Metabolic traits ruling the specificity of the immune response in different cancer types

(1)

Metabolic

traits

ruling

the

specificity

of

the

immune

response

in

different

cancer

types

Nina

C

Flerin

1,2

,

Federica

Cappellesso

1,2

,

Samantha

Pretto

1,2

and

Massimiliano

Mazzone

1,2

Cancerimmunotherapyaimstoaugmenttheresponseofthe patient’sownimmunesystemagainstcancercells.Despite effectiveforsomepatientsandsomecancertypes,the therapeuticefficacyofthistreatmentislimitedbythe compositionofthetumormicroenvironment(TME),whichisnot well-suitedforthefitnessofanti-tumoralimmunecells. However,theTMEdiffersbetweencancertypesandtissues, thuscomplicatingthepossibilityofthedevelopmentof therapiesthatwouldbeeffectiveinalargerangeofpatients.A possiblescenarioisthateachtypeofcancercell,grantedbyits ownmutationsandreminiscentofthefunctionsofthetissueof origin,hasaspecificmetabolismthatwillimpingeonthe metaboliccompositionoftheTME,whichinturnspecifically affectsTcellfitness.Therefore,targetingcancerorTcell metabolismcouldincreasetheefficacyandspecificityof existingimmunotherapies,improvingdiseaseoutcomeand minimizingadversereactions.

Addresses

1_Laboratory_of_Tumor_Inflammation_and_{Angiogenesis,}_Center_for Can-cerBiology,VIB,Leuven,B3000,Belgium

2

LaboratoryofTumorInflammationandAngiogenesis,Centerfor CancerBiology,DepartmentofOncology,KULeuven,Leuven,B3000, Belgium

Correspondingauthor:

Mazzone,Massimiliano(massimiliano.mazzone@kuleuven.vib.be)

CurrentOpinioninBiotechnology2021,68:124–143

ThisreviewcomesfromathemedissueonSystemsbiology

EditedbyKarstenHillerandDirkBrenner

https://doi.org/10.1016/j.copbio.2020.10.011

0958-1669/_ã2020TheAuthor(s).PublishedbyElsevierLtd.Thisisan openaccessarticleundertheCCBY-NC-NDlicense( http://creative-commons.org/licenses/by-nc-nd/4.0/).

Introduction

Immunotherapy is a promising treatment option for

patientswithdifferentcancersresistant to conventional

therapies.TherapeuticregimenssuchasadoptiveTcell

transfer(ACT),cancervaccinesandimmunecheckpoint

inhibitors (ICI) (e.g. a-PD-1, a-PDL-1 or a-CTLA-4

antibodies), harness the ability of the immune system

torecognizeandrejectthetumor[1].Tcellsarethemajor

playersinthesetherapeuticoptionsduetotheirantigen

specificity,robustkillingcapacityandlongevity.Despite

theseintrinsicTcellcharacteristics,immunotherapyhas

beenlargelyineffectiveforthemajorityofpatients.Itis

believed that this low efficacy is due to the highly

immunosuppressive tumor microenvironment (TME)

which limitsthe infiltrationof theinvigorated immune

cellsdeliveredbyimmunotherapy. Cancercell

metabo-lismcontributesgreatlytothisimmunosuppressive

envi-ronmentbydepletingnutrientsandsecretingmetabolites

thatinhibitanti-tumoraleffectorTcells[2].

Cancercellsarewellknownfortheirabilitytomodulate

their metabolic processes in order to fuel their

uncon-trolled proliferation and create an environment which

prohibitstheinfiltrationofeffectorTcellswhile

recruit-ingimmunosuppressive celltypes suchas regulatoryT

cells (Treg) or tumor-associated macrophages (TAMs)

[3]. Research suggests thatthe metabolic reprograming

ofcancercellsisoneofthemajorfactorsresponsiblefor

the development of therapy resistance [4]. Therefore,

targeting metabolic processes of cancer cells and/or T

cellsincombinationwithimmunotherapycouldboostthe

efficacyofthis therapeuticoption.

An additional parameter to consider when designing

therapeutics targeting cancer-metabolism or

immuno-metabolism is the organ in which the tumor develops.

Different tissues support different cellular functions

which are accompanied by distinct metabolic needs.

Anexampleoftissuewithaveryspecificcellular

metab-olismistheprostate. Epithelialcellsin theprostateare

wellcharacterizedtoproducecitrateinsteadofoxidizing

it in the TCA cycle as it is common in other tissues

(Figure 1) [5].Furthermore, different oncogenic

addic-tionsoroncosuppressivecues(recurrentinsomebutnot

inothertumorhistotypes)canrewirethemetabolismof

cancer cells towards a specific direction. For example,

copynumbersofoncogenicKRASmutantinlungcancer

(LC) correlate with augmented glucose uptake and

increasedchannelingofglucose-derivedmetabolitesinto

theTCAcycle[6].Thesetwoconsiderationsleadtothe

obvious conclusion that cancer types originating in the

sameorgancanbecharacterizedbyverydifferent

meta-bolic profiles requiring a different treatment approach

(Box1).

Thepresenceoftumorinfiltratinglymphocytes(TILs)in

solid tumors is strongly correlated to improved disease

outcome.TcellsrepresentthemajorcelltypeinTILs.Of

(2)

clinicalresultsandarethereforeconsideredessentialfor

the fight against cancer.On the other hand, Tregcells

contributeto theimmunosuppressive TMEandinhibit

anti-tumorTcellfunction.Unsurprisingly,inmostcases

the presence of Treg cells in the tumor hasa negative

effectonpatients’diseaseoutcome.AlthoughTcellsare

undoubtedly important players in the body’s immune

responseagainstallcancertypes,thisresponsehasbeen

showntovarybetweendifferentcancers.Consideringthe

greatmetabolicheterogeneitybetweencancertypes,and

evenbetweencancerswiththesameorigin,itis

reason-abletoconsiderthatdifferencesinTcellresponsearea

direct consequence of these metabolic characteristics.

Furthermore, generally, effector T cells share many of

thesamemetabolicfeaturesasmostcancercells(suchas

increased aerobic glycolysis and anabolic metabolism).

Thisleadstometaboliccompetitionbetweenthesecells

whichultimatelyimpactsdiseaseprogression.For

exam-ple, cancer cells deplete glucose in the TME at the

expense ofTcellsthereforeinhibiting theiranti-cancer

functions[7].

In this review we discuss the link betweencancer cell

metabolismandTcellresponseusingexamplesfromthe

mostrecentresearchdoneonmelanoma,lung,breastand

prostate cancers which are some of the most common

primary cancersites.Recently describedpossible

meta-bolic targetsfor thetreatment of thesecancers are

out-lined inTable 1.

Melanoma

Melanomaisoneofthemostaggressivecancersduetoits

ability to disseminate from a small primary tumor.

According to the American Cancer Society there will

be over 100000 cases of melanoma diagnosed in the

US in 2020 making it the cancer type with the fifth

highest incidence[8].Melanomaisahighlygenetic

het-erogenoustypeof tumor.The mostcommon oncogenic

mutation is BRAF (in about 45–50% of melanoma

tumors), followed byNRAS (30%) and NF1 (10–15%),

allgenesinvolvedinthemitogen-activatedproteinkinase

(MAPK) pathway,while theremaining 5–10% of

mela-noma areconsidered as triplewild-type tumors(TWT)

andaredrivenbyothermutations.Inphysiological

con-ditions, melanocytesarequiescent cells withalow

pro-liferativerate.Theirmainroleistoproduceandtransfer

melanintothekeratinocytesinordertoprotecttheskin

againstUVradiations.Inhealthytissue,melanocytesrely

heavily on oxidative phosphorylation and are able to

survive reactive oxygen species (ROS) and UV-induce

damage thanks to melanin and a good antioxidant

response (Figure1).

Figure1

(a) (c)

(d) (b)

Current Opinion in Biotechnology

Metabolicfeaturesinhealthytissue.(a)Metabolisminhealthybreasttissueischaracterizedbylowconcentrationsofintracellularserine,high productionandsecretionoftheadipokineleptin.Additionally,breasttissuehaslowamountsofROSdueinpartbyhighantioxidantresponse.(b) MelanocytesinhealthyskinhavealowproliferativerateandmetabolismthatisdominatedbyOXPHOS.Highantioxidantresponseandmelanin pigmentationprotectthesecellsfromUVradiationandROS.(c)Lungepitheliumreliesmainlyonaglycolyticmetabolismandproduceshighlevels oflactateeveninpresenceofoxygen.Moreover,lungepitheliumischaracterizedbyhighlipidssynthesisandexpressionofthemetabolism sensorAMPK.(d)ProstateepitheliumisknownforaccumulationofZincandhighcitratesynthesisaswellasacetylatedpolyaminesecretion.

(3)

Melanomacellmetabolism

Triggering of theMAPK pathway in melanoma cancer

cellsleadstoanupregulationofgeneexpressionrelated

toglycolysiseveninrathernormoxicconditions,resulting

inan aggressivephenotype.InBRAF-mutated

melano-mas,theMAPK-activatedp90ribosomalS6kinase(RSK)

directlyphosphorylatesPFKFB2,anenzymeinvolvedin

theglycolysispathway.Phosphorylation-deficientmutant

of PFKFB2 decreased aerobic glycolysis and reduced

tumorgrowthinmousemodelsofmelanoma[9].Hence,

itisnotsurprisingthatthemostcommontherapiesforthe

treatmentofmelanomapatients,namelytheinhibitorsof

BRAForMEK(adownstreamregulatorofMAPK),have

anevengreaterbeneficialeffectwhencombinedandthis

effectcanbepartlyduetoablockincellmetabolism(and

glycolysis).Thisandothermelanomaspecific metabolic

characteristicsaredepictedinFigure 2.

