ICANCERRESEARCH56,5557—5562,December15, 19961 Advances in Brief

Altered @16INK4andRetinoblastoma Status in Non..Small Cell Lung : Potential Synergistic Effect with Altered Protein on Proliferative Activity

Ichiro Kinoshita, Hirotoshi Dosaka.Akita,' Takayuki Mishina, Kenji Aide, Motoi Nishi, Hiromitsu Hiroumi, Fumihiro Hommura, and Yoshikazu Kawakami

First Department of Medicine, Hokkaido University School of Medicine Ii. K., H. D-A.. T. M., K. A., H. H., F. H., Y. K.J, and Department of Public Health, Sapporo Medical University School of Medicine [M. N.], Sapporo 060. Japan

Abstract (4, 6), which is supported by observation of an inverse correlation between alteration of both in several types of tumors (7—13). @―4protein (pl6) and (pith), like p53 pro In NSCLCs, whereas the RB is shown to be inactivated in Win, are Important tumor suppressors that regulate the . We 6—32%of the tumors (14—18),the frequency of alteration involving immunohistochemically examined fresh-frozen specimens of 114 resected non-small cell lung (NSCLCs) for loss of p16 and pRB expression, the CDKN2 gene is controversial. Of NSCLC cell lines, homozygous together with aberrant accumulation of p53 proteIn and the proliferative of the CDKN2 gene is observed in 20—50% and point activity determined by the Ki-67 Index. Three pith-positive tumors were mutation in 6—21% (2, 3, 10—12, 19—22). Recent reports showed uninterpretable for p16 status. Of the remaining 111 tumors, 30 (27%) DNA hypermethylation with CDKN2 gene silencing in about 30% of lacked p16 expression, and 10 (9%) lost pith expression. No tumors NSCLC cell lines (12). For primary NSCLCs, although there is a showed coincident loss of both proteins, supporting the hypothesis that study reporting 30% of intragenic mutations of the CDKN2 gene (23), they function In a single pathway. Of 25 tumors, including 4 p16-negative most studies showed that the abnormalities of the CDKN2 gene are tumors, examined by Southern blot analysis, only 2 p16-negative tumors rare ( 10, 21, 22, 24) and are largely restricted to metastatic sites (21, were considered to have reduced gene dosage consistent with possible 22). Nevertheless, the rare occurrence of CDKN2 alterations in pri homozygous deletion of the CDKN2 gene encoding pi6, suggesting that mary NSCLCs may be underestimated, not only because epigenetic immunohistochemistry is a sensitive and suitable method to screen for p16 alteration. Loss of p16 expression did not correlate with any clinical alterations, including DNA hypermethylation, has not been investi factors or p53 status, whereas loss of pRB expression correlated with gated thus far, but also because homozygous deletion may be ob heavy smoking (P 0.03 by Fisher's exact test and P = 0.01 by the scured by contamination due to normal stromal cells (13). Moreover, multivariate logistic regression analysis). Proliferative activity was con in NSCLCs, little is known about the clinical and cell biological siderably higher in p53-positive tumors than in p53-negative tumors implications of the inactivation of the pl6/pRB pathway and its (P < 0.001). Loss of p16 or pith expression was associated with a further association with the status of p53 protein (25), another tumor sup increase in proliferative activity In the p53-positive tumors (P 0.009) but pressor working at the G@checkpoint (26). not with proliferative activity in the p53-negativetumors. These results Therefore, in the present study, we immunohistochemically exam suggest that alteration of the pl6IpRB pathway is relatively frequentiy med primary NSCLCs for the loss of p16 and pRB expression, involved in the development and progression of NSCLCs and that its together with aberrant accumulation of p53 protein and the prolifer effect on the proliferative activity is potentially synergistic with altered p53 proteIn. ative activity determined by Ki-67 index. We show a relatively frequent loss of p16 in both early and late stages of NSCLCs. The Introduction reciprocal loss of p16 and pRB provides further support for the hypothesis that both proteins function in a single pathway. Further The CDKN2 and the RB2tumor suppressor play a pivotal role more, we demonstrate the potential synergistic effect of the altered in the regulation of the cell cycle at the G1 checkpoint. In normal cells, p16/pRB pathway with altered p53 protein on the proliferative active hypophosphorylated pRB prevents cells from transition across activity. the G1 checkpoint by sequestering factors such as , which is essential for DNA replication, but inactivation of pRB by Materials and Methods CDK-mediated phosphorylation allows cells to progress to the through the G1 checkpoint (1). pl61@4 protein (p16), the Tissue Specimens. Primary tumor specimens from 114 NSCLCs were product of the CDKN2 gene, which has been recently found to be obtained by surgery from the Hokkaido University Medical Hospital, Sapporo deleted in a variety of tumor cells (2, 3), inhibits CDK4- and CDK6- Minami-Ichijo Hospital, and National Sapporo Minami Hospital during 1990 and 1995. Tumors were snap-frozen in liquid nitrogen and stored at —80CCin mediated phosphorylation of pRB (4, 5). Therefore, p16 and pRB are OCT compound (Miles, Elkhart, IN) for immunohistochemistry. For 25 spec suggested to function in a single regulatory pathway of the cell cycle imens, portions of the tumor were also snap-frozen and stored at —80°Cfor genomic DNA preparation. Adjacent non-frozen blocks were fixed in 10% Received 7/15/96; accepted 10/29/96. neutral buffered formalin and embedded in paraffin, and then several sections The costs of publication of this article were defrayed in part by the payment of page from each specimen were stained with H&E to observe histopathology and charges. This article must therefore be hereby marked advertisement in accordance with normal contamination. According to the 1981 WHO classification, they were 18 U.S.C. Section 1734 solely to indicate this fact. I To whom requests for reprints should be addressed, at First Department of Medicine, histopathologically diagnosed as adenocarcinoma (n 74), squamous cell Hokkaido University School ofMedicine, North 15,West 7, Kita-ku, Sapporo 060, Japan. carcinoma (n = 34), and large cell carcinoma (n 6). Phone:8l-(ll).7l6-ll6l,ext.5911;Fax:81-(ll)-727-6724. Immunohistochemistry for p16, pRB, and p53 Proteins. In a pilot study 2 The abbreviations used are: RB, retinoblastoma; pRB, retinoblastoma protein; to choose the mmmunostainingprotocol for p16 protein on frozen materials, CDK, -dependent kinase; p16. @16@'@K4proteinproduct of the CDKN2/MTSI gene; NSCLC, non-small cell lung cancer; GAPDH, glyceraldehyde-3-phosphate we used cytospinned p16-positive and -negative cell lines, NCI-H209 and dehydrogenase. HL6O,respectively (8), and tested three rabbit polyclonal anti-pl6 primary 5557

