[CANCER RESEARCH 42, 5139-5146,
0008-5472/82/0042-OOOOS02.00
December 1982]
Neoplastic Transformation and Defective Control of Cell Proliferation
and Differentiation1
John J. Wille, Jr., Peter B. Maercklein, and Robert E. Scott2
Section of Experimental Pathology, Departments of Anatomic Pathology and Cell Biology. Mayo Clinic/Foundation.
ABSTRACT
The control of proliferation of nontransformed 3T3 T-proadipocytes in vitro can be mediated at three states in the G,
phase of the cell cycle. These states are induced by the
commitment of cells to differentiate (GD); by growth factor
deprivation at low density or "contact inhibition" at high density
(Gs); and by nutrient deprivation (GN). To determine if neoplastic transformation of proadipocytes is associated with a selec
tive defect in one or more of these G, growth arrest processes,
we developed and studied eight cloned and several noncloned
tumorigenic proadipocyte cell lines. We report that all trans
formed proadjjipocyte cell lines are tumorigenic and all lack
the ability to arrest at GD and differentiate. By contrast, ~90%
of transformed proadipocyte cell lines retain their ability to
growth arrest at Gs at low density when deprived of growth
factors, and ~90% growth arrest at GN when deprived of
nutrients. These observations suggest that neoplastic transfor
mation of proadipocytes is primarily associated with abrogation
of growth control mediated at GD. However, whereas most
transformed proadipocytes arrest at Gs at low density when
deprived of serum, all transformed proadipocyte cell lines do
not efficiently arrest at Gs at high density due to "contact
inhibition." This suggests that neoplastic transformation of
proadipocytes results from a primary defect in growth control
mediated at GD and from an additional defect at Gs. These
results are discussed with respect to their possible significance
for the biological mechanisms of the initiation and promotion of
carcinogenesis.
Rochester, Minnesota 55905
role of defects in Gìcell cycle-dependent control mechanisms
in neoplastic transformation (24). More specifically, studies
were performed to determine if transformation is associated
with selective defects in the control of cell proliferation and/or
differentiation. The rationale for this approach is based on the
fact that stem cells are the targets of most carcinogenic agents
(26) and that stem cells also have the capacity to proliferate
and to differentiate. We have used proadipocytes in these
studies because they represent a model in vitro cell line which
also has stem cell-like characteristics.
Transformed proadipocytes were developed and cloned and
then studied with respect to their cell cycle growth control and
differentiation properties. In particular, transformed proadipo
cytes were characterized with respect to their (a) saturation
density, (b) ability to growth arrest at Gìfollowing serum or
isoleucine deprivation, and (c) ability to arrest at GD and
differentiate. The data show that: (a) most transformed proad
ipocyte cell lines retained their ability to Gt growth arrest at GN
as a result of nutrient deficiency; (b) at low density, most
transformed proadipocytes can arrest at Gs as a result of serum
deprivation; (c) at high density, transformed proadipocytes
showed variable defects in their ability to growth arrest at Gs
as a result of contact inhibition; and (d) all transformed proad
ipocytes showed defects in their ability to arrest at GD and
differentiate. These results suggest that neoplastic transfor
mation of proadipocytes is associated with a primary defect in
growth control mediated at GD and with an additional defect in
growth control at Gs, which is mediated at high cell density.
MATERIALS
AND METHODS
INTRODUCTION
Methods
The regulation of growth in many cell types is physiologically
associated with the induction of cell differentiation (15). For
example, in hematopoietic (19) and epithelial (16) stem cells,
control of proliferation is associated with the expression of a
differentiated phenotype. In many such cell types, the control
of both cell proliferation and cell differentiation is mediated in
the Gìphase of the cell cycle (10, 12, 28), and we recently
reported (14, 18, 32-34) that in cultured proadipocytes the
control of both cell proliferation and differentiation can be
regulated at a distinct state in G,, designated GD. In particular,
it was shown that the GDarrest state was distinct from other Gt
states, including those induced by nutrient deprivation (GN) or
serum starvation and contact inhibition (Gs), by a variety of
criteria (14, 18, 32,33, 38).
The central question examined in this paper relates to the
Cell Culture. Studies were performed on BALB/c 3T3 T-proadipocytes and on 3T3 (clone A31 ) cells. In addition, a variety of transformed
cell lines was studied. These included methylcholanthrene (MCA3T3)and SV40-transformed
3T3 cells (SV3T3) and recently transformed
3T3 T-proadipocytes
(see below). The 3T3 T-proadipocytes
were
previously characterized by Dr. L. Diamond, who kindly provided these
cells (11 ). Nontransformed and transformed proadipocytes were grown
in DMEM3 containing 10% PCS at 37° in a 5% CO2 atmosphere;
nontransformed and transformed 3T3 cells (provided by Dr. G. Todaro)
were grown as above in DMEM containing 10% calf serum. Stock
cultures of each cell line were initiated every 2 months from frozen
specimens, and they were maintained at low density unless otherwise
designated to avoid selection of cells which are not density inhibited.
All cell lines were shown to be free of Mycoplasma contamination as
previously described (7, 18, 29-33).
Neoplastic Transformation of 3T3 T-Proadipocytes and 3T3 Cells
1This study was funded in part by NIH Grants CA 28240 and CA 21722, and
by the Mayo Foundation.
2 To whom requests for reprints should be addressed.
Received December 14, 1981; accepted September 9. 1982.
