Pathology International

2002;

52

: 103–109

Original Article

Blackwell Science, LtdOxford, UKPINPathology International1320-54632002 Japanese Society of Pathology

522February 20021321

pRb and p53 pathways in lung cancersQ. Xue

et al.10.1046/j.1320-5463.2002.01321.x

Original Article103109BEES SGML

Correspondence: Takashi Nakajima, MD, Second Department ofPathology, Gunma University School of Medicine, 3-39-22, Showa-machi, Maebashi, Gunma 371-8511, Japan. Email: nakajima@med.gunma-u.ac.jp

Received 30 March 2001. Accepted for publication 16 November2001.

Aberrant expression of pRb, p16, p14ARF, MDM2, p21 and p53 in stage I adenocarcinomas of the lung

Qi Xue, Takaaki Sano, Kenji Kashiwabara, Masako Saito, Tetsunari Oyama and Takashi Nakajima

Second Department of Pathology, Gunma University School of Medicine, Maebashi, Japan

Cancers are always associated with cell cycle abnormali-ties. To clarify the cell cycle abnormalities present in lungadenocarcinomas, various cell cycle regulatory proteins ofboth the pRb and p53 pathways were studied immunohis-tochemically in 50 cases of stage I adenocarcinoma of thelung. In regard to the pRb pathway, most adenocarcinomasshowed frequent expression of both p16 and pRb proteins,and aberrant expression in the pRb pathway was observedin about one-quarter of stage I adenocarcinomas. In regardto the p53 pathway, the frequency of immunohistochemicalpositivity was 8% for p14ARF, 64% for MDM2, 20% for p53and 24% for p21. In this pathway, the immunohistochemicalprofile of p14ARF-negative/MDM2-positive/p53-negative ischaracteristic of stage I adenocarcinoma of the lung. Aninverse relationship was found between MDM2 and p53 pro-tein and was associated with the differentiation status ofstage I adenocarcinoma of the lung. Our results suggestthat the disruption of the pRb and p53 pathways is fre-quently observed in the early stages of lung adenocarci-noma and might play an important role in the growth anddifferentiation of adenocarcinoma of the lung.

Key words:

adenocarcinoma, immunohistochemistry, lung, p53pathway, pRb pathway

Most human cancers are caused by deregulation of cell cyclecontrols that are mediated by various cell cycle regulatoryproteins, the majority of which are tumor suppressor pro-teins.

1–4

Abnormalities at the G

1

and G

2

checkpoints are themost critical for the disruption of cell cycle control, and thesecheckpoints are functionally influenced by two major tumorsuppressor proteins, pRb and p53.

1

The cell cycle regulatoryprotein p16, encoded by the

CDKN2A

gene, is a cyclin-dependent kinase (CDK) inhibitor and negatively regulatesthe cyclin D-dependent pRb phosphorylation from G

1

to S-

phase by the sequestration of E2F1.

2

At the G

1

checkpoint,pRb phosphorylation is stimulated by cyclin D1 and inhibitedby p16. Therefore, the p16–CDK4–cyclin D1–pRb cascade isimportant for cell cycle regulation in the G

1

checkpoint, andthis suppressor mechanism is thought to function in a singleregulatory pathway of the cell cycle, recently called the Rbpathway.

1,3

The tumor suppressor p53 plays an important role in suchprocesses as cell cycle control, DNA repair and apoptosis.

4

Mutations leading to deregulation of p53 are the most fre-quent genetic abnormalities in human cancers. In cell cyclecontrol, it is well established that p53 influences the Rb path-way through the induction of p21.

5

Recent studies haveshown that the inactivation and activation of p53 are regu-lated by MDM2 and the checkpoint kinase Chk2, respec-tively.

6

MDM2 binds directly to the p53 transactivation domainand downregulates the p53 transcriptional activity.

7,8

Chk2causes G

2

arrest by inhibiting cdc25C, which is independentof p53.

