13
Molecular and Cellular Pathobiology H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra When Driven by the Same Regulatory Sequences Matthias Drosten 1 , Lucía Sim on-Carrasco 1 , Isabel Hern andez-Porras 1 , Carmen G. Lechuga 1 , María T. Blasco 1 , Harrys K.C. Jacob 1 , Salvatore Fabbiano 2,3 , Nicoletta Potenza 4 , Xos e R. Bustelo 2,3 , Carmen Guerra 1 , and Mariano Barbacid 1 Abstract Genetic studies in mice have provided evidence that H-Ras and K-Ras proteins are bioequivalent. However, human tumors display marked differences in the association of RAS oncogenes with tumor type. Thus, to further assess the bioequivalence of oncogenic H-Ras and K-Ras, we replaced the coding region of the murine K-Ras locus with H-Ras G12V oncogene sequences. Germline expression of H-Ras G12V or K-Ras G12V from the K-Ras locus resulted in embryonic lethality. However, expression of these genes in adult mice led to different tumor phenotypes. Whereas H-Ras G12V elicited papillomas and hematopoietic tumors, K-Ras G12V induced lung tumors and gastric lesions. Pulmonary expression of H-Ras G12V created a senescence-like state caused by excessive MAPK signaling. Likewise, H-Ras G12V but not K-Ras G12V induced senescence in mouse embryonic broblasts. Label-free quantitative analysis revealed that minor differences in H-Ras G12V expression levels led to dras- tically different biological outputs, suggesting that subtle differences in MAPK signaling confer nonequivalent functions that inuence tumor spectra induced by RAS oncoproteins. Cancer Res; 77(3); 70718. Ó2016 AACR. Introduction Mammals contain three Ras loci, H-Ras, K-Ras, and N-Ras, that encode highly related proteins (14). Ras proteins are small GTPases that function as mitogenic switches to control the trans- mission of extracellular signals to the nucleus (1). They share extensive homology at their N-terminal half, a region involved in nucleotide binding and interaction with downstream effectors (2). Their unique features reside in their carboxy-terminal half that includes the hypervariable region and a terminal domain known as the CAAX box (2). These structural motifs have been implicated in differential transport, posttranslational processing, and subcellular localization of the different Ras proteins (3, 4). Early knockout data revealed signicant functional differ- ences for the three Ras loci. Whereas H- and N-Ras were dispensable for embryonic development, K-Ras was essential (57). These observations did not establish whether these differences were due to the intrinsic properties of their cognate Ras proteins or their patterns of expression. This issue was solved, at least in part, when Potenza and colleagues replaced mouse K-Ras alleles by chimeric K/H-Ras alleles encoding functional H-Ras proteins (8). These mice developed to adult- hood despite the absence of K-Ras, indicating that the require- ment for K-Ras alleles during embryonic development is pri- marily due to their pattern of expression. Yet, these mice displayed cardiovascular defects, thus raising the possibility that H-Ras and K-Ras proteins might have differential signaling properties, at least in certain tissues (8). RAS genes have also attracted interest due to their involvement in tumor development (1, 2). The overall incidence of each RAS oncogene varies signicantly among tumor types (9). Whereas KRAS is frequently mutated in pancreatic, colorectal, and lung adenocarcinomas, HRAS oncogenes are found in a limited per- centage of tumors from the salivary gland, urinary track, cervix, or skin. On the other hand, NRAS oncogenes are present in mela- nomas and hematopoietic tumors. To date, the molecular basis for this incidence bias is still unresolved (9). Mutant RAS genes also induce different phenotypes when expressed in the germline of patients suffering from RASopathies, a series of developmental defects that result from constitutive activation of RAS signaling pathways (10, 11). Oncogenic muta- tions in HRAS lead to relatively mild developmental defects in Costello syndrome patients (12, 13). In contrast, those KRAS mutations present in human tumors have not been found in RASopathy patients, suggesting that such mutations may cause embryonic lethality (14). 1 Molecular Oncology Programme, Centro Nacional de Investigaciones Oncol ogicas (CNIO), Madrid, Spain. 2 Centro de Investigaci on del C ancer and Instituto de Biología Molecular y Celular del C ancer, CSIC-Universidad de Salamanca, Salamanca, Spain. 3 Centro de Investigaci on Biom edica en Red de C ancer (CIBERONC), Salamanca, Spain. 4 Department of Environmental, Bio- logical and Pharmaceutical Sciences and Technologies (DiSTABiF), Second University of Naples, Caserta, Italy. Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/). Current address for S. Fabbiano: Department of Cell Physiology and Metabolism, Centre M edical Universitaire (CMU) and Diabetes Centre, Faculty of Medicine, University of Geneva, Geneva 1206, Switzerland. Corresponding Authors: Matthias Drosten, Molecular Oncology Programme, Centro Nacional de Investigaciones Oncol ogicas, Melchor Fern andez Almagro 3, Madrid 28029, Spain. Phone: 349-1732-8000; Fax: 349-1732-8033; E-mail: [email protected]; and Mariano Barbacid, [email protected] doi: 10.1158/0008-5472.CAN-16-2925 Ó2016 American Association for Cancer Research. Cancer Research www.aacrjournals.org 707 on February 24, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from Published OnlineFirst November 21, 2016; DOI: 10.1158/0008-5472.CAN-16-2925

H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra … · Molecular and Cellular Pathobiology H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra When Driven by the

  • Upload
    others

  • View
    0

  • Download
    0

Embed Size (px)

Citation preview

Page 1: H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra … · Molecular and Cellular Pathobiology H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra When Driven by the

Molecular and Cellular Pathobiology

H-Ras and K-Ras Oncoproteins Induce DifferentTumor Spectra When Driven by the SameRegulatory SequencesMatthias Drosten1, Lucía Sim�on-Carrasco1, Isabel Hern�andez-Porras1, CarmenG. Lechuga1,María T. Blasco1, Harrys K.C. Jacob1, Salvatore Fabbiano2,3, Nicoletta Potenza4,Xos�e R. Bustelo2,3, Carmen Guerra1, and Mariano Barbacid1

Abstract

Genetic studies in mice have provided evidence that H-Rasand K-Ras proteins are bioequivalent. However, human tumorsdisplay marked differences in the association of RAS oncogeneswith tumor type. Thus, to further assess the bioequivalenceof oncogenic H-Ras and K-Ras, we replaced the coding regionof the murine K-Ras locus with H-RasG12V oncogene sequences.Germline expression of H-RasG12V or K-RasG12V from the K-Raslocus resulted in embryonic lethality. However, expressionof these genes in adult mice led to different tumor phenotypes.Whereas H-RasG12V elicited papillomas and hematopoietic

tumors, K-RasG12V induced lung tumors and gastric lesions.Pulmonary expression of H-RasG12V created a senescence-likestate caused by excessive MAPK signaling. Likewise, H-RasG12V

but not K-RasG12V induced senescence in mouse embryonicfibroblasts. Label-free quantitative analysis revealed thatminor differences in H-RasG12V expression levels led to dras-tically different biological outputs, suggesting that subtledifferences in MAPK signaling confer nonequivalent functionsthat influence tumor spectra induced by RAS oncoproteins.Cancer Res; 77(3); 707–18. �2016 AACR.

IntroductionMammals contain three Ras loci, H-Ras, K-Ras, and N-Ras, that

encode highly related proteins (1–4). Ras proteins are smallGTPases that function as mitogenic switches to control the trans-mission of extracellular signals to the nucleus (1). They shareextensive homology at their N-terminal half, a region involved innucleotide binding and interaction with downstream effectors(2). Their unique features reside in their carboxy-terminal halfthat includes the hypervariable region and a terminal domainknown as the CAAX box (2). These structural motifs have beenimplicated in differential transport, posttranslational processing,and subcellular localization of the different Ras proteins (3, 4).

Early knockout data revealed significant functional differ-ences for the three Ras loci. Whereas H- and N-Ras weredispensable for embryonic development, K-Ras was essential(5–7). These observations did not establish whether thesedifferences were due to the intrinsic properties of their cognateRas proteins or their patterns of expression. This issue wassolved, at least in part, when Potenza and colleagues replacedmouse K-Ras alleles by chimeric K/H-Ras alleles encodingfunctional H-Ras proteins (8). These mice developed to adult-hood despite the absence of K-Ras, indicating that the require-ment for K-Ras alleles during embryonic development is pri-marily due to their pattern of expression. Yet, these micedisplayed cardiovascular defects, thus raising the possibilitythat H-Ras and K-Ras proteins might have differential signalingproperties, at least in certain tissues (8).

RAS genes have also attracted interest due to their involvementin tumor development (1, 2). The overall incidence of each RASoncogene varies significantly among tumor types (9). WhereasKRAS is frequently mutated in pancreatic, colorectal, and lungadenocarcinomas, HRAS oncogenes are found in a limited per-centage of tumors from the salivary gland, urinary track, cervix, orskin. On the other hand, NRAS oncogenes are present in mela-nomas and hematopoietic tumors. To date, the molecular basisfor this incidence bias is still unresolved (9).

Mutant RAS genes also induce different phenotypes whenexpressed in the germline of patients suffering from RASopathies,a series of developmental defects that result from constitutiveactivation of RAS signaling pathways (10, 11). Oncogenic muta-tions in HRAS lead to relatively mild developmental defects inCostello syndrome patients (12, 13). In contrast, those KRASmutations present in human tumors have not been found inRASopathy patients, suggesting that such mutations may causeembryonic lethality (14).

