The American Journal of Pathology, Vol. 182, No. 4, April 2013

ASIP

2013

AJP

CME Program

ajp.amjpathol.org

MINI-REVIEWMechanistic Insights into Self-Reinforcing ProcessesDriving Abnormal Histogenesis During the Development ofPancreatic CancerJuan L. Iovanna,* David L. Marks,yz Martin E. Fernandez-Zapico,yz and Raul Urrutiaz

From the Cancer Research Center of Marseille,* Inserm U1068, CNRS, UMR7258, Institute Paoli-Calmettes, Aix-Marseille University, Marseille, France;and the Division of Oncology Research,ythe Schulze Center for Novel Therapeutics, and the Laboratory of Epigenetics and Chromatin Dynamics,z

Translational Epigenomics Program, Center for Individualized Medicine, Gastroenterology Research Unit, Departments of Biophysics, Medicine,Biochemistry and Molecular Biology, Mayo Clinic, Rochester, Minnesota

CME Accreditation Statement: This activity (“ASIP 2013 AJP CME Program in Pathogenesis”) has been planned and implemented in accordance with the Essential Areas andpolicies of the Accreditation Council for Continuing Medical Education (ACCME) through the joint sponsorship of the American Society for Clinical Pathology (ASCP) and theAmerican Society for Investigative Pathology (ASIP). ASCP is accredited by the ACCME to provide continuing medical education for physicians.

The ASCP designates this journal-based CME activity (“ASIP 2013 AJP CME Program in Pathogenesis”) for a maximum of 48 AMA PRA Category 1 Credit(s)�. Physiciansshould only claim credit commensurate with the extent of their participation in the activity.

CME Disclosures: The authors of this article and the planning committee members and staff have no relevant financial relationships with commercial interests to disclose.

Accepted for publication

C

P

h

December 24, 2012.

Address correspondence toRaul Urrutia, M.D.,Division of Gastroenterology,Department of InternalMedicine, 200 First St. SW,Mayo Graduate School,Rochester, MN 55905.E-mail: urrutia.raul@mayo.edu.

opyright ª 2013 American Society for Inve

ublished by Elsevier Inc. All rights reserved

ttp://dx.doi.org/10.1016/j.ajpath.2012.12.004

Pancreatic ductal adenocarcinoma, one of the most feared lethal and painful diseases, is increasing inincidence. The poor prognosis of pancreatic ductal adenocarcinomaeaffected patients primarily is owingto our inability to develop effective therapies. Mechanistic studies of genetic, epigenetic, and cell-to-cellsignaling events are providing clues to molecular pathways that can be targeted in an attempt to cure thisdisease. The current review article seeks to draw inferences from available mechanistic knowledge to builda theoretical framework that can facilitate these approaches. This conceptual model considers pancreaticcancer as a tissue disease rather than an isolated epithelial cell problem, which develops and progresses inlarge part as a result of three positive feedback loops: i) genetic and epigenetic changes in epithelial cellsmodulate their interaction with mesenchymal cells to generate a dynamically changing process ofabnormal histogenesis, which drives more changes; ii) the faulty tissue architecture of neoplastic lesionsresults in unsynchronized secretion of signalingmolecules by cells, which generates an environment that ispoor in oxygen and nutrients; and iii) the increased metabolic needs of rapidly dividing cells serve as anevolutionary pressure for them to adapt to this adverse microenvironment, leading to the emergence ofresistant clones. We discuss how these concepts can guidemechanistic studies, as well as aid in the designof novel experimental therapeutics. (Am J Pathol 2013, 182: 1078e1086; http://dx.doi.org/10.1016/j.ajpath.2012.12.004)

Supported by grants from Ligue Contre le Cancer, INSERM, and INCA(Institute National du Cancer) (J.L.I.); the NIH (DK52913), the MayoClinic Center for Cell Signaling in Gastroenterology (P30DK084567), andthe Translational Epigenomic Program, Center for Individualized Medicine,Mayo Clinic (R.U.); and by the Mayo Clinic SPORE in Pancreatic Cancer(P50 CA102701 and CA136526 to M.E.F.-Z.).

The incidence of pancreatic ductal adenocarcinoma (PDAC)is increasing with more than 44,000 predicted new cases in theUnited States and 65,000 in Europe,1,2 with a 5-year survivalof less than 5%. PDAC arises from epithelial cells throughan accumulation of genetic and epigenetic alterations inoncogenes and tumor suppressors,3,4 which contribute toform precursor lesions5,6 known as pancreatic intraepithelialneoplasias (PanINs) (Figures 1 and 2). Less frequently, PDACmay progress from two types of cystic lesions:mucinous cystic

stigative Pathology.

.

neoplasms and intraductal papillary mucinous neoplasms. Inthis process, tumor cells proliferate and secrete molecules thatdrive their communication with surrounding cells. In the

Figure 1 Self-reinforcing processes that drive abnormal histogenesis during the development of pancreatic cancer. Diagrammatic representation ofpositive feedback loops that contribute to pancreatic carcinogenesis involves the progressive genetic and epigenetic changes in epithelial cells, whichmodulate their interaction with mesenchymal cells to generate a dynamically changing process of abnormal histogenesis to further drive more changes. Thefaulty tissue architecture of neoplastic lesions results in unsynchronized secretion of signaling molecules by cells, which generates an oxygen- and nutrient-poor environment as a result of aberrant angiogenesis. Finally, the increased metabolic needs of rapidly dividing cells serve as an evolutionary pressure forthem to adapt to this adverse microenvironment, leading to the emergence of resistant clones.

Abnormal Histogenesis

fashion of a self-reinforcing loop, surrounding cells alsoproliferate and secrete new substances, which initiate newcommunications among themselves, with other noncancer celltypes within the tumor (Figure 3).

This extended cellular network generates a dynamic tumormicroenvironment that influences genetically heterogeneous

Figure 2 Histologic correlates of abnormal histogenesis during pancreatic cagenic animals were stained using the Masson trichromic method, in which epitheliseries of micrographs show that from the beginning, pancreatic cancer developmenmesenchyma. A: PanIN1A lesion formed by cells with normal-shaped nuclei butlesion showing papillary projections formed by cells with normal-looking nuclei wlesion. C: PanIN2 lesion with abnormally shaped nuclei and typical papilar projecpiling up of nuclei with incipient atypia showing anisokaryosis, poikilokaryosis, anduct-like structure. E: PanIN3 lesion with extensive atypia showing the surroundcancerous lesions embedded in a dense desmoplasia.

