Cyclocreatine accumulation leads to cellular swelling in C6 glioma multicellular spheroids:...

[CANCER RESEARCH 55, 153-158, January 1, 1995]

Cyclocreatine Accumulation Leads to Cellular Swelling in C6 Glioma MulticellularSpheroids: Diffusion and One-Dimensional Chemical Shift Nuclear MagneticResonance Microscopy1

Yael S. Schiffenbauer, Catherine Tempel, Rinat Abramovitch, Gila Meir, and Michal Neeman2

Department of Hormone Research, Weizmann Institute of Science, Rehovot 76100, Israel

ABSTRACT

Cyclocreatine, an analogue of creatine, inhibits tumor cell proliferationin vitro and in vivo. The effects of Cyclocreatine in large C6 gliomamulticellular spheroids were mapped here by magnetic resonance microscopy. Diffusion-weighted images of C6 glioma spheroids resolved the

bright viable rim and the dark necrotic center. Sequential sets of diffusionimages, following Cyclocreatine administration, showed increasing self-

diffusion coefficients of the intracellular water in the viable rim(0.49 x 10~s cm2/s for untreated spheroids, 0.62 x 10~5 cm2/s after 48 h

perfusion with 20 HIMCyclocreatine). This fact correlated with cellularswelling apparent in histológica! sections. The radial distribution of cy-

clocreatine and soluble lipids across perfused C6 spheroids was measuredby one-dimensional chemical shift imaging. Cyclocreatine accumulation

was prominent throughout the viable cell layer, with no Cyclocreatineaccumulation in the necrotic center. In both cyclocreatine-treated and

control spheroids the lipid signal was highest in the necrotic center andlower in the inner viable cell layer.

INTRODUCTION

Resistance to radiation and chemotherapy of the hypoxic/quiescentcell fraction in solid tumors compromises the efficacy of most currentanticancer treatment modalities. The multicellular spheroid model wasdeveloped for the study of the microenvironmental heterogeneity insolid tumors leading to this type of resistance (1, 2). The spheroids arespherical aggregates of cells with an outer proliferating layer, a layerof quiescent but viable cells, and an inner necrotic center. The kineticsof spheroid growth are similar to the growth of tumors in vivo.Initially, the spheroids are composed of an homogenous population ofproliferating cells. As growth progresses, the proportion of nonpro-

liferating cells increases. Eventually, the inner region of the spheroidbecomes necrotic because of the limited diffusion of nutrients, such asoxygen and glucose, and the accumulation of toxic waste products (1).The spherical symmetry of the spheroid lends itself to mathematicalmodeling and simplifies data analysis (3).

Mapping physiological properties in large multicellular spheroidshas been the goal of many recent studies (4-8). Among these,

microelectrodes were used for mapping pH and oxygen gradients (4).A bioluminescence method for metabolic imaging made it possible todetermine the spatial distribution of different metabolites such asglucose, láclate,and ATP in frozen tissue sections with an estimatedresolution of 20 firn (5). ESR microscopy was used for mappingcellular integrity and oxygen gradients (6), and proton NMR3

Received 6/13/94; accepted 10/31/94.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1This work was supported by a research grant from The Frenkel Foundation and

equipment was funded by The Wolfson Foundation and The Israel Academy of Sciences.M. N. is an incumbent of the Helena Rubinstein Career Development Chair in CancerResearch.

2 To whom requests for reprints should be addressed, at Department of Hormone

Research, Weizmann Institute of Science, Rehovot 76100, Israel.3 The abbreviations used are: NMR, nuclear magnetic resonance; CY, Cyclocreatine

(l-carboxymethyl-2-iminoimidazolidine); PCY, phosphocyclocreatine; CHESS, chemical

shift selective; TR, repetition time.

microscopy was used for mapping diffusion properties of water indifferent regions of EMT-6 spheroids (7, 8). Compartmentation ofwater in the viable rim of mouse mammary tumor (EMT-6) spheroids

was determined by NMR microscopy through application of a diffusion filter (£)(8). Mathematical models of mass transfer in spheroids(3) could be critically evaluated using the combined information fromthese methods and used for assessing changes during treatment.

