Comparative analysis for the presence of IgG anti-aquaporin-1 in patients with NMO-Spectrum disorders

In e na ional Jou nal o Molecula Sciences A icle Compa a i e Analysis o he P esence o IgG An i-Aquapo in-1 in Pa ien s wi h NMO-Spec um Diso de s Ismael Sánchez Goma 1, Ma ía Díaz Sánchez 2, An onio José Uclés Sánchez 2, José Luis Casado Chocán 2, Nela Suá ez-Luna 1, Reposo Ramí ez-Lo ca 1, Ja ie Villadiego 1, Juan José Toledo-A al 1,3 and Mi iam Eche a ía 1,4,* 1Ins i u o de Biomedicina de Se illa (IBiS), Hospi al Uni e si a io Vi gen del Rocío/CSIC/Uni e sidad de Se illa, Se ille 41013, Spain; [email p o ec ed] (I.S.G.); [email p o ec ed] (N.S.-L.); epo ami [email p o ec ed] (R.R.-L.); [email p o ec ed] (J.V.); [email p o ec ed] (J.J.T.-A.) 2Unidad de Ges ión Clínica de Neu ociencias, Se icio de Neu ología del Hospi al Uni e si a io Vi gen del Rocío, Se ille 41013, Spain; [email p o ec ed] (M.D.S.); [email p o ec ed] (A.J.U.S.); [email p o ec ed] (J.L.C.C.) 3Cen o de In es igación Biomédica en Red sob e En e medades Neu odegene a i as (CIBERNED), Mad id 28029, Spain 4Cen o de In es igación Biomédica en Red sob e En e medades Respi a o ias (CIBERES), Mad id 28029, Spain *Co espondence: [email p o ec ed]; Tel.: +34-955-923-036 Academic Edi o : Kenichi Ishibashi Recei ed: 31 May 2016; Accep ed: 19 July 2016; Published: 23 July 2016 Abs ac : De ec ion o IgG an i-Aquapo in-4 (AQP4) in se um o pa ien s wi h Neu omyeli is op ica synd ome diso de s (NMOSD) has imp o ed diagnosis o hese p ocesses and di e en ia ion om Mul iple scle osis (MS). Recen indings also claim ha a subg oup o pa ien s wi h NMOSD, se um nega i e o IgG-an i-AQP4, p esen an ibodies an i-AQP1 ins ead. Explo e he p esence o IgG-an i-AQP1 using a p e iously de eloped cell-based assay (CBA) highly sensi i e o IgG-an i-AQP4. Se um o 205 pa ien s diagnosed as NMOSD (8), mul iple scle osis (94), op ic neu i is (39), idiopa hic myeli is (29), o he idiopa hic demyelina ing diso de s o he cen al ne ous sys em (9), o he neu ological diseases (18) and heal hy con ols (8), we e used in a CBA o e ixed HEK cells ans ec ed wi h hAQP1-EGFP o hM23-AQP4-EGFP, ea ed wi h T i on X-100 and un ea ed. ELISA was also pe o med. Analysis o se um wi h ou CBA indica ed absence o an i-AQP1 an ibodies, whe eas in cells p e ea ed wi h de e gen , noisy signal made eliable de ec ion impossible. ELISA showed posi i e esul s in ew se ums. The low numbe o NMOSD se ums included in ou s udy educes i s powe o conclude he speci ici y o AQP1 an ibodies as new bioma ke s o NMOSD. Ou s udy does no sus ain de ec ion o an i-AQP1 in se um o NMOSD pa ien s bu u he expe imen s a e expec ed. Keywo ds: AQP1; AQP4; NMO-IgG; HEK cells; neu omyeli is op ica 1. In oduc ion Neu omyeli is op ica (NMO) is an au oimmune, in lamma o y and demyelina ing disease o he cen al ne ous sys em (CNS) ha p ima ily a ec s he op ic ne es and spinal co d. Pa ien s wi h his disease, also known as De ic synd ome, p esen ecu en episodes o op ic neu i is and acu e myeli is ha conduce o a loss o uni- o bila e al ision and o a sensi i e and mo o comp omise below he a ec ed medulla le el wi h equen loss o sphinc e s con ol [ 1 , 2 ]. Al hough i was conside ed o a long ime as a a ian o mul iple scle osis (MS), new pa hological and se ological es s ha e con ibu ed In . J. Mol. Sci. 2016,17, 1195; doi:10.3390/ijms17081195 www.mdpi.com/jou nal/ijms In . J. Mol. Sci. 2016,17, 1195 2 o 11 o iden i y his diso de as a di e en disease [ 3 ]. The main e idence o such dis inc ion was p o ided by Lennon e al. [ 3 – 5 ] when hey disco e ed he p esence o speci ic immunoglobulins (IgG-NMO) in se um o NMO pa ien s, IgG ha we e usually absen in classical o ms o MS. The an igen ecognized o IgG-NMO is Aquapo in-4 (AQP4), he mos abundan ly exp essed aquapo in in he CNS [ 6 – 8 ], highly localized in as ocy es memb ane acing blood essels, in ependymal cells o b ain en icles, and laye s o he meninges su ounding he b ain and spinal co d [ 4 ]. Recen wo ks p esen con incing e idences suppo ing a di ec in ol emen o AQP4 au oan ibodies (IgG-NMO) in he de elopmen o NMO disease [ 4 , 9 – 11 ]. These an ibodies ha e also been iden i ied in pa ien s wi h limi ed o ms o his disease ( ecu en idiopa hic op ic neu i is, e c.), bu wi h a high isk o su e ing subsequen clinical episodes. In his sense, las yea , he In e na ional Panel o NMO diagnosis was con ened o de elop e ised diagnos ic c i e ia o include hese limi ed o ms o he disease in o de o acili a e a diagnosis consensus. The new nomencla u e de ines he uni ying e m NMO spec um diso de (NMOSD) ha is s a i ied u he by he p esence o absence o IgG-NMO and he p esence o an isola ed o se e al clinical e en s [12]. Un o una ely, li le mo