[CANCER RESEARCH 50, 5171-5176, August 15, 1990]
Progressive State Selection of Cells in Low Serum Promotes High Density Growth
and Neoplastic Transformation in NIH 3T3 Cells1
Adam Yao, Andrew L. Rubin, and Harry Rubin2
Department of Molecular and Cell Biology [A. Y., H. R.J, and Virus Laboratory [A. L. R., H. K.J, University of California, Berkeley, California 94720
ABSTRACT
A subline of NIH 3T3 cells maintained by frequent passage (every 2
to 3 days) in 10% calf serum (CS) at low population density reached a
low saturation density in 2% CS and produced no transformed foci on
prolonged incubation at confluency in 2% CS. Within 3 frequent low
density passages in 2% CS, the saturation density and focus-forming
capacity in that serum concentration began an increase which was contin
ued in subsequent passages. The saturation density and focus-forming
capacities of the cells in both 2% and 1% CS were further enhanced by
passage in 1% CS. The cells could then be passaged in 0.5% CS and
then in 0.25% CS, which would support no multiplication of cells previ
ously passaged only in 10% CS. The cells passaged in 0.25% CS gradually
increased their saturation density and focus-forming capacity in that
extremely low serum concentration during 24 low density passages,
although their initial growth rate did not increase. They also attained a
colony-forming efficiency in 0.25% CS of about 30%, as compared to
less than 1% for cells passaged in 10% CS. Cells passaged, cloned, and
passaged again in 2% CS yielded clonal populations which differed from
one another in saturation density and focus-forming capacity in 2% CS.
We conclude that NIH 3T3 cells diversify phenotypically at a high rate
in their capacity to multiply and produce foci in limiting concentrations
of serum, and we propose that progressive selection of these heteroge
neous states accounts for the acquired capacity to function effectively in
low concentrations of serum growth factors. Since lymph and presumably
extracellular fluid in vivo contain low concentrations of growth factors
which govern the multiplication of normal cells, the adaptation we observe
in vitro may be related to tumor production in the animal.
INTRODUCTION
Spontaneous neoplastia transformation of normal mouse em
bryo cells (1, 2), normal rat liver epithelial cells (3), and estab
lished mouse cell lines (4) occurs regularly when they are
cultured at high population density. It also has been found in
established mouse cell lines suspended in soft agar (5) or
passaged in low concentrations of serum (6). The common
denominator favoring transformation in these cases appears to
be a moderate constraint of growth and/or metabolism which
calls forth an adaptive response by the cells permitting increased
growth under constraint (4-6). The NIH 3T3 cell line, which
is particularly inclined to spontaneous transformation, is com
monly used as a target for transfection by tumor DNA, since it
readily undergoes accelerated transformation by the ras class of
oncogenes (7).
Transformation, whether induced spontaneously or by trans
fection, is detected by the development of foci containing cells
which continue to multiply after the surrounding cells have
become density inhibited. NIH 3T3 cells subjected to a regimen
of frequent passage for months at low density in high serum
concentration produce few if any foci when grown to confluency
in low serum concentration (4). Conversely, frequent passage
in low serum concentrations generates cell populations with
Received 4/6/90.
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 1734 solely to indicate this fact.
1This research was supported by grants from the USPHS and the Council for
Tobacco Research.
2To whom requests for reprints should be addressed.
high focus-forming capacities. These populations are also ca
pable of achieving high saturation densities in the low serum
concentrations (4, 6). The increase in saturation density in
volves the entire cell population in a relatively short period of
time, so it could not have resulted from the selection of a rare
mutant. However, the capacity to produce transformed foci
initially involves only a small fraction of the cell population.
No preexisting cells which could produce transformed foci in
low serum were found in repeated tests of the standard passage
of cells over 3 months of thrice weekly passages in high serum.
