1
Scien i ic RepoR s | 5:12910 | DOi: 10.1038/s ep12910
www.na u e.com/scien i ic epo s
Human iPSC de i ed disease
model o MERTK-associa ed
e ini is pigmen osa
Dunja Luko ic1, Ana A e o Cas o1, Ana Belen Ga cia Delgado1,
Ma ía de los Angeles Ma ín Be nal1, Noelia Luna Pelaez1, And ea Díez Llo e 1, Rocío Pe ez
Espejo1, Kunka Kamena o a1, Lau a Fe nández Sánchez3, Nicolás Cuenca3, Ma a Co ón4,5,
Almudena A ila Fe nandez4,5, Anni So kio6, Heli Sko man6, Ca men Ayuso4,5,
Sla en E ceg1,2 & Shomi S. Bha acha ya1
Re ini is pigmen osa (RP) ep esen s a gene ically he e ogeneous g oup o e inal dys ophies
a ec ing mainly he od pho o ecep o s and in some ins ances also he e inal pigmen epi helium
(RPE) cells o he e ina. Clinical symp oms and disease p og ession leading o mode a e o se e e
loss o ision a e well es ablished and despi e signi ican p og ess in he iden i ica ion o causa i e
genes, he disease pa hology emains unclea . Lack o his unde s anding has so a hinde ed
de elopmen o e ec i e he apies. He e we epo success ul gene a ion o human induced
plu ipo en s em cells (iPSC) om skin ib oblas s o a pa ien ha bo ing a no el Se 331Cys s*5
mu a ion in he MERTK gene. The pa ien was diagnosed wi h an ea ly onse and se e e o m o
au osomal ecessi e RP (a RP). Upon di e en ia ion o hese iPSC owa ds RPE, pa ien -speci ic RPE
cells exhibi ed de ec i e phagocy osis, a cha ac e is ic pheno ype o MERTK de iciency obse ed
in human pa ien s and animal models. Thus we ha e c ea ed a ai h ul cellula model o a RP
inco po a ing he human gene ic backg ound which will allow us o in es iga e in de ail he disease
mechanism, explo e sc eening o a a ie y o he apeu ic compounds/ eagen s and design ei he
combined cell and gene- based he apies o independen app oaches.
Re ini is pigmen osa (RP; OMIM 268000) wi h a p e alence o 1 in 3,500 indi iduals is he mos com-
mon o m o he edi a y e inal diso de a ec ing he wo king age g oup. RP is cha ac e ized by p o-
g essi e dys unc ion and dea h o mainly he od pho o ecep o cells (PR) o he e ina howe e in some
cases e inal pigmen epi helium (RPE) cells a e also in ol ed, o en esul ing in pe manen blindness. So
a 54 genes ha e been implica ed in his disease coding o p o eins in ol ed in a my iad o unc ions
such as pho o ansduc ion signaling cascade, e inoid cycle, cell-cell adhesion o he cy oskele on1. The
disease is inhe i ed in all he e Mendelian o ms, he au osomal ecessi e (a RP) being he mos common
wi h o e 50% o cases. La gely due o he high gene ic he e ogenei y and una ailabili y o disease issue,
pa hology o he disease emains elusi e.
Pa ien -de i ed induced plu ipo en s em cells (iPSCs) p o ide an unp eceden ed oppo uni y o eca-
pi ula e disease pa hogenici y wi hou he need o gene ic manipula ion and c ea ion o gene a ge ed
1CABIMER (Cen o Andaluz de Biología Molecula y Medicina Regene a i a), A da. Ame ico Vespucio s/n, Pa que
Cien í ico y Tecnológico Ca uja, 41092, Se illa, Spain. 2S em Cell he apies in Neu odegene a i e diseases Lab,
and Na ional S em Cell Bank-Valencia Node, Biomolecula and Bioin o ma ics Resou ces Pla o m PRB2,ISCIII,
Resea ch Cen e “P incipe Felipe”, c/ Edua do P imo Yú e a 3, 46012, Valencia, Spain. 3Depa men o Physiology,
Gene ics and Mic obiology, Uni e si y o Alican e, Campus de San Vicen e del Raspeig, 0369 Alican e, Spain.
4Depa men o Gene ics and Genomics, IIS-Fundación Jiménez Díaz, 28040 Mad id, Spain. 5Cen e o Biomedical
Ne wo k Resea ch on Ra e Diseases (CIBERER), ISCIII, Mad id, Spain. 6BioMediTech, Uni e si y o Tampe e,
Bioka u 12, 33520 Tampe e, Finland. Co espondence and eques s o ma e ials should be add essed o S.S.B.
