| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Virology |
Laboratory of Cellular Carcinogenesis and Tumor Promotion, National Cancer Institute [M. N., J. M. G., C. D. W.], Center for Information Technology [P. M., V. P.], NIH, Bethesda, Maryland, 20892, and Department of Dermatology, New England Medical Center and Tufts University School of Medicine, Boston, Massachusetts 02111 [Y. L., E. A.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
HPVs are small DNA tumor viruses that replicate in differentiating epithelial cells of the epidermis and anogenital tract. The E6 and E7 viral genes are expressed at low levels in proliferating basal cells, but transcription is activated as cells enter the terminal differentiation pathway. The E6 and E7 proteins alter cell differentiation (7 , 8) , reactivate host DNA synthesis (9) , and stimulate cell cycle progression (10) . This allows the virus to use host DNA synthetic enzymes to replicate the viral genome (reviewed in Ref. 11 ). Although many HPV types induce benign warts and papillomas, infection with certain high-risk types (HPV-16, HPV-18, HPV-31, and HPV-45) is a major risk factor for the development of cervical cancer (12) . The E6 and E7 genes are retained and expressed in most cervical carcinomas, and continued expression is required to maintain the malignant phenotype (13) .
An early step in HPV-associated carcinogenesis is perturbation of cellular gene expression by the HPV E6 and E7 oncoproteins. E6 binds to several cellular proteins including E6BP (14) and E6AP, a protein-ligase of the ubiquitin pathway of proteolysis (15) . E6/E6AP complexes target the p53 tumor suppressor protein for rapid degradation by the proteasome (15) . p53 integrates responses to genotoxic stress and DNA damage with cell cycle control and apoptosis. Therefore, loss of p53 function leads to increased genetic instability. p53 is a transcriptional activator, and numerous p53-responsive genes have been identified (16) . The HPV-16 E7 protein binds pRb and members of the pRb family (17) . Interaction occurs primarily with the hypophosphorylated form of pRb, causing release of active E2F transcription factors, which in turn stimulates expression of multiple genes involved in cell cycle progression. The E7 protein also binds and alters the function of AP-1 transcription factors, which regulate cellular gene expression (18 , 19) . E6 and E7 exert overlapping effects on cell cycle control, and in combination, they efficiently immortalize human keratinocytes (20, 21, 22, 23, 24) .
The p53 and pRb proteins are important transcriptional regulators. Thus, inactivation of their function by E6 and E7 is likely to alter keratinocyte gene expression significantly. We used cDNA array technology to identify global changes in gene expression in differentiating cultures of primary human cervical keratinocytes infected with retroviruses encoding HPV-16 E6 and E7. cDNA microarrays allow high-throughput, parallel expression analysis of thousands of cellular genes in a single experiment (25 , 26) . Our results showed that E6 and E7 influenced expression of a large number of cellular genes, including TGF-ß2. TGF-ß2 was induced during keratinocyte differentiation, and both E6 and E7 blocked this induction and altered expression of multiple TGF-ß-regulated genes. We describe inhibition of expression and release of biologically active TGF-ß2 from cervical keratinocytes by E6 and E7 and identify the biological consequences of TGF-ß2 inhibition.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cell Proliferation Assays.
Primary keratinocytes were added to 100-mm culture dishes at a density
of 0.5–1.0 x 105 cells/dish.
After 24 h, cells were treated with various concentrations of
recombinant human TGF-ß2 or TGF-ß1 (R&D Systems, Minneapolis, MN).
Fresh medium with recombinant cytokine was added every other day, and
cells were counted after 8 days using a Coulter counter. All
proliferation experiments were repeated three times in duplicate
dishes.
Cell Immortalization Assay.
Keratinocytes expressing HPV-16 E6/E7 were seeded on 150-mm plastic
dishes at a density of 0.5–1.0 x 104 cells/dish. Cultures at higher passage
numbers that showed evidence of crisis were seeded at densities of
2.5–5.0 x 104 cells/dish. After
24–48 h, cells were treated with various concentrations of recombinant
human TGF-ß2 (50, 100, 200, and 1000 pg/ml medium) every second day,
and TGF-ß2 treatment was continued until the first rapidly and
continuously growing colonies appeared (usually after 10–14 days). For
quantification of immortalization, colonies were allowed to grow to a
diameter of 
1 cm in the absence of TGF-ß2. Selected colonies were
picked and cultured for three more passages to confirm that they were
immortal.
RNA Isolation and Purification.
