(CANCER RESEARCH 45, 1516-1524,
April 1985]
Quantitation of the Rate of Spontaneous Generation and Carcinogen-induced
Frequency of Anchorage-independent Variants of Rat Trachea! Epithelial
Cells in Culture
David G. Thomassen,1 Paul Nettesheim, Thomas E. Gray, and J. Carl Barrett2
Epithelial Carcinogenesis Group, Laboratory of Pulmonary Pathobiology, National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina 27709
ABSTRACT
plastic development of any given cell type must be examined,
because exceptions to the correlation between anchorage inde
pendence and tumorigenicity do exist, and examples of anchor
age-independent, nontumorigenic variants (11, 38), as well as
anchorage-dependent, tumorigenic variants (12, 24, 36) have
The rate of spontaneous generation and the frequency of
carcinogen-induced anchorage-independent variants of preneoplastic rat trachéalepithelial (RTE) cells in culture were quantitated. Anchorage-independent variants of different RTE cell lines
arose spontaneously by a stochastic process at rates of 0.5 x
10~4 to 5.4 x 1CT4 variants/cell/generation, as determined by
been described.
Studies in our laboratory have been concerned with the de
scription and analysis of neoplastic development in trachéal
epithelial cells (for review, see Ref. 27). We have previously
described an epithelial cell culture system which uses RTE3 cells
fluctuation analyses. These variants were also induced by the
mutagen A/-methyl-A/'-nitro-A/-nitrosoguanidine with a frequency
of ~1(T3 variants/surviving cell. The rates of spontaneous
for quantitating the induction of variants having an enhanced in
vitro growth potential following exposure to carcinogens (15, 30,
39). These enhanced growth variants become immortal, anchor
age independent, and finally neoplastic, forming adeno- or squamous cell carcinomas upon injection into compatible host animals
(21, 22, 30, 39, 40). In the RTE cell system, as was found in
studies with fibroblasts, anchorage independence also is com
monly associated with neoplastic potential, or at least with a late
preneoplastic stage (21, 22, 30,40). We have recently described
the quantitation of the induction of the early transformed phenotype of RTE cells (30,39) but the evolution of late preneoplas
tic and neoplastic variants of RTE cells has not been quantitated.
In this report we describe studies with preneoplastic variants
of RTE cells to determine the spontaneous rate and carcinogeninduced frequency of change from anchorage dependence to
anchorage independence. Our results, the first using epithelial
cells in culture, are compared with the available quantitative
studies in the literature using fibroblasts in culture.
change and the frequencies of induction are the first reported
for epithelial cells and are similar to some, although not all, rates
and frequencies of change to anchorage independence for fibroblast-like cells in culture. In addition, these rates and frequencies
are similar to those for mutations at some known gene loci. The
induced frequency of this late change in neoplastic progression
is, however, considerably lower than the frequency of induction
of the initial, preneoplastic changes in RTE cells in culture (~3 x
10~2/surviving cell). These quantitative determinations are useful
in defining the mechanisms of late changes occurring during the
progression of RTE cells to the neoplastic state.
INTRODUCTION
There is considerable evidence that neoplastic transformation
is a multistep process (3, 14, 32), but definitive information on
the number and nature of the cellular changes involved is lacking.
Quantitative studies on the transitions between normal and
preneoplastic cells and between preneoplastic and neoplastic
cells can provide information which is useful in identifying and
characterizing cellular changes involved in neoplastic transfor
mation. Using cell culture systems, it has been possible to
quantitate the induction of early, preneoplastic changes in normal
cells (5, 6, 17, 30, 39), as well as the progression of cells from
the preneoplastic to the neoplastic state (4, 8,11, 34, 38).
Studies carried out in fibroblast culture systems have identified
anchorage-independent growth as a common feature of many
neoplastic cells (2,10,11), leading to its use as an in vitro marker
of neoplastic growth potential. Although studies examining the
change from anchorage dependence to anchorage independence
have been mostly qualitative, a number of quantitative studies
on the inducibility or the spontaneous rate of change to anchor
age independence have been reported (7, 8,11, 23, 33, 34, 37,
38). However, the role of anchorage independence in the neo1Recipient of National Research Award
1 F32 ESO5224-01
MATERIALS AND METHODS
Cell Culture Methods. RTE cells were obtained from 8-week-old male
Fischer 344 rats (specific pathogen free), as described previously (15).
Preneoplastic RTE cell lines, termed EGVs, were isolated from carcino
gen-treated or control cultures of RTE cells (Table 1) as large colonies
of small, tightly packed, proliferating epithelial cells having increased
nuclearcytoplasmic
ratios (38). These variants proliferate in vitro under
conditions nonpermissive for normal RTE cells, and can progress in vitro
to cells capable of forming carcinomas when injected into suitable hosts.
Anchorage-independent clones of EGVs (Table 1) were established by
isolating colonies from agarose. All EGV cell lines were maintained in
culture as previously described (15, 39), using Ham's F-12 medium
(GIBCO), supplemented with 5% fetal bovine serum (GIBCO or Micro
biological Associates, Walkersville, MD), insulin (1.0 /ig/ml; Sigma Chem
ical Co., St. Louis, MO), hydrocortisone(0.1 Mg/ml; Sigma Chemical Co.),
penicillin (100 units/ml), and streptomycin (100 f/g/ml; GIBCO) at 37°C
in a humid air atmosphere containing 5% CO2.
Assays for Anchorage-independent
Growth. Tests for anchorage
from the NIH.
Present address: Laboratory of Experimental Pathology, Division of Cancer Etiol
ogy, National Cancer Institute, Frederick, MD 21701.
2 To whom requests for reprints should be addressed.
3The abbreviations used are; RTE, rat trachéalepithelial; EGV, enhanced growth
variant; CFE, colony-forming efficiency; MNNG, N-methyl-W'-nitro-N-nitrosoguani-
Received 7/20/84; revised 12/11/84; accepted 12/28/84.
CANCER
dine; GIBCO, Grand Island Biological Co., Grand Island, NY.
RESEARCH
VOL.
45 APRIL
1985
1516
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1985 American Association for Cancer Research.
