[CANCER RESEARCH 49. 6731-6737. December 1. 1989]
Model for the Formation of Double Minutes from Prematurely Condensed
Chromosomes of Replicating Micronuclei in Drug-treated Chinese
Hamster Ovary Cells Undergoing DNA Amplification1
Subrata Sen,2 Walter N. Hittelman, Larry D. Teeter, and M. Tien Kuo
Divisions of Laboratory Medicine. Hematopathology Program [S. S.J, Medicine fH'. A/. //./. and Pathology fL. D. T., M. T. K.], The University of Texas M. D. Anderson
Cancer Center, Houston, Texas 77030
ABSTRACT
Double minutes (DM) have been associated with gene amplification in
drug-resistant cells and tumor cells. However, the mechanisms by which
DM are formed have not been elucidated. \Ye present here a model to
describe a possible mechanism of DM formation based on the observa
tions made in two independent early drug-selected multidrug-resistant
cell lines and from in vitro somatic cell fusion experiments between
synchronized S- and \I-phase cells. The multidrug-resistant cell lines
contain both DM and amplified mdr (P-glycoprotein) gene. Cytogenetic
analyses of cells at early stages of selection revealed the presence of a
number of micronuclei in a subpopulation of these cells. These micronu
clei were often asynchronous in their progression through the cell cycle.
As a result, premature condensation of micronuclear chromatin was often
observed in metaphase plates. The pulverized chromatin pattern seen in
certain instances of S-phase prematurely condensed chromosomes dis
plays a striking resemblance to DM structures. These DM-like structures
are linked by replicating DNA as revealed by DNA labeling experiments.
Somatic cell hybrids between S- and M-phase cells when grown in vitro
demonstrated that S-phase prematurely condensed chromatin indeed
gives rise to extra chromosomal structures in the successive cell genera
tions. It is hypothesized that distinct DM-like structures may arise from
the partially replicated and prematurely condensed S-phase chromosomes
following their liberation as extra chromosomal entities after replication
and/or recombination in the succeeding division cycle(s). The enrichment
for DM containing specific genes in drug-resistant cells may result from
the subsequent drug selections.
INTRODUCTION
DM3 and expanded chromosomal
segments known as HSR
are two distinct chromosomal abnormalities that have been
described in a number of cell lines either selected for resistance
to cytotoxic drugs or derived from tumors (1-4). Typically,
DM are 0.3- to 0.5-¿imchromatin particles (5), but their size
varies from limits of light microscopic resolution to that of a
human G group chromosome (6). In addition to this size
heterogeneity, the number of DM per cell has been found to
vary greatly among different cells, i.e., ranging from a few (7)
to more than 1000 per cell (8). Cytogenetic observations of
squashed preparations from metaphase cells revealed clusters
of DM, suggesting close spatial association of DM structures
in the nucleus. Ultrastructural studies suggested that DM are
chromosomal segments but lack functional centromeres (5, 9).
The absence of centromeres in DM may result in their unequal
segregation during mitosis and, therefore, numerical variations
in DM among daughter cells.
DM and HSR have been shown to contain amplified genes
in drug-resistant cells. DM-containing drug-resistant cells are
usually unstable. When these cells are grown in the absence of
drug, the resistance rapidly decreases concomitant with the loss
of DM in these cells. Interconversion between DM and HSR
has been suggested (6, 10, 11). In tumor cells, amplified oncogenes in DM have also been noted (3). These amplified oncogenes may render growth advantage to these cells.
Despite the significant roles of DM in drug-resistant cells
and tumor cells, the detailed mechanisms for DM formation
have not been elucidated. Recently, Carroll et al. (12, 13) have
reported the presence of submicroscopic autonomously repli
cating circular molecules in cells with amplified genes and
proposed that DMs are formed from these extrachromosomal
DNA molecules.
In this paper, a model is presented to describe the possible
route of DM formation. Our model is primarily based on
Cytogenetic observations made in two early passage drug-resist
ant cell lines which contain DM and an amplified mdr (Pglycoprotein) gene and also display DM-like structures (11). In
a significant number of these cells which were being continu
ously treated with drug, thin fibrillar structures were distinctly
visible around the DM. These structures resembled the appear
ance of S-PCC. These structures were virtually absent from the
parental CHO line from which the drug-resistant cells were
selected. Chromosome labeling experiments of the drug-treated
cells revealed these structures to indeed represent S-PCC. Cell
hybrids resulting from fusion of S- and M-phase cells subse
quently showed that S-PCC structures can give rise to extra
chromosomal DM-like entities. The paper describes studies
which led to the development of a model suggesting involve
ment of the S-PCC phenomenon in the formation of DMs.