Here we report the use of NMR microscopy for the study of theresponse of C6 glioma spheroids to treatment with CY. CY, ananalogue of creatine, can be supplemented by diet and can replacecreatine in the creatine kinase reaction (9). The stable PCY productcan serve as an efficient buffer for intracellular ADP during ischemieinsults to normal tissues (10). CY has recently been shown to act asan inhibitor of tumor growth in vivo (11-12) and inhibits proliferation

of tumor cells in vitro (12, 13). In a previous study we found by31P-NMR spectroscopy that the transport of CY in C6 rat glioma cells

was sodium dependent and led to extensive intracellular accumulationof PCY (13). In the study presented here, diffusion mapping was usedfor determination of changes in cellular packing and morphology. Amethod for metabolic mapping of a single perfused spheroid bychemical shift NMR imaging has been developed and applied formapping CY + PCY distribution. In this experiment the sphericalsymmetry of the spheroid enabled us to obtain the radial distributionof nutrients from single projections of the spheroid.

MATERIALS AND METHODS

Cell Culture and Spheroid Preparation. C6 rat glioma cells were cultured in DMEM supplemented with 5% PCS (Biological Industries Israel) andantibiotics: 50 units/ml penicillin, 50 mg/ml streptomycin, and 125 mg/mlfungizone (Biolab Ltd.). Aggregation of cells into large spheroids of about1-1.5 mm was initiated by plating cells from confluent cultures in bacterio

logical plates. After 48 h the spheroids were transferred to a 250-ml spinnerflask (Bélico).One week later the spheroids were transferred to a 500-ml

spinner flask and the medium was changed every other day for approximately1 month. Other details of spheroid culture were essentially as reportedpreviously (8, 14).

Perfusion of a Single Spheroid during NMR Measurements. The spheroid was positioned on a porous filter in a 3-mm Teflon tubing and placed insidea 5-mm NMR tube. Growth medium was supplied at a rate of 1.15 ml/minusing 1-mm internal diameter polyethylene tubing. Another tubing placed

above the medium level was used to deliver a filtered gas mixture (95% air, 5%C02) into the NMR tube. The rate of the return flow was adjusted to be fasterthan the inflow line, thus reoxygenating the used medium with no foaming ofthe serum-containing medium. The medium reservoir (30 ml) was maintainedat 39°C.An additional return line was added as a precaution in case of failure

in the main return line. All tubing was presterilized by -y-irradiation (15). NMR

measurements were done on spheroids of 1 mm in diameter or larger.Setup of NMR Microscopy Parameters. NMR microscopy measurements

were performed on a 400-MHz narrow-bore AMX Bruker spectrometerequipped with a microimaging attachment. A 5-mm Helmholtz coil was used(9 JASfor a hard 90°pulse). Nonactively shielded gradients were shaped with

a preemphasis unit (Bruker). A delay of 1-2 ms was incorporated into all of the

pulse sequences for gradient stabilization. Sample temperature was maintainedat 36°Cusing the variable temperature unit of the spectrometer. A horizontal

XY slice through the center of the spheroid was selected by using a presaturated

153

Research. on January 24, 2016. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

NMR MICROSCOPY OF CYCLOCREATINE UPTAKE IN C6 SPHEROIDS

gradient echo-imaging pulse sequence. In this sequence the candidate XY slice(with respect to position and thickness) was saturated by application of a 90°

2-ms sine pulse and a slice selection gradient along the Z axis. The excitedspins of this slice were then dephased by a 2-ms magnetic field gradient pulsealong the Z axis. This preparation was repeated for each phase-encoding step(0.4-mm slice, 256 X 256 data points, 10-mm field of view, 100-ms repetitiondelay, 1-min total acquisition). This scout image of a ZY plane showed the

perfusate flow and the spheroid, with the candidate slice appearing as a darkband. The offset for the presaturation pulse was adjusted until the slice wascentered on the spheroid equatorial. The same pulse length, shape, offset, andgradient strength were then used for the diffusion and chemical shift imagingexperiments.