ANCHORAGE-INDEPENDENT
VARIANTS
Table1
Origins and CFEsin agarose of EGVsand anchorage-independentvariants of rat
trachea! epithelial cells
OF TRACHEAL
EPITHELIAL
CELLS
(15, 38). Surviving cells were allowed to form colonies of >100 cells, and
cells were replated into agarose medium as above. The total number of
anchorage-independent
variants in treated and control cultures was
in
agarose"
lineEGV-1EGV-1
Cell
(0.01xg/mlfSubclone
by MNNG
clone19EGV-3EGV-6EGV-9EGV-1
EGV-1Induced
of
determined by correcting the number of colonies observed for the
vivo tumorigenicity
recovery efficiency of anchorage-independent cells in agarose. The fre
test"NontumorigenicNontumorigenicKeratinic
(%)<0.01<0.01<0.01<0.01<0.01<0.011-55-305-10In
quency of variants induced by MNNG was calculated by subtracting the
corrected frequency of colonies observed in agarose without treatment
from the corrected frequency observed following MNNG treatment.
cysts'*Squamous
rads)Induced
by 7-ray (150
Mg/ml)Induced
by MNNG (0.1
cellcarcinoma8NontumorigenicKeratinic
RESULTS
^g/ml)Spontaneous
by MNNG (0.1
18EGV-1
transformantClone
A1EGV-9A1EGV-6A1OriginInduced
isolatedfrom
of EGV-1
cystsKeratinic
cystsCystsSquamous
agaroseClone
isolatedfrom
of EGV-9
agaroseClone
of EGV-6 isolated
cell
from agaroseCFE
carcinoma
a At early passages (< passage 10).
6 Six passages after isolation cells were injected s.c. into nude mice at 2 x 10"
cells/site. Nontumorigenic,no visible lesion at site of injection up to 3 months after
injection.
c MNNG treatment for 4 h (39).
d Benign cysts formed at site of injection.
e Latency period of 12 weeks.
' Latency period of 4 weeks.
independence
were done in 60-mm-diameter
gridded
tissue culture
dishes (Costar, Cambridge, MA), using SeaPlaque agarose (Marine Col
loids, Rockland, ME), as previously described (38). Agarose top layers
of 0.26 to 0.3% were used with bottom layers of 0.6% agar. The
antimycotic agent Fungizone (0.25 ¿/g/ml;GIBCO) was included in the
medium in some experiments. Tests for anchorage independence were
always performed using <105 cells/60 mm dish, and were scored after
2 weeks. Only colonies of 240 cells were scored.
Determination of Spontaneous Rate of Change from Anchorage
Dependence to Anchorage Independence. The rate of spontaneous
change to anchorage independence was determined by a Luria-Delbrück
fluctuation analysis (19), based on the occurrence of cells with the ability
to grow in agarose in a series of parallel cultures established from
anchorage-dependent
cell populations. To ensure that no anchorageindependent cells were included in initial anchorage-dependent popula
tions, EGV cell lines having very low CFEs in agarose (<0.01%) were
used, and approximately 1 cell/well was plated in microtiter wells (Lux
Scientific Corp., Newbury Park, CA). Colonies were expanded to 10* to
107 cells and up to 106 cells were tested for colony formation in agarose
using <105 cells/dish. The spontaneous rate of change to anchorage
independence, n (variants/cell/generation),
was estimated using the for
mula derived from the work of Luria and Delbrück(19) by Kondo ef al.
(16),
M-AHn(3.46
Optimization of Conditions for Anchorage-independent
Growth Assays. Colony formation by anchorage-independent
cells can be affected by the type and concentration of semisolid
medium used, by cell density, by additives to the medium, and
by the serum concentration (18, 25, 26, 28, 29, 31). We chose
to optimize conditions to obtain the maximum CFE and colony
size of anchorage-independent variants without affecting colony
formation by anchorage-dependent cells. The concentration of
agarose over the range of 0.24 to 0.35% in the top layer (cell
suspension) was examined for its effect on the CFE and colony
size of anchorage-independent RTE cells. Colony size at 2 weeks
increased with decreasing concentrations of agarose; CFEs var
ied <2-fold from 0.24 to 0.3% agarose with a maximum at
0.26%, while at concentrations higher than 0.37, a decrease in
CFE (5-fold) was observed (data not shown). We used 0.26 or
0.3% top agarose in our experiments. Colonies increased in size
from 2 to 3 weeks, but a constant number of colonies was
observed after 2 weeks, at which time all the cultures were
scored.
To maximize the number of cells which could be tested per
dish without affecting the CFE of anchorage-independent var
iants, we examined cell density effects on growth in agarose.
Reconstruction experiments were performed using constant
numbers of the anchorage-independent variants EGV-1 A1 or
EGV-9A1, mixed with variable numbers (0 to 5 x 105) of anchor
age-dependent EGV-1 clone 19 cells per dish (Table 2). CFEs
varied <2-fold up to 105 cells/dish, while 3- to 4-fold decreases
Table2
EHectof cell density on colony formation in agarose
Cells were plated as indicated in agarose medium containing 0.3% agarose (top
layer), 20% fetal bovine serum, 1.0 ^g insulin/ml, and 0.1 /*g hydrocortisone/ml.
pendent
EGV-1clone
Anchorage-inde
19 cells/
pendent
cellsEGV-1
dish03x
A1CEGV-9A1"Anchorage-de
1031
10"3x
x
10"1
10s2x10»03x
x
= M-ln 2,
in which C is the number of parallel clonal populations whose cells were
tested for growth in agarose, and N is the average total number of cells
per expanded clonal populations (16, 19). The average total number of
anchorage-independent cells per expanded clonal population (M) was
calculated from the equation
M
No. of colonies observed in agarose
±2.918.0
±3.66.0
0.720.0
±
Three dishes/group.
6 CFE relative to value with no anchorage-dependentcells present.
0 2000 cells/dish.
" Mean ±SO.
" 500 cells/dish.
Variants
Induced by MNNG. EG variants were plated in plastic tissue culture
dishes and were treated with MNNG in 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid-buffered F12 medium for 4 h as previously described
CANCER
2.07.018.0
±
0.717.6
±
1031
±2.623.6
X1043x
5.531.0
±
1041
±2.532.0
1052x
x
6.17.0
±
105Colonies/dish"8.0
±1.0Relative
CFE in agarose of cells from isolated colonies/fraction
of total cells of each sample tested in agarose
Estimation of the Frequency of Anchorage-independent
CFE61.001.000.882.252.250.
±1.1"8.0
RESEARCH
VOL. 45 APRIL 1985
1517
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1985 American Association for Cancer Research.
ANCHORAGE-INDEPENDENT
VARIANTS
were seen at 2 x 105 cells/dish. At 5 x 105 cells/dish there were
too many cells in the culture to score the colonies. Thus, all
subsequent experiments were done using <105 cells/60-mm
OF TRACHEAL
Clone8EGV-1
fold increases in CFE were observed at 20% versus 5% serum
for both cell lines, with 20% serum always yielding the highest
CFE. We therefore used 20% serum in our agarose medium.
Our optimized medium (20% serum, 1.0 /zg insulin/ml, 0.1 /¿g
hydrocortisone/ml, 0.26% agarose) did not support colony for
mation by anchorage-dependent cells. No colonies formed from
a total of >5 x 105 primary RTE cells, or from 5x10* cells from
each of 15 freshly isolated subclones of EGV-1 cells (7.5 x 10s
total cells).