6

Moreover, the product of the

CDKN2A

gene,p14ARF, has been shown to bind directly to MDM2, resultingin the stabilization of both p53 and MDM2.

9

Therefore, thep14ARF–MDM2–p53 cascade is another cell cycle regula-tory system, called the p53 pathway, that is different from thepRb pathway and may have a close connection to G

2

cellcycle arrest.

10,11

As p16 and p14ARF are upstream of bothpathways and are both transcribed from the

CDKN2A

gene,the pRb and p53 pathways are closely connected and arelinked to each other through E2F1.

11

In human cancers, abnormalities in the pRb and p53 path-ways are frequent events. The p53 point mutation is the mostfrequent genetic change, with more than 60% of human can-cers showing this abnormality in the p53 pathway.

12,13

Mostsmall-cell lung cancers show a loss of function both for pRband p53, as well as an unknown tumor suppressor proteinlocated on chromosome 3p.

14–16

Even in non-small-cell lungcancers (NSCLC), previous reports have frequently sug-gested this functional loss of pRb and p53.

17–20

Althoughadenocarcinoma and squamous cell carcinoma have bothbeen categorized as NSCLC, these cancers show differ-

104 Q. Xue

et al

.

ences in the expression of several cell cycle regulators, suchas pRb, p53, p16 and MDM2.

17–20

In order to clarify the char-acteristic features of cell cycle regulator expression in lungadenocarcinomas, we analyzed the expression of pRb, p16,p14ARF, MDM2, p53 and p21 in 50 cases of stage I lung ade-nocarcinomas using immunohistochemical procedures.

MATERIALS AND METHODS

Lung cancer tissues

A total of 50 cases of adenocarcinoma of the lung, evaluatedby the pathological international TNM classification as stageI, were randomly selected from our pathology files. None ofthe cases had undergone any chemotherapy before opera-tion. After operation, the resected lung tissues were fixed with15% formalin and processed for routine pathological diagno-sis. A paraffin block containing the representative tumor his-tology of each case was selected, and paraffin sections of4

µ

m thickness were prepared. These were subsequentlyused for immunohistochemical study, as well as for the re-evaluation of tumor histology, following hematoxylin andeosin staining, using the new WHO classification (1999).

Clinicopathological information was obtained from ourpathology files. The cases comprised 24 men and 26 women,and had an age distribution of 41–82 years (mean =64 years). According to the histological grade of tumor differ-entiation, 31 of the 50 cases were re-evaluated as well-differentiated adenocarcinoma, while the remaining caseswere composed of 14 moderately differentiated and fivepoorly differentiated tumors.

Immunohistochemistry of cell cycle regulatory proteins

Immunohistochemical methods for pRb, p16, p14ARF, p53and p21 have been described previously in our recent stud-ies.

21–24

For the present study, the details of the primary anti-bodies used are as follows: G3-245 mouse monoclonal IgG1antibody against pRb (1 : 200; PharMingen, San Diego, CA,USA); JC8 mouse monoclonal IgG2a antibody against p16(1 : 100; Dr J. Koh, Massachusetts General Hospital CancerCenter, Boston, MA, USA); DO-7 mouse monoclonal IgG2bantibody against p53 (1 : 100; Novocastra Laboratories,Newcastle, UK); C-18 goat polyclonal antibody againstp14ARF (1 : 100; Santa Cruz Biotechnology, Santa Cruz,CA, USA); 4D10 mouse monoclonal IgG1 antibody againstp21 (1 : 40; Novocastra Laboratories). After antigen retrievaltreatment with each appropriate buffer, the sections wereincubated with each primary antibody at 4

°

C overnight. Forthe immunohistochemistry of MDM2, antigen retrieval wascarried out by boiling in 0.01 mol/L sodium phosphate–citric

acid buffer, pH 8.0, for 10 min. The sections were subse-quently incubated with a primary antibody against MDM2(1 : 1000; SMP14 mouse monoclonal IgG antibody; SantaCruz Biotechnology) at 4

°

C overnight. After treatment with abiotinylated secondary antibody, the sections were incubatedwith a streptavidin–biotin–peroxidase complex (Vectastain;Vector Laboratories, Burlingame, CA, USA) for 30 min atapproximately 20

°

C. Visualization was carried out with 0.02%3-3

′

-diaminobenzidine tetrahydrochloride as chromogen con-taining 0.005% H

2

O

2

in 50 mmol/L ammonium acetate–citricacid buffer, pH 6.0. The slides were then lightly counter-stained with hematoxylin. For positive controls, paraffinsections of various cancers, known to be positive for eachantibody, were used.