It is possible that rare genetic variants capable of focus forma
tion arose and were selected during the passages in low serum
concentrations. An alternative to this conventional explanation
is that physiological fluctuations occurred in the capacity of all
the cells to metabolize and function in low serum concentra
tions. Persistent culture of cells in low serum might then select
more efficient metabolic patterns or states in many if not all
cells of a population. Further fluctuation would make available
still more efficient states, thereby permitting progressive in
creases in the functional capacity of the cells. Since the fluctua
tion and selection are statistical processes, a particular fraction
of the cell population might be able to outstrip the multiplica
tion of the rest of the cells under conditions of added constraint,
such as confluency, thereby producing discrete foci of increased
density. We have termed this successive selection of metabolic
patterns "progressive state selection" (6, 8). We assume that
each step involves selection from an immense number of alter
native states so that there is a very low probability of precisely
retracing the pathway if conditions favoring unconstrained
function, e.g., high serum concentration, are restored.
A prerequisite for progressive state selection is that sponta
neous transient phenotypic variation occurs among cells and is
manifest in population heterogeneity for a particular character
istic. Such heterogeneity has been reported for many character
istics of cells in culture and in tumors. Most significant for the
present case is the observation that every cell in a population
of NIH 3T3 cells passaged and cloned in low serum concentra
tion gives rise to a clonal population with a capacity for focus
formation and for a sequence of changes in that capacity which
differ from those of the other clonal populations (6).
We wished to explore further the biological mechanism of
adaptation to growth and focus formation in low concentrations
of serum. Previous studies reported the lingering effects on
growth and focus formation in 2% CS3 of preconditioning the
cells by prolonged growth in 2% CS (6). Here we proceed in
stepwise fashion to passage cells in even lower serum concen
trations and measure changes in the growth and focus-forming
capacities of the cells. We also extend our observations on
clonal heterogeneity. The results show that progressive in
creases occur in growth capacities of individual cells and of the
entire population and that they may have as their basis the
progressive selection of fluctuating cellular states.
3The abbreviations used are: CS, calf serum; MCDB 402, molecular, cellular,
and developmental biology medium 402.
5171
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1990 American Association for Cancer Research.
ADAPTATION
MATERIALS
TO LOW SERUM AND NEOPLASTIC TRANSFORMATION
AND METHODS
NIH 3T3 cells are the product of five successive selections of flat
areas in low density seedings of Swiss mouse embryo cells (9). They
were provided by S. A. Aaronson of the National Cancer Institute and
passaged once in MCDB 402 (10) with 10% CS before storage in liquid
nitrogen, in that medium plus 10% dimethyl sulfoxide. After thawing,
the cultures for the serum adaptation experiments were maintained in
MCDB 402 plus 10% CS and 1 Mg/m' fungizone by passaging succes
sively at repeated intervals of 2, 2, and 3 days at 5 x IO4, 5 x IO4, and
2 x IO4 cells, respectively, per 60-mm plastic culture dish (Falcon,
Oxnard, CA). The clonal isolation experiment was done separately.
Before and after cloning the cells were passaged three times per week
at IO5, 10' and 5 x IO4 cells per dish, respectively, and fungizone was
not used. All incubations were done at 37°Cin humidified incubators
with an atmosphere of about 5% CO2, and cell numbers were deter
mined after trypsin treatment by electronic counting. Adaptation to
growth in lower serum concentrations was achieved by reducing the
serum concentration to the appropriate level during passages carried
out with the thrice weekly regimen noted above. Focus formation was
tested by seeding cells at 105/dish in 2% CS, changing the medium
twice a week, and incubating for 2 weeks. Since the cultures became
confluent at 3 to 4 days, they remained at the same confluent density
for 10 days or more. The cultures were then fixed in methanol and
stained with 4% Giemsa stain, and foci were counted without magnifi
cation against a white background. Numerical counts of foci are pre
sented in tables, and their appearance is illustrated in photographs to
exhibit differences in size and staining intensity of the foci, as well as
visual properties of the background cell sheet. Growth curves and
saturation densities were derived by seeding 5 x 10" or IO5cells in the
appropriate serum concentration, feeding twice a week, and harvesting
the cells for electronic counting at designated times. Colony-forming
efficiency was determined by seeding 100 cells on plastic in the appro
priate serum concentrations and incubating at 37°Cfor 12 and 16 days,
when the colonies were fixed in methanol, stained with Giemsa, and
counted.