(email: shomi.bha acha ya@cabime .es) o S.E. (email: se ceg@cip .es)
Recei ed: 15 May 2015
Accep ed: 10 July 2015
Published: 11 Augus 2015
OPEN
www.na u e.com/scien i ic epo s/
2
Scien i ic RepoR s | 5:12910 | DOi: 10.1038/s ep12910
animal models. Human iPSCs, simila o emb yonic s em cells (ESC), can be expanded inde ini ely in
i o and di e en ia ed in o any ype o ma u e cell in he human body, wi hou he e hical and immu-
nogenici y issues associa ed wi h ESC2. These cells a e also aluable o de eloping he apeu ic s a egies,
d ug oxici y sc eens and de elopmen o disease models, in addi ion o p o iding a sou ce o cell
ansplan a ion he apy. RPE cells and pho o ecep o s (PR) ha e been success ully gene a ed om iPSCs
(iPSC-RPE and iPSC-PR espec i ely) by a ious g oups in s epwise di e en ia ion p o ocols mimicking
e inal de elopmen by in oducing Wn signaling inhibi o s (DKK1), Nodal an agonis Le y A, No ch
pa hway inhibi o (DAPT-gamma sec e ase inhibi o ), o IGF-13–5. In con as , only RPE cells ha e been
gene a ed spon aneously in o e g own iPSC/ESC cul u es wi hou he addi ion o exogenous ac o s,
since de i a i es o neu oec ode m appea by de aul in non-induced cul u es6,7. Gene a ed RPE cells in
hese s udies display a ully ma u e pheno ype and physiological ac i i y in i o such as phagocy osis,
sec e ion o ascula endo helial g ow h ac o (VEGF) and pigmen epi helium-de i ed ac o (PEDF)
and epi helial ba ie o ma ion. Cellula models o he edi a y e inal dys ophies ha e been success ully
c ea ed in i o in Bes disease, and RP whe e pa ien s’ ib oblas s we e ep og ammed o iPSC and
hen con e ed o RPE8,9 o pho o ecep o cells10, exp essing he disease pheno ype. iPSC- de i ed RPE
(iPSC-RPE) cells ha e also been shown o ha e a p o ec i e e ec when injec ed sub- e inaly in o he Royal
College o Su geons (RCS) a s11 and RPE65-de ec i e mice12. Mo eo e , iPSCs ha e me clinical-g ade
equi emen s13 as a sou ce o RPE g a s and ha e ecen ly been injec ed in pa ien s a ec ed by he
exuda i e o m (we - ype) o age- ela ed macula degene a ion (AMD)14. I has been a gued ha in his
o m o AMD he dys unc ion and loss o RPE cells is he main cause o isual impai men in he elde ly.
Me y osine kinase ecep o (MERTK) belongs o he Ty o3/Axl/Me (TAM) ecep o y osine kinase
amily o p o eins dis inguished by a conse ed in acellula kinase domain and ex acellula adhesion
molecule-like domain. TAM ecep o s egula e a a ie y o p ocesses such as cell p oli e a ion/su i al,
adhesion, mig a ion, in lamma o y esponse, in a cell- mic oen i onmen - and ligand- speci ic manne 15.
In p e ious s udies MERTK was ound o be dis up ed in RCS a s16,17, a classic model o e inal degen-
e a ion inhe i ed as an au osomal ecessi e ai , and ound o cause ea ly- onse e ini is pigmen osa in
pa ien s18. RPE cells ail o phagocy ize he shed ou e segmen (OS) ma e ial o PR, a ci cadian ac i -
i y pe o med by RPE cells which se es o enew he damaged lipid and p o ein componen s o ligh
exposed PR, while new memb anous discs a e o med (disc biogenesis) and inse ed in he basal pa o
he OS. As a esul , RCS a s exhibi OS associa ed deb is accumula ion in he sub e inal space, abno mal
OS leng h, e en ually leading o he onse o PR degene a ion by he P20 s age. Usually comple e degen-
e a ion occu s by P60. Simila pheno ype is obse ed in me kd mice19 indica ing ha he RPE phagocy ic
de ec is he unde lying molecula mechanism o disease in humans ca ying MERTK mu a ions. Indeed,
he educed e inal hickness and deb is de ec ed in he sub- e inal space in pa ien s ha bo ing he
MERTK –splice-si e -mu a ion esembles he obse ed pheno ype in he RCS a 20. The dis inc i e clini-
cal p esen a ion o RP is he only disease mani es a ion o pa ien s ha bo ing MERTK mu a ions wi hou
any sys emic disease o de ec s o phagocy osis by mac ophages, indica ing a specialized unc ion o his
p o ein in he RPE cells. In con as o he de ailed clinical unde s anding o he disease, he mechanism
by which MERTK ac s du ing he phagocy osis emains pa ially un eiled. Ou e segmen s a e known
o bind o he in eg in ecep o α β 521 ollowed by ocal adhesion kinase (FAK) ac i a ion in he apical
memb ane o RPE22 while he MERTK ac i a ion occu s ia Gas6/P o ein S, TUB, TULP1 ligand bind-
ing23,24. The la e is hough o ac i a e au ophospho yla ion a y osine Y-749, Y-753 and Y-754 in he
y osine kinase domain, which in u n ac i a es he molecula cascade a ge ing ac in o non-muscle
myosin II o coo dina e he cy oskele al ea angemen s necessa y o phagocy ic inges ion25.
In his s udy we desc ibe he gene a ion o an in i o model o RP caused by he mu a ion o MERTK
gene using human iPSC echnology. We epo he gene a ion o a cellula model o MERTK-associa ed
RP, which ecapi ula es he disease pheno ype desc ibed in animals and in pa ien s and p o ides a ool
o ad ance ou unde s anding o he signaling pa hways ac i a ed by MERTK. Impo an ly his can be
s udied in de ail in he con ex o he human gene ic backg ound, a c ucial ea u e o he cellula model.
Equally impo an is ha i p o ides us wi h a unique oppo uni y o he es ing o no el he apeu ic
agen s in he p esence o he human backg ound, which has he po en ial o signi ican ly educe he cos
o p eclinical s udies in animal models. The de i a ion o iPSCs om a RP pa ien ’s ib oblas s and he
gene a ion o a RPE cell line exp essing he molecula de ec makes an impo an con ibu ion in ou
abili y o unde ake ansla ional esea ch.
Resul s
RP caused by MERTK gene mu a ion. Skin punch biopsies we e aken om a pheno ypically
well-cha ac e ized RP pa ien and an una ec ed indi idual a e he signing o in o med consen . Fo y-
i e- yea -old pa ien bo n o a consanguineous ma iage (pa ien II:2; amily RP-0503, Supplemen a y
Fig. S1A) was diagnosed wi h ea ly-onse au osomal ecessi e RP (a RP). Genome-wide linkage anal-
ysis using he Illumina HumanCy o-12 SNP a ay ollowed by homozygosi y mapping was pe o med
in his Spanish amily wi h 3 a ec ed and 2 una ec ed siblings ha e ealed a la ge egion o 32 Mb
wi h a maximum loga i hm o he odds (LOD) sco e o 3.6 on ch omosome 2 encompassing MERTK.
