| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Oncology, Markers, Clinical Correlates |
Pathology Unit of the "F.Addarii" Institute of Oncology, Department of Hematology-Oncology, University of Bologna, 40138 Bologna, Italy [M. F., A. A., A. D., B. C., C. B., W. F. G.], Department of Pathology, Brigham and Women’s Hospital and Adult Oncology, Dana-Farber Cancer Institute, Harvard Medical School, Boston, Massachusetts 02115 [M. L.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The p27 gene was first identified in G1 cells arrested by transforming growth factor-ß (6, 7, 8) . Its protein is highly expressed in quiescent cells, where it preferentially binds to and inactivates cyclin A–E/Cdk2 complexes, thereby preventing cell entry into the S-phase (9) . The function of p27 as a negative regulator of the cell cycle, suggests its putative role as a tumor suppressor gene. In vivo, p27 knockout mice develop generalized hyperplasia and pituitary tumors (10, 11, 12) . In addition, heterozygous mice that are haploinsufficient for p27 have increased susceptibility for tumors (13) . However, no homozygous deletions and only rare point mutations have been found in the p27 gene in humans (14, 15, 16) . Reduced p27 levels in colon and lung carcinomas result from an accelerated proteolytic degradation via the ubiquitin-proteasome pathway (17) . Lost or decreased p27 protein expression has been correlated to aggressive behavior in human cancers of the breast, gastrointestinal tract, prostate, and lungs (18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29) .
The progression of human HCC has been putatively related with various cell cycle regulators (30 , 31) . In particular, overexpression of the p27 protein has been observed in confluent differentiating cells of the HBG hepatoma cell line and during experimental hepatic regeneration, providing further evidence that p27 may play a regulatory role in the liver in vivo (32 , 33) . It recently has been reported that p27 expression could be an independent prognostic marker for disease-free survival in HCC (34) .
The present study analyzed the expression of p27 in 54 human HCC specimens and in corresponding nonneoplastic tissues. Tissues from 17 cirrhoses with different etiologies were also included as nonneoplastic controls. We show that p27 is not expressed by normal hepatocytes, but that a subset of HCCs are high expressers of p27; the patients bearing these tumors fared better, independently of other variables.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
5 cm (n = 19).
Twenty-nine patients (54%) had HCV infection, 13 (24%) HBV infection,
and 5 (9%) had a history of alcohol abuse. The remaining 7 (13%) had
chronic liver diseases unrelated to HCV, HBV, or alcohol. The follow-up
interval ranged from 4 to 154 months from the date of surgery (mean, 50
months). None of the patients received postoperative adjuvant therapy.
Cirrhotic tissues from 17 patients with no concomitant HCC were also
selected as nonneoplastic controls. Among these cirrhoses six were HCV
correlated, six were HBV correlated, and five were associated with
alcohol abuse. After surgical resection, specimens from
neoplastic and nonneoplastic surrounding tissue were fixed in 10%
buffered formalin. In 20 patients, snap-frozen immediately adjacent
tissue (normal and tumoral) was available for protein extraction.
Informed consent was obtained from all patients.
IHC.
Tissue sections (4 µm) were cut, placed on silane-pretreated
slides, deparaffinized, and rehydrated through graded alcohol. Antigen
retrieval was performed by microwave heating at high power (750 W) in
10 mM sodium citrate buffer (pH 6) for four cycles of 5 min
each. Slides were then allowed to cool for 30 min prior to incubation
for 1 h at room temperature with the p27 monoclonal antibody
(Transduction Laboratories, Lexington, KY) diluted 1:200. Staining was
performed with the Envision monoclonal System (Dako, Carpinteria, CA).
The reaction was developed with 3,3diaminobenzidine. Slides were
counterstained with Mayer’s hematoxylin and mounted. Nonimmune mouse
serum was used as negative control. Lymphocytes were used as the
internal positive control for p27. Identical reaction times allowed
accurate comparison of all samples. IHC staining was evaluated
independently by two pathologists (M. F. and A. D.).
p27 immunostaining was scored as the ratio of strongly positive nuclei
to the total number of cells as described previously
(18, 19, 20, 21)
: score 0, no staining; score 1, <50% positive
nuclei; score 2, 50% positive nuclei. At least 20 high-power fields
were randomly chosen, and 2000 cells were counted.