The American Journal of Pathology - ajp.amjpathol.org

tumor cells and selects for highly proliferative and resistantclones. Epithelial and mesenchymal cells each contribute toremodeling the stroma into a dense fibrotic tissue (desmo-plasia) enriched in fibrillar collagens, stromal cells, and othermigratory cell populations. Accordingly, each component ofthe developing tumor takes an active role in the process of

ncer progression. Neoplastic pancreatic tissue from p48-cre/KrasG12D trans-al cells are labeled in red and the extracellular matrix is labeled in blue. Thist involves the tight interaction between epithelial cells and its surroundingshowing incipient nuclear piling up and increased cytoplasm. B: PanIN1Bith a mucin-containing cytoplasm that displaces nuclei to the base of thetions occupying the duct lumen. D: PanIN2 to 3 lesion showing the typicald papilar projections. Note that a ring of extracellular matrix surrounds theing ECM as a dense ring that deforms the ductular structures. F: Multifocal

1079

Figure 3 Epithelialemesenchymal interactionsduring the development of pancreatic cancer. Thefunctional co-evolution between pancreatic epithelialcells and their stromal counterparts from early pre-neoplastic to frank neoplastic lesions is shown.Several growth factors are secreted at abnormalamounts, times, and places to generate abnormalsignaling cascades that drive the communicationbetween both the epithelial compartment and thetissue microenvironment. As explained in the text,different growth factors exert their function either ina paracrine or autocrine manner to help form thetumors, desmoplasia, and generate an oxygen- andnutrient-poor tumor bed, impacting the pathobiologyof pancreatic cancer and contributing to its resistanceand aggressiveness. FGF, fibroblast growth factor;MMP, matrix metalloproteinase; PDGF, platelet-derived growth factor; SDF1, serum derived factor-1;SHH, Sonic hedgehog; VEGF, vascular endothelialgrowth factor.

Iovanna et al

carcinogenesis. Thus, dissecting the temporal and spatialsequence of events that drives these processes should providenew ways for therapeutically transforming pancreatic cancerfrom a rapidly fatal to a chronic and treatable, or even curable,disease.

Stromal Cells Are Key Players in PancreaticCancer Development

The stroma co-evolving with tumor cells during PDACprogression is composed of several cell types,7e9 includingfibroblasts, pancreatic stellate cells (PSCs), endothelial cells,bone marrowederived cells, as well as inflammatory andimmunoregulatory populations. Although fibroblasts con-tribute to pancreatic fibrosis, PSCs also contribute to thisprocess by secreting fibrillary collagens (type 1, type 3) andfibronectin.10 PSCs are located between acini and endothelialcells, where they influence the differentiation and function ofacinar cells or work as pericytes. PSCs represent less than 5%of cells within the normal gland10 but in tumors they prolif-erate to outnumber malignant epithelial cells.

These cells normally exist in a quiescent state marked bythe presence of vitamin Aerich lipid droplets, desmin, andglial-fibrillary-acidic protein. On tissue stress and injury,PSCs become activated myofibroblast-like cells, lose theirdroplets, express a-smooth muscle actin, and secrete extra-cellular matrix (ECM) proteins.8,11 The proliferation of thesea-smooth muscle actinepositive cells and the deposition ofcollagen in human PanIN and PDAC compared with normalpancreas is observed clearly in micrographs of human patientsamples and mice models. The idea that activated myofi-broblasts in the pancreas arise from PSCs has many propo-nents but still remains under debate12 because a few studiessuggest that they can originate from circulating bone mar-rowederived stem cells.13,14 Nevertheless, the currentlyavailable data make it impossible to ignore the fact thatstromal cells, both resident and migratory, are as important as

1080

epithelial cancer cells in the formation and pathobiology ofPDAC tumors.

Epithelial and Mesenchymal Interactions AreCritical in the Histogenesis of PanINs

Even at the PanIN stage, abnormal epithelial cells exert influ-ence on their surrounding microenvironment. PSCs are foundwithin PanINs from human pancreatic samples and mousemodels15,16 where they are stimulated by signals from othercells, including epithelial populations17,18(Figures 1 and 3).These signals include platelet-derived growth factor, fibro-blast growth factor, and transforming growth factor b1(TGF-b1). TGF-b stimulates collagen I, collagen III, andfibronectin biosynthesis, whereasfibroblast growth factor andplatelet-derived growth factor induce PSC proliferation.17e19

When orthotopically injected into mouse pancreata, TGF-b1eoverexpressing PDAC cells20 increase desmoplasia,showing that secreted factors from epithelial cells can inducethe mesenchymal reaction characteristic of pancreatic cancer.Sonic hedgehog, which is expressed during development andsilenced in mature cells, becomes re-expressed again inPanINs, increasing in concentration during the progression ofPanIN2 and 3 to PDACs.21 Expression of Sonic hedgehog inmice using the pdx1 promoter leads to PanIN formation.21 Incontrast, elastase 1edriven expression of this factor inpancreatic acinar cells does not result in PDAC but causes anexpansion of nestin-positive mesenchymal cells.22 Together,these studies provide relevant examples of how communi-cation between both malignant epithelial cells and theirnonmalignant mesenchymal counterparts are established denovo and aid the progression from a preneoplastic into a frankneoplastic phenotype.Other stromal responses include the effect of angiogenic

factors secreted by tumor cells that act on inflammatorycells.23e26 In exchange, these inflammatory cells triggera reinforcing positive feedback loop by producing IL-6, thus

ajp.amjpathol.org - The American Journal of Pathology

Abnormal Histogenesis

inducing proliferation of premalignant cells and theirprotection from apoptosis.27 These protumoral effects aresynergized by granulocyte-macrophage colony-stimulatingfactor,28 which is produced by both murine and human PanINcells.28 Granulocytes, mast cells, and macrophages locatedwithin human PDAC produce late protumoral factors, such asTGF-b29 and vascular endothelial growth factor. Therefore,an intense exchange of information between tumor cells,stromal cells, and immune cells contribute to arm PDACwithits aggressive nature.30 Interestingly, however, few studiesdirectly have investigated whether stromal cells from PanINssecrete growth factors and cytokines. Serum-derived factor-1is secreted by PSCs from both human and mouse PanINs,although the receptor for this factor, CXCR4, is expressedin PanINs and cancer epithelial cells.31e33 Thus, mesen-chymal serum-derived factor-1 stimulates CXCR4 on PDACcancer cells, leading to signals that promote their survival,proliferation, and migration.32e34 Combined, this knowledgesupports the conclusion that paracrine signaling from PSCs toPDAC cells stimulates the malignancy of PDAC cellsbeginning at the PanIN stage.

