Eur. J. Biochem. 20Y, 707-715 (1992)

0FEBS 1992

8-Adrenergic stimulation of C6 glioma cells: effects of cAMP overproduction on cellular metabolites A rnultinuclear NMR study Isabelle PIANET, Paul CANIONI, Julie LABOUESSE and Michel MERLE Institut de Biochimie Cellulaire et Neurochimie du Centre National de la Recherche Scientifiquc, Universite de Bordeaux 11, France (Received June 18,1992) - EJB 92 0856

We used 31P-NMR spectroscopy to investigate the response of living C6 glioma cells to stimulation by a /I-adrenergic agonist, isoproterenol. In the presence of 3-isobutyl-l-methylxanthine, stimulation induced an accumulation of CAMP,making possible the NMR detection of the second messenger in living cells grown on microcarrier beads and perfused in the NMR tube. The cAMP signal rose to a maximum level within 20 - 25 min of stimulation; thereafter it decreased to the detection threshold within 60 min. At the same time, 40% increases of phosphomonoester and diphosphodiester signals were observed, whereas no significant change in phosphocreatine and nucleotide signals was detected. The kinetics of changes of the cellular content in phosphorylated metabolites were analyzed after recording 31P-NMR spectra of cell perchloric acid extracts as a function of time of stimulation. CAMPaccumulation in stimulated cells was evidenced by a near linear increase of its NMR signal as a function of incubation time (from 0 to 60 rnin). Concomitantly with the production of CAMP, the data showed 30% decreases ofphosphocrcatine and ATP levels within 60 min of stimulation, and an unexpected redistribution of pyrimidine and purine nucleoside triphosphates. At the same time, levels of phosphomonoesters (phosphorylcholine and phosphorylethanolamine) and phosphodiesters (glycerophosphorylcholine and glycerophosphorylethanolamine) rose (50% increase). 3C-NMR spectra of cell perchloric acid extracts prepared after isoproterenol stimulation of cells incubated in the presence of [I-' 3C]glucose indicated a higher glucose content in stimulated cells, whereas the resonance of ribose C1 was diminished. Moreover, the resonances of C1 of ethanolamine and choline (and their derivatives) were increased in spectra of stimulated cells, whereas that of C3 of serine was decreased. In addition, the 13C-NMRdata indicated that neither the pattern of glutamale carbon enrichment nor the glutamate/glutamine ratio was modified in stimulated cells. On the other hand, the heteronuclear coupling pattern of the lactate (methyl group) resonance in 'H-NMR spectra of cell incubation media indicated that no change occurred in the carbon flux through the pentosephosphate shunt under stimulation. The results of this multinuclear NMR approach are discussed in terms of metabolic responses of C6 cells to P-adrenergic stimulation and cAMP overproduction.

In hormone and neurotransmitter action on metabolic pathways, short- and long-term effects are to be differentiated, both of which are nowadays explorcd at the molecular level. The latter arc recognized to occur a few hours after stimulation and generally correspond to the modification of the gene transcription of key enzymes of the considered pathway (Nimmo and Cohen, 1987); they are studied at both the mRNA and the protein level. Thc former imply metabolic modifications induced by changes occurring, firstly, at the Correspondence to Michel Merle, Institut de Biochimie Cellulaire et Neurochimie du CNRS, 1 rue Camille Saint-Saens, F-33077 Rordeaux Cedex, France Ahhreviations. GroPCho, glycerophosphoryl choline; GroPEtn, glycerophosphoryl elhanolamine; iBuMeXan, 3-isobutyl-I-methylxanthine; P[CH,]P, methylcnc diphosphonic acid; PCho, phosphoryl cholinc; PEtn, phosphoryl ethanolamine; R3, purine trinucleotide; Y,, pyrimidine trinucleotide; 51, p or yNTP, a, p or 11 phosphate of NTP; 51 or BNDP, a or /I phosphate of' NDP.

plasma membrane level or in its close vicinity (changes in intracellular second messenger or critical cation concentration) followed, subsequently, by a cascade of cellular events comprising protein phosphorylation, changes in cell compartmentalization of sequestred proteins or ions, etc. (Koshland, 1984). Short-term effects are the consequence of an immediate disturbance of the stationary resting state of the cellular metabolism, and take place within tens of minutes following stimulation. They are essentially due to the modification of enzyme activities that disturbs the established flux of metabolites before any change in key enzyme syntheses occurs. Short-term effects are often metabolically pleiotropic, involve several interconnected pathways and are difficult to apprehend as a whole (Koshland, 1984). fi-Adrenergic regulation is one of the best studied hormonal/neurotransmitter stimulations. Its pharmacology has been extensively developed, and its molecular transduction mechanism is one of the best understood (Strosberg, 1991).

708 One of the well known targets of fl-adrenergic regulation is the metabolism of glycogen, an important energy store for many cell types. p-Adrenergic receptors are distributed throughout many organs and are also present in the nervous system (Pittman et al., 1980). In the brain, they are carried mainly by glial cells which are considered to be the metabolic partners of neurons in their energy needs, both for neurotransmission and the rapid removal of products of this activity (Minneman et al., 1981; McCarthy, 1983; Hertz and Schousboe, 1986). Investigation of the short-term effects of padrenergic stimulation on glial metabolism as a whole therefore deserves consideration. We chose to study this aspect of cellular regulation with a glial cell line, C6 glioma, on extracts of isoproterenol-stimulated cells in the presence of 3-isobutyl-1-methylxanthine (iBuMeXan), an inhibitor of phosphodiesterases, by using multinuclear NMR spectroscopy. Indeed, a multinuclear approach including 31P, I3C, and 'H-NMR presents the specific advantage of checking many aspects of the cellular metabolism owing to the large range of compounds whose resonances can be detected, thus making it possible to reveal an eventual unexpected metabolic change that could not have been observed using another method focused on a particular metabolic pathway. Furthermore, 'P-NMR allows measurements on living cells and makes possible the direct analysis of the second messenger which is at the origin of the metabolic disturbance. The present study demonstrates that, in the presence of iBuMeXan, C6 cells are able to cope, at least in the short term, with an overproduction of CAMP, probably be sequestering it. This unexpected property of C6 cells could be one of the explanations of their ability to minimize the disturbance of the cellular metabolism observed here in noncontrolled cAMP overproduction, which could mimic an acute, pathological stress state.

