Oncogene (1999) 18, 1457 ± 1464
ã 1999 Stockton Press All rights reserved 0950 ± 9232/99 $12.00
http://www.stockton-press.co.uk/onc
Bcl-2 overexpression results in reciprocal downregulation of Bcl-XL and
sensitizes human testicular germ cell tumours to chemotherapy-induced
apoptosis
Emma L Arriola, Ana M Rodriguez-Lopez, John A Hickman and Christine M Chresta
CRC Molecular and Cellular Pharmacology Research Group, School of Biological Sciences, The University of Manchester,
G.38 Stopford Building, Oxford Road, Manchester M13 9PT, UK
Testicular germ cell tumours are hypersentive to
chemotherapy and cell lines derived from these tumours
are chemosensitive in vitro. We have previously shown
that these cell lines express undetectable levels of the
suppressor of apoptosis Bcl-2 and relatively high levels of
the apoptosis inducer Bax (Chresta et al., 1996). To
determine whether the absence of Bcl-2 in these cell lines
makes them highly susceptible to drug-induced apoptosis,
Bcl-2 was expressed ectopically in the 833K testicular
germ cell tumour cell line. Stable overexpressing clones
were isolated and three clones were studied further.
Surprisingly, Bcl-2 overexpressing cells were sensitized
to chemotherapy-induced apoptosis compared to the
parental and vector control cells. Analysis of potential
mechanisms of sensitization revealed there was a
reciprocal downregulation of the endogenously expressed
Bcl-XL in the Bcl-2 overexpressing clones. Downregulation of Bcl-XL to the same extent using antisense
oligonucleotides enhanced etoposide-induced apoptosis
by twofold. Our results indicate that Bcl-2 and Bcl-XL
have di€erent abilities to protect against chemotherapyinduced apoptosis in testicular germ cell tumours. In
contrast to ®ndings in some tumour cell types, Bcl-2 did
not act as a gatekeeper to prevent entry of p53 to the
nucleus.
Keywords: testicular tumours; apoptosis; Bcl-2; Bcl-XL
Introduction
Disseminated testicular cancer, in contrast to the
majority of solid tumours, is highly sensitive to
combination chemotherapy and is cured in greater
than 80% of cases (Einhorn, 1990). Recently it has
been suggested that many chemotherapeutic agents
exert their cytotoxicity through activtion of apoptosis
(Dive and Hickman, 1991). Signi®cantly, chemosensitive testicular germ cell tumours (TGCT) are
hypersentitive to drug-induced apoptosis (Boersma et
al., 1997; Chrestra et al., 1996).
The precise mechanisms that control apoptosis have
not been elucidated. However, studies in several
eukaryotic systems have demonstrated that apoptosis
is regulated by genetic programs involving both
activators and suppressors of cell death. Bcl-2 was
Correspondence: CM Chresta
Received 19 January 1998; revised 9 September 1998; accepted 10
September 1998
the ®rst member of a rapidly expanding family of
proteins which have been demonstrated to regulate
apoptosis in response to chemotherapy both in vivo and
in vitro (Fisher et al, 1993; reviwed in Kroemer, 1997;
Reed, 1997). Enforced overexpression of Bcl-2, or its
homologue Bcl-XL, have been demonstrated to delay
apoptosis and in some cases to provide a true survival
advantage after many diverse stimuli, including
chemotherapeutic agents, g-irradiation and growth
factor deprivation (Lock and Stribinskiene, 1996;
Simonian et al., 1997b; Walker et al., 1997; Yin and
Schimke, 1995). Conversely, overexpression of the
apoptosis promoting members of the Bcl-2 family,
Bax and Bak, have been demonstrated to counteract
the protective e€ect of Bcl-2 and Bcl-XL (reviewed in
Farrow and Brown, 1996; Korsmeyer, 1995). Genetic
manipulation has shown that the relative ratio of these
dimers can determine the susceptibility of the cell to
apoptosis in response to a wide variety of stimuli
(Farrow and Brown, 1996; Korsmeyer, 1995; Oltvai
and Korsmeyer, 1994). From early studies employing
immunoprecipitation and yeast two-hybrid analysis it
was noted that the Bcl-2 family of proteins form
homodimers and heterodimers and it was suggested
this was important for function (Oltvai et al., 1993;
Yang et al., 1995). However, more recently Bcl-2
family members have been demonstrated to interact
with non-Bcl-2 family members (e.g. raf-1, APAF1)
and to regulate apoptosis independently of heterodimerization with each other (Knudson and Korsmeyer, 1997; Simonian et al., 1996, 1997a; Wang et al.,
1996; Zou et al., 1997). In addition, following an
apoptotic stimuli Bax and Bcl-XL are redistributed to
intracellular membranes an event which is important
for function (Hsu et al., 1997; Wolter et al., 1997).
