Biop-C: A Method for Chromatin Interactome Analysis of Solid Cancer Needle Biopsy Samples

Biop‐C:AMethodforChromatinInteractomeAnalysisofSolidCancerNeedleBiopsy

Samples

SambhaviAnimesh

,RuchiChoudhary

1,2*

,XinYiNg

,JoshuaKaiXunTay

,Wan‐Qin

Chong

,BoonCherGoh

1,3,5

,MelissaJaneFullwood

1,2^



CancerScienceInstituteofSingapore,CentreforTranslationalMedicine,National

UniversityofSingapore,Singapore

SchoolofBiologicalSciences,NanyangTechnologicalUniversity,Singapore

DepartmentofHaematology‐Oncology,NationalUniversityCancerInstituteofSingapore,

NationalUniversityHealthSystem,Singapore

DepartmentofOtolaryngology‐Head&NeckSurgery,NationalUniversityHospital,

Singapore

DepartmentofPharmacology,YongLooLinSchoolofMedicine,NationalUniversityHealth

System,Singapore,Singapore



*Equalcontribution(Co‐firstauthors)

^Correspondingauthor



Abstract

Amajorchallenge in understanding the3Dgenome organization of cancer samples  is the lack ofa

methodadaptedtosolidcancerneedlebiopsysamples.HerewedevelopedBiop‐C,amodifiedinsitu

Hi‐C method, and applied it to characterize three nasopharyngeal cancer patient samples. We

identified Topologically‐Associated

 Domains (TADs), chromatin i

nteraction loops, and Frequently

Interactingregions(FIREs) atkey oncogenes in nasopharyngeal cancerfromBiop‐C heat maps.Our

resultsdemonstratetheutilityofourBiop‐Cmethodininvestigatingthe3Dgenomeorganizationin

solid cancers, and the importance of 3D genome organization in regulating oncogenes

in

nasopharyngealcan

cer.

Keywords‐Hi‐C,Biop‐C,chromatinorganization,3Dgenomeorganization,nasopharyngeal

cancer



.CC-BY-NC 4.0 International licenseperpetuity. It is made available under a

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in

The copyright holder for thisthis version posted January 12, 2021. ; https://doi.org/10.1101/2021.01.11.426176doi: bioRxiv preprint

Background

Thethree‐dimensional(3D)genomeorganization of thenucleusplaysavital role in

the regulation of transcription [1,2]. Alterations in 3D genome organization structures

includingTopologically‐AssociatedDomainboundariesandchromatinloopshavebeenshown

toleadtooncogeneexpressionandcancerprogression[1,3–6].

High‐throughputchromosomeconformationcapturetechnologiessuchasHi‐Chave

beenusedtoinvestigatethree‐dimensional(3D)chromatinconformation[7,8].Thestandard

Hi‐Capproachgenerallyrequiresapproximately1millioncells.Consequently,mostprevious

analyses of cancer samples have been restricted to human cancer cell lines [9–11], but

recently,Hi‐Chasbeenconductedonclinicalsamplesfromliquid cancerssuchasT‐cellAcute

LymphoblasticLeukaemia(T‐ALL)[12]andDiffuseLargeB‐cellLymphoma(DLBCL)  [13]and

onesolidcancer‐gastriccancer [14].Forthesecancers,itispossibletoobtain1millioncells

–forexample,gastriccancerscangrowtoalargesize.However,therearemanycancersfor

whichonlyneedlebiopsiesareavailable[15].

Toallowinterrogationofsampleswithmorelimitedquantitiesofstartingmaterials,

therehavebeenseveraleffortstoreducethenumberofcellsrequiredtojust1Kor500cells

usingmodifiedprotocolssuchas

small‐scaleinsit

uHi‐C(sisHi‐C)[16]andLow‐C[13].However,

when dealing with solid cancers, a second challenge is that the tissue requires special

preparationinordertodissociatethetissueintosingle cellsforHi‐Canalysis.Thecoreneedle

biopsiesposethechallengeofboth

l

imitedcellnumbersaswellastherequirementfortissue

dissociation,whichmightleadtofurtherlossordegradationofchromatinforanalysis.Solid

cancersrepresentapproximately90%ofadulthumancancers[17],therefore,aneasy‐to‐use

methodforpreparingHi‐C librariesfromneedlebiopsycancersamples would advanceour

understanding of how chromatin organization contributes to cancer pathogenesis in solid

cancers.

Here,wepresentBiop‐C,amodifiedinsituHi‐Cmethodforthechromatinanalysisin

solidcancertissuesfromneedlebiopsysamples.TheBiop‐Cmethodhasbeendesignedtobe

usedonsmalltissuesamplesobtainedfromneedlebiopsies.Todemonstratetheutilityof

this method, we analyzed three Nasopharyngeal Cancer (NPC) patient samples. NPC is an

epithelialmalignancy of thenasopharyngealmucosa, and isan aggressive subtypeof head

and neck cancer [18,19,20]. NPCs are further subdivided into three subtypes viz Non‐

keratinizing

 undifferentiated carcinoma, Non‐keratinizing differentiated car

cinoma and 

Keratinizingsquamouscell carcinoma. Depending onthe treatment given, thestageof the

cancer,andthesitewherethecancerpresentsat,NPCscanbesmallandanalysedusingcore

needlebiopsies.

