Oncogene (1997) 14, 1735 ± 1746
 1997 Stockton Press All rights reserved 0950 ± 9232/97 $12.00
p53 mediated sensitization of squamous cell carcinoma of the head and neck
to radiotherapy
Kathleen F Pirollo1,3, Zhengmei Hao1, Antonina Rait1.3, Yng-Ju Jang1, Willard E Fee Jr1,
Patricia Ryan2, Yawen Chiang2, and Esther H Chang1,3
1
2
Department of Surgery, Division of Otolaryngology, 300 Pasteur Drive, Stanford University, Stanford, California 94305-5328;
Genetic Therapy Incorporated/Novartis, 19 First®eld Avenue, Gaithersburg, Maryland 20878, USA
Radiation resistant squamous cell carcinoma of the head
and neck cell line JSQ-3 carries a mutant form of tumor
suppressor gene p53. Treatment of these cells with an
adenoviral vector containing wild-type p53 (Av1p53) was
able to inhibit their growth in vitro and in vivo while
having no e€ect on normal cells. More signi®cantly,
introduction of wtp53 also reduced the radiationresistance level of this cell line in vitro, in a viral dosedependent manner. Furthermore, this radiosensitization
also carried over to the in vivo situation where the
response of JSQ-3 cell-induced mouse xenografts to
radiotherapy was markedly enhanced after treatment
with Av1p53. Complete, long-term regression of the
tumors for up to 162 days was observed when a single
dose of Av1p53 was administered in combination with
ionizing radiation, demonstrating the e€ectiveness of this
combination of gene therapy and conventional radiotherapy. This sensitization of tumors to radiation therapy
by replacement of wtp53 could signi®cantly decrease the
rate of recurrence after radiation treatment. Since
radiation is one of the most prevalent forms of adjunctive
therapy for a variety of cancers, these results have great
relevance in moving toward an improved cancer therapy.
Keywords: p53; adenovirus; head and neck; cancer;
radiosensitization
Introduction
The treatment of tumors which fail radiation therapy
has been a major clinical challenge, not only for head
and neck tumors, but for other forms of cancers as
well. Each year in the US approximately 40 000
individuals will be diagnosed with squamous cell
carcinoma of the head and neck (SCCHN) and upper
aerodigestive tract (American Cancer Society, 1993).
This disease not only has a profound e€ect upon
speech, swallowing and physical appearance, but has
an overall survival rate of only approximately 50%, a
rate which has remained relatively unchanged for more
than 30 years (American Joint Committee on Cancer,
1983).
Despite advances in current treatments, patients with
advanced disease have a poor prognosis. More than
two-thirds of the individuals in whom the diagnosis of
Correspondence: EH Chang
3
Present address: Dept. of Otolaryngology, Lombardi Cancer Center,
The Research Building/E420, Georgetown University Medical
Center, 3970 Reservoir Rd. NW, Washington, D.C. 20007
Received 13 December 1996; accepted 27 February 1997
SCCHN is made will present with stage III or stage IV
disease (Dimery and Hong, 1993) at the primary and/
or nodal sites. Despite optimal local therapy, 50 ± 60%
of these patients will ultimately develop local recurrence and 30% or more will develop distant metastatic
disease. Treatment of recurrent disease no amenable to
surgery is palliative at best and without long-term
bene®t (Dimery and Hong, 1993). Furthermore, the
long-term outlook for those surviving their ®rst
malignancy is overshadowed by a 10 ± 40% rate of a
second primary malignancy (Dimery and Hong, 1993).
Treatment often requires aggressive adjunctive therapy
after surgery, with radiation being the treatment or
choice, either as a single modality, or as part of a
multimodal course of treatment. Unfortunately, a
signi®cant number of SCCHN have been found to be
resistant to radiotherapy, a major factor in the high
rate of recurrence observed for this type of malignancy.
The development of an e€ective method for sensitizing
SCCHN tumors to radiotherapy therefore, will have a
profound e€ect on the treatment of this disease.
Mutant (mt) forms of the tumor suppressor gene
p53 have been associated in a number of studies with
poor clinical prognosis for malignancies such as
carcinoma of the breast, lung, prostate, and bladder,
soft tissue sarcoma, Wilm's tumor and various
leukemias and lymphomas (reviewed in Lowe, 1994;
Chang et al., 1995; Newcomb, 1995). p53 may also
play a role in the development and progression of
SCCHN. Depending upon the tissue source and
method of detection of abnormal p53, both the gene
and its expression have been identi®ed in 33% to 100%
of head and neck cancers. Mutations have been found
in exons 4 through 9 with a hot spot in the codon
238 ± 248 region (reviewed in Field et al., 1993;
Brachman, 1994). The presence of abnormal p53 may
also correlate with a history of heavy smoking and
drinking which are the primary environmental factors
associated with SCCHN (Field et al., 1993). Studies
indicate that the presence of p53 mutations may also
be indicative in SCCHN of a higher frequency of, and
shorter median time to, recurrence of the tumor
(Brachman, 1994). In a recent study Shin et al. also
reported that higher levels of mtp53 expression in
primary SCCHN, due presumably to the stabilization
and longer half-life of the mutant form of the protein,
were associated with both earlier recurrence and
development of second primary tumors and could be
an important adverse prognostic factor for survival
(Shin et al., 1996).
A role for wild type (wt) p53 in the control of
cellular proliferation by induction of programmed cell
death (apoptosis) (Lowe et al., 1993, 1994; Clarke et
p53 sensitizes head and neck cancer to radiotherapy
KF Pirollo et al
1736
al., 1993; Yonish-Rouach et al., 1991) and the
regulation of cell cycle events (Kastan et al., 1991,
1992; Kuerbitz et al., 1992) has been established.
Wtp53 has been found to induce apoptosis in mouse
myeloid leukemia cells (Yonish-Rouach et al., 1991).
Furthermore, two separate groups of investigators
using p53 knockout mice which lack wtp53 have
demonstrated that thymocytes from these animals are
resistant to g-radiation, or etoposide (a topoisomerase
II inhibitor) induced apoptosis, whereas thymocytes
from control mice with wtp53 showed radiosensitivity
to g-irradiation (Lowe et al., 1993; Clarke et al.,
1993). Kastan and his colleagues (Kastan et al., 1991,
1992; Kuerbitz et al., 1992) demonstrated that after
exposure to g-irradiation, cells containing wtp53
display a block in the G1 phase of the cell cycle,
while cells expressing only mutant (mt), or no p53,
continued to progress through the cell cycle. It has
more recently been suggested that p53 may play a role
in the G2/M checkpoint as well (Powell et al., 1995;
Russell et al., 1995; Stewart et al., 1995; Agarwal et
al., 1995).
The correlation between p53 and apoptosis, in
conjunction with apparently normal development of
mice lacking wtp53 (Donehower et al., 1992), and the
observations of a post-irradiation G1 block, suggests
that wtp53 functions in the regulation of the cell after
DNA damage or stress rather than during proliferation
and development. As the presence of mtp53 has also
been shown to correlate with increased radiation
resistance in some human tumors and cell lines
(Agarwal et al., 1995; Donehower et al., 1992; Lee
and Bernstein, 1993; O'Conner et al., 1993; McIlwrath
et al., 1994) and a high percentage of head and neck
tumors fail radiation therapy, it is conceivable that a
cause and e€ect relationship exists between the lack of
functional wtp53 found in a large number of SCCHN
and this radiation resistance. The replacement of wtp53
may, therefore, ameliorate the high level of radiation
resistance frequently observed in these cancers.
