generation of albino xenopus tropicalis using zinc-finger nuclease

7/31/2019 Generation of Albino Xenopus Tropicalis Using Zinc-finger Nuclease

1/8

Original Article

Generation of albino Xenopus tropicalis using zinc-fingernucleases

Keisuke Nakajima, Taeko Nakajima, Minoru Takase and Yoshio Yaoita*

Division of Embryology and Genetics, Institute for Amphibian Biology, Graduate School of Science, Hiroshima University,

Higashihiroshima, 739-8526, Japan

To generate albino lines of Xenopus tropicalis, we injected fertilized eggs with mRNAs encoding zinc-finger nuc-

leases (ZFNs) targeting the tyrosinase coding region. Surprisingly, vitiligo was observed on the skin of F0 frogs

that had been injected with ZFN mRNAs, indicating that both tyrosinase genes in the genome were disrupted in

all melanocytes within the vitiligo patches. Mutation analysis using genomic DNA from the skin revealed that

two mosaic F0 frogs underwent spatially complex tyrosinase gene mutations. The data implies that the ZFN-

induced tyrosinase gene ablations occurred randomly over space and time throughout the entire body, possibly

until the young tadpole stage, and that melanocyte precursors lacking functional tyrosinase proliferated and

formed vitiligo patches. Several albino X. tropicalis, which are compound heterozygotes for biallelic tyrosinase

mutations, were obtained by mating the mosaic F0 frogs. To our knowledge, this is the first report of the albino

vertebrates generated by the targeted gene knockout.

Key words: albino frog, targeted gene knockout, tyrosinase, Xenopus tropicalis, zinc-finger nuclease.

Introduction

Gene knockout by homologous recombination in

embryonic stem (ES) cells involves the characterization

of altered phenotypes as a means of clarifying thefunctions of the ablated genes. Because of the diffi-

culty in establishing ES cell lines from other species,

this technology has only been used in the mouse gen-

ome (Chisaka & Capecchi 1991) and more recently, in

the rat genome (Tong et al. 2011). The facile gene

knockout technique, which is used to analyze gene

function, is desired for use in other animals. The

knockdown technique, which uses antisense morpholi-

nos (Nasevicius & Ekker 2000), is effective for only the

first few days of early development and is not useful

for analyzing juvenile or adult phenotypes. TILLING is

time-consuming because it requires the detection of a

heterozygous point mutation at a locus of interest

using a chemically mutagenized sperm DNA library,

DNA sequencing to confirm the presence of nonsense

mutations and artificial insemination with the corre-

sponding sperm (Wienholds et al. 2002, 2003).

Zinc-finger nucleases (ZFNs) are fusion proteins

comprised of a zinc-finger DNA-binding sequence and

the nuclease domain from the restriction enzyme Fok I(Kim & Chandrasegaran 1994). ZFNs bind to target

nucleotide sequences and introduce double-strand

breaks, which are repaired by homologous recombina-

tion (Bibikova et al. 2001) and non-homologous end-

joining (NHEJ). The target site may continue to be cut

by the ZFN until NHEJ produces small deletions or

insertions and alters the nucleotide sequence of the

ZFN recognition site. When NHEJ results in a frame

shift, a nonsense mutation, a deletion or a missense

mutation that affects essential amino acids, the target

gene product loses its function. ZFN-induced muta-

genesis has been widely used for targeted mutagene-

sis in the fruit fly (Beumer et al. 2008), nematode

(Morton et al. 2006), zebrafish (Doyon et al. 2008;

Meng et al. 2008), frog (Young et al. 2011), sea urchin

(Ochiai et al. 2010), silk worm (Takasu et al. 2010),

Ciona intestinalis (Kawai et al. 2012) and medaka

(Ansai et al. 2012).

Albinism is a condition found throughout the animal

kingdom, including vertebrates. Tyrosinase is essential

for melanin biosynthesis and converts tyrosine to dop-

aquinone in the initial step of the melanin-producing

pathway. Tyrosinase-negative oculocutaneus albinism

*Author to whom all correspondence should be addressed.

Email: [email protected]

Received 12 August 2012; revised 7 September 2012;

accepted 9 September 2012.

2012 The Authors

Development, Growth & Differentiation 2012 Japanese

Society of Developmental Biologists

Develop. Growth Differ. (2012) doi: 10.1111/dgd.12006

The Japanese Society of Developmental Biologists


2/8

is an autosomal-recessive disease and characterized

by a clinically undetectable level of melanin in skin, hair

and eyes (Kinnear et al. 1985). This form of human

albinism is comparable to the c-albino locus form in

mouse that is caused by the loss of function of the

tyrosinase gene (Beermann et al. 1990; Schedl et al.1993). Therefore, it is reasonable that ZFN-induced

knockout of the tyrosinase gene produces the albino

phenotype.

Our goal is to promote further studies of Xenopus

tropicalis (X. tropicalis) by establishing and distributing

albino lines of X. tropicalis that may facilitate research

techniques, such as in situ hybridization expression

analysis and transplantation experiments (Kashiwagi

et al. 2010). In the present study, we obtained albino

X. tropicalis by injecting mRNAs encoding ZFNs

that target the tyrosinase-coding region into fertilized

eggs.