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1996 American Association for Cancer Research. p16 AND pRB IN NSCLCs

resectedNSCLCsCharacteristicsp16expressionaPpRB Table 1plo and pRB expressionin 114 expression+—+—ppRB

expression+

00.04Age—71 1030 8.50.29SexMale60210.677670.54Female219283Smoking(mean ±SD)62.7 ±9.764.6 ±9.60.3662.9 ±9.766.3 ±

years)―0—192390.853400.03205620689Histological(pack

typeAdeno 53190.076680.52Squamous267322Large2460pT

statusT120100.362830.72T,@6120767pN

status'N048210.376840.09N,_3319346pM

statusM075280.8696100.46M16280pStageI—LI48220.176850.26III—IV338365

p53 status Normal 46 19 0.53 6430.06407 Altered —@--______a Three pRB+ tumors could not be evaluated for p16 status (see text). b Smoking history data were not available for three cases.

C Lymph nodes were not resected for two cases.

antibodies, one against the entire human p16 protein (PharMingen, San Southern Blot Analysis. Genomic DNA was extracted from 25 NSCLC Diego, CA), and the other two against the NH2 terminus or COOH terminus tumor specimens, Southern blotted after digestion with EcoRI, and hybrid of human p16 proteins (Santa Cruz Biotechnology, Santa Cruz, CA), in ized to [a-32P]dCTP-labeled human CDKN2 cDNA (7, 10) and cardiac various concentrations, and three fixatives: methanol, acetone, and 10% actin cDNA. The final wash was performed twice in 0.2X SSC, 0.1% SDS buffered formalin. The protocol using anti-entire-human p16 antibody at 50°Cfor 20 mm. The blots were then exposed to Kodak XAR-5 film (PharMingen) as a primary antibody and methanol as a fixative yielded the (Kodak, Rochester, NY) and quantitatively analyzed by a densitometer. most nucleus-specific staining with the highest intensity in NCI-H209 cells, The autoradiographs were developed at various times to ensure that this whereas all of the combinations of antibodies and fixatives generated analysis was performed in the linear range of the film. The CDKN2 proper nuclear staining and weak cytoplasmic staining in the NCI-H209 signal intensity was normalized against the cardiac actin signal intensity. cells. For additional experiments, therefore, 5-p@m frozen sections of Because normal cells were present in tumor specimens, we considered a NSCLC specimens were fixed in methanol at 4°Cfor 5 mm and were tumor to have reduced gene dosage consistent with possible homozygous reacted with the anti-entire-human p16 primary antibody (PharMingen) at deletion of the CDKN2 gene in the tumor cells when the normalized a I:200 dilution or with control rabbit immunoglobulin at 4°Covernight. CDKN2 signal intensity in the tumor sample was reduced to less than 80% Immunostaining was performed by the biotin-streptavidin immunoperoxi of that in the normal lung control, and the reduced rate was similar to the dase method with 3,3'-diaminobenzidine as a chromogen (SAB-PO ; proportion of the tumor cells in the sample, essentially as described Nichirei, Tokyo, Japan). Methyl green was used for counterstain. Immu previously (10). Additional Southern blot analysis for the GAPDH gene nohistochemical analyses to detect normal expression of pRB and aberrant was performed to confirm the appropriateness of the cardiac actin gene as accumulation of p53 protein using frozen tissue sections were performed an internal control. with mouse monoclonal primary antibodies PMG3-245 (PharMingen) and Ki-67 Immunostaining and Index Scoring. Frozen tissue sections were PAbl8Ol ( Science, Manhasset, NY), respectively, as described mmmunohistochemically analyzed using Ki-67 monoclonal antibody (Dako, previously (15, 27). Tokyo, Japan) as described previously (29). Cells showing any nuclear staining Tumors were scored as p16-negative (p16—)if all malignant cells had no were considered to be Ki-67 positive. The Ki-67 index (percentage of Ki-67- nuclear staining, whereas surrounding normal stromal cells showed adequate positive cells) was determined by scoring at least 500 tumor cells in three or nuclear staining as a positive internal control (13). Tumors were regarded as more well-preserved areas at X400. Sections with fewer than 500 tumor cells p16-positive (p16+) if any malignant cells had nuclear staining. Small lym were excluded. phocytes, which showed no nuclear staining of p16, were used as a negative Statistical Analysis. The associations between loss of p16 or pRB and internal control (9, 28). These criteria for the evaluation of p16 staining were categorical variables were analyzed by the x@ test or Fisher's exact test as the same as described previously for pRB staining (14, 16), which were also appropriate (30). The associations between loss of p16 or pRB and age were used to classify pRB-positive (pRB+) and -negative (pRB—) tumors in the analyzed by Student's t test. To examine the effect of more than one factor present study. Tumors were considered to be p53-positive (p53+) and thus to simultaneously on loss of p16 or pRB, the multivariate logistic regression contain the putative p.53 mutations when more than 10% of tumor cells showed analysis was used (31). The differences of Ki-67 indices between two groups nuclear staining, and the others were scored as p53-negative (p53—),as were determined by Mann-Whitney's U test. Multiple comparisons of Ki-67 described previously (27). indices among groups divided by p53 and pl6/pRB status were performed by 5558

