journa l homepage: www.e lsev ier .com/ locate / lungcan
eview
re adenosquamous lung carcinomas a simple mix of adenocarcinomas andquamous cell carcinomas, or more complex at the molecular level?
ristell Bastidea,∗, Nicolas Ugolina, Céline Levaloisa, Jean-Francois Bernaudinb, Sylvie Chevillarda
CEA, DSV, IRCM, SREIT, Laboratoire de Cancérologie Expérimentale, BP6, Fontenay-aux-Roses Cedex F-92265, FranceService d’Histologie-Biologie Tumorale, Equipe de Recherche ER 2 UPMC, Université Paris 6, Hôpital Tenon, Paris F-75020, France
r t i c l e i n f o
rticle history:eceived 12 May 2009eceived in revised form7 September 2009ccepted 2 November 2009
eywords:ung cancerdenosquamous carcinoma
a b s t r a c t
Adenocarcinomas (AC), squamous cell carcinomas (SCC) and adenosquamous carcinomas (ASC) are threehistological subtypes of non-small-cell lung carcinomas (NSCLC). ASC are morphologically mixed tumoursthat contain the two cell components AC and SCC. To understand if they are a “simple” mix of AC andSCC or if they present molecular specificities, as compared with the molecular characterization of bothcomponents, we performed a comparative transcriptome analysis on a series of nine ASC, five AC and fiveSCC induced in rats by radon exposure. We found that 72, 40 and 39 genes were differentially expressedwhen comparing AC SCC, ASC SCC and AC ASC, respectively. Moreover, when classifying the three histo-logical subtypes, using genes that discriminated AC and SCC, we observed that all ASC were classified as
intermediate between the AC and SCC, some being closer to AC, others to SCC. These results indicated that,regarding gene expression, ASC could be considered as a mix of AC and SCC, both in various proportions.However, they also exhibit molecular specificities since we found specific genes discriminating ASC SCCand AC ASC. In conclusion, the ASC mixed lung tumours are more complex than simple mixes of AC andSCC components. Neuroendocrine differentiation and ERK proliferation pathways seemed preferentiallyderegulated in ASC compared to AC and SCC respectively, pathways that are worthy of being explored
lly ex
because they could partia
. Introduction
Lung cancer is the leading cause of cancer death worldwide [1,2].he non-small-cell lung carcinomas (NSCLC) are the most frequentype of lung tumours in humans, with two major histological sub-ypes: adenocarcinomas (AC) and squamous cell carcinomas (SCC).he adenosquamous carcinomas (ASC), a less frequent subtype inuman lung cancer, are also classed as NSCLC. ASC form morpholog-
cally mixed tumours composed of the two cell components AC andCC in various proportions, each one representing at least 10% of thehole tumour [3,4]. Although infrequent in humans, this composite
umour is more aggressive than “pure” AC and SCC, with frequentymph-node metastasis at diagnosis, and a poor prognosis [5–7].
orphological characterizations of AC and SCC are well described,C being glandular with alveolar, tubular or papillary structureshile SCC are squamous with or without keratin differentiation.
any researches have analyzed genetic and molecular alterations
n AC and SCC, but few molecular studies have been conducted oneterogeneous ASC tumours. One paper described identical Tp53utations in the two histological components AC and SCC cells of
12 lung ASC [8], suggesting that the two cell components origi-nated from the same clone or genetically related cellular clones.To investigate the molecular characterization of ASC and specifi-cally to determine whether these complex tumours correspond toa simple mix of AC and SCC cells or if they present their own molec-ular specificities, we used transcriptome analysis to compare geneexpression profiles of a series of morphologically well characterizedASC, AC and SCC developed in rats after radon inhalation [9].
2. Materials and methods
2.1. Biological samples
A large series of lifespan experiments carried out in our lab-oratory on out-bred Sprague–Dawley rats after radon progenyinhalation provided primary rat lung carcinomas [9]. Animals werehandled according to the French Legislature and the EuropeanDirectives regarding the care and use of laboratory animals. Afterradon exposure, rats were kept until death or were euthanized
when moribund. All pulmonary lobes were systematically paraffin-embedded to microscopically detect tumours with a global viewand to propose a definitive histological diagnosis according tothe European Late Effects Project Group (EULEP) classification[10]. Before embedding, macroscopically detectable tumours were
unctured and were kept frozen for further RNA extraction andolecular biology studies. For the transcriptomic study, nine ASC,
ve AC and five SCC were selectively collected in 19 different rats.ormal rat lung, liver, spleen, muscle and brain tissues, kidneypithelial cell line (NRK-52E) and osteosarcoma cell lines (UMR-06 and UMR-108) were also collected to constitute an externalat reference (RR) for microarray hybridization. Total RNA wassolated from each frozen tumour (Exp), and from normal tissuesnd cell lines, using the RNA-plus lysis solution, according to theanufacturer’s protocol (Q-Biogene, Illkirsch, France). To prepare
he external rat reference for microarray analysis, RNAs from eachormal tissue and cell line were mixed in equimolar quantities.
.2. cDNA synthesis and labelling
To avoid RNA degradation during microarray hybridization, weeveloped a technique to hybridize labelled cDNA. For each tumourExp) and for the reference (RR), 5 �g of total RNA were amplifiedsing the MessageAmpTM aRNA kit (Ambion). Amplified RNA waseverse transcribed as previously described [11]. The second strandf cDNA was synthesized and labelled after overnight incubationt 37 ◦C in 1× Hexanucleotide buffer (Roche Diagnostics) contain-ng 5U Klenow DNA polymerase (Roche Diagnostics), unlabelleducleotides (final concentration 200 �M each dATP, dCTP, dGTP,nd 130 �M dTTP), and either Cy3 or Cy5 conjugated dUTP (finaloncentration 70 �M; Amersham).