It has been established that NPC has a comparatively low mutational

burden, and

oncogenicityis driven by epigen

eticregulation.Typically, NPC associatedwith Epstein ‐Barr

Virus (EBV) are characterized as having comparatively low DNA mutation rates but

widespread DNA hypermethylation and overexpression or mutation of DNA methylation

enzymes,histonemodificationenzymesandchromatinremodellingenzymes[16,17].Hence,

unravellingthe3Dconformational structure will providefurtherinsightintotheepigenetic

regulatory mechanisms that promulgate the NPC phenotype. Here we obtained a

comprehensive understanding of the 3D genome organization of nasopharyngeal cancer

through Biop‐C analysis, which revealed complex 3D genome organization patterns at

oncogenesimportantinNPC.



.CC-BY-NC 4.0 International licenseperpetuity. It is made available under a

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in

The copyright holder for thisthis version posted January 12, 2021. ; https://doi.org/10.1101/2021.01.11.426176doi: bioRxiv preprint



ResultsandDiscussion



AGenome‐wideMapof3DGenomeOrganizationinNasopharyngealCancer

We applied "Biop‐C" to analyze three NPC tissue samples, i.e., "S009", "S010", and

"S012".Thetumorcoreswerecollectedbyneedlebiopsiesandimmediatelyflash‐frozenin

liquidnitrogen(Fig.S1A).Thetumorcoreswereapproximatelyweighedintherange3‐10mg,

andtheclinicalcharacteristicsofthesamplesarelistedinTableS1.Topreparethetissuefor

analysisof3Dgenomeorganization,weusedaliquidnitrogen‐cooledminimortarandpestle

for the pulverization of tissue within a microtube. This approach kept the biopsy sample

frozenandreducedtheriskofsampledegradation.Additionally,thisapproachalsoprovided

theflexibilityofperformingtheworkflowfromsampleacquisitiontoHiClibrarypreparation

inasinglemicrotube,whichfurtherminimizedpotentialsamplelosses.

Afterpulverization,weprocessedthesampleswithacommerciallyavailableArimain

situHi‐Ckit.Briefly,thechromatinwasfixedwithformaldehydeinthenucleusanddigested

with a restriction enzyme. Then overhangs were filled in with biotinylated nucleotides

followed by proximity ligation. After ligation, crosslinks were reversed, and the DNA was

purifiedfromprotein.Further,thepurifiedDNAwasshearedto~350bpmeanfragmentsize.

Finally, the sequencing libraries were generated using low input swift bioscience Illumina‐

compatibleadapters(Fig.1A).TheusageoftheminimortarandpestlefollowedbyHi‐Cisthe

keyinnovationoftheBiop‐Cmethod.Whilethisimprovementinsampleprocessingisasmall

change,it is highlyeffectivein generating high‐qualitychromatininteraction libraries from

needlebiopsyclinicalsamples(Fig.S1B).

Finally,wesequencedBiop‐ClibrariesdeeplybyIlluminanext‐generationsequencing

usingaHiSeq4000machine.Eachlibrarycontainedbetween450Mand922Mcontacts(Table

S2).We obtainedmore than 200million mapped/validjunction reads(>50% oftotal read

pairs)foreachlibrary,reflectingthatourBiop‐Cdatasetsareadequatelycomplex[23](Fig.

1B,TableS2).Furthermore,thelowratiosofthenumberoftranstociscontactsindicatehigh

libraryqualityforallsamples[24](TableS2).Notably,insomesamples(e.g.,S009),onlyone

laneofHiSeq4000sequencingwassufficienttoobtainahighqualityBiop‐Clibrary.

We used Juicer for processing the resulting data, and the package Arrowhead was

usedtoannotateTopologically‐AssociatedDomains(TADs)genome‐wide,whilethepackage

HiCCUPSwasusedtocallloops[9].HeatmapswerevisualizedwithJuicebox [25].Wewere

abletosuccessfullycallTADsandloopsfromourlibraries(Fig.1B,TableS3&S4).TheTADs

wereclearlyidentifiableataresolutionof10kb.PatientS009had1309TADs,whilepatient

S010had1453TADsandpatientS012had1516TADs(Fig.1B).Loopscouldbeidentifiedat 

5kb,10kb,and25kbresolutionsinalldatasets,andtheseloopswereallmergedtogetherfor

subsequent analyses. Patient S009 had 4730 merged loops, while patient S010 had 6539

mergedloops,andpatientS012had2546mergedloops.

Moreover, because Frequently Interacting Regions ("FIREs") are a new type of

chromatin interaction landmark associated with super enhancers and tissue specific

chromatininteractions[40],weusedFIREcallerRpackage[45]tocallFrequentlyinteracting

regions(FIREs).Weidentified2783FIREsinNPCsampleS009,1393FIREsinsampleS010and

2906FIREsinsampleS012fromourBiop‐Cdata.WealsocalledFIREsfromNPCcelllineHK1

andidentified3585FIREs(Fig.1B).

