Current Biology 17, 431–437, March 6, 2007 ª2007 Elsevier Ltd All rights reservedDOI 10.1016/j.cub.2007.01.049
A Virus Causes Cancer by Inducing
Massive Chromosomal Instability
through Cell Fusion
Dominik M. Duelli,1Hesed M. Padilla-Nash,2
David Berman,3Kathleen M. Murphy,3,4
Thomas Ried,2and Yuri Lazebnik1,*
1Cold Spring Harbor Laboratory
One Bungtown Road
Cold Spring Harbor, New York 11724
2Genetics Branch at the National Cancer Institute
National Institutes of Health
Building 50, Room 1408, MSC-8010
Bethesda, Maryland 20892
3Department of Pathology
Johns Hopkins University Medical School
1550 Orleans Street, CRB2 Room 545
Baltimore, Maryland 21231
4Molecular Diagnostics Laboratory
Johns Hopkins University Medical School
600 North Wolfe Street, Park Building, Room SB202
Baltimore, Maryland 21287
Chromosomal instability (CIN) underlies malignant
properties of many solid cancers and their ability to
escape therapy, and it might itself cause cancer [1, 2].
CIN is sustained by deficiencies in proteins, such as
the tumor suppressor p53 [3–5], that police genome
integrity, but the primary cause of CIN in sporadic can-
mutations that deregulate telomere maintenance, or
mitosis, yet such mutations have not been identified
in the majority of sporadic cancers . Alternatively,
CIN could be caused by a transient event that destabi-
lizes the genome without permanently affecting mech-
anisms of mitosis or proliferation [5, 8]. Here, we show
that an otherwise harmless virus rapidly causes mas-
sive chromosomal instability by fusing cells whose
cell cycle is deregulated by oncogenes. This synergy
between fusion and oncogenes ‘‘randomizes’’ normal
diploid human fibroblasts so extensively that each
analyzed cell has a unique karyotype, and some pro-
duce aggressive, highly aneuploid, heterogeneous,
many viruses are fusogenic, this study suggests that
viruses, including those that have not been linked to
carcinogenesis, can cause chromosomal instability
and, consequently, cancer by fusing cells.
To test the hypothesis that viruses can induce chromo-
somal instability (CIN) by fusing cells, we compared
three aliquots from the same population of normal
ways. We derived one population (Figure 1A.1, ‘‘D’’) by
cotransducing an aliquot of D551 cells with two onco-
genes, adenoviral E1A, which deregulates the cell
cycle by inactivating the tumor suppressor RB1, and
human HRAS1, which prevents E1A from triggering
apoptosis. Human cells cotransduced only with E1A
and HRAS1 do not form tumors in animal models of
We derived the second population (Figure 1A.2, ‘‘DV’’)
by first infecting another aliquot of D551 cells with
MPMVE, a primate retrovirus that has been found in
people and has no cytostatic or cytotoxic effect .
The infected cells remained diploid and mononuclear
because normal cells that are fused by the addition of
MPMVEcease proliferation and die within a few weeks,
whereas cells that become infected do not fuse to
each other, which prevents the emergence of new fused
cells . Therefore, we waited for about a month until
the fused cells died, and then we cotransduced E1A
and HRAS. Thus, the resulting DV cells expressed the
oncogenes and were infected with the virus but were
The third population also expressed both oncogenes
and was infected with the virus, but the cells were hy-
brids (Figure 1A.3, ‘‘DVH1,’’ ‘‘DVH2,’’ and ‘‘DVH3’’) that
we created by transducing the third aliquot of D551 cells
with either E1A or HRAS1 and then fusing the resulting
cells to each other by MPMVE.
A Synergy of Cell Fusion and Oncogenes Induces
Within the few generations required for obtaining
enough cells for analysis, we found that the hybrid
populations (DVH1–DVH3) were highly heterogeneous
in their DNA content, as detected by flow cytometry,
whereas the control cells (D and DV) were not (Figure S1
in the Supplemental Data available online). Because
DVH and DV both expressed the oncogenes and the
virus but only DVH cells were formed by fusion, we con-
cludedthatthe heterogeneity was aconsequence ofcell
while the populations were expanding, perhaps as a
result of selective pressure of tissue culture (Figure S1,
compare left and right histograms). Consistent with the
reports that CIN causes apoptosis , the rate of apo-
ptosis in the hybrids was higher during the expansion
than in D or DV cells (Figure S1). As expected, expres-
sionof E1Awith HRAS increased the rate ofproliferation
(Figure S1, right column).
