REPORT

Antisense Oligonucleotide-Based Therapyin Human Erythropoietic Protoporphyria

Vincent Oustric,1,11 Hana Manceau,1,11 Sarah Ducamp,1,11 Rima Soaid,1 Zoubida Karim,1,2

Caroline Schmitt,1,2,3 Arienne Mirmiran,1 Katell Peoc’h,1 Bernard Grandchamp,1,2,4 Carole Beaumont,1,2

Said Lyoumi,1,5 Francois Moreau-Gaudry,6,7 Veronique Guyonnet-Duperat,6,7 Hubert de Verneuil,6,7

Joelle Marie,8 Herve Puy,1,2,3,4 Jean-Charles Deybach,1,2,3,4,* and Laurent Gouya1,3,5,9,10

In 90% of people with erythropoietic protoporphyria (EPP), the disease results from the inheritance of a common hypomorphic FECH

allele, encoding ferrochelatase, in trans to a private deleterious FECHmutation. The activity of the resulting FECH enzyme falls below the

critical threshold of 35%, leading to the accumulation of free protoporphyrin IX (PPIX) in bone marrow erythroblasts and in red cells.

Themechanism of low expression involves a biallelic polymorphism (c.315�48T>C) localized in intron 3. The 315�48C allele increases

usage of the 30 cryptic splice site between exons 3 and 4, resulting in the transcription of an unstablemRNAwith a premature stop codon,

reducing the abundance of wild-type FECHmRNA, and finally reducing FECH activity. Through a candidate-sequence approach and an

antisense-oligonucleotide-tiling method, we identified a sequence that, when targeted by an antisense oligonucleotide (ASO-V1),

prevented usage of the cryptic splice site. In lymphoblastoid cell lines derived from symptomatic EPP subjects, transfection of

ASO-V1 reduced the usage of the cryptic splice site and efficiently redirected the splicing of intron 3 toward the physiological acceptor

site, thereby increasing the amount of functional FECH mRNA. Moreover, the administration of ASO-V1 into developing human

erythroblasts from an overtly EPP subject markedly increased the production of WT FECH mRNA and reduced the accumulation of

PPIX to a level similar to thatmeasured in asymptomatic EPP subjects. Thus, EPP is a paradigmaticMendelian disease inwhich the in vivo

correction of a common single splicing defect would improve the condition of most affected individuals.

Erythropoietic protoporphyria (EPP, MIM 177000) is a rare

inherited disorder caused by the partial mitochondrial defi-

ciency of ferrochelatase (FECH, EC 4.99.1.1.), the last

enzyme in the heme biosynthesis pathway1 (See also

Balwani et al., 2012 in GeneReviews in theWeb Resources).

FECH is an inner mitochondrial membrane enzyme that

catalyzes the insertion of ferrous iron into free proto-

porphyrin IX (PPIX) to form heme. FECH deficiency in

medullar erythroid cells leads to overproduction and accu-

mulation of photoactive PPIX in erythrocytes and subse-

quently in the plasma, skin, bile, and feces.1 The most

common clinical manifestation of the disease is lifelong

acute photosensitivity of sun-exposed areas that first

appears in early childhood. Although it is generally a

benign disease, hepatic complications such as choleli-

thiasis or, in about 2% of cases, cirrhosis with rapidly fatal

liver disease, can occur.2–4 Cases of EPP have been reported

in Europe, the USA, China, and Japan. So far, no case of

EPP has been reported in individuals of African descent.

Constitutively, FECH undergoes alternative splicing of

exon 4 as a result of partial utilization of a cryptic intron

3 acceptor splice site located 63 bp upstream of the major

intron 3/exon 4 boundary, resulting in the pseudo-exon

1Institut National de la Sante et de la RechercheMedicale, U1149, Centre de Rec

F-75018 Paris, France; 3Assistance Publique-Hopitaux de Paris, Centre Francai

Colombes, France; 4Assistance Publique-Hopitaux de Paris, Laboratoire de B5Universite de Versailles Saint Quentin en Yvelines, F-78035 Versailles, Fra

Biotherapies des Maladies Genetiques et Cancers, Laboratoire d’Excellence du

F-33000 Bordeaux, France; 8Centre de Genetique Moleculaire, Centre Natio

Gif-sur-Yvette, Universite Paris-Sud, 91400 Orsay, France; 9Assistance Publiq

Hopital Ambroise Pare, F-92100 Boulogne Billancourt, France; 10Laboratory of11These authors contributed equally to this work

*Correspondence: jc.deybach@wanadoo.fr

http://dx.doi.org/10.1016/j.ajhg.2014.02.010. �2014 by The American Societ

The Am

inclusion of 63 nucleotides of intron 3 about 20% of the

time. The aberrant mRNA contains an in-frame premature

stop codon (Figure 1) and is degraded by nonsense-medi-

ated mRNA decay.5–7 The hypomorphic C allele of a com-

mon intronic SNP, c.315�48T>C (rs2272783, Figure 1), is

responsible for increasing the usage of the �63 bp cryptic

intron 3 acceptor splice site from 20% (for the T allele) to

60% (for the C allele), resulting in a reduction of FECH

activity. When, in combination with a severe FECH muta-

tion on the other allele, the overall FECH activity falls

below a critical threshold of about 35% of normal, PPIX

accumulates, and photosensitivity occurs.8,9Several studies

in the USA, Europe, and Asia have confirmed that this

mechanism is generally operative in EPP.10–15 According

to the French Centre of Porphyria database, 94% of French

EPP subjects show this striking inheritance pattern.