Diffusion NMR Microscopy. Diffusion studies were performed using aspin echo pulse sequence (8, 16). The attenuation of signal due to diffusion isdescribed by a general "h value":

InA(gd)

A(gd = 0)= -bD

where D is the apparent diffusion coefficient. In addition, the imaginggradients alone can impose significant diffusion weighing describedby a diffusion filter £(8). In the experiments reported here theduration of the frequency-encoding gradient was adjusted so as to

attenuate signals from extracellular water (8, 17). The acquisitionconditions were TR of 1 s; a soft sine shaped 2 ms 90°pulse, a hard180°pulse; diffusion time A = 9 ms; diffusion gradient time 8 = 3

ms; and in-plane resolution of 23 p.m. The frequency-encoding gra

dient was 7.8 G/cm; slice thickness of 400 p,m and diffusion gradients(grf)of 0, 6, 12, and 18 G/cm.

One-Dimensional Chemical Shift Imaging. Projections of water, CY,and lipids from the spheroids were obtained by one-dimensional chemical

shift imaging based on a spin echo pulse sequence (18). In this experiment,chemical shift information was encoded in one axis, and spatial informationwas obtained by phase encoding along the other axis (Fig. 1). A slicethrough the center of the spheroid was selected by 2-ms sine 90°and 180°

pulses and 10 G/cm gradient (slice thickness of 400 /¿m).The sphericalsymmetry of the spheroid enabled us to deduce the radial distribution fromsuch one-dimensional projections. The experimental parameters used inthis study, namely, 256 phase-encoding steps and 6-mm field of view,

resulted in a spatial resolution of 23 /¿m/point. Projection of intracellularwater was obtained by addition of diffusion gradients to suppress themedium and extracellular water. For observation of CY, the water signalwas suppressed by three 10-ms CHESS pulses, followed by gradients at the

beginning of the pulse sequence (19).

CHESS 90 180 Spin echo

Fig. 1. Chemical shift imaging. A one-dimensional chemical shift-imaging pulse

sequence with CHESS suppression of water was used in order to get the distribution ofmetabolites in the spheroid. One-dimensional projections were obtained by phase encoding (256 traces, 23-ju,m resolution). Slice selection was done by 2-ms sine pulses and 10G/cm gradient (400-^.m slice thickness). CHESS pulses were 10-ms sine positioned on the

resonance frequency of water.

BFig. 2. CY effects on multicellular spheroids. Light microscope pictures of histological

sections of C6 spheroids stained with eosin and hematoxylin. A, control untreatedspheroids; B, C6 glioma spheroids treated with 20 niM CY for 48 h. Note the swelling ofnuclei and cells through the viable rim following CY treatment.

Histology. Spheroids treated with 5 and 20 mM CY for 48 h were fixed in

Bouin solution for 24 h, washed in 70% ethanol, embedded, sectioned, and

stained with eosin and hematoxylin as previously descibed (8, 14).

Data Processing. NMR data were analyzed on a personal Iris work station(Silicon Graphics, Inc.). Statistical significance was determined from Student's

f tests. Projections of a slice (400 ¡im)through the center of the spheroid werefitted by a nonlinear least square routine to a model of three concentric layerswith different signal densities. An approximation of cylinder geometry wasused for layers with a radius significantly larger than the slice thickness. Layerswith radii close to or smaller than the slice thickness, such as the necroticcenter, were modeled by a sphere, truncated to the slice thickness. The totalradius of the spheroid was imposed on the fitting routine using the experimental data. The position of the other boundaries were allowed to vary in theoptimization procedure. This model reproduced the experimental data muchbetter than a model of only two concentric layers.

RESULTS

Cyclocreatine Treatment Causes Cellular Swelling. Histological sections of C6 spheroids exposed to CY for 48 h showed cellularswelling and swelling of the nuclei (Fig. 2). The nuclei also appearedless dense following treatment. Nuclei swelling was dependent on theconcentration of CY and was less pronounced for spheroids treatedwith 5 mM CY than for spheroids treated with 20 mM CY. The meandiameter of the spheroids shown in Fig. 2 was 600 /xm, with a viablerim thickness of about 250 jam and a necrotic center of 100 /xm. Theswelling was apparent in all spheroid diameters observed.