Genetic alterations that prevent or delay apoptosis
can render tumour cells relatively resistant to the
cytotoxic e€ects of chemotherapy agents. High levels of
Bcl-2 and lack of Bax expression have been reported in
a wide variety of human cancers and may be associated
with poor response to chemotherapy in the clinic
(reviewed in Reed, 1994). We have previously shown
that testicular tumours have high levels of Bax and low
to undetectable levels of Bcl-2 compared to drug
resistant tumours of the human bladder (Chresta et
al., 1996). This may be related to the origin of these
tumours, from immature cells of the testis, which
similarly do not express Bcl-2 (Hockenbery et al., 1991;
Lu et al., 1993). To determine whether the absence of
Bcl-2 is a major determinant of chemosensitivity in
these tumours we have investigated the e€ect of
ectopically overexpressing Bcl-2 in a human TGCT
Bcl-2 overexpression in testis tumour cells
EL Arriola et al
1458
cel line 833K. Surprisingly, when the response of the
Bcl-2 overexpressors to chemotherapy was investigated
it was found that Bcl-2 sensitized the cells to
chemotherapy-induced apoptosis and increased cell
death measured by a colony forming assay.
Results
The endogenous level of Bcl-2 in TGCT's (GH,
GCT27, 833K, NT2 and 2102EP) is signi®cantly
lower than in apoptosis resistant RT4 bladder tumour
cells (Figure 1). To test the hypothesis that absence of
Bcl-2 expression is responsible for the apoptosis
susceptible phenotype of TGCT, we ectopically overexpressed Bcl-2 in TGCT (833K) cells. Ten G418
resistance colonies were analysed for Bcl-2 expression.
Of these, three clones expressed Bcl-2 at levels
equivalent to those found in apoptosis-resistant RT4
(Figure 1). Three clones overexpressing Bcl-2 and three
vector control clones were selected for further study.
The cell cycle distribution, cell doubling time and
cloning eciency of the Bcl-2 overexpresing cells were
similar to that of the vector control clones (Table 1).
Thus overexpression of Bcl-2 in TGCT does not
prevent entry into S-phase or colony formation as
has been found in some tumour types (O'Reilly et al.,
1996; Pietenpol et al., 1994).
Overexpression of Bcl-2 enhances etoposide-induced
apoptosis
In contrast to ®ndings in many tumour cell systems
(Fisher et al., 1993; Hashimoto et al., 1995; Kamesaki
et al., 1993; Simonian et al., 1997b), overexpression of
Bcl-2 in TGCT resulted in an increase in etoposideinduced apoptosis measured by both analysis of
morphology and release of low molecular weight
DNA fragments (indicative of endonuclease activation ± see Materials and methods). The percentage of
apoptosis was increased approximately fourfold from
11% in the parental and vector control cells to 38% in
the Bcl-2 overexpressing clones 4 h after drug
treatment (Figure 2a). The level of topoisomerase II
associated single strand breaks produced by a 1 h
treatment with 4 mM etoposide was approximately
equal in the parental and transfected cells (results
not shown). This suggests that Bcl-2 is enhancing
apoptosis at a step in the cell death pathway
subsequent to etoposide-induced DNA damage. Bcl-2
overexpression also promoted apoptosis induced by
the non-genotoxic agent N-methylformamide and the
vinca alkaloid, vincristine. Apoptosis induced by a
20 h treatment with 300 mM NMF was increased by
1.7-fold by Bcl-2 overexpression, whilst that induced
by a 20 h treatment with 10 mg/ml Vincristine was
increased by 1.4-fold.
Bcl-2 has previously been demonstrated to delay but
not inhibit chemotherapy-induced death under certain
circumstances (Lock and Stribinskiene, 1996; Yin and
Schimke, 1995). In order to determine if the
enhancement of etoposide induced apoptosis was just
an alteration in the kinetics of cell death we used a
long-term survival assay, measuring the clonogenic
potential of the cells. As can be seen in Figure 2b,
following etoposide treatment, all Bcl-2 overexpressing
clones showed a 2 ± 3-fold greater loss of clonogenicity
compared to vector controls (and parental cells).
Enhancement of cytotoxicity was not limited to the
topoisomerase poison etoposide. Bcl-2 overexpression
also enhanced sensitivity to cis-platin by 2.2-fold
(Figure 2c).
Analysis of subcellular localization of Bcl-2 and p53 in
Bcl-2 overexpressing TGCT
Figure 1 Comparison of Bcl-2 levels in TGCT cell lines, Bcl-2
overexpressing clones and in an apoptosis resistant bladder
tumour cell line RT4. Levels of Bcl-2 were determined by
immunoblotting. 20 mg of whole cell lysate was separated by
polyacrylamide gel electrophoresis and analysed by immunoblotting. TGCT-GH, GCT27, 833K, NT2, 2102EP, 833K.Neo and
833K.Bcl-2 clones. RT4 is a human bladder tumour cell line
Table 1
Growth characteristics of the Bcl-2 overexpressing clones
Cell type
833K.Neo
833K.Bcl2
Doubling time
Cloning eciency
%Cells in G1 : S : G2/M
24 h
30%
61 : 19 : 20
24 h
23%
61 : 17 : 22
The Bcl-2 clone used in these studies has been
sequenced and found to be wild type. It has also
been shown to be functional in other cell types. It
suppresses apoptosis in a Xenopus oocyte cell free
assay and inhibits dexamethasone-induced apoptosis in
the human lymphoblastoid cell line CEM-C7A
(Cosulich et al., 1996 and Dr P Hedge personal
communication). Although the actual mechanism by
which Bcl-2 functions is as yet not fully understood,
the location of this protein in the mitochondria appears
to be important for function (Kluck et al., 1997; Yang
et al., 1997; Zhu et al., 1996). To determine if the
ectopically expressed Bcl-2 was localized in the
mitochondria we used immunostaining for Bcl-2 and
co-stained with a dye which localizes to active
mitochondria (mito-tracker). The staining pattern in
the TGCT overexpressing Bcl-2 transfects was compared to that of RT4, a bladder tumour cell line which
constitutively expresses Bcl-2. Figure 3d,g show that
Bcl-2 co-localizes with the mitochondrial marker in
both the Bcl-2 overexpressing TGCT cells and in the
RT4 bladder carcinoma. RT4 was used as a positive
control because we have demonstrated, using antisense,
that Bcl-2 protects against drug-induced apoptosis in
this cell type (Arriola, manuscript in preparation).
