[CANCER RESEARCH 42, 1761-1768,
0008-5472/82/0042-OOOOS02.00
May 1982]
Effect of Magnesium Content on Density-dependent Regulation of the
Onset of DMA Synthesis in Transformed 3T3 Cells1
H. Rubin
Department of Molecular Biology. University of California, Berkeley. California 94 720
tration of Mg2+ in the medium is reduced to less than Veothe
ABSTRACT
A spontaneously transformed clone of BALB/c 3T3 cells
became more transformed after more than 90 passages as
indicated by increased rounding of cells, multiplication to a
higher saturation density, and increased ability to form colonies
when suspended in agar. When the extracellular concentration
of Mg2+ was sharply reduced, the highly transformed cells
flattened, assumed the shape of nontransformed cells, and
became regularly arranged in cohesive arrays. If crowded when
deprived of Mg2*, they lost more intracellular Mg2+ than did
nontransformed and early passage-transformed
cells and re
mained at constant cell density for at least 10 days. The
intracellular content of neither Na+ nor K* changed con
sistently with Mg2+ deprivation, but the Ca2+ content increased
more than 2-fold. The sensitivity of the onset of DNA synthesis
to inhibition by Mg2+ deprivation increased with the extent of
crow( 'ng of the cultures. This was demonstrated by varying
population density within a single culture dish as well as from
culture to culture. The loss of intracellular Mg2+ in low concen
trations of extracellular Mg2+ increased with cell crowding as
did the inhibition of DNA synthesis per fractional loss of intra
cellular Mg2+. Neither deprivation of K+ or Ca2+ nor addition of
cyclic adenosine 3':5'-monophosphate
produced a densitydependent inhibition of DNA synthesis.
The results indicate that a reduction of the Mg2+ content of
highly transformed cells restores density-dependent inhibition
of the onset of DNA synthesis, which is a characteristic property
of nontransformed cells. The differences in Mg2* retentiveness
with population density may reflect differences of intracellular
distribution and binding of Mg+, which could in turn explain
some of the regulatory effects of population density on metab
olism and growth.
INTRODUCTION
BALB/c 3T3 cells undergo spontaneous transformation
when repeatedly transferred in culture over a prolonged period
(8). They change from a flat appearance and systematic sideby-side arrangement to a thinner, more elongated or rounded
appearance and a random, overlapping arrangement. Clonal
isolates of the altered cells multiply in much lower concentra
tions of serum than do clonal isolates of the flat cells, form
large colonies when suspended in agar, and are no longer
subject to a strict density-dependent inhibition of growth. The
Ca2+ content of the transformed cells is only one-half to onethird that of the nontransformed
cells (10). When the concen-
1 Supported by Research Grants CA-15744 from the National Cancer Institute
and DE-ATO3-79EV10277
from the United States Department of Energy.
Received July 27, 1981; accepted January 22, 1982.
MAY
1982
physiological level of 0.8 rnw for 1 day or longer, the trans
formed cells take on the flattened appearance and systematic
arrangement of the nontransformed cells (8). Their requirement
for serum increases markedly as does their Ca2+ content (10).
A transient period of density-dependent inhibition of the onset
of DNA synthesis was observed, but this was followed, begin
ning at about 7 days, by adaptation of the cells to the low
concentration of Mg2+, which was marked by a return to their
transformed appearance and resumption of multiplication. The
reductions in extracellular Mg2* which restore normal appear
ance and growth behavior were accompanied
of total intracellular Mg2* of less than 10%.
by a reduction
With continued passage of the transformed cells (high-pas
sage cells), they became more rounded in appearance, grew
with higher efficiency in agar, and reached higher saturation
densities. The degree of inhibition of growth by Mg2* depriva
tion became more marked than it had been with the earlier
passage cells, and there was a reduced capacity to adapt to
the lowered Mg2+ levels as evidenced by their failure to return
to their former transformed appearance. We studied the effect
of population density on the inhibition of the onset of DNA
synthesis in the high-passage cells which follows deprivation
of Mg2+. We found that the degree of inhibition of DNA synthe
sis by deprivation of extracellular Mg2+ increases sharply with
population density. The present paper is an analysis of this
effect and its significance for understanding the role of Mg2+
in growth regulation and transformation.
MATERIALS
AND METHODS
Cell Types and Culture Methods. Clone A31 of BALB/c 3T3 mouse
embryo cells was recloned by J. Bartholomew to obtain a uniform
culture of flat cells. This clone was obtained in January 1979 and
maintained by weekly transfers at low population density first in Duibecco s modification of Eagle's medium with 10% calf serum and later
in MCDB 402 (14) with calf serum. From 2 to 5 x 10* cells were
seeded
per 100-mm
plastic
Petri dish and transferred
before
the
culture was fully confluent. After 4 months of transferring, small areas
of slender, elongated, or somewhat rounded cells appeared, and such
areas became more numerous with successive transfers. When seeded
at very low density, most of the colonies were of the flat, systematically
arranged, parental type, but about 2% consisted of thin, randomly
arranged cells (8). The latter colonies were considered transformed,
and one of them (clone 14) was the source of the cells used in the
present experiments. It was picked on July 5, 1979, and weekly
transfers were begun, seeding 1 x 105 cells/100-mm
dish. These
were continued for a total of 65 transfers until September 30, 1980,
when more rapid growth of the cells necessitated a switch to a twiceweekly schedule of transfers. The increase in cell number averaged
about 50-fold in the weekly transfers and was probably limited in part
by medium depletion. On the twice weekly schedule, the cell number
increased 20- to 30-fold after 3 days and 50- to 100-fold after 4 days,
for a doubling
time of less than 1 day. Cells used in the present
1761
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1982 American Association for Cancer Research.