Water Diffusion in C6 Glioma Spheroids. Diffusion-weighted

images of perfused C6 glioma spheroids clearly resolved the two154

Research. on January 24, 2016. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

NMR MICROSCOPY OF CYCLOCREATINE UPTAKE IN C6 SPHEROIDS

Fig. 3. Diffusion-weighted images of C6 glioma spheroids. Proton NMR images of aperfused spheroid (1.1 mm in diameter), obtained on a 400-MHz Bruker AMX spectrometer at 36°C.A slow-diffusion filter (£= 2.27) was applied to attenuate extracellular

water. Diffusion gradients of 0 G/cm (A), 6 G/cm (ß),12 G/cm (C), and 18 G/cm (D)were used. Frequency encoding gradient = 7.8 G/cm (6-mm field of view); slice selectiongradient = 10 G/cm (slice thickness = 400 /xm). In-plane pixel resolution = 23 ^im.TR = 1 s; A = 9 ms; 8 = 3 ms. Total accumulation time = 2 h.

major parts of the spheroid: the viable rim appearing bright and thedark necrotic center (Fig. 3). The signal in the spheroid rim decayedslowly with increasing diffusion gradient, relative to the signal in thenecrotic center. From sets of images such as shown in Fig. 3, thediffusion coefficient in each region of the spheroid could be determined. By attenuating the extracellular water using a "diffusion filter"(£),we could select the intracellular water. The self-diffusion coefficient for the viable rim measured at a high-diffusion filter (£= 2.27)was 0.53 X 10~5 cm2/s, while measurements performed using alow-diffusion filter (£= 1.53) gave values of 0.69 X 10~5 cm2/s.

These results imply some compartmentation of water in the viable rimof C6 glioma spheroids, although to a smaller extent than foundpreviously for the EMT-6 spheroids (8). The water self-diffusioncoefficient in the necrotic center was significantly faster than in theviable rim (Table 1).

CY Effects on Water Diffusion. A time course of CY uptakewas obtained by accumulating sequential sets of diffusion images(four diffusion gradient values for each time point) every 2 h (Fig.4). These studies were performed using a high-diffusion filter(f = 2.27) in order to select for the intracellular water. The initialdiffusion coefficient for intracellular water was 0.49 (±0.09,SE) X 10~5 cm2/s (n = 2). Spheroids were then perfused with

medium containing 20 iriM CY. We observed increasing self-diffusion coefficients of the intracellular water in the viable rim ofC6 spheroids after CY treatment. The diffusion coefficient ofintracellular water after 48 h of CY treatment increased significantly and reached a value of 0.62 (±0.07) X IO"5 cm2/s (two

spheroids; P = 0.0001 by Student's i test for paired regions of

interest in each spheroid, relative to the initial value). This increased diffusion coefficient correlated with the swelling of cellsobserved in histological sections. An increased diffusion coefficient should result in larger dephasing of the spin echo and, thus,in decreased signal intensity in diffusion-weighted images, unlessthere is also a change in intracellular volume. However, we observed a consistent enhancement of the signal (of about 20%)despite the higher diffusion coefficient (Fig. 4). Since these experiments were conducted under conditions in which only intracellular water was observable, this signal enhancement was mostprobably caused by increased T2 relaxation for intracellular waterin combination with increased intracellular partial volume due tocellular swelling. The total volume of the spheroids did not change.These results are all consistent with water imbibition and cellularswelling occurring at the expense of the extracellular volumefraction.

Chemical Shift Imaging. The distributionof CY in the spheroidswas determined by one-dimensional chemical shift imaging (Fig.1). Large C6 glioma multicellular spheroids (1 mm in diameter)were treated with 20 mM CY for 48 h prior to the measurement.The outline of the spheroid projection was obtained from a diffusion-weighted projection of intracellular water (Fig. 5, left). Theprojection of intracellular water (Fig. 6, bold line) clearly resolvedthe viable rim and the necrotic center. By presaturation of the watersignal we could resolve two additional chemical species (Fig. 5,right): a cluster of signals due to CY and PCY (the peaks atapproximately 4 ppm) and a signal at 1.2 ppm. The residual watersignal was also visible at 4.7 ppm.