Sequence analysis allowed iden i ica ion o a no el homozygous ameshi mu a ion c.992_993delCA
(p.Se 331Cys s*5) in MERTK, seg ega ing wi h he disease in he amily (Supplemen a y Fig. S1A).
www.na u e.com/scien i ic epo s/
3
Scien i ic RepoR s | 5:12910 | DOi: 10.1038/s ep12910
All a ec ed membe s o his amily we e diagnosed su e ing om RP du ing he i s decade o li e.
Pa ien II:2 complained o nigh blindness, loss o pe iphe al isual ield and isual acui y (VA) a he age
o 7. Oph halmological examina ion a 32 yea s o age showed se e e cons ic ion o isual ields in bo h
eyes and loss o VA o 20/70 in igh eye and inge coun ing in he le eye. The undus showed ypical
changes o RP wi h pale disc, na owed e inal essels and bone-spicule pigmen a ions in he pe iphe al
e ina. Fo his s udy, a heal hy indi idual un ela ed o his amily and no amily his o y o blindness was
aken as a heal hy con ol. The heal hy indi idual had a VA o 20/20 a he ime o skin biopsy.
P ima y ib oblas cell lines we e es ablished om he pa ien II:2 and heal hy indi idual in pa -
allel and subjec ed o di ec genomic sequencing o con i m he geno ype (Fig. 1A). The dele ion
c.992_993delCA was con i med in he pa ien ’s DNA. The MERTK gene is composed o 19 coding exons
wi h wo immunoglobulin-like C2- ype (Ig-like) domains, wo ib onec in ype III (FN3)-like domains,
a ansmemb ane (TM) domain and a y osine kinase (TK) domain (UniP o Q12866). The dele ion o
CA in exon 7 (NM_006343.2), a posi ion 992_993 causes he ameshi esul ing in a p ema u e s op
codon a posi ion 335 (Supplemen a y Fig. S1B). The p edic i e s op codon esul s in a unca ed p o ein
wi h 2 Ig-like domains and pa o he i s FN3-like domain wi h no kinase ac i i y (Fig.1B).
Gene a ion o iPSC om heal hy indi idual and RP pa ien . The p ima y ib oblas s de i ed
om he skin biopsies we e expanded o h ee o- i e passages be o e being ep og ammed by a Sendai
i al cons uc con aining ou genes: Oc 3/4, Sox2, Kl 4 and cMyc (Supplemen a y Fig. S2A). This con-
s uc desc ibed as highly e icien in ansduc ion was selec ed o being non-in eg a i e and eplica ion
de icien . Thi y days a e ansduc ion, iPSC colonies we e selec ed by hei mo phology ( e ac i e
edges, high nuclea /cy oplasmic a io) and in si u s aining o plu ipo ency ma ke s (TRA-1-60 o TRA-
1-81). Th ee iPSC lines pe indi idual we e sub-cul u ed and analyzed a cellula and gene ic le el o
con i m success ul ep og amming. Plu ipo ency was assessed by immunocy ochemis y o plu ipo-
ency ma ke s such as oc ame -binding ansc ip ion ac o 4 (OCT4), NANOG, SRY (sex de e mining
egion Y)-box 2 (SOX2), and in si u s aining o TRA-1-81. The alkaline phospha ase is known o be mo e
ac i e in iPSCs and he colo ime ic assay depic ing i s ac i i y con i med ha he selec ed iPSC colonies
a e indeed plu ipo en (Supplemen a y Fig. S3A). To es he abili y o gene a ed iPSC lines o gene a e
de i a es o h ee ge m laye s in i o, he iPSCs we e ansplan ed subcu aneously in o he immunode-
icien (SCID) mice. Eigh weeks a e injec ion, umo was o med and ex ac ed. His ological sec ions
showed ha he umo con ained de i a i es o all h ee ge m laye s including gu -like issues (endo-
de m), neu al cells and e ina (ec ode m) and s ia ed muscle, bone, ca ilage and adipose issue (mes-
ode m) (Supplemen a y Fig. S3B). The heal hy indi idual’s iPSCs we e indis inguishable om pa ien ’s
iPSCs wi h espec o plu ipo ency cha ac e iza ion. All selec ed lines we e ansgene- ee a passages
7–10 and ka yo ypically no mal o e 30 passages (Supplemen a y Fig. S2B and C, espec i ely). DNA
Figu e 1. (A). MERTK DNA sequence ch oma og am a ound c.992_993 in heal hy and a ec ed pa ien
ca ying a dele ion o CA (c.992_993delCA). DNA sequencing, pe o med on heal hy and pa ien ’s de mal
ib oblas s and iPSCs’ gDNA, con i med he mu a ion in ea ly-onse RP pa ien . (B). The wild ype sequence
codes o 999 amino acid p o ein wi h Immunoglobulin Ig-like C2- ype (Ig-like) ( esidues 100–194 and
203–281), ib onec in ype III-like (FN3) ( esidues 284–368 and 384–470) ansmemb ane (TM) ( esidues
502–524) and kinase domain ( esidues 587–854). The mu a ion esul s in a p edic i e unca ed p o ein o
334 amino acids wi h expec ed molecula weigh o 36.6 kDa.
www.na u e.com/scien i ic epo s/
4
Scien i ic RepoR s | 5:12910 | DOi: 10.1038/s ep12910
inge p in ing was pe o med wi h iPSC lines and p o ed hei gene ic iden i y o pa en al ib oblas s
(Supplemen a y Fig. S2D).