Materials and methods

Animals

The Ivory Coast line of X. tropicalis was provided by

the National Bio-Resource Project (NBRP) of the

MEXT, Japan. The fertilized eggs were obtained from

pairs of male and female X. tropicalis after an injection

of human chorionic gonadotropin. The tadpoles were

reared in dechlorinated tap water (28C) and fed Sera

Micron (Sera). The frogs were maintained at 24C. All

animals were maintained and used in accordance withthe guidelines established by Hiroshima University for

the use and care of experimental animals.

Construction of zinc-finger nucleases

The design, assembly and MEL-1 assay of ZFNs were

performed by Sigma-Aldrich as previously described

(Urnov et al. 2005; Doyon et al. 2008). The obligate

heterodimer form of the ZFN was used to reduce off-

target effects (Miller et al. 2007).

Single-strand annealing assay

Inverse polymerase chain reaction (PCR) was per-

formed to amplify the entire pRL-CMV vector (Pro-

mega), and LucF and LucR (Table 1) primers were

used to obtain the pRL-CMV-single-strand annealing(SSA) vector containing the 5 and 3 inactive frag-

ments of the Renilla luciferase gene with a 640-bp

region of homologous overlap (Ochiai et al. 2010). The

double-stranded ZFN target sequences (Table 1) were

synthesized and inserted into the EcoR I and Xho I

sites of the pRL-CMV-SSA vector.

The cultured frog kidney cell line A6 was cotransfect-

ed with the ZFN-expressing plasmids, pRL-CMV-SSA

reporter vector and pGL3-Control vector (Promega)

(31.25 ng each) using 0.3 lL of FuGENE 6 Transfec-

tion Reagent (Roche) in each well of 24-well plates.

After three days at 25C, dual-luciferase assays were

conducted using the Dual-Luciferase Reporter Assay

System (Promega) according to the manufacturers

instructions.

RNA microinjection

Zinc-finger nuclease mRNAs (20 pg each) were

injected into X. tropicalis embryos at the one-cell stage

along with 200 pg of mCherry mRNA (Clontech). The

fluorescent product of the latter was used to identify

successfully injected embryos and confirm that the

injected RNA had been translated (Young et al. 2011).

Embryos were raised in 0.19 MMR with 0.1% BSAand 50 lg/mL gentamycin at 22C.

Mutation analysis

Genomic DNA was extracted from tadpole tail fins and

frog skin using SimplePrep reagent for DNA (TaKaRa).

A 1305-bp fragment of X. tropicalis tyrosinase DNA

was amplified by PCR from genomic DNA using TyrF

and TyrR primers (Table 1). The amplified fragment

was inserted into the pGEM-T Easy vector (Promega),

Table 1. Oligonucleotide sequences used in the present study

Oligonucleotide Sequence (5-3) Description

Set-1F AATTGCCCTCAGTTTCCATTCTCTGGGGTTGACGATAGA ZFN set-1 target sequence

Set-1R TCGATCTATCGTCAACCCCAGAGAATGGAAACTGAGGGC

Set-2F AATTGCCCTGGCACAGGTACTTCCTGCTGCACTGGGAACATGAG ZFN set-2 target sequence

Set-2R TCGACTCATGTTCCCAGTGCAGCAGGAAGTACCTGTGCCAGGGC

Set-3F AATTCACAGGTACTTCCTGCTGCACTGGGAACATGAGATTCAG ZFN set-3 target sequence

Set-3R TCGACTGAATCTCATGTTCCCAGTGCAGCAGGAAGTACCTGTG

LucF GACCTCGAGTGACATGGTAACGCGGCCTC Construction of pRL-CMV-SSA

LucR GACGAATTCAGAATCCTGGGTCCGATTC

TyrF GTGAGGAGCAGCATGGAA Mutation analysis

TyrR GCACCCCTACAACAGCCTTC

2012 The Authors

Development, Growth & Differentiation 2012 Japanese Society of Developmental Biologists

2 K. Nakajima et al.


3/8

and the nucleotide sequence was subsequently deter-

mined. The position of a nucleotide in the tyrosinase

gene is denoted as the number of nucleotides from

the translation start site in the open reading frame,

and the A nucleotide of the initiation codon is defined

as position 1.

Results

To disrupt the X. tropicalis tyrosinase gene, ZFN

target-sequence candidates were selected; the

sequences considered were those that would not

result in off-target sites in the genome with four or

fewer mismatched nucleotides. This off-target search

was performed with five to seven spacer nucleotides

because ZFNs preferentially bind to such target

sequences (Handel et al. 2009). Several ZFNs directed

at these target sites were constructed in the expres-

sion vector. Three sets of ZFNs, set-1, set-2 and set-

3, had higher activities in the yeast MEL-1 assay than

other ZFNs (Doyon et al. 2008) (data not shown). Fig-

ure 1A shows the target sites of the three ZFNs as

well as a comparison of the X. tropicalis tyrosinase

amino acid sequence to those of Rana nigromaculata

(the Japanese pond frog), Bufo bufo, Gallus gallus and

Homo sapiens. The identity profile for the five tyrosi-

nase genes revealed a few conserved regions

(Fig. 1B), one of which (residues 367408) may be a

component of the catalytic center (Lerch 1983) and

may be involved in copper binding (Huber et al. 1985).