The signal at 1.2 ppm in 'H-NMR spectra of tumors is known to

be composed of soluble lipids and láclate.Differentiation betweenthe two was done here by measuring the attenuation of signalintensity due to diffusion in a pulsed field gradient spin echoexperiment. Láclateshould have a diffusion coefficient very closeto that of water, while the diffusion coefficient of soluble lipidsshould be a few orders of magnitude lower. By application ofdiffusion gradients up to 9 G/cm, we obtained 70% attenuation ofthe water but no detectable attenuation of the signal at 1.2 ppm.Based on these diffusion properties we tentatively assigned thissignal to soluble lipids. The signal at 4 ppm was assigned to CY bychanging the content of CY in the perfusate and observing theappearance of this signal. We could not resolve the area betweenthe signals of CY and PCY.

From chemical shift images such as shown in Fig. 5, we obtainedfor each chemical substance the projections along a slice throughthe center of the spheroid (Fig. 6). The projections were analyzedby fitting them to a model of three concentric layers with differentrelative concentrations. The obtained densities showed accumulation of soluble lipids in the center necrotic region of the CY-treatedspheroid. The intermediate region showed very low levels ofsoluble lipids, and no lipid signal could be observed in the outerlayer of the spheroid. Very similar projections of lipids wereobserved in control untreated spheroids. An almost reciprocaldistribution was observed for CY: the inner necrotic center showed

Table 1 Self-diffusion coefficients of water in the different regions of Co spheroids

Viable rim (£=Viable rim (f =

Necrotic centerFree medium2.27)°

1.53)n3Spheroids 6 ROIs each

2 Spheroids 6 ROIs each3 Spheroids 6 ROIs each3 Experiments 6 ROIs eachDiffusion

coefficient (cm/s)0.528X 10~5

0.688 X 10~50.947 X 10~52.250 X 10~5''SD

(cm2/s)0.034

X 10~50.028 X IO"50.168 X 10~50.330 X IO"5*

' Diffusion filter £= />¡X Dw, where b-t is the contribution of the imaging gradients to the b value and Dw is the self-diffusion coefficient of water.''The diffusion coefficient of the medium could not be determined accurately from these data sets due to the rapid nonlaminar perfusion in the center of the NMR tube.

155

Research. on January 24, 2016. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

NMR MICROSCOPY OF CYCLOCREATINE UPTAKE IN C6 SPHEROIDS

Fig. 4. Time course of CY effect on intracellular waterof C6 spheroids. Data were obtained by accumulatingsequential sets of four diffusion images as shown in Fig. 3.Spheroid diameter was 1.0 mm. The third image of eachdiffusion set is shown (diffusion gradient = 12 G/cm).

Each diffusion set was accumulated for 2 h. Note theincreasing signal intensity during PCY accumulation. Images (256 X 256) with in-plane resolution of 23 u,m wererecorded with TR = 1 s; A = 9 ms; S = 3 ms.

almost no CY + PCY, while the intermediate and outer regionsspanning the entire viable rim showed high concentration ofCY + PCY (Fig. 6).

DISCUSSION

The toxicity mechanism of CY in C6 glioma cells appears to berelated to the control of cellular osmotic pressure. Cellular expansionfollowing CY treatment was observed in histological sections of large

C6 glioma spheroids. This swelling was apparent also in NMRdiffusion measurements that showed an increase in the self-diffusion

coefficient of intracellular water, implying massive water imbibition.The observed increase in signal intensity of diffusion-weightedimages, despite the elevated self-diffusion coefficient, is in accord

with an increase in the partial volume of intracellular water. In aprevious study we found that CY is transported via an active sodium-

dependent mechanism (13). The cotransport of sodium with CY, and

Fig. 5. One-dimensional chemical shift imaging

of a perfused C6 spheroid loaded with 20 mM CYfor 48 h. Proton chemical shift information is encoded in the horizontal axis, while spatial information is encoded in the vertical axis. Left, diffusion-weighted projection of intracellular water. Righi,projections of lipids and CY + PCY. Water signalwas suppressed as described in "Materials andMethods." The three major signals, from left to

right, are residual water (4.7 ppm); CY + PCY(approximately 4 ppm); soluble lipids (1.2 ppm).Data were obtained on a Bruker 400 AMX spectrometer; slice thickness = 400 firn; resolution

along the phase encoding axis, 23 fim; spectralwidth = 4000 Hz; 1000 data points along thechemical shift axis. TR = 1 s; total accumulationtime = 6 h. Line broadening of 5 Hz and sine

square window functions were applied along thechemical shift and phase-encoding axes, respectively.