RNA was extracted from monolayer cultures that were growing or that
were induced to differentiate as described above. For RNA extraction,
cells were washed with PBS and lysed with Trizol (Life Technologies),
and DNA was removed by digestion with RNase-free DNaseI (Life
Technologies) or DNase RQ1 (Promega, Inc., Palo Alto, CA.). DNase,
proteins, and other contaminants were removed by two extractions with
buffer-saturated phenol/chloroform and precipitation with 1.5 volumes
of isopropanol. The integrity of RNA was analyzed by electrophoresis,
and samples were stored at -80°C.
cDNA Synthesis and cDNA Array Hybridization.
The human and human Cancer cDNA arrays on nylon membranes were
purchased from Clontech, Inc. (Palo Alto, CA; ATLAS cDNA arrays). The
protocol for cDNA synthesis was modified from the manufacturer’s
recommendations. Briefly, 32P-labeled cDNAs were
synthesized from DNase-treated total RNA. We used total RNA because
polyadenylated RNA often resulted in severe background and variability
in signal intensity from experiment to experiment. Five to 10 µg of
RNA in a volume of 5 µl were denatured in the presence of 1 µl of
first-strand primer mix at 70°C for 5 min in a thermocycler (GeneAmp
PCR System 9700; PE Applied Biosystems, Foster City, CA). RNA was
incubated for 5 min at 48°C to allow primer annealing. Reverse
transcription was performed using 5 µl of
[
-32P]dATP (3000 Ci/mmol; Amersham Life
Science, Cleveland, OH), 50 µM dNTP mix lacking dATP, 1
µl 100 mM DTT, and 200 units Superscript MMLV Reverse
Transcriptase (Life Technologies) in a total volume of 20 µl. The
reaction was incubated at 48°C for 60 min, and an additional 200
units of Superscript were added after the first 30 min.
Non-incorporated radioactive nucleotides were removed by spin column
purification. Approximately 15–30 x 106 cpm were incorporated in the final product.
Radioactive cDNA probes were denatured for 20 min at 68°C by adding
one-tenth volume of 1 M NaOH, 10 mM EDTA,
neutralized with 1 M sodium phosphate buffer (pH 7.0), and
7.5 µg of Cot-1 DNA and 5 µg poly(A)-DNA were added. Membranes were
prehybridized at 68°C for 6–8 h. Hybridization was performed
overnight at 68°C in a total volume of 7.5–10 ml of ExpressHyb
hybridization solution and a final probe concentration of
2.5–5 x 106 cpm/ml. Membranes
were washed three times at 68°C with 2x SSC, 1.0% SDS, and twice
with 0.2x SSC, 0.5% SDS for 10 min each, sealed in plastic pouches,
and exposed to phosphor imager screen cassettes. Screens were exposed
for 1, 3, and 10 days and analyzed using the Storm Phosphor Imager
System 860 (Molecular Dynamics, Sunnyvale, CA). Computer analysis and
normalization of array hybridization data were performed using the
P-SCAN software.
RNA from keratinocytes infected with vector retroviruses was compared with RNA from cells infected with one or both viral oncogenes of HPV-16. Cells from three to five different individuals were pooled to reduce biological variation between samples. For both the Normal Human and Human Cancer cDNA arrays, experiments were repeated three times for each oncogene combination (total, 24 array hybridizations). Comparisons between growing and differentiating cells, or between untreated cultures and cultures treated with TGF-ß2, were each performed twice using independent samples (total, 4 hybridizations). Differential gene expression data were considered significant if the average factor of differential mRNA expression exceeded 2.0.
RT-PCR.