DECEMBER 1982
3 The abbreviations
SST, smooth-surface
used are: DMEM, Dulbecco's
tumorigenesis;
modified Eagle's medium;
FCS. fetal calf serum.
5139
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1982 American Association for Cancer Research.
J. J. Wille, Jr. et al.
by the SST Method. Both 3T3 cells and 3T3 T-proadipocytes were
transformed by the SST method of Boone (6). Briefly, 3T3 cells or 3T3
T-proadipocytes were grown on smooth-surfaced plates and implanted
s.c. in syngeneic mice. The incidence of tumor development was
recorded, and then tumor fragments were minced and/or trypsinized
and placed in culture. Penicillin, streptomycin, and Fungizone were
included in the culture medium for the first 2 weeks; thereafter, cells
were grown in antibiotic-free medium. Three 3T3 T-transformed cell
lines and 23 3T3-transformed cell lines were obtained from the tumors
thus formed. One 3T3 T-transformed cell line with a saturation density
only moderately greater than the nontransformed parent 3T3 T-clone
was extensively cloned as described below.
Spontaneously transformed 3T3 T-proadipocytes
were also isolated
by maintaining the cells at high density for extended intervals without
passage but with repeated feeding. This procedure selected for cells
incapable of growth inhibition by contact or density-dependent control
processes.
The cloning
of the transformed
3T3 T-cell line was achieved
by
plating trypsinized cells at very low density onto glass cover chips
which had been prepared by crushing glass coverslips and selecting
for fragments smaller than 0.1-mm diameter by filtration through geo
nutrient deficiency state, the effects of various potentially mitogenic
agents were tested as above. These included: (a) fresh DMEM + 10%
FCS; (£>)
an isoleucine concentrate; or (c) 30% dialyzed FCS (3, 32).
Nutrient-arrested cells are induced to synthesize DNA only when iso
leucine or DMEM is added. Dialyzed serum is not mitogenic for nutrientarrested cells.
Go Arrest and Adipocyte Differentiation. Two different methods
which induce GD arrest and differentiation in nontransformed 3T3 T
proadipocytes were used to study transformed proadipocytes; (a) highdensity cells were fed 30% FCS and insulin (50 /ig/ml) repeatedly over
a 21-day interval (31, 32); and (b) low-density cells were cultured in
heparinized (30 units/ml) medium containing 25% human plasma (18,
32, 33).
With the first method, nontransformed
cells typically grow to a
density of ~1 x 105 cells/sq cm and then become resistant to further
growth stimulation. Greater than
content and are designated to be
to 40% of these cells differentiate
In the second method, greater
90% of these cells have a G, DNA
at the Go arrest state. Thereafter, up
into fat cells within 2 to 3 weeks.
than 95% of nontransformed 3T3 T-
proadipocytes growth arrest in G, following culture in heparinized
DMEM containing 25% human plasma for ~2 days. These cells are
logical screen material. After the cells had adhered to the glass
fragments, individual chips containing a single cell were microscopi
cally located, isolated, and transferred to Retri dishes. Cells on chips
were cultured in DMEM containing 10% PCS until approximately 100
to 500 cells were present for each clone. They were then passaged by
standard procedures. A total of 5 SST clones were isolated. They were
designated M2, M3, M4, M6, and M10. Frozen stock cultures of all of
these cell lines have been preserved except M3, which was recently
lost due to contamination. To assure that the SST-derived cell lines
were derived from the implanted 3T3 T-parent cell line, karyotype
thymidine into DNA upon repeated feedings with DMEM-30% FCS +
50/ig insulin per ml (32). In both high- and low-density nontransformed
and transformed cells, mitogenic responsiveness to 3-methyl-1 -isobutyl
xanthine (5 x 10~4 M) was also used to determine if they were at the
analyses were performed. The mean number of chromosomes in the 5
M-clones [65 ± 27 (S.E.)] was not significantly different from the
parent 3T3 T-proadipocyte line (62 ±25). In addition, all of the cloned
Go state. This assay has been shown to reproducibly
distinguish
between cells arrested at GD and cells arrested at Gs and GN (32). The
extent of proliferation induced by 3-methyl-l-isobutyl
xanthine was
transformed proadipocyte lines had a very similar unimodal frequency
distribution of chromosome number which was indistinguishable from
the parent 3T3 T-proadipocyte
cell line. Spontaneously transformed
determined as the percentage of cells containing nuclei labeled with
[3H]thymidine during a 48-hr interval following drug addition relative to
cells were also cloned, and 3 such lines were isolated: T2, T3, and T8.
The tumorigenicity
of these clones was assayed in vivo, using
syngeneic mice (The Jackson Laboratory, Bar Harbor, Maine). Cell
clones were trypsinized, washed once in DMEM, and resuspended in
DMEM at a concentration of 1 x 107 cells/ml. One ml of this cell
also designated to be at the GD state. Thereafter, adipocyte differentia
tion occurs within 5 to 10 days when 80 to 100% of the cells develop
into mature adipocytes. The biological characteristics of both nontrans
formed and transformed cells were studied to determine if they were
arrested at GD. In the first method, autoradiography
(3) was used to
demonstrate loss of ability of high-density cells to incorporate [3H]-
untreated cells.
Differentiation was assayed morphologically by the appearance of
large lipid droplets in the cell cytoplasm and by the increased activity
of glycerol-3-phosphate
dehydrogenase (17). These assays have been
described in detail previously (32) and have been shown to be excellent
criteria for adipocyte differentiation (22, 33).
suspension was then injected s.c. on the dorsum of male BALB/c mice
(29). The development of tumors was assayed over a 50-week period.