4.7156

ppm 4.7 3.9 1.2

Research. on January 24, 2016. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

NMR MICROSCOPY OF CYCLOCREATINE UPTAKE IN C6 SPHEROIDS

a

2ues

eo»

-0.6 -0.4 -0.2 O 0.2R (mm)

0.4 0.6

Fig. 6. Projections of CY and lipids from a large C6 spheroid. Projections through thecenter of a C6 spheroid show CY + PCY (A, data points) and lipids (B, data points). Theprojection of intracellular diffusion-weighted water is superimposed on both graphs (boldline). The projections were obtained from a one-dimensional chemical shift imaging

experiment as in Fig. 6, and made symmetrical around the center of the spheroid.Simulated curves for a three-layer model are shown for the signals of CY + PCY and for

the soluble lipids (thin lines in A and B, respectively).

the intracellular trapping of the phosphorylated CY, present the cellswith an increased osmotic and energy load which will affect quiescentcells that are energetically compromised at least to the same extentthat it will affect the outer proliferating cell layer.

The distribution of CY and soluble lipids in the different regionsof the spheroid was obtained from projections of a one-dimensional chemical shift-imaging experiment. The multicellular spher

oid is a unique model of tumor because of its almost perfectspherical symmetry. The deviation from perfect spherical symmetry was <5%. Thus, a projection of the whole spheroid or a thinslice through the center contains the whole radial information. Thereconstruction of this radial information can be done directly byback projection or by Abel transform (20). Alternatively, thisinformation can be obtained by simulation to a model of a numberof concentric layers, as done in this study. Using this model we cancorrectly account for the geometry of the sample associated withthe actual experimental slice thickness.

The model used for the simulations of the projections consisted ofthree concentric layers. By comparison with histological sections webelieve that the inner region represents the necrotic core, while theintermediate and the outer regions represent the viable rim of thespheroid. The signal at 1.2 ppm showed an extremely low-diffusion

coefficient and was, therefore, assigned to soluble lipids. In bothCY-treated and control spheroids, this signal was highest in the

necrotic region, with almost no accumulation in the viable rim. Thehigh content of lipids in the necrotic center could arise from theaccumulation of products of membrane degradation, in which the fattyacid chains are mobile enough to produce narrow 'H-NMR signals.Lipid signals were observed previously in 'H-NMR spectra of tumors,

including C6 glioma (21), and were assigned to the membrane oftumor cells in some cases (22) and to cytoplasmic lipid vesicles inother cases (21, 23).

The creatine kinase reaction is sensitive to pH and could be modulated by acidic conditions in the center of the spheroid. If these acidicconditions also prevail inside the cells, the phosphorylation of CYwould be reduced by shifting the equilibrium of the creatine kinasereaction. Extracellular pH gradients of 0.1 to 0.5 unit were measuredwith microelectrodes in multicellular spheroids (24). The extent of thepH gradient depended on the particular cell type, metabolite concentration, and buffer capacity of the medium. Despite the expectedextracellular pH gradients, CY accumulation was within our sensitivity, homogeneous throughout the entire viable rim, implying effectiveregulation of intracellular pH in inner cell layers. This finding is inaccord with that found in a previous 31P-NMR spectroscopy study, in

which there was no intracellular pH change in inner layers of spheroids (25). The uniform accumulation of CY + PCY throughout theviable rim is very promising with regard to the potential use of CY,since the quiescent cells occupying the intermediate region are usuallyresistant to most of the conventional treatments.