Three ZFNs were designed to cleave the well-con-served region of the tyrosinase gene upstream of a

putative copper-binding region; this approach was

intended to ensure that these ZFNs would cut the tar-

get site in most cases, even if a non-inbred line of

X. tropicalis was used. The target sites of these ZFNs

are located within the exons of the gene. To reconfirm

the activity of the ZFNs, an SSA assay (Ochiai et al.

2010) was performed by introducing the ZFN-express-

ing vectors and an SSA reporter construct with a tar-

get site into Xenopus A6 cells (Rafferty 1969). Because

set-2 had the highest activity in this assay (Fig. 2) and

the MEL-1 assay, set-2 was used in subsequent

experiments (Fig. 1C).

The set-2 mRNAs were injected along with mCherry

mRNA into the animal hemisphere of fertilized one-

cell-stage embryos. This microinjection procedure did

not have a significant effect on embryo development.

The co-injection of mCherry mRNA facilitated the iden-

tification of successfully injected embryos at hatching

(Young et al. 2011). mCherry-expressing embryos (574

total) were raised, and 41 froglets had some vitiligo

patches. All melanocytes in vitiligo patches should

have biallelic mutations of the tyrosinase gene because

heterozygous mutants on tyrosinase gene locus

appeared normal and homozygous mutants showed

albino phenotype (Fig. 4BD). Nine mosaic frogs were

sexually matured and used for mating (Fig. 3).

Ten albino offspring were obtained from the mating

of mosaic F0 frogs (Fig. 4B,C and Table 2), and thetyrosinase gene mutations were analyzed in six albino

tadpoles (Fig. 4E). Two albino offspring from a cross

using male m1 and female f2 frogs (m1 9 f2) harbored

the same mutations: a one-base deletion at position

653 on one chromosome along with a two-base dele-

tion at position 653654 and a one-base substitution

on the other chromosome. Three albino offspring from

the m2 9 f1 cross also had the same mutations: a

one-base deletion at position 653 and a 367-base

deletion from position 645 to 1011. One albino off-

spring from m2 9 f2 possessed the same tyrosinase

mutations as three albino offspring from m2 9 f1,

although the origin of the one-base deletion is different

between offspring from the m2 9 f2 and m2 9 f1

crosses (see discussion). All mutations resulted in

frame shifts and premature translational termination,

which led to products without a putative copper-bind-

ing region.

Cel-1 assay (Miller et al. 2007) using PCR fragments

of tyrosinase gene revealed that three tadpoles were

heterozygous in six wild-type-colored offspring

obtained from the m1 9 f2 cross (data not shown).

The mutant tyrosinase gene was isolated from one

heterozygous tadpole and demonstrated by sequenc-

ing to contain a two-base deletion at positions 653654 and a one-base substitution on one locus as

observed in albino siblings (Fig. 4E). One tadpole was

shown by Cel-1 assay to be heterozygous in five wild-

type-colored offspring from the m2 9 f1 cross, and

had a 367-base deletion, which was found in albino

siblings. To estimate the mutation rate in the F0 germ

cells, m1 was outcrossed to a wild-type female. Off-

spring from this cross were raised to the neurula

stage, and analyzed by Cel-1 assay for mutations on

the tyrosinase gene locus. Genotyping offspring from a

cross using m1 and a wild-type female clarified that 11

of 19 embryos had inherited a two-base deletion at

positions 653654 and a one-base substitution on one

locus, which were observed in albino and hetero-

zygous offspring from the m1 9 f2 cross (Fig. 4E).

The presence of F0 frogs with several vitiligo patches

demonstrates that the tyrosinase gene was mutated in

a biallelic manner in melanocytes and that the ZFNs

ablated the tyrosinase gene target in a spatially distinct

pattern in each F0 frog. To analyze these changes,

total cellular DNA was extracted. For the m1 frog, skin

samples were collected from the following areas with

vitiligo: a right finger, right forearm and left toe.

2012 The Authors


Albino Xenopus tropicalis 3


4/8

Samples were also collected from wild-type-colored

areas on the left forearm and on a right toe (Fig. 5A).

For the m2 frog, skin samples were collected from the

following areas with vitiligo: a left toe, left thigh and

posterior-most back area. Samples from wild-type-col-

ored areas were collected from a left finger and right

toe (Fig. 5B). All observed mutations are listed in Fig-

ure 5 and varied among albino offspring from the

AATCTCATGTTCCCAGTGCAGCAGGAAGTACCTGTGCCAGGGCACAATTAGAGTACAAGGGTCACGTCGTCCTTCATGGACACGGTCCCGTGTT

Homo sapiens MFNGSCQCTRNYFVSPWSERDDVGTFPFQPGLPANSLLINQCSGRGSLQGCPSRDGSWPPCCEKEMLNKSSVCARPFHGASTQFSWLLCYLVALLM 96

Gallus gallus MFNGRCRCTRNYFVSPWDERDDVGSFPFQPGLPAQSLLIRQCTGRNSREGCPTGDGDWPPCCEKRLLSQTNACVRPFQGTSPQLIVLLLGMAFLFM 96

Bufo bufo MFNGQCQCTRNYFVIPWNERDDVGSFPFQPGISSNSLVVDQCTGRGSAEGCPSGDGAWAPCCEKSLLVEASSCARPFQGHASPLVLLLVAAAFIMISGRM 100

Rana nigromaculata MFNGQCQCTRNYFVIPWNERDDIGSFPFQAGVPSNSPVVDQCVGRGSVEGCSSNDGPWVPCCEKSLLVQATSCARPFQGYSAHLVLLLTSTALLVTTSEM 100