However,thereisstillalargefraction of patientsthat relapse

orbecometherapyresistant.Recentpapershavelinkedthe

resistantphenotypeofmelanomatumorswithasubsetof

cancercellscharacterizedbyanenhancedoxidative

metab-olism. Melanoma cells resistant to MAPK inhibitors

developanoxidativephenotype,relyingmoreonglutamine

consumptionandoxidation[10].Moreover,the

upregula-tionof ATP-Citrate Lyase(ACLY)specificallyactivates,

throughepigeneticregulation,theMITF–PGC1aaxisto

promotemitochondrialbiogenesisandmelanomagrowth.

Interestingly,thecombinationofMAPKandACLY

inhi-bitionwasabletoreducetumorgrowth[11].

Inaddition, Gabra et al. showedthat dietary glutamine

supplementationreducestumorgrowthandimprovesthe

response to BRAF inhibitors. Specifically, it leads to

increased intra-tumoral a-ketoglutarate concentration

thatdriveshypomethylationof H3K4me3, thereby

sup-pressingepigeneticallyactivatedoncogenicpathways in

melanoma[12]

The most relevantcause of morbidity and mortality in

melanoma is metastatic disease. Therefore, a major

research focus is on targeting the metastatic spread of

melanomacells.Theaccumulationoflactateinthetumor

microenvironment(TME), consequentto theincreased

glycolyticfluxhasanimportantroleinmetastasis

forma-tion.Analysisofpatient-derivedxenograftsandsyngeneic

models revealed that cells with the greatest metastatic

potentialwereableto utilizelactateasacarbon source,

throughtheupregulationofmonocarboxylatetransporter

1(MCT1).Consistently,theinhibitionofthistransporter

had little effect ontumor growth but induced a strong

impairmentonmetastasisformation[13].Otherfindings

linkedAMP-activatedProteinKinase(AMPK)activation

in metastatic lesions with an increased oxidative

phos-phorylation (OXPHOS) and glutamine consumption.

Consequently,thesecells showedahigherresistance to

Box1 Mainconclusionsandfutureimpactoftissueandtumor specificityof(immune)metabolisminthetreatmentofdifferent typesofcancer:

Immunotherapyrepresentsasignificantbreakthroughinmodern medicine.Ithasimprovedpatients’livesinmanydifferentdisease areas,fromautoimmunedisorderssuchasasthmaandatopic der-matitiswhereitcontributestobetterqualityoflife,tocancerwherein somecancertypesitcancontributetocancercureandextended lifespan.Despiteitsgreatsuccessinasubsetofcancertypes, immunotherapyisstilllargelyineffectiveinmanysolidtumortypes andinlargepatients’subsets.Inthequestforareasonwhy immu-notherapyiseffectiveinsomecancertypes(belongingtoaspecific histotypeorcarryingadefinedgeneticbackgroundormutations), butnotinothers,scientistshaveturnedtocellularmetabolismfor answers.Generalcharacteristicsofcancercellmetabolismarewell described,howeverdependingontheorganoforiginandsome recurrentmutationsinoncogenesoroncosuppressors,cancercell metabolismcanvarysignificantly.Thisisgreatlyinfluencedby gen-eralorganmetabolismand/oractivationofmetabolicpathways/ dependencythataredownstreamthegeneticalterations. Conse-quently,cancercellmetabolisminfluencestheinfiltrationandthe activityofimmunecellsattheprimarytumorsite.Researchisstill lackingontheconnectionbetweencancercellandimmunecells metabolism.ConsideringTcellsarethemaintargetsof immu-notherapy,theyaretheobviouschoiceofcelltypestofocuson. Whatisalsoobviousisthatonesolutionwillnotworkforallcancer types,thereforewemuststudymetaboliccharacteristicsandidentify targets,inbothTcellsorcancercells,thatcouldbeexploitedin ordertoboostimmuneinfiltrationandresponsewithindifferenttypes ofsolidtumors.Thiskindofapproachwouldlikelybeusefulin combinationwithconventionalimmunotherapeuticssuchasimmune checkpointinhibitorsoradoptivecelltransfer.

Therecentresearchonimmune-andcancercellmetabolismin differentcancertypes,withdifferentorganoforigin,presentedhere pointstodifferentmetabolicpathwaysaspromisingnewdirections forcancertherapy.Forexample,inmelanoma,recentfindings demonstratethatInosinesupplementationinvivocanincreasethe efficacyofanti-PD-1andACTtreatments[34]aswellaspointsto methionineasanutrientspecificforCD8+_T_cells_{functionality}_[₃₅_].

Inlungcancer,newreportpresentsevidenceoftheroleof deace-tylaseSirtuin-2(Sirt2)asametabolicimmunecheckpoint,which negativelyregulateskeyenzymeofglycolysisandoxidative phos-phorylationinCTLs[54].Inbreastcancer,itwasshownthat targetingtheleptin-STAT3-FAOpathwayresultsinenhancedanti tumoralCD8+_T_cell_response_in_addition_to_inhibiting_cancer_cells [77].Inprostatecancer,arecentreportsuggeststhepotentialofa therapeuticeffectofIDOinhibitionincombinationwithdifferent typesoftherapies,suchaswithothermetabolictargetsinadditionto immunotherapy[96].

Currentlymuchoftheresearchisdoneinasmallsubsetofcancer models,partlyduetotheeaseofmanipulationinsomeanimal modelsandtheavailabilityofwelldescribedcancercelllines. However,itisclearthattheresearchdoneononecancertypeisnot necessarilyapplicabletoothercancertypes,orevensubtypeswithin thesamebroadercancertypecategory.Thoughmoretime con-suming,spontaneouslyoccurringtumors(enforcedbyexposureto carcinogenicagentsordiet)mightresemblesomeoftheaspects thatareseeninhumans,toabetterextentthangraftedcancercells (evenwhenorthotopic)oroncogene-drivengeneticallyengineered mousemodels(GEMMs).Therefore,moreresearchfocusingon tumorswithdifferentorgansoforiginsthatresemblemoreclosely what’shappeningin‘cancerevolution’isurgentlynecessaryinorder todeterminethemosteffectivetherapeuticforpatientsinflictedby thisdeadlyandheterogeneousdisease.

(4)

Table1

Potentialmetabolictargetsforthetreatmentofmelanoma,lung,breastandprostatecancer

Cancer type Cancer subtype Potential Metabolic Target

Roleindisease Pharmacologicalinhibitor Clinicalobservations Proposed therapeutic setting Reference Melanoma Braf mutated RSK InhibitionofRSK decreasesthemetabolic fluxincancercell reducingtumorgrowth

LJH685 None Unknown [9]

ACLY InhibitionofACLY sensitized

toMAPKinhibitionby suppressing MITF–PGC1aaxis.

HCandSB-204990 None Incombination withMAPK inhibitor

[11]

GRM1+ glutamate InhibitionofGLSand glutamaterelease promotesapoptoticcell death

CB-839+riluzole None Unknown [18]

MAPKi resistant

OXPHOS InhibitionofOXPHOS metabolismreducethe numberofbrain metastasisandimprove micesurvival

IACS-010759 None Unknown [15]

Unspecified

OXPHOS Inhibitionofoxidative metabolismincancer cellsimproveTcells activityandresponseto a-PD-1 None Incombination witha-PD-1 [21] LDHA (cancer cells)

Inhibitionofcancercells glycolysispromotesT cell-mediatedapoptosis

GSK2837808A None Incombination withACT

[19]

LDHA(T cells)

PretreatmentofTcellwith LDHiandIL-21promotes stemcellmemoryTcells formationandimproves ACT

NCI-737 None Incombination

withACT [26] MCT1/ MCT4 Inhibitionoflactate transportrestricted cancercellsglycolysis andimprovesTcells function

Diclofenac None Incombination witha-PD-1 andCTLA-4

[20]

MCT1 InhibitionofMCT1on cancercelldoesnotaffect primarytumorgrowthbut reducesthenumberof metastasis

AZD3965 None Unknown [13]

Enolase1 Addingpyruvaterestore glycolyticandoxidative metabolismofCD8+Tcell improvingtheireffector functions

pyruvate None Unknown [25]

FATP1/ SLC7A1

FATP1inhibition decreasesfattyacid uptakeincancercellsand reducesmelanoma growthandinvasion

Lipofermata/CB16.2 None Unknown [16]

CD36 BlockingofCD36 impairedTregand reinforcesanti-tumor immunity a-CD36monoclonal antibody None Aloneorin combination witha-PD-1 [29] Glutamine metabolism Glutamineantagonism suppressoxidativeand glycolyticmetabolismin cancercells,while promotesTcell anti-tumorfunctions

JHU083 None Aloneorin

combination withACTor a-PD-1

(5)

Table1 (Continued) Cancer type Cancer subtype Potential Metabolic Target

Roleindisease Pharmacologicalinhibitor Clinicalobservations Proposed therapeutic setting

Reference

SphK1 InhibitionoftheSphK1/ PPARgaxisinTcells improvesTcellmediated tumorresponse

PF543 None Incombination

witha-PD-1 [32]

UCP2 OverexpressionofUCP2 incancercellsincrease CD8+_T_cell_and_dendritic cellanti-tumorimmunity

rosiglitazone None Incombination witha-PD-1

[23]

PPAR-a StimulationofPPAR-a increasesFAOand improvesTILs functionality Fenofibrate(PPAR-a agonist) None Incombination witha-PD-1 [27] GCLC DeletionofGclcinTreg cellsreducesGSH productionanddisrupt Tregfunctionality

buthioninesulfoximine None Unknown [31]

Glutaminedietary supplementation increasesglutamineand itsdownstream metaboliteaKGdriving hypomethylationof H3K4me3 Glutamine supplementation None Aloneorin combination withBraf inhibitor [12] ROS Cell-permeantformof cysteineincreases glutathionesynthesis, thereforereducesROS andrestoreself-renewal inexhaustedTcells

N-acetylcysteine None Incombination withACT

[28]

Inosinesupportsthe functionofTeffcellsinthe absenceofglucose

Inosinesupplementation None Incombination witha-PD-1 andACT

[34]

SLC43A2 Itsinhibitionincreases methionineavailabilityfor theTcellsresultingin increaseddimethylation ofH3K79me2and improvedTcellimmunity.