H. Rubin
experiments were from passages 92 to 122.
In experiments requiring measurement of the cation content of cells,
100-mm plastic culture dishes were used, with a measured surface
area of 56 sq cm. In other experiments, 60- and 35-mm dishes were
used, with surface areas of 21 and 9.1 sq cm, respectively.
Cells were transferred by washing with a Tris-buffered 0.9% NaCI
solution without divalent cations and detached from the dish with
0.01 % crystalline trypsin containing 0.5 mM EDTA in the buffered 0.9%
NaCI solution. The seeding densities are noted in the legends to the
charts and tables. Calf serum which had been exhaustively dialyzed
against 0.15 M NaCI was used in all the ion variation experiments. All
incubations were at 37°in an atmosphere of about 5% CO2 in humid
ified air.
Experimental Procedure. Cells were seeded at the densities noted
in the figure and table legends in MCDB 402 with 5 or 10% calf serum.
They were incubated for the indicated time and then washed twice with
a Tris-buffered 0.9% NaCI solution free of divalent cations, and the
medium was replaced with one containing the appropriate Mg2* con
centration plus 10% dialyzed calf serum. The complete medium to
which no Mg2+ had been deliberately added contained between 0.005
and 0.01 mM Mg2*. It also contained 130 mM Na*, 4 HIM K +, and 1.6
mM Ca2* except in the experiment in which the concentrations of K*
and Ca24 were varied. In this case, the background concentration of
K* was 0.2 mM and that of Ca2+ was 0.02 mM. The medium was
always changed 17 hr before measurements of [3H]thymidine incor
poration and intracellular cation content to minimize effects of medium
depletion on the results.
Cell Counting, Labeling, and Protein Determinations. Cells were
counted in a Model F Coulter Counter. The rate of DNA synthesis was
assessed by labeling the cells for 60 min with 1 fiCi [3H]thymidine per
ml (20 Ci/mmol, 1 Ci = 3.7 x 10'° y-rays) in fresh MCDB 402
containing the usual concentrations of all ions. They were then pre
pared for either scintillation counting and protein determinations or
autoradiography by methods already described (3, 11). A comparison
was made of the scintillation counting and autoradiographic methods
for measuring relative rates of DNA synthesis in cultures incubated for
2 days in varying concentrations of Mg2*. The results in Table 1 show
excellent agreement between the methods in the range of Mg2* con
centrations in the medium from 0.026 to 0.80 mM. In the lowest Mg2*
concentration of 0.016 mM, the scintillation counts showed less of a
reduction than did the counts of labeled nuclei by autoradiography.
However, it was noted that the extreme periphery of the cultures in
0.016 mM Mg2* had regions with sparse cell populations and that
as measured by scintillation counting of cell extracts represents a
reduction in the number of cells in the S period of DNA synthesis rather
than a reduction in the rate of DNA chain elongation. If the latter were
the case, there would be less of a reduction in the number of labeled
nuclei than in scintillation counts, since even lightly labeled nuclei, if
they occurred in low Mg2+, could be counted as well as the very heavily
labeled nuclei of the control. It is also noteworthy that the inhibitory
effects are fully reversible when Mg2* is restored to the cultures (15),
indicating no damage has been done.
Measurement of Cations by Atomic Absorption Spectrophotometry. The methods used for measuring the cation content of cells and
medium are minor modifications of a procedure which has been de
scribed and validated in detail (11, 13). Briefly, each culture was
washed 4 times with 9 ml of a CO2-free, 0.25 M sucrose solution
(approximately pH 7) per wash. It was then left for 5 sec in 5 ml of a
carbonated 0.25 M sucrose solution (pH 4) and drained of fluid for 5
sec. Protons of the carbonated solution rapidly exchange with essen
tially all the cations at the surface without removing any protein from
the surface (13). The washed cells were harvested by carefully scraping
each dish with a well-rinsed rubber policeman and suspending the
cells in distilled, deionized water. When all 4 major cations were to be
measured, a minimum of 6 culture dishes was used to provide enough
material for accurate analysis. When only Mg2* was to be measured,
one or 2 dishes were sufficient. The cell suspensions were sonicated
(Bronwill Biosonik IV), and aliquots were taken for measurement of
protein by the method of Lowry et al. (4), using serum albumin (Sigma
Chemical Co.) as standard, and of the relevant cations by atomic
absorption spectrophotometry.
All samples and standards for atomic
absorption spectrophotometry
contained 15 mM La3*, 4 mM Cs*, and
100 mM HCI to minimize chemical and ionization interferences. Dupli
cates of each sample were each read twice against suitable standards
using curvature correction and the 100-average mode of the PerkinElmer Model 403 atomic absorption spectrophotometer.
The concen
tration of cations in the medium was also determined in each experi
ment.