In this regard, complementary in vitro studies allow us todraw mechanistic information that explains the behavior ofstromal cells at the PanIN stage. Conditioned media fromPSCs, for instance, stimulates tumor cell proliferation andmigration.31,35,36When co-injected into nudemice, PSCs alsomediate protumoral effects on epithelial cells through thesecretion of epidermal growth factor, platelet-derived growthfactor, and fibroblast growth factor, whereas other factorssuch as TGF-b1, IL-1, and IL-6 act in an autocrine manner toexacerbate PSC activation.37,38 Activated PSCs produceseveral types of collagen9,39 and diffusible ECM componentsthat also help to stimulate PDAC cell proliferation andmigration.40e42 However, this desmoplasia is not homoge-neous in composition, which likely affects PanINs differentlydepending on whether they are located near areas of denselypacked fibers or juxtaposed to looser tissue layers. Thisheterogeneously remodeled ECM results from the localizedsecretion of powerful proteolytic machinery. Matrix metal-loproteinases, the best-studied ECM proteases, are secretedby many cells from the tumor, including PSCs, tumorepithelial cells, and leukocytes, to both degrade fibers andrelease growth factors sequestered in the ECM.43

All these studies revealed an exquisite communicationnetwork that mediates reciprocal signals among most cellswithin the tumor as well as messages going from cells to theECMandback.Suchfindingsguide us to the idea that drugs thatinhibit stromal cell activation or stromal/tumor cell communi-cation may have promising therapeutic value in PDAC.

Exploring the Concept of PDAC as a Disease ofAbnormal Tissue Morphogenesis

One of the striking characteristics of cancer is the tendency torecapitulate, in part, the morphologic features of the organ

The American Journal of Pathology - ajp.amjpathol.org

from which they originate. In many cases, these features arepathognomonic so that a well-trained pathologist can predictthe site of a primary tumor by looking at a metastasis. The factthat cancers do not display an unlimited variety of histologicprofiles but rather present with similar phenotypes indicatesthat our genes and the epigenome, which regulates theirtemporal and spatial pattern of expression, not only give riseto the normal phenotype, but also shape its neoplastic coun-terpart. This idea is supported by the fact that pancreaticepithelial cells display characteristics of progenitor pop-ulations found in early development of pluripotent stem cellswith similar features to those described in other forms ofcancer.44e46 For example, a number of signaling pathways(eg, hedgehog, Notch, and Wnt) and transcription factors(eg, Gata4, Nanog) associated with embryonic developmentare expressed in cancer cells.47e50 Cooperation betweenpancreatic tumors and surrounding tissue that leads to unde-sirable outcomes such as hypoxia and chemoresistance can beseen as resulting in part from a faulty morphogenetic cascadethat tries but fails to recapitulate normal histogenesis.

During normal development, a precise series of geneexpression events generate signals (eg, angiogenic factor) andoutcomes (eg, angiogenesis) in a controlled manner at theappropriate level, place, and time. In contrast, during carci-nogenesis, many molecules are produced simultaneously,leading to confusing morphogenic instructions that preventnormal programs from occurring or, if they do occur, theybegin and terminate at the wrong time and place. This lack ofsynchronization of morphogenetic cascades gives rise toa distorted architecture, which causes epithelial and mesen-chymal cells to miscommunicate. Thus, once abnormalmorphogenesis has begun, it causes additional architecturalproblems, behaving as another of the self-reinforcing mech-anisms that are critical to the formation of PDAC. Numerousstudies have shown the re-emergence of mesenchymal/epithelial developmental signaling pathways in PDAC. Earlyin development, for instance, hedgehog is absent from thedorsal and ventral pancreatic endoderm,51 but its forcedexpression in this tissue influences the stroma.51 Sonichedgehog expression in PDAC cells also stimulates theformation and remodeling of the tumor microenvironment ina manner that is abnormal. Thus, HH expression in PDACmay be thought of as a faulty recapitulation of an earlydevelopmental program of tissue morphogenesis. Similarexamples are foundwithfibroblast growth factors 1, 7, and 10,serum-derived factor-1, and many members of the TGF-b/activin/nodal/inhibin family of cytokines. Thus, together, thisknowledge serves as evidence that PDAC presents asa process of histogenesis that progresses abnormally asa result of signals arising and terminating inappropriately.

This concept is worthy of being compared and contrastedwith previous theories of carcinogenesis. Indeed, until the1970s, most informed investigators agreed that pancreaticcancer recapitulated developmental programs, and in thatsense our considerations are not new. However, most of theseearly theories viewed neoplastic lesions as arising from

1081

Iovanna et al

dedifferentiated cells, but current considerations in thisregard are different because dedifferentiated cells imply thatan adult population has reverted its phenotype to that found inan early developmental stage. If that was the case, neoplasticcells would have all of the instructions necessary for not onlyinitiating the formation, but also the completion, of an organ,develop normal architectural relationships within the organ,organize an efficient system of nutrient supply (vessels), andappropriate disposal of its products (secretory ducts). Inaddition, their genes would lack mutations and be expressedin a sequential manner as needed. Obviously, this is not thecase with pancreatic cancer, leading some investigators toalternatively propose that PDAC arises through a process oftransdifferentiation. Examples of this phenomenon in thepancreas are found in benign intraductal papillary mucinousneoplasms where the normal ductular cells transdifferentiateinto those resembling the gastrointestinal epithelium. Inter-estingly, this type of differentiation also has been proposedfor pancreatic ductal adenocarcinoma, primarily based onstudies in animal models, although the evidence supportingtransdifferentiation in human PDAC still remains contro-versial. Other investigators have invoked abnormal differ-entiation of stem cells as the origin for this type of cancer. It isnoteworthy, however, that what we call stem cells in solidtumor are different from those found in embryonic devel-opment with a differentiation potential, which is currentlyunclear.