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1996 American Association for Cancer Research. @ a @d@b ?@

p16 AND pRB IN NSCLCs

the ANOVA factorial with Fisher's protected least significant difference. The with clinical and clinicopathological characteristics (Tables 1 and 2). significance level chosen was P < 0.05, and all tests were two-sided. All Loss ofpl6 expression did not correlate with any factors, whereas loss immunohistochemical studies were done without knowledge of the clinical of pRB expression correlated with heavy smoking (pack years 20; data. P = 0.03 by Fisher's exact test and P = 0.01 by the multivariate Results logistic regression analysis). Aberrant accumulation of p53 protein, observed in 47 (41%) tumors ofthe 114 tumors, did not correlate with To determine the loss of p16 and pRB expression in tumor cells the loss of pl6 or pRB (Table 1). with the surrounding stromal cells as a normal internal control in Ki-67 antibody recognizes a nuclear non-histone protein expressed NSCLCs, we subjected fresh-frozen tissues from 114 resected throughout the cell cycle but absent in G0 (32). Therefore, the pro NSCLCs to immunohistochemistry for these proteins. Three pRB + portion of Ki-67-labeled tumor cells (Ki-67 index) provides a measure tumors were uninterpretable for p16 expression due to lack of of the proliferative activity. To elucidate the cell biological signifi nuclear staining of p16 in all tumor and stromal cells. Of the cance of alterations of p16 and pRB in combination with p53, we remaining 111 tumors, 30 (27%) lacked p16 expression (p 16—), determined the Ki-67 index in the same tumor specimens as analyzed and 10 (9%) lost pRB expression (pRB—; Table 1). No tumors for p16, pRB, and p53 proteins. The results are summarized in showed coincident loss of p16 and pRB expression. The inverse Table 3. The Ki-67 indices were assessed in 105 cases and ranged correlation between loss of both proteins was statistically signifi from 1 to 92% (median, 13.4%; mean, 20.9%). Increased values of cant (P = 0.04). Typical p16—/pRB+ and p16+/pRB— tumors are Ki-67 indices strongly correlated with the altered expression of shown in Fig. 1. All pRB —tumors had strong and homogeneous p53 protein (P < 0.001) but did not correlate with p16 or pRB nuclear staining of p16 in most tumor cells (Fig. 1A), although 23 status. Because p16 and pRB are suggested to function in a single of 7 1 pRB +/pl6 + tumors showed weak and heterogeneous nu regulatory pathway of the cell cycle (4, 6), we divided tumors into clear p!6 staining. two groups: the tumors retaining both p16 and pRB (p16+ and To investigate to what extent the loss of p16 protein reflected the pRB+; pl6/pRB-normal) and the tumors lacking either p16 or homozygous deletion of the CDKN2 gene, we examined 25 tumors, pRB (p16— or pRB—; p16/pRB-altered). The difference in Ki-67 including 4 pl6— tumors, by Southern blot analysis. Two of the 4 indices between the two groups of tumors was not statistically p16-tumors were considered to have reduced gene dosage consistent significant. When p53+ and p53— tumors were analyzed sepa with possible homozygous deletion of the CDKN2 gene (samples 1 rately, however, loss of p16 or pRB expression correlated with and 5; Fig. 2), whereas none of the 21 p16+ tumors showed such higher Ki-67 indices in the p53+ tumors (P = 0.009) but not in the reduced gene dosage. p53—tumors. This correlation was also shown to be significant in After completion of immunohistochemical analysis of p16 and a multiple comparison analysis; that is, p53+ and pl6/pRB-altered pR.B, status of these proteins was statistically analyzed to correlate it tumors had significantly higher Ki-67 indices than p53+ and