MATERIALS
AND METHODS
Cell Culture. The CHO cell line and its drug-resistant variants were
maintained as monolayer cultures at 37°Cin 5% COj in air in Dulbecco's modified Eagle's medium (Gibco Laboratories, Grand Island, NY)
supplemented with 10% fetal bovine serum (Hazelton, Denver, PA)
containing 0.1% neomycin (Pharma-Tek). Procedures for selection of
these drug-resistant mutants by vinblastine and Adriamycin have been
described earlier (11). Two of the low levels of drug-resistant mutants
selected with 1.0-Mg/ml concentrations of vinblastine (VBR 1.0) and
Adriamycin (ADR 1.0) at early passages (within 10 to 15 passages of
their first exposure to this concentration of the drugs) were used for
our study.
Cytogenetic Procedures. The frequency of cells with mitotic anoma
lies was analyzed in squash preparations. Cells were trypsinized, pel
leted by centrifugation, and resuspended gently in a hypotonie solution
(medium:water, 3:1) for 10 min. Cells were collected by centrifugation,
fixed with methanohacetic acid (3:1, v/v), and stained with 2% acetoorcein stain. Squash preparations of these cell pellets were made on
6731
Received 12/9/87; revised 6/7/89. 8/28/89: accepted 9/1/89.
The costs of publication of this article were defrayed in part by the payment
of page charges. This article must therefore be hereby marked advertisement in
accordance with 18 L'.S.C. Section 1734 solely to indicate this fact.
1Supported in part by grants from the Robert A. Welch Foundation (G 831
to M. T. K.) and the NIH (GM28573 and CA43621 to M. T. K: CA2793I and
45746 to W. N. H.).
! To whom requests for reprints should be addressed, at the University of
Texas M. D. Anderson Cancer Center. Division of Laboratory Medicine. Box 72,
1515 Holcombe Blvd.. Houston. TX 77030.
3The abbreviations used are: DM. double minutes: HSR. homogenously stain
ing regions; dThd. thymidine; PCC. prematurely condensed chromosomes; SPCC. S-phase prematurely condensed chromosomes; CHO. Chinese hamster
ovan cells; VBR, vinblastine resistant cells; ADR. Adriamycin resistant cells;
BrdUrd, bromodeoxyuridine; PBS, phosphate-buffered saline: S/M hybrid, Sand M-phase-fused cells: M/M hybrid. M- and M-phase-fused cells.
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1989 American Association for Cancer Research.
MECHANISM OF DM FORMATION
clean slides after covering with 22-mm2 coverslips.
ratories, and was used in consultation with Dr. Joshua Epstein of the
Department of Hematology at the University of Texas-M. D. Anderson
Cancer Center. Replicative behavior of the extrachromosomal struc
tures in the hybrids was also studied following autoradiography of the
slides made from cells grown in |'H]dThd after fusion.
DNA Labeling Analyses. To detect labeled multinucleated cells,
semiconfluent monolayer cultures were labeled with 6 nC\/m\ of ['H|dThd (specific activity. 6.7 Ci/mmol; ICN, Irvine, CA) for 20 min, and
the cells were directly fixed in methanohacetic acid (3:1, v/v). Air-dried
slides were then processed for autoradiography.
In the case of labeling DNA associated with DM, semiconfluent
cultures were similarly incubated with 6 ^Ci/ml of |'H]dThd for 20
min, washed twice with regular medium containing unlabeled dThd
(0.1 HIM), and again incubated for 10 more min at 37°C.Colcemid
(0.16 ^M) was included in the medium throughout the entire 30-min
period of labeling and washing the cells. Cells were harvested, and airdried slides were then processed in one of the following ways, (a) One
set of slides was first autoradiographed to reveal the grain distribution
on metaphase cells. The silver grains were removed, and the same cells
were examined for DM-like structures around the same area, (b) An
other set of slides was first stained with Giemsa (5% in 0.01 M NaPO4,
pH 6.8) to observe the DM-like structures in the metaphase cells and
then autoradiographed to see if replicating DNA could be localized in
the vicinity of DM.