Methionine

supplementationorBCH

SLC43A2ishihgly expressedincancer cellscomparedtoT cellsandnormal tissue.Itcorrelates withpoorsurvival

Incombination withai-PDL-1 [35] Lung NSCLC GLDC GLDCinhibitionimpairs pyruvatemetabolism, inducingtumorregression

GLDC-shAON None Aloneorin combination withcisplatin

[37]

PRODH PRODHinhibitionreduces tumorgrowth,block epithelialtomesenchymal transitionandinhibits inflammatorycytokines releasesuchasCXCL1, LCN2andIL17C

L-THFA PRODHhighly expressedinAdC tumorscomparedto normaltissue

Unknown [48]

Malin Downregulationofmalin inducesnuclearglycogen accumulationand decreasehistone acetylation,thus promotingtumorgrowth

None HighmalinmRNA expression correlateswith bettersurvival Unknown [50] IGF-1R Inhibitionor downregulationofIGF-1R synergizewithPD-1 inhibitiontoboostCD8 +Tcells,leadingto reducedtumorgrowth

PQ401 HighIGF-1plasma levels/highIGF-R1 tumorexpressionis associatedwith resistanceto a-PD-L1immunotherapy Incombination witha-PD-L1 [55] NSCLC AdC MGST1 MGST1knockdown promotescancercells apoptosisand

suppressedtumorgrowth invivo

None Highexpressedin humanAdC.The expression correlateswithpoor overallsurvival.

(6)

Table1 (Continued) Cancer type Cancer subtype Potential Metabolic Target

Roleindisease Pharmacologicalinhibitor Clinicalobservations Proposed therapeutic setting Reference NSCLC SCC GLS GLSinhibitionovercomes resistancetomTOR inhibition CB-839 None Incombination withmTOR inhibitor (MLN128) [47] NSCLC EGFR mutant JNK/ Glucose metabolism JNKdirectactivationor indirectactivationthrough inhibitionofglucose metabolism,induces EGFR-mutantNSCLC cellsapoptosisand overcomeresistanceto tyrosinekinaseinhibitors.

Anisomycin/ 2DG Expressionof phosphorylatedJNK issignificantly decreasedinEGFR mutantNSCLC comparedtoEGRF WT Unknown [39] NSCLS KRAS mutant

KRAS KRASinhibitioninduces aninflamedtumor microenvironment, leadingtotumor regression

AMG510 Inclinicaltrial (NCT03600883) inducedobjective partialresponseor stabledisease Aloneorin combination with carboplatinor immunotherapy [40]

ASNS InhibitionofASNS decreasestumorgrowth

L-asparaginase None Incombination withAKT inhibitor

[41]

AMPK AMPKdeletion

suppressestumorgrowth

None Unknown [45] NSCLC KRAS mutant KEAP1 deficient Pentose phosphate pathway

Inhibitionofthispathway abrogatestumorgrowth

6-AN None Unknown [42]

NSCLC KRAS mutant LKB1 deficient

Atg7 DeletionofAtg7 abrogatesbothtumor initiationandgrowth

None Unknown [43] NSCLC KRAS mutant LKB1 deficient KEAP1 mutant

GLS Thissetofmutations promotea glutamine-addictivemetabolism, renderingthetumor sensitivetoglutaminase inhibition CB-839 None Unknown [44] NSCLC SMARCA4 mutant

OXPHOS MutationsintheSWI/SNF complexinduce dependenceonOXPHOS, sensitizingtumorstoits inhibition

IACS-010759 Highexpressionof OXPHOS-related genesinSMARCA4 mutantpatients

Unknown [49]

SCLC

DHODH DHODHinhibition attenuatesprimarytumor growthanddelaysliver metastasis

Brequinar None Aloneorin

combination withcisplatin/ etoposide [52] MEK5/ ERK5 DepletionofMEK5/ERK5 pathwayperturbs differentlipidmetabolism pathwaysandsensitizes SCLCcellstomevalonate inhibition None Unknown [53] Breast TNBC Cholesterol synthesis pathway enzymes

Inhibitionofthispathway leadstoareductionof TNBCpropagation

Simvastatin(inhibitorof thekeyenzymeinthe cholesterolbiosynthesis pathway, 3-hydroxy-3- methyl-glutaryl-coenzymeA(HMG-CoA) reductase) None Unknown [66] Estrogen-related receptora (ERRa) Cpd29 None Incombination withGLS1and G6PDinhibitors [69]

(7)

oxidative stress and, therefore, a better survival [14].

Moreover, a transcriptomic analysis of brain metastasis

compared with patient-paired extracranial metastasis

revealedanincreasein OXPHOSinbrain lesions.

Con-sistently, inhibition of OXPHOS in murine models

reduces the number of metastasis in the brain but not

inthelungs[15].

Uponmetastaticdiseaseinitiation,melanomacancercells

initiatetheircross-talkwiththeadipocytespresentinthe

subcutaneous region[16].Adipocytes transferlipids via

FATP1/SLC7A1transporteroverexpressedinmelanoma

cells,boostingcancercellproliferationandinvasion[16].

FattyacidsarealsoimportantinthecaseofMift-mutated

melanomas,whicharepartoftheTWTgroup,wherethe

reductionofthistranscriptionfactordecreasesthelevelof

fattyacid saturation,impacting oncancer cell

prolifera-tion[17].

Inadditiontofattyacidsandglucose,glutamatealsoplays

animportant rolein melanoma.Hyperactivationof

glu-tamatemetabotropicreceptor(GRM1)inmalignant

mel-anoma is an oncogenic driver and GRM1- activated

melanomaswerefoundtoexhibitsignificantlyincreased

expression levels of glutaminase (GLS). In an in vivo

modelof human melanomacells engraftedin

immuno-deficient mice, concurrent inhibition of glutaminolysis

and glutamate release was shown to suppress tumor

progression[18].

To conclude, although glycolysis and its products

have a strong impact on melanoma tumor

aggres-siveness and progression, recent findings highlighted

theimportanceofothermetabolic pathways,likefatty

acid oxidations or glutaminolysis, that can

consider-ably affect cancer cells proliferation and therefore

tumor progression. Table1 (Continued) Cancer type Cancer subtype Potential Metabolic Target

Roleindisease Pharmacologicalinhibitor Clinicalobservations Proposed therapeutic setting

Reference

ER-and HER+

AKR1B10 promotesmetastasisby limitingtheoxidative stressandenablingfatty acidoxidationinthe metastaticenvironment

None None Unknown [70]

Obesity promoted BC

STAT3 DeletionofSTAT3leadsto increasedCD8+_T_cell effectorfunction

None None Unknown [77]

Not specified

Atg5 DeletionofAtg5resultsin increasedglycolysisin CD8+Tcellsandashiftto amemoryphenotype resultingindecreased tumorburdenine0771 mouseBCmodel.

None None Suggesteduse

combination with

immunotherapy [82]

Prostate IDO IncreasedIDOinhibits effectorTcellfunctions. IDOexpressioncouldbea mechanismofimmune evasionisprostatecancer

Severaldescribed (Epacadostat, Navoximod..) IncreasedIDO activityinpatients afterimmunotherapy Incombination with immunotherapy [95]

Hypoxia Hypoxicregionsoftumors arecharacterizedwith absenceofTcell infiltration.

hypoxiaactivated pro-drugTH-302 None Incombination with immunotherapy [93] Polyamine expression andMTAP Increasedpolyamine expressiontogetherwith theinhibitionofMTAP.

BENSpm(polyamine analogtoboost acetylatedpolyamine secretion)andMTDIA (inhibitorofMTAP)

None Suggesteduse togetherwith Docetaxel

[89]

SIRT7 SIRT7activationinhibits glycolysisandleadstothe upregulationofFAO. DeletionofSIRT7leadsto increasedCD8+_T_cell effectorfunctions.

ID:97491[100] None Unknown [91]

DRP1 DRP1Expression promotescancercell survivalundermetabolic stressconditions

(8)

Tcellmetabolisminmelanoma

TogetherwithBRAFandMEKinhibitors,

immunother-apyhasbecomeanimportanttreatmentoptionfor

mela-nomapatients.Thediscoveriesofimmunotherapyandin

particular ICI, had a strong impact on the survival of

melanomapatients.However,thereisstillalargenumber

of patients that do not benefit from this treatment. In

melanoma, T cells constitute a high percentage of the

immuneinfiltrate,buttheyoftenretainapro-tumoralor

exhaustedphenotype.Recentadvanceshaveshownhow

targeting melanoma cell metabolism can change the

TME, improving thefitness of theendogenous Tcells

and, therefore, the efficacy of immunotherapies

(Figure 3).Casconeet al., forexample, havelinkedthe

augmentedglycolysisofafractionofcancercellswiththe

resistance to ACT in melanoma patients [19].