@ ,. ii@@

@ , ‘.i'@.rI%•@ @ @s' @@ # @4 4 ‘a •@* , ,‘@. @ ‘@% ‘ .‘@ @ I . 1tij@t@@ @‘ ,@ @ ‘@ •:, ‘,.._@ 4 @. \ 1@4@\. _@ ,dI, ‘4.'. ,‘@

t. (., — -& @ , 1' I ;‘U “p •‘ • @ ,4@ -. @ ‘. I •& @;@@‘1i . .@‘.m', I @ I t , øø;,o ..V £ @ ‘ m, .@ @ •# 4.,'@,C@' OISY @1 .@- .@ .@ % @ .@uuIJ:@t@• ,,,,,. b @• .a I . . 0 kb,. @‘ . a— • . 0 @- pap, @# @.. , . ‘ ‘ S I • . . 0,., f.@, ‘@r'(s @ .,

@ . flb• .,. P@ .;g, ;; ,@. — l',,• ( .@ @ ‘@.fj;4i@, @ . ..-@ “% ,@,.

@ .“.p@.. @1o@,:― I . 7 .-... .. ) .@ 4j'@ ., . ., , I p@.,i, . ,1 @)

@ ‘a •,“ a• @, J, . ‘, ., I @S ‘ ‘ I@' ‘.‘ .@ ,%-@., ‘I tl @t. S a ‘ @ - .@. rn ‘•. “7@:‘t:@' C @- i@ ti.'@ Fig. 1. Immunostaining patterns for p16 (A and C) and pRB (B and D) in primary NSCLCs. Representative p16+/RB— shows strong and homogeneous nuclear staining of p16 (A) and no nuclear staining of pRB (B) in tumor cells. Representative pl6—IRB+ squamous cell carcinoma displays no nuclear staining of p16 (C) and intense nuclear staining of pRB (D) in tumor cells. Note that admixed stromal cells show distinct nuclear staining (arrows) of p16 and pIth in all specimens, which provide positive internal controls for both proteins. A—D,X400. 5559

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1996 American Association for Cancer Research. p16 AND pRE IN NSCLCs

Ni 1 2 3 4 N25 6 78 @ CDKN2 p. i@ @uI.@ @1@

@ C'Actin ,@. @. @* @,

Relative Dosage 0.32 0.75 1.19 0.96 059 0.83 1.10 0.83 p16IHC — — + + — + +

Fig. 2. Southern blot analysis of the CDKN2 gene with EcoRI-digested genomic DNA from tumor samples (Lanes 1—8)and normal lung tissues (Lanes NI and N2). The left (Lanes NJ and 1—4)andthe right (Lanes N2 and 5—8)examples are derived from two separate radiography sheets. Alto shown is the internal control cardiac actin (C-Actin) band to which the CDKN2 signal intensity was normalized for densitometer analysis. Relative CDKN2 Dosage, the normalized CDKN2 signal intensity in the tumor samples relative to that in the normal lung control (= 1.00). pitS IHC, the pretence (+) or absence (—)of p16 protein expression determined by immunohistochemistry. The relative CDKN2 dosage in sample 1 it reduced to 32%; tumor and normal stromal cells comprise about 70 and 30% ofthis sample, respectively, consistent with possible homozygous deletion ofthe CDKN2 gene in the tumor cells. The relative dosage in sample 5 is reduced to 59%; tumor and normal stromal cells comprise about 40 and 60% of this sample, respectively, also consistent with possible homozygous deletion. The relative dosage in sample 2 is 75%; tumor and normal stromal cells comprise about 60 and 40% of this sample, respectively. Therefore, we did not assess the tumor of the sample as having reduced gene dosage consistent with possible homozygous deletion, although the sample might have hemizygous deletion. Several sections from each specimen surrounding the portion of the tumor used for Southern blot analysis were reviewed, and the percentage of normal contamination was shown to be approximately consistent. Additional Southern blot analysis of the GAPDH gene was performed to evaluate the appropriateness of the cardiac actin gene as an internal control, and it was ensured.

