aspsw

m

BIO

LOG

YO

FTH

EPA

NCREA

S

GASTROENTEROLOGY 2013;144:1220–1229

Roles for KRAS in Pancreatic Tumor Development and Progression

Marina Pasca diMagliano1,2,3

Craig D. Logsdon4,5

Departments of 1Surgery and 2Cell and Developmental Biology, and 3Comprehensive Cancer Center, University of Michigan, Ann Arbor, Michigan; and 4Departments

of Cancer Biology and 5Medical Oncology, The University of Texas MD Anderson Cancer Center, Houston, Texas

twttric

The Kras gene is mutated to an oncogenic form inmost pancreatic tumors. However, early attempts touse this molecule as a specific biomarker of the dis-ease, or inhibit its activity as a cancer therapy, failed.This left a situation in which everyone was aware ofthe association between this important oncogene andpancreatic cancer, but no one knew what to do aboutit. Recent findings have changed this picture—manyassumptions made about KRAS and its role in pan-creatic cancer were found to be incorrect. Several fac-tors have contributed to increased understanding ofthe activities of KRAS, including creation of geneti-cally engineered mouse models, which have allowedfor detailed analyses of pancreatic carcinogenesis inan intact animal with a competent immune system.Cancer genome sequencing projects have increasedour understanding of the heterogeneity of individualtumors. We also have a better understanding of whichoncogenes are important for tumor maintenance andare therefore called “drivers.” We review the advancesand limitations of our knowledge about the role ofKras in development of pancreatic cancers and theimportant areas for future research.

Keywords: Kras; Inflammation; Pancreatic Cancer; MouseModel.

Pancreatic cancer progresses through a series of pre-cursor lesions, the most common of which are known

s pancreatic intraepithelial neoplasia (PanIN). Progres-ion requires specific genetic changes and, at least inancreatic tumors, each stage seems to be associated withpecific mutations. Oncogenic KRAS was first associatedith pancreatic cancer at least 24 years ago.1,2 At that

time, Kras was shown to be mutated to an oncogenicform, most commonly KrasG12D, in most pancreatic tu-

ors. Expression of the human oncogenic KrasG12D in themouse pancreas duplicated, at least approximately, these

precursor stages. These genetically engineered mouse

models allow for the study of the earliest phases of pan-creatic cancer development, as their gene expression canbe manipulated. However, mouse models have inherentlimitations, beyond the biologic differences between miceand humans. One limitation is that they express onco-genic KRAS in all the cells of the pancreas, unlike humanpancreatic tumors. Another is the concurrent, rather thansequential, introduction of the genetic alterations associ-ated with each stage of spontaneous tumor development.Therefore, the use of genetic mouse models needs to bebalanced by other approaches, such as using human pan-creatic cancer cell lines, primary human cells, and humanxenograft tumors. Judicious use of all of these modelsprovides the best picture of the initiation and progressionof pancreatic cancer, and has allowed us to appreciate theroles of KRAS in these processes.

Individual vs Population of KRASMoleculesIndividual KRAS proteins function as binary mo-

lecular switches. When bound to guanosine triphosphate(GTP), they interact with signaling molecules that regu-late cell activities such as proliferation, differentiation,apoptosis, and cell migration (Figure 1).3 Binding of GTPo KRAS is extremely low in the absence of interactionsith guanine nucleotide exchange factors, which increase

he rate of GTP loading. Many receptors for growth fac-ors, cytokines, hormones, neurotransmitters, and otheregulators are able to activate RAS, either by directly orndirectly increasing access of guanine nucleotide ex-hange factors.

Abbreviations used in this paper: GTP, guanosine triphosphate;MAPK, mitogen-activated protein kinase; MEK, MAP kinase kinase;PanIN, pancreatic intraepithelial neoplasia; PDAC, pancreatic ductaladenocarcinoma.

© 2013 by the AGA Institute0016-5085/$36.00

http://dx.doi.org/10.1053/j.gastro.2013.01.071

K

tfisd

tp

otbpi

Kpsefmebsglaitoi

BIO

LOG

YO

FTH

EPA

NCREA

S

May 2013 KRAS IN PANCREATIC TUMOR DEVELOPMENT AND PROGRESSION 1221

Once a cell has responded to an incoming signal, KRASactivity is no longer needed, so GTP is hydrolyzed toguanosine diphosphate. KRAS itself has some limitedGTP hydrolysis activity, which is increased by interactionswith specific GTPase-activating proteins. Although indi-vidual KRAS molecules act as binary switches, the popu-lation of KRAS molecules in a cell acts more like a rheo-stat than a switch. Greater numbers of KRAS moleculesbound to GTP lead to a greater overall signal. In otherwords, at the level of the cell, KRAS is neither “on” nor“off”—the number of active molecules determines the lev-els of the resulting signal.

Specific point mutations in KRAS (primarily those thataffect KRAS– GTPase-activating protein interactions) re-duce GTP hydrolysis and thereby cause KRAS to remainactive.4 They are considered to be oncogenic because whenthey were first investigated, they were observed to trans-form cells in the absence of other manipulations, and cellsthat expressed these mutant forms of KRAS formed tu-mors in vivo. It has been estimated that approximately30% of all tumors have oncogenic mutations in RAS

Figure 1. Basic KRAS biology and the positive feedback loop betweenRAS and inflammation. KRAS binds either GTP or guanosine diphos-hate (GDP). When occupied by GDP, KRAS does not activate down-tream signaling pathways and is considered “off.” KRAS is activated byxtracellular signals coming from the environment in the form of growthactors, cytokines, damage molecules (DAMPs), hormones, or other

olecules. These molecules indirectly interact with guanine nucleotidexchange factors (GEFs) to replace GDP for GTP and causing KRAS toe “on.” Active KRAS will then interact with a large number of down-tream signaling pathways. Some of these pathways in turn lead toeneration of signals that activate KRAS through a positive feedback

oop. Examples are inflammatory mediators that are activated by KRASnd in turn lead to further activation of KRAS. Normal KRAS is rapidly

nactivated thanks to the effect of the GTPase-activating proteins (GAPs)hat help hydrolyze GDP to GTP. In the presence of an oncogenic formf KRAS the return of KRAS to an “off” state is delayed, and a patholog-

cal response ensues that can lead to cancer.