Fifty to 100 µg of total RNA were purified using lithium chloride
precipitation. An equal volume of 5 M LiCl, 0.1
M Tris-HCl (pH 7.4) was added to samples, and RNA was
precipitated overnight at 4°C, centrifuged at 15000 rpm, washed with
70% ethanol, dried by evaporation, and resuspended in diethyl
pyrocarbonate-treated water. For cDNA synthesis, 5–10 µg of purified
total RNA were reverse transcribed for 1 h at 42°C in a volume
of 50 µl, containing 500 µM individual dNTPs; 10
µM DTT, 1.25 µM oligo-dT primer
(T16–18), 20–40 units RNasin (Promega), and 75
units Superscript Reverse Transcriptase II (Life Technologies) in 1x
RT buffer. RT reactions were diluted 5–10-fold, and for each PCR,
one-tenth (or 2 µl) of diluted RT-mix was included. Amplifications
were performed in a volume of 20 µl in 96-well plates using a GeneAmp
9700 thermocycler, in the presence of 50 µM dNTPs, 0.5
µM of each oligonucleotide primer, 1.75 mM
magnesium chloride, and 1 unit of AmpliTaq Gold (PE Applied
Biosystems). To avoid saturation or plateau effect of amplification,
PCR was limited to a total of 20–25 cycles. Each reaction was
performed twice using independent reverse transcription reactions to
confirm reproducibility. For direct quantification of PCR products,
1.0–2.0 µCi of radioactive [-32P]dCTP was
incorporated, and PCR products were separated on 6% denaturing
polyacrylamide/50% urea gels. DNA was detected and quantified by
autoradiography or using a STORM Phosphor Imager (Molecular Dynamics,
Inc.). Oligonucleotide primers were designed using the GCG (Genetics
Computer Group, Madison, Wisconsin) Program Package. Table 1
summarizes all primer combinations used for RT-PCR analysis. Northern
blot analyses were performed as described (30)
.
|
Immunoblotting.
Western blot analysis was performed as described (10)
using polyclonal antibodies to HPV-16 E7 (Zymed, San Francisco, CA) and
p53 (DO-1; Santa Cruz Biotechnology, Santa Cruz, CA).
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
5% of all genes on arrays). These genes were clustered into
functional groups, revealing important differences in global gene
expression induced by HPV16 E6 and E7 (Table 2)
. Data in column 2 (E6/E7) represent the average of three independent
experiments and were highly reproducible. Growth factors and cytokines
were a prime target for HPV oncogenes, including insulin like growth
factor-1 receptor and heparin-binding epidermal growth factor. E6 and
E7 also decreased expression of genes involved in tissue remodeling and
wound healing including the MMPs 2, 3, 7, 9, and 10–13, fibronectin,
and PAI-1 and PAI-2. Interestingly, E6 and E7 altered expression of
multiple apoptosis-related genes known to decrease susceptibility to
programmed cell death including survivin (inhibitor of apoptosis
hIAP4). As expected, E6 and E7 also stimulated expression of multiple
genes that regulate cell cycle progression including cyclins A and B,
cyclin-dependent kinases 1, 2, and 3, and DNA synthetic enzymes as
described previously (31)
. Expression of DNA repair
enzymes including RAD 50, RAD 51, and RAD 52 were also induced by HPV.
|
; TGF-ß2). Many genes that
were down-regulated by HPV-16 E6/E7 were also identified as
transcriptionally regulated by TGF-ß2.
The cDNA array experiments showed that effects of E6 and E7 on cellular
gene expression were most pronounced under conditions promoting
keratinocyte differentiation (removal of growth factors and addition of
1.4 mM calcium). E6 and E7 induced similar changes in
proliferating cells, although the magnitude of changes were smaller
(data not shown). This suggested that E6 and E7 might alter
keratinocyte gene expression in part by delaying or blocking
differentiation. Therefore, cDNA arrays were used to compare cellular
gene expression in proliferating keratinocytes versus cells
that were induced to undergo differentiation for 24 h. A large
number of genes (15% on arrays) showed altered mRNA expression
during the process of differentiation (Table 2)
. Differentiation was
confirmed by decreased expression of keratins 5, 14, and 6 (data not
shown), which are down-regulated during terminal differentiation
in vivo. Most genes that were down-regulated by E6/E7 were
up-regulated during terminal differentiation, and one of the most
strongly induced was TGF-ß2. Differentiation-dependent induction was
also observed for many TGF-ß-responsive genes such as MMP 7, 10–12,
Fas/APO-1, and Fas ligand. These results suggest that E6 and E7 may
alter expression of TGF-ß and TGF-ß-inducible genes indirectly
through their ability to block or delay keratinocyte differentiation.
Confirmation of Differential Gene Expression by Northern Analysis
and RT-PCR.