Glycolysis-related protein, in fact, are among the most

studiedtargets incombinationaltherapieswiththe

ulti-mate goal to improve the outcome of immunotherapy

treatments. Consistently, restricting glycolysis using

diclofenac, as a non-specific inhibitor of both lactate

transportersMCT1andMCT4,reducescancercell

pro-liferationpreventingthemfromusinglactateasacarbon

source. ThisinhibitoraffectsalsoTcellmetabolismby

decreasing theirproliferation,but nottheiractivationor

cytokineproduction.Consequently,diclofenactreatment

increases the response to a-PD-1, improving mice

sur-vival [20]. Moreover, Najjar et al.compared an arrayof

melanomacelllinescharacterizedbyglycolyticor

oxida-tivemetabolism.Theydemonstratethatalowoxidative

metabolism results in aless hypoxicmicroenvironment

thathelpstheinfiltrationandactivationofCD8+Tcells,

Figure2

(a) (c)

(d) (b)

Tissuespecificcancermetabolism.(a)Breastcancerspecificmetabolism.ER+BCsubtypeischaracterizedbyReverseWarburgmetabolism– influencingneighboringstromalcellstoupregulateOXPHOSandchannelmetabolitestosupportcancercells.Regardingfattyacids,ER+BC prefersFASandFAOwhileTNBCreliesmoreonuptakeandstorageofexogenousfattyacids.HER+BCreliesheavilyonFASandhashighrates offattyacidstorage.Additionally,TNBCexhibitsclassicalWarburgmetabolismandisalsodependentoncholesterolsynthesis.ER+BChas upregulatedtriacylglycerols(TG),whileTNBCandHER+BCisknowntodownregulateTG.(b)Melanomametabolism.Melanomaprimarytumors haveincreasedglycolysis(mainlydrivenbyMAPK),increasedglutaminolysis(duetohyperactivationofGRM1)andaugmentedoxidative phosphorylationepigeneticallypotentiatesbyACLY.Moreover,dietaryglutaminesupplementationhasbeenshowedtoreducetumorgrowth. MelanomametastaseshavebeenshowntohaveincreasedlactateuptakeduetotheupregulationofMCT1,augmentedOXPHOSandincreased fattyacidmetabolism.(c)Lungcancermetabolism.SCLCischaracterizedbyanincreasedMEK5/ERK5kinasesactivitywhichpromotes proliferationandcontrolslipidmetabolism.NSCLChasbeenshowntoaccumulateglycogeninthenucleusandexpresshigherGLDC, mechanismsthatrewirecancermetabolismandsustainitsgrowth.Differentsubtypesharboralsometabolicdifferencesandinparticular,in adenocarcinoma(AdC)higherexpressionofMGST1hasbeendetectedandassociatedwithpoorprognosis,whereassquamouscellcarcinomas (SCC)arecharacterizedbyhigherglycolysisandmTORactivity,inhibitionofwhichleadstoapreferentialglutaminemetabolism.Oncogenic mutationsareoftenfoundinNSCLC,inducingdifferentmetabolicpathways.KRASactivatingmutationsinduceglutamineconsumptionand increasedaminoacidsuptakeandsynthesis.WhenKRASmutationsareaccompaniedbyKEAP1lossoffunction,cancercellsshowahigherPPP andAAuptake,whereas,inpresenceofLKB1lossoffunction,theyshowanincreasedautophagytosustainanelevatedmacromolecular biosynthesis.Additionally,inKRASmutant,LKB1deficientNSCLC,activatingmutationofKEAP1,inducehigherglutamineconsumption. Moreover,activatingmutationsinEGFRpromoteaglycolyticaddictivemetabolism,whereaslossoffunctionmutationsinSMARCA4promote dependencyonOXPHOS.(d)Prostatecancermetabolism.MalignantprostatecellsconverttocitrateoxidationandadecreaseinZinc

concentrations,togetherwithincreaseddemandforintracellularpolyamines.Androgenreceptorsignalinginprostatecellsupregulatesexpression ofDRP1whichleadstoincreasedOXPHOSandlipogenesis.

(9)

ultimatelyreducingtumorgrowthandimprovinga-PD-1

response[21].Incontrastwiththesefindings,a

proteo-mic analysis of different cohorts of melanoma patients

treated with a-PD-1 or ACT, highlighted an increased

lipidandoxidativemetabolisminpatientsthatresponded

better to the treatment. Additionally, augmented

mito-chondrial metabolismwas linked with ahigher antigen

presentationandIFNgsignaling[22].Accordingly,

dele-tionofAcetyl-CoAacetyltransferasetoimpedeoxidative

metabolism, strongly reduced MHC class 1 expression

andincreasedtumorgrowth[22].

Independentlyfromglycolysis,anotherstudyshowshow

Uncoupling protein 2 (UCP2) expression in melanoma

patientscorrelateswithTcellinfiltrationandprolonged

survivalrate.Itsoverexpressionincancercellswasfound

to change the cytokines composition of the TME,

increasing CD8+ T cellinfiltration in adendritic

cells-dependentmannerandfinallysensitizingtumorstoPD-1

treatment[23].Thiseffectseemstobeindependentfrom

the uncoupling function of UCP2, but the underlying

mechanismisstillunclear.

Gut microbiota composition can also influence the

responseto ICI.Analysis of serumandstoolmicrobiota

oftwodifferentcohortsofmelanomapatientsrevealedan

association between the two short chain fatty acids

(SCNA)butyrateandpropionatewithclinicaloutcomes.

A high concentration of these SCNA correlates with

resistancetoa-CTLA4.Inmice,treatmentwithbutyrate

reducestheefficacyofa-CTLA-4treatmentleadingtoa

decreaseinCD80andCD86expressiononthesurfaceof

dendriticcellsandadecreaseinmemoryTcellsnumber

[24].

Ontheotherhand,targetingthemetabolismofTcells

can also have a strong impact on their function and

polarization. TILs in human melanoma display an

impairment of both glycolytic and oxidative

metabo-lism,due tothe compromisedactivityofenolaseI,an

Figure3

(a) _(c)

(d) (b)

MetabolicfactorsinfluencingTcellresponsesindifferenttumortypes.(a)Tcellmetabolisminresponsetobreastcancer.Leptinexpressedby mammaryadipocytesactivatesSTAT3inCD8+_T_cells,_which_induces_FAO_and_inhibits_glycolysis_leading_to_decreased_effector_function._Deletion ofAtg5inTcellsresultsinashifttoamemoryphenotypeandanincreaseofproinflammatorycytokineexpression.InhibitionofFABP5leadsto increasedTregsuppressivefunction.(b)Tcellmetabolisminmelanoma.CompromisedactivityofEnolaseIleadstoimpairedglycolysisand oxidativemetabolismandreducedeffectorcellfunction.ExhaustedTcellspresentmitochondrialdysfunction.FAOandmethioninearenecessary tomaintainTILfunctionality.LDHAinhibitiontogetherwithIL-21treatmentofTcellsbeforeACTresultsinincreasedTscmpolarizationandamore effectiveanti-tumoralresponse.CD36enhanceslipiduptakeinTregcells.TargetingtheS1P-PPARgpathwayorGSHproductionmediatedby GCLCdecreasesTregpolarizationandfunction.(c)TcellmetabolisminLungcancer.Sirt-2actsasametabolicimmunecheckpoint.Itsinhibition enhancesTcellsmetabolicfitnessandeffectorfunction.Shorttermstarvation(STS)increasesCD8/Tregratioandsynergizeswithanti-PD-L1 treatment.(d)Tcellmetabolisminprostatecancer.IDOexpressionleadstoadecreaseintryptophanandsubsequentlyintheinhibitionof effectorTcellfunction.

(10)

importantglycolyticenzyme.Ultimately,restorationof

its function improves both glycolytic and oxidative

metabolism, boosting effector T cell functions [25].

Interestingly, a transcriptomic analysis aimed to

ana-lyze the differenteffectsof IL-2andIL-21 on Tcell

metabolism, highlighted LDHA as one of the most

differently expressed genes. IL-2 treatment increases

T cell glycolytic flux and polarizes them towards an

effector phenotype, whereas IL-21 maintains a more

quiescentmetabolismandpromotesstemcellmemory

T cells (Tscm). Accordingly, LDHA inhibition

com-bined with IL-21 treatment of T cells before ACT,

increases the formation of Tscm cells, resulting in a

more profound anti-tumor responses and a prolonged

host survival [26]. In addition, CD8+ T cells isolated

frommurineandhumanmelanomatumorsdemonstrate

an increasedfatty aciduptake and oxidation [27]. In

particular, catabolism of fatty acids is necessary to

maintain TIL functionality [27]. Another study

showedhowexhaustedTcellshavedefectinoxidative

phosphorylation.Anantioxidanttreatmentwasenough

to restoreTcells functions[28].

Tregcellsinfiltratingdifferenttypesoftumors,including

melanoma,haveanenhancedlipiduptakemediatedbythe

transporterCD36,whichalsofuelsoxidativemitochondrial

fitnessviaPPAR-b.Accordingly,geneticdeletionor

phar-macologicaltargetingofCD36decreaseTregnumberin

thetumor,createalesssuppressiveTMEandsynergize

witha-PD-1,thusleading toareducedtumor growth [29].

Interestingly, notonly fattyaciduptakeis importantfor

Tregcellsproliferation,butalsoFAS.Bothglycolyticand

oxidative pathwaysare necessaryto fuel FASand allow

TregexpansionintheTME.Theabilityofthiscelltypeto

utilizedifferent metabolicroutesconfersthemagrowth

advantageascomparedtoconventionalTcells[30].