family members, HRAS, NRAS, and KRAS5; oncogenic

ras is found in nearly every pancreatic tumor.6,7 Manystudies have indicated that Kras activity increases aftertransfection with oncogenic forms of Kras, indicating thatthey are constitutively active.8 –10 However, KRAS activa-ion is a complex phenomenon; GTP binding is not suf-cient to define active KRAS. Other factors, includingubcellular localization, can influence its association withownstream effectors.11,12 Therefore, KRAS-GTP might

not always stimulate effector signaling.13,14 This observa-ion has important implications for cancer treatment andrevention.An additional layer of complexity was added with the

bservation that Kras alone might not be sufficient toransform a cell. When oncogenic Kras was evaluated as aiomarker, numerous studies reported that healthy peo-le have cells with oncogenic Kras in different organs,

ncluding the pancreas,15–17 colon,16 and lungs,18 at ratesfar exceeding the rates of cancer development. More re-cently, mice that express oncogenic Kras, either in thewhole body or in specific organs, develop cancers fromonly a small fraction of the cells that contain the onco-genic Kras,19,20 It can therefore be assumed that other,genetic or epigenetic, factors are required to initiate car-cinogenesis, even when a mutation in the Kras oncogenehas been acquired. One key factor might be the level ofKRAS activity. In fact, in genetically engineered mice thatexpress oncogenic KRAS at physiologic levels (because asingle allele is mutated, the cells presumably express equalamounts of oncogenic and nononcogenic KRAS), only asmall fraction of the cells are transformed. In addition,the overall level of KRAS activity is lower than expected,whether it is because not all of the oncogenic KRASmolecules bind GTP or because, even when GTP bound,they do not always activate effectors (possibly because ofinappropriate subcellular localization or negative-feed-back mechanisms).11 Accordingly, expression of onco-genic Kras from its endogenous locus in mice is insuffi-cient to activate downstream signaling pathways, such asthe mitogen-activated protein kinase (MAPK) pathway.13

Supporting the hypothesis that a specific level of KRASactivation is required to initiate transformation, upstreamstimuli were shown to accelerate the development of can-cer in mice (Figure 2).13,21–24 These studies indicate thatreaching a threshold level of KRAS activity might beessential to initiate the carcinogenesis process, and thatthe presence of a mutant copy of Kras is not sufficient toreach this threshold. Therefore, signals that act upstreamof Kras, such as epidermal growth factor and inflamma-tory stimuli, might play an important role during thecarcinogenesis process. In this light, it is noteworthy thatmany reagents shown to accelerate formation of pancre-atic ductal adenocarcinomas (PDACs) in mice that expressoncogenic Kras, also directly or indirectly activate KRAS.Likewise, many reagents shown to prevent or increase thetime to development of PDAC in mouse models reduceRAS activity, directly or indirectly (Figure 3). A secondaspect to be considered is the ability of the cells to with-

stand high levels of Kras activation. It is conceivable that

pststpoi

afol

nfoThmutissrvtKa

BIO

LOG

YO

FTH

EPA

NCREA

S

1222 MAGLIANO AND LOGSDON GASTROENTEROLOGY Vol. 144, No. 6

the oncogenic stress associated with expression of onco-genic Kras might result in apoptosis or senescence,25,26

and factors that allow cells to overcome the senescencebarrier, such as inflammation25 or loss of tumor suppres-sor genes, such as p16 or p53,27 allow the transformation

rocess. Some controversy still persists, however, as othertudies have indicated that oncogenic KRAS is sufficiento inhibit the onset of senescence and repress the expres-ion of p16.28 However, neither the presence of inflamma-ion nor the loss of tumor suppressor genes is sufficient,er se, to initiate pancreatic cancer in the absence ofncogenic Kras.29 –31 These observations support a model

n which Kras plays a unique role in the onset of PDAC.Human and mouse PDAC cells have elevated KRAS

ctivity and correspondingly high activation of KRAS ef-ector genes. In mice that express endogenous levels ofncogenic Kras, tumor cells contain significantly higher

evels of KRAS–GTP than nontumor tissues.7 In miceengineered to express elevated levels of oncogenic KRAS,

Figure 2. Multiple changes in gene expression influence KRAS activity,the KRAS downstream pathway and cell growth occur during develop-ment of a “mature” PDAC cell. RAS activity is still affected by upstreamfactors, but this component might be less important in the presence of amutant form of KRAS. The downstream effectors of KRAS are highlyactivated and might eventually become independent of KRAS. Finally,expression of several regulators of cell cycle are altered in cancer cells;the cyclins listed in the figure are all expressed at levels exceeding thoseof normal pancreatic cells.

approaching the elevated levels found in tumors, wide-

spread formation of precursor lesions and rapid progres-sion to invasive tumors are observed, in the absence of anyother manipulations.13 Taken together, these results sup-port the concept that a relatively high level of oncogenicKras activity is necessary for cellular transformation.6

Therefore, the mechanisms that increase the levels ofoncogenic Kras in cells contribute to transformation andcancer development.

What Is the Role of Oncogenic KRAS inPancreatic Tumor Formation?People acquire oncogenic mutations in Kras in

lung, pancreas, colon, and other tissues as they age.15–18

This is a major reason that Kras was not found to be agood biomarker for cancer. So why do only a few peopledevelop PDAC or other cancers? A major new insight hasbeen that oncogenic KRAS is not always in the activestate.14 In other words, the presence of oncogenic KRAS isnot sufficient to transform cells and additional genetic orenvironmental factors that raise the threshold of KRASactivity and remove barriers to tumorigenesis might berequired.

Oncogenic and nononcogenic forms of Kras can beactivated by many factors. However, oncogenic KRAS hasslower kinetics of return to its guanosine diphosphate–bound status than nononcogenic forms.9 The slower ki-

etics provide extra time for activated KRAS to generate aeedback loop that sustains its activity, such as activationf nuclear factor��B, cyclo-oxygenase�2, and others.hese finding apply to cancer prevention strategies, whichave focused on inhibiting KRAS effectors, under theodel that oncogenic KRAS is always active. However, the

nderstanding that oncogenic KRAS needs to be activatedo cause cancer means that inhibition of KRAS activationtself is a reasonable preventative strategy. This model isupported by reports that drugs that reduce cancer risk,uch as nonsteroidal anti-inflammatory agents (eg, aspi-in, celecoxib, etc) and various antioxidants, inhibit acti-ation of KRAS (Figure 2). This concept also indicateshat attempting to block cancer initiation downstream ofRAS will be a challenge because many effectors arectivated by KRAS (Figure 1).

Kras as a Component of a ComplexSystemKras is clearly important component in the patho-

genesis of pancreatic cancer. Oncogenic mutations in Krasare also frequently detected in lung, colon, and othertumor types. Increased Kras activity is required for devel-opment of hepatocellular carcinoma, via alterations inGTPase-activating proteins, instead of oncogenic muta-tions. Increased levels of activity of KRAS effectors, ordecreased levels or activity of KRAS inhibitors (or theirregulatory molecules) could also lead to hepatic tumori-genesis.