.CC-BY-NC 4.0 International licenseperpetuity. It is made available under a

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in

The copyright holder for thisthis version posted January 12, 2021. ; https://doi.org/10.1101/2021.01.11.426176doi: bioRxiv preprint

Next,weexaminedthechromatininteractionsaroundimportantoncogenesinNPC,

such as MYC [26] and epidermal growth factor receptor (EGFR)[27–29] . The c‐Myc is

overexpressed in 76% ofthe NPC patients.The patients with c‐Myc positive tumors had a

longerdisease‐freeperiod[26].EGFRishighlyexpressedinnearly85%ofNPCpatientsand

associated with a significantly poorer prognosis in patients with advanced nasopharyngeal

cancerthaninpatientswithoutEGFRoverexpression[28,29].WefoundthatEGFRismarked

bythreesuperenhancersinHK1cellline,twoofthesethreeSE arelocalizedupstream,i.e.,

nearthestarttranscriptionsite,andthethirdoneislocatedinanintron(Fig.S1D).

Finally,forcomparisonwithatypicalHi‐Clibrary,wegeneratedtworeplicatesofHK1

NPC cell line by traditional in situ Hi‐C (Arima kit, [9,30]). Visual inspection of coverage

normalized Biop‐C heat maps of three NPC tissue samples, and Hi‐C maps of HK1 cell line

showed that the libraries were largely similar to each other, although we also noted that

certainlocicontaineddifferencessuggestingpatientheterogeneitywhichweexplorefurther

in the next section of this manuscript (Fig 1C,1F,1G, S1C, S1I‐J). Overall, our successful

detectionofTADs,loopsandFIRESsuggestthatourBiop‐Cmethodcangeneratehigh‐quality

genome‐widechromatinconformationmapsfromthesolidtumorneedlebiopsysamples.



Patientheterogeneityinchromatininteractions

The question of patient heterogeneity is relatively unexplored in chromatin

interactionanalyses.InourpreviousresearchinvestigatingchromatininteractionsattheTP53

and MYC loci, we observed that some chromatin interactions at MYC and TP53 could be

detected in bone marrow andperipheral blood samples but not all chromatin interactions

thatareobservedinK562cellsweredetectedinclinicalsamples[31].

To investigate potential patient heterogeneity, we compared the TADs and Loops

between the patients' Biop‐C heatmaps using the  Jaccard Index.We calculated Jaccard's

similaritycoefficient(JaccardIndex,JI)tomeasuretheoverlapbetweenthecalledTADsand

loops in three  Biop‐C matrices. The resulting JI value indicates the fraction of shared TAD

boundariesandloopsbetweenthepatients.Weobservedthat38‐42%oftheTADsand53‐

64%ofloopsaresharedinthethreepatients(Fig.S1L).

Next, we examined individual  specific chromatin interactions. We found the

chromatin interactions for the genomic locations EGFR, PTPN1, DDIT4, MIR205HG, PDGFA,

MALAT1,CAV2,NOTCH1,TEAD1,TP63,RUNX1,CCAT1,MYC,andYAP1(Fig.S1C)aresimilar

andFOXA1,MIPOL1,SP4,SGCZ,MROH9,FMO1,FMO2, FMO1,FMO4,andFMO6Pgeneare

different(Fig.S1I,J).Inoneexample,weobservedthattwoloopsi.e.loop1andloop2are

present near FOXA1 and MIPOL1 genesin S009 and S012, which are thought tobe tumor

suppressorsinnasopharyngealcancer.However,theseloopswereabsentinS010(Fig1F,G).

InanotherexampleweobservedalooponlyinS009andabsentinS010andS012nearmiR383,

whichisconsideredanexcellentdiagnosticbiomarker forheadandneckcancer[35](Fig.S1J).

Next,aswehadobservedpatient ‐specificchromatininteractions,weinvestigatedthe

tissue‐specificityofthesechromatininteractions.Thus,wecharacterizedthesimilaritiesand

thedifferencesbetweentheNPClandscapeandothertissuetypes.WecomparedtheBiop‐C

heatmapsandtheHi‐Cheatmapfrom HK1withthepreviously published Hi‐C heatmaps in

humancelllinessuchasK562(chronicmyelogenousleukemiacellline),HAP1(near‐haploid

cell line), IMR90 (fetal lung fibroblast cell line), KBM7 (chronic myelogenous leukemia),

HUVEC(humanumbilicalveinendothelialcellline),RPE1(retinalpigmentepitheliumcellline),

GM12878 (lymphoblastoid cell line), NHEK (normal human epidermal keratinocytes), HeLa

.CC-BY-NC 4.0 International licenseperpetuity. It is made available under a

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in

The copyright holder for thisthis version posted January 12, 2021. ; https://doi.org/10.1101/2021.01.11.426176doi: bioRxiv preprint

(human cervical carcinoma cell line), HCT116 (colon cancer cell line), HMEC (mammary

epithelialcelllineatarepresentativefourgenomicregions(Fig.S1E‐H).

WeobservedthattheHiCheatmapsofK562,HAP1,IMR90,KBM7,HUVEC,andRPE1

showasimilarpatternastheBiop‐CheatmapsofS009,S010,S012,andHK1Hi‐Cheatmap

for the genomic locations around MYC and CCAT1, but differences in the profile were

observedinGM12878,NHEK,HeLa,HCT116,andHMEC(Fig.S1E).Thegenomeorganization

atRUNX1locus was sim ilar  toourpatient's Biop‐C heatmapsin K562,HAP1,IMR90, RPE1,

NHEK1, HeLa, HCT116,and HMECbut different in KBM7, HUVEC, and GM12878 (Fig. S1F).