MATERIALS AND METHODS Materials iBuMeXan, isoproterenol, luciferin and luciferase were from Sigma (USA). The C6 glioma cell line was a gift from Dr Legault (Collkge de France, Paris, France). [l-'3C]Glucose was purchased from the Centre d'Energie Atomique (Saclay, France).

Cell culture Cells were seeded at a density close to lo4 cells/cm2 in 10cm-diameter Falcon dishes or close to 2 x lo4 cells/cm2 on microcarrier beads (Biosilon) as described previously (Pianet et al., 1991). The experiments were carried out 6 days after plating, with confluent cells at a density close to 4 x lo5 cells/ cm2.

occupied by the bcad during perfusion (around 10 ml) equal to twice the volume of the packed beads. NMR analyses were done at 161.9 MHz with a Briiker AM 400 spectrometer equipped with a 13C/31Pdouble-tuned probe. After recording control spectra, 50 pM isoproterenol and 200 pM iBuMeXan were added to the perfusion medium. The acquisition parameters were as follows: 40-ps pulse width (65" flip angle), 205ms acquisition time and 2-s interscan delay. This delay made it possible to acquire spectra (250 or 500 scans) within 8.5 min or 17 min under medium conditions of signal saturation (the longitudinal relaxation time of metabolite phosphorus nuclei ranges over 0.2-5 s at 161.9 MHz; Merle et al., unpublished results); thus one can follow, under favorable conditions of metabolite observation, the kinetics of changes induced by receptor stimulation. A 15-Hz exponential line-broadening was applied before Fourier transformation. Peak areas were determined as compared to the signal at 18.4 ppm of 2 pmol methylene diphosphonic acid (P[CH,]P) enclosed in a sealed capillary.

NMR measurements on cell perchloric acid extracts The extracellular cell medium was replaced 4 h before stimulation experiments to ensure both a good cell energy status and a metabolic steady-state. For experiments done in the presence of [l-'3C]glucose, cell medium was again replaced at time zero by Dulbecco's modified Eagle's medium containing 5.5 mM [l-13C]glucose; 50 pM isoproterenol and 200 pM iBuMeXan were present (or absent for control samples) in the extracellular medium. Stimulation was stopped at the indicated time by removing the medium. Cells were then rinsed twice with a 0.9% NaCl solution at 4°C and frozen in liquid nitrogen. The extraction procedure was done as described previously (Pianet et al., 1991). Before NMR analyses, samples were adjusted to pH 7.3. 31P-NMR spectra were recorded using a 10-mm broadband NMR probe. Before the analysis, 0.5 pmol P[CH2]P was added to the extract. The acquisition parameters were: 5-ps pulse width (36" flip angle), 0.4-s acquisition time and 0.51-s repetition time. A 2.3-W proton-decoupling power was applied during acquisition; 50000 scans were recorded for each sample in order to obtain a good signal/noise ratio even for minor peaks. A 5-Hz exponential line-broadening was applied before Fourier transformation. The cell content in a metabolite i ( M J was determined using Eqn (1):

Mi = [(Ai/Aref/Fc) x 2MreJP

(1) where A j and Arcrrepresent the area of the resonance peak of the metabolite i and that of P[CH,]P, respectively; Mref,the amount of P[CH2]P added to the sample (0.5 pmol) and P the cell protein content. F,, the correction factor for the metabolite i, was determined from spectra of known amounts of standard of phos31P-NMR measurement on living cells phorylated metabolites and P[CH2]P acquired under the same C6 cells cultured on microcarrier beads were used for 31P- conditions as cell perchloric acid extracts, using Eqn 2. NMR measurements in the following way. The beads (3.5 g), Fc = (aiici) x (2Cref/aref) (2) covered by cells (approximately 3 x lo8 cells), were transferred into a 20-mm NMR tube adapted for cell perfusion (Pianet et where aj and aref(or ci and crCf)correspond to peak areas (or al., 1991). The perfusion medium (200 ml) was the culture concentrations) of metabolite i and P[CH2]P, respectively. medium supplemented with 25mM Hepes. It was thermo- The factor 2 in Eqns (1) and (2) corresponds to the two statted at 37"C, oxygenated by bubbling 95% 0 2 / 5 % C 0 2 equivalent P atoms in P[CH2]P. The F, values were: 0.71 for and buffered at pH 7.3. The medium was cycled with a flow GroPEtn, 0.73 for PCho, PEtn and GroPCho, 0.80 for Pi, rate of 12ml/min chosen in order to maintain the volume 0.89 for phosphocreatine, 0.91 for cAMP and 1.41 for PNTP.

709 13C-NMR spectra of cell extracts were recorded at 100.6 MHz in the 30-mm probe using 7-ps pulse width (55' flip angle), 0.65s acquisition time and 1.65s repetition time, under proton-decoupling conditions with a bilevel decoupling power of 6 W during acquisition and 1 W during relaxation. Each spectrum was derived from the acquisition of 40000 free induction decays, using an 8-Hz exponential line-broadening. Quantitative analyses of spectra were based on the peak area (at 63.7 ppm) of a standard of ethylene glycol added to the sample. PCH2P

Proton-NMR spectra of the extracellular medium Aliquots of the medium of cells incubated in the presence of [l-13C]glucose were filtered through a Chelex 100 (Na' form) column. After neutralization to pH 7.3 and freeze-drying, the samples were dissolved in D 2 0 . Proton-NMR analyses were performed in a 5-mm broad-band probe at 400 MHz using 45" flip angle and 6-s relaxation delay. The residual water signal was reduced by homonuclear presaturation. Spectra recorded from 64 scans were referenced to the lactate resonance (doublet centered at 1.32 ppm). The lactate-C3 "C/ "C ratio (R)corresponding to the ratio between the peak areas of the resonance doublet of 'H-13C coupling in [3"C]lactate, on the one hand, and the resonance of the methyl group of [3-"C]lactate, on the other, was determined. The relative value for the glucose flux (X)through the pentose phosphate pathway was then calculated using the expression X = 3(1 -R)/(3 2R) (Willy et al., 1986).