Tumors that developed were examined histologically to confirm their
neoplastic phenotype. Nontransformed 3T3 and 3T3 T/proadipocytes
were used as controls.
Gs and GN Arrest. The capacity of transformed proadipocytes to
growth arrest was determined both by measurement of decreased
[3H]thymidine incorporation and by measurement of cellular DMA con
tent, using flow microfluorimetric analysis (29). Growth arrest of trans
formed cells in serum-deficient medium at Gs was induced by culture
of low-density cells in DMEM containing 0.5% PCS for 4 days as
described (32). To prove that such cells were indeed at the growth
factor deficiency arrest state, studies were performed to show that
such cells could be restimulated to grow only by addition of growth
factors to the cells (27, 32, 39). For these assays, a comparison was
made of the relative mitogenic effect of: (a) 30% FCS + 50 /ig insulin
per ml; or (b) 10% DMEM added as a x 10 concentrate. Gs arrest was
also assayed by measurement of the saturation density of transformed
cell lines 4 to 6 days after 80% confluent cells were refed DMEM-10%
FCS. Maximum cell density was determined by dispersing cells with
trypsin and counting in a hemocytometer or in a Coulter apparatus.
Comparable results were obtained by both methods. Transformed 3T3
T-cells were also cultured in isoleucine-deficient
DMEM containing
dialyzed 10% FCS for 3 days to induce arrest at the GN state (32, 39).
To establish that the clones were actually growth arrested at the
5140
Materials
Heparinized medium containing
prepared from citrate-anticoagulated
human platelet-poor plasma was
human blood drawn by venipunc-
ture from healthy volunteers as described in detail elsewhere (18).
Briefly, venous blood was drawn and placed immediately in sodium
citrate at a final concentration of 0.38%. Citrate-anticoagulated
blood
was sedimented by centrifugaron at 100 x g for 30 min at 22° and
then at 25,000 x g for 30 min at 22°to remove platelets. This citrateanticoagulated platelet-poor plasma was frozen at —¿70°
overnight and
subsequently thawed at 4°.The freeze-thawed plasma was sedimented
at 100 x g for 10 min at 4° and stored at 4° following Millipore
filtration. It was then added to heparinized DMEM at the appropriate
concentration. Optimum GD arrest and adipocyte differentiation were
typically induced by culture of proadipocytes in heparinized (30 units/
ml) DMEM containing 25% human plasma. DMEM was purchased from
Grand Island Biological Co., Grand Island, N. Y. Serum was purchased
from KC Biologicals, Lenexa, Kansas. Mithramycin, penicillin, strepto
mycin, insulin, and 3-methyl-1-isobutyl
xanthine were purchased from
Sigma Chemical Co., St. Louis, Mo. Sodium heparin (Panheparin) was
purchased from Abbott Laboratories, North Chicago, III. [3H]Thymidine
(40 to 60 Ci/mmol) was purchased from New England Nuclear, Boston,
Mass.
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1982 American Association for Cancer Research.
VOL. 42
Proliferation-Differentiation
RESULTS
3T3 T-proadipocytes are noninitiated, nontransformed cells
because they are not tumorigenic when injected in suspension
in syngeneic mice and they show very low tumorigenic potential
when implanted in mice on smooth-surfaced plates, which
promotes the tumorigenesis of initiated cells (6, 7, 30). Further,
pretreatment of 3T3 T-proadipocytes with a tumor promoter
prior to implantation does not increase tumor incidence,
whereas pretreatment of initiated 3T3 T cells with phorbol
myristate promotes tumorigenesis (30). 3T3 T-proadipocytes
are also distinguished from 3T3 cells because only 3T3 Tproadipocytes have the potential to arrest at GD and differen
tiate.
These observations provide the basis for 2 types of studies
described in this paper: (a) studies were performed to deter
mine the effect of oncogenic transformation of 3T3 T-proadi
pocytes on their ability to arrest at GD and differentiate relative
to their ability to regulate growth at the Gs and GNarrest states;
(£>)studies were performed to determine if nontransformed
3T3 cells and chemically and virally transformed 3T3 cells can
be induced to arrest at GD and whether they differ in their G,
growth arrest properties from one another and from neoplastically transformed 3T3 T-proadipocytes.
Neoplastically Transformed
3T3-Proadipocytes.
Even
though 3T3 T-proadipocytes are noninitiated, nontransformed
cells, an occasional tumor does develop when such cells are
implanted in mice on smooth surface plates. Three such tumors
have been produced, and from these 3 cell lines have been
developed. Two cell lines had a saturation density approxi
mately twice that of the 3T3 T-parent, while one cell line had
an only moderately greater saturation density (6.5 x 10" cells/
sq cm) than did nontransformed 3T3 T-proadipocytes (5 x 104
Defects in Neoplasia
Table 1 lists the tumorigenic potential of the clones of neoplastically transformed proadipocytes that were isolated. These
data show that suspensions of all clones induce a significant
tumor incidence in syngeneic mice. Experiments were there
fore performed to answer the following questions: (a) do any of
the cloned transformed cell lines show evidence of contact
inhibition of growth; (b) can any of the cloned transformed cell
lines be growth arrested at Gs by growth factor deprivation; (c)
can any of the cloned transformed cell lines be growth arrested
at GN by isoleucine deprivation; and (d) do the cloned trans
formed cell lines show defects in their ability to arrest at GD
and differentiate?