1Molecular Oncology Programme, Centro Nacional de InvestigacionesOncol�ogicas (CNIO), Madrid, Spain. 2Centro de Investigaci�on del C�ancer andInstituto de Biología Molecular y Celular del C�ancer, CSIC-Universidad deSalamanca, Salamanca, Spain. 3Centro de Investigaci�on Biom�edica en Red deC�ancer (CIBERONC), Salamanca, Spain. 4Department of Environmental, Bio-logical and Pharmaceutical Sciences and Technologies (DiSTABiF), SecondUniversity of Naples, Caserta, Italy.

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

Current address for S. Fabbiano: Department of Cell Physiology andMetabolism,Centre M�edical Universitaire (CMU) and Diabetes Centre, Faculty of Medicine,University of Geneva, Geneva 1206, Switzerland.

Corresponding Authors: Matthias Drosten, Molecular Oncology Programme,Centro Nacional de Investigaciones Oncol�ogicas, Melchor Fern�andez Almagro 3,Madrid 28029, Spain. Phone: 349-1732-8000; Fax: 349-1732-8033; E-mail:[email protected]; and Mariano Barbacid, [email protected]

doi: 10.1158/0008-5472.CAN-16-2925

�2016 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org 707

on February 24, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 21, 2016; DOI: 10.1158/0008-5472.CAN-16-2925

Page 2: H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra … · Molecular and Cellular Pathobiology H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra When Driven by the

Similar results have been observed in mouse strains carryinggenetically engineered Ras mutations. Whereas expression ofresident K-Ras oncoproteins in the germline leads to early embry-onic death, expression of an endogenous H-Ras oncogene iswell tolerated and leads to developmental defects very similarto those observed in Costello patients (15–18). Likewise, germ-line expressionof thepartially activatedK-RasV14I-mutant isoformresults in phenotypic defects that closely resemble those of Noo-nan patients (19).

Here, we provide genetic evidence that the wild-type H-Ras andK-Ras proteins are bioequivalent in spite of their different struc-tural and biological properties. We have also compared thespectrum of tumors elicited upon expression of the H-RasG12V

and K-RasG12V oncoproteins from the same mouse K-Ras locus.These studies have revealed that these oncoproteins induce adifferent spectrum of tumors primarily due to differences in theintensity of MAPK signaling that results from subtle differences intheir levels of expression.

Materials and MethodsMouse strains

Generation of K-Rasþ/LSLH-RasG12V mice (KHRasV12) is describedin Supplementary Information. KrasKI (8), HrasKI (8), H-Ras�/�

(5), K-Rasþ/LSLG12Vgeo (KV12; ref. 16), and p53lox/lox (20) geneti-cally engineered strains have been described. hUBC-CreERT2,Elas-tTA, and tetO-Cre transgenic strain have also been reported(21, 22). Activation of the inducible CreERT2 recombinase wasachieved by feeding the mice with a tamoxifen-containing diet(Harlan Laboratories). For activation of H-RasG12V andK-RasG12V expression in lung tissue, mice were infected withAdeno-Cre particles as described previously (23). Trametinib(Mekinist) was purchased from Sellek Chemicals and wasadministered orally daily (1 mg/kg) for 8 weeks. All mice weremaintained in a mixed 129Sv/J x C57BL/6j background andhoused in a barrier facility according to standards establishedby the European Union. All animal experiments were approvedby the CNIO, the Carlos III Health Institute (Madrid, Spain),and the Comunidad de Madrid Ethical Committees and per-formed in accordance with the ARRIVE (Animal Research:Reporting on In Vivo Experiments) guidelines.

Histopathology, IHC, and SA-b-Gal stainingTissues were fixed in 10% buffered formalin and embedded

in paraffin. Hematoxylin and eosin (H&E) staining andimmunohistochemical analyses were performed on 3-mm par-affin sections. For IHC, the following antibodies were used:pErk1/2 (9101, Cell Signaling Technology), Active Caspase-3(MAB835, R&D Systems), SPC (AB3786, Abcam), CC10 (T-18,Santa Cruz Biotechnology), and CD3e (clone 145-2C11,Abcam). Senescence-associated b-galactosidase (SA-b-Gal)activity in cultured mouse embryonic fibroblasts (MEF) aswell as in cryosections of lungs was detected by X-Gal stainingas described previously (24).

Southern and Western blot analysisSouthern blot analysis is described in Supplementary Infor-

mation. Western blot analysis of protein extracts obtained fromtotal lung tissue or MEFs was performed as described previously(25). Primary antibodies used included: Pan-Ras (OP40, MerckMillipore), H-Ras (clone 18, BD Transduction Laboratories),

pErk1/2 (9101), p53 (2524), pAkt (9271), Akt (9272; all fromCell Signaling Technology), p19Arf (clone 5-C3-1, Upstate Bio-technology), Erk1 (C16), p16INK4a (M-156; all from Santa CruzBiotechnology), and GAPDH (G8795, Sigma-Aldrich).

Statistical analysisAll statistical analyses were performed using an unpaired Stu-

dent t test. P values <0.05 were considered to be statisticallysignificant (� P < 0.05, ��� P < 0.001).

ResultsH-Ras and K-Ras are bioequivalent proteins

Germline ablationof K-Ras results in embryonic lethality (6, 7).Yet, expression of H-Ras proteins under the control of K-Ras–regulatory sequences results in viable mice, thus illustrating thatH-Ras can replace K-Ras isoforms, for most biological activities(8). Yet, these mice, designated as HrasKI, displayed dilatedcardiomyopathy and arterial hypertension when they reachedadulthood (8). These cardiovascular defects were initially attrib-uted to the absence of K-Ras proteins in heart tissue. Subsequentstudies, however, revealed that germline expression of constitu-tively active H-RasG12V led to similar cardiovascular defects in amouse model of Costello syndrome (17). Thus, we reasoned thatthese cardiovascular defectsmight bedue toH-Ras overexpressiondue to the presence of four H-Ras–coding alleles (the knocked-inHrasKI alleles and the endogenous H-Ras alleles). To reduce theload of H-Ras expression, we crossed HrasKI mice with H-Ras�/�

animals. HrasKI;H-Ras�/� mice displayed normal heart functionand no hypertension (Fig. 1A and B). In addition, these micedisplayed normal cardiomyocyte areas and did not accumulatefibrosis in their heart (Fig. 1C and D). These results indicate thatthe cardiovascular defects of HrasKI mice were due to H-Rasoverexpression and not to differential activities between H-Rasand K-Ras proteins.

Germline expression of the H-RasG12V oncoprotein from theK-Ras locus results in embryonic lethality

Next, we interrogated whether their oncogenic isoforms,K-RasG12V and H-RasG12V, also have similar properties. To thisend, we knocked in a cDNA encoding an H-RasG12V oncogenewithin the first coding exon of the K-Ras locus (Fig. 2A). We alsoknocked in a lox-STOP-lox (LSL) cassette upstreamof the initiatorcodon to prevent expression of H-RasG12V (Supplementary Fig.S1A–S1C). Expression of the H-RasG12V cDNA clone and a geno-mic DNA fragment containing the four H-Ras coding exonsresulted in similar expression levels, indicating that the H-Rasintronic sequences do not play a significant role in regulatingH-Ras expression (Fig. 2B and C). For simplicity, the wild-typeK-Rasþ/þmice and the targeted K-Rasþ/LSLG12Vgeo andK-Rasþ/LSLH-

RasG12V strains will be referred to hereafter as Kþ, KV12, andKHRasV12, respectively.

Previous studies have indicated that expression of a K-RasG12V

oncogene in the mouse germline results in embryonic lethality(16). In contrast, expression of the H-RasG12V oncogene, even inhomozygosity, is well tolerated during embryonic development(17, 18). To determine whether this differential effect is anintrinsic property of the K-RasG12V and H-RasG12V oncoproteins,we crossed KHRasV12 mice to EIIA-Cre transgenics (26) to expressthe H-RasG12V oncoprotein from the K-Ras locus in the germline.These embryos were no longer viable and died right after implan-tation, a time similar to that observed for embryos expressing an

Drosten et al.

Cancer Res; 77(3) February 1, 2017 Cancer Research708

on February 24, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 21, 2016; DOI: 10.1158/0008-5472.CAN-16-2925

Page 3: H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra … · Molecular and Cellular Pathobiology H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra When Driven by the

endogenous K-RasG12V oncoprotein (16). Hence, the ability ofmice, as well as of Costello syndrome patients, to tolerate expres-sion of an oncogenic H-RasG12V protein during embryonic devel-opment is dictated by the expression pattern of the H-Ras locus.

Expression of H-RasG12V from the K-Ras locus in adult miceNext,we explored theoncogenic properties of theK-RasG12V and

H-RasG12V isoforms expressed from the K-Ras locus in adult mice.To this end,we inserted in theKV12 andKHRasV12 strains a transgeneencoding the inducible CreERT2 recombinase under the control ofthe human ubiquitin C promoter (21). We exposed KV12;hUBC-CreERT2þ/T and KHRasV12;hUBC-CreERT2þ/T mice to a continuoustamoxifen diet to ensure efficient recombination of the targetedK-Ras alleles. Under these conditions, KV12;hUBC-CreERT2þ/T

mice had a median survival of 24 weeks (Fig. 2D). In agreementwith previous studies (27), these mice displayed multiple lesionsin their lungs as well as abundant gastric papillomas (Fig. 2E). Noother tissue displayed significant alterations at the histopathologiclevel. KHRasV12;hUBC-CreERT2þ/T mice submitted to the sametamoxifen treatment died 4 to 5 weeks earlier (Fig. 2D). Thesemice did not develop detectable lesions in either lungs orstomach (Fig. 2E) despite expression of the mutant H-RasG12V

protein in these tissues (Supplementary Fig. S1D). Instead, theydisplayed papillomas on their footpads after approximately 2months of treatment (Supplementary Fig. S1E). Moreover, theyhad enlarged spleens and thymic tumors (Fig. 3). Tamoxifen-treated control Kþ;hUBC-CreERT2þ/T mice did not display

detectable lesions for up to one year of age. These observationsindicate that oncogenic signaling initiated by K-RasG12V andH-RasG12V oncoproteins expressed under the same regulatorysequences has substantially different consequences on tumorformation.