Karyotype analysis of the expanded populations
revealed that the chromosomal heterogeneity was sub-
stantially higher in the hybrids than in the control cells.
Karyotypes of control cells (D and DV) were relatively
stable, uniform, and usually hypodiploid with relatively
few chromosomal abnormalities (Figures 1B and 1C).
In contrast, the karyotypes of the hybrids were highly
aneuploid, from hypodiploid to octaploid (Figures 1B
and 1E), and were heterogeneous with respect to both
the number of chromosomes and the number of
numerical and structural aberrations, which included
chromosome breakage and endoreduplication (Figures
1B–1E; also Figure S2).
The differences detected by spectral karyotyping
(SKY), which reveals aberrations undetectable by con-
ventional karyotyping , were even more remarkable
(Figure 1D). Whereas the cotransduced cells had few
structural abnormalities or deviations in chromosome
number, all the hybrids were diverse to the degree that
yotype. A distinguishing feature ofthe hybrids was com-
plex rearrangements involving multiple chromosomes
(Figure 1D; also Tables S1 and S2 and Figure S2). The
clonal chromosomal imbalances in the hybrids were
Figure 1. A Synergy between Virus-Mediated Cell Fusion and Oncogenes Induces Genomic Heterogeneity
hygromycin (HR) and with Ha-RasV12 (H-RAS) in a vector that confers resistance to puromycin (PR) (1). To control for the effects unrelated to cell
fusion, we produced cells that were infected with MPMVEand cotransduced with E1A and H-RAS but were not fused (2; see text for details).
Alternatively (3), we transduced D551 cells with either E1A or H-Ras by using the same vectors as in (1) and (2) and then fused them to each other
with MPMVE(Supplemental Experimental Procedures). Three colonies of cells resistant to both selection drugs (DVH1, DVH2, and DVH3) were
chosen randomly from one plate. (B) A synergy of virus-mediated cell fusion and oncogenes induces massive chromosomal instability. The in-
dicated cell populations of 108each were obtained as described in Figure 1A and Figure S1 and were analyzed via karyotyping for chromosome
number (left column), numerical aberrations (middle column), and structural abnormalities (right column). The number of cells used for each cell
line was between 19 and 24. Note that the data for D551 are plotted on a different scale than the data for other cells. The heterogeneity of the
chromosome number in the analyzed cells is presented in (C). The concentric circles indicate the chromosome number per cell, each symbol
corresponds to one cell, and the spokes correspond to the cell’s identifying number. Note the contrast between the uniformity of DV cells
and the diversity of the hybrids. (D) A SKY image from one representative cell from each population is presented. The white arrows refer to chro-
mosome losses. See Table S1 for the karyotype annotations.
reminiscent of karyotypes of cancer cells . Because
all this genomic diversity was derived from just a few
cells and changed over time (Figure S1, compare left
and right histograms) we concluded that MPMVEin-
duced massive chromosomal instability by fusing cells
‘‘primed’’ with the oncogenes. The degree of CIN and
the diversity of chromosomal aberrations by far ex-
ceeded those reported in cells that became aneuploid
after drug-induced failure of cytokinesis .
The number of chromosomal aberrations in the hy-
brids was not simply a result of the hybrids’ having
more chromosomes than the control cells. In fact, the
number of numerical aberrations per chromosome was
inversely proportional to the number of chromosomes
per cell (Figure S3, left column), which could be ex-
plained if cell fusion induces chromosomal instability
by means other than merely duplicating the genome.
A Synergy of Cell Fusion and Oncogenes
To test whether chromosomal instability promoted
tumorigenicity, as the evolutionary model of cancer
hypothesizes , we transplanted the hybrid and
control cells into nude mice (Figure 2). As expected ,
cotransduced cells (Figures 2A and 2D) failed to induce
(Figure 2A, ‘‘DV’’). In contrast, one of three hybrid
populations produced tumors (Figure 2A, ‘‘DVH3,’’ and
Surprisingly, histopathologic examination revealed
that these fibroblast-derived tumors did not exhibit the
typical fusiform cell shape and fascicular cellular
arrangement characteristic of fibroblast-derived malig-
nant tumors (fibrosarcomas). Rather, the tumors com-
prised more primitive looking cells that had epithelial-
like morphology and were arranged in nests and cords.