Fig. 1A
shows one sector of a typical cDNA array comparing gene
expression in vector-infected controls (left) and E6/E7
infected keratinocytes (right). This sector contained
TGF-ßs 1, 2, and 3 as well as multiple cytokines and growth factors,
and it showed the most striking patterns of differential gene
expression. Differences in the relative signal intensity of each
TGF-ß isoform in response to E6, E7, or E6/E7 are shown in Fig. 1B
. Expression of TGF-ß2 mRNA was down-regulated 5–7-fold
in keratinocytes expressing E6/E7 and 2–3-fold in cells expressing E6
or E7 individually (Fig. 1B
). Expression of other TGF-ß
isoforms including TGF-ßs 1 and 3 were not significantly changed by
E6, E7, or E6/E7. Down-regulation of TGF-ß2 mRNA was confirmed using
Northern blot hybridization (Fig. 1C
), semiquantitative
RT-PCR (Fig. 1D
), and RNase protection (data not shown). All
three methods detected similar levels of down-regulation of TGF-ß2
RNA. However, no differences in TGF-ß1 or TGF-ß3 expression were
detected. Interestingly, retroviral expression of the low risk HPV-6b
E6 and E7 proteins decreased TGF-ß2 RNA expression with much lower
efficiency, and effects were not always reproducible (Fig. 1D
).
|
). Fig. 2
shows representative RT-PCR analyses for several of these genes
including the MMPs 2, 3, and 10, cyclins A and B, PAI-1 (a
well-established TGF-ß-inducible gene), TRAIL, and GAPDH. These
experiments confirmed that each gene was differentially regulated by
E6/E7 and indicated that regulation was generally more dramatic in
response to E6/E7 than to E6 or E7 alone. The results also confirmed
that many but not all of the E6/E7 regulated genes were TGF-ß
responsive. Importantly, addition of recombinant TGF-ß2 to infected
cultures did not alter expression of E6 or E6/E7 mRNAs, which in
retrovirus-infected cells are regulated by the mouse mammary tumor
virus long terminal repeat (see below). Western blot analyses confirmed
that retrovirus-infected keratinocytes actually expressed the HPV-16 E7
protein (Fig. 2B
). Although the E6 protein was not detected
with several commercially available antibodies, E6 or E6/E7-infected
cells had dramatically reduced levels of p53, a known consequence of E6
expression.
|
). As observed for mRNA expression, the combination of
HPV-16 E6 and E7 was significantly more efficient at reducing TGF-ß2
than either oncogene alone. No differences were detected for TGF-ß1
protein expression (data not shown). Interestingly, localization and
expression patterns for TGF-ß1 and TGF-ß2 proteins were strikingly
different. About 33% of the cell-associated TGF-ß2 protein was
localized in the extracellular matrix and was removed by a brief
trypsin digest prior to cell extraction. The remainder was
intracellular and resistant to trypsin treatment. The ratio of
matrix-associated to intracellular TGF-ß2 was increased in
keratinocytes expressing E7 or E6/E7 (Fig. 3A
). In contrast,
TGF-ß1 protein was almost exclusively intracellular and not
associated with the extracellular matrix; trypsin digest did not reduce
the amount of cellular TGF-ß1 detected by ELISA (Fig. 3B
).
This observation might be related to differences in activation of
TGF-ß1 and TGF-ß2 observed in vitro.
|
). One E6 mutant that retained the ability
to degrade p53 (E6mut F125V) decreased levels of intracellular TGF-ß2
as efficiently as wild-type HPV-16 E6. To further investigate the
importance of p53, we constructed retroviruses encoding a
dominant-negative p53 miniprotein (29)
. Expression of this
dominant-negative mutant decreased intracellular TGF-ß2 as strongly
as wild-type E6.
HPV-16 E6 and E7 Decrease Secretion of Biologically Active
TGF-ß2.
Conditioned medium from keratinocytes infected with E6, E7,
E6/E7, or vector-only viruses (pLXSN) was used for detection of
biologically active TGF-ß1 and TGF-ß2. Medium was subsequently
acid-treated to activate and detect latent TGF-ß1 and TGF-ß2
because the ELISA only detects the mature, active cytokine. Secretion
of biologically active TGF-ß2 was strongly induced in differentiating
keratinocytes relative to growing cells (Fig. 4A
). E6 and E7 down-regulated secretion of biologically active
TGF-ß2 in both rapidly growing and differentiating cultures.
Differentiating keratinocytes secreted an average of 112 pg of total
TGF-ß2/106 cells/24 h. Most of the secreted
protein was biologically active TGF-ß2 (Fig. 4B
). In
differentiating cells, secretion of active TGF-ß2 was significantly
reduced by HPV-16 E6 and E7 oncogenes. The
combination of E6 and E7 was most efficient (7-fold reduction),
suggesting a cooperative interaction. E6 and E7 also decreased
secretion of latent TGF-ß2 (Fig. 4B
). In contrast to
differentiating cultures, growing keratinocytes activated only 15% of
the secreted TGF-ß2 protein, and low levels of biologically active
TGF-ß2 were detected (10–24 pg/106 cells/24 h;
data not shown). In most experiments with growing cells, expression of
E6 and E7 resulted in a complete loss of active TGF-ß2 secretion
detectable by ELISA. In contrast, cervical keratinocytes secreted
TGF-ß1 exclusively as the latent, inactive form and E6 and E7 did not
alter secretion (data not shown).
|
Down-Regulation of Biologically Active TGF-ß2 Secretion Involves
the pRb and p53 Tumor Suppressor Genes.