RESULTS
Growth and Focus Formation by Cells Passaged in Low Serum
Concentrations. Cells were transferred every 2 to 3 days at low
density in 10% or 2% CS, and changes in their capacity to
produce foci at high density in 2% CS were tested at each
passage. Cells maintained in 2% CS gained the capacity to
produce foci in 2% at the third successive passage, which was
done 7 days after the start of the experiment (Fig. 1). Cells
which were passaged in parallel in 10% CS did not gain the
capacity to produce foci in 2% CS, even after months of
passaging (not shown). There was a transient decrease in focusforming capacity between the third and fourth passage among
cells passaged in 2% CS, which was accompanied by a decrease
in their saturation density in 2% CS from 1.5 x 10* to 1.0 x
IO6 cells/dish. However, the upward trend was resumed with
cells from the fifth and sixth passages (Fig. 1). Close inspection
of the photograph reveals that there are light, intermediate, and
dark foci especially evident in the dishes derived from the fifth
and sixth passages, suggesting that different degrees of trans
formation occur or that the transforming event occurred later
in the light than in the dark foci.
Having established that focus formation in 2% CS could be
elicited by a relatively short sequence of passages in 2% CS, the
effect of such pretreatment on further adaptation to 1% CS was
determined. In Fig. 2, it can be seen that cells passaged 12 times
in 2% CS or 6 times in 10% CS plus 6 times in 1% CS multiplied
to a higher saturation density in both 2% and 1% CS than did
cells which had been passaged only in 10% CS. Those which
had been passaged 6 times in 2% CS plus 6 times in 1% CS
multiplied to a higher saturation density than any of the other
cultures. A similar rank ordering was encountered when focus
formation rather than saturation density was measured (Fig. 3,
Tables 1 and 2). Cells passaged only in 10% CS produced no
foci in 2% or 1% CS. Those passaged only in 2% CS or in 10%
CS followed by 1% CS produced similar numbers of dark foci
though the latter produced many more light foci in both serum
concentrations. Cells that had been through the 2% plus 1%
sequence produced the largest number and size of foci in both
serum concentrations. It is apparent that prior adaptation in
2% CS facilitated further adaptation when passaged in 1% CS,
as measured by growth or focus formation in both serum
concentrations. The increase in size of the foci suggested that
it was not just the proportion of transformed cells but also the
degree of transformation that increased with graded decreases
in serum concentration.
We then wished to determine whether similar treatment
would allow cells to adapt to concentrations of serum less than
1%, which failed to support any net growth by cells that had
been maintained exclusively in 10% CS. We also wished to
determine whether such adaptation, were it to occur, would
result in a change in the initial rate of exponential multiplica-
A.
4
6
8
10 12 14 16
4
6
8 10 12 14 16
Days in culture
Fig. I. Focus formation in 2% CS of cells from successive passages in 2% CS.
Cells which had been passaged at 2-3-day intervals in 10% CS were shifted to
successive passages in 2% CS. At each successive passage, some of the cells were
set aside for focus formation in 2% CS. The assay dishes stained at 14 days were
from successive passages made at (left to right): top row, 2, 4, and 7 days; bottom
row, 9. 11, and 14 days.
Fig. 2. Saturation density in 2% and 1% CS of cells which had been repeatedly
passaged in various serum concentrations. Cells were divided in four categories
and passaged either 12 times in 10% CS (A), 12 times in 2% CS (•),6 times in
10% CS and 6 times in 1% CS (•),or 6 times in 2% CS and 6 times in 1% CS
(A). Each was then seeded on 6 dishes at 10' cells in 2% CS (A) or 1% CS (B).
The medium was changed twice a week, and cell counts were made at the times
indicated. Bars, range of cell counts. Where none are shown, the range was less
than the size of the symbols.
5172
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1990 American Association for Cancer Research.
ADAPTATION TO LOW SERUM AND NEOPLASTIC TRANSFORMATION
Fig. 3. Focus formation in 2% and 1% CS
of cells which had been repeatedly passaged in
various serum concentrations. Top row, cul
tures assayed in 2% CS; bottom row, 1% CS.
The cells were passaged as follows (left to
right): 12 times in 10% CS, 12 times in 2%
CS, 6 times in 10% CS and 6 times in 1% CS,
or 6 times in 2% CS and 6 times in 1% CS.
