)
. The substrate [Man1–6(Man1–3)Man1–6](Man1–3)Manß1–4GlcNAcß1–4GlcNAc-PA was kindly provided by Dr. Y. Chiba (KIRIN Brewery, Yokohama) and used as a substrate at a concentration of 10 µM; the incubation period was 3 h at 37°C.
Northern Blot Analysis.
Poly(A)+RNA was extracted from three choriocarcinoma cell lines using Fast Track 2.0 kit (Invitrogen). As a normal tissue control, we used human placenta Poly(A)+RNA purchased from Clontech. Poly(A)+RNA (2 µg) was denatured in formamide/formaldehyde and electrophoresed on 1% agarose/formaldehyde gels. The separated RNA was transferred to a Hybond N+ nylon membrane (Amersham). Probes for the genes of related enzymes were prepared by PCR Radioactive Labeling System (Life Technologies, Inc.) with [-32P]dCTP (Amersham). The temperature cycle was as follows: 94°C for 30 s, 60°C for 75 s, and 72°C for 2 min for 30 cycles after an initial denaturation for 10 min at 95°C on a Zymoreactor II (Atto, Tokyo, Japan). The sense/antisense primers for labeling PCR were designed according to the published human sequences as follows:
(a) GnT-I, 5'-GGTGGAGAAAGTGAGGACCAATG-3'/5'-ACTGGAAGGTGACAATACCCCG-3';
(b) GnT-II, 5'-ATACCTCAGACTGCTGCTGGACTC-3'/5'-CAGGTCTCTGGGACAATCTCTAGG-3';
(c) GnT-III, 5'-ACGGCGTCCTTTTCCTCAAG-3'/5'-CGGTTCTCATACTGTCTGAAG-3';
(d) GnT-IVa, 5'-GGCTATCACACCGATAGCTGGAG-3'/5'-TCCACCATTCCTTCTGCAACACC-3';
(e) GnT-IVb, 5'-ACAACCCTCAGTCAGACAAGGAGG-3'/5'-GGTACCCTCAGAAGCCCGCAGCTT-3';
(f) GnT-V, 5'-TGCGAGGGGATGCTACAGAGAATC-3'/5'-CCTTGTTGAGGTGCTGGAAGAAAG-3';
(g) GalT, 5'-GGTGTTTTTCACAGCCACGG-3'/5'-TGTCATCATCTTCTCCTCCCCAG-3';
(h) -Man’ase II, 5'-CGGAAGAAGAAAAGAAGTCGGTC-3'/5'-CCTCATTGGAGTGCCCATTG-3';
(i) glyceraldehyde-3-phosphate dehydrogenase (G3PDH), 5'-CCAAAATCAAGTGGGGCGATG-3'/5'-CAGGAGGCATTGCTGATGATCTTG-3'; and
(j) ß-actin, 5'-TGAAGTGTGACGTGGACATCCG-3'/5'-GCTCAGGAGGAGCAATGATCTTG- 3'.
The blots were hybridized with each probe in Rapid-hyb buffer (Amersham) at 65°C for 3 h and then washed three times in 2x SSC/0.1% SDS at 65°C for 15 min, once in 1x SSC/0.1% SDS at 65°C for 30 min, and 3 times in 0.1x SSC/0.1% SDS at 65°C for 15 min. To compare the expression level of enzymes among tissue and cell lines, the emissions of the corresponding enzyme messages were measured using a BAS 2000 bioimaging analyzer (Fuji Film, Tokyo, Japan) and normalized to that of ß-actin mRNA.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-Man’ase II are involved in the biosynthesis of abnormal biantennary sugar chains. As shown in Fig. 2
, the activities of these enzymes were measured in normal placentae and three choriocarcinoma cell lines (JAR, JEG-3, and BeWo) as models of choriocarcinoma tissue. Compared with the values of GnT-I and -II activities in seventeen human cell lines [3,130
33,720 (mean, 13,260) pmol/h/mg protein and 3,23015,820 (mean, 9,510) pmol/h/mg protein, respectively5
], normal placentae have high GnT-I activity (25,340 pmol/h/mg protein) but contain low GnT-II activity (1,670 pmol/h/mg protein). Possibly, the low GnT-II activity is the enzymatic basis of the formation of monoantennary sugar chain found in the hCG produced by a normal placenta. GnT-IV activities in choriocarcinoma cells were increased from 16- to 66-fold over those in normal placentae. The GnT-IV activities of these cancer cells (1,920 to 7,930 pmol/h/mg) are the highest among the various human cancer cell lines investigated in our laboratory (the values of the 17 human cell lines were 805,210 pmol/h/mg protein, mean = 750 pmol/h/mg protein). Such a strong activity of GnT-IV should affect the branch formation of sugar chains in these cancer cells.