+

Biochemical assays ATP in cell extracts was determined by bioluminescence (Lemasters and Hackenbrock, 1978) using a LKB Wallach 1250 luminometer, and proteins by the method described by Lowry et al. (1951) using bovine serum albumin as a standard. RESULTS Effect of isoproterenol on C6 glioma cells: 31P-NMR study of perfused cells

A typical spectrum of living cells before P-adrenergic receptor stimulation is shown in Fig. 1 A. The main resonances were assigned to NTP (yNTP at - 5 ppm; aNTP at - 10 ppm; PNTP at - 18 ppm); Pi (at 3 pprn), phosphocreatine (at -2.3 ppm), phosphomonoesters (around 3.5 ppm) and diphosphodiesters (around - 12 ppm). Stimulation of cells was initiated by adding to the perfusate the P-adrenergic agonist, isoproterenol, and the phosphodiesterase inhibitor, iBuMeXan. Taking into account the flow rate and the void volume of both the perfusion chamber and the circuit, around 50 ml, spectrum B of Fig. 1 was recorded after 17 - 34 rnin (26 rnin mean time) of stimulation. As compared to the control spectrum, this spectrum included a new peak at around -1 ppm, assigned to the cAMP phosphate group. In addition, it showed significant increases of phosphomonoester and diphosphodiester resonances, whereas the resonances of other metabolites appeared unchanged. The time variations of the phosphorylated metabolite resonance intensities following fl-adrenergic stimulation were analyzed for 80 min. As shown in Fig. 2, the resonance of cAMP increased during the first 25 min, then decreased back to the detection limit within 60 min. On the same time scale,

/

ow

'

~

' 10

I

'

' 0

'

l

'

'

~

I

'

-10

Chemical shifts (ppm)

Fig. 1. Effects of /?-adrenergic stimulation on the 31P-NMRspectrum of C6 glioma cells; in vivo experiments. The B-adrenergic stimulation of approximately 3 x lo8 C6 glioma cells anchored on microcarrier beads was initiated by addition of 50 pM isoproterenol and 200 pM iBuMeXan to the perfusate. 31P-NMR spectrum of cells were recorded before (spectrum A, 500 scans) and after 26 min of j3-adrenergic stimulation (spectrum B, 500 scans), using a 40-ps pulse width (65" flip angle) and a 2-s repetition time. A 15-Hz line-broadening was applied to the data before Fourier transformation. Spectra were plotted without base-line correction. Spectrum B was normalized to spectrum A, so peak intensities are directly comparable. In both spectra, the Pi resonance corresponds mostly to the signal of extracellular Pi in the perfusale. Abbreviations used : PCH2P, methylenediphosphonic acid; PME, phosphomonoesters; PCr, phosphocreatine; DPDE, diphosphodiesters; TSO, isoproterenol; aNTP, BNTP and yNTP, a. j3 and y phosphates of NTP.

the signals of phosphomonoesters and diphosphodiesters vaned in a similar way. In contrast, no change in NTP resonances could be found during stimulation, whereas a slight decrease in phosphocreatine resonance, around 13% (i. e. at the limit of the area determination error), was observed after 22 min and remained stable. No change in Pi signal was observed. In fact, this signal corresponded mainly to extracellular Pi (0.9 mM in the medium used). As a consequence of the transient aspect of the resonance intensity changes, a quantitative analysis of spectra was not attempted. Such an approach would have required the determination of the saturation factor for each metabolite, particularly that of CAMP. during cell stimulation; this was indeed not possible. Effect of P-adrenergic stimulation on the phosphorylated metabolite contents of C6 glioma cells Perchloric acid extracts of cells grown on culture dishes were prepared at various times of cell incubation with isoproterenol and iBuMeXan. Fig. 3 A shows a 31P-NMR spectrum obtained from control cells. This spectrum displays

'

~

710

P PE t n

I

PEijr ----

o

n 0

m

(0

PCr

I3

I

B +IS0

3

Li

m

2i

a W

0

A

Stimulation time ( m i n )

Fig. 2. Variations of the 31P resonance intensity of phosphorylated metabolites as a function of p-adrenergic stimulation time. Peak areas on spectra acquircd under the conditions described in Fig. 1 were dctermined using, as an intcrnal reference, 2 bmol methylcnc diphosphonic acid enclosed in a sealed capillary. Areas were plotted taking as 100 arbitrary units the area of thc pNTP resonance in the control spectra. The data correspond to a typical one of four experiments. Considering the signal/noisc ratio obtained after the 15-Hz line-broadening, the absolute error on each point could be estimated to be around 10 arbitrary units. (*) PME, phosphomonoesters; (0)DPDE, diphosphodiesters; ( W ) NTP; (0) PCr, phosphocreatine; (e) CAMP.

the resonances of numerous phosphorylated compounds: phosphomonoesters [phosphorylethanolamine (PEtn), phosphorylcholine (PCho), Pi, phosphodiesters [glycerophosphorylethanolamine (GroPEtn) and glycerophosphorylcholine (GroPCho)], phosphocreatine, NTP, NDP and diphosphodiesters. The inserts in the figure correspond to extensions of the NTP resonance regions in order to reveal the relative contributions of purine and pyrimidine nucleotides (Cohen, 1983). Fig. 3 B shows a spectrum obtained after 60 rnin of jreceptor stimulation. The cAMP signal can be seen at around - 1 ppm. As compared to spectrum A, several significant relative changes are observed, including higher levels of phosphomonoesters and phosphodiesters and a smaller level of phosphocreatine. In addition, the overall patterns of NTP inultiplets appear modified, indicating a higher contribution of pyrimidine nucleotides. In contrast, diphosphodiester signals appear very similar in both spectra. The quantitative analyses of spectra of cell perchloric acid extracts prepared after different times of incubation with the agonist made it possible to describe the time courses of cell metabolite content changes. As shown in Fig. 4, a nearly linear increase of cAMP was observed during 60 rnin of cell stimulation. The initial rate of the process, 0.07 nmol min-' Ing-', was of the order of magnitude of that previously determined after a I-min cell stimulation, 0.23 nmol min-' mg-1, the cAMP concentration being measured, in this case, by radioimmuno assay (Pianet et al., 1989). On the same time scale, the levels of GroPCho, GroPEtn and PCho rose in the same way whereas that of PEtn increased during the first 20-

- IS0 r

v

r

0

---r I

,

,

I

-1B Chemical s h i f t s !ppm!

I

,

I

,

-za

Fig. 3. Comparative proton-decoupled '1'-NMR spectra o f perchloric acid extracts of control and P-adrenergic-stimulated C6 glioma cells. Perchloric acid extracts were prepared as a function of time of cell incubation in the presencc of the fi-adrenergic agonist. Spectrum A corresponds to control cells and spectrum B t o cells afler 60 min of stimulation. The inserts represent expanded regions of the yNTP and pNTP resonances, to show the contributions of purine and pyrimidine nucleotides. Abbreviations used: TSO, isoproterenol; PCh. phosphorylcholine; PEtn, phosphorylethanolamine; GPCh, glycerophosphorylcholine; GPEtn, glyccrophosphorylethanolamine; xNDP and SNDP, a and phosphates of NDP; rNTP, IjNTP and yNTP, a, p and */ phosphates of NTP; Rj,purine NTP; Y , , pyrimidine NTP; Pi, inorganic phosphate.