Autoradiography was performed with Kodak NTB-2 nuclear track
emulsion (diluted 1:2 in water) and developed in D19-B solution. The
exposure time was about 10 days at 4°C.To remove silver grains from
autoradiographs, the slides were rinsed in water for about 5 to 10 min,
treated with 0.22 M potassium ferricyanide solution for 3 min, trans
ferred to 1.2 M sodium thiosulfate for 3 to 5 min, and then rinsed in
three changes of water for 1 min each. Slides were then stained with
Giemsa or ethidium bromide (5 ng/m\ in citrate-phosphate buffer, pH
5.5) and observed under light or fluorescence microscope.
Cell Synchronization, Cell Fusion, and Preparation of Chromosome
Slides. Mitotic CHO cells were obtained by gentle shake off of subconfluent cultures treated with 0.05 Mg/ml of Colcemid for 3.5 h. Only
preparations with mitotic indices greater than 95% were utilized for all
subsequent experiments. Synchronized S-phase cells were obtained by
washing mitotic cells free of Colcemid and allowing the cells to divide
and proceed into the next cell cycle. Preliminary experiments showed
that the labeling index of these populations reached 95 to 97% by 7 h.
Thus, for cell fusion experiments, cells synchronized to and released
from mitosis were incubated for 7 h with 5 Mg/ml of BrdUrd, prior to
fusion with mitotic CHO cells. The procedure for cell fusion has been
previously described (14). Briefly, about equal numbers of S- and Mphase cells were washed twice separately and twice after mixing in
Hanks' balanced salt solution. The cell pellet was resuspended in 0.5
ml of medium without serum containing UV-inactivated Sendai virus
and incubated at 4°Cfor 15 min to allow agglutination. Fifty jul of 20
HIMMgCl2 and 50 n\ of 0.5 ng/m\ of Colcemid were then added, and
the fusion mixture was incubated at 37°Cfor 45 min. At the end of the
incubation time, an aliquot of cells was withdrawn for chromosome
preparation, and the rest was replated in equal aliquots of 10 ml of
culture medium with or without 0.5 ¿iCi/mlof ['HjdThd (6.7 Ci/mmol;
ICN) for different time intervals. The cells were harvested at 24 h post
plating and onwards. The first harvest immediately after fusion is
referred to as time zero, and all others are indicated by the number of
hours of their growth in culture before being harvested for slide prepa
ration. Air-dried slides were prepared following treatment of cells with
a hypotonie solution of 0.075 M KC1 and their fixation in methanol:glacial acetic acid.
Detection of BrdUrd-substituted Chromosomes. The differential stain
ing method of Perry and Wolff (15) and a modified immunological
detection method of Speit and Vogel (16) were used to monitor the fate
of BrdUrd-labeled chromatids of the S-phase parent cell in the hybrids.
For immunostaining, slides were treated with methanohNaOH (0.1 M)
(5:2, v/v) for 5 to 10 min, to denature the DNA, rinsed in PBS, treated
with Triton X-100 for 5 min, and rinsed in PBS again. Slides were then
blocked in PBS-containing 5% serum and incubated with a 1:200
dilution of a mouse monoclonal anti-BrdUrd antibody (IU-4) for 1 h.
Following another rinse and block, slides were stained with fiuorescein
isothiocyanate-conjugated rabbit antimouse IgG. Slides were counterstained with 4,6-diamidino-2-phenylindole. The anti-BrdUrd antibody,
IU-4, was a kind gift from Dr. Joe Gray, Lawrence Livermore Labo
RESULTS
Cytogenetic observations in two cell lines, VBR1.0 and
ADR 1.0, have led us to propose the present model for double
minute formation. These cell lines were being selected to survive
in media containing low concentrations of vinblastine and Adriamycin, respectively. These two cell lines display the multidrug
resistance phenotype and contain amplified mdr (P-glycoprotein) gene (H). A significant subpopulation (20 to 30%) of cells
in both VBR1.0 and ADR 1.0 displayed the presence of DM (or
DM-like structures) in their metaphase spreads. The number of
DM varied from 0 to about 50 in both cell lines.
In many instances, these extrachromosomal
structures
showed resemblance with S-PCC. For these structures to be
products of PCC, it was hypothesized that heterophasic nuclei
in micronucleated cells had to occur following exposure of the
cells to the drugs and that premature condensation of the
chromatin from the micronucleus lagging behind the primary
nucleus in its entry into mitosis had taken place. We, therefore,
sought to score the frequency of cells with abnormal metaphase
plates and micronuclei.