.3. Microarrays
The rat 10K 50 mers Oligo Set (6100 genes of known function,560 ESTs, 100 replicas and 169 Arabidoposis negative controls)Ocimum Biosolutions) was spotted on the CEA genomic platformCEA, Evry, France) on GAPS coated slides (Corning) and exposedo UV radiation as previously described [11]. Gene annotationas updated monthly from the database of Stanford University
http://source.stanford.edu) and functional grouping of genes waserformed by using key words of the NCBI Entrez Gene databasehttp://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=gene) andpecific bibliographic data.
.4. Hybridization
For each tumour, competitive hybridization was performedsing equal quantities of Cy3-labelled cDNA from Exp and Cy5-
abelled cDNA from RR. After alcohol precipitation with 20 �g eachf poly(A), yeast tRNA and rat Cot-1 DNA (Rat Hybloc, Appliedenetics Laboratories), the cDNA pellet was dissolved in 50 �l ofybridization solution (50% formamide, 10× SSC and 0.2% SDS) andeated for 5 min at 95 ◦C. The preparation was dropped onto pre-ybridized arrays (incubated 1 h at 42 ◦C in a buffer 1% BSA, 5× SSC,.1% SDS). After overnight incubation at 42 ◦C in a humidified cham-er, slides were washed for 4 min in 1× SSC–0.2% SDS at 42 ◦C, 0.1×SC–0.2% SDS, 0.1× SSC and distilled water at room temperature,nd were dried by centrifugation. Competitive hybridization waserformed for each tumour in duplicated dye-swap experiments.
.5. Microarray scanning and data analysis
Microarrays were scanned using a confocal laser scanner (Scan-rray Express, PerkinElmer). Independent images were acquired
or Cy3 (532 nm) and Cy5 (635 nm). The spot coordinates wereetermined by automatic positioning of a grid using an image-nalysis spot-tracking software (patent US 10/173,672 June, 19,002; CA 2,389,901 June, 20, 2002). The signal intensity of eachpot was defined as previously described [11,12].
cer 68 (2010) 1–9
Data normalization and expression quantification were also per-formed as described [11], expression ratios being log-transformed(base 2) and median-centred. To identify genes that were differen-tially expressed between two groups of tumours (AC versus SCC,AC versus ASC or ASC versus SCC), we used a supervised microar-ray analysis. We did a permutation t-test to select a first list ofgenes. Then, to avoid choosing an arbitrary cut-off of p-value forthe t-test, an EM algorithm was used to precisely calculate, for eachtumour, the probability of over- and under-expression for each ofthese genes. We finally selected genes with a difference of morethan 95% between these two probabilities, showing no contradic-tion in the sense of gene expression deregulation when comparingall the tumours of a given subtype, and showing a difference inthe sense of deregulation when comparing the two subtypes oftumours. For example, to discriminate AC from ASC, we retainedgenes over-expressed in all AC and under-expressed in all ASC,and conversely, under-expressed in all AC and over-expressed inall ASC. Finally, using the selected genes, tumours were classi-fied using the ascendant hierarchical classification algorithm ofthe web-available TIGR MultiExperiment Viewer (MeV) software(http://www.tm4.org/mev.html).
For the first strand cDNA synthesis, 100 ng of amplified RNA wasreverse transcribed using random hexamers (pd(N)6; BoehringerMannheim) and the reverse transcriptase MMLV (Invitrogen).PCR was performed in duplicate using the 7300 real-time PCRsystem (Applied Biosystems) by using the ABsolute QPCR ROX Mix(Abgene). Quantification of each gene expression was calibratedas previously described [13–15], using a reference standard curveobtained by serial dilutions of a PCR product prepared from RR.The geometric mean expression of three housekeeping genes,Actb, Gusb, and Hmbs, allowed to normalize the expression ofthe interesting genes by using the web-available geNorm soft-ware (http://medgen.ugent.be/∼jvdesomp/genorm) [16]. The PCRprimers and probes were purchased from Applied Biosystemsfor Hp, Fgg, Krt19, Cdk2, Dsp, Prkaa2, Gng12, Gnas, Notch2,Rab13, Sftpa1, Amph1, Muc1, Mst1, Actb, Gusb, and Hmbs genes(TaqMan Gene Expression Assays: assay ID Rn00561393 m1,Rn00561196 m1, Rn01496867 m1, Rn01529542 m1,Rn01434652 m1, Rn00576935 m1, Rn01425123 m1,Rn00569454 m1, Rn00577522-m1, Rn00709515 m1,Rn00824545 m1, Rn00824616 m1, Rn01462585 m1,Rn00577395 m1, Rn00667869 m1, Rn00566655-m1 andRn00565886-m1 respectively).
2.7. K-Ras gene mutation
On the series of tumours, K-Ras mutations at codons 12, 13 and61 were screened for on PCR-amplified cDNA using the forwardprimer GCCTGCTGAAAATGACTGAGTAT and the reverse primerAAAGAAAGCCCTCCCCAGTT (1.7 mM MgCl2—Tm 60 ◦C). Sequenc-ing reactions were carried out on both strands by the PremiumRead service of Cogenics (Meylan, France) and compared with thereferential sequence NM 031515 from GenBank database (NCBI).