Table 1 Focus formation in 2% CS of cells which had been repeatedly passaged
in various concentrations ofCS
Cells were passaged thrice weekly in the indicated concentrations of CS for
the indicated number of passages and assayed for focus formation in 2% CS.
10'
Foci from cells passaged in
10% CS
(12)"
Dark foci
Light foci
Mottled background
" Number of passages.
0,0
0,0
2% CS
(12)°
10% CS +
1%CS
(6 + 6)°
12,22
3,7
10, 14
26,34
2% CS +
I%CS
(6 + 6)°
34,46
27,37
tr
Table 2 Focus formation in 1% CS of cells which had been repeatedly passaged
in various concentrations ofCS
Cells were passaged thrice weekly in the indicated concentrations of CS for
the indicated number of passages and assayed for focus formation in 1% CS.
Foci from cells passaged in
10% CS
(12)°
Dark foci
Light foci
Mottled background
" Number of passages.
0,0
0,0
2%CS
(12)°
5,9
2,6
10% CS +
1%CS
(6 + 6)°
7, 11
22,26
2% CS +
1%CS
(6 + 6)°
17,25
25,29
tion as well as the saturation density of the cells, since this
would inform us whether the adaptation could have resulted
from selection of genetically preadapted individual cells during
frequent low density passages. The cells that had been passaged
6 times in 2% CS plus 6 times in 1% CS were passaged 6 more
times in 0.5% CS. Their saturation density in 0.25% CS was
determined, and a passage series in that serum concentration
was begun. At passages 9, 12, 15, and 24, aliquots of the
passaged cells were used to establish full growth curves and
measure focus-forming capacities in 0.25% CS. It can be seen
in Fig. 4 that there was a gradual increase in the saturation
density of the cells with successive passage in 0.25% CS, with
the largest increase coming between passages 15 and 24.
The focus-forming capacity for the cells in 0.25% CS after
various numbers of passages in that serum concentration is
shown in Table 3 and Fig. 5. The steady increase in both
numbers and staining intensity of the foci with increasing
passages is apparent in Fig. 5. The figure also shows an increase
in staining of the background cells as well as the foci, especially
at passage 24, consonant with the increased saturation density
seen in Fig. 4. Failure to detect any change in the initial
io12
14
16
18 20
Days in culture
Fig. 4. Growth curves in 0.25% CS of cells from successive passages in 0.25%
CS. Cells previously passaged in 10% CS were passaged 6 times in 2% CS, 6
times in 1% CS, 6 times in 0.5% CS, and then 24 times in 0.25% CS. After
varying numbers of the successive passages in 0.25% CS, aliquots of cells were
set aside to carry out growth curves in 0.25% CS. The passage number in 0.25%
CS from which cells were obtained is indicated next to each curve.
exponential growth rates argues against selection of preexisting
mutants for growth to high density or focus formation, since,
as already noted, the passages were done at very low densities
in which the cells were kept in continuous exponential multi
plication.
We studied the speed and efficiency of colony formation after
seeding 100 cells on plastic in three concentrations of CS, using
cells that had previously been passaged in those three concen
trations (Table 4). Colony formation at very low cell densities
provides a more rigorous test of growth capacity than does
5173
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1990 American Association for Cancer Research.
ADAPTATION TO LOW SERUM AND NEOPLASTIC TRANSFORMATION
Table 3 Focus formation in 0.25% CS of cells from successive transfers
in 0.25% CS
Cells were passaged 6 times in 2% CS, 6 times in 1% CS, and 6 times in 0.5%
CS before initiating a passage series in 0.25% CS. They were assayed for focus
formation in 0.25% CS at the indicated passage number in 0.25% CS.
5 weeks in 10% CS, and their growth and focus-forming capac
ities were measured. Table 5 shows that the focus-forming
capacity of three clones (numbers 3, 5, and 7) decreased mark
edly, while that of two clones (numbers 6 and 8) increased
during 5 weeks of passaging in 10% CS. These changes in focusforming capacity of the clones are strikingly illustrated in the
photograph of Fig. 6, where the clones after 5 weeks of thrice
weekly passage are arranged in order of decreasing focus for
mation at the first passage.