Virtually all of the cells expressed vimentin, an interme-
diate filament associated with mesenchymal differentia-
tion but also with aggressive epithelial-derived tumors
(carcinomas). An aggressive phenotype was also indi-
cated by alternating regions of highly proliferative and
necrotic tumor cells (Figure 2C).
The remarkable observation that fibroblasts pro-
duced tumors with aberrant—focally carcinomatous—
differentiation could be explained if chromosomal
instability were sufficiently extensive to generate cells
with diverse phenotypes. An alternative explanation is
that the tumors represented an accidental contamina-
tion with an unrelated epithelial cancer cell line. How-
ever, the microsatellite analysis estimated the chance
Figure 2. A Synergy of Cell Fusion and Oncogenes Produces Cancers
Cells (5 3 106) of the indicated populations that were expanded to 108cells (Figure S1) were injected subcutaneously into both flanks of 6- to 8-
week-old irradiated nude female mice, and the mice were monitored for the appearance of tumors, indicated by blue arrows in (A). The growth of
tumors is plotted in (B). Each line in (B) is a history of one injection that gave rise to a tumor, and the total tumor frequency is indicated in (A). The
picture in (A) was taken 6 weeks after injection. Two tumors produced by injection of DVH3 cells (DVH3-t1 and DVH3-t2) were analyzed by his-
tology and immunohistochemistry. An example is shown in (C). Asterisks show necrotic regions; arrows (insets) show mitotic figures. One of the
mice injected with DV developed a primary (of mouse origin) anal skin cancer that was away from the injection site and was detected at day 120
after injection. All mice were monitored for more than 10 months after injection until their death, or until the tumor burden was the maximal al-
lowed for the animal’s welfare.
Cell Fusion Can Lead to Genomic Instability and Cancer
of a contamination at less than one in 109(Figure S4).
Thus, the morphological and immunohistological fea-
tures of the tumors raised the possibility that at least
some epithelial tumors might not be of epithelial origin
and illustrated the randomizing power of cell fusion.
The ability of cell fusion to change epigenetic regulation
of cells  or expression of oncogenes could also con-
tribute to the change in morphology.
of mutations acquired before cell fusion. For several
reasons, we found this possibility unlikely. Oncogenic
mutations were unlikely to have occurred in D551 cells
because we made all cell lines simultaneously from the
same D551 population (Figure 1A), but only DVH3 cells
became carcinogenic. Transduction of E1A and HRAS
could promote mutagenesis, but consistent with the
previous report , culturing the control cells (D and
DV) for at least a month failed to make them tumorigenic
(Figure 2), which made it unlikely that culturing the pa-
fore fusion (Supplemental Experimental Procedures)
would result in oncogenic mutations. Finally, tumors
formed in mice by human fibroblasts cotransduced
with E1A and HRAS, along with other oncogenes, such
as MYC  or hTERT , were fibrosarcomas, whereas
tumors formed by DVH3 were not (Figure 2C).
Carcinogenic Hybrids Evolve
Several observations suggested that, as envisioned by
theevolutionary model ofcarcinogenesis , the tumor
cells were a minute fraction of the parental cells and
were produced by CIN and then selected in the animal
for their ability to form tumors. One such observation
was that of a shift from the karyotypic diversity caused
by CIN in the parental cells (Figure 3A and 3C, ‘‘DVH3’’)
to a relatively uniform chromosome number in tumor
cells (Figure 3A and 3C, ‘‘DVH3-t1’’ and ‘‘DVH3-t2’’).
Additionally, there were recurrent and clonal numerical
and structural chromosome aberrations detected in
both analyzed tumors but not in parental cells by con-
ventional karyotyping (Figure 3A) or SKY (Figure 3B). A
third observation was that the tumors that developed
in the same animal as a result of injecting two identical
aliquots of cells had distinct genomes even though
they shared marker chromosomes (Figure 3B; compare
DVH3-t1 and DVH3-t2). Finally, the outcome of the injec-
tions was not predetermined because some tumors
grew aggressively, others regressed, and some did not
form at all.
The finding that the tumors were formed by a minor
subpopulation of DVH3 cells further supported our con-
clusion that the tumorigenic subpopulation emerged
during the clone progression rather than before the hy-
brid was created.