Di e en ia ion o iPSCs in o RPE cells. We induced pa ien ’s and he heal hy indi idual’s iPSCs
o di e en ia e owa d RPE cells in pa allel, by emo ing he plu ipo ency ac o bFGF and cul u ing
he cells in suspension un il da k pa ches appea ed in he loa ing agg ega es as p e iously desc ibed6
(Supplemen a y Fig. S4). The pigmen ed a eas o he agg ega es we e mechanically excised, ypsinized
and pla ed as a monolaye cell cul u e. A e 3–4 weeks in cul u e he iPSC-de i ed RPE (iPSC-RPE) cells
appea ed o exhibi polygonal, cobbles one-like mo phology, which is cha ac e is ic o ma u e RPE cells.
The a eas wi h co ec mo phology we e li ed and eseeded on pe meable cul u e inse s o yield a uni-
o m cellula monolaye . To each ull unc ional ma u i y, he cells we e cul u ed o addi ional 30–80
days un il high pigmen a ion. Figu e 2A,B depic s he ma u e iPSC-RPE cells om bo h indi iduals
showing indis inguishable polygonal mo phology and pigmen a ion. Toluidine blue s aining (Fig.2C,D)
Figu e 2. Di e en ia ion o heal hy and MERTK p.Se 331Cys s*5 iPSCs in o RPE cells. B igh ield
mic og aph o heal hy (A) and MERTK p.Se 331Cys s*5 (B) iPSC-RPE showing ypical RPE mo phology,
polygonal shape and pigmen a ion. Semi hin sec ions o iPSC-RPE cells s ained wi h oluidine blue. Bo h
heal hy (C) and MERTK p.Se 331Cys s*5 (D) iPSC-RPE cells o m a monolaye o cuboid cells highly
pola ized wi h abundan apical mic o illi (a ows) and melanosomes (a owheads). Elec on mic og aph o
cul u ed iPSC-RPE, heal hy (E) and MERTK p.Se 331Cys s*5 (F,G,H,I). (G) High magni ica ion o melanin
g anules showing di e en s ages o de elopmen . (H) In e cellula junc ional complexes which include he
igh junc ions, adhe ens junc ions (a ows) and memb ane in e digi a ions (a owheads). (I) The basemen
memb ane o he cells igh ly bound o he answells suppo ilm (a owheads) and small basal in oldings
(a ows). A,B scale ba 50µm; C,D scale ba 10µm.
www.na u e.com/scien i ic epo s/
5
Scien i ic RepoR s | 5:12910 | DOi: 10.1038/s ep12910
and ansmission elec on mic oscopy o he iPSC-RPE cells (Fig.2E,F) show ha he apical su ace o
he cell monolaye is a he op, whe e abundan mic o illi can be obse ed. Melanosomes a e ep e-
sen ed wi h black ound and o al shapes (Fig. 2G) and ha e apical dis ibu ion. Mi ochond ia ha a e
ellipsoidal in shape and seen below he nuclei a e la gely displaced owa d he basal-la e al pa o he
cells (Fig. 2E,F a owheads), he na u al posi ion o hese o ganelles in i o. In e cellula junc ional
complexes a e isible in app op ia ely aligned sec ions o iPSC-RPE o bo h indi iduals. The in eg i y
and unc ion o he RPE monolaye depend on i s basola e al cellula junc ions, which include he igh
junc ions, adhe ens junc ions and memb ane in e digi a ions (Fig.2H). The basemen memb ane o he
cells is igh ly bound o he answells suppo ilm and small basal in oldings we e obse ed (Fig.2I).
To alida e he iden i y o gene a ed RPE cells, RT-PCR o selec ed RPE speci ic genes was pe o med
o compa e he exp ession le els h oughou he di e en ia ion p ocess (Fig.3A,i)). This included ma k-
e s such as cellula e inaldehyde-binding p o ein (CRALBP), Bes ophin 1 (BEST1), e inal pigmen
epi helium-speci ic p o ein 65kDa (RPE65), MERTK, and plu ipo ency ma ke NANOG. The exp ession
p o ile was compa ed in heal hy and MERTK p.Se 331Cys s*5 pa ien ’s ib oblas s, wo iPSCs lines and
hei espec i e iPSC-RPE. Figu e 3A,i) shows he expec ed pa e n o exp ession o RPE ansc ip s
o BEST1, CRALBP and RPE65 wi h compa able s ong exp ession le els in iPSC-RPE om bo h indi-
iduals while, as expec ed, being absen om iPSCs and ib oblas s. BEST1 was an excep ion, which
shows a weak exp ession in ib oblas s. NANOG, a plu ipo ency ma ke , is exp essed in iPSCs bu
silenced a e he cells di e en ia e in o RPE. The da a desc ibed in Fig. 3A (i) we e co obo a ed by
quan i a i e analyses (Fig.3A(ii)). The exp ession o RPE speci ic genes CRALBP, RPE65, BEST1 and
mic oph halmia-associa ed ansc ip ion ac o (MITF) a e aised o e a 1000 old in iPSC-RPE cells
om bo h indi iduals wi h espec o he ib oblas s, while CRALBP and RPE65 a e jus de ec able by
qRT-PCR in iPSCs.
On checking exp ession o MERTK by RT-PCR using p ime s spanning exons 14–17 and also by
qRT-PCR assay using p ime s spanning exon junc ion 18–19 (Fig.3A, Supplemen a y Fig. S5) indica e
ha MERTK mRNA is de ec able in bo h heal hy and MERTK p.Se 331Cys s*5 iPSCs and ises o e
1000 imes wi h espec o he o iginal ib oblas s in iPSC-RPE o bo h indi iduals. MERTK exp ession
in iPSC has no been p e iously desc ibed and is in co ela ion wi h i s exp ession in human ESC7.