Evaluation of the immunohistochemistry

For the evaluation of immunostaining, the criteria used bypreviously published investigators were partly followed in thisstudy.

20,24–26

For the immunohistochemistry of pRb, p53 andp21, only nuclear staining was observed. In contrast, p16,p14ARF and MDM2 showed both nuclear and cytoplasmicimmunostaining. However, only nuclear staining was consid-ered in scoring immunoreactivity; cytoplasmic immunoreac-tivity was not considered to be evident of protein expression.The immunoreactivity for all the cell cycle regulatory proteinsinvestigated in this study was evaluated as strongly positive(++) when more than 50% of tumor nuclei were stained,positive (+) when more than 10% of the nuclei were stained,and negative (–) when less than 10% of the tumor nucleiwere stained or when the tumor cells completely lackedimmunoreactivity.

Statistical analysis

The

χ

2

-test and Fisher’s exact probability test were used forstatistical analysis of the results.

P

-values less than 0.05were considered to be significant.

RESULTS

Immunohistochemical analysis

Immunohistochemical analysis revealed that p16 immunore-activity was present in the nuclei, as well as in the cytoplasm,of the adenocarcinoma cells, and that 38 of 50 adenocarci-nomas (76%) were positive for p16 (Table 1). Of the 38 ade-nocarcinomas, 28 showed strong p16 staining, recorded as++ (Fig. 1a). However, p16 immunoreactivity differed fromarea to area, so that heterogeneous expression of p16 was

pRb and p53 pathways in lung cancers 105

the predominant feature of the adenocarcinomas, except in11 cases that were strongly positive for p16 (++) but com-pletely negative for pRb immunoreactivity (Fig. 2).

Immunoreactivity for pRb was located exclusively inthe nuclei of the tumor cells and almost all of the nucleishowed strong and diffuse immunoreactivity for pRb. Of the50 adenocarcinomas, 38 (76%) showed pRb immuno-reactivity, including 13 cases with intense and diffuse stain-ing, recorded as ++ (Table 1). Eleven adenocarcinomas withno pRb immunoreactivity showed strong p16 staining,recorded as ++ (Fig. 1). Although statistically not significant,an inverse expression pattern between pRb and p16 wasobserved in 22 (44%) adenocarcinomas.

In this study, immunoreactivity for p14ARF was presentexclusively in the nuclei of the tumor cells. As shown inTable 1, however, only four tumors (8%) were immunohis-tochemically positive for p14ARF, which was not significantlyrelated to p16 immunostaining. MDM2 immunoreactivity wasseen strongly in the nuclei of the tumor cells, in addition tofaint cytoplasmic staining. Of the 32 adenocarcinomas withMDM2 positivity, 17 were scored as strongly positive forMDM2. Although all of the adenocarcinomas positive forMDM2 (except one) lacked any immunoreactivity for p14ARF,no significant statistical correlation was observed between

p14ARF and MDM2 immunoreactivity (

P

= 0.12667). Immu-noreactivity for p53 was observed in 10 tumors, of which nineshowed strong and diffuse nuclear staining. As shown inTable 2, there was a strong inverse relationship betweenMDM2 and p53, and a less-strong relationship betweenp14ARF and p53. In this study, p21 immunoreactivity wasobserved in 12 (24%) of 50 adenocarcinomas; however, sta-tistical analyses did not show any significant correlationbetween p21 and p53 or MDM2 (data not shown).