Analyses of the squash preparations of directly fixed cells
revealed that a significantly higher proportion of both VBR 1.0
and ADR1.0 lines display mitotic anomalies compared with the
parental CHO line. The anomalies identified included lagging
chromosomes during anaphase separation, multipolar mitoses
(Fig. \A), and resulting cells with micronuclei (Fig. IB). The
frequency of such cells was found to be about 3.2% in VBR 1.0,
3.1% in ADR 1.0, and 0.6% in the CHO line (Table 1).
In order to determine whether DM-like structures were re
lated to S-PCC, in cells being continuously exposed to drug
during early stages of selection, we performed DNA labeling
experiments with ADR 1.0 and VBR 1.0 cells. The cells were
briefly labeled with ['HjdThd, harvested, exposed to hypotonie
solution, fixed, and then autoradiographed.
Twenty-one
ADR 1.0 and 24 VBR 1.0 metaphase cells that contained clusters
of silver grains were photographed. Examples of such labeled
cells are shown in Fig. l, D and F. After removal of silver
grains, the slides were restained with Geimsa or Ethidium
Bromide, and the same metaphase plates were reexamined. We
found that, in all the cells analyzed, DM-like structures were
present in the same areas as the replicating DNA (Fig. 1, C and
E). A detailed examination of the distribution of silver grains
revealed that the replicating DNA is situated around but not
overlying the condensed chromatin resembling DM bodies.
This correlation strongly suggested a relationship between Sphase PCC and DM-like structures seen in these cells.
It was further hypothesized that the replicating DNA in SPCC containing metaphase cells must be very labile and, as this
S-phase PCC-containing metaphase cell progresses through the
following cell cycles, chromosomal deletions at the site of
replicating DNA may disconnect DM bodies from the replicat
ing DNA and result in the formation of extrachromosomal
entities, such as DM (Fig. 1C) (see below). In this context, one
would predict that not all the DM or DM-like structures seen
in every metaphase plates would be associated with replicating
DNA. Indeed, in the above experiment, when metaphase plates
were chosen for examination, on the basis of the presence of
6732
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1989 American Association for Cancer Research.
MECHANISM OF DM FORMATION
Fig. 1. Composite figures showing the se
quence of events leading to the formation of
DM-like structures in VBR1.0 and ADR 1.0
cells. A. an abnormal metaphase cell showing
multipolar movement of chromosomes and
lagging chromosomes during segregation, li.
micronuclei (arrows) containing replicating
DNA associated with unlabeled nucleus in the
same cell. C to F, autoradiographic analyses of
replicating DNA associated with S-PCC. Cells
were labeled according to the procedures de
scribed in "Materials and Methods." Autora
diographic pictures were taken (D, F), and the
same plates were degrained and restained with
Giemsa (C) and ethidium bromide (E). Note
that pulverized chromatin resembling DM-like
structures is seen in fand £which are adjacent
to the silver grains seen in /) and /. G, a
metaphase plate of VBRI.O showing DM (ar
row).
Table 1 Frequency of cells with micronuclei and mitotic anomalies scored in
squash preparations
of cells may become aneuploid due to chromosome nondisjunction and form multipolar metaphase plates (Stage 2), leading
withmicronucleiand
to the formation of micronuclei in the subsequent cell cycle
cellsscored814698776Cells
mitoticanomaly5::24Frequency(%)0.63.153.09
(Stage 3). Occasionally, micronuclei are not in the same phase
typeCHOVBRI.OADR
Cell
of cell cycle as the main nucleus in the polykaryon. When such
a cell carrying heterophasic nuclei progresses to metaphase,
condensation of chromosomes results in the formation of PCC
1.0Total
from micronuclei. In the case of S-PCC, the micronuclear
chromatin forms a pulverized configuration connected by dif
fuse chromatin fibers. Most often this diffuse chromatin con
DM, only 4 of 25 ADR 1.0 and 5 of 28 VBRI.O metaphase
tains replicating DNA (Stage 4). When this metaphase cell
plates were found to contain replicating DNA around the DM.
On the basis of these observations, the following model is proceeds to anaphase, incompletely replicated condensed chro
proposed for the formation of DM (see drawing in Fig. 2). mosomes may give rise to DM-like structures (Stage 5). The
chromatin bodies, generated through S-PCC, may eventually
When animal cells are treated with cytotoxic agents, a fraction
6733
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1989 American Association for Cancer Research.