In summary, NMR microscopy studies of CY action on C6 glioma,cultured as three-dimensional multicellular spheroids, revealed an

interesting mode of action. Diffusion NMR microscopy studiesshowed marked facilitation of intracellular water mobility and elevated intracellular water content spanning the entire viable rim.Chemical shift imaging of CY content showed uniform accumulationof the drug throughout the viable rim. Thus, it appears that themechanism of CY inhibition of cell division in this system may be dueto osmotically driven cellular swelling. The feasibility of mappingmetabolite content in spheroids by NMR microscopy as shown hereopens the door for the study of microenvironmental effects on metabolic activity in this model. The noninvasive nature of the methodallows studies of treatments and acquisition of full time courses fromsingle preparations, reducing the error associated with biologicalvariation between samples. The expected sensitivity gains expectedthrough improved hardware and pulse sequences will probably contribute further to the spectrum of biologically relevant metabolites thatcan be followed.

ACKNOWLEDGMENTS

We would like to thank Professor Mildred Cohn and Professor KevinBrindle for stimulating discussions. Cyclocreatine was generously supplied byProfessor Mildred Cohn.

REFERENCES

1. Sutherland, R. M. Cell and environment interactions in tumor microregions: themulticeli spheroid model. Science (Washington DC), 240: 177-183, 1988.

2. Nederman, T., and Twentyman, P. Spheroids for studies of drug effects. In: H. Acker,J. Carlsson, R. Durand, and R. M. Sutherland (eds.), Spheroids in Cancer Research:Methods and Perspectives, pp. 84-102. New York: Springer-Verlag, 1984.

3. Casciari, J. J., Sotirchos, S. V., and Sutherland, R. M. Mathematical modelling ofmicroenvironment and growth in EMT6/Ro multicellular tumour spheroids. CellProliferation, 25: 1-22, 1992.

4. Mueller-Klieser, W. Microelectrode measurements of oxygen tension distributions in

multicellular spheroids cultured in spinner flasks. In: H. Acker, J. Carlsson, R.Durand, and R. M. Sutherland (eds.), Spheroids in Cancer Research: Methods andPerspectives, pp. 134-149. New York: Springer-Verlag, 1984.

5. Walenta, S., Dotsch, J., and Mueller-Klieser, W. ATP concentrations in multicellular

spheroids assessed by single photon imaging and quantitative bioluminescence. Eur.J. Cell Biol., 52: 389-393, 1990.

157

Research. on January 24, 2016. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

NMR MICROSCOPY OF CYCLOCREATINE UPTAKE IN C6 SPHEROIDS

6. Dobrucki, J. W., Sutherland, R. M., and Swartz, H. M. Nonperturbing test forcytotoxicity in isolated cells and spheroids, using electron paramagnetic resonance.Magn. Reson. Med., 19: 42-55, 1991.

7. Sillerud, L. O., Freyer, J. P., Neeman, M., and Mattingly, M. A. Proton NMRmicroscopy of multicellular tumor spheroid morphology. Magn. Reson. Med., 16:380-389, 1990.

8. Neeman, M., Jarret, K. A., Sillerud, L. O., and Freyer, J. Self diffusion of water inmulticellular spheroids measured by magnetic resonance microimaging. Cancer Res.,SI: 4072-4079, 1991.

9. Bessman, S. P. The creatine-creatine phosphate energy shuttle. Annu. Rev. Biochem.,54: 831-862, 1985.

10. McLaughlin, A. C., Cohn, M., and Kenyon, G. L. Specificity of creatine kinase forguanidino substrates: kinetic and proton nuclear magnetic relaxation studies. J. Biol.Chem., 247; 4382-4388, 1972.

11. Miller, E. E., Evans, E. A., and Cohn, M. Inhibition of rate of tumor growth bycreatine and cyclocreatine. Proc. Nati. Acad. Sci. USA, 90: 3304-3308, 1993.

12. Lillie, J. W., O'Keefe, M., Valinski, H., Hamlin, A. H, Jr., Lee Varban, M., and

Kaddurah-Daouk, R. Cyclocreatine (l-carboxymethyI-2-iminoimidazolidine) inhibits

growth of a broad spectrum of cancer cells derived from solid tumors. Cancer Res.,53: 3172-3178, 1993.

