Fishing for cures: The alLURE of using zebrafish to develop precision oncology therapies

14 Pages·2017·1.55 MB·English

by  Matteo Astone

www.nature.com/npjprecisiononcology
REVIEW ARTICLE OPEN
fi
Fishing for cures: The alLURE of using zebra sh to develop
precision oncology therapies
MatteoAstone1,Erin N. Dankert1,Sk.KayumAlam1 and Luke H.Hoeppner1
Zebrafishhaveproventobeavaluablemodeltostudyhumancancerbiologywiththeultimateaimofdevelopingnewtherapies.
Danio rerioare amenableto invivoimaging, high-throughputdrug screening, mutagenesis, and transgenesis, and theyshare
histologicaland genetic similaritieswith Homosapiens.The significanceof zebrafishin thefieldofprecision oncology israpidly
emerging.Indeed,modelingcancerinzebrafishhasalreadybeenusedtoidentifytumorbiomarkers,definetherapeutictargetsand
provideaninvivoplatformfordrugdiscovery.Newzebrafishstudiesarestartingtopavethewaytodirectindividualizedclinical
applications.Patient-derivedcancercellxenograftmodelshavedemonstratedthefeasibilityofusingzebrafishasareal-timeavatar
ofprognosisanddrugresponsetoidentifythemostidealtherapyforanindividualpatient.Geneticcancermodelinginzebrafish,
nowfacilitated by rapidly evolvinggenome editing techniques, represents another innovativeapproach to recapitulate human
oncogenesisanddevelopindividualizedtreatments.Utilizingzebrafishtodesigncustomizableprecisiontherapieswillimprovethe
clinicaloutcomeof patientsafflicted withcancer.
npjPrecision Oncology (2017) 1:39 ;doi:10.1038/s41698-017-0043-9
INTRODUCTION etc.)requireasimilartimeframeasmurinemodels.Thesmallsize,
Precision medicine in oncology arises from recognition that external development, and transparency of zebrafish embryos
patient-specific clinical, genetic, and molecular features dictate make them amenable to fluorescent live imaging to monitor
effectiveness of a given treatment. Therefore, precision oncology physiological processes (e.g., development, morphogenesis,
seeks to identify the most effective therapy for an individual angiogenesis, etc.) and pathological phenomena (e.g., cancer
patient, based on characterization of their cancer. The develop- initiation, tumorigenesis, metastasis, etc.). Taken together, the
mentofgenomictechnologiesandmoleculardiagnosticsenables attractive features of the zebrafish model system underscore the
detection of cancer biomarkers. These relevant abnormalities reasons it has gained prominence in the study of cancer and
associated with specific cancers lead to the identification of servesasanexcellentadditiontoothercommononcologymodels
actionable targets. Diagnostic (associated with the presence of a andplatforms.
specific pathophysiological state), prognostic (associated with This review aims to provide an overview of how current
diseaseoutcome),andpredictive(associatedwithdrugresponse) zebrafishcancerstudieslaythegroundworkforutilizationofthis
cancer biomarkers guide clinical treatment decisions and direct model organism in precision oncology, highlighting specific
the use of drugs that modulate the activity of the specific studies oriented to the development of zebrafish-based patient-
actionabletarget.1,2 specific approaches for cancer treatment. The challenges and
Zebrafish (Danio rerio) have rapidly emerged as a promising shortcomingsofzebrafishcancerstudiesarepresentedasareasof
animal model of human cancer. Histological, molecular, and thefieldrequiring advancements andgrowth.
genetic similarities to Homo sapiens facilitate zebrafish studies of
human malignancies. Zebrafish are amenable to in vivo fluor-
escent imaging, chemical and genetic screens, transgenesis, and ZEBRAFISH:FROMMODELINGCLASSICCANCERRESEARCHTO
high-throughput mutagenesis assays, which have brought zebra- PRECISION ONCOLOGY
fishtothecenterstageoffutureadvancesinthefieldofprecision The application of the zebrafish model to precision oncology
oncology.3 A variety of attributes have contributed to the remainsinitsinfancy,andtherearenotyetexamplesofdirectuse
emergence of zebrafish as an attractive vertebrate model of zebrafish to guide patient-specific cancer treatments in the
organism. Zebrafish are easy and inexpensive to maintain and clinic. However, the field has matured enough to move toward
breedwithhighfecundity,whichfacilitateslargestudiesandhigh- this aim in the near future. Modeling cancer in zebrafish has
throughput in vivo assays. Another advantage of working with provided important insights that contribute to the development
zebrafish embryos is their conserved vertebrate features develop of precision oncology as well as straightforward examples of
rapidly and genetic studies restricted to embryos can be advantages andfeasibility ofdirect clinicalutilization (Table 1).
completed in days to weeks rather than weeks to months as is Classic cancer modeling via mutagenesis, transgenesis, and
often the case with mammalian models.3,4 However, it is xenotransplantation has contributed in numerous ways to
important to note that zebrafish sexual maturation takes about precision oncology (Fig. 1, left). Zebrafish cancer models have
three months, so generation studies (i.e., transgenics, knockouts, facilitated(i)theidentificationandinvivovalidationofmolecular
1TheHormelInstitute, UniversityofMinnesota,Austin,MN55912,USA
Correspondence:LukeH.Hoeppner([email protected])
Received:13August2017Revised:6November2017Accepted:7November2017
PublishedinpartnershipwithTheHormelInstitute,UniversityofMinnesota
Fishingforcures
MAstoneetal.
2
6 8
ef. 5 5 6 8 3 4 5,7 0 1 7 7,4 9 6
R 2 2 6 2 7 7 7 6 9 7 7 4 5 4 5
fiSpeciccontributiontoprecisiononcology Invivovalidationoftargeteddrugsforthetreatmentofmelanoma. Discoveryoftwonewpotentialdrugsforthetreatmentofmelanoma. fiValidationofazebrashxenograftmodelasadrugscreeningplatformforthetreatmentofmelanoma. fiInvivoidenticationofatargeteddrugforthetreatmentofglioma. Invivovalidationoftargeteddrugsforthetreatmentofglioblastomaviaangiogenesisinhibition. Discoveryofanewsmallmoleculeradiationsensitizerforthetreatmentofglioblastoma. InvivovalidationofadrugtargetingGSCsforthetreatmentofglioblastoma. fiProofofprinciplefortheuseofazebrashorthotopicxenograftmodelasadrugscreeningplatformforthetreatmentofglioblastoma.fiValidationofazebrashorthotopicxenograftmodelasadrugscreeningplatformforthetreatmentofglioblastoma.fiProofofprinciplefortheuseofazebrashorthotopicxenograftmodelasadrugscreeningplatformforthetreatmentofpediatricbraintumors. Invivovalidationofatargeteddrugforthetreatmentofpancreaticcancer. DevelopmentofProhema,currentlyinPhaseIIclinicaltrialsforuseinleukemiaandlymphomapatientsreceivingbloodtransplantations.fiProofofprinciplefortheuseofazebrashxenotransplantationmodelasapreclinicalplatformforapersonalizedtherapy. Invivovalidationoftargeteddrugsforthetreatmentofthyroidcancer. Discoveryoftwoclassesofpotentialtargeteddrugsforthetreatmentofhepatocellularcarcinoma.
