[CANCER RESEARCH 50. 552-557. February I, 1990)
Regulation of the Cytidine Phospholipid Pathways in Human Cancer Cells and
Effects of l-/?-D-Arabinofuranosylcytosine: A Noninvasive 3IP Nuclear
Magnetic Resonance Study
Peter F. Daly,1 Gerhard Zugmaier, David Sandier, Mary Carpen, C. E. Myers, and J. S. Cohen2
Pediatrie Oncology Branch [P. F. D., M. C.] and Medicine Branch /G. Z.. D. S., C. E. M., J. S. C.J. National Cancer Institute, National Instituten of Health. Bethesda,
Maryland 20892
ABSTRACT
Using "I' nuclear magnetic resonance spectroscopy we have noninvasively observed metabolic control through the cytidine pathways of phosphatidylcholine and phosphatidylethanolamine synthesis in intact ac
tively metabolizing MDA-MB-231 human breast cancer cells. Perfusion
with the phospholipid precursors ethanolamine or choline (2 HIM)indi
cates that the cytidylyltransferase enzymes are rate limiting for both
pathways. Complete inhibition of choline kinase with ethanolamine al
lowed the observation of the utilization of phosphocholine by the ratelimiting enzyme choline-phosphate cytidylyltransferase. The rate was
dependent on the phosphocholine concentration. Inhibition of glycerophosphorylcholine phosphodiesterase with accumulation of substrate was
also observed and allows an estimate of the flux through the degradative
pathways.
The human lymphoma cell line MOLT-4 was also found to contain
high levels of phosphocholine and phosphoethanolamine.
The levels of
these precursors in the MOLT-4 line are lowered by 40% after 6 h when
perfused with high dose l-/3-D-arabinofuranosylcytosine
(Ara-C) (400
Mm) but are unaffected by 2 «imAra-C or dideoxycytidine. High dose
Ara-C also resulted in lysis in 8-10 h. However, the MDA-MB-231
cell
line which is not sensitive to Ara-C showed no change in its spectrum
when perfused with Ara-C. A potential mechanism based on classic
phospholipid metabolism for the lytic effect of high dose Ara-C is
discussed.
INTRODUCTION
In vivo "P-NMR3
spectroscopy of humans has revealed a
prominent PME peak in tumors of breast (1), liver (2), lung
(3), bone (4), muscle (5), neural tissue (6), and skin (7) not
present in or in much lower concentration in the tissue of
origin. The PME resonance in cancer cells (8, 9), tumors in
nude mice (10), and brain tissue (11) has been resolved into two
dominant peaks. These have been identified as the precursors
PC and PE in the cytidine pathways for phosphatidylcholine
and phosphatidylethanolamine
synthesis (8, 9, 12). The catabolites, GPC and GPE are also frequently seen in "P-NMR
spectra of cells.
The three step pathway, choline to PC to CDP-choline to
phosphatidylcholine
is catalyzed by choline kinase (EC
2.7.1.32), choline-phosphate cytidylyltransferase (EC 2.7.7.15),
and phosphocholine transferase (EC 2.7.8.2), respectively. The
cytidylyltransferase enzyme is rate limiting (13, 14), but the
Received 12/15/88: revised 5/22/89; accepted 10/25/89.
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 with 18 U.S.C. Section 17.14solely to indicate this fact.
1Present address: Pittsburgh NMR Institute. .1260 Fifth Avenue. Pittsburgh.
PA 15213.
1To whom requests for reprints should be addressed, at Clinical Pharmacology
Branch, National Cancer Institute. Building 10. Room 6N119. Bethesda. MD
20892.
1The abbreviations used are: NMR. nuclear magnetic resonance: CDP-choline.
cytidine diphosphocholine; CDP-ethanolamine. cytidine diphosphocthanolamine;
DPDE, diphosphodiester; GPC, glycerophosphorylcholine: GPE, glycerophosphorylethanolamine; HC-3. hemicholinium-3; IMEM. improved minimal essen
tial media: PC. phosphocholine: PE, phosphoethanolamine; PME. phosphomonoester; Ara-C. l-/i-i>-arabinofuranosylc\ tosine; FCS. fetal calf serum.
question remains as to whether choline kinase may alter the
rate of the cytidylyltransferase by controlling the level of its
substrate, PC (13, 15). Analogously, ethanolamine is catalyzed
to PE then to CDP-ethanolamine and finally to phosphatidyl
ethanolamine by ethanolamine kinase (EC 2.7.1.82), ethanolamine-phosphate cytidylyltransferase (EC 2.7.7.14), and ethanolaminephosphotransferase
(EC 2.7.8.1). Degradation of
these phospholipids occurs via phospholipases A1 (EC 3.1.1.32)
and A2 (EC 3.1.1.4) to GPC and GPE, and then to choline and
ethanolamine by GPC phosphodiesterase (EC 3.1.4.2). Phos
phatidylcholine and phosphatidylethanolamine
are in a con
stant state of rapid turnover and are maintained by a balance
of these anabolic and catabolic pathways; or increased in
amount by a greater net flux through the synthetic pathways.