The development over the last few years of methods
for the introduction of genes into cells in vivo has made
such therapeutic interventions a possibility. Of these
gene delivery systems, adenovirus vectors have been
found to be particularly advantageous over traditional
transfection methods and retroviral vectors for a
number of reasons: (i) they have suciently high
MOI
C
JSQ-3
10 20 40 80 160 320
MOI
C
cloning capacity to accommodate most cDNAs; (ii)
they can infect both replicating and quiescent cells; (iii)
they can be grown to high titers and easily puri®ed (iv)
they can be rendered replication defective; and (v) they
can deliver DNA with high eciency both in vitro and
in vivo (reviewed in Bramson et al., 1995). In addition,
adenovirus vectors are often preferable over retroviruses in that they are not subject to the possible
adverse side e€ects which may be encountered due to
genomic integration and deletion. Adenovirus also
possesses a natural tropism for the epithelium of the
aerodigestive tract which makes them particularly
attractive in gene therapy for cancers of the head and
neck region. Therefore, in spite of some disadvantages,
such as the induction of immunogenicity, the use of an
adenoviral construct to introduce wtp53 is a promising
system for treating local SCCHN tumors and even the
microscopic residual disease which is a major cause of
the frequent post-surgical recurrence of this type of
cancer.
In this study we have examined the ability of wtp53,
introduced by means of an adenoviral construct
(Av1p53), to sensitize SCCHN to radiation treatment
both in an in vitro cell culture system and in an in vivo
xenograft mouse model. Our ®ndings suggest that this
is an e€ective means of sensitizing these tumors to
radiation resulting in a greatly reduced rate of
recurrence. Moreover, the sensitization of residual
tumors through the combination of wtp53 and
radiotherapy has the potential for use as a new
therapeutic modality in the treatment of those tumors
which initially failed radiation therapy, not only in
cancers of the head and neck but in other forms of
cancer as well.
Results
E€ect of Av1p53 on in vitro cell growth
To determine if replacement of wtp53 could a€ect the
growth of SCCHN tumor cells in an in vitro situation,
exponentially growing SCCHN cell lines JSQ-3, SQ20B and SCC61, were seeded at 36104 cells/well of a
24-well tissue culture plate. Twenty-four hours later, at
approximately 50% con¯uency, the cells were treated
with Av1p53 in doses varying from 10 to 320
SQ20-B
10 20 40 80 160 320
MOI
p53
p53
p53
LacZ
LacZ
LacZ
MOI
p53
LacZ
C
SK-OV-3
10 20 40 80 160 320
MOI
C
SK-BR-3
10 20 40 80 160 320
MOI
p53
p53
LacZ
LacZ
SCC61
C 10 20 40 80 160 320
H500
C 10 20 40 80 160 320
Figure 1 Growth inhibition of tumor cell lines by wtp53. SCCHN cell line (JSQ-3, SQ-20B, SCC 61); ovarian (SK-OV-3); and
breast (SK-BR-3) carcinoma cell lines; and a normal skin ®broblast cell line (H500), were treated with increasing doses of either an
adenoviral vector carrying wtp53 (p53) or the control adenoviral vector which carries the LacZ gene in place of p53 (LacZ). Two
virus doses were given 24 h apart. Four days after the second treatment the cells were stained with Giemsa stain (MOI=Multiplicity
of Infection, the number of virus plaque forming units per cell)
p53 sensitizes head and neck cancer to radiotherapy
KF Pirollo et al
Multiplicity of Infection (MOI) (56105 ± 1.66107
PFU). As a control, cells were also treated in an
identical fashion with Av1LacZ4, the viral vector
carrying the b-galactosidase gene in place of p53.
Forty-eight hours after the ®rst virus treatment, a
second dose of virus was administered. Four days after
the second virus treatment the cells were ®xed and
stained with Giemsa stain. Visual observation of the
Figure 2 Comparison of cell growth inhibition and killing by Av1p53 and Av1LacZ4. Light microscopic observation of SCCHN
tumor cell line JSQ-3 treated as in Figure 1 was made. (a) untreated cells; (b) cells after treatment with 10 MOI Av1p53; (c) after 20
MOI Av1p53; (d) after 40 MOI Av1p53; (e) after 80 MOI Av1p53; (f) after 160 MOI Av1p53; (g) after treatment with 80 MOI of
Av1LacZ4; and (h) after 320 MOI Av1LacZ4 (Magni®cation 6200)
1737
p53 sensitizes head and neck cancer to radiotherapy
KF Pirollo et al
1738
status of the cells was made with regard to morphology
and to approximate the amount of cell killing. As
shown in Figures 1 and 2, signi®cant inhibition of cell
growth with doses of Av1p53 as low as 20 MOI was
observed in the JSQ-3 cells. These radiation resistant
cells, which were derived from a tumor which failed
radiotherapy, have been shown to carry a mutant form
of p53 (Jung et al., 1992). Vacuolization was observed
down to 10 MOI and by 80 MOI very few live cells
were evident (Figure 2). However, the only notable
inhibition of cell growth by Av1LacZ4 in this cell line
was with doses of 160 and 320 MOI (Figures 1 and 2)
and that is still relatively insigni®cant in comparison to
the e€ect of Av1p53.
Radioresistant cell line SQ-20B, which was also
derived from a tumor which failed radiotherapy,
exhibits an alteration in the p53 gene expression
(Jung et al., 1992). These cells showed a level of
growth inhibition with Av1p53 similar to that observed
in JSQ-3 cells (Figure 1). However, more growth
inhibition by Av1LacZ4 was observed at high doses in
these cells than in the JSQ-3 cells, indicating some nonspeci®c e€ect in this cell line. In contrast, both the p53
and LacZ containing adenovirus constructs had only
minimal e€ect on the growth of radiosensitive SCCHN
cell line SCC61, and that only at the highest doses
(Figure 1). The p53 status of this cell line has not been
determined.
More signi®cantly, treatment with either Av1p53 or
Av1LacZ4 had no e€ect on the growth of a normal
human ®broblast cell line (H500), which has been
shown to contain only wtp53 (Srivastava et al., 1992).
Examination under higher magni®cation, as had been
done for JSQ-3 cells, con®rmed this ®nding (data not
shown). The clear area observed at 160 and 320 MOI
with Av1p53 in these cells is an artifact which occurred
during Giemsa staining (Figure 1).
Also shown in Figure 1 is the e€ect of treatment
with Av1p53 or Av1LacZ4 on human ovarian and
breast carcinoma cell lines SK-OV-3 and SK-BR-3,
respectively, both of which have mutant or no p53
(Johnson et al., 1991; Elstner et al., 1995). The
treatment protocol for these cell lines was identical to
that used for the SCCHN lines. Both cell lines display
a strong response to the replacement of wtp53. SK-BR3 cells display not only growth inhibition, but almost
complete cell killing at a dose of Av1p53 as low as
10 MOI. This cell line was also signi®cantly more
sensitive to the LacZ containing adenoviral construct.
SK-OV-3, while also highly sensitive to Av1p53,
showed more speci®city in its response. Here,
signi®cant growth inhibition and cell killing with
Av1LacZ4 were observed only at the highest doses.