The le el o exp ession o MERTK in heal hy indi idual was highe han in MERTK p.Se 331Cys s*5
pa ien . The di e ence in exp ession pa e n o MERTK in iPSC-RPE om heal hy indi idual and RP
pa ien is expec ed since i is known ha euka yo es possess a nonsense-media ed mRNA decay pa hway,
which deg ades mRNAs con aining nonsense mu a ions be o e hey a e ansla ed in o non unc ional
polypep ides26. The possibili y emains ha al e na i e ansc ip ha does no include he mu a ion
si e is ampli ied by he p ime s and p obe used in he assay. In he case ha RNA ansc ibed om
he c.992_993delCA allele escapes nonsense media ed decay, a unca ed p o ein o 334 amino acids
(expec ed size o 36,6 kDa) lacking he kinase domain would be p oduced. O e all, we can con i m ha
he gene a ed iPSC-RPE cell lines o igina ed om wo di e en indi iduals exhibi he ma u e RPE gene
exp ession pa e n and a e indis inguishable a his poin .
Immunocy ochemical s aining e ealed a s ong exp ession o he isual cycle ma ke s CRALBP, RPE65,
as well as BEST1 and zonula-occludens-1 (ZO-1), a igh -junc ion ma ke , indis inguishable in bo h
indi iduals (Fig.3B,i)). Apico-la e al and apical localiza ion o ZO-1 and sodium/po assium-dependen
ATPase (Na+/K+ ATPase), espec i ely, is a hallma k o RPE cells and was ound in bo h heal hy and
MERTK p.Se 331Cys s*5 iPSC-RPE cells (Supplemen a y Fig. S6).
MERTK exp ession in iPSC-RPE. Immunocy ochemical de ec ion o MERTK in iPSC-RPE mon-
olaye om heal hy indi idual showed i s exp ession apically as p e iously desc ibed27 (Fig.3B,ii)). In
con as , iPSC-RPE cells de i ed om he RP pa ien ca ying he mu a ion p.Se 331Cys s*5 a e com-
ple ely de icien in MERTK (Fig. 3B,ii)). The an ibody used de ec ed he N- e minal agmen o he
p o ein implying ha any unca ed p o ein e aining he N- e minal domain is absen in he iPSC-RPE
om he RP pa ien . This da a co obo a es he pa ien -speci ic o igin o gene a ed iPSC-RPE monolay-
e s and addi ionally con i ms he co ec pola iza ion o gene a ed RPE cells.
Exp ession o MERTK and ma u e RPE speci ic ma ke s was con i med by wes e n blo h ough-
ou he di e en ia ion p ocess o wo iPSC lines pe geno ype (Fig. 3C). The exp ession pa e n o
RPE-speci ic ma ke s is ma ched wi h human RPE (hRPE) ex ac om cada e ic dono by loading he
same o al p o ein amoun . I espec i e o he indi idual, whe he heal hy o RP pa ien , RPE-speci ic
ma ke s BEST1 and CRALBP a e absen om ib oblas s and iPSCs bu a e de ec ed in iPSC-RPE mon-
olaye s a simila le els. MERTK is de ec ed as a double band, which mig a es wi h molecula weigh o
180 and 130 kDa in heal hy indi idual’s monolaye and na i e RPE while being absen in iPSC-RPE om
RP pa ien ha bo ing he MERTK mu a ion. Human emb yonic kidney cells (HEK 293T) a e known o
exp ess MERTK28 as a single 180 kDa band while being nega i e o RPE ma ke s. The MERTK exp es-
sion is de ec ed also in iPSC om con ol indi idual upon highe p o ein load (Supplemen a y Fig. S7).
A 180 kDa band co esponding o MERTK is also desc ibed in human ESC and ESC-RPE7, suppo ing
he iew ha MERTK is exp essed in plu ipo en s em cells and co obo a ing he iden i y o he 180 kDa
band. The obse ed molecula weigh o MERTK by wes e n blo ing does no co espond o he p edic-
i e molecula weigh o 110 kDa o he 999 amino acid long MERTK. The p o ein was de ec ed a signi -
ican ly la ge sizes anging om 165–205 kDa in di e en cell ypes due o pos ansla ional modi ica ion

www.na u e.com/scien i ic epo s/
6
Scien i ic RepoR s | 5:12910 | DOi: 10.1038/s ep12910
Figu e 3. Fo legend see nex page.
www.na u e.com/scien i ic epo s/
7
Scien i ic RepoR s | 5:12910 | DOi: 10.1038/s ep12910
such as glycosya ion (14 glycosyla ion si es a e de ec ed in UniP o da abase), ubiqui ina ion o phos-
pho yla ion15,28. The wo bands co esponding o 180 and 130 kDa pep ides (Fig.3C), demons a e ha
hese wo iso o ms a e speci ic o RPE as con i med in hRPE ex ac as opposed o a single 180 kDa
band obse ed in HEK 293T cells. The wes e n blo con i ms he absence o MERTK and any unca ed
p o ein p ese ing he N- e minal domain (Supplemen a y Fig. S7) in RP-a ec ed pa ien co obo a ing
he iden i y o he pa ien -de i ed iPSC-RPE and indica ing ha he de ec i e, kinase lacking, MERTK
is p obably deg aded in pa ien ’s RPE.
Pa ien ’s iPSC ecapi ula e he disease pheno ype. MERTK de icien pheno ype desc ibed in
animal models is de ec i e phagocy osis, a daily p ocess o OS up ake, which enews he ligh –damaged
memb anes o PR. The p ocess o OS phagocy osis can be mimicked in cul u ed RPE cells by eeding he
isola ed OS de i ed om bo ine o pig e ina6,7. We exposed he cul u ed iPSC-RPE monolaye s om
bo h indi iduals o luo escen ly labeled bo ine and pig OS and moni o ed hei inges ion by con ocal
imaging. Figu e4 shows he in e naliza ion o OS by he iPSC-RPE om he heal hy indi idual while
Figu e 3. RPE cha ac e is ic ma ke gene and p o ein exp ession in iPSC-RPE om heal hy indi idual
and RP pa ien . (A. i) BEST1, CRALBP, RPE65, MERTK gene exp ession in ib oblas s, iPSCs and iPSC-
RPE. Two iPSC lines (iPSCc1 and iPSCc2) and hei espec i e iPSC-RPE om each indi idual a e analyzed.