The reproducibility of the atomic absorption measurements for Mg2*
was determined by working up separately 7 sets of 2 cultures, each of
which had been incubated for 2 days in either a physiological or a very
low concentration of Mg2*. The results in Table 2 show excellent
reproducibility of the determinations for Mg2* with an S.D. of 2.9% for
cells incubated in 0.818 mM Mg2* and 3.8% for cells in 0.018 mM
Mg2*. The 45% reduction of intracellular Mg2* reduced the rate of
[3H]thymidine incorporation 230-fold. Multiple determinations made
these regions had a much higher proportion of labeled cells than did
the confluent regions toward the center of the dish. Since the periphery
of the dish was not included in the counting of labeled nuclei but was
included in the NaOH extraction for scintillation counting, this may
account for the disproportionate reduction in DNA synthesis as mea
sured by autoradiography of cells in 0.016 mM Mg2*. The results affirm
that the reduction in [3H]thymidine incorporation by Mg2* deprivation
with the other 3 major cations indicate a similar degree of reproduci
bility in their measurement (13). In the case of Na*, there is a problem
Table 1
Scintillation counting versus autoradiography in measuring the reduction of
l^HJthymidine incorporation by Mg2* deprivation
Effect of Magnesium Deprivation on Growth and Cation
Concentration. Cultures of clone 14 cells at passage 92 were
grown to a high density, and one group of the cultures was
continued in a medium with a physiological concentration of
Mg2+ (1.0 HIM), while another group was switched to a medium
with a low concentration (0.013 mM) of Mg2+. Although
crowded, the cultures in 1.0 mM Mg2+ continued to grow during
Cells from passage 114 were seeded at 2 x 103 cells/sq cm and incubated
for 3 days. The medium was then replaced with media of various Mg2* concen
trations, and the cultures were incubated another 2 days when another medium
change with the same Mg2' concentrations was made. Seventeen hr later, the
cells were labeled with [3H]thymidine for 1 hr and prepared for scintillation
counting or for autoradiography.
countingExtracellular
(mM)0.8
Mg2*
(control)
0.040
0.026
0.016cpm//ig
1762
Scintillation
protein108.20ofcontrol1.00.44
of la
beled nu
of
clei26.1511.29control1.0
47.77
0.43
2.10
0.019
045
0.017
1.24Fraction0.011Autoradiography%
<0.01Fraction<0.001
with contamination of the cells from small amounts of medium and
other sources, but proper precautions were observed.
RESULTS
the 10-day course of the experimental measurements, while
those in 0.013 mM Mg2+ did not grow at all (Chart 1a). The
rate of [3H]thymidine incorporation in cultures in 1.0 mM Mg2+
decreased slightly over the 10-day period, although the me
dium was always changed 17 hr before incorporation was
measured (Chart 1b). The cultures in 0.013 mM Mg2+ showed
only a slight decrease at 1 day in the rate of [3H]thymidine
incorporation,
but they dropped to a level about 1/500 that of
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1982 American Association for Cancer Research.
VOL. 42
Density-dependent
Regulation Effects of Low Mg2
Table 2
A statistical evaluation of the accuracy of the method for determining Mg2' concentrations in cells
incubated in media with physiological and very low concentrations of Mg2'
Thirty-two cultures prepared with a seeding of 1.8 x 103 cells (passage 111) per sq cm were incubated
for 3 days, and 16 of the cultures were switched to media with physiological and 16 to media with very low
concentrations of Mg2*. Two days later, the media were replaced with media of the same Mg2* concentra
tions and incubated for another 17 hr. Two cultures in each group were labeled with [3H]thymidine and
prepared for scintillation counting. The remaining cultures in each group were divided into 7 groups of 2
each, and each pair was harvested separately and prepared for atomic absorption spectrophotometry.
Mg2*0.818
Extracellular
rtiMIntracellular
Mg2'
protein)0.0779
(¿imol/mg
mMIntracellular
Protein
midine
(jig/sq
(cpm/fig
protein)80.4
cm)
0.275"
0.0772
0.0837
0.0802
0.0823
0.0818
0.0806
0.0805 ±0.0023a[3H]Thy-
[3HJThymiMg2'
(/imol/mg
protein)0.0461
(/ig/sq
dine (cpm/
protein)57.2
cm)
fig
63.10.018
0.0466
0.0455
0.0437
0.0444
0.0417
0.0438
0.0445 ±0.0017Protein
Average ±S.D.
the control cultures by 3 days. In the following days, the Mg2*deprived cultures increased slightly in the rate of [3H]thymidine
incorporation, but they never exceeded 1/100 the rate of that
in the controls on the same day. The Mg2*-deprived cells were
more flattened in appearance than were the control cultures,
but the degree of flattening was limited by the high degree of
crowding.
The concentrations of the 4 major cations in the cells are
shown in Chart 1c. The concentration of Mg2* in the deprived
cultures was reduced relative to the control cultures by about
30% on Day 1, and the relative reduction later ranged between
15 and 20%. Intracellular Na+ was increased by Mg2* depri
vation on Day 1 but later was reduced to about the same level
as the control. Intracellular K* was decreased in the Mg2*deprived cultures relative to the control on Day 1 but remained
relatively constant in the following days when the control levels
gradually decreased to a point below the deprived cultures.
Intracellular Ca2* was almost doubled by Mg2+ deprivation on
Day 1, and this relative increase grew steadily larger during
the course of the experiment to a point more than 3 times as
high as the control value at 10 days.
The results showed that the maximum inhibitory effect of
Mg2* deprivation on [3H]thymidine incorporation
was ex
pressed by 3 days. The concentrations of Na+ and K+ were at
various times higher and lower in the Mg?*-deficient than the
Mg2*-sufficient cultures, showing no correlation with DNA syn
thesis. Mg2* deprivation caused a rise in the Ca2* concentra
tions of the transformed cells to a level characteristic of nontransformed cells (10). Raising the Ca2* content of the cells to
the same level by sharply raising the Ca2* concentration of the
medium (in low concentrations of P¡to avoid precipitate for
mation) did not cause a change in their appearance (16). Thus,
there is no indication that the effects of Mg2* deprivation on
the appearance and growth behavior of the cells are exerted
indirectly by changing the cellular content of the other 3 major
cations. The reduction of 15 to 20% in the Mg2* content of the
cells from Days 3 to 10 through its concerted effect on cellular
metabolism (2) could be responsible for the altered appearance
of the cells, the complete inhibition of protein accumulation,
MAY
1982
and the profound inhibition of DNA synthesis.