There was considerable diversity in the saturation densities
of the clones after 5 weeks of thrice weekly passage (Fig. 7),
with those producing the most foci generally having the highest
saturation density. However, they all had the same initial
growth rates up to 2 days, after which areas of confluency made
for variable decreases in growth rate. This supported the con
clusion of the previous section that there was no opportunity
for selection of conventional mutants with high saturation
density or focus-forming capacity during frequent low density
passage.
No. of passages in 0.25% CS
12
Dark foci
2,3
Light foci
10, 13
Mottled background
" TN, too numerous to count.
8, 12
40,46
13, 18
TN
15
28,29
TN
24
TN°
TN
DISCUSSION
We establish here that cells can adapt by degrees to graded
decreases in serum concentration. The extent of adaptation is
dependent on the previous passage history of the cells. Cells
adapted to 2% CS reach a higher saturation density and focusforming capacity in both 2% and 1% CS than those passaged
in 10% CS. Cells passaged in 1% CS have a still higher
saturation density, the more so if they had been preadapted to
2% CS before their passage in 1% CS. Full adaptation to 0.25%
CS required many repeated passages in that serum concentra
tion after the cells had been successively adapted to 2%, 1%,
and 0.5% CS. It seems that there are almost as many levels of
adaptation as there are permutations of passage frequency, cell
density, and serum concentration. Indeed, there is nothing in
our evidence to rule out the possibility that the adaptation is a
continuous rather than stepwise process. It may, therefore, be
more appropriate to speak of degrees than of levels of adapta
tion. Increased saturation density and capacity for focus for
mation usually go hand in hand. Nevertheless, we have observed
that the cells of some cultures which grow to high density fail
to make foci among themselves unless left for periods far
beyond the usual 2-week assay. Such cells, however, do multiply
to form distinct foci within 2 weeks when a small number are
seeded in the presence of a large number of cells which provide
a flat monolayer background to set off the foci. Focus forma
tion, therefore, depends both on the ability to grow to high
saturation density and on a clear-cut difference in this capacity
between the focus formers and the surrounding cells. We indeed
find that heterogeneity in both focus-forming capacity and
saturation density exists among clones obtained from a cell
population which had been passaged and cloned in 2% CS. The
fact that there is sufficient heterogeneity to generate foci even
Fig. 5. Focus formation in 0.25% CS by cells from successive passages in
0.25% CS. The cells described in the legend to Fig. 4 were used at various passage
levels in assays for focus formation. Cells were from (left to right): top row,
passages 0 and 9; bottom row. passages 12. 15. and 24.
increase of mass populations and indicates how individual cells
and their progeny respond to the various conditions. Two major
features of colony formation present themselves. First, cells
adapted to 0.25% were a few days slower to develop large
colonies in 10% CS than were cells that had been passaged in
the higher serum concentrations, although all three groups grew
to semiconfluency at 16 days. Second, only a small fraction of
the cells passaged in 10% and in 2% CS managed to form
colonies in 0.25% CS, even at 16 days, and those colonies were
small. By contrast, cells passaged in 0.25% CS had a colonyforming efficiency of about 30% when assayed in that CS
concentration, and those colonies reached a large size at 16
days. It may also be noted that colony formation in 2% CS was
equivalent for cells passaged in 10% and in 2% CS, confirming
the earlier observation that adaptation from 10% to 2% CS
increases the saturation density and focus-forming capacity of
the cells in 2% CS, but not their initial growth rate.
Focus Formation and Growth of Clones from Cells Passaged
and Cloned in High and Low Serum Concentrations. We have
previously reported that cells passaged repeatedly in 2% CS
and cloned in 2% CS gave rise to clones, all of which produced
foci in 2% CS but each of which was unique in the number of
foci produced (6). The clones were passaged thrice weekly for
Table 4 Colony formation in 10%, 2%, and 0.25% CS of cells passaged in those serum concentrations
Cells were transferred 3 times a week at low density in 10% CS or 2% CS for 51 passages. In addition, cells which had been adapted to 0.5% CS were transferred
in 0.25% CS for 28 passages (see Table 3 and Fig. 5). One hundred cells from each of the 3 series were then seeded on 4 dishes in each serum concentration and 2
dishes were fixed and stained at 12 and 16 days.