The samples we e loaded on aga ose gels p epa ed and un unde he same expe imen al condi ions.
(ii) Quan i a i e e e se ansc ip ion polyme ase chain eac ion o analyze exp ession o genes cha ac e is ic
o RPE. Fold exp ession in undi e en ia ed iPSCs and iPSC-RPE om heal hy and MERTK p.Se 331Cys s*5
pa ien no malized o hei o igina ing de mal ib oblas s (DF). Each ba ep esen s he a e age ± SEM
o a leas h ee independen biological eplica es (* p ≤ 0.05, **p ≤ 0.05, ***p ≤ 0.05). (B. i) Exp ession o
RPE cha ac e is ic p o eins by iPSC-RPE. Immunocy ochemis y agains RPE65, CRALBP, BEST1, ZO-1 in
heal hy and MERTK p.Se 331Cys s*5 iPSC-RPE. Images we e aken wi h Leica con ocal mic oscope TCS
SP5 using HCX PL APO lambda blue 63X/ 1.4 OIL objec i e. Scale ba 10µm. (ii) MERTK exp ession in
iPSC-RPE. Immunocy ochemis y agains MERTK. Apical sec ion and e ical sec ion simula ion showing
apical MERTK dis ibu ion in he heal hy indi idual. MERTK p.Se 331Cys s*5 iPSC-RPE s ain nega i e o
MERTK a all sec ions. Images we e aken wi h Leica con ocal mic oscope TCS SP5 using HCX PL APO
lambda blue 63X/ 1.4 OIL objec i e, scale ba 10 μ m. (C). Wes e n blo analyses o RPE-speci ic ma ke
p o ein exp ession in ib oblas s (DF), wo iPSC lines (iPSCc1 and iPSCc2) and hei espec i e iPSC-RPE
om each indi idual a e analyzed. The exp ession o CRALBP and BEST1 is de ec ed only in iPSC- RPE
and hRPE cells. β ac in was used as loading con ol. C opped blo s a e om gels un unde he same
expe imen al condi ions and loaded wi h he same samples.
Figu e 4. In i o phagocy osis assay o pho o ecep o ou e segmen s (OS) by iPS-RPE om heal hy
indi idual and MERTK p.Se 331Cys s*5 pa ien . Basola e al sec ions ac oss he iPSC-RPE cells oge he
wi h e ical sec ion simula ion. Heal hy iPS-RPE in e nalize FITC-labeled OS (g een) while pa ien ’s one
do no . F-ac in is s ained by phalloidin ( ed) o isualize cell mo phology. Images we e aken wi h Leica
con ocal mic oscope TCS SP8 using HCX PL APO lambda blue 63X/ 1.4 OIL objec i e, scale ba 25 μ m.
www.na u e.com/scien i ic epo s/
8
Scien i ic RepoR s | 5:12910 | DOi: 10.1038/s ep12910
his was comple ely lacking in he case o he RPE gene a ed om he RP pa ien . The labeled phalloidin
depic s he abundan F-ac in in he mic o illi and he OS a e de ec ed in he egion o mic o illi and
inside he cells de i ed om heal hy indi idual. Ve ical sec ioning demons a ing he inges ed OS pha-
gosomes inside he cells is shown on Supplemen a y Fig. S8. The pa ien ’s iPSC-RPE monolaye showed
only a ew OS in he a ea o mic o illi. Bo h iPS cell lines de i ed om he RP pa ien exhibi ed he same
de ec in phagocy osis o OS upon di e en ia ion o RPE.
We ha e hus gene a ed pa ien -speci ic RPE cells ha lack he exp ession o MERTK and conse-
quen ly ecapi ula e he MERTK de icien pheno ype i.e. absence o phagocy osis. These da a show ha
pa ien -speci ic, ma u e and unc ional RPE can be gene a ed om con ol and RP pa ien s ia iPSCs.
The monolaye s de i ed om he wo indi iduals as epo ed he e a e mo phologically iden ical and
indis inguishable by RPE-speci ic ma ke exp ession while p ese ing pa ien ’s iden i y as obse ed by
dis inc i e MERTK exp ession and i s associa ed unc ion.
Discussion
Re inal dys ophies comp ise a g oup o eye diseases caused by mu a ions in a spec um o genes wi h
a speci ic unc ion in pho o ecep o s o RPE cells o in many cases mu a ions in ubiqui ously exp essed
genes ha exclusi ely gi e ise o a e inal pheno ype. While e inal issue is inaccessible o he s udies
o molecula mechanism, un ela ed issues such as blood is no uly ep esen a i e o molecula e en s
in he e ina. Animal models emain an in aluable ool o unde s and disease mechanism, howe e in
some cases hey a e no ully ep esen a i e o he disease pa hophysiology, owing o di e ences in neu-
onal de elopmen , physiology and he gene ic backg ound. In some cases, such as he Ushe -1 mouse
model, he animal lacks he disease pheno ype29, o he CHM gene de icien mice (model o cho oi-
de emia) which a e le hal30, iPSC-based models ep esen a iable al e na i e o s udy disease biology.
Pa ien -speci ic iPSC o e a enewable cell sou ce ha can be coaxed owa d a desi ed cell ype, which
o e s a unique oppo uni y o gain aluable insigh in o he disease mechanism in addi ion o d ug
sc eening, oxici y es ing and explo e new he apies in he con ex o he human gene ic backg ound.