Characterization
of Anchorage-independent
Variants of
EGV Cell Lines. Individual colonies of different EGVs which
formed in agarose were isolated and characterized for their ability
to grow in agarose (Table 4). All the anchorage-independent
variants had an increased CFE in agarose compared to the
parental cells; however, the CFEs were relatively low (<10%),
and in some cases were highly variable. For example, the aver
age CFE in agarose for anchorage-independent variants of EGV1 cells was 3.6%, but the CFE for different agarose clones
ranged from 0.05 to 8.0%, a 160-fold difference. This repre
sented the extreme case, but demonstrates the heterogeneity
of the response for this phenotype. This observation does not
Table3
Effect o' insulin and hydrocortisone and serum concentration on colony formation
on plastic and in agarose
Cells were plated as indicated in 0.3% agarose.
CFE in agarose of dif
in agarose
(%)2.70.058.0 ferent (%f3.6
subclones
A1EGV-1
A3EGV-1
A4
EGV-1
A5EGV-1
A6EGV-1
A7EGV-6A1EGV-6A2
decrease in the CFE of EGVs on plastic (Table 3) or primary RTE
cells on feeder cells (15), depending on the serum concentration.
Anchorage-independent EGV cells failed to grow in agarose if
hydrocortisone was omitted from the medium (>100- to 1000-
agarose and on plastic (Table 3). Although CFEs on plastic
decreased 2- to 200-fold as the serum concentration increased
to 20%, the opposite effect was seen in agarose. Two- to 10-
CELLS
Table 4
CFEin agarose of anchorage-independentvariants of EGVs(recovery efficiencies)
dish.
Insulin and hydrocortisone were previously demonstrated to
be important growth factors for primary RTE cells (15). Omission
of either or both of these additives resulted in a 1.5- to 10-fold
fold decreases in CFE), even at high concentrations of serum
(Table 3). Insulin had no effect on colony formation in agarose
but, since it affected growth on plastic, we chose to use both
insulin and hydrocortisone in the growth medium.
We examined the effect of serum concentration over the range
of 5 to 20% on the CFE of EGV-1A1 and EGV-9A1 cells in
EPITHELIAL
EGV-6A3EGV-6A5EGV-1
18A1EGV-1
18A2
EGV-1
18A3EGV-1
18A4CFE
6.05.00.055.00.85
1.31.80.231.06
0.520.2Av.
3.22.2
±
±1.90.5
±0.4
8 Sets of clonal anchorage-independentvariants of EGV-1, EGV-6, EGV-118
cells were establishedfollowing isolation of individualcolonies from agarose.
" Recovery efficiency x 100 ±SD.
preclude meaningful quantitative measurements of the rate of
generation of the anchorage-independent phenotype but should
be kept in mind as a large source of variation in these experi
ments. With this caution in mind, estimates af the spontaneous
rate of generation of anchorage-independent variants of different
EGVs were made and corrected for the average recovery effi
ciency or CFE of the agarose clones in order to compare the
rates obtained from different EGVs to each other and to other
cell types.
Determination of Spontaneous Rate of Change from An
chorage Dependence to Anchorage Independence in Clones
of Preneoplastic RTE Cells. Fluctuation tests based upon LuriaDelbrückanalysis (19) were performed to determine the spon
taneous rate of change from anchorage dependence to anchor
age independence in clones of EGVs. Single-cell colonies of
EGVs were isolated to ensure that starting populations did not
include anchorage-independent cells. Any colonies originating
from anchorage-independent cells would have high CFEs in
agarose (>1%) in subsequent tests and would be easily identifi
able. Each culture was expanded to 5 x 104 to 5 x 106 cells,
and <106 cells of each culture were tested for colony formation
in agarose using <1 05 cells/60-mm-diameter dish. After 2 weeks,
the number of colonies per dish was determined for each culture.
To estimate the spontaneous rate of change to anchorage
independence, the total number of anchorage-independent cells
in each expanded culture (M) was calculated. The spontaneous
rate of change, M (variants/cell/generation), was then estimated
from the equation (16)
Anchorage-in
fetalbovineserum55101515205510152020Hydro-Insulin
of
cortisone(1.0
dependentcellsEGV-1A1EGV-9A1%
eg/
(0.1 »jg/ CFE in aga
onplastic
mi)
(%)+
ml)
rose
(%)17.02.215.010.04.84.418.04.825.01.5<0.1<0.1
M-AMn(3.46
= M-ln 2
0.4N.D8+ +
1.6+
0.97+
<0.003+
2.7+
A representative fluctuation test using EGV-1 cells to deter
mine the spontaneous rate of change from anchorage depend
ence to anchorage independence is shown in Table 5. The
number of cultures established (C) with ~1 cell/culture was 10,
and the cells in these cultures were expanded to 2.0 to 6.1 x
106 total cells. The average number of cells per culture (A/) was
4.23 x 106. From each original culture, 1 x 106 cells (10 dishes
at 105/dish) were plated in agarose, and the number of colonies
+
+
+
18N.D.-I- +
27+
19+
-I+
+
30<0.01CFE
in agarose was determined after 2 weeks. The total number of
colonies forming in agarose per cells sampled was determined
8 N.D., not determined.
CANCER
RESEARCH
VOL. 45 APRIL 1985
1518
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1985 American Association for Cancer Research.
ANCHORAGE-INDEPENDENT
VARIANTS
OF TRACHEAL
EPITHELIAL
CELLS
Table 5
Fluctuation test of EGV-1 (passage8) to determine the spontaneousrate of change from anchoragedependence to
anchorage independence
For calculation of spontaneous rate of change see Table 6, Experiment 1.
cell
col
vari-ance:mean1.21.630.931.41.01.20.68
cate
no./culture
in agarose/dish
onies in
colo
agarose87417766051178
factor63.637.334.893.964.005.024.18
cells/dish)3,6,6,7,7,7,7,8,8,150,1,2,4,5.54,5,7,8,9,10,
(10s
(x10»)3.634.404.403.962.004.524.18
Culture1234567
nies/culture2691253722382055326
10,10,132,2,3,5,7,7,7,8,9,100,0,
1,2,20,0,0,1,1,1,2,3,34,6,7,7,8,8,8,8,9,13
89
10Final
6.103.74
1,1,2,2,3,3,4,4,5,70,0,1,1,1,2,2,3,3,33216
5.37Colonies 0,1, 1,2,2,2,3,3,4Total
6.103.74
19560107Intrarepli1.090.86
18Sampling5.97Total
av. = 4.23 x
10«
cells/
culture (N)
0.75
176.6(av.)
s2 = 14,840
S2/X= 84
(interculture variance:mean ratio)
8 Ten cultures of 10scells were tested in agarose, but some dishes were lost due to contamination.
" Number of cells per culture divided by the number of cells tested for growth in agarose.
and multiplied by the sampling factor (number of cells per culture
divided by the number of cells tested) to determine the total
number of colonies per culture. The number of anchorageindependent cells per original culture (M) was then calculated
after correcting the number of colonies per culture for the recov
ery efficiency of anchorage-independent cells in agarose. For
example, the average CFE in agarose in subclones EGV-1 cells
isolated from agarose = 3.6% (Table 4); thus,
M = (colonies/sample
set) x (sampling factor)
x (recovery efficiency)"1 = 4905.
clones were expanded
4 x 106 cells) (Table
separate experiments
variant/cell/generation,
to different final cell numbers (5 x 104 to
6). The estimated rates of change in 6
ranged from 0.5 x 10~4 to 2.0 x 10~4
with an average of 0.95 ±0.5 x 10~4
variant/cell/generation. These results indicate that estimates of
n are not greatly affected by the degree of clonal expansion over
the range tested.