WB Analysis.
In the 20 cases where snap-frozen tissue was available, histological
analysis of the tissues used for WB confirmed the presence of only a
minimal stromal or necrotic component in the tumor samples. Two-hundred
mg of each sample were promptly homogenized at 15,000 rpm with a
Polytron homogenizer (PT 3000; Brinkman, Westbury, NY) in 1 ml of
ice-cold lysis buffer containing PBS, 0.1% Triton X-100, 5
mM EDTA, 1 mM phenylmethylsulfonyl
fluoride, 1 mM DTT, 5 µg/ml leupeptin, 1 µg/ml
aprotinin, 5 µg/ml antipain, 1 µg/ml pepstatin, and 10 µg/ml
trypsin-chymotrypsin inhibitor (all from Sigma, St. Louis, MO). Lysates
were kept on ice for 30 min and then centrifuged at 13,000 rpm for 5
min at 4°C. Supernatants were collected, and protein quantification
was performed using the DC (detergent compatible) protein assay
(Bio-Rad, Hercules, CA). One hundred µg of extracted proteins were
heated at 95°C for 5 min in lysis buffer containing 1:5 bromphenol
blue (Sigma) and electrophoresed in a 10% SDS-PAGE gel for
2 h at 45 mA. Proteins were then transferred to Hybond 0.45 µm
nitrocellulose membrane (Amersham, Buckinghamshire, England) at
constant voltage (40 V) overnight at 4°C in transfer buffer (25
mM Tris, 190 mM glycine, 20% methanol). Blots
were then incubated with either the polyclonal anti-Cdk2 antibody
(Santa Cruz Laboratories, Santa Cruz, CA) or the anti-p27 monoclonal
antibody, diluted 1:1000 and 1:2500, respectively, for 1 h at room
temperature. Signals were revealed with the Envision system (Dako), the
reagent of which was diluted 1:500 for 1 h and developed
using the ECL detection kit (Amersham).
IP-WB.
For each sample, 300 µg of protein homogenate was diluted in 1 ml of
lysis buffer and precleared with 50 µL of a 1:1 slurry of protein
A-Sepharose (Sigma) in PBS on ice for 30 min. Samples were centrifuged
for 2 min at 2000 rpm at 4°C, and the supernatants were incubated
with the polyclonal anti-Cdk2 antibody, diluted 1:50, overnight at
4°C. After IP, complexes were recovered on protein A-Sepharose for
2 h in ice. Proteins were subsequently collected by brief
centrifugation and then submitted to WB analysis for the detection with
the anti-p27 monoclonal antibody. Control for the specificity of the IP
method was performed by immunoprecipitating all of the cell lysates
with the anti-p27 monoclonal antibody followed by probing of the blots
with the anti-Cdk2 polyclonal antibody.
Densitometry.
The intensities of each band on WB and IP-WB were quantitated on a RBR
Altair (Florence, Italy) densitometer using the Autogel image analysis
software program. The values of p27 bands were evaluated after
normalization for the intensity of a reference sample that was added to
each run and scored in AU. IP-WB band intensities were grouped as 0
(0–0.3 AU), 1+ (0.3–1 AU), and 2+ (>1 AU) using a non-IP WB p27 band
as reference (2+) standard.
Statistical Analysis.
Contingency table methods were used to analyze the univariate
association between p27 and other variables (age, sex, tumor grade,
tumor size, and liver status) and the univariate correlation between
survival status and all of the above-mentioned variables. Significance
was evaluated by Pearson’s 
2
test and
confirmed by Fisher’s exact test. Logistic regression was also
performed to assess survival with respect to the variables. Survival
curves were calculated using the Kaplan-Meier method, and the log-rank
test was used for analysis. Patients who died of other causes during
the follow-up period were treated as censored data in the survival
analysis. Univariate and multivariate RRs were calculated using Cox
proportional hazards regression. The RRs for age are represented as age
65 versus <65 years. For sex, the RRs are given as male
versus female. For grade and dimensions, Edmondson’s grade
I-II and tumor size <5 cm were used as baselines, respectively. All
calculations were performed using the SPSS 8.0 statistical software
package, and the results were considered statistically significant at
P < 0.05.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, a and
b). The localization of the p27 protein was nuclear in the
rare positive hepatocytes and in all of the mature bile-duct cells,
whereas it was mixed nuclear/cytoplasmic in the neo-ductules and
inflammatory cells. No difference was found in the cirrhotic tissues
adjacent to HCC compared with cirrhoses without HCC. Foci of both
small- and large-cell hepatocellular dysplasia showed p27 expression
comparable to the other cirrhotic nodules. WB for p27 was positive in
all normal liver samples (Fig. 2)
.