Many molecular alterations are needed for PanIN le-

sions to progress to metastatic tumors. Hundreds of

eipd

dprscacoemog

mmHota

popioedTcKPsr

kfeaaacawcfu

BIO

LOG

YO

FTH

EPA

NCREA

S

May 2013 KRAS IN PANCREATIC TUMOR DEVELOPMENT AND PROGRESSION 1223

changes in gene expression occur in pancreatic cancer cellscompared with normal pancreatic cells.7 Analyses of genexpression changes in human pancreatic cancer cells havedentified several genes whose increased or decreased ex-ression would affect KRAS activity (Figure 2) (C. Logs-on, unpublished observation).KRAS is a regulator of the MAPK pathway. In fully

eveloped PDAC, several molecules within the MAPKathway are over- or underexpressed that are normallyegulated by active KRAS (C. Logsdon, unpublished ob-ervation). Likewise, expression of cyclins and other cellycle regulators increases in cancer cells. These alterationsre likely selected for— higher levels of KRAS activity in-rease cell proliferation. Notwithstanding the complexityf changes that promote the activity of KRAS and itsffectors in pancreatic cancer, several studies in mouseodels or human cell lines suggest that at least a subset

f pancreatic tumors continuously require KRAS forrowth. This aspect will be discussed here in more detail.

Oncogenic Kras in Pancreatic CancerProgression and MaintenanceThere is much evidence that Kras is required for

pancreatic tumorigenesis, but until recently it was as-sumed that nothing could be done to prevent the onco-genic effects of Kras activation, and that the only optionwas to slow tumor growth and progression. Researchersassumed that Kras must be important at all stages ofPDAC. However, like the assumption that oncogenic Krasmust be always on, this might not actually be the case. Tostudy the role of Kras in growth and progression ofPDAC, in the absence of good inhibitors, studies haverelied on small hairpin RNA-mediated knockdown in celllines, inhibitors of downstream targets of Kras, or morerecently, on mouse models that allow reversible expres-sion of oncogenic Kras.

During Early Stages of Pancreatic CancerPancreatic adenocarcinoma cells have highly un-

stable genomes; each individual tumor is estimated toaccumulate several hundred mutations.7 The progressionfrom a normal pancreatic cell into a metastatic cancer cellrequires multiple steps, each of which requires changes ingene expression. Clearly, PDAC is not the same fromformation until metastasis. Early human PanIN lesionshave Kras mutations, often in absence of additional ge-netic alterations.32,33 However, progression requires inac-tivation of tumor suppressors, such as p16.25,28 In hu-

ans, sporadic mutations in Kras might occur, and inost cases be cleared from the tissue by cell senescence.owever, cells that escape senescence through acquisition

f additional mutations might become transformed. Inhe pancreas of mice, these mutations can be introducedt the same time in a large number of cells.

The widely used KC (Pdx1-Cre;LSL-KrasG12D or Ptf1a-Cre; LSL-KrasG12D), KPC (most commonly Pdx1-Cre;LSL-

KrasG12D;LSL-Trp53R172H or Ptf1a-Cre; LSL-KrasG12D;LSL-

Trp53R172H, but also used to describe models with loss-offunction or conditional inactivation of the tumor sup-pressor p53) and KC;Ink4a (Pdx1-Cre;LSL-KrasG12D;Ink4af/f or Ptf1a-Cre;LSL-KrasG12D;Ink4af/f) mouse modelsof pancreatic cancer were developed via tissue-specific,Cre-mediated expression of oncogenic Kras.29,30,34 Thesemodels recapitulate the initiation and progression of pan-creatic cancer. However, because Kras expression, onceinduced, is irreversible, these models are not suitable forinvestigating whether Kras is required beyond tumorigen-esis—for growth and progression. Some tumors that de-velop in mice have been reported to be dependent on (oraddicted to) a single oncogene product; tumors that de-velop in mouse models of lung adenocarcinoma, breastcancer, and melanoma depend on oncogenic Kras tomaintain their mass and continue growing.35–37

Researchers have recently created mice in which pan-creatic expression of oncogenic Kras can be reversed,called inducible Kras* or iKras mice.38,39 These mice ex-

ress oncogenic Kras under the control of a tetracyclineperator; the rtTa transcription factor is expressed in aancreas-specific manner from the Rosa26 locus. Admin-

stration of doxycycline to iKras* mice leads to expressionf oncogenic Kras and withdrawal of the drug inactivatesxpression of the transgene. Unlike KC mice, iKras miceo not express KrasG12D from the endogenous Kras locus.he expression levels of oncogenic Kras, however, areomparable with endogenous levels, and activation ofras in adult animals leads to formation of sporadicanINs only with long latency and low penetrance, pos-ibly because a threshold of KRAS activity needs to beeached before the oncogene has an effect on the tissue.

Induction of acute pancreatitis with the cholecysto-inin agonist caerulein leads to rapid and widespreadormation of PanINs, likely through initiation of the Krasffector loop described here. Induction of acute pancre-titis in wild-type mice leads to acinar damage, includingcinar to ductal metaplasia, infiltration of immune cells,nd edema; the peak level of damage occurs within 24 h ofaerulein administration. Tissue repair ensues rapidly,nd the pancreas resumes its normal histology within 1eek, although slightly higher proliferation of acinar

ells, compared with healthy pancreata, is observed for aew weeks. In contrast, KC and iKras* pancreata fail tondergo tissue repair after caerulein administration.38,40

In these mice, acinar to ductal metaplasia progresses toform dysplastic ductal structures, surrounded by exten-sive fibrosis, within 1 week. Within 3 weeks, virtually allthe ductal structures show characteristics of PanINs. Overtime, higher-grade PanIN lesions populate the pancreasand, finally, carcinoma in situ develops.

To understand the role of Kras signaling in thesechanges, the effects of inactivating oncogenic Kras expres-sion at different time points were examined. The effects ofKras inactivation were found to be time dependent (Fig-ure 4). Inactivation in low-grade PanIN lesions results inmost cells that line the dysplastic ducts to activate expres-

sion of genes of the acinar lineage, and inactivate ductal

BIO

LOG

YO

FTH

EPA

NCREA

S

1224 MAGLIANO AND LOGSDON GASTROENTEROLOGY Vol. 144, No. 6

genes, in a process that is the opposite of acinar to ductalmetaplasia. At the early stages of lesion formation, KRASactivity is therefore necessary for the transformation pro-cess to continue; it seems to prevent tissue repair andregulate the differentiation status of the epithelial cells, asproposed previously.40 Kras inactivation in high-gradePanIN lesions has different effects. The precancerous cells

Figure 3. Molecules that increase KRAS activity tend to promote PDACdevelopment, while and those that inhibit KRAS activity tend to preventPDAC. This would be predicted by the KRAS-inflammatory feedbackloop model.

Figure 4. Oncogenic Kras in pancreatic cancer maintenance. Activatprogression to metastatic pancreatic cancer is dependent on acquisitioKras in PanINs leads to either redifferentiation of PanIN cells to normaoncogenic Kras in advanced disease leads to tumor regression through aa small subset of cancer cells survives Kras inactivation and persists in a

Other mechanisms of relapse might also exist.

that appear require Kras for survival, and undergo apo-ptosis once the transgene is inactivated. These resultsindicate that Kras is important not only for tumor for-mation, but also during early stages of tumor progression.

