TNFα-mediated activation of NF-κB downregulates sodium-iodide symporter expression in thyroid cells

RESEARCH ARTICLE TNFα-media ed ac i a ion o NF-κB down egula es sodium-iodide sympo e exp ession in hy oid cells Ma ´ cia Fa ia 1,2,3☯ , Ri a Domingues 1,4☯ , F ancisca Paixão 1,4 , Ma ia João Bugalho 1,5 , Paulo Ma os 2,3 , Ana Luı ´sa Sil aID 1,4 * 1Se ic¸o de Endoc inologia, Diabe es e Me abolismo do CHULN-Hospi al San a Ma ia, Lisboa, Po ugal, 2BioISI-Biosys ems and In eg a i e Sciences Ins i u e, Faculdade de Ciências da Uni e sidade de Lisboa, Lisboa, Po ugal, 3Depa amen o de Gene ´ ica Humana, Ins i u o Nacional de Sau ´de Dou o Rica do Jo ge, Lisboa, Po ugal, 4ISAMB-Ins i u o de Sau ´de Ambien al, Faculdade de Medicina da Uni e sidade de Lisboa, Lisboa, Po ugal, 5Faculdade de Medicina da Uni e sidade de Lisboa, Lisboa, Po ugal ☯These au ho s con ibu ed equally o his wo k. *sil a. [email protected] Abs ac The sodium-iodide sympo e (NIS) media es anspo o iodide ac oss he basola e al memb ane o hy oid cells. NIS exp ession in hy oid cance (TC) cells allows he use o adioac i e iodine (RAI) as a diagnos ic and he apeu ic ool, being RAI he apy he sys emic ea men o choice o me as a ic disease. S ill, a signi ican p opo ion o pa ien s wi h ad anced TC lose he abili y o espond o RAI he apy and no e ec i e al e na i e he apies a e a ailable. De ec i e NIS exp ession is he main eason o impai ed iodide up ake in TC and NIS down egula ion has been associa ed wi h se e al pa hways linked o malignan ans o ma ion. NF-κB signaling is one o he pa hways associa ed wi h TC. In e es ingly, NIS exp ession can be nega i ely egula ed by TNF-α, a bona ide ac i a o o NF-κB wi h a cen al ole in hy oid au oimmuni y. This p omp ed us o cla i y NF-kB’s ole in his p o- cess. We con i med ha TNF-αleads o down egula ion o TSH-induced NIS exp ession in non-neoplas ic hy oid ollicula cell-de i ed models. No ably, a simila e ec was obse ed when NF-κB ac i a ion was igge ed independen ly o ligand- ecep o speci ici y, using pho bol-my is a e-ace a e (PMA). TNF-αand PMA down egula ion o NIS exp ession was e e ed when NF-κB-dependen ansc ip ion was blocked, demons a ing he equi emen o NF-kB ac i i y. Addi ionally, TNF-αand PMA we e shown o ha e a nega i e impac on TSH-induced iodide up ake, consis en wi h he obse ed ansc ip ional down egula ion o NIS. Ou da a suppo he in ol emen o NF-κB-di ec ed ansc ip ion in he modula ion o NIS exp ession, whe e up- o down- egula ion o NIS depends on he combined ou pu o NF-κB o se e al con e ging pa hways. A be e unde s anding o he mechanisms unde ly- ing NIS exp ession in he con ex o no mal hy oid physiology may guide he de elopmen o pha macological s a egies o inc ease he e iciency o iodide up ake. Such s a egies would be ex emely use ul in imp o ing he esponse o RAI he apy in e ac o y-TC. PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0228794 Feb ua y 12, 2020 1 / 11 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 OPEN ACCESS Ci a ion: Fa ia M, Domingues R, Paixão F, Bugalho MJ, Ma os P, Sil a AL (2020) TNFα-media ed ac i a ion o NF-κB down egula es sodium-iodide sympo e exp ession in hy oid cells. PLoS ONE 15(2): e0228794. h ps://doi.o g/10.1371/jou nal. pone.0228794 Edi o : Sal a o e Papa, Uni e si y o Leeds, Facul y o Medicine and Heal h, UNITED KINGDOM Recei ed: No embe 6, 2019 Accep ed: Janua y 22, 2020 Published: Feb ua y 12, 2020 Copy igh : ©2020 Fa ia e al. This is an open access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal au ho and sou ce a e c edi ed. Da