However,thegenomicpatternsforthePTPN1locuswerefoundtobe similarinalltheHi‐C

andpatient'sBiop‐Cheatmaps(Fig.S1G).Ontheotherhand,thegenomicpatternsforthe

MALAT1locusdidnotshowanysimilarityandweredifferentinalltheHi‐C,andBiop‐Cheat

maps(Fig.S1H).Weconcludethatcertainchromatininteractionsinnasopharyngealcancer

arecommonacrosstissuetypes(PTPN1)andcertainregionsaretissuespecific(MALAT1).

The observation of patient heterogeneity  and tissue specificity in TADs appears to

contradictearlierobservationsthatTADsareprimarilyconservedacrossdifferenthumancell

typesandpossiblyevenacrossdifferentspecies[9,30,36].However,Sauerwaldanalyzed137

Hi‐C samples from 9 studies and observed

significant TAD variations across human cell and

tissuetypes[37]

,suggestingthatwhilethere arecommonTADsandloops,therearealsoTADs

andloopsthatvaryacrosspatientsamplesandtissuetypes.



SuperenhancersareassociatedwithFrequentlyInteractingRegionsandlooptogenes

Super enhancers (SE) are regions of the DNA which enhances the transcription of

targetgenes.Thesearecomprisedofgroupofenhancerswhichareatcloseproximitytoeach

otherandaremarkedbyhighenrichmentofH3K27achistonemodification[38].Inprevious

research,weandothershaveshownthatsuperenhancerscanregulatedistantgenesvialong‐

rangechromatininteractions[31,39].Moreover,frequentlyinteractingregions(FIREs)have

been reported to form at genomic regions also enriched by super enhancers [40].

Consequently, we wished to understand the relationship between super enhancers and

chromatininteractionsinnasopharyngealcancer.

Asthebiopsysamples

weretoosmallforustoobtainbothH3K27acChIP‐Seqdataas

wellasBiop‐Cdata,wei

dentifiedSEfromNPCcelllinesHK1,C666‐1andHNE1usingtheChIP‐

Seqdatafrom[41]andwefoundthat298SEwerecommoninallthethreecelllines(Table

S5). We reasoned that these  "common" su per enhancers that are present in all cell lines

examined will most likely  also be present in the patient samples examined.We then used

these common super enhancers andassociated them withchromatin loopsobtained from

theBiop‐Cdataofthepatient

samplesaswellasfro

mtheHiCdataofHK1cellline.Asaresult,

wefoundthattheseSE arehighlyassociatedwithchromatinloops.InsampleS009,S010and

S012,theassociationofSEwithchromatinloopswere54%,57%and41%respectively(Fig.

2A).

Wefurtherobservedthatmorethan90%ofthesechromatinloopsareassociatedwith

genesontheothersideofthechromatinloop.InS009,95%ofSEassociatedchromatinloops

arealsolinkedtodistant genes.InS010aswellasS012,91%ofSEassociatedloopsarelinked

togenes(Fig.2B).

Weal

soassociatedtheseSEtochromatinloopsinoneoftheNPCcellline

HK1andfoundthat92%ofcommonsuperenhancersareassociatedwithchromatinloopsin

HK1and91%oftheseloopslinkSEtodistantgenes(Fig.2A,B,TableS7).Forexample,DDIT4

genewhich

isknowntobeover‐expressedinNPCcelllin

es[41]islinkedtoadistantSEin

patientsampleS010,S012andNPCcelllineHK1(Fig.2C).Wealsorepeatedtheaboveanalysis

.CC-BY-NC 4.0 International licenseperpetuity. It is made available under a

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in

The copyright holder for thisthis version posted January 12, 2021. ; https://doi.org/10.1101/2021.01.11.426176doi: bioRxiv preprint

with HK1 super enhancers and found that 85% of HK1 SE were associated with chromatin

loops in HK1 cell line and 94% of these super enhancers were linked to distant genes via

chromatin loops (Supplementary figure S2A, S2B). Next, we categorized SE which were at

closeproximity(lessthan15kbfromagene)toageneandwerecalled"proximal"SEandthe

oneswhichareawayfromthegeneandassociatedviachromatinloopswerecalleddistalSE.

Outof298SEtestedwefound27proximalSEand147distal SEinS009,155distalSEinS010,

110distalSEinS012and260distalSEinHK1(SupplementarytableS7,S8).48geneswere

associatedwithproximalSE(SupplementarytableS8)while356genesinS009,421genesin 

S010,291inS012and944genesinHK1wereassociatedwithdistalSE(Supplementarytable

S7).TherewerealsosomegenesthathavebothproximalanddistalSE:InS009wefound10

genes;inS010aswellasS012wefound6genesandinHK1wefound18such genes(Table

S8).Forexample,MACF1geneinallthe3NPCsamplesaswellasHK1celllinehasdistalas

wellasproximalSE(TableS8).WeconcludethatSEarehighlyassociatedwithchromatinloops

todistalgenesinNPC.

Subsequently,wewantedtoassociatetheseSEwithFIREscalled(TableS6)fromthe

Biop‐CdatafrompatientsamplesaswellasHK1HiCdata.Wecouldrecognize24common

SEinS009,26inS010,31inS012and56inHK1whicharewithinaFIRE.Uponcombiningthe

chromatinloopsdatawithFIREcalling,wefound23SEinS009,32insampleS010,30inS012

and69SEinHK1celllinethatloopstoaFIRE(Fig.2D,TableS9).WealsofoundSE whichfall

withinthesameFIREasanoncogenelikePTPN1whichisalsoknowntobeover‐expressedin

NPCcelllines[41](Fig.2E)aswellasSEwhichloopstoadistantFIREcontainingARF6gene

whoseover‐expressioncanbecorrelatedwithmetastasisandinvasioninseveralcancers[42].