A common feature of KC and iKras mice is that theirlesions rarely progress to adenocarcinoma unless addi-tional mutations are introduced. This is consistent withthe observation that, in patients, pancreatic adenocarci-noma does not occur without the accumulation of mul-tiple genetic alterations, probably over the course of de-cades.41 Loss, inactivation, or mutation of multiple tumorsuppressors (such as Tp53 and p16) is commonly detectedin human pancreatic tumors. In mice, these tumor sup-pressors are spontaneously lost at different rates, depend-ing on the level of inflammation and/or Kras activity. Forexample, in KC mice, which express endogenous levels ofoncogenic Kras, the tumor suppressor Tp53 tends to bemutated or lost at late stages of tumor development.34 Incontrast, in mice engineered to express high levels ofoncogenic Kras in the pancreatic cells, such as the Elas-tase-CreER;cLGL-KrasG12D (LGL) model, p16 is rapidlylost.42 Possibly, in presence of high levels of oncogenicKras there is a higher selective pressure for pancreatic cellsto lose p16 and therefore escape Kras-induced senes-cence.26 The high frequency of p16 loss might explain whyLGL mice develop PDAC at a faster rate than LSL mice.

To speed cancer development in models with low Krasactivity, mutant alleles of tumor suppressors can be in-

of oncogenic Kras in the normal pancreas leads to PanIN formation;f additional genetic alterations, such as mutation of p53. Inactivation ofncreatic lineages such as acinar cells, or to apoptosis. Inactivation oftosis and probably additional mechanisms yet to be identified. However,rmant state; these cells can lead to tumor relapse on Kras inactivation.

ionn ol papop

do

Kmta

rsrtbtadvoaocr

tet

docn

ta

mmoctpt

BIO

LOG

YO

FTH

EPA

NCREA

S

May 2013 KRAS IN PANCREATIC TUMOR DEVELOPMENT AND PROGRESSION 1225

troduced, which would resemble development of pancre-atic cancer in humans. iKras mice carrying a loss-of-function allele in p53 (called iKras*p53 mice) rapidlydevelop pancreatic adenocarcinoma with high pen-etrance,38,39 and can be used to determine the effects of

ras inactivation in invasive tumors. Studies with theseice have shown that inactivation of Kras leads to rapid

umor regression through loss of tumor cell proliferationnd viability.38,39 These observations have been extended

to the metastases that develop with high frequency iniKras* mice, engineered to express a point mutation inp53 (p53R172H mice).43 In vivo imaging studies showedapid regression of the primary tumor and liver metasta-es on Kras inactivation. Although these tumors appear toegress steadily over time, accurate in-depth characteriza-ion of the tissue after regression, as well as observationsased on primary tumor cells, indicated that not all theumor cells were eliminated on Kras inactivation. In fact,

fraction of the tumor cells appeared to persist in aormant state and resume rapid growth on K-ras reacti-ation.43 It is possible that Kras inhibition is eventually bever-ridden by generation of resistant clones, which haveccumulated additional mutations in the same pathwaysr have activated other oncogenic pathways. Lung adeno-arcinoma and melanoma cells have also been shown toequire oncogenic Kras.36,37

In the Tumor MicroenvironmentMice that have been genetically engineered to ex-

press oncogenic Kras have also been used to study the roleof Kras on the pancreatic tumor microenvironment,which contains extensive inflammatory stroma. Duringthe earliest stages of PanIN formation, the lesions accu-mulate proliferating cells of mesenchymal origin thatmight comprise fibroblasts and pancreatic stellate cells.PanIN formation and progression is also accompanied bythe infiltration of immune cells.44 It is interesting to notehat changes in the phenotype of the stellate cells occurarlier than noticeable changes in other components ofhe pancreas.45 Therefore, even low levels of Kras activity

generate signals that influence the microenvironment.Unlike most other solid tumors, pancreatic tumors are

considered to be hypovascularized, although blood vesselsare present within the tumor microenvironment. It isunclear whether the vascularity per cancer cell is lowerthan in other tumors or whether the abundant nonvas-cular extracellular matrix contributes to the low level ofoverall vascularity. Stellate cells produce angiogenic fac-tors.46,47 In addition to the cellular components, thestroma comprises components of the extracellular matrix,such a collagen fibers and hyaluronic acid.48,49 Little isknown about how the formation of the stroma is regu-lated and maintained.

Perhaps the most convincing evidence for the effect ofKras on the tumor microenvironment is found when Krasis inactivated in pancreata bearing low-grade PanINs. Un-der these conditions the stroma is remodeled. The acti-

vated fibroblasts that populate the stroma stop expressing s

markers of activation, exit the cell cycle, and are elimi-nated from the pancreas via an unknown mechanism.Inactivation of Kras also leads to resolution of the chronicinflammation associated with pancreatic cancer.

Kras therefore regulates production of factors thatmaintain an active stroma (Figure 5). These factors andtheir activities have not been completely identified, butappear to include sonic hedgehog, interleukin-6, andprostaglandin E,50 each of which is expressed in a Kras-

ependent manner.38 Sonic Hedgehog, one of the ligandsf the Hedgehog signaling pathway, is expressed by pan-reatic tumor cells51,52 and functions in a paracrine man-er,53 activating Hedgehog signaling in the stroma and

potentially mediating its maintenance.54 The inflamma-ory cytokine IL-6 is overexpressed in pancreatic tumorsnd is important for development of PanINs in mice.55

Prostaglandin E acts directly on stellate cells through theprostaglandin E receptor 4, to stimulate production of thestroma.50 These, and likely several other factors, are gen-erated by high levels of sustained Kras activity.

The immune cells that infiltrate the pancreas alsoappear to be regulated by Kras. In mouse models ofpancreatic cancer, PanINs are infiltrated by immunecells, including those that suppress the immune re-sponse, such a regulatory T cells, myeloid-derived sup-pressor cells,44 and mast cells.56 Tumor cells secretecytokines, such as granulocyte-macrophage colony-stimulating factor,57,58 which promote infiltration of

yeloid-derived suppressor cells that inhibit anti-tu-or immune responses. Kras inactivation leads to an

verall reduction in the number of infiltrating immuneells. So, the inflammatory environment of pancreaticumors also appears to be regulated by Kras, in aaracrine manner, forming part of a Kras–inflamma-ion positive-feedback loop that requires additional

Figure 5. Kras mediates interactions between the tumor cells and thesurrounding stroma. Tumor cells expressing oncogenic Kras secretemolecules that act in a paracrine manner on surrounding components ofthe stroma, such as fibroblasts, innate and adaptive immune cells (blackarrows). These cells in turn promote tumor growth (dashed arrows); onlya small subset of the signals mediating the interaction between cancercells and components of the stroma have been identified.

tudy.