Because EPP is primarily a seasonal disease, and because

affected individuals present mainly non-life-threatening,

dermatological symptoms, integrative gene therapy is not

appropriate. Rather, correcting this single common

splicingmutation by antisense oligonucleotide (ASO) ther-

apy is an attractive strategy that should improve the condi-

tion of the majority of EPP subjects independently of the

herches sur l’Inflammation, F-75018 Paris, France; 2Universite Paris Diderot,

s des Porphyries, Hopital Louis Mourier, 178 Rue des Renouillers, F-92701

iochimie Hormonale et Genetique, Hopital Bichat, F-75018 Paris, France;

nce; 6Institut National de la Sante et de la Recherche Medicale, U1035,

Globule Rouge, F-33000 Bordeaux, France; 7Universite Bordeaux Segalen,

nal de la Recherche Scientifique, UPR 3404, Avenue de Terrasse, 91198

ue-Hopitaux de Paris, Laboratoire de Biochimie Hormonale et Genetique,

Excellence GR-Ex, 75000 Paris, France

y of Human Genetics. All rights reserved.

erican Journal of Human Genetics 94, 611–617, April 3, 2014 611

ttttttcttttttattgagtagAAAACATTTCTCAGGTTGCTAAExon 3 Exon 4 3'5'

-63 bp

Normal splicing

Intron 3

Exon 3 Exon 4stopC

Aberrant splicing

Figure 1. Schematic Representation of Exon 3–4 Splicingof FECHThe c.315-48T>C transition modulates the splicing efficiency of aconstitutive cryptic acceptor splice site. �63 bp: position of thecryptic acceptor splice site. (adapted from Gouya et al., 2002).

nature of their FECH mutation. ASOs, which are generally

used for inhibiting gene expression, can also be used for

modulating pre-mRNA splicing by targeting splice sites or

by targeting positive or negative elements that affect

splice-site selection.16

In this study, we applied two strategies to identify ASOs

that repress the partial inclusion of intron 3 in FECH

mRNA: first, an ASO-tiling method to search for regulating

cis-acting elements residing within the 130 bp stretch

upstream of exon 4; and second, an orientated strategy

in which we used three ASOs targeting the putative

315�48C-activated cryptic branch point, the cryptic

acceptor splice site, or both the cryptic splice site and the

315�48C nucleotide (Figure 2A and Table 1). For all of

these experiments, we used locked nucleic acids (LNAs)

that demonstrate nuclease resistance and enhance affinity

for hybridization to complementary RNA.17 This class of

ASOs is highly effective for modifying pre-mRNA splicing

patterns.18 To investigate the effects of individual ASOs

on splicing, we constructed two FECH minigenes (FECH-

C-pcDNA3 and FECH-T-pcDNA3) comprising exon 3

(121 bp), a shortened intron 3 (300 bp, with either the

c.315�48C or the c.315�48T genotype) and exon 4

medium at a final concentration of 30 mM. Twenty-four hours lateRNAplus reagent (MP Biomedical). Human-specific exon 3–4 FECH pand aberrantly spliced mRNA, the PCR products were analyzed by elewas measured with Image J Software. The migration positions of the nproduct are indicated on the right. Ratios of the aberrantly spliced R