Bcl-2 overexpression in testis tumour cells
EL Arriola et al
1459
Figure 2 Comparison of chemotherapy-induced apoptosis and loss of clonogencity in parental, vector control and Bcl-2
overexpressing clones. (a) Percentage apoptosis 4 h after 1 h treatment with etoposide (25 mM). Apoptosis was quanti®ed by the
®lter binding assay, as described in Material and methods. Results are the mean and standard deviation of three independent
experiments. (b and c) Long-term survival assays for etoposide and cis-platin (cDDP) cytotoxicity. Cells were treated for 1 h with
the indicated concentrations of etoposide or cDDP. Viability is calculated relative to the vehicle (DMSO) treated control by colony
formation 14 days after drug treatment. ~ 833K.Neo; & 833K parental; &, ~, * 833K.Bcl-2, clones D2, B5 and B6 respectively.
Results are the mean and standard deviation of three independent experiments, triplicate plates were used in each individual
experiment
Recently, Bcl-2 has been shown to inhibit translocation of p53 to the nucleus and this has been suggested
to be important for function (Beham et al., 1997).
However, we did not ®nd this to be the case in either
RT4, where Bcl-2 inhibits apoptosis (Figure 3i) or in
833K.Bcl-2 clones, where Bcl-2 promotes apoptosis
(Figure 3f).
E€ect of Bcl-2 overexpression on other Bcl-2 family
members
Several reports have suggested reciprocity in the
expression of Bcl-2 and Bcl-XL Chao et al., 1995;
Han et al., 1996). We therefore determined if there was
any evidence for alteration in the expression of Bcl-XL
in the Bcl-2 overexpressing clones. In addition, as Bcl-2
has been reported to exert its e€ects through
interaction with pro-apoptotic members of the Bcl-2
family Bak and Bax, we analysed if the expression of
these partners had changed. As can be seen in Figure 4,
Bax protein was uniformly expressed in all lines and
similar results were found for Bak (data not shown).
However, in the Bcl-2 overexpressing cells there was a
reciprocal downregulation of the expression of Bcl-XL
that was directly related to the levels of Bcl-2
overexpression (Figure 4a).
Downregulation of Bcl-XL causes sensitization to
etoposide-induced apoptosis
To investigate the importance of Bcl-XL in 833K cells,
Bcl-XL was downregulated using an antisense phosphorothioate oligodeoxynucleotide to the coding region
of the Bcl-XL mRNA (Dibbert et al., 1998). Cells were
treated with the transfecting agent alone (lipofectin) or
with a scrambled (SC) sequence oligonucleotide as
controls. 833K cells in which Bcl-XL expression had
been downregulated (Figure 5a) were treated with
Etoposide (IC50) for 4 h. Apoptosis was measured by
caspase activation as described in Materials and
methods. Bcl-XL downregulation resulted in a twofold
increased sensitivity of 833K cells to Etoposide induced
Bcl-2 overexpression in testis tumour cells
EL Arriola et al
1460
Figure 3 Subcellular localization of Bcl-2, Bcl-XL and p53 in 833K.neo, 833K.Bcl-2 and RT4. Cells grown as monolayers on
microscope slides were ®xed in methanol acetone (1 : 1) and then examined by ¯uorescence microscopy. Panels a, b, and c are 833K
cells (TGCT) transfected with vector control. Panels d, e and f are 833K cells transfected with Bcl-2 and panels g, h and i are RT4
cells (TCC) which endogenously overexpress Bcl-2. Panels a, d and g have been stained with Hoechst 3258 to show the position of
the nucleus (blue) and with mito-tracker red CMXRos to show the position of mitochondria (red). Bcl-2 was detected using a mouse
antihuman antibody and a FITC conjugated secondary antibody (green). Each ®eld was analysed independently at the appropriate
wavelength for FITC and MitoTracker and then the two images overlaid. The orange-yellow staining (depending on the level of Bcl2) shows that Bcl-2 co-localizes with the mitochondria in both 833K.Bcl-2 and RT4. Panels b, e and h show nuclear staining using
Hoechst 33258 (blue) and a punctate cytoplasmic staining for Bcl-XL in all three cell types. Bcl-XL was detected using a rabbit
polyclonal antibody and a CyTM3-conjugated secondary antibody (red). Note the much reduced expression of Bcl-XL in the Bcl-2
overexpresing clones (panel e). Panels c, f and i show the nuclear localization of p53 in all cell types. p53 was detected using a mouse
monoclonal Ab and a CyTM3-conjugated secondary antibody (red)
apoptosis compared to cells pretreated with the
transfecting agent alone or with the SC oligonucleotide, as seen in Figure 5b.