Fluctuation tests were also performed on EGV-6, EGV-118,
and EGV-3 cells (Table 7). Two tests were performed using EGV6 cells. In the first test, one set of 12 clones was expanded first
to ~5 x 105 cells, followed by testing of one-half of the cells in
The values of C (10), N (4.23 x 106) and M (4905) were then
used to estimate for EGV-1 cells the value of M = 0.85 x 10~4
agarose. The remaining cells were further expanded to 2.5 x
106 cells and retested for growth in agarose. The estimated rates
anchorage-independent variants/cell/generation.
This analysis can also be used to determine whether the
variants arise by a stochastic process, i.e., to distinguish be
tween the random generation of variants in our cultures and the
préexistenceof variants or their induction by the assay conditions
(19). As discussed later, this determination requires an exami
nation of the variance:mean ratios for the distribution of colonies
in agarose between samples from one culture (intrareplicate
ratio) and between cultures (interculture ratio). The ratio of the
variance to the mean between samples from the same culture
(intrareplicate variance:mean ratio) indicates the precision of our
assay conditions and sampling procedures. Since fluctuations in
these samples should be due only to random errors in sampling,
the variance:mean ratio should be ~1 according to the Poisson
distribution (19). For the experiment Illustrated in Table 5, the
intrareplicate variancermean ratios ranged from 0.68 to 1.63
(average, 1.07). By contrast, the interculture variance:mean ratio
was 84, indicating that the variability in the cultures was not due
to sampling but to the stochastic generation of anchorageindependent cells during the growth of the cultures (see "Discus
sion").
of change in this experiment were similar, regardless of when
the cells were assayed (~5 x 10~4variant/cell/generation) (Table
To determine the effect of varying the final cell number per
culture on the measurement of the spontaneous rate of change
to anchorage independence, a series of fluctuation tests was
performed using EGV-1 and EGV-1 clone 19 cells, in which
CANCER
RESEARCH
7). In the second experiment with EGV-6 cells, the estimated
rate was slightly lower (1.4x10~4).
One experiment was per
formed with spontaneous EGV-118 cells and a rate of 1.3 x
10~4 variant/cell/generation was observed.
The results with EGV-3 cells were different from those ob
tained with other cell lines. Although colonies clearly formed in
agarose, the cells from these colonies when isolated on plastic
had a very limited proliferative potential, and quickly enlarged
and ceased proliferating. Thus, a recovery efficiency for anchor
age-independent variants of EGV-3 cells could not be obtained.
We used a recovery efficiency (0.021) equal to the average of
those obtained for EGV-1, EGV-6, and EGV-118 cells (Table 4)
to estimate a spontaneous rate of 1.9 x 10~4 variant/cell/gen
eration for EGV-3.
We determined the interculture variance/mean ratios for each
experiment summarized in Tables 6 and 7 and observed values
ranging from 3.0 to 1261 (average value, 471.3; median value,
84). By contrast the average intrareplicate variance:mean ratios
(Tables 6 and 7) were close to 1, ranging from 0.72 to 1.9 (mean
value, 1.33) and were always 27 to 750-fold lower than their
corresponding interculture variance:mean ratios, providing evi
dence for the random generation of anchorage-independent var
iants in cultures of each of these EGV cell lines.
VOL. 45 APRIL 1985
1519
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1985 American Association for Cancer Research.
ANCHORAGE-INDEPENDENT
VARIANTS
OF TRACHEAL
EPITHELIAL
CELLS
Table6
Fluctuation tests with EGV-1 and EGV-1 clone 19 cells to determine the spontaneous rate of change from anchorage
dependence to anchorage independence
Ofcultures(C)1018103376Av.
Experi
cells/culture
total
ment1'^3"4"5"6"No.
(A/)4.23
10"1.8X
x
10*2.14
10"3x
x
10»1.8X
1055x
104Av.
age-inde
pendent
intra-replicatevariance:mean cell/gen
cells
peroriginalculture(Mf4905513021332617514Av.
eration(X10-*)Me0.852.00.810.540.5
sam
pling fac
tor84.88362.14211Recoveryefficiency^0.0360.0360.0360.0360.0360.036Anchorratio08473119.315.23.53.0Variants/
ratio1.07N.D.90.72N.A.N.A.N.A.Interculturevariance:mean
a Inverse of fraction of total cejls tested in agarose per culture.
0 CFE in agarose of cells from representative colonies isolated from agarose and grown out on plastic.
°M = (colonies/sample set) x (sampling factor) x (recovery efficiency)'1.
" Calculated after correction for sampling.
"Ã-Ã-W-ln
(3.46 H-N.C) = M-ln 2 from Ref. 16.
'Culturesfrom
EGV-1 cells, passage 8.
9 N.D., not determined; N.A., not applicable.
"Cultures from EGV-1 clone 19, a subclone of EGV-1 cells.
Tabte7
Fluctuation tests on EGV-6, EGV-118, and EGV-3 cells to determine spontaneous rate of change from anchorage
dependence to anchorage independence
intra-replicalevariancemean
cell/gen
sam
culturevariance:mean
eration(X10-4)
of cul
cells/
pling fac
ra
tor822.51.783.865.8Recovery
ratio"186126128036.9682Variants/
Ou)95.55.01.41.31.9"
efficiency60.0220.0220.066"0.0050.021"Anchorage-independentcells/original
lineEGV-6'EGV-6'EGV-6EGV-1
Cell
tures
(C)12128710Total
(N)5.28
culture
culture
(Mf3,89819,6002,0056,62013,919Av.
tio1.91.71.11.31.5Inter
1052.5x10«1.18X
x
18EGV-3No.
10"3.86
10"4.93
x
x 10«Av.
'' Inverse of fraction of total cells tested in agarose per culture.
" CFE in agarose of cells from representative clones isolated from agarose and grown out on plastic.
c M = (colonies/sample set) x (sampling factor) x (recovery efficiency)"1.
" Calculated after correction for sampling.
"jiAMn (3.46 /i-N-C) = M-ln 2.