612 The American Journal of Human Genetics 94, 611–617, April 3, 2

(193 bp). We transfected these into Cos7 cells by using lip-

ofectamine reagent (Invitrogen). Two days after transfec-

tion, we analyzed exon 3–4 FECH mRNA by RT-PCR with

a pair of human primers (Table S1). These primers did

not amplify the endogenous Cos7 FECH transcript

(Figure 2B). The extent of intron 3 inclusion was expressed

as the ratio of the intensity of the aberrantly spliced band

to the sum of the intensities of the aberrantly and correctly

spliced bands. FECH minigene transfection in Cos7 cells

recapitulated the splicing pattern as shown in lymphoblas-

toid cell lines (LBCL) of 315�48C/C or T/T subjects.7 The

FECH-C-pcDNA3 minigene revealed a ratio of about

60%, and the FECH-T-pcDNA3 minigene one of 20%

(Figure 2B). For the ASO-tiling experiments, we cotrans-

fected the FECH-C-pcDNA3 minigene individually with

each of the ASOs (Figure S1A and Table S2; 13 ASOs, over-

lapping by 5 nt). None of the 13 LNA-ASOs used in the

initial ‘‘walk’’ along intron 3 reduced intron 3 retention

(Figure S1B). Only the ASO targeting position �45 to –63

(V1-LNA-ASO), which was obtained via the orientated

strategy and was complementary to a sequence including

the 315�48 polymorphism and the 30 cryptic acceptor

splice site, reduced intron 3 inclusion to about 20% of

the total RNA, a level similar to the level obtained with

the FECH-T-pcDNA3 construction (Figure 2B). A dose-

response effect was already evident at 50 nM ASO concen-

tration,and was maximal at 125 nM (Figure 2B). To further

refine the targeting sequence, we microwalked around

the �45 position by using 18 additional 15-mer LNA-

ASOs (Figure S2A and Table S2). Again, none displayed

any effect, demonstrating that both the 315�48 region

and the �63 acceptor splice site have to be targeted to

achieve the repression of intron 3 inclusion (Figure S2B

and S2C). Finally, we tested whether the splicing was sen-

sitive to the length of the ASO. We used eight additional

Figure 2. Repression of the Splicing ofCryptic Exon 3–4 by a LNA-ASO TargetingBoth the Cryptic Acceptor Splice Site andthe 315�48C Region(A) Position of the LNA-ASOs targeting theputative cryptic branch point (at posi-tion �97 to �116), the cryptic acceptorsplice site (position �54 to �74), and the315�48 region (position �45 to �63).Horizontal arrows represent the positionsof the PCR primers used for splicing anal-ysis. Putative cryptic branch point:(B) Inhibition of abnormal FECH splicingby three LNA-ASOs. Cos7 cells were trans-fected with 800 ng of plasmid and 2 ml ofLipofectamine 2000 (Life Technologies).Three hours later, the cells were trans-fected with 50, 125, or 250 nM LNA and2 ml of Lipofectamine 2000. Emetin (SigmaAldrich) was added simultaneously in the

r, the cells were harvested, and total RNA was extracted via therimers were used for qualitative PCR that amplified the correctly

ctrophoresis on a 3% agarose gel, and the density of the PCR bandsormal exon 3–4 81 bp amplimer and the aberrantly spliced 144 bpNA to the total RNA are indicated at the bottom.

014

Table 1. LNA-ASOs Used in the Candidate-Sequence Approach

Position Sequence

�45 to �63 50-gcEgcLtgEgaEatPttZt-30

�54 to �74 50-gEaaZgtZttLtaLtcEatEa-30

�97 to �116 50-aEaaLatZtcZcaPgcZgc-30

For each ASO in intron 3, the position corresponds to the distance of the final 50

and 30 ASO nucleotides from the first base of exon 4. The sequences are shownon the complementary strand. LNA bases are as follows: ‘‘E’’ for adenine, ‘‘P’’for guanine, ‘‘Z’’ for thymine, and ‘‘L’’ for cytosine.

LNA-ASOs, which were one or two bases longer than V1-

LNA-ASO at either one or both extremities (Figure S2D

and Table S2). All these LNA-ASOs reduced intron 3 inclu-

sion, but none was more effective than the V1-LNA-ASO

(Figure S2E).

We further investigated whether reduction of intron 3

retention could also induce excess FECH mRNA produc-

tion. Therefore, we measured FECH splicing in lympho-

blastoid cell lines (LBCL) of an overt EPP-subject (subject

II1, FECH genotype c.709delT mutation/315�48C) and

her asymptomatic brother (subject II3, FECH genotype

c.709delT mutation/315�48T, Figure 3A). The optimal

concentration (125 nM) of V1-LNA-ASO or mock-ASO (a

complementary sequence of the ASO �45 to –63: 50-AAAACATTTCTCAGGCTGC-30) were transfected with lipofect-

amine reagent. Three hours after transfection, emetine

(Sigma-Aldrich) was added to the medium to block NMD.

Forty-eight hours after ASO transfection, endogenous

exon 3–4 splicing was analyzed with exon 3–4 RT-PCR

A B

0

10

20

30

40

50

24H

WT FECH

expr

essi

on

I 1 2

II 1 2 3

FC = 3.3M +T T

FC = 6+ +T C

FC = 1.7M +T C

FC = 1M +T C

FC = 3.3M +T T

Cp =0.0022

p =0.0022

NS

NS

0.20

0.25

0.30

0.35

0.40

Abno

rmal

FECH

mR

NA

to to

tal m

RN

A

T C T C T T

mock-LNA-ASO without emetin. WT FECH RNA was analyzed by RRNA. n ¼ 6 independent experiments. The Mann-Whitney statistica

The Am

(Table 1; Figure 3B). As expected, intron 3 retention in

the LBCL of subject II3 (transfected with the mock ASO)

was about 25% lower than in the LBCL of subject II1

(abnormal mRNA/total mRNA ratio 0.30 versus 0.38, p ¼0.0022; Figure 3B). Interestingly, when the LBCL of subject

II1 was transfected with V1-LNA-ASO, the proportion of

abnormally spliced RNA decreased from 0.38 to a level

similar to that measured in the asymptomatic II3 subject

(0.32 versus 0.30, not significant; Figure 3B). Wild-type

FECHmRNAwas thenmeasured by RT-qPCR using primers

specific to the wild-type exon 3–4 boundary (Table S1).

Twenty-four hours after transfection, normal FECH

mRNA was 30% more abundant in the LBCLs of the

asymptomatic (II3) than the overt (II1) subject

(Figure 3C),7 and 24 hr later, the normal mRNA of the cells

of the overt II1 subject treated with the V1-LNA-ASO

was 1.8- and 1.7-fold higher than that of the II1

mock-transfected and II3 -V1-LNA-ASO-transfected cells,

respectively (Figure 3C).