No. of colonies formed in
Prior pas
in10%
sage
10% CS
daysSC,
SO1
2%CS
daysSC,
days44,
2
CS
SC
37
38, 38 (B)
2%CS
SC29,
SC,
SC, SC
41,41 (D)16
0.25% CS12
15(D)16
SC, SC1
' SC, semiconfluent; S, small colonies; B. broad colonies; D, dense colonies.
0.25% CS
days47,42
42,40
38,-1
days0,0
2
d6ays1.
0,0
13, 8 (S)1
1 (S)
4, 5 (S)
23,39
5174
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1990 American Association for Cancer Research.
ADAPTATION TO LOW SERUM AND NEOPLAST1C TRANSFORMATION
10'
Table 5 Focus formation by clones at the first and fifth weeks of their weekly
passage at low density
Cells which had been passaged for 6 weeks in 2% CS were cloned in 2% CS
and the clones were grown to large population for passaging in 10% CS. They
were tested for focus formation in 2% CS at the first passage and again after 5
weeks of thrice weekly passage, and growth curves were conducted with the latter
(see Fig. 7). Clones are listed in order of decreasing focus formation at first
passage.
A.
Clone
no.123456789Passage 1TN"TNTN31311612105»Passage
5TNTN8573'401623"
B.
10
10 =
' TN, too numerous to count.
*Very pale foci.
IO12 14
0
24
6
10 12 14
Days in culture
Fig. 7. Growth curves in 2% CS of clones which had been passaged and cloned
in 2% CS. Cells previously in 10% CS were passaged weekly for 5 weeks in 2%
CS and cloned in 2% CS. Nine clones were isolated and passaged for 5 weeks in
10% CS. After 5 weeks of thrice weekly passage, growth curves in 2% CS were
established starting with 5 x 10* cells/dish. See Fig. 6 and Table 5 for foci
produced by these cells. A, clone 1 (G), clone 2 (O), clone 3 (A), clone 4 (+), and
clone 5 (•).B, clone 6 (•),clone 7 (A), clone 8 (x). and clone 9 (TD).The growth
curves are separated into two panels to facilitate discrimination among them.
Fig. 6. Focus formation in 2% CS by clones which had been passaged and
cloned in 2% CS. The cells were aliquots of clones passaged thrice weekly for 5
weeks and were also used in the growth curves of Fig. 7. The numbers correspond
to those of Fig. 7 and Table 5 and represent decreasing degrees of transformation
in passage 1, as seen in Table 5. The rank order of focus formation has obviously
changed in passage.
during development of a clonal population is a clear indication
that heterogeneity is generated at a high rate during develop
ment of the clones.
The continuous phenotypic diversification in properties
which determine the capacity for growth to high density and
focus formation provides sufficient variety for selection of those
properties. However, the cells adapting to reduced concentra
tions of serum were never allowed to reach population densities
during their passage regimes that would permit direct selection
for high saturation density. During the 24 successive passages
of adaptation to 0.25% CS, there was no increase in the initial
exponential growth rate of the cells as their saturation density
and focus-forming capacity increased (Fig. 4 and Table 3). This
suggests that selection operated on variations in the metabolic
state of the cells responsible for the increased saturation density
and focus-forming capacity, rather than the growth capacity
itself. Selection on the basis of fluctuating metabolic states
permits adaptation of a large fraction if not the whole popula
tion. And indeed, we have previously shown that an entire
population of the NIH 3T3 cells can acquire the capacity to
grow to high density in reduced serum concentrations within a
period too short to allow for selection of a minority element of
the population (6). We have termed this adaptive process pro
gressive state selection. The concept of progressive state selec
tion implies that the process might be reversible, i.e., that
deadaptation would occur if the cells were returned to the
original conditions of high serum concentration. And indeed
this has been found to be at least partly true, since cells which
were adapted to 2% CS gradually lost their capacity for focus
formation in 2% CS during repeated passages in 10% CS (6).
However, some of the capacity for focus formation in 2% CS
was retained even after many cell divisions in 10% CS, and the
saturation density was only slightly reduced (6). The partial
stability of adaptation to 2% CS may result from the fact that
such cells also function well in 10% CS and that there is only
weak selection pressure on the metabolic state of the cells when
the serum concentration is raised, i.e., a serum downshift in
this range is a more selective condition than an upshift.