Expression of H-RasG12V from the K-Ras locus in adult miceinduces hematopoietic disorders

Postmortem characterization of tamoxifen-treated KHRasV12;hUBC-CreERT2þ/T mice at humane endpoint revealed enlargedspleens infiltrated with myeloid cell populations in 100% of theanimals (Fig. 3A andB). This phenotypewas not observed inKV12;hUBC-CreERT2þ/T or Kþ;hUBC-CreERT2þ/T mice. Flow cytome-try analyses of these infiltrates revealed a dramatic expansion ofCD11bþ and Gr1þ/CD11bþ double-positive cells indicative ofmyeloproliferative disease (Fig. 3C; Supplementary Fig. S2A). Thisincrease in the myeloid cell compartment was accompanied by aconcomitant decrease in CD3þ T cells and CD19þ B cells (Supple-mentary Fig. S2A). Analysis of committed progenitors in the bonemarrow of mice exposed to the tamoxifen diet for 3 monthsrevealed a significant increase in the commonmyeloid progenitors(CMP; Lin�/IL7Ra�/Sca-1�/c-Kitþ/FcgRlow/CD34þ) as well as aslight expansion of the granulocyte–macrophage progenitor pop-ulation (GMP; Lin�/IL7Ra�/Sca-1�/c-Kitþ/FcgRhigh/CD34þ; Sup-plementary Fig. S2B). The common lymphoid progenitors (CLP;Lin�/IL7Raþ/Sca-1low/c-Kitlow), themegakaryocyte–erythroid pro-genitors (MEP; Lin�/IL7Ra�/Sca-1�/c-Kitþ/FcgRlow/CD34�), and

10

0

20

30

40

50

Eje

ctio

n fra

ctio

n (%

)

BA

Mea

n ar

teria

l pre

ssur

e (m

m H

g)

***

C

100

200

300

400

Car

diom

yocy

te a

rea

(μm

2 )

HrasKI;H-Ras–/–HrasKID

*** ******

100

80

60

110

90

70

Figure 1.

H-Ras and K-Ras proteins are functionally bioequivalent.A,Heart function as determined by percentage of ejection fraction in the left ventricle of KrasKI (n¼ 9; solidbar), HrasKI (n ¼ 10; open bar), or HrasKI;H-Ras�/� (n ¼ 9; gray bar) mice. Data, mean � SD. B, Mean arterial pressure of KrasKI (n ¼ 6; solid bar), HrasKI(n ¼ 7; open bar), or HrasKI;H-Ras�/� (n ¼ 7; gray bar) mice. Data, mean � SD. ��� , P < 0.001 (unpaired Student t test). C, Cardiomyocyte area in the left ventricleof KrasKI (n ¼ 4; solid bar), HrasKI (n ¼ 4; open bar), or HrasKI;H-Ras�/� (n ¼ 5; gray bar) mice. Data, mean � SD. ��� , P < 0.001 (unpaired Student t test).D, Masson trichrome staining for fibrosis in sections obtained from the left ventricle of representative HrasKI and HrasKI;H-Ras�/� mice. Scale bars, 200 mm.

Expression of H-RasG12V from the K-Ras Promoter

www.aacrjournals.org Cancer Res; 77(3) February 1, 2017 709

on February 24, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 21, 2016; DOI: 10.1158/0008-5472.CAN-16-2925

Page 4: H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra … · Molecular and Cellular Pathobiology H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra When Driven by the

LSK cells (Lin�/Sca-1þ/c-Kitþ), a population enriched in hemato-poietic stem cells, did not display significant changes. These obser-vations indicate that H-RasG12V, but not K-RasG12V, is capable ofinducing myeloproliferative disease in adult mice via expansion ofa specific subset of committed progenitors in their bone marrow.

We also detected large tumor masses in the thymus in morethan 85% of KHRasV12;hUBC-CreERT2þ/T animals, a pathologynot observed in KV12;hUBC-CreERT2þ/T or Kþ;hUBC-CreERT2þ/T

mice (Fig. 3D). Histopathologic characterization revealed thepresence of T-cell lymphoblastic lymphoma (T-ALL), as deter-mined by a uniform expansion of CD3þ T lymphocytes. More-over, tumors displayed an increase in double-negative (DN)CD4/CD8�, single-positive CD4þ, and single-positive CD8þ cells (Fig.3E). In particular, when DN cells were further characterized, wedetected elevated DN1 (CD44þ/CD25�) and DN2 (CD44þ/CD25þ) populations (Supplementary Fig. S3A). The characteristichyperactivation of the Notch pathway in T-ALL was also observedas demonstrated by immunostaining of Hes1 (SupplementaryFig. S3B; ref. 28). We also detected abundant lymphocyte infil-trates in a variety of organs, including lung, liver, kidney, and eye

(Supplementary Fig. S3C). These data indicate that in addition tomyeloproliferative disease, most mice expressing the H-RasG12V

oncoprotein from the K-Ras locus also developed T-ALL, a tumortype not induced by the K-RasG12V isoform.

H-RasG12V driven from the K-Ras locus inducespancreatic tumors

Expression of a resident K-RasG12V oncogene during late embry-onic development in pancreatic acinar cells results in pancreaticintraepithelial neoplasias (PanIN lesions) that occasionally prog-ress to pancreatic ductal adenocarcinomas (PDAC; ref. 22). Nosuch lesionswere observed inH-Rasþ/G12V orH-RasG12V/G12Vmicein which the H-RasG12V oncoprotein is expressed from its ownlocus (17). To determine whether H-RasG12V could induce pan-creatic lesions when expressed from the K-Ras locus, we crossedKHRasV12 mice to Elas-tTA/tetO-Cre transgenic animals known toselectively express Cre recombinase in acinar cells from E16.5onwards (22). Analysis of serial pancreatic tissue sections ofKHRasV12;Elas-tTA/tetO-Cre animals at one year of age revealedthe appearance of focal low- and high-grade PanIN lesions

D

C

Hyg0 H-RasG12VK HRasV12

STOPHyg0 1*

G12V

K V12

ATG

ATG

STOP

0 1K +ATG

A B

E

Lung

Sto

mac

h

K + K V12 K HRasV12

*

0

20

40

60

80

100

0 10 20 30 40 50Age (weeks)

Sur

viva

l (%

)H-Ras Genomic DNA

H-Ras cDNA

ATG

ATG

TGA

TGA

1 2 3 4H-Ras

GAPDH

H-Ras–/– Cells

*

*

**

*

K +

K V12

K HRasV12

Figure 2.

Phenotypic consequences of the expression of H-RasG12V and K-RasG12V oncoproteins from the K-Ras locus in adult mice. A, Schematic representation of K-Raswild-type (Kþ), K-RasLSLG12Vgeo (KV12), and K-RasLSLH-RasG12V (KHRasV12) alleles. The H-RasG12V cDNA (blue) was knocked-in into exon 1 of the endogenous K-Ras locus(red) along with transcriptional termination sequences (STOP) and a hygromycin-expressing selection cassette (Hyg) flanked by loxP sequences (closedtriangles) upstream of exon 1, at the same location as in the KV12 allele. The exon 1 encoding the oncogenic G12V mutation is indicated by an asterisk. Thethree alleles share the same ATG initiator codon sequences. B, Schematic representation of the coding exons (1–4) present in the endogenous H-Ras locusand the H-Ras cDNA (cDNA). The noncoding domains of exons 1 and 4 are indicated by open boxes. C,Western blot analysis of H-Ras expression in H-Ras�/�MEFsinfected with empty lentiviruses (vector), lentiviruses expressing the 1.6 kbp DNA fragment encompassing the H-Ras genomic sequences shown in B (genomic),or the H-Ras cDNA (cDNA). GAPDH expression served as a loading control. D, Survival of Kþ;hUBC-CreERT2þ/T (green dots; n ¼ 6), KV12;hUBC-CreERT2þ/T

(red dots; n ¼ 19), or KHRasV12;hUBC-CreERT2þ/T (blue dots; n ¼ 16) mice subjected to a continuous tamoxifen diet at 4 weeks of age (solid arrow). E, H&Estaining of lung and stomach sections obtained from Kþ;hUBC-CreERT2þ/T, KV12;hUBC-CreERT2þ/T, or KHRasV12;hUBC-CreERT2þ/T mice treated for 4 months withtamoxifen diet. Asterisks indicate adenomas (lung) or papillomas (stomach) in KV12;hUBC-CreERT2þ/T animals (KV12). Arrowheads show the normal stratifiedepithelium of the forestomach in Kþ;hUBC-CreERT2þ/T (Kþ) and KHRasV12;hUBC-CreERT2þ/T (KHRasV12) mice (scale bars, 200 mm).

Drosten et al.

Cancer Res; 77(3) February 1, 2017 Cancer Research710

on February 24, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 21, 2016; DOI: 10.1158/0008-5472.CAN-16-2925

Page 5: H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra … · Molecular and Cellular Pathobiology H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra When Driven by the

indistinguishable from those present in KV12;Elas-tTA/tetO-Cremice (Supplementary Fig. S4A). However, the number of lesionswas significantly lower than in similar mice expressing theK-RasG12V oncoprotein (Fig. 4A).