Discussion
We and others have previously shown that testicular
tumours are exquisitely sensitive to drug-induced
apoptosis (Burger et al., 1997; Chresta et al., 1996).
TGCT cells express signi®cantly lower levels of Bcl-2
than other solid tumour types (Burger et al., 1997;
Chresta et al., 1996, and Figure 1). In this study, we
have determined if the absence of Bcl-2 in TGCT cells
plays a key role in their susceptibility to drug-induced
apoptosis. A TGCT (833K) cell line was engineered to
overexpress Bcl-2 to similar or greater levels than
found in resistant tumours of transitional carcinoma of
the bladder.
Surprisingly, when the sensitivity of the Bcl-2
overexpressing transfects to etoposide or cis-platin
was studied, they were found to be more sensitive
than the vector controls. The long-term survival of
833K.Bcl-2 was on average 2 ± 3-fold lower than that of
833K.Neo after etoposide or cis-platin treatment and
the cells were fourfold more sensitive to etoposideinduced apoptosis in short term apoptosis assays (4 h).
Susceptibility of tumour cells to drug induced
apoptosis has been suggested to be controlled by the
relative levels and dimerizations of the Bcl-2 family
members. We have investigated several parameters
which could potentially a€ect the sensitivity of the
Bcl-2 overexpressing transfects to chemotherapy.
Bcl-2 has previously been demonstrated to exert a
growth inhibitory activity on tumour cells and to
prevent re-entry into the cell cycle from a G0-like state
(O'Reilly et al., 1996; Pietenpol et al., 1994). Thus,
Bcl-2 itself can result in a reduction in colony
formation in clonogenic assays. However, in contrast
to the earlier studies, Bcl-2 overexpression per se did
not prevent colony formation in TGCT (Table 1 and
Figure 2). Rather, it appears that the e€ects of Bcl-2
Bcl-2 overexpression in testis tumour cells
EL Arriola et al
1461
Figure 4 Reciprocal expression of Bcl-2 and Bcl-XL in TGCT cells with ectopic overexpression of Bcl-2. (a) Relative levels of Bcl-2
and Bcl-XL, determined by scanning densitometry of immunoblots. (b) Levels of Bcl-2, Bax and Bcl-XL determined by
immunoblotting. 20 mg of whole cell lysates were separated by polyacrylamide gel electrophoresis and analysed by immunoblotting
Figure 5 Sensitization of 833K cells to etoposide-induced
apoptosis by downregulation of Bcl-XL. 833K cells were treated
with 450 nM Bcl-XL antisense (AS) or scrambled (SC) or
transfecting agent contol (Lip) during 70 h and then incubated
with 10.5 mM etoposide (T) or vehicle control (C) for 4 h. (a)
Levels of Bcl-XL determined by immunoblotting. 20 mg of whole
cell lysates were separated by polyacrylamide gel electrophoresis
and analysed by immunoblotting. Membranes were stained with
India Ink to verify equal loading of protein. (b) Determination of
caspase activtion. 20 mg of whole cell lysates were incubated with
substrate and ¯uorescence was measured as described in Materials
and methods
are exerted through an enhancement of apoptosis not
a decrease in colony forming ability. We have shown
that Bcl-2 enhances apoptosis just 4 h after drug
treatment and that the degree of enhancement in
etoposide-induced apoptosis by Bcl-2 overexpression
(approximately fourfold ± from 11 ± 38% at 4 h after
drug treatment) is in good agreement with the
increment in cytotoxicity measured by clonogenic
assays. This is the ®rst observation to our knowledge, of increased sensitivity to drug-induced apoptosis by Bcl-2 overexpression.
Why should Bcl-2 enhance apoptosis? One trivial
explanation is that the Bcl-2 clone used in our study is
mutated. However, this has been ruled out because the
Bcl-2 cDNA used in this work has been demonstrated
to be wild-type by sequencing and functional as a
suppressor of drug-induced apoptosis in other systems
(e.g. CEM-C7 human lymphoblastic leukaemic cells)
(Cosulich et al., 1996) and Drs P Hedge and Ged
Brady personal communication). Thus the apoptosis
activating e€ects of Bcl-2 appear to depend on the cell
type where it is expressed.
Recently two studies have indicated Bcl-2 may not
always function as a suppressor of apoptosis. Firstly,
Chen et al. have described activation of apoptosis by
Bcl-2 following cleavage by caspase-3 which converts
Bcl-2 into Bax-like death e€ector (Cheng et al., 1997).
However, there was no evidence for proteolysis of Bcl2 in the overexpressing 833K cells following drug
treatment suggesting this is not the mechanism of
enhacement of apoptosis in the TGCT cell line (data
not shown). Secondly, Uhlmann et al. showed that a
Bcl-2 cDNA clone lacking the extensive 3' and 5'
untranslated regions (UTR's) possessed a potent cell
death activity in transient transfects of several cell
types (human embryonal kidney, foetal lung, breast
adenocarcinoma and prostate carcinoma (Uhlmann et
al., 1998). This e€ect appears to be related to the very
high expression of Bcl-2 that occurs in transient (but
not stable) transfects when the negative regulatory
UTR regions are deleted (Uhlmann et al., 1998). Our
results are similar to those of Uhlmann et al., however,
Bcl-2 overexpression in testis tumour cells
EL Arriola et al
1462
we ®nd Bcl-2 to behave as an activator of drug-induced
apoptosis in stable transfects. In addition in the TGCT
cells the e€ects of Bcl-2 overexpression are only
manifested when cells are also treated with a
chemotherapeutic agent.