Contact-inhibition or Density-dependent Growth Inhibition
in Cloned Transformed Proadipocytes. Figs. 1 and 2 compare
the morphology of nontransformed 3T3 cells and 3T3 T-proad
ipocytes to 8 clones of neoplastically transformed proadipo
cytes at their respective saturation densities (Table 1 gives the
saturation density of these cultures). All transformed clones
display a higher saturation density than do the nontransformed
Table 1
Tumorigenic potential and saturation densities of nontransformed and
transformed proadipocytes
Cell types
Tumorigenicity
incidence
Saturation density
(cells/sq cm)
0/10
5.0 X 10*
Nontransformed
3T3T
Transformed, spontaneous method
10s1.8
X
1051.7
X
1051.3
X
T2T3T8Transformed,
methodM2M3M4M6M103/55/53/52/55/55/55/55/51.5
SST
cells/sq cm). This latter cell line was cloned, and the clones
were further studied to determine their cell cycle growth control
properties and their ability to arrest at GD and differentiate.
1056.0
X
10*1.5
X
1051x
057.0
.4 x 1
X 104
Table 2
Ability of neoplastically transformed proadipocytes to G, growth arrest due to either growth factor (Gs) or
nutrient fGJ deprivation
(%)Cell
Labeled nuclei
arrest3Growth
arrest"Growth
restimulationarrest
typeNontransformed
3T3TGs
arrest0
restimulationABC100
A
B
95
0
6GN
75
2
Transformed, spontaneous
method
T2T3T8Transformed,
methodM2M3M4M6M1015124100602787572984525805281c_010—0001304870823310010095959071825195856075551518412096
SST
°Growth arrest was determined
by autoradiographic
measurement
of [3H)thymidine (5 fiCi/ml)
incor
poration into DNA during a 48-hr interval begun 3 days after cells were fed DMEM-0.5% FCS. Growth
restimulation was determined as above when arrested cells were refed either (A) 30% PCS + 50 fig insulin
per ml; or (B) 10% DMEM added as a 10x concentrate.
b Growth arrest was determined as in footnote a, after cells were cultured in isoleucine-deficient DMEM10% dialyzed PCS for 3 days. Growth restimulation was determined as in footnote a, when arrested cells
were refed either (A) fresh DMEM + 10% PCS; (B) an isoleucine concentrate; or (C) 30% dialyzed PCS.
c —¿.
not determined.
DECEMBER 1982
5141
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1982 American Association for Cancer Research.
J. J. Wille, Jr. et al.
proadipocytes. T-clones showed the highest cell densities,
while M-clones showed a wide range of saturation densities.
For example, M-clones M3 and M10, have only slightly elevated
saturation densities, whereas the remainder of the M- and Tclones showed high saturation densities. The elevated satura
tion densities observed in confluent monolayer cultures of
transformed proadipocytes are also a feature shared with other
cells transformed
by chemicals (MCA3T3) or by virus
(SV403T3). Since loss of the ability to show density-dependent
growth inhibition, i.e., contact inhibition, has been shown to be
associated with defects in growth factor-dependent
control
processes mediated at Gs, these data suggest that most trans
formed proadipocytes do have a defect in the control of Gs
arrest at high density.
Arrest of Cloned Transformed Proadipocytes at Gs when
Cultured ¡nGrowth Factor-deficient Medium. The data in
Table 2, however, show that all transformed clones tested, with
the exception of T8, can be growth arrested at Gs when
cultured in DMEM-0.5% PCS, i.e., a growth factor deficiency
medium. Growth arrest induced under these conditions is
shown to be reversed by addition of 30j% PCS plus insulin
(Column A) but not by the addition of nutrients (Column B). This
establishes that they were indeed arrested due to a deficiency
of serum factors. Spontaneously transformed proadipocytes
(T2, T3, and T8) are more variable in their response to growth
factor deprivation than are cells transformed by the SST
method (M2, M3, M4, M6, and M10). Conversely, SST-transformed cells are more variable in their response to added
growth factors following arrest.
To establish that growth arrest induced by serum deprivation
actually occurred in the G, phase of the cell cycle in trans
formed cells as occurs in 3T3 T-cells and 3T3 cells, flow
microfluorometric
analyses were performed. The data pre
sented in Table 3 show that clones M2, M3, M4, and M6
growth arrested in serum-deficient medium are highly enriched
in Gìcells. In the other clones, selective growth arrest also
occurs in Gt, but the enrichment is less pronounced.
Arrest of Cloned Transformed Proadipocytes in GN when
Cultured in Isoleucine-deficient Medium. The data in Table 2
also show that all 8 transformed 3T3 T-proadipocyte clones
growth arrest when cultured in isoleucine-deficient
medium
and that this arrest can be reversed by addition of nutrients
(Columns A and B) but not by the sole addition of serum growth
factors (Column C). Further, Table 3 shows the cell cycle
distribution of nutrient-arrested transformed cell lines. All of
the clones, except M10, arrest in the d phase of the cell cycle
when cultured in isoleucine-deficient medium. It should, how
ever, also be stressed that the M10 cell line also shows an
unusual cell cycle population profile even in the rapidly growing
state (Table 3).
