| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
FJ
EXPRESS SUMMARY ARTICLE The Full-length version of this article is also available, published online April 18, 2001 as doi:10.1096/fj.00-0728fje. |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* University of Manchester Academic Unit of Obstetrics and Gynaecology, St. Mary’s Hospital, Manchester M13 OJH, U.K.; and
Department of Laser Oncology, Paterson Institute of Cancer Research, Christie Hospital, Manchester M20 4BX, U.K.
3Correspondence: University of Manchester Academic Unit of Obstetrics and Gynaecology, St. Mary’s Hospital, Whitworth Park, Manchester M13 OJH, U.K. E-mail: mdsisinh{at}fs1.scg.man.ac.uk
SPECIFIC AIMS
We have devised an in vivo experimental model to investigate the role of the human papilloma virus (HPV) type 16 E6 and E7 oncoproteins in the development of radiation resistance in advanced cervical carcinoma. HPV negative human C33A cervical carcinoma cells were transfected with E6 and E7 cDNAs and the transfectants subcutaneously (s.c.) transplanted into scid mice. Tumor growth and radiation resistance under normal and hypoxic conditions were assessed in male and female recipients. Patterns of gene expression were investigated in radiation-resistant vs. radiation-sensitive tumors.
PRINCIPAL FINDINGS
1. Constitutive expression of the E6 protein promotes rapid tumor
development
E6-expressing tumors reproducibly developed more quickly than E7
tumors, which in turn developed more quickly than vector-only control
tumors (Fig. 1a
), although no difference in growth rate was observed in
vitro between E6, E7, or control cell lines. Once established, however,
tumor growth rates were similar. There was no significant difference in
tumor growth observed between male and female recipients.
|
2. Tumors expressing high levels of the E6 protein were resistant
to ionizing radiation
Treatment of tumors with ionizing radiation showed that E6 tumors
were vastly more radiation resistant than either E7 or control tumors
(Fig. 1b
) There was no significant difference in the
radiation response observed between E7 and control tumors or between
male and female recipients.
3. The radiation-resistant E6 tumor phenotype was not solely
produced by increased hypoxia
Artificial induction of hypoxia by clamping the tumor blood supply
before and during radiotherapy in E7 and control tumors did not produce
the same degree of radiological resistance found in E6 disease (Fig. 1c
). Although clamped hypoxic E7 tumors showed an increase
in radiological resistance when compared to unclamped, this was
observed only at lower doses of radiation.
4. Differential screening of 1.2 human cancer cDNA arrays with cDNA
probes made from E6 and control tumors identified differentially
expressed genes with potential roles in producing the E6 tumor
phenotype
Several cDNAs were shown to be more highly expressed in E6 tumors,
including PDGF (platelet-derived growth factor), ERCC1 (excision repair
cross complementation 1); NF kappa p65, FGF receptor 1, interferon
gamma, and, to a lesser extent, VEGF-B (vascular endothelial growth
factor B). High-level expression of mRNA for the DNA repair gene ERCC1
was shown to be restricted to high-level E6-expressing,
radiation-resistant tumors, although this was not evident in high
E6-expressing cells in vitro. Western immunoblots demonstrated that the
ERCC1 protein was also up-regulated in high E6-expressing tumors (Fig. 2c
).
|
5. Growth of E6-expressing cells under anoxic conditions in vitro
for 24 h produced an increase in the expression of ERCC1 mRNA
The observed anoxia-dependent in vitro up-regulation of ERCC1 was
not as pronounced as that observed in the radiation-resistant E6
tumors, although high E6-expressing cells showed a more pronounced
induction than was observed in control cells.
CONCLUSIONS AND SIGNIFICANCE
Our data demonstrate that high-level expression of the HPV16 E6
oncoprotein in tumors derived from transplanted human C33A cervical
carcinoma cells confers an aggressive radiation-resistant phenotype.
This was not observed with E7-expressing or control vector-only tumors.
Some human solid tumors are considerably less well oxygenated than
normal tissues; this can lead to resistance to radiotherapy and
anticancer chemotherapy, as well as predisposing to increased tumor
metastases. However, the radiation-resistant E6 tumor phenotype cannot
be accounted for solely by levels of radiobiological hypoxia, since
artificial induction of tumor hypoxia before and during radiation
treatment increased the resistance of E7 tumors at lower radiation
doses, as expected, although not to the same extent as that observed
for E6 tumors irradiated ‘in air’. The detection of equal levels of
Hif1
protein in radiation-resistant and -sensitive E6 tumors
confirmed their hypoxic status and indicated that the E6 tumor
phenotype must result from other factors.
Changes in gene expression associated with E6 tumors included elevated mRNA expression of several factors known to be involved in angiogenesis—for example, PDGF, NF kappa, VEGF-B, FGF receptor 1, and the proinflammatory cytokine interferon gamma. This may explain the earlier onset of macroscopic disease with E6 tumors, presumably by increased stimulation of neovascularization, which in turn facilitates more rapid growth in the initial stages of development. Analysis of the levels of p53 protein in these tumors predictably indicated that lower levels of E6 expression were coincident with higher levels of p53.