1234567890 Results SmallmoleculeinhibitorsofMEKandPI3K/mTORsuppressthemelanocytehyperplasiaphenotype. TwoFDA-approvedcompoundscooperatewithMEKinhibitorstosuppressthegrowthoftransformedmelanocytes.fiTargetedinhibitionofknownpathwaysbyspecicdrugsiseffectiveincounteractingcancercellsmigrationandproliferation. AKT1/2inhibitorsuppressesgliomagenesis,inhibitscellularproliferation,andinducesapoptosisinestablishedgliomas. JNK,ERK,andPI3Kinhibitorssuppressangiogenesisinducedbyglioblastomacells. Anovelsmallmoleculeradiationsensitizerenhancesthetumorgrowth-inhibitoryeffectsofionizingradiation. Asyntheticcompound,Nordy,suppressesangiogenesis,fitumorinvasion,andproliferationofthezebrashGSCxenograft. Adrugwithaknownanti-cancereffectincellcultureinhibitsproliferationandinvasioninthexenograftmodel. Currentlyusedglioblastomatherapeuticsdecreasefixenotransplanttumorburdenandsignicantlyrescuesurvival.flAcytotoxicchemotherapeuticagent(5-uorouracil)andatyrosinekinaseinhibitorsuppressERBB2-drivengliomas. Aknownsmallmoleculeinhibitor,U0126,targetingtheKRASsignalingpathway,repressesproliferationandmigrationofcancercells. Chemicalsthatenhanceprostaglandin(PG)E2synthesisincreaseHSCnumbers. AbonemarrowsamplederivedfromaT-ALLpatientNOTCH1harboringamutationrespondstoNOTCH1fiinhibitorinthezebrashxenograftmodel. CombinatorialtreatmentwithBRAFandMEKinhibitorsrescuenormalfolliculararchitecture,restorethyroidhormoneproduction,andreduceepithelialmesenchymaltransition. Twoc-JunN-terminalkinase(JNK)inhibitorsandtwoanti-βdepressantssuppress-catenin-inducedlivergrowth.
gy al al al al voal al al al al al al al
precisiononcolo Approach Pharmacologictestinvivo Invivodrugscreening Pharmacologictestinvivo Pharmacologictestinvivo Pharmacologictestinvivo Invitrodrugscreeninginvipharmacologictest Pharmacologictestinvivo Pharmacologictestinvivo Pharmacologictestinvivo Pharmacologictestinvivo Pharmacologictestinvivo Invivodrugscreening Pharmacologictestinvivo Pharmacologictestinvivo Invivodrugscreening
fiebrashcancermodelsto Transgene/injectedcells G12Vmitfa:HRAS G12Vmitfa:HRAS Humanuvealmelanomacellsgeneratedfromprimarytumorsandmetastasis ptf1a:Gal4;UAS:GFP-UAS:DAAkt1 Humanglioblastomacells Humanglioblastomacells Gliomastemcells(GSCs)isolatedfromahumanglioblastomacellline Humanglioblastomacells Patient-derivedgliomacells Mouseependymoma,glioma,andchoroidplexuscarcinomacells Humanpancreaticadenocarcinomacells — Patient-derivedT-ALLcells V600Etg:BRAF βfabp10a:pt--catenin
z
nof enic enic enic os enic enic
ontributio Model Transgline Transgline XT Transgline XT XT XT XT XT XT erXT WTembry XT Transgline Transgline
Table1.Thec Cancertype Melanoma Glioma Brainpediatrictumors Pancreaticcanc Leukemiaandlymphoma T-ALL Thyroidcancer Hepatocellularcarcinoma
npjPrecisionOncology(2017) 39 PublishedinpartnershipwithTheHormelInstitute,UniversityofMinnesota
Fishingforcures
MAstoneetal.
3
n
ef. 2 8 9 0 1 2 2 3 ee
R 6 8 8 8 9 9 9 8 7 9 6 b
e
fiSpeciccontributiontoprecisiononcology fiValidationofazebrashorthotopicxenograftmodelasadrugscreeningplatformforthetreatmentofretinoblastoma. Invivovalidationofanewmoleculartargetandanti-metastatictargeteddrugsforthetreatmentofpancreaticcancer. Invivovalidationofanti-metastatictargeteddrugsforthetreatmentofprostatecancer. Invivovalidationofananti-metastatictargeteddrugforthetreatmentofprostatecancer. Invivovalidationofananti-metastatictargeteddrugforthetreatmentofmelanoma. Invivovalidationofapotentialanti-metastaticprecisiononcologytreatmentforbreastcancerfipatientswithARF1amplication. Invivovalidationofanti-metastatictargeteddrugsforthetreatmentofbreastcancer. Invivovalidationofananti-metastatictargeteddrugforthetreatmentofbreastcancer. fiProofofprinciplefortheuseofzebrashxenograftfortheevaluationofcancerpatientprognosis. Invivovalidationofanewmoleculartargetandananti-metastatictargeteddrugforthetreatmentofEwingsarcoma.fiValidationofazebrashxenotransplantationmodelasaplatformfortheanalysisofmetastaticbehaviorofprimaryhumantumorspecimen. fiofeachmodeltotheprecisiononcologyeldhav
s
Results Orthotopicxenograftofretinoblastomacellspermitsquantitativeanalysisofcancercellsproliferationandtheanti-cancereffectofdrugssystemicallyadministered. miR-10Asuppressionbyknockdownorretinoidacidreceptorantagonistsblocksmetastasis. PharmacologicinhibitorsofSYKkinase,currentlyinphase–IIItrialsforotherindications,preventmetastaticdissemination. ThesmallmoleculeVPC-18005,targetingERG,exhibitsanti-metastaticactivityagainstprostatecancercellsaberrantlyexpressingERG. TheFDA-approvedanti-DNAvirusagentcidofovirinhibitsmetastasisofFGF2-driventumorcells.fiSpecicinhibitionofArf1bysmallmoleculeLM11impairsmetastaticcapabilityofbreastcancercells. NovelcompoundsdesignedtoantagonizeP2×7receptorinhibitinvasionofbreastcancercells. InhibitionofsignalingbetweenhumanCXCR4andfizebrashligandsbythesmallmoleculeIT1timpairsbreastcancerearlymetastases.flTransplantedprimarycellbehaviorreectstheclinical’courseofthepatientsmedicalhistory. TheSIRT1/2inhibitorTenovin-6prohibitstumorgrowthandspreadofcancercells. Xenograftsofprimaryhumantumorsshowrapidinvasivenessandmicrometastasisformationaftertransplantationintheyolkororganotopicallyintheliver. shcancermodelsinprecisiononcologyresearch.Contribution
fi
a
al al al al al al al al al br
Approach Pharmacologictestinvivo Pharmacologictestinvivo Pharmacologictestinvivo Pharmacologictestinvivo Pharmacologictestinvivo Pharmacologictestinvivo Pharmacologictestinvivo Pharmacologictestinvivo Xenograftandimaging Pharmacologictestinvivo Xenograftandimaging theutilityofzeblasticleukemia
Transgene/injectedcells Humanretinoblastomacells Pancreaticcarcinomacellsandfragmentsofresectedtumortissue Humanprostatecancercells Humanprostatecancercells Mousemelanomacells Triple-negativebreastcancercells Triple-negativebreastcancercells Triple-negativebreastcancercells Primarycultureofbreastcancerbonemetastasis HumanEwingsarcomacells Tumorexplantsfrompancreas,colon,andstomachcarcinoma inthisreviewthatexemplifyT-ALLon,T-cellacutelympho
edati
del describnsplant
Table1.continued CancertypeMo RetinoblastomaXT PancreaticductalXTadenocarcinoma(metastasis) ProstatecancerXT(metastasis) XT MelanomaXT(metastasis) BreastcancerXT(metastasis) XT XT XT EwingsarcomaXT(metastasis) GastrointestinalXTtumors(metastasis) AsummaryofstudiesXThighlightedxenotra
PublishedinpartnershipwithTheHormelInstitute,UniversityofMinnesota npjPrecisionOncology(2017) 39
Fishingforcures
MAstoneetal.