Effect of Population Density on the Incorporation of
[3H]Thymidine in Cultures Maintained in Various Concentra
tions of Mg2*. Cells at passage 107 were seeded at different
densities and shifted to various Mg2* concentrations at 1 day,
and the rates of [3H]thymidine incorporation were determined
3 days and, in some cases, 6 days later. These rates are
plotted as a function of the cell densities (expressed as protein
per sq cm) at the time of labeling in Chart 2. At 3 days, the
cells in the physiological concentration of 0.8 mM Mg2* had
high rates of incorporation of [3H]thymidine at densities of less
than 30 /ig protein per sq cm and had a 3-fold-lower rate at
higher cell densities. At 6 days, when all the cultures in this
concentration of Mg2* regardless of seeding density had
reached the same high density, the rate of [3H]thymidine incor
poration was at the 3-fold-reduced level. In cultures maintained
in low concentrations of Mg2* for 3 days, there was a drastic
reduction of [3H]thymidine incorporation with increasing cell
density, although there was no such reduction at equivalent
densities in cultures maintained in 0.8 mM Mg2*. A family of
curves was generated, in which each successive reduction in
extracellular Mg2* caused density-dependent inhibition of [3H]
thymidine to become manifest at a lower cell density. When
the cells were cultured in 28 ¡UM
Mg2* for 6 days rather than 3
days, higher cell densities were required to inhibit the rate of
[3H]thymidine incorporation, suggesting that there was some
adaptation by the cells to Mg2* deprivation. The steepness of
the curve relating cell density inversely to [3H]thymidine incor
poration was just as great at 6 as at 3 days.
Effects of Population Density on Cellular Mg2* Content
and [3H]Thymidine Incorporation at Various Extracellular
Mg2* Concentrations. A comparison was made of the relation
ship between intracellular Mg2* content and [3H]thymidine
incorporation in cultures from passage 114 at different popu
lation densities and extracellular Mg2* concentrations. The rate
of [3H]thymidine incorporation was about 5 times higher in the
less crowded than the more crowded cultures in 0.8 mM Mg2*
and was reduced only 3.5-fold in the less crowded cultures by
lowering the Mg2* of the medium to the lowest concentration
1763
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1982 American Association for Cancer Research.
H. Rubin
150
1000 p
e 125
800 i
100
5. 100
50
n.
o>
25
10
: (b)
S
'O* r
-3E io' -
0.1
£, 10°r
-o
IO'1
4
6
Days
^q/sq
10
cm
100
Chart 2. Effects of population density on the incorporation of [3H]thymidine in
cultures maintained in various concentrations of Mg2*. Cells from the 107th
passage were seeded on 35-mm dishes at densities of 0.55. 1.1, 2.2, and 5.5
x 103 cells/sq cm. They were incubated for 24 hr, and the medium was replaced
by media with the appropriate concentrations of Mg2'. The medium was again
changed 2 days later and labeled with [3H)thymidine (¡3H¡TdR)
after a further 17hr incubation. This procedure was repeated on Days 5 and 6 for cultures in 800
and 28fiM Mg2*. The cultures were extracted with NaOH for scintillation counting
and protein determination. The lines connect counts obtained from cultures of
different densities at 3 or 6 days incubated in the particular concentration of
Mg2* (JIM) indicated at the fop of each curve. The values on the abscissa
0.10
represent the protein concentrations at the time of labeling the cultures. The
various symbols represent the initial seeding densities (cells/sq cm) of the
cultures. V and T, 0.55 x 103;DandB,
1.1 x 103; A and A. 2.2 x 103;Oand
•,5.5 x 103. The open symbols are values obtained at 3 days, and the filled
0.08
«-o
0 £0.06
symbols are values obtained at 6 days.
o E
ui
0.04
reducing extracellular
10 O
5
10
Days
Chart 1. Effects of Mg2* deprivation on growth rate (a), [3H]thymidine
(/3H/7"dfl) incorporation (b), and cation content (c) of transformed BALB/c 3T3
cells. Cells from the 92nd passage were seeded on 100-mm plastic Retri dishes
at a density of 1.8 x 103 cells/sq cm and incubated for 4 days. One half of the
cultures was then switched to fresh medium with 10% dialyzed serum containing
0.91 rriM Mg2*, and the other half was switched to medium with 0.014 mM Mg2*.
At each of the times indicated by symbols, 2 cultures in each group were labeled
with [3H]thymidine and harvested for scintillation counting and protein determi
nations. At the same time. 6 of the cultures were washed, scraped from the dish,
and prepared for determination of the 4 major cations by atomic absorption
spectrophotometry
and protein content by the method of Lowry ef al. (4).
Seventeen hr prior to these determinations, all the cultures of the experiment
had a medium change with the appropriate Mg2* concentration. In Chart 1, a and
b, each open symbol represents one culture which had been labeled with [3HJthymidine and dissolved in NaOH. In Chart 1a, the filled symbols represent a
pool of 6 cultures prepared for atomic absorption spectrophotometry as do all
the open symbols in Chart 1c. Mg2* concentrations in the medium: O, •,0.91
mM; A, A, 0.014 mM.
used (Chart 3, a and b). Thus, the less crowded cultures
incorporated [3H]thymidine at a higher rate even in Mg2*-deficient medium than did the more crowded cultures in Mg2*sufficient medium. In the more crowded cultures, there was
only a slight decrease in [3H]thymidine incorporation when the
extracellular Mg2* was reduced to 0.036 mM, but any further
reduction of extracellular Mg2* resulted in a sharp decrease in
[3H]thymidine incorporation. In the lowest concentration of
Mg2+, the more crowded cultures incorporated [3H]thymidine
at a rate 300 times lower than the control for the group and
1500 times lower than the control for the less crowded cultures.