WeconcludethatSEareassociatedwithFIREsinNPC.

Taken together, our new Biop‐C method is suitable for interrogating needle biopsy

patient samples, and more generally, situations of limited tumor sampling when surgical

biopsiesmaybetechnicallydifficult.UsingBiop‐C,weexaminedchromatininteractionsin3

nasopharyngeal cancer patient samples, which allowed us to identify super  enhancers

associatedwithFIREsandwhichlooptoimportantoncogenes.Wealsodemonstratedpatient

heterogeneityinchromatininteractionsinthesepatientsamples,aswellastissuespecificity.

Hi‐C libraries from an NPC cell line, HK1, showed differences compared with chromatin

interactionsinthepatientsamples.Thesedifferencescouldariseduetodifferentsubtypeof

NPC. Our results indicate the necessity of interrogating chromatin interactions in actual

patientcancers,whichBiop‐Cenables.Ourresultsdemonstratetheutilityandimportanceof

Biop‐C as a method for understanding cancer 3D genome organization. In the future, we

anticipate that the versatility of Biop‐C will also allow us to interrogate perturbations of

chromatingeneregulationinpatientsundergoingtherapeuticinterventions.

Methods

DetailedmethodsaregivenintheSupplementaryMaterials.Briefly,thebiopsysampleswere

obtainedfromNPCpatientsusingan18gaugeneedleandflashfrozen.Thefrozensamples

were subjected to the Biop‐C method for the generation of the Biop‐C matrix. HK1 were

culturedat5%CO2at37°CinRoswellParkMemorialInstitute(RPMI)1640media(Hyclone)

supplemented with 10% heat‐inactivated Fetal Bovine Serum (FBS; Hyclone) and 1%

penicillin/streptomycin(Hyclone).TheHi‐CheatmapsofHK1werepreparedusingArimaHiC

kitandsequenced150basespaired‐endontheIlluminaHiSeq4000.Hi‐Cdatawerealigned

andprocessedbyJuicer(version1.5.7)[9].Thereferencegenomewashg19.Theheatmaps

arevisualizedusingJuicebox[25].



.CC-BY-NC 4.0 International licenseperpetuity. It is made available under a

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in

The copyright holder for thisthis version posted January 12, 2021. ; https://doi.org/10.1101/2021.01.11.426176doi: bioRxiv preprint

Datadeposition

WeareintheprocessofsubmittingourdatatoGEO,dbgapandcontrolledshortreadarchive.

Dataavailability

Allrelevantdatasupportingthekeyfindingsofthisstudyareavailablewithin thearticle and

itsSupplementaryInformationfilesorfromthecorrespondingauthoronreasonablerequest.



Acknowledgements

ThisresearchissupportedbytheNationalResearchFoundation(NRF)Singaporethroughan

NRFFellowshipawardedto M.J.F(NRF‐NRFF2012‐054)and NTUstart‐upfundsawardedto

M.J.F.ThisresearchissupportedbytheRNABiologyCenterattheCancerScienceInstituteof

Singapore, NUS, as part of funding under the Singapore Ministry of Education Academic

Research Fund Tier 3 awarded to Daniel Tenen (MOE2014‐T3‐1‐006). This research is

supported by the National Research Foundation Singapore and the Singapore Ministry of

EducationunderitsResearchCentresofExcellenceinitiative.



Authorcontributions

S.A. and M.J.F conceived the research idea. M.J.F, S.A., and R.C. contributed to the study

design.S.A. and R.C.performedbioinformaticsanalysis of theHi‐C.S.A.performedmanual

curationoftheHi‐Cdata.R.CperformedFIREcallingandChIP‐Seqanalysisonpublisheddata

toidentifyS.E.sandtheirloopingpatternsinNPC.X.Y,J.T.,W‐Q.CandB.C.G.providedNPC.

clinicalsamples.S.A.,R.C,andM.J.Freviewedthedata.S.A.,R.C,B.C.G.andM.J.Fwrotethe

manuscript.Allauthorsreviewedandapprovedthemanuscript.



Competinginterests

M.J.FdeclarestwopatentsonmethodologiesrelatedtoChIA‐PET.Nootherconflictsof

interestaredeclared.



References:

1.HniszD,WeintrauAS,DayDS,ValtonAL,BakRO,LiCH,etal.Activationofproto‐

oncogenesbydisruptionofchromosomeneighborhoods.Science.2016;351:1454–8.

2.HniszD,SchuijersJ,LiCH,YoungRA.RegulationandDysregulationofChromosome

StructureinCancer.Annu.Rev.CancerBiol.2018;2:21–40.



3.LiL,BarthNKH,PilarskyC,TaherL.Cancerisassociatedwithalterationsinthethree‐

dimensionalorganizationofthegen

ome.Cancers.2019;11.

4.ValtonAL,DekkerJ.TADdisruptionasoncogenicdriver.Curr.Opin.Genet.Dev.Elsevier

Ltd;2016.p.34–40.