13. Steinfeld, Y., Meir, G., Tempel, C., Neeman, M., and Cohn, M. Cyclocreatine toxicityin multicellular tumor spheroids: 31P NMR spectroscopy and diffusion NMR microscopy. Proc. Soc. Magn. Reson. Med., 12: 257, 1993.

14. Freyer, J. P., and Sutherland, R. M., Regulation of growth saturation and developmentof necrosis in EMT6/Ro multicellular spheroids by the oxygen and glucose supply.Cancer Res., 46: 3504-3512, 1986.

15. Neeman, M., and Degani, H. Metabolic studies of estrogen and tamoxifen-treated

human breast cancer cells by nuclear magnetic resonance spectroscopy. Cancer Res.,49: 589-594, 1989.

16. Stejskal, E. O., and Tanner, J. E. Spin diffusion measurements: spin echoes in thepresence of a time-dependent field gradient. J. Chem. Phys., 42: 288—292, 1965.

17. Van Zijl, P. C. M., Moonen, C. T. W., Faustino, P., Pekar, J., Kaplan, O., and Cohen,J. S. Complete separation of intracellular and extracellular information in spectra ofperfused cells by diffusion-weighted spectroscopy. Proc. Nati. Acad. Sci. USA, 88:3228-3232, 1991.

18. Meyerhoff, D. J. Magnetic resonance spectroscopic imaging. In: Gillies R. J. (ed.),NMR in Physiology and Medicine, pp. 169-184. New York: Academic Press, Inc.,

1994.19. Van Zijl, P. C. M., and Moonen, C. T. W. Solvent suppression strategies for in vivo

magnetic resonance spectroscopy. NMR, Basic Principles and Progress, 26: 67-108,

1991.20. Majors, P. D., and Caprihan, A. Fast radial imaging of circular and spherical objects

by NMR. J. Magn. Res., 94: 225-233, 1991.21. Sam-Lai, E., Cucurella, M. G., Anis, C., Massarelli, R., Ziegler, A., Fouilhe, N.,

Decorps, M., and Remy, C. Determination by 'H NMR of the subcellular origin of the

mobile lipid resonance in an intracerebral rat glioma. Proc. Soc. Magn. Reson., 2:1330, 1994.

22. Mountford, C. E., Saunders, J. K., May, G. L., Holmes, K. T., Williams, P. G., Fox,R. M., Tattersall, M. H., Barr, J. R., Russell, P., and Smith, I. C., Classification ofhuman tumours by high-resolution magnetic resonance spectroscopy. Lancet, 1:651-653, 1986.

23. Callies, R., Sri-Pathmanathan, R. M., Ferguson, D. Y. P., and Brindle, K. M. Theappearance of neutral lipid signals in the 'H NMR spectra of an myeloma cell line

correlates with the induced formation of cytoplasmic lipid droplets. Magn. Reson.Med., 29: 546-552, 1993.

24. Acker, H., Carlsson, J., Holtermann, G., Nederman, T., and Nylen, T. Influence ofglucose and buffer capacity in the culture medium on growth and pH in spheroids ofhuman thyroid carcinoma and human glioma origin. Cancer Res., 47: 3504—3508,

1987.25. Freyer, J. P., Schor, P. L., Jarrett, K. A., Neeman, M., and Sillerud, L. O. Cellular

energetics measured by phosphorous NMR are not correlated with chronic nutrientdeprivation in multicellular tumor spheroids. Cancer Res., 51: 3831-3837, 1991.

158

Research. on January 24, 2016. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

1995;55:153-158. Cancer Res   Yael S. Schiffenbauer, Catherine Tempel, Rinat Abramovitch, et al.   Chemical Shift Nuclear Magnetic Resonance MicroscopyGlioma Multicellular Spheroids: Diffusion and One-Dimensional Cyclocreatine Accumulation Leads to Cellular Swelling in C6

  Updated version

  http://cancerres.aacrjournals.org/content/55/1/153

Access the most recent version of this article at:

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  .permissions@aacr.orgDepartment at

To request permission to re-use all or part of this article, contact the AACR Publications

Research. on January 24, 2016. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from