3. Results
The lifespan experiments carried out in our laboratory afterradon progeny inhalation in rat provided three types of primary
lung carcinomas [9]: squamous cell carcinomas (SCC), adenocarci-nomas (AC) and adenosquamous carcinomas (ASC), which presentmolecular similarities with human non-small cell lung cancers[17]. Each tumour type could have heterogeneous microscopicphenotypes (Fig. 1). SCC could have different stages of squamous
ig. 1. Histological subtypes of lung carcinomas induced in rat after radon inhalatB and C) typical patterns of papillary and tubular adenocarcinomas (AC) respectiveonfluent (D and E) or merged (F and G).
ifferentiation with or without keratin pearls (Fig. 1A). AC revealedlandular phenotypes with diverse papillary or tubular formationsFig. 1B and C). ASC were heterogeneous because of the variousatios of AC and SCC components and because the two cell compo-ents could be confluent (Fig. 1D and E) or merged (Fig. 1F and G).
A comparative transcriptome analysis was carried out on thesehree different types of tumours in order to look for specific genexpression signatures discriminating AC from SCC, AC from ASC andSC from SCC, each type being considered as a unique entity despite
heir microscopic heterogeneity. Interestingly, we found that 72enes discriminated AC from SCC, 40 genes distinguished ASC fromCC and 39 genes constituted the signature AC ASC (Fig. 2). Amonghe differentially expressed genes, we found several genes involvedn the tumorigenesis. These included genes implicated in cell adhe-ion, angiogenesis, cell division, cell cycle regulation, epithelial andeuroendocrine differentiation, and tumour-suppressor functions
Tables 1–3).
The functions and pathways of the 40 deregulated genes dis-riminating ASC from SCC (Table 2) led us to search for K-Rasutation in our series of tumours (see Section 4) but any mutationas detected (data not shown).
) Typical patterns of squamous cell carcinomas (SCC) with keratin pearls (arrow);–G) patterns of adenosquamous carcinomas (ASC): SCC and AC components can be
In order to discover if ASC correspond to a mixture in var-ious proportions of the two subtypes AC and SCC, in terms ofgene expression, we classified the three subtypes of tumours usingthe specific gene signature discriminating AC from SCC (Fig. 2A).Interestingly, ASC classified as a group between AC and SCC, sometumours being closer to AC (ASC-1 and 2) than to SCC, while othersseem closer to SCC (ASC-9) than to AC.
To confirm the gene expression data obtained by microarrayanalysis, the expression levels of several representative genes wereanalyzed by quantitative real-time RT-PCR (Table 4). As a whole, thedifferential expression profiles identified in the three signatures bythe microarray analysis were reproduced by the RT-PCR analysis.
4. Discussion
The aim of the transcriptome analysis of nine ASC compared
with five AC and five SCC was to discover if these tumours,histologically composed of AC and SCC cells, constitute a mix atthe molecular level or if they present their own specificities. SinceASC are diagnosed if each glandular and squamous componentrepresents at least 10% of the tumour, they form a heterogeneous
4 K. Bastide et al. / Lung Cancer 68 (2010) 1–9
Table 1Genes differentially expressed between adenocarcinomas (AC) and squamous cell carcinomas (SCC).
Gene symbol Accession no.a Gene name Functionb Mean ACc Mean SCCd p-Value
b Gene function summarized from literature sources and/or NCBI Entrez Gene dac Mean of the AC log values centered and normalized.d Mean of the SCC log values centered and normalized.
roup with many variations on the proportion of each AC and SCComponent [18].
Firstly, in order to compare the ASC gene expression profile
ith that of AC and SCC, we analyzed how ASC classify among
he AC and SCC. A signature AC SCC, composed of 72 genes, wasdentified (Table 1). Comparative transcriptome analyses have beenerformed on human AC and SCC [19–24] and many of the genesiscriminating AC and SCC were found in our AC SCC signature such
able 2enes differentially expressed between adenosquamous carcinomas (ASC) and squamou
Notch2 NM 024358 Notch gene homolog 2Cyp11b NM 012538 Cytochrome P450, family 11, subfamily B, polypeptidGata6 NM 019185 GATA binding protein 6Hoxc4 M37567 Homeo box C4Mrgprf M35297 MAS-related GPR, member FOprl1 NM 031569 Opioid receptor-like 1Tas2r105 NM 023999 Taste receptor, type 2, member 105Dyt1 NM 153303 Dystonia 1Ap1gbp1 AF169549 AP1 gamma subunit binding protein 1Myo1b NM 053986 Myosin IbUnc13d AF159356 Unc-13 homolog DScrn2 NM 001012142 Secernin 2Fbxo22 NM 001037770 F-box only protein 22Cab39 XM 217464 Calcium binding protein 39Sost NM 030584 Sclerostin
Crlz1 NM 001012036 Charged amino acid rich leucine zipper 1Mum1 NM 001108736 Melanoma associated antigen (mutated) 1RGD 1564852 XM 228973 Similar to hypothetical protein FLJ14503LOC500097 AF213506 Similar to T-cell receptor beta chain V region C5 precCDNA clone BC086519 CDNA clone IMAGE: 7302535RGD 735106 NM 198766 Similar to RIKEN cDNA 0610011N22 gene
a GenBank accession number.b Gene function summarized from literature sources and/or NCBI Entrez Gene databasec Mean of the ASC log values centered and normalized, d: mean of the SCC log values c
as the tumour-suppressor gene Dlc1 (deleted in liver cancer 1),Lmo7 associated with actinin for cell adhesion, and the desmosomecomponent Dsp (desmoplakin). Many genes coding for keratins
were also observed in the rat signature in accordance with thehuman literature: keratins 1, 7, 8 and 18 were more expressed in AC[25–27] whereas the keratins 5, 14, 19 and 20 were over-expressedin SCC [25,26,28–30]. These data confirmed that AC and SCC canbe distinguished by epithelial differentiation markers in the two
ig. 2. Hierarchical clustering of the tumours according to the genes differentially exC) AC ASC signature. The upper dendrograms represent classification of the tumou
pecies. Moreover, many genes implicated in proliferation, angio-enesis and metastasis were differentially expressed between ratC and SCC, including the growth factor Vegfa, the cyclin Cdk2, theetalloproteinase Mmp9, claudin 7 (Cldn7), fibrinogen (Fgg) and
aptoglobin (Hp). All of these genes are also described as specificarkers to distinguish the two types of human tumours [31–36].