|
|
f). Twenty-six of 54 (48%) were low
expressers (score 1; Fig. 1
e) and 14 of 54 (26%) expressed
high levels of p27 (score 2; Fig. 1
, c and d).
Diffuse cytoplasmic immunostaining was also noted in 24 cases. Of
these, 8 were high nuclear expressers and 14 were low nuclear
expressers. In two cases, the p27 immunostaining was exclusively
cytoplasmic. Among the 14 high p27 expressers, 5 were grade I-II and 9
were grade III-IV HCCs. Among the low expressers, 13 were grade I-II
and 27 were grade III-IV (Table 1)
.
Twenty HCC samples showed either p27 or Cdk2 bands with similar
intensity by WB (Fig. 2)
.
|
. IP-WB revealed the presence of
a p27 band in all 20 normal samples examined, with similar
densitometric values ranging from 0.6 to 1.9 AU. Conversely, samples
from the tumor counterparts in the same cases displayed different
intensities of the p27 bands (see "Materials and Methods" for AU
intervals). In particular, 6 of 20 HCC (30%) samples showed no signal
(densitometric value, 0), 10 of 20 (50%) displayed a weak band
(densitometric value, 1+), and the remaining 4 (20%) had a strong p27
band (densitometric value, 2+; Fig. 2
). Detection of the p27 band after
IP with Cdk2 revealed a positive highly statistically significant
correlation (P = 0.00) to high expression of p27
by IHC (Fig. 3)
. In particular, the four
cases that displayed a strong band (densitometric value, 2+) in IP-WB
analysis were all high p27 expressers in IHC (score 2). In addition,
the six cases that displayed a densitometric value of 0 after IP were
low p27 expressers by IHC (score for five cases, 0; score for one case,
1). Control IP-WB performed by immunoprecipitating tumor cell lysates
with p27 followed by probing with Cdk2 revealed bands with similar
intensities in all cases (data not shown).
|
). No significant correlation was
found between viral or alcoholic etiology (data not shown). Concerning
survival, only 2 of 14 (14%) patients in the high-expresser group died
of disease versus 22 of 40 (55%) in the low-expresser group
(Table 2)
. When all variables were
compared separately to survival status, only p27 (P =
0.008), the presence of cirrhosis (P = 0.007), and age
(P = 0.013) significantly influenced survival (Table 2)
. Logistic regression confirmed that cirrhotic status implies a
5.9-fold higher risk of mortality with respect to noncirrhotic status.
Low or absent p27 expression had a 10:1 higher risk of mortality
than high expression.
|
), as did those of cirrhosis
versus non-cirrhosis (P = 0.015) and age
65 years versus <65 years (P = 0.021).
When a multivariate Cox proportional hazard model was constructed
(including gender, age, tumor grade, tumor size, liver status, and p27
expression), p27 was the strongest independent predictor of survival
(P = 0.027; RR = 5.313), the second predictor
being liver status (Table 3)
.
|
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
1 million cases a year, and
its prognosis is poor (36)
. Pathological features
such as tumor grade and size are the most frequently used prognostic
parameters for HCC (37, 38, 39)
. Proliferative index, p53
mutations, and the expression of cell cycle-related genes have also
been proposed as negative predictors of biological behavior, and
recently p27 has been proposed for independent prediction of clinical
recurrence (31
, 34
, 40
, 41)
. Our data confirmed that HCC patients who express p27 fare better than those who do not. This is in keeping with recent data suggesting a strong predictive value for p27 in HCC and other human epithelial tumors (18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29 , 34) . Univariate analysis showed that p27 expression does not correlate with any of the clinicopathological parameters analyzed in this study. In multivariate analysis, p27 expression was the strongest predictor of survival for HCC patients, independent of the other variables. Not unexpectedly, a significant association between presence of cirrhosis and poor prognosis was also found (42) . In fact, in cirrhotic patients a more extensive surgical resection increases the risk of liver failure, and development of a further HCC is also more likely.