The increased tumorigenicity of the explanted cells
lection of tumorigenic cells in the animal. The explanted
cells always produced tumors if they were injected into
idly as those induced by DVH3 cells (compare Figures
4A and 2B) and were more aggressive, as indicated by
their acquired ability to invade through the peritoneal
wall (Figure 4C) and disseminate to visceral organs (Fig-
ures 4C–4F). In addition, these tumors included areas of
clear cell differentiation (Figure 4B), and these areas
were not observed in the original tumors, indicating an
evolution of the histological properties. Thus, we con-
cluded that a synergy between cell fusion and onco-
genes produced CIN that persisted in the progeny,
resulted in cells capable of forming tumors, and
was likely to fuel their progression to a more aggressive
Our findings provide a simple model that explains how
CIN can occur in sporadic cancers without mutations
that directly affect chromosome segregation. Several
observations suggest that this mechanism can function
in vivo. Indeed, the key step of the proposed mecha-
nism, fusion of cells by viruses, is known to occur in
the body . Cell fusion has been detected in human
cancers and might be explained by the activity of
viruses, exogenous or endogenous [17–19]. Remark-
ably, the karyotypes produced by fusion in our system
are indistinguishable from karyotypes ofhuman cancers
, which implies that determining whether an aneu-
ploid cancer cell is not a hybrid may be difficult. Finally,
although the fate of cells fused by viruses in the body is
not known, even though fusogenic viruses have been
explored as therapeutic tools , there is no reason
to assume that fusion would cause CIN in the dish but
not in the body. Considering that the frequency of can-
isastronomically small,thecarcinogenic outcome ofthe
quent asthatfound inour studytocontributetocarcino-
genesis. However, testing whether this mechanism in-
deed functions in vivo would require animal models
and approaches to detecting proliferating hybrids in
We estimate that at least 18 of 29 virus families that in-
fect human cells have species that fuse cells, which im-
plies that CIN and, by implication, cancer might be
caused by many viruses, including those that are con-
sidered commensal. Pre-existing mutations that affect
the cell cycle may not be required for CIN even if only
one of the fusion partners is a stem cell; stem cells are
frequently recruited to sites of infection because their
plastic cell-cycle regulation allows proliferation after fu-
are fusogenic would provide both components required
for induction of CIN literally in one package. Viruses
might also collaborate. For example, one can speculate
that, whereas human papillomavirus (HPV) deregulates
the cell cycle with oncoproteins E6 and E7 , herpes
viruses or other fusogenic viruses associated with HPV
infection can fuse the cells and thus cause CIN. Impor-
tantly, in principle the fusogenic virus does not need to
be in the fused cells, which implies that the model that
we propose could be an example of a ‘‘hit-and-run’’
type of viral carcinogenesis .
Why cell fusion induces chromosomal instability that
is maintained throughout the generations of the cells is
not known. One possibility is that cell fusion has the
same effect as the drugs thatinhibit cytokinesis , spe-
cifically, thatitcausestetraploidyand theensuing spon-
taneous chromosomal loss. However, we find that the
 and does not require propagation in mice. This differ-
ence can be explained by the fact that cell fusion does
not merely mimic failed cytokinesis or endoreduplica-
tion, which simply duplicate the genome. Instead, the
destabilizing effect of cell fusion could be a result of
differences between the parental cells ; such differ-
ences could include acquired mutations, telomere
length, the stage of DNA replication, epigenetic regula-
tion, and other properties.
For example, the difference in the cell-cycle position
of parental cells often results in premature chromosome
. The chromosome fragments produced by PCC are
randomly incorporated into chromosomes of daughter
cells , which might explain the massive chromo-
somal aberrations of the hybrids. The hybrids may also
retain properties of the parental cells, which can further
expand their diversity and account for emerging proper-
ties, such as the ability to metastasize  or become
drug resistant . The experimental model that we
Figure 3. Cancer Cells Are Selected in the Animal from a Diverse Population Generated by CIN
Cells explanted from tumors (DVH3-t1 and DVH3-t2) were analyzed for DNA content by flow cytometry ([A], left column), and for chromosome
number and structural aberrations by conventional banding with DAPI ([A], middle and right columns, respectively) and by spectral karyotyping
(B). If it could be identified, the modal number of chromosomes is noted in the middle column in (A). For DVH3-t1 and DVH3-t2, 20 spreads were
analyzed for each. The yellow arrows indicate a numerical (+2) and a structural [del(X)] aberration that are clonal and recurrent in the tumor cells.