e han 20% o pa ien s included in he g oup o demyelina ing diso de s o he NMOSD a e se onega i e o an i-AQP4 an ibodies [ 12 , 13 ]. Besides AQP4, human as ocy es in he CNS also exp ess Aquapo in-1 (AQP1), and speci ically a la ge exp ession o his p o ein has been de ec ed in a eas wi h p ocli i y o de elop NMO-like lesions as in he spinal co d, op ic ne es and in he whi e ma e o he b ain [ 14 – 17 ]. In he las ew yea s, se e al g oups [ 17 – 20 ] ha e ac i ely sea ched o he p esence o an i-AQP1 an ibodies in he se um o pa ien s wi h ch onic demyelina ing p ocess o he CNS using di e en me hods, and some con lic ing indings ha e aised in his espec . On one side, wo g oups [ 17 – 19 ], one based in p o ein immunoblo ing and ELISA [ 17 , 18 ] and he o he in esul s ob ained om a cell based assay (CBA) using ixed HEK cells p e ea ed wi h i on [ 19 ], ha e demons a ed ha a subg oup o pa ien s, some o hem se onega i e o an i-AQP4, p esen an ibodies an i AQP1 in se um. Such esul s allowed hese au ho s o sugges ha his an ibody may be aken as ano he new classi ying bioma ke o hese demyelina ing diso de s. In con as o his, Schanda e al., [ 20 ] ailed o con i m he p esence o AQP1 an ibodies in NMOSD pa ien s using a li e cell immuno luo escence assay. The e o e, gi en his scena io, we decided o explo e in a coho o 205 se ums, be ween heal hy and pa ien s, ec ui ed om he Neu ology Se ice o Hospi al Uni e si a io Vi gen del Rocío, he bigges hospi al o Andalusia, Spain; o he p esence o an i-AQP1 an ibodies, wi h he aim o helping o es ablish a simple assay o iden i y wi h speci ici y he p esence o hese an ibodies and con ibu e o cla i y whe he o no AQP1-an ibody, should be conside ed as a new p o agonis o he o igin o NMOSD pa hology. In he p esen wo k we show he esul s ob ained using a p o ocol o ixed CBA wi h HEK cells, as published in ou p e ious wo k [ 21 ], o iden i y p esence o an i-AQP4 an ibodies, a e adap ing he p o ocol in o de o look o AQP1 exp ession as well. Se um o pa ien s we e analyzed: 8 NMOSD, 94 MS, 39 op ic neu i is (ON), 29 idiopa hic myeli is, 9 o he idiopa hic demyelina ing p ocesses o CNS, 18 o he neu ological diso de s and 8 con ols, and he assay was pe o med o e AQP1- o AQP4- ans ec ed HEK293 cells, ei he pe meabilized wi h i on X-100 o un ea ed. We conclude om ou esul s ha he CBA as desc ibed in he p esen wo k o by o he s [ 20 , 21 ] does no gi e a speci ic signal ha can allow o he de ec ion o an i-AQP1 an ibodies in he se um pa ien s. Howe e , he low numbe o NMOSD se ums included in ou analysis educes he s eng h o mo e de ini i e conclusions, and we hink ha i would be s ill wo h o dedica e some addi ional e o s o cla i y whe he o no an i-AQP1 an ibodies a e p esen in se um o pa ien s, and i so, whe he he p esence o hese an ibodies may be used as a new bioma ke , ha could be widely used as a complemen o he an i-AQP4 an ibody assay ha s ill emains as a pa adigm bioma ke o NMOSD diagnosis. 2. Resul s and Discussion Recen wo ks ha e p o ided con incing e idences ha implica e he di ec pa icipa ion o an i-AQP4 an ibodies in he pa hogenesis and diagnosis o he NMOSD [ 9 – 11 ]. De ec ion o IgG-an i-AQP4 in he se um o pa ien s has become a decisi e bioma ke o diagnosis o mos In . J. Mol. Sci. 2016,17, 1195 3 o 11 NMOSD pa ien s bu un o una ely s ill a ound 20% o NMOSD pa ien s esul in se um-nega i e o hese an ibodies [ 12 , 13 ]. Finding ano he bioma ke o hese diso de s in hese NMO-AQP4 se um nega i e pa ien s would be highly desi able and has led in es iga o s o u n hei a en ion o AQP1, ano he p o ein o he same amily o in eg al memb ane p o eins, whose exp ession has also been obse ed in he CNS in a na u al way, bu i is, ascina ingly, o e exp essed when inju y o he ne ous issue occu s [ 14 – 16 ]. Two p e ious wo ks, Tza os e al., [ 17 ] and Long e al., [19], indica ed ha an i-AQP1 an ibodies a e p esen in he se um o pa ien s wi h NMOSD. We decided he e o e o explo e he p esence o hese an ibodies in he se um o a la ge epe oi e o pa ien s o which de ec ion o IgG-an i-AQP4 was p e iously es ed using a me hod desc ibed in ou labo a o y [ 21 ], ha was p e iously alida ed con as ing ou esul s agains hose ob ained o he same se ums, by he only cen e o e e ence o diagnosis o an i-NMO an ibodies in Spain (Hospi al Clínic, Ba celona, Spain). We ini ia ed he sc eening using he same p o ocol o CBA p e iously desc ibed [ 21 ] bu now, HEK cells we e ans ec ed wi h human AQP1-EGFP plasmid o p oduce a la ge exp ession o AQP1 in he cell memb ane. Readou o AQP1 exp ession was ollowed by di ec appea ance o g een luo escen signal coming om he EGFP used o AQP1 (Figu e 1, hAQP1-EGFP). As shown in Figu e 1, o al absence o AQP1 an ibodies was ob ained a e analysis