hulcp

cmam

is

m

abitafto

Bdlo

stpo

BIO

LOG

YO

FTH

EPA

NCREA

S

1226 MAGLIANO AND LOGSDON GASTROENTEROLOGY Vol. 144, No. 6

Tumor MetabolismPancreatic tumors create a hypoxic microenviron-

ment; they are not well vascularized and contain largeamounts of desmoplasia, which contribute to the lowlevels of oxygen.59 Cancer cells are able to adapt to the

ypoxic conditions by several mechanisms, such as byp-regulating hypoxia inhibitory factor��,60 which al-

ows them to change their metabolic pathways. Thehanges in cancer cell metabolism, which have been pro-osed to be “hallmarks of cancer,”61 include increased

glucose metabolism via the glycolytic pathway and re-duced reliance on the Krebs cycle.62 The specific metabolichanges that occur in cancer cells and their regulatoryechanisms are complex and beyond the scope of this

rticle (for a comprehensive review, see reference63), butight be targeted therapeutically.K-ras was recently shown to regulate metabolic changes

n cancer cells by controlling factors that regulate tran-cription of metabolic genes.39 Reagents might therefore

be designed to target Kras, or its effectors, that alterpancreatic cancer metabolism, and impair the ability ofthe cancer cells to maintain high levels of glycolysis.64

Kras DependencyAlthough Kras is known to be involved in cell

transformation, pancreatic tumor formation, and earlystages of tumor progression in mice, little is known aboutits requirement for progression of human pancreatic tu-mors. Mouse models were engineered to depend on Krasfor quick development of tumors. Human tumors takedecades to progress. Because small molecule inhibitors ofKras were not available at the time, studies to determinewhether pancreatic tumor progression requires (is ad-dicted to) KRAS activity have used RNA interference-based approaches. These studies have been conductedusing pancreatic cancer cell lines (mostly derived frompatients) that have been carried for years, in differentlaboratories. Of 7 lines analyzed, 4 were found to be Krasdependent, in that they did not undergo apoptosis whenKras levels were reduced.65

One factor associated with Kras dependency of humanpancreatic cancer cells was overexpression of KRAS pro-tein (which usually correlated with amplification of itsgenomic locus). The Kras gene is amplified in human and

ouse pancreatic tumors.13 Another factor was increasedexpression of epithelial genes and functions of their prod-ucts. Kras-independent lines expressed more genes thatencoded mesenchymal factors, and fewer that encodedepithelial factors, than Kras-dependent cells. Expressionof small hairpin RNAs against Kras slowed growth ofestablished pancreatic xenograft tumors in mice.66

Human PDACs have been subdivided based on geneexpression patterns into quasi-mesenchymal, classic, andexocrine-like (the least characterized) subtypes.67 As thenomenclature indicates, the mesenchymal subtype hasmesenchymal rather than epithelial characteristics. Inter-

estingly, in vitro studies using cell lines with the charac-

teristics of classic and quasi-mesenchymal subtypesshowed that classic PDACs depend on Kras, whereas thoseof the mesenchymal subtype are not. More importantly,these differences were maintained in tumor xenografts. Ittherefore appears that at least a subset of human pancre-atic tumors depends on Kras for progression. Additionalstudies are necessary to elucidate the extent of Kras de-pendency of human pancreatic tumors. Kras dependencyhas not been investigated using primary human tumorxenografts, or in orthotopic tumors implanted into thepancreas.

Effectors of KrasKRAS inhibitors might be in the pipeline,68 in fact,

inhibitors of nononcogenic RAS proteins were describedrecently69,70; however, they have not been historically

vailable and the current drug candidates are still proba-ly years away from the clinic. Farnesyl transferase inhib-

tors, designed to prevent membrane association andhereby activation of K-ras, were found to be nonspecificnd affect the activities of many other proteins. The oralarnesyl transferase inhibitor R115777 did not increasehe median survival time of patients with locally advancedr metastatic pancreatic adenocarcinoma.71 It is likely

that nonfarnesylated Kras can still undergo prenylation,via geranylgeranylation, to associate with the cell mem-brane.

An alternative approach is target effectors of Kras thatare involved in tumor development. Members of severaldifferent Kras signaling pathways participate in formationand growth of pancreatic tumors in mice, such as thekinase Akt. Expression of a dominant-active form of Akt(caAkt) in the pancreas of mice caused expansion of duc-tal structures and activation of progenitor gene expres-sion, but did not lead to progression of PanINs orPDAC.72 Expression of BRAFV600E (the oncogenic form of

RAF) in the pancreas of mice, but not PI3CAH1047R (theominant-active form of phosphatidylinositol 3 kinase)

ed to formation of PanIN lesions. In addition, expressionf BRAFV600E, but not PI3CAH1047R, in combination with a

cancer-associated mutant form of p53 (TP53R270H), led todevelopment of lethal PDACs in mice.67 However, a recenttudy has challenged these conclusions by showing thathe PI3K pathway is sufficient and necessary to initiateancreatic carcinogenesis; thus, the relative contributionf different Kras effectors needs further study.73 There-

fore, activation of MAPK signaling, but not phosphatidyl-inositol 3 kinase signaling via Akt, recapitulates the ef-fects of Kras activation in mice.

So which of these effector pathways is required formaintenance of established tumors? Inhibition of theMAPK signaling with the inhibitor PD325901 had a cy-tostatic effect on KPC tumors, orthotopically implantedin immune-competent, syngeneic mice. Mice givenPD325901 survived longer than controls, but died soonafter cessation of the drug. This finding contrasts, at leastin part, with the observation that Kras inactivation in

tumors affects cell survival in iKras* mice.38,39

a

t

ttsii

pfip

BIO

LOG

YO

FTH

EPA

NCREA

S

May 2013 KRAS IN PANCREATIC TUMOR DEVELOPMENT AND PROGRESSION 1227

RAF inhibitors had no effect on pancreatic cancer celllines, and interfered with the cytostatic effects of MAPkinase kinase (MEK) inhibitors when the compoundswere applied to cells in combination. However, combinedinhibition of MEK and AKT had a synergistic, antiprolif-erative effect on a large panel of human pancreatic cancercell lines. Similar results were obtained with the MEK andAKT inhibitors GDC0941 and AZD6244, respectively.66 Inmice, the combination of these inhibitors caused regres-sion of xenograft tumors, whereas administration of eachreagent alone only slowed tumor growth. The logical nextstep appears to be to determine the efficacy of this treat-ment regimen in mice that spontaneously develop pan-creatic tumors or mice with orthotopic tumors. Studiesare also needed to determine the effects of these inhibitorson survival, and of discontinuing their administration.

Another, somewhat less-studied effector of pathway isthe Ral guanine nucleotide exchange factor–Ral smallGTPase signaling network (for review, see references74

and75), which is frequently activated in pancreatic tumorsnd involved in their progression.76 Inhibitors of cyclin-

dependent kinase-5 have been shown to suppress KRAS–Ral signaling and block pancreatic tumor formation andprogression in mice.77 The Kras-related C3 botulinumsubstrate 1 (Rac1) is an effector of Kras signaling from theepidermal growth factor receptor. Rac1 regulates the re-arrangement of the actin cytoskeleton and cell motility(for review, see reference78). Recently, Rac1 was shown tobe dispensable during development of the normal pan-creas, but required for formation of pancreatic tumors inmice.79 These factors might be developed as therapeuticargets.