5.FlavahanWA,DrierY,LiauBB,Gillespie

SM,Venteich

erAS,Stemmer‐RachamimovAO,et

al.InsulatordysfunctionandoncogeneactivationinIDHmutantgliomas.Nature.

2016;529:110–4.

6.FudenbergG,GetzG,MeyersonM,MirnyLA.Highorderchromatinarchitectureshapes

thelandscapeofchromosomalalterationsincancer.Nat.Biotechnol.2011;29:1109–13.

7.Lieberman‐AidenE,VanBerkumNL,WilliamsL,ImakaevM,RagoczyT,TellingA,etal.

Comprehensivemappingoflong‐rangeinteractionsrevealsfoldingprinciplesofthehuman

genome.Science.2009;326:289–93.

8.vanBerkumNL,Lieberman‐AidenE,WilliamsL,ImakaevM,GnirkeA,MirnyLA,etal.Hi‐

C:Amethodtostudythethree‐dimensionalarchitectureofgenomes.J.Vis.Exp.

2010;39:1869.

.CC-BY-NC 4.0 International licenseperpetuity. It is made available under a

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in

The copyright holder for thisthis version posted January 12, 2021. ; https://doi.org/10.1101/2021.01.11.426176doi: bioRxiv preprint

9.RaoSSP,HuntleyMH,DurandNC,StamenovaEK,BochkovID,RobinsonJT,etal.A3D

mapofthehumangenomeatkilobaseresolutionrevealsprinciplesofchromatinlooping.

Cell2014;159:1665–80.

10.HaarhuisJHI,vanderWeideRH,BlomenVA,Yáñez‐CunaJO,AmendolaM,vanRuiten

MS,etal.TheCohesinReleaseFactorWAPLRestrictsChromatinLoopExtension.Cell.

2017;169:693‐707.

11.DarrowEM,HuntleyMH,DudchenkoO,StamenovaEK,DurandNC,SunZ,etal.Deletion

ofDXZ4onthehumaninactiveXchromosomealtershigher‐ordergenomearchitecture.

Proc.Natl.Acad.Sci.U.S.A.2016;113:E4504–12.

12.KloetgenA,ThandapaniP,NtziachristosP,GhebrechristosY,NomikouS,LazarisC,etal.

Three‐dimensionalchromatinlandscapesinTcellacutelymphoblasticleukemia.Nat.Genet.

2020;52:388–400.

13.DíazN,KruseK,ErdmannT,StaigerAM,OttG,LenzG,etal.Chromatinconformation

analysisofprimarypatienttissueusingalowinputHi‐Cmethod.Nat.Commun.2018;9:1–

13.

14.OoiWF,NargundAM,LimKJ,ZhangS,XingM,MandoliA,etal.Integratedpaired‐end

enhancerprofilingandwhole‐genomesequencingrevealsrecurrentCCNE1andIGF2

enhancerhijackinginprimarygastricadenocarcinoma.Gut.2020.;69:1039–52.

15.LinF,ZhangJ,LiuH.Handbookofpracticalfineneedleaspirationandsmalltissue

biopsies.Handb.Pract.FineNeedleAspirationSmallTissueBiopsies.2017;1–496.

16.DuZ,ZhengH,HuangB,MaR,WuJ,ZhangX,etal.Allelicreprogrammingof3D

chromatinarchitectureduringearlymammaliandevelopment.Nature.2017;547:232–5.

17.SocietyAC.AmericanCancerSociety.CancerFacts&Figures2020.Am.CancerSoc.

2020;1–52.

18.YuMC,YuanJM.Epidemiologyofnasopharyngealcarcinoma.Semin.Cancer

Biol.2002;421–9.

19.DaiW,ZhengH,CheungAKL,LungML.Geneticandepigeneticlandscapeof

nasopharyngealcarcinoma.ChineseClin.Oncol.2016;5:4–4.

20.LiM,CheungAKL,YeeKoJM,LungHL,ChengY,DaiW.Theinterplayofhostgenetic

factorsandEpstein‐Barrvirusinthedevelopmentofnasopharyngealcarcinoma.Chin.J.

Cancer.2014;566–8.

21.TsangCM,LuiVWY,BruceJP,PughTJ,LoKW.Translationalgenomicsofnasopharyngeal

cancer.Semin.CancerBiol.2020;61:84–100.

22.DaiW,CheungAKL,

KoJMY,ChengY,ZhengH,NganRKC

,etal.Comparativemethylome

analysisinsolidtumorsrevealsaberrantmethylationatchromosome6pinnasopharyngeal

carcinoma.CancerMed.2015;4:1079–90.

23.LajoieBR,DekkerJ,KaplanN.TheHitchhiker’sguidetoHi‐Canalysis:Practical

guidelines.Methods.2015;72:65–75.

24.NaganoT,VárnaiC,SchoenfelderS,

JavierreBM,Winget

tSW,FraserP.Comparisonof

Hi‐Cresultsusingin‐solutionversusin‐nucleusligation.GenomeBiol.2015;16:175.

25.DurandNC,RobinsonJT,ShamimMS,MacholI,MesirovJP,LanderES,etal.Juicebox

ProvidesaVisualizationSystemforHi‐CContactMapswithUnlimitedZoom.Cell

Syst.2016;3:99–101.