Loss of the tumor suppressor p53 accelerated tumor formationin KHRasV12;Elas-tTA/tetO-Cre animals (22). These mice died ofPDAC tumors, although they survived longer than those miceexpressing the K-RasG12V oncoprotein (30 vs. 17 weeks averagesurvival, a 75% increase; Fig. 4B; Supplementary Fig. S4A). Com-parative analysis of KHRasV12;p53lox/lox;Elas-tTA/tetO-Cre andKV12;p53lox/lox;Elas-tTA/tetO-Cre mice at 10 weeks of age revealedthat the H-RasG12V-expressing animals displayed significantlyfewer PanIN lesions and PDAC tumors than those expressingK-RasG12V (Fig. 4C). These quantitative differences are likely to be

due to a reduction in the number of initiating events, as thenumber of acinar–ductal metaplasias was significantly higher inacinar cell explants of KV12 mice than in those derived fromKHRasV12 animals (Fig. 4D; Supplementary Fig. S4B).

H-RasG12V failed to induce lung adenocarcinomas due toexcessive MAPK signaling

Systemic expression of H-RasG12V from the K-Ras locus didnot yield lung lesions, including hyperplasias or benign ade-nomas. To rule out non–cell-autonomous effects, we infectedKHRasV12 and KV12 mice with Adeno-Cre particles to selectivelyinduce oncogene expression in lung tissue. Six months afterinfection, none of the KHRasV12 mice displayed detectablelesions in their lungs, whereas all KV12 animals had developed

A B

C

D

K + K V12 K HRasV12K + K V12 K HRasV12

Gr1

CD11b

6.10 2.48

1.57

4.87 6.85

2.35

7.23 51.9

4.43102

103

104

105

102 103 104

K + K V12 K HRasV12

102 103 104 102 103 104

H&

EC

D3

IHC

H&

E

K + K HRasV12E

CD

4

K HRasV12

K +

104

103

0

78.2

4.60 2.45

CD8

102

0 102 103 104

25.5

16.4 16.4

104

103

0

102

14.7

41.6

89.9 85.9 36.5

Figure 3.

H-RasG12V, but not K-RasG12V, induceshematopoietic tumors in adult mice. A,Representative images of spleens obtained fromKþ;hUBC-CreERT2þ/T, KV12;hUBC-CreERT2þ/T,or KHRasV12;hUBC-CreERT2þ/T mice treated for4 months with tamoxifen diet. B, H&E staining ofspleen sections obtained from Kþ;hUBC-CreERT2þ/T, KV12;hUBC-CreERT2þ/T, or KHRasV12;hUBC-CreERT2þ/T mice treated for 4 monthswith tamoxifen diet. White and red pulp areindicated by arrows and arrowheads,respectively (scale bars, 200 mm). C, Flowcytometry analysis of Gr1þ and CD11bþ cells inspleens obtained from Kþ;hUBC-CreERT2þ/T,KV12;hUBC-CreERT2þ/T, or KHRasV12;hUBC-CreERT2þ/T mice treated for 4 months withtamoxifen diet. D, H&E (top and middle) andCD3 immunohistochemical staining (bottom) inthymus sections obtained from Kþ;hUBC-CreERT2þ/T or KHRasV12;hUBC-CreERT2þ/T micetreated for 4 months with tamoxifen diet. Scalebars, 5 mm (top), 25 mm (middle, bottom).E, Flow cytometry analysis of CD4þ and CD8þ

cells in thymuses obtained from Kþ;hUBC-CreERT2þ/T or KHRasV12;hUBC-CreERT2þ/T micetreated for 4 months with tamoxifen diet.

Expression of H-RasG12V from the K-Ras Promoter

www.aacrjournals.org Cancer Res; 77(3) February 1, 2017 711

on February 24, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 21, 2016; DOI: 10.1158/0008-5472.CAN-16-2925

Page 6: H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra … · Molecular and Cellular Pathobiology H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra When Driven by the

multiple lesions, including lung adenocarcinomas (Supple-mentary Fig. S4C and S4D).

Tamoxifen treatment of KV12;hUBC-CreERT2þ/T andKHRasV12;hUBC-CreERT2þ/T mice for short periods of timeallowed us to analyze the immediate events that followedexpression of the K-RasG12V and H-RasG12V oncoproteins inlung tissue. Four days after tamoxifen treatment, H-RasG12V

induced phosphorylation of the downstream Erk1/2 kinases(pErk1/2) in more than 20% of the cells, whereas expression ofK-RasG12V only resulted in pErk1/2 immunostaining in lessthan 5% of the cells (Fig. 5A). Similar results were obtained 7days after treatment. These findings were not due to a differ-ential number of cells expressing the K-RasG12V and H-RasG12V

oncoproteins as the levels of Cre-mediated recombination inthe lungs of KV12;hUBC-CreERT2þ/T and KHRasV12;hUBC-CreERT2þ/T mice were similar (Supplementary Fig. S4E andS4F). More importantly, when mice were analyzed 2 weeks aftertamoxifen treatment, the number of pErk1/2-expressing cells inthe lungs of KHRasV12 mice decreased dramatically, whereasthose present in KV12 lungs displayed a modest increase (Fig.5A). These results were further confirmed by Western blotanalysis (Fig. 5B). These observations were selective for lungcells as we did not observe significantly differential numbers ofpErk-expressing cells in other tissues with the possible excep-tion of cells in the basal layer of the skin and in the white pulpof the spleen (Supplementary Fig. S5A). To determine whether

these results were due to differential expression levels of thetwo mutant proteins, we compared their relative levels ofexpression in the lungs of KV12;hUBC-CreERT2þ/T andKHRasV12;hUBC-CreERT2þ/T mice by a label-free quantificationapproach (29). This technique allowed us to determine theirrelative amounts based on the detection of the shared peptide(6-LVVVGAVGVGK-16) in which the underlined residue corre-sponds to the activating valine (Supplementary Fig. S5B). Asillustrated in Fig. 5C, H-RasG12V is expressed at about 5-foldhigher levels than K-RasG12V. The higher levels of H-RasG12V

expression also resulted in increased levels of total GTP-boundactive Ras complexes as determined by RBD pull-down assays(Fig. 5D).

Next, we examined whether the increased levels of GTP-boundH-RasG12V may have an effect on the proliferation of H-RasG12V-expressing cells that might prevent the appearance of hyperplasticlesions. To this end, we used the recombinant K-RasH-RasG12V

allele as a molecular marker for the presence of H-RasG12V-expressing lung cells. Exposure of 4-week-old KHRasV12;hUBC-CreERT2þ/T mice to a tamoxifen diet for 4 weeks resulted in theeffective recombination of the K-RasLSLH-RasG12V allele, thus indi-cating that a significant fraction of lung cells expressed the H-RasG12V oncoprotein (Fig. 6A). However, when these KHRasV12;hUBC-CreERT2þ/T mice weremaintained in a diet lacking tamox-ifen for an additional 8-weekperiod, those cells that contained therecombined K-RasH-RasG12V allele diagnostic of H-RasG12V

C

A

D

B

0

10

100

1

Lesi

ons

per

mou

se

K V12 K HRasV12

0

10

1

Lesi

ons

per

mou

se

K V12;p53 lox/lox K HRasV12;p53 lox/lox

0

100

200

300

– EGF

Met

apla

sias

/wel

l

TGFα EGF+TGFα

***

******

K V12

K HRasV12400

0

20

40

60

80

100

0 10 20 30 40 50Age (weeks)

Sur

viva

l (%

)

K HRasV12;p53 lox/lox

K V12;p53 lox/lox

Figure 4.

K-RasG12V and H-RasG12V oncoproteins expressed from the K-Ras locus induce pancreatic lesions. A, Number of low (P1) and high (P2/3) grade PanINs aswell as PDACs per mouse in 1-year-old KV12;Elas-tTA/tetO-Cre (red circles, n ¼ 13) or KHRasV12;Elas-tTA/tetO-Cre (blue circles; n ¼ 10) mice. Data, mean(horizontal bars) � SD. B, Survival of KV12;p53lox/lox;Elas-tTA/tetO-Cre (KV12;p53lox/lox; red dots; n ¼ 22) or KHRasV12;p53lox/lox;Elas-tTA/tetO-Cre mice(KHRasV12;p53lox/lox; blue dots; n ¼ 19). C, Number of low (P1) and high (P2/3) grade PanINs as well as PDACs per mouse in 10 week-oldKV12;p53lox/lox;Elas-tTA/tetO-Cre (red circles; n ¼ 9) or KHRasV12;p53lox/lox;Elas-tTA/tetO-Cre (blue circles; n ¼ 8) mice. Data, mean (horizontal bars) � SD.D,Acinar cell explants isolated from pancreata of 6- to 8-week-old KV12;Elas-tTA/tetO-Cre (KV12; red bars; n¼ 6) and KHRasV12;Elas-tTA/tetO-Cre (KHRasV12; blue bars;n ¼ 7) mice were incubated in the absence or presence of the indicated growth factors. The number of metaplasias was determined after 5 days in culture.Data, mean � SD. ��� , P < 0.001 (unpaired Student t test).

Drosten et al.