In development of therapeutics for pancreatic cancer,he ability of Kras to induce inflammation in surroundingissues should also be considered. Nuclear factor��B andignal transducer and activator of transcription 3 signal-ng are regulated by Kras, and can be targeted with specificnhibitors.13,80 A different way to approach Kras inhibi-

tion would be to block factors that promote KRAS activ-ity, such as inflammatory factors or epidermal growthfactor receptor signaling. However, these types of reagentsmight be better for blocking tumor formation than pro-gression.

Targeting of downstream effectors of KRAS has beentested in other tumor types, such as colon and lungcancer81,82; whether the importance for specific effector

athways is different in different tumor types, or whetherndings from other malignancies can be translated toancreatic cancer will have to be addressed in the future.

A Therapeutic Target?Experimental inhibition of Kras activity slows

growth or even causes regression of pancreatic tumors inmice. Even some human pancreatic cancer cells requireKRAS activity to form tumors. Therefore, strategies toinhibit KRAS directly, or inhibit its effectors, are under

active investigation. But will inhibiting oncogenic KRAS

be sufficient to cure patients with pancreatic cancer? Inmice, inactivation of Kras leads to tumor regression, andthe animals remain healthy, with no evidence of relapse,for a relatively long time.38,39,43 The mechanism of tumorregression involves apoptosis induction, and likely othercomponents that have not been fully elucidated. However,individual tumor cells survive inactivation of Kras, pre-sumably by maintaining a dormant state. Recent unpub-lished observations indicate that this dormancy can bemaintained for several months (M. Collins and MarinaPasca di Magliano, unpublished results). When Kras isreactivated in cancer cells in mice, the cells begin toproliferate rapidly and the mice die within a few days.43

Whether these cells could also resume proliferation inabsence of reactivation of Kras through a different mech-anism is not known. Therefore, the risk of cancer relapseafter withdrawal of KRAS inhibitors is likely to be high.Full tumor eradication will require identification of themechanisms that allow a subset of tumor cells to surviveKras inactivation, and development of methods to targetthem.

Only a subset of human tumors remains Kras-indepen-dent over time, based on studies of human cell lines. It isimportant to determine whether mouse models of pan-creatic cancer resemble a specific subset of human tumors.We also need to develop criteria to determine which hu-man cell lines depend on Kras, so that individual patientscan be treated appropriately.

Future DirectionsThe current experimental evidence indicates that a

high level of sustained KRAS activity is required for pan-creatic tumorigenesis. Whether the expression of an on-cogenic form of Kras is sufficient to obtain these highlevels of activation in tumor cells, or whether mutant RASrequires further activation by upstream signals to achieveits fully active GTP-bound state to drive cancer growthremains an unresolved issue. The observation that, at leastinitially, KRAS activity can be modulated by upstreamsignals provides a rationale to investigate new approachesto pancreatic cancer prevention. Those could conceivablyinclude reducing factors that activate KRAS, by prevent-ing inflammation or altering diet and lifestyle. A secondfocus of research might be the interaction between tumorcells and their surrounding stroma. Recent studies inmouse models indicate that KRAS activity leads to inflam-mation and changes in the stroma. That is not to say thatinflammation and the stroma are not part of what sus-tains tumorigenesis; in fact, there is likely to be a positive-feedback loop that involves KRAS and the stroma andsustains tumor growth. Experiments with Kras inactiva-tion indicate that Kras activity is not only required forpancreatic tumor formation and development but, at leastin mice, the requirement for Kras is continuous. KRAS

might therefore be a good therapeutic target after all.

1

1

1

1

BIO

LOG

YO

FTH

EPA

NCREA

S

1228 MAGLIANO AND LOGSDON GASTROENTEROLOGY Vol. 144, No. 6

References

1. Smit VT, Boot AJ, Smits AM, et al. KRAS codon 12 mutations occurvery frequently in pancreatic adenocarcinomas. Nucleic Acids Res1988;16:7773–7782.

2. Almoguera C, Shibata D, Forrester K, et al. Most human carcino-mas of the exocrine pancreas contain mutant c-K-ras genes. Cell1988;53:549–554.

3. Spaargaren M, Bischoff JR, McCormick F. Signal transduction byRas-like GTPases: a potential target for anticancer drugs. GeneExpr 1995;4:345–356.

4. Scheffzek K, Ahmadian MR, Kabsch W, et al. The Ras-RasGAPcomplex: structural basis for GTPase activation and its loss inoncogenic Ras mutants. Science 1997;277:333–338.

5. Fernandez-Medarde A, Santos E. Ras in cancer and developmentaldiseases. Genes Cancer 2011;2:344–358.

6. Biankin AV, Waddell N, Kassahn KS, et al. Pancreatic cancergenomes reveal aberrations in axon guidance pathway genes.Nature 2012;491:399–405.

7. Jones S, Zhang X, Parsons et al. Core signaling pathways inhuman pancreatic cancers revealed by global genomic analyses.Science 2008;321:1801–1806.

8. Koeffler HP, McCormick F, Denny C. Molecular mechanisms ofcancer. West J Med 1991;155:505–514.

9. Pylayeva-Gupta Y, Grabocka E, Bar-Sagi D. RAS oncogenes: weav-ing a tumorigenic web. Nat Rev Cancer 2011;11:761–774.

0. Karnoub AE, Weinberg RA. Ras oncogenes: split personalities. NatRev Mol Cell Biol 2008;9:517–531.

1. Campbell PM, Singh A, Williams FJ, et al. Genetic and pharmaco-logic dissection of Ras effector utilization in oncogenesis. Meth-ods Enzymol 2006;407:195–217.

2. Campbell SL, Khosravi-Far R, Rossman KL, et al. Increasing com-plexity of Ras signaling. Oncogene 1998;17:1395–1413.

3. Daniluk J, Liu Y, Deng D, et al. An NF-�B pathway-mediatedpositive feedback loop amplifies Ras activity to pathological levelsin mice. J Clin Invest 2012;122:1519–1528.

14. Haojie Huang JD, Yan Liu, Jun Chu, et al. Oncogenic K-Ras re-quires activation for enhanced activity. Oncogene 2013 Jan 21.[Epub ahead of print].

15. Yan L, McFaul C, Howes N, et al. Molecular analysis to detectpancreatic ductal adenocarcinoma in high-risk groups. Gastroen-terology 2005;128:2124–2130.

16. Lu X, Xu T, Qian J, et al. Detecting K-ras and p53 gene mutationfrom stool and pancreatic juice for diagnosis of early pancreaticcancer. Chin Med J (Engl) 2002;115:1632–1636.

17. Parsons BL, Meng F. K-RAS mutation in the screening, prognosisand treatment of cancer. Biomark Med 2009;3:757–769.

18. Yakubovskaya MS, Spiegelman V, Luo FC, et al. High frequency ofK-ras mutations in normal appearing lung tissues and sputum ofpatients with lung cancer. Int J Cancer 1995;63:810–814.