Moreover,Tregcellsdisplayahighlyantioxidant

pheno-type mediated by reduced glutathione (GSH). Specific

deletionofGclc,asubunitoftheglutamatecysteineligase

involvedinGSHsynthesis,revealedhowtargetingGSH

productionin Tregcells reducesserine metabolismand

impactsonthefunctionalityofthisTcellsubset.Invivo

Gclc deletion induces a strong autoimmune effect and

improves anti-tumor immunity. In particular, ACT of

GclcdepletedTregcellstogetherwithTeffector(Teff)

cellshadagreateranti-tumoralactivitycomparedtoACT

withwild-typeTregandTeffcells,duetotheirreduced

immunosuppressiveeffect[31].

Furthermore, itis knownthatSphingosine 1-phosphate

(S1P)controlsTcellegressionfromlymphoidorgansand

TregandTh17cellpolarization.Chakrabortyetal.

dem-onstrated that T cells lacking sphingosine kinase 1

(SphK1), the enzyme involved in the production of

S1P,maintainTcentralmemoryphenotypeanddecrease

their polarization towardsTreg cells. Theyproved that

these cells exhibitan increasedOXPHOS and lipolysis

mediatedbyPPARgandthattargetingtheS1P-PPARg

axisimproves theanti-tumor response[32].

Finally,inhibitionofglutaminemetabolismbytheuseof

apro-drugof6-diazo-5-oxo-L-norleucine(DON), a

glu-tamineantagonistthatbecomesactiveintheTMEupon

enzymatic cleavage, has a strong impact on the overall

metabolismofmelanomacellsandaffectstheir

prolifera-tion.Whereasthesamemoleculehastheoppositeeffect

onTcellsbyincreasingtheirproliferationandreducing

exhaustion and anergy. Tcells, differently from cancer

cells,areabletoutilizeacetateascarbonsourcetofuelthe

TCA, thanks to the up-regulation of acyl–coenzyme A

synthetaseshort-chainfamilymember1(ACSS1).

There-fore,the use ofthis inhibitorin mouse modelsshows a

strongdecreaseintumorgrowthandasynergisticeffect

witha-PD1 treatment[33].

Moreover, Teff cells are able to utilize inosine as an

alternativeenergysourcewhenglucoseismissing,fueling

theirproliferationandeffectorfunctions.Onthecontrary

differentcancercelllinesseem tobeunableto use the

same metabolite.Consistently,inosinesupplementation

invivoincreasestheefficacya-PD-1andACTtreatments

[34].

Analternativestudyhighlightsinsteadtheimportanceof

methionine as anutrient specific for CD8Tcells

func-tionality. The deprivation of methionine in the TME

caused by its highly consumption from cancer cells

decreases dimethylation at lysine 79 of histone H3

(H3K79me2) in Tcells, leading to a low expression of

STAT5 and impaired T cell immunity. Consistently,

pharmacological or genetical inhibition of SLC43A2 in

the tumor improves response to immune-checkpoint

therapy[35].

Inconclusion,recentworkbroughttoattentionthe

het-erogeneity of melanoma metabolism highlighting the

effectof cancercellmetabolism onTcellfunctionality

andviceversa.Inaddition,theselastfindingsunderline

theimportanceofstudyingmetabolicvulnerabilities

spe-cifictocancercells,inordertoopennewpossibilitiesfor

the creation of specific therapies targeting cancer cells

whileatthesametimepromotinganti-tumorimmunecell

functions. While contradictory results leave space for

further study regarding the contribution of oxidative

metabolism to immunotherapy resistance, growing

evi-dencesindicatetheimportanceoftargetingglycolysisand

tofurtherstudytheroleoffattyacidsinthepolarization

and functionality of Tcells, in particular of Tregcells,

opening thewayfor newpossible treatments.

Lung

LCdisplaysthehighestincidenceworldwide(2millionin

(11)

deathsworldwide (1.7million in2018),duetoa

combi-nationoflatediagnosisandpoortreatmentoptions.This

tumortypeishighlyheterogeneousandcategorizedinto

two main groups: Non-Small Cell Lung Cancer

(NSCLC), which represents 80–85% of LC patients,

and Small Cell Lung Cancer (SCLC), that affects the

neuroendocrinetissueofthelungsandoccursin15–20%

of the cases.Based on thehistology NSCLCis further

subcategorized into Adenocarcinoma (AdC), Squamous

cellcarcinoma (SCC) and Large cellcarcinoma (LCC).

SmokingisthefirstriskfactorforLC,andisthecauseof

80–90% of the cases. This holds true particularly for

SCLC,thatdevelopsalmostexclusivelyinheavy

smok-ers, and for the SCC subtype among the NSCLC. In

healthy conditions,lung epitheliumformsaconduit for

gasexchange,but itisalsoengagedin morespecialized

energy-consuming activities such as airway clearance,

throughphagocytosisandciliarymotility,andproduction

ofpulmonarysurfactant,supportedbyahighcellularlipid

synthesis.Forthesereasons,lungepitheliumisoneofthe

tissues withthe highestglucose consumption

accompa-niedbyanelevatedlactateproduction,tominimizelocal

oxygen utilization, thus enhancing its delivery to the

other tissues. Moreover, energy production and usage

arefine-tunedbythemetabolicregulatorAMPK,highly

expressed in this tissue (Figure 1) [36]. LC specific

metabolicfeaturesarerepresented inFigure2.

In2015 threedifferentdrugs, namelynivolumab,

pem-brolizumab(targetingPD-1),andatezolizumab(targeting

PD-L1), received FDA approval for the treatment of

advancedstageNSCLC,leadingtoanimproveinsurvival

andinsomecasestotumorregression.However,80%of

NSCLCpatientsdonotrespondtoimmunotherapywith

someevenundergoinghyperprogression.Therefore,itis

becomingimportanttoidentifybiomarkerspredictiveof

response or resistance to immune checkpoint therapy,

andwaystoimprovesuchresponseorsensitizethetumor

toICI.

Lungcancermetabolism

NSCLC, as many other tumor types, relies on a high

glucoseandpyruvateconsumptiontosustainitsgrowth.

Amongalltheglycolyticpathways,glycinedecarboxylase

(GLDC),whichcatabolizesglycinetoyield

5,10-methy-lenetetrahydrofolate (MeTHF), is frequently

upregu-latedinvarioustypes ofcancerincludinglung,prostate

andbrain.ParticularlyinLC,itsupportstumorgrowth,as

shownbythefactthatGLDCinhibitionimpairspyruvate

metabolism,thusinducingtumorregressionbothinvitro

andinvivo[37].Increasedproliferationandmitochondrial

activity can be responsible for a higher production of

ROS, deleterious products for cellhomeostasis. In this

context,anddifferentlyfromwhataforementionedinthe

contextofmelanomas,GlutathioneS-transferases(GSTs)

are important for cells detoxification. In particular, the

microsomalisoform1(MGST1)issignificantlyincreased

inhumanAdC,comparedto normallungtissue, andits

expressionisassociatedwithpoorprognosis.Accordingly,

MGST1 knockdown, by acting on the mitochondrial

apoptoticrelatedproteins,inducesapoptosisinAdCcells

andsuppressesinvivotumorgrowth[38].

NSCLCisalso characterizedbyamultiplicity of

recur-rentoncogenicdrivermutations,underliningahigh

vari-ability evenamong the same histological NSCLC

sub-type. Different genetic alterations can induce diverse

metabolic adaptation and immune response, that can

beexploitedfortherapeuticapplication.Agreatexample

inthissenseisprovidedbytheEpidermalGrowFactor

Receptor(EGFR),atyrosinekinasereceptor,altered in

15–20%ofAdC.EGFR-inducedglycolysissustains

can-cerproliferationandmaintains,atthesametime,EGFR

stability,thuscreatingapositiveloop[39].Therefore,the

useofTyrosineKinaseInhibitorshasbeeneffectivefor

thetreatment ofpatientsharboringEGFRmutations.

KRAS, a member of the Ras family small GTPases, is

involved in the transmission of extracellular mitogenic

signals,and its aberrant activationsustains cellular

pro-liferation and survival through RAF-MEK and PI3K/

AKTpathways.Therefore,KRASmutationsdrive

multi-plecancers,includingLC,wheretheyarefoundinmore

than30%ofNSCLC,andmostlyinAdC.Despiteitshigh

incidence, differentlyfrom EGFR,KRAS isdifficult to

target. Indeed,the firstKRAS effectiveinhibitor isthe

smallmolecule AMG510,which wasrecently proved to

inducetumor regressionalone andin combination with

ICIinamodelofcolorectalcancerandisnowinclinical

trial for multiple KRAS-mutant tumors, including

NSCLC, for which, based on the preliminary data, it

resulted in partial response or stable disease [40]. In

anattempttofindalternativetherapeuticstrategiestoits

directinhibition, the role of mutantKRAS in NSCLC

metabolismisbeingdeeplystudied.TheworkofGwinn

D.M.etal.,identifiedKRASasakeyregulatorofnutrient

deprivationresponse.ThroughPIK3-NRF2,KRAS

acti-vates ATF4 inducingthe expression of its downstream

targets,suchasglutaminetransporterSNAT1and

aspar-agine synthetase (ASNS), thus regulating amino acid

uptakeand asparaginebiosynthesis[41].KEAP1, akey

player in ROS detoxification, is a negative regulator of

NRF2. Interestingly KRAS-driven LC are frequently

paired with KEAP1 loss of function mutations. In this

context,KEAP1lossrewiresKRAS-mutanttumor

metab-olism,ononehandbyreinforcingATF4-mediatedamino

acid uptake and on the other byinducing the pentose

phosphatepathway(PPP),inhibitionofwhichabrogates

tumor growth [42]. KRAS activation is also commonly

coupledwith lossof LKB1function.Patients

character-ized by KRAS-mutant LKB1-deficient (KL) develop

moreaggressivetumors,showhigh frequencyof

metas-tasisandareresistanttoimmunotherapy.LossofLKB1

(12)

a nutrient deprived environment, by promoting

macro-molecular biosynthesis and inducing autophagy to fuel

thereprogrammedmetabolism.Consistently,deletionof

the autophagy essential gene Atg7 abrogates initiation

and growth of KL tumor,suggesting anewtherapeutic

strategyforthissubsetofNSCLC[43].Furthermore,in

KRAS-driven AdC, LKB1 loss induces often KEAP1

activation (KLK tumors). By acting cooperatively, this

set of mutations maintains energetic andredox

homeo-stasisinaglutaminedependentmanner,thusenhancing

KLK tumors sensitivity to glutaminase inhibition [44].