Cancer Res; 77(3) February 1, 2017 Cancer Research712

on February 24, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 21, 2016; DOI: 10.1158/0008-5472.CAN-16-2925

Page 7: H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra … · Molecular and Cellular Pathobiology H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra When Driven by the

expression completely disappeared (Fig. 6A). These results wereselective for lung cells as the K-RasH-RasG12V recombined alleleremains present in other tissues, such as thymus, after the 8-weekperiod in theabsenceof tamoxifen (Fig. 6A). Todeterminewhetherthe disappearance of the lung cells was due to excessive H-RasG12V

signaling, we treated KHRasV12;hUBC-CreERT2þ/T mice with a non-limiting dose (1 mg/kg) of trametinib, a MEK inhibitor known toeffectively block MAPK signaling driven by Ras oncogenes (30).Trametinibwasprovidedduring the8weeks inwhichmicewerenolonger exposed to tamoxifen. As illustrated in Fig. 6A, these miceretained the K-RasH-RasG12V recombined allele, indicating thatexpressionof theH-RasG12V oncoproteinwasno longer deleteriousfor lung cells in the presence of the MEK inhibitor.

H-RasG12V expression in lung cells induces a senescence-likearrest partially mediated by p53

The disappearance of H-RasG12V-expressing lung cells was notdue to apoptosis, as we could not detect increased levels of activecaspase-3 in the lungs of tamoxifen-treated KHRasV12 mice (Sup-

plementary Fig. S5C). Likewise, we did not detect SA-b-Galexpression, a standard marker for oncogene-induced senescence(OIS). Yet, we observed the induction of other senescence mar-kers, such as p16Ink4a and p19Arf (Figs. 5B and 6B). These resultssuggest that H-RasG12V expression may induce some sort ofnoncanonical senescence-like state. Finally, we interrogatedwhether this phenomenon could be mediated by p53. To thisend, we introduced p53lox alleles into KV12 and KHRasV12 mice. Asillustrated in Supplementary Fig. S6A and S6B, KHRasV12;p53lox/lox

mice developed some hyperplasias and occasional adenomas in30% of the animals, all of which stained positive for type IIalveolar cell markers (Supplementary Fig. S6C). These resultsindicate that induction of a senescence-like phenotype in lungtissue by expression of the H-RasG12V oncoprotein from the K-Raslocus is only partially mediated by p53.

H-RasG12V expressed from the K-Ras locus induces OIS in MEFsFinally, we decided to compare the effect of expressing the

K-RasG12V andH-RasG12V oncoproteins from the sameK-Ras locus

A pERK IHC

0

10

20

30

pER

K+

Cel

ls (%

)

*** ***

Days in tamoxifen diet

Day 0 Day 4 Day 7 Day 14

K V12

K HRasV12

B

C D

pErk1/2

H-Ras

GAPDH

Erk1/2

p16Ink4a

Days in tamoxifen diet

p19Arf

KV12 K HRasV12

Akt

pAkt

K V12 K HRasV12

14Days in tamoxifen diet0 14 0

GAPDH

Ras-GTP

Ras

4 7 140 4 7 14 0

40

Nor

mal

ized

inte

nsity

(×10

7 )

0

0.8

1.2

0.4

K V12 K HRasV12 K+

K V

12K

HR

asV

12

Figure 5.

H-RasG12V expressed from the K-Ras locus induces robust downstream signaling in lung tissue.A, Top, immunohistochemical staining for pErkþ cells in lung sectionsobtained from KV12;hUBC-CreERT2þ/T or KHRasV12;hUBC-CreERT2þ/T mice subjected to tamoxifen diet for the indicated time (scale bars, 250 mm); bottom,quantification of the percentage of pERKþ cells in the lung sections shown above obtained from KV12;hUBC-CreERT2þ/T (KV12; red bars; n ¼ 3) or KHRasV12;hUBC-CreERT2þ/T (KHRasV12; blue bars; n ¼ 3) mice. Data, mean � SD. ��� , P < 0.001 (unpaired Student t test). B, Western blot analysis of H-Ras, pErk1/2, Erk1/2,pAkt, Akt, p16Ink4a, and p19Arf expression in total lung extracts obtained from KV12;hUBC-CreERT2þ/T or KHRasV12;hUBC-CreERT2þ/T mice exposed to tamoxifendiet for the indicated time. GAPDH expression served as a loading control. C, Relative quantification of the levels of expression of K-RasG12V and H-RasG12V

oncoproteins by label-free quantification in lungs obtained from Kþ;hUBC-CreERT2þ/T (n¼ 6), KV12;hUBC-CreERT2þ/T (n¼ 6), or KHRasV12;hUBC-CreERT2þ/T (n¼ 6)mice treated for 2 months with tamoxifen diet. The tryptic peptide LVVVGAVGVGK was used to detect expression of both K-RasG12V and H-RasG12V proteins.Data, mean � SD. D, Total Ras-GTP levels in lungs obtained from KV12;hUBC-CreERT2þ/T or KHRasV12;hUBC-CreERT2þ/T mice treated with tamoxifen dietfor the indicated time. GAPDH was used as a loading control.

Expression of H-RasG12V from the K-Ras Promoter

www.aacrjournals.org Cancer Res; 77(3) February 1, 2017 713

on February 24, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 21, 2016; DOI: 10.1158/0008-5472.CAN-16-2925

Page 8: H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra … · Molecular and Cellular Pathobiology H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra When Driven by the

inMEFs. Previous studies have shown that overexpressionof theseoncoproteins in MEFs readily induced OIS (31). However, whenthey were expressed from their own endogenous promoters,they failed to induce OIS (16, 17). We induced expression ofeither K-RasG12V or H-RasG12V oncoproteins by infecting immor-talized KV12 and KHRasV12 MEFs with Adeno-Cre particles, respec-tively. As expected, K-RasG12V expression did not cause significantchanges indownstream signaling (Fig. 7A). In contrast, expressionof H-RasG12V resulted in a sustained increase in the phosphory-lation of the downstream substrates pErk and pAkt (Fig. 7A).Moreover, whereas K-RasG12V expression had nodetectable effectson proliferation, expression of H-RasG12V effectively inducedOIS,leading to complete cessation of cell proliferation and SA-b-Galexpression (Fig. 7B and C).

H-RasG12V expressing MEFs displayed increased levels of p53(Fig. 7A). Yet, efficient depletionof p53by short hairpinRNAshadno effect on cell-cycle arrest or induction of senescence, indicatingthat abrogation of p53 expression was not sufficient to overcomeH-RasG12V–inducedOIS (Supplementary Fig. S7AC). Likewise,weobserved no significant increase in the expression levels ofp16INK4a. However, ectopic expression of adenoviral E1A, anoncoprotein known to target the p53 and Rb pathways (32), wasable to bypass OIS and to restore efficient proliferation of MEFs(Fig. 7B and C; Supplementary Fig. S7D). Furthermore, E1Acooperated with H-RasG12V, but not with K-RasG12V, to transformimmortal MEFs (Fig. 7D; Supplementary Fig. S7E). In an effort toprovide a mechanistic explanation for the dramatic differentialeffects induced by the K-RasG12V versus the H-RasG12V oncopro-teins when expressed from the same locus, we established their

relative levels of expression using label-free quantification tech-niques. As illustrated in Fig. 7E, H-RasG12V was more efficientlyexpressed (about 2.5-fold) than K-RasG12V.

To determine whether these minor differences might beresponsible for the drastically differential outputs, we decidedto explore how differential expression levels of the same onco-protein, H-RasG12V, affected the biological behavior of MEFs.Whereas H-RasG12V expressed from its own promoter had noeffect of MEF properties, including immortalization, prolifer-ation, or senescence (17), H-RasG12V expression from the K-Raspromoter in KHRasV12 led to rapid OIS (Fig. 7B and C). Westernblot analysis revealed that the amount of H-Ras protein (bothH-Ras and H-RasG12V) in KHRasV12 MEFs was 3-fold higher thanin H-RasG12V/G12V MEFs (Fig. 7F). As KHRasV12 MEFs also expressa wild-type H-Ras protein from its endogenous locus, theamount of H-RasG12V expressed from the K-Ras locus inKHRasV12 MEFs compared with that expressed from its ownlocus in H-RasG12V/G12V MEFs is a mere 2-fold higher (1.9 �0.7) than that expressed from its own locus. These observationsestablish that relatively subtle increases in the levels of expres-sion of the H-RasG12V oncoprotein can result in dramaticallydifferent biological consequences.

DiscussionEarly genetic studies revealed that the three Ras loci have

different biological properties. Whereas K-Ras is essential forembryonic development, H-Ras and N-Ras are dispensable(5–7). To determine whether these differences are due to the

A

B

ThymusLung

Off tamoxifen: 8 wk

kbp

7.4

6.0K-Ras LSLHRasV12

K-Ras HRasV12

Trametinib: –– + ––8 wk8 wk ––

Rel

ativ

e m

RN

A

expr

essi

on le

vels

0

2

p19Arfp16Ink4a

1

3

4***

*** K V12

K HRasV12

tamoxifen: 4 wk4 wk4 wk 4 wk4 wk

Figure 6.

H-RasG12V expressed from the K-Ras locus induces a senescence-like arrest in lung tissue that results in subsequent elimination of cells. A, Southern blot analysis ofDNA isolated from lung tissue obtained from 4-week-old KHRasV12;hUBC-CreERT2þ/T mice exposed to a tamoxifen diet for 4 weeks either immediately ending thetreatment (left) or after mice were maintained for 8 additional weeks in a regular diet (center and right). The latter mice were either left untreated during theadditional 8-week period (center) or treated daily with 1 mg/kg of trametinib, a selective MEK inhibitor (right). DNA isolated from thymus tissue was used as control.The migration (open arrowheads) and sizes (solid arrowheads) of the diagnostic SphI þ KpnI DNA fragments for the recombined K-RasH-RasG12V allele thatexpresses theH-RasG12V oncoprotein and the nonrecombinedK-RasLSLH-RasG12V allele that does not allowH-RasG12V expression are indicated. Note the disappearanceof the recombined K-RasH-RasG12V allele in lung but not thymus tissue. Lung tissue frommice treatedwith trametinib also retains the recombined K-RasH-RasG12V allele.B, Relative expression levels of p16Ink4a and p19ArfmRNAs in total lung extracts from KV12;hUBC-CreERT2þ/T (KV12; red bars; n ¼ 3) or KHRasV12;hUBC-CreERT2þ/T

(KHRasV12; blue bars; n ¼ 3) mice exposed to tamoxifen diet for the indicated time, as determined by qRT-PCR analysis. GAPDH expression levels were used fornormalization. Data, mean � SD. ��� , P < 0.001 (unpaired Student t test).