Xenopus tropicalis MFNGLCQCTRNYFVTPWNERDDVGTFPFQPGTASSTLVVDQCVGRGSLQGCSSGDGSWVPCCEKSLLSEATSCARPFQGRCVHLFFFLCCFALPVMNREM 100

Homo sapiens RWIEYGGLLADMSVYYHMWVFLDYINIDNFMPTSGNKMQGYTGIPIVYDSSITHKALTLYAFFKDKEPASLDFINRRVLLRRETCNPGWFGFKCNGCNFG 196

Gallus gallus RWVNSGGLLTDRSAYYHMWVFLDYVNINRFMPNSGNNMQAYTGTAIVYDKSPINKALNLYALFKDKESITLQFINRRTRLRRETCNQGSFGFKCEGCNFG 196

Bufo bufo SWVLGGELFADRSSYYHLWVFLDYVNIDAFLPNSGNNMQEYTGTAIVYDPSITRKALNLYAIFKDKEASTMRFIEKRIMRKRVTCNPSTFGFKCDGCNYG 200

Rana nigromaculata AWVSGDELFADRSAYYHIWVFLDYASIDAFMPNSGNNMQEYTGTAIVYDPSITHKALNLYAILKDKEATTMRFIDKRIMNRRVTCNPGTYGFRCEGCNYG 200

Xenopus tropicalis RWLADEGIFVDRSAYYHMWVFLDYVNIDAFMPNSGNNMQAYTGTVIVYDRSTTHKALNLYAVFKSKEASSMRFIEKRIMTRRETCNPGRFGFKCDACNYG 200

Homo sapiens PGEPTGNCLSQHSNYEELRSCVIQWSSFFSAPSLLNPNTPHQGGMYEDTCIDCKEADRWDWYPITFNEDGTLKQIEQEWRLLFLRHWPLFAPAEHAFDID 296

Gallus gallus PGESTANCLAQQSNYEESQTCIVQWSSFFSAPSLLNPNTPHQGGMYEDTCIVCDEADRWDWYPITFNEDGTIKQIEREWLLLFARHWPLFGPAEHAFDID 296

Bufo bufo PGENTGNCLIRLRNYEEPQSCVVQWSSFFSAPSLLATSTPHIGGFFEDTCVDCQQADRWDWFPITFNEDGTVKQLEHEWQLLFFRHWPLFGPAEHAFDIN 300

Rana nigromaculata PGENTGNCIIRLSNYEEPRSCIIQWSSFFSAPSLLNNSTPHTGGFFEDTCLECQQADRWDWFPITFNDDGTVKQIEREWLLLYFRHWPLFGPAEHAFDID 300

Xenopus tropicalis PGEGTGNCLIRQANYEEPRSCVIQWSAFFSAPSLLSTTTPHVGGLLEDTCIDCGQADRWDWFPITFNEDGTLKQIEHEWHLLFYRHWPVFAPAEHAFDID 300

Homo sapiens ISDVFAHHLLFIPDNASGQVQSMTGNMYIHLANHMSSQSADAIGTLPSAFGELTNRFSFNAAKDMSGSEYQTLSLCFEVDASSPLRPTRSKDHNGPNRRL 396

Gallus gallus ISDVFAHHLIFIPDNASGQVQSMSGNMYIHLANHLGSQSINSIATHPDAFGELTNRFSYNAMKDMSGSEYQTLTLCFEVESSSPLRPTRSKDNNGPNRLI 396

Bufo bufo ISDVFAHHLVFIPDNASGQVSSMSGNLYVHLSNHMNSQSSNAIATRTNAFGELTNRFSFNANRDMPGTEYDTLSLCAEVDASTPLSPTRNRDHRGPSRQL 400

Rana nigromaculata LSDVFAHHLVFIPDNASGQVSSMSGNLFVHLSNHMSSQSRNAIGSTPVAFGELTNRFSFNASRDMAGTEYDTLSLCAEVDASTPLRPTRNRDHRGPSRLL 400

Xenopus tropicalis ISDVFAHHLVFVPDNASGQVSSMSGNLFVHLSNHMNSQSRNAIGTRPDAFGELTNRFSFNASRDMPETEYNTLSLCLEVEATTPLRPTRSRDHGGPNRFL 400