The intracellular concentration of Mg2* in the less crowded
cultures was lowered to about 75% of the control value by
1764
1
Protein,
Mg2* to its lowest concentration;
in the
more crowded cultures, it was lowered to 50% of the control
value (Chart 3c). A sharp drop in Mg2+ content of the more
crowded cultures occurred when extracellular Mg2* was re
duced from 0.036 to 0.026 mM, and this is where the sharp
decrease in DMA synthesis also occurred. Several significant
features of the cellular response to Mg2* are brought out when
the rate of [3H]thymidine incorporation is plotted against the
intracellular content of Mg2* (Chart 3d). The data show that
the more crowded cultures are much more sensitive than are
the less crowded cultures to Mg2* loss. When the Mg2* content
of the less crowded cells was reduced to its lowest level, the
rate of [3H]thymidine incorporation was 20 times higher than
that of the more crowded cells with the same Mg?* content.
The results suggest that a higher proportion of the Mg2*
content of the less crowded cultures is available for carrying
out the functions required to initiate DMA synthesis, i.e., that
the ratio of free to bound Mg2* differs with the degree of
crowding.
Autoradiographic Studies of the Effect of Population Den
sity on the Response of Cells to Mg2* Deprivation. One of
the most striking illustrations of the effect of population density
on the rate of multiplication of cells is provided by the so-called
"wound healing" experiment (3). This allows one to estimate
relative rates of multiplication of cells at different local densities
sharing the same medium on the same dish. The wound healing
experiment itself could not be done in low concentrations of
Mg2* because cell-cell adhesion was increased by Mg2* dep
rivation and because the migration of cells into the wound was
slowed down. However, when a strip of cells was removed from
a transformed culture (passage 122) and the culture was
incubated in 0.8 mM Mg2* for a day, individual cells and groups
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1982 American Association for Cancer Research.
VOL. 42
Density-dependent
Regulation Effects of Low Mg'
Table 3
(o)
Autoradiographic
study of the effects of population density on the response of
cells to Mg" deprivation
See legend to Fig. 1 for procedure. More than 2000 nuclei were counted in
the confluent regions, and more than 1000 were in the isolated colonies.
'S.
10'
nucleiExtracellular
S
o.
J
5
L
Mg2*
(mM)0.80
_LU
o.
to'
K
V
I-
10°
0.040
0.026
0.016Confluent
I
I
-(c)
10.0
10.9
8.3Ratio
9.71
36.3
>166.0
Lack of Density Dependence in the Inhibition of DNA
Synthesis by Deprivation of K ' or Ca: ' or the Addition of
Cyclic Adenosine 3':5'-Monophosphate. Decreasing the con
centration of K* in the medium to 0.5 mM or less does not
flatten the cells (passages 109 and 110) and inhibits [3H]-
C
fi
5 - °-06
0.04
thymidine incorporation more in sparse than in dense cultures
(Table 4). Omission of Ca2* from the medium, which leaves the
I
0.01
e
1.03
0.29
<0.05Isolated
pendent inhibition of the transformed cells which is not due
merely to an artefactual depletion of the medium by high
density cultures in low Mg2* concentration.
I I I I
»- ,. 0.08
si
of labeled nu
clei in isolated colo
nies to those in con
fluent region3.47
the wound had a high proportion of labeled cells (Fig. 1c). This
shows that the Mg2* deprivation induces a true density-de
10-
irg i
re
gion7.2
col
onies25.0
10'
10«
(M
% of labeled
lililÃ0.1
Extrocellulor Mgz*, mM
1.0
io'
: (d)
io
synthesis in the sparse cultures and actually a slight stimulation
in the crowded cultures, possibly associated with cell detach
ment which reduces the population density of the cells remain
ing on the dish. /V6,O2-Dibutyryl adenosine 3':5'-monophos-
io'
10°
10-
lili
0.02
Intracellular
concentration at 0.02 mM from contamination carried by other
constituents of the medium, reduces cellular adhesion to the
substratum, thereby causing the cells to become rounder and
even more transformed in appearance than cells in normal
concentrations of Ca2*. There is a slight inhibition of DNA
0.04^
0.06
Mg2*, ^mol/mg
0.08
protein 0.1
Charts. Effects of population density on cellular Mg2* content and [3H]thymidine ([3H]TdR) incorporation with variation in extracellular Mg2*. Cells from
passage 114 were seeded at 3.6 x 103 and 0.9 x 103 cells/sq cm and incubated
for 6 and 72 hr, respectively, before being switched to media with various Mg2*
phate, which was fairly effective at flattening early passage
transformed cells, flattened only a small proportion of these
late passage cells although it inhibited their growth. It was
much more inhibitory to [3H]thymidine incorporation in sparse
than in crowded cultures. Mg2* deprivation was unique among
these treatments in inhibiting the high density cultures more
than the low density cultures.
concentrations. The former group, although seeded at a higher density, had no
chance for multiplication before incubation in media of various Mg2* concentra
tions and therefore represented the lower density population.
increased in number approximately 20-fold before switching
The latter group
to various Mg2*
concentrations and therefore represented the higher density population. After 2
days of incubation in the various Mg2* concentrations, a fresh replacement of
the same media was made, and 17 hr later, the cultures were labeled with
[3H]thymidine or harvested for atomic absorption spectrophotometry.