LKB1 is also the main positive regulator of AMPK.

Interestingly, in a KRAS-mutantmouse model,genetic

deletion of AMPK does notphenocopy LKB1 loss and

promotesinsteadtumorregression,pointingoutthat,in

nutrientrestrictiveconditions,aminimallevelofAMPK

activityisrequiredfor NSCLCgrowth[45].

SCC are highly aggressive, highly glycolytic and,

com-pared to AdC, show differential expression of a set of

genes involved in glucose and glutamine catabolism as

well as nucleotide and glutathione biosynthesis [46].

Moreover, SCC, unlike AdC, have frequent mutations

inthePIK3/AKTpathway,thatresultinanoveractivation

ofmTORandtumorprogression.However,single

thera-piesinhibitingthispathwayhavelimitedclinicalefficacy.

Asshownbyaninvivometabolicandmolecularprofiling,

SCCcanadapttochronicmTORinhibitionand

glycoly-sis suppression via theGSK3a/b pathway, which

upre-gulates glutaminolysis through the induction of GLS,

providing a proof of concept for combinatorial GLS

and mTORinhibition [47].

Epigenetics plays also an important role in metabolic

rewiringandtumorprogression.InNSCLCtheactivation

of Proline Dehydrogenase (PRODH) mediated by the

chromatinremodelingfactorLSH,inducestumorgrowth,

epithelial to mesenchymal transition (EMT) and the

expression ofinflammatory cytokinesincludingCXCL1

and IL17C[48].Another exampleistheSWI/SNF

chro-matinremodeling complex,whichisalso foundaltered,

mostly in adenocarcinoma. In particular, the chromatin

rearrangement mediated by the inactivation of

SMARCA4,partofthecomplex,leadstoenhanced

oxy-gen consumption and respiratory capacity. As a result,

SMARCA4 mutant xenografts show a dependence on

OXPHOS and a marked sensitivity to its inhibition

[49]. Moreover, unlike physiological lung tissue, in

NSCLCdecreasedabundanceofmalin,anE3ubiquitin

ligase, preventsnuclear translocationof glycogen

phos-phorylaseandinducesnuclearaccumulationofglycogen,

thus impairinghistoneacetylationandsustaining tumor

growth[50].

Because of the high heterogeneity of LC, it may be

important to unveil therelationshipbetweenmolecular

andmetaboliccellfeaturesonalargerscale.Inthissense,

aninvitroextensivecharacterizationofmetabolicfluxes

inahighvarietyofNSCLCcelllines,provedtobeuseful

inpredictingdependencyonspecificmetabolicpathways

andsensitivitytometabolicdrugssuchaspemetrexed,an

antimetabolitecurrentlyinuseforNSCLCtherapy[51].

SCLC is frequently characterized by loss of function

mutation in tumor suppressor genes TP53 and RB1,

but,unlikeNSCLC,itrarelypresentsoncogenicdriving

mutations, thus limiting therapeutic options. In this

regard, through a genetic screening approach, Li et al.

uncovered SCLC sensitivity toward disruption of the

pyrimidine biosynthesis pathway and identified

dihy-droorotatedehydrogenase(DHODH)asanewpromising

therapeutictarget[52].Anotherpathwaythathasbeen

studied in SCLC is the MEK5–ERK5 axis. This dual

kinase axis is known to be responsible for increased

growth andmetastasis andlower overallsurvivalin

dif-ferent tumor types includingbreast, prostate andcolon

cancer. This is also true for SCLC, where MEK5 and

ERK5 playa criticalrole in cell survival bycontrolling

lipid metabolism. Notably, the loss of MEK5/ERK5

perturbs cholesterol synthesis, making SCLC sensitive

to mevalonateinhibitionstrough statins[53].

Tcellmetabolisminlungcancer

AnimportantparameterfortheresponsetoICIseemsto

be the evaluation of PD-1/PD-L1 expression on both

cancer and immune cells, particularly TILs and CD8+

cytotoxic T cells (CTLs). In this context it has been

shown that in human NSCLC the presence of CD8+

TILs expressing high levels of PD-1 is predictive of

response and survival upon a-PD-1 treatment. Despite

thenumerousstudiesanalyzingtheinfiltrationofTcells

in LC, littleisknown abouttheirmetabolicfeaturesin

thecontextof specificnutrientsdeprived

microenviron-ment.Ahighlyrelevantrecentdiscoverypointstotherole

of deacetylase Sirtuin-2 (Sirt2) as a metabolic immune

checkpoint, which negatively regulates key enzyme of

glycolysisandoxidativephosphorylationinCTLs[54].

Moreover, the fact that Sirt2 KO can enhance T cells

fitness and effector function, leading to an effective

antitumor immune response, strongly supports T cells

metabolic manipulation as a promising strategy to

improvetheefficacyofthecurrentimmunotherapy

regi-mens[54].

AsapossiblewaytoimproveimmunotherapyinNSCLC,

Ajonaet al.showed thatshort termstarvationsensitizes

tumors to a-PD-1 treatment by reducing circulating

insulin-likegrowthfactor1(IGF-1)anddownregulating

IGF-1 receptor (IGF-1R)signaling in tumor cells

(Fig-ure3).Moreover,acombinationofIGF-1Rinhibitionand

PD-1 blockade impaired LC progression by boosting

intra-tumoral CD8+/Treg ratio, supporting the clinical

evaluation of IGF-1 modulators together with a-PD-1

(13)

SCLC is characterized by a high mutational burden,

providinga strongrationale for the use of ICI. Indeed,

despitethelimitedaddedsurvivalbenefit,FDArecently

approved the use of atezolizumab (a-PD-L1) together

with chemotherapy as first line treatment in extensive

stage SCLC (i.e. SCLC spreading farfrom theprimary

site,inthelungsthemselvesorelsewhere).However,like

NSCLC,alargenumberofpatientsdonotrespondtothe

treatment.Inthisscenario,studieshavebeenconducted

toincrease Tcells infiltrationbyacting oncancercells,

but manipulation of T cells metabolism has not been

exploitedsofar.

To conclude, an extensive effort has been done to

describetheheterogeniclungcancermetabolism.

How-ever,Tcellmetabolicfeaturesassociatedtothese

differ-entmicroenvironmentsarestillpoorlyunderstood,

under-lingtheneedforfurtherinvestigation. Inthisdirection,

singlecellRNAsequencingaimingtoperformlargescale

NSCLCphenotypingofstromalcellsingeneralorTcells

could be useful to improve the outcome of current

immunotherapies and to define new treatment options

[56,57].

Breast

Breast cancer (BC) is the most commonly diagnosed

cancerin women and hasthesecond highest incidence

overallworldwide[58].Cancersarisinginthebreasttissue

are diversein their(epi)genetic andmorphological

fea-tures as wellas metabolic profiles.BCcan beclassified

intothreedistinctmoleculargroupsbasedonexpression

of hormonal receptors (estrogen and/or progesterone

receptorpositive),humanepidermalgrowthfactor

recep-torHER2positive,andtriplenegativetumors(TNBC)

-lacking estrogen receptor, progesterone receptor and

HER2[59].Thisclassificationisthemajorcharacteristic

that determines the therapeutic approach to treat the

disease. However, it has been shown that even within

eachofthosegroupstherecanbeconsiderablemolecular

heterogeneity.

The correlation between the abundance of CTLs and

favorable disease outcome has recently been

demon-stratedinacohortof187BCpatients[60].Furthermore,

higherCD8+Tcellsinfiltrationcanbeusedasa

predic-tivefactorofresponsetonon-adjuvantchemotherapyfor

patientswith HER+and TNBC.Inregard toTreg cell

infiltration, there is a considerable disparity among the

differentBCsubtypes.MoststudiesfindthathighTreg

cellinfiltrationcorrelateswithpoordiseaseoutcome[61].

However,thereare somestudieswhich havefoundthe

oppositetobetrueinTNBC[62].

Breastcancercellmetabolism

Different types of BC also differ in their metabolic

profiles (Figure 2). TNBC subtype primarily utilizes

the Warburg effect metabolism characterized by high

glucoseuptakeandlactatesecretioneveninthepresence

of oxygen [63]. Additionally, TNBC relies heavily on

uptakeand storage of exogenous fattyacids whileER+

BCsubtypeprefersfattyacidsynthesis(FAS)and

oxida-tion.Furthermore,ER+BCsubtypedisplaysthereverse

Warburg metabolic effect – where neighboring stromal

cellsareinducedtoundergoaerobicglycolysisandthen

transferthecatabolitestothecancercellsforthemtouse

inOXPHOStosupporttheirproliferation.HER2+cancer

subtypecanbecharacterized byauniqueWarburglike

metaboliceffectwhichisheavilyreliantonFASandhas

beenshowntohaveincreasedlevelsofstoredfattyacids

[64].Consideringthat lipidmetabolism differs

substan-tiallybetweendifferentsubtypesEirikssonetal.soughtto

determine if the differences in the lipidome could be

usedasamarkertodistinguishthedifferentsubtypesfor

diagnostic purposes [65]. Analysis of the lipidome of

differentBCsubtype celllinesshowedthatcells ofthe

luminal BC subtype (ER and PgR positive) have

increasedabundanceoftriacylglycerols(TG) with

mod-erate or multiple unsaturated fatty acid chains, while

these lipids are significantly downregulated in HER2

andTNBCsubtypes[65].Furthermore,concerninglipid

metabolism, cholesterol biosynthesis was shown to be

essential for breast cancer stem cell propagation [66].