Down-regulation of TGF-ß2 secretion was dependent on the ability of
the E6 or E7 proteins to bind and inactivate the p53 and pRb proteins,
respectively. A dominant-negative p53 miniprotein (p53 DD) decreased
secretion of active TGF-ß2 as efficiently as wild-type HPV-16 E6
(Fig. 4C
). In contrast, overexpression of wild-type p53
resulted in a 2-fold induction of TGF-ß2 secretion. A similar
induction was observed in senescent keratinocytes (in vitro
passages >4). These data point to a functional role for the
p53 tumor suppressor gene in regulation of TGF-ß2
expression. The low-risk E6 and E7 proteins of HPV-6b, which do not
degrade or inactivate the p53 and pRb proteins, also failed to alter
secretion of biologically active TGF-ß2 (Fig. 4C
). HPV-6b
E6 and E7 also did not alter expression of cellular genes regulated by
TGF-ß (data not shown). A mutant HPV-16 E7 protein unable to bind the
retinoblastoma protein [E7-p24gly (33)
] did not diminish
secretion of biologically active TGF-ß2. Interestingly, infection of
keratinocytes with a retrovirus overexpressing the E2F-1 transcription
factor, which signals downstream of pRb, was not sufficient to decrease
TGF-ß2 secretion.
TGF-ß2 Inhibits Growth of E6/E7-expressing Keratinocytes.
TGF-ßs strongly inhibit growth of keratinocytes. We found that
keratinocytes expressing HPV-16 E6 and E7 retained sensitivity to
growth inhibition by recombinant TGF-ß2 in three independent
experiments (Fig. 5
). In fact, cells expressing E6 and E7 were slightly more sensitive to
TGF-ß2 than primary cultures of normal human keratinocytes. Addition
of recombinant TGF-ß2 at concentrations produced endogenously by
differentiating keratinocytes (10–200 pg/ml) inhibited growth of
E6/E7-expressing cells by 25–75%. Cells expressing HPV-16 E6/E7 were
inhibited 50% by 100 pg/ml TGF-ß2 compared with uninfected
keratinocytes, which were inhibited 50% by 200 pg/ml (Fig. 5
). E6/E7
positive cells grew faster than control cells, but this difference was
lost after physiological doses of TGF-ß2 were added to the culture
medium. An HPV-immortalized cell line examined at low passage retained
sensitivity to growth inhibition by TGF-ß2. Cells with reduced growth
potential were particularly sensitive to low concentrations of
TGF-ß2. Keratinocytes infected with HPV-16 E6/E7 and approaching
crisis because of prolonged culture in vitro (Fig. 5
) were
growth inhibited 50% by 5–10 pg/ml of TGF-ß2. Interestingly, E6/E7
expressing cells in crisis or senescing cultures of normal
keratinocytes up-regulated expression of TGF-ß2 2-fold (Fig. 4C
).
|
). Growth inhibition and severe crisis preceded formation of
colonies arising from cells that survived crisis. Colonies of small,
rapidly growing cells were picked and subcultured to confirm that they
were immortal. Of 20 colonies selected, 18 grew for three more in
vitro passages and were established as immortal cell lines. Fig. 6B
summarizes data from six independent experiments.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
7% of genes on the array were either increased or
decreased in expression by >2-fold. These genes were classified into
several functional groups including growth factors/cytokines,
apoptosis-related genes, genes involved in tissue remodeling, and cell
cycle genes. One of the genes that was strongly down-regulated by E6/E7
was TGF-ß2. Expression of TGF-ß2 is increased during epidermal
differentiation (34)
, and our array experiments confirmed
that TGF-ß2 RNA was up-regulated 7-fold in differentiating cervical
keratinocytes. We also demonstrated that many of the genes that were
perturbed by E6 and E7 in differentiating keratinocytes were
TGF-ß-responsive genes and/or genes regulated during keratinocyte
differentiation. Thus, our results suggest that E6 and E7 indirectly
alter expression of diverse cellular genes by decreasing expression of
TGF-ß2. Our results also suggest that differential expression of
TGF-ß2 may represent a consequence of the ability of E6 and E7 to
interfere with aspects of keratinocyte differentiation. However, we
also observed down-regulation of TGF-ß2 expression in growing cells
that did not differentiate, and E6 and E7 also blocked up-regulation of
TGF-ß2 expression by all-trans retinoic acid. Thus,
down-regulation of TGF-ß2 expression by E6 and E7 is not restricted
to differentiating cells.