Drosten et al.

Cancer Res; 77(3) February 1, 2017 Cancer Research714

on February 24, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 21, 2016; DOI: 10.1158/0008-5472.CAN-16-2925

Page 9: H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra … · Molecular and Cellular Pathobiology H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra When Driven by the

intrinsic properties of the different Ras protein isoforms or theirpattern of expression, Potenza and colleagues modified theendogenous K-Ras alleles so they direct the synthesis of H-Rasinstead of K-Ras proteins (8). Thesemice developed to adulthood,indicating thatH-Ras could effectively compensate for the absenceof the K-Ras proteins. Yet, these mice displayed cardiovasculardefects, including dilated cardiomyopathy associatedwith arterialhypertension, attributed to the lack of K-Ras proteins in hearttissue (8). However, as illustrated in this study, these cardiovas-cular defectswere adirect consequence of theoverexpressionofH-Ras proteins, as these mice carry four H-Ras–expressing alleles,two endogenous H-Ras alleles, and two HrasKI alleles. Indeed,elimination of the endogenous H-Ras alleles completely pre-vented these cardiovascular defects. Moreover, HrasKI;H-Ras�/�

mice appear to be completely normal, indicating that the H-Rasand K-Ras proteins are fully bioequivalent, at least under normal

homeostatic conditions. These observations are surprising con-sidering the different properties of theH-Ras andK-Ras proteins insubcellular transport, posttranslational processing, and localiza-tion within the plasma membrane (3, 4). Whether their differ-ential properties may play a role under stress conditions remainsto be determined.

The intense focus on RAS biology is due, at least in part, to theirinvolvement in human cancer. Over the years, scientists have beenbaffled by the distinct incidence of the various RAS oncogenes intumors (9). In an attempt to shed some light on this issue, weexpanded the early work of Potenza and colleagues (8), by gen-erating anewconditionalmouse strain,KHRasV12, that expresses theH-RasG12V oncoprotein from the endogenous K-Ras locus. Germ-line expression of H-RasG12V from its own locus had no significanteffect on embryonic development. Yet, these animals displayedfacial and cardiovascular defects reminiscent of patients with

p16Ink4a

p53

pErk1/2

H-Ras

0 2 4 6 8 10Adeno-Cre (days)

Erk1/2

Akt

pAkt

GAPDH

0 2 4 6 8 10

K-RasH/N-Ras

AE

1AV

ecto

r

Days

Absorbance

0

8

16

20

01086420 86420 10

40

K V12 K HRasV12B

Foci

/pla

te

GAPDH

Akt

Erk1/2

pAkt

pErk1/2

H-Ras

Days

0 2Adeno-Cre:

F

K V12 K HRasV12 C

0

10

20

30

0

25

50

75

SA

-β-G

al+

Cel

ls (

%)

100

V E1A V E1A

K V12 K HRasV12 K V12 K HRasV12

V E1A E1AV

Nor

mal

ized

inte

nsity

107 )

D

E K V12 K HRasV12

0

3

1

2

Figure 7.

H-RasG12V, but not K-RasG12V, expressed from the K-Ras locus induces senescence in MEFs. A, Western blot analysis of the indicated proteins expressed in KV12 orKHRasV12 MEFs infected with Adeno-Cre particles for the indicated time. GAPDH expression served as a loading control. B, Growth curve of KV12 or KHRasV12 MEFsinfected with Adeno-GFP (open circles) or Adeno-Cre particles (closed circles) for the indicated time after stable infection with empty retroviruses (vector) orretroviruses expressing the Ad5 E1A oncoprotein (E1A). Data, mean � SD. C, Percentage of SA-b-Gal–positive cells in KV12 or KHRasV12 MEFs stably infectedwith empty retroviruses (V) or retroviruses expressing theAd5 E1A oncoprotein (E1A) 4 days after infectionwith Adeno-Cre particles. Data,mean� SD. ��� , P <0.001(unpaired Student t test). D, Focus formation in KV12 or KHRasV12 MEFs stably infected with empty retroviruses (V) or retroviruses expressing Ad5 E1A (E1A) 14 daysafter infection with Adeno-Cre particles. Data, mean � SD. ��� , P < 0.001 (unpaired Student t test). E, Relative quantification of the levels of expression of theK-RasG12V and H-RasG12V oncoproteins by label-free quantification in KV12;hUBC-CreERT2þ/T (red bars; n ¼ 3) or KHRasV12;hUBC-CreERT2þ/T MEFs (blue bars;n ¼ 3) 4 days after infection with Adeno-Cre particles. The tryptic peptide LVVVGAVGVGK was used to detect expression of K-RasG12V and H-RasG12V proteins.Data, mean� SD. F,Western blot analysis of H-Ras protein expression (H-Ras and H-RasG12V) driven by the K-Ras locus in KHRasV12 MEFs upon infectionwith Adeno-Cre particles for the indicated time and by the H-Ras locus in H-RasG12V MEFs. Expression of Ras effector proteins pErk1/2, Erk1/2, pAkt, and Akt was alsoanalyzed. GAPDH expression served as a loading control.

Expression of H-RasG12V from the K-Ras Promoter

www.aacrjournals.org Cancer Res; 77(3) February 1, 2017 715

on February 24, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 21, 2016; DOI: 10.1158/0008-5472.CAN-16-2925

Page 10: H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra … · Molecular and Cellular Pathobiology H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra When Driven by the

Costello syndrome (17, 18). However, germline expression ofH-RasG12V from the endogenous K-Ras locus led to early embry-onicdeathat a time similar to that observed for embryos expressingan endogenous K-RasG12V oncoprotein (16). Thus, expression ofK-Ras and H-Ras oncoproteins from the K-Ras locus is equallydeleterious for embryonic development. These observations illus-trate that the viability of H-Rasþ/G12V and H-RasG12V/G12V micemust be due to a more restricted pattern of expression, lowerexpression levels or a combination of both (17, 18).

Expression of K-RasG12V in embryonic acinar cells resulted inthe formation of low- and high-grade PanIN lesions andoccasional PDAC tumors (24). Expression of the H-RasG12V

oncoprotein from the K-Ras locus also resulted in the formationof low- and high-grade PanIN lesions albeit at lower frequen-cies, with no detectable PDAC tumors, at least by one yearof age. Whether the high-grade PanINs will eventually progressto yield PDAC tumors in older animals remains to be deter-mined. Recent studies have indicated that K-RasG12V, but notH-RasG12V, promotes tumor formation by suppressing nonca-nonical Wnt signaling (33). This property was used to repressthe transforming activity of K-RasG12V in pancreatic xenografttumor models with prostratin, a PKC activator (33). It will beinteresting to explore whether this natural product has differ-ential activity in pancreatic lesions driven by K-RasG12V andH-RasG12V oncoproteins.

Ubiquitous expression of H-RasG12V in adult KHRasV12 mice ledto a variety of tumors, mainly papillomas and hematopoieticmalignancies. More importantly, the pattern of tumors observedin these mice was significantly different to that present in KV12

animals, thus indicating that H-RasG12V and K-RasG12V oncopro-teins have different transforming capabilities in different tissues.Tamoxifen-treated adult KHRasV12;hUBC-CreERT2þ/T mice devel-oped hematologic malignancies not observed in KV12;hUBC-CreERT2þ/T animals, including myeloproliferative disease, a dis-ease that resulted from expansion of the CMP population, as wellas T-ALL. Previous studies have reported that expression of aresident K-RasG12D oncogene in the hematopoietic compartmentcaused fatal myeloproliferative disease (34, 35) as well as T-ALLupon bone marrow transplantation (36). In these studies, how-ever, K-RasG12D expression was induced by anMx1-Cre transgeneknown to express low levels of Cre recombinase activity duringembryonic development. Thus, it is possible that these differentresultsmight be due to the differential susceptibility of embryonichematopoietic precursors to K-Ras oncoproteins.

