genistein inhibits nonoxidative ribose synthesis in mia pancreatic adenocarcinoma cells: a new...

Genistein Inhibits Nonoxidative Ribose Synthesis in MIAPancreatic Adenocarcinoma Cells: A New Mechanism of

Controlling Tumor Growth

Laszlo G. Boros, Sara Bassilian, Shu Lim, and Wai-Nang Paul Lee

Harbor-UCLA Research and Education Institute, UCLA School of Medicine, Torrance, California, U.S.A.

Summary: Genistein is a plant isoflavonoid bearing potenttumor growth–regulating characteristics. This effect of genis-tein has been attributed partially to its tyrosine kinase–regulating properties, resulting in cell-cycle arrest and limitedangiogenesis. Genistein has been used in chemotherapy-resistant cases of advanced leukemia with promising results.Here we demonstrate that genistein primarily affects nucleicacid synthesis and glucose oxidation in tumor cells using the[1,2-13C2]glucose isotope as the single tracer and gas chroma-tography/mass spectrometry to follow various intracellular glu-cose metabolites. The ribose fraction of RNA demonstrated arapid 4.6%, 16.4%, and 46.3% decrease in isotope uptakethrough the nonoxidative branch of the pentose cycle and asharp 4.8%, 24.6%, and 48% decrease in 13CO2 release from

glucose after 2, 20, and 200 �mol/L genistein treatment, re-spectively. Fatty acid synthesis and the 13C enrichment of ace-tyl units were not significantly affected by genistein treatment.De novo glycogen synthesis from media glucose was not de-tected in cultured MIA cells. It can be concluded from thesestudies that genistein controls tumor growth primarily throughthe regulation of glucose metabolism, specifically targetingglucose carbon incorporation into nucleic acid ribose throughthe nonoxidative steps of the pentose cycle, which represents anew paradigm for the antiproliferative action of a plant phyto-chemical. Key Words: Pentose cycle—Ribose synthesis—Genistein—Nonoxidative glucose metabolism—Glucose oxi-dation.

Genistein, the isoflavonoid of the soy plant, has potenttumor growth–regulating characteristics (1–3). Genisteinhas exhibited tyrosine kinase (4) and protein kinase (PK)(5) inhibiting properties, resulting in cell-cycle arrest (6)and limited angiogenesis (7) in several tumor models.Genistein also enhances the effect of various plant phy-tochemicals including limonene, curcumin, epigallocat-echin gallate, or sulforaphene (8). Genistein counteractsthe growth-promoting effect of many human growth fac-tors through their signaling pathways that include signalcoupling to transcription factors that depend on trigger-ing of Met-receptor and protein kinase transducers (9).The invasive transformation of several human epithelialcarcinoma cell lines in response to transforming growth

factor � (TGF-�) treatment (10) is characterized by in-creased nucleic acid ribose synthesis through the non-oxidative reactions of the pentose cycle (11). Experimen-tal studies strongly indicate that genistein inhibits cellgrowth by modulating TGF-� signaling pathways spe-cifically (12). Genistein has recently been reported as aclinically effective and well-tolerated anticancer drug inadvanced chemotherapy-resistant cases of acute child-hood lymphoblastic leukemia, as well as adult chroniclymphocytic leukemia (13). Pancreatic tumor cells alsorespond to genistein treatment, as stimulated growth andp42 activation were inhibited by genistein in MIA pan-creatic adenocarcinoma cell cultures (14,15). Becausegenistein provides the basis for an effective treatmentstrategy for therapy-resistant human malignancies in-cluding pancreatic cancer, we were interested in themechanism of how genistein regulates nucleic acid,amino acid, lipid syntheses, and glucose oxidation inpancreatic adenocarcinoma cells. This was accomplishedusing biologic mass spectrometry of important metabo-

Manuscript received March 1, 2000; revision accepted April 26,2000.