Depen
dence on extracellular Mg2* concentration of protein content (a) and [3H]thymidine incorporation (£>),intracellular Mg2* content (c), and intracellular Mg2'
content of [3H]thymidine incorporation (d). O, higher-density cultures; •,lowerdensity cultures.
detached from the confluent region and reattached to the dish
in the denuded strip, producing a region of low population
density. If these cultures were maintained in physiological
concentrations of Mg2* and then exposed to [3H]thymidine for
autoradiography, a relatively high proportion of cells in both
the confluent region of the culture and the wound was labeled
(Table 3; Fig. 1a). When the cultures were switched to a very
low Mg2* concentration, the confluent region contained very
few or no labeled cells (Table 3). In Fig. 1b, some of the nuclei
are darkened by the histological stain, but none of them is
autoradiographically labeled. However, the isolated colonies of
MAY 1982
DISCUSSION
During the course of repeated weekly and biweekly passages
of the spontaneously transformed clone 14 over a period of
some 20 months, progressive changes in appearance of the
cells and their growth behavior occurred. At first, the cells were
slender and somewhat retracted along their lateral edges, but
with time, the cells in sparse cultures assumed an almost
spherical shape. They grew more rapidly and to a higher
population density than they had when first isolated and formed
colonies more quickly and with higher efficiency when sus
pended in agar. During the first year of passage, only a tran
sient density-dependent
inhibition of growth could be estab
lished by decreasing the Mg2* concentration of the medium to
about 0.015 mM which caused a reduction of total intracellular
Mg2* of no more than 10% (10). Although Mg2*-deficient
medium induced a marked flattening of the cells and a change
from a random overlapping to a systematic, nonoverlapping
cell-cell arrangement, the slowdown in the onset of DNA syn
thesis relative to that in Mg2*-sufficient medium was only about
1765
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1982 American Association for Cancer Research.
H. Rubin
"
?£ -'V:.
»__ Ä
Fig. 1. Local effects of population density on [3H]thymidine incorporation
in Mg2'-deprived
cells. Cells of passage 122 were seeded at 2.4 x I03/sq
cm in 60-mm
dishes and incubated for 3 days in MCDB 402 with 10% calf serum. A strip of cells about 8 mm wide across the diameter of the dish was scraped off with a rubber
policeman, the detached cells were removed, and fresh medium was substituted. The cultures were incubated for 30 hr to allow individual cells and groups from the
confluent regions to detach and reattach to the denuded region of the swath. They were then washed, and media with various concentrations of Mg2' were added.
After 2 days of further incubation, fresh media with the appropriate Mg2* concentrations were added, and 17 hr later, the cells were labeled with [3H]thymidine and
prepared for autoradiography. a, confluent region of culture in 0.8 mM Mg2*; b, confluent region of culture in 0.016mM Mg2*; c, colony of cells in swath in 0.016 mM
Mg2*. The darkened regions seen in some nuclei of b are caused by staining with Harris hematoxylin stain and are not to be confused with the uniform black of
exposed silver grains in the emulsion over the autoradiographically
labeled cells of a and c.
Table 4
Effects of deprivation of K *, Ca2*. or Mg2* or addition of N6.O2-d/oufyry/ adenosine 3':5'-monophosphate
on ¡^Hfthymidine incorporation
into cultures of spontaneously transformed cells at different densities
Cells from passages 109 and 110 were seeded at 5.5 x 103/sq cm and incubated 8 hr (low density) or 3 days (high density) before being
switched to media containing normal medium; medium deficient in either K*. Ca2*. or Mg2*; or normal medium containing either 0.6 or 1.0 mM
/V6,O2-dibutyryl adenosine 3':5'-monophosphate.
After 2 more days, fresh medium of the same composition replaced the old medium, and 17
hr later, the cultures were labeled with [3H]thymidine and prepared for scintillation counting. The high density culture controls had 29.4 pg
protein per sq cm. and the low density controls had 9.5 jig per sq cm.
[3H]ThymidineExtracellularControlLowK*Low
density cul
tures1cpm//jg
ef
fectColumn
cultures3
density
cpm//jg
Fraction
2/
protein61.651.4
ofcontrol1.00.84 protein192.598.3
ofcontrol1.00.51 Column
41.01.65
(mM)1.61.6
0.44.04.0 1.60.021.6 0.80.80.027
7.1103.00.65 0.121.680.011
4.8154.631.30.0250.800.16
4.802.100.069
Ca2*LowMg2'Control
0.0170.6
4.0Ca2* 1.6Mg2*0.80.8
medium + cyclic adenosine
3' :5'-monophosphateK*40.7
1.0High
3-fold after 3 days of treatment and did not exceed 10-fold at
6 days. Measurements beyond 6 days were complicated by
adaptation of the cells to the Mg2+-deficient medium, where
upon they resumed their original transformed morphology and
1766
0.2337.0
0.0040.60
34.02Fraction
0.56Low
11.925.2 0.060.13
9.94
0.051Density
0.0674.62
10.99
behavior (8). It also became evident that serum concentration
and population density were important variables in determining
the response to Mg2* deprivation (10).
As the progressive changes in appearance and growth deCANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1982 American Association for Cancer Research.
VOL. 42
Density-dependent
veloped, the transformed cells became more susceptible to
inhibition by deprivation of Mg2+, so that the rate of [3H]thymidine incorporation was reduced by as much as 500-fold in
crowded cultures. That this large reduction was not an artefact
due to effects on uptake or phosphorylation of the [3H]thymidine was evident from the fact that the reduction in number of
labeled nuclei detected by autoradiography was proportional
to the reduction in incorporation of [3H]thymidine into acidinsoluble material measured in cell extracts. Although the cells
sometimes became vacuolated under the conditions of Mg2+
deprivation which profoundly inhibited the initiation of DMA
synthesis, they were fully viable as indicated by full restoration
of DMA synthesis beginning about 12 hr after addition of
physiological concentrations of Mg2* (15).