o 205 se ums ob ained om pa ien s and con ols (Table 1). Independen ly o whe he he se um came om a pa ien diagnosed as NMOSD o no , he esul was always he same: absence o luo escen signal. In . J. Mol. Sci. 2016, 17, 1195 3 o 11 NMOSD pa ien s bu un o una ely s ill a ound 20% o NMOSD pa ien s esul in se um-nega i e o hese an ibodies [12,13]. Finding ano he bioma ke o hese diso de s in hese NMO-AQP4 se um nega i e pa ien s would be highly desi able and has led in es iga o s o u n hei a en ion o AQP1, ano he p o ein o he same amily o in eg al memb ane p o eins, whose exp ession has also been obse ed in he CNS in a na u al way, bu i is, ascina ingly, o e exp essed when inju y o he ne ous issue occu s [14–16]. Two p e ious wo ks, Tza os e al., [17] and Long e al., [19], indica ed ha an i-AQP1 an ibodies a e p esen in he se um o pa ien s wi h NMOSD. We decided he e o e o explo e he p esence o hese an ibodies in he se um o a la ge epe oi e o pa ien s o which de ec ion o IgG-an i-AQP4 was p e iously es ed using a me hod desc ibed in ou labo a o y [21], ha was p e iously alida ed con as ing ou esul s agains hose ob ained o he same se ums, by he only cen e o e e ence o diagnosis o an i-NMO an ibodies in Spain (Hospi al Clínic, Ba celona, Spain). We ini ia ed he sc eening using he same p o ocol o CBA p e iously desc ibed [21] bu now, HEK cells we e ans ec ed wi h human AQP1-EGFP plasmid o p oduce a la ge exp ession o AQP1 in he cell memb ane. Readou o AQP1 exp ession was ollowed by di ec appea ance o g een luo escen signal coming om he EGFP used o AQP1 (Figu e 1, hAQP1-EGFP). As shown in Figu e 1, o al absence o AQP1 an ibodies was ob ained a e analysis o 205 se ums ob ained om pa ien s and con ols (Table 1). Independen ly o whe he he se um came om a pa ien diagnosed as NMOSD o no , he esul was always he same: absence o luo escen signal. Figu e 1. Immuno luo escence assay in HEK cells exp essing AQP1 o AQP4 wi h se um o pa ien s. The luo escence signal om HEK cells exp essing human AQP1 (hAQP1-EGFP) o AQP4 (hAQP4-EGFP) used o GFP is shown in g een (le column). In he cen al panels he immune eac ion p oduced by se um pa ien s ei he om he Neu omyeli is op ica spec um diso de (+NMOSD) o no (−NMOSD) o e AQP1 exp essing cells e ealed absence o an i-AQP1 an ibodies in analyzed se ums. In con as , se um om a posi i e an i-AQP4 an ibody pa ien (+NMOSD) showed clea eac i i y o e AQP4 exp essing cells (yellow signal), while nega i e esul s a e ob ained in he absence o IgG an i-AQP4 in a pa ien (−NMOSD). The me ge image o bo h luo escen signals is shown in he igh column. Scale ba = 50 µm. Figu e 1. Immuno luo escence assay in HEK cells exp essing AQP1 o AQP4 wi h se um o pa ien s. The luo escence signal om HEK cells exp essing human AQP1 (hAQP1-EGFP) o AQP4 (hAQP4-EGFP) used o GFP is shown in g een (le column). In he cen al panels he immune eac ion p oduced by se um pa ien s ei he om he Neu omyeli is op ica spec um diso de (+NMOSD) o no ( ´ NMOSD) o e AQP1 exp essing cells e ealed absence o an i-AQP1 an ibodies in analyzed se ums. In con as , se um om a posi i e an i-AQP4 an ibody pa ien (+NMOSD) showed clea eac i i y o e AQP4 exp essing cells (yellow signal), while nega i e esul s a e ob ained in he absence o IgG an i-AQP4 in a pa ien ( ´ NMOSD). The me ge image o bo h luo escen signals is shown in he igh column. Scale ba = 50 µm. In . J. Mol. Sci. 2016,17, 1195 4 o 11 The high speci ici y and sensi i i y o he CBA used he e was u he e i ied in pa allel expe imen s in which HEK cells we e ans ec ed wi h human AQP4 (Figu e 1, hAQP4-EGFP), and by doing so, AQP4 an ibodies we e clea ly de ec ed in he se um o NMOSD pa ien s, while hey we e ne e obse ed in pa ien s wi h o he diseases and heal hy con ols. Then, assuming ha he disc epancy wi h he esul s ob ained by o he s [ 17 – 19 ] was due o di e ences in he pe meabiliza ion p ocedu e, we ea ed HEK cells wi h i on X-100 as indica ed by Long e al. [ 19 ]. Using hei pe meabiliza ion p o ocol, we ob ained a la ge luo escence signal, bu i was highly unspeci ic, and coming om all cells in he pla e ega dless o whe he hey exp essed o no AQP1 o AQP4 (Figu e 2). These noisy esul s ob ained in ou hands wi h such a p ocedu e led us o conclude ha i is no accu a e enough o he de ec ion nei he o AQP1 no AQP4 an ibodies in he se um o pa ien s. In . J. Mol. Sci. 2016, 17, 1195 4 o 11 The high speci ici y and sensi i i y o he CBA used he e was u he e i ied in pa allel expe imen s in which HEK cells we e ans ec ed wi h human AQP4 (Figu e 1, hAQP4-EGFP), and by doing so, AQP4 an ibodies we e clea ly de ec ed in he se um o NMOSD pa ien s, while hey we e ne e obse ed in pa ien s wi h o he diseases and heal hy con ols. Then, assuming