he expression profiles of all these genes indicated that AC couldave a higher metastatic potential than SCC, as previously described
n human [37]. Overall, we showed that the present AC SCC signa-ure, which effectively classified the AC and SCC rat lung tumoursFig. 2A), was consistent with published data on human tumours.hen, using this gene expression signature we checked the clas-ification of all the tumours, AC, ASC and SCC. Interestingly, theSC clustered as a group between AC and SCC tumours (Fig. 2A),ome tumours being closer to the AC group while others wereloser to the SCC group, probably according to the diverse pro-ortions of AC and SCC in the ASC tumours. The histological dataeing non-quantitative, it was not possible to confirm the molecu-
ar classification. Anyway, this analysis clearly indicated that, whenonsidering expression of genes of the AC SCC signature, the mixSC appeared to be a true mixture of AC and SCC, in various pro-ortions.
Secondly, we investigated if ASC had molecular specificitiesompared with AC and SCC and we looked for AC ASC and ASC SCC
ene expression signatures. Two signatures were found (Fig. 2Bnd C) that did not share any common genes (Tables 2 and 3) andhat have less than four common genes with the AC SCC signature.hese specific signatures, never described before, indicated for therst time that ASC present molecular specificities compared with
ed between the three types of tumours. (A) AC SCC signature, (B) ASC SCC signature,nes of the AC SCC signature were used to classify ASC with AC and SCC (A).
each AC and SCC. Compared with the SCC, the genes Gata6 andHoxc4, implicated in differentiation of the respiratory epitheliumand keratinocytes [38,39], were less expressed in ASC, in agreementwith a smaller squamous component in mixed ASC than in “pure”SCC. The Notch2 gene, expressed in non-neuroendocrine lung cells,induced transcriptional repression of genes implicated in neuroen-docrine differentiation [40]. Its under-expression in ASC indicatedthat mixed tumours could belong to the minority of NSCLC thatpresent a neuroendocrine profile [41]. Many genes of the ASC SCCsignature belonged to the MAPK pathway (Fgf4, Gng12, Gnas, Mosand Ptprf) and were deregulated in a sense that could preferentiallyactivate ERK in all the heterogeneous ASC tumours compared withSCC (Fig. 3). The MAPK cascade is known to activate cell prolifera-tion and is implicated in malignant transformation, JNK, p38MAPKand ERK being more specifically involved in NSCLC carcinogenesis[42–45]. Since K-Ras is a major component of the ERK cascade andwas previously found mutated in human NSCLC [46], we searchedfor the K-Ras mutation in our series of tumours, but found none.Because K-Ras and other main genes of the ERK pathway (c-Mos,B-Raf, Egfr) were exclusively activated in human lung tumours[47–50], future investigations on these other genes would be use-ful to determine the precise role of the ERK pathway on mixedtumours.
On the AC ASC signature, as expected, the differential expres-
sion of Muc1 and Sftpa1 indicated that AC had a higher glandulardifferentiation than ASC. Moreover, we observed deregulation ofMst1, Chrna7 and Amph1 expression, genes that are known to beimplicated in the phenotype of human neuroendocrine tumours[51–53]. These data implied that neuroendocrine differentiation
K. Bastide et al. / Lung Cancer 68 (2010) 1–9 7
Table 3Genes differentially expressed between adenocarcinomas (AC) and adenosquamous carcinomas (ASC).