MERTK mu a ions accoun o app oxima ely 1% o a RP and we e epo ed in se e al amilies wi h
e inal dys ophy20,31–35; wi h common clinical ea u es such as ea ly onse o symp oms, usually in he
i s decade o li e, loss o pe iphe al ision as ypically associa ed wi h od-cone ype o dys ophy and
accumula ion o OS by-p oduc deb is in he sub- e inal space de ec ed by op ical cohe ence omog a-
phy. These clinical pa ame e s can be used as dis inc i e ea u es o MERTK-associa ed RP acili a ing
ea ly iden i ica ion o pa ien s. Ea ly de ec ion o pa ien s should p o ide a be e choice o he apeu ic
op ions. Al oge he 20 mu a ions ha e been desc ibed in he MERTK gene causing RP, he majo i y
being missense/nonsense mu a ions (11), splicing-si e mu a ions (3), small dele ions (3), inse ions (1)
and g oss dele ions (2) as epo ed in he HGMD public da abase (las accessed Janua y 8, 2015). We
show he e he success ul gene a ion o iPSC om a RP pa ien bea ing a no el p.Se 331Cys s*5 mu a-
ion in MERTK which has no been desc ibed be o e. Adul de mal ib oblas s we e ep og ammed
wi h non-in eg a i e i us deli e ing plu ipo ency genes and he gene a ed iPSCs main ained genomic
in eg i y h oughou he cul u e pe iod o o e 30 passages. I should be no ed ha genomic al e a ions
including nume ical and s uc u al al e a ions a e no uncommon in iPSC cul u es. In ou expe ience
he use o Sendai i us as a ehicle o gene deli e y esul ed in he absence o genomic al e a ions in
con as o in eg a i e i al s a egies. As demons a ed he e, iPSCs es ablished om he con ol and
he a ec ed pa ien we e iden ical in ela ion o ep og amming e iciency and exhibi ed simila kine ics
and e iciency o di e en ia ion. Bo h iPSC-RPE we e mo phologically indis inguishable, he polygo-
nal mo phology and pigmen a ion was displayed by con ol and pa ien RPE. The RPE speci ic ma ke
exp ession such as CRALBP, RPE65, BEST1 and ZO-1 was simila in bo h iPSC-RPE monolaye s. A
dis inc i e unc ional ea u e o he na i e RPE is he cellula pola i y based on he pola ized localiza ion
o RPE ma ke s. We show pola ized dis ibu ion o ZO-1, a igh junc ion ma ke , and Na+/K+ ATPase
pump p o ein exp essed on he apical memb ane. This is a dis inc i e RPE ea u e since Na+/K+ ATPase
is epo ed o be exp essed basola e ally in o he anspo ing epi helia36. MERTK dis ibu ion is also
shown o be apical as p e iously desc ibed in RPE cells27. Ano he RPE unc ional cha ac e is ic is he
p esence o igh junc ions essen ial o he o ma ion o blood- e ina ba ie . We obse e by elec on
mic oscopy s uc u es co esponding o igh junc ions and apico-la e al localiza ion o ZO-1 by luo-
escen mic oscopy. In conclusion i can be s a ed ha MERTK de iciency does no in e e e wi h hese
unc ional pa ame e s.
In o de o con i m ha he di e ence be ween heal hy and diseased cells is due o disease mu a ion
and no due o in e -lineage a iabili y desc ibed o iPSCs37,38 we s udied in pa allel wo iPSC lines
pe indi idual. The disease-speci ic de ec is hen expec ed o be p esen in all disease iPSC lines while
absen om all heal hy iPSC lines which is con i med by ou indings. Analysis o he MERTK exp ession
aises he possibili y ha he pa ien ’s mRNA pa ially escapes nonsense–media ed decay and pe haps he
ansc ip is ansla ed in o an abno mal 334 amino acid p o ein lacking he kinase domain. Howe e ,
we do no de ec he unca ed p o ein in pa ien ’s iPSC-RPE ex ac s by wes e n blo no by immuno-
cy ochemis y in he iPSC-RPE monolaye , sugges ing ha i may be deg aded a e being ansla ed.
Ou esul s p o e ha pa ien -speci ic iPSC ep oduce he disease pheno ype eliably and his should
allow ai h ul modeling o degene a i e disease o he RPE in he u u e wi h mo e sub le pheno ypes
and comple ely unknown disease mechanisms. In addi ion, he gene a ed iPSC-RPE ep esen s clinically
www.na u e.com/scien i ic epo s/
9
Scien i ic RepoR s | 5:12910 | DOi: 10.1038/s ep12910
eliable se ing o es he apies in o de o e e se o delay he disease onse in pa ien s ha bo ing
MERTK mu a ions.
Ma e ials and Me hods
Gene ic analysis. A consanguineous Spanish amily (RP-0503) consis ing o h ee a ec ed siblings
(Supplemen a y Fig. S1A) was ec ui ed by he Fundacion Jimenez Diaz Hospi al (Mad id, Spain). DNA
was ex ac ed om pe iphe al blood leukocy es collec ed in EDTA ubes using an au oma ed DNA
ex ac o (BioRobo EZ1; Qiagen; Hilden, Ge many). In o med consen was ob ained om all indi id-
uals in ol ed, all p ocedu es we e e iewed and app o ed by he E hics Commi ee o he Fundacion
Jimenez Diaz Hospi al and adhe ed o he ene s o he Decla a ion o Helsinki.
Whole-genome single nucleo ide polymo phism (SNP) mic oa ay analysis was pe o med using
Illumina HumanCy o-12 SNP a ay, con aining 298,199 SNPs ma ke s (Illumina, Inc, San Diego, CA).