p1 6/pRB-normal tumors (P = 0.02) as well as p53 —and pl6/pRB way (4, 6). The invariably strong nuclear p16 staining ofpRB— tumor normal tumors (P < 0.001) and p53—and p16/pRB-altered tumors cells also supported a negative feedback model between p16 and pRB, (P < 0.001). i.e., pRB inactivation leading to increased p16 expression (4, 33). Furthermore, we demonstrated, for the first time, the association of Discussion loss of p!6 or pRB expression with increased proliferative activity in The present study showed a relatively frequent loss of p16 expres p53+ tumors of NSCLCs. sion in primary NSCLCs. The reciprocal loss of p16 and pRB, By immunohistochemical analysis, we could determine the loss of p16 consistent with previous studies (7—13), provided further support for expression in tumor cells, using the surrounding stromal cells retaining the plausible hypothesis that both proteins function in a single path p16 expression as a normal internal control. Thus, this analysis may more sensitively detect the inactivation of the CDKN2 gene than molecular

Table 2 Multivariate logistic regression analysis for the correlation between clinical genetic analyses using homogenates of tumor specimens for three major characteristics and loss ofpl6 or pRB reasons: (a) homozygous deletion, the principal mechanism of CDKN2 of pl6 of pRB inactivation, may be obscured by contamination by stromal cells when genetic techniques, especially those based on PCR, are used (13); (b) ratioPAge103b0.28102b0.18Sex1.430.632.170.09Smoking1Characteristics―Loss Odds ratioPLots Odds epigenetic changes such as DNA hypermethylation, another common mechanism of CDKN2 inactivation (12), may not be detected by DNA .040.933.950.01Histology1.750.160.610.16pT0.730.540.720.50pN0.740.552.300.06pMI based analysis (13); and (c) both homozygous deletion and DNA hyper methylation of the CDKN2 gene result in loss of p16 expression. As expected, the present immunohistochemicalstudy showed loss of p16 in .090.920.540.25 27% (30 of 111) of primary NSCLCs, although previous genetic studies a Selected from Table 1. detected the abnormality of the CDKN2 gene in 0—7%of such tumors b For an increase of 1 year in age.

statusnMedianMeanRangeSDPp16+ Table3 Ki-67 indices in NSCLCgroups divided by pitS,pRB, and p53

27.30.57pRB+p16—75 3013.4 14.519.1 25.41.2—82.8 1.0—92.018.9 95 pRB18.30.50p16+ 1012.6 15.920.8 22.11.0—92.0 1.6—57.222.1 and pRB+― p16—or25.10.35PS3— pRB@―65 4012.6 15.918.6 24.61.2—82.8 1.0—92.019.0

p53+6222.6<0.001when 438.4 19.015.7 28.41.0—78.8 3.0—92.019.5 p53— p16+ and pRB+ 40 p16—orpRB 228.6 7.016.3 14.61.2—74.0 1.0—78.818.6 21.40.53 when p53+ p16+ and pRB+ 2515.222.43.0—82.819.50.0091826.436.812.0—92.024.4 p16—or pRB

a Tumors retaining both p16 and pRB. b Tumors lacking either p16 or pRB. 5560

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1996 American Association for Cancer Research. p16 AND pRB IN NSCLCs