Previous attempts have been made by others to adapt other
cell types for growth in low serum concentrations or in no
serum. Ten cell lines out of 56 of clone 929 mouse cells were
successfully adapted to a serum-free medium (11). The cells
went through an extensive critical period (from 4 to 21 weeks)
during which they either succumbed or adjusted to the new
milieu. There was wide variation in the response of the lines,
despite their origin from the same clone. Attempts by others to
adapt Swiss 3T3 cells to reduced concentrations have failed
(12). It is apparent that there is considerable variability within
and between cell lines in capacity to adapt.
The adaptation of the NIH 3T3 cells to growth in low serum
concentrations resembles certain long term changes induced as
a response to restrictive environmental conditions in other
systems. The earliest to be described were the Dauermodifika
tionen or enduring modifications in Paramecium aurelia (13,
14). In these organisms, resistance could be induced to high
concentrations of antiserum to surface antigens by exposing
them to low concentrations of the antiserum. This resistance
persisted for many generations after the antiserum was removed
but disappeared after mating of resistant organisms to either
sensitive or to other resistant organisms. The resistance was
shown to be due to a change in the type of surface antigen
which, although coded by nuclear genes, was controlled by the
5175
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1990 American Association for Cancer Research.
ADAPTATION
TO LOW SERUM AND NEOPLASTIC TRANSFORMATION
state of the cytoplasm (15). Similar long lasting but nongenetic
changes were found to underlie adaptation in bacteria to mod
erate concentrations of drugs or to changes in the sources of
carbon and nitrogen (16). Animal cells also adapt to low doses
of purine and pyrimidine analogues and remain adapted for
many generations after removal of the drugs but eventually
revert to sensitivity either spontaneously (17) or after treatment
with a variety of compounds (18). Variation in capacity to
multiply in low concentrations of serum is generated at a very
high rate within clones of mouse cells in culture (19). Similar
high rates of variation occur among the surface antigens of
human urinary bladder tumor cells in culture (20). Bladder
tumor cells with low surface antigen content can be identified
with fluorescent antibodies and selected by fluorescence-acti
vated cell sorting to generate populations which retain this
property (20). The molecular basis for these enduring modifi
cations is unknown. However, it has been proposed that bacte
rial adaptation involves integrated change in metabolic patterns
of the entire cell (21). We have proposed the model of progres
sive state selection for such integrated change based on the high
rate of phenotypic diversification in cultures of animal cells (5,
6, 22-25). When cells are put under moderate metabolic con
straint, as in the present case by lowering the concentrations of
serum growth factors, selection occurs of those metabolic pat
terns better suited than others to function under that constraint.
But metabolic fluctuation persists, and further reduction in
serum concentration results in successive selection of more
efficient patterns. Interestingly, when a different type of adap
tive pressure is applied to the cells, in the form of a restriction
of the energy-supplying precursor glutamine, transformation is
suppressed (26). This suggests that the metabolic responses
elicited by this condition are different from those engendered
by limiting concentrations of serum. The fact that we show
quasistable adaptation proportional to the constraints of var
iously lowered serum concentration supports the concept of
progressive state selection as the biological basis for adaptive
growth.
ACKNOWLEDGMENTS
The technical assistance of Alisa Sneade-Koenig and manuscript
preparation by Dawn Davidson are greatly appreciated.
REFERENCES
1. Todaro, G. J., and Green, H. Quantitative studies of the growth of mouse
embryo cells in culture and their development into established lines. J. Cell
Biol., 77:299-313, 1963.
2. Aaronson, S. A., and Todaro, G. J. Basis for the acquisition of malignant
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
potential by mouse cells cultivated in vitro. Science (Wash. DC), 162: 10241026, 1968.
Tsao, M-S., Sheperd, J., and Batist, G. Phenotypic expression spontaneously
transformed cultured rat liver epithelial cells. Cancer Res., 50: 1941-1947,
1990.
Rubin, H., and Xu, K. Evidence for the progressive and adaptive nature of
spontaneous transformation in the NIH 3T3 cell line. Proc. Nati. Acad. Sci.
USA, 86: 1860-1864, 1989.
Rubin, H. Uniqueness of each spontaneous transformant from a clone of
Balb 3T3 cells. Cancer Res., 48: 2512-2518, 1988.