Gene symbol Accession no.a Gene name Functionb Mean ACc Mean ASCd p-Value
MiscellaneousArrb2 NM 012911 Arrestin, beta 2 Signal transduction −0.402 0.236 0.005Bspry NM 022261 B-box and SPRY domain containing Signal transduction 0.175 0.050 0.012Tiprl XM 341145 TIP41, TOR signalling pathway regulator-like Signal transduction −0.142 0.297 0.009Gpiap1 NM 001012185 GPI-anchored membrane protein 1 Activator of transcription −0.421 −0.090 0.009Apbb1 NM 080478 Amyloid beta precursor protein-binding, family B, memb 1 Regulator of transcription −0.073 0.301 0.007Zmynd11 BC065308 Zinc finger, MYND domain containing 11 Repressor of transcription 0.161 −0.308 0.006Cope XM 214309 Coatomer protein complex, subunit epsilon Protein transport −0.030 −0.292 0.014Psmd5 XM 216041 Proteasome 26S subunit, non-ATPase, 5 Protein degradation −0.160 0.302 0.008Sirt3 XM 215124 Sirtuin3 Protein deacetylation −0.156 0.115 0.009Wbp1 XM 216198 WW domain binding protein 1 Protein interaction 0.995 0.216 0.006LOC501110 NM 001024361 Similar to Glutathione S-transferase A1 Detoxification 0.440 −0.496 0.005Mrs2l NM 024001 MRS2-like, magnesium homeostasis factor Magnesium transport 0.261 −0.914 0.005Atp2b1 NM 053311 ATPase, Ca++ transporting, plasma membrane 1 Calcium transport 0.007 0.416 0.003Ryr1 AF112256 Ryanodine receptor 1, skeletal muscle Calcium transport −0.119 0.376 0.008Trhr NM 013047 Thyrotropin releasing hormone receptor Thyroid regulation 0.034 0.263 0.006RGD1309350 XM 215112 Similar to transthyretin (4L369) Thyroid hormones transport 0.364 −0.048 0.008Stard3nl NM 001008298 STARD3 N-terminal like Cholesterol transport 0.442 −0.017 0.010Echdc2 XM 216479 Enoyl Coenzyme A hydratase domain containing 2 Fatty acid metabolism 0.208 −0.706 0.004Acsl1 D90109 Acyl-CoA synthetase long-chain family member 1 Fatty acid metabolism 0.552 −0.056 0.009Lep NM 013076 Leptin Fatty acid catabolism −0.053 0.116 0.009Prlh NM 022222 Prolactin releasing hormone Hormonal activity −0.253 −0.027 0.008Cryaa NM 012534 Crystallin, alpha A Crystalline protein 0.247 −0.078 0.007Ctla4 NM 031674 Cytotoxic T-lymphocyte-associated protein 4 Immune response 0.466 −0.114 0.007
Unknown functionPmf31 AB020504 PMF32 protein Unknown −0.267 −0.691 0.006RGD 1305138 XM 235689 Similar to expressed sequence AW556797 Unknown 0.243 −0.533 0.004RGD 1308468 XM 227253 Similar to expressed sequence C87860 Unknown 0.07 −0.423 0.004RGD 1560183 XM 224552 Similar to hypothetical protein FLJ14624 Unknown −0.091 0.218 0.006RGD 1562351 XM 575345 Similar to chromosome 7 open reading frame 23 Unknown 0.790 −0.330 0.005RGD 1561459 XM 341884 Similar to RIKEN cDNA 1810020D17 Unknown 0.427 −0.316 0.010
a GenBank accession number.b Gene function summarized from literature sources and/or NCBI Entrez Gene database (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=gene).c Mean of the AC log values centered and normalized.d Mean of the ASC log values centered and normalized.
Table 4Confirmation of microarray data by real-time RT-PCR.
a Mean of relative expression for each group of tumours obtained by RT-PCR.b Mean of log values centered and normalized for each group of tumours obtained by microarray analyses.
ig. 3. Representation of deregulated genes related to the ERK pathway when compreen indicate over-and under-expressed genes in ASC compared with SCC, respect
ould be more pronounced in ASC than in AC. The deregulationf this type of genes, on the both AC ASC and previous ASC SCCignatures, emphasizes a specific neuroendocrine gene expressionrofile in ASC. Variations on neuroendocrine differentiation haveeen already described between AC and SCC [54], but it was the firstime that this profile was underlined in ASC. The relevance for prog-osis of neuroendocrine differentiation observed in some humanSCLC is controversial [55–57]. Consequently, the impact of theeuroendocrine differentiation on ASC aggressiveness remained toe better investigated. In addition, the over-expression of cytoplas-ic calcium transporters and calcium entry facilitators (Atp2b1,
yr1, Chrna7 and Trhr) that was observed in rat ASC comparedo AC, could also explain why ASC are more aggressive than AC,ince calcium mobilization allows initiation of many mitogenic cas-ades [58,59] and is essential for the proliferation of SCLC cells60]. Finally, in the AC ASC signature, three genes of fatty acid
etabolism were identified (Echdc1, Acsl1 and Lep). Fatty acidsrovide energy for biosynthesis, such as nucleic acid or aminocid synthesis, that are necessary for cell proliferation and tumourell growth [61]. The differential expression profile of all theseenes in ASC and AC could indicate a major role of the mito-enic pathways regulated by calcium in ASC and suggested thathe development of AC and ASC could require different proliferationathways.
. Conclusion
Overall, using transcriptome analysis, this study indicated thatSC could be considered as a mix of various proportions of AC andCC, when comparing all ASC. However, they also exhibit molecularpecificities since we found specific deregulated genes discriminat-ng ASC SCC and AC ASC. In conclusion, ASC lung tumours are moreomplex than simple mixes of AC and SCC components. The neu-oendocrine differentiation and ERK proliferation pathways, thateemed to be preferentially promoted in these tumours, need fur-her investigation to determine whether these specificities have aole in the clinical aggressiveness of the ASC.
onflict of interest statement
None declared.
cknowledgements
This work was supported in part by the EC RISC-RAD contract –I6R-CT-2003-508842, AREVA NC and Electricité de France.
[
ASC and SCC. The coloured genes are deregulated in the ASC SCC signature: red and
References
[1] Alberg AJ, Brock MV, Samet JM. Epidemiology of lung cancer: looking to thefuture. J Clin Oncol 2005;23(14):3175–85.
[2] Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer JClin 2005;55(2):74–108.
[3] Stancu M, Libbey PN. In: Weitberg AB, editor. Cancer of the lung: from molecularbiology to treatment guidelines. Totowa, NJ: Humana; 2002. p. 35–80.