A body of recent evidence suggests that the location
of Bcl-2-family members is important for function
(Hsu et al., 1997; Wolter et al., 1997; Zhu et al., 1996).
Bcl-2 itself is usually associated with several membrane
fractions, including the nucleus, endoplasmic reticulum
and mitochondria, the latter location apparently being
important for function (Kroemer et al., 1997; Zhu et
al., 1996). To determine if the overexpressed Bcl-2 was
targeted to mitochondria, dual staining for Bcl-2 and
mitochondria was performed. As can be seen in Figure
3, Bcl-2 and the mitochondria speci®c dye co-localize.
Thus, failure to protect against drug-induced death is
not a result of lack of targeting to appropriate
membranes. We have also shown that the subcellular
location of Bcl-2 is the same in testicular tumour cells,
where it promotes apoptosis (TGCT), and bladder
tumour cells where it inhibits apoptosis (TCC-RT4)
(Figure 3). Activation rather than suppression of
apoptosis could also occur if the transfected Bcl-2 in
the testis tumours could not function, for example, due
to lack of an essential post-translational modi®cation
step (Haldar et al., 1995, 1996), but could still compete
for molecules to which it (and presumably Bcl-XL)
usually bind to suppress the death pathway. Whilst we
cannot rule out that Bcl-2 is inactive due to
postranslational modi®cation, the immunostaining
data suggest it is localized correctly.
A further mechanism by which Bcl-2 could regulate
apoptosis in TGCT is by a€ecting wild-type p53
function, either by competing with p53 for binding to
binding protein 2 (53BP2) or by restricting access of
p53 to the nucelus (Beham et al., 1997; Naumovski and
Cleary, 1996). 53BP2 is located in the cytoplasm where
it can presumably regulate entry of p53 into the
nucleus. However, as can be seen in Figure 3, p53 is
predominantly in the nucleus in both TGCT cells with
exogenously overexpressed Bcl-2 and RT4 cells with
high endogenous levels of Bcl-2. Thus neither Bcl-2 nor
53BP2 appear to be able to restrict entry of p53 to the
nucleus. This is in contrast to studies in human
prostate cancer cells where overexpressed Bcl-2 acts
as a `gatekeeper' to prevent entry of p53 into the
nucleus (Beham et al., 1997).
Finally, apoptosis susceptibility is regulated by the
relative levels of Bcl-2 family members. We therefore
determined if the balance of protein partners had been
altered. The expression of the pro-apoptotic proteins
Bak and Bax was unchanged. However, as can be seen
in Figure 4, the expression of Bcl-XL shows reciprocal
expression to that of Bcl-2 (compare clone D2 with B6
and B5). Similar ®ndings have been reported in the
thymus and spleen of Bcl-2 and Bcl-XL transgenic mice
suggesting that Bcl-2 and Bcl-XL expression is
coordinated in an inverse fashion (Chao et al., 1995).
Although Bcl-2 and Bcl-XL can substitute for each other
in many in vitro assays it is clear that during
development the expression of the two proteins are
di€erentially regulated (Krajewski et al., 1994, 1995).
This raises the important question of whether the two
molecules have distinct functions. It has been shown
that in certain cell types the two proteins show di€erent
potencies in repression of apoptosis. This was ®rst
demonstrated by Gottschalk et al. who showed Bcl-XL
but not Bcl-2 could repress immunosuppressant-induced
apoptosis in murine B-cells (Gottschalk et al., 1994).
More recently Bcl-XL has been demonstrated to be a
more ecient repressor of chemotherapy-induced
apoptosis in murine lymphoid cells engineered to
overexpress either Bcl-2 or Bcl-XL (Simonian et al.,
1997b). Interestingly, the di€erential protection depends
on the mechanism of action of the drug employed, BclXL was up to 50% more ecient than Bcl-2 in
protection against etoposide and cis-platin induced
apoptosis but had similar ecacy against g-irradiation
or vinca-alkaloid-induced apoptosis (Simonian et al.,
1997b). Here we show that the TGCT cells with
downregulated Bcl-XL were approximately fourfold
more sensitive to etoposide induced apoptosis and 1 ±
2-fold more sensitive to the non-genotoxic agent NMF
and the vinca alkaloid Vincristine. The importance of
Bcl-XL rather than Bcl-2 as a determinant of
chemoresistance is also implicated by studies of drug
resistant HL-60 human promyelocytic leukaemia cells
and B9 human myeloma cells (Han et al., 1996;
Schwarze and Hawley, 1995). Our results, in which we
achieve a twofold sensitization to apoptosis by downregulation of Bcl-XL (Figure 5), similarly support a key
role for Bcl-XL in modulation of drug sensitivity. In
testicular tumours Bcl-2 could not substitute for Bcl-XL
in the suppression of drug-induced apoptosis.