4
Fig. 1 Applications of the zebrafish model in precision oncology. Classic cancer research using zebrafish has contributed to precision
oncologythroughtheestablishmentofnumerouscancermodels,leadingnotonlytosignificantadvancementsincancerbiology,butalsoto
thedefinitionoftargeteddrugssuitableforpersonalizedcancertreatments(blue,left).Possibleapplicationsofzebrafishintheclinictodrive
personalizedtherapiesforspecificpatientshavealsobeenshown.Thefeasibilityofthisapproachhasbeendemonstratedthroughtheuseof
patient-derivedzebrafishxenograftsandgenerationoftransgeniczebrafishmodelingmutationsortranslocationsdefiningaspecificpatient’s
tumor(red,right)
players in tumorigenesis and metastasis, (ii) the definition of only14%ofpatientswithmetastaticdiseasesurviveforfiveyears.
actionable alterations and therapeutic targets, and (iii) the Unlike many other tumor types, new cases and mortality of
discoveryoftumorbiomarkersandgeneticsignaturesaspotential melanoma are still rising.11,12 While some oncogenic driver
diagnostic and prognostic indicators. Moreover, several studies mutations, such as BRAF and NRAS, have been identified in
have exemplified the potential of zebrafish models to contribute melanoma,theefficacyoftherapiesislimitedandtheprognosisof
moresignificantly anddirectly toprecisiononcologythrough(iv) metastatic melanoma patients remains poor.13 Many sponta-
identifying and testing drugs for targeted inhibition of specific neous, oncogene-driven zebrafish models of melanoma exist. In
pathways/alterations by utilizing zebrafish as an in vivo drug 2005,Pattonetal.14expressedBRAFV600Einmelanocytesusingthe
screening platform. A number of small molecules that might microphtalmia-associated transcription factor a (mitfa) promoter.
represent new targeted drugs for individualized medicine have These fish developed nevi, but required a p53M214K mutant
been identified through this approach. Notably, the rapidly zebrafish background for melanoma development in ~5% of
increasing number of patient-derived cancer cell xenografts5–10 zebrafishbyfourmonths.14Acrestin:EGFPreporter,recapitulating
placeszebrafishontheroadtowarditsclinicalapplicationforthe
the embryonic neural crest expression patter of crestin, showed
treatment of individual cancer patients. Various studies have
thatafatechangeoccursatmelanomainitiationinthismodel,as
demonstrated the applicability of these models in (v) evaluating asinglemelanocytereactivatestheneuralcrestprogenitorstate.15
patient prognosis in vivo and (vi) directing individualized Similarly, human oncogenic NRASQ61K expression under the
treatments in real-time based on responses to drugs of patient control of mitfa promoter resulted in a transgenic fish that
cancer cell xenografts (Fig. 1, right). Taken together, modeling
requiredp53lossoffunctionforthegenesisofmelanoma.16The
cancer in zebrafish has evolved to the extent that precision
first p53 mutation-independent model was developed through
oncologyapplications are emerging.
expression of human oncogenic HRASG12V driven by the same
mitfapromoterfragment.Inthismodel,however,melanomadoes
GENETICMODELS OFCANCER not arise at a high frequency and takes several months to
To date, innumerable zebrafish genetic models of cancer have develop.17Instead, when HRASG12V expression is driven in
melanocyte progenitor cells by the kita (c-kit in humans)
been generated, and the number continues to rapidly increase.
promoter, melanoma occurs spontaneously by 1–3 months in
Genetic cancer models have been developed using various
~20% of fish.18 Recently, a novel zebrafish transgenic model of
strategies, including transient, stable, and double transgenesis
uvealmelanomawascreatedbyexpressingoncogenicGNAQQ209P
and various inducers of mutagenesis (Table 2). Their use in
in the melanocyte lineage using again the mitfa promoter. The
precision oncology is gaining momentum. Here, we will discuss
corresponding p53 inactivation was also required for the
themostsignificantreports exemplifying this evolution.
malignant progression in this system.19 Importantly, zebrafish
models of melanoma closely resemble human cancer, both in
Melanoma termsofhistopathological features and molecular signatures.20
Melanoma research offers many concrete examples of genetic Thesemodelshaveconfirmedtheroleofrelevantoncogenesin
zebrafish models used for the definition of new therapeutic melanomagenesis and progression. Moreover, they have proven
targets and as an in vivo platform for drug screening. Melanoma to be outstanding tools to test and screen for other genes that
accounts for the death of over 70% of skin cancer patients and promote melanoma onset and might represent new therapeutic
npjPrecisionOncology(2017) 39 PublishedinpartnershipwithTheHormelInstitute,UniversityofMinnesota
Fishingforcures
MAstoneetal.
5
Table2. Geneticmodelsofcancer
Cancertype Geneticsystem Transgenes/mutatedgenes Ref.