(10, 21, 22, 24), except for a study reporting 30% of intragenic mutations with altered p53 protein. The present finding was consistent with (23). experimental observations that deregulated growth can be inhibited by Southern blot analysis in this study suggested reduced gene dosage the introduction of either the RB (38) or p53 gene (39) in SAOS-2 consistent with possible homozygous deletion of the CDKN2 gene in cells, devoid of normal pRB and p53 protein. two of four p16— tumors analyzed, which agrees with the observa In conclusion, the present study suggests that alteration of the tions that one-half of NSCLC cell lines lacking p16 expression show pl6/pRB pathway is relatively frequently involved in the development homozygous deletion of the CDKN2 gene and that most of the rest and progression of NSCLCs and that its effect on the proliferative sustain DNA hypermethylation (12). It is also true that Southern blot activity may be synergistic with altered p53 protein. In the future, it analysis may have lower sensitivity and specificity for detecting will be necessary to determine whether the status of these proteins can homozygous deletions in tumor specimens due to contaminating nor be a for the clinical outcome and a target of therapy in mal cells and aneuploidy of tumor cells (13). Again, we emphasize NSCLCs. Alteration of both pl6/pRB and p53 pathways may be that the immunohistochemical detection of p16 is a sensitive and linked to poor prognosis because of its association with extremely suitable method to screen for p16 alterations resulting from both high proliferative activity (29, 36, 37). homozygous deletion and DNA hypermethylation. The frequency of p16-negative cases in this study was lower than Acknowledgments those in two recent immunohistochemical studies, reported as 47% (8 The authors thank Dr. D. Beach of Cold Spring Harbor Laboratory, Cold of 17; Ref. 13) and 49% (30 of 61 ; Ref. 34) of NSCLCs with the same Spring Harbor, NY, for providing human CDKN2cDNA, and Dr. M. Sakai of method of scoring p16 expression. This discrepancy may be attributed First Department of Biochemistry, Hokkaido University School of Medicine, to the difference of fixation and preservation of materials; we used for providing eDNA of the human GAPDH gene. We also thank Prof. H. Kato fresh rapidly frozen tissues and fixed them in methanol, although the of the Second Department of Surgery, Hokkaido University School of Mcdi two previous studies used archival formalin-fixed paraffin-embedded cine, Dr. T. Hirata of the National Sapporo Minami Hospital, and Dr. T. tissues. p!6 is considered to be highly susceptible to damage and to Morikawa of the Sapporo Minami-ichijo Hospital for providing surgical spec degenerate, depending on the condition of the fixation and/or preser imens analyzed in this study. vation of archival tissues (13). Therefore, some of our 23 (2 1%) tumors with heterogeneous and weak staining of p16 might be eval References uated false-negatively for the p16 expression should formalin-fixed I. Weinberg, R. A. The retinoblastoma protein and cell cycle control. Cell, 81: 323—330, paraffin-embedded tissues be used. Moreover, one of the two immu 1995. 2. Kamb, A., Gruis, N. A., Weaver-Feldhaus, J., Liu, Q., Harshman, K., Tavtigian, S. V., nohistochemical studies did not mention the nuclear staining of nor Stockert, E., Day, R. S., III, Johnson, B. E., and Skolnick, M. H. A cell cycle regulator mal stromal cells as a positive internal control in p16-negative tumor potentially involved in genesis of many tumor types. Science (Washington DC), 264: 436—440, 1994. specimens (34). 3. Nobon, T., Miura, K., Wu, D. J., Lois, A., Takabayashi, K., and Carson, D. A. Our observation of the loss of p16 expression in the early Deletions of the cyclin-dependent kinase-4 inhibitor gene in multiple human cancers. clinical stage including T1/N0 tumors is intriguing. Nakagawa et Nature (Land.), 368: 753—756,1994. 