Homo sapiens VLGALLATLVAGVMAAGLLWSWIRSAQELYSKIYDQFSDPDSDQLYSYDYGLDKSSIFFDGNRYLPIFPVMYSERNHGIPANAEPYVEQLPRHRRLWQEF 496

Gallus gallus ILGSLVATIIGGIVAAGVLWPWIQHAQKLYPILFDQFSGLAPEQLYEYDYGLERSSIFFEGNRYLPIFPVMYNERNHGIPANAAPYVELMPRHRRLWREF 496

Bufo bufo VISAIVATILGGVVAAGVLWQWTQRAKELYRSLFDEISD--AEALYDYDYGLDRSQAFFEGNTYLPIFPVMYFGRNHGIPANVEPFVDLSPQHRRLWQEF 498

Rana nigromaculata IITAIVATILGGLVAAGLLWQWIQRAQELYPLLFDEISG--SEALYDYDYGLDRSQVFFEGNTFLPIFPVMNYERNHGVPANAEPYVDLSPQHRRLWQEF 498

Xenopus tropicalis VITAIVATILGGVVAAGVLWQWIQRAQELYPLLFDEISG--SEALYDYDYGLDRSAAFFEGNTYLPIFPVMYYGRNHGIPANAEPYIDVSAGHRRLWQEF 498

LsneipasomoH HSQY-LSHYDEKEMLLPQKEEPL-QKRKHRCLLS 529

Gallus gallus FHSQYSVNNYDESETLLPQIEPST-GKRKKRCAL- 529

Bufo bufo LNSQY-TSQYDEAEMLLPQAEESV-ARKKRRCALS 531

Rana nigromaculata LNSQY-TAQYDEAEMLLPRTEESIKTKRKRRCTLS 532

Xenopus tropicalis LHSQY-TPQYDEAEMLLPQTEESP-FKRKRRCALG 531

Putative copper-binding region

ZFN set-2

ZFN set-3

ZFN set-1

Window length = 31

0 500

0

20

40

60

80

100Putative copper-binding region

ZFN set-1 ZFN set-2, 3

Identity(%)

Amino acid number

(A)

(B) (C)

Fig. 1. The locations of zinc-finger nuclease (ZFN) target sites in the Xenopus tropicalis tyrosinase gene. (A) A comparison of the tyrosi-

nase amino acid sequences of X. tropicalis (NP_001096518), Rana nigromaculata (Q04604), Bufo bufo (CAR95491), Gallus gallus(NP_989491) and Homo sapiens (AAK00805). Conserved amino acids are indicated by yellow boxes. The vertical arrows denote the

boundaries of exons and introns. (B) Identity profile for five tyrosinase genes. Average identity values are plotted against the sequence

position in the peptide chain. Each average is plotted in the middle of the 31 residues from which it is derived. (A,B) The three sites tar-

geted by ZFNs and a putative copper-binding region are indicated by purple and green lines, respectively. (C) The location of the ZFN

set-2 target site is shown in the genomic structure of the X. tropicalis tyrosinase gene. Exons and introns are indicated by black boxes

and lines, respectively. The colored box shows the sequence recognized by each zinc finger of ZFN set-2R and L.

2012 The Authors




5/8

m1 9 f2, m2 9 f1 and m2 9 f2 crosses. These differ-

ences indicate that the m1 and m2 frogs had at least

four different mutations in the tyrosinase gene that

occurred at spatially distinct areas and that the mosaic

F0 frogs are genetic chimeras.

Discussion

We succeeded in generating albino X. tropicalis by the

ablation of the tyrosinase gene. Vitiligo patches were

observed on the skin of F0 frogs that had been

injected with ZFN mRNAs immediately after fertiliza-

tion, indicating that the tyrosinase gene was destroyed

in a biallelic manner in the melanocytes within the viti-

ligo patches. Each of the F0 frogs had a different pat-

tern of vitiligo that should be related to the melanocyte

precursor cells that were affected by the biallelic abla-

tion of the tyrosinase gene. The different sizes of the

vitiligo patches lead us to speculate that large vitiligo

regions are indicative of an early melanocyte precursor

losing both functional copies of the tyrosinase gene

and subsequently undergoing many cell divisions. A

small melanin-free area indicates damage to the tyrosi-

nase gene in a late precursor cell. The mosaic frog m2

had an albino leg, a mosaic leg and several vitiligo

patches on the trunk. Because skin melanocytes

migrate from the neural crest and spread over the

surface of the body, the albino leg indicates that the

biallelic ablation of the tyrosinase gene occurred in a

melanocyte precursor that was eventually distributed

over the entire left leg. There are two possible explana-

tions for the complex mosaic patterns observed on the

F0 frogs. The first explanation is that the ablation of

the second allele occurred in only one melanocyte

precursor, which affected many different locations and

proliferated to form the vitiligo patches. The other is

that a second ablation occurred independently in

several melanocyte precursor cells at different times.

Because the ZFN mRNAs induced several mutations

in individual frogs, as described in the results (Fig. 5),

the latter explanation is more likely.

Fig. 2. The functional activity of zinc-finger nucleases (ZFNs)

estimated using the single-strand annealing (SSA) assay. A6 cells

were transfected, incubated for 3 days at 25C and evaluated for

luciferase activity. The fold activation resulting from the cleavage

of the ZFN target site and single-strand annealing to repair lucif-

erase gene is represented by the relative luciferase activity levels

in cells transfected with the ZFN expression constructs, an SSA

reporter gene with the target site and a reference gene for trans-fection. These activity levels are relative to cells transfected with

the ZFN expression constructs, an SSA reporter gene without

the target site and a reference gene. Data are expressed as the

means standard error of the mean (SEM) (n = 3).