Additionally,inhibitionofcholesterolsynthesisenzymes

leads to significant reduction in TNBC mammosphere

growthandpropagation[66],thusprovidingevidencefor

thepotentialoftargetingcholesterolsynthesispathwayin

the treatment of TNBC patients. Lipid metabolism

reprograming has also been extensively linked to the

developmentofresistancetokinaseinhibitorsinBC[67].

ArecentstudyhasshownthatEMTinBCisassociated

withtheattenuationofsuccinatedehydrogenase(SDH)

[68].Thisenzymeispartofthemitochondrialelectron

transport chain and therefore an important player in

energymetabolism.With thisstudyRøsland et al.

pro-vide further proof of the importance of mitochondrial

processesinEMTandhighlightitspotentialasa

thera-peutictarget.Additionally,inhibitionofestrogen-related

receptora(ERRa)inTNBCcelllinesleadstoreduced

pyruvate entryinto themitochondria resulting incells

becoming more dependent on glutamine and glucose

oxidationandultimatelyleadstocancercellsbeingmore

sensitive to GLS and PPP inhibitors [69]. This

high-lights the need to evaluate the interplay between the

differentmetabolicfactorsandthepotentialofacascade

effectbetweendifferentmetabolicpathwayswhen

con-sidering therapeutic approaches targeting cancer

metabolism.

Sincethedevelopmentofmetastasiscorrelateswithpoor

prognosis,itisimportanttoidentifyfactorsthat

contrib-uteto amoreinvasive BCphenotype.Aldo-keto

reduc-tase AKR1B10 has been identified as one such factor.

(14)

stress and enabling fatty acid oxidation (FAO) in the

metastaticenvironment[70].Thismetabolicalteration

has been shown to be important only for metastatic

colonizationbutnotforprimarytumorgrowth,therefore

AKR1B10 inhibition represents a potential treatment

option for advanced BC. Additionally, high expression

of AKR1B10 could be used to identify patients with a

higher risk of metastasis. Even though the connection

betweenmetastasisand AKR1B10expressionhassofar

onlybeenshowninBC,AKR1B10ishighlyexpressedin

othercancertypes,includinglung[71]andpancreas[72],

suggesting the possibility to explore this factor as a

therapeutic targetin multiplecancertypes.

Consideringtheemergingnewresearch,fattyacid

metab-olism stands out as the group of metabolic processes

which have been shown to be most important for BC

cellgrowthanddevelopmentofresistancetomany

exist-ingtherapeuticapproaches.Furtherresearchisnecessary

todeterminehowtargetingdifferentmetabolicpathways

candiminishcancercellproliferationwhile atthesame

timeboostingTcellfitnesswithintheTME,ultimately

leadingtoimproveddiseaseprognosis.

Tcellmetabolisminbreastcancer

OutofthedifferentBCsubtypes,TNBCisconsideredas

theonetomostlikelyrespond toICItreatment.Thisis

largelyduetothefactthatTNBCgenerallyhasahigher

TILinfiltration[73],aswellashighexpressionofPD-L1

[74],whichisadirecttargetforthisclassof

immunother-apy.Infact,basedontheresultsofarecentclinicaltrial

(NCT02425891) [75], the FDA granted accelerated

approval for a-PD-L1 treatment using a monoclonal

antibody in combination with chemotherapy for

treat-ment ofPD-L1positive,unresectable, locallyadvanced

ormetastaticTNBC.Moreonthecurrentstateof

immu-notherapy in TNBC has been recently reviewed

else-where[76].

STAT3activationandCD8+Tcellfunctionshavebeen

correlatedinobesity-associatedBC.DeletionofSTAT3

inobesemice,whoareknowntospontaneouslydevelop

tumors,wasshownto decreaseFAO, increaseglycolysis

and improve CD8+ Tcelleffectorfunctions,leadingto

reduced BCincidence inthesemice[77].Additionally,

leptin, which is abundant in mammary adipocytes and

surroundingfattissue,wasshowntoinhibitCD8+Tcell

effector functions by activating STAT3, consequently

inducing FAO and reducing glycolysis. Furthermore,

leptin is one of thekey adipokines producedby breast

tissue adipocytes (Figure 1) thatcan promote BCstem

cell self-renewal and chemo-resistance[78]. Therefore,

targetingtheleptin-STAT3-FAOpathwaymayresultin

enhancedantitumoralCD8+Tcellresponseinaddition

to inhibiting cancer cells in obesity associated BC

(Figure 3).

Lipid chaperone fatty acid binding protein 5 (FABP5)

expression is increased in tumor infiltrating Treg cells

isolatedfromhumanBC[79]andmurinesubcutaneous

E.G7tumormodel[80].Curiously,invitrostudiesshow

that inhibition of FABP5leads to mitochondrial

altera-tionsand increasedsuppressivefunctionof Treg[80].

Becausethereisalimitedamountoflipidsinthetumor,it

isbelievedthatTregcellsincreaseFABP5expressionas

anadaptiveresponsetothelow-lipidenvironment.This

andotherindicationssuggestthatthemetabolicneedsof

Tregcells,bothinvivoandinvitro,arelikelydynamicand

depend ontheiractivationstage as wellas thenutrient

microenvironment atthesiteof activation[80].

Abioinformaticsstudy,whichanalyzedgeneexpression

datafrom4921cancerpatients,determinedthatelevated

expression of the hepatic lipase LIPC correlates with

higher infiltration of both myeloid and lymphoid cells

in several cancertypesincluding breast,melanomaand

NSCLC [81].In contrast high expression of aldehyde

dehydrogenase seven family, member A1 (ALDH7A1)

correlateswithlowerinfiltrationofimmuneeffectorcells.

Themechanismgoverningthesecorrelationsisstilltobe

investigated.

Autophagy is a cellular process linked to immune

response in various infections and recentlyalsoto

anti-tumoralCD8+Tcellimmunity.Micewithadeficiencyin

the genes essential for autophagy Atg5, Atg14, or

Atg16L1, have a significant decrease in the growth of

breast,prostateandcolorectaltumors[82].Additionally,

TcellswithadeletionintheAtg5geneexhibitashifttoa

memoryphenotypealongwithanincreaseof

proinflam-matory cytokinesIFNg andTNFa. Invitro andinvivo

mechanisticstudiesdoneinsyngeneice0771modelofBC

indicate that the suppression of autophagy caused by

deletion of Atg5 leads to increased glycolysis in CD8+

Tcellsandadecreaseintheconcentrationsof

S-adeno-sylmethionine (SAM) [82]. This metabolic alteration

consequently leads to changes in histone methylation,

whichisassociatedwithincreasedexpressionofeffector

andmetabolic genes,resultingin amemory Tcell

phe-notype.DisruptionofautophagyinTcellshastherefore

been proposed as a therapeutic strategy that could

enhanceimmunotherapeuticeffortinhumancancers.

With the breakthrough of CRISPR/Cas9 technology,

genomic screens represent a promising approach for

new targetdiscovery andtheadvancement of

immuno-therapy. One such screen was done on CD8+ T cells

selectingfortumorinfiltrationinmousemodelsofTNBC

in the context of immunotherapy [83]. This screen

identified severalgenes associatedwith higherCD8+ T

cellinfiltration,activationandeffectorfunctions.Oneof

these genes is Dhx37, encoding for an RNA helicase,

which, when deleted in CD8+ T cells, improved ACT

(15)

syngeneicmouseTNBCcelllineE0771.Theuseofmore

focusedscreeninglibraries,suchasthosetargeting

meta-bolic genes, could potentially lead to the discovery of

evenmore promising druggabletargets affectingT cell

infiltrationand fitnesswithintheTME.

Prostate

Prostatecanceristhesecondmostcommonlydiagnosed

cancertypeinmen, behindLC, andrepresentsamajor

cause of male morbidity and mortality. In 2018 alone,

therewereover1.2millionnewdiagnosisreported

world-wide and close to 360000 deaths related to prostate

cancer[58].Despitesomeheterogeneityinprostate

can-ceramongthepatientpopulation,thereisageneralized

phenotypethatiscommontoallpatientsincludingsome

metabolic alterations observed in prostate cancer cells.

Thesemetabolic eventsaredistinctbetweenthe

differ-entstagesofdisease-earlyversuslatestage.TheGleason

gradingsystemiscommonlyusedtodiagnoseandclassify

the severity of prostate cancer using histopathological

examination and scoringof thecells biopsied based on

itsresemblancetohealthyprostatecells.Score6isgiven

tolow-gradecancers,score7tointermediate,andscore8–

10 indicates high-grade and most aggressive prostate

cancer.Healthy prostateepitheliumcellsare knownfor

thehighaccumulationofZincaswellascitratesynthesis

andalackofcitrateoxidation(Figure 1).