Choline and ethanolamine are both the start of synthesis and
the end of degradation for these two phospholipids. Because
choline, (trimethylethanolamine)
and ethanolamine are struc
turally similar, ethanolamine is a competitive inhibitor of cho
line kinase (12). Two isoenzymes of ethanolamine kinase exist,
one which is choline inhibited and one which is not (16).
Choline and ethanolamine are also product inhibitors of GPC
phosphodiesterase. These properties of choline and ethanola
mine can be used to manipulate the enzymes in the above
pathways and the effects on intact metabolizing cells observed
by "P-NMR.
Because of the prominence of PE and PC in human tumors
( 17) we have studied the significance of the high concentrations
of these metabolites to the control of phospholipid synthesis in
tumor cells. In this report we show a lack of significant accu
mulation of CDP-choline or CDP-ethanolamine in the NMR
spectra as compared to the large accumulation of PE and PC
when excess choline or ethanolamine are added to the perfusion
media of MDA-MB-231 cells indicating the cytidylyltransferase
enzymes are rate limiting for both pathways. In addition, by
inhibiting choline kinase with ethanolamine the formation of
PC was stopped. This allowed the subsequent direct observation
of the rate of utilization of PC by the rate-limiting cytidylyl
transferase enzyme, which is known to equal the rate of phos
phatidylcholine production in this pathway (18, 19). The rate
was fastest at high PC concentrations.
The PME peaks in the MOLT-4 human lymphoma cell line
were also found to be predominantly the phospholipid precur
sors PE and PC. Treatment of the MOLT-4 cells with high
dose Ara-C resulted in decreases in the PE and PC peaks which
preceded by hours the lysis of the cells. Since Ara-CTP, the
active metabolite, is a CTP analogue and CTP is the key
regulatory cofactor for the cytidylyltransferase enzymes a pos
sible mechanism of this chemotherapeutic on these enzymes is
discussed.
MATERIALS AND METHODS
Agarose Gel Threads. MDA-MB-231 human breast cancer cells were
grown as monolayers in NIH IMEM supplemented with pcnicillin552
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1990 American Association for Cancer Research.
CYTIDINE PHOSPHOLIPID
streptomycin (100 units/ml, 10 mg/liter), and 5% PCS under a 95%
air/5% CO2 environment. IMEM contains 400 ¿IM
choline and no
ethanolamine but was modified to contain 15 MMcholine and is desig
nated as "IMEM (normal choline)." IMEM (normal choline) plus 10
i«M
ethanolamine simulates normal serum levels for humans of these
two amines (12). MDA-MB-231 cells were grown to log phase (3050% confluency) or to 95% confluency, harvested with 0.5% trypsin0.2% EDTA (Gioco), and washed twice in HEPES buffered IMEM
with 5% PCS and then suspended in agarose. Cells were suspended in
gel threads by mixing 1.3 ml of cells (3-5 x 10") with 1.3 ml of 1.8%
agarose at 37°Cand extruding the mixture through SOO-^m internal
diameter tubing into a screw cap Wilmad NMR tube (10 mm). These
threads were then concentrated by placing an insert with inlet and outlet
tubing for perfusate into the NMR tube. A more detailed description
of the method has been published (20). Perfusate from a 500-ml
reservoir entered through 0.5-mm tubing which opened at the bottom
of the NMR tube and flowed at a rate of 0.5 ml/min upwards through
the threads exited via openings in the insert to a waste bottle. The
perfusate consisted of buffer A (50 miviHEPES-sodium salt at pH 7.5,
105 mivi NaCl, 5 mM KC1, 2 mivi MgCl2) and 11 miviglucose.
Basement Membrane Gel Threads. MDA-MB-231 cells or MOLT-4
human lymphoma cells were embedded by mixing 0.1 ml of cells with
2 ml of liquid basement membrane material (Collaborative Research
Incorporated, Bedford, MA) and extruding this mixture into 400-nm
diameter threads. Cells were allowed to grow in these threads in Petri
dishes in IMEM (normal choline) plus 10 ^M ethanolamine and 5%
PCS under a 5% CO2 environment and then were transferred to an
NMR tube and perfused under sterile conditions with the same cell
culture media. Details of this method have been recently reported (21).