Thus, these considerations lead us to propose thatpancreatic cancer cells neither go back in differentiation nordo they cross the differentiation barrier to another cell type.Rather, pancreatic cancer cells appear to have their owndifferentiation pathway, which in most part uses over-lapping gene networks with those turned on and off duringdevelopment to produce structures that are phenotypicallysimilar yet distinct from those found in embryonic devel-opment. In this context, we suggest that this neoplasiadevelops as a sui generis case of abnormal morphogenesis.This inference should help to develop a holistic view of thedisease as a process in which unsynchronized molecular andcellular events taking place in most of the structures thatform the tumor, and not exclusively in the epithelial cells,self-reinforce a path toward progression. When, where, andhow to intervene with these processes, although a significantchallenge, should be aided by the conceptual frameworkdiscussed in this article.

Angiogenesis, Hypoxia, and NutrientDeprivation Behave as a Positive FeedbackLoop Driving the Selection of Cells with HighMalignant Potential

Abnormal architectural relationships among structuresforming a defined tissue are one of the most obvious conse-quences of abnormal morphogenesis either during normaldevelopment (eg, thalidomide toxicity) or in cancer. In this

1082

section, we discuss how an increased demand of nutrients andoxygen by cells from the PDAC tumor bed is met by a relativedecrease in their supply imposed by both architectural andfunctional changes that occur during the remodeling of pre-neoplastic lesions as they becomemalignant. In particular, werecognize that a faulty architectural design results in anotherpositive feedback loop that leads to the establishment of moreaberrant structures. The law of supply and demand, imposedby a decrease in the offering of nutrients and an increase inenergetic needs by the pancreatic cancer, lends itself as a goodexample of this positive feedback loop. Indeed, by the timePDAC is diagnosed, most patients have invasive tumors withextensive desmoplasia, which are unexpectedly hypovascu-larized52,53 and poorly perfused, thereby generating lowoxygen and nutrient availability at tumor sites. Reducedvascularity also is found in chronic pancreatitis, suggestingthat hypoxic conditions begin early during the development ofPDAC and only increase in frank adenocarcinoma.54 Thesefeatures, along with the increased metabolic needs of rapidlyproliferating cells, function as a selection pressure towardtheir adaption to conditions that would be suboptimal tonormal cells, thus promoting tumor cell survival and theircharacteristic resistance to chemotherapy.55

Hypoxia Plays a Pivotal Role in the Selection ofAggressive Pancreatic Cancer Cell Clones

The effect of hypoxia on pancreatic cancer cells has beenmodeled extensively in vitro. For instance, cell linessurviving hypoxia show enhanced invasive and migratorypropensities,56,57 indicating that this condition favors cancerprogression. Although the precise mechanisms by whichhypoxia affects this process remain incompletely understood,changes in oxygen tension switch metabolism, modulateautophagy-mediated survival and death, modulate genomicinstability, and promote additional cell responses that furtherfacilitate malignant clonal selection.Mechanistically, oxygendeprivation triggers the stabilization of hypoxia-induciblefactor 1a (HIF-1a), which dimerizes with HIF-1b, under-goes nuclear translocation, and binds to a multitude ofhypoxia-responsive elements present in the cancer genome.These events activate a complex geneticeepigenetic programthat seeks to counteract the deleterious impact of decreasedoxygen tension.58 HIF-1a is overexpressed in PDAC,59

where it regulates gene networks involved in autophagy,glycolysis, angiogenesis, epithelial-to-mesenchymal transi-tion, and metastasis.58,60 This is a good example of howtranscription and chromatin-induced changes in geneexpression contribute to pancreatic carcinogenesis. Anotherexample is revealed by studies on the small chromatinbinding protein, nuclear protein 1. Nuclear protein 1 isoverexpressed in PDAC,61,62 where it is induced by hypoxia,glucose deprivation, and other stresses.63,64 Genetic inacti-vation of nuclear protein 1 sensitizes PDAC cells to hypoxiaand nutrient deprivation and induces their death by

ajp.amjpathol.org - The American Journal of Pathology

Abnormal Histogenesis

autophagy.64 In this manner, nuclear protein 1 works asa sensor of tumor cell stresses that, when activated, protectstumor cells against damage caused by a hostile environment.Thus, in summary, PDAC cells become primed for survivalunder hypoxic environments by deploying a defined geneexpression network, which ultimately results in a moreresistant phenotype, illustrating another positive feedbackloop that drives PDAC progression.

A Metabolic Switch Is a Defining Feature ofHighly Aggressive Pancreatic Cancer

Hypoxia not only affects epithelial cells, but also stimulatesthe activation of mesenchymal cells that secrete angiogenicfactors such as vascular endothelial growth factor in anattempt to increase O2 through enhanced vascular perme-ability and angiogenesis.65,66 Despite this mechanism,PDAC still fails to increase its vascularity as robustly asother tumors, suggesting that the function of proangiogenicfactors is counterbalanced by antiangiogenic molecules suchas endostatin.65 Thus, both angiogenic and antiangiogenicsubstances secreted in different amounts at various timesduring the development of pancreatic cancer ultimately leadto a functional equilibrium characterized by the hypo-vascularity that is typical of this malignancy.

