FEBS Letters 581 (2007) 2625–2632

dXNP, a Drosophila homolog of XNP/ATRX, induces apoptosis viaJun-N-terminal kinase activation

Nam Gon Leea,1, Yoon Ki Honga,1, Si Young Yua, Seung Yeop Hana, Dongho Geumb,*,Kyoung Sang Choa,*

a Department of Biological Sciences, Konkuk University, Seoul 143-701, Republic of Koreab Graduate School of Medicine, Korea University, Seoul 136-707, Republic of Korea

Received 14 February 2007; revised 27 April 2007; accepted 3 May 2007

Available online 11 May 2007

Edited by Francesc Posas

Abstract XNP/ATRX, a causative gene of X-linked a-thalasse-mia/mental retardation syndrome, encodes an SNF2 familyATPase/helicase protein. To better understand the role ofXNP/ATRX in development, we isolated and characterized aDrosophila XNP/ATRX homolog, dXNP, which contains highlyconserved SNF2 and helicase domains. Ectopically expresseddXNP induced strong apoptosis in the developing eye and wing,but did not affect cell cycle progression or the expression of wing-less and engrailed, essential regulators of development. ThedXNP-induced apoptosis was strongly suppressed by DJNKK/hemipterous mutation, and dXNP increased JNK activity.Taken together, these results suggest that dXNP regulates apop-tosis via JNK activation.� 2007 Federation of European Biochemical Societies. Publishedby Elsevier B.V. All rights reserved.

Keywords: Apoptosis; ATRX; Drosophila; dXNP; JNK; XNP

1. Introduction

Mutations in XNP/ATRX cause several X-linked mental

retardation syndromes that also feature facial dysmorphism,

urogenital defects, and a-thalassemia [1,2]. In males, somatic

mutations of this gene also cause a-thalassemia myelodysplasia

syndrome (OMIM 300448) [3]. Female carriers, however, are

physically and intellectually normal [1].

XNP/ATRX, which was first isolated as a gene poten-

tially coding for a nuclear protein [4], encodes a member of

the SNF2 family proteins with ATPase and helicase domains

that are most similar to those of RAD54 [5,6]. XNP/ATRX

is also known to associate with nuclear matrix and chroma-

tin [7,8]. Recently, it was shown that XNP/ATRX forms

a complex, which displays ATP-dependent chromatin

Abbreviations: ATRX, a-thalassemia X-linked mental retardation;JNK, Jun-N-terminal kinase; XNP, X-linked nuclear protein

*Corresponding authors. Fax: +82 2 3436 5432 (K.S. Cho); +82 2 9295696 (D. Geum).E-mail addresses: geumd@korea.ac.kr (D. Geum), kscho@konkuk.

ac.kr (K.S. Cho).

1The authors equally contributed to this work.

0014-5793/$32.00 � 2007 Federation of European Biochemical Societies. Pu

doi:10.1016/j.febslet.2007.05.005

remodeling activities including triple-helix DNA displace-

ment and alteration of mononucleosome disruption pat-

terns [9], and was found to be a part of the NuRD complex

[10].

A series of studies have shown that XNP/ATRX is involved

in diverse cellular processes, including DNA methylation, tran-

scription, cell cycle, and apoptosis [11–14]. The methylation

pattern of several highly repeated sequences changes in the

genomic DNA of ATR-X syndrome patients’ lymphocytes

[11]. The fusion protein of XNP/ATRX and GAL4 DNA

binding domain strongly inhibits the expression of thymidine

kinase promoter driven luciferase [12]. BrdU incorporation

and histone H3 phsophorylation, which are cell cycle markers

for S phase and G2/M phase, respectively, are altered in trans-

genic XNP/ATRX embryos [13]. In addition, loss of XNP/

ATRX protein by conditional gene-targeting in mice causes

a 12-fold increase in neuronal apoptosis during early stages

of corticogenesis, and resulted in hypocellularity in the mutant

brain [14].

Despite the XNP/ATRX-related cellular processes have been

revealed, their molecular mechanisms and the role in develop-

ment are still remain to be determined. Here, to uncover the

function of XNP/ATRX in developing tissues, we isolated

and characterized Drosophila XNP/ATRX homologue (dXNP).