In conclusion, our data suggest a scenario in which
Bcl-XL is an important protector against chemotherapy-induced apoptosis in TGCT. Bcl-2 does not appear
to be able to substitute for Bcl-XL, whether this is
because it is not activated in these cells or necessary cofactors, e.g. APAF's are limiting is a subject of future
studies. The mechanism by which Bcl-2 promotes
apoptosis may in part be the reciprocal downregulation of Bcl-XL. Certainly downregulation of Bcl-XL by
AS oligonucleotides can enhance drug-induced apoptosis. However, the overexpressed Bcl-2 may also
compete with Bcl-XL for other proteins necessary for
the suppression of apoptosis. Our results suggest the
role Bcl-2 family members have in protection against
apoptosis is cell type speci®c. As TGCT are one of the
few curable solid tumours it also suggests caution in
the interpretation of some authors that tumours with
Bcl-XL rather than Bcl-2 are likely to be the more
dicult to treat (Simonian et al., 1997b). It remains to
be seen whether drug resistant TGCT will be found to
express higher levels of Bcl-XL. Our results suggest BclXL will form a useful target to augment chemotherapy
in this tumour type.
Materials and methods
Cell culture
Characteristics of cell lines used are as detailed previously
(Chresta et al., 1996). All cell lines were grown routinely as
monolayers in RPMI 1640 medium with 10% (v/v) heat
inactivated fetal calf serum, 2 mM glutamine, at 378C in a
humidi®ed atmosphere of 5% CO2. Transfected clones of
833K were grown under identical conditions in the
presence of 250 mg/ml G418. Cells were used over a
maximum of 12 passages in order to minimize changes
which might occur as a result of long term culture.
Bcl-2 overexpression in testis tumour cells
EL Arriola et al
Transfection
833K cells were transfected with the pcDNA3 plasmid
containing a cDNA of the open reading frame of human
bcl-2 (a kind gift from S Cosulich) or vector control
(833K.neo cells). Transfections were carried out by
lipofection with the TfxTM-50 reagent (Promega) as
described by the manufacturer. After 2 ± 3 weeks selection, single cell clones were isolated, expanded and frozen
stocks were made.
Bcl-XL antisense treatment of 833K TGCT cells
An antisense phosphorothioate oligodeoxynucleotide to the
coding region of the Bcl-XL mRNA was used (Dibbert et
al., 1998) kindly provided by Dr Uwe ZangemeisterWittke, Department of Oncology, University Hospital
Zurich, Switzerland. Sequence of antisense Bcl-XL, 5'TGT ATC CTT TCT GGG AAA GC-3'; and scrambled
Bcl-XL, 5'-TAA GTT CCG ATG CGA CTT GT-3'. Cells
were transiently transfected using 7 mg/ml lipofectin (Gibco
BRL) in 1% serum containing medium (Optimem). Cells
were treated with the oligonucleotides (antisense or
scrambled) or vehicle control (water) (450 nM) for 70 h.
They were then either treated with etoposide or vehicle
control (DMSO) for 4 h and lysates prepared for caspase
assays and immunoblotting, to determine Bcl-XL levels.
Immunoblotting
Cells were detached by scraping into ice cold versene and
collected by centrifugation. Cells were rinsed in ice cold
PBS containing the following protease inhibitors: aprotinin
(1%), benzamidine (1 mM), phenylmethylsulphonyl¯uoride
(1 mM), trypsin/chymotrypsin inhibitor (10 mg/ml) then
resuspended in 200 ml of PBS with protease inhibitors
and sonicated using a Soniprep 150 (MSE) for 10 s at 12
microns, allowed to cool, then sonicated again for 10 s at
16 microns. Protein concentration was determined using a
BioRad protein assay kit and 20 mg of protein used per
sample. Samples were boiled for 5 min in SDS sample
bu€er and then electrophoresed on 1 mm mini polyacrylamide gels at 120 V for 2 h. Proteins were transferred to
Immobilon-P membrane at 80 V for 1.5 h.
allowed to adhere overnight. They were then treated for
1 h with the indicated drug or appropriate vehicle control.
The drug was then washed from the cells with two changes
of PBS and the cells reincubated in fresh medium for 10 ±
14 days. Colonies were stained with methylene blue and
colonies of greater than 50 cells were counted. Results are
the mean and standard deviation of three independent
experiments and triplicate plates were used in each
individual experiment.
DNA fragmentation ®lter binding assay (FBA)
Apoptosis was quanti®ed using the ®lter binding assay
(FBA) which is an assay of endonuclease activation
(Bertrand et al., 1991). The assay measures the percentage
of low molecular weight DNA by size separation through a
2 mm ®lter; at pH 10, high molecular weight DNA is
retained on the ®lter. Etoposide-induced topoisomerase
associated DNA damage does not result in elution of DNA
through the ®lter. This type of DNA damage is not
revealed until after treatment with proteases and incubation of DNA at pH 12.1 to denature the DNA. We have
previously shown apoptosis quanti®ed by this assay
corresponds with percentage apoptosis determined by
counting trypan-blue-negative/Hoescht 33258 positive cells
(Chresta et al., 1996). Cells in early logarithmic phase
growth were labelled for 24 h with 0.015 mCi
[14C]thymidine/ml, wahsed and reincubated in fresh
medium for 1 h. Cells were then treated with 25.7 mM
etoposide for 1 h, washed with PBS (378C) and reincubated
in fresh medium at 378C until harvested at 4 h.