|
Activities of the other enzymes were not increased significantly in the choriocarcinoma cells when compared with normal placentae. Therefore, the major difference in the enzymes involved in the formation of abnormal biantennary sugar chains between normal placentae and the choriocarcinoma cells is the extraordinarily increased GnT-IV activity.
Northern Blot Analysis of mRNAs for the Enzymes Related to the Branch Formation of Asparagine-linked Sugar Chains in a Normal Placenta and in Choriocarcinoma Cell Lines.
Compared with the rather similar patterns of the enzyme activities among the three choriocarcinoma cells, the pattern of mRNA expression levels of the enzymes in each cell line is quite different (Fig. 3)
. In JAR cells, the mRNAs of GnT-IVa and -Man’ase II were significantly increased as compared with those of a normal placenta. In JEG-3 cells, only GnT-IVa mRNA was present in an increased concentration. In BeWo cells, the mRNAs of GnT-III and GnT-IVa were extremely enhanced. In any case, the mRNA of GnT-IVa, but not of GnT-IVb was strongly expressed in all of the three choriocarcinoma cell lines.
|
was generally correlated to the actual enzyme activity (Fig. 2)
, although there were slight discrepancies in the case of -Man’ase II in JAR cells and GnT-I in the three cell lines for unknown reasons. Although G3PDH is generally used as a control for measuring mRNA expression level, it could be overexpressed in established cell lines. Therefore, we used the ß-actin mRNA to normalize the expression levels of glycosyltransferases and glycosidase mRNAs. |
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Current knowledge of the biosynthesis of the complex type sugar chains is as follows. The glucosylated high mannose type sugar chains, added cotranslationally to a nascent polypeptide chain, are processed to Man5GlcNAc2 when the glycosylated polypeptide is transported to the Golgi apparatus. When the glycopeptide reaches the medial Golgi, GlcNAcß1–2 is added to the Man1–3 arm of this heptasaccharide by the catalytic action of GnT-I. Then the two -mannosyl residues are removed from the Man1–6 arm by the action of -Man’ase II to form the agalacto-monoantennary sugar chain (pathway I in Fig. 4
). More highly branched complex-type sugar chains are then formed from this monoantennary sugar chain as follows. Agalacto-biantennary sugar chain will be formed by the action of GnT-II (pathway II). Then the 2,4-branched triantennary sugar chain will be formed by the catalytic action of GnT-IV (pathway V).
|
1–6 residue in Fig. 1
), which occupies approximately one-fourth of all of the sugar chains of the hCGs from pregnant women (25)
. It was confirmed that a small amount of GnT-IV activity is expressed in placentae. However, the sugar patterns of hCGs from pregnant women indicated that the activity of GnT-IV is not strong enough to produce triantennary sugar chains.