The magenta arrows indicate clonal aberrations found in the tumor cells. The heterogeneity of the chromosome number in the analyzed cells is
presented in(C).Theconcentriccirclesindicatethe chromosome numbers, andeachsymbolcorresponds toonecell.Thecellsare numbered by
Cell Fusion Can Lead to Genomic Instability and Cancer
developed may help to shed light on the mechanisms
of genomic and, perhaps, epigenetic instability and
may help with studying the consequences of these
The mechanism of carcinogenesis and tumor pro-
gression that we propose is consistent with the reality
that the diversity of cancer cells and their ability to
tive and suggest that early detection or prevention are
likely to remain the most successful ways of conquering
this disease. For example, immunizations against HPV
are strikingly effective in preventing cervical cancer,
and immunization against hepatitis B virus (HBV) sub-
stantially decreases the incidence of liver cancer. Our
study suggests that many viruses can contribute to car-
cinogenesis in a previously unrecognized way. There-
fore, if our model is correct, protecting the body against
these viruses or preventing the cell fusion that they
cause may decrease the frequency of cancers and pre-
vent their progression.
Supplemental Experimental Procedures, four figures and three
tables are available with this article online at http://www.current-
We thank Deb Aufiero for skilled technical assistance, Lisa Bianco
for expert assistance with testing tumorigenicity in mice, Phil Renna
of CSHL Media Arts forphotography, andPamelaMoody of the Cold
Spring Harbor Laboratory Flow Cytometry Facility for help with flow
cytometry. We also thank Linda Barenboim-Stapelton for producing
the SKY probes used in this study. We thank Prasad Jallepalli for his
comments on the manuscript and helpful discussion, Ravi Sachida-
nandam for help with analyzing data, and Paloma Anderson for
assistance with the manuscript. This work was funded by National
Institutes of Health grant R21 GM69357 to Y.L., D.M.D. was sup-
ported by National Institutes of Health training grant CA009176.
T.R. and H.P.N. were supported in part by the Intramural Research
Program of the National Institutes of Health, National Cancer Insti-
tute, Center for Cancer Research.
Figure 4. Tumors Produced by the Hybrids Are Transplantable Invasive Carcinomas
The explanted tumor cells (DVH3-t1 and DVH3-t2, Figure 3) were expanded to 4 3 107cells and injected at 5 3 106into the flanks of nude mice
with six injections for each explant, and the growth of the tumors was monitored (A). Each line in (A) represents a history of one injection. The
history of all injections is plotted. Upon histologic analysis, some of the tumors showed marked regional heterogeneity, including the areas
of clear cell differentiation shown here. (B). A preparation of one of the mice injected with DVH3-t1 (C) illustrates the spread of the cancer into
the peritoneum and some visceral organs, including the intestine (C and E) and the liver (D and F). (E) A small segment of intestinal mesentery
(arrow) is superficially attached to (and dwarfed by) a large tumor nodule. (F) A tumor nodule (circled) invades the outside of the liver. The pattern
ofthe tumorspread was similar inallsix injected mice, exceptthatthe spread was moreextensive andthe sizeof thetumor noduleswas larger in
DVH3-t1 than in DVH3-t2.
Received: November 15, 2006
Revised: January 9, 2007
Accepted: January 10, 2007
Published online: February 22, 2007
1. Jallepalli, P.V., and Lengauer, C. (2001). Chromosome segrega-
tion and cancer: Cutting through the mystery. Nat. Rev. Cancer
2. Duesberg, P., Li, R., Fabarius, A., and Hehlmann, R. (2005). The
chromosomal basis of cancer. Cell Oncol. 27, 293–318.
3. Margolis, R.L. (2005). Tetraploidy and tumor development. Can-
cer Cell 8, 353–354.
4. Fujiwara, T., Bandi, M., Nitta, M., Ivanova, E.V., Bronson, R.T.,
and Pellman, D. (2005). Cytokinesis failure generating tetra-
ploids promotes tumorigenesis in p53-null cells. Nature 437,
5. Storchova, Z., and Pellman, D. (2004). From polyploidy to aneu-
ploidy, genome instability andcancer.Nat.Rev. Mol. Cell Biol.5,
6. Kops, G.J., Weaver, B.A., and Cleveland, D.W. (2005). On the
road to cancer: Aneuploidy and the mitotic checkpoint. Nat.