Fig. 3. Mosaic F0 frogs. Photographs of male (m1m4) and

female (f1f5) F0 frogs with vitiligo patches.

2012 The Authors




6/8

One medaka founder injected with ZFN mRNAs was

reported to contain six mutations in the target gene

(Ansai et al. 2012). Our mutation analysis using skin

revealed that the m1 and m2 F0 frogs underwent mul-

tiple mutations of the tyrosinase gene that occurred in

a spatially complex manner. m1 had the common

mutation I in the left wild-type-colored forearm and in

vitiligo patches on the right forearm and a right finger.

It also had mutation II in the left wild-type-colored fore-

arm and mutation III in an unpigmented left toe. The

m2 frog harbored mutations IV and V in the vitiligo

region on its back and mutation VI in a wild-type-col-

ored right toe. It is worth emphasizing that only mela-

nocytes produce melanin and that the skin includes

many other types of cells, such as basal cells, fibro-

blasts, vessels and exocrine gland cells. For this rea-

son, it is sometimes difficult to detect tyrosinase gene

mutations in skin with vitiligo patches (e.g., left leg of

m2) and sometimes possible to find mutations in wild-

type-colored skin (e.g., left forearm of m1). These data

demonstrate that ZFN-induced gene knockout involves

multiple mutation events and spatiotemporally distinct

patterns. These complex chimeric patterns may have

occurred in the m1 and m2 frogs because the transla-

tion of ZFN mRNA and accumulation of ZFN protein in

nuclei take time; thus, ZFNs are most active during

late development rather than during early develop-

ment. This hypothesis is supported by the finding that

the fluorescence of the translational product encoded

by the co-injected mCherry mRNA peaked a few days

later and could be detected for at least 2 weeks under

our experimental conditions (data not shown). How-

ever, differences in the stability of mCherry and ZFN

mRNAs as well as the proteins should be considered.

Embryos develop to the swimming stage within a few

Al m2 x f1 (3)----------------------------------ACGGTCCCGTGTTT 642

TTAGAGTACAAGGGTCACGTCGTCC-TCATGGACACGGTCCCGTGTTT 677

Al m1 x f2 (2)TTAGAGTACAAGGGTCACGTCGT- - GTCATGGACACGGTCCCGTGTTT 676


Ph. TTAGAGTACAAGGGTCACGTCGTCCTTCATGGACACGGTCCCGTGTTT 678

F V P W H R Y F L L H W E H E I

ZFN set-2RZFN set-2L

Al m2 x f2 (1)----------------------------------ACGGTCCCGTGTTT 642


Allotype

A, B

B

A

A

B

B

B

WT m1 x f2 (1) TTAGAGTACAAGGGTCACGTCGT- - GTCATGGACACGGTCCCGTGTTT 676

TTAGAGTACAAGGGTCACGTCGTCCTCATGGACACGGTCCCGTGTTT 677T

(A)

(E)

(B) (C) (D)

Fig. 4. Mutation analysis in albino F1. (AD) Photographs of F1 froglets. (A) A wild-type froglet. (B) An albino froglet from a cross using

m1 and f2 frogs. (C) An albino froglet from the m2 9 f1 cross. (D) A wild-type-colored froglet that is a heterozygous offspring from the

m1 9 f2 cross. (E) The tyrosinase gene mutations in F1. The wild-type tyrosinase amino acid sequence is shown at the top. The nucleo-

tide sequences of tyrosinase genes in offspring from the m1 9 f2, m2 9 f1 and m2 9 f2 crosses are compared with the wild-type

sequence. Numbers of characterized offspring are shown in parentheses. The phenotypes of F1 (Ph.) are indicated as wild-type (WT)

and albino (Al). Deletions and a substitution are indicated by red dashes and a red letter, respectively. The recognition sequences of

ZFN set-2 are underlined with purple lines. There are two allotypes, A and B, for the tyrosinase gene locus in the Ivory Coast line of Xen-

opus tropicalis. Allotype B has a nine-base deletion in the intron from position 1102 to 1110 that is not present in allotype A. The tyrosi-

nase allotype is indicated on the right-hand side.

Table 2. The occurrence of albino F1 tadpoles from the mating

of mosaic F0 frogs. The ratio of albino tadpoles to total hatched

tadpoles for each mating is indicated.

F0 female

f1 f2 f3 f4 f5

F0 male m1 0/22 4/88

m2 3/962 1/322

m3 2/218 0/147

m4 0/1047 0/362

2012 The Authors




7/8

days and to NF-stage 49 within 2 weeks. It is possible

that the ablation of the tyrosinase gene occurs in a

spatially random fashion until the young tadpole stage.

Therefore, vitiligo patches are observed when melano-

cyte precursors undergo biallelic ablation, proliferateand spread out. Even if the mutation of the target

gene is detected in a part of the body, this does not

necessarily mean that the F0 germ cells contain the

knockout gene because the F0 is a spatially and

genetically complex chimera.