These results indicate that the replacement of wtp53 is
not just speci®c for SCCHN and may be an e€ective
treatment for various types of cancer.
The replacement of wtp53 was shown by these
experiments to e€ect the growth of JSQ-3 cells in a p53
speci®c manner. As mentioned above, these cells
possess both a mtp53 gene and a radiation resistant
phenotype. These cells therefore, are an ideal model
with which to examine the usefulness of combining
wtp53 gene therapy with radiotherapy in the treatment
of SCCHN. Accordingly, the remainder of the studies
described in this report employ this SCCHN tumor cell
line.
In vitro sensitization to radiation treatment by Av1p53
The presence of a mutated form of p53 has been shown
to correlate with increased radiation resistance in some
human tumors and cell lines (Lee and Bernstein, 1993;
Lowe et al., 1993). Therefore, the replacement of wtp53
may ameliorate the high level of radiation resistance
frequently observed in head and neck cancers. We
therefore examined the e€ect that replacement of
wtp53 has on the radiation survival level of radiation
resistant SCCHN cell line JSQ-3.
JSQ-3 cells were treated one time with 5, 10 or
20 MOI of Av1p53 for either 24 or 36 h and the D10
values (radiation dose required to reduce survival to
10%) of the cells were compared to that of untreated
JSQ-3 cells. As shown in Figure 3, infection with
Av1p53 was able to decrease the radioresistance level
of JSQ-3 cells in a dose and time dependent manner. In
all instances, a greater e€ect was observed when the
cells were exposed to g-rays 36 h post-virus treatment
compared to 24 h. However, with 20 MOI the D10
value was reduced from the highly resistant level found
in the control cells (6.02+0.11 Gy) to 4.40 Gy even at
24 h post-treatment. While the D10 values with 20 MOI
of Av1p53 at 24 and 36 h are still somewhat higher
than that of the radiosensitive SCCHN cell line SCC61
(D10=2.97+0.25), D10 values of 4.40 Gy (24 h) and
4.30 Gy (36 h) are similar to that of a radiosensitive
human ®broblast cell line H500 (D10=4.50+0.05) and
are now clearly in the range considered to be
radiosensitive.
Moreover, the large di€erences in radioresistance
levels seen before and after treatment with 20 MOI,
and even 10 MOI (D10=4.63+0.12 Gy), of Av1p53 are
statistically signi®cant (P50.001). In terms of survival,
a decrease of almost 2 Gy represents a dramatic
increase in sensitization to the killing e€ects of
ionizing radiation. By comparison, the radiation
Figure 3 E€ect of replacement of wtp53 on the radiation
resistance level of an SCCHN cell line. Radiation resistant
SCCHN cell line JSQ-3 was treated for 24 or 36 h with 5, 10 or
20 MOI of an adenoviral vector carrying wtp53 (Av1p53), or 20
MOI of the control vector containing the LacZ gene in place of
p53 (Av1LacZ4), prior to exposure to varying doses of girradiation. The cells were replated and stained with Giemsa
approximately 2 weeks later. Survival was measured as colony
forming ability, with colonies of 50 or more normal appearing
cells being counted. Radiation resistance levels are given as D10
values, the radiation dose (in Gy) required to reduce survival to
10%. Control represents untreated JSQ-3 cells. Error bars
represent the standard error of the mean (s.e.m.) of 2 ± 9 values
p53 sensitizes head and neck cancer to radiotherapy
KF Pirollo et al
resistance level of the cells treated with 20 MOI of
Av1LacZ4 was only minimally reduced, to a value
approximately that observed with only 5 MOI of the
p53 construct. Survival curves for the above data are
given in Figure 4.
As stated above, the radiation resistance level of the
Av1p53 treated cells decreased with increasing time, up
to 36 h after one viral treatment. However, by 48 h
post-infection, the D10 values began to increase. This
rise was inversely proportional to Av1p53 dose, the
amount of increase being smaller with increasing MOI
(data not shown). Conversely, if two treatments with
Av1p53 were given 24 h apart the decrease in radiation
resistance observed at 36 h was maintained, or even
ampli®ed 24 h after the second round of infection (data
not shown). These ®ndings indicate that those cells
which had not been transduced with the virus after one
treatment continue to grow and maintain their initial
radiation resistant phenotype thereby leading to an
overall increase in resistance of the culture. A second
virus treatment however, leads to an increased number
of e€ected cells in the population, resulting in increased
sensitization to ionizing radiation.
Expression of exogenous wtp53 protein in infected cells
To demonstrate that the virally transduced wtp53 is
being expressed, the level of p53 protein in JSQ-3 cells
was examined by Western blot analysis, using the
pantropic anti-p53 monoclonal antibody Ab-2, 36 h
post-infection with increasing doses of Av1p53. Thirtysix hours was chosen based upon the timing of the
maximal reversal of radioresistance after one viral
treatment as discussed above. As shown in Figure 5,
infection with increasing MOI of Av1p53 results in a
concomitant increase in expression of the exogenous
wtp53 (top band), as compared to uninfected JSQ-3
cells (C). An increase is also observed in the lower
band, the endogenous mtp53 band, as viral dose
increases. A similar observation was made by Liu et
al. (1994) after treatment of two di€erent SCCHN cell
lines with a wtp53-adenoviral construct under control
of the CMV promoter. Data indicates that, at least in
murine cell lines, wild-type, but not mutant, p53 is
capable of transactivating its own promoter through a
sequence nearly identical to the NF-kB-binding site
(Dee et al., 1993). This increase in endogenous p53
may therefore be attributable to this transcriptional
autoregulation. An increase in p53 expression often
results in an induction of the WAF-1 gene, an
important downstream e€ector of p53 function (ElDeiry et al., 1994), which is involved in control of the
cell cycle (Xiong et al., 1993; Harper et al., 1993). We
are currently examining the e€ect of introduction of
wtp53 on WAF-1 expression in these cells. These
results demonstrate that infection with Av1p53 results
in production of immunoreactive p53 protein within
the transduced cells in a dose dependent manner.
Enhancement of the e€ect of radiotherapy by wtp53
replacement in vivo
The results described above indicated that treatment
with Av1p53 was able to individually a€ect both tumor
growth and the level of radiation resistance in vitro.
The clinical relevance of this form of gene therapy
would be greatly enhanced if synergism between
replacement of wtp53 and radiation therapy could be
demonstrated in vivo. Such an e€ect may result in a
reduction of the dose of radiation required for e€ective
MOI
C
5
10 20 30 40 50 60 70 80 90 100 150
Figure 5 Western analysis of p53 expression in JSQ-3 treated
cells transduced with increasing doses of Av1p53. p53 protein
expression was examined 36 h after viral infection C represents,
uninfected JSQ-3 cells. The amount of Av1p53 for viral infection
is given as multiplicity of infection (MOI)
Figure 4 Survival curves, after graded doses of g-irradiation, for JSQ-3 cells treated for 24 or 36 h with Av1p53 or Av1LacZ4 (see
legend to Figure 3). Curves are plotted as the log of the surviving fraction vs radiation dose. Points are plotted as the s.e.m. of 2 ± 9
experiments
1739
p53 sensitizes head and neck cancer to radiotherapy
KF Pirollo et al
1740
treatment of head and neck cancer and a greatly
decreased rate of recurrence after radiotherapy.