ha he disc epancy wi h he esul s ob ained by o he s [17–19] was due o di e ences in he pe meabiliza ion p ocedu e, we ea ed HEK cells wi h i on X-100 as indica ed by Long e al. [19]. Using hei pe meabiliza ion p o ocol, we ob ained a la ge luo escence signal, bu i was highly unspeci ic, and coming om all cells in he pla e ega dless o whe he hey exp essed o no AQP1 o AQP4 (Figu e 2). These noisy esul s ob ained in ou hands wi h such a p ocedu e led us o conclude ha i is no accu a e enough o he de ec ion nei he o AQP1 no AQP4 an ibodies in he se um o pa ien s. Figu e 2. Immuno luo escence assay in HEK cells ea ed wi h T i on X-100. Panels a e as indica ed in Figu e 1, bu he e expe imen s we e done using a p o ocol o pe meabiliza ion wi h i on X-100 as indica ed in he ex (ma e ial and me hods sec ion). High backg ound o luo escen signal (middle column panels) was de ec ed o e all cells in he pla e ega dless exp ession o no o AQPs in hem, con i ming unspeci ic eac ion. Scale ba = 20 µm As seen in Figu es 1 and 2, an appa en ly no mal exp ession o he luo escen p o ein hAQP1-EGFP was being p oduced in he ans ec ed HEK cells. Howe e , o u he disca d he possibili y ha AQP1 was no being inse ed in an app op ia e way in o he plasma memb ane o cells, hinde ing he na u al ecogni ion by an ibodies which could explain he lack o immune eac ion and consequen de ec ion o AQP1 an ibodies, we pe o med expe imen s using a comme cial p ima y an ibody o AQP1. Resul s om hese analyses a e shown in Figu e 3. Figu e 2. Immuno luo escence assay in HEK cells ea ed wi h T i on X-100. Panels a e as indica ed in Figu e 1, bu he e expe imen s we e done using a p o ocol o pe meabiliza ion wi h i on X-100 as indica ed in he ex (ma e ial and me hods sec ion). High backg ound o luo escen signal (middle column panels) was de ec ed o e all cells in he pla e ega dless exp ession o no o AQPs in hem, con i ming unspeci ic eac ion. Scale ba = 20 µm As seen in Figu es 1and 2, an appa en ly no mal exp ession o he luo escen p o ein hAQP1-EGFP was being p oduced in he ans ec ed HEK cells. Howe e , o u he disca d he possibili y ha AQP1 was no being inse ed in an app op ia e way in o he plasma memb ane o cells, hinde ing he na u al ecogni ion by an ibodies which could explain he lack o immune eac ion and consequen de ec ion o AQP1 an ibodies, we pe o med expe imen s using a comme cial p ima y an ibody o AQP1. Resul s om hese analyses a e shown in Figu e 3. In . J. Mol. Sci. 2016,17, 1195 5 o 11 In . J. Mol. Sci. 2016, 17, 1195 5 o 11 Figu e 3. Immuno luo escence assay wi h comme cial an ibody o AQP1. HEK cells exp essing AQP1 in g een (le column) immune eac ed wi h a comme cial IgG-an i-AQP1 ( op middle panel, ed signal), using ixed cells and wi hou i on pe meabiliza ion. A clea and speci ic immune eac ion was de ec ed only in cells exp essing AQP1 (me ge signal in o ange, op igh panel). Absence o IgG an i-AQP1 was e ealed in all es o condi ions es ed: wi h an an i-human IgG seconda y an ibody; wi h human se um om a pa ien wi h NMOSD (+NMOSD); and wi h human se um om a pa ien nega i e o NMOSD (−NMOSD). Me ge o luo escence signal can only be seen when he comme cial an ibody o AQP1 was used (o ange/yellow signal, op igh panel). Scale ba = 15 µm. A ed luo escen signal wi h he IgG-an ihuman AQP1 comme cial was obse ed clea ly labeling all cells exp essing hAQP1-EGFP and s ained in g een, esul ing in a pe ec o e lapping me ge o luo escence signals. The eby, hese esul s demons a e wo impo an indings: Fi s , ha exp ession o AQP1 in an an igenic way occu s in HEK cells and second, ha human se ums ei he om he NMOSD subjec s o no , did no p esen IgG an i-AQP1 o a leas no enough le els, o allow di e en ia ion among hese pa ien s by ou CBA p ocedu e. We conclude hen ha he absence o an i-AQP1 an ibodies obse ed in all he 205 se ums analyzed wi h he CBA assay is no due o he lack o AQP1 exp ession in he cells, bu ins ead is likely due o a ue absence o hese an ibodies in he se um, a leas a le els de ec able wi h ou p o ocol o CBA. A p e ious s udy [20], using a li e CBA assay wi h a p o ocol ha allowed au ho s o a nice de ec ion o AQP4 an ibodies in human se um, also ailed o con i m he p esence o AQP1 an ibodies in NMOSD in ag eemen wi h esul s p esen ed he e by us. As indica ed by Schanda e al., [20] we also ag ee ha AQP1 an ibodies in pa ien s mus eac agains in acellula epi opes o AQP1, o he wise pa ien s would su e o se e e anemia due o he hemoly ic ac ion o an ibodies eac ing o e he Col on blood g oup an igen p esen on he e y h ocy es memb ane, and ha is no Figu e 3. Immuno luo escence assay wi h comme cial an ibody o AQP1. HEK cells exp essing AQP1 in g een (le column) immune eac ed wi h a comme