In summary, we identified a sequence at nucleo-

tides �45 to �63 that, when targeted by V1-LNA-ASO,

reduced intron 3 inclusion in LBCL with a 315�48T/C

genotype to a level comparable to that of the 315�48T/T

genotype and increased wild-type mRNA production in

the cells of an overt subject to a higher level than that

measured in an asymptomatic EPP subject. Taken together,

these results suggest that the V1-LNA-ASO has con-

siderable therapeutic potential. The sequence at posi-

tion �45 to �63 is intronic with regard to physiological

exon 3–4 splicing, but it becomes exonic when the �63

cryptic splice site is used. The mechanisms underlying

48H

p=0.015

Figure 3. Restoration of WT FECHmRNAProduction in the LBCLs of EPP SubjectsPedigree of the EPP-affected family. M in-dicates the c.709delT deleterious FECHmutation. T indicates the 315�48T allele.C indicates the 315�48C allele. SubjectsI1 and II3 were asymptomatic carriers forthe c.709delT mutation (RefSeq accessionnumber NM_000140.3; FECH genotypeM/315�48T), and subjects II1 and II2were overt EPP subjects (FECH genotypeM/315�48C). FC: FECH activity in nmolof Zn-mesoporphyrin/mg of protein/hr.(B) The LBCLs were transfected with125 nM V1-LNA-ASO or mock-LNA-ASO(�43 to �65 sense sequence), and emetinwas added 24 hr later. Total RNA was ex-tracted 48 hr after transfection. Splicingof intron 3 to exon 4 was analyzed as inFigure 2B. The ratios of the intensity ofthe aberrantly spliced RT-PCR product tothe sum of the intensities of the aberrantlyand correctly spliced products are pre-sented as box plots showing the median,quartiles, and 90th and 10th percentiles.n ¼ 6 experiments.(C) Total RNA was extracted 24 and 48 hrafter transfection with V1-LNA-ASO or

T-qPCR with primers specific to the normally spliced exon 3–4l test was used with Prism 4 software (Prism4 GraphPad software).

erican Journal of Human Genetics 94, 611–617, April 3, 2014 613

Table 2. Clinical, Biochemical, and FECH Genotypic Data in Three EPP Subjects and One Control from Whom the CD34þ ErythroidPrecursors Were Extracted

Subject FECH Mutation 315�48 Genotype Total Erythrocyte Protoporphyrins FECH Activity Phenotype

A c.1038T>G p.Tyr346Stop T/C 43 1.8 symptomatic

B c.1078�2A>G T/C 45 1.8 symptomatic

C c.1157A>C p.His386Pro T/T 2.5 2.4 asymptomatic

D WT T/T 1.5 5.6 asymptomatic

Total erythrocyte PPIX is expressed in mmol/l RBC. FECH activities are presented in nmol of Zn-Mesoporphyrin/mg of protein/hr. WT: wild-type. The FECH RefSeqaccession number is NM_000140.3.

splicing redirection from the exon 3–4 boundary of the

cryptic to the physiological exon are complex and not

yet fully understood. Blocking exclusively either the 30

aberrant splice site (six ASOs: �50 to �64, �51 to �65,

�52 to �66, �53 to �67, �61 to �75, and �54 to �74)

or the 315�48 region (11 ASOs: �34 to �48, �35 to �49,

�36 to �50, �37 to �51, �38 to �52, �39 to �53, �41

to �55, �45 to �59, �46 to �60, �47 to �61, and �48

to�62) was not sufficient to restore proper splicing. In

the competition between the cryptic and the physiological

splice sites, redirection of splicing toward the physiological

site has to include blocking both the cryptic splice site and

the 315�48 region, suggesting that this region might

include an exonic splicing enhancer (ESE) of cryptic

splicing or an intronic splicing inhibitor (ISI) of exon 3–4

splicing.

Because medullar erythroblasts are the relevant tissue to

be therapeutically targeted in EPP, we tested the V1-ASO

in differentiating CD34þ-derived, erythroid progenitors

isolated from the peripheral whole blood from two overt

EPP subjects (A and B), one asymptomatic subject (D),

and one control subject (C, Table 2). Our study was con-

ducted in accordance with the World Medical Association

Declaration of Helsinki ethical principles for medical

research involving human subjects and its subsequent

amendments. All subjects gave informed consent before

undergoing investigation. Overt subjects A and B had a

classical history of skin photosensitivity that began during

childhood, a high level of erythrocyte-free protoporphyrin,

and 25%–30% residual FECH activity in peripheral blood

mononuclear cells. The FECH genotypes of the overt

subjects consisted of a deleterious FECH mutation in trans

to the hypomorphic 315�48C allele (Table 2). The non-

sense mutation of subject A most likely triggered mRNA

degradation, and the mutation in subject B resulted in

aberrant exon 10 splicing that conserved the correct

reading frame. The third subject (subject D) was an asymp-

tomatic EPP, and her FECH genotype consisted of a delete-

rious FECH nonsense mutation in trans to the 315�48T

allele. Her erythrocyte PPIX levels were slightly above the

normal limit (X2), and her FECH activity in peripheral

bloodmononuclear cells was ~50%belownormal (Table 2).