The ectopically expressed dXNP by UAS-GAL4 system in

developing tissues induced apoptosis in a cell autono-

mous manner. Genetic and biochemical analysis showed that

the dXNP-induced apoptosis is mediated by activation of

JNK.

2. Materials and methods

2.1. Drosophila strainsGlass multimer reporter (GMR)-GAL4, patched (ptc)-GAL4, apterous

(ap)-GAL4, UAS-DIAP1, UAS-puckered, UAS-lacZ, wingless (wg)-lacZ, engrailed (en)-lacZ, dXNP1 (EP635), and dXNP2 (UY3132) wereobtained from the Bloomington Drosophila Stock Center. MS1096-GAL4 and hemipterous1 were gifts from Dr. M. Freeman (MRC Lab-oratory of Molecular Biology, UK) and Dr. S. Noselli (CNRS,France), respectively.

2.2. Ectopic gene expression using UAS-GAL4 systemTo ectopically overexpress dXNP, we used the modular misexpres-

sion system based on a P-element vector carrying a GAL4-regulatedpromoter [15–17]. In dXNP1 and dXNP2 lines, the P-elements are in-serted in the flanking region of dXNP so that GAL4-dependent tran-scription begins within the P-element and extends out into the

blished by Elsevier B.V. All rights reserved.

2626 N.G. Lee et al. / FEBS Letters 581 (2007) 2625–2632

dXNP. GMR-GAL4, MS1096-GAL4, ptc-GAL4, and ap-GAL4 direc-ted dXNP expression in the developing eye, the whole developing wing(strongly in the dorsal part), the border of anterior/posterior compart-ment of wing, and the dorsal compartment of wing, respectively.

External eye morphologies were observed by scanning electronmicroscopy (Korea Basic Science Institute, Korea).

Fig. 1. Drosophila XNP and its mutants. (A) Comparison of Drosophila XDrosophila and Human XNP. Abbreviations: SNF2 N-terminal domain (SNorthern blot analysis of dXNP transcripts during Drosophila development.third instar larva; P, pupa; M, adult male; F, adult female. (C) Real time RT-Upper panel shows the relative transcript levels determined by real time RT-Pthe body (**P < 0.01 vs. body by paired Student’s t-test; n = 3). Lower paneRT-PCR products were resolved on 1% agarose gels and visualized by ethidrawn as boxes and coding regions are highlighted in black. The P-elemenexpression of dXNP using UAS-GAL4 system. (E) Determination of expressdXNP alleles using RT-PCR method. Actin was used as an internal controlGAL4/+), GMR > dXNP1 (GMR-GAL4/+; dXNP1/+), and GMR > dXNP2 (Gthe induction of dXNP mRNA expression in the eye imaginal discs of GMR-GGAL4/+, F) or GMR > dXNP (GMR-GAL4/+; dXNP2/+, G).

2.3. Acridine orange (AO) staining and BrdU labeling experimentEye and wing imaginal discs of stage L3 larvae were dissected in

phosphate-buffered saline (PBS). The discs were then incubated for5 min in 1.6 · 10�6 M solution of acridine orange (Aldrich, USA),and rinsed briefly in PBS. The samples were examined under an Axiop-hot2 fluorescent microscope (Carl Zeiss, Germany).

NP with its human counterpart. Protein domains conserved betweenNF2_N) and helicase superfamily C-terminal domain (HELICc). (B)18S rRNA was used as a loading control (lower panel). E, embryo; L,PCR analysis of dXNP transcripts in the body and head of Drosophila.CR analysis, normalized with Actin, and expressed as mean ± S.E.% of

l shows the representative picture for the RT-PCR products of dXNP.dium bromide staining. (D) Genomic structures of dXNP. Exons aret insertion sites of dXNP1 and dXNP2 are indicated. (E–G) Ectopic

ion levels of dXNP transcript in the eye imaginal discs of GAL4 driven(lower panel). The genotypes of the samples are GMR-GAL4 (GMR-MR-GAL4/+; dXNP2/+). (F,G) RNA in situ hybridization shows thatAL4 driven dXNP alleles. Samples were prepared from control (GMR-

N.G. Lee et al. / FEBS Letters 581 (2007) 2625–2632 2627

5-Bromo-20-deoxyuridine (BrdU) labeling experiment of wing imag-inal discs were performed using the in situ cell proliferation kit (Roche,Germany) according to the manufacturer’s protocol.