Neoplastic Transformation of Proadipocytes and Loss of
Ability to Go Arrest and Differentiate. Table 4 presents the
results of studies to determine whether transformation of proad
ipocytes is associated with loss of the ability to growth arrest
at the Go state (Table 4, Left Columns A and B) and to
differentiate (Table 4, Right columns A and B) into mature fat
cells. The data show that all transformed proadipocyte clones
have lost the potential to arrest at GD when assayed with 2
different GD arrest assay methods. Further, Table 4 shows that
adipocyte differentiation cannot be induced in any of the trans
formed proadipocyte clones. Because it is possible that cells
5142
Table 3
Effect of growth factor deprivation and isoleucine deprivation on the cell cycle
distribution of nontransformed and transformed proadipocytes
Cell cycle distribution
Cell type and growth states
G,
S
(%)
G2-
Nontransformed3T3TRGaSAIA3T3RQSAIATransformed,
methodT2RGSAIAT3RGSAIAT8RGSAIATransformed,
spontaneous
methodM2RGSAIAM3RGSAIAM4RGSAIAM6RGSAIAM10RGSAIA5592804893863950644059714359
SST
RG. rapidly growing; SA, serum arrested; IA, isoleucine arrested.
which did not express morphological differentiation might still
express evidence of enzymatic differentiation, we tested each
transformed clone for its ability to differentiate, as assayed
enzymatically by an increase in the activity of glycerol-3-phosphate dehydrogenase. Table 4 (Column C) shows that, whereas
nontransformed 3T3 T-proadipocytes show a marked increase
in activity, no detectable increase in enzyme activity was ob
served in any of the transformed proadipocyte clones.
Effect of Neoplastic Transformation on Ability of 3T3 Cells
to Gs, GN, and GD Arrest and Differentiate. Table 5 summa
rizes the results of studies which compared the ability of 3T3
cells, methylcholanthrene-transformed
3T3 cells (MCA3T3),
and SV40-transformed 3T3 cells (SV3T3) to arrest at Gs, GN,
and GD and to differentiate. The results show that neither
initiated 3T3 cells nor transformed MCA3T3 or SV3T3 cells
can arrest at GD and differentiate. The data, however, show
that 3T3 cells can Gìgrowth arrest at Gs when cultured in
serum-deficient medium (they also contact inhibit at Gs when
cultured ¡nmedium containing 10% calf serum). In addition,
3T3 cells are shown to arrest at GN in isoleucine-deficient
medium. By contrast, MCA3T3 and SV3T3 cells show signifi-
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1982 American Association for Cancer Research.
VOL.
42
Proliferation-Differentiation
Defects in Neoplasia
Table 4
Defective Ga arrest and differentiation
potential in neoplastically
Go characteristics (% of la
beled nuclei)0
transformed proadipocytes
Adipocyte differentiation
Cell type
Nontransformed
3T3T
95
40 ±51-
150
90
241.2 ±5.3
Transformed, spontaneous
method
T2T3T8Transformed,
methodM2M3M4M6M10134321455465706700000a0000000000000000000004.5
SST
±3.0
±0000.50.2
Go arrest was determined by 2 methods: Method A, by the loss of the ability of Go-arrested cells to
proliferate at high density when cultured in DMEM-30% PCS + 50 /ig insulin per ml. Data are presented as
the percentage of cells which could not be restimulated to grow when refed this medium after high-density
arrest; Method B, by mitogenic responsiveness of Go-arrested cells when stimulated with 3-methyl-1isobutyl xanthine (5 x 10"* M). The extent of proliferation is given as the percentage of cells containing
nuclei labeled with [3H]thymidine (5 fiCi/ml) during a 48-hr interval following drug addition.
" Method A, morphological differentiation at confluence in medium containing 30% PCS + 50 /¿g
insulin per
ml (differentiated foci/25 sq cm); Method B, morphological differentiation at low density in heparinized
medium containing 25% human plasma (percentage of differentiated cells); Method C. enzymatic differen
tiation, glycerol-3-phosphate dehydrogenase activity, of cells cultured at low density in heparinized medium
containing 25% human plasma for 5 days (¿imolNADH oxidized per min per mg protein x 10~2 in treated
minus untreated cells). The sensitivity of this latter assay permits detection of 400 pmol NADH oxidized per
min per mg protein.
0 Mean ±S.E.
0 —¿,
not determined.
Table 5
Potential of chemically and virally transformed 3T3 T-cells to differentiate and
control cell cycle-dependent proliferation
Go arrest is expressed as the percentage of Gt arrested cells which showed
mitogenic responsiveness to 3-methyl-1-isobutyl
xanthine when cultured in hep
arinized DMEM containing 25% human plasma. Differentiation was scored mor
phologically. GS arrest is expressed as the increase in the percentage of G, cells
in the population following culture in DMEM-0.5% PCS for 5 days relative to that
observed in rapidly growing cells. GN arrest is similarly expressed as the increase
in the percentage of G, cells in the population following culture in isoleucinedeficient DM EM-10% dialyzed PCS for 3 days relative to that observed in rapidly
growing cells. All cell cycle analyses were performed by flow microfluorimetry.
tion000Go
arrest000Gs
arrest462213GN arrest38282
3T3MCA3T3SV3T3Differentia
cant defects ¡ntheir ability to G, growth arrest at Gs following
culture in serum-deficient medium and they do not contact
inhibit (data not shown). The defects in the ability to Gs arrest,
in general, are much more pronounced in SV3T3 cells than in
MCA3T3 cells. MCA3T3 and SV3T3 cells also show defects in
their ability to GN arrest. As with respect to GN arrest potential,
SV3T3 cells show a much more pronounced defect in the
ability to GN arrest than do MCA3T3 cells.