' Same set of cultures sampled at 2 expansion levels.
a No recovery efficiency available. Assuming average of recoveries for anchorage-independent variants of EGV-1, EGV6, and EGV-118 cells.
h Experiment was performed at a different time and with a different lot of serum than the other experiments, and
recovery efficiencies of anchorage-independent EGV-6 cells were increased.
Determination of Frequency of Carcinogen-induced An
chorage-independent Variants. We determined the inducibility
of anchorage-independent variants of EGV-1 clone 19 and EGV6 cells by the carcinogen MNNG. To ensure that anchorageindependent variants were not selected for or against by MNNG
treatment, we measured the cytotoxic effect of MNNG on an
chorage-dependent EGV-1 clone 19 and EGV-6 cells and an
chorage-independent EGV-1 A1 and EGV-6A1 cells. Over a dose
range of 0.01 to 0.1 u.g MNNG/ml (70 to 7% relative survival),
the survival curves for the anchorage-dependent and -independ
ent cell lines were similar (data not shown). Thus, MNNG would
not be expected to select for anchorage-independent variants
frequency of induced variants was decreased to 5 x 10~4.
by differential survival.
MNNG treatment increased the frequency of anchorage-inde
pendent variants in cultures of EGV-1 clone 19 cells relative to
control cultures by 3- to 7-fold (Table 8). At doses of MNNG
differences, because the average is influenced most by cultures
with the highest frequencies of spontaneous variants which arise
due to the high rate of spontaneous change to anchorage
independence in these cells.
which were 50 to 60% cytotoxic, the induced frequency of
anchorage-independent variants was ~1 x 10~3/surviving cell
DISCUSSION
after correcting for the frequency of variants in control cultures.
At a much higher dose which resulted in 95% cytotoxicity, the
In this communication, a quantitative analysis of the appear
ance of anchorage-independent variants in epithelial cells is
CANCER
RESEARCH
Similarly, the induced frequency of anchorage-independent var
iants of EGV-6 cells was ~10~3/surviving cell. The absolute
frequencies of variants observed in control and treated cultures
of EGV-6 cells were greater than those seen in cultures of EGV1 clone 19 cells; however, after correction for the frequency of
variants in control cultures, the frequencies of induced anchor
age-independent variants of both cell lines were similar. Absolute
increases in the frequency of anchorage-independent variants of
EGV-6 cells, were observed in 5 of 8 clonal cultures treated with
MNNG, with observed increases of 3- to 60-fold. The average
increase (1 J-fold) does not reflect the degree of these individual
VOL. 45 APRIL 1985
1520
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1985 American Association for Cancer Research.
ANCHORAGE-INDEPENDENT
VARIANTS
OF TRACHEAL
EPITHELIAL
CELLS
Table 8
Induction of anchorage-independentvariants of EGV-1 clone 19 and EGV-6cells by MNNG
frequency
of anchorrected
ofcolonies/10"
no.
ofanchorage-inde age-inde
pendentvariants
cells
cells
Gig/
pendent var
survival*1.000.421.000.470.051.000.29Colonies/10s
efficiency00.0360.0360.0360.0360.0360.0660.066Cor
(X10-4)"9.612511.9
iants
(X10-4)4.414214716.728.6Induced
CellsEGV-1EGV-1EGV-6Experi
ment121MNNG
plateeT1.574.90.725.02.561118.9Recovery
ml)00.0300.0240.0700.024Relative
plated441362013971167286Frequency
" CFE on plastic in controls: 0.45 (Experiment 1) and 0.26 (Experiment 2).
" Cells were replated 5 days posttreatment when colonies contained 100 to 400 cells.
c CFE in agarose of cells from representative colonies that were isolated from agarose and grown out on plastic.
'' Frequency In treated cultures minus frequency in control cultures.
described. The induced frequency and spontaneous rate of
change from anchorage dependence to anchorage independence
was quantitated in clones of enhanced growth variants of RTE
cells. Fluctuation tests based upon Luria-Delbruck analysis (19)
were performed to make 2 critical determinations: (a) to deter
mine whether anchorage-independent variants of RTE cells arise
by a stochastic process; and (b) to estimate the rate of that
change. As described by Luna and Delbrück(19) and illustrated
by Crawford ef al. (11 ), if variants preexist in the cultures or are
induced by the assay conditions, then their distribution among
independent cultures should be Poisson, and the ratio of the
variance of the mean (interculture variancermean) will be ~1. In
contrast, if the variants arise by a stochastic process, they will
arise randomly in independent cultures; their distribution will have
a high variance and will not be Poisson, and the ratio of the
variance to the mean (interculture variancermean ratio) will be
much greater than 1. The interculture variance:mean ratios for
all fluctuation tests summarized in Tables 6 and 7 were much
greater than 1, suggesting that the anchorage-independent var
iants did in fact arise randomly and de novo in the various
cultures. In contrast, the intrareplicate variance:mean ratios for
those same experiments were close to 1, as expected for random
sampling from individual cultures. These results further support
the proposed random generation of variants in these experi
ments.
The estimated spontaneous rates of change to anchorage
independence ranged from 0.5 to 5.5 x 10~4 variants/cell/gen
eration, with an average of ~2 x 10~4. Several factors may
influence the estimation or reliability of these estimated values.
It is possible that anchorage-independent variants have a
growth advantage on plastic relative to anchorage-dependent
cells, so that increased clonal amplification results in unequal
amplification of any anchorage-independent variants present.
Such a growth advantage would result in inflated estimates of
the spontaneous rate of change to anchorage independence.
However, this does not seem to be the case at least for EGV-1
and EGV-6 cells, since 10-fold differences in the levels of clonal
amplification of these EGVs did not uniformly affect estimated
rates of change to anchorage independence.
Another determinant in these studies is the recovery efficien
cies of the cells, i.e., the CFE of anchorage-independent cells in
CANCER
RESEARCH
agarose. It is necessary to correct the number of colonies
observed in agarose by the efficiency with which the cells grow
in agarose, in order to compare the absolute rate of generation
of these altered cells in different cell populations. This corrects
for differences in assay conditions (for example, serum lots)
between different cell types, and hopefully yields insight into
underlying similarities or differences in the molecular mechanisms
of this phenotypic change. The recovery efficiencies for the EGVs
studied here (Table 4) were relatively low and variable compared
to other studies using fibroblast-like cells (7, 8,11, 23, 33, 34,
37, 38). Low recovery efficiencies result in large corrections for
estimates of total numbers of anchorage-independent variants,
and are a potentially large source of error. The uniformity in the
rates of change observed here in different experiments with the
same EGVs and for different EGVs could be due in part to the
large correction factors used or could be indicative of truly
uniform rates of change by different EGVs.
The variability observed in the recovery efficiencies should not
have a marked effect on estimates of rate, since the very low
values do not cause big decreases in the recovery efficiencies.