MECHANISM OF DM FORMATION
Fig. 2. A proposed model for double min
ute formation. Normal cell (Stage 1) exposed
to cytotoxic drug resulting in formation of
mitotic abnormality in metaphase (Stage 2).
When this eell progresses into the following
cell cycle, micronuclei (A/A') are formed (Stage
3). Micronuclei and main nucleus (A') may
asynchronously progress into metaphase, re
sulting in the formation of S-PCC in micronuclcar rinomatiti (Stage 4). Chromosomal deg
radation, replication, and/or recombination in
S-PCC leads to release of chromosomal bodies
like DM (Stage 5).
Asynchronous
Progression
S-PCC Formation
mosomes in one of the parents. Results from these independent
experiments show that, at 24 h post fusion, about 6% of the
hybrids displayed extra chromosomal fragments. This fre
quency was reasonably high, since the frequency of S-PCC in
the original fusion between S-phase and mitotic cells at time
zero ranged between 7 and 9%. With extended periods of
growth in the absence of any selective pressure, however, the
number of hybrids with such extra chromosomal structures
showed a distinct decline. This implies that, under normal
growth conditions in cultures, hybrids with extra chromosomal
structures are at a selective growth disadvantage compared with
the parental cells. M/M hybrids, on the other hand, exhibit
almost no induction of extra chromosomal structures. An oc
casional M/M hybrid with such structure is seen. This is prob
ably a secondary consequence of segregation anomalies taking
place in the tetraploid cells generated after fusion.
Having seen the generation of extrachromosomal structures
hybridsExperimentExperiment
Table 2 Fate of S-PCC in S/M and M/M
in the dividing S/M hybrids in culture, the authors were inter
of
ested in determining how a partially replicated chromosome
metaphases
(i.e., S-phase PCC) replicates in the next cell cycle and, specif
with PCC and/or
of
metaphases
extrachromosomal
ically, if, in succeeding cell generations following fusion, these
(i)024384402444024384402444024384602446No.
=
scored125130112128125105101100115110122114130110ion103105103596847No.
fragments9853187431196621%7.26.14.42.30.9863.62.40.70.995.85.71.91.4
structures first replicated their unreplicated regions or autono
1S/MM/MExperiment
mously replicated their replicated segments from the previous
cell cycle. To address these issues, the S-phase cells were labeled
with BrdUrd immediately prior to fusion, and following fusion,
the hybrids were grown either in the absence of the analogue or
in some cases in the presence of 'H-labeled thymidine. The
BrdUrd-labeled chromatids can be distinguished from unsubstituted chromatids with the help of immunostaining and/or
2S/MM/MExperiment
Hoechst dye-mediated differential staining techniques. The rep
licative performance of BrdUrd-substituted chromatids follow
ing premature condensation in hybrids could also be monitored
due to their differential staining characteristics. Thus, if the
BrdUrd-substituted PCC structures were to replicate once or
twice in the absence of the analogue and incorporate thymidine,
.1S/MM/MTime
then they would have either differentially stained or uniformly
stained sister chromatids in the ensuing cell generations. Also
the spatial and numerical arrangement of sister chromatids
would indicate if the previously replicated regions reinitiated
replication or not, and if the replicated segments fell apart.
In anti-BrdUrd antibody-stained and differentially stained
be established as distinct DM structures following their repli
cation and/or recombination in the next division cycle. Ampli
fication of their DNA in the population of cells under selection
could take place through anomalous segregation of these struc
tures in the succeeding cell generations.
To test the validity of this hypothesis for DM formation,
hybrid cells containing S-phase PCC were generated by fusing
BrdUrd-labeled S phase cells with M phase cells (S/M hybrids).
As a control, mitotic cells were fused with mitotic cells (M/M
hybrids). The hybrid cells were grown in culture and examined
at subsequent mitoses.
Table 2 illustrates the fate of S-phase prematurely condensed
chromosomes in S/M hybrids and in M/M hybrids. S/M hy
brids were distinguished from other homophasic hybrids on the
basis of differential staining of the BrdUrd-substituted chro-
6734
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1989 American Association for Cancer Research.