40 min, then slightly decreased. In this analysis, the amount of PEtn in cells was probably slightly overestimated owing to the fact that other resonances (probably sugar-phosphate signals) overlapped that of PEtn (as shown in Fig. 3). Concerning the contents in energy metabolites. the phosphocreatine level decreased until a plateau. corresponding to 60-65% of its initial value, was reached within 40 min of stimulation (Fig. 4). The NTP content of cells during stimulation (19.2 & 1.3 nmol/mg protein) was close to the control value (21.3 f 3.1 nmoljmg); however, as seen in Fig. 4, the composition of NTP evoked from 63.1 *3.2% and 36.9 3.2% to 51.6 2.5% and 48.3 k 2.5% for purine and pyrimidine nucleotides (Rjand Y3), before and after 60-min stimulation, respectively. The determination by biochemical assay of ATP in the extracts evidenced a decrease in cellular ATP (Fig. 4). The evaluation of cellular GTP ([GTP] = [purine nucleotides] - [ATP]) indicated that the maintenance of the cell NTP store was due to an increase in both GTP and pyrimidine nucleotides. On the other hand, the Pi content of stimulated cells (25.3 & 2.5 nmoljmg) was similar to that of control cells (28.8 8.3 nmol/mg).

+

71 1 LOO

A

75

c" '1 r. m 3

3 P 0

50 rr r-

a m

-x (0

Y

25

5 t

je 01

4

v

(0

0

0

20

0

0 W

-

E

0

t

n

zn

40

60

1

1

40

60

S t i m u l a t i o n time (min)

Fig. 4. Time course of variations in phosphorylated metabolite contents of C6 glioma cells as a function of the time of cell incubation with the p-adrenergic agonist. The phosphorylated metabolite contents in cells were determined from quantitative analyses of "P-NMR spectra of cell perchloric acid extracts after correction for partial saturation and NOE, using methylcne diphosphonic acid (0.5 pmol) added to the extract as an internal standard. The percentage of purine (and pyrimidine) NTP (right scale) was determined from the analyses of the a, and y phosphate resonances of NTP, the pattern of which made possible the determination of the contribution of the two types of nuclcotidcs to the signal. The amount of ATP in the extracts was determined by a bioluminescence assay and that of GTP was calculated by the difference: purine NTP-ATP. The data correspond to the mean values for four experiments. The relative error on each point was around lOo/u. Abbreviations used: PCho. phosphorylcholine; PEtn, phosphorylethanolamine; GPCho, glycerophosphorylethanolamine; GPEtn, glycerophosphorylcholine; PCr, phosphocreatine; R,, purine NTP: Y3, pyrimidine NTP.

8-Adrenergic stimulation of C6 glioma cells in the presence

of II-'3C]glucose

Fig. 5 shows two spectra of perchloric acid extracts prepared from 3 x 10' cells incubated for 30 min with 5.5 mM [l'3C]glucose either in the absence (spectrum A), or in the presence (spectrum B) of isoproterenol and iBuMeXan. Significant differences between these t w o spectra can be observed. In the low-field region, increases (55 f 11% increase, mean value f SD for four experiments) of carbon 3 (Cl) resonances

of the a and /i'anomers of glucose (at 92.6 and 96.5 ppm, respectively) were observed in spectra of stimulated cells, indicating a higher glucose concentration in these cells, as compared to the control. In contrast, two resonances (at 87.3 and 84.4 ppm) assigned to the C1 of ribose 5-phosphate (and its derivatives, ribose-5-P in nucleotides) were found to be less intense in spectra of stimulated cells. However, the most striking variations concerned the 60 - 70-ppm region. An increase (160 20%) in the resonance o f choline C1 (and/or choline derivatives, PCho, GroPCho, at 66.4 ppm) was observed under stimulation and an intense resonance of ethanolamine C1 (and derivatives, PEtn, GroPEtn, at 60.9 pprn), undetectable in spectra of control cells, was displayed in spectra of stimulated cells. In contrast, a decrease (35 f 8%) in the resonance of serine C3 (at 61.1 ppm) was noticed in stimulated cell spectra. These latter assignments were ensured by recording I3C-NMR spectra of standards (PCho, GroPCho, PEtn, GroPEtn). In the same spectral region, two other resonances (at 65.3 and 64.8 ppm), not yet assigned, appeared enhanced in spectra from stimulated cells (45 f 9% and 120 20% increases, respectively). In the high-field region, the major resonances were assigned to the C2, C3 and C4 of glutamate (at 55.2, 27.6 and 34.0 ppm, respectively) an glutamine (at 54.7, 27.0 and 32.0 ppm. respectively). A detailed analysis of the intensities of Glu and Gln carbon resonances (involving corrections for partial saturation and NOE) was carried out. The relative 'jC enrichments of glutamate carbons and the Glu/Gln enrichment ratio are reported in Table 1. They were not modified under stimulation. The extracellular media of isoproterenol-treated cells or control cells, incubated in the presence of [1-' 3C]glucose was analyzed by 'H-NMR (spectra not shown). The I3C enrichment of lactate C3 was determined in both cases by analyzing the heteronuclear coupling pattern, making it possible to calculate the fraction of glucose metabolized through the pentose phosphate pathway. The data, reported in Table 1, show no significant effect of the stimulation on this fraction.