cial IgG-an i-AQP1 ( op middle panel, ed signal), using ixed cells and wi hou i on pe meabiliza ion. A clea and speci ic immune eac ion was de ec ed only in cells exp essing AQP1 (me ge signal in o ange, op igh panel). Absence o IgG an i-AQP1 was e ealed in all es o condi ions es ed: wi h an an i-human IgG seconda y an ibody; wi h human se um om a pa ien wi h NMOSD (+NMOSD); and wi h human se um om a pa ien nega i e o NMOSD ( ´ NMOSD). Me ge o luo escence signal can only be seen when he comme cial an ibody o AQP1 was used (o ange/yellow signal, op igh panel). Scale ba = 15 µm. A ed luo escen signal wi h he IgG-an ihuman AQP1 comme cial was obse ed clea ly labeling all cells exp essing hAQP1-EGFP and s ained in g een, esul ing in a pe ec o e lapping me ge o luo escence signals. The eby, hese esul s demons a e wo impo an indings: Fi s , ha exp ession o AQP1 in an an igenic way occu s in HEK cells and second, ha human se ums ei he om he NMOSD subjec s o no , did no p esen IgG an i-AQP1 o a leas no enough le els, o allow di e en ia ion among hese pa ien s by ou CBA p ocedu e. We conclude hen ha he absence o an i-AQP1 an ibodies obse ed in all he 205 se ums analyzed wi h he CBA assay is no due o he lack o AQP1 exp ession in he cells, bu ins ead is likely due o a ue absence o hese an ibodies in he se um, a leas a le els de ec able wi h ou p o ocol o CBA. A p e ious s udy [ 20 ], using a li e CBA assay wi h a p o ocol ha allowed au ho s o a nice de ec ion o AQP4 an ibodies in human se um, also ailed o con i m he p esence o AQP1 an ibodies in NMOSD in ag eemen wi h esul s p esen ed he e by us. As indica ed by Schanda e al., [ 20 ] we also ag ee ha AQP1 an ibodies in pa ien s mus eac agains in acellula epi opes o AQP1, o he wise pa ien s would su e o se e e anemia due o he hemoly ic ac ion o an ibodies eac ing o e he Col on blood g oup an igen p esen on he e y h ocy es memb ane, and ha is no he In . J. Mol. Sci. 2016,17, 1195 6 o 11 case in any o ou pa ien s. Al e na i e con o ma ional ways o exp ession o AQP1 in as ocy es compa e o he way he p o ein is exp essed in HEK293 cells, as indica ed by Schanda e al., [ 20 ] could s ill o e some explana ion o he lack o posi i e esul s when using CBA assays o de ec ion o AQP1 an ibodies, bu all hese explana ions will need u he expe imen al analysis using di e en expe imen al app oaches. In ha sense, we ca ied ou expe imen s o ELISA o u he clea his appa en disc epancy in he ield and, al hough in ew se ums alues o an i-AQP1 an ibodies we e de ec ed (Figu e 4), he o e all compa a i e analysis o he le els o an i-AQP1 an ibodies among all g oups analyzed e ealed no s a is ical di e ences among hem, making impossible a his s a e o us any discussion abou p esence o AQP1 an ibodies in hose se ums. The e o e, ou indings wi h he ELISA a e inconsis en wi h an associa ion be ween p esence o an i-AQP1 an ibodies and NMOSD pa hology. We a e awa e ha a la ge numbe o samples in ou analysis would be desi able o inc ease he powe o ou conclusions, and o allowing us de e mina ion o he cu -o le el o an i-AQP1 an ibody in ELISA, bu such aim was un o una ely impossible a he ime o he s udy. Ou gene al imp ession, gi en hese esul s, is ha AQP1 an ibodies does no ep esen ano he speci ic bioma ke o he NMOSD bu i s a e p esence in se um o some pa ien s as indica ed by o he au ho s [ 17 – 19 ] may p obably be associa ed wi h an inc eased au oimmune humo al esponse as p e iously obse ed in pa ien s wi h MS in which p esence o an inuclea (ANA), an ica diolipin (ACA) and an i-Ro (SS-A) an ibodies ha e been de ec ed [ 22 , 23 ]. Like hese an ibodies, he p esence o an i-AQP1 an ibodies may be indica i e o an unde lying au oimmune disease; bu mo e han associa ed o pa icula aspec s o any o hese diseases, i appea s as an epiphenomenon o a mo e di use immunological dys unc ion. Ne e heless, u he analysis wi h he ELISA, o o he immune assays, especially hose wi h de ec ion o he p o ein in suspension, accompanied wi h mo e analysis including la ge numbe o pa ien s on each subg oup o hese di e se pa hologies as well as mo e cases o he NMOSD would be necessa y o s eng h ou incipien hypo hesis. In . J. Mol. Sci. 2016, 17, 1195 6 o 11 he case in any o ou pa ien s. Al e na i e con o ma ional ways o exp ession o AQP1 in as ocy es compa e o he way he p o ein is exp essed in HEK293 cells, as indica ed by Schanda e al., [20] could s ill o e some explana ion o he lack o posi i e esul s when using CBA assays o de ec ion o AQP1 an ibodies, bu all hese explana ions will need u he expe imen al analysis using di e en expe imen al app oaches. In ha sense, we ca ied ou expe imen s o ELISA o u he clea his appa en disc epancy in he ield and, al hough in ew se ums alues