For the oligonucleotide treatment, liposomal transfec-

tions of V1-LNA-ASO were ineffective because of their

high toxicity and poor transfection efficiency (data not

614 The American Journal of Human Genetics 94, 611–617, April 3, 2

shown). We therefore used a free-uptake method in

which morpholino-ASOs were added directly to the

culture medium at a final concentration of 45 mM (see

GeneReviews in Web Resources).19,20 R. Kole and col-

leagues developed the original method to direct mor-

pholino-ASOs to erythroid precursors in order to restore

human beta-globin gene expression in human IVS2-654

thalassemic erythroid cells.20 The authors also showed

that adding labeled morpholinos to the culture medium

from days 8 to 17 resulted in strong nuclear staining. The

morpholinos were prepared and purified by Gene Tools,

LLC (Philomath, USA). Two morpholino-ASOs, targeting

the sequence at position�45 to�63 in either the antisense

(PMO-V1) or the complementary sequence, respectively,

were used for the control mock oligonucleotide (PMO-

mock). The oligonucleotides were labeled with fluorescein

isithiocyanate (FITC).

The phenotype differentiation of erythroid cells was

monitored by flow cytometry (FACS Canto II, BD Bio-

sciences). Cells were stained with the relevant isotype or

antibodies (CD34PE 1/100 (A07776), CD36-FITC 1/100

(IM0766U), and CD71-FITC 1/100 (IM0483) from

Beckman Coulter and GPA-PE 1/50 (MHGLA04) from

Invitrogen). Hemoglobinization was monitored daily by

benzidine staining. Cell morphology was established after

May-Grunwald-Giemsa staining of cytospin preparations

by light microscopy. Time-kinetic analysis of PPIX accu-

mulation was determined by flow cytometry at four stages

of differentiation: CFU (colony-forming unit) erythroblast

(CFU E), pro- and intermediate-erythroblast (pro-E, int E),

and late erythroblast (late E). These stages were established

on the basis of cell morphology (MGG coloration;

Figure S3) and the expression of the CD71 and GPA surface

markers, which correspond with the major stages of

development: CFU E (CD71lo, GPA�), pro-E (CD71hi,

GPA�/lo), int E (CD71hi, GPAhi), and late E (CD71lo,

GPAhi). PPIX accumulation in these cells was analyzed by

flow cytometry with an APC filter and the violet laser of

the Canto II (PPIX excitation, 405 nm; emission,

660 nm). Because of the limited number of cells available,

FECH activity could not be used for these experiments. WT

FECH mRNA and ALAS2 mRNA were quantified by

RT-qPCR (Table S1).

To determine the optimal conditions for ASO treat-

ments, we established a culture of CD34þ cells isolated

014

Table 3. WT FECH and ALAS2 mRNA Quantifications in PrimaryCultures of EPP Erythroblasts

ALAS2 mRNA WT FECH mRNA

Subject A (M/C) mock 0.27 0.32

V1 0.40 0.51

Subject B (M/C) mock 0.86 0.96

V1 1.5 1.9

Control subject C (WT/T) mock 1 1

V1 0.75 0.73

Subject D (M/T) mock 1.1 1.1

V1 1.28 0.93

Subjects A, B, C, and D are described in Table 2. Abbreviations are as follows:M, deleterious FECH mutation; WT, wild-type FECH; T, 315�48T allele; and C,315�48C allele. Total RNA was extracted at the late E stage, and morpholino-ASO targeting position�45 to�63 or mockmorpholino-ASOwas added sevendays after the CD34þ cells were isolated fromwhole blood. ALAS2 andWT FECHRNA were analyzed by RT-qPCR with primers specific to the normal exon 3–4spliced RNA of FECH. Both quantifications were normalized to two genes (B2M[MIM 109700] and HPRT1 [MIM 308000]), and the results are expressed as apercentage from the mock condition of subject C.

from a healthy control. We assayed the time course of

FECH mRNA expression by RT-qPCR, and FECH mRNA

increased dramatically from the CFU E stage to the late E

stage (23-fold induction; Figure S4). Erythroid differentia-

tion was confirmed on the basis of cell morphology

(Figure S3), FACS analyses, and the progressive appearance

of hemoglobinized cells ranging from 0% at the CFU E

stage to >95% at the late E stage. Finally, morpholino

oligonucleotides were added before the induction of

heme production at the CFU E stage (after ~7 days in

culture) and maintained at the same concentration

(45 mM) until mature cells were obtained.

Nor

mal

ized

MFI

3.03.5

2.52.01.51.00.50

CFU E Pro E

AsymptomatiDOvert EPP subjectB

CFU E Pro E Int E Late E

Nor

mal

ized

MFI

3.03.5

2.52.01.51.00.50

ControlC

CFU E Pro E

Nor

mal

ized

MFI

3.03.5

2.52.01.51.00.50

Overt EPP subjectA

CFU E Pro E Int E Late E

Nor

mal

ized

MFI

3.03.5

2.52.01.51.00.50

mockMorpholino V1

(all cytokines were from Miltenyi Biotec). Morpholino oligonucleotCFU E stage (after ~7 days in culture) and maintained at the same conanalysis of PPIX accumulation was determined by flow cytometry atgraphs represent the ratios between the mean fluorescence intensityin erythroid cells from two overt EPP subjects (subjects A and B), onsubject (subject D), who were each treated with PMO-V1 (triangles)