Clinicopathological relationship to the immunohistochemical results

No significant relationship was observed between any ofthe immunohistochemical results and the age distribution.Immunoreactivity for p53 was observed in three (11%) and

Table 1

Immunohistochemical results of p16, pRb and p14 in lungadenocarcinomas

p16 (+/++) p16 (–)Percentage (76) (24)

P

-value

p14ARF (+/++) 8 3 1p14ARF (–) 92 35 11 NS

pRb (+/++) 76 27 11pRb (–) 24 11 1 NS

NS, not significant.

Figure 1

Inverse relationship between p16 and pRb expression in well-differentiated papillary adenocarcinoma. Tumor cells were

(

a

)

stronglypositive for p16 immunostaining, but (

b

) completely negative for pRb immunoreactivity (H).

a b

Figure 2

Heterogeneous expression of p16 in well-differentiatedadenocarcinoma. Adenocarcinoma tissue at the right side showedstrong p16 immunoreactivity (H).

106 Q. Xue

et al

.

seven (29%) tumors from 26 women and 24 men, respec-tively (

P

= 0.16). Meanwhile, MDM2 immunoreactivity wasobserved in tumors from 19 women (73%) and 13 men (54%)(

P

= 0.16). Compared with the histological differentiation ofthe adenocarcinomas, all but one of the tumors were stronglypositive for MDM2 but negative for p14ARF, and belongedto well-differentiated adenocarcinomas (Fig. 3a,b). MDM2staining was observed in 25 (81%) of 31 well-differentiated

adenocarcinomas, and decreased to six (43%) of 14 moder-ately differentiated and one (20%) of five poorly differentiatedtumors (Table 3). Conversely, the percentage of p53 immu-noreactivity decreased with adenocarcinoma differentiation(Fig. 3c). Of the five poorly differentiated adenocarcinomas,three (60%) were strongly and diffusely positive for p53(Fig. 3d), and the moderately and well-differentiated tumorsrevealed p53 positivity in four and three tumors, respec-tively (

P

= 0.015). No significant correlation was observedbetween the histological differentiation of the tumor cellsand expression of the other cell cycle regulatory proteins(Table 3).

DISCUSSION

Disruption of the pRb pathway is a frequent molecular eventin lung cancers. Aberrant expression of the p16 protein isthe most frequent abnormality in the pRb pathway ofNSCLC.

17,19,20

Homozygous deletion and promoter methyla-

Figure 3

The expression of cell cycle regulatory proteins in well- and poorly differentiated adenocarcinoma. A well-differentiated bronchi-oloalveolar carcinoma was

(

a

)

immunohistochemically negative for p14ARF,

(

b

)

strongly positive for MDM2 and

(

c

)

negative for p53, exceptfor a few cells. However,

(

d

)

three of five poorly differentiated adenocarcinomas showed strong and diffuse immunostaining for p53, suggestingthe point mutation of the gene was present (H).

a b

c d

Table 2

Immunohistochemical results of p53, p14ARF, MDM2 andp21 in lung adenocarcinomas

p53(+/++) p53 (–)Percentage Percentage (20) (80)

P

-value

p14ARF (+/++) 8 3 1p14ARF (–) 92 7 39 < 0.05

MDM2 (+/++) 64 1 31MDM2 (–) 36 9 9 < 0.05

p21 (+/++) 24 3 9p21 (–) 76 7 31 NS

NS, not significant.

pRb and p53 pathways in lung cancers 107

tion are the major causes of loss of p16 function, andprevious immunohistochemical results have shown p16expression to be reduced in 27–67% of NSCLC.

17,20,27,28

Moreover, the abnormal p16 expression is more frequentlyfound in squamous cell carcinoma than in adenocarcinomasamong NSCLC.

17,20

Our results found that more than two-thirds of the adenocarcinomas were immunohistochemicallypositive for p16, and the staining pattern was heterogeneousin about half of the adenocarcinoma cases. It remains to beclarified whether the heterogeneous expression of p16 inadenocarcinomas may be attributed to gene expression(caused by an unknown mechanism) or by epigenetic pro-moter methylation. Furthermore, several immunohisto-chemical studies, including ours, have already shown thereciprocal relationship between pRb and p16 expres-sion.