HPV-16 E6 Down-Regulates TGF-ß2 Production by a p53-dependent
Pathway.
The HPV-16 E6 oncoprotein binds to several cellular proteins that have
been implicated in cell cycle regulation or gene expression (14
, 15)
. In particular, E6 proteins of high-risk HPV types bind to
the p53 tumor suppressor protein and target it for degradation by a
ubiquitin-mediated pathway (15)
. We used HPV-16 E6 mutants
that differed in their ability to bind and degrade p53 to demonstrate a
direct correlation between inactivation of p53 and decreased TGF-ß2
protein production. In particular, the E6-F2V mutant that interacts
with E6BP but fails to degrade p53 (28)
did not decrease
levels of TGF-ß2. The low-risk HPV-6b E6 protein, which does not
efficiently degrade p53, also failed to decrease TGF-ß2. We confirmed
this association by showing that a dominant-negative p53 mutant
inhibited TGF-ß2 production as effectively as the wild-type HPV-16 E6
protein. Furthermore, retroviral expression of a wild-type
p53 gene in cervical keratinocytes increased TGF-ß2
protein levels. We also showed that TGF-ß2 secretion was induced
2-fold in cells undergoing replicative senescence, a process that
involves activation of p53 and p53-responsive genes (16)
.
Our data indicate that TGF-ß2 is a
p53-responsive gene. However, the TGF-ß2 promoter does not contain
p53 response elements, suggesting that regulation might be indirect. Of
more general interest, our results infer that p53 inactivation by
mutations, which occur frequently in premalignant lesions and
carcinomas, might also block production of biologically active
TGF-ß2. This might confer a selective growth advantage on cells with
mutated p53 and contribute to malignant progression.
Interactions between E6/E7, p53, and TGF-ß in Keratinocyte
Differentiation.
Differentiation in HPV-16-infected keratinocytes and premalignant
lesions is altered and partially uncoupled from proliferation
(10
, 35) . In particular, HPV-16 E6 has been shown to
impair serum- and calcium-induced keratinocyte differentiation
(7
, 9)
. This function correlates with the ability of E6
mutants to degrade p53, although additional functions of E6 might also
be involved (7
, 9)
. The biochemical activities of p53 and
p53-mediated genes like p21-waf1 are increased during
keratinocyte differentiation (9
, 37)
, and overexpression
of wild-type p53 in keratinocytes causes premature differentiation and
cell flattening (38)
. In turn, TGF-ß isoforms control
stability and phosphorylation patterns of p53 protein and activate
p53-dependent expression of p21/waf-1 and GADD45 (39)
.
p53-null keratinocytes show impaired responsiveness to negative growth
regulators, including TGF-ß (40)
.
The expression of HPV-16 oncoproteins and TGF-ß2 protein coincide in the differentiating, suprabasal layers of HPV-infected cervical epithelia (9 , 34) . Expression of mitotic regulatory genes (e.g., cyclins A and B and cyclin-dependent kinases) are elevated in HPV-infected keratinocytes (31) but repressed in growth-inhibited and/or differentiating keratinocytes. These genes are also repressed in response to p53-mediated growth arrest or by TGF-ß. Therefore, HPV-mediated inhibition of TGF-ß2 expression, which supports epithelial differentiation (41) , may contribute to growth stimulation and defective differentiation observed in HPV-infected keratinocytes. TGF-ß has also been reported to induce replicative senescence in normal keratinocytes (42) and apoptosis in HPV-immortalized keratinocytes (43) . Inhibition of p53- and TGFß- responsive proapoptotic genes (Fas/APO-1, FasL, TNF-receptor, TRAIL, and others) or induction of protective genes (survivin or inhibitor of apoptosis 4) may help to explain the partial inhibition of differentiation-related cell death by HPV-16 E6 (8) .
HPV-16 E7 Protein Down-Regulates TGF-ß2 Production by a
pRb-dependent Pathway.