The tyrosinase mutations in the albino F1 tadpoles

were analyzed in the present study. The mutation in

the m1 germ cells is a two-base deletion at positions

653654 and a one-base substitution on the allotype

A locus because this mutation was shared in all het-

erozygous offspring produced from a cross using m1

and a wild-type female. This means that the mutation

of f2 is a one-base deletion at position 653 on allotype

B, as albino offspring from the m1 9 f2 harbored both

mutations. The deletion of 367 bases is derived from

m2 because this large deletion is contained besides

the f2 mutation (a one-base deletion) in an albino off-

spring from m2 9 f2, and observed in offspring from

m2 9 f1 and m2 9 f2 but not in albino offspring from

m1 9 f2. In the same manner, the one-base deletion

at position 653 on allotype A is derived from the f1

frog. The one-base deletion at position 653 must have

occurred independently two times because this

deletion is located on allotype A in f1, while this muta-

tion is on allotype B in f2.

We obtained a small number of albino offspring by

mating the F0 frogs. One possible explanation is that

only some of the germ cells are mutated in the maleand female F0 frogs. In this scenario, the ability to

obtain albino F1 offspring decreases synergistically.

The ratio of albino siblings to total siblings from

m1 9 f2 is 4.5% and the mutation frequency in the

germ cells of m1 is 58% from the outcross experi-

ment, which suggests that the mutation rate in the f2

germ cells is approximately 10%. These mutagenesis

efficiencies are comparable to the mutation rate (12

24%) reported by Young et al. (2011). However, the

possibility that albino F1 offspring have reduced sur-

vival rates in early development cannot be excluded.

We demonstrated that F0 frogs injected with ZFN

mRNAs exhibit spatially variable disruption of the target

gene throughout the body. Our data suggest that the

most efficient gene-knockout corresponds to high lev-

els of ZFN activity as soon as possible in the embryos.

We hope that the albino X. tropicalis generated in this

study will contribute to further developmental studies.

References

Ansai, S., Ochiai, H., Kanie, Y., Kamei, Y., Gou, Y., Kitano, T.,

Yamamoto, T. & Kinoshita, M. 2012. Targeted disruption of

TCGTCCTTCATTCG- --- TCAT

TCGTCC--CGTTCGTGTCAT - -

TCGTCCTTCATTCG----TTCG

-

-

TCAT

--

---

TCCTT

-------- --

TCG- -- TCAT A

CA AGGCA AGGCA AGG

AA G

----

----------- TCGT

(A) (B)

Fig. 5. Mutation analysis in mosaic F0 frogs. The tyrosinase gene mutations in m1 (A) and m2 (B) frogs. (A, B) The observed mutations

in m1 and m2 skin at the indicated locations, and the mutations in sperm that are predicted from mutations in the F1 offspring (Fig. 4).

The spacer sequence is indicated by a bracket. The wild-type nucleotide sequence is shown at the top, and the predicted mutation in

sperm is shown at the bottom. Deletions and substitutions are indicated by red dashes and red letters, respectively. An insertion is indi-

cated by a green letter. Numbers in the photograph indicate the numbers of mutated and wild-type clones among those analyzed. For

the analyzed clones, the frequencies of mutations I, II, III, IV, V and VI along with the wild-type sequence for the tyrosinase gene are

denoted by red, yellow, green, orange, purple, magenta and blue, respectively, in each circle graph. Open and closed squares indicate

that genomic DNA was isolated from the skin in areas with vitiligo patches and wild-type-colored areas, respectively.

2012 The Authors




8/8

exogenous EGFP gene in medaka using zinc-finger nucleas-

es. Dev. Growth Differ. 54, 546556.

Beermann, F., Ruppert, S., Hummler, E., Bosch, F. X., Muller,

G., Ruther, U. & Schutz, G. 1990. Rescue of the albino phe-

notype by introduction of a functional tyrosinase gene into

mice. EMBO J. 9, 28192826.

Beumer, K. J., Trautman, J. K., Bozas, A., Liu, J. L., Rutter, J.,Gall, J. G. & Carroll, D. 2008. Efficient gene targeting in Dro-

sophila by direct embryo injection with zinc-finger nucleases.

Proc. Natl. Acad. Sci. USA 105, 1982119826.

Bibikova, M., Carroll, D., Segal, D. J., Trautman, J. K., Smith, J.,

Kim, Y. G. & Chandrasegaran, S. 2001. Stimulation of

homologous recombination through targeted cleavage by

chimeric nucleases. Mol. Cell Biol. 21, 289297.

Chisaka, O. & Capecchi, M. R. 1991. Regionally restricted devel-

opmental defects resulting from targeted disruption of the

mouse homeobox gene hox-1.5. Nature 350, 473479.

Doyon, Y., Mccammon, J. M., Miller, J. C., Faraji, F., Ngo, C.,

Katibah, G. E., Amora, R., Hocking, T. D., Zhang, L., Rebar,

E. J., Gregory, P. D., Urnov, F. D. & Amacher, S. L.2008.

Heritable targeted gene disruption in zebrafish using

designed zinc-finger nucleases. Nat. Biotechnol. 26, 702

708.

Handel, E. M., Alwin, S. & Cathomen, T. 2009. Expanding or

restricting the target site repertoire of zinc-finger nucleases:

the inter-domain linker as a major determinant of target site

selectivity. Mol. Ther. 17, 104111.