The excision repair cross-complementation enzyme 1 (ERCC1) also showed marked mRNA up-regulation in two of three primary in vivo passage E6 tumors. This was not seen in any primary passage control or E7 tumors. As subsequent in vivo passages were carried out with pooled primary tumors, it would be expected that the E6 inoculum would contain clones expressing both low and high levels of ERCC1. Was high-level expression of ERCC1 consistent with increased E6 tumor radiation resistance? Northern blots of in vitro polyclonal/monoclonal cells and second-passage polyclonal tumors probed with ERCC1 indicated that high-level expression was indeed restricted to X-ray-resistant E6 tumors, whereas the two E6 tumors that showed low-level expression of ERCC1 demonstrated a partial response to X-ray treatment. Reprobing this blot with E6 showed a clear correlation between the levels of E6 and ERCC1 expression. Western immunoblots confirmed that the ERCC1 protein was also elevated in high E6-expressing, radiation-resistant tumors. This would suggest that tumors derived from high E6-expressing clones have a radiation-resistant survival advantage attributable, at least in part, to enhanced DNA repair capabilities. In support of this, we also inoculated animals with monoclonal low- and high-level E6-expressing C33A cells, and it is clear that high-level expression of E6 promotes the more rapid tumor development and radiation resistance associated with high ERCC1 expression (data not shown).
A recent in vitro study has also indicated that ERCC1 RNA levels may be a molecular marker of cisplatin resistance in cultured cervical carcinoma cells, although the authors concluded that the level of ERCC1 protein did not always correlate with this observation. Our in vivo data demonstrate that higher ERCC1 protein levels are associated with radiation resistance in E6 tumors.
What is the mechanism responsible for the up-regulation of ERCC1 expression? We have shown that the codon 273 Arg-Cys mutant p53 in C33A cells is degraded by E6, and it has been demonstrated that mutant p53 can still function to stabilize the genome, preventing the development of hyperploidy. Since it is known that expression of the HPV16 E6 protein can produce genetic instability, we analyzed the genomic copy number of ERCC1 in E6 tumors and cell lines. Southern blot analysis demonstrated no amplification of the ERCC1 gene in high E6-expressing tumors, which indicates that the observed up-regulation must be due to either increased transcriptional activation or increased half-life of the ERCC1 transcript.
Some DNA repair proteins play a role in the radiation resistance of tumors. We have also analyzed the mRNA expression levels in E6 and vector-only tumors of several genes contained on the 1.2 k cancer array known to be involved in the DNA repair process. DNA-PK, Ku70, XRCC4, and ATM were expressed in both tumor types and showed no significant change in expression levels. Rad51C truncated protein showed decreased expression in the E6 tumor whereas Rad51 was up-regulated. This was not, however, as pronounced as that seen for ERCC1.
ERCC1 is part of an endonuclease complex that is involved in both nucleotide excision repair (NER) of bulky chemical adducts and recombination repair (RR) of highly genotoxic interstrand DNA cross-links. A role for ERCC1 in X-ray sensitivity is consistent with previous findings that the ERCC1/ERCC4 complex-catalyzed, RR-dependent pathway for the removal of DNA interstrand cross-links is associated with radiation resistance under hypoxic conditions. Furthermore, it has been shown that antisense inhibition of ERCC1 expression suppresses the synergism observed between the cytotoxic drug cisplatin and the radiation-sensitizing drug gemcitabine. Therefore, these data support a role for ERCC1 in the resistance of E6-expressing tumors to X-ray treatment.
We found no evidence of increased ERCC1 expression in high E6-expressing cells when grown under aerobic conditions in vitro. Indeed, a recent study failed to demonstrate a consistent correlation between the expression levels of several DNA repair genes and radiation sensitivity in vitro. The overall conclusion from this work was that the in vitro environment may not accurately reproduce the in vivo requirements necessary for induction of DNA repair enzymes involved in tumor radiation resistance. Since the work of Murray et al. has shown that ERCC1 is clearly involved in the radiobiological resistance of cells under hypoxic conditions, we investigated the effect of hypoxia on the induction of ERCC1 expression in vitro. Growth of high E6-expressing C33A cells under anoxic conditions in vitro produced a modest but more pronounced induction of ERCC1 than in vector-only control cells. Clearly, the tumor environment is an important factor in the induction of genes involved in radiation resistance, and our data are consistent with this hypothesis.
Based on our experimental observations, we propose the following scheme
for the development of radiation resistance in relapsed cervical
carcinoma after radiotherapy (<5% 5 year survival). It has been shown
that low-dose X-irradiation induces higher levels of E6 expression in
cultured cervical carcinoma cells, presumably by induction of a
p53-mediated killing response in cells expressing lower levels of E6.
Thus, cells that survive the first round of radiotherapy may have
increased expression of E6. Intracellular levels of E6 mRNA can vary
according to the HPV integration status. The HPV early protein E2
functions as a negative regulator of E6/E7 transcription; thus,
disruption of the HPV E2 open reading frame, which is a common feature
of HPV integration, could produce increased expression of E6. Indeed,
it has been observed that detection of integrated HPV in primary tumors
of cervical carcinoma patients was strongly associated with treatment
failure. Our results indicate that high-level expression of E6 will
facilitate high-level expression of ERCC1 under hypoxic tumor
conditions. Since ERCC1 has been shown to be involved in the repair of
radiation damage by RR and alkylating agent damage by NER, our data may
also provide an explanation of the phenomenon of radiation-induced
chemoresistance in cervical carcinomas. We are currently
investigating this hypothesis by analyzing tumor biopsies from cervical
carcinoma patients that relapse after radiotherapy (Fig. 2
).
FOOTNOTES
1 To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.00-0728fje ; to cite this
article, use FASEB J. (April 18, 2001) 10.1096/fj.00-0728fje 
2 Both of these authors contributed equally to this work.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