2.4. RT-PCR, RNA in situ hybridizations, and Northern blot analysisTotal RNAs from flies were isolated by TRIZOL-bromo-chloropro-

pane protocol. For RT-PCR, one microgram of RNA templates werereverse-transcribed with 200 U of Moloney murine leukemia virus(MMLV) reverse transcriptase (Promega, WI) using manufacturer’sinstructions. RT-PCR was performed using the following primer pairs:dXNP (5 0ATGGGAAAGAAAAACCCCAACG30 for 5 0 and 5 0TTG-CGACTTTTGCGAGCCG3 0 for 3 0), Actin (5 0TTAGCTCAGCCT-CGCCACTTGCG3 0 for 5 0 and 5 0GAGCACAGCCTGGATGGCC3 0

for 3 0), hep (5 0GATTGCAGAAGTGCGACAAA3 0 for 5 0 and5 0GGCTTCTTGGACAGCTTGAG3 0 for 3 0), bsk (5 0AAATGC-TCGCCACTTTGAGT3 0 for 5 0 and 5 0TGGCTGTAACCGTTG-CATAA30 for 3 0), DJun (5 0ACACACCCGATTTGGAGAAG3 0 for5 0 and 5 0TGATGAATAAATAAGTTTAAGGCGTA3 0 for 3 0), andDfos (5 0CAGGACACGACCGATACTT3 0 for 5 0 and 5 0ATT-TGAGGTATTCGCGTTGC3 0 for 3 0). RT products were amplified28 times by PCR.

In situ hybridization and Northern blot analysis was performedusing standard method using the probe made from the PCR productthat corresponds to the 3 0 region of the dXNP amplified with the prim-ers (5 0TTCTCATGGCTCCTCCAGCC3 0 for 5 0 and 5 0TCATCCTC-GTCGGTTTCGTC30 for 3 0).

2.5. Real time RT-PCRA 2 ll aliquot of each RT sample of head and body was subjected to

real time PCR in a 20 ll reaction mixture containing 4 mM MgCl2,10 pmol of upstream and downstream primers and 2 ll of 10· Light-

Fig. 2. The effect of dXNP overexpression in developing tissues. Scanning ewing (D–L) resulted from overexpression of dXNP gene are shown. Asterisgenotypes of the samples are (A) GMR-GAL4/+, (B) GMR-GAL4/+; dXNP1

(E,H) MS1096-GAL4/Y; dXNP1/+, (F,I) MS1096-GAL4/Y; dXNP2/+, (J) ptc

Cycler FastStrart� DNA Master SYBR Green 1 (Roche, Germany).Data were analyzed with LightCycler Software version 3.5. Same prim-ers were used as conventional RT-PCR.

2.6. Immunohistochemistry and X-gal stainingDrosophila eye or wing imaginal discs were dissected and fixed in 4%

paraformaldehyde in PBS and then washed with PBT (PBS + 0.1%Tween 20) and blocked in PBT with 2% Normal Goat Serum. Thediscs were incubated overnight with rabbit anti-active-Drice antibody(1:200) [18], rabbit anti-phospho-JNK antibody (Promega, 1:200) orrabbit anti-phospho-histone 3 antibody (Promega, 1:200) and thenwashed in PBT and subsequently further incubated for 1 h at roomtemperature in FITC or rhodamine-labeled goat anti-rabbit IgG(H + L) secondary antibody (1:200 in PBT). The samples were ob-served by a Axiophot2 fluorescent microscope (Carl Zeiss, Germany).

For 5-bromo-4-chloro-3-indolyl-b-DD-galactopyranoside (X-Gal)staining, eye and wing imaginal discs were fixed in 4% formaldehydein PBS for 4 min, washed and then incubated in standard X-Gal stain-ing solution [4.9 mM X-Gal, 3.1 mM K4Fe(CN)6, 3.1 mM K3Fe(CN)6,1 mM MgCl2, 150 mM NaCl, 10 mM Na2HPO4, 10 mM NaH2PO4,0.3% Triton X-100] for 30 min at 37 �C before observation.