Melanoma Transgenicline mitfa:BRAFV600E;p53mutantbackground 14
Melanoma Transgenicline mitfa:NRASQ61K;p53mutantbackground 16
Melanoma Transgenicline mitfa:HRASG12V 17
Melanoma Transgenicline kita:HRASG12V 18
Melanoma Transgenicline mitfa:GNAQQ209P;p53mutantbackground 19
Braintumors Transgenicline krt5:Gal4VP16;UAS:mCherry-KRASG12V 27
Braintumors Transgenicline gfap:Gal4VP16;UAS:mCherry-KRASG12V 27
Glioma Transgenicline ptf1a:Gal4;UAS:GFP-UAS:DAAkt1 28
MPNST Mutantlines Heterozygousmutationsin11ribosomalproteingenes 31
MPNST Deletion 15.2Mbdeletioninchromosome1 32
MPNST Mutantline tp53M214K 33
Neurofibromas/MPNST Mutantlines mlh1−/−,msh6−/−,msh2−/− 34
MPNST Mutantlines Heterozygousmutationsin17ribosomalproteingenes 35
Pancreaticcancer Transgenicline ptf1a:Gal4-VP16;UAS:mutatedKRAS 36,38–40
T-ALL Transgenicline rag2:mMyc 44
T-ALL Transgenicline hsp70:Cre;rag2:lox-dsRED2-lox-EGFP-mMyc 45
T-ALL Transgenicline rag2:ICN1-EGFP 46
Thyroidcancer Transgenicline tg:BRAFV600E 49
Hepatocellularcarcinoma Transgenicline Mifepristone-inducedCre-mediatedrecombination:fabp10:loxP-mCherry-loxP-EGFP-krasV12 55
Hepatocellularcarcinoma Transgenicline fabp10a:pt-β-catenin 56
Colonadenoma Mutantline apcmcrmutantinjectedwithmRNAencodingoncogenicV5-KRASG12D 98
Significantzebrafishgeneticcancermodels,includingallthosediscussedinthereview,havebeensummarizedMPNSTmalignantperipheralnervesheath
tumors,T-ALLT-cellacutelymphoblasticleukemia
targets, and even, in the near future, tumor biomarkers for p53−/−zebrafishembryosdemonstrateagenesignatureenriched
personalized cancer therapy. An excellent example has been for markers of multipotent neural crest cells. A chemical genetic
reported by Ceol and colleagues. They have used transgenic screen was, therefore, performed to identify small molecule
zebrafishoverexpressingBRAFV600Eonap53mutantbackground suppressors of the neural crest lineage. A positive result was
totestgenesinarecurrentlyamplifiedregiononchromosome1. obtained with the inhibitors of dihydroorotate dehydrogenase,
The histone methyltransferase SETDB1 has been found to whose activity as an anti-melanoma agent was then confirmed
cooperatewithBRAFV600Eandacceleratemelanoma.Itsrelevance in vitro and through mouse xenograft models.26 To fully realize
inhumanmalignantmelanomahasalsobeendemonstrated,and the utility of zebrafish in precision oncology, translating these
therefore, SETDB1 has been revealed as a novel oncogene in types of drug identification and validation studies to a patient
melanoma.21 RAC and RSK1, whose hyperactivation has been
samplesizeofoneistheultimategoal,suchthattreatmentscan
detected in human melanoma, have been shown in distinct betailoredtotheindividualpatientbasedonzebrafishsurrogates
studies to contribute to melanoma progression when constitu-
tively activated in mitfa:HRASG12V and mitfa:BRAFV600E;p53−/− oftheindividual’stumor.
transgenicbackgrounds,respectively.22,23Listeretal.24haveused
a temperature-sensitive mitfa mutant to show the oncogenic Neurological tumors
activity of Mitfa transcription factor in BRAFV600E transgenic Neurological tumors have also been modeled via transgenic
zebrafish and the regression of BRAFV600E mitfa melanoma after expression of oncogenes, demonstrating the potential to define
Mitfa activity abrogation, thus presenting Mitfa as a promising relevant actionable alterations driving cancer progression and to
therapeutic target. The use of zebrafish to identify novel successfullytestspecificdrugstargetingthosealterations.27,28The
oncogenes begins to exemplify how this model organism will focus of most zebrafish studies on brain tumors is malignant
be utilized to overcome tumor heterogeneity through precision
glioma, which accounts for 70% of malignant primary brain
oncology.
tumors, and in particular glioblastoma, the most aggressive
Thesignificanceofzebrafishmelanomamodelsintranslational
primary brain cancer, accounting for 70% of malignant glio-
medicine and precision oncology is not limited to the discovery
mas.29,30Transgenicmodelsofmalignantperipheralnervesheath
and characterization of potential therapeutic targets, as various
studieshavealreadyshowntheefficacyofzebrafishinidentifying, tumors have also been described.27,31–35 Jung and colleagues
established a transgenic zebrafish that overexpressed dominant
discovering, and testing drugs for the development of new
melanomatreatments.SmallmoleculeinhibitorsofMEKandPI3K/ active,humanAKT1attheptf1adomainleadingtogliomagenesis.
mTOR,knownplayersinmelanoma,havebeenvalidatedinvivoas Pharmacological tests identified AKT1/2 inhibitor as a targeted
targeteddrugssuppressingmelanocytehyperplasiaphenotypein drug capable of effectively suppressing gliomagenesis, inhibiting
HRASG12V transgenic embryos.25 Moreover, a zebrafish screen of cellular proliferation, and inducing apoptosis in established
FDA-approved compounds led to the discovery of two new gliomas.28 The scope of available brain tumor models offers
potential drugs cooperating with MEK inhibitors to suppress the promiseforusingzebrafishtotailorspecifictreatmentapproaches
growthoftransformedmelanocytes.25Transgenicmitfa:BRAFV600E; toindividual neurological cancerpatients.
PublishedinpartnershipwithTheHormelInstitute,UniversityofMinnesota npjPrecisionOncology(2017) 39
Fishingforcures
MAstoneetal.