In choriocarcinoma cells, in which GnT-IV activity is remarkably increased, an additional pathway may work as shown in Fig. 4
. The enzyme level is strong enough to form 2,4-branched triantennary sugar chains (pathway V). On the basis of the report of Gleeson and Schachter (12)
and ourselves (17)
, the highly expressed GnT-IV may also work on the agalacto-monoantennary sugar chain to form an abnormal biantennary sugar chain (pathway IV). By using a hen oviduct membrane preparation as an enzyme source, Allen et al. (26)
found that the pathways III and VI in Fig. 4
do occur to form the agalacto abnormal biantennary sugar chains. However, these pathways seem to unlikely to work in choriocarcinoma inasmuch as no intermediate hybrid-type oligosaccharide was detected in the urinary hCGs. Therefore, we concluded that pathway IV is the most likely one to form the abnormal biantennary sugar chains in choriocarcinoma cells. By in vitro experiments, we found that GnT-II can act on the agalacto abnormal biantennary sugar chains to form triantennary sugar chains (pathway VII). If pathway VII actually works in vivo, certain amounts of the triantennary sugar chains will be produced from the accumulated abnormal biantennary sugar chains, even if the normal pathway (pathways II and V) becomes restricted. Actually, from two to four times more triantennary sugar chains (structures D in Fig. 1
) than biantennary ones (structures B) were detected in the hCGs from choriocarcinoma patients (25)
, which is hard to explain without the contribution of the pathway VII.
The evidence that no abnormal biantennary sugar chain was detected in the hCGs from patients with invasive mole, despite having triantennary sugar chains, may well be explained by competition between GnT-II and GnT-IV for the common substrate, monoantennary structure. Although the weak GnT-II activity in normal placentae seems dominative against weaker GnT-IV activity, the ratio of sugar chains having GnT-II product was decreased in choriocarcinoma compared with normal placentae, which suggests that GnT-II activity was overwhelmed by enhanced GnT-IV activity in choriocarcinoma. In invasive mole, the moderate GnT-IV activity may not be strong enough to compete with GnT-II on monoantennary structures (pathway IV against II), however, it is sufficient for the pathway V. Alternatively it may also be explained by taking the pathway VII into account. Because the hCGs from invasive mole contained smaller amount of triantennary sugar chains than choriocarcinoma hCGs (15) , less GnT-IV may be expressed in invasive mole than in choriocarcinoma cells. Under such a circumstance, pathway IV may become the rate-limiting step, and the abnormal biantennary sugar chains will be most completely converted to the triantennary sugar chains by pathway VII.
It was very interesting to know which GnT-IV gene contributed the elevated GnT-IV activity in choriocarcinoma tissues. Among seventeen human cell lines tested previously, we found that the cells with high GnT-IV activity always expressed GnT-IVa message abundantly, whereas those with very weak enzyme activity showed the almost undetectable IVa message. On the other hand GnT-IVb presented housekeeping gene-like expression. Our results shown in Fig. 2
and 3
were consistent with this previous observation. Therefore, we think that the basal level of GnT-IV activity in the cells could be conveyed by GnT-IVb expression and the GnT-IVa gene plays an essential role in elevated GnT-IV activity.
Our present explanation of the altered glycosylation of hCGs in tumor cells can be summarized as follows. Normal placentae express a basal level of GnT-IV activity that is too low to form the abnormal biantennary and triantennary sugar chains. Such a low GnT-IV activity may come from the constitutively expressed GnT-IVb gene because the GnT-IVa gene seems inactive in normal tissue. Invasive mole should express a moderate amount of GnT-IVa and produce triantennary structures through normal pathway as well as through pathway VII. Choriocarcinoma cells express substantial amounts of the abnormal biantennary as well as the triantennary sugar chains because of the ectopic overexpression of GnT-IVa.
To confirm this theory, measurement of glycosyltransferases’ activities and their expression levels in invasive mole must be performed in the future. Investigation of the localization of the glycosyltransferases in the Golgi apparatus is also important, because studies on the substrate specificities of the solubilized enzymes may not always represent actual reactions in the Golgi apparatus of living cells.