High expression of nuclear p27 was generally found in bile-duct cells of both cirrhotic and noncirrhotic livers but only rarely in hepatocytes. Despite the fact that hepatocytes have a very low proliferative rate, only low levels of p27 were observed by IHC in normal tissues. Lack of correlation between p27 protein levels and proliferative rates has been extensively demonstrated in a variety of tumors. Quiescent cells in normal tissues generally express nuclear p27. However, it is unclear whether p27 is expressed to maintain cells in G0 or as a result of the induction/maintenance of differentiation. In fact, colonic epithelial cells express p27 in superficial terminally differentiated cells, whereas benign breast epithelial cells commonly express high levels of p27 (18 , 21) . It is possible that hepatocytes are kept in the resting state by other cell-cycle inhibitory proteins. One possible explanation is that similarly to what happens in esophageal dysplasia, up-regulation of p27 in HCC occurs in the presence of increased proliferative signals (27) . Its up-regulation may then result in a more controlled growth pattern and consequently in less aggressive behavior.
Induction of p27 in HCC results in its binding to Cdk2 (2) . This further underscores the hypothesis that increased expression of p27 occurs as an attempt to block Cdk2 activity in tumor cells. Our results showed that high p27 expression in tumor cells by IHC correlated with increased p27 binding to Cdk2 on tumor samples by IP-WB. The failure to show differences in the p27 levels by WB without IP among different normal and tumor samples was probably attributable to the presence of inflammatory, stromal, and bile-duct elements contaminating the cell lysates, as seen previously (20) . Whereas the variable levels of Cdk2-bound p27 observed at IP-WB among our tumor samples may reflect an increase of p27 in the smaller proportion of cells that are truly malignant. On technical grounds, our data revealed that IHC is the most reliable technique for the detection of p27. Indeed, IHC was particularly sensitive for cellular localization of the p27 protein and assessment of its active fraction.
The significance of cytoplasmic p27 ICH localization is still a matter of debate and generally is not considered in the assessment of the p27 IHC score. However, in adenocarcinomas associated with Barrett’s esophagus, the presence of cytoplasmic p27 was found to correlate with poor prognosis (27) . These clinical data were corroborated recently by experimental studies on thyroid and liver cells showing that cytoplasmic sequestration of the p27 protein by cyclins D1 and D3 prevents it from penetrating the nucleus and exerting its inhibitory effect on the G1-S transition (43 , 44) . In our series of HCCs, reliable cytoplasmic p27 staining was observed in 44% of cases, mostly in association with the nuclear signal. Biochemical studies using subcellular fractionation of protein lysates are needed to confirm the role of cytoplasmic localization in preventing p27 function in HCC.
The life expectancy of patients with HCC is generally poor because of a lack of truly effective therapeutic agents and the rarity of early diagnosis. Therapy consists of either surgery (liver resection or transplantation) or local treatments (chemo-embolization, alcoholization). Predictive studies recently have shown that the flavonoid antioxidant silibinin induces G1 arrest in prostate cancer cells via an increased binding of p27 to Cdk2 (45) . Furthermore, the anticancer agent indole-3-carbinol in combination with tamoxifen was found to arrest breast cancer cells in the G1 phase via a decrease in Cdk2-specific enzymatic activity (46) . In the meantime, we recommend IHC detection of p27 on routine tissue sections as a reliable new method to help predict survival of individual HCC patients and possibly to help direct future therapeutic strategies.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 This research was supported by the Ministero
della Ricerca Scientifica e Tecnologica (MURST) and Universita’ degli
Studi di Bologna. 