3. Results

3.1. Drosophila XNP and its mutants

The dXNP gene, which corresponds to the Berkeley Dro-

sophila Genome Project annotated sequence CG4548, was iso-

lated by PCR-based cloning. The conceptually translated

dXNP protein (1311 amino acids) contains the characteristic

lectron micrographs of the external eyes (A–C) and the phenotypes ofks in (K,L) indicate reduced distance between L3 and L4 veins. The/dXNP1, (C) GMR-GAL4/+; dXNP2/dXNP2, (D,G) MS1096-GAL4/Y,-GAL4/+, (K) ptc-GAL4/+; dXNP1/+, and (L) ptc-GAL4/+; dXNP2/+.

2628 N.G. Lee et al. / FEBS Letters 581 (2007) 2625–2632

motifs, a SNF2 family N-terminal domain (SNF2_N) and heli-

case superfamily C-terminal domain (HELICc) (Fig. 1A). The

C-terminal region of dXNP, which contains these domains,

shows 66% homology (48% identity) to the human counter-

part. As these domains have been known to be important in

chromatin remodeling, it is very likely that dXNP may func-

tion as a chromatin remodeling factor just as mammalian

XNP/ATRX. Consistent with the sequence results, we detected

a single �4.0 kb-transcript of dXNP in all developmental

stages from Northern blot analysis (Fig. 1B). Real time RT-

PCR showed that dXNP is expressed in the whole body, and

Fig. 3. Induction of apoptosis by expression of dXNP. AO stainings of eye (A(D) and wing (H,L) imaginal discs are shown. The genotypes of the samples a+; dXNP2/+, (D) GMR-GAL4/UAS-lacZ, (E) MS1096-GAL4/Y, (F) MS109GAL4/Y; +/UAS-lacZ, (I) ptc-GAL4/+, (J) ptc-GAL4/+; dXNP1/+, (K) ptphenotypes caused by ectopic expression of dXNP in the developing eye can bGAL4/+; dXNP1/+, (O) GMR-GAL4/+; dXNP1/+; UAS-DIAP1/+. (P,Q) Ant(P) and GMR-GAL4/+; dXNP2/+ (Q).

its expression is elevated approximately 2-fold in the head

compared with that of the body (220 ± 10.9% of the body,

p < 0.01) (Fig. 1C), implying that dXNP may play an impor-

tant role in the head.

With the help of the Berkeley Drosophila Genome Project,

we found two P-element insertion lines, dXNP1 and dXNP2,

which contain P-elements in either the 5 0 upstream or first exon

of the dXNP gene (Fig. 1D). The directions of the P-elements

are oriented to induce gene expression, implying that these mu-

tants can be used to study the gain-of-function of dXNP [16].

As expected, strong expression of dXNP was observed in the

–C) and wing (E–G and I–K) imaginal discs, and X-gal stainings of eyere (A) GMR-GAL4/+, (B) GMR-GAL4/+; dXNP1/+, (C) GMR-GAL4/6-GAL4/Y; dXNP1/+, (G) MS1096-GAL4/Y; dXNP2/+, (H) MS1096-c-GAL4/+; dXNP2/+, and (L) ptc-GAL4/UAS-lacZ. (M–O) The eyee suppressed by coexpression of DIAP1. (M) GMR-GAL4/+, (N)GMR-i-active Drice antibody staining of eye imaginal discs of GMR-GAL4/+

N.G. Lee et al. / FEBS Letters 581 (2007) 2625–2632 2629

eye imaginal discs of dXNP1 and dXNP2 lines carrying the

GMR-GAL4 driver by RT-PCR (Fig. 1E) and RNA in situ

hybridization (Fig. 1F and G).

3.2. dXNP induces apoptosis

To verify the function of dXNP in Drosophila development,

we examined the effect of overexpression of dXNP on develop-

ing tissues using the UAS-GAL4 system. Overexpression of

dXNP in the developing eye under the control of GMR-

GAL4 destroyed all ommatidia (Fig. 2A–C). In addition, we

examined the effect of overexpression of dXNP in a specific

compartment of the wing. When dXNP is ectopically overex-

pressed in the entire wing pouch with MS1096-GAL4 driver,

the wings were rumpled and diminished (Fig. 2D–I). And,

the ectopically expressed dXNP in the border of anterior/pos-

terior compartment of wing by ptc-GAL4 resulted in a reduc-

tion of the distance between L3 and L4 veins (Fig. 2J–L).