+

DlSCUSSlON In spite of its non-invasive character, 31P-NMR spectroscopy applied to isolated living cells is virtually unable to detect the synthesis of CAMP in response to stimulation of cell receptors, since the concentration of the second messenger is in the micromolar range, a consequence of a tight control due to the relative rates of the reactions catalyzed by adenylyl cyclase and phosphodiesterases (Butcher, 1984; Beavo, 1987). In the presence of phosphodiesterase inhibitors, however, the intracellular concentration of cAMP can be increased beyond the detection threshold, thus making NMR investigations possible. Under such conditions, cAMP in melanoma cells has been detected by 'P-NMR measurements after melanotropin stimulation (Degani et al., 1991). However the low intensity of metabolite signals in isolated living cells does not allow one to record high-resolution spectra in a time compatible with the kinetics of the cellular response to stimulation. Therefore, NMR analyses performed on cell perchloric acid extracts are useful for complementing data. Under near physiological conditions, cAMP synthesis does not affect the cellular ATP (Shimizu et al., 1970; Krishna et al., 19701, whereas activation of adenylyl cyclase in the presence of phosphodiesterase inhibitors induces the accumulation of cAMP in cells at the expense of the ATP store, as observed in adipocytes and melanoma cells (Chung et al., 1985; Degani et

712

A

6

21

l

'

- IS0 2

I 'i ~

4

'

l

'

90.0

~

'

80.B

"

~

'

l

'

'

'

l

'

'

'

50.0 Chemical s h i f t s (ppm) 60.0

7m.m

"

'

48.16

'

"

'

'

30.0

I

l

'

'

'

~

'

20. 0

F ig.5 Effeets of fl-adrenergic stimulation of C6 glioma cells on the 13C enrichment of metabolites after 30 min of incubation in the presence of [l-'3C]glucose. Proton-dccouplcd 13C-NMR spectra of cell perchloric acid extracts were recorded arter 30 min of incubation in the presence of 5.5 mM [l-'3C]glucose, either in the absence (spectrum A), or the presence of the /3-adrenergic agonist (spcctrum B). Both spectra show the resonances of "C-enriched carbons, except the signals in the 72-77-ppm region arising from inositol (peaks 15) and that at 63.7 ppm from ethylene glycol (peak 13) used as a reference for chemical shift determination. Peak assignments in the 60-70-ppm region were ensured by recording spectra of pure standards of metabolites. Peak assignment: (1) alanine C3; (2) lactate c 3 ; (3) ghtamine c3; (4) glutamate c 3 ; (5) glutamine C4; (6) glutamate C4; (7) aspartate C3; (8) aspartate C2; (9) glutamine C2; (10) glutamate C2; (1 1) ethanolamine (and derivatives) C1; (12) serinc C3; (14) choline (and derivatives) C1; (16) and (17) ribose-5-phosphate anomers C1; (18) and (20) glucose anomers C1; (21) and (19) unassigned resonances, probably from sugars. The expanded region of spectrum B, plotted with a line broadening of 1 Hr instead of 8 Hz for the whole spectrum, shows the very close chemical shifts of cthanolaminc C1 (peak 11) and serinc C3 (peak 12) rcsonances. Abbreviation used: ISO, isoproterenol.

Table 1. Analysis of the '3C-enrichment of glutamate, glutamine and lactate in control and stimulated C6 glioma cells incubated during 30 min with isoproterenol in the presence of [l-'3C]glucose. The relative 13C enrichments of glutamate C2, C3 and C4, expressed as a percentage of the total glutamate enrichment, were calculated from peak areas in 13C-NMR spectra of cell pcrchloric acid extracts. after correction for partial saturation with TI values = 1.48, 0.77 and 1.00 s, respectively, and NOE values = 3.0, 3.1 and 2.9, respectively. The GluiCln ratio was determined from peak area ratios of homologous carbons of glutamate and glutamine. The ratio 13C/12Cfor lactate C3 was the ratio between the peak areas of the resonance doublet corresponding to the 'H-13C coupling in [3-13C]lactatc, on thc one hand, and the resonance of the methyl group of unlabelled lactate in 'H-NMR spectra of cell media, on the other. The relative value for the glucose flux through the pentose phosphate pathway (x)was calculated using the expression X = 3(1 -R)/(3 2R), with R representing the I3C/'*C ratio for lactate C3. The results are expressed as mean value k SD for three independent experiments.

+

Cells

Glutamate "C enrichment c2

C3

c4

21.7 0.2 20.2 f0.2

36.9 & 0.9 37.2 f 2.0

GluiGln

Laclate C3

X

2.9 F 0.4 2.4 f 0.6

0.59 f 0.07 0.53 0.06

0.30 & 0.08 0.34 0.04

%

Control Stimulated

41.4 i 1.7 42.6 k 2.0

al., 1991). The present work devoted to a cell line of glial origin analyzes the same type of situation, both from the point of view of the second messenger and from that of the consequences of such a cAMP overproduction on metabolites of different pathways: carbohydrates, high-energy phosphorylated metabolites, metabolites entering the Krebs cycle and corresponding also to neurotransmitters, lipid and phospholipid precursors or degradation products.

It appears that the time course of the cAMP resonance intensity changes determined from 31P-NMR spectra of living cells is different from that of the accumulation of the second messenger in cells, as determined from spectra of cell extracts (Figs 2 and 4, respectively). Although, cAMP can be excreted across the C6 cell membrane by an energy-dependent transport (Doore et al.. 1975), the observed difference could not be explained by a more efficient efflux of cAMP from cells per-

71 3 fused in the NMR tube than from cells used for the extracts. Indeed, in both cases, the composition of the medium was the same, and the cell number/total medium volume ratio was very similar: 300 x lo6 cells/200 ml medium for perfused cells and 30 x lo6 cells/lO ml medium for cells used in the extract (cells used in the extract being washed twice with 10 ml saline solution). Therefore, as only the signals of phosphorylated metabolites mobile in the intracellular medium can be detected in vivo (Lundberg et al., 1990), this result suggests that an efficient mechanism to mobilize (then cancel out) overproduced CAMPdoes exist. The involvement of this mechanism could bring down the free cAMP concentration before a decrease in the rate of synthesis of the second messenger by induction of an homologous desensitization process (as discussed by Sibley and Leflcowitz, 1985) could be observed. It would be too speculative to discuss the nature of the molecular species involved for such a cAMP mobilization. However, a recent study by Muller and Bandlow (1991) provided evidence for the existence, in yeast, of two new types of membraneanchored CAMP-binding proteins. The existence of similar structures able to bind cAMP in C6 glioma cells could be envisaged. Though cAMP synthesis generates pyrophosphate, this metabolite was not detected, indicating a high level of pyrophosphatase activity. The same observation was made by Degani et al. (1991) for melanoma cells. In contrast, under stimulation, a large increase in diphosphodiester resonances was observed in spectra of living cells, but not in spectra of cell perchloric acid extracts. Considering this discrepancy, no satisfactory explanation may be found, apart from the possibility of hydrolysis of undefined diphosphodiesters during the extraction process. As observed for other cell types (Chung et al., 1985; Degani et al., 1991), a decrease in ATP store was correlated to cAMP synthesis, its amplitude closely reflecting cAMP production, 4 and 3.3 nmol/mg after 60 min of stimulation, respectively. This result emphasizes the fact that, in the presence of iBuMeXan, adenosine was trapped in cAMP at the expense of ATP, and that the nucleoside store was not replenished from de now synthesis. During stimulation, the phosphocreatine level was also depleted (a drop of around 6.5 nmol/mg within 60 min). This decrease suggested that, regarding the lesser pool of adenosine nucleotides available for energy recycling, phosphocreatine acts as a buffer to maintain the higher potential ATP level. The transfer of the highenergy phosphate from phosphocreatine to ATP is governed by creatine kinase, generally described as working under equilibrium. Therefore, under p-adrenergic stimulation, the maintenance of equilibrium would require a net increase in cytosolic ADP concentration to balance the [phosphocreatine]/[creatine]decrease. Owing to the low cellular ADP pool (less than 10% of ATP), and, a.fortiori, to the smaller NMR-visible cytosolic ADP pool, the NMR data from living cells did not make it possible, however, to check the equilibrium assumption. A striking event linked to P-adrenergic stimulation is the simultaneous increase of nucleoside triphosphates other than ATP. To our knowledge, this result has not already been reported. After 60 min of stimulation, the estimated cellular GTP was around twice that in control cells (2.1 and 1 nmol/ mg, respectively) and, at the same time, the pyrimidine trinucleotide content increased from 6.3 to 9 nmol/mg. Taking into account the depletion of the ATP store, these changes reflected a redistribution of the metabolites involved in the energy charge, the pyrimidine trinucleotides evolving from