a A ailabili y S a emen : All ele an da a a e wi hin he manusc ip and i s Suppo ing In o ma ion iles. Funding: This wo k was suppo ed by he Fundac¸ão pa a a Ciência e a Tecnologia, g an [PTDC/BIAMOL/31787/2017] om Po ugal. MF is ecipien o FCT ellowship PD/BD/114388/2016. The unde s had no ole in s udy design, da a collec ion and analysis, decision o publish, o p epa a ion o he manusc ip . In oduc ion The well-di e en ia ed hy oid ca cinomas (DTCs) a ise om hy oid ollicula cells and ep- esen he mos equen o ms o hy oid cance (TC), including he papilla y hy oid cance (PTC) and ollicula hy oid cance (FTC) sub ypes [1]. The majo i y o DTCs a e associa ed wi h a a o able p ognosis. Howe e , abou 30% o pa ien s wi h ad anced o ms o DTC become esis an o adioac i e iodine (RAI) he apy, he s anda d ea men o me as a ic disease [2]. The lack o e icien he apeu ic op ions al e na i e o RAI makes he clinical managemen o hese pa ien s challenging, educing he 10-yea su i al a e om app oxima ely 90% o 10% [2,3]. The main eason o impai ed iodide up ake in e ac o y-TC is he de ec i e unc ional exp ession o he sodium iodide sympo e (NIS) [4,5]. NIS belongs o he human solu e ca ie (SLC) amily o anspo e s, is highly exp essed a he basola e al memb ane o hy oid ollicula cells and is esponsible o he ac i e anspo o iodide ac oss he plasma memb ane in o hy oid ollicles [6]. The p ima y egula o o NIS exp ession in hy oid gland is he hy oid s imula ing ho mone (TSH) [7,8]. TSH-induced accumula ion o cyclic AMP leads o he binding o he PAX8 ansc ip ion ac o o he NIS ups eam enhance (NUE) elemen on he NIS gene p o- mo e , a p ima y equi emen o he ull ac i a ion o NIS exp ession [9,10]. Despi e TSH- de i ed signaling being he key egula o o NIS exp ession in hy oid issue, o he signaling pa hways may ha e an impac on his p ocess. NIS exp ession le els and iodine up ake in DTC a e educed when compa ed o no mal issue [11,12] and his down egula ion has been associa ed wi h he o e ac i a ion o se e al pa hways linked o hy oid malignancy [13]. NF-κB signaling has been implica ed in cance -associa ed p ocesses o se e al human malig- nancies, including hy oid cance . Inc eased NF-kB ac i a ion has been desc ibed in PTC, FTC and anaplas ic TC, as being associa ed wi h esis ance o apop osis and main enance o he malignan pheno ype [14–16]. Also, in p e ious s udies, we ha e shown ha o e exp es- sion o umo - ela ed RAC1b, a highly ac i a ed splice a ian o he GTPase RAC1 [17,18], has a signi ican ole in PTC umo igenesis by inducing esis ance o p og ammed cell dea h h ough NF-κB ac i a ion [19]. The NF-κB pa hway is also esponsible o con olling se e al aspec s o cell g ow h and in lamma ion [20]. One majo pa hway esponsible o NF-κB ac i a ion is he canonical NF-κB pa hway, which in ol es p e e en ially he he e odime p65/p50 and is igge ed in esponse o nume ous s imuli including p o-in lamma o y cy okines such as umo -nec osis- ac o -α(TNF-α) and bac e ial lipopolysaccha ide (LPS) [20–22]. The unde s anding o he mechanisms unde lying NIS exp ession in he pe spec i e o no mal hy oid physiology may guide he de elopmen o s a egies o enhance he e iciency o iodide up ake, pa icula ly in he neoplas ic con ex . In ac , in addi ion o i s ole in TC, NF-κB has also been implica ed in no mal hy oid physiology, being necessa y o no mal hy oid s uc u e de elopmen and unc ion, and in ol ed in he exp ession o se e al hy oid speci ic genes, such as PAX8, Tg, TTF-1, TPO and also NIS [23,24]. Indeed, i was shown ha he p65 subuni o NF-κB ac s in syne gy wi h he