[4] Travis WD, Colby TV, Corrin B, Shimosato Y, Brambilla E. Histological typing oflung and pleural tumors. Berlin: Springer-Verlag; 1999.
[5] Hsia JY, Chen CY, Hsu CP, Shai SE, Wang PY. Adenosquamous carcinoma of thelung Surgical results compared with squamous cell and adenocarcinoma. ScandCardiovasc J 1999;33(1):29–32.
[6] Gawrychowski J, Brulinski K, Malinowski E, Papla B. Prognosis and survival afterradical resection of primary adenosquamous lung carcinoma. Eur J Cardiotho-rac Surg 2005;27(4):686–92.
[7] Nakagawa K, Yasumitu T, Fukuhara K, Shiono H, Kikui M. Poor prognosis afterlung resection for patients with adenosquamous carcinoma of the lung. AnnThorac Surg 2003;75(6):1740–4.
[8] Kanazawa H, Ebina M, Ino-Oka N, Shimizukawa M, Takahashi T, Fujimura S, ImaiT, Nukiwa T. Transition from squamous cell carcinoma to adenocarcinoma inadenosquamous carcinoma of the lung. Am J Pathol 2000;156(4):1289–98.
[9] Monchaux G, Morlier JP, Altmeyer S, Debroche M, Morin M. Influence of expo-sure rate on lung cancer induction in rats exposed to radon progeny. Radiat Res1999;152(6 Suppl.):S137–40.
10] Hahn FF, Boorman GA. Neoplasia and preneoplasia of the lung. In: Bannasch P,Gössner W, editors. Pathology of neoplasia and preneoplasia in rodents, EULEPColor Atlas, vol. 2. Stuttgart/New York: Schattauer; 1997. pp. 29–42.
11] Chevillard S, Ugolin N, Vielh P, Ory K, Levalois C, Elliott D, Clayman GL, El-NaggarAK. Gene expression profiling of differentiated thyroid neoplasms: diagnosticand clinical implications. Clin Cancer Res 2004;10(19):6586–97.
12] Klising-Sireul E, Rigaud O, Ory K, Ugolin N, Lebeau J, Levalois C, Lectard B,Chevillard S. Transcriptional response of wild-type and ataxia telangiectasialymphoblasts following exposure to equitoxic doses of ionizing radiation. JRadiat Res (Tokyo) 2006;47(3–4):259–72.
13] Bishay K, Ory K, Olivier MF, Lebeau J, Levalois C, Chevillard S. DNA damage-related RNA expression to assess individual sensitivity to ionizing radiation.Carcinogenesis 2001;22(8):1179–83.
14] Chevillard S, Muller A, Levalois C, Laine-Bidron C, Vielh P, Magdelenat H.Reverse transcription-polymerase chain reaction (RT-PCR) assays of estro-gen and progesterone receptors in breast cancer. Breast Cancer Res Treat1996;41(1):81–9.
15] Oudard S, Levalois C, Andrieu JM, Bougaran J, Validire P, Thiounn N, Poupon MF,Fourme E, Chevillard S. Expression of genes involved in chemoresistance, prolif-eration and apoptosis in clinical samples of renal cell carcinoma and correlationwith clinical outcome. Anticancer Res 2002;22(1A):121–8.
16] Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, De Paepe A,Speleman F. Accurate normalization of real-time quantitative RT-PCR data bygeometric averaging of multiple internal control genes. Genome Biol 2002;3(7).RESEARCH0034.
17] Dano L, Guilly MN, Muleris M, Morlier JP, Altmeyer S, Vielh P, El-Naggar AK,Monchaux G, Dutrillaux B, Chevillard S. CGH analysis of radon-induced rat lungtumors indicates similarities with human lung cancers. Genes ChromosomesCancer 2000;29(1):1–8.
18] Ishida T, Kaneko S, Yokoyama H, Inoue T, Sugio K, Sugimachi K. Adenosquamouscarcinoma of the lung clinicopathologic and immunohistochemical features.Am J Clin Pathol 1992;97(5):678–85.
Furukawa Y, Kawamura M, Kobayashi K, Imai K, Nakamura Y. Expression pro-files of non-small cell lung cancers on cDNA microarrays: identification of genesfor prediction of lymph-node metastasis and sensitivity to anti-cancer drugs.Oncogene 2003;22(14):2192–205.
20] Yamagata N, Shyr Y, Yanagisawa K, Edgerton M, Dang TP, Gonzalez A, Nadaf S,Larsen P, Roberts JR, Nesbitt JC, Jensen R, Levy S, Moore JH, Minna JD, Carbone
ng Can
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
K. Bastide et al. / Lu
DP. A training-testing approach to the molecular classification of resected non-small cell lung cancer. Clin Cancer Res 2003;9(13):4695–704.
21] Kettunen E, Anttila S, Seppanen JK, Karjalainen A, Edgren H, Lindstrom I,Salovaara R, Nissen AM, Salo J, Mattson K, Hollmen J, Knuutila S, Wikman H. Dif-ferentially expressed genes in nonsmall cell lung cancer: expression profilingof cancer-related genes in squamous cell lung cancer. Cancer Genet Cytogenet2004;149(2):98–106.
22] Hofmann HS, Bartling B, Simm A, Murray R, Aziz N, Hansen G, Silber RE, BurdachS. Identification and classification of differentially expressed genes in non-smallcell lung cancer by expression profiling on a global human 59.620-elementoligonucleotide array. Oncol Rep 2006;16(3):587–95.