37% to 48.5% of the NTP pool. As generally described, nucleoside diphosphates and triphosphates are interconvertible through the activity of nucleoside diphosphate kinase whose activity was described to be regulated by the energy charge of the adenylate pool (Thompson and Atkinson, 1971). Assuming that creatine kinase worked at equilibrium under stimulation, and that the concentration of AMP remained negligible as compared to that of ATP (in our experiments, this assumption was justified because the AMP and IMP resonances were in the range of the noise in spectra of cell extracts as shown in Fig. 3), the adenylate energy charge defined as ([ATP] 1/2]ADP])/([ATP] + IADP] [AMP] [CAMP]) was rather lower under stimulation than under resting conditions, even if CAMP,which could be considered as a dead-end product in the presence of iBuMeXan, it not taken into account for the evaluation. Such a situation suggests that, either as a direct or an indirect consequence of stimulation, the activity (or the specificity) of the nucleoside diphosphate kinase was modified, leading to an equilibrium constant in favor of guanine and pyrimidine trinucleotide syntheses. This conclusion can be taken together with the segregation of functions among nucleoside triphosphates, ATP being involved in providing energy for a wide variety of biological processes and GTP, UTP and CTP in providing energy for anabolic processes, as already discussed by Pall (1985). In most tissues, glycogen metabolism is known to be regulated, under 8-adrenergic stimulation, by phosphorylation of key enzymes; in brain, these regulations mainly concern glial cells (Nathanson, 1977). In our study on C6 cells, no glycogen was detected in extracts of control cells when using I3C-NMR, suggesting a rather low glycogen store in the cells, as already reported for C6 cells at confluency (Keller et al., 1981). The comparison of the "C-NMR signals of glucose C1 from control and stimulated cells indicated, however, a higher glucose content in stimulated cells. Taking into account that, in C6 cells, the control point for glucose utilization is believed to be membrane transport (Keller et al., 1981), this result would be in agreement with an induced inhibition of glycolysis, the rate of glycolysis becoming limiting, in the case of stimulated cells, as compared to that of the transport process, or at least, of the same order of magnitude. However, a positive effect on glucose transport may also contribute to the increase in intracellular glucose, and in fact, stimulation of glucose uptake by isoproterenol in glial cells was recently demonstrated (Hsu and Hsu, 1990). The relative activity of the pentose phosphate pathway in C6 glioma cells was unchanged under P-adrenergic stimulation (Table 1). However, the level of 13C enrichment of the C1 of ribose 5-phosphate (and/or its derivatives) was lowered. This metabolite and NADPH are the main products of the pentose phosphate pathway which presents two reaction branches connected with intermediates involved in glycolysis. Starting from [l-'3C]glucose, the oxidative branch leads to unenriched ribose 5-phosphate whereas the nonoxidative branch leads to ribose 5-phosphate enriched on C1 or on both C1 and C5. The lower I3C enrichment of ribose 5-phosphate C1 cannot be related to an increase in the activity of the oxidative branch, since from the analysis of 13Cenrichment of lactate C3, we observed that the net flux (relative to glycolysis) through the pentose phosphate way, i. e. the flux through the irreversible oxidative branch, was unchanged under stimulation (Table 1). Glutamate is a typical marker of Krebs cycle activity since this amino acid is in equilibrium with 2-oxoglutarate. Starting from [l-13C]glucose,the relative I3C enrichments of glutamate