MECHANISM OF DM FORMATION
Fig. 3. Composite figures showing the generation of
extra chromosomal structures and their replication in
representative S/M phase hybrid cells harvested at 44 to
46 h post fusion in different experiments. A. a metaphasc
plate of a hybrid cell showing anti-BrdUrd antibodystained extrachromosomal structures. The S-phase parent
cell chromatin was labeled with BrdUrd prior to fusion
that gives rise to these structures. Specificity of the staining
is indicated by the distinct differential staining of the two
chromatids in the metaphase chromosomes of a cell la
beled with BrdUrd as shown on the righi. B. Hoechst
fluorescence plus Giemsa-mediated differential staining of
the extrachromosomal structures in a hybrid cell derived
from BrdUrd-labeled S-phase parent. Differentially
stained structures shown with arrows indicate that they
have replicated at least once following fusion in the ab
sence of BrdUrd. C. a hybrid cell between a mitotic and
BrdUrd-labeled S-phase cell when grown in the presence
of [3H]dThd shows differential staining of some of the
extra chromosomal structures (—»).
O, radioactive precur
sor incorporation in them indicating their replication dur
ing in vitro growth in culture.
•'rv- «?: <
V,
«c.
6735
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1989 American Association for Cancer Research.
MECHANISM OF DM FORMATION
hybrid cells (Fig. 3, A to C), extra chromosomal structures seen
in S/M hybrids at 24 h and later after fusion were derived from
S-phase prematurely condensed chromatin. Fig. 3A shows the
extrachromosomal structures positively stained with the antiBrdUrd antibody, verifying their derivation from the S-phase
parent cell chromatin that was prelabeled with BrdUrd in early
S phase prior to fusion. Differential staining of the chromatids
of these structures (Fig. 3B) reinforced this conclusion and
indicated further that the prematurely condensed replicated
segments rereplicated in the subsequent S phase. Moreover, the
rereplicated regions appeared to become distinct DM-like struc
tures. The replicative potential of these structures was further
confirmed by their ability to incorporate [JH]dThd as shown in
Fig. 3, C and D. The differential staining pattern of the chro
matids was not equally distinct in all the extra chromosomal
structures of S/M hybrids. Whether or not this reflects differ
ences in the replication behavior of some of these structures is
not clear.
DISCUSSION
This report describes a simple model for the origin of DM
through PCC of S-phase micronuclei in cells exposed to drugs.
The model assumes that micronuclei are initially formed fol
lowing mitotic anomalies. Vinblastine and Adriamycin, the two
selective drugs studied with our cell lines, despite having differ
ent targets of action inside the cells (e.g., microtubule for
vinblastine and chromosomes for Adriamycin), share the com
mon feature of inducing multiple nucleation. Micronuclei and
mitotic anomalies can be seen not only in vinblastine- and
Adriamycin-treated cells, but also in animal cells treated with
a number of other cytotoxic agents. Micronuclei-inducing
agents need not be mitotic poisons, e.g., Colcemid or Vinca
alkaloids. In fact, micronuclei formation has long been consid
ered an index for standard mutagenicity assay (17). Among the
20 or so known agents that have been shown to induce DM
formation or gene amplification in cultured cells, many, if not
all, are capable of inducing micronuclei. Furthermore, if mi
cronuclei formation were an important initial event in the
production of DM, as proposed in this model, one would predict
that DM-containing cells are aneuploid. This has indeed been
predominantly seen, and DMs have been observed rarely in
normal diploid cells.
DNA labeling experiments were performed to show that the
polykaryons that formed in the drug-treated cells often traverse
the cell cycle in distinct asynchrony. Such asynchrony in drugtreated polykaryons has been observed previously, and occa
sional induction of PCC also has been reported (18, 19). The
mechanism of asynchrony among nuclei in the same cell is still
not clearly known.
The presence of pulverized chromatin of S-phase nuclei as
an intermediate step in DM formation as described in the
present model is consistent with the idea that DM are produced
through chromosomal deletions. We have proposed that some
times deletions occur in those regions of the genome which are
induced to prematurely condense while undergoing DNA rep
lication. One possibility to explain this phenomenon is that the
replicating DNA which is in diffused chromatin configuration
is very sensitive to endogenous nucleases. Alternatively, pre
mature condensation of chromatin may interfere with the nor
mal DNA replication machinery, leaving gaps and unligated
DNA strands. As metaphase cells containing S-phase PCC
further proceed through the cell division cycle, chromosome
structural changes (continuing condensation/decondensation)
and movements (segregations) might also facilitate breakage of
the replicating DNA. Furthermore when S-PCC structures
reinitiate replication in the next cycle and/or undergo recom
bination at the chromosomal sites with nicked DNA strands,
extrachromosomal structures could be generated in the process.