Rubin, A. L., Yao, A., and Rubin, H. Relation of spontaneous transformation
in cell culture to adaptive growth and clonal heterogeneity. Proc. Nati. Acad.
Sci. USA, 87: 482-486, 1990.
Varmus, H. E. The molecular genetics of cellular oncogenes. Annu. Rev.
Genet., 18: 553-612, 1984.
Rubin, H. On the nature of enduring modifications induced in cells and
organisms. Am. J. Physiol., 25«:L19-L24, 1990.
Jainchill, J. L., Aaronson, S. A., and Todaro, G. J. Murine sarcoma and
leukemia viruses: assay using clonal lines of contact-inhibited mouse cells. J.
Virol., 4: 549-553, 1969.
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, 17: 656-670, 1981.
McQuilken, W. T. Evans, V. J., and Earle, W. R. The adaptation of additional
lines of NCTC clone 929 (strain L) cells to chemically defined protein-free
medium NCTC 109. J. Nati. Cancer Inst., 19: 885-907, 1957.
Brooks, R. F., Richmond, F. N., Riddle, P. N., and Richmond, K. M. V.
Apparent heterogeneity in the response of quiescent Swiss 3T3 cells to serum
growth factors: implications for the transition probability model and parallels
with "cellular senescence" and "competence." J. Cell. Physiol., ¡21:341-
350, 1984.
13. .loilos. V. Experimentelle Protisten Studien. I. Unterschungen überVaria
bilitätund Vererbung bei Infusoren. Arch. F. Protistenk., 43: 1-222, 1921.
14. Sonneborn, T. Recent advances in the genetics of Paramecium and Euplotes.
Adv. Genet., /: 263-358, 1947.
15. Beale, G. Antigen variation in Paramecium aurelia. Genetics, 37: 62-74,
1952.
16. Dean, A. C. R., and Hinshelwood, C. Reaction patterns of a coliform
organism. Prog. Biophys. Biophys. Chem., 5: 1-40, 1955.
17. Fox, M., and Radicic, M. Adaptational origin of some purine-analogue
resistant phenotypes in cultured mammalian cells. Mutât.Res., 49:275-296,
1978.
18. Harris, M. Induction of thymidine kinase in enzyme deficient Chinese
hamster cells. Cell, 29: 483-492, 1982.
19. Brooks, R. F., and Riddle, P. N. Differences in growth factor sensitivity
between individual 3T3 cells arise at high frequency: possible relevance to
cell senescence. Exp. Cell Res., 174: 378-387, 1988.
20. Coon, J. S., Watkins, J. R., Pauli, B. U., and Weinstein, R. S. Flow cytometric
analysis of heterogeneity in blood group-related antigen expression in a
human urinary bladder carcinoma cell line, 647V. Cancer Res., 45: 30143021, 1985.
21. Dean, A. C. R., and Hinshelwood, C. Integration of cell reactions. Nature
(Lond.), / 99: 7-11, 1963.
22. Rubin, H., and Hatie, C. Increase in the size of chick embryo cells upon
cultivation in serum-containing medium. Dev. Biol., 17: 603-616, 1968.
23. Rubin, H. Early origin and pervasiveness of cellular heterogeneity in some
malignant transformations. Proc. Nati. Acad. Sci. USA, */: 5121-5125,
1984.
24. Gründe!,R., and Rubin, H. Maintenance of multiplication rate stability by
cell populations in the face of heterogeneity among individual cells. J. Cell.
Sci., 91: 571-576, 1988.
25. Rubin, H. The significance of biological heterogeneity. Cancer Met. Rev., in
press, 1990.
26. Rubin, A. L. Suppression of transformation by, and growth adaptation to,
low concentrations of glutamine in NIH 3T3 cells. Cancer Res., 50: 28322839, 1990.
5176
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1990 American Association for Cancer Research.
Progressive State Selection of Cells in Low Serum Promotes
High Density Growth and Neoplastic Transformation in NIH 3T3
Cells
Adam Yao, Andrew L. Rubin and Harry Rubin
Cancer Res 1990;50:5171-5176.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/50/16/5171
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 18, 2017. © 1990 American Association for Cancer Research.