These results suggest that dXNP is essential for proper tissue

growth during development.

As dXNP overexpression reduces the compartment size of

wing, we tested whether overexpression of dXNP causes apop-

tosis by acridine orange (AO) staining. As expected, we found

strong staining signals in all dXNP overexpressed tissues

(Fig. 3A–C, E–G, and I–K). The AO staining patterns are

found in the region showing GAL4-driven gene expression,

which is reported by lacZ expression (Fig. 3D, H, and L), sug-

gesting that dXNP induced apoptosis in a cell autonomous

manner. Furthermore, dXNP-induced destruction of the eye

was almost completely inhibited by coexpression of Drosophila

inhibitor of apoptosis protein 1 (DIAP1) (Fig. 3M–O), suggest-

ing that the apoptosis induced by dXNP is mediated by caspase

activation. Supporting this idea, we found strong activation of

Drice, an essential caspase of Drosophila [18], in the eye imag-

inal disc of dXNP over-expressed fly (Fig. 3Q) but not in the

control fly (Fig. 3P).

Fig. 4. dXNP does not affect cell cycle and expression of develop-mental regulators. Cell cycle was determined by anti-phospho-histoneH3 antibody staining (A,B) or BrdU incorporation analysis (C,D), andtranscriptional level of compartment determining gene was determinedby X-gal staining (E–H). (A,C) MS1096-GAL4/Y, (B) MS1096-GAL4/Y; dXNP2/+, (D) MS1096-GAL4/Y; dXNP1/+, (E) MS1096-GAL4/Y;wg-lacZ/+, (F) MS1096-GAL4/Y; wg-lacZ/+; dXNP1/+, (G)MS1096-GAL4/Y; en-lacZ/+, (H) MS1096-GAL4/Y; en-lacZ/+; dXNP1/+.

3.3. dXNP does not affect the cell differentiation and cycle

The reduction of tissue compartment also can be a result of

defective cell cycle or gene expression. Therefore, we tested

whether overexpressed dXNP cause impairment of these cellu-

lar processes. However, there was no difference between the

control and dXNP overexpressed tissues in antibody staining

for phospho-Histone H3 (Fig. 4A and B), a marker for mitosis,

or incorporation of BrdU, a marker for S phase (Fig. 4C and

D). These results imply that cell cycle is not affected by dXNP

overexpression. Next, we examined whether the expression of

genes which are essential for development is altered by elevated

dXNP level. To test this possibility, we checked the expression

of two compartment determining factors, wingless and en-

grailed, in dXNP-overexpressed wing discs using reporter gene

system. As shown in Fig. 4E–H, we did not find any difference

in the expression level of these genes in dXNP-overexpressed

tissues, implying that dXNP overexpression do not affect these

gene expression as well as cell cycle.