Huber, M., Hintermann, G. & Lerch, K. 1985. Primary structure

of tyrosinase from Streptomyces glaucescens. Biochemistry

24, 60386044.

Kashiwagi, K., Kashiwagi, A., Kurabayashi, A., Hanada, H., Nak-

ajima, K., Okada, M., Takase, M. & Yaoita, Y. 2010. Xeno-

pus tropicalis: an ideal experimental animal in amphibia. Exp.

Anim. 59, 395405.

Kawai, N., Ochiai, H., Sakuma, T., Yamada, L., Sawada, H., Ya-

mamoto, T. & Sasakura, Y. 2012. Efficient targeted muta-

genesis of the chordate Ciona intestinalis genome with zinc-finger nucleases. Dev. Growth Differ. 54, 535545.

Kim, Y. G. & Chandrasegaran, S. 1994. Chimeric restriction

endonuclease. Proc. Natl. Acad. Sci. USA 91, 883887.

Kinnear, P. E., Jay, B. & Witkop Jr, C. J. 1985. Albinism. Surv.

Ophthalmol. 30, 75101.

Lerch, K. 1983. Neurospora tyrosinase: structural, spectroscopic

and catalytic properties. Mol. Cell Biochem. 52, 125138.

Meng, X., Noyes, M. B., Zhu, L. J., Lawson, N. D. & Wolfe, S. A.

2008. Targeted gene inactivation in zebrafish using engi-

neered zinc-finger nucleases. Nat. Biotechnol. 26, 695701.

Miller, J. C., Holmes, M. C., Wang, J., Guschin, D. Y., Lee, Y.

L., Rupniewski, I., Beausejour, C. M., Waite, A. J., Wang, N.

S., Kim, K. A., Gregory, P. D., Pabo, C. O. & Rebar, E. J.

2007. An improved zinc-finger nuclease architecture for

highly specific genome editing. Nat. Biotechnol. 25, 778

785.

Morton, J., Davis, M. W., Jorgensen, E. M. & Carroll, D. 2006.

Induction and repair of zinc-finger nuclease-targeted double-

strand breaks in Caenorhabditis elegans somatic cells. Proc.Natl. Acad. Sci. USA 103, 1637016375.

Nasevicius, A. & Ekker, S. C. 2000. Effective targeted gene

knockdown in zebrafish. Nat. Genet. 26, 216220.

Ochiai, H., Fujita, K., Suzuki, K., Nishikawa, M., Shibata, T., Sa-

kamoto, N. & Yamamoto, T. 2010. Targeted mutagenesis in

the sea urchin embryo using zinc-finger nucleases. Genes

Cells 15, 875885.

Rafferty Jr, K. A. 1969. Mass culture of amphibian cells: Methods

and observations concerning stability of cell type. In: Biology

of Amphibian Tumors. (ed. Mizell M) New York: Springer-

Verlag, pp. 5281.

Schedl, A., Montoliu, L., Kelsey, G. & Schutz, G. 1993. A yeast

artificial chromosome covering the tyrosinase gene confers

copy number-dependent expression in transgenic mice. Nat-

ure 362, 258

261.Takasu, Y., Kobayashi, I., Beumer, K., Uchino, K., Sezutsu, H.,

Sajwan, S., Carroll, D., Tamura, T. & Zurovec, M. 2010. Tar-

geted mutagenesis in the silkworm Bombyx mori using zinc

finger nuclease mRNA injection. Insect Biochem. Mol. Biol.

40, 759765.

Tong, C., Huang, G., Ashton, C., Li, P. & Ying, Q. L. 2011. Gen-

erating gene knockout rats by homologous recombination in

embryonic stem cells. Nat. Protoc. 6, 827844.

Urnov, F. D., Miller, J. C., Lee, Y. L., Beausejour, C. M., Rock, J.

M., Augustus, S., Jamieson, A. C., Porteus, M. H., Gregory,

P. D. & Holmes, M. C. 2005. Highly efficient endogenous

human gene correction using designed zinc-finger nucleas-

es. Nature 435, 646651.

Wienholds, E., Schulte-Merker, S., Walderich, B. & Plasterk, R.

H. 2002. Target-selected inactivation of the zebrafish rag1gene. Science 297, 99102.

Wienholds, E., Van Eeden, F., Kosters, M., Mudde, J., Plasterk,

R. H. & Cuppen, E. 2003. Efficient target-selected mutagen-

esis in zebrafish. Genome Res. 13, 27002707.

Young, J. J., Cherone, J. M., Doyon, Y., Ankoudinova, I., Faraji,

F. M., Lee, A. H., Ngo, C., Guschin, D. Y., Paschon, D. E.,

Miller, J. C., Zhang, L., Rebar, E. J., Gregory, P. D., Urnov,

F. D., Harland, R. M. & Zeitler, B. 2011. Efficient targeted

gene disruption in the soma and germ line of the frog Xeno-

pus tropicalis using engineered zinc-finger nucleases. Proc.

Natl. Acad. Sci. USA 108, 70527057.

2012 The Authors



generation of albino xenopus tropicalis using zinc-finger nuclease

Documents