Depending on the characteristic of thetumor, prostate

cancerpatientsaresubjectedtooneormoretreatments,

including surgery, androgen deprivation therapy and

chemotherapy. However, there are very limited

treat-mentoptionsformetastaticdiseaseandimmunotherapy

hasnotbeenproveneffectiveinprostatecancer.Prostate

cancerisclassifiedasa‘cold’tumorwithminimalTcell

infiltration, with several factors contributing to its

non-inflamedphenotype.Prostate cancerischaracterizedby

relatively few somatic mutations, compared to other

cancertypes,andconsequentlylowexpressionof

neoan-tigensthatcouldtriggeraTcellresponse.Otherfactors

contributingto thisnoninflamedphenotypeofprostate

cancer are downregulationof MHC Class I expression,

PTENlossandadeficiencyinIFN1signaling[84].Itis

possiblethattheuseofdrugstargetingmetabolismcould

synergizeandrenderthesetumorssusceptibleto

immu-notherapysuchasICI.DespitethisunfavorableTME,it

has recently been shown that it is possible to isolate

functional and anti-tumor reactive TILs from clinical

prostatecancersamples[85],thereforeprovidinga

ratio-nalefor further investigation into TIL immunotherapy

forprostatecancer.

Prostatecancercellmetabolism

Incancercellsthereisametabolicshiftinthisrespectand

malignant prostate cells convert to a citrate oxidizing

phenotypeaccompaniedbyadecreaseinzinc

concentra-tions(Figure 2). In line with this, prostatecancer cells

havebeenshowntoutilizetheTCAcycleandOXPHOS

morecomparedtobenignprostatecells.Prostatecancer

cellmetabolism is unique in that it does not generally

display the Warburg effect in glucose metabolism that

mostother cancers adhere to. The Warburg effect and

increasedglucosemetabolismtogetherwithlactate

secre-tioncanonlybeobservedinthelatestagesofdiseaseand

afteraccumulatingmultiplemutations[5].Usinginvivo

hyperpolarizedMRIimagingfollowedbytissue

histopa-thology, Granlund et al. demonstrate that tumors with

increased Gleason grades had significantly increased

levelsoflactate[86].Furthermore,increasedmetabolism

ofpyruvatetolactateisaugmentedintumorswith

homo-zygousdeletionof thePTENgene.

The remodeling of OXPHOS in prostate cancer has

recently been linked to mitochondrial DNA (mtDNA)

mutations and differentially expressed mitochondrial

genes,especiallyinhigh-gradecancers[87].Specifically,

increasedmutations inmitochondrialComplexI

encod-inggenes,correlatewithasignificantincreaseinsuccinate

oxidation.

Prostateepithelialcellsexhibitahighrateofacetylated

polyaminesecretionintheprostatelumen.Thisbehavior

ismaintainedbyprostatecancer cells,who additionally

also require high concentrations of intracellular

polya-mines leading to an increased demand for connected

metabolicpathways[88].Aninterestingnewtherapeutic

approach has been proposed to target this metabolic

vulnerabilityofprostatecancer[89].Inorderto cause

apoptosis,metabolic stressisinducedinprostatecancer

cells by using a polyamine analog N1,N11

-bisethylnor-spermine (BENSpm) which increases

spermidine/sper-mineN1-acetyltransferase(SSAT)activityresultingina

higheracetylated polyamineexport. At the same time,

cellsaretreatedwithasmallmoleculeinhibitor(MTDIA)

ofmethylthioadenosinephosphorylase(MTAP),thekey

enzymeinthemethioninesalvagepathway(MSP),which

isnecessaryforthecelltoalleviatethisstress,ultimately

leading to cell death. This approach was shown to be

effectivein differentprostatecancercellmodelsinvivo

andinvitro[89].

Androgenreceptorsignalingleadstoincreased

prolifera-tion and survival of prostate cancer cells. This is why

androgendeprivation therapy(ADT)isone ofthemain

treatmentoptionsformetastaticprostatecancerorupon

relapsetohelpcontrolthespreadof thedisease.

Andro-genreceptorhasalsobeenshowntoplayacentralrolein

metabolic reprogramingof prostatecancercells.

Andro-gen receptor signaling upregulates the expression of

Dynaminrelatedprotein1(DRP1)inthemitochondria,

leading to increased pyruvate transport into the

mito-chondriaandresultinginaugmentedOXPHOSand

lipo-genesis [90]. Moreover, high expression of DRP1 in

(16)

prognosis[90].Anothermolecularplayerinvolvedinthe

promotionofprostatecancerprogressioninconnectionto

androgen receptor (AR) signaling is Sirtuin7 (SIRT7)

[91]. SIRT7 depletion significantly reduces prostate

cancercellproliferationandandrogen-inducedautophagy

invitroandinvivo.Furthermore,upregulationofSIRT7

correlateswiththeexpressionof ARaswellas

prostate-specific antigen (PSA)and couldthereforebeusedas a

prognosticmarkerforprostatecancerinadditiontobeing

apotentialdrugtarget.

Tcellmetabolisminprostatecancer

Hypoxiainprostatetumorshasbeenassociatedwithpoor

prognosis[92].Tcellsarelargelyexcludedfromhypoxic

zoneswithinthetumorthereforedrugsreversinghypoxia

represent a promising treatment strategy to increase T

cellinfiltrationandimprove theirfitness intheprostate

TME. In fact, hypoxia-activated prodrug TH-302 was

showntoeliminatehypoxiainthetransgenic

adenocarci-nomaofthemouseprostate(TRAMP)derived

TRAMP-C2 mouse model [93]. Additionally, treatment of

TRAMP-C2 mice with a combination of both TH-302

andimmunecheckpointblockade(a-CTLA-4and

a-PD-1) actssynergisticallytocure over80% ofthetumors.

Indoleamine 2,3-dioxygenase(IDO)is therate limiting

enzyme of themetabolic reactionin the metabolismof

tryptophan, which is known to be essential for T cell

activation. Excessive expression of IDO in cancercells

leadstothedepletionoftryptophanandtheaccumulation

of associatedmetabolites,which areknownto inhibitT

cellexpansionandactivationaswellasaidinthe

recruit-mentofTregcellsandothersuppressivecelltypes[94].

IDOexpressionwasshowntocorrelatewithhigherstages

of prostate cancer progression and was indicative of

resistance to immunotherapy [95]. Furthermore, IDO

expression is higher after anti-tumoral vaccine or

a-PD-1treatment.Takentogethertheseresults suggest

thatIDOexpressioncouldbeamechanismofresistance

to immunotherapy, therefore future research including

IDO inhibitors in combination with immunotherapy

should be explored. Considering that IDO inhibition

has previously been studied as a potential addition to

immunotherapy in differentcancertypeswithoutmuch

success [96], it is perhaps worth exploring therapeutic

effect of IDO inhibition in combination with different

typesoftherapies,potentiallyothermetabolictargetsin

addition toimmunotherapy.

Several clinical trials are exploring the possibility of

immunotherapy for the treatment of prostate cancer.

The firstcancer vaccinefor thetreatment of metastatic

castrate resistant prostate cancer was approved by the

FDA in 2010 and has been shown to prolong overall

survivalbyanaverageof4months[97].Numerousother

cancer vaccines and different immunotherapeutic

approaches have been tested but ultimately none have

ledto adramatic improvementin prognosisfor patients

with advancedprostatecancer[98].Therefore,research

intonewandimprovedtherapeuticoptionsareurgently

needed. Metabolic targets in both cancer and immune

cells representapromisingavenuetoexplore.

Conclusion

The ideaoftargetingmetabolismin cancertherapyhas

been around for a long time. In fact, many standard

chemotherapy drugs have metabolic targets. However,

systemicadministrationofmanyofthesemetabolicdrugs

comewithsignificantofftargettoxicitysincetheytarget

pathways that all healthy and cancer cells require for

proliferation(suchasnucleotidesynthesis).Overthepast

decadeimmunotherapyhasestablisheditselfasa

prom-isingtreatmentoptionforcancerpatients.However,ithas

becomeclearthatonitsown,immunotherapyintheform

ofICIandACTisnotsuccessfulformostpatients.The

majorlimitationoftheseapproachesisthefactthatTcell

(and other anti-tumor immune cell) fitness is generally

compromised within the harsh TME. Moreover, the

abilityofimmunotherapytounleashtheimmunesystem

against the tumor comes with the draw back to elicit

systemicimmune reactions. Forexample, patients

trea-tedwithICImayexperiencecolitis,hepatitis,

myocardi-tis, pneumonitis or neurotoxic effects all due to the

excessively activated immune system [99]. In an

esti-mated0.3–1.3%ofcasesthesecomplicationscanbefatal.

CancercellsdepletecrucialnutrientsfromtheTMEand

secretefactorsthatactivelyinhibit effectorTcell

func-tion.Asdiscussedinthisreview,themetabolicfeaturesof

the TME can vary significantly between the different

tissuesoforiginandevenbetweencancersubtypeswithin

thesametissue.Consequently,Tcellresponsein

differ-entcancersdiffersaswellduetothediverseadaptationof

TcellsandtheirfitnesswithinthespecificTME.

Target-ing specific metabolic factors, in cancer cells and/or T

cells, basedonthecancertissue oforiginand histotype

represents apromisingtreatment optionto complement

immunotherapy. There are several approaches to

con-sider in this respect. Metabolic drugscould be usedin

combination with immunotherapyas an exvivo

precon-ditioning approach in order to induce T cellmetabolic

adaptation to thespecificTMEand increasetheir

infil-trationandfitnesswhichinturnwouldleadtoimproved

therapeuticefficiencyofACT.BytreatingTILsor

engi-neered T cells ex vivo before ACT, off target toxicity

related to metabolic reprograming could be avoided.

Alternatively,drugstargetingmetabolismcouldbeused

as part of the systemic immunotherapeutic regimen,

wherepatientswouldonlyneedtobetreatedforashort

periodoftime,allowingtherewiringoftheTME,

there-fore creating an opportunity for T cells, re-invigorated

through an immunotherapeutic approach, to enter the

TMEandactagainstcancercells.Thiswouldallowfora

more tissue specific response, possibly preventing