6
Pancreaticcancer Liver cancer
TheGal4/UAStransgenicsystem,basedontheabilityoftheGal4 Liver cancer is thesecond leading cause of cancer-related death.
transcriptional activator to drive the expression of multiple Hepatocellular carcinoma (HCC) accounts for 90–95% of liver
transgenes under the regulation of UAS (upstream activator cancercases.50Severalzebrafishmodelsoflivercancerhavebeen
sequence) regulatory elements, is widely used to model KRAS- developedutilizingdifferentexpressionsystems(reviewedwellby
initiated pancreatic cancer in zebrafish.36 Pancreatic cancer is a Lu et al.50), including extensive contributions by Dr. Gong’s
deadlygeneticdisease,withadismal~9%fiveyearsurvivalrate.37 group.51–55 As a recent example, they developed a transgenic
The majority of pancreatic cancers are pancreatic ductal system for a liver-specific, mifepristone-inducible expression of
adenocarcinomas (PDACs) and over 90% of them carry an oncogenic krasV12 via permanent genomic recombination
activating point mutation in the KRAS gene.36 Genetic models mediatedbytheCre-loxPsystem,55whichwillfacilitatethestudy
basedontheGal4/UASsystemenableassessmentoftheeffectsof oflivertumorsthatoriginatefromasinglecellorasmallnumber
differentKRASmutationsandtheabilityofotherproteinstoalter ofprecursorcellsthroughclonalexpansion.Inall,20–40%ofHCC
the response to oncogenic KRAS, potentially leading to the are defined by an activating mutation in the gene encoding β-
identification of new targets for precision oncology therapeutic catenin. Evason and colleagues created a transgenic zebrafish
strategies.38,39Inthisregard,theinvolvementofavarietyofcore expressinghepatocyte-specificactivatedβ-catenin.Theyusedthe
signaling pathways, including TGFβ, Wnt, Notch, and Hedgehog, modeltoscreenfordruggablepathwaysthatmediateβ-catenin-
in pancreatic cancer development has also been investigated induced liver growth and identified two c-Jun N-terminal kinase
usingGal4/UAS system.36,38,40 (JNK) inhibitors and two anti-depressants as potential targeted
therapeutics.56 As is true in other tumor types discussed in this
Leukemia section,zebrafishhavecontributedtodrugdevelopmenttotreat
livercancer.
Leukemia, the ninth most common cancer type is a cancer of
blood-forming tissues usually involving dysfunction of white
blood cells.41 Leukemia has been modeled mainly through
TRANSPLANTATION CANCERMODELS
transgenesis. A zebrafish model of T-cell acute lymphoblastic
Withthefirstexperimentreportedin2005,57xenotransplantation
leukemia (T-ALL), the most common type of childhood leuke-
of human cells into zebrafish represent a young frontier in
mia,42,43wascreatedintheearly2000sexpressingamousec-Myc
zebrafishcancermodeling.However,thefieldhasevolvedrapidly,
transgene fused to green fluorescent protein (GFP) under the
and xenograft zebrafish models utilizing various injection sites,
control of a zebrafish rag2 promoter.44 Visualization of GFP+
developmental stages, and transplanted specimens (i.e., human
leukemic cells has demonstrated leukemia originates in the
cell lines, patient-derived primary cancer cells, patient-derived
thymus, disseminates to the gill arches and surrounding retro-
tumor tissue explants) have been developed58,59 (Table 3).
orbital soft tissue, and then spreads to skeletal muscle and
Engraftment of a diverse range of human, murine, and zebrafish
abdominal organs.44 Feng and colleagues have subsequently
tumor cells has been demonstrated. Zebrafish transplantation
improved this model by developing conditional, heat-inducible
modelsofferthepossibilitytostudymanyhallmarksofcancerand
activationof the c-Myconcogene resulting ingreater penetrance
steps of cancer progression, such as self-renewal, tumor-induced
of T-ALL and increased control of disease onset.45 Similarly,
angiogenesis, invasion and dissemination, interaction between
anotherzebrafishmodelofT-ALLhasbeencreatedbyexpressing
tumor and host, and drug responses.58,59 Cancer specimen
thetruncatedhumanNOTCH1proteinfusedtoEGFP(ICN1-EGFP)
transplantation into embryos is certainly the most commonly
under the control of the zebrafish rag2 promoter.46 While these used zebrafish developmental stage for undeniable advantages,
transgeniczebrafishdevelopedT-ALLby5monthsofage,onsetof
includingtheeaseofproducingandinjectingmanyembryosina
leukemiawasdramaticallyacceleratedwhencrossedtozebrafish
short amount of time. Furthermore, the immature state of the
overexpressing anti-apoptotic protein, Bcl2. The oncogenic
immune system of embryos avoids the requirement of immune
synergy between NOTCH1 and Bcl2 in this model suggests
suppressingagentsorirradiation.58,59Fluorescentlylabeledtumor
geneticmodifierscreensmayrevealothergenesthatinteractwith
cells have been transplanted at developmental stages varying
NOTCH1 to promote T-ALL.46 All of these rag2 promoter-driven
fromtheblastulastage to72hpost fertilization (hpf)ininjection
transgenic zebrafish models are amenable to drug and genetic
sitessuchasblastodisc,yolksac,bloodstream,perivitellinespace,
screening to identify individualized treatment strategies for
and orthotopic sites, including the hindbrain ventricle and
leukemia and lymphoma patients. Indeed, a zebrafish screen for vitreous cavity.9,58–62 The transplanted cells can be studied for
therapeutics that alter the number of hematopoietic stem cells upto21dayspostfertilization(dpf),atwhichpointthezebrafish
(HSCs) has led to the development of Prohema, a derivative of
has developed a fully functional innate and adaptive system.59
prostaglandinE2(PGE2),currentlyinPhaseIIclinicaltrialsforuse
While embryonic xenotransplantation offers numerous advan-
in leukemia and lymphoma patients receiving blood
tages,alimitationisthatmanyofthetumortypesbeingmodeled
transplantations.47,48
occurpredominantly in adults.
Xenotransplantation in juvenile and adult zebrafish seeks to
Thyroidcancer overcome the limitation of translating embryonic zebrafish
Stable transgenic expression of oncogenic BRAF (BRAFV600E) in models to mature human cancer patients. Transplantation of
thyroidepithelialcellshasrecentlybeenshowntoinducethyroid human cancer cells in 30dpf zebrafish has been established by
cancerinadultzebrafish.CombinatorialtreatmentwithBRAFand Stoletov and colleagues in 2007 by injecting cells into the
MEK inhibitors rescue normal follicular architecture, restore peritoneal cavity and treating the fish with dexamethasone to
thyroidhormoneproduction,andreduceepithelialmesenchymal prevent rejection.63 The study of cancer cell transplantation in
transition stimulated by BRAFV600E. The model has demonstrated adult fish requires immune suppression by irradiation or
invivothegeneticrequirementforTwistexpressiondownstream dexamethasone pre-conditioning, and the use of transparent
of BRAFV600E, as ablation of twist3 by CRISPR-Cas9 suppressed transgenic zebrafish allows the rapid identification of the
BRAF-mediated oncogenesis.49 The in vivo validation of targeted transplanted cells.58,59 Casper fish, a cross between the nacre
molecular therapies for the treatment of thyroid cancer demon- androymutantlines,lackingalltypesofpigments,arecommonly
strates the applicability of the zebrafish system to precision usedforthispurpose.64Morerecently,Tangetal.65developedan
oncologyapproaches. optically clear immunocompromised transgenic mutant zebrafish
npjPrecisionOncology(2017) 39 PublishedinpartnershipwithTheHormelInstitute,UniversityofMinnesota
Fishingforcures
MAstoneetal.