Address correspondence and reprint requests to Dr. L. G. Boros,Harbor-UCLA Research and Education Institute, UCLA School ofMedicine, 1124 West Carson street RB1, Torrance, CA 90502, U.S.A.E-mail: [email protected]

PancreasVol. 22, No. 1, pp. 1–7© 2001 Lippincott Williams & Wilkins, Inc., Philadelphia

1

lites formed from a uniquely labeled glucose molecule intumor cell cultures in the presence of increasing doses ofgenistein.

METHODS

Cell line and cultureMIA pancreatic adenocarcinoma cells (American

Type Culture Collection) were grown in minimum es-sential medium (MEM) in the presence of 10% fetalbovine serum (FBS), at 37°C in 95% air/5% CO2. Tocompare glucose utilization rates, ribose synthesis, lac-tate production, and glutamine oxidation, 75% confluentcultures of MIA cells were incubated in [1,2-13C2]glucose-containing media (180 mg/dL, 50% isotopeenrichment). Cultures for the study were selected withthe same cell number (6 × 107), which was achievedusing standard cell-counting techniques. Media glucoseand lactate levels were measured using a Cobas Mirachemistry analyzer (Roche). Glucose oxidation was mea-sured by media 13C/12C ratios in released CO2 by aFinnegan Delta-S ion ratio mass spectroscope (GC/C/IRMS). 13CO2 release was used to estimate glucose car-bon utilization through oxidation by the cell lines andexpressed as atom percentage excess (APE), which is thepercentage of 13C produced by the cultured cells abovebackground in calibration standard samples (16).

RNA ribose was isolated by acid hydrolysis of cellularRNA after Trizol purification of cell extracts. Ribose waspurified using a tandem set of Dowex1/Dowex50 ion-exchange columns (Sigma). Ribose was derivatized to itsaldonitrile acetate form using hydroxyl amine in pyridineand acetic anhydrate. We monitored the ion clusteraround the m/z256 (carbons 1–5 of ribose, chemical ion-ization; CI), m/z217 (carbons 3–5 of ribose), and m/z242(carbons 1–4 of ribose, electron-impact ionization; EI) tofind molar enrichment and positional distribution of 13Clabels in ribose (17,18).

Stable [1,2-13C2]D-glucose isotope was purchasedwith >99% purity and 99% isotope enrichment for eachposition (Isotec, Inc., Miamisburg, OH, U.S.A.). For iso-tope-incubation and drug-treatment studies, fibroblastswere seeded in T-75 tissue-culture flasks after adjustingthe number of cells to the values reported earlier. Duringthe study, the cultures were supplied with 50% [1,2-13C2]glucose dissolved in otherwise glucose and sodiumpyruvate–free DMEM with 10% FBS. The final glucoseconcentration was adjusted to 180 mg/100 mL. Glucosemass isotope analysis of the medium before cell incuba-tions showed that the actual labeled glucose enrichmentwas 48% in the culture media, and this number was usedfor further calculations to determine maximal labeled

glucose enrichment in the molecules we studied. Singlylabeled ribose molecules (m1) recovered from RNA onthe first carbon position were used to measure the ribosemolar fraction produced by direct oxidation of glucosethrough the G6PD pathway, after subtracting the fractionof the singly labeled product that came from the TKpathway, calculated by the published m1 = m2(m3/m4formula (18). Doubly labeled ribose molecules (m2) onthe first two carbon positions were used to measure themolar fraction produced by transketolase. Doubly la-beled ribose molecules (m2) on the fourth and fifth car-bon positions were used to measure the molar fractionproduced by triose phosphate isomerase and TK. Isoto-pomers with three labels were used to estimate riboseproduction by combining recycled products of the G6PDreaction through the TK and transaldolase reactions. Iso-topomers with four labels (m4) were used to estimatesynthesis through TK and triose phosphate isomerase.Figure 1a shows the possible rearrangement of labelsfrom glucose to ribose as detected by gas chromatogra-phy/mass spectrometry.