Determination of caspase activation
The substrate used was Ac-DEVD-AMC from Calbiochem/Novabiochem which was made to 5 mM stock in
DMF. On the day of the assay, substrate stock was diluted
in water to 500 mM and kept on ice. 20 mg of cell lysate
were diluted to equal volumes (20 ml) in NP40 lysis bu€er
(NET) and 1/10th of the volume of substrate was added.
Exactly 10 min later, the reaction was stopped by adding
the mix into 2 ml of water. Fluorescence was read on the
¯uorimeter at an excitation wavelength of 380 nm and
emission wavelength of 460 nm.
Antibodies
Immunostaining
The antibodies used for this study were, Bcl-2: or FITCconjugated anti-human Bcl-2 monoclonal Ab-124 (1 : 25 for
immunostaining) from Dako (High Wycombe, UK); Bax:
anti-human Bax polyclonal antibody from Pharmingen
(1 : 50 for immunostaining) (San Diego, USA); Bcl-X: antihuman Bcl-X polyclonal antibody from Transduction
Laboratories (1 : 50 for immunostaining) (Exeter, UK);
p53: anti-human p53 DO1 monoclonal antibody from
Oncogene Science (Cambridge, USA) (1 : 200 for immunostaining). Goat anti-rabbit or rabbit anti-mouse horse
radish peroxidase conjugated secondary antibodies from
Dako (High Wycombe, UK) were used as described by the
manufacturers and bands were visualized using ECL
reagents from Amersham (UK). For immuno¯uorescence,
CyTM-conjugated donkey anti-mouse or goat anti-rabbit
secondary antibodes were used (Jackson Immunoresearch,
Pennsylvania, USA). For mitochondria staining, Mito
Tracker Red CMXRos from Molecular Probes (Oregon,
USA) was used.
106 cells/ml were seeded onto eight chamber slides
(Falcon) and incubated at 378C overnight to adhere.
For experiments in which mitochondrial staining was
studied the cells were incubated with a 1 : 250 dilution of
MitoTracker in culture medium for 30 min at 378C. They
were rinsed once with PBS and ®xed in methanol:acetone
solution (1 : 1) for 10 min at room temperature. Fixed
cells were rinsed twice with PBS then incubated with
antibody diluted in PBS with 0.1% fetal calf serum for
1 h at room temperature before rinsing with PBS and
incubating with ¯uorescent conjugated secondary antibody for 1 h at room temperature. Cells were then
washed with PBS and DNA stained with Hoescht 33258
(1 mg/ml in PBS) for 5 min.
Abbreviations
FBA, ®lter binding assay; NMF, N-methylformamide;
TGCT, testicular germ cell tumours; TCC, transitional
cell carcinomas.
Cytotoxicity assay
The drugs used were Etoposide (Sigma) and cis-Platin
(Sigma). Drug sensitivity was measured by clonogenic
assay. 1000 cells were plated in 6-well plates (2 cm2) and
Acknowledgements
We thank Dr Sabina Cosulich for providing the pcDNA3Bcl-2 plasmid and Dr Uwe Zangemeister-Wittke, Depart-
1463
Bcl-2 overexpression in testis tumour cells
EL Arriola et al
1464
ment of Oncology, University Hospital Zurich, Switzerland
for providing AS Bcl-XL oligonucleotides. We would also
like to thank Dr John Masters, University College London
for helpful advice and discussions. This project has been
supported by the Cancer Research Campaign, UK and by
a scholarship granted to EL Arriola from the Government
of Mexico, CONACyT.
References
Beham A, Marin MC, Fernandez A, Herrmann J, Brisbay S,
Tari AM, Lopez-Berestein A, Lozano G, Sarkiss M and
McDonnell TJ. (1997). Oncogene, 15, 2767 ± 2772.
Bertrand R, Sarang M, Jenkin J, Kerrigan D and Pommier
Y. (1991). Cancer Res., 51, 6280 ± 6285.
Boersma AWM, Nooter K, Burger H, Kortland CJ and
Stoter G. (1997). Cytometry, 27, 275 ± 282.
Burger H, Nooter K, Boersma AWM, Kortland CJ and
Stoter G. (1997). Int. J. Cancer, 73, 592 ± 599.
Chao DT, Linette GP, Boise LH, White LS, Thompson CB
and Korsmeyer SJ. (1995). J. Exp. Med., 182, 821 ± 828.
Cheng EHY, Kirsch DG, Clem RJ, Ravi R, Kastan MB,
Bedi A, Ueno K and Hardwick JM. (1997). Science, 278,
1966 ± 1968.
Chresta CM, Masters JRW and Hickman JA. (1996). Cancer
Res., 56, 1834 ± 1841.
Cosulich SC, Green S and Clarke PR. (1996). Curr. Biol., 6,
997 ± 1005.
Dibbert B, Daigle I, Braun D, Schranz C, Weber M, Blaser
K, Zangemeister-Wittke U, Akbar AN and Simon HU.
(1998). Blood, 92, 778 ± 783.
Dive C and Hickman JH. (1991). Br. J. Cancer, 64, 192 ± 196.