+

+

+

714

c2, c3 and c4 for control and stimulated cells were very similar. In both cases, the enrichment of glutamate C2 was higher than that of C3 and close to that of C4 (Table 1). Such a result indicates a large supply of 13Cthrough the anaplerotic way involving pyruvate carboxykdse activity. Indeed, when oxaloacetate is formed from pyruvate, the "C-enriched pyruvate C3 becomes oxaloacetate C3 and, in the next cycle, 2-oxoglutarate C2, whereas at the same time a fraction of the previous 13C enrichment of oxaloacetate on C2 and C3, is lost. The fact that the pattern of glutamate 13C enrichment was unchanged under B-adrenergic stimulation suggests that the relative contributions of the ways providing carbon to the Krebs cycle were the same. More particularly, since the ratio C4/C3 was unchanged under stimulation, the isotopic dilution at the acetyl-CoA level, related to the utilization of a carbon source different from the pyruvate produced by glycolysis (for example fatty acids) was roughly the same, suggesting that acetyl-CoA was produced in the same way in both control and stimulated cells. Glutamine synthase is a key enzyme in the central nervous system since glutamine is the precursor of neurotransmitters such as glutamate and 4-aminobutyrate. Its activity, known to be specifically astroglial (Martinez-Hernandez et al., 1977), is increased by trophic factors released by surrounding neurons, the increase being linked to the differentiation state of astrocytes (Hayashi et al., 1988). This enzyme therefore represents a potential strategic target for hormonal regulation. For example, a CAMP-dependent decrease in glutamine synthase activity within hours (half-time of the process 14 18 h) was described for 3T3 adipocytes, this type of regulation being classified, as discussed in the introduction, as a longterm regulation (Miller and Burns, 1985). We found that the Glu/Gln enrichment ratio was unchanged under stimulation, indicating that the relative fluxes of the metabolic processes involved in the Glu and Gln balance were not modified during short-term regulation. Finally, 31P-NMR spectra of cell extracts evidenced a net increase of YEtn and PCho as well as GroPEtn and GroPCho during stimulation. The simultaneous increases in contents of these compounds could suggest either a modification of the phospholipid turnover or a net increase of their precursors, ethanolamine and choline, which arc known to be synthesized via serine from glyceraldehyde 3-phosphate. In fact, these two possibilities cannot be considered as exclusive. A modification in phospholipid turnover was proposed by Degani et al. (1991) to explain the increase in PEtn and PCho under stimulation of melanocytes by melanotropin stimulating hormone, though they did not report an increase in GroPEtn and GroPCho. Such a modification could be the consequence of either an inhibition of the synthesis orland an activation of the degradation of phospholipids. The two-step synthesis (PEtn CDP-Etn + PtdEtn) is catalyzed by ethanohmine-phosphate cytidylyltransferase and ethanolamine phosphotransferase, respectively. An analogous pathway is used for phosphatidylcholine synthesis. The reaction step catalyzed by cytidylyltransferase is rate-limiting (Sundler and Akesson, 1975). One could speculate that inhibiting this enzyme, as a consequence of 8-adrenergic stimulation, firstly induces an increase in PEtn and PCho levels and, as a secondary effect, an increase of GroPEtn and GroPCho levels, knowing that these metabolites can be directly synthesized from PEtn and PCha (Mampandry et al., 1991). However, activation of phospholipid degradation by phospholipases is not to be excludcd. Rccently, a Ca2 -dependent modulation of the activity of choline-phosphate cytidylyltransferase in C6 glioma +

cells was demonstrated (George et al., 1991), pointing OUt the possibility for receptor-mediated regulation of this metabolic pathway. The 13C-NMR data indicated an increase in the I3C cnrichinent of ethanolamine and choline (and their derivatives) under stimulation, particularly in the case of ethanolamine whose C1 resonance was nearly undetectable in control spectra. At the same time, the I3C enrichment of serine was decreased. These results strongly suggest a higher rate of ethanolamine and choline synthesis from serine, which could probably be due to an activation of serine decarboxylase. Such an effect of B-adrenergic stimulation has not previously been reported, and in fact, very little is known about isoproterenol effects on ethanolamine, choline and derivative syntheses. However, the three methylation steps of PEtn, leading to PCho, are known to be inhibited under stimulation (Hirata et al., 1979; Marin-Cao et al., 1983). Such an effect, coupled to the activation of serine decarboxylase, would result in a relatively higher de n o w synthesis of ethanolamine than of choline (derivatives). In fact, such assumptions are in agreement with our NMR data which indicate a relatively higher 3C enrichment of the ethanolamine derivatives. In conclusion, our work shows the possibility of using 31PNMR spectroscopy to monitor the cAMP signal inside living cells, after stimulation of the 8-adrenergic receptors by isoproterenol. Although the experiments were done under non-physiological conditions (presence of a phosphodiesterase inhibitor), new information concerning the effects of stimulation (or more precisely the consequence of a severe increase in cAMP concentration) was obtained by analyzing data acquired from both living cells and cell perchloric acid extracts. They demonstrate that, even in the absence of regulation of the intracellular cAMP concentration by active cyclic nucleotide phosphodiesterases, C6 glioma cells appear to bc able to cancel out the overproduction of the second messenger. They minimize the disturbances of the energy metabolism. ATP depletion (primarily linked to adenosine trapping in CAMP) is compensated by maintaining the total NTP store. Hence, the carbohydrate and intermediary metabolisms of stimulated cells are only slightly disturbed, with only increased intracellular glucose concentration. Howcver, an as yet unreported stimulation of ethanolamine and choline syntheses is observed in stimulated cells, which may be the primary explanation of the observed increase in PEtn and PCho as well as GroPEtn and GroPCho.

We would like to thank G. Raffard for his expert technical assistance and B. Matusiak for cell cultures. This work was supported by grants from the Centre Nutionul de la Recherche Scientfique, the lnstitut National de la Santt et de la Recherche MSdiicale, the UniversitC. de Bordeaux Ii, the Rtgion Aquitaine, the Ligue contre Ie Cancer and the Associution pour la Recherche contre le Cancer.

REFERENCES Beavo, J. A. (1987) Multiple isozymes of cyclic nucleotide phosphodiesterase, Adv. Cyclic Nucleotide Protein Phosphorylation Res. 22, 1-38. Butcher, R. W. (1984) Phosphodiesterase after twenty years: an introduction, Adv. Cyclic Nucleotide Protein Phosphorylation Res. 16, 1 - 12. Chung, F. Z., Weber, H. W. & Appleman, M. M. (1985) Extensive but reversible depletion of ATP via adenylate cyclase in rat adipocytes, Proc. Natl Acad. Sci. USA 82, 1614-1617. Cohen, S. M. (1983) Simultancous 13C and "P NMR studies of perfused rat liver. Effect of insulin and glucagon and a 3CNMR assay of the Mg' ', J. Biol. Chern. 258, 14294- 14308.

'