Whatever the mechanisms involved, our results presented in
Table 2 clearly show that S/M heterophasic cells give rise to a
higher frequency of DM-like structures in subsequent cell gen
erations than the M/M homophasic cells. These results support
the idea that S-PCC segregate through cell division cycles,
reinitiate DNA synthesis, and fall free.
DM are extrachromosomal bodies that may contain ampli
fied DNA sequences encoding gene products as targets for the
actions of specific drugs. The model describing the early steps
in the formation of DM can be extended to accomplish gene
amplification in the following manner. The partially replicated
PCC structures often segregate to one of the daughter cells as
double entities during the first anaphase, making the recipient
daughter cell 4n with respect to the replicated segment. Follow
ing the next round of replication, these regions of the genome
would become 8n. Further random segregation of the DM-like
structures could allow the copy number of their sequences to
gradually increase in subsequent cell generations.
The enrichment of specific DNA sequences in DM in drugresistant cells might be a result of continuous selection pressure
which favors the growth of cells containing DM harboring
specific genes conferring drug resistance. The drug treatment
thus seems to have the dual function of first, initiating partially
replicated genomes that could serve as substrates for generation
of DM and thereafter selecting cells with extra copies of the
gene which enhances growth and survival. It is conceivable that
the same model could be applied to the formation of DM in
tumor cells containing amplified oncogenes. Several oncogeneencoded proteins are involved in regulation of cell proliferation,
and cells with DM containing these oncogenes may have growth
advantages over the others (3).
Multiple mechanisms may be involved in gene amplification
in drug-resistant cells. In some instances of drug-selected cell
lines and of cells established from tumor samples, DM may
arise from small circular episomal DNA as proposed by Wahl
and his «workers (13, 20, 21). In other instances, however, it
is plausible that large DNA segments are liberated from chro
mosomes to form DM, as proposed here, especially in view of
the fact that, in multidrug-resistant CHO cells, amplification
units have been found to be over IO6 base pairs long (22). In
fact, the model described in this paper can explain the formation
of both microscopic and submicroscopic precursors of DM.
This model does not address the question of how the amplified
DNA is organized in DM. Whether these abnormal chromo
somal structures contain DNA rereplicated within one cell
cycle, as has been suggested, remains to be determined (23).
ACKNOWLEDGMENTS
The authors are thankful to Dian Miller and Gloria Clinkscales for
typing the manuscript and to Phylisha Agbor for technical assistance.
Subrata Sen is thankful to Dr. Sanford A. Stass, Director of the
Hematopathology Program and Dr. Emil J. Freireich, Director of the
Adult Leukemia Research Program in the Institute, for their support
and encouragement during the concluding stages of this study, and to
his wife, Dr. Pramila Sen, for her help in preparation of this manuscript.
REFERENCES
1. Hamlin. J. L., Milbrandt. J. D.. Heintz. N. H., and Azizkahn. J. D. DNA
sequence amplification in mammalian cells. Int. Rev. Cytol.. 90: .11-82.
1984.
6736
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1989 American Association for Cancer Research.
MECHANISM OF DM FORMATION
2. Stark. G. R.. and Wahl, G. M. Gene amplification. Annu. Rev. Biochem..
S3.-447-491, 1984.
3. Stark. G. R. DNA amplification in drug resistant cells and in tumors. Cancer
Sun.. 5: 1-23. 1986.
4. Schimke, R. T. Gene amplification in cultured cells. J. Biol. Chem.. 263:
5989-5992. 1988.
5. Barker. P. E., and Stubblefield. E. Ultrastructure of double minutes from a
human tumor cell line. J. Cell. Biol.. 83: 663-666, 1979.
6. Cowcll, J. K. Double minutes and homogenously staining regions. Gene
amplification in mammalian cells. Annu. Rev. Genet.. 16: 21-59. 1982.
7. Morgan. R. T.. Woods. L. K.. Moore, G. E., McGraven, L., Quinn, L. A.,
and Sample. T. V. A human gall bladder carcinoma cell line. In Vitro, 17:
503-510, 1981.
8. Levan, G., Mandahl. H., Bregula. U.. Klein, G., and Levan. A. Double
minute chromosomes are not centromeric regions of the host chromosome.
Hereditas. 83: 83-90. 1976.