7
6
ef. 3 6 7 8 9 1 2 3 4 5,7 0 7 1 2 3 4 5 6 7 8 9 0 1 2 2 3
R 6 6 6 6 6 7 7 7 7 7 6 7 6 6 9 5 8 8 8 8 8 8 8 9 9 9 8 9
d
n
a y
Injectionsite Peritonealcavity Yolksac Earlyembryo Pericardium Peritonealcavity Intothecerebrumviatheintranasalroute Yolksac Yolksac Blastodiscoryolksac Yolksac Hindbrainventricle Perivitellincavity(embryos)cardiacchamber(adults) Vitreouscavity Vitreouscavity Midbrain-hindbrainboundar Yolksac Pericardium Pericardium Pericardium Intraperitoneally Yolksac Yolksac Yolksac DuctofCuvier Perivitellincavity Yolksac DuctofCuvier Yolksac
s
Developmentalstage –2535dpf 48hpf 3hpf(blastula) 48hpf Adult 30dpf –1248hpf 48hpf–3.54.5hpf(blastula) 48hpf 48or72hpf 48hpfembryosand6mpfadult 48hpf 48hpf 36hpf 48hpf 48hpf 72hpf 72hpf Adult 48hpf 48hpf 48hpf 48hpf 48hpf 48hpf 48hpf 48hpf
tumorsandmetastasis nomalesions mousechoroidplexus acellline ellsandneurospheres) onadenocarcinoma ancreatic
Transplantedcells Severalhumancancercells Humanuvealmelanomacellsgeneratedfromprimary Humancutaneousmelanomacells HumanmelanomacellsderivedfrommetastaticmelafiZebrashmelanomacells Mouseglioblastomacells,mouseependymomacells,carcinomacells Mouseschwannomacells Humanglioblastomacells Humanglioblastomacells Gliomastemcellsisolatedfromahumanglioblastom Humanglioblastomacells Humanpancreaticadenocarcinomacells Humanandmouseretinoblastomacells Humanretinoblastomacells Patient-derivedgliomacells(serum-grownadherentc Leukemiacellsandpatient-derivedleukemiacells fiHumanbreastadenocarcinoma,brosarcomaandcolcells Humanrenalcelladenocarcinomacellsandhumanpadenocarcinomametastasiscells Humanlungadenocarcinomacells fiZebrashlymphomacells Humanprimarypancreaticadenocarcinomacells Humanprostatecancercells Humanprostatecancercells Mousemelanomacells Triple-negativebreastcancercells Triple-negativebreastcancercells Triple-negativebreastcancercells HumanEwingsarcomacells
models Cancertype Varioustypes Uvealmelanoma Cutaneousmelanoma Melanoma Melanoma Brainpediatrictumors Vestibularschwannoma Glioblastoma Glioblastoma Glioblastoma Glioblastoma Pancreaticcancer Retinoblastoma Retinoblastoma Glioblastoma T-ALL Varioustypes Pancreaticcancer Lungadenocarcinoma T-ALL Pancreaticductaladenocarcinoma Prostatecancer Prostatecancer Melanoma Triple-negativebreastcancer Triple-negativebreastcancer Triple-negativebreastcancer Ewingsarcoma
er
c s s
ationcan Cellline Patient-derivedcells Cellline
nt
Table3.Transpla Transplantationcancermodels Transplantationmetastasismodels
PublishedinpartnershipwithTheHormelInstitute,UniversityofMinnesota npjPrecisionOncology(2017) 39
Fishingforcures
MAstoneetal.
8
line for optimized cell transplantation and direct visualization of
ef. 0 fluorescently labeled cancer cells in the adult fish. Although
R 1 6 8 7
juvenile and adult zebrafish transplantation models more closely
matchthedevelopmentalstateandageofhumansafflictedwith
cancer,therequirementforimmune-deficientzebrafishrepresents
thedownside.
Melanoma
site um Cuvier tBdroeartnhsEpnlteamanntbadtriycooonnlliemcaogdauneedlssitnoajedscuttluetddyzdedifbrfeurargefinsehtffihcuhamcayvaeninubmveeeealnalnmouemslaeand.ovmaanas
Injection Pericardi Yolksac Yolksac Ductof icyneohlllkibliointfieo4sn8gohefpnkefnrzaoetwberdnafipfrsaohtmhewmparbyimsrybaorysy.stTpuhemecyiofirchsadavrneudgsshmopewrtoanvsettashsaeetfsfteiancrttgioveettehidne
counteracting cancer cells migration and proliferation, thus
al demonstrating the applicability of the zebrafish xenograft model
nt for drug screening and discovery.66 Other xenograft zebrafish
e
m models have been used to explore relevant pathways in
p
Develostage 72hpf 48hpf 48hpf 48hpf kemia mcRieteycl,ea6nn7tolyma,naad,dsTuruGcghF-tβaresainNtmodceenalltluisnlyasrcteermleluslifasotrarnptchlaeestlitocointygM-taeEnrKdmtianudhmmiboiirntiogisretsrn.a6i8--
east mor cleu tpiolannteodf wanitthi-maezleabnroamfisahmdreulgasnoimnaacdeullltlinceashpaesrbzeeebnradfiesvheltorpaends-.