On the contrary, some tissues are susceptible to K-RasG12V,but not to H-RasG12V

–induced neoplastic lesions, such as thelung alveoli and the stomach epithelium. The lack of lungtumors in KHRasV12 mice appears to be a consequence of over-activation of the MAPK signaling cascade by the H-RasG12V

oncoprotein. Indeed, H-RasG12V expression in type II alveolarcells of KHRasV12 mice induced a significantly more robustphosphorylation of the Erk kinases than that observed uponinduction of the K-RasG12V oncoprotein in KV12 animals. Thisincreased signaling induced a noncanonical, senescence-likestate that prevented proliferation of the H-RasG12V–expressinglung alveolar cells. Ablation of p53 in KHRasV12 mice resulted inlimited induction of hyperplastic lesions along with few ade-nomas, but not in overt lung tumor development, thus indi-cating that the senescence-like state was only partially mediatedby p53. These senescent H-RasG12V

–expressing cells are shortlived as they could no longer be detected 8 weeks after the

induction of H-RasG12V expression. Likewise, IHC examinationof lung tissue of KHRasV12 mice 2 weeks after the induction of H-RasG12V expression revealed a drastic reduction in the numberof pErk-positive cells. The fate of these cells remains unclear,although the absence of active caspase-3 expression suggeststhat they did not undergo apoptosis. Finally, exposure ofKHRasV12 mice to the MEK inhibitor trametinib prevented elim-ination of the H-RasG12V–expressing cells, thus providing fur-ther evidence that the senescence-like state of these H-RasG12V

expressing lung cells was due to excessive MAPK signaling.It has been proposed that the abundance of Ras isoforms can be

affected by differences in protein translation efficiency (37). K-RasmRNA is less efficiently translated than H-Ras due to the prefer-ential usage of rare codons (37). Germline replacement of rarecodons in the K-Ras locus resulted in mice that expressed higherlevels of K-Ras proteins (38). Interestingly, these mice displayedincreased resistance to urethane-mediated carcinogenesis, sug-gesting that increased levels of K-Ras oncoproteins had adverseeffects on lung tumorigenesis (38). Whether these observationswere due to the induction of a senescence-like state as describedhere for the H-RasG12V oncoprotein remains to be determined.Precise quantitative analysis of the relative amounts of H-RasG12V

and K-RasG12V proteins in lung tissue revealed that H-RasG12V waspresent at 5-fold higher levels than K-RasG12V. Thus, it is possiblethat these oncoproteinsmay have similar signaling properties andtheir differential effects in lung tissue could be primarily due toquantitative differences in their levels of expression. Whether thedifferential results obtained in hematopoietic cells in which onlyH-RasG12V was capable of inducing tumors is also due to quan-titative differences remains to be determined. For instance, it ispossible that the hematopoietic precursors responsible for initi-ating myeloproliferative disease and T-ALL may require higherlevels of MAPK signaling, such as those provided by H-RasG12V

expression. Yet, at this time, we cannot rule out that theH-RasG12V

and K-RasG12V oncoproteins may utilize differential signalingpathways in these hematopoietic cells.

Previous studies have illustrated the presence of H-Ras onco-genes in lung tumors of HrasKI mice exposed to urethane (39).Indeed, these mice developed more lung tumors than wild-typeanimals exposed to the same carcinogenic insult, a resultattributed to the frequent activation of H-Ras oncogenes(39). This apparent conundrum could be explained by thepresence of lower levels of H-Ras protein in HrasKI mice.Indeed, HrasKI is a chimeric allele made of H-Ras and K-Rassequences that contains K-Ras–derived rare codons in two ofthe four coding exons. This fact may result in limited transla-tion efficiency (9, 36, 37). Thus, HrasKI mice may express lowerlevels of H-Ras as compared with those present in KHRasV12

animals that exclusively use H-Ras–derived codons. Alterna-tively, urethane may induce mutations and/or epigenetic altera-tions that prevent the development of the senescence-like stateinduced by H-RasG12V expression in KHRasV12 mice.

Finally, in vitro studies also provided strong support for adirect relationship between H-RasG12V- and K-RasG12V–inducedMAPK signaling and biological output. Expression of a residentK-RasG12V oncoprotein in primary MEFs led to immediateimmortalization bypassing the replication-induced senescencecharacteristic of wild-type cells (16). In contrast, expression ofthe H-RasG12V oncoprotein from the K-Ras locus resulted incanonical OIS leading to complete cessation of cell prolifera-tion and expression of SA-b-Gal. This senescence phenotype

Drosten et al.

Cancer Res; 77(3) February 1, 2017 Cancer Research716

on February 24, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 21, 2016; DOI: 10.1158/0008-5472.CAN-16-2925

Page 11: H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra … · Molecular and Cellular Pathobiology H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra When Driven by the

was accompanied by a robust increase in the phosphorylationof Erk1/2 and Akt. Quantitative analysis of the relative levelsof expression of the H-RasG12V and K-RasG12V oncoproteins inKHRasV12 and KV12 MEFs, respectively, revealed 2.5-fold higherlevels of expression of H-RasG12V, thus raising the possibilitythat differences other than levels of expression may account forthe differential biological outputs induced by these oncopro-teins. However, comparative analysis of the levels of expressionof H-RasG12V in KHRasV12 MEFs, which induces irreversible OISversus H-RasG12V/G12V MEFs in which there are no significantbiological changes, revealed a meager 2-fold difference in itslevels of expression. Thus, minor changes in the levels ofexpression of the H-RasG12V oncoprotein can induce signifi-cantly different outputs. These results, taken together, under-score the need to better understand the molecular mechanismsthat regulate the intensity of MAPK signaling to control thedetrimental effects induced by Ras oncoproteins during neo-plastic development.

It is difficult to reckon why the wild-type H-Ras and K-Rasisoforms are bioequivalent in spite of their differential propertiesrelating to cellular trafficking, posttranslational processing, andsubcellular localization, whereas their oncogenic counterpartsdisplay differential transforming properties. It is conceivable thatcells tolerate an ample range of Ras signaling during normalhomeostasis providing that the differential signaling intensitiesbetween the different Ras isoforms stay below a critical thresh-old. In contrast, elevated signaling induced by Ras oncopro-teins may activate "emergency" signals that either result in theactivation of negative feedback loops or in the induction ofOIS, ultimately leading to cell death. Activation of OIS and/orfeedback loops might be cell type dependent, leading to themanifestation of different phenotypes ranging from senescenceto malignant transformation. Finally, it should be noted thatour observations do not eliminate the possibility that differentoncogenic Ras isoforms may also engage differential signalingpathways. Understanding the differential outputs of H-Ras andK-Ras oncogenes in different cell types will require a moreprecise knowledge of the molecular mechanisms that controltheir key effector pathways.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Authors' ContributionsConception and design: M. Drosten, X.R. Bustelo, C. Guerra, M. BarbacidDevelopment of methodology: M. Drosten, M.T. Blasco, H.K.C. Jacob,X.R. Bustelo, M. BarbacidAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): M. Drosten, L. Sim�on-Carrasco, C.G. Lechuga,S. Fabbiano, N. Potenza, X.R. BusteloAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis):M.Drosten, L. Sim�on-Carrasco, I. Hern�andez-Porras,M.T. Blasco, H.K.C. Jacob, S. Fabbiano, X.R. Bustelo, C. Guerra, M. BarbacidWriting, review, and/or revision of the manuscript:M. Drosten, X.R. Bustelo,C. Guerra, M. BarbacidStudy supervision: M. Drosten, X.R. Bustelo, C. Guerra, M. BarbacidOther (designed and performed experimental work): I. Hern�andez-PorrasOther (mass spectrometric analyses): H.K.C. Jacob

AcknowledgmentsWe thank Alan Balmain (UCSF) for stimulating discussions and for

critically reading the manuscript. We also thank Marta San Roman, RaquelVillar, Beatriz Jim�enez, Nuria Cabrera, Patricia Villanueva, and JenniferCondo for excellent technical assistance. We value the support of JuanAntonio C�amara and Francisca Mulero (Molecular Imaging Core Unit,CNIO) for the echocardiographic analyses, Alba de Martino (HistopathologyCore Unit, CNIO) for her help with mouse histopathology, as well asEduardo Zarzuela and Javier Mu~noz (Proteomics Core Unit, CNIO) formass spectrometric analyses.

Grant SupportThis work was supported by grants from the European Research

Council (ERC-AG/250297-RAS AHEAD), EU-Framework Programme(LSHG-CT-2007-037665/CHEMORES, HEALTH-F2-2010-259770/LUNG-TARGET, and HEALTH-2010-260791/EUROCANPLATFORM), SpanishMinistry of Economy and Competitiveness (SAF2011-30173 andSAF2014-59864-R), Autonomous Community of Madrid (S2011/BDM-2470/ONCOCYCLE to M. Barbacid), Castilla-Le�on Autonomous Govern-ment (CSI101U13, BIO/SA01/15), the Spanish Ministry of Economy andCompetitiveness (SAF2012-31371, RD12/0036/0002), Worldwide CancerResearch (14-1248), the Sol�orzano Foundation, and the Ram�on ArecesFoundation (X.R. Bustelo). Spanish government–sponsored funding toX.R. Bustelo is partially supported by the European Regional Develop-ment Fund. M. Barbacid is the recipient of an Endowed Chair from theAXA Research Fund.

The costs of publication of this articlewere defrayed inpart by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received October 27, 2016; accepted October 28, 2016; published Online-First November 21, 2016.

References1. Malumbres M, Barbacid M. RAS oncogenes: the first 30 years. Nat Rev

Cancer 2003;3:459–65.2. Karnoub AE,Weinberg RA. Ras oncogenes: split personalities. Nat RevMol

Cell Biol 2008;9:517–31.3. Cox AD, Der CJ, Philips MR. Targeting RAS membrane association: back

to the future for anti-RAS drug discovery? Clin Cancer Res 2015;21:1819–27.

4. Prior IA, Hancock JF. Ras trafficking, localization and compartmentalizedsignalling. Semin Cell Dev Biol 2012;23:145–53.

5. Esteban LM, Vicario-Abej�on C, Fern�andez-Salguero P, Fern�andez-MedardeA, Swaminthan N, Yienger K, et al. Targeted genomic disruption of H-rasand N-ras, individually or in combination, reveals the dispensability ofboth loci for mouse growth and development. Mol Cell Biol 2001;21:1444–52.

6. Johnson L, Greenbaum D, Cichowski K, Mercer K, Murphy E, Schmitt E,et al. K-ras is an essential gene in themouse with partial functional overlapwith N-ras. Genes Dev 1997;11:2468–81.