715 Degani, H., DeJordy, J. 0. & Salomon, Y. (1991) Stimulation of cAMP and phosphomonoester production by melanotropin in mclanoma cells: 31P NMR studies, Proc. Natl Acad. Sci. USA 88, 1506- 1510. Doore, €3. J., Bashor, M. M., Spitzer, N., Mawe, R. C. & Saier, M. H. Jr (1975) Regulation of adenosine 3’: 5’-monophosphatc efflux from rat glioma cells in culture, J . Biol. Chem. 250,4371 -4372. George, T. P., Cook, H. W., Bycrs, D. M., Palmer, F. B. St. C. & Spence, M. W. (1991) Channeling of intermediates in the CDPcholinc pathway of phosphatidylcholine biosynthesis in cultured glioma cells is dependent on intracellular Ca’ +,J . Biol. Chem. 266, 12419-12423. Hayashi, H., Hayashi, R., Tanii, H., Hashimoto, K. & Patel, A. J. (1988) Thc influence of neuronal cells on the development of glulamine synthetase in astrocytes in vitro, Dev. Brain Res. 41, 37-42. Hertz, I,.& Schousboe, A. (1986) Role of astrocytes in compartmentation of amino acid and energy metabolism, in Astrocyfes (Fedoroff, S. & Vernadakis, A., eds) vol. 2, pp. 179-208, Academic Press. Hirata, F., Strittmatter, W. J. & Axelrod, J. (1979) /I-Adrenergic rcccptor agonists increase phospholipid methylation, membrane fluidity, and /3-adrenergic receptor-adenlyate cyclase coupling, Proc. Natl Acad. Sci. USA 76, 368 -372. Hsu, C. C. & Hsu. C. S. (1990) Effect of isoproterenol on thc uptake of [‘4C]glucose into glial cells, Neurosci. Res. 9, 54-58. Keller, K., Lange, K. & Noske, W. (1981) D-Glucose transport in cultured cells of neural origin: the membrane as possible control point of glucose utilization, J . Neurochem. 36, 1012- 1017. Koshland, D. E. Jr (1984) Control of enzyme activity and metabolic pathways, Trends Biochem. Sci. 9, 155- 159. Krishna, G . , Forn, J., Voigt, K., Paul, M. & Gcssa, G. L. (1970) Dynamic aspects of neurohormonal control of cyclic 3 ’ 3 ’ AMP synthcsis in brain, Adv. Biochem. Psychopharmacol. 3, 155- 172. Lcmasters, J. J. & Hackensbrock, C. R. (1978) Firefly luciferase assay Tor ATP production by mitochondria, Method~vEnzymol. 57, 3650. Lowry, 0. H., Rosenbrough, N., Farr, A. & Randall, R. J. (1951) Protein measurcment with the Folin phenol reagent, J . Biol. Chem. 193,265 -275. Lundberg, P., Harmsen, E., Ho, C. & Vogel, H. J. (1990) Nuclear magnetic rcsonancc of cellular metabolism, Anal. Biochem. 191, 193 - 222. Mampandry, C. A., Freysz, L., Kanfcr, J. N.. Drcyfus, H. & Massarelli, R. (1991j Effect of monomethylethanolamine, dimethylethanolamine, gangliosides, isoproterenol, and 2hydroxyethylhydrazine on the conversion of ethanolamine to methylated products by cultured chick brain neurons, .I. Neurochem. 56, 1845- 1850. Marin-Cao, D., Alvarez-Chira, V. & Mato, J. M. (1983) 8-Adrenergic control of phosphatidylcholine synthesis by transmethylation in hepatocytes from juvenile, adult and adrenalectomized rats, Biochem. J . 216,675 - 680. Martinez-Hernandez, A., Bell, K. P. & Norenberg, M. D. (1977) Glutamine synthetase: glial localization in brain, Science 1Y5, 1356-1358.

McCarthy, K. D. (1983) An autoradiographic analysis of j-adrenergic receptors on immunocytochemically defined astroglia, J . Pharmacol. Exp. Ther. 226,282 - 290. Miller, R. E. & Burns, D. M. (1985) Regulation of glutamine synthetase in cultured 3T3-Ll adipocytes by insulin, hydrocortisone, and cyclic AMP, Curr. Top. Cell. Regul. 26, 65 - 78. Minneman, K. P., Pittman, R. N. & Molinoff, P. B. (1981) 8Adrencrgic receptor subtypes: properties, distribution and regulation, Annu. Rev. Neurosci. 4, 419 - 461, Miiller, G. & Bandlow, W. (1991) Two lipid-anchored cAMP binding proteins in the yeast Saccharomyces cerevisiae are unrelated to the R subunit of cytoplasmic protein kinase A, Eur. J . Biochem. 202, 299 - 308. Nathanson, J. A. (1977) Cyclic nucleotides and nervous system function, Physiol. Rev. 57, 157-256. Nimmo, FI. G. & Cohen, P. T. W. (1987) Application of recombinant DNA technology to studies of metabolic regulation, Biochem. J. 247,l- 13. Pall, H. L. (1985) GTP: a central regulator of cellular anabolism, Curr. Top. Cell. Regul. 25, 1-20. Pianet, I., Merle, M. & Labouesse, J . (1989) ADP and, indirectly, ATP arc potent inhibitors of CAMP production in intact isoproterenolstimulated C6 glioma cells, Biochem. Biophys. Res. Commun. 163. 11 50 - 11 57. Pianet, I., Merle, M., Labouesse, J. & Canioni, P. (1991) Phosphorus31 nuclear magnetic resonance of C6 glioma cells and rat astrocytes: evidence for a modification of the longitudinal relaxation time of ATP and Pi during glucose starvation, Eur. J. Biochem. 195, 87 -95. Pittman, R. N., Minneman, K. P. & Molinoff, P. B. (1980) Ontogeny of 81- and 82-adrenergic receptors in rat cerebellum and cerebral cortex, Brain Res. 188, 357-368. Shimizu, H., Creveling, C. R. & Daly, J . W. (1970) Effect ofmembrane depolarization and biogenic amines on the formation of cyclic AMP in incubated brain slices, Adv. Biochem. Psychopharmacol. 3, 135-154. Sibley, D. R. & Lelkowitz, R. J. (1985) Molecular mechanisms of receptor desensitization using the p-adrenergic receptor-coupled adenylate cyclase systcm as a model, Nature 31 7 , 124- 129. Strosberg, A. D. (1991) Structure/function relationship of proteins belonging to the family or receptors coupled to GTP-binding proteins, Eur. J . Biochem. 196, 1 - 10. Sundlcr, K.& Akesson, B. (1975) Regulation of phospholipid biosynthesis in isolated ral adipocytes. Effect of different substrates, J . Biol. Chem. 25Q13359-3367. Thompson, F. M. & Atkinson, D. E. (1971) Response of nucleoside diphosphate kinase to the adenylate energy store, Biochem. Biophys. Res. Commun. 45, 1581 - 1585. WiIly, J. A,, Williams, W. F. & Schleich, T. (1986) Dynamic assessment of hexose rnonophosphate shunt activity in rabbit lens by proton NMR spectroscopy, Biochem. Biophys. Res. Commun. 138, 1068- 1073.

Beta-adrenergic stimulation of C6 glioma cells: effects of cAMP overproduction on cellular metabolites. A multinuclear NMR study.

We used 31P-NMR spectroscopy to investigate the response of living C6 glioma cells to stimulation by a beta-adrenergic agonist, isoproterenol. In the ...
1MB Sizes 0 Downloads 0 Views