However, this variability does represent an important feature of
these cultures, which may relate to their potential to undergo
terminal differentiation (27, 39). As a population, each EGV
continues to be proliferative, while clonal isolates can have a
greatly reduced proliferative potential. An extreme example of
this phenomenon is seen with EGV-3 cells. Although EGV-3 cells
grew in agarose, the cells from these colonies when isolated had
a limited life span on plastic. The cells enlarged greatly, ceased
growing, and appeared squamous-like as though they were
terminally differentiated. Some of the anchorage-independent
variants from other EGVs having very low recovery efficiencies
may also represent cells having a high propensity to terminally
differentiate. Consistent with these findings is the differentiated
nature of tumors or cysts observed in vivo following injection of
these EGVs into animals (Table 1). These results should be
contrasted with results obtained in similar studies using fibroblasts, where little variability in recovery efficiencies is seen,
perhaps due to the lack of terminal differentiation in these
cultures.
It may be significant to note that the highest observed rate of
change to anchorage independence was with EGV-6 cells. This
VOL. 45 APRIL 1985
1521
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1985 American Association for Cancer Research.
ANCHORAGE-INDEPENDENT
VARIANTS
OF TRACHEAL
EPITHELIAL
CELLS
Table 9
Comparison of spontaneousrates and frequencies of carcinogen-inducedchange to anchorage independence
Spontaneous rate of change"
Cells
Semisolid medium
Corrected for
recovery0
Not correeled for
recovery"
Maximum frequency of car
cinogen-inducedchange"
Not correeled for
recovery"*
Corrected for
recovery0
X10-4*>1
10-'N.D.'1.7x10-« x
10-42
x
2X10-3"1-2X10-2"N.D.2
10-51-5
x
1(T59N.D.1
x
10-"5.5
x
Preneoplastic
RTEPreneoplastic
(CAK)Mouse mouse fibroblasts
2")Preneoplastic
embryo fibroblasts (1 °or
fibroblasts(CHEF)Preneoplastic
hamster
Ref.
paper39734238113337
X10-»N.D.3.4
(BHK)Preneoplastic
hamster fibroblasts
(BHK)Preneoplastic
hamster fibroblasts
(FOL*)Humanhamster fibroblasts
10-3N.D.'4
x
X10-7N.D.N.D.1
10"2'>2
x
10-42.6
x
fibroblastsHuman
x 10-'*2
x IO"2This
fibroblastsAgaroseAgaroseAgaroseMethylcelluloseAgarAgarAgarAgarAgar0.5-5
* Variants/cell/generation.
6 Variants/viable cell.
c Corrected for recovery of anchorage-independentcells in semisolid medium.
" Values as originally reported when not corrected for recovery of anchorage-independentcells.
' Mean ±SD.
'N.D., not determined.
9 Recalculatedfrom original data using recovery efficiency in agarose of 0.01.
" Recalculatedfrom original data using recovery efficiency in methylcelluloseof 0.15.
' No detectable induction by carcinogens; no data given.
' Recalculatedfrom original data and notes on scoring in subsequent paper (20) using recovery efficiency in agar of 0.15.
* Recalculatedfrom original data using recovery efficiency in agar of 0.004.
cell line was the only one of those in this study which at a low
passage number formed tumors after injection into nude mice.
Because the latency period of this tumor was relatively long (~12
weeks), it is assumed that either a small number of tumorigenic
cells were in the population at the time of injection, or that
tumorigenic cells arose in the animal (38). The higher rate of
appearance of anchorage-independent variants in this cell line
may have contributed to its increased rate of neoplastic progres
sion relative to the other cell lines in this study. Anchorageindependent variants of this cell line had a reduced latency period
(~4 weeks) for tumor formation in nude mice, suggesting that
conversion to this phenotype may increase the tumorigenic
potential of the cells.
We also determined the inducibility of anchorage independ
ence in EGV-1 clone 19 cells and EGV-6 cells using the carcin
ogen MNNG. The maximum frequency of induction in both cell
types was ~1 x 10~3 variant/surviving cell at relative survivals
of 30 to 50%. It is interesting to compare the frequency of
induction of anchorage-independent variants in EGVs with the
frequency of induction of EGVs from normal RTE cells. At
equitoxic doses of MNNG, the frequency of MNNG-induced
EGVs from normal RTE cells was 3 to 5 x 10"2/surviving cell
(39), in contrast to the frequency of anchorage-independent
variants induced from preneoplastic EGVs of ~10~3/surviving
cell. The 30- to 50-fold difference in the frequency of induction
of 2 different cell variants by the same carcinogen in the same
type of cells suggests that these variants may arise by different
mechanisms and may involve different molecular targets. One
fact that should be considered, however, is that the dose of
MNNG which causes 50% cell death in primary RTE cells which
are grown on feeder cells (~0.3 u.g MNNG/ml) (15, 39) is 10
times higher than the 50% cytotoxic dose for EGV cells which
are grown on plastic. This is due in part to the presence of feeder
cells which reduce the cytotoxicity of MNNG to both RTE and
EGV cells (15). However, even when grown on feeder cells EGVs
are still much more sensitive to the cytotoxic effects of MNNG
CANCER
RESEARCH
(50% cytotoxic dose, -0.06
/ug/ml on feeder cells).4 While the
cytotoxic response of EGVs to MNNG is increased relative to
RTE cells, it is not known if their susceptibility to heritable genetic
damage is also increased. This could be examined by determin
ing the effect of equitoxic doses of MNNG on the same locus in
both RTE and EGV cells (e.g., induction of thioguanine- or
ouabain-resistant mutants).
Comparisons of rates and frequencies of preneoplastic events
in carcinogenesis with mutational events at specific gene loci
might provide insights into the role of mutagenesis in carcino
genesis (4, 38). The known rates of spontaneous mutation
affecting expression of single defined genes of mammalian cells
range from >10~7/cell/generation for ouabain resistance (13) to
about 1Q~*/cell/generation for adenine phosphoribosyltransferase deficiency in human diploid fibroblasts which have only a
single functional alÃ-elefor that enzyme (35). Similarly, the known
frequencies of induced mutation range from >10~6 for ouabain
resistance (1,9) to ~10~2 for adenine phosphoribosyltransferase
deficiency in the heterozygous human fibroblasts noted above
(35). Thus, the observed spontaneous rates and induced fre
quency of change to anchorage independence described here
for preneoplastic RTE cells are compatible with some of the
higher values reported in the literature for mutations at known
gene loci. However, these comparisons can be misleading with
out an understanding of (a) what fraction of all genotypic changes
in loci coding for these changes result in phenotypic changes;
(b) the dominance of the expression of these changes; and (c)
the effective target sizes for the changes being compared.
Comparisons can be also made of the rates and frequencies
of change to anchorage independence observed here with similar
values observed for other cells. Such comparisons must be made
cautiously, however, since different semisolid media, different
carcinogens, and different methods of calculation have been
used in various studies. Table 9 summarizes the spontaneous
4D. G. Thomassen,unpublisheddata.