23] Boelens MC, van den Berg A, Vogelzang I, Wesseling J, Postma DS, TimensW, Groen HJ. Differential expression and distribution of epithelial adhesionmolecules in non-small cell lung cancer and normal bronchus. J Clin Pathol2007;60(6):608–14.
24] Angulo B, Suarez-Gauthier A, Lopez-Rios F, Medina PP, Conde E, Tang M, Soler G,Lopez-Encuentra A, Cigudosa JC, Sanchez-Cespedes M. Expression signaturesin lung cancer reveal a profile for EGFR-mutant tumours and identify selec-tive PIK3CA overexpression by gene amplification. J Pathol 2008;214(3):347–56.
25] Young GD, Winokur TS, Cerfolio RJ, Van Tine BA, Chow LT, Okoh V, Garver Jr RI.Differential expression and biodistribution of cytokeratin 18 and desmoplakinsin non-small cell lung carcinoma subtypes. Lung Cancer 2002;36(2):133–41.
26] Wikman H, Kettunen E, Seppanen JK, Karjalainen A, Hollmen J, Anttila S,Knuutila S. Identification of differentially expressed genes in pulmonary ade-nocarcinoma by using cDNA array. Oncogene 2002;21(37):5804–13.
27] Su YC, Hsu YC, Chai CY. Role of TTF-1, CK20, and CK7 immunohistochemistryfor diagnosis of primary and secondary lung adenocarcinoma. Kaohsiung J MedSci 2006;22(1):14–9.
28] Miettinen M, Sarlomo-Rikala M. Expression of calretinin, thrombomod-ulin, keratin 5, and mesothelin in lung carcinomas of different types: animmunohistochemical analysis of 596 tumors in comparison with epithelioidmesotheliomas of the pleura. Am J Surg Pathol 2003;27(2):150–8.
29] Camilo R, Capelozzi VL, Siqueira SA, Del Carlo Bernardi F. Expression of p63,keratin 5/6, keratin 7, and surfactant-A in non-small cell lung carcinomas. HumPathol 2006;37(5):542–6.
30] Murata K, Ota S, Niki T, Goto A, Li CP, Ruriko UM, Ishikawa S, Abu-ratani H, Kuriyama T, Fukayama M. p63—key molecule in the early phaseof epithelial abnormality in idiopathic pulmonary fibrosis. Exp Mol Pathol2007;83(3):367–76.
31] Beckman G, Eklund A, Frohlander N, Stjernberg N. Haptoglobin groups and lungcancer. Hum Hered 1986;36(4):258–60.
32] Dobashi Y, Shoji M, Jiang SX, Kobayashi M, Kawakubo Y, Kameya T. Activecyclin A-CDK2 complex, a possible critical factor for cell proliferation in humanprimary lung carcinomas. Am J Pathol 1998;153(3):963–72.
33] Yuan A, Yu CJ, Chen WJ, Lin FY, Kuo SH, Luh KT, Yang PC. Correlation of totalVEGF mRNA and protein expression with histologic type, tumor angiogenesis,patient survival and timing of relapse in non-small-cell lung cancer. Int J Cancer2000;89(6):475–83.
34] Amatschek S, Koenig U, Auer H, Steinlein P, Pacher M, Gruenfelder A, DekanG, Vogl S, Kubista E, Heider KH, Stratowa C, Schreiber M, Sommergruber W.Tissue-wide expression profiling using cDNA subtraction and microarrays toidentify tumor-specific genes. Cancer Res 2004;64(3):844–56.
35] Jones JM, McGonigle NC, McAnespie M, Cran GW, Graham AN. Plasma fib-rinogen and serum C-reactive protein are associated with non-small cell lungcancer. Lung Cancer 2006;53(1):97–101.
36] Moldvay J, Jackel M, Paska C, Soltesz I, Schaff Z, Kiss A. Distinct claudinexpression profile in histologic subtypes of lung cancer. Lung Cancer2007;57(2):159–67.
37] Kotoulas CS, Foroulis CN, Kostikas K, Konstantinou M, Kalkandi P, DimadiM, Bouros D, Lioulias A. Involvement of lymphatic metastatic spread in non-small cell lung cancer accordingly to the primary cancer location. Lung Cancer2004;44(2):183–91.
38] Guo M, Akiyama Y, House MG, Hooker CM, Heath E, Gabrielson E, Yang SC, Han
Y, Baylin SB, Herman JG, Brock MV. Hypermethylation of the GATA genes inlung cancer. Clin Cancer Res 2004;10(23):7917–24.
39] Rieger E, Bijl JJ, van Oostveen JW, Soyer HP, Oudejans CB, Jiwa NM, WalboomersJM, Meijer CJ. Expression of the homeobox gene HOXC4 in keratinocytes of nor-mal skin and epithelial skin tumors is correlated with differentiation. J InvestDermatol 1994;103(3):341–6.
[
[
cer 68 (2010) 1–9 9
40] Ito T, Udaka N, Yazawa T, Okudela K, Hayashi H, Sudo T, Guillemot F,Kageyama R, Kitamura H. Basic helix-loop-helix transcription factors regu-late the neuroendocrine differentiation of fetal mouse pulmonary epithelium.Development 2000;127(18):3913–21.
41] Carnaghi C, Rimassa L, Garassino I, Santoro A. Clinical significance of neuroen-docrine phenotype in non-small-cell lung cancer. Ann Oncol 2001;12(Suppl.2):S119–23.