Using the late-passage cells, a clear-cut limit on population
density could be established by Mg2+ deprivation (Chart 1).
Although the cells remained fully viable, only a small minority
showed any signs of adaptation by resuming their original
transformed morphology. The increased sensitivity of the cells
to inhibition by Mg2+ deprivation and the loss of capacity to
adapt were associated with a greater loss of Mg2+ from the
Mg2+-deprived cells. After 16 months of weekly, and later
light of the demonstration
Regulation Effects of Low Mg2 +
that Mg2+ deprivation
induces a
requirement for high serum concentration to support growth of
the transformed cells (10).
Deprivation of the other major cations causes neither a
normalization of appearance of the transformed cells nor a
requirement for serum (8, 10). Here, it is shown that deprivation
of K+ or Ca2+ is more inhibitory to sparse than crowded cells,
an effect opposite to that of Mg2+ deprivation. Addition of
W6,O2-dibutyryl
adenosine
3':5'-monophosphate,
which
causes some irregular flattening of a small fraction of the
population, is much more inhibitory to sparse than to dense
populations. Retinoic acid and dimethyl sulfoxide cause a
phenotypic reversion of some transformed cells, including a
reduction of their saturation densities (1, 5). However, retinoic
acid does not revert the majority of late-passage, transformed
BALB/c 3T3 cells used here, and the effect of dimethyl sulf
oxide on the saturation density of these cells has not been
tested.
There is some indication as exemplified in Chart 3d that the
intracellular distribution of Mg2* differs in cultures with differ
ences in their population density. There, it can be seen that
relatively large reductions of the intracellular Mg2+ content of
biweekly, passaging (passage 92), this loss was 15 to 20%,
and by 18 months (passage 114), it reached 40 to 50% when
the extracellular Mg2+ level was reduced to about 0.015 mw.
the cultures with lower population density cause only a slight
reduction in the rate of [3H]thymidine incorporation, while any
reduction of Mg2+ content in the cultures of higher population
There was no correlation between the intracellular concentra
tions of monovalent cations under these conditions and the
rate of DNA synthesis in the cells. Although the intracellular
concentration of Ca2+ was raised in cells deprived of Mg2 +,
similar increases of ¡ntracellular Ca2* produced by raising the
extracellular concentration of Ca2+ failed to reproduce the
normalizing effects of Mg2+ deprivation on their appearance
density markedly inhibits that incorporation. A 25% reduction
of intracellular Mg2+ in the less crowded cultures lowers the
rate of [3H]thymidine incorporation only about 3-fold; a similar
reduction of intracellular Mg2+ in the more crowded cultures
and behavior (16). The results therefore indicate that the effects
are produced directly by altering the intracellular level of Mg?+
rather than indirectly through effects on other cations of the
cell.
The late-passage cells showed a strong dependence on
population density in their response to Mg2+ deprivation. In
physiological concentrations of Mg?+, the transformed cells
were only slightly inhibited by increasing population density,
and those densities had to be high (about 50 /ig protein per sq
cm of substratum) to exert a 3-fold inhibition in the rate of DNA
synthesis (Chart 2). As the Mg2* concentration of the medium
was reduced to lower and lower levels, the cells showed an
ever increasing sensitivity to the inhibitory effects of population
density. Thus, in 28 /IM Mg2 +, the rate of DNA synthesis was
reduced some 30-fold as the density of the culture was in
creased from 5 to 18 fig protein per sq cm of substratum. The
effect of Mg2+ deprivation in producing a density-dependent
lowers their rate about 20-fold. With this degree of intracellular
Mg2* reduction, the less crowded cultures incorporate [3H]thymidine at a rate slightly higher than that of the more crowded
cultures with a full complement of Mg2+. The results suggest
that there is an inverse relationship between population density
and the fraction of ¡ntracellular Mg2+ available for the reactions
which determine the onset of DNA synthesis.
The results raise the question of the mechanism by which
Mg2* deprivation imposes density-dependent inhibition of DNA
synthesis on transformed cells. The Mg2+-deprived cells flatten
out and appear to form tight lateral associations with each
other which resemble the associations formed between nontransformed cells in confluent cultures. These associations
may limit the degree of movement of the cells and restrict the
movement of the cell membrane. Therefore, the imposition of
density-dependent inhibition by Mg2+ deprivation may be the
consequence of the change in cell shape and cell-cell adhe
siveness which occurs in the Mg2+-deprived cells. Even if these
conjectures prove to be valid, they would merely set the search
for a mechanism one step back to the question of what Mg2*-
inhibition of DNA synthesis was most graphically seen in autoradiographs of cultures which had adjacent areas of high and
low density (Fig. 1, b and c). In such deprived cultures, the
dense regions had virtually no labeled nuclei, while the adjacent
small patches of more loosely arranged cells had relatively high
frequencies of labeled cells. With regard to the inhibitory effects
of population density therefore, the Mg?*-deprived transformed
dependent reactions control the changes in shape and adhe
siveness of the cells. There is a significant and direct effect of
Mg2+ concentration on protein synthesis (15), and this could
cells were behaving like their nontransformed counterparts (9).
These experiments show that the normalizing effects of Mg?+
its normalizing effect.