4. Serrano, M., Hannon, G. J., and Beach, D. A new regulatory motif in cell-cycle a!. (22) showed that all homozygous deletions and point mutations control causing specific inhibition of cycin D/CDK4. Nature (Land.), 366: 704—707, of the CDKN2 gene are observed in NSCLC cell lines or tumors 1993. from patients with advanced stage disease. Okamoto et a!. (21) 5. Hannon, G. J., and Beach, D. p15INK4B is a potential effector of TGF-@3-induced cell cycle arrest. Nature (Lund.), 371: 257—261,1994. reported that homozygous deletion and insertions are detected only 6. Lukas, J., Parry, D., Aagaard, L., Mann, D. J., Bartkova, J., Strauss, M., Peters, G., in metastatic NSCLC tumors. These genetic studies suggest that and Bartek, J. Retinoblastoma protein-dependent cell-cycle inhibition by the tumour suppressor p16. Nature (Land.), 375: 503—506,1995. structural alterations of the CDKN2 gene are associated with 7. Okamoto, A., Demetrick, D. J., Spillare, E. A., Hagiwara, K., Hussain, S. P., Bennett, progression of NSCLCs. Two immunohistochemical studies (34), W. P., Forrester, K., Gerwin, B., Serrano, M., Beach, D. H., and Harris, C. C. including the present study, however, showed no associations Mutations and altered expression of pl6INK4 in human cancer. Proc. Nail. Acad. Sci. USA, 91: 11045—11049, 1994. between loss of p16 and clinical stage in primary NSCLCs, sug 8. Otterson, G. A., Kratzke, R. A., Coxon, A., Kim, Y. W., and Kaye, F. J. Absence of gesting that the loss of p16 may be a relatively early event in the p16IN@@protein is restricted to the subset of lung cancer lines that retains wildtype development and progression of a subset of NSCLCs. A possible RB. Oncogene, 9: 3375—3378,1994. 9. Tam, S. W., Shay, J. W., and Pagano, M. Differential expression and cell cycle explanation for this discrepancy between the genetic and immuno regulation of the cyclin-dependent kinase 4 inhibitor pl&―4.Cancer Ret., 54: histochemical studies is that the loss of p16 in the early stage of 5816—5820, 1994. 10. Shapiro, G. I., Edwards, C. D., Kobzik, L., Godleski, J., Richards, W., Sugarbaker, primary NSCLC results primarily from epigenetic changes such as D. J., and Rollins, B. J. Reciprocal Rb inactivation and p16'@4 expression in primary DNA hypermethylation (35). This hypothesis is supported by the lung cancers and cell lines. Cancer Ret., 55: 505—509,1995. observation that transcriptional inactivation without genetic alter 11. Kelley, M. J., Nakagawa, K., Steinberg, S. M., Mulshine, J. L., Kamb, A., and Johnson, B. E. Differential inactivation of CDKN2 and Rb protein in non-small-cell ation occurs in NSCLC cell lines, even from patients with the early and small-cell lung cancer cell lines. J. Nail. Cancer Inst., 87: 756—761, 1995. stage (22). 12. Otterson, G. A., Khleif, S. N., Chen, W., Coxon, A. B., and Kaye, F. J. CDKN2gene We demonstrated the possible correlation of the loss of pRB with silencing in lung cancer by DNA hypermethylation and kinetics of p16'@4 protein induction by 5-aza 2'deoxycytidine. Oncogene, 11: 1211—1216,1995. heavy smoking in NSCLC, although previous studies have not shown 13. Geradts, J., Kratzke, R. A., Niehans, G. A., and Lincoln, C. E. Immunohistochemical such correlation (15, 17). The clinical significance of pRB alteration detection of the cyclin-dependent kinase inhibitor 2/multiple I (CDKN2JMTSJ) product p16@4A in archival human solid tumors: correlation with remains to be determined in larger studies of NSCLC. retinoblastoma protein expression. Cancer Ret., 55: 6006—6011,1995. This is the first study demonstrating the association of proliferative 14. Xu, H-J., Hu, S-X., Cagle, P. T., Moore, 0. E., and Benedict, W. F. Absence of activity with alterations of p16 and pRB in combination with p53 retinoblastoma protein expression in primary non-small cell lung carcinomas. Cancer Ret., 51: 2735—2739,1991. protein status by measuring the Ki-67 index in primary NSCLCs. The 15. Reissmann, P. T., Koga, H., Takahashi, R., Figlin, R. A., Holmes, E. C., Piantadosi, considerably higher proliferative activity in p53+ tumors than in S., Cordon-Cardo, C., Slamon, D. J., and the Lung Cancer Study Group. Inactivation PS3—tumors was consistent with previous studies (36, 37). Loss of of the retinoblastoma susceptibility gene in non-small-cell lung cancer. Oncogene, 8: 1913—1919,1993. p16 or pRB expression was associated with a further increase in the 16. Xu, H-J., Quinlan, D. C., Davidson, A. G., Hu, S-X., Summers, C. L., Li, J., and proliferative activity in the p53 + tumors, but there was no significant Benedict, W. F. Altered retinoblastoma protein expression and prognosis in early stage non-small-cell lung carcinoma. J. Nail. Cancer Inst., 86: 695—699, 1994. change in the p53— tumors, suggesting that derangement of the 17. Shimizu, E., Coxon, A., Otterson, G. A., Steinberg, S. M., Kratzke, R. A., Kim, pl6/pRB pathway might synergistically increase cell proliferation Y. W., Fedorko, J., Oie, H., Johnson, B. E., Mulshine, J. L., Minna, J. D., Gazdar, 5561