A ays we e p ocessed acco ding o he manu ac u e ’s p o ocols. Linkage analysis was pe o med, using
he Gene Hun e p og am ( e sion 2.1 5) in he easyLINKAGE plus so wa e package ( e sion 5.08), o
iden i y egions ha may con ain he causa i e mu a ion. As he a ec ed siblings we e bo n o a consan-
guineous ma iage, only he homozygous egions we e aken in o accoun .
Bidi ec ional au oma ic sequencing was pe o med o sc een he MERTK gene o mu a ions. Exons
and exon–in on bounda ies o MERTK we e analyzed using oligonucleo ide p ime pai s designed using
P ime 3 so wa e (h p:// odo.wi.mi .edu/). Sequences and annealing empe a u es a e a ailable om
he au ho s on eques . The PCR condi ions and p ime sequences a e a ailable on eques . The PCR
p oduc s we e enzyma ically pu i ied wi h ExoSAP-i (USB, A yme ix) and bo h s ands sequenced
using Big Dye Te mina o Cycle Sequencing Ki e sion 3.1 (Applied Biosys ems, Ca lsbad, CA). The
PCR p oduc s we e pu i ied in a 96-well mul isc een il e pla e (Mon age SEQ96 Sequencing Reac ion
Cleanup Ki ; Millipo e, Bed o d, MA) and esol ed on an au oma ed sequence (ABI 3130xl Gene ic
Analyze , Applied Biosys ems).
Di ec sequencing in ib oblas s and iPSCs: Genomic DNA om ib oblas s and iPSCs was isola ed
using he QIAamp DNA Blood mini ki (Quiagen). P ime s used o ampli ica ion and di ec ed sequencing
o MERTK ups eam and downs eam o c.992_993 we e as ollows: 5′ CGAAGAGGT TCTAAGAGAGG3′
and 5′ CCAT TT TCATCAGTCGCCTC3′ (annealing empe a u e 55 °C).
De i a ion o skin ib oblas s. Skin biopsy o a heal hy indi idual and an a ec ed pa ien (wi h a
con i med molecula diagnosis o ea ly-onse RP) was aken unde s e ile condi ions ollowing in o med
consen . The skin biopsy sample was placed in ~25 ml o s e ile PBS a oom empe a u e (20–25 °C) o
shipmen . Inside a issue cul u e hood, he PBS was eplaced wi h PBS con aining 10× penicil lin/s ep-
omycin (500 U/ml penicillin and 500 µg/ml s ep omycin) and 10× Fungizone (25 µg/ml ampho e icin
B) and le a oom empe a u e o 15 min, wi h occasional mixing by in e ing he ube. This s ep was
epea ed wice and inally he biopsy sample was washed wi h PBS o 15 min. The sample was cu in o
small pieces and cul u ed in 60-mm plas ic dishes (8–10 pieces pe dish) in DMEM (10% FBS, 2 mM
Glu aMAX, 50 U/ml penicillin and 50 mg/ml s ep omycin) a 37 °C unde 5% CO2. A e 3–4 weeks, ib o-
blas s ou g owing om he biopsy pieces co e mos o he dish and cells we e passaged and pla ed in a
T-75 lask. A e eaching 80% con luency ib oblas s we e u he expanded and c yos o ed in liquid N2.
Gene a ion and main enance o iPSC lines. The iPSC we e de i ed om pa ien ’s ib oblas s using
Sendai i us (Cy o une iPS ep og amming Ki , Li e Technologies) acco ding o manu ac u e ’s ins uc-
ions. Single colony subcloning was pe o med o 5 passages and he p esence o Sendai i us ansgenes
was de ec ed by RT-PCR (Supplemen a y Fig. S2) wi h p ime s desc ibed in Supplemen a y Table S1,
du ing passages 7–10. Cells se aside du ing ep og amming on day 7 we e used as posi i e con ol. PCR
was ca ied ou using 500 ng o cDNA in he p esence o MyTaq Red DNA Polyme ase (Bioline) using
manu ac u e ’s ins uc ions. PCR p oduc s we e analyzed using 2% aga ose gel elec opho esis. iPSCs we e
g own on i adia ed (45Gy) human o eskin ib oblas s (ATCC CRL 2429) in iPSCs medium con aining
KO DMEM, KSR 20%, Glu amax 2 mM, non essen ial aminoacids 0.1 mM, β -me cap oe hanol 0.23 mM,
basic FGF 10 ng/mL, and peniciline/s ep omicine. Cells we e mechanically passaged e e y 6–8 days.
Di e en ia ion owa d RPE cells. To ob ain RPE cells he iPSCs colonies we e mechanically
emo ed and cul u ed in low a achmen pla es as loa ing agg ega es in he iPSC medium wi h 15% KSR
and dep i ed o bFGF. When da k pa ches o med on he agg ega es hey we e mechanically excised,
ea ed wi h ypsin and pla ed on Ma igel (BD, #354277) coa ed plas ic cul u e dishes. Depending on
he con luency o pla ing, 3–5 weeks la e , monolaye s o cells wi h RPE cha ac e is ics (polygonal shape,
pigmen a ion) appea ed. A e excising he agg ega ed po ions, he RPE-like monolaye was dissocia ed
wi h ypsin–EDTA (0.05%), passed h ough a 70 μ m s aine and pla ed a 200,000 cells/cm2 on Ma igel
coa ed answell inse s (Co ning 3470, MilliCell). The RPE cells we e cul u ed o 30–60 days un il hey
eached high pigmen a ion. Two iPS cell lines pe each geno ype we e induced o di e en ia e owa d
RPE and no di e ences we e obse ed wi h espec o kine ics o di e en ia ion, deg ee o pigmen a ion
o cha ac e is ic RPE mo phology be ween he iPS cell lines om one indi idual no be ween heal hy
and a ec ed geno ype.