The varied effects of Mg?* concentration
deprivation are not due to some nonspecific inhibitory action
on the cells, since the cells display a subtle regulatory aspect
of normal cell behavior. This takes on added significance in
of transformed cells should be viewed alongside the hypothesis
that Mg2+ plays an important role in regulating the metabolism
and growth of normal cells (2, 6). Reduction of the Mg2*
MAY
1982
play a major role in determining the response of the cells.
However, there are many Mg2+-dependent and -regulated re
actions in cells (7), and it is possible that the concerted control
of a group of reactions is required for Mg2+ deprivation to exert
on the properties
1767
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1982 American Association for Cancer Research.
H. Rubin
content of nontransformed cells inhibits just those processes
which are inhibited by restricting the concentration of serum in
the medium or increasing the population density of the cells
(7). Stimulation of DMA synthesis by addition of serum is
preceded by an increase of the Mg2* content of the cells which
reaches a level of about 15% higher than that of the unstimulated control by the time DMA synthesis reaches its peak (12).
It is possible therefore that transformation is brought about, at
least in part, by a defect in regulating the distribution of Mg2+
in cells, which is in turn responsible for many of the phenotypic
traits of transformed cells. Such a defect could be the result of
structural or metabolic changes which increase the concentra
tion of free Mg2+ in the cell at the expense of bound reserves.
The normalizing effects of Mg2+ deprivation could then be
ascribed to a restoration of free intracellular Mg?+ to its normal
lower concentrations. If the transformation is related to a defect
in the Mg2+-regulatory system of the cell, it raises the questions
of how the availability of Mg2* is normally controlled in the cell
and, particularly, how it is altered by population density or
application of growth-stimulatory agents. We know very little at
either the molecular or structural level about how the distribu
tion of Mg2* within the cell is regulated, and we may have to
learn much more about this aspect of cell physiology before
we can understand the basis of the malignant transformation.
ACKNOWLEDGMENT
I wish to thank Berbie Chu for her excellent technical assistance in all phases
of this work.
REFERENCES
1. Dion, L. D., Blalock, J. E., and Gifford, G. E. Vitamin A-induced density
dependent inhibition of L-cell proliferation. J. Nati. Cancer Inst., 58. 795-
1768
801, 1977.
2. Garfinkel. D., Kohn, M. C., and Achs, M. J. Computer simulation of metab
olism in pyruvate perfused rat heart. V. Physiological implications. Am. J.
Physiol., 237. R181-R186, 1979.
3. Gurney. T. Local stimulation of growth in primary cultures of chick embryo
fibroblasts. Proc. Nati. Acad. Sei. U. S. A., 62: 906-911, 1969.
4 Lowry. O. H.. Rosebrough, A. J.. Farr, A. L., and Randall, R. J. Protein
measurements with the Polin phenol reagent. J. Biol. Chem., 793. 265-275,
1951.
5. Matsuhisa, T., Mori, Y., and Tamura. H. Phenotypic reversion of SV-40transformed 3T3 cells by dimethyl sulfoxide. Cell Biol. Int. Rep.. 5: 179186, 1981.
6. Rubin. H. A central role for magnesium in coordinate control of metabolism
and growth in animal cells. Proc. Nati. Acad. Sei. U. S. A., 72. 3551-3555,
1975.
7. Rubin, H. Magnesium removal reproduces the coordinate effects of serum
removal or cortisol addition on transport and metabolism in chick embryo
fibroblasts. J. Cell. Physiol., 89: 613-626, 1976.
8. Rubin. H. Growth regulation, reverse transformation, and adaptability of 3T3
cells in decreased Mg2* concentration. Proc. Nati. Acad. Sei. U. S. A., 78:
328-332. 1981.
9. Rubin, A. H., and Bowen-Pope, D. F. Coordinate control of BALB/c 3T3 cell
survival and multiplication by serum or calcium pyrophosphate complexes.
J. Cell. Physiol., 98: 81-94. 1979.
10. Rubin. H., Vidair. C., and Sanui. H. Restoration of normal appearance,
growth behavior, and calcium content to transformed 3T3 cells by magne
sium deprivation. Proc. Nati. Acad. Sei. U. S. A., 78: 2350-2354,
1981.
11. Sanui, H.. and Rubin, A. H. Membrane bound and cellular cationic changes
associated with insulin stimulation of cultured cells. J. Cell. Physiol., 96:
265-278. 1978.
12. Sanui, H., and Rubin, A. H. Membrane bound and cellular ion changes
associated with serum stimulation of DNA synthesis in mouse 3T3 cells. J.
Cell Biol., 79: 84a. 1978.
13. Sanui, H., and Rubin, A. H. Measurement of total, intracellular, and surface
bound cations in animal cells grown in culture. J. Cell. Physiol., !00: 215226, 1979.
14. Shipley, G. D., and Ham, R. G. Improved medium and culture conditions for
clonal growth with minimum serum protein and for enhanced serum-free
survival of Swiss 3T3 cells. In Vitro (Rockville). ) 7 656-670, 1981.
15. Terasaki, M., and Rubin, H. Relations between intracellular Mg2', protein
synthesis, and the phenotype of transformed BALB/c 3T3 cells. J. Cell
Biol., 9i:4a, 1981.
16. Vidair, C., and Rubin, H. Regulation of intracellular Ca2* and its effects on
the phenotype of transformed
91: 12a, 1981.
and nontransformed
CANCER
3T3 cells. J. Cell Biol.,
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1982 American Association for Cancer Research.
VOL. 42
Effect of Magnesium Content on Density-dependent Regulation
of the Onset of DNA Synthesis in Transformed 3T3 Cells
H. Rubin
Cancer Res 1982;42:1761-1768.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/42/5/1761
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 17, 2017. © 1982 American Association for Cancer Research.