ansc ip ion ac o PAX8 o he p omo- ion o NIS exp ession, as he esul o LPS s imula ion [25]. Con e sely, despi e being widely ecognized as a po en ac i a o o he canonical NF-κB pa hway [26], TNF-αwas shown o ha e a nega i e impac in NIS exp ession [27]. This is consis en wi h he obse ed educ ion in RAI up ake in DTCs wi h concu en hy oidi is [28], as TNF-αis a cen al ac o in hy- oid au oimmuni y. This led us o aise he ques ion o whe he his nega i e impac o TNF-αon NIS exp es- sion in ol es ac i a ion o NF-κB. TSH-induced NIS exp ession is hampe ed by TNF-α-media ed NF-kB ac i a ion PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0228794 Feb ua y 12, 2020 2 / 11 Compe ing in e es s: The au ho s ha e decla ed ha no compe ing in e es s exis . Ma e ials and me hods Cell lines and cul u e condi ions The TSH- esponsi e cell lines PCCL3 and FRTL5, de i ed om Fische a ’s hy oid ollicula no mal epi helium, we e main ained in Coon’s F-12 modi ied liquid medium (Me ck) supple- men ed wi h 5% e al bo ine se um (FBS) (Gibco), 10 μg/mL o insulin (SIGMA), 5 μg/mL o Apo-T ans e in (Apo-T) (SIGMA) and 0.1 mU/mL o TSH (SIGMA). Cells we e main ained a 37˚C in a 5% humidi ied CO2 en i onmen and disca ded a e 20 passages. When app op i- a e, cells we e cul u ed in s a a ion medium (F12 Coon’s Modi ica ion medium supplemen ed wi h 0.2% ( / ) FBS and 5 μg/mL o Apo-T) o TSH-medium (s a a ion medium supple- men ed wi h 1 mU/mL o TSH). S imula ion wi h TNF-α(10 ng/mL, SIGMA) o pho bol 12-my is a e 13-ace a e (PMA; 10 ng/mL; SIGMA) was pe o med (60 minu es o 24 hou s) in se um-s a ed cells o 72h, subsequen ly subjec ed o TSH s imula ion o 24 hou s (TSH- medium). When applicable, cells we e ea ed wi h he NF-kB selec i e inhibi o BMS-345541 (10 μM, SIGMA) o 6 hou s. RNA ex ac ion, cDNA syn hesis and RT-qPCR To al RNA was ob ained om cul u ed cells using he eady- o-use eagen T ipleX ac o (GRiSP Resea ch Solu ions) ollowing manu ac u e ’s ins uc ions. cDNA was syn he ized om 2 μg o RNA using andom p ime s and Re e Aid Re e se T ansc ip ase (The mo Sci- en i ic), ollowing he manu ac u e ’s p o ocol. NIS exp ession le els we e quan i ied by SYBR G een based quan i a i e e e se ansc ip- ion polyme ase chain eac ion (RT-qPCR) on he Ligh Cycle 480II (Roche), using Xpe Fas SYBR (G isp). HPRT1 was used as endogenous con ol gene. NIS le els we e no malized o endogenous con ol exp ession le el. NIS no malized alues we e hen exp essed ela i e o ha o a e e ence sample (pool o TSH s imula ed a cell lines). Exp ession alues co espond o a bi a y uni s ep esen ing old di e ences ela i e o he e e ence sample. P ime s used (syn hesized by NZy ech, Po ugal) we e he ollowing: NIS F (5’-TCCTCACAGGCCGTATC TCA) and NIS R (5’–GAAGGAACCCTGGAGGACAC); HPRT1 F (5’-GCTGAAGATTTGGAA AAGGTG) and HPRT1 R (5’-AATCCAGCAGGTCAGCAAAG). IKBαexp ession le els we e quan i ied h ough he same me hod as NIS using IKBα he speci ic p ime s IKBαF (5’-TCC TCACAGGCCGTATCTCA) and IKBαR (5’-GACACGTGTGGCCGTTGTAG). HS-YFP–based iodide in lux assay Iodide in lux was accessed in PCCL3 cells s ably exp essing he halide-sensi i e yellow luo es- cen p o ein (HS-YFP-H148Q/I152L - Y-PCCL3 cells) [29,30]. Cells we e seeded in 8-well chambe slides (Ibidi) and se um-s a ed o 24h. Subsequen ly, cells we e s imula ed o 96h wi h 1 mU/mL TSH and hen ea ed wi h TNF-α(10 ng/mL, 24h) o PMA (10 ng/mL, 24h). Cells we e washed wi h PBS and hen incuba ed o 15 min a 37˚C wi h in lux-solu ion (137 mM NaCl, 2.7 mM KCl, 0.7 mM CaCl2, 1.1 mM MgCl2, 1.5 mM KH2PO4, 8.1 mM Na2HPO4, and 10 mM glucose, pH 