7. KoeraK,NakamuraK,NakaoK,Miyoshi J, ToyoshimaK,Hatta T, et al. K-rasis essential for the development of the mouse embryo. Oncogene1997;15:1151–9.

8. Potenza N, Vecchione C, Notte A, De Rienzo A, Rosica A, Bauer L, et al.Replacement of K-RaswithH-ras supports embryonic development despiteinducing cardiovascular pathology in adult mice. EMBO Rep 2005;6:432–7.

9. Prior IA, Lewis PD,Mattos CA. A comprehensive survey of Rasmutations incancer. Cancer Res 2012;72:2457–67.

10. Tidyman WE, Rauen KA. The RASopathies: developmental syndromesof Ras/MAPK pathway dysregulation. Curr Opin Genet Dev 2009;19:230–6.

11. Aoki Y, Niihori T, Inoue S, Matsubara Y. Recent advances in RASopathies.J Hum Genet 2016;61:33–9.

12. Gripp KW, Lin AE. Costello syndrome: a Ras/mitogen activated proteinkinase pathway syndrome (rasopathy) resulting from HRAS germlinemutations. Genet Med 2012;14:285–92.

Expression of H-RasG12V from the K-Ras Promoter

www.aacrjournals.org Cancer Res; 77(3) February 1, 2017 717

on February 24, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 21, 2016; DOI: 10.1158/0008-5472.CAN-16-2925

Page 12: H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra … · Molecular and Cellular Pathobiology H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra When Driven by the

13. Aoki Y, Niihori T, Kawame H, Kurosawa K, Ohashi H, Tanaka Y, et al.Germline mutations in HRAS proto-oncogene cause Costello syndrome.Nat Genet 2005;37:1038–40.

14. Kratz CP, Schubbert S, Bollag G, Niemeyer CM, Shannon K, ZenkerM. Germline mutations in components of the Ras signaling path-way in Noonan syndrome and related disorders. Cell Cycle 2006;5:1607–11.

15. Tuveson DA, Shaw AT, Willis NA, Silver DP, Jackson EL, Chang S, et al.Endogenous oncogenic K-ras(G12D) stimulates proliferation and wide-spread neoplastic and developmental defects. Cancer Cell 2004;5:375–87.

16. Guerra C,Mijimolle N,Dhawahir A, Dubus P, BarradasM, SerranoM, et al.Tumor induction by an endogenous K-ras oncogene is highly dependenton cellular context. Cancer Cell 2003;4:111–20.

17. Schuhmacher A J, Guerra C, Sauzeau V, Ca~nameroM, Bustelo XR, BarbacidM. A mouse model for Costello syndrome reveals an Ang II-mediatedhypertensive condition. J Clin Invest 2008;118:2169–79.

18. Chen X, Mitsutake N, LaPerle K, Akeno N, Zanzonico P, Longo VA, et al.Endogenous expression of HrasG12V induces developmental defects andneoplasms with copy number imbalances of the oncogene. Proc Natl AcadSci USA 2009;106:7979–84.

19. Hern�andez-Porras I, Fabbiano S, Schuhmacher AJ, Aicher A, Ca~namero M,C€amara JA, et al. K-RasV14I recapitulates Noonan syndrome in mice. ProcNatl Acad Sci USA 2014;111:16395–400.

20. Jonkers J,Meuwissen R, Van derGuldenH, PeterseH, van der ValkM, BernsA. Synergistic tumor suppressor activity of BRCA2 and p53 in a conditionalmouse model for breast cancer. Nat Genet 2001;29:418–25.

21. Ruzankina Y, Pinzon-Guzman C, Asare A, Ong T, Pontano L, Cotsarelis G,et al. Deletion of the developmentally essential gene ATR in adult miceleads to age-related phenotypes and stem cell loss. Cell Stem Cell 2007;1:113–26.

22. Guerra C, Schuhmacher AJ, Ca~namero M, Grippo PJ, Verdaguer L, P�erez-Gallego L, et al. Chronic Pancreatitis is essential for induction of pancreaticductal adenocarcinoma by K-Ras oncogenes in adult mice. Cancer Cell2007;11:291–302.

23. Blasco RB, Francoz S, Santamaría D, Ca~namero M, Dubus P, Charron J,et al. c-Raf, but not B-Raf, is essential for development of K-Rasoncogene-driven non-small cell lung carcinoma. Cancer Cell 2011;19:652–63.

24. Itahana K, Campisi J, Dimri GP. Methods to detect biomarkers of cellularsenescence: the senescence-associated b-Galactosidase assay.MethodsMolBiol 2007;371:21–31.

25. Drosten M, Sum EY, Lechuga CG, Sim�on-Carrasco L, Jacob HK, García-Medina R, et al. Loss of p53 induces cell proliferation via Ras-independentactivation of the Raf/Mek/Erk signaling pathway. Proc Natl Acad Sci USA2014;111:15155–60.

26. Lakso M, Pichel JG, Gorman JR, Sauer B, Okamoto Y, Lee E, et al. Efficientin vivomanipulation of mouse genomic sequences at the zygote stage. ProcNatl Acad Sci USA 1996;93:5860–5.

27. Mainardi S, Mijimolle N, Francoz S, Vicente-Due~nas C, S�anchez-García I,Barbacid M. Identification of cancer initiating cells in K-ras driven lungadenocarcinoma. Proc Natl Acad Sci USA 2014;111:255–60.

28. Wendorff AA, Koch U, Wunderlich FT, Wirth S, Dubey C, Br€uning JC, et al.Hes1 is a critical but context-dependent mediator of canonical Notchsignaling in lymphocyte development and transformation. Immunity2010;33:671–84.

29. Cox J,HeinMY, LuberCA,Paron I,NagarajN,MannM.Accurate proteome-wide label-free quantification by delayed normalization and maximalpeptide ratio extraction, termed MaxLFQ. Mol Cell Proteomics 2014;13:2513–26.

30. Wright CJ, McCormack PL. Trametinib: first global approval. Drugs2013;73:1245–54.

31. Serrano M, Lin AW, McCurrach ME, Beach D, Lowe SW. Oncogenic rasprovokes premature cell senescence associated with accumulation of p53and p16INK4a. Cell 1997;88:593–602.

32. Frisch SM, Mymryk JS. Adenovirus-5 E1A: paradox or paradigm. Nat RevMol Cell Biol 2002;3:441–52.

33. WangMT,HolderfieldM,Galeas J,Delrosario R, ToMD,BalmainA, et al. K-Ras promotes tumorigenicity through suppression of non-canonical Wntsignaling. Cell 2015;163:1237–51.

34. Braun BS, Tuveson DA, Kong N, Le DT, Kogan SC, Rozmus J, et al.Somatic activation of oncogenic Kras in hematopoietic cells initiates arapidly fatal myeloproliferative disorder. Proc Natl Acad Sci U S A2004;101:597–602.

35. Chan IT, Kutok JL, Williams IR, Cohen S, Kelly L, Shigematsu H, et al.Conditional expression of oncogenic K-ras from its endogenous promoterinduces a myeloproliferative disease. J Clin Invest 2004;113:528–38.

36. Kindler T, Cornejo MG, Scholl C, Liu J, Leeman DS, Haydu JE, et al. K-RasG12D-induced T-cell lymphoblastic lymphoma/leukemias harborNotch1 mutations and are sensitive to g-secretase inhibitors. Blood 2008;112:3373–82.

37. Lampson BJ, Pershing NL, Prinz JA, Lacsina JR, Marzluff WF, Nicchitta CV,et al. Rare codons regulate KRas oncogenesis. Curr Biol 2013;23:70–5.

38. Pershing NL, Lampson BL, Belsky JA, Kaltenbrun E, MacAlpine DM,Counter CM. Rare codons capacitate Kras-driven de novo tumorigenesis.J Clin Invest 2015;125:222–33.

39. ToMD,Wong CE, Karnezis AN, Del Rosario R, Di Lauro R, Balmain A. Krasregulatory elements and exon 4A determine mutation specificity in lungcancer. Nat Genet 2008;40:1240–4.

40. Gao S, Ho D, Vatner DE, Vatner SF. Echocardiography inmice. Curr ProtocMouse Biol 2011;1:71–83.

Cancer Res; 77(3) February 1, 2017 Cancer Research718

Drosten et al.

on February 24, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 21, 2016; DOI: 10.1158/0008-5472.CAN-16-2925

Page 13: H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra … · Molecular and Cellular Pathobiology H-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra When Driven by the

2017;77:707-718. Published OnlineFirst November 21, 2016.Cancer Res   Matthias Drosten, Lucía Simón-Carrasco, Isabel Hernández-Porras, et al.   When Driven by the Same Regulatory SequencesH-Ras and K-Ras Oncoproteins Induce Different Tumor Spectra

  Updated version

  10.1158/0008-5472.CAN-16-2925doi:

Access the most recent version of this article at:

  Material

Supplementary

  http://cancerres.aacrjournals.org/content/suppl/2016/11/19/0008-5472.CAN-16-2925.DC1

Access the most recent supplemental material at:

   

   

  Cited articles

  http://cancerres.aacrjournals.org/content/77/3/707.full#ref-list-1

This article cites 40 articles, 13 of which you can access for free at:

  Citing articles

  http://cancerres.aacrjournals.org/content/77/3/707.full#related-urls

This article has been cited by 8 HighWire-hosted articles. Access the articles at:

   

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

  Subscriptions

Reprints and

  [email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/77/3/707To request permission to re-use all or part of this article, use this link

on February 24, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 21, 2016; DOI: 10.1158/0008-5472.CAN-16-2925