To ascertain the e€ectiveness of the combination
therapy, 4 ± 6 week old female athymic nude mice were
injected subcutaneously on the lower back above the
tail with 1.56106 JSQ-3 cells in 50 ml PBS. Five days
later tumors of approximately 12 mm3 were evident at
the injection site. The animals were divided into groups
(eight mice/group) and injected once directly into the
tumor, with either 2.56108, 56108, 16109, or 1.56109
PFU of Av1p53. 1.56109 PFU of the control
Av1LacZ4 was injected into one group and one goup
was left untreated. Thirty-six to forty-eight hours post
injection, the tumor area only was exposed to a 2.5 Gy
dose of ionizing radiaton. Thereafter, the animals were
given 2.5 Gy of radiation every 48 h to a total dose of
20 Gy. For comparison, a group of untreated mice and
mice which had been treated with Av1p53, or
Av1LacZ4, received no radiation.
The ®ndings presented in Figure 6a ± c demonstrate
that the combination of radiation and wtp53 replacement is markedly more e€ective in controlling tumor
growth than either treatment individually. In the
animals treated with radiation alone, the tumors
initially regressed (Figure 6a). However, in a manner
analogous to that observed in the clinical situation,
these tumors began to recur by approximately 5 weeks
post-radiation. Similarly, those treated with radiation
and a high dose (1.56109 PFU) of the control
Av1LacZ4 also regressed, and then began to regrow
at about 7 weeks post-treatment.
In contrast, treatment with the combination of
radiation and the adenoviral construct containing
wtp53 resulted in complete eradication of the tumors.
With the exception of those tumors treated with the
lowest dose of Av1p53 (2.56108 PFU) (Figure 6b),
these tumors did not recur even 5 months after the last
dose of radiation was administered. This observation is
represented in Figure 6b where the plots for Av1p53 at
56108, 16109 and 1.56109 PFU go to zero between
days 42 to 65 and remain at zero for the duration of
the experiment. This is in striking contrast to the plots
for 1.56109 PFU of Av1LacZ4 (Figure 6a) and
2.56108 PFU of Av1p53 (Figure 6b), both of which
begin to regrow at about 65 ± 72 days post-virus
infection (approximately 50 days post-irradiation).
In the unirradiated tumors (Figure 6c), treatment
with the three lower doses of Av1p53 alone had only
minimal inhibitory e€ect on tumor growth. Although
the highest dose of Av1p53 (1.56109 PFU) was able to
induce signi®cant tumor regression for a prolonged
period of time, these tumors also began to recur and
reached approximately 50% of their initial volume by
162 days after Av1p53 treatment (Figure 6c). Four
separate in vivo experiments of this nature have been
performed with consistent results.
Histochemical staining of xenograft tumors after
combination therapy
Figure 6 E€ect of the combination of wtp53 replacement and
radiation therapy on tumor growth in a xenograft mouse model.
Radiation resistant SCCHN cell line JSQ-3 was subcutaneously
injected on the lower back of 4 ± 6 week old female a thymic nude
(athymic NCr-nu) mice. The tumor size was measured before viral
injection with either Av1p53 or Av1LacZ4 (Day 0); before each
radiation dose; and weekly thereafter. Data are calculated as
percent of original tumor volume (Day 0) and plotted as
fractional tumor volume [f(s/so)] +s.e.m. V represents the day
of viral injection; R represents eight 2.5 Gy doses of radiation, a
total of 20 Gy
Subcutaneous xenograft tumors and surrounding
tissues were removed, ®xed, sectioned, and stained
with hematoxilin and eosin (H&E) for evaluation of
their histological appearance, with or without the
combination of Av1p53 and radiation treatment. The
control untreated, unirradiated tumor is shown in
Figure 7a. The xenograft exhibited characteristics of
squamous cell carcinoma, with keratin pearls and
desmosomes. Two injections (48 h apart) with 16108
PFU Av1p53 without radiation was able to induce
some morphological changes as indicated by the
condensation of the chromatin leaving empty space
(vacuoles) within the con®nes of the nuclear membrane. Some tumor cells also appeared more differ-
p53 sensitizes head and neck cancer to radiotherapy
KF Pirollo et al
entiated after Av1p53 treatment (Figure 7b). After
exposure to 20 Gy, some post-irradiation e€ects,
necrosis and pyknosis, are identi®ed in both the nonvirally treated tumor (Figure 7c) and in the Av1LacZ4
(16108 PFU) injected tumor (Figure 7d). However, a
number of live tumor cells are still visible. The presence
of these live tumor cells would indicate that these
tumors had a high probability of recurrence.
A synergistic e€ect is clearly seen, however, when
gene replacement therapy and radiation are combined.
Only very few pyknotic cells (mostly cell debris) at the
site previously bearing tumor are present after
treatment with 56107 PFU of Av1p53 in combination
with 20 Gy of radiation (Figure 7e). Even more
striking is the appearance of the tumor tissue
remaining after two treatments with 16108 PFU of
Av1p53 and 20 Gy (Figure 7f). In the very small
residual tumor tissue, there are no live tumor cells and
only necrotic tissue is present.
The results of these studies suggest a synergistic
e€ect when wtp53 replacement is used in concert with
standard radiotherapy (compare panels b, c and f).
Discussion
SSCHN can be dicult to treat with a high rate of
local recurrence. This is due primarily to the presence
of microscopic residual tumor after surgery at the site
of the primary tumor and the resistance of a signi®cant
percentage of these tumors to radiation, the most
common form of adjuvant therapy for this disease.
Figure 7 Histochemical analysis of subcutaneous xenograft tumors before and after treatment with radiation and/or Av1p53.
Subcutaneous tumors were excised from the animals before or after treatment with radiation and/or Av1p53 or control Av1LacZ4.
The tumors were subsequently ®xed in Histochoice (AMRESCO), sectioned and stained with H&E. (a) control tumor, untreated,
unirradiated; (b) tumor after two injections of 16108 PFU of Av1p53, without irradiation; (c) tumor after 20 Gy of ionizing
radiation, without viral treatment. (d) tumor after two injections of 16108 PFU of control Av1LacZ4 plus 20 Gy of ionizing
radiation; (e) tumor after a combination of 20 Gy ionizing radiation plus two injections of 56107 PFU of Av1p53; (f) tumor after a
combination of 20 Gy of ionizing radiation plus two injections of 16108 PFU Av1p53 (Magni®cation 6400)
1741
p53 sensitizes head and neck cancer to radiotherapy
KF Pirollo et al
1742
Since it appears that most anti-cancer agents work by
inducing apoptosis (Kerr et al., 1994), inhibition of, or
changes in the pathway can conceivably lead to failure
of therapeutic regimens. The role of the p53 tumor
suppressor gene in inducing apoptosis in response to
stress is well established. Moreover, a direct link has
been suggested between mutations in p53 and
resistance to cytotoxic cancer treatments, (both
chemo- and radiotherpay) (Lowe, 1995). It has also
been suggested that the loss of wtp53, which is essential
for the e€ective activation of apoptosis following
irradiation or treatment with chemotherapeutic
agents, may contribute to the cross-resistance to anticancer agents observed in some tumor cells (Johnson et
al., 1991). For example, various laboratories have
A
established a correlation between the presence of
mtp53 and chemoresistance in mouse ®brosarcomas
(Lowe et al., 1994) and in primary tumor cultures from
human gastric, esophageal and breast carcinomas as
welll as b-cell chronic lymphoblastic leukemia (Koechli
et al., 1994; Silber et al., 1994; Nabeya et al., 1995).