3IP Magnetic Resonance Spectra and Data Analysis. 31P-NMR spectra
were recorded on a Varian XL-400 at 162 MHz at 37°C.For quanti
tative spectra the parameters were a pulse width of 48-^s, 90°flip angle,
32.5-s repetition time, a spectral width of 8000 Hz, acquisition time of
0.5 s, 8192 points, and 100 transients. These parameters were tested in
phantoms and gave quantitative results without selective saturation.
The same parameters also gave quantitative agreement when intact
cells were compared to extracts of the same cell line (20). For more
rapid scanning the parameters were a pulse width of 32 ¿is,flip angle
of 66°,repetition time of 2.0 s, acquisition time of 0.5 s, 8192 points,
PATHWAY REGULATION
pH 7.5 and 37°Cin agarose were perfused for 18 to 24 h with
buffer A with glucose and "P-NMR quantitative spectra were
recorded every 54 min. No significant change in any of the
resonances was observed during this time period. By contrast,
Fig. 1 shows the changes in the spectrum when perfused with
2 mivi choline. The increase in PC is rapid and large, and
equally significant is the lack of appearance or increase of any
peak in the DPDE region where CDP-choline resonates. The
DPDE peaks in cell lines have been identified as UDP-glucose,
NADP, NADPH, and UDP-jV-acetylglucosamine/galactosamine (11, 22-23). Similarly, perfusion with ethanolamine
caused a rapid increase in PE but no appearance of a CDPethanolamine peak (Fig. 2). For both choline and ethanolamine
infusion there was a rapid increase in PC or PE for the first 5
h and then the rate of increase declined. A saturation curve (1
—¿
e~") for the PC increase gave a time constant of 2.5 h and
for the PE increase of 3.0 h (Fig. 3).
Leakage of PC through the Membrane. MDA-MB-231 cells
were grown in basement membrane gel threads and perfused
with IMEM (normal choline) plus 10 urn ethanolamine. Cho
line kinase was inhibited by the addition of 100 ¿/MHC-3, a
specific choline kinase inhibitor. This resulted in a 40% reduc
tion in the level of the PC resonance but no change in the PE
resonance (Fig. 4). To determine if any of the decrease of the
and 1800 transients. This gave relative changes over time but not
quantitative results because of selective saturation of the PME, P¡.and
phosphodiester peaks. 31P chemical shifts are reported with a-ATP
standardized to -11.3 ppm which is equivalent to acidic phosphoric
acid solution at 0 ppm. Except where stated otherwise, the fi-ATP peak
remained stable with only a 10% variation around a mean value and
functions as an internal standard. The area under the /i-ATP peak in
our cell lines has been measured in acid extracts with an externally
added 1 mM phosphoric acid standard to be 2 mM with a standard error
of 0.3 mM which is the same as measured in cells by other techniques
(13).
Spectra were analyzed either by peak height since the line width of
the resonances did not change by more than 3% over the time course
of our experiments: by integration and estimates of the area on a Varian
ADS-4000 data station; or by curve fitting by transferring the data to
a DEC POP 10 computer and using the MLAB program assuming
Lorentzian-shaped peaks. Because they are only 0.5 ppm apart integra
tion or curve fitting to separate the areas of the PC and PE peaks gave
irreproducible results in intact cells and varied as much as 50% either
by the same person on different days, or by three different people
analyzing the same data. However, peak height measurements were
reproducible with less than 5% variation either by the same person or
different people. Therefore, the results for PE and PC are reported as
changes in peak height. However, integration or curve fitting of the
GPC peak were reproducible by the same person or different people
with less than 5% variation.
Fig. I. "P spectra at 162 MHz of MDA-MB-231 cells suspended in agarose
and perfused with buffer A, 11 HIMglucose, and 2 mM choline at 1. (A). 4 (B)
and 7 (C) h showing the large increases in PC, GPC, and GPE. No change in the
DPDE area due to CDP-choline formation ¡s
seen. Spectra were acquired al 37°C
with a 32.5-s repetition time and a 90' flip angle.
RESULTS
Observation of Rate-limiting Steps in Phosphatidylcholine and
Phosphatidylethanolamine
Synthesis. MDA-MB-231 cells at
553
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1990 American Association for Cancer Research.
CYTIDINE PHOSPHOLIPID
PATHWAY REGULATION
1.2
PE
i
PK
0.9
Q
W
ß-ATP
0-8
N
0.3
Effect
of HC-3
io
TIME (Hr)
Fig. 4. Relative changes of peak height ratios of PE//Ì-ATPand PC//J-ATP
after addition of 100 UMHC-3 to the perfusate (buffer A plus 11 mM glucose) of
MDA-MB-231 cells suspended in agarose.
present in the perfusate but no detectable amounts of PC. This
indicates that the large decreases in PC observed are due
predominantly to intracellular utilization and not leakage
through the membrane. This agrees with studies using radio
active isotopes (18-19).