17,20,27–30

This relationship is characteristic of small-celllung and cervical cancers, in which pRb function is alreadydestroyed, resulting in an increased E2F1 level and the sub-sequent overexpression of p16.

26,27,31

Although the recipro-cal relationship between pRb and p16 may also explain theheterogeneous expression of pRb in adenocarcinoma, thisphenomenon has not been observed in squamous cell car-cinoma of the lung.

24

In the p53 pathway, p14ARF is a key upstream protein andMDM2, an E3 ligase that targets both p53 and itself forubiquitination, is downregulated by p14ARF.

9,11,32–35

MDM2creates a negative feedback loop as p53 activates expressionof MDM2, keeping p53 levels low during normal growth anddevelopment.

33

Moreover, there is an autoregulatory feed-back loop in which p53 downregulates the expression ofp14ARF.

34–36

Because of the interactive nature of cell cycleregulation, many studies have focused on the abnormalexpression of these cell cycle regulator proteins to elucidatethe disruptions in the p53 pathway in various cancers, includ-ing NSCLC.

18–27,30,37

Previous reports have shown that p14ARF is down-regulated in about half of NSCLC and the frequency ofimmunohistochemical MDM2 positivity varies from 24 to78%.

18,19,24,37,38

Moreover, p53 is immunohistochemicallypositive in up to 50% of NSCLC, the majority of which showdiffuse p53 staining, suggesting p53 gene mutation.

18,19,24

Recent advances in molecular biology have argued whetherloss of p14ARF function is caused by genetic or epigenetic

changes.

36,37,39–41

Although p14ARF mutation is infrequentin NSCLC, downregulation of p14ARF is reportedly causedby the deletion of the

INK4A

/

ARF

locus and by promotermethylation in various cancers.

39–42

Although the mechanismfor decreased p14ARF immunoreactivity in adenocarcinomaof the lung is unclear, immunohistochemical downregulationof p14ARF is more prominent in adenocarcinoma than insquamous cell carcinoma of the lung.

24

Previous studies on NSCLC have not defined conclusivelythe relationship between MDM2 overexpression and its clini-copathological events.

18,38,43

MDM2 mRNA overexpressionis a favorable prognostic factor in NSCLC and immunohis-tochemical MDM2-positive and p53-negative patients tend tohave better prognoses.

38,43

However, several reports contra-dict these results.

18,19

In 1996, Gorgoulis

et al

. noted that theMDM2-positive/p53-negative immunohistochemical profile ismore often seen in adenocarcinomas, especially well-differentiated adenocarcinomas, than in other histologicaltypes of lung cancer.

18

In the present study, we confirmed thesignificant correlation of the MDM2-positive/p53-negativeimmunohistochemical profile with the well-differentiated ade-nocarcinoma phenotype, which may give a better prognosisfor the patient, as suggested previously.

38,42

In addition, ourstudy suggests that the immunohistochemical profile forMDM2 and p53 might be closely related to the differentiationgrade of lung adenocarcinomas.

In NSCLC, immunohistochemical overexpression of p53and mutation of p53 occur frequently, being reported in morethan 50% of cases.

18,19,24

The incidence of p53 point muta-tions generally increases with higher grades of malignancyand increased tumor stages in various human cancers.

44,45

Recent molecular and immunohistochemical studies haveshown that inactivation of the

INK4A

/

ARF

locus frequentlycoexists with p53 mutation, and that coinciding overexpres-sion of MDM2 and p53 is significantly associated with lymphnode metastasis.

18,19,37,43

However, several previous studiescontradict these findings. These show that immunohis-tochemical p53 overexpression or mutation inversely corre-lates with the inactivation of p14ARF and/or with MDM2overexpression.