Rev. Cancer 5, 773–785.
7. Jefford, C.E., and Irminger-Finger, I. (2006). Mechanisms of
chromosome instability in cancers. Crit. Rev. Oncol. Hematol.
8. Duelli, D., and Lazebnik, Y. (2003). Cell fusion: A hidden enemy?
Cancer Cell 3, 445–448.
9. Seger, Y.R., Garcia-Cao, M., Piccinin, S., Cunsolo, C.L., Do-
glioni, C., Blasco, M.A., Hannon, G.J., and Maestro, R. (2002).
Transformation of normal human cells in the absence of telome-
rase activation. Cancer Cell 2, 401–413.
10. Duelli, D.M., Hearn, S., Myers, M.P., and Lazebnik, Y. (2005). A
primate virus generates transformed human cells by fusion.
J. Cell Biol. 171, 493–503.
11. Zhivotovsky, B., and Kroemer, G. (2004). Apoptosis and geno-
mic instability. Nat. Rev. Mol. Cell Biol. 5, 752–762.
12. Schrock, E., du Manoir, S., Veldman, T., Schoell, B., Wienberg,
J., Ferguson-Smith, M.A., Ning, Y., Ledbetter, D.H., Bar-Am, I.,
Soenksen, D., et al. (1996). Multicolor spectral karyotyping of
human chromosomes. Science 273, 494–497.
13. Heim, S., and Mitelman, F. (1995). Cancer Cytogenetics, Second
Edition (New York: John Wiley & Sons).
14. Nowell, P.C. (1976). The clonal evolution of tumor cell popula-
tions. Science 194, 23–28.
15. Mason, D.X., Keppler, D., Zhang, J., Jackson, T.J., Seger, Y.R.,
Matsui, S., Abreo, F., Cowell, J.K., Hannon, G.J., Lowe, S.W.,
et al. (2006). Defined genetic events associated with the sponta-
neous in vitro transformation of ElA/Ras-expressing human
IMR90 fibroblasts. Carcinogenesis 27, 350–359.
16. Chen, E.H., and Olson, E.N. (2005). Unveiling the mechanisms of
cell-cell fusion. Science 308, 369–373.
17. Bjerregaard, B., Holck, S., Christensen, I.J., and Larsson, L.I.
(2006). Syncytin is involved in breast cancer-endothelial cell
fusions. Cell. Mol. Life Sci. 63, 1906–1911.
18. Jacobsen, B.M., Harrell, J.C., Jedlicka, P., Borges, V.F., Varella-
Garcia, M., and Horwitz, K.B. (2006). Spontaneous fusion with,
and transformation of mouse stroma by, malignant human
breast cancer epithelium. Cancer Res. 66, 8274–8279.
19. Andersen, T., Boissy, P., Sondergaard, T., Kupisiewicz, K., Ples-
ner, T., Rasmussen, T., Haaber, J., Kolvraa, S., and Delaisse,
J.M. (2007). Osteoclast nuclei of myeloma patients show chro-
mosome translocations specific for the myeloma cell clone: A
new type of cancer-host partnership? J. Pathol. 211, 10–17.
20. Vile, R.G. (2006). Viral mediated cell fusion: Viral fusion—the
making, or breaking, of a tumour? Gene Ther. 13, 1127–1130.
21. Ogle, B.M., Cascalho, M., and Platt, J.L. (2005). Biological impli-
cations of cell fusion. Nat. Rev. Mol. Cell Biol. 6, 567–575.
22. Munger, K., Hayakawa, H., Nguyen, C.L., Melquiot, N.V., Duens-
ing, A., and Duensing, S. (2006). Viral carcinogenesis and geno-
mic instability. EXS 96, 179–199.
23. Vignery, A. (2005). Macrophage fusion: are somatic and cancer
cells possible partners? Trends Cell Biol. 15, 188–193.
24. Rao, P.N., and Johnson, R.T. (1972). Premature chromosome
condensation: A mechanism for the elimination of chromo-
somes in virus-fused cells. J. Cell Sci. 10, 495–513.
25. Duelli, D.M., and Lazebnik, Y.A. (2000). Primary cells suppress
oncogene-dependent apoptosis. Nat. Cell Biol. 2, 859–862.
Cell Fusion Can Lead to Genomic Instability and Cancer