The HPV-16 E7 protein binds to and inactivates pRb and related proteins
p107 and p130 (17)
. We have shown that two different E7
proteins that fail to bind pRb were not effective in decreasing
secretion of TGF-ß2. Specifically, the low-risk HPV-6b E7 protein and
an HPV-16 E7 mutant [E7-p24gly (33)
], which do not bind
pRb, also do not perturb TGF-ß2 production. However, we found that
retroviral expression of the E2F-1 transcription factor, which acts
downstream of pRb, was not sufficient to decrease TGF-ß2 expression,
suggesting that additional factors are involved. E2F-1 overexpression
had strong effects on the growth of keratinocytes, which did not
involve changes in TGF-ß2 expression. These data indicate that a
functional pRb protein is necessary for regulation of TGF-ß2 protein
expression. Interestingly, pRb has been reported to increase expression
of the TGF-ß2 gene by binding to the
transcription factor ATF-2, which in turn interacts with the promoter
of the TGF-ß2 gene (44)
.
Furthermore, pRb binds to c-jun and alters activity of
members of the AP-1 family of transcription factors, which serve an
important role in epithelial differentiation (18
, 19) .
E6 and E7 Regulate TGF-ß2 Activity at the Transcriptional Level.
Regulation of TGF-ß activity is complex and may be mediated by
changes in transcription, secretion, or activation of the latent
peptide. Our results suggest that E6 and E7 decreased levels of
biologically active TGF-ß2 mainly by inhibiting expression of
TGF-ß2 RNA (5–7-fold). Levels of total cell-associated TGF-ß2,
secretion of latent peptide, and levels of active TGF-ß2 generally
decreased by a similar factor. We found significant differences between
localization of the different TGF-ß isoforms. TGF-ß1 was almost
exclusively intracellular, whereas 30–50% of TGF-ß2 was found in
the extracellular matrix. TGF-ßs are activated in association with
matrix proteins; therefore, the absence of TGF-ß1 from the
extracellular matrix might explain in part why most TGF-ß1 remained
in the latent state.
Inhibition of Keratinocyte Growth and Immortalization by TGF-ß2.
Previous studies have reported conflicting results regarding the
sensitivity of E6/E7-expressing keratinocytes to TGF-ß growth
inhibition (36
, 45)
. Our results show that cervical
keratinocytes expressing E6 and E7 were growth inhibited by low
concentrations of recombinant TGF-ß2 (10–200 ng/ml), similar to
levels secreted by differentiating keratinocytes in the absence of E6
and E7. Furthermore, the induction of TGF-ß-responsive genes by
recombinant TGF-ß2 confirmed that the downstream signaling and
response to TGF-ß were intact in cells expressing E6 and E7. These
results suggest that inhibition of TGF-ß2 expression by HPV E6 and E7
is biologically significant. Recombinant TGF-ß2 at levels secreted by
differentiating keratinocytes also significantly inhibited
immortalization of primary cervical keratinocytes by HPV-16
E6/E7 genes. Senescent keratinocytes or E6/E7-expressing
cells approaching crisis were strongly growth inhibited by recombinant
TGF-ß2 at levels secreted by differentiating keratinocytes.
Furthermore, expression and secretion of TGF-ß2 increased 2-fold
(Fig. 4C
) in these cells. Autocrine TGF-ß has been
implicated in blocking growth and immortalization of keratinocytes in
response to vitamin D3 or retinoic acid (46
, 47)
. TGF-ß
isoforms actively induce cellular senescence (42)
and
programmed cell death (43)
, processes that directly
precede and counteract immortalization of keratinocytes. Therefore,
decreased secretion of TGF-ß2 in HPV-infected keratinocytes may
result in loss of an autocrine or paracrine pathway that inhibits
cellular immortalization in vitro.
Importance of TGF-ß2 in Cervical Cancer.
The role of TGF-ßs in epithelial carcinogenesis is complex.
During multistage carcinogenesis in the skin, loss of the tumor
suppressor function of TGF-ßs 1 and 2 accelerates malignant
progression. Loss of TGF-ß expression in mouse skin papillomas is
associated with suprabasal proliferation (48)
, increased
genomic instability (49)
, and rapid malignant progression
(5)
. The loss of TGF-ß responsiveness through somatic
mutations of TGF-ß receptor subunits I and II is common in many human
cancers (reviewed in Ref. 4
), including cervical carcinoma
(50)
. Loss or altered expression of TGF-ß1 has been
described as an early event in cervical carcinogenesis and frequently
observed in cervical intraepithelial neoplasia (51)
.