One of the major consequences of intratumoral hypoxia isthe metabolic switch from the primary use of oxidativephosphorylation to reliance on the glycolysis that occurs tomeet the requirements of tumor proliferation under lowoxygen and low nutrient supply.67 Although normal cells relyprimarily (90%) on oxidative phosphorylation, approximately50% of cellular energy is produced by glycolysis in tumorcells, with the remainder being generated in the mitochondria,even in the presence of ample oxygen to fuel mitochondrialrespiration, a phenomenon known as the Warburg effect.68

This metabolic transformation is a consequence of acquiredmutations that lead to cancer, coupled with hypoxic andnutrient-deficient conditions and faulty architectural relation-ships among cells and vessels, all of which contribute to theselection of tumor clones able to survive this hostile envi-ronment. Congruently, PDAC tumors in vivo show alterationsin pathways involved in thismetabolic switch,69which even atthe PanIN2 and 3 stages lead to increased expression of theglucose transporter 1 (GLUT1), a protein that is critical forglucose uptake by tumor cells.70 Subsequently, hypoxia alsofunctions as a key mechanism that increases glycolysis. Forexample, thehypoxia-induced stabilization ofHIF-1a not onlyregulates angiogenesis, but also directly up-regulates severalgenes involved in this process, such as GLUT1, hexokinase 1and 2, lactate dehydrogenase A, and lactate transporter mon-ocarboxylate transporter 4.71

Defined genetic alterations, such as loss of p53 function,also stimulate glycolysis at several levels.71 Similarly, acti-vation of Ras inhibits pyruvate kinase and stimulates GLUT1translocation to the plasma membrane.71 Consequently,

The American Journal of Pathology - ajp.amjpathol.org

turning off the expression of an inducible KrasG12D oncogenein PDAC cells leads to reduced glucose uptake, decreasedexpression of rate-limiting glycolytic enzymes, decreasedlactate production, and reduced channeling of glucosemetabolites into nonoxidative anabolic pathways such ashexosamine biosynthesis.72 These changes in metabolicprogramming by KrasG12D are mediated, at least in part, bymitogen-activated kinases and Myc.72 Therefore, shiftingsubstrates from energy production to molecular synthesisseems to provide tumor cells with advantages that help toaccount for its increased survival and aggressiveness.Glycolytic pathways produce ATP and pyruvate thatgenerate bioproducts, which enter the pentose phosphatepathway to generate ribose-5-phosphate and NADPH, keyintermediates in nucleotide biosynthesis. During glycolysis,conversion of glucose to fructose provides an essentialsubstrate for the nonoxidative pentose phosphate pathway.Pyruvate, another major intermediate, is converted to lactateby lactate dehydrogenase and secreted into the extracellularenvironment. Lactate secreted by tumor cells could beabsorbed by stromal cells and converted to pyruvate that isused either for oxidative phosphorylation or secreted againand then taken up by tumor cells for use as glycolytic fuel.71,73

A recent study investigated the expression of glycolyticproteins in PDAC tumors and associated stromal cells to showthat both primary PDAC tumors and metastases express highlevels of proteins involved in glycolysis (eg, GLUT1 andhexokinase 2) compared with normal tissue.69 GLUT1 alsois increased in tumor-associated stromal cells, althoughto a lesser degree than in the tumor cells.69 In totality, theepithelial cell and tumor microenvironment relationship notonly regulates morphogenesis by paracrine/autocrine mech-anisms and drives angiogenesis, but also actively participatesin the recycling of metabolites within the tumor. This recy-cling of metabolites, as in the case of the lactate cycle, iscritical tomaintain the energetic needs of the tumor, leading tothe prediction that a better understanding of these interactionsbetween cancer cells and the surrounding populations can beexploited for diagnostic (eg, labeled metabolites) or thera-peutic efforts (eg, metabolic manipulations).

Conclusions

The particularly bad prognosis that is observed for patientswith PDAC is at least in part owing to the pancreatic trans-formed cells that develop in cooperation with surroundingstromal cells. Activated PSCs and immune cells may berecruited in response to pancreatic injury and/or mutantepithelial cells, and initially may provide growth factors andcytokines that nurture preneoplastic epithelial cells, selectingfor clones with key mutations, such as KRAS. Because PSCsproliferate and secrete ECM proteins, an adverse desmo-plastic microenvironment is formed, characterized byhypoxia and nutrient starvation as a consequence of itshypovascularization, which selects for aggressively invasive,

1083

Iovanna et al

chemotherapy-resistant tumor cells. Hypovascularization-induced hypoxia activates HIF-1a accumulation in tumorcells, which together with Kras activation increases auto-phagy and induces a metabolic switch to glycolysis. The lowvascularization of tumors also impedes the delivery ofchemotherapeutic drugs. On the other hand, the presence ofhigh intratumoral concentrations of antiapoptotic cytokinesalso enhances tumor cell survival. Tumor cells show somecharacteristics of embryonic cells and, through interactionswith stromal cells, participate in a warped recapitulation oftissue morphogenesis. Together, these findings suggest thatthe development of therapeutic approaches that target stromalcells, stromal/tumor communication, hypoxia, and/or glyco-lytic metabolism may bring novel treatments for PDAC.

References

1. Cascinu S, Falconi M, Valentini V, Jelic S, on behalf of the EGWG:Pancreatic cancer: ESMO Clinical Practice Guidelines for diagnosis,treatment and follow-up. Ann Oncol 2010, 21:v55ev58

2. Siegel R, Naishadham D, Jemal A: Cancer statistics, 2012. CA CancerJ Clin 2012, 62:10e29

3. Lomberk G, Mathison AJ, Grzenda A, Urrutia R: The sunset ofsomatic genetics and the dawn of epigenetics: a new frontier inpancreatic cancer research. Curr Opin Gastroenterol 2008, 24:597e602

4. McCleary-Wheeler AL, Lomberk GA, Weiss FU, Schneider G,Fabbri M, Poshusta TL, Dusetti NJ, Baumgart S, Iovanna JL,Ellenrieder V, Urrutia R, Fernandez-Zapico ME: Insights into theepigenetic mechanisms controlling pancreatic carcinogenesis. CancerLett 2013, 328:212e221

5. Maitra A, Fukushima N, Takaori K, Hruban RH: Precursors to inva-sive pancreatic cancer. Adv Anat Pathol 2005, 12:81e91

6. Scarlett C, Salisbury E, Biankin A, Kench J: Precursor lesions inpancreatic cancer: morphological and molecular pathology. Pathology2011, 43:183e200

7. Neesse A, Michl P, Frese KK, Feig C, Cook N, Jacobetz MA,Lolkema MP, Buchholz M, Olive KP, Gress TM, Tuveson DA:Stromal biology and therapy in pancreatic cancer. Gut 2011, 60:861e868

8. Omary MB, Lugea A, Lowe AW, Pandol SJ: The pancreatic stellatecell: a star on the rise in pancreatic diseases. J Clin Invest 2007, 117:50e59

9. Shields MA, Dangi-Garimella S, Redig AJ, Munshi HG: Biochemicalrole of the collagen-rich tumour microenvironment in pancreatic cancerprogression. Biochem J 2012, 441:541e552