o an i-AQP1 an ibodies we e de ec ed (Figu e 4), he o e all compa a i e analysis o he le els o an i-AQP1 an ibodies among all g oups analyzed e ealed no s a is ical di e ences among hem, making impossible a his s a e o us any discussion abou p esence o AQP1 an ibodies in hose se ums. The e o e, ou indings wi h he ELISA a e inconsis en wi h an associa ion be ween p esence o an i-AQP1 an ibodies and NMOSD pa hology. We a e awa e ha a la ge numbe o samples in ou analysis would be desi able o inc ease he powe o ou conclusions, and o allowing us de e mina ion o he cu -o le el o an i-AQP1 an ibody in ELISA, bu such aim was un o una ely impossible a he ime o he s udy. Ou gene al imp ession, gi en hese esul s, is ha AQP1 an ibodies does no ep esen ano he speci ic bioma ke o he NMOSD bu i s a e p esence in se um o some pa ien s as indica ed by o he au ho s [17–19] may p obably be associa ed wi h an inc eased au oimmune humo al esponse as p e iously obse ed in pa ien s wi h MS in which p esence o an inuclea (ANA), an ica diolipin (ACA) and an i-Ro (SS-A) an ibodies ha e been de ec ed [22,23]. Like hese an ibodies, he p esence o an i-AQP1 an ibodies may be indica i e o an unde lying au oimmune disease; bu mo e han associa ed o pa icula aspec s o any o hese diseases, i appea s as an epiphenomenon o a mo e di use immunological dys unc ion. Ne e heless, u he analysis wi h he ELISA, o o he immune assays, especially hose wi h de ec ion o he p o ein in suspension, accompanied wi h mo e analysis including la ge numbe o pa ien s on each subg oup o hese di e se pa hologies as well as mo e cases o he NMOSD would be necessa y o s eng h ou incipien hypo hesis. Figu e 4. De ec ion by ELISA assay o an i-AQP1 an ibodies in se um o pa ien s. Se ums om di e en pa hologies acco ding o he classi ica ion shown in Table 1 we e analyzed including a g oup wi h he NMO spec um diso de (+NMOSD). Abb e ia ions o pa hologies a e as indica ed in Table 1. Insigni ican di e ences we e ob ained among g oups by analysis o a iance o nonpa ame ic da a using he K uskal–Wallis es (p = 0.067). Medians and in e qua ile ange a e ep esen ed and he numbe o pa ien s analyzed pe g oup we e as ollows: 8 NMOSD; 6 ON . (Op ic neu i is ecu en ); 2 Sjö gen; 9 LETM (Longi udinally ex ensi e ans e se myeli is); 3 Myeli is epe i ion; 18 MS (Mul iple scle osis); 5 ON (Op ic neu i is); 4 Con ol (Heal hy). Ci cle, co espond o an ou side da a poin . Figu e 4. De ec ion by ELISA assay o an i-AQP1 an ibodies in se um o pa ien s. Se ums om di e en pa hologies acco ding o he classi ica ion shown in Table 1we e analyzed including a g oup wi h he NMO spec um diso de (+NMOSD). Abb e ia ions o pa hologies a e as indica ed in Table 1. Insigni ican di e ences we e ob ained among g oups by analysis o a iance o nonpa ame ic da a using he K uskal–Wallis es (p= 0.067). Medians and in e qua ile ange a e ep esen ed and he numbe o pa ien s analyzed pe g oup we e as ollows: 8 NMOSD; 6 ON . (Op ic neu i is ecu en ); 2 Sjö gen; 9 LETM (Longi udinally ex ensi e ans e se myeli is); 3 Myeli is epe i ion; 18 MS (Mul iple scle osis); 5 ON (Op ic neu i is); 4 Con ol (Heal hy). Ci cle, co espond o an ou side da a poin . In . J. Mol. Sci. 2016,17, 1195 7 o 11 3. Ma e ials and Me hods 3.1. Subjec s and Se um Recollec ion The s udy includes 205 subjec s (135 emale) classi ied in o 7 g oups based on hei medical diagnosis (Table 1). G oup 1 comp ised 8 pa ien s wi h NMOSD acco ding o he cu en diagnos ic c i e ia [ 12 ]. G oup 2 consis ed o 94 pa ien s wi h MS acco ding o 2010- e iewed McDonald c i e ia [ 24 ]. In ela ion o he MS cou se, we u he di ided his g oup in o 85 pa ien s wi h emi ing- elapsing MS, 7 wi h p ima y p og essi e MS and 2 wi h seconda y p og essi e o m. G oup 3 was composed o 39 pa ien s wi h op ica neu i is. Twen y-nine pa ien s wi h myeli is o med g oup 4. Conce ning o he leng h o he spinal co d lesion, we dis inguished be ween pa ien s whose lesions ex ended mo e han 3 e eb al segmen s (longi udinally ex ensi e myeli is) and hose wi h spinal plaques ex ended up o 3 e eb al segmen s (10 and 16 pa ien s, espec i ely). Besides, we ca ego ized he pa ien s o he la e wo g oups depending whe he hey su e ed an isola ed episode o se e al elapses. G oup 5 comp ised 9 pa ien s wi h o he demyelina ing idiopa hic diso de s o he CNS and g oup 6 o med by subjec s wi h o he neu ological diso de s, bo h g oups desc ibed in Table 1. Besides he MRI (magne ic esonance imaging) scans and he ce eb ospinal luid analysis, pe iphe al blood exam ha included blood coun , biochemis y, e y h ocy e sedimen a ion a e, i amins, hy oid ho mones, long chain a y acids, angio ensin con e ing enzyme, immunological and se ological examina ions we e