Einhorn LH. (1990). J. Clin. Oncol., 8, 1777 ± 1781.
Farrow SN and Brown R. (1996). Curr. Opin. Gen. Dev., 6,
45 ± 49.
Fisher TC, Milner AE, Gregory CD, Jackman AL, Aherne
GW, Hartley JA, Dive C and Hickman JA. (1993). Cancer
Res, 53, 3321 ± 3326.
Gottschalk AR, Boise LH, Thompson CB and Quintans J.
(1994). Proc. Natl. Acad. Sci. USA, 91, 7350 ± 7354.
Haldar S, Chintapalli J and Croce CM. (1996). Cancer Res.,
56, 1253 ± 1255.
Haldar S, Jena N and Croce CM. (1995). Proc. Natl. Acad.
Sci. USA, 92, 4507 ± 4511.
Han Z, Chatterjee D, Early J, Pantazis P, Hendrickson EA
and Wyche JH. (1996). Cancer Res., 56, 1621 ± 1628.
Hashimoto H, Chatterjee S and Berger NA. (1995). Cancer
Res., 55, 4029 ± 4035.
Hockenbery DM, Zutter M, Hickey W, Nahm M and
Korsmeyer SJ. (1991). Proc. Natl. Acad. Sci. USA, 88,
6971 ± 6965.
Hsu YT, Wolter KG and Youle RJ. (1997). Proc. Natl. Acad.
Sci. USA, 94, 3668 ± 3672.
Kamesaki S, Kamesaki H, Jorgensen TJ, Tanizawa A,
Pommier Y and Cossman J. (1993). Cancer Res., 53,
4251 ± 4256.
Kluck RM, BossyWetzel E, Green DR and Newmeyer DD.
(1997). Science, 275, 1132 ± 1136.
Knudson CM and Korsmeyer SJ. (1997). Nature Genetics,
16, 358 ± 363.
Korsmeyer SJ. (1995). Trends Genet., 11, 101 ± 105.
Krajewski S, Bodrug S, Krajewska M, Shabaik A, Gascoyne
R, Berean K and Reed JC. (1995). Am. J. Pathol., 146,
1309 ± 1319.
Krajewski S, Krajewska M, Shabaik A, Wang H, Irie S, Fong
L and Reed JC. (1994). Cancer Res., 54, 5501 ± 5507.
Kroemer G. (1997). Nature Med., 3, 614 ± 620.
Kroemer G, Zamzami N and Susin SA. (1997). Immunol
Today, 18, 44 ± 51.
Lock RB and Stribinskiene L. (1996). Cancer Res., 56, 4006 ±
4012.
Lu QL, Poulsom R, Wong L and Hanby AM. (1993). J.
Pathol., 169, 431 ± 437.
Naumovski L and Cleary ML. (1996). Mol. Cell. Biol., 16,
3884 ± 3892.
O'Reilly LA, Huang D and Strasser A. (1996). EMBO J., 15,
6979 ± 6990.
Oltvai ZN and Korsmeyer SJ. (1994). Cell, 79, 189 ± 192.
Oltvai ZN, Milliman CL and Korsmeyer SJ. (1993). Cell, 74,
609 ± 619.
Pietenpol JA, Papadopoulos N, Markowitz S, Willson JKV,
Kinzler KW and Vogelstein B. (1994). Cancer Res., 54,
3714 ± 3717.
Reed JC. (1994). J. Cell Biol., 124, 1 ± 6.
Reed JC. (1997). Nature, 387, 773 ± 776.
Schwarze MMK and Hawley RG. (1995). Cancer Res., 55,
2262 ± 2265.
Simonian PL, Grillot D and Nunez G. (1997a). Blood, 90,
1208 ± 1216.
Simonian PL, Grillot DAM, Merino R and Nunez G. (1996).
J. Biol. Chem., 271, 22764 ± 22772.
Simonian PL, Grillot DAM and Nunez G. (1997b).
Oncogene, 15, 1871 ± 1875.
Uhlmann EJ, Subramanian T, Vater CA, Lutz R and
Chinnadurai G. (1998). J. Biol. Chem., 273, 17926 ± 17932.
Walker A, Taylor ST, Hickman JA and Dive C. (1997).
Cancer Res., 57, 1939 ± 1945.
Wang HG, Rapp UR and Reed JC. (1996). Cell, 87, 629 ±
638.
Wolter KG, Hsu YT, Smith CL, Nechushtan A, Xi XG and
Youle RJ. (1997). J. Cell Biol., 139, 1 ± 12.
Yang E, Zha J, Jockel J, Boise LH, Thompson CB and
Korsmeyer J. (1995). Cell, 80, 285 ± 291.
Yang J, Liu XS, Bhalla K, Kim CN, Ibrado AM, Cai JY,
Peng TI, Jones DP and Wang XD. (1997). Science, 275,
1129 ± 1132.
Yin DX and Schimke RT. (1995). Cancer Res., 55, 4922 ±
4928.
Zhu W, Cowie A, Wasfy GW, Penn LZ, Leber B and
Andrews DW. (1996). EMBO J., 15, 4130 ± 4141.
Zou H, Henzel WJ, Liu XS, Lutschg A and Wang XD.
(1997). Cell, 90, 405 ± 413.