26.YuY,DongW,LiX,YuE,ZhouX,LiS.Significanceofc‐MycandBcl‐2ProteinExpression

inNasopharyngealCarcinoma.Arch.Otolaryngol.‐HeadNeckSurg.2003;129:1322–6.

27.RuanL,LiXH,WanXX,YiH,LiC,LiMY,etal.AnalysisofEGFRsignalingpathwayin

nasopharyngealcarcinomacellsbyquantitativephosphoproteomics.Proteome

.CC-BY-NC 4.0 International licenseperpetuity. It is made available under a

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in

The copyright holder for thisthis version posted January 12, 2021. ; https://doi.org/10.1101/2021.01.11.426176doi: bioRxiv preprint

Sci.2011;9:35.

28.ChuaDTT,NichollsJM,ShamJST,AuGKH.Prognosticvalueofepidermalgrowthfactor

receptorexpressioninpatientswithadvancedstagenasopharyngealcarcinomatreatedwith

inductionchemotherapyandradiotherapy.Int.J.Radiat.Oncol.Biol.Phys.;2004;59:11–20.

29.FujiiM,YamashitaT,IshiguroR,TashiroM,KameyamaK.Significanceofepidermal

growthfactorreceptorandtumorassociatedtissueeosinophiliaintheprognosisofpatients

withnasopharyngealcarcinoma.AurisNasusLarynx.2002;29:175–81.

30.DixonJR,SelvarajS,YueF,KimA,LiY,ShenY,etal.Topologicaldomainsinmammalian

genomesidentifiedbyanalysisofchromatininteractions.Nature.2012;485:376–80.

31.CaoF,FangY,TanHK,GohY,ChoyJYH,KohBTH,etal.Super‐enhancersandbroad

h3k4me3domainsformcomplexgeneregulatorycircuitsinvolvingchromatininteractions.

Sci.Rep.2017;7:1–14.

32.Ammous‐BoukhrisN,AyadiW,DerbelM,Allaya‐JaafarN,CharfiS,DaoudJ,etal.FOXA1

ExpressioninNasopharyngealCarcinoma:AssociationwithClinicopathological

CharacteristicsandEMTMarkers.BiomedRes.Int.2020;2020.

33.LeongMML,CheungAKL,KwokTCT,LungML.Functionalcharacterizationofacandidate

tumorsuppressorgene,MirrorImagePolydactyly1,innasopharyngealcarcinoma.Int.J.

Cancer.2020;146:2891–900.

34.KwokA,CheungL,LungHL,MunJ,KoY,ChengY,etal.Chromosome14transferand

functionalstudiesidentifyacandidatetumorsuppressorgene,Mirrorimagepolydactyly1,

innasopharyngealcarcinoma.ProcNatlAcadSciUSA.2009;106(34):14478‐83.

35.LiuC,YuZ,HuangS,ZhaoQ,SunZ,FletcherC,etal.Combinedidentificationofthree

miRNAsinserumaseffectivediagnosticbiomarkersforHNSCC.EBioMedicine.2019;50:135–

43.

36.JinF,LiY,DixonJR,SelvarajS,YeZ,LeeAY,etal.Ahigh‐resolutionmapofthethree‐

dimensionalchromatininteractomeinhumancells.Nature.2013;503:290–4.

37.SauerwaldN,SinghalA,KingsfordC.Analysisofthestructuralvariabilityoftopologically

associateddomainsasrevealedbyHi‐C.NARGenomicsBioinforma.2020;2:1–13.

38.PottS,LiebJD.Whataresuper‐enhancers?Nat.Genet.2015.8–12.

39.SeeYX,WangBZ,FullwoodMJ.ChromatinInteractionsandRegulatoryElementsin

Cancer:FromBenchtoBedside.TrendsGenet.2019;35:145–58.

40.SchmittAD,HuM,JungI,XuZ,QiuY,TanCL,etal.ACompendiumofChromatinContact

MapsRevealsSpatiallyActiveRegionsintheHumanGenome.CellRep.2016;17:2042–59.

41.KeL,ZhouH,WangC,XiongG,XiangY,LingY,etal.Super‐enhancerspromote

transcriptionaldysregulationinnasopharyngealcarcinoma.Proc.Natl.Acad.Sci.U.S.A.

2017;77:6614–26.

42.LiR,PengC,ZhangX,WuY,PanS,XiaoY.RolesofArf6incancercellinvasion,

metastasisandproliferation.LifeSci.2017;1;182:80‐84.

43.KeL,ZhouH,WangC,XiongG,XiangY,LingY,etal.Nasopharyngealcarcinomasuper‐

enhancer–drivenETV6correlateswithprognosis.Proc.Natl.Acad.Sci.U.S.A.

2017;114:9683–8.

44.KentWJ,SugnetCW,FureyTS,RoskinKM,PringleTH,ZahlerAM,etal.TheHuman

GenomeBrowseratUCSC.GenomeRes.2002;12:996–1006.

45.Crowley,C.,Y.Yang,Y.Qiu,B.Hu,H.Won,B.Ren,M.HuandY.Li(2019)."FIREcaller:an

RpackagefordetectingfrequentlyinteractingregionsfromHi‐Cdata."bioRxiv:619288.