conduc ed in all pa ien s in o de o exclude al e na i e e iologies o hei inal diagnosis. Finally, g oup 7 was composed o 8 heal hy con ols. Table 1. Demog aphic and clinical a iables o 205 pa ien s. Diagnosis Numbe o Pa ien s (205) Gende Female/Male Mean Age a Inclusion ˘SD ( ange) AQP4+ An ibodies AQP1+ An ibodies 1. NMOSD 8 7/1 57.14 ˘13.52 (40–80) 6 0 2. MS 94 66/28 39.87 ˘11.84 (18–76) 0 0 * RRMS 85 59/26 0 0 * PPMS 7 5/2 0 0 * SPMS 2 2/0 0 0 3. Idiopa hic ON 39 27/12 39.55 ˘13.02 (14–68) 0 0 * Isola ed episode 30 22/8 0 0 * Recu en idiopa hic ON 9 5/4 0 0 4. Idiopa hic myeli is 29 19/10 45.13 ˘13.58 (21–69) 0 0 * Isola ed episode: 26 17/9 0 0 >3 e eb al segmen s 10 5/5 0 0 <3 e eb al segmen s 16 12/4 0 0 * Recu en idiopa hic myeli is 3 2/1 0 0 5. OIDD o he CNS 9 5/4 48.88 ˘10.37 (26–60) 0 0 * ADEM 2 0/2 0 0 * In a en o ial CIS 4 2/2 0 0 * RIS 3 3/0 0 0 6. O he neu ology diso de s 18 8/10 51.35 ˘12.79 (26–79) 0 0 * Myeli is associa ed wi h lupus 3 3/0 0 0 * Myeli is associa ed wi h sa coidosis 1 0/1 0 0 * ON associa ed wi h Sjög en synd ome 1 1/0 0 0 * Mul i ocal mo o neu opa hy 3 0/3 0 0 * CIDP 1 1/0 0 0 * He edi a y spas ic pa apa esis 1 1/0 0 0 * Spinal in ac ion 2 1/1 0 0 * Ischemic op ic neu opa hy 6 1/5 0 0 7. Heal hy con ols 8 6/2 36.42 ˘8.12 (27–47) 0 0 AQP: Aquapo in; CNS: cen al ne ous sys em; SD: S anda d de ia ion; NMOSD: Neu omyeli is op ica synd ome diso de ; ON: Op ic neu i is; MS: Mul iple scle osis; RRMS: Remi ing elapsing mul iple scle osis; SPMS: Seconda y p og essi e mul iple scle osis; PPMS: P ima y p og essi e mul iple scle osis; ADEM: Acu e dissemina ed encephalomyeli is; CIDP: Ch onic in lamma o y demyelina ing polyneu opa hy, CIS: Clinically isola ed synd ome, RIS: Radiological isola ed synd ome; OIDD: O he idiopa hic demyelina ing diso de s o he CNS. * co espond o u he di ision o he MS g oup. In . J. Mol. Sci. 2016,17, 1195 8 o 11 3.2. Plasmid Cons uc ion, Cell Cul u e and Cell T ans ec ion The pCMV6-AC-AQP1-GFP (human AQP1-GFP) was pu chased om O iGene Technologies Inc. (Rock ille, MD, USA) as a plasmid eady o use in mammalian cells and he human M23-AQP4 iso o m was ampli ied by PCR om he comme cial ec o pDNR-LIB cDNA (Taka a Bio Eu ope/Clon ech, Sain -Ge main-en-Laye, F ance) using speci ic p ime s and cloned in o pEGFP-N1 (Taka a), o ans ec ion and exp ession in he HEK293T cell line as p e iously desc ibed [ 20 ]. Bo h AQP-EGFP cons uc s allow syn hesis o ei he AQP1 o AQP4 usion luo escen p o eins wi h he enhanced-g een- luo escen -p o ein-EGFP bound o i s ca boxyl ends. HEK293T cells we e cul u ed in DMEM wi h 10% FCS and 1% penicillin/s ep omycin (37 ˝ C, 5% CO 2 ). Cells (2 ˆ 10 5 ) we e seeded in 35 mm dishes o ans ec ion wi h Lipo ec amine 2000 (In i ogen, Ca lsbad, CA, USA) as desc ibed be o e [ 21 , 25 ].T ans ec ed cells we e main ained o 25–30 passages un il luo escence signal o AQPs exp ession dec ease below 90%. 3.3. Immuno luo escence Assay Based in a me hod desc ibed p e iously we e alua ed p esence o AQP1- and AQP4-Abs in he se um o pa ien s using a p o ocol ha we called basic assay [ 21 ]. The p o ocol combines exp ession o a usion g een luo escen p o ein (AQP1- o AQP4-EGFP) wi h he use o a ed luo escen goa an i-human seconda y an ibody ha , by dual labeling (g een and ed luo escence), cons i u es a me hod wi h ex emely high sensi i i y and speci ici y o iden i y posi i e pa ien s o any o hose an ibodies. B ie ly, 24 h be o e s a ing he immune-assay, HEK293T pla ed cells a abou 80% o con luence we e ans ec ed wi h ei he AQP1- o AQP4-EGFP cons uc s. Then wo di e en p e ea men s we e pe o med sepa a ely, one ha we called “pe meabiliza ion” p o ocol, and ano he one called “wi hou pe meabiliza ion”, equi alen o he o iginal p o ocol desc ibed ini ially [ 21 ]. In he pe meabiliza ion p o ocol he cells we e ixed wi h pa a o maldehyde 4% (5 min) and hen washed wi h i on X-100 2% (SIGMA, S . Louis, MO, USA) in PBS (PBTx, 10 min), be o e incuba ing (1 h) in FCS 10% wi h 1 mg/mL BSA in PBTx o he blocking s ep. A e wa d, incuba ion wi h he se um’s pa ien (1:10 dilu ion o de ec ion o an i-AQP1 and 1:50 dilu ion o de ec ion o an i-AQP4, 1 h a oom empe a u e) o wi h an an i-Aquapo in 1 an ibody (ab 117970, ABCAM, Camb idge, UK) (1:500 dilu ion) aised agains a ull leng h ecombinan human Aquapo in 1 p oduced in HEK293T cells, was ollowed by h ee imes washes wi h PBS and hen 30 min o incuba ion wi h Alexa Fluo 568 goa an i-human seconda y an ibody (In i ogen, Ca lsbad, CA, USA). In he me hod in which he cells we e no pe meabilized, he T i on X-100 was emo ed om each s ep. So, a e ixing he cells wi h pa a o maldehyde 4% (5 min), a washing s ep in PBS o 5 min was ollowed o a blocking pe iod (1 h) in 10% FCS wi h 1 mg/mL BSA in PBS, and inally he cells we e ixed (1 min), a e he seconda y an ibody incuba ion, wi h a mix u e o e hanol 95% and ace ic acid 5%. Nuclei we e s ained wi h 4 1 ,6 1 -diamidino-2-phenylindole (DAPI, 1:1000) and a Leica DM IRBE con ocal mic oscope (Leica, We zla , Ge many) was used o obse e he slides. Fi e pho os (40 ˆ ) pe sample we e andomly aken and he NIH ImageJ so wa e (NIH, Be hesda, MD, USA) used o densi ome y analysis o luo escence. 