DISCUSSION
Numerous attempts to establish the mechanisms of carcinogenesis have focused on the study of cells grown in tissue
culture. These include cell lines derived from BALB/c (8, 21),
Swiss (25), AKR (20), and C3H (5) mouse embryos. Typically,
density-inhibited cell lines which do not grow in soft agar and
which do not produce tumors in syngeneic animals when
injected in suspension have been used as the prototype of
DECEMBER 1982
nontransformed cells (1, 2, 9, 20, 23, 25, 34, 36, 37). However,
BALB/c 3T3 and many other mouse embryo cell lines are not
normal cells. They show karyotypic abnormalities, they are
tumorigenic when implanted on plastic plates in syngeneic
mice (6, 30), and they can grow in soft agar under certain
culture conditions (35). In fact, it has been suggested that most
established mouse embryo cell lines represent initiated cells
(6, 30) with respect to the 2-stage model of carcinogenesis.
According to this model of carcinogenesis,
the state of
initiation can be induced by a variety of physical and chemical
agents and can be maintained for an extended interval of time
without overt expression of the neoplastic phenotype. Subse
quently, initiated cells can be induced to express the trans
formed phenotype if exposed to certain noncarcinogenic tu
mor-promoting agents, such as phorbol myristate acetate. Pro
motion, the second step in carcinogenesis, is therefore defined
as a process whereby populations of initiated cells are induced
to expand and form neoplasms. This 2-stage model of carci
nogenesis has been demonstrated in many in vivo studies (4),
and also in vitro (13).
Since many previous studies have been performed on mouse
embryo cell lines which appear to be initiated, it is difficult to
interpret the conclusion of previous studies with respect to
mechanisms of carcinogenesis. In addition, since such cell
lines also lack the capacity to differentiate, which is a universal
characteristic of normal stem cells which serve as the target
for most carcinogenic agents, it is difficult to predict if the
results of previous studies on growth control abnormalities
induced in cells during transformation are of physiological
significance.
It is our contention that in vitro studies on carcinogenesis
should use cells which are not initiated and which can differ5143
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1982 American Association for Cancer Research.
J. J. Wille, Jr. et al.
Table 6
Summary of the potential of noninitiated, initiated, and transformed cells to regulate cell cycle-dependent
growth arrest and differentiation
Growth arrest and differentiation
potentialities
arrestHigh
line3T3T3T3M
Cell
tionNoninitiated
Go arrest
Lowdensity
arrest+
+Initiated
+
—¿Transformed
-TransformedTransformed
TMCA3T3SV3T3Differentia-Phenotype
and T 3T3
++
density
GN
+
++
+±
+-Gs
entiate under physiological conditions, such as, the proadipocyte cell line we used. Even though 3T3 T-proadipocytes do
show karyotypic abnormalities, they are not initiated (30), and
they share many characteristics in common with normal stem
cells in that they can differentiate, they can remain quiescent
in the Gìphase of the cell cycle, and they can proliferate (32).
With this system, we have previously reported that the inte
grated physiological control of cell proliferation and differentia
tion is mediated at the G0 state (32), and we have suggested
that growth regulation mediated at other G, arrest states,
designated Gs and GN, represents auxiliary growth control
mechanisms which limit cell growth under stressful environ
mental conditions, such as those resulting from the lack of
growth factors and nutrients, respectively.
Experiments reported in this paper have therefore compared
the cell cycle-dependent control processes in noninitiated 3T3
T-proadipocytes,
in initiated 3T3 cells, and in a variety of
transformed 3T3 T-proadipocytes and 3T3 cells. In particular,
we compared the ability of these cell lines to: (a) arrest at Go
and differentiate; (b) arrest at Gs at low density following serum
deprivation or at high density following contact inhibition; or
(c) arrest at GN following nutrient deprivation. Table 6 sum
marizes the results. It shows that noninitiated 3T3 T-proadi
pocytes can arrest at Gs and GN and can arrest at GD and
differentiate. By contrast, initiated 3T3 cells lack the capacity
to arrest at GD and differentiate but can arrest at Gs and GN.
Further, the results show that transformed cells also lack the
capacity to arrest at GD and differentiate and in addition show
variable defects in their ability to arrest at Gs and GN- For
example, transformed M- and T-clones of 3T3 T-proadipocytes
can arrest at Gs and GN at low density when deprived of growth
factors and nutrients, respectively, but are able to overcome
growth control mediated at Gs at high cell densities. Methylcholanthrene-transformed
3T3 cells show even more pro
nounced defects in that they cannot arrest at GD nor differen
tiate, they show a decreased capacity to arrest at Gs at low
density in serum-depleted medium, and they do not Gs arrest
at high density even though they do maintain the limited ability
to arrest at GN. Finally, SV40-transformed 3T3 cells lack the
capacity to arrest growth efficiently in G, under any of these
conditions, and they also fail to differentiate. We interpret these
data to suggest that neoplastic transformation is associated
with a primary defect in the control of cell proliferation and
differentiation which is mediated at the GD arrest state, and
with additional but variable defect in the control of cell prolif
eration which is mediated at the Gs arrest state.
It is our hypothesis that initiation of carcinogenesis may be
associated with induction of defects in the mechanism respon
5144
-
-
sible for the control of Go arrest and differentiation and that
promotion may be associated with induction of defects in
mechanisms governing growth control mediated at Gs. The
data in Table 6 support this hypothesis by showing that initiated
cells selectively lack the ability to arrest at GD and to differen
tiate and that their conversion to the fully transformed state
involves development of additional defects in the backup
growth control mechanisms at Gs.