3.4. dXNP induces apoptosis through JNK signaling pathway

As activation of the JNK signaling pathway is frequently

correlated with induction of apoptosis, and chromatin remod-

eling proteins has been implicated in the JNK-induced apopto-

sis [19], we tested whether JNK signaling pathway mediates

dXNP-induced apoptosis. First, we examined the genetic inter-

action between hep1, a Drosophila JNKK, and dXNP. As de-

scribed previously, overexpressed dXNP destroys ommatidia

of compound eye (Fig. 5B and E), while hep1, a hypomorphic

mutant of hep, male exhibits normal compound eye (Fig. 5A

and D). However, interestingly, the dXNP-induced destruction

Fig. 5. dXNP induces apoptosis through the activation of JNK signaling pathway. Scanning electron micrographs of the external eyes (A–F), and AOstaining (G-I) and anti-phospho-JNK antibody staining (J,K) of the eye imaginal discs are shown. (A, D, and G) hep1/Y (B, E, and H) GMR-GAL4/+;dXNP2/+, (C, F, and I) hep1/Y; GMR-GAL4/+; dXNP2/+, (J) GMR-GAL4/+, (K) GMR-GAL4/+; dXNP2/dXNP2. (L) Real time RT-PCR analysis ofChic and Puc transcripts in the heads of GMR-GAL4/+ (GMR) and GMR-GAL4/+; dXNP2/+ (X2) flies. Upper panel shows the relative transcript levelsdetermined by real time RT-PCR analysis, normalized with Actin, and expressed as means ± S.E.% of the heads of GMR-GAL4/+ (*P < 0.05 and**P < 0.01 vs. body by paired Student’s t-test; n = 3). Lower panel shows the representative picture for the RT-PCR products of Chic, Puc and Actin.RT-PCR products were resolved on 1% agarose gels and visualized by ethidium bromide staining. (M) Determination of expression levels of thecomponents of JNK signaling pathway in the eye imaginal discs using RT-PCR. The genotypes of the samples are GMR-GAL4 (GMR-GAL4/+),GMR > dXNP1 (GMR-GAL4/+; dXNP1/+), and GMR > dXNP2 (GMR-GAL4/+; dXNP2/+). Actin was used as an internal control (lower panels).

2630 N.G. Lee et al. / FEBS Letters 581 (2007) 2625–2632

N.G. Lee et al. / FEBS Letters 581 (2007) 2625–2632 2631

of the compound eye was strongly suppressed by the hep1

mutation (Fig. 5C and F). Furthermore, as shown in

Fig. 5G-I, the dXNP-induced apoptosis was also inhibited by

hep1, suggesting that JNK signaling pathway acts as a vital

downstream effector of dXNP in the regulation of apoptosis.

Consistent with these results, we also observed that puckered,

a negative regulator of JNK, suppressed the dXNP-induced

wing defect (Fig. S1). In addition, we found that overexpres-

sion of dXNP increases phosphorylation of JNK in both eye

(Fig. 5J and K) and wing (Fig. S2) imaginal discs, and the

expression of direct targets of JNK signaling pathway such

as chickadee (Chic) and puckered (Puc) (Fig. 5L). The expres-

sions of Chic and Puc were increased 1.8-fold (179 ± 29.8%

of the WT, p < 0.05, n = 3) and 2.2-fold (224 ± 14.6% of the

WT, p < 0.01, n = 3), respectively, in the dXNP overexpressed

samples. These results suggest that dXNP regulates the JNK

signaling pathway.

Because XNP has been known as a chromatin remodeling

factor, we tested whether dXNP increases the activity of

JNK signaling pathway by controlling transcription of its com-

ponents. To do this, the transcript levels of bsk, hep, Dfos, and

DJun were measured by quantitative RT-PCR in the eye discs

of the GMR-GAL4 controls and dXNP overexpression flies.

However, as shown in Fig. 5M, their transcript levels were

not altered by overexpression of dXNP, implying that activa-

tion of the JNK signaling pathway is not resulted from an in-

crease in the transcriptional level of signaling components but

probably by the post-transcriptional regulation, such as phos-

phorylation.

4. Discussion

In the present study, we have isolated and characterized

Drosophila XNP (dXNP). Overexpression of dXNP induced

strong apoptosis without affecting cell cycle and the expression

level of essential regulators of development. Co-expression of

DIAP1 suppressed dXNP-induced rough eye phenotype

(Fig. 3M–O). As it has been reported that DIAP1 degrades

two proteins, Drosophila caspase DRONC [20] and TRAF1

[21] as its E3 substrate, our result suggest that the caspase cas-

cade is implicated in the apoptosis induced by dXNP. Support-

ing this idea, Drice, a downstream caspase of DRONC [22],

was activated in dXNP-overexpressing tissues (Fig. 3Q). De-

spite the detailed mechanism by which XNP/ATRX involved

in apoptosis is unclear, our study clearly showed that the apop-

tosis induced by dXNP is mediated by caspase cascade. It was

interesting to note that XNP/ATRX is localized in promyelo-

cytic leukemia (PML) nuclear bodies [23], which have been

implicated in apoptosis and senescence [24]. Moreover, the

accumulated XNP/ATRX at PML nuclear bodies interacts

with Daxx, a multifunctional apoptosis-related protein

[23,25], and strong apoptosis was observed in the XNP/

ATRX-null mice [14], implying that XNP/ATRX is a potent

regulator of apoptosis. Since, as a chromatin remodeling fac-

tor, XNP has been associated with DNA methylation and

transcription [11,13], it is possible that dXNP regulates apop-

tosis through direct control of transcription of proapoptotic

factors. However, our study argues that dXNP induces apop-

tosis by activating JNK signaling pathway. dXNP-induced

apoptosis was almost completely suppressed by hep/DJNKK

mutation, suggesting that the apoptosis might be resulted from

the activation of JNK signaling pathway. Moreover, we also

found that overexpressed dXNP increases phosphorylation of

JNK (Fig. 5K) and the expression of direct targets of JNK

pathway (Fig. 5L). These results demonstrate that dXNP reg-

ulates apoptosis via regulating the JNK signaling pathway

rather than via direct regulation of the transcription of apop-

totic factors.