Transplantedcells CulturedcirculatingtumorcellsisolatedfromthebloodofametastaticbrcancerpatientMousemammaryepithelialcellstransformedwithoncogenicRasandtuexplantsfrompancreas,colonandstomachcarcinoma Pancreaticcarcinomacellsandfragmentsofresectedtumortissue Breastcancercellsandprimarycultureofbreastcancerbonemetastasis T-ALLelsdiscussedinthereviewhavebeenoutlinedT-cellacutelymphoblasti aiaccmnmstfaGsNAbdeTpurhnohaetoannhtsoueirSeehfduooeovpevmdgeefmmCumsaiineudemadabniayrlioprrarsdaaeb.ontisieax7iietgndcoi.ntlestll07oe.olnoeoteeiehuut)e1vnonnmrt.ylgm,gninlasio7sGlgbteovsti2lH.nespodfco7iSavoigheeost(t3rtdaAosifbCaGernuiinonrmf.l-sewalaelsoAzdmadumsvSgi,tflTcasearinsenitCuualthttrrta.boiivigtob7iuaomcosmnzoevcnetr5e)hinitmynebedeaidnfo,hoer7tishrmlfil,tugyg,g6aonperlpmGhtasisschesdasrretel.ahseSoTea7tseanp-rendnozC4t.agifndlva6irnnhnteiiigcmksie9gantelTamvtbsreeeimttrnioodhoaaafiurerdanavasaiaugsnieadelrcgfixyrbinfiovgeyfgodiceetnesletbsoblnhoyiandnnottrobhetiruengphe,ohgtoragneegssducenueglrgiaiefti.olcneotgittstfiirhelnrenohaiiiger,uasccaMoxrdseltmewl,iofagtpemsucctsriotccuchhhaovlesa,mipuoomrsyargulalnaiemre,zlcarmonohcsveoboeexhocnihroeltwdnepdevbenifshfwrinelhtoerienhlaxntaivoaaasrxonlaaurreiv,isgnfivacaetvshgnenntrhistdanni(ettarodeioooJhrhroasossNbeno.z,otrvvpftengwrveetNfhebKegNaldoairbaelasa,eeolnmaomofnrinntrnasewousrfsfiaErdttodbngmcnopbdgefiRestedyutlidmllydsihidKhnaaipasddoel-,hesanel,coshaorelmllutcpobrrntwidxinoaehbaempnzpryagreSentemnetaeliepoiltnciiloTdnfvssbooesvcrshyaVoGlitieagedrneeonbowhfigasFnPnatedcruagefisatro-alIcunenoaseebβe3bvsanomeilwlifdddrhnnK1eeeyycstr-fi
d
o gliomaxenograftmodelshaveproventobevaluablesystemsfor
Cancertype Breastcancer Varioustypes Pancreaticductaladenocarcinoma Breastcancer andmetastasism fmtoaaenrcsotMdhitrliieonatjatnugroteyvepvleesaibncvrriacailaoelanliuynntsbcirmetremuarspimuoncllla.otee7ns0rlc.ltuGixDnslealugiirronrelvcybociaglvtatnraasactrltrfegoataremnntmcseadepdolclladspetnerhlilnotlsesalriehtfaihemaoparvnaibveetersiiybon.oebntnoee,inecwznieohnbriojcerebhacvtfitaeeslinedhnaelidanbdrsrvtahbatinoeey
cer hindbrain ventricle at 48–72hpf to develop a xenograft assay to
n
Patient-derivedcells antationca dpainbirhsoicligibotryivetesorsfpioraoandln.ifrdeTurhgaeptwirouiiontthriliittaaiyznekdonfociwtnohnvmeaapsanioostunsi-nacydaisnnwceaitmrsheepdfafeexcmcteitnoniongngstcrrageafllttlieocdbumllabtosuydtroeetmlh.t6oea0
ued nspl Anotherstudydescribedaplatformtostudytheefficacyofdrugs
ontin shtra fgolirotmhea,traenadtmcehnotrooifdpepdleiaxturisccbararcininotummaorcse.lMlsowuseereepteranndsypmlaonmtead,
ble3.c fiezebra poERrrtiBnhBco2itpo-ldepriictvhaelanlyt tighnletioosmethamesobdwreaelirsnecoasfnuzcbecebersuasffisueslhdlyjtuoivneahsnibsielietsess.ddArsbuyagpetrrfeofiacotaficnoygf,
a h
T T zebrafishwithacytotoxicchemotherapeuticagent(5-fluorouracil)
npjPrecisionOncology(2017) 39 PublishedinpartnershipwithTheHormelInstitute,UniversityofMinnesota
Fishingforcures
MAstoneetal.
9
or a tyrosine kinase inhibitor.71 When feasible, orthotopic Patient-derived transplantationmodels
xenograft models in zebrafish brain tissue offer the advantage Xenotransplantation of human cancer cells directly derived from
ofmorefaithfullyreplicatinghumandiseasebyutilizingthesame individual patients (patient-derived xenograft, PDX) represents a
anatomical tumor microenvironment. fascinating and forthcoming opportunity for the development of
zebrafish-based patient-specific clinical approaches for cancer
Pancreaticcancer treatment. Such patient-derived xenografts in zebrafish offer a
Xenotransplantation of pancreatic cancer cells in zebrafish has platformforreal-timeinvivoevaluationofpatientprognosisand
drug responses, aimed at identifying the most appropriate
also been proposed for the screening of new anti-cancer
individualized therapy (Fig. 1, right). Although only several
compounds. Guo and colleagues established a pancreatic
adenocarcinomaxenograftmodelinzebrafishembryosandadults examplesof direct transplantation of patient-derived cancercells
inzebrafishhavebeenreportedthusfar,6,8therapidlyincreasing
andfoundthataknownsmallmoleculeinhibitor,U0126,targeting
number of zebrafish xenograft cancer models suggests that
theKRASsignalingpathway,repressesproliferationandmigration
ofthetransplantedcancercellsinzebrafishlarvae.77Theseresults zebrafish xenografts are on the road to a clinical application in
precisiononcology.
suggest this model could be used to identify new therapies for
Welker and colleagues standardized a patient-derived ortho-
pancreaticcancer.
topic zebrafish xenograft model of glioblastoma. They trans-
plantedtwopatient-derivedglioblastomacelllines,serum-grown
Retinoblastoma
adherent and neurospheres, into the midbrain region of
Two studies have shown an orthotopic transplantation zebrafish embryonic zebrafish. In vivo tumor growth and cancer cell
model may represent a powerful tool for the development of proliferation, migration, and differentiation were described, with
specific drugs for the treatment of retinoblastoma, the most different characteristics in adherent and neurosphere glioblas-
common intraocular childhood cancer, which often invades the toma cell lines. Furthermore, currently used glioblastoma ther-
brain and metastasizes.61,62 Injection of retinoblastoma cells into apeuticsdecreasedxenotransplanttumorburdenandsignificantly
the vitreous cavity of the zebrafish embryo has permitted rescued survival. These results provide proof of principle for the
quantitative analysis of the tumor cells’ proliferative potential useofthemodel asaplatform for drug screening.9
andtheanti-cancereffectofsystemicallyadministereddrugs.This A preclinical human cancer xenotransplantation platform has
model offers a potential screening platform for retinoblastoma been recently developed in zebrafish to inform therapeutic
anti-cancerdrugs.62 decisions in T-ALL patients (Fig. 2a).5 The authors previously
Fig. 2 Precision oncology approach to leukemia drug screening using zebrafish. a Flow chart demonstrating the timeline used. Patient-
derivedleukemia cells were xenotransplanted into zebrafish embryos,which were administered variousdrugs. Leukemia cell number was
usedtoassessdrugefficacyinthezebrafishavatarcorrespondingtoanindividualleukemiapatient.Assessmentofdrugefficacyiscompleted
within8days,leadingtoafast,effective,andindividualizedcancertreatment.hpf:hourspost-fertilization,hpi:hourspost-injectionofcells,
hpt: hours post-treatment. b Bright-field and fluorescence images of zebrafish injected with patient-derived leukemia cells. Embryos were
treatedwithvehicle(control),Rapamycin(Rap)orCompoundE(CE).Imagesweretakenat72hpt.Scalebarsare500µM.cAbaselinenumber
ofleukemiacellswasdeterminedat96hpi.Anincreaseinthenumberofleukemiacellswhencomparedtothebaselinedatademonstrates
cellproliferationinthezebrafishmodel.Inpatientsampleone,datademonstratesasignificant(p<0.0001)responsetotheNotchinhibition
(CE).ThepatientsamplewassubsequentlysequencedandagainoffunctionmutationintheNotchpathwaywasfound.Patientsampletwo
didnotdemonstratesignificantresults,suggestingthemutationwasnotintheNotchpathway,whichwassubsequentlyconfirmedthrough
sequencing.Reproducedwithpermissionandadaptedfrom:Bentley,V.L.etal.Haematologica100,70–76(2015)5
PublishedinpartnershipwithTheHormelInstitute,UniversityofMinnesota npjPrecisionOncology(2017) 39
Fishingforcures
MAstoneetal.