Lactate from the cell-culture media (0.2 mL) was ex-tracted by ethyl acetate after acidification with HCl. Lac-tate was derivatized to its propylamine-HFB form, andthe m/z328 (carbons 1–3 of lactate, chemical ionization;CI) was monitored for the detection of m1 (recycledlactate through the PC) and m2 (lactate produced by theEmbden–Meyerhoff–Parnas pathway) for the estimationof pentose cycle activity (17). Fragmental lactate analy-sis was not necessary because the [1,2-13C2]glucosetracer labels lactate on the third (m1) or the second andthird (m2) carbon positions, which are clearly distin-guished in the molecular ion, as shown in Fig. 1b.

Glutamate tissue culture medium was first treated with6% perchloric acid. Proteins were removed by centrifu-gation, and the supernatant was neutralized with potas-sium hydroxide. The neutralized supernatant was passedthrough a 3-mL Dowex-50 (H+) column. Amino acidswere eluted from the Dowex-50 column with 15 mL 2Nammonium hydroxide, and the solution was evaporatedto dryness by blowing air. To separate glutamate furtherfrom glutamine, the amino acid mixture was passedthrough a 3-mL Dowex-1 (acetate) column. Glutaminewas washed with 10 mL water, and glutamate was col-lected with 15 mL 0.5N acetic acid. Glutamate fractionfrom tissue culture medium was converted to its trifluo-roacetyl butyl ester (TAB) (19,20). Under EI conditions,ionization of TAB-glutamate gives rise to two fragments,m/z198 and m/z152, corresponding to C2–C5 and C2–C4of glutamate. Glutamate labeled on the 4–5 carbon po-sitions indicates pyruvate dehydrogenase activity,whereas glutamate labeled on the 2–3 carbon positions

L. G. BOROS ET AL.2

Pancreas, Vol. 22, No. 1, 2001

indicates pyruvate carboxylase activity for the entry ofglucose carbons to the TCA cycle. The TCA cycle me-tabolite �-ketoglutarate is in equilibrium with glutamate,which is released by the cells into the medium (Fig. 1c).

The m2/m1 ratio in glutamate is proportional with theactivity of glucose oxidation as 13CO2 is released from�-ketoglutarate during each completed cycle. Anaple-rotic flux is calculated based on the m2/m1 ratios ofglutamate (20).

Fatty acids were extracted after saponification of cellpellets in 30% KOH and 100% ethanol using petroleumether. Fatty acids were converted to their methylated de-rivative using 0.5N methanolic-HCL. Palmitate wasmonitored at m/z270, and the enrichment of acetyl unitsas well as the synthesis of the new lipid fraction in MIAcells in response to genistein treatment was determinedusing the mass isotopomer distribution analysis (MIDA)approach of different isotopomers of palmitate, as re-ported previously (21).

Glycogen glucose was extracted after sonication ofcell pellets and digestion with amyloglucosidase fromAspergillus niger (Boehringer, Mannheim, Germany).Glucose was purified using a tandem set of Dowex1/Dowex50 ion-exchange columns and then derivatized toits aldonitrile-acetate form. Glucose molecular ion wasdetected in the m/z328 ion cluster. Electron-impact ion-ization gives rise to two glucose fragments, m/z187 (car-bons 3–6 of glucose) and m/z242 (carbons 1–4 of glu-cose), to determine positional distribution of 13C labels(22).

Gas chromatography/mass spectrometry (GC/MS)Mass spectral data were obtained on the HP5973 mass

selective detector connected to an HP6890 gas chromato-graph. The settings are as follows: GC inlet, 230°C;transfer line, 280°C; MS source, 230°C; MS Quad,150°C. An HP-5 capillary column (30 m length, 250 �mdiameter, 0.25 �m film thickness) was used for glucose,ribose, glutamate, and lactate analysis. A Bpx70 column(25 m length, 220 �m diameter, 0.25 �m film thickness;SGE Incorporated, Austin, TX, U.S.A.) was used forfatty acid analysis with specific temperature program-ming for each compound studied.