42] Adjei AA. The role of mitogen-activated ERK-kinase inhibitors in lung cancertherapy. Clin Lung Cancer 2005;7(3):221–3.
43] Greenberg AK, Basu S, Hu J, Yie TA, Tchou-Wong KM, Rom WN, Lee TC. Selectivep38 activation in human non-small cell lung cancer. Am J Respir Cell Mol Biol2002;26(5):558–64.
44] Coe BP, Lockwood WW, Girard L, Chari R, Macaulay C, Lam S, Gazdar AF, MinnaJD, Lam WL. Differential disruption of cell cycle pathways in small cell andnon-small cell lung cancer. Br J Cancer 2006;94(12):1927–35.
45] Vicent S, Lopez-Picazo JM, Toledo G, Lozano MD, Torre W, Garcia-Corchon C,Quero C, Soria JC, Martin-Algarra S, Manzano RG, Montuenga LM. ERK1/2 isactivated in non-small-cell lung cancer and associated with advanced tumours.Br J Cancer 2004;90(5):1047–52.
46] Aviel-Ronen S, Blackhall FH, Shepherd FA, Tsao MS. K-ras mutations in non-small-cell lung carcinoma: a review. Clin Lung Cancer 2006;8(1):30–8.
47] Kosaka T, Yatabe Y, Endoh H, Kuwano H, Takahashi T, Mitsudomi T. Mutationsof the epidermal growth factor receptor gene in lung cancer: biological andclinical implications. Cancer Res 2004;64(24):8919–23.
48] Shigematsu H, Lin L, Takahashi T, Nomura M, Suzuki M, Wistuba II, Fong KM, LeeH, Toyooka S, Shimizu N, Fujisawa T, Feng Z, Roth JA, Herz J, Minna JD, GazdarAF. Clinical and biological features associated with epidermal growth factorreceptor gene mutations in lung cancers. J Natl Cancer Inst 2005;97(5):339–46.
49] Shibata T, Hanada S, Kokubu A, Matsuno Y, Asamura H, Ohta T, SakamotoM, Hirohashi S. Gene expression profiling of epidermal growth factorreceptor/KRAS pathway activation in lung adenocarcinoma. Cancer Sci2007;98(7):985–91.
50] Zacharatos P, Kotsinas A, Tsantoulis P, Evangelou K, Kletsas D, AsimacopoulosPJ, Doussis-Anagnostopoulou I, Pezzella F, Gatter K, Papavassiliou AG, Kittas C,Gorgoulis VG. Relationship of the K-ras/c-mos expression patterns with angio-genesis in non-small cell lung carcinomas. Mol Med 2001;7(9):590–7.
51] Plummer 3rd HK, Dhar M, Schuller HM. Expression of the alpha7 nicotinicacetylcholine receptor in human lung cells. Respir Res 2005;6:29.
52] Bumming P, Nilsson O, Ahlman H, Welbencer A, Andersson MK, Sjolund K,Nilsson B. Gastrointestinal stromal tumors regularly express synaptic vesi-cle proteins: evidence of a neuroendocrine phenotype. Endocr Relat Cancer2007;14(3):853–63.
53] Willett CG, Smith DI, Shridhar V, Wang MH, Emanuel RL, Patidar K, Graham SA,Zhang F, Hatch V, Sugarbaker DJ, Sunday ME. Differential screening of a humanchromosome 3 library identifies hepatocyte growth factor-like/macrophage-stimulating protein and its receptor in injured lung Possible implications forneuroendocrine cell survival. J Clin Invest 1997;99(12):2979–91.
54] Linnoila RI, Mulshine JL, Steinberg SM, Funa K, Matthews MJ, Cotelingam JD,Gazdar AF. Neuroendocrine differentiation in endocrine and nonendocrine lungcarcinomas. Am J Clin Pathol 1988;90(6):641–52.
55] Ionescu DN, Treaba D, Gilks CB, Leung S, Renouf D, Laskin J, Wood-Baker R, Gown AM. Nonsmall cell lung carcinoma with neuroendocrinedifferentiation—an entity of no clinical or prognostic significance. Am J SurgPathol 2007;31(1):26–32.
56] Sterlacci W, Fiegl M, Hilbe W, Auberger J, Mikuz G, Tzankov A. Clinical relevanceof neuroendocrine differentiation in non-small cell lung cancer assessed byimmunohistochemistry: a retrospective study on 405 surgically resected cases.Virchows Arch 2009;455(2):125–32.
57] Pelosi G, Pasini F, Sonzogni A, Maffini F, Maisonneuve P, Iannucci A, Terzi A,De Manzoni G, Bresaola E, Viale G. Prognostic implications of neuroendocrinedifferentiation and hormone production in patients with Stage I nonsmall celllung carcinoma. Cancer 2003;97(10):2487–97.
58] Berridge MJ, Lipp P, Bootman MD. The versatility and universality of calciumsignalling. Nat Rev Mol Cell Biol 2000;1(1):11–21.
for cancer cell proliferation and survival. Nat Rev Cancer 2008;8(5):361–75.60] Gudermann T, Roelle S. Calcium-dependent growth regulation of small cell lung
cancer cells by neuropeptides. Endocr Relat Cancer 2006;13(4):1069–84.61] Deberardinis RJ, Sayed N, Ditsworth D, Thompson CB. Brick by brick:
metabolism and tumor cell growth. Curr Opin Genet Dev 2008;18(1):54–61.