23,38,42

Our present study shows that immu-nohistochemical inactivation of both p14ARF and p53 asso-ciated with immunohistochemical overexpression of MDM2seems to be a characteristic feature of stage I lung adeno-

Table 3

Percentage of cases showing immunohistochemical positivity for p53, p14ARF, MDM2 and p21 according to the differentiation ofadenocarcinomas

No. tumors pRb p16 p14ARF MDM2* p53* p21

Well differentiated 31 77 84 3 81 10 26Moderately differentiated 14 71 57 14 43 29 21Poorly differentiated 5 80 80 20 20 60 20

*

P

< 0.05

108 Q. Xue

et al

.

carcinoma. This finding may result from the fact that our ade-nocarcinomas included 34 cases of well-differentiated tumorwith low frequency of immunohistochemical p53 overexpres-sion. Although speculative, tumors with the p14ARF-negative/MDM2-positive/p53-negative immunohistochemicalprofile shown in this study might mimic the mechanism of nor-mal growth and development.

33

In conclusion, aberrant expression in the pRb pathwaywas observed immunohistochemically in approximately one-quarter of the stage I adenocarcinoma cases. In the p53pathway, the p14ARF-negative/MDM2-positive/p53-negativeimmunohistochemical profile was a feature characteristic ofstage I adenocarcinoma of the lung. An inverse relationshipwas found between the MDM2 and p53 proteins and was asso-ciated with the differentiation status of stage I adenocarcinomaof the lung. Our results suggest that disruption of the pRband p53 pathways is frequently observed in the early stagesof lung adenocarcinoma and might play an important role inthe growth and differentiation of adenocarcinoma of the lung.

ACKNOWLEDGMENTS

This work was supported in part by a Grant-in-Aid from theMinistry of Education, Science, Sports and Culture, and fromthe Ministry of Health and Welfare. The authors thank MrFutoshi Hara and Mr Toshiaki Hikino for technical assistance.

REFERENCES

1 Serrano M, Hannon GJ, Beach D. A new regulatory motif incell-cycle control causing specific inhibition of cyclin D/CDK4.

Nature

1993;

366

: 704–707.2 Sheer CJ. The Pezcoller Lecture. Cancer Cell Cycles Revisited.

Cancer Res.

2000;

60

: 3689–3695.3 Koh J, Enders GH, Dynlacht BD, Harlow E. Tumour-derived p16

alleles encoding proteins defective in cell-cycle inhibition.

Nature

1995;

375

: 506–510.4 Prives C, Hall PA. The p53 pathway.

J. Pathol.

1999;

187

: 112–126.

5 Chen X, Ko LJ, Jayaraman L, Prives C. p53 levels, functionaldomains, and DNA damage determine the extent of the apop-totic response of tumor cells.

Genes Dev.

1996;

10

: 2438–2451.

6 Hirao A, Kong YY, Matsuoka S

et al.

DNA damage-inducedactivation of p53 by the checkpoint kinase Chk2.

Science

2000;

287

: 1824–1827.7 Zauberman A, Barak Y, Ragimov N, Levy N, Oren M.

Sequence-specific DNA binding by p53: identification of targetsites and lack of binding to p53–MDM2 complexes.

EMBO J.

1993;

12

: 2799–2808.8 Barak Y, Juven T, Haffner R, Oren M. mdm2 expression is

induced by wild type p53 activity.

EMBO J.

1993;

12

: 461–468.9 Pomerantz J, Schreiber-Agus N, Liegeois NJ

et al.

The Ink4atumor suppressor gene product, p19Arf, interacts with MDM2and neutralizes MDM2′s inhibition of p53. Cell 1998; 92: 713–723.

10 Chin L, Pomerantz J, Depinho RA. The INK4a/ARF tumor sup-pressor: one gene-two products-two pathways. Trends Bio-chem. Sci. 1998; 23: 291–296.

11 Stott FJ, Bates S, James MC et al. The alternative product fromthe human CDKN2A locus, P14ARF, participates in a regulatoryfeedback loop with p53 and MDM2. EMBO J. 1998; 17: 5001–5014.