7.4). Each well was assayed indi idually o iodide in lux by eco ding luo escence con inuously in a Leica TCS-SPE con ocal mic oscope a e addi ion o isomola PBS-NaI solu ion (1 mM inal NaI concen a ion). Decay o YFP luo escence was ollowed o 600 seconds, acqui ing an image e e y 10s. To con i m ha he iodine in lux obse ed was spe- ci ically media ed by NIS, addi ional assays we e pe o med in he p esence o ClO4- (a com- pe i i e inhibi o o iodide up ake by NIS), in which cells we e incuba ed wi h 1mM ClO4-, o 10min, be o e PBS-NaI solu ion was added. Images we e s acked and analyzed wi h Image J, de ining whole ield egions o in e es (ROIs) o pixel in ensi y measu emen s ha excluded TSH-induced NIS exp ession is hampe ed by TNF-α-media ed NF-kB ac i a ion PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0228794 Feb ua y 12, 2020 3 / 11 sa u a ed cells. Cell luo escence eco dings we e no malized o he ini ial a e age alue mea- su ed be o e addi ion o I−. To al p o ein lysa es and wes e n blo . Y-PCCl3 cells we e se um-s a ed o 24h, s imula ed o 96h wi h 1 mU/mL TSH and hen ea ed wi h TNF-α(10 ng/mL, 24h) o PMA (10 ng/mL, 24h). P o ein ex ac s we e p epa ed in Laemmli sample bu e (supplemen ed wi h 1 U/μl Benzonase, Sigma-Ald ich) and esol ed, acco ding o s anda d p o ocols, in 12% SDS-PAGE and ans e ed o PVDF memb anes (Bio-Rad). The p ima y an ibody abbi poly- clonal an i-NIS (P o ein ech) was used in Wes e n blo a 1:1000. An ibody mouse monoclonal an i-PCNA (Me ck) was used a 1:10000. De ec ion was ca ied ou using seconda y pe oxidase- conjuga ed an i- abbi o an i-mouse IgG (Bio-Rad) an ibodies ollowed by chemiluminescence. Da a and s a is ical analysis The da a used o suppo he indings o his s udy a e included wi hin he a icle. S a is ical analysis was ca ied ou using G aphPad P ism s a is ical so wa e (San Diego, CA). Quan i a- i e esul s a e shown as means ±SD o a leas h ee independen obse a ions. S a is ical compa isons o da a se s we e made using unpai ed wo- ailed S uden ’s - es and s a is ical signi icance was accep ed when p alues <0.05. In he halide-sensi i e YFP-based unc ional assay, he signal decay caused by YFP luo es- cence quenching was i ed o an exponen ial decay unc ion o de i e he maximal slope ha co - esponds o ini ial in lux o I − in o he cells. Maximal slopes we e con e ed in o a es o a ia ion o he in acellula I − concen a ion (in nmol/min) using he equa ion d[I − ]/d = K d [d(F/F 0 )/d ], whe e K d is he a ini y cons an o YFP o I − , and F/F 0 is he a io o he cell luo escence a a gi en ime e sus he ini ial luo escence [31,32]. Resul s and discussion TNF-αis a p o-in lamma o y cy okine widely de ined as a po en ac i a o o he canonical NF-κB pa hway. TNF-αhas been p e iously epo ed o induce a nega i e impac on NIS exp ession by down egula ing NIS mRNA le els up o 70% in FRTL5 a hy oid cells [27]. Howe e , whe he his e ec was dependen o NF-κB ac i a ion emained elusi e. NIS exp ession le els a e ypically educed in malignan hy oid issue ela i e o no mal issue [11,12]. This esul s om mul iple mechanisms elici ed by se e al signaling pa hways in ol ed in hy oid umo igenesis [33]. Consis en ly, mos o he DTC-de i ed cell lines a ail- able ail o exp ess hy oid-speci ic genes, including NIS and he TSH ecep o [34], which ham- pe s he s udy o TSH-media ed egula ion on NIS exp ession in hese neoplas ic cellula models (S1 Fig). Ne e heless, he s udy o he mechanisms unde lying NIS unc ional egula ion using non-neoplas ic, TSH- esponsi e, hy oid cell models ha e been shown o allow