Frequent loss or decreased expression of TGF-ß2 also has been
described in cervical dysplasia (52)
. The latter
observations are consistent with the idea that TGF-ßs function as
tumor suppressors in cervical epithelia and also confirm that loss of
TGF-ß expression in HPV-infected keratinocytes is not restricted to
observations in cell culture. Thus, down-regulation of TGF-ß2
expression by HPV oncoproteins may represent a novel mechanism that
contributes to growth stimulation, immortalization, and carcinogenesis.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 To whom requests for reprints should be
addressed, Biology Department, Clarkson University, Potsdam, NY 13699.
Phone: (315) 268-6641; Fax: (315) 268-6610; E-mail: woodworth{at}clarkson.edu 
2 The abbreviations used are: TGF, transforming
growth factor; HPV, human papillomavirus; pRb, retinoblastoma protein;
GAPDH, glyceraldehyde phosphate dehydrogenase; E6BP, E6 binding
protein; E6AP, E6-associated protein, RT-PCR, reverse
transcription-PCR; MMP, matrix metalloproteinase; dNTP, deoxynucleotide
triphosphate; PAI, plasminogen activator inhibitor.
Received 11/24/99. Accepted 5/30/00.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A A Borbely, M Murvai, K Szarka, J Konya, L Gergely, Z Hernadi, and G Veress Survivin promoter polymorphism and cervical carcinogenesis J. Clin. Pathol., March 1, 2007; 60(3): 303 - 306. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. de Boer, E. S. Jordanova, G. G. Kenter, A. A. Peters, W. E. Corver, J. B. Trimbos, and G. J. Fleuren High Human Papillomavirus Oncogene mRNA Expression and Not Viral DNA Load Is Associated with Poor Prognosis in Cervical Cancer Patients Clin. Cancer Res., January 1, 2007; 13(1): 132 - 138. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. R. Pett, M. T. Herdman, R. D. Palmer, G. S. H. Yeo, M. K. Shivji, M. A. Stanley, and N. Coleman Selection of cervical keratinocytes containing integrated HPV16 associates with episome loss and an endogenous antiviral response PNAS, March 7, 2006; 103(10): 3822 - 3827. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. A. Borbely, M. Murvai, J. Konya, Z. Beck, L. Gergely, F. Li, and G. Veress Effects of human papillomavirus type 16 oncoproteins on survivin gene expression J. Gen. Virol., February 1, 2006; 87(2): 287 - 294. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Baez, A. Cantor, S. Fonseca, M. Marcos-Martinez, L. A. Mathews, C. A. Muro-Cacho, and T. Munoz-Antonia Differences in Smad4 Expression in Human Papillomavirus Type 16-Positive and Human Papillomavirus Type 16-Negative Head and Neck Squamous Cell Carcinoma Clin. Cancer Res., May 1, 2005; 11(9): 3191 - 3197. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. You, Y. Liu, M. J. Carey, C. L. Lowery, and P. L. Hermonat Defective 3A Trophoblast-Endometrial Cell Adhesion and Altered 3A Growth and Survival by Human Papillomavirus Type 16 Oncogenes Mol. Cancer Res., November 1, 2002; 1(1): 25 - 31. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. J. Berger, A. Baege, T. Guillemette, J. Deeds, R. Meyer, G. Disbrow, R. Schlegel, and R. Schlegel Insulin-Like Growth Factor-Binding Protein 3 Expression Increases during Immortalization of Cervical Keratinocytes by Human Papillomavirus Type 16 E6 and E7 Proteins Am. J. Pathol., August 1, 2002; 161(2): 603 - 610. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Nees, J. M. Geoghegan, T. Hyman, S. Frank, L. Miller, and C. D. Woodworth Papillomavirus Type 16 Oncogenes Downregulate Expression of Interferon-Responsive Genes and Upregulate Proliferation-Associated and NF-{kappa}B-Responsive Genes in Cervical Keratinocytes J. Virol., May 1, 2001; 75(9): 4283 - 4296. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
) and matrix-associated (
)
TGF-ß2 were determined by ELISA. B, levels of
intracellular and matrix-associated TGF-ß1. C,
relative levels of total cell-associated TGF-ß2 in keratinocytes
infected with retroviruses encoding vector-only (PLXSN);
HPV-16 E6 wild type (wt); a HPV-16 E6 mutants that binds
E6BP but does not degrade p53 (E6mutF2V), or efficiently degrades p53
(E6mutF125V); a dominant-negative p53 mini protein (p53
DD); HPV-16 E7 and 16 E6/E7. Values represent the means of
three individual experiments; bars, SD.
).
B, detection of biologically active (