Wu and El-Deiry (1996), Lutzker and Levine (1996)
and Aas et al. (1996) have recently associated p53
status with the e€ectiveness of chemotherapeutic agents
in ovarian teratocarcinomas, testicular teratocarcinomas and breast cancer (respectively). In addition,
chemosensitivity, along with apoptosis, was restored
by expression of wtp53 in non-small cell lung
carcinoma mouse xenografts carrying mtp53 (Fujiwara et al., 1994). However, the presence of some
p53 sensitizes head and neck cancer to radiotherapy
KF Pirollo et al
con¯icting reports, (Wahl et al., 1996; Hawkins et al.,
1996) indicate that the relationship between p53 and
chemoresistance may have a tissue or cell type-speci®c
component.
The lack of G1 block, due to the presence of mtp53,
has been shown, particularly in cells of hematopoietic
lineage, to correlate with an increase in radiation
resistance (Lee and Bernstein, 1993; O'Conner et al.,
1993). Moreover, McIlwrath et al. (1994) measured the
G1 arrest in 12 human tumor cell lines displaying a
wide range of radioresponsiveness. They found a
signi®cant inverse correlation between the level of
ionizing radiation-induced G1 arrest and radiation
resistance, i.e. cell lines lacking G1 arrest were
radioresistant. Since abnormal p53 has been detected
in SCCHN at a high frequency, there is likely to be a
correlation between this and the high percentage of
radiation resistance observed in these tumors. However, the relationship between p53 status and the
radiation resistant phenotype is not entirely clear and
may not completely or directly correlate for ®broblast
cells or all tumor cell lines (Jung et al., 1992; Kasid et
al., 1993).
The introduction of wtp53 has been reported to
suppress, both in vitro and in mouse xenograft models,
the growth of various types of malignancies including
SCCHN. In in vivo studies in nude mice with
established subcutaneous SCCHN nodules, Liu et al.,
demonstrated that peritumoral injection of a wtp53
carrying adenovirus was able to signi®cantly reduce
tumor growth (Liu et al., 1994). Moreover, Clayman et
al. was able to prevent the establishment of SCCHN
tumors in nude mice for 12 weeks by introducing
wtp53-adenovirus no more than 48 h after subcutaneous implantation of 2.56106 cells (Clayman et al.,
1995). The induction of apoptosis by the transduced
wtp53 was shown by these investigators to play a role
inhibiting tumor formation in this xenograft model
(Liu et al., 1995). Other investigators have observed
similar growth inhibition by introduction of wtp53 in,
Figure 8 Construction of adenoviral vector Av1p53 as described
in Materials and methods. (a) is a diagram of the construction of
plasmid pAVS6.p53; (b) is a diagram of the homologous
recombination of pAVS6.p53 with Ad dl327 to yield Av1p53
for example, prostate (Yang et al., 1995; Srivastava et
al., 1995), colon (Shaw et al., 1992), and lung tumor
cells (Fujiwara et al., 1993, 1994). Many of these
studies employed adenoviral vectors as the vehicles for
introduction of the wtp53 (Liu et al., 1994, 1995;
Fujiwara et al., 1994; Clayman et al., 1995; Yang et al.,
1995; Srivastava et al., 1995; Shaw et al., 1992). The
use and advantages of adenovirus for gene therapy
have been recently reviewed (Bramson et al., 1995;
Blau et al., 1995; Stewart, 1995). Due to its natural
anity for the epithelium of the aerodigestive tract,
adenovirus is the best vehicle currently available for
gene therapy studies of head and neck cancer.
In the studies described above, we employed an
adenoviral vector carrying wtp53 (Av1p53) under
control of the RSV promoter, to determine if
replacement of wtp53 could sensitize radiation
resistant SCCHN tumors to radiotherapy. Consistent
with previous reports (Liu et al., 1994, 1995; Fujiwara
et al., 1994; Clayman et al., 1995; Yang et al., 1995;
Srivastava et al., 1995; Shaw et al., 1992), we also
found that introduction of wtp53 was capable of
inhibiting cell growth in tumor cell lines known to have
an aberrant form of the gene. In contrast, Av1p53 had
virtually no e€ect on a normal human ®broblast cell
line which carries only wtp53. More signi®cantly,
replacement of wtp53 was able to revert the radiation
resistant phenotype of SCCHN cell line JSQ-3 in a
dose and time dependent manner. These results
support the hypothesis that the e€ectiveness of
therapeutic agents is determined, at least in part, by
a p53-dependent mechanism (Lowe, 1995). Since it is
known that wtp53 can promoter apoptosis (Oren,
1994), it is possible that the increase in radiation
sensitivity observed here after introduction of wtp53
may be related to a restoration of the apoptotic
pathway in these cells. Hallahan et al. also observed
a decrease in tumor growth, an induction of apoptosis
and a sensitization to radiation treatment after
infection of xenografts of SCCHN cell line SQ-20B
with an adenovirus containing the gene for TNF-a
under control of the radiation inducible promoter Egr
(Hallahan et al., 1995). Therefore, induction of
apoptosis by cytotoxic agents, such as TNF-a, or the
restoration of the apoptotic pathway by replacement of
wtp53, may play a signi®cant role in controlling tumor
growth.
From a clinical standpoint, the sensitization of the
tumor to radiotherapy after treatment with wtp53 is
signi®cant. As demonstrated in our in vivo studies, the
combination of gene therapy with wtp53 and conventional radiotherapy was markedly more e€ective than
either treatment alone. This combined modality was
able to regress completely and prevent recurrence of
established tumors for over 5 months when a radiation
dose of only 20 Gy was administered concomitantly
with as little as 56108 PFU of wtp53-adenovirus. In
the clinical setting, radiation doses of 65 to 75 Gy for
gross tumor and 45 to 50 Gy for microscopic disease
are commonly employed in the treatment of cancer of
the head and neck (Awan et al., 1991). Even so, 50 ±
60% of these patients will ultimately develop local
recurrence (Dimery and Hong, 1993). Given the
adverse side e€ects associated with high doses of
radiation, radiosensitization of these tumors, resulting
in a decrease in the amount of radiation necessary to
1743
p53 sensitizes head and neck cancer to radiotherapy
KF Pirollo et al
1744
be e€ective, would be of immense bene®t to the
patient. Furthermore, restoration of wtp53 function
in tumors which had previously failed radiotherapy
would sensitize these tumors and thus allow further
therapeutic intervention.
Besides their many advantages, there are also
disadvantages to the potential use of adenoviral
vectors for gene therapy. The current generation of
vectors are immunogenic and induce a non-speci®c
in¯ammatory response and antiviral cellular immunity
(Smith et al., 1993; Setoguchi et al., 1994; Yang et al.,
1994, 1995). Therefore, repeated administration may
not be feasible. In this respect our ®ndings that, when
used in concert with conventional radiotherapy, one
administration of Av1p53 was sucient to have a
therapeutic e€ect is of great consequence.