12 Takahashi T, Nau MM, Chiba I et al. p53: a frequent target forgenetic abnormalities in lung cancer. Science 1989; 246: 491–494.

13 Nigro JM, Baker SJ, Preisinger AC et al. Mutation in the p53gene occurs in diverse human tumor types. Nature 1989; 342:705–708.

14 Todd S, Franklin WA, Varella-Garcia M et al. Homozygous dele-tions of human chromosome 3p in lung tumors. Cancer Res.1997; 57: 1344–1352.

15 Harbour JW, Lai SL, Whang-Peng J, Gazdar AF, Minna JD,Kaye FJ. Abnormalities in structure and expression of thehuman retinoblastoma gene in SCLC. Science 1988; 241: 353–357.

16 Takahashi T, Takahashi T, Suzuki H et al. The p53 gene is veryfrequently mutated in small-cell lung cancer with a distinctnucleotide substitution pattern. Oncogene 1991; 6: 1775–1778.

17 Kratzke RA, Greatens TM, Rubins JB et al. Rb and p16INK4a

expression in resected non-small-cell lung tumors. Cancer Res.1996; 56: 3415–3420.

18 Gorgoulis VG, Rassidakis GZ, Karameris AM et al. Immunohis-tochemical and molecular evaluation of the mdm-2 gene prod-uct in bronchogenic carcinoma. Mod. Pathol. 1996; 9: 544–554.

19 Gorgoulis VG, Zacharators P, Kotsinas A et al. Alteration of thep16–pRb pathway and the chromosome locus 9p21–22 in non-small-cell lung carcinoma: relationship with p53 and the MDM2protein expression. Am. J. Pathol. 1998; 153: 1749–1765.

20 Kashiwabara K, Oyama T, Sano T, Fukuda T, Nakajima T. Cor-relation between methylation status of the p16/CDKN2 geneand the expression of p16 and Rb proteins in primary non-small-cell lung cancers. Int. J. Cancer 1998; 79: 215–220.

21 Saito T, Nakajima T, Mogi K. Immunohistochemical analysis ofcell cycle-associated proteins p16, pRb, p53, p27 and Ki-67 inoral cancer and precancer with special reference to verrucouscarcinomas. J. Oral Pathol. Med. 1999; 28: 226–232.

22 Kato H, Yoshikawa M, Fukai Y et al. An immunohistochemicalstudy of p16, pRb, p21 and p53 proteins in human esophagealcancers. Anticancer Res. 2000; 20: 345–349.

23 Sano T, Hikino T, Xue Q et al. Immunohistochemical inactiva-tion of p14ARF concomitant with MDM2 overexpressioninversely correlated with p53 overexpression in oral squamouscell carcinoma. Pathol. Int. 2000; 50: 709–716.

24 Xue Q, Sano T, Kashiwabara K, Oyama T, Nakajima T. AberrantExpression of pRb, p16, p14ARF, MDM2, p21 and p53 in squa-mous cell carcinomas of Lung. Jpn. J. Cancer Res. 2001; 92: 1–8.

25 Burns KL, Ueki K, Jhung SL, Koh J, Louis DN. Molecular geneticcorrelates of p16, cdk4 and pRb immunohistochemistry in glio-blastomas. J. Neuropathol. Exp. Neurol. 1998; 57: 122–130.

26 Sano T, Oyama T, Kashiwabara K, Fukuda T, Nakajima T.Expression status of p16 protein is associated with human pap-illomavirus oncogenic potential in cervical and genital lesions.Am. J. Pathol. 1998; 153: 1741–1748.

27 Shapiro GI, Edwards CD, Kobzik L et al. Reciprocal Rb inacti-vation and p16INK4 expression in primary lung cancers and celllines. Cancer Res. 1995; 55: 505–509.

28 Taga S, Osaki T, Ohgami A et al. Prognostic value of the immu-nohistochemical detection of p16INK4 expression in nonsmallcell lung carcinoma. Cancer 1997; 80: 389–395.