The Am

Normal FECH and ALAS2 mRNAs were quantified at the

late E stage, when FECH mRNA expression was highest. At

this stage of differentiation, the normal FECH mRNA in

ASO-treated cells from subjects A and B had increased by

59% and 98%, respectively, in comparison with mock-

treated cells (Table 3). FECH mRNA in V1-treated cells of

subject A remained 50% lower than in those of control

subject C. This could have been due to the type of the mu-

tations for subject A: a 1 bp deletion introducing a prema-

ture STOP codon potentially leading to a null allele. FECH

expression could have been improved primarily because of

the redirection of splicing but also because of an overall

improvement in heme biosynthesis. Indeed, ALAS2, en-

coding the key regulator of erythroid heme biosynthesis,

was overexpressed in V1-treated cells when compared

with mock-treated cells at the same stage of differentiation

(48% and 74% improvement for subjects A and B, respec-

tively; Table 3), probably leading to increased ALAS2

activity. ALAS2 expression is also posttranscriptionally

regulated by the IRE/IRP system, and in a situation where

iron is normally available, we expect the ALAS2 mRNA

translation not to be repressed. In the cells of EPP subjects

who were treated with the V1 morpholino, we observed a

cumulative reduction in the amount of PPIX as erythroid

differentiation progressed and culminated at the late E

end-point stage (44% and 58% reduction for subjects A

and B, respectively; Figure 4). Remarkably, PPIX accumula-

tion decreased in the cultured progenitor cells of overt EPP

subjects A and B to a level similar to that found in the

asymptomatic subject D, although the amount of PPIX

remained higher than in the normal subject C (Figure 4).

This result was in agreement with the observation that

the total amount of erythrocyte PPIX that was measured

in peripheral RBCs from asymptomatic EPP subjects was

Int E Late E

c EPP subject

subject

Int E Late E

Figure 4. Correction of PPIX Over-production in Developing Primary Eryth-roblasts of EPP Subjects: ErythroidDifferentiation of CD34þ Cells Isolatedfrom Peripheral Whole Blood in the Pres-ence of 45 mM FITC-Labeled V1 or MockMorpholinosTotal mononuclear cells were isolated fromwhole peripheral blood (60 ml) by Ficolldensity-gradient centrifugation (lympho-cyte separation medium; PAA Labora-tories), and CD34þ cells were purifiedimmunomagnetically by positive selection(MACS CD34 MicroBead Kit; Miltenyi Bio-tec). CD34þ cells were cultured in IscoveDulbecco’s modified Eagle medium (Invi-trogen) containing 15% BIT 9500 (Stem-Cell Technologies), 100 ng/ml stem cellfactor, 2 U/ml erythropoietin, 10 ng/mlinterleukin-6, and 10 ng/ml interleukin-3

ides were added before the induction of heme production at thecentration (45 mM) until mature cells were obtained. Time kineticfour stages of differentiation: CFU E, pro E, int E, and late E, and(MFI) values of the accumulated PPIX to MFI of FITC-positive cellse healthy control subject (subject C), and one asymptomatic EPPor PMO-mock (squares).

erican Journal of Human Genetics 94, 611–617, April 3, 2014 615

Figure 5. Erythrocyte Protoporphyrin IX in 40 EPP-AffectedFamiliesTotal erythrocyte PPIX was measured in 40 EPP-affected families,each having one overt EPP subject, one asymptomatic EPP subject,and one healthy subject. The results were expressed as a histogramshowing the mean and standard deviation. Means were comparedwith the ‘‘t’’ test with Prism 4. Total erythrocyte PPIX level was2.5-fold higher in peripheral RBCs from asymptomatic EPPsubjects than in healthy subjects.

2.5-fold higher than in healthy subjects; however, this

value remained at subclinical levels (Figure 5). It is thus

possible that the reduced amount of PPIX observed in

erythroid cells of affected individuals could become suffi-

ciently low to suppress skin sensitivity in vivo.