It is unlikely that 100% of the cells in the xenograft
tumors were transduced by the virus. How, therefore,
do we account for the e€ectiveness of our in vivo
studies? It has been postulated that adenoviral vectors
containing wtp53 control cell growth partially through
a `bystander e€ect' which may be related to the
induction of apoptosis by the wtp53. It is conceivable, therefore, that this `bystander e€ect' may be a
contributory factor to the e€ectiveness of the combination therapy. However, very little is known at this time
concerning the mechanism and pathway involved in
this process for p53. It has been speculated that some
as yet unknown apoptotic signal may be contained
within the vesicles which result from apoptosis and
which are ultimately phagocytized by neighboring cells
(Seachrist, 1994). Alternatively, this apoptotic signal
may be transferred through gap junctions, as is
believed to be the case for phosphorylated gancyclovir
with the HSV-TK gene (Seachrist, 1994; Pitts, 1994).
Work is currently in progress to examine this question.
The presence of p53 mutations have also been linked to
resistance to chemotherapeutic agents (Lowe, 1995;
Koechli et al., 1994; Silber et al., 1994; Nabeya et al.,
1995). Furthermore, Fujiwara demonstrated that
expression of wtp53 in p53-de®cient lung cancer cells
was able to restore chemosensitivity (Fujiwara et al.,
1994). Therefore, the use of wtp53 gene therapy, not as
a single modality, but in combination with conventional radio- and/or chemotherapy holds promise as a
new, more e€ective means of treating not only a wide
variety of primary malignancies but also as a means to
sensitize residual or recurring tumor to further
therapeutic intervention.
Materials and methods
Cell culture
Human head and neck cell lines JSQ-3, derived from a
tumor of the nasal vestibule (Weichselbaum et al., 1988),
SQ-20-B, derived from a squamous cell carcinoma of the
larynx (Weichselbaum et al., 1986) and SCC61 derived
from an oral (tongue) tumor (Weichselbaum et al., 1986)
were generous gifts from Dr Ralph Weichselbaum,
University of Chicago. These cell lines along with human
non-cancerous skin ®broblast cell line H500, were
maintained in Minimum Essential Medium with Earle's
salts (EMEM), supplemented with 10% heat-activated fetal
bovine serum; 50 mg/ml each of penicillin, streptomycin
and neomycin; 2 mM L-glutamine; 0.1 mM non-essential
amino acids and 1 mM sodium pyruvate. Human breast
(SK-BR-3) and ovarian (SK-OV-3) carcinoma cell lines
(obtained from ATCC, Rockville, MD) were maintained in
McCoy's 5A medium supplemented with 10% fetal bovine
serum, 50 mg/ml each of penicillin, streptomycin and
neomycin and 2 mM L-glutamine.
For in vitro viral infection, the cells, at approximately 30 ±
50% con¯uency, were incubated with either Av1p53 or
Av1LacZ4, diluted to the appropriate viral concentration
with 150 ± 500 mL Phosphate bu€ered saline (PBS), at 378C
with gentle rocking. At the end of 2 h, fresh medium was
added without removal of the virus.
In vitro radiobiology
Cellular response to radiation was evaluated by the cell
survival assay. Exponentially growing monolayer cultures of
each cell line were virally infected as described above. The
cells were harvested 24, 36, or 48 h later, suspended in fresh
medium and irradiated at room temperature with graded
doses of 137Cs g-rays at a dose of approximately 36 Gy/min in
a JL Shepard and Associates Mark I irradiator. Afterward,
the cells were diluted and plated at a concentration of 500 to
5000 cells per well in a 6-well tissue culture plate. Two to
three days after plating, the plates were supplemented with
0.5 ml of fetal bovine serum plus 5 mg/ml hydrocortisone.
Approximately 7 ± 14 days later, the plates were stained with
1% crystal violet and colonies (comprising 50 more cells of
normal appearance) were scored. Survival curves were
plotted as the log of the survival fraction versus the
radiation dose using Sigma-Plot Graphics program. D10
(the dose required to reduce survival to 10%) values were
calculated from the initial survival data.
Protein analysis
Cells for p53 protein analysis were trypsinized, pelleted,
rinsed with PBS and lysed in RIPA bu€er (1% NP40, 0.5%
sodium deoxycholate, 0.1% SDS, 30 mg/ml aprotinin and
1 mM sodium orthovanidate in PBS) (Santa Cruz Biotechnology, Inc). After shearing with a 26 gauge needle, 100 mg/
ml Phenylmethylsulfonyl ¯uoride (PMSF) was added, the
lysate incubated on ice for 30 ± 60 min and centrifuged at
13 000 g for 20 min at 48C to pellet insoluble material.
Protein concentration was determined using the micro-BCA
Protein Assay Kit (Pierce Biochemicals).
Forty mg of protein lysate was mixed with an equal
volume of 26 protein sample bu€er (0.05 M Tris (pH 6.8),
3% SDS, 20% Glycerol, 6% 2-Mercaptoethanol and 0.001%
Bromophenol blue) boiled for 5 min, loaded on a 10% (4%
stacking gel) SDS/Polyacrylamide gel and electrophoresed at
200 V for 8 h. The protein was transferred to nitrocellulose
membrane as previously described (Janat et al., 1994).
Preparation of membrane and incubation with the primary
and secondary antibodies was performed essentially as
described in a protocol supplied by Santa Cruz Biotechnology, Inc, with the exception that incubation with the primary
antibody (Anti-p53 antibody Ab-2, Oncogene Research
Products) was extended to 2 h, with wash times of 15 min
per wash. The washings after addition of the secondary
antibody (Anti-mouse IgG-HRP, Santa Cruz Biotechnology,
Inc) were also lengthened to 15 min per wash.
Visualization of the protein was accomplished using the
ECL Western Blotting Kit (Amersham) according to the
manufacturer's protocol.
Adenoviral vectors
The adenoviral vectors employed in the study, Av1p53 and
Av1LacZ4, were constructed at Genetic Therapy, Inc/
Novartis, Gaithersburg, MD and are replication de®cient,
Ela/Elb, E3 deletion mutants containing a wtp53 gene, and
a LacZ (b-galactosidase) gene, respectively.
p53 sensitizes head and neck cancer to radiotherapy
KF Pirollo et al
Construction of Av1LacZ4 has previously been described
(Smith et al., 1993; Trapnell, 1993; Mittereder et al., 1994).
Av1p53 was constructed as follows: Plasmid pBSK-SN3,
obtained from PharmaGenics (Allendale, NJ), contains a
1.8 kb XbaI fragment that includes the wild-type p53 open
reading frame as well as 5' and 3' untranslated regions cloned
into the XbaI site of Bluescript SK (Strategene, La Jolla,
CA). pBSK-SN3 was digested with SmaI and partially
digested with NcoI to generate a 1322 bp fragment containing the p53 open reading frame. The fragment was gel
puri®ed and ligated into plasmid pBg in place of the bgalactosidase gene between the NcoI the XhoII sites to yield
plasmid pp53. Av1p53 was constructed by ligation of a
1.4 Kb NotI ± SmaI fragment of pp53, which contains the
open reading frame of wtp53 and approximately 140 bp of 3'
untranslated region, to the EcoRV fragment of PAVS6
(Mittereder et al., 1994) to create pAVS6p53 (Figure 8a).
pAVS6p53 contains the adenoviral inverted terminal repeat
(ITR) and packaging signal, the Rous Sarcoma Virus (RSV)
promoter, the adenoviral tripartite leader, the human p53
open reading frame, the SV40 polyadenylation signal, the
adenoviral homologous recombination region and the
ampicillin resistance gene. This plasmid was subsequently
linearized with NotI, followed by homologous recombination
with the large ClaI fragment of wild-type adenovirus Add1327 according to the same procedure as described for the
generation of Av1LacZ4 from pAVS6n-LacZ (Smith et al.,
1993; Trapnell, 1993; Mittereder et al., 1994) (Figure 8b).