Flux through Choline-Phosphate Cytidylyltransferase En
zyme. Since PC exists at a steady state concentration it is
necessary to completely inhibit formation to isolate and directly
observe utilization. Ethanolamine is a competitive inhibitor of
choline kinase in the MDA-MB-231 cell line (12). Perfusion
with buffer A and 2 mM ethanolamine and no choline caused
an exponential decrease (e~") of the PC peak (Fig. 3). where r
0.4
= 0.11/h. The half-life of PC was 6.3 h. At the end of 14 h, the
perfusate was changed to buffer A with 2 mM choline and no
ethanolamine. During the first 2 h after this change no increase
was seen in the PC peak indicating complete inhibition of
choline kinase. By comparison, 2 mM choline without the
previous presence of excess ethanolamine in the perfusate con
sistently caused a 30% increase of PC within 2 h. Repeating
the same experiment using 4 mM ethanolamine in the perfusate
gave the same results with an exponential half-life of 6.1 h.
This compares well with the half-life of 6-10 h observed for
HeLa cells (19).
The point represented by the normalized intensity of 1.0 in
Fig. 3 occurs with the PC/ß-ATParea ratio is 1.6 indicating a
PC concentration of 3.2 mM at log phase growth. We have
consistently reproduced this result both in harvested cells sus
pended in agarose and in cells actively growing under sterile
conditions in basement membrane gel threads. The equation
for PC concentration in Fig. 3 then is 3.2e~01" mM/h and the
0.2
derivative with respect to time of the PC decrease gives the flux
through the Cytidylyltransferase to be 0.35e~°'" mM/h. The
Fig. 2. "P spectra at 162 MHz of MDA-MB-231 cells suspended in agarose
and perfused with buffer A, 11 mM glucose, and 2 mM ethanolamine at 1 (A), 4
(B), and 7 (C) h showing the decrease in PC and large increases in PE. GPE, and
GPC. No change in the DPDE area due to CDP-ethanolamine formation is seen.
Spectra were acquired at 37"C with a 32.5-s repetition time and a 90' flip angle.
1.2
Effect
of Ethanolamine
en
x.
u
K
O
rate at / = 0 before PC concentration decreases is 0.35 mM/h,
which should be the steady state rate during log phase growth.
At this rate over a 15-h period (the doubling time in culture)
TIME (Hr)
this pathway could produce 5.3 mM of phosphatidylcholine.
Fig. 3. Relative changes of peak height ratios of PE/0-ATP and PC/0-ATP
over time in MDA-MB-231 cells suspended in agarose and perfused with buffer
Furthermore there is a degradation of phosphatidylcholine
A, 11 mM glucose, and 2 mM ethanolamine. Absolute millimolar changes are
estimated to be 0.12 mM/h (see below) during the course of the
discussed in the text. Spectra were collected every 54 min for 20 h. and fitted
data for Fig. 3. This would give a net production of 3.4 mM
with exponential curves. The signal to noise was 33 implying a measurement
over a 15-h doubling period at log phase. This is compatible
error of 0.03 units on the y axis.
with the total concentration of 3-5 mM of phosphatidylcholine
PC peak could be due to leakage through the membrane, 1-h measured in l g of BHK and HeLa cells, intestinal mucosa,
samples were obtained of the perfusate from the cells and and fetal lung (13).
Inhibition of GPC-Phosphodiesterase. Ethanolamine, choline,
concentrated 30-fold by lyophilization. With a signal to noise
of 400, quantitative "P-NMR spectra revealed the 1.25 HIMP¡ and HC-3 all inhibit GPC phosphodiesterase (12). Figs. 1 and
10
554
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1990 American Association for Cancer Research.
CYTIDINE PHOSPHOLIPID
2 show the accumulation of GPE and GPC when this enzyme
is inhibited. Fig. 5 is a graph of the increase in the GPC//3ATP peak height ratio. The buildup was linear and 2 mivi
ethanolamine had a slope of 0.08 units/h, 2 mM choline had a
slope of 0.10, and 100 Mm HC-3 had a slope of 0.16. The
scatter of data points for the GPC/0-ATP ratio during eth
anolamine inhibition results from the 10% variation of peak
heights in ^i-ATP, whereas the absolute increase in peak height
for GPC showed much less scatter. Data analyzed in terms of
area normalized to the ß-ATParea (2 mM) gave an estimated
increase in GPC of 0.18 mivt/h for HC-3, and 0.12 mM/h for
both choline and ethanolamine. Each GPE or GPC molecule
observed to accumulate after inhibition of GPC-phosphodiesterase is the result of removing two fatty acids from one
membrane phospholipid. Therefore, the 0.12 mM/h buildup of
GPC observed is an estimate of the rate of phosphatidylcholine
degradation.