2 To whom requests for reprints should be
addressed, at Istituto Oncologico "F.Addarii," Viale Ercolani 4/2,
40138 Bologna, Italy. Phone: 011-39-051-6364546; Fax:
011-39-051-6364403; E-mail: grigioni{at}med.unibo.it
3 The abbreviations used are: Cdk,
cyclin-dependent kinase; HCC, hepatocellular carcinoma; HCV, hepatitis
C virus; HBV, hepatitis B virus; IHC, immunohistochemistry; WB, Western
blot; IP, immunoprecipitation; AU, arbitrary units; RR, risk ratio.
Received 1/24/00; revised 6/20/00; accepted 7/14/00.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  Y. Yuan, J. Wang, J. Li, L. Wang, M. Li, Z. Yang, C. Zhang, and J. L. Dai Frequent Epigenetic Inactivation of Spleen Tyrosine Kinase Gene in Human Hepatocellular Carcinoma. Clin. Cancer Res., November 15, 2006; 12(22): 6687 - 6695. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Varghese, C. Agarwal, A. Tyagi, R. P. Singh, and R. Agarwal Silibinin Efficacy against Human Hepatocellular Carcinoma Clin. Cancer Res., December 1, 2005; 11(23): 8441 - 8448. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. B. Thomas and J. L. Abbruzzese Opportunities for Targeted Therapies in Hepatocellular Carcinoma J. Clin. Oncol., November 1, 2005; 23(31): 8093 - 8108. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Carloni, F. Vizzutti, and P. Pantaleo Farnesyltransferase Inhibitor, ABT-100, Is a Potent Liver Cancer Chemopreventive Agent Clin. Cancer Res., June 1, 2005; 11(11): 4266 - 4274. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. P. Claudio, G. Russo, C. A. C. Y. Kumar, C. Minimo, A. Farina, S. Tutton, G. Nuzzo, F. Giuliante, G. Angeloni, V. Maria, et al. pRb2/p130, Vascular Endothelial Growth Factor, p27(KIP1), and Proliferating Cell Nuclear Antigen Expression in Hepatocellular Carcinoma: Their Clinical Significance Clin. Cancer Res., May 15, 2004; 10(10): 3509 - 3517. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Denicourt and S. F. Dowdy Cip/Kip proteins: more than just CDKs inhibitors Genes & Dev., April 15, 2004; 18(8): 851 - 855. [Full Text] [PDF]
|
|
|
|
|
|
C. Armengol, G. Tarafa, L. Boix, M. Sole, R. Queralt, D. Costa, O. Bachs, J. Bruix, and G. Capella Orthotopic Implantation of Human Hepatocellular Carcinoma in Mice: Analysis of Tumor Progression and Establishment of the BCLC-9 Cell Line Clin. Cancer Res., March 15, 2004; 10(6): 2150 - 2157. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Fiorentino, A. Altimari, M. Ravaioli, E. Gruppioni, E. Gabusi, B. Corti, M. Vivarelli, P.-P. Bringuier, J.-Y. Scoazec, W. F. Grigioni, et al. Predictive Value of Biological Markers for Hepatocellular Carcinoma Patients Treated with Orthotopic Liver Transplantation Clin. Cancer Res., March 1, 2004; 10(5): 1789 - 1795. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Matsuda, T. Ichida, T. Genda, S. Yamagiwa, Y. Aoyagi, and H. Asakura Loss of p16 Contributes to p27 Sequestration by Cyclin D1-Cyclin-dependent Kinase 4 Complexes and Poor Prognosis in Hepatocellular Carcinoma Clin. Cancer Res., August 1, 2003; 9(9): 3389 - 3396. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Fiorentino, A. Altimari, A. D'Errico, E. Gabusi, P. Chieco, M. Masetti, and W. F. Grigioni Low p27 Expression Is an Independent Predictor of Survival for Patients with Either Hilar or Peripheral Intrahepatic Cholangiocarcinoma Clin. Cancer Res., December 1, 2001; 7(12): 3994 - 3999. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. GOUKASSIAN, A. DIEZ-JUAN, T. ASAHARA, P. SCHRATZBERGER, M. SILVER, T. MURAYAMA, J. M. ISNER, and V. ANDRES Overexpression of p27Kip1 by doxycycline-regulated adenoviral vectors inhibits endothelial cell proliferation and migration and impairs angiogenesis FASEB J, September 1, 2001; 15(11): 1877 - 1885. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | Cell Growth & Differentiation |