If the results of subsequent studies substantiate this hypoth
esis, a potential explanation for the 2 stages of carcinogenesis
will be available. In particular, it would be possible to explain
how initiated cells can remain occult for long intervals and still
remain sensitive to the effects of tumor-promoting agents. That
is, if initiation results in the induction of defects in the integrated
control of cell proliferation and differentiation at a GD-like state
while not affecting the backup growth control mechanisms
mediated at Gs and GN, initiated cells could remain quiescent
for long intervals. Such cells could subsequently be induced to
express the transformed phenotype and become tumorigenic
when exposed to promoting agents which cause initiated cells
to overcome backup growth-regulatory processes.
ACKNOWLEDGMENTS
The authors thank Karen Connelly for excellent secretarial assistance, Bryan
Hoerl for technical assistance in developing differentiation assays, and Michael
Zschunke for conducting the enzyme assays.
REFERENCES
1. Aaronson, S. S.. and Todaro, G. H. Basis for the acquisition of malignant
potential by mouse cells cultivated in vitro. Science (Wash. D. C.). 162:
1024-1026.
1968.
2. Altman, P. L., and Katz, D. D. Cell biology. Vol. 1. pp. 11-23. Bethesda.
Md.: Federation of American Societies for Experimental Biology, 1976.
3. Benz, E. W., Jr., Getz, M. J., Wells, D. J.. and Moses. H. L. Nuclear RNA
polymerase activities and poly(A)-containing mRNA accumulation in cultured
AKR mouse embryo cells stimulated to proliferate. Exp. Cell Res.. 108:
157-165, 1977.
4. Berenblum, I. Aspects of chemical carcinogenesis: skin. In: F. F. Becker
(ed.), Cancer, Ed. 2, Vol. 1, pp. 451-472. New York: Plenum Press, 1982.
5. Bertram, J. S. Effects of serum concentration on the expression of carcino
gen-induced transformation in the C3H/10T'/2 CL8 cell line. Cancer Res..
37: 514-523, 1977.
6. Boone, C. W., and Jacobs, J. B. Sarcomas routinely produced from putatively nontumorigenic Balb/3T3 and C3H/10T'/2 cells by subcutaneous
inoculation attached to plastic plates. J. Supramol. Struct.. 5: 131-137,
1976.
7. Boone, C. W.. and Scott, R. E. Plate-induced tumors of BALB/3T3 cells
exhibiting foci of differentiation into pericytes, chondrocytes, and fibroblasts.
J. Supramol. Struct., 14: 233-240. 1980.
8. Brown, K. D., and Holley. R. W. Epidermal growth factor and the control of
proliferation of Balb 3T3 and benzo(a)pyrene-transformed
Balb 3T3 cells.
J. Cell. Physiol., 100: 139-146, 1979.
9. Butel, J. S., and Estes, M. K. Properties of cells transformed by DNA tumor
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1982 American Association for Cancer Research.
VOL.
42
Proliferation-Differentiation
viruses. In Vitro (Rockville), 11: 142-150, 1975.
10. Clauson, O. P. F. Regenerative proliferation of mouse epidermal cells
following application of a skin irritant (cantharidin). Cell Tissue «¡net.,12:
135-144, 1979.
11. Diamond, L., O'Brien, T. G.. and Povera, G. Inhibition of adipose conversion
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
of 3T3 fibroblasts by tumor promoters. Nature (Lond.). 269: 247-249,
1977.
Epifanova, D. I., and Terskikh, V. V. On the resting periods of the cell life
cycle. Cell Tissue Kinet.. 2: 75-93, 1969.
Farber, E. Chemical carcinogenesis. Am. J. Pathol., 706. 271-296, 1982.
Florine, D. L., Hoerl. B. J., and Scott, R. E. Control of cell proliferation and
differentiation at a distinct G, arrest state: GD. Fed. Proc., 40: 3173, 1981.
Ham, R. G., and Veomett. M. J. Mechanisms of Development, pp. 317-428.
St. Louis. Mo.: C. V. Mosby Co., 1980.
Iverson, O. H., Bjerkness, R., and Devik, F. Kinetics of cell renewal, cell
migration and cell loss in the hairless mouse dorsal epidermis. Cell Tissue
Kinet., 1: 351-367, 1968.
Kozak, L. P., and Jensen, J. T. Genetic and developmental control of multiple
forms of L-glycerol-3-phosphatedehydrogenase.
J. Biol. Chem.,249. 77757781, 1976.
Krawisz, B. R., and Scott, R. E. Coupling of proadipocyte growth arrest and
differentiation. I. Induction by heparinized medium containing human plasma.
J. Cell Biol., 94: 394-399, 1982.
Lajtha, L. G., Lord, B. I.. Dexter, T. M., Wright, E. G., and Allen, T. A.
Interrelationship of differentiation and proliferation control in hematopoietic
stem cells. In: G. F. Saunders (ed.). Cell Differentiation and Neoplasia, pp.
179-209. New York: Raven Press. 1978.
Moses, H. L., Proper, J. A.. Volkenant. M. E., Wells, D. J., and Getz. M. J.
Mechanisms of growth arrest of transformed cells in culture. Cancer Res.,
38. 2807-2812,
1978.
Nicholson, G. L. Trans-membrane control of the receptors on normal and
tumor cells. II. Surface changes associated with transformation and malig
nancy. Biochim. Biophys. Acta, 458. 1-72, 1976.