10. Bachem M, Schneider E, Gross H, Weidenbach H, Schmid R,Menke A, Siech M, Beger H, Grunert A, Adler G: Identification,culture, and characterization of pancreatic stellate cells in rats andhumans. Gastroenterology 1998, 115:421e432

11. Apte MV, Park S, Phillips PA, Santucci N, Goldstein D, Kumar RK,Ramm GA, Buchler M, Friess H, McCarroll JA, Keogh G, Merrett N,Pirola R, Wilson JS: Desmoplastic reaction in pancreatic cancer: roleof pancreatic stellate cells. Pancreas 2004, 29:179e187

12. Kim EJ, Simeone DM: Advances in pancreatic cancer. Curr OpinGastroenterol 2011, 27:460e466

13. Scarlett CJ, Colvin EK, Pinese M, Chang DK, Morey AL,Musgrove EA, Pajic M, Apte M, Henshall SM, Sutherland RL,Kench JG, Biankin AV: Recruitment and activation of pancreaticstellate cells from the bone marrow in pancreatic cancer: a model oftumor-host interaction. PLoS One 2011, 6:e26088

14. Watanabe T, Masamune A, Kikuta K, Hirota M, Kume K, Satoh K,Shimosegawa T: Bone marrow contributes to the population of

1084

pancreatic stellate cells in mice. Am J Physiol Gastrointest LiverPhysiol 2009, 297:G1138eG1146

15. Aichler M, Seiler C, Tost M, Siveke J, Mazur PK, Da Silva-Buttkus P,Bartsch DK, Langer P, Chiblak S, Dürr A, Höfler H, Klöppel G,Müller-Decker K, Brielmeier M, Esposito I: Origin of pancreatic ductaladenocarcinoma from atypical flat lesions: a comparative study intransgenic mice and human tissues. J Pathol 2012, 226:723e734

16. Collins MA, Bednar F, Zhang Y, Brisset J-C, Galban S, Galban CJ,Rakshit S, Flannagan KS, Adsay NV, Pasca di Magliano M: Onco-genic Kras is required for both the initiation and maintenance ofpancreatic cancer in mice. J Clin Invest 2012, 122:639e653

17. Apte MV, Haber PS, Darby SJ, Rodgers SC, McCaughan GW,Korsten MA, Pirola RC, Wilson JS: Pancreatic stellate cells are acti-vated by proinflammatory cytokines: implications for pancreaticfibrogenesis. Gut 1999, 44:534e541

18. Bachem MG, Schunemann M, Ramadani M, Siech M, Beger H,Buck A, Zhou S, Schmid-Kotsas A, Adler G: Pancreatic carcinomacells induce fibrosis by stimulating proliferation and matrix synthesisof stellate cells. Gastroenterology 2005, 128:907e921

19. Schneider E, Schmid-Kotsas A, Zhao J, Weidenbach H, Schmid RM,Menke A, Adler G, Waltenberger J, Grunert A, Bachem MG: Identi-fication of mediators stimulating proliferation and matrix synthesis ofrat pancreatic stellate cells. Am J Physiol Cell Physiol 2001, 281:C532eC543

20. Lohr M, Schmidt C, Ringel J, Kluth M, Muller P, Nizze H,Jesnowski R: Transforming growth factor-beta1 induces desmoplasiain an experimental model of human pancreatic carcinoma. Cancer Res2001, 61:550e555

21. Thayer SP, di Magliano MP, Heiser PW, Nielsen CM, Roberts DJ,Lauwers GY, Qi YP, Gysin S, Fernández-del Castillo C, Yajnik V,Antoniu B, McMahon M, Warshaw AL, Hebrok M: Hedgehog is anearly and late mediator of pancreatic cancer tumorigenesis. Nature2003, 425:851e856

22. Fendrich V, Oh E, Bang S, Karikari C, Ottenhof N, Bisht S, Lauth M,Brossart P, Katsanis N, Maitra A, Feldmann G: Ectopic overexpressionof Sonic Hedgehog (Shh) induces stromal expansion and metaplasia inthe adult murine pancreas. Neoplasia 2011, 13:923e930

23. Boreddy SR, Sahu RP, Srivastava SK: Benzyl isothiocyanatesuppresses pancreatic tumor angiogenesis and invasion by inhibitingHIF-alpha/VEGF/rho-GTPases: pivotal role of STAT-3. PLoS One2011, 6:e25799

24. Chu GC, Kimmelman AC, Hezel AF, DePinho RA: Stromal biology ofpancreatic cancer. J Cell Biochem 2007, 101:887e907

25. Niedergethmann M, Hildenbrand R, Wostbrock B, Hartel M,Sturm JW, Richter A, Post S: High expression of vascular endothelialgrowth factor predicts early recurrence and poor prognosis aftercurative resection for ductal adenocarcinoma of the pancreas. Pancreas2002, 25:122e129

26. Su Y, Loos M, Giese N, Hines OJ, Diebold I, Gorlach A, Metzen E,Pastorekova S, Friess H, Buchler P: PHD3 regulates differentiation,tumour growth and angiogenesis in pancreatic cancer. Br J Cancer2010, 103:1571e1579

27. Corcoran RB, Contino G, Deshpande V, Tzatsos A, Conrad C,Benes CH, Levy DE, Settleman J, Engelman JA, Bardeesy N: STAT3plays a critical role in KRAS-induced pancreatic tumorigenesis. CancerRes 2011, 71:5020e5029

28. Pylayeva-Gupta Y, Lee KE, Hajdu CH, Miller G, Bar-Sagi D: Onco-genic Kras-induced GM-CSF production promotes the development ofpancreatic neoplasia. Cancer Cell 2012, 21:836e847

29. Aoyagi Y, Oda T, Kinoshita T, Nakahashi C, Hasebe T, Ohkohchi N,Ochiai A: Overexpression of TGF-beta by infiltrated granulocytescorrelates with the expression of collagen mRNA in pancreatic cancer.Br J Cancer 2004, 91:1316e1326

30. Wachsmann MB, Pop LM, Vitetta ES: Pancreatic ductal adenocarci-noma: a reviewof immunologic aspects. J InvestMed2012, 60:643e663