Regarding the in vivo activity of hCG with abnormal sugar chains, Nishimura et al. (1) demonstrated the low activity of an hCG sample purified from the urine of a choriocarcinoma patient. It was later explained, however, that the decrease of the hormonal activity was caused by the loss of sialic acid residues (27) . No precise investigation has been performed thus far for the change in in vivo activity of hCG with sugar chains of abnormal branching structures. We would like to examine the relation of in vivo activity and branch structures of hCG in the future.
|
FOOTNOTES |
|---|
1 This work was supported by New Energy and Industrial Technology Development Organization (NEDO) as a part of the Research and Development Projects of Industrial Science and Technology Frontier Program. S. T. and M. T. M are research fellows of NEDO. 
2 Present address: Department of Bioproduction, Faculty of Bioindustry, Tokyo University of Agriculture, 196 Yasaka, Abashiri, Hokkaido 099-2493, Japan.
3 To whom requests for reprints should be addressed, at Central Laboratories for Key Technology, KIRIN Brewery Co., Ltd., 1–13-5 Fuku-ura, Kanazawa-ku, Yokohama 236-0004, Japan. Phone: 81-45-788-7228; Fax: 81-45-788-4047; E-mail: makotot{at}kirin.co.jp
4 The abbreviations used are: hCG, human chorionic gonadotropin; GnT, N-acetyl-glucosaminyltransferase; GnT-I, UDP-N-acetylglucosamine:-1,3-D-mannosideß-1,2-GnT; GnT-II, UDP-N-acetylglucosamine:-1,6-D-mannosideß-1,2-GnT; GnT-III, UDP-N-acetylglucosamine:ß-D-mannosideß-1,4-GnT; GnT-IV, UDP-N-acetylglucosamine:-1,3-D-mannosideß-1,4-GnT; GnT-V, UDP-N-acetylglucosamine:-1,6-D-mannosideß-1,6-GnT; GalT, UDP-Gal:N-acetylglucosaminide-ß-1,4-galactosyltransferase; -Man’ase II, -mannosidase II; PA, 2-aminopyridine; FBS, fetal bovine serum; G3PDH, glyceraldehyde-3-phosphate dehydrogenase.
5 S. Takamatsu, unpublished observations.
Received 1/13/99. Accepted 6/15/99.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
subunit. Biochem. Biophys. Res. Commun., 122: 1260-1267, 1984.[Medline]
subunit of urinary human chorionic gonadotropin and the alpha protein secreted by HeLa cells. Biochem. Biophys. Res. Commun., 98: 942-951, 1981.[Medline]
1,3-D-manno-side ß1,4-N-acetylglucosaminyltransferase (N-acetylglucosaminyltransferase-IV) from bovine small intestine. J. Biol. Chem., 272: 22721-22727, 1997.1,3-D-mannoside-ß1,4-N-acetylglucosaminyltransferase IV. J. Biol. Chem., 273: 11556-11562, 1998.1,3-D-mannnoside ß1,4-N-acetylglucosaminyitransferase. Glycobiology, 9: 303-310, 1999.1,3-D-mannoside ß1,4-N-acetylglucosaminyltransferase family. Glycoconj. J., 15: 1115-1123, 1998.[Medline]
This article has been cited by other articles:
|
|
 |
|
  K. S Lau and J. W Dennis N-Glycans in cancer progression Glycobiology, October 1, 2008; 18(10): 750 - 760. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. Gornik, L. Royle, D. J Harvey, C. M Radcliffe, R. Saldova, R. A Dwek, P. Rudd, and G. Lauc Changes of Serum Glycans During Sepsis and Acute Pancreatitis Glycobiology, December 1, 2007; 17(12): 1321 - 1332. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Mendelsohn, P. Cheung, L. Berger, E. Partridge, K. Lau, A. Datti, J. Pawling, and J. W. Dennis Complex N-Glycan and Metabolic Control in Tumor Cells Cancer Res., October 15, 2007; 67(20): 9771 - 9780. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