he iden i ica ion o key egula o y e en s ha can be u he ansla ed in o he malignan con ex [35–38]. In his s udy, we s a ed by assessing TNF-αimpac on TSH-induced NIS exp ession using he PCCL3 cell line as an addi ional cellula model. This, like he FRTL5 cell line, is also a non-neo- plas ic a hy oid ollicula cell-de i ed model ep esen a i e o TSH- esponsi e sys em o NIS exp ession and iodide up ake [8]. Se um-s a ed cells we e subjec ed o TSH s imula ion and hen ea ed wi h TNF-α o 1h o 24h. The e ec s on NIS mRNA le els we e accessed by qRT- PCR (Fig 1A). We con i med ha TNF-αled o down egula ion o TSH-induced NIS exp ession in FRLT5. This e ec was also obse ed in PCCL3 cells, eaching a s a is ically signi ican dec ease wi h a 24h ea men (Fig 1A). Nex , we asked whe he NIS down egula ion was a speci ic ou - come o TNF-αs imula ion o i i may esul om s imula ion wi h o he NF-κB ac i a o s. In ac , in con as o ha obse ed wi h TNF-α ea men , NIS le els we e shown o inc ease upon s imula ion o FRTL5 cells wi h LPS, ano he well-de ined ac i a o o he canonical NF-κB TSH-induced NIS exp ession is hampe ed by TNF-α-media ed NF-kB ac i a ion PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0228794 Feb ua y 12, 2020 4 / 11 pa hway. This inc ease was ac ually shown o be a esul o he syne gis ic ac ion o PAX8 and he p65-NF-κB subuni in he NUE egula o y egion [25]. This sugges s ha di e en igge s o NF-κB ac i a ion may esul in opposi e NIS exp ession ou comes. These may be a consequence o he conce ed ac ion be ween he signaling cues gene a ed by ligand- ecep o in e ac ion and he NF-κB ac i e dime s p oduced. Thus, we hen es ed he e ec o NF-κB ac i a ion on NIS exp ession when his is igge ed independen ly o ligand- ecep o speci ici y by s imula ing he cells wi h pho bol-my is a e-ace a e (PMA). This has been shown o induce canonical NF-κB- dependen ansc ip ion by ac ing h ough he di ec ac i a ion o p o ein-kinase-C [39]. We obse ed ha , simila ly o TNF-α, PMA down egula es NIS mRNA exp ession in bo h PCCL3 and FRTL5 cells (Fig 1B). I should also be no ed ha , albei PMA has been shown o a ec NIS p o ein ac ing as a down egula o o NIS unc ion and iodine up ake in human b eas cells [39,40], no e ec s a NIS ansc ip ional con ol le el ha e been epo ed so a . Ou nex s ep was o cla i y whe he his nega i e impac on NIS exp ession was in ac dependen on he NF-κB-media ed ansc ip ional ac i i y. Thus, we assessed he e ec o TNF-αand PMA in he p esence o BMS-345541, an allos e ic inhibi o o he IkB Kinase (IKK) complex [41] whose ac i i y is manda o y o canonical NF-κB ac i a ion [20]. We Fig 1. E ec o TNF-αand PMA on NIS ansc ip ional exp ession in non-neoplas ic, TSH- esponsi e hy oid ollicula cell lines. NIS mRNA le els we e assessed by RT-qPCR and co espond o a bi a y uni s ep esen ing old di e ences ela i e o a e e ence sample, co ec ed o HPRT le els used as endogenous con ol gene. Impac on NIS ansc ip le els o TNFα(A) and PMA (B) s imuli. FRTL5 and PCCL3 we e subjec ed o a 96h s a a ion pe iod ollowed by s imula ion wi h TSH (1 mU/mL o 24h), in he p esence o ei he TNFα(10 ng/mL; 1h and 24h) o PMA (10 ng/mL; 1h and 24h). As con ols, non-s imula ed cells and cells s imula ed wi h TSH o 24h in he absence o TNFαand PMA we e included. Plo ed alues a e he mean ±SD (e o ba s) o h ee independen assays. Compa isons o TNF-αand PMA s imula ion o non-s imula ed condi ions we e made using wo- ailed S uden ’s - es s (�p�0.05; ��p�0.01). h ps://doi.o g/10.1371/jou nal.pone.0228794.g001 TSH-induced NIS exp ession is hampe ed by