10
testedtheinvitrosensitivityofthreeT-ALLcelllines,withspecific of zebrafish embryos, coupled with the emerging opportunities
mutationsinPTENandNOTCH1genes,tothreedifferentinhibitors offered by the xenograft models, represents an exceptional
(targeting mTOR, AKT, and NOTCH1) and demonstrated that the frontier to model and visualize the entire process of metastasis
same cell lines were sensitive to the same drugs upon atsingle-cell resolution.
xenotransplantation in the zebrafish embryos. The relevance of The xenotransplantion procedure is well optimized and
thezebrafishxenotransplantation modelasapreclinicalplatform automated quantitative assays are available to study invasion
for a personalized therapy was demonstrated by xenotransplant- andmetastasis ofcancer cells.79,80 Zebrafishxenograft modelsof
ing two primary patient-derived bone marrow samples into human cancer cell invasion, metastasis, and responsiveness to
zebrafish embryos and treating with the three inhibitors. One pharmacological or genetic intervention have been correlated to
patientsamplerespondeddrasticallytoNOTCH1inhibitor(Fig.2b, tumorigenicity of analogous human tumor cells in mouse
c), suggesting a mutation in the NOTCH pathway, which was xenograftmodels.79,81Furthermore,Tulottaetal.82demonstrated
subsequentlyconfirmedtobeaNOTCH1mutationprevalentinT- cross communication between zebrafish and human ligands and
ALL. The ability to assess a patient’s responsiveness to such a receptors, which enables the study of the interactions between
targetedtreatmentinazebrafishavatar(i.e.,likeness,surrogateor human cancer cells and host microenvironment during the
embodiment of an individual) within 1 week following biopsy, metastaticprocesses.
highlights how the zebrafish xenotransplantation response can Outstanding examples of zebrafish metastasis models have
directpersonalized therapy inreal-time.5 recently emerged (Table 3). Stoletov et al.83 used real-time
intravital imaging to study the dynamic process of intravascular
locomotion and extravasation of fluorescent human cancer cells
TRANSPLANTATION METASTASIS MODELS injected into the pericardium of 48hpf zebrafish embryos, thus
Most cancer deaths are caused by metastasis, as opposed to providing new insights into the underlying molecular regulation,
primarytumors.Metastasesresultfromthespreadofcancerfrom whichinvolvesβ1integrin,TwistandVEGFA.Auetal.10elegantly
the primary site to distant organs where new tumors form. showed the migration dynamics of clusters of circulating tumor
Metastatic cancers have acquired the capacity to escape the cells isolated from the blood of breast cancer and melanoma
primary malignant lesion site through intravasation into the patients. The zebrafish xenograft metastasis model represents a
bloodstream, migration, extravasation, and colonization of a valuable model to test the metastatic potential of human
distant site.78 Metastases are associated with poor prognosis precision oncology target genes. We recently adapted Stoletov’s
due to the difficulty of treating such a complex and diffuse model to demonstrate neuropilin-2 promotes extravasation and
process. Moreover, unlike most other cancer processes, such as metastasis of human pancreatic cancer and renal cell carcinoma
tumor initiation, proliferation, apoptosis, invasion etc., metastasis cells in zebrafish (Fig. 3) by interacting with endothelial α5
cannot be well modeled in vitro and the development and integrin. We demonstrate synergy of the zebrafish extravasation
utilization of in vivo models of the dynamic sequence of steps modelwithmammalianmetastasismodelsbyalsoexhibitingthe
from the local invasion to the distant colonization remains metastatic potential of neuropilin-2 in mice (Fig. 3).84 Thus,
challenging. The transparency and ease of genetic manipulation zebrafishmetastasismodelsserveasanexcellentinvivoplatform
Fig.3 Humancancercellxenograftmodelsofextravasationinzebrafishandmetastasisinmice.a–hHuman786-Orenalcellcarcinomacells
overexpressingretroviralcontrolvector(a–d)orneuropilin-2(NRP-2;e–h)weretransientlylabeledwithcelltrackerorangedye,microinjected
into the pericardium of 3dpf Tg(Fli-GFP) zebrafish, and imaged 1day later. a–d control 786-O cells stay in the ISVs. e, f 786-O cells
overexpressingNRP-2extravasatefromtheISVs.i–j2×106luciferase-labeled786-OcellssuspendedinPBSweresubcutaneouslyinjectedinto
therightflankoffemalenudemice.Priortothetumorgrowingto10%ofbodyweight,thesubcutaneoustumorsweresurgicallyresected.
Luciferaseimagingwasperformedonthemicefor4–6monthstomonitormetastasis,andthe786-ONRP-2knockdowngroup(top)exhibited
significantlyfewerlungmetastasesthanthecontrolcohort(bottom).k–rHumanASPC-1pancreaticcancercellsweretransducedwithcontrol
shRNA(k–n)orNRP-2shRNA(o–r),transientlylabeled,microinjected,andimagedasdescribedabove.k,lExtravasatedcontrolshRNAASPC-1
cells. m, l Actively extravasating control shRNA ASPC-1 cells. o–r NRP-2 knockdown ASPC-1 cells stay in the ISV. s–t Male SCID mice were
orthotopically injected with 2×106GFP-labeled ASPC-1 pancreatic cancer cellssuspended inPBS,and after15dayslivermetastases were
assessedbyxenogenimaging.Reproducedwithpermissionandadaptedfrom:CaoY.etal.CancerRes73,4579–4590(2013)84
npjPrecisionOncology(2017) 39 PublishedinpartnershipwithTheHormelInstitute,UniversityofMinnesota