Data analysis and statistical methodsIn vitro experiments were carried out using three cul-

tures each time for each treatment regimen. Mass spec-tral analyses were carried out by three independent au-tomatic injections of 1-�L samples by the automaticsampler and accepted only if the standard sample devia-tion was <1% of the normalized peak intensity. Statisti-cal analysis was performed using the parametric un-paired, two-tailed independent sample t test with 99%confidence intervals (� ± 2.58�), and p < 0.01 was con-sidered to indicate significant differences in glucose car-bon metabolism in MIA pancreatic adenocarcinoma cellstreated with increasing doses of genistein.

FIG. 1. Possible 13C rearrangement in intermediates of the pentosecycle (A), lactate (B), or glutamate (C) using [1,2-13C2]glucose as thesingle tracer. For measuring the activity of each synthesis pathway, theratio of 12C- versus 13C-labeled molecules and the position of 13C-labeled carbons is determined using mass isotopomer analysis by gaschromatography/mass spectrometry. Absolute glucose utilization andlactate production also are measured from the cell-culture media toobtain absolute values for glucose carbon utilization through the spe-cific pathways of intermediate metabolism.

CHANGES AFTER GENISTEIN IN MIA CELLS 3


FIG. 2. Metabolic profile of cultured MIA pancreatic adenocarcinoma cells in the presence of increasing doses of genistein after 72 hours oftreatment. Genistein was administered in 2-, 20-, and 200 �mol/L doses, as shown on the X axis. Glucose carbon incorporation into nucleic acid riboseis expressed as molar enrichment (�mn), and the data indicate that nucleic acid ribose rapidly reaches the theoretic maximal enrichment in controluntreated cultures incubated with 48% of [1,2-13C2]glucose (A). Pentose cycle activity calculated by the m1/m2 ratios in lactate indicates that theactivity of the PC is about 2.5% that of glycolysis with steady recycling of pentose cycle metabolites into the glycolysis, which is not affected byincreasing doses of genistein treatment (B). TCA cycle anaplerotic flux relative to glucose oxidation indicates a significant dose-dependent decreaseafter 72 hours of genistein treatment (C), which was associated with a significant decrease in 13CO2 release from [1,2-13C2]glucose (D). The fractionof newly synthesized palmitate (E) and the 13C enrichment of acetyl units (F) did not change significantly after genistein treatment (mean ± SD;n � 9).

L. G. BOROS ET AL.4


RESULTS

For the studies reported herein, we incubated MIApancreatic adenocarcinoma cells in the presence of [1,2-13C2]glucose, treated with increasing amounts of genis-tein (2, 20, and 200 �mol/L) for 72 hours. Genisteindoses were selected for the study based on reports thatplasma levels of genistein are around 18.5 �mol/L or 5�g/mL (23). The possible rearrangements of 13C-labeledcarbons from [1,2-13C2]glucose in ribose, lactate, andglutamate are given in Fig. 1. Ribose metabolizedthrough glucose-6-phosphate-dehydrogenase (G6PD) re-tains one 13C-labeled carbon on the first position,whereas the nonoxidative reactions of the pentose cycleinvolving transketolase and transaldolase produce riboselabeled on other positions (Fig. 1a). Based on the numberof 13C-labeled carbons and their positions, known asmass isotopomers, it was evident that MIA pancreaticadenocarcinoma cells synthesized all (100%) nucleicacid ribose directly through the nonoxidative transketo-lase reaction, with a negligible amount of ribose arrivingthrough G6PD, as shown in Fig. 2a. Increasing doses ofgenistein decreased ribose synthesis in a dose-dependentfashion from glucose, primarily through the nonoxidativereactions of the pentose cycle.