MOLT-4 Human Lymphoma Cells. The MOLT-4 cells grew
as microspheroids in the basement membrane gel threads and
NMR spectra were obtained after the gel was observed to be
confluent. Quantitative spectra of this cell line repeatedly dem
onstrated two prominent PME peaks at 3.0 and 2.5 ppm. The
peak at 3.0 ppm could not be observed if ethanolamine was
removed from the media. The peak at 2.5 ppm was not seen if
500 MMHC-3 was added to the medium 24 h before observation.
These results indicate the 2.5 ppm peak to be PC produced by
choline kinase and the 3.0 ppm peak to be PE produced by
ethanolamine kinase.
Effects of Ara-C. When 400 MMAra-C was added to MOLT4 cells in suspension culture lysis of 70% of the cells was
observed 8-10 h later as determined by trypan blue exclusion
test. Fig. 6 shows spectra obtained using the rapid scanning
parameters on MOLT-4 lymphoma cells perfused with 400 MM
Ara-C present in the media. The cells were stable for 42 h while
being perfused with IMEM (normal choline) plus 10 MMeth
anolamine and 5% fetal calf serum with only a 10% variation
observed in the peak heights. However, after addition of 400
MMAra-C to the perfusion media there was a decrease of PE
plus PC peak height by 40-50% while the ATP, P¡,and DPDE
peaks remained stable. After 8-10 h there was a collapse of the
ATP, Pi(i„„,
and DPDE peaks due to cell lysis. By contrast the
identical experiments performed twice on MDA-MB-231 cells
which do not form Ara-CTP showed no changes in the spectra
PATHWAY REGULATION
Before ARA-C
Stable for 42 h.
Fig. 6. "P spectra of MOLT-4 lymphoma cells in basement membrane gel
perfused with IMEM. 15 ^M eholine. IO MMethanolamine, and 5rr FCS. The
initial decrease in the PME peaks after 400 MMAra-C infusion and before the
lysis and loss of other intracellular compounds is clearly seen. Spectra were
acquired with a 2-s repetition time and a 66' flip angle.
a.
after 10 h of perfusion with 400 MMAra-C (Fig. 7).
A graph of the changes in intensity of the PE peak and /3ATP peak in MOLT-4 cells for two separate experiments is
shown in Fig. 8. Time 0 represents the addition of Ara-C to the
perfusate. The ATP and PE peaks were stable for at least 12 h
preceding the addition of Ara-C. The curve for /3-ATP peak
intensity showed no statistically significant change for the first
8 h after Ara-C was added. However, the curve of the PE peak
heights showed a significant change in slope (P much less than
0.01) after the addition of Ara-C. After 8 h the ATP peaks
rapidly collapsed due to cell lysis (not shown on graph). These
effects are dose dependent and when 2 MMof Ara-C or 2 MM
dideoxycytidine are added to the perfusion media of the MOLT4 cells no changes in the spectrum were observed during a 24h observation.
DISCUSSION
•¿tÃu
N
O
TIME (hr)
Fig. 5. Relative changes of peak height ratio of GPC//J-ATP after addition of
2 mM choline (+), 2 mM ethanolamine (A), or 100 nM HC-3 (O) to the perfusate
(buffer A plus 11 mM glucose) of MDA-MB-231 cells suspended in agarose.
Absolute millimotar changes are discussed in the text. The signal to noise was 20
for tf-ATP implying a measurement error of 0.05 on the y axis.
The main results of these experiments are the direct and
noninvasive observations of the accumulation of substrate for
the rate-limiting enzymes of the de novo phosphatidylcholine
and phosphatidylethanolamine
pathways; the observation of
flux through the choline-phosphate cytidylyltransferase enzyme
in its undamaged in situ intracellular environment, and the
effects of Ara-C on these phospholipid pathways in a lymphoma
cell line.
Rate-limiting Steps in Synthesis. Our data indicate that in
both the choline and the ethanolamine pathways in the MDAMB-231 cell line the cytidylyltransferase enzymes are rate
limiting, since increased concentrations of choline and ethanol
amine caused accumulation of their substrates PC and PE but
555
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1990 American Association for Cancer Research.
CYTIDINE PHOSPHOLIPID PATHWAY REGULATION
phosphotransferase enzymes their steady state concentration of
less than 100 ^m stays below the level of detection by "P-NMR
even when their rate of formation has increased (13. 14, 16).