TNF-α-media ed NF-kB ac i a ion PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0228794 Feb ua y 12, 2020 5 / 11 obse ed ha , in he p esence o BMS-345541, he dec ease in NIS ansc ip le els induced by bo h TNF-αand PMA was e e ed, wi h NIS mRNA eaching le els simila o hose o TSH s imula ion alone (Fig 2A). This obse a ion is consis en wi h NF-κB ac i a ion being equi ed o NIS down egula ion. NF-κB ac i a ion by TNF-αand PMA was u he moni o ed using IκBαmRNA le els as a eadou o NF-κB ansc ip ional ac i i y. In ac , one o he ea ly e en s esul ing om canonical NF-κB ac i a ion is he inc ease o IκB-αmRNA le els [20] and i has been p e iously shown ha IκBαmRNA le els co ela e obus ly wi h he ac i a ion o NF-κB in se e al cellula models, making i a b oadly e ec i e eadou o NF-κB ac i a ion [42]. A signi ican inc ease in IκB-α mRNA le els was obse ed upon s imula ion o bo h PCCL3 and FRTL5 cells wi h TNF-αand PMA. Mo eo e , his inc ease was s ongly inhibi ed by BMS-345541, indica ing a speci ic esponse om he NF-κB canonical pa hway (Fig 2B). Finally, we asked whe he TNF-αand PMA would also ha e a nega i e impac on iodide up ake. To add ess his ques ion, we es ablished an iodide in lux assay in PCCL3 cells, modi- ied o s ably exp ess he halide-sensi i e yellow luo escen p o ein (YFP) F46L/H148Q/I152L mu an (HS-YFP) [17,31,43] which has been alida ed o NIS ac i i y assessmen [29,30]. The HS-YFP inside hese modi ied PCCL3 cells (Y-PCCL3) is quenched by iodide, enabling changes in iodide in acellula concen a ion o be moni o ed in hy oid cells by li e cell imag- ing. Y-PCCL3 cells we e hus subjec ed o TSH s imula ion in he p esence o TNF-αand PMA. A nine- old as e decay a e o HS-YFP luo escence was obse ed upon TSH s imula- ion, compa ed o ha de ec ed in he absence o TSH (Fig 3A and 3B), consis en wi h TSH Fig 2. NF-κB impac on TNF-α- and PMA-media ed down egula ion o NIS ansc ip ional exp ession. NIS (A) and IKBα(B) mRNA le els we e quan i ied by RT-qPCR. TSH-s imula ed cells we e ea ed o 24h wi h TNF-αand PMA ei he in he absence ( ehicle) o p esence o he NF-κB inhibi o BMS-345541 (10 μM; 6h). Plo ed alues a e he mean ±SD (e o ba s) o h ee independen assays, compa ed wi h he g oup ea ed wi h only TSH. Compa isons we e made using wo- ailed S uden ’s - es s (�p�0.05; ��p�0.01; ��� p�0.001). h ps://doi.o g/10.1371/jou nal.pone.0228794.g002 TSH-induced NIS exp ession is hampe ed by TNF-α-media ed NF-kB ac i a ion PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0228794 Feb ua y 12, 2020 6 / 11 Fig 3. E ec o TNF-αand PMA on TSH-induced iodide up ake and NIS p o ein le els in Y-PCCL3 cells. Cells PCCL3 cells s ably exp essing he YFP-halide senso we e subjec ed o a 24h s a a ion pe iod and hen ea ed wi h TSH o 96h (TSH c l), in he p esence o absence o ei he TNF-αo PMA Addi ional iodide in lux assays we e pe o med in he p esence o bo h (24h ea men ), and subjec ed o iodide in lux assays and Wes e n blo . TSH and ClO4- (1 mM, 10 min), a compe i i e inhibi o o iodide up ake by NIS. Rep esen a i e HS-YFP luo escence decay TSH-induced NIS exp ession is hampe ed by TNF-α-media ed NF-kB ac i a ion PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0228794 Feb ua y 12, 2020 7 / 11 enhancemen o NIS exp ession ha ul ima ely p omo es NIS-media ed iodide up ake. We con i med ha he obse ed TSH-induced iodide in lux was media ed by NIS since i was comple ely abolished in he p esence o pe chlo a e (ClO4-) (Fig 3A and 3B), a compe i i e inhibi o o iodide up ake by NIS [44]. In addi ion, bo h TNF-αand PMA s imuli induced