31. Gao Z, Wang X, Wu K, Zhao Y, Hu G: Pancreatic stellate cellsincrease the invasion of human pancreatic cancer cells through the

ajp.amjpathol.org - The American Journal of Pathology

Abnormal Histogenesis

stromal cell-derived factor-1/CXCR4 axis. Pancreatology 2010, 10:186e193

32. Koshiba T, Hosotani R, Miyamoto Y, Ida J, Tsuji S, Nakajima S,Kawaguchi M, Kobayashi H, Doi R, Hori T, Fujii N, Imamura M:Expression of stromal cell-derived factor 1 and CXCR4 ligand receptorsystem in pancreatic cancer: a possible role for tumor progression. ClinCancer Res 2000, 6:3530e3535

33. Thomas RM, Kim J, Revelo-Penafiel MP, Angel R, Dawson DW,Lowy AM: The chemokine receptor CXCR4 is expressed in pancreaticintraepithelial neoplasia. Gut 2008, 57:1555e1560

34. Marchesi F, Monti P, Leone BE, Zerbi A, Vecchi A, Piemonti L,Mantovani A, Allavena P: Increased survival, proliferation, andmigration in metastatic human pancreatic tumor cells expressingfunctional CXCR4. Cancer Res 2004, 64:8420e8427

35. Hwang RF, Moore T, Arumugam T, Ramachandran V, Amos KD,Rivera A, Ji B, Evans DB, Logsdon CD: Cancer-associated stromalfibroblasts promote pancreatic tumor progression. Cancer Res 2008,68:918e926

36. Vonlaufen A, Joshi S, Qu C, Phillips PA, Xu Z, Parker NR, Toi CS,Pirola RC, Wilson JS, Goldstein D, Apte MV: Pancreatic stellate cells:partners in crime with pancreatic cancer cells. Cancer Res 2008, 68:2085e2093

37. Aoki H, Ohnishi H, Hama K, Ishijima T, Satoh Y, Hanatsuka K,Ohashi A, Wada S, Miyata T, Kita H, Yamamoto H, Osawa H, Sato K,Tamada K, Yasuda H, Mashima H, Sugano K: Autocrine loop betweenTGF-beta1 and IL-1beta through Smad3- and ERK-dependent path-ways in rat pancreatic stellate cells. Am J Physiol Cell Physiol 2006,290:C1100eC1108

38. Aoki H, Ohnishi H, Hama K, Shinozaki S, Kita H, Yamamoto H,Osawa H, Sato K, Tamada K, Sugano K: Existence of autocrineloop between interleukin-6 and transforming growth factor-beta1 inactivated rat pancreatic stellate cells. J Cell Biochem 2006, 99:221e228

39. Erkan M, Weis N, Pan Z, Schwager C, Samkharadze T, Jiang X,Wirkner U, Giese N, Ansorge W, Debus J, Huber P, Friess H,Abdollahi A, Kleeff J: Organ-, inflammation- and cancer specifictranscriptional fingerprints of pancreatic and hepatic stellate cells. MolCancer 2010, 9:88

40. Grzesiak JJ, Bouvet M: The alpha2beta1 integrin mediates themalignant phenotype on type I collagen in pancreatic cancer cell lines.Br J Cancer 2006, 94:1311e1319

41. Koenig A, Mueller C, Hasel C, Adler G, Menke A: Collagen type Iinduces disruption of E-cadherin-mediated cell-cell contacts andpromotes proliferation of pancreatic carcinoma cells. Cancer Res 2006,66:4662e4671

42. Menke A, Philippi C, Vogelmann R, Seidel B, Lutz MP, Adler G,Wedlich D: Down-regulation of E-cadherin gene expression bycollagen type I and type III in pancreatic cancer cell lines. Cancer Res2001, 61:3508e3517

43. Korc M: Pancreatic cancer-associated stroma production. Am J Surg2007, 194:S84eS86

44. Hezel AF, Kimmelman AC, Stanger BZ, Bardeesy N, DePinho RA:Genetics and biology of pancreatic ductal adenocarcinoma. Genes Dev2006, 20:1218e1249

45. Micalizzi D, Farabaugh S, Ford H: Epithelial-mesenchymal transitionin cancer: parallels between normal development and tumor progres-sion. J Mammary Gland Biol Neoplasia 2010, 15:117e134

46. Strizzi L, Hardy KM, Seftor EA, Costa FF, Kirschmann DA,Seftor REB, Postovit L-M, Hendrix MJC: Development and cancer: atthe crossroads of Nodal and Notch signaling. Cancer Res 2009, 69:7131e7134

47. Harris PJ, Speranza G, Dansky Ullmann C: Targeting embryonicsignaling pathways in cancer therapy. Expert Opin Ther Targets 2012,16:131e145

48. Karafin M, Cummings C, Fu B, Iacobuzio-Donahue C: The develop-mental transcription factor Gata4 is overexpressed in pancreatic ductaladenocarcinoma. Int J Clin Exp Pathol 2009, 3:47e55

The American Journal of Pathology - ajp.amjpathol.org

49. Lomberk G, Fernandez-Zapico ME, Urrutia R: When developmentalsignaling pathways go wrong and their impact on pancreatic cancerdevelopment. Curr Opin Gastroenterol 2005, 21:555e560

50. Lonardo E, Hermann PC, Mueller M-T, Huber S, Balic A, Miranda-Lorenzo I, Zagorac S, Alcala S, Rodriguez-Arabaolaza I, Ramirez JC,Torres-Ruiz R, Garcia E, Hidalgo M, Cebrian DA, Heuchel R, Lohr M,Berger F, Bartenstein P, Aicher A, Heeschen C: Nodal/activinsignaling drives self-renewal and tumorigenicity of pancreatic cancerstem cells and provides a target for combined drug therapy. Cell StemCell 2011, 9:433e446

51. Apelqvist A, Ahlgren U, Edlund H: Sonic hedgehog directs specialisedmesoderm differentiation in the intestine and pancreas. Curr Biol 1997,7:801e804