19. Tuveson DA, Shaw AT, Willis NA, et al. Endogenous oncogenicK-ras(G12D) stimulates proliferation and widespread neoplasticand developmental defects. Cancer Cell 2004;5:375–387.

20. Guerra C, Mijimolle N, Dhawahir A, et al. Tumor induction by anendogenous K-ras oncogene is highly dependent on cellular con-text. Cancer Cell 2003;4:111–120.

21. Carriere C, Young AL, Gunn JR, et al. Acute pancreatitis markedlyaccelerates pancreatic cancer progression in mice expressingoncogenic Kras. Biochem Biophys Res Commun 2009;382:561–565.

22. Ardito CM, Gruner BM, Takeuchi KK, et al. EGF receptor is requiredfor KRAS-induced pancreatic tumorigenesis. Cancer Cell 2012;22:304–317.

23. Navas C, Hernandez-Porras I, Schuhmacher AJ, et al. EGF receptorsignaling is essential for k-ras oncogene-driven pancreatic ductaladenocarcinoma. Cancer Cell 2012;22:318–330.

24. Appleman VA, Ahronian LG, Cai J, et al. KRAS(G12D)- andBRAF(V600E)-induced transformation of murine pancreatic epithe-lial cells requires MEK/ERK-stimulated IGF1R signaling. Mol Can-

cer Res 2012;10:1228–1239.

25. Guerra C, Collado M, Navas C, et al. Pancreatitis-induced inflam-mation contributes to pancreatic cancer by inhibiting oncogene-induced senescence. Cancer Cell 2011;19:728–739.

26. DeNicola GM, Tuveson DA. RAS in cellular transformation andsenescence. Eur J Cancer 2009;45(Suppl 1):211–216.

27. Bardeesy N, Aguirre AJ, Chu GC, et al. Both p16(Ink4a) and thep19(Arf)-p53 pathway constrain progression of pancreatic adeno-carcinoma in the mouse. Proc Natl Acad Sci U S A 2006;103:5947–5952.

28. Lee KE, Bar-Sagi D. Oncogenic KRas suppresses inflammation-associated senescence of pancreatic ductal cells. Cancer Cell2010;18:448–458.

29. Aguirre AJ, Bardeesy N, Sinha M, et al. Activated Kras and Ink4a/Arf deficiency cooperate to produce metastatic pancreatic ductaladenocarcinoma. Genes Dev 2003;17:3112–3126.

30. Hingorani SR, Wang L, Multani AS, et al. Trp53R172H andKrasG12D cooperate to promote chromosomal instability andwidely metastatic pancreatic ductal adenocarcinoma in mice. Can-cer Cell 2005;7:469–483.

31. Guerra C, Schuhmacher AJ, Canamero M, et al. Chronic pancre-atitis is essential for induction of pancreatic ductal adenocarci-noma by K-Ras oncogenes in adult mice. Cancer Cell 2007;11:291–302.

32. Kanda M, Matthaei H, Wu J, et al. Presence of somatic mutationsin most early-stage pancreatic intraepithelial neoplasia. Gastroen-terology 2012;142:730–733.e9.

33. Hong SM, Vincent A, Kanda M, et al. Genome-wide somatic copynumber alterations in low-grade PanINs and IPMNs from individu-als with a family history of pancreatic cancer. Clin Cancer Res2012;18:4303–4312.

34. Hingorani SR, Petricoin EF, Maitra A, et al. Preinvasive and inva-sive ductal pancreatic cancer and its early detection in the mouse.Cancer Cell 2003;4:437–450.

35. Podsypanina K, Politi K, Beverly LJ, et al. Oncogene cooperation intumor maintenance and tumor recurrence in mouse mammarytumors induced by Myc and mutant Kras. Proc Natl Acad Sci U S A2008;105:5242–5247.

36. Fisher GH, Wellen SL, Klimstra D, et al. Induction and apoptoticregression of lung adenocarcinomas by regulation of a K-Rastransgene in the presence and absence of tumor suppressorgenes. Genes Dev 2001;15:3249–3262.

37. Chin L, Tam A, Pomerantz J, et al. Essential role for oncogenic Rasin tumour maintenance. Nature 1999;400:468–472.

38. Collins MA, Bednar F, Zhang Y, et al. Oncogenic Kras is requiredfor both the initiation and maintenance of pancreatic cancer inmice. J Clin Invest 2012;122:639–653.

39. Ying H, Kimmelman AC, Lyssiotis CA, et al. Oncogenic Kras main-tains pancreatic tumors through regulation of anabolic glucosemetabolism. Cell 2012;149:656–670.

40. Morris JP 4th, Cano DA, Sekine S, et al. Beta-catenin blocksKras-dependent reprogramming of acini into pancreatic cancerprecursor lesions in mice. J Clin Invest 2010;120:508–520.

41. Yachida S, Jones S, Bozic I, et al. Distant metastasis occurs lateduring the genetic evolution of pancreatic cancer. Nature 2010;467:1114–1147.

42. Ji B, Tsou L, Wang H, et al. Ras activity levels control the devel-opment of pancreatic diseases. Gastroenterology 2009;137:1072–1082.

43. Collins MA, Brisset JC, Zhang Y, et al. Metastatic pancreaticcancer is dependent on oncogenic Kras in mice. PLoS One 2012;7:e49707.

44. Clark CE, Hingorani SR, Mick R, et al. Dynamics of the immunereaction to pancreatic cancer from inception to invasion. CancerRes 2007;67:9518–9527.

45. Won JH, Zhang Y, Ji B, et al. Phenotypic changes in mousepancreatic stellate cell Ca2� signaling events following activationin culture and in a disease model of pancreatitis. Mol Biol Cell

2011;22:421–436.

4

4

4

5

5

5

5

5

5

5

5

5

5

6

6

6

6

6

6

6

6

BIO

LOG

YO

FTH

EPA

NCREA

S

May 2013 KRAS IN PANCREATIC TUMOR DEVELOPMENT AND PROGRESSION 1229

46. Xu Z, Vonlaufen A, Phillips PA, et al. Role of pancreatic stellatecells in pancreatic cancer metastasis. Am J Pathol 2010;177:2585–2596.

7. Erkan M, Adler G, Apte MV, et al. StellaTUM: current consensusand discussion on pancreatic stellate cell research. Gut 2012;61:172–178.

8. Jacobetz MA, Chan DS, Neesse A, et al. Hyaluronan impairsvascular function and drug delivery in a mouse model of pancre-atic cancer. Gut 2013;62:112–120.

9. Provenzano PP, Cuevas C, Chang AE, et al. Enzymatic targeting ofthe stroma ablates physical barriers to treatment of pancreaticductal adenocarcinoma. Cancer Cell 2012;21:418–429.

0. Charo C, Hwang RF, Arumugam T, et al. PGE2 activation of the EP4receptor regulates stellate cell activity in the pancreatic stromalmicroenvironment. Pancreas 2013;42:467–474.