3.4. ELISA o AQP1 (Aquapo in-1) 3.4.1. P epa a ion o AQP1 P o ein Homogena e HEK293T cells we e ans ec ed wi h a pcDNA3-AQP1 o exp ession o human AQP1 p o ein as p e iously used [ 26 , 27 ]. A e 24–48 h o ans ec ion cells we e washed wi h cold PBS and ea ed wi h ypsin 0.25% (GIBCO, Paisley, UK) o collec ion. The cell pelle was esuspended in 1 mL o cold PBS and cen i uged a 300 ˆ g o 5 min a 4 ˝ C. Fo whole-cell p o ein ex ac , pelle was dissol ed in 500 µ L o lysis bu e : 137 mM NaCl, 20 mM T is (pH: 8); 1% IGEPAL-CA630 (Sigma Ald ich, S . Louis, MO, USA), a nonionic, non-dena u ing de e gen ; 10% Glyce ol and 10 µ L/mL o comple e p o ease inhibi o s cock ail (Sigma Ald ich). The homogena e was le on ice 15 min, o ex, and hen In . J. Mol. Sci. 2016,17, 1195 9 o 11 cen i uged a 16,000 ˆ g o 15 min a 4 ˝ C, and ex ac ed p o eins emain in he supe na an . P o ein concen a ion was analyzed wi h he B ad o d me hod (BioRad P o ein Assay, BioRad, Be keley, CA, USA) and kep a ´20 ˝C un il loading in o pla es o ELISA assay. 3.4.2. Adhesion o AQP1 P o ein o ELISA Assay Gene al guidelines o ELISA assay ha e been desc ibed elsewhe e [ 28 ]. P o eins p epa ed as be o e we e dilu ed a 20 µ g/mL inal concen a ion in 0.01 M bu e ca bona e and 50 µ L pe well o p o ein suspension we e loaded in o a 96 well pla e o ELISA (Mic owell MaxiSo p, Nunc, Wal ham, MA, USA), a e wa ds he pla e was co e ed wi h a plas ic ilm and le o e nigh a 4 ˝ C. The nex day he solu ion was emo ed and he pla e washed h ee imes by illing he wells wi h 200 µ L PBS1X + 0.05% Tween and once wi h PBS1X. Blocking: To block he emaining p o ein-binding si es in he coa ed wells 200 µ L o Supe Block Blocking Bu e (The moScien i ic, Van aa, Finland) we e added pe well and incuba ed a oom empe a u e o 1 h, main aining he pla e co e wi h plas ic ilm. Then, blocking solu ion was emo ed and he pla e was washed h ee imes by illing he wells again wi h 200 µ L PBS1X + 0.05% Tween and once wi h PBS1X. 3.4.3. Incuba ion wi h P ima y and Seconda y An ibodies Two p ima y an ibodies, 100 µ L pe well, we e used; a comme cial an ibody an i-AQP1 (ab15080, ABCAM) dilu ed 1:10,000 in PBS wi h 2% BSA, ha se es as a con ol o se he assay condi ions, and he pa ien se ums wi hou dilu ion. The incuba ion was allowed o p oceed o e nigh a 4 ˝ C and he nex day pla es we e washed as indica ed o emo ing he blocking solu ion men ioned abo e. Then, incuba ion wi h he seconda y an ibodies o 1 h a oom empe a u e was ca ied ou . Ho se adish pe oxidase conjuga ed goa an i- abbi IgG an ibody dilu ed (1:5000) in PBS wi h 2% BSA o he AQP1 comme cial an ibody, and ho se adish pe oxidase conjuga ed chicken an i-human IgG an ibody o he pa ien se um an ibodies we e used. Wash o pla es a he end was again ca ied ou as be o e. 3.4.4. Signal De ec ion: Pe Well, 100 µL o 3,31,5,51-Te ame hylbenzidine (TMB) TMBOne solu ion (P omega, Madison, WI, USA) was added and incuba ed a oom empe a u e o 15 min o allow enzyma ic eac ion and de eloping o colo ed subs a e. Then, 100 µ L o HCl 1N we e added pe well o s op he eac ion and abso bance a 450 nm was measu ed in a pla e eade sys em (Mul iskan Spec um-The mo, Van aa, Finland). 3.5. S a is ical Analysis Da a a e p esen ed as mean ˘ s anda d e o o he mean, and analyzed using he S a is ical Package o Social Sciences (SPSS Inc., Chicago, IL, USA), e sion 19.0. Da a wi h a non-no mal dis ibu ion we e analyzed using analysis o a iance (ANOVA) o non-pa ame ic da a wi h he K uskal–Wallis H es . 4. Conclusions Ou s udy does no show sus ained de ec ion o an i-AQP1 in se um o NMOSD pa ien s analyzed by ou ixed cell based assay o ELISA p o ocol. To ou unde s anding, hese an ibodies do no seem o allow con i ma ion o speci ic immune diso de s associa ed wi h NMOSD. Acknowledgmen s: G an s om “La Jun a de Andalucía, Conseje ía de Inno ación Ciencia y Emp esa” (P08-CTS-03574) and Conseje ía de Salud (PI0298-2010), and om he “Ins i u o de Salud Ca los III” (Exp. PI12/01882) o Mi iam Eche a ía unded his wo k. We hank Genzyme Founda ion in mul iple scle osis o gi ing o Mi iam Eche a ía one o hei 2012 ellowships. We hank Juan Manuel P aena o his help wi h da a plo ing and s a is ical analysis, and Robe o Ga cia Swinbu n o his help wi h English co ec ion o he manusc ip .