Our study is a proof-of-concept for the ability of ASO

therapy to restore normal exon 3–4 pre-mRNA splicing

and increase normal FECH mRNA production while

decreasing PPIX accumulation to a subclinical level in

cultured erythropoietic cells from EPP subjects, and it

opens a new field in treating EPP. The correction of

FECH exon 3–4 splicing is an attractive therapeutic

approach for EPP because the 315-48C allele is present in

more than 90% of overt EPP subjects. Moreover, a modest

increase in FECH activity is sufficient to shift the subject’s

status from overt to asymptomatic.8 In contrast to integra-

tive gene therapy, antisense therapy has several advan-

tages. (1) Because the correction occurs in medullar

erythroblasts and persists in circulating erythrocytes with

a lifespan of about 120 days, the effects of the oligonucle-

otide treatment are probably prolonged; (2) the splicing

correction occurs in the endogenous gene that is tran-

scribed in its physiological environment, thus preventing

excessive or inappropriate expression; and (3) a pharmaco-

logical treatment is easier to administer than somatic gene

therapy and can easily be interrupted in the case of adverse

effects. Targeting pre-mRNA splicing as a therapeutic strat-

egy in Mendelian disorders was proposed several years ago

for Duchenne muscular dystrophy,21 spinal muscular

atrophy22 and b-thalassemia (MIM 613985). In 1993,

Dominski et al. demonstrated that correct splicing can

be restored in vitro by an ASO targeting the b-globin pre-

616 The American Journal of Human Genetics 94, 611–617, April 3, 2

mRNA.23 Two laboratories were later able to improve

hemoglobin synthesis in vivo and to reduce cell damage

in humanized IVS2–654 thalassemic mice by using either

a morpholino oligomer conjugated to a peptide or an

antisense RNA vector.24,25 The challenge facing antisense

therapeutic strategies is to develop efficient ways to target

ASO to specific cells, in our case to the erythroid pro-

genitors. For systemic administration, it is important to

enhance the specificity of the treatment while reducing

the concentration of ASOs so as to limit toxicity.Different

strategies, including cell-penetrating peptides, aptamers,

and cationic liposome-ASO complexes, have aimed to

enhance ASO targeting. For erythropoietic protoporphyria

therapy, an interesting target could be transferrin receptor

1 (CD71), which is expressed at a very high level during

the differentiation stage, when FECH incorporates iron

into PPIX to form heme. Several peptides and aptamers

that bind with high affinity to human CD71 and display

endocytotic properties are already available,26,27 sug-

gesting that this or similar compounds should be tested

in a humanizedmouse model of EPP and that it might sub-

sequently hold promise in the treatment of individuals

with EPP.

Supplemental Data

The Supplemental Data include four figures and two tables and

can be found with this article online at http://www.cell.com/ajhg.

Acknowledgments

We thank the anonymous family members for their cooperation

and Vasco Pereira Da Silva, Sylvie Simonin, and Anne Marie

Robreau-Fraolini for their expert laboratory assistance. This work

was supported by the Association Francaise contre les Myopathies

(L.G., grant reference N� 16534) and by the ANR- 11-LABX-0051

grant. V.O. was supported by grants from the Association Francaise

contre les Myopathies and the Societe Francaise d’Hematologie.

The authors declare competing financial interests. V.O., H.P.,

J.C.D., and L.G. have patents pending with the European Patent

Office for the ASOs and the targeting approach.

Received: December 5, 2013

Accepted: February 18, 2014

Published: March 27, 2014

Web Resources

The URLs for data presented herein are as follows:

GeneReviews, Balwani, M., Bloomer, J., and Desnick, R. (2012).

Erythropoietic Protoporphyria, Autosomal Recessive, http://

www.ncbi.nlm.nih.gov/books/NBK100826/

Online Mendelian Inheritance in Man (OMIM), http://www.

omim.org

References

1. Puy, H., Gouya, L., and Deybach, J.C. (2010). Porphyrias.

Lancet 375, 924–937.

014

2. Bloomer, J., Bruzzone, C., Zhu, L., Scarlett, Y., Magness, S., and

Brenner, D. (1998). Molecular defects in ferrochelatase in

patients with protoporphyria requiring liver transplantation.

J. Clin. Invest. 102, 107–114.

3. Meerman, L. (2000). Erythropoietic protoporphyria. An

overview with emphasis on the liver. Scand. J. Gastroenterol.

Suppl. 2000, 79–85.

4. Lyoumi, S., Abitbol, M., Rainteau, D., Karim, Z., Bernex, F.,

Oustric, V., Millot, S., Letteron, P., Heming, N., Guillmot, L.,

et al. (2011). Protoporphyrin retention in hepatocytes and

Kupffer cells prevents sclerosing cholangitis in erythropoietic

protoporphyria mouse model. Gastroenterology 141, 1509–

1519, 1519.e1–1519.e3.

5. Gouya, L., Deybach, J.C., Lamoril, J., Da Silva, V., Beaumont,

C., Grandchamp, B., and Nordmann, Y. (1996). Modulation

of the phenotype in dominant erythropoietic protoporphyria

by a low expression of the normal ferrochelatase allele. Am. J.

Hum. Genet. 58, 292–299.

6. Gouya, L., Puy, H., Lamoril, J., Da Silva, V., Grandchamp, B.,

Nordmann, Y., and Deybach, J.C. (1999). Inheritance in eryth-

ropoietic protoporphyria: a common wild-type ferrochelatase

allelic variant with low expression accounts for clinical mani-

festation. Blood 93, 2105–2110.

7. Gouya, L., Puy, H., Robreau, A.M., Bourgeois, M., Lamoril, J.,

Da Silva, V., Grandchamp, B., and Deybach, J.C. (2002). The

penetrance of dominant erythropoietic protoporphyria is

modulated by expression of wildtype FECH. Nat. Genet. 30,

27–28.

8. Gouya, L., Martin-Schmitt, C., Robreau, A.M., Austerlitz, F., Da

Silva, V., Brun, P., Simonin, S., Lyoumi, S., Grandchamp, B.,

Beaumont, C., et al. (2006). Contribution of a common

single-nucleotide polymorphism to the genetic predis-

position for erythropoietic protoporphyria. Am. J. Hum.

Genet. 78, 2–14.

9. Tahara, T., Yamamoto, M., Akagi, R., Harigae, H., and Taketani,

S. (2010). The low expression allele (IVS3-48C) of the ferroche-

latase gene leads to low enzyme activity associated with eryth-

ropoietic protoporphyria. Int. J. Hematol. 92, 769–771.

10. Risheg, H., Chen, F.P., and Bloomer, J.R. (2003). Genotypic

determinants of phenotype in North American patients

with erythropoietic protoporphyria. Mol. Genet. Metab. 80,

196–206.