This ClaI fragment contains all the adenoviral genes except
the ITR, the Ela enhancer/packaging region and the Ela
gene. Viral stocks were propagated in 293 cells and titered by
plaque assay (Mittereder et al., 1994). Speci®cally, the
pAVS6p53 expression plasmid was linearized with NotI and
gel puri®ed. The Ad-d1327 ClaI fragment (2.5 mg) and 5 mg
of linearized pAVS6p53 plasmid were introduced into 293
cells by calcium phosphate transfection (BRL kit). Viral
DNA from ten plaques was isolated and digested with BglII
and ClaI to generate a 1.4 kb p53 fragment. Digested DNA
was fractionated on a 1% agarose gel and transferred to
nitrocellulose for Southern hybridization with a radiolabeled
p53 speci®c probe. The crude viral lysate (CVL) from a p53
positive plaque was used to make a large scale preparation of
Av1p53 which was puri®ed by CsCl banding using standard
methods. This preparation of Av1p53 was then plaque
puri®ed twice by limiting dilution in 293 cells. Ten plaques
were isolated and ampli®ed on 293 cells to generate CVL.
The viral DNA prepared from each CVL was used in a PCR
reaction to amplify a 1.6 kb fragment containing the p53
gene. One of these p53 positive plaques was then used to
make large scale preparations of Av1p53.
Growth of human xenograft tumors in vivo
1.0 ± 1.56106 human SCCHN tumor cells (JSQ-3) in 50 ml
PBS were injected subcutaneously on the lower back above
the tail of 4 ± 6 week old female athymic nude (athymic
NCr-nu)mice. When xenografts of the predetermined size
were evident at the injection site, the animals were divided
into groups. Each group (4 ± 8 mice/group) was injected
one time, directly into the tumor, with 2.56108 to 1.56109
PFU of the Av1p53 or Av1LacZ4 (in 50 ml PBS). Control
tumors were not injected. 36 ± 48 h post injection, the
animals were immobilized in a lead chamber which
shielded the body except for the tumor area which was
exposed to a 2.0 ± 2.5 Gy dose of ionizing radiation using a
Philips RT 250, 250 Kv X-Ray machine using a 0.5 mmCu
®lter with a dose rate of 84 cGy/min. Thereafter, the
animals were given 2.0 ± 2.5 Gy of radiation every 48 h to a
total dose of 20 ± 25 Gy. For comparison, one group of
untreated tumors, as well as tumors which had been
injected with Av1p53 or Av1LacZ4, received no irradiation.
The size of each tumor was measured prior to viral
infection and to each radiation treatment and weekly
thereafter. Data was calculated as the percent of the original
tumor volume present prior to the viral injection at day 0.
Acknowledgements
We wish to thank Dr R Weichselbaum for his kind gift of
the SCCHN cell lines, Diana Mo€at and Marlene Hammer
(GTI/Sandoz) for help in preparation of the virus, Dr Will
Jacob for coordination, Dr Joe B Harford for his critical
review of the manuscript, and Amy Rutherford for
assistance in its preparation. This work was supported in
part by NCI grant R01 CA45158 (EC) and National
Foundation for Cancer Research grant HU 0001 (EC),
and was performed in the Liem Siou Liong Molecular
Biology Laboratory in the Division of OtolaryngologyHead and Neck Surgery at the Stanford University
Medical Center.
References
Aas T, Borresen AL, Geisler S, Smith-Sorensen B, Johnson
H, Varhaugh JE, Akslen LA and Lonning PE. (1996).
Nature Med., 2, 811 ± 814.
Agarwal ML, Agarwal A, Taylor WR and Stark GR. (1995).
PNAS USA, 92, 8493 ± 8497.
American Cancer Society. (1993). American Cancer Society
facts and ®gures. Publ. No. 93 ± 400. Washington, DC:
American Cancer Society.
American Joint Committee on Cancer. (1983). Manual for
staging of cancer, 2nd ed. Philadelphia: Lippincott,
pp. 25 ± 54.
Awan AM, Vokes EE and Weichselbaum RR. (1991).
Hematology/Oncology Clinics of North America, 5, 635 ±
655.
Blau HM and Springer ML. (1995). N.E. J. Med., 333, 1204 ±
1207.
Brachman DG. (1994). Seminars in Oncology, 21, 320 ± 329.
Bramson JL, Graham FL and Gauldie J. (1995). Curr. Opin.
in Biotech., 6, 590 ± 595.
Chang F, Syrjanen S and Syrjanen K. (1995). J. Clin. Oncol.,
13, 1009 ± 1022.
Clarke AR, Purdie CA, Harrison DJ, Morris RG, Bird CC,
Hooper ML and Wyllie AH. (1993). Nature, 362, 849 ±
852.
Clayman GL, El-Naggar AK, Rith JA, Zhang WW,
Goepfert H, Taylor DL and Liu TJ. (1995). Cancer Res.,
55, 1 ± 6.
Dee A, Wu H, Reinke V and Lozano G. (1993). Mol. Cell.
Biol., 13, 3415 ± 3423.
Dimery IW and Hong WK. (1993). JNCI, 85, 95 ± 111.
Donehower LA, Harvey M, Slagle BL, McArthur MJ,
Montgomery CA Jr, Butel JS and Bradley A. (1992).
Nature, 356, 215 ± 221.
El-Deiry WS, Harper JW, O'Conner PM, Velculesco VE,
Cannan CE, Jackman J, Pietenpol JA, Burrell M, Hill DE,
Wang Y, Wiman KG, Mercer WE, Kastan MB, Kohn
KW, Elledge SJ, Kinzler KW and Vogelstein B. (1994).
Cancer Res., 54, 1169 ± 1174.
Elstner E, Linker-Israeli M, Said J, Umiel T and Vos S.
(1995). Cancer Res., 55, 2822 ± 2830.
1745
p53 sensitizes head and neck cancer to radiotherapy
KF Pirollo et al
1746
Field JK, Pavelic ZP, Spandidos DA, Stambrook PJ, Jones
AS and Gluckman JL. (1993). Arch. Otolarying. Head and
Neck Surg., 119, 1118 ± 1122.
Fujiwara T, Grimm EA, Mukhopadhyay T, Cai DW, OweSchaub LB and Roth JA. (1993). Cancer Res., 53, 4129 ±
4133.
Fujiwara T, Grimm EA, Mukhopadhyay T, Zhang WW,
Owen-Schaub LB and Roth JA. (1994). Cancer Res., 54,
2287 ± 2291.
Hallahan DE, Mauceri HJ, Seung LP, Dunphy EJ, Wayne
JD, Hanna NN, Toledana A, Hellman S, Kufe DW and
Weichselbaum RR. (1995). Nature Medicine, 1, 786 ± 791.