Flux through the Cytidylyltransferase Enzyme. The complete
inhibition of choline kinase by ethanolamine allowed us to
isolate and observe the rate of utilization of PC as a substrate
by the rate-limiting Cytidylyltransferase enzyme. At high PC
concentrations the rate of PC utilization was rapid and this
slowly decreased as the PC concentration became low (Fig. 3).
Since small changes in PC will not significantly change the
rate, this may explain why 4-fold changes or greater in PC and
PE pool sizes are seen in tumors when compared to the normal
tissue of origin (1-7, 17). Overall, our results are consistent
with a mechanism where the Cytidylyltransferase enzymes are
rate limiting, but the kinase enzymes effect the rate by control
ling the concentration of PC and PE.
MOLT-4 Cells and Effects of Ara-C. Ara-C had no effect on
MDA-MB-231 breast cancer cells but caused a rapid reduction
of the steady state levels of PC and PE in the MOLT-4
lymphoma cell line prior to cell lysis. This is due either to
inhibition of the kinase enzymes or acceleration of the Cytidy
lyltransferase enzymes. Although there is no obvious mecha
nism for the active metabolite Ara-CTP to inhibit the kinase
enzymes, it can be a cofactor for the Cytidylyltransferase en
zymes (24-25). CTP is the key regulatory cofactor for the ratelimiting cytidylyltransferases. Ara-CTP differs from CTP only
in the configuration of the hydroxyl group at the 2' position of
PEth
ß-ATP
Fig. 7. 3IP spectra of MDA-MB-231 cells in basement membrane gel perfused
with IMEM, 15 JIMcholine, 10 H.Mcthanolaminc. and i^i FCS. The addition of
400 I¿MAra-C did not affect the spectra. Spectra were acquired with a 2-s
repetition time and a 66°flip angle.
Effect
of AraC
1.2
1.0-
0.8-
0.6-
0.*
-12
-4
TIME (h)
Fig. 8. Relative changes in peak height (intensity) of the PE and /i-ATP
resonances in MOLT-4 cells after addition of high dose (h.d.) Ara-C at time 0.
Open symbols represent ji-ATP and darkened symbols represent PE. Two separate
experiments are shown. The triangles represent the first experiment and the
squares represent the second experiment. The stability of the ATP peaks for 8-h
post-Ara-C contrast with the rapid decrease in the PE peak. Experimental
conditions are the same as Fig. 6. Dashed line, 8-point moving average of the 0~
ATP peak intensities; solid line, 8-point moving average of the PE peak intensities.
not equivalent amounts of their products CDP-choline and
CDP-ethanolamine. The lack of formation of significant con
centrations of CDP-choline and CDP-ethanolamine following
high concentrations of choline and ethanolamine is in agree
ment with a previous study on hepatocytes (14). Since CDPcholine and CDP-ethanolamine are rapidly metabolized by the
the sugar moiety being in a ß
rather than <*configuration. This
2' hydroxyl group appears unimportant in the formation of
CDP-choline or the rate of flux through these pathways since
deoxy-CTP has been shown to be as effective a cofactor in these
pathways in concentrations equal to that of CTP (16). Studies
on leukemia cells from patients treated with high dose Ara-C
show the intracellular CTP concentration to be 300 A/Mand the
intracellular Ara-CTP concentration to be 400 ^M (26). In
addition increased Ara-CDP-choline and Ara-CDP-ethanolamine were formed in leukemia and ovarian carcinoma cells
although at concentrations below the sensitivity of "P-NMR
to detect (24, 25). A sudden doubling of key cofactors capable
of accelerating the rate-limiting enzymes would therefore be
present in high dose Ara-C therapy but not in low dose (<1
MM)Ara-C therapy.
The short 8-10-h time period before lysis of the MOLT-4
cells due to 400 ¿IM
Ara-C is consistent with membrane destabilization rather than an effect on DNA and could be explained
by excess synthesis of phospholipids. Excess phospholipids
(10-100 MM)are lytic to cells (27-29). In the spectra shown
(Fig. 6) at least 500 MMconcentrations of excess phospholipids
are being produced. Oleic acid also accelerates production of
phosphatidylcholine 5-fold in HeLa cells without a counterbal
ancing degradation and causes lysis of 60% of the cells within
9 h (30). By contrast phorbol esters cause a balanced increase
in both synthesis and degradation without excess production
and no lysis is observed (31).
A major problem in the study of phospholipid pathways has
involved methodology (32). Although the three-step de novo
pathways appear simple they present unique problems in that
many of the enzymes are intimately membrane bound yet act
on both water soluble and insoluble substrates, their end prod
uct is the membrane, and they are involved in the growth of the
cell. Invasive techniques damage the membrane and destroy the
normal intracellular environment so that measurements made
on isolated enzymes are potentially artifactual. These experi-
556
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1990 American Association for Cancer Research.