52. Olive KP, Jacobetz MA, Davidson CJ, Gopinathan A, McIntyre D,Honess D, Madhu B, Goldgraben MA, Caldwell ME, Allard D,Frese KK, DeNicola G, Feig C, Combs C, Winter SP, Ireland-Zecchini H, Reichelt S, Howat WJ, Chang A, Dhara M, Wang L,Ruckert F, Grutzmann R, Pilarsky C, Izeradjene K, Hingorani SR,Huang P, Davies SE, Plunkett W, Egorin M, Hruban RH,Whitebread N, McGovern K, Adams J, Iacobuzio-Donahue C,Griffiths J, Tuveson DA: Inhibition of Hedgehog signaling enhancesdelivery of chemotherapy in a mouse model of pancreatic cancer.Science 2009, 324:1457e1461

53. Sofuni A, IijimaH,Moriyasu F, NakayamaD, ShimizuM, Nakamura K,Itokawa F, Itoi T: Differential diagnosis of pancreatic tumors usingultrasound contrast imaging. J Gastroenterol 2005, 40:518e525

54. Koong AC, Mehta VK, Le QT, Fisher GA, Terris DJ, Brown JM,Bastidas AJ, Vierra M: Pancreatic tumors show high levels of hypoxia.Int J Radiat Oncol Biol Phys 2000, 48:919e922

55. Hockel M, Vaupel P: Tumor hypoxia: definitions and current clinical,biologic, and molecular aspects. J Natl Cancer Inst 2001, 93:266e276

56. Rausch V, Liu L, Apel A, Rettig T, Gladkich J, Labsch S,Kallifatidis G, Kaczorowski A, Groth A, Gross W, Gebhard MM,Schemmer P, Werner J, Salnikov AV, Zentgraf H, Büchler MW,Herr I: Autophagy mediates survival of pancreatic tumour-initiatingcells in a hypoxic microenvironment. J Pathol 2012, 227:325e335

57. Ridgway PF, Ziprin P, Alkhamesi N, Paraskeva PA, Peck DH,Darzi AW: Hypoxia augments gelatinase activity in a variety ofadenocarcinomas in vitro. J Surg Res 2005, 124:180e186

58. Majmundar AJ, Wong WJ, Simon MC: Hypoxia-inducible factors andthe response to hypoxic stress. Mol Cell 2010, 40:294e309

59. Buchler P, Reber HA, Buchler M, Shrinkante S, Buchler MW,Friess H, Semenza GL, Hines OJ: Hypoxia-inducible factor 1 regulatesvascular endothelial growth factor expression in human pancreaticcancer. Pancreas 2003, 26:56e64

60. Bennewith K, Dedhar S: Targeting hypoxic tumour cells to overcomemetastasis. BMC Cancer 2011, 11:504

61. Su S-B, Motoo Y, Iovanna JL, Berthezene P, Xie M-J, Mouri H,Ohtsubo K, Matsubara F, Sawabu N: Overexpression of p8 is inverselycorrelated with apoptosis in pancreatic cancer. Clin Cancer Res 2001,7:1320e1324

62. Su S-B, Motoo Y, Iovanna JL, Xie M-J, Mouri H, Ohtsubo K,Yamaguchi Y, Watanabe H, Okai T, Matsubara F, Sawabu N:Expression of p8 in human pancreatic cancer. Clin Cancer Res 2001, 7:309e313

63. Cano CE, Hamidi T, Sandi MJ, Iovanna JL: Nupr1: the Swiss-knife ofcancer. J Cell Physiol 2011, 226:1439e1443

64. Hamidi T, Cano CE, Grasso D, Garcia MN, Sandi MJ, Calvo EL,Dagorn J-C, Lomberk G, Urrutia R, Goruppi S, Carracedo A,Velasco G, Iovanna JL: Nupr1-aurora kinase A pathway providesprotection against metabolic stress-mediated autophagic-associated celldeath. Clin Cancer Res 2012, 18:5234e5246

65. Erkan M, Reiser-Erkan C, Michalski CW, Deucker S, Sauliunaite D,Streit S, Esposito I, Friess H, Kleeff J: Cancer-stellate cell interactionsperpetuate the hypoxia-fibrosis cycle in pancreatic ductal adenocarci-noma. Neoplasia 2009, 11:497e508

1085

Iovanna et al

66. Masamune A, Kikuta K, Watanabe T, Satoh K, Hirota M,Shimosegawa T: Hypoxia stimulates pancreatic stellate cells to inducefibrosis and angiogenesis in pancreatic cancer. Am J Physiol Gastro-intest Liver Physiol 2008, 295:G709eG717

67. Vasseur S, Tomasini R, Tournaire R, Iovanna JL: Hypoxia inducedtumor metabolic switch contributes to pancreatic cancer aggressive-ness. Cancer 2010, 2:2138e2152

68. Warburg O: On the origin of cancer cells. Science 1956, 123:309e31469. Chaika NV, Yu F, Purohit V, Mehla K, Lazenby AJ, DiMaio D,

Anderson JM, Yeh JJ, Johnson KR, Hollingsworth MA, Singh PK:Differential expression of metabolic genes in tumor and stromalcomponents of primary and metastatic loci in pancreatic adenocarci-noma. PLoS One 2012, 7:e32996

70. Pizzi S, Porzionato A, Pasquali C, Guidolin D, Sperti C, Fogar P,Macchi V, De Caro R, Pedrazzoli S, Parenti A: Glucose transporter-1

1086

expression and prognostic significance in pancreatic carcinogenesis.Histol Histopathol 2009, 24:175e185

71. Kroemer G, Pouyssegur J: Tumor cell metabolism: cancer’s Achilles’heel. Cancer Cell 2008, 13:472e482

72. Ying H, Kimmelman AC, Lyssiotis CA, Hua S, Chu GC, Fletcher-Sananikone E, Locasale JW, Son J, Zhang H, Coloff JL, Yan H,Wang W, Chen S, Viale A, Zheng H, Paik JH, Lim C, Guimaraes AR,Martin ES, Chang J, Hezel AF, Perry SR, Hu J, Gan B, Xiao Y,Asara JM, Weissleder R, Wang YA, Chin L, Cantley LC,DePinho RA: Oncogenic Kras maintains pancreatic tumors throughregulation of anabolic glucose metabolism. Cell 2012, 149:656e670

73. Koukourakis MI, Giatromanolaki A, Harris AL, Sivridis E: Compar-ison of metabolic pathways between cancer cells and stromal cells incolorectal carcinomas: a metabolic survival role for tumor-associatedstroma. Cancer Res 2006, 66:632e637

ajp.amjpathol.org - The American Journal of Pathology