Harper JW, Adami GR, Wei N, Kayomarssi K and Elledge
SJ. (1993). Cell, 19, 805 ± 816.
Hawkins DS, Demers GW and Galloway DA. (1996). Cancer
Res., 56, 892 ± 898.
Janat MF, Srivastava S, Devadas K, Chin GA, Pirollo KF
and Chang EH. (1994). Mol. Cell. Di€., 2, 241 ± 253.
Johnson P, Gray D, Mowat M and Benchimol S. (1991).
Mol. Cell. Biol., 11, 1 ± 11.
Jung M, Notario V and Dritschilo A. (1992). Cancer Res., 52,
6390 ± 6393.
Kasid U, Pirollo K, Dritschilo A and Chang E. (1993). Adv.
Cancer Res., 61, 195 ± 233.
Kastan MB, Onyekwere O, Sidransky D, Vogelstein B and
Craig RW. (1991). Cancer Res., 51, 6304 ± 6311.
Kastan MB, Zhan Q, El-Deiry WS, Carrier F, Jacks T,
Walsh WV, Plunkett BS, Vogelstein B and Fornace AJ.
(1992). Cell, 7, 587 ± 597.
Kerr JFR, Winterford CM and Harmon BV. (1994). Cancer,
73, 2013 ± 2026.
Koechli OR, Schaer GN, Seifert B, Hornung R, Haller U,
Eppenberger U and Muller H. (1994). Lancet, 344, 1647 ±
1648.
Kuerbitz SJ, Plunkett BS, Walsh WV and Kastan MB.
(1992). PNAS USA, 89, 7491 ± 7495.
Lee JM and Bernstein A. (1993). PNAS USA, 90, 5742 ±
5746.
Liu TJ, Zhang WW, Taylor DL, Roth JA, Goepfert H and
Clayman GL. (1994). Cancer Res., 54, 3662 ± 3667.
Liu TJ, El-Naggar AK, McDonnell TJ, Steck KD, Wang M,
Taylor DL and Clayman GL. (1995). Cancer Res., 55,
3117 ± 3122.
Lowe SW, Ruley HE, Jacks T and Housman DE. (1993).
Cell, 74, 957 ± 967.
Lowe SW, Schmitt EM, Smith SW, Osborne BA and Jacks T.
(1993). Nature, 362, 847 ± 848.
Lowe SW, Bodis S, Bardeesy N, McClatchey A, Remington
L, Ruley HE, Fisher DE, Jacks T, Pelletier J and Housman
DE. (1994). Cold Spring Harbor Symposium on Quantitative Biology, 59, 419 ± 426.
Lowe SW, Bodis S, McClatchey A, Remington L, Ruley HE,
Fisher DE, Housman DE and Jacks T. (1994). Science,
266, 807 ± 810.
Lowe SW. (1995). Current Opin. Onc., 7, 547 ± 553.
Lutzker SG and Levine AJ. (1996). Nature Med., 2, 804 ±
810.
Mittereder N, Yei S, Bachurski C, Cuppoletti J, Whitsett J,
Tolstoshev P and Trapnell B. (1994). Human Gene Ther.,
5, 717 ± 729.
McIlwrath AJ, Vasey PA, Ross GM and Brown R. (1994).
Cancer Res., 54, 3718 ± 3722.
Nabeya Y, Loganzo F, Maslak P, Lai L, Deoliveira AR,
Schwarth GK, Blundell ML, Altorki NK, Kelsen DP and
Albino AP. (1995). Int. J. Cancer, 64, 37 ± 46.
Newcomb EW. (1995). Luek. Lymphoma, 17, 211 ± 221.
O'Conner PM, Jackman J, Jondel D, Bhatia K, Magrath I
and Kohn KW. (1993). Cancer Res., 53, 4776 ± 4780.
Oren M. (1994). Semin. Cancer Biol., 5, 221 ± 227.
Powell SN, DeFrank JS, Connell P, Eogan M, Pre€er F,
Dombkowski D, Tang W and Friend S. (1995). Cancer
Res., 55, 1643 ± 1648.
Pitts JD. (1994). Mol. Carcinogen., 11, 127 ± 130.
Russell KJ, Wiens LW, Demers GW, Galloway DA, Plon SE
and Groudine M. (1995). Cancer Res., 55, 1639 ± 1642.
Seachrist L. (1994). JNCI, 86, 82.
Setoguchi Y, Ja€e HA, Chu CS and Crystal RG. (1994). Am.
J. Respir. Cell. Mol. Biol., 10, 369 ± 377.
Shaw P, Bovey R, Tardy S, Sahil R, Sordat B and Costa J.
(1992). PNAS USA, 89, 4495 ± 4499.
Shin DM, Lee JS, Lippmann SM, Lee JJ, Tu ZN, Choi G,
Heyne K, Shin HJC, Ro JY, Goepfert H, Ki Hong W and
Hittelman WN. (1996). JNCI, 88, 519 ± 529.
Silber R, Degar B, Costin D, Newcome EW, Mani M,
Rosenberg CR, Morse L, Drygas JC, Canellakis ZN and
Potmesil M. (1994). Blood, 84, 3440 ± 3446.
Smith TA, Meha€ey MG, Kayda DB, Saunders JM, Yei S,
Trapnell BC, McClelland A and Kaleko M. (1993). Nature
Gen., 5, 397 ± 402.
Srivastava S, Tong YA, Devadas K, Zou ZQ, Sykes VW,
Chen Y, Blattner WA, Pirollo K and Chang EH. (1992).
Oncogene, 7, 987 ± 991.
Srivastava S, Katayose D, Tong YA, Craid CR, McLeod
DG, Moul JW, Cowan KH and Seth P. (1995). Urology,
46, 843 ± 848.
Stewart FM. (1995). J. Cell. Biochem., 58, 416 ± 423.
Stewart N, Hicks GG, Paraskewvas F and Mowat M. (1995).
Oncogene, 10, 109 ± 115.
Trapnell BC. (1995). Adv. Drug Delivery Rev., 12, 185 ± 199.
Wahl AF, Donaldson KL, Fairchild C, Lee FYF, Foster SA,
Demers GW and Galloway DA. (1996). Nature Med., 2,
72 ± 79.
Weichselbaum RR, Dahlberg W, Beckett M, Karison T,
Miller D, Clark J and Erwin TJ. (1986). PNAS USA, 83,
2684 ± 2688.
Weichselbaum RR, Beckett MA, Schwartz JL and Dritschilo
A. (1988). Int. J. Radiation Oncology Biol. Phys., 15, 575 ±
579.
Wu SG and El-Deiry WS. (1996). Nature Med., 2, 255 ± 256.
Xiong Y, Hannon GJ, Zhang H, Casso D, Kobayashi R and
Beach D. (1993). Nature, 366, 701 ± 704.
Yang C, Cirielli C, Capogrossi MC and Passaniti A. (1995).
Cancer Res., 55, 4210 ± 4213.
Yang Y, Nunes FA, Berencsi K, Furth EE, Gonczol E and
Wilson JM. (1994). PNAS USA, 91, 4407 ± 4411.
Yang Y, Li Q, Ertl HC and Wilson JM. (1995). J. Virol., 69,
2004 ± 2015.
Yonish-Rouach E, Resnitzky D, Lotem J, Sacks L, Kimchi A
and Oren M. (1991). Nature, 352, 345 ± 347.