Pentose cycle activity relative to the Embden–Meyerhoff–Parnas pathway was measured by the m2/m1ratios recovered in lactate, as shown in Fig. 1b. Pentosecycle activity did not show a significant decrease in MIAtumor cells after 3 days of genistein treatment, as indi-cated by similar distribution of 13C and the m2/m1 ratiosin lactate (Fig. 2b). However, the moderate decrease ofthe m1 lactate isotopomer (Fig. 2b) indicates that genis-tein also may decrease G6PD and the activity of thepentose cycle in pharmacologic doses. The amount and13C label enrichment of lactate released by the tumorcells in the culture media (13.4–14.9 mmol/L/72 h) didnot differ significantly between the treatment groups, in-dicating that reduced glucose carbon incorporation intoRNA ribose was the result primarily of decreased non-oxidative synthesis through transketolase and transaldol-

ase and not due to decreased glucose transport into tumorcells. Glucose consumption was similar in both the con-trol and treated cell cultures, ranging from 13.7 to 14.4mg during the 72-hour study period.

Glutamate mass isotopomers derived from [1,2-13C2]-glucose are shown in Fig. 1c. Tricarboxylic acid (TCA)cycle anaplerotic flux was measured using the equilib-rium between glutamate and �-ketoglutarate. The 13C-label rearrangement from pyruvate through pyruvate car-boxylase or pyruvate dehydrogenase results in differentglutamate mass isotopomers that can be easily detectedand measured using fragmental mass isotopomer analy-sis. Glutamate stable isotope label rearrangement indi-cated a significant decrease in TCA cycle carbon fluxafter genistein treatment, as shown in Fig. 2c. 13C-labeledcarbon dioxide release from glucose was decreased dosedependently after genistein treatment, which is the resultof genistein effect on the TCA cycle (Fig. 2d).

The fraction of newly synthesized palmitate was de-creased only moderately after increasing doses of genis-tein treatment (Fig. 2e), and 13C enrichment of acetylunits for fatty acid synthesis also showed only a moder-ate decrease (Fig. 2f).

A summary of metabolic changes in MIA pancreaticadenocarcinoma cells is given in Table 1 as percentagechange of metabolic activity in response to increasingdoses of genistein treatment compared with untreatedcontrol tumor cell cultures. The percentage changes inmetabolic activity indicate that genistein treatment af-fects cell metabolism primarily through decreasing TCAcarbon flux, direct glucose oxidation, and the nonoxida-tive synthesis of ribose to build nucleic acid. The effectof genistein treatment on lipid synthesis, the formation ofacetyl units from glucose, and glycogen synthesis (datanot shown) are less prominent and probably do not playa significant role in controlling tumor cell growth.

DISCUSSION

It is well established that tumor cells recruit glucosecarbons intensively, and their exceptional dependence on

TABLE 1. Percentage changes in the metabolic activity of MIA pancreatic adenocarcinoma cells in response to increasingdoses of genistein after 72 hours of culturing

Genistein dose 13CO2releaseNonoxidative

ribose synthesisTCA cycle

glucose fluxPentose cycle

activityAcetyl-CoA

13C enrichmentNew lipidsynthesis

2 �M 95.2 (±5.3) 95.4 (±2.4) 95.4a (±3.1) 91.6 (±2.8) 95.5 (±3.6) 100.8 (±4.4)20 �M 75.4a (±13.5) 83.6a (±2.42) 92.6a (±4.32) 92.3 (±2.9) 95 (±3.1) 101.7 (±3.8)

200 �M 52a (±14.4) 57.7a (±5.5) 66.6a (±1.5) 82.7 (±4.3) 88.9a (±6.4) 94.2 (±6.4)

Data shown as percentage of untreated control cultures (100%). The most significant metabolic effects of genistein treatment are detected in thesynthesis of the nucleic acid ribose moiety, CO2 production, and TCA cycle carbon flux. The new fraction of lipid synthesis, acetyl unit glucose carbonenrichment, and pentose cycle activity are not affected by physiologic levels of genistein. Mean (±SD).

a p < 0.01; n � 9.



glucose is related to their increased energy requirementsand intensive intermediate metabolism. Using stable 13C-glucose isotope tracers in metabolic studies in cultures ofhuman cells enables us to study a broad range of intra-cellular glucose intermediates and investigate their labeldistribution to determine carbon flow through variousmetabolic pathways simultaneously. Such approach hasbeen successful in determining pentose cycle activity, thecontribution of the two branches of the pentose cycle tonucleic acid ribose synthesis, TCA cycle activity, andlipid synthesis simultaneously in HepG2 and MIA pan-creatic adenocarcinoma cells (17,18). The application ofstable isotopes to tumor cell metabolism has revealedthat glucose carbons are the primary source of nucleicacid ribose, lactate, and glutamate and that glucose car-bon metabolism follows a nonoxidative path in the pen-tose cycle in response to TGF-� treatment in lung epi-thelial carcinoma cells (11). As genistein is known toinhibit EGF and TGF-� signaling in various tumors (3,4,12), we were interested in revealing how genistein af-fects tumor cell glucose-utilization pathways and inhibitstumor cell proliferation by analyzing metabolic pathwaysubstrates and end products using biologic mass spec-trometry.

This report uses the powerful mass isotope distributionanalysis (MIDA) method for the characterization of thedifferent pathways of glucose metabolism simulta-neously in tumor cells in response to genistein treatmentin culture. Data acquired through this method indicatethat genistein regulates tumor cell proliferation by alter-ing the rate of glucose oxidation and the synthesis ofnucleic acid ribose through the nonoxidative steps of thepentose cycle. As ribose is a close metabolite of glucose,and ribose is essential for de novo nucleic acid synthesisas well as the salvage pathways of purine and pyrimidinebases, it is likely that inhibiting the formation of ribosefrom glucose through the nonoxidative steps of the pen-tose cycle is the underlying mechanism by which genis-tein regulates tumor cell growth. The nonoxidative reac-tions of the pentose cycle have an important role in regu-lating the balance between the two branches of thepentose cycle and its overall output, as measured by GSHproduction, and thus influence sensitivity to cell deathsignals (24). This is the first report of a naturally occur-ring plant phytochemical that controls nucleic acid syn-thesis through the synthesis of its ribose moiety, whichmay have major implications in future drug-developmentefforts for new anticancer drugs.

Acknowledgment: We thank the volunteers of the Inflam-matory Breast Cancer Research foundation and Mr. Marvin L.Davis of Palos Verdes Estates, CA, for their editorial help. This

work was supported by a grant from the Clinical Nutrition andResearch Unit of California and the National Institutes ofHealth of the United States.

REFERENCES

1. Su SJ, Yeh TM, Lei HY, et al. The potential of soybean foods asa chemoprevention approach for human urinary tract cancer. ClinCancer Res 2000;6:230–6.

2. Alhasan SA, Pietrasczkiwicz H, Alonso MD, et al. Genistein-induced cell cycle arrest and apoptosis in a head and neck squa-mous cell carcinoma cell line. Nutr Cancer 1999;34:12–9.

3. Okura A, Arakawa H, Oka H, et al. Effect of genistein on topo-isomerase activity and on the growth of [Val 12]Ha-ras-transformed NIH 3T3 cells. Biochem Biophys Res Commun 1988;157:183–9.

4. El-Zarruk AA, van den Berg HW. The anti-proliferative effects oftyrosine kinase inhibitors towards tamoxifen-sensitive and tamoxi-fen-resistant human breast cancer cell lines in relation to the ex-pression of epidermal growth factor receptors (EGF-R) and theinhibition of EGF-R tyrosine kinase. Cancer Lett 1999;142:185–93.

5. Waltron RT, Rozengurt E. Oxidative stress induces protein kinaseD activation in intact cells involvement of Src and dependence onprotein kinase C. J Biol Chem 2000;275:17114–21.

6. Lian F, Bhuiyan M, Li YW, et al. Genistein-induced G2-M arrest,p21WAF1 upregulation, and apoptosis in a non-small-cell lungcancer cell line. Nutr Cancer 1998;31:184–91.

7. Zhou JR, Gugger ET, Tanaka T, et al. Soybean phytochemicalsinhibit the growth of transplantable human prostate carcinoma andtumor angiogenesis in mice. J Nutr 1999;129:1628–35.

8. Bradlow HL, Telang NT, Sepkovic DW, et al. Phytochemicals asmodulators of cancer risk. Adv Exp Med Biol 1999;472:207–21.

9. Tacchini L, Dansi P, Matteucci E, et al. Hepatocyte growth factorsignal coupling to various transcription factors depends on trigger-ing of Met receptor and protein kinase transducers in human hep-atoma cells HepG2. Exp Cell Res 2000;256:272–81.

10. Hojo M, Morimoto T, Maluccid M, et al. Cyclosporine inducescancer progression by a cell-autonomous mechanism. Nature1999;397:530–4.

11. Boros LG, Torday JS, Lim S, et al. TGF-�2 promotes glucosecarbon incorporation into nucleic acid ribose through the non-oxidative pentose cycle in lung epithelial carcinoma cells. CancerRes 2000;60:1183–5.

12. Kim H, Peterson TG, Barnes S. Mechanisms of action of the soyisoflavone genistein: emerging role for its effects via transforminggrowth factor beta signaling pathways. Am J Clin Nutr 1998;68:1418S–25S.

13. Uckun FM, Messinger Y, Chen CL, et al. Treatment of therapy-refractory B-lineage acute lymphoblastic leukemia with an apop-tosis-inducing CD19-directed tyrosine kinase inhibitor. Clin Can-cer Res 1999;5:3906–13.

14. Douziech N, Calvo E, Laine J, et al. Activation of MAP kinases ingrowth responsive pancreatic cancer cells. Cell Signal 1999;11:591–602.

15. Douziech N, Lajas A, Coulombe Z, et al. Growth effects of regu-latory peptides and intracellular signaling routes in human pancre-atic cancer cell lines. Endocrine 1998;9:71–83.

16. Kasho VN, Cheng S, Jensen DM, et al. Feasibility of analyzing[13C]urea breath tests for Helicobacter pylori by gas chromatog-raphy-mass spectrometry in the selected ion monitoring mode. Ali-ment Pharmacol Ther 1996;10:985–95.

17. Lee WN, Boros LG, Puigjaner J, et al. Mass isotopomer study ofthe non-oxidative pathways of the pentose cycle with [1,2-13C2]glucose. Am J Physiol 1998;274:E843–51.

18. Boros LG, Puigjaner J, Cascante M, et al. Oxythiamine and de-hydroepiandrosterone inhibit the non-oxidative synthesis of riboseand tumor cell proliferation. Cancer Res 1997;57:4242–8.

L. G. BOROS ET AL.6


19. Lee WN, Edmond J, Bassilian S, et al. Mass isotopomer study ofglutamine oxidation and synthesis in primary culture of astrocytes.Dev Neurosci 1996;18:469–77.

20. Leimer KR, Rice RH, Gehrke CW. Complete mass spectra ofN-TAB esters of amino acids. J Chromatogr 1977;141:121–44.

21. Lee WN. Stable isotopes and mass isotopomer study of fatty acidand cholesterol synthesis: a review of the MIDA approach. AdvExp Med Biol 1996;399:95–114.

22. Katz J, Lee WN, Wals PA, et al. Studies of glycogen synthesis andthe Krebs cycle by mass isotopomer analysis with [U-13C]glucosein rats. J Biol Chem 1989;264:12994–3004.

23. Peterson G. Evaluation of the biochemical targets of genistein intumor cells. J Nutr 1995;125:784S–9S.

24. Banki K, Hutter E, Colombo E, et al. Glutathione levels and sen-sitivity to apoptosis are regulated by changes in transaldolase ex-pression. J Biol Chem 1996;271:32994–3001.



genistein inhibits nonoxidative ribose synthesis in mia pancreatic adenocarcinoma cells: a new...

Documents