CYTIDINE PHOSPHOLIPID PATHWAY REGULATION
ments demonstrate the usefulness of "P-NMR spectroscopy in
noninvasively studying phospholipid metabolism.
REFERENCES
1. Degani. H.. Horwitz, A., and Itzchak. \. Breast tumors: evaluation with P31 MR spectroscopy. Radiology. 161: 53-55. 1986.
2. Oberhaensli. R., Hilton-Jones. D.. Bore. P.. Hands. L., Rampling. R., and
Radda, G. Biochemical investigation of human tumours in vivo with phosphorus-31 magnetic resonance spectroscopy. Lancet, 2: 8-11. 1986.
3. Onodera. K., Okubo. A.. Yasumoto, K., Suzuki. T., Kimura, G.. and Nomoto.
K. 31P nuclear magnetic resonance analysis of lung cancer: the perchloric
acid extract spectrum. Gann. 77: 1201-1206. 1986.
4. Ross. B.. Helsper. J. T.. Cox, J., Young, 1. R., Kempf. R.. Makepeace, A.,
and Pennock, J. Osteosarcoma and other neoplasms of bone, magnetic
resonance spectroscopy to monitor therapy. Arch. Surg., 722: 1464-1469,
1987.
5. Griffiths. J. R.. Cady. E.. Edwards. R. H.. McCready. V. R.. Wilkie. D. R..
and YViltshaw, E. 3IP-NMR studies of a human tumor in situ. Lancet. /:
1435-1436,
1983.
6. Maris, J. M., Evans, A. E., McLaughlin. A. C., D'Angio, G. J., Bolinger. L..
Manos, H., and Chance. B. 3IP nuclear magnetic resonance spectroscopic
investigation of human neuroblastoma in situ. N. Eng. J. Med. 312: 15001505, 1985.
7. Ng. T., Majors, A., and Meany, T. In vivo MR spectroscopy of human
subjects with a 1.4-T whole-body MR imager. Radiology. 158: 517-520,
1986.
8. Navon, G., Navon, R., Shulman. R., and Yamane, T. Phosphate metabolites
in lymphoid. Friend erythroleukemia. and HeLa cells observed by highresolution 3I P nuclear magnetic resonance. Proc. Nati. Acad. Sci. USA, 75:
891-895. 1978.
9. Pettegrew. J. \V., Glonek. T.. Baskin. F.. and Rosenberg, R. N. Phosphorus31 NMR of neuroblastoma clonal lines: effect of cell confluency state and
dibutyryl cyclic AMP. Neurochem. Res., 4: 795-802. 1979.
10. Lyon. R. C.. Tschudin. R. G.. Daly. P. F., and Cohen, J. S. A versatile
multinuclear probe designed for in vivo NMR spectroscopy: applications to
subcutaneous human tumors in mice. Magn. Reson. Med.. 6: 1-14. 1988.
11. Glonek. T., Kopp. S. J.. Kot. E.. Pettegrew, J. W.. Harrison. W. H., and
Cohen, M. M. P-31 nuclear magnetic resonance analysis of brain: the
perchloric acid extract spectrum. J. Neurochem.. 39: 1210-1219, 1982.
12. Daly. P. F., Lyon, R. C.. Faustino. P. J., and Cohen, J. S. Phospholipid
metabolism in cancer cells monitored by 3IP NMR spectroscopy. J. Biol.
Chem., 262: 14875-14878, 1987.
13. Pelech, S. L.. and Vance, D. E. Regulation of phosphatidylcholine biosyn
thesis. Biochim. Biophys. Acta, 779: 217-251, 1984.
14. Sundler, R., and Akesson. B. Regulation of phospholipid biosynthesis in
isolated rat hepatocytes. J. Biol. Chem.. 251: 3359-3367, 1975.
15. Warden. C. H., and Friedkin, M. Regulation of choline kinase activity and
phosphatidylcholine biosynthesis by mitogenic growth factors in 3T3 fibroblasts. J. Biol. Chem., 260: 6006-6011. 1985.
16. Ansell, G. B.. and Spanner. S. Phosphatidylserine.
phosphatidylethanolamine. and phosphatidylcholine.
In: G. B. Ansell and S. Spanner (eds.).
Phospholipids. pp. 1-49. New York: Elsevier Biomedicai Press, 1982.
17. Daly, P. F.. and Cohen. J. C. Magnetic resonance spectroscopy of tumors
and potential in vivo clinical applications: a review. Cancer Res.. 49: 770779, 1989.
18. Ko. K. W., Cook, H. W.. and Vance. D. E. Reduction of phosphatidylcholine
turnover in a Nb 2 lymphoma cell line after prolactin treatment. J. Biol.