9. Hamkalo. B. A.. Farnham. P. J.. Johnston. R.. and Schimke, R. T. Ultrastructural features of minute chromosomes in a methotrexate resistant mouse
3T3 cell line. Proc. Nati. Acad. Sci. USA, 82: 1126-1130. 1985.
10. Martinsson, T., Dahlof. B., Weltergren, Y., Leffler. H., and Levan, G.
Pleiotropic drug resistance and gene amplification in a SEWA mouse tumor
cell line. Exp. Cell. Res.. 75«:382-394. 1985.
11. Sen, S., Teeter. L. D,. and Kuo, T. Specific gene amplification associated
with consistent chromosomal abnormality in independently established multidrug resistant Chinese hamster ovary cells. Chromosoma (Beri.), 95: 117125, 1987.
12. Carroll. S. M.. Gaudray, P., De Rose, M. L., Emery, J. F., Meinkoth. J. L.,
Nakkim, E., Subler. M.. Von Hoff. D. D., and Wahl. G. M. Characterization
of an episome produced in hamster cells that amplify a transfected CAD
gene at high frequency. Functional evidence for a mammalian replication
origin. Mol. Cell Biol.. 7: 1740-1750. 1987.
13. Carroll, S. M.. De Rose, M. L., Gaudray. P.. Moore, C. M.. Needham
Vandevanter, D. R.. Van Hoff. D. D., and Wahl, G. M. Double minute
chromosomes can be produced from precursors derived from a chromosomal
deletion. Mol. Cell Biol.. 8: 1525-1533. 1988.
14. Hittclman. W. N. Premature chromosome condensation for the detection of
mutagenic activity. In: T. C. Hsu (ed.). Cytogenetic Assays of Environmental
Mutagen. pp. 353-384. Totwa. NJ: Osman Allanheld. 1982.
15. Perry. P.. and Wolff, S. New Giemsa method for differential staining of sister
chromatids. Nature (Lend.). 251: 156-158. 1974.
16. Speit. G.. and Vogel. W. Detection of bromodeoxyuridine incorporation in
mammalian chromosomes by a bromodeoxyuridine antibody. Chromosoma
(Beri.). 94: 103-106. 1986.
17. Jenssen. D., and Ramel, C. The micronucleus test as part of a short term
mutagenicity test program for the prediction of carcinogenicity evaluated by
143 agents tested. Mutât.Res., 75: 191-202, 1980.
18. Stubblefield. E., DNA synthesis and chromosomal morphology of Chinese
hamster cells cultured in media containing jV-deacetyl-A'-methylocolchicine
(Colcemid). In: J. R. C. Harris (ed.). Cytogenetics of Cells in Culture, pp.
223-248. New York: Academic Press. 1964.
19. Obe, G., and Beek. B. Premature chromosome condensation in micronuclei.
In: P. N. Rao. R. T. Johnson, and K. Sperling (eds.). Premature Chromosome
Condensation. Applications in Basic. Clinical, and Mutation Research, pp.
113-130. New York: Academic Press, 1982.
20. Von Hoff. D. D.. Needham-Vandevanter. D. R., Yucel, J., Windle, B. E.,
and Wahl. G. M. Amplified human myc oncogenes localized to replicating
submicroscopic circular DNA molecules. Proc. Nati. Acad. Sci. USA, 85:
4804-4808. 1988.
21. Wahl, G. M. The importance of circular DNA in mammalian gene amplifi
cation. Cancer Res., 49: 1333-1340, 1989.
22. DeBruijn. M. L.. Vander Blick. A. M.. Biedler. J. L., and Borst, P. Differ
ential amplification and disproportionate expression of five genes in three
multidrug resistant Chinese hamster lung cell lines. Mol. Cell Biol.. 6:47124722. 1986.
23. Schimke, R. T., Sherwood, S. W., Hill, A. B.. and Johnson, R. N. Overreplication and recombination of DNA in higher eukaryotes: potential
consequences and biological implications. Proc. Nati. Acad. Sci. USA, 83:
2157-2161.1986.
6737
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1989 American Association for Cancer Research.
Model for the Formation of Double Minutes from Prematurely
Condensed Chromosomes of Replicating Micronuclei in
Drug-treated Chinese Hamster Ovary Cells Undergoing DNA
Amplification
Subrata Sen, Walter N. Hittelman, Larry D. Teeter, et al.
Cancer Res 1989;49:6731-6737.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/49/23/6731
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. © 1989 American Association for Cancer Research.