11. Wiman, A., Floderus, Y., and Harper, P. (2003). Novel muta-

tions and phenotypic effect of the splice site modulator

IVS3-48C in nine Swedish families with erythropoietic proto-

porphyria. J. Hum. Genet. 48, 70–76.

12. Saruwatari, H., Ueki, Y., Yotsumoto, S., Shimada, T., Fuku-

maru, S., Kanekura, T., and Kanzaki, T. (2006). Genetic anal-

ysis of the ferrochelatase gene in eight Japanese patients

from seven families with erythropoietic protoporphyria.

J. Dermatol. 33, 603–608.

13. Kong, X.F., Ye, J., Gao, D.Y., Gong, Q.M., Zhang, D.H., Lu,

Z.M., Lu, Y.M., and Zhang, X.X. (2008). Identification of a

ferrochelatase mutation in a Chinese family with erythro-

poietic protoporphyria. J. Hepatol. 48, 375–379.

14. Whatley, S.D., Mason, N.G., Holme, S.A., Anstey, A.V., Elder,

G.H., and Badminton, M.N. (2010). Molecular epidemiology

of erythropoietic protoporphyria in the U.K. Br. J. Dermatol.

162, 642–646.

The Am

15. Balwani, M., Doheny, D., Bishop, D.F., Nazarenko, I., Yasuda,

M., Dailey, H.A., Anderson, K.E., Bissell, D.M., Bloomer, J.,

Bonkovsky, H.L., et al.; Porphyrias Consortium of the

National Institutes of Health Rare Diseases Clinical Research

Network (2013). Loss-of-function ferrochelatase and gain-of-

function erythroid-specific 5-aminolevulinate synthase muta-

tions causing erythropoietic protoporphyria and x-linked

protoporphyria in North American patients reveal novel

mutations and a high prevalence of X-linked protoporphyria.

Mol. Med. 19, 26–35.

16. Kole, R., and Leppert, B.J. (2012). TargetingmRNA splicing as a

potential treatment for Duchenne muscular dystrophy.

Discov. Med. 14, 59–69.

17. Kurreck, J., Bieber, B., Jahnel, R., and Erdmann, V.A. (2002).

Comparative study of DNA enzymes and ribozymes against

the same full-length messenger RNA of the vanilloid receptor

subtype I. J. Biol. Chem. 277, 7099–7107.

18. Roberts, J., Palma, E., Sazani, P., Ørum, H., Cho, M., and Kole,

R. (2006). Efficient and persistent splice switching by system-

ically delivered LNA oligonucleotides in mice. Mol. Ther. 14,

471–475.

19. Sazani, P., Kang, S.H., Maier, M.A., Wei, C., Dillman, J.,

Summerton, J., Manoharan, M., and Kole, R. (2001). Nuclear

antisense effects of neutral, anionic and cationic oligonucle-

otide analogs. Nucleic Acids Res. 29, 3965–3974.

20. Suwanmanee, T., Sierakowska, H., Lacerra, G., Svasti, S.,

Kirby, S., Walsh, C.E., Fucharoen, S., and Kole, R. (2002).

Restoration of human beta-globin gene expression in murine

and human IVS2-654 thalassemic erythroid cells by free

uptake of antisense oligonucleotides. Mol. Pharmacol. 62,

545–553.

21. Wilton, S.D., Lloyd, F., Carville, K., Fletcher, S., Honeyman, K.,

Agrawal, S., and Kole, R. (1999). Specific removal of the

nonsense mutation from the mdx dystrophin mRNA

using antisense oligonucleotides. Neuromuscul. Disord. 9,

330–338.

22. Hua, Y., Vickers, T.A., Okunola, H.L., Bennett, C.F., and

Krainer, A.R. (2008). Antisense masking of an hnRNP A1/A2

intronic splicing silencer corrects SMN2 splicing in transgenic

mice. Am. J. Hum. Genet. 82, 834–848.

23. Dominski, Z., and Kole, R. (1993). Restoration of correct

splicing in thalassemic pre-mRNA by antisense oligonucleo-

tides. Proc. Natl. Acad. Sci. USA 90, 8673–8677.

24. Svasti, S., Suwanmanee, T., Fucharoen, S., Moulton, H.M.,

Nelson, M.H., Maeda, N., Smithies, O., and Kole, R. (2009).

RNA repair restores hemoglobin expression in IVS2-654 thal-

assemic mice. Proc. Natl. Acad. Sci. USA 106, 1205–1210.

25. Xie, S.Y., Li, W., Ren, Z.R., Huang, S.Z., Zeng, F., and Zeng, Y.T.

(2011). Correction of b654-thalassaemia mice using direct

intravenous injection of siRNA and antisense RNA vectors.

Int. J. Hematol. 93, 301–310.

26. Lee, J.H., Engler, J.A., Collawn, J.F., and Moore, B.A. (2001).

Receptor mediated uptake of peptides that bind the human

transferrin receptor. Eur. J. Biochem. 268, 2004–2012.

27. Wilner, S.E., Wengerter, B., Maier, K., de Lourdes Borba

Magalhaes, M., Del Amo, D.S., Pai, S., Opazo, F., Rizzoli,

S.O., Yan, A., and Levy, M. (2012). An RNA alternative to

human transferrin: A new tool for targeting human cells.

Mol. Ther. Nucleic Acids 1, e21.

erican Journal of Human Genetics 94, 611–617, April 3, 2014 617