VOL. 45 APRIL 1985
1522
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1985 American Association for Cancer Research.
ANCHORAGE-INDEPENDENT
VARIANTS OF TRACHEAL
EPITHELIAL CELLS
rates and frequencies of carcinogen induced change to anchor
age independence which have been reported. Many of the values
reported in the literature are underestimates of the absolute
values, because they were calculated without correcting for
recovery efficiencies. We have recalculated the rates and fre
quencies which were not originally corrected for recovery of
anchorage-independent cells, using the methods described in
genetic loci (e.g., oncogenes) will greatly enhance our under
standing of the mechanism of chemical carcinogenesis.
We would like to thank Dr. John Drake, for his suggestions
Alma Gonzalez, for preparation of this manuscript.
this communication and using as recovery efficiencies the CFEs
of anchorage-independent variants given in the various reports
REFERENCES
ACKNOWLEDGMENTS
when available.
The estimates of the rates of spontaneous change to anchor
age independence fall into 3 classes: (a) 2 values (7, 38) using
mouse fibroblasts fall into the range of values reported here for
RTE cells; both of these estimates were made using agarose
and used diploid or near-diploid cells; (b) an estimate (23) with a
line of near-diploid, preneoplastic hamster fibroblasts (BHK) using
agar was 100-fold lower than the values reported here for RTE
1. Baker, R. M., Bumette, D. M., Mankovitz, R., Thompson, L. H., Whitmore, G.
F., Siminovitch, L., and Till, M. E. Ouabain-resistant mutants of mouse and
hamster cells in culture. Cell, 1:9-21,1974.
2. Barrett, J. C., Crawford, B. D., Mixter, L. O., Schectman, L. M., Ts'o, P. 0.
P., and Pollack, P. Correlation of in vitro growth properties and tumorigenicity
of Syrian hamster cell lines. Cancer Res., 39:1504-1510,1979.
3. Barrett, J. C., Crawford, B. D., and Ts'o, P. 0. P. The role of somatic mutation
in a multistage model of carcinogenesis. In: N. Mishra, V. C. Dunkel, and M.
Mehlman (eds.), Mammalian Cell Transformation by Chemical Carcinogens,
pp. 467-500. Princeton Junction, NJ: Senate Press, Inc., 1980.
4. Barrett, J. C., Hesterberg, T. W., and Thomassen, D. G. Use of cell transfor
mation systems for carcinogenicity testing and mechanistic studies of carci
nogenesis. Pharmacol. Rev., 36: 53s-70s, 1984.
5. Barrett, J. C., and Thomassen, D. G. Use of quantitative cell transformation
assays in risk estimation. In: V. B. Vouk, G. C. Butler, D. G. Hoel, and D. B.
Peakall (eds.), Methods of Estimating Risk in Human and Chemical Damage
in Non-Human Biota and Ecosystems, SCOPE, SGOMSEC 2, IPCS Joint
Symposia 3, New York: John Wiley & Sons, Inc., in press, 1984.
6. Barrett, J. C., and Ts'o, P. O. P. Mechanistic studies of neoplastic transfor
mation of cells in culture. In: H. Gelboin and P. 0. P. Ts'o (eds.), Polycyclic
cells; and (c) a determination (11) made using preneoplastic
hamster fibroblasts (FOL+) was 1000-fold lower than the values
reported here for RTE cells. Although this determination was
also made in agar, the estimated rate of spontaneous change
was nearly equal to the square of the values reported for RTE
and CAK mouse cells. Interestingly, the FOL+ cells are subtetraploid, whereas the CAK mouse cells are near diploid (38), and
the RTE cells used here are hyperdiploid. Whether the differ
ences in the estimated rates are due to the use of agar versus
agarose, the ploidy of the cells, or some combination of these
factors, will require further studies of this type.
The calculated frequencies of induced change to anchorage
independence summarized in Table 9 fall into 2 classes. Two
publications (7, 8) report frequencies in the range reported here
for EGV cell lines of RTE cells. The other publications (33, 34,
37) report values 10 to 100 times greater than those reported
here. These higher values are similar to the frequencies of
induction of early preneoplastic changes in RTE (39) or Syrian
hamster embryo (6) cells. Whether the changes to anchorage
independence in different cell types represent changes at the
same or different loci in the various cells remains to be deter
mined. The variable frequencies of induction and the variable
rates of spontaneous change to anchorage independence sug
gest that there may be fundamental differences between different
cell types in the molecular basis of certain transformed phenc-
7.
8.
9.
10.
11.
12.
13.
14.
15.
types.
In summary, we have quantitated the development of anchor
age-independent variants of preneoplastic RTE cells to improve
16.
our understanding of the origin and development of this late,
presumed neoplasia-related phenotype. We have demonstrated
that anchorage-independent variants of preneoplastic RTE cells
arise by a stochastic process at a high rate, >10"4 variant/cell/
17.
18.
generation. In addition, these variants are readily inducible with
a frequency of >10~3 variant/cell by exposure to carcinogens.
19.
The rates and frequencies observed are similar to some though
not all rates and frequencies of change to anchorage independ
ence reported for fibroblasts in culture. In addition, these rates
and frequencies are similar to the rates and frequencies of some
mutations at known gene loci. Quantitative somatic cell genetic
studies on the development of both early and late changes
involved in the neoplastia development of fibroblast and epithelial
cells in culture in conjunction with molecular studies at defined
CANCER RESEARCH
and advice, and
20.
21.
22.
Hydrocarbons and Cancer, pp. 235-267. New York: Academic Press, Inc.,
1978.
Bellett, A. J. D., and Younghusband, H. B. Spontaneous, mutagen-induced
and adenovirus-induced anchorage-independent,
tumorigenic variants of
mouse cells. J. Cell. Physiol., 101: 33-48,1979.
Bouck, N., and diMayorca, G. Somatic mutation as the basis for malignant
transformation of BHK cells by chemical carcinogens. Nature (Lond.), 264:
722-727,1976.
Buchwald, M. Mutagenesis at the ouabain-resistance locus in human diploid
fibroblasts. Mutât.Res., 44: 401-412,1977.
Colbum, N. H., Vorder Bruegge, W. F., Bates, J. R., Gray, R. H., Rossen, J.
D., Kelsey, W. H., and Shimada, T. Correlation of anchorage-independent
growth with tumorigenicity of chemically transformed mouse epidermal cells.
Cancer Res., 38: 624-634, 1978.
Crawford, B., Barrett, J. C., and Ts'o, P. O. P. Neoplastic conversion of
preneoplastic Syrian hamster cells; rate estimation by fluctuation analysis. Mol.
CellBiol., 3:931-945, 1983.
Danielson, K. G., Anderson, L. W., and Hosick, H. L. Selection and character
ization in culture of mammary tumor cells with distinctive growth properties in
vivo. Cancer Res., 40: 1812-1819,1980.