1. Thayer SP, di Magliano MP, Heiser PW, et al. Hedgehog is an earlyand late mediator of pancreatic cancer tumorigenesis. Nature2003;425:851–856.

2. Berman DM, Karhadkar SS, Maitra A, et al. Widespread require-ment for Hedgehog ligand stimulation in growth of digestive tracttumours. Nature 2003;425:846–851.

3. Yauch RL, Gould SE, Scales SJ, et al. A paracrine requirement forhedgehog signalling in cancer. Nature 2008;455:406–410.

4. Olive KP, Jacobetz MA, Davidson CJ, et al. Inhibition of Hedgehogsignaling enhances delivery of chemotherapy in a mouse model ofpancreatic cancer. Science 2009;324:1457–1461.

5. Lesina M, Kurkowski MU, Ludes K, et al. Stat3/Socs3 activationby IL-6 transsignaling promotes progression of pancreatic intra-epithelial neoplasia and development of pancreatic cancer. Can-cer Cell 2011;19:456–469.

6. Chang DZ, Ma Y, Ji B, et al. Mast cells in tumor microenvironmentpromotes the in vivo growth of pancreatic ductal adenocarcinoma.Clin Cancer Res 2011;17:7015–7023.

7. Pylayeva-Gupta Y, Lee KE, Hajdu CH, et al. Oncogenic Kras-in-duced GM-CSF production promotes the development of pancre-atic neoplasia. Cancer Cell 2012;21:836–847.

8. Bayne LJ, Beatty GL, Jhala N, et al. Tumor-derived granulocyte-macrophage colony-stimulating factor regulates myeloid inflam-mation and T cell immunity in pancreatic cancer. Cancer Cell2012;21:822–835.

9. Hidalgo M, Maitra A. The hedgehog pathway and pancreatic can-cer. N Engl J Med 2009;361:2094–2096.

0. Chaika NV, Gebregiworgis T, Lewallen ME, et al. MUC1 mucinstabilizes and activates hypoxia-inducible factor 1 alpha to regu-late metabolism in pancreatic cancer. Proc Natl Acad Sci U S A2012;109:13787–13792.

1. Hanahan D, Weinberg RA. Hallmarks of cancer: the next genera-tion. Cell 2011;144:646–674.

2. Semenza GL. HIF-1: upstream and downstream of cancer metab-olism. Curr Opin Genet Dev 2010;20:51–56.

3. Cantor JR, Sabatini DM. Cancer cell metabolism: one hallmark,many faces. Cancer Discov 2012;2:881–898.

4. Neesse A, Michl P, Frese KK, et al. Stromal biology and therapy inpancreatic cancer. Gut 2011;60:861–868.

5. Singh A, Greninger P, Rhodes D, et al. A gene expression signa-ture associated with ”K-Ras addiction” reveals regulators of EMTand tumor cell survival. Cancer Cell 2009;15:489–500.

6. Hofmann I, Weiss A, Elain G, et al. K-RAS mutant pancreatictumors show higher sensitivity to MEK than to PI3K inhibition invivo. PLoS One 2012;7:e44146.

7. Collisson EA, Trejo CL, Silva JM, et al. A central role for RAF¡ME-K¡ERK signaling in the genesis of pancreatic ductal adenocarci-

noma. Cancer Discov 2012;2:685–693.

68. Cox AD, Der CJ. Ras family signaling: therapeutic targeting. Can-cer Biol Ther 2002;1:599–606.

69. Maurer T, Garrenton LS, Oh A, et al. Small-molecule ligands bindto a distinct pocket in Ras and inhibit SOS-mediated nucleotideexchange activity. Proc Natl Acad Sci U S A 2012;109:5299–5304.

70. Patgiri A, Yadav KK, Arora PS, et al. An orthosteric inhibitor of theRas-Sos interaction. Nat Chem Biol 2011;7:585–587.

71. Macdonald JS, McCoy S, Whitehead RP, et al. A phase II study offarnesyl transferase inhibitor R115777 in pancreatic cancer: aSouthwest oncology group (SWOG 9924) study. Invest New Drugs2005;23:485–487.

72. Elghazi L, Weiss AJ, Barker DJ, et al. Regulation of pancreasplasticity and malignant transformation by Akt signaling. Gastro-enterology 2009;136:1091–1103.

73. Eser S, Reiff N, Messer M, et al. Selective requirement of PI3K/PDK1 signaling for Kras oncogene-driven pancreatic cell plasticityand cancer. Cancer Cell 2013 Feb 27. Epub ahead of print.

74. Vigil D, Cherfils J, Rossman KL, et al. Ras superfamily GEFs andGAPs: validated and tractable targets for cancer therapy? Nat RevCancer 2010;10:842–857.

75. Neel NF, Martin TD, Stratford JK, et al. The RalGEF-Ral effectorsignaling network: the road less traveled for anti-Ras drug discov-ery. Genes Cancer 2011;2:275–287.

76. Lim KH, O’Hayer K, Adam SJ, et al. Divergent roles for RalA andRalB in malignant growth of human pancreatic carcinoma cells.Curr Biol 2006;16:2385–2394.

77. Feldmann G, Mishra A, Hong SM, et al. Inhibiting the cyclin-dependent kinase CDK5 blocks pancreatic cancer formation andprogression through the suppression of Ras-Ral signaling. CancerRes 2010;70:4460–4469.

78. Etienne-Manneville S, Hall A. Rho GTPases in cell biology. Nature2002;420:629–635.

79. Heid I, Lubeseder-Martellato C, Sipos B, et al. Early requirement ofRac1 in a mouse model of pancreatic cancer. Gastroenterology2011;141:719–730.

80. Fukuda A, Wang SC, Morris JP 4th, et al. Stat3 and MMP7contribute to pancreatic ductal adenocarcinoma initiation andprogression. Cancer Cell 2011;19:441–455.

81. Ebi H, Corcoran RB, Singh A, et al. Receptor tyrosine kinases exertdominant control over PI3K signaling in human KRAS mutantcolorectal cancers. J Clin Invest 2011;121:4311–4321.

82. Engelman JA, Chen L, Tan X, et al. Effective use of PI3K and MEKinhibitors to treat mutant Kras G12D and PIK3CA H1047R murinelung cancers. Nat Med 2008;14:1351–1356.

Received October 9, 2012. Accepted January 22, 2013.

Reprint requestsAddress requests for reprints to: Marina Pasca di Magliano, PhD,

Department of Surgery, University of Michigan, 4304 CC, SPC 5936,1500 E. Medical Center Drive, Ann Arbor, Michigan 48109. e-mail:marinapa@umich.edu; fax: (734) 647-9654; or Craig D. Lonsdon,PhD, The University of Texas MD Anderson Cancer Center,Department of Cancer Biology, 1515 Holcombe Boulevard, UnitNumber 953, Houston, Texas 77030. e-mail:clogsdon@mdanderson.org; ; fax: (713) 563-8986.

Conflicts of Interest

The authors disclose no conflicts.