a meaning ul dec ease in TSH-induced iodide in lux in Y-PCCL3 cells (Fig 3A and 3B). Consis- en wi h he iodide in lux a es and he obse ed down egula ion o NIS exp ession a an- sc ip ional le el, a conside able educ ion in NIS p o ein le els was also no ed in he p esence o bo h TNF-αand PMA (Fig 3C). T ansc ip ional and pos - ansc ip ional egula ion o NIS exp ession in ol es di e en molecula playe s and signaling pa hways ha a e also implica ed in umo p og ession and agg essi eness. NF-κB amily o ansc ip ion ac o s is esponsible o egula ing many cellu- la p ocesses such as in lamma o y esponses, cellula g ow h and cell dea h, and has been also implica ed in TC de elopmen . We ha e p e iously epo ed NF-kB ac i a ion as a key mecha- nism h ough which he umo - ela ed RAC1b exe s i s oncogenic ac ion in he con ex o PTC [19]. Addi ionally, we ha e also ecen ly shown an in e se co ela ion be ween RAC1b o e exp ession and NIS le els in DTCs [45]. This p omp ed us o explo e, in he p esen s udy, he impac o he NF-κB signaling on NIS exp ession. Ou da a suppo ha he canonical NF- κB pa hway may also be in ol ed in he down egula ion o NIS. The in e play be ween he NF-κB ac i a ion and NIS exp ession, howe e , equi es u he cla i ica ion. This is pa icula ly impo an when we ake in o accoun ha he e ec s on NIS ansc ip ion induced by TNF-αand PMA a e opposi e o hose desc ibed o LPS [25]. One possible explana ion o hese appa en an agonis ic e ec s o NF-κB ac i a ion in NIS an- sc ip ion may be he ac i a ion and in e play o addi ional pa hways ha conce edly modula e he a ini y o NF-κB ansc ip ion ac o s o di e en p omo e s. Thus, NF-κB-d i en an- sc ip ion may lead o di e en esul s depending on he o e all signaling in ol ed in i s ac i a- ion and egula ion. Unde s anding he mechanisms unde lying physiological NIS unc ional egula ion may allow he iden i ica ion o egula o y e en s ha can be pha macological a ge ed o enhance he iodide up ake e iciency and imp o e he esponse o RAI he apy in RAI- e ac o y hy- oid cance s. Suppo ing in o ma ion S1 Fig. E ec o TSH on NIS ansc ip ional exp ession in hy oid ollicula cell-de i ed cell lines. NIS mRNA le els we e assessed by RT-qPCR and co espond o a bi a y uni s ep esen ing old di e ences ela i e o a e e ence sample, co ec ed o endogenous con ol exp ession le els. HPRT1 and TBP we e used as endogenous con ol genes o a and human, espec i ely. RT-qPCR p e o med as p e iously desc ibed [45]. The di e en cell lines we e subjec ed o a 96h s a a ion pe iod ollowed by s imula ion wi h TSH (1 mU/mL o 48h). Plo ed alues a e he mean ±SD (e o ba s) o h ee independen assays. (TIF) aces (A) eco ded con inuously o 600 seconds, acqui ing an image e e y 10 s, a e exposu e o 1mM NaI (as desc ibed in [30]). Fluo escence (F) was plo ed o e ime as pe cen age o luo escence a ime 0 (F0). Iodide in lux a es (B) calcula ed by i ing he cu es o he exponen ial decay unc ion o de i e he maximal slope ha co esponds o ini ial in lux o I − in o he cells. Endogenous NIS p o ein exp ession upon TSH, TNF-αo PMA ea men was moni o ed Wes e n Blo using an i-NIS p ima y an ibody. Endogenous PCNA exp ession was used as loading con ol. Da a a e means ±SEM o h ee independen assays. Compa isons we e made using one- ailed S uden ’s - es s (�p�0.05; ��p�0.01; ���p�0.001). h ps://doi.o g/10.1371/jou nal.pone.0228794.g003 TSH-induced NIS exp ession is hampe ed by TNF-α-media ed NF-kB ac i a ion PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0228794 Feb ua y 12, 2020 8 / 11 Au ho Con ibu ions Concep ualiza ion: Ana Luı ´sa Sil a. Fo mal analysis: Ma ´ cia Fa ia, Ana Luı ´sa Sil a. Funding acquisi ion: Paulo Ma os, Ana Luı ´sa Sil a. In es iga ion: Ma ´ cia Fa ia, Ri a Domingues, F ancisca Paixão. Me hodology: Ma ´ cia Fa ia, Ri a Domingues, Paulo Ma os. P ojec adminis a ion: Ana Luı ´sa Sil a. Resou ces: Ma ia João Bugalho, Paulo Ma os, Ana Luı ´sa Sil a. Supe ision: Ana Luı ´sa Sil a. Valida ion: Ma ´ cia Fa ia, Ri a Domingues. Visualiza ion: Ana Luı ´sa Sil a. W i ing – o iginal d a : Ana Luı ´sa Sil a. W i ing – e iew & edi ing: Ma ia João Bugalho, Paulo Ma os. Re e ences 1. D alle H, Machens A, Basa J, Fa ou echi V, F anceschi S, Hay ID, e al. Follicula cell-de i ed hy oid cance . Na Re Dis P ime . 2015; 1: 15077. 2. Coope DS, Dohe y GM, Haugen BR, Kloos RT, Lee SL, Mandel SJ, e al. Managemen guidelines o pa ien s wi h hy oid nodules and di e en ia ed hy oid cance . Thy oid O J Am Thy oid Assoc. 2006; 16: 109–142. 3. Schmid A, Iglesias L, Klain M, Pi oia F, Schlumbe ge MJ. Radioac i e iodine- e ac o y di e en ia ed hy oid cance : an uncommon bu challenging si ua ion. A ch Endoc inol Me ab. 2017; 61: 81–89. h ps://doi.o g/10.1590/2359-3997000000245 PMID: 28225999 4. Spi zweg C, Bible KC, Ho baue LC, Mo is JC. Ad anced adioiodine- e ac o y di e en ia ed hy oid cance : he sodium iodide sympo e and o he eme ging he apeu ic a ge s. Lance Diabe es Endoc i- nol. 2014; 2: 830–842. h ps://doi.o g/10.1016/S2213-8587(14)70051-8 PMID: 24898835 5. Du an e C, Haddy N, Baudin E, Leboulleux S, Ha l D, T a agli JP, e al. Long- e m ou come o 444 pa ien s wi h dis an me as ases om papilla y and ollicula hy oid ca cinoma: bene i s and limi s o adioiodine he apy. J Clin Endoc inol Me ab. 2006; 91: 2892–2899. h ps://doi.o g/10.1210/jc.2005- 2838 PMID: 16684830 6. Doha ´n O, Ca asco N. Ad ances in Na+/I−sympo e (NIS) esea ch in he hy oid and beyond. Mol Cell Endoc inol. 2003; 213: 59–70. h ps://doi.o g/10.1016/j.mce.2003.10.059 PMID: 15062574 7. Ra e a S, Reyna-Ney a A, Fe andino G, Amzel LM, Ca asco N. The Sodium/Iodide Sympo e (NIS): Molecula Physiology and P eclinical and Clinical Applica ions. Annu Re Physiol. 2017; 79: 261–289. h ps://doi.o g/10.1146/annu e -physiol-022516-034125 PMID: 28192058 8. Kogai T, Endo T, Sai o T, Miyazaki A, Kawaguchi A, Onaya T. Regula ion by hy oid-s imula ing ho - mone o sodium/iodide sympo e gene exp ession and p o ein le els in FRTL-5 cells. Endoc inology. 1997; 138: 2227–2232. h ps://doi.o g/10.1210/endo.138.6.5189 PMID: 9165005 9. Kogai T, B en GA. The sodium iodide sympo e (NIS): egula ion and app oaches o a ge ing o can- ce he apeu ics. Pha macol The . 2012; 135: 355–370. h ps://doi.o g/10.1016/j.pha m he a.2012.06. 007 PMID: 22750642 10. Taki K, Kogai T, Kanamo o Y, He shman JM, B en GA. A hy oid-speci ic a -ups eam enhance in he human sodium/iodide sympo e gene equi es Pax-8 binding and cyclic adenosine 3’,5’-monophos- pha e esponse elemen -like sequence binding p o eins o ull ac i i y and is di e en ially egula ed in no mal and hy oid cance cells. Mol Endoc inol Bal im Md. 2002; 16: 2266–2282. h ps://doi.o g/10. 1210/me.2002-0109 PMID: 12351692 11. Laza V, Bida JM, Caillou B, Mahe ´C, Lac oix L, File i S, e al. Exp ession o he Na+/I- sympo e gene in human hy oid umo s: a compa ison s udy wi h o he hy oid-speci ic genes. J Clin Endoc inol Me ab. 1999; 84: 3228–3234. h ps://doi.o g/10.1210/jcem.84.9.5996 PMID: 10487692 TSH-induced NIS exp ession is hampe ed by TNF-α-media ed NF-kB ac i a ion PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0228794 Feb ua y 12, 2020 9 / 11