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1996 American Association for Cancer Research. p16 AND pRB IN NSCLCs

A. F., and Kaye, F. J. RB protein status and clinical correlation from 171 cell lines J. M., Bogdany, J. K., Shiel, M. J., Haluska, F. G., Fountain, J. W., and Albino, A. P. representing lung cancer. extrapulmonary small cell carcinoma, and . Loss of expression of the p16/cyclin-dependent kinase inhibitor 2 tumor suppressor Oncogene, 9: 2441—2448,1994. gene in melanocytic lesions correlates with invasive stage of tumor progression. 18. Higashiyama, M., Doi, 0., Kodama, K., Yokouchi, H., and Tateishi, R. Retinoblas Cancer Ret., 55: 2713—2718, 1995. toma protein expression in lung cancer: an immunohistochemical analysis. Oncology. 29. Pence, J. C., Kerns, B. J., Dodge, R. K., and Iglehart, J. D. Prognostic significance of 51:544—551,1994. the proliferation index in surgically resected non-small-cell lung cancer. Arch. Surg., 19. Liu, Q., Neuhausen, S., McClure, M., Frye, C., Weaver-Feldhaus, J., Gruis, N. A., 128: 1382—1390,1993. Eddington, K., Allalunis-Turner, M. J., Skolnick, M. H., Fujimura, F. K., and Kamb, 30. Mehta, C. R., and PaId, N. R. A network algorithm for performing Fisher's exact test A. CDKN2 (MTSJ) tumor suppressor gene mutations in human tumor cell lines. in r X c contingency tables. J. Am. Stat. Assoc., 78: 427—434, 1983. Oncogene, 10: 1061—1067,1995. 31. Cox, D., and Snell, E. Analysis of Binary Data, Ed. 2. London: Chapman and Hall, 20. Washimi, 0., Nagatake, M., Osada, H., Ueda, R., Koshikawa, T., Seki, T., Takahashi, 1989. T., and Takahashi, T. In vivo occurrence of p16 (MTS 1) and p15 (MTS2) alterations 32. Genies, J., Lemke, H., Baisch, H., Wacker, H-H., Schwab, U., and Stein, H. Cell preferentially in non-small cell lung cancers. Cancer Ret., 55: 514—517, 1995. cycle analysis of a cell proliferation-associated human nuclear antigen defined by the 21. Okamoto, A., Hussain, S. P., Hagiwara, K., Spillare, E. A., Rusin, M. R., Demetrick, monoclonal antibody Ki-67. J. Immunol., 133: 1710—1715,1984. D. J., Serrano, M., Hannon, G. J., Shiseki, M., Zariwala, M., Xiong, Y., Beach, D. H., 33. Li, Y., Nichols, M. A., Shay, J. W., and Xiong, Y. Transcriptional repression of the Yokota, J., and Harris, C. C. Mutations in the p16@4/MTSI/CDKN2, pJ5JNK4B/ D-type cyclin-dependent kinase inhibitor p16 by the retinoblastoma susceptibility MTS2, and p!8 genes in primary and metastatic lung cancer. Cancer Res., 55: gene product pRb. Cancer Ret., 54: 6078—6082,1994. 1448—1451,1995. 34. Sakaguchi, M., Yothitaka, F., Hirabayashi, H., Yoon, H-E., Komoto, Y., Oue, T., 22. Nakagawa, K., Conrad, N. K., Williams, J. P., Johnson, B. E., and Kelley, M. J. Kusafuka, T., Okada, A., and Matsuda, H. Inversely correlated expression of p16 and Mechanism of inactivation of CDKN2 and MTS2 in non-small cell lung cancer and Rb protein in non-small cell lung cancers: an immunohistochemical study. Int. J. association with advanced stage. Oncogene, ii: 1843—1851,1995. Cancer., 65: 442—445,1996. 23. Hayashi, N., Sugimoto. Y., Tsuchiya, E., Ogawa, M., and Nakamura, Y. Somatic 35. Belinsky, S. A., Nikula, K. J., Baylin, S. B., and Issa, J-P. J. Increased cytosine mutations of the MTS (multiple tumor suppressor) I/CDK4I (cyclin-dependent ki DNA-methyltransferase activity is target-cell-specific and an early event in lung nase-4 inhibitor) gene in human primary non-small cell lung carcinomas. Biochem. cancer. Proc. Natl. Acad. Sci. USA, 93: 4045—4050, 1996. Biophys. Ret. Commun., 202: 1426—1430, 1994. 36. Ebina, M., Steinberg, S. M., Mulshine, J. L., and Linnoila, R. I. Relationship of p53 24. Cairns, P., Mao, L., Merlo, A., Lee, D. J., Schwab, D., Eby, Y., Tokino, K.. van der overexpression and up-regulation ofproliferating cell nuclear antigen with the clinical Riet, P., Blaugrund. J. E., and Sidransky, D. Rates of p16 (MTSI) mutations in course of non-small cell lung cancer. Cancer Ret., 54: 2496—2503,1994. primary tumors with 9p loss. Science (Washington DC), 265: 415—417, 1994. 37. Harpole, D. H., Jr., Herndon, J. E. I., Wolfe, W. G., Iglehart, J. D., and Marks, J. R. 25. Greenblatt, M. S., Bennett, W. P., Hollstein, M., and Harris, C. C. Mutations in the A prognostic model of recurrence and death in stage I non-small cell lung cancer p53 tumor suppressor gene: clues to cancer etiology and molecular pathogenesis. utilizing presentation, histopathology, and oncoprotein expression. Cancer Ret., 55: Cancer Ret., 54: 4855—4878,1994. 51—56,1995. 26. Cordon-Cardo, C. Mutation of cell cycle regulators: biological and clinical implica 38. Huang, H-J. S., Yee, J-K., Shew, J-Y., Chen, P-L., Bookstein, R., Friedmann, T., Lee, lions for human neoplasia. Am. J. Pathol., 147: 545—560,1995. E. Y-H. P., and Lee, W-H. Suppression of the neoplastic phenotype by replacement 27. Dosaka-Akita, H., Shindoh, M., Fujino, M., Kinoshita, I., Akie, K., Katoh, M., and of the RB gene in human cancer cells. Science (Washington DC), 242: 1563—1566, Kawakami, Y. Abnormal p53 expression in human lung cancer is associated with 1988. histologic subtypes and patient smoking history. Am. J. Clin. Pathol., 102: 660—664, 39. Crook, T., Marston, N. J., Sara, E. A., and Vousden, K. H. Transcriptional activation 1994. by p53 correlates with suppression of growth but not transformation. Cell, 79: 28. Reed, J. A., Loganzo, F., Jr., Shea, C. R., Walker, G. J., Flores, J. F., Glendening, 817—827,1994.

5562

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1996 American Association for Cancer Research. Altered p16INK4 and Retinoblastoma Protein Status in Non-Small Cell Lung Cancer: Potential Synergistic Effect with Altered p53 Protein on Proliferative Activity

Ichiro Kinoshita, Hirotoshi Dosaka-Akita, Takayuki Mishina, et al.

Cancer Res 1996;56:5557-5562.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/56/24/5557

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/56/24/5557. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1996 American Association for Cancer Research.