Elmore. E., and Barrett, J. C. Measurement of spontaneous mutation rates at
the Na*/K* ATPase locus (ouabain resistance) of human fibroblasts using
improved growth conditions. Mutât.Res., 97: 393-404,1982.
Foulds, L. Neoplastic Development, Vol. 1. New York: Academic Press, Inc.,
1969.
Gray, T. E., Thomassen, D. G., Mass, M. J., and Barrett, J. C. Quantitäten of
cell proliferation, colony formation, and carcinogen induced cytotoxicity of rat
trachéalepithelial cells grown in culture on 3T3 feeder layers. In Vitro (Rockville), 19: 559-570,1983.
Kondo, S. A theoretical study on spontaneous mutation rate. Mutât.Res., 14:
365-374, 1972.
Kulesz-Martin, M., Koehler, B., Hennings, H., and Yuspa, S. H. Quantitative
clonal assay for carcinogen altered mouse epidermal cells. Proc. Am. Assoc.
Cancer Res., 22: 79,1981.
Leavitt, J. C., Crawford, B. D., Barrett, J. C., and Ts'o, P. O. P. Regulation of
requirements for anchorage-independent growth of Syrian hamster fibroblasts
by somatic mutation. Nature (Lond.), 269: 63-65,1977.
Luria, S. E., and Delbrück,M. Mutations of bacteria from virus sensitivity to
virus resistance. Genetics, 28: 491-511,1943.
Mäher, V., Rowan, L. A., Silinskas, K., Kateley, S., and McCormick, J.
Frequency of UV-induced neoplastic transformation of diploid human fibreblasts is higher in xeroderma pigmentosum cells than in normal cells. Proc.
Nati. Acad. Sci. USA, 79: 2613-2617,1982.
Marchok, A. C., Rhoton, J. L., Griesemer, R. A., and Nettesheim, P. Increased
in vitro growth capacity of trachéal epithelium exposed in vivo to 7,12dimethylbenz(a)anthracene. Cancer Res., 37:1811-1821,1977.
Marchok, A. C., Rhoton, J. C., and Nettesheim, P. In vitro development of
oncogenicity in cell lines established from trachéalepithelium pre-exposed in
vivo to 7,12 dimethylbenz(a)anthracene. Cancer Res., 38: 2030-2037,1978.
VOL. 45 APRIL 1985
1523
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1985 American Association for Cancer Research.
ANCHORAGE-INDEPENDENT
VARIANTS
RESEARCH
EPITHELIAL
CELLS
human fibroblasts. Proc. Nati. Acad. Sci. USA, 78: 3053-3057,1981.
32. Pilot, H. C. Fundamentalsof Oncology. New York: Marcel Dekker, Inc., 1978.
33. Silinskas, K. C., Kateley. S. A., Tower, J. E., Mäher,V. M., and McCormick, J.
J. Inductionof anchorage-independentgrowth in humanfibroblasts by propane
sultone. Cancer Res., 41: 1620-1627,1981.
34. Smith, B. L., and Sager, R. MultiSteporigin of tumor-forming ability in Chinese
hamster embryo fibroblast cells. Cancer Res., 42: 389-396, 1982.
35. Steglich, C. S., and DeMars. R. Mutations causing deficiency of APRT in
fibroblasts cultured from humans heterozygous for mutant APRT alÃ-eles
Somatic Cell Genet., 8:115-141, 1982.
36. Strauss, D. S., Jonasson, J., and Harris, H. Growth in vitro of tumor cell x
fibroblast hybrids in which malignancy is suppressed. J. Cell Sci., 25: 73-86
1976.
37. Sutherland, B. M., Cimino, V. S., Delihas, M., Shih, A. G., and Oliver, R. P.
Ultravioletlight-inducedtransformation of human cells to anchorage-independ
ent growth. Cancer Res., 40: 1934-1939,1980.
38. Thomassen,D. G., and DeMars,R. Ctonalanalysisof the stepwise appearance
of anchorage-independenceand tumorigenicity in a permanent line (CAK) of
mouse cells. Cancer Res., 42: 4054-4063,1982.
39. Thomassen, D. G., Gray, T., Mass, M. J., and Barrett, J. C. High frequency of
carcinogen-inducedearly, preneoplasticchanges in rat trachéal
epithelial cells
in culture. Cancer Res., 43: 5956-5963,1983.
40. Terzaghi, M., and Nettesheim, P. Dynamics of neoplastic development in
carcinogen-exposedtrachéal
mucosa. Cancer Res., 39: 4003-4010,1979.
23. MarÃ-n,G. Origin and rate estimation of the spontaneous loss of anchorage
dependenceof growth in BHK21 cells. Exp. Cell Res., 125: 31-36,1980.
24. Marshall, C. J.. Franks, L. M., and Carbone!!. A. W. Markers of neoplastia
transformation in epithelial cell lines derived by human carcinomas. J. Nati.
Cancer Inst., 58: 1743-1751,1977.
25. Montagnier, L. Factors controlling the multiplication of untransformed and
transformed BHK 21 cells under various environmental conditions. In: Wolstenholme and Knight (eds.), pp. 33-44. CIBA Foundation Symposium on
Growth Control in Cell Cultures. London: Churchill, 1971.
26. Montagnier, L. Soft agar techniques. In: P. F. Kruse and M. K. Patterson,
(eds.), Methods and Applications of Tissue Culture, pp. 276-280. New York:
Academic Press, Inc., 1973.
27. Nettesheim, P., and Barrett, J. C. Trachéalepithelial cell transformation: a
model system for studies on neoplastic progression. Crit. Rev. Toxico!., 72:
215-239,1984.
28. Neugut, A. I., and Weinstein, l. B. The use of agarose in the determination of
anchorage-independentgrowth. In Vitro (Rockville), 75: 351-335,1979.
29. Neugut, A. I., and Weinstein, I. B. Growth limitation of BHK-21 cells and its
relation to folate metabolism. In Vitro (Rockville), 75: 363-367,1979.
30. Pai, S. B., Steele,V. E., and Netteshiem,P. Quantitätenof early cellularevents
during neoplastic transformation of trachéal
epithelial cell cultures. Carcinogenesis (Lond.), 4: 369-374,1983.
31. Peehl, D. M., and Stanbridge, E. J. Anchorage-independentgrowth of normal
CANCER
OF TRACHEAL
VOL. 45 APRIL 1985
1524
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1985 American Association for Cancer Research.
Quantitation of the Rate of Spontaneous Generation and
Carcinogen-induced Frequency of Anchorage-independent
Variants of Rat Tracheal Epithelial Cells in Culture
David G. Thomassen, Paul Nettesheim, Thomas E. Gray, et al.
Cancer Res 1985;45:1516-1524.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/45/4/1516
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. © 1985 American Association for Cancer Research.