Chem.. 261: 7846-7852. 1986.
19. Vance, D. E.. Trip, E. M.. and Paddon, H. B. Poliovirus increases phospha
tidylcholine biosynthesis in HeLa cells by stimulation of the rate-limiting
reaction catalyzed by CTP: phosphocholine cytidylyltransferase. J. Biol.
Chem.. 255: 1064-1069, 1980.
20. Cohen, J. S., Lyon, R. C., Chen, C., Faustino. P. J., Batist, G., Shoemaker,
M.. Rubalcaba. E., and Cowan. K. H. Differences in phosphate metabolite
levels in drug-sensitive and -resistant human breast cancer cell lines deter
mined by "P magnetic resonance. Cancer Res.. 46: 4087-4090. 1986.
21. Daly, P. F., Lyon. R. C., Straka. E. J.. and Cohen. J. S. 31P NMR spectros
copy of cancer cells proliferating in a basement membrane gel. FASEB J.. 2:
2596-2604. 1988.
22. Corbett, R. J.. Nunnally. R. L.. Giovanella. B. C.. and Antich. P. P. Char
acterization of the "P nuclear magnetic resonance spectrum from human
melanoma tumors implanted in nude mice. Cancer Res.. 47: 5065-5069.
1987.
23. Desmoulin. F.. Galons, J. P., Canioni. P.. Marvaldi, J.. and Cozzone, P. J.
31P nuclear magnetic resonance study of a human colon adenocarcinoma
cultured cell line. Cancer Res.. 46: 3768-3774, 1986.
24. Lauzon. G. J., Paran, J. H.. and Paterson. A. R. P. Formation of \-ß-oarabinofuranosylcytosine diphosphate choline in cultured human leukcmic
RPMI 6410 cells. Cancer Res., ¿S:1723-1729. 1978.
25. Jamieson, G. P., Snook, M. B., Bradley, T. R.. Bertoncello, I., and Wiley, J.
S. Transport and metabolism of 1-0-D-arabinofuranosylcytosine in human
ovarian adenocarcinoma cells. Cancer Res.. 49: 309-313. 1989.
26. Avramis. V. I., Biener, R.. Krailo. M.. Finkelstein, J., Ettinger, L.. Willoughby. M.. Siegel. S. E.. and Holcenberg. J. S. Biochemical pharmacology
of high dose l-J-D-arabinofuranosylcytosine in childhood acute leukemia.
Cancer Res., 47: 6786-6792. 1987.
27. Tanaka, Y., Mashino.
erythrocyte hemolysis
sophosphatidylcholine.
K., Inoue. K.. and Nojima. S. Mechanism of human
induced by short-chain phosphatidylcholines
and lyJ. Biochem.. 94: 833-840. 1983.
28. Reman, F. C.. Demel, R. A.. DeGier. J.. VanDeenen. L. L. M., Eibl, H.. and
Westphal. O. Studies on the lysis of red cells and bimolccular lipid leaflets
by synthetic lysolecithins. lecithins, and structural analogs. Chem. Phys.
Lipids. .5:221-233, 1969.
29. Tanaka. Y., Inoue. K., and Nojima. S. Interaction of dilauroylglycerophosphocholine with erythrocytes: prehemolytic events and hemolysis. Biochem.
Biophys. Acta, 600: 126-139, 1983.
30. Pelech, S. L., Cook. H. W.. Paddon, H. B., and Vance, D. E. Membranebound CTP:phosphocholine cytidylyltransferase regulates the rate of phos
phatidylcholine synthesis in HeLa cells treated with unsaturaled fatly acids.
Biochim. Biophys. Acta, 795:433-440, 1984.
31. Pelech. S. L.. Paddon. H. B.. and Vance. D. E. Phorbol esters stimulate
phosphatidylcholine biosynthesis by translocation of CTP:phosphocholine
cytidylyltransferase from cytosol to microsomes. Biochim. Biophys. Acta,
795:447-451, 1984.
32. Esko, J. D.. and Ractz. C. R. Synthesis of phospholipids in animals cells. In:
P. D. Boyer (ed.). The Enzymes. Vol. 16. pp. 207-253. New York: Academic
Press, 1983.
557
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1990 American Association for Cancer Research.
Regulation of the Cytidine Phospholipid Pathways in Human
Cancer Cells and Effects of 1- β-d-Arabinofuranosylcytosine: A
Noninvasive 31P Nuclear Magnetic Resonance Study
Peter F. Daly, Gerhard Zugmaier, David Sandler, et al.
Cancer Res 1990;50:552-557.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/50/3/552
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1990 American Association for Cancer Research.