.CC-BY-NC 4.0 International licenseperpetuity. It is made available under a

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in

The copyright holder for thisthis version posted January 12, 2021. ; https://doi.org/10.1101/2021.01.11.426176doi: bioRxiv preprint



Fig1



.CC-BY-NC 4.0 International licenseperpetuity. It is made available under a

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in

The copyright holder for thisthis version posted January 12, 2021. ; https://doi.org/10.1101/2021.01.11.426176doi: bioRxiv preprint



Fig 1: Biop ‐C method enables the  study of high‐resolution 3D Genome organization and

chromatininteractomeinsolidtumorsamples(A)SchematicoverviewoftheBiop‐Cmethod.

Thefrozenneedlebiopsytissueispulverizedinaliquidnitrogen‐cooledminimortar,andthen

chromatin was fixed with formaldehyde in the nucleus. The samples were processed with

commercial Arima genomics HiC Kit. Fixed cells were permeabilized using  a lysis buffer

suppliedinArimaHiCkitandthen digestedwitharestrictionenzyme.Followingrestriction

digestion,thesiteswerefilledinwithbiotinylatednucleotides.TheresultingDNAbluntends

weresubsequentlyligated.Afterligation,crosslinkswerereversedtoremoveproteinsfrom

DNA. Hi‐C material was then sonicated using a Covaris Focused‐Ultrasonicator M220

instrument to achieve fragment sizes of 300‐500 bp. The sonicated DNA was double‐size

selectedusingAmpureXPbeads,andthesequencinglibrariesaregeneratedusinglowinput

SwiftBioscienceIllumina‐compatibleadapters(B)TableofstatisticsofBiop‐CandHi‐Cdata.

#Hi‐C contacts indicates number of mapped/valid junction reads of each library. #TAD

indicates

the number of TADs called from Biop‐C and Hi‐C data at 10kb resolution. #Loops

indicatesthetotalnumberofloopscalledfromBiop‐CandHi‐Cdataat5kb,10kb,and25kb

resolutionandthenmerged.#FiREsindicatesthenumberofFIREscalledfromBiop

‐CandHi‐

Cdataat10kbresolution.(C)Biop‐CandHi‐Che

atmapshowing3Dgenomeorganisationat

CCAT1 andMYC gene.

 Genes are indicated in blackcolor. The “common superenhancers”,

showninbluecolor,indicatethesuperenhancerspresentinallthreeNPCcelllines‐HK1,C66‐

1,andHNE1celllines.ThesuperenhancerspresentinHK1celllinesareindicatedinredcolor.

ThesuperenhancerdatasetsareobtainedfromKeetal.2017[43](D)Venndiagramshowing

theoverlapofTADboundariesbetweenthepatientsinferredusingtheArrowheadalgorithm

[9]at10kbresolution.(E)VenndiagramdepictingtheoverlapofLoopsbetweenthepatients.

TheloopswerecalledusingtheHiCCUPsalgorithm[9]at5kb,10Kb,and25Kbresolution.The

loopsofthedifferentresolutionsweremergedforthisanalysis.(F)ZoomedviewofBiop‐C

Juicebox‐visualized heat map of S009, S010, and S012 showing patient heterogeneity in

chromatin interaction at FOXA1 and MIPOL1 gene. (G) UCSC genome browser [44]

screenshotsofgenomiccoordinateschr14:36,459,966‐42,220,035.Loops1and2arepresent

inS009&S012,whileloop3ispresentonlyin S010.Loop1,Loop2,andLoop3arerepresented

ingreen,blue,andblackcolors,respectively.







.CC-BY-NC 4.0 International licenseperpetuity. It is made available under a

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in

The copyright holder for thisthis version posted January 12, 2021. ; https://doi.org/10.1101/2021.01.11.426176doi: bioRxiv preprint



Fig2



.CC-BY-NC 4.0 International licenseperpetuity. It is made available under a

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in

The copyright holder for thisthis version posted January 12, 2021. ; https://doi.org/10.1101/2021.01.11.426176doi: bioRxiv preprint



Fig2:Associationofsuperenhancerswithchromatininteractionsandgenes.(A)Graphshowingthe 

numberofcommonSEassociatedwithchromatininteractionsinNPCpatientsampleS009,S010,S012

and cell line HK1 (blue) and number of SE which are not associated with chromatin interactions

(orange)(B) graph showing number

of SEassociated with distant genes via chr

omatin loops in NPC

patientsampleS009,S010,S012andcelllineHK1(darkblue)andnumberofSEwhichdonotlinkto

distantgenesviachromatinloops  (lightblue)(C)Biop‐CheatmapsforNPCsampleS009,S010,S012

andHic heatmap for

cell li

neHK1 fortheDDIT4 ge ne locus where aSElinkstothegenethrougha

chromatinloop(D)graphshowingthenumberofSEassociatedwithFIREsviachromatinloop(green)

andSEwhicharewithinFIRE(yellow)(E)Biop‐Cheatmaps forNPCsampleS009,S010,S012

andHic

heatmapforcelllineHK

1showingaSE(blue:comm onSE,red:HK1SE)andPTPN1genewithinthe

sameFIRE(F)Biop‐CheatmapsforNPCsampleS009,S010,S012andHicheatmapforcelllineHK1

showingaSEloopingtoFIREattheARF6genelocus.



.CC-BY-NC 4.0 International licenseperpetuity. It is made available under a

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in

The copyright holder for thisthis version posted January 12, 2021. ; https://doi.org/10.1101/2021.01.11.426176doi: bioRxiv preprint