Then, how dXNP can activate JNK? One possible scenario

is that dXNP activates JNK by controlling the transcription

of upstream regulator of JNK. Although we showed that the

transcriptional level of hep/DJNKK is not altered by dXNP

overexpression, we can not exclude the possibility that the

transcription of other upstream regulator(s) of JNK signaling

pathway is changed. Further study to find dXNP targets will

determine this possibility. Alternatively, other interesting sce-

nario is that Daxx can mediate the dXNP-induced JNK activa-

tion and apoptosis. Recently, it has been reported that

Ras-association domain family 1C (RASSF1C), a binding

partner of Daxx, is released from the nucleus to cytoplasm

and activates JNK signaling pathway, when Daxx is degraded

in response to DNA damage [26]. Because Daxx associates

with XNP or RASSF1C in PML nuclear bodies [9,26], overex-

pression of XNP may abrogate the balance between Daxx-

XNP and Daxx-RASSF1C complexes and increase free

RASSF1Cs which, in turn, activate the JNK signaling path-

way. Since Drosophila RASSF homolog has been identified

and characterized recently [27], it may be interesting to test

whether Daxx and RASSF are involved in the dXNP-induced

JNK activation and apoptosis.

In summary, we have provided genetic and histochemical

evidences that dXNP controls apoptosis in a JNK-dependent

manner, which may explain the underlying mechanism of some

symptoms of ATR-X syndrome, such as mental retardation. In

addition, we here provide evidence for the first time that

dXNP, which was previously known as a chromatin remo-

deler, could also play a role as a signal transducer. Elucidating

the molecular mechanism through which dXNP regulates JNK

signaling and apoptosis might be helpful to understand the

developmental role of XNP and this complex disorder.

Acknowledgements: This work was supported by the faculty researchfund of Konkuk University in 2005. We thank Jeehye Park (KAIST,Korea) for critical reading of this manuscript.

Appendix A. Supplementary data

Supplementary data associated with this article can be

found, in the online version, at doi:10.1016/j.febslet.2007.

05.005.

References

[1] Gibbons, R. (2006) Alpha thalassaemia-mental retardation, Xlinked. Orphanet J. Rare Dis. 1, 15.

[2] Villard, L. and Fontes, M. (2002) Alpha-thalassemia/mentalretardation syndrome, X-Linked (ATR-X, MIM #301040, ATR-X/XNP/XH2 gene MIM #300032). Eur. J. Hum. Genet. 10, 223–225.

[3] Gibbons, R.J., Pellagatti, A., Garrick, D., Wood, W.G., Malik,N., Ayyub, H., Langford, C., Boultwood, J., Wainscoat, J.S. andHiggs, D.R. (2003) Identification of acquired somatic mutations

2632 N.G. Lee et al. / FEBS Letters 581 (2007) 2625–2632

in the gene encoding chromatin-remodeling factor ATRX in thealpha-thalassemia myelodysplasia syndrome (ATMDS). Nat.Genet. 34, 446–449.

[4] Gecz, J., Pollard, H., Consalez, G., Villard, L., Stayton, C.,Millasseau, P., Khrestchatisky, M. and Fontes, M. (1994) Cloningand expression of the murine homologue of a putative human X-linked nuclear protein gene closely linked to PGK1 in Xq13.3.Hum. Mol. Genet. 3, 39–144.

[5] Stayton, C.L., Dabovic, B., Gulisano, M., Gecz, J., Broccoli, V.,Giovanazzi, S., Bossolasco, M., Monaco, L., Rastan, S. andBoncinelli, E. (1994) Cloning and characterization of a newhuman Xq13 gene, