Metabolism of alternative substrates and the bioenergetic status of EMT6 tumor cell spheroids.

In order to evaluate the ability of EMT6/Ro multicellular spheroids to utilize various pathways of energy production, (13)C and (31)P MRS have been employed to monitor the metabolism of glucose, glutamine, acetate and propionate. EMT6/Ro spheroids perfused with culture medium containing 5.5 mM glucose maintain stable levels of nucleotide triphosphates (NTP) and phosphocreatine (PCr) for up to 48 h, even in the absence of glutamine. The metabolism of 1-(13)C-glucose was almost entirely to 3-(13)C-lactate (88 +/- 12%, n = 7), even though the perfusion medium was equilibrated with 95% O(2). Labeling was also observed in other glycolytic metabolites, primarily alanine and alpha-glycerolphosphate. A low level of (13)C labeling in glutamate, indicative of mitochondrial oxidative metabolism (TCA cycle), was consistently detected when spheroids were perfused with 1-(13)C-glucose, almost exclusively in the C4 position of glutamate. Labeling of glutamate C2 and C3 was always less than 20% of the labeling in C4 and was usually undetectable. No evidence of adjacent carbon labeling in individual glutamate molecules (indicative of multiple cycles of label incorporation) was found, even in high-resolution (13)C NMR spectra of extracts from cells or spheroids. Despite the predominantly glycolytic metabolism of glucose, the mitochondrial substrate glutamine (2 mM, in the presence of < or =0.5 mM glucose from fetal bovine serum), supported stable levels of NTP and PCr in the tumor cells for up to 12 h. In the presence of 2.5 mM acetate, the bioenergetic status of cells in EMT6 spheroids declined slowly but measurably, and no incorporation of label from 2-(13)C-acetate into other metabolites was detected either in intact perfused spheroids or in high-resolution spectra of extracts. In contrast, when the anaplerotic TCA cycle substrate 3-(13)C-propionate replaced acetate, the high-energy phosphate levels in EMT6/Ro spheroids were somewhat reduced, but stabilized at a new lower level. Incubation of spheroids with 3-(13)C-propionate (with natural abundance glucose and glutamine) resulted in label detectable in the C2 and C3 of glutamate, but the primary labeled compound was methylmalonate, an intermediate in propionate metabolism. Addition of vitamin B(12), a cofactor for methylmalonyl CoA reductase, to the growth medium 24 h prior to perfusion with propionate resulted in the elimination of the methylmalonate resonance. A variety of 2- and 3-labeled metabolites were detected, including succinate, malate and glutamate. Labeling of C2 and C3 of lactate implicated cytoplasmic malic enzyme activity.

Inhibition of tumor cell proliferation by dexamethasone: 31P NMR studies of RIF-1 fibrosarcoma cells perfused in vitro.

The impact on tumor cell metabolism of a substantial reduction in cell proliferation rate without acute cytotoxicity was examined in cultured RIF-1 tumor cells following treatment with an antiproliferative steroid, dexamethasone (DEX). After 48 h exposure to 4 mM DEX, acute cell viability was essentially unchanged: cells were 93 +/- 2% trypan blue excluding in both control and treated cultures (all values are mean +/- SD). The fraction of actively proliferating cells in the S phase (as indicated by incorporation of 5-bromodeoxyuridine) was only 4 +/- 3%, compared with 13 +/- 3% in age-matched control cultures (n =4, paired t-test: p < 0.004) and 23 +/- 7% at the beginning of the treatment. Three days of DEX treatment resulted in a limited increase in the level of apoptosis (programmed cell death): cells did not become rounded or detached, but the fraction expressing apoptotic DNA fragmentation (susceptible to nick end labeling by terminal deoxy-nucleotidyl transferase) was 15 +/- 7%, vs 2 +/- 1% in control cultures (p < 0.02). Despite a 75% inhibition of cell proliferation, DEX caused only a modest change in the 31P NMR spectra of RIF-1 cells in vitro. The ratio of phosphocreatine to nucleoside triphosphates (NTP) was 30% higher, on average, in treated than in control cells (n = 8, paired t-test, p < 0.02), even when both treated and control cell densities were low. The level of total phosphomonoester (relative to NTP) was lower at low cell density, but this was independent of whether cells were growing rapidly (control low density) or were growth inhibited by DEX. Neither the ratio of phosphocholine to NTP nor the intracellular pH was significantly different in DEX-treated cells.

Glucose metabolism in RIF-1 tumors after reduction in blood flow: an in vivo 13C and 31P NMR study.

Low pH appears to enhance the effectiveness of therapeutic hyperthermia. 13C and 31P NMR spectroscopy have been employed to examine the possibility that elevating glucose in a solid tumor while simultaneously reducing tumor blood flow would induce a more profound acidosis than either treatment alone. When blood flow in RIF-1 tumors was acutely reduced by administration of hydralazine and additional glucose was delivered locally by intratumoral injection, tumor acidosis (as determined by 31P NMR spectroscopy) during the period of reduced blood flow was not enhanced, relative to administration of hydralazine alone. Tumor NTP/P1 ratios decreased significantly within 20 min of hydralazine administration, whether or not glucose was injected, although NTP/P1 ratios were slightly higher in tumors that received extra glucose. Tumor lactate concentrations were not significantly different in glucose-supplemented tumors, despite glucose concentrations that were 4 to 5 times higher. When the added glucose was labeled with 13C, no correlation was detected between the pH in an individual tumor and the intensity of the 3-[13C]-lactate resonance in the same tumor.

31P NMR spectroscopic studies of the effects of cyclophosphamide on perfused RIF-1 tumor cells.

To determine whether direct cellular effects of chemotherapy are responsible for 31P NMR spectral changes observed in treated tumors in vivo, RIF-1 fibrosarcoma cells were examined in vitro before, during, and after treatment with 4-hydroperoxycyclophosphamide (4-HC), an activated form of cyclophosphamide. When RIF-1 cells were treated with 4-HC in a metabolically stable but nonproliferating state, the 31P NMR spectra were identical with those of untreated cells for up to 70 h. When actively proliferating RIF-1 cells were treated with 4-HC, the intensities of the nucleotide triphosphate resonances, which increased linearly during control cell growth, remained constant for 50 h or longer. These studies demonstrate that the bioenergetic improvement observed following treatment of RIF-1 tumors in vivo [S.-J. Li, J.P. Wehrle, S.S. Rajan, R.G. Steen, J.D. Glickson, and J. Hilton, Cancer Res. 48, 4736 (1988)] does not result from direct effects of cyclophosphamide metabolites on RIF-1 cell metabolism, but rather from indirect effects of treatment on tumor or host physiology.

Gel-entrapment of perfluorocarbons: a fluorine-19 NMR spectroscopic method for monitoring oxygen concentration in cell perfusion systems.

Oxygenation is a major determinant of the physiological state of cultured cells. 19F NMR can be used to determine the oxygen concentration available to cells immobilized in a gel matrix by measuring the relaxation rate (1/T1) of perfluorocarbons (PFC) incorporated into the gel matrix. In calcium alginate gel beads without cells the relaxation rate (1/T1) of the trifluoromethyl group of perfluorotripropylamine (FTPA) varies linearly with oxygen concentration, with a slope of 1.26 +/- 0.15 x 10(-3) s-1 microM-1 and an intercept of 0.50 +/- 0.04 s-1. During perfusion with medium equilibrated with 95%/5% O2/CO2, changes in PFC T1s indicate that the average oxygen concentration was reduced from 894 +/- 102 microM in the absence of cells to 476 +/- 65 microM and 475 +/- 50 microM in the presence of 0.7 x 10(8) EMT6/Ro and RIF-1 murine tumor cells per milliliter of gel, respectively. The presence of 0.2 microliters of FTPA/ml of gel had no effect on the energy status of the cells as indicated by 31P NMR spectra. To calculate oxygen gradients within the beads from the average PFC T1 of the sample, a mathematical model was used assuming that oxygen is the limiting nutrient for cell metabolism and that the cellular oxygen consumption rate is independent of oxygen concentration. Data for EMT6/Ro cells were fit using experimentally determined perfusion parameters together with literature values for cell volume and oxygen consumption rate.(ABSTRACT TRUNCATED AT 250 WORDS)

Determination of absolute phosphate metabolite concentrations in RIF-1 tumors in vivo by 31P-1H-2H NMR spectroscopy using water as an internal intensity reference.

The absolute metabolite quantification method of Thulborn and Ackerman [J. Magn. Reson. 55, 357 (1983)] in which the tissue water proton signal is used as an internal intensity standard and its more recent variation in which NMR peak intensities are referenced to that of the natural abundance deuterium signal of water [Li et al., SMRM Abstr. 2, 825 (1988); Song et al., Magn. Reson. Med. 25, 45 (1992) have been implemented to obtain absolute phosphate metabolite concentrations in subcutaneous RIF-1 tumors during untreated growth and following treatment with 5-fluorouracil. The equivalence of these two hydrogen isotopes as intensity standards and the validity of their use in the determination of absolute metabolite concentrations in vivo by NMR has been demonstrated. On matched in vivo and extract tumor samples (n = 5), excellent agreement has been obtained between nucleoside triphosphate concentrations determined by NMR and those derived by HPLC analysis for the control tumors. Following 3 days of untreated growth, absolute concentrations of phosphate metabolites in RIF-1 tumors (n = 10) decreased significantly, except for the Pi concentration which did not vary. For the treated tumors (n = 10) there were no changes in metabolite concentrations except for a decrease in the PCr and, possibly, Pi concentrations. The PCr/Pi ratio in the latter tumors did not change. These observations suggest that changes in absolute metabolite concentrations may be more sensitive indices of response to therapy than changes in metabolite peak amplitude ratios, a parameter commonly used to express in vivo NMR data.

31P NMR spectroscopic study of the effects of gamma-irradiation on RIF-1 tumor cells perfused in vitro.

In order to examine the mechanisms underlying radiation-induced changes in phosphorus metabolite levels observed in RIF-1 tumors in vivo, RIF-1 cells in culture were perfused for up to 70 h following gamma-irradiation with 0-25 Gy and monitored continuously by 31P NMR spectroscopy at 8.5 T. Cells immobilized in the sample volume by incorporation into calcium alginate beads were bioenergetically stable, but did not replicate at the cell density used. Following an initial increase in PCr and NTP, which occurred in both control and irradiated cells, a dramatic decline in high-energy phosphates was detected beginning 24-30 h after irradiation with 15 or 25 Gy. In contrast, unirradiated cells or cells treated with 10 Gy remained metabolically stable for up to 72 h. The metabolic changes induced by irradiation of the cultured cells, which reflected cell death and lysis, were distinctly different from those observed in RIF-1 tumors in vivo during the same postirradiation time interval--an increase in high-energy relative to low-energy phosphates. This suggests that the spectral changes in vivo do not result from direct modification of cellular energy metabolism by radiation injury.

1H NMR spectroscopy of subcutaneous tumors in mice: preliminary studies of effects of growth, chemotherapy and blood flow reduction.

This study evaluates a number of methods for obtaining 1H NMR spectra with adequate suppression of lipid and water resonances in two subcutaneously implanted transplantable tumor models (RIF-1 and EMT6/SF). Spin-echo spectra with long TEs (270 ms; water suppressed by presaturation) eliminated lipid resonances from 1H spectra of RIF-1 and decreased lipid contamination in spectra of EMT6/SF; however, spectral sensitivity was substantially reduced. A shorter TE (135 ms) increased sensitivity but did not result in adequate suppression of the lipid peaks. In RIF-1, but not EMT6/SF, adequate lipid suppression was achieved by: (i) spatially selective presaturation of lipid, which in this tumor (but not in EMT6/SF) was localized in a thin region along the periphery of the tumor, followed by a 1-D spin-echo chemical shift imaging pulse sequence (TE = 135 ms); and (ii) 2-D spin-echo chemical shift imaging (TE = 270 ms; approximately 2 x 2 x 9 mm3 voxels). Preliminary 1H studies of the RIF-1 tumor indicate that: (i) there are no significant changes in metabolite levels relative to tumor water during 4 days of untreated tumor growth; (ii) tumor response to chemotherapy with 5-fluorouracil results in a decrease in intensity of all metabolite 1H resonances relative to tumor water, with total choline decreasing the most and lactate the least; and (iii) acute tumor blood flow reduction induced by administration of hydralazine results in doubling of the lactate intensity relative to water.(ABSTRACT TRUNCATED AT 250 WORDS)

Spatial heterogeneity of the metabolic response of RIF-1 tumors to a vasoactive agent evaluated in vivo by one-dimensional 31P chemical-shift imaging.

Localized 31P NMR spectroscopy was used to evaluate the spatial heterogeneity of the metabolic response of RIF-1 tumors to hydralazine. Volume localized 31P spectra were obtained from subcutaneous RIF-1 tumors using one-dimensional chemical-shift imaging, before and 20 min after treatment with 5 mg/kg hydralazine, administered intravenously. Following treatment all of the tumors showed an overall decrease in the ratio of nucleoside triphosphate (NTP) to inorganic phosphate (Pi) and a decrease in pH. However, spatial localization revealed that the reduction in NTP/Pi was not uniform within some tumors. This was partly due to regional differences in the levels of metabolites existing before treatment. Normal tissue adjacent to the tumor did not show a significant decrease in high-energy metabolites or pH.

Selective suppression of lipid resonances by lipid-soluble nitroxides in NMR spectroscopy.

The ability of lipid-soluble nitroxides to suppress selectively the peaks of lipid resonances in 31P, 1H, and 13C NMR spectra was investigated in serum as part of studies aimed at using these contrast agents for magnetic resonance imaging and magnetic resonance spectroscopy in vivo. Nitroxides are especially interesting potential contrast agents because they can reversibly be converted in cells to diamagnetic hydroxylamines, with conversion rates that are dependent on the redox potential and the intracellular concentration of oxygen; the characterization of nitroxide-dependent changes in NMR spectra may therefore be a useful means to measure oxygen-dependent redox metabolism in vivo. The fatty acid analogs, doxyl stearates, suppressed the methyl resonance of choline and the methyl and methylene peaks of lipids in the 1H NMR spectra of serum samples. As a consequence, lactate peaks, which were not readily detected became clearly resolved and could be evaluated quantitatively. The 31P resonance of phosphatidylcholine in the 31P NMR spectrum was suppressed by 5-doxyl stearate and 4-(N,N-dimethyl-N-hexadecyl)ammonium-2,2,6,6-tetramethylpiperidine-1-oxy l,iodid e (Cat16). In the 13C NMR spectrum, the resonances of the methyl groups of choline and the lipids also were broadened significantly by addition of 5-doxyl stearate. Differential suppression of lipid resonances can be employed to facilitate quantitation of lactate.

Energy metabolism, pH changes, and lactate production in RIF-1 tumors following intratumoral injection of glucose.

The metabolic consequences of increased glucose availability were examined in subcutaneous RIF-1 tumors in vivo, using 13C and 31P NMR spectroscopy. Significant increases in the levels of nucleotide triphosphates and phosphocreatine relative to low energy phosphates and in tumor pH were observed within 30 min following injection of 1 g/kg of glucose directly into the tumor. These changes did not occur following an equivalent intratumoral dose of the non-metabolizable sugar alcohol, mannitol. When [1-13C]-glucose was administered, [3-13C]-lactate and [3-13C]-alanine were the only labeled metabolites detected in the in vivo 13C NMR spectra during the period of bioenergetic improvement. Biochemical analysis revealed a substantial increase in tumor and plasma glucose concentration, but no increase in either tumor or plasma lactate, consistent with the absence of acidosis. Evaluation of the distribution of glucose in the tumor by quantitative autoradiography of [1-14C]-2-deoxyglucose administered with the glucose indicated that, on average, 7 mM of the added glucose distributed over the entire tumor within 10 min. The significant improvement in overall metabolic status of the tumors following glucose administration is attributed to the existence of substrate limited regions within the tumor.

Tumor bioenergetics and blood flow in RIF-1 murine tumors treated with 5-fluorouracil.

Treatment of RIF-1 solid tumors with 5-fluorouracil (5-FU, 100 or 200 mg/kg, ip) caused substantial regression of the tumors, with regrowth initiated on Day 6 (100 mg/kg) or Day 9 (200 mg/kg). Blood perfusion in the tumor, estimated by uptake of 86Rb+, was significantly increased after treatment with 5-FU, while Rb+ uptake in normal tissues (skin, muscle) was unaffected. The increase in tumor perfusion during the first few days following treatment was significantly greater in animals treated with the higher dose of 5-FU. Perfusion-dependent 86Rb+ uptake returned to control levels by the 9th day after treatment with 100 mg/kg of 5-FU, but remained elevated on Days 9-12 after the higher dose. By the 1st day following treatment with 5-FU, in vivo 31P NMR spectra of treated tumors indicated significantly higher ratios of phosphocreatine to Pi, higher pH, and lower ratios of Pi to nucleoside triphosphates compared to untreated age-matched controls. These changes persisted for 9 days following the lower 5-FU dose and for at least 12 days following the higher dose. Treatment with 5-FU induces profound, dose-dependent changes in tumor bioenergetics, which may result, at least in part, from changes in tumor perfusion after cytoreduction.

In vivo 13CNMR spectroscopy of glucose metabolism of RIF-1 tumors.

An efficient method for measuring in vivo 13C NMR spectra of tumors has been developed and employed to monitor glucose metabolism in radiation-induced fibrosarcomas (RIF-1) subcutaneously implanted in C3H/HeN mice. [1-13C]Glucose was injected directly into the tumors at a dose of 1 g/kg body wt. Spectra were obtained with a Bruker AM 360-WB spectrometer (8.4 T/8.9 cm bore) employing a homebuilt probe equipped with a four-turn solenoidal coil (1.5 cm outer diameter) for detection of 13C signals and a Helmholtz coil (two 3-cm turns separated by a 3-cm gap, oriented orthogonally to the 13C coil) for 1H decoupling. In addition to the natural abundance 13C resonances of the tumors, signals were detected from the alpha- and beta-anomers of labeled glucose. Within 15 min following injection of labeled glucose [3-13C]lactate and [3-13C]alanine were detected. Lactate labeling approached steady state levels within about 50 min after glucose injection: in contrast, alanine labeling increased continuously over the duration of the experiment (70 min). Sixty minutes after glucose injection, the ratio of the intensity of [3-13C]lactate to the principal lipid methylene resonance (30 ppm from external tetramethylsilane), which served as an internal intensity reference, was correlated with tumor size, whereas the corresponding ratio of the [3-13C]alanine resonance was not. Labeling of glutamate was below the level of detection in the in vivo spectra; however, labeling of C4-glutamate at a level approximately 50-fold lower than the level of [3-13C]lactate was detected in perchloric acid extracts. Incorporation of 13C label into C2- and C3-glutamate and C2-lactate was also observed.

Metabolic heterogeneity in RIF-1 tumours detected in vivo by 31P NMR spectroscopy.

The spatial distribution of phosphate metabolites and pH within subcutaneously implanted RIF-1 tumours was determined by 1-dimensional phosphorus chemical shift imaging. 31P spectra from two to three 4 mm thick cross-sectional slices were obtained for each tumour. Quantitative morphometry was used to estimate the amount of necrosis within those slices. A spatially heterogeneous distribution of phosphate metabolites and pH was detected in most of the tumours. Levels of necrosis ranged from 0-24% for the tumours in this study. There was no significant correlation between the extent of necrosis over this range and levels of metabolites or pH, suggesting that factors besides necrosis can contribute to spectral heterogeneity.

Potentiation of interleukin 1 alpha mediated antitumor effects by ketoconazole.

In the present studies, the regulatory role of adrenal hormones on the antitumor activity of recombinant human interleukin 1 alpha (IL-1 alpha) was investigated. Ketoconazole, a potent but transient inhibitor of adrenal steroid hormone biosynthesis, inhibited IL-1 alpha induced increases in plasma corticosterone. In s.c. RIF-1 tumors (C3H/HeJ mice) ketoconazole potentiated IL-1 alpha induced hemorrhagic necrosis (59Fe labeled RBC uptake) and prolonged intervals of low tumor perfusion (86Rb+ uptake) and attendant depletion of tumor high energy phosphate reserves as determined by in vivo 31P nuclear magnetic resonance spectroscopy. In normal muscle and skin the ketoconazole-IL-1 alpha combination had no effect on RBC content and little or no effect on tissue perfusion. Ketoconazole potentiation of IL-1 alpha induced tumor pathophysiologies was accompanied by time and ketoconazole dose dependent potentiation of RIF-1 tumor clonogenic cell killing. Although ketoconazole at 40 mg/kg and IL-1 alpha at 25 micrograms/kg alone each produced approximately 50% clonogenic cell kill, a combined treatment (IL-1 alpha 1 h after ketoconazole) resulted in surviving fractions of approximately 1.5%. In vitro, ketoconazole and IL-1 alpha induced only additive clonogenic cell kill in primary RIF-1 explant cultures. The effect of elevated plasma corticosterone levels, induced by ketamine-acepromazine anesthesia, on IL-1 alpha responsiveness was also studied in the RIF-1 tumor model. In C3H/HeJ mice, anesthesia increased plasma corticosterone levels within 30 min, abrogated the IL-1 alpha effect on tumor perfusion, and prevented depletion of tumor high energy phosphate metabolite reserves. Our results are consistent with the hypothesis that IL-1 alpha mediated adrenal hormone responses exert a profound negative feedback on IL-1 alpha antitumor activities. Our data also indicate that adrenal steroid hormone biosynthetic pathways could provide a focus for modulation strategies to increase the efficacy of cytokine based therapeutic interventions.

Proton nuclear magnetic resonance methyl and methylene linewidths from plasma decrease during postprandial lipemia.

Narrow proton nuclear magnetic resonance (1H-NMR) linewidths from plasma have been associated with the presence of malignancy (Fossel et al., New Engl. J. Med. (1986) 315, 1369-1376). In that study, subjects and controls were not fasted. In the present study, 1H-NMR methyl and methylene linewidths were measured in plasma from normolipemic individuals without cancer both during fasting and every 90 min after eating a fat meal. Plasma lipoprotein levels were measured in order to relate results to postprandial lipemia. Methyl, methylene, and average 1H-NMR linewidths were strongly positively correlated with high-density lipoprotein levels and inversely correlated with triacylglycerol-rich lipoprotein levels in both the fasting and postprandial states. Linewidths decreased postprandially, reaching a nadir at the peak of plasma triacylglycerol levels. This study demonstrated that postprandial lipemia can lead to narrowing of plasma methyl and methylene resonances comparable to that reported for subjects with cancer.

31P NMR spectroscopic study of bioenergetic changes in radiation-induced fibrosarcoma-1 after radiation therapy.

The effects of localized gamma-irradiation on the in vivo 31P NMR spectra of RIF-1 tumors grown subcutaneously in C3H/HeN mice have been examined before and during the week after treatment. Increases in the ratio of phosphocreatine (PCr) to inorganic phosphate (Pi) and in tumor pH, and decreases in the ratio of Pi to the beta phosphorus resonance of the nucleotide triphosphates (beta NTP) were observed in irradiated tumors. The time course of changes in the 31P spectrum following treatment was the opposite of the pattern during untreated growth, and the magnitude and duration of the changes increased with increasing radiation dose, decreasing clonogenic cell survival and increasing growth delay. To examine the possibility that nontherapeutic systemic effects of the tumor irradiation were responsible for the changes observed, a number of animals bearing two tumors were examined. One tumor on each mouse was selectively irradiated. Changes in tumor volume, Pi/beta NTP, PCr/Pi, the ratio of phosphomonoesters to beta NTP, and tumor pH were all significantly different in the treated compared to the untreated tumor on each animal, indicating that these changes in 31P NMR spectra were a response to radiation therapy and not a systemic response to radiation toxicity.