29 Sakaguchi M, Fujii Y, Hirabayashi H et al. Inversely correlatedexpression of p16 and Rb protein in non-small cell lung can-

pRb and p53 pathways in lung cancers 109

cers: an immunohistochemical study. Int. J. Cancer 1996; 65:442–445.

30 Geradts J, Fong KM, Zimmerman PV, Maynard R, Minna JD.Correlation of abnormal RB, p16ink4a, and p53 expression with3p loss of heterozygosity, other genetic abnormalities, and clin-ical features in 103 primary non-small cell lung cancers. Clin.Cancer Res. 1999; 5: 791–800.

31 Yuan J, Knorr J, Altmannsberger M et al. Expression of p16 andlack of pRB in primary small cell lung cancer. J. Pathol. 1999;189: 358–362.

32 Weber JD, Kuo ML, Bothner B, et al. Cooperative signals gov-erning ARF–mdm2 interaction and nucleolar localization of thecomplex. Mol. Cell. Biol. 2000; 20: 2517–2528.

33 Vousden KH. p53. Death star. Cell 2000; 103: 691–694.34 Zhang Y, Xiong Y, Yarbrough WG. ARF promotes MDM2 deg-

radation and stabilizes p53: ARF-INK4a locus deletion impairsboth the Rb and p53 tumor suppression pathway. Cell 1998; 92:725–734.

35 Weber JD, Taylor LJ, Roussel MF, Sherr CJ, Bar-Sagi D. Nucle-olar Arf sequesters Mdm2 and activates p53. Nat. Cell. Biol.1999; 1: 20–26.

36 Robertson KD, Jones PA. The human ARF cell cycle regulatorygene promoter is a CpG island which can be silenced by DNAmethylation and down-regulated by wild-type p53. Mol. Cell.Biol. 1998; 18: 6457–6473.

37 Vonlanthen S, Heighway J, Tschan MP et al. Expression ofp16INK4a/p16 and p19ARF/p16 is frequently altered in non-small cell lung cancer and correlates with p53 overexpression.Oncogene 1998; 17: 2779–2785.

38 Higashiyama M, Doi O, Kodama K et al. MDM2 gene amplifi-cation and expression in non-small-cell lung cancer: immuno-histochemical expression of its protein is a favourableprognostic marker in patients without p53 protein accumulation.Br. J. Cancer 1997; 75: 1302–1308.

39 Esteller M, Tortola S, Toyota M et al. Hypermethylation-associ-ated inactivation of p14ARF is independent of p16INK4A methy-lation and p53 mutational status. Cancer Res. 2000; 60: 129–133.

40 Iida S, Akiyama Y, Nakajima T et al. Alteration and hyperme-thylation of the p14ARF gene in gastric cancer. Int. J. Cancer2000; 87: 654–658.

41 Newcomb E, Alonso M, Sung T, Miller DC. Incidence of p14ARF

gene deletion in high-grade adult and pediatric astrocytomas.Hum. Pathol. 2000; 31: 115–119.

42 Sanchez-Cespedes M, Reed AL, Buta M et al. Inactivation ofthe INK4A/ARF locus frequently coexists with TP53 mutationsin non-small cell lung cancer. Oncogene 1999; 18: 5843–5849.

43 Ko JL, Cheng YW, Chang SL, Su JM, Chen C, Lee H. MDM2mRNA expression is a favorable prognostic factor in non-small-cell lung cancer. Int. J. Cancer 2000; 89: 265–270.

44 Guang SG, Ogura T, Sekine T et al. Association between p53mutation and clinicopathological features of non-small cell lungcancer. Jpn. J. Clin. Oncol. 1997; 27: 211–215.

45 Fukuyama Y, Mitsudomi T, Sugio K, Ishida T, Akazawa K,Sugimachi K. K-ras and p53 mutations are an independentunfavourable prognostic indicator in patients with non-small-celllung cancer. Br. J. Cancer 1997; 75: 1125–1130.