Pairault, J., and Green, H. A study of the adipose conversion of suspended
3T3 cells by using glycerophosphate
dehydrogenase as differentiation
marker. Proc. Nati. Acad. Sei. U. S. A., 76. 5138-5142,
1979.
Paranjpe, M. S., Eaton, S. D. A., and Boone, C. W. Neoplasms produced
from C3H/10TVÕ, cells attached to plastic plates: saturation density, anchor
age dependence and serum requirement of in vitro lines correlates with
growth aggressiveness in vitro. J. Cell. Physiol., 96. 63-72, 1978.
Pardee. A. B., Dubrow, R., Hamlin, J. L.. and Kletzien, R. F. Animal cell
cycle. Annu. Rev. Biochem.. 47: 715-750, 1978.
Paul, D., Henahan, M., and Walter, S. Changes in growth control and growth
requirement associated with neoplastic transformation in vitro. J. Nati.
Defects in Neoplasia
Cancer Inst. 53: 1499-1503,
1974.
26. Pierce, G. B. Differentiation of normal and malignant cells. Fed. Proc., 29.
1248-1254, 1970.
27. Pledger, W. J., Stiles, C. D., Antoniades, H. N., and Sehers. C. D. Induction
of DNA synthesis in Balb/3T3 cells by serum components: regulation of the
commitment process. Proc. Nati. Acad. Sei. U. S. A., 74:4481 -4485, 1977.
28. Rosendaal, M., Hodgson. G. S., and Bradley. T. R. Organization of haemopoietic stem cells: the generation-age hypothesis. Cell Tissue Kinet.. 12:
17-29, 1979.
29. Scott, R. E.. Boman, B. M.. Swartzendruber, D. E., Zschunke, M. A., and
Hoerl, B. J. Cell cycle-associated modulation in cAMP-dependent plasma
membrane phosphorylation. Exp. Cell Res.. 733. 73-82. 1981.
30. Scott, R. E., and Boone, C. W. Promotion of smooth surface tumorigenesis
by phorbol myristate acetate in Balb/3T3 cells and Balb/3T3 T proadipocytes. J. Nati. Cancer Inst.. 66. 733-736. 1981.
31. Scott, R. E., and Florine, D. L. Reciprocal coupling of cell growth and
differentiation during the cell cycle. Fed. Proc.. 39. 3200, 1980.
32. Scott, R. E., Fiorine. D. L., Wille, J. J., Jr., and Yun, K. Coupling of growth
arrest and differentiation at a distinct state in the Gìphase of the cell cycle:
Go Proc. Nati. Acad. Sei. U. S. A., 79. 845-849, 1982.
33. Scott, R. E., Hoerl, B. J., Wille, J. J., Jr., Florine, D. L., Krawisz, B. R.. and
Yun, K. Coupling of proadipocyte growth arrest and differentiation. II. A cell
cycle model for the physiological control of cell proliferation. J. Cell Biol.,
94: 400-405, 1982.
34. Stiles, C. D., Desmond. W., Chuman, L. M., Sata, G., and Sauer, M. H.. Jr.
Growth control of heterologous tissue culture cells in the congenially
athymic nude mouse. Cancer Res., 36. 1353-1360.
1976.
35. Todaro, G. J., DeLarco. J. E., Marquardt. H.. Bryant. M. L., Sherwin, S. A.,
and Sliski, A. H. Polypeptide growth factors produced by tumor cells: a
possible growth advantage for the producer cells. In: G. Sato and R. Ross
(eds.). Cold Spring Harbor Conference on Cell Proliferation. VI. Hormones
and Cell Culture, pp. 113-127. Cold Spring Harbor, N. Y.: Cold Spring
Harbor Laboratory, 1979.
36. Todaro, G. J.. Lazar, G. K., and Green, H. The initiation of cell division in a
contact-inhibited mammalian cell line. J. Cell. Physiol.. 66 325-334, 1966.
37. Whittenberger, B., and Glaser, L. Inhibition of DNA synthesis in cultures of
3T3 cells by isolated surface membranes. Proc. Nati. Acad. Sei. U. S. A.,
74: 2251-2255,
1977.
38 Wille, J. J., Jr.. and Scott, R. E. Topography of the predifferentiation
Go
growth arrest state relative to other growth arrest states in the Gìphase of
the cell cycle. J. Cell. Physiol., 772: 115-122, 1982.
39. Yen, A., and Pardee, A. B. Arrested states produced by isoleucine depriva
tion and their relationship to the low serum produced arrested state in Swiss
3T3 cells. Exp. Cell Res., 114: 389-395, 1978.
1
Figs. 1 and 2. Morphological comparison of nontransformed 3T3 cells and 3T3 T-proadipocytes (Fig. 1) with transformed clones of 3T3 T-proadipocytes (Fig. 2).
Phase photomicrographs showing typical appearance of cultures at their respective saturation densities. Fig. 1: A. 3T3; ß,3T3 T; Fig. 2: A. M2: B. M3; C, M4; D, M6;
E. M10; F. T2; G, T3; H, T8. x 280.
DECEMBER
1982
5145
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1982 American Association for Cancer Research.
i
J. J. Wille, Jr. et al.
5146
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1982 American Association for Cancer Research.
VOL. 42
Neoplastic Transformation and Defective Control of Cell
Proliferation and Differentiation
John J. Wille, Jr., Peter B. Maercklein and Robert E. Scott
Cancer Res 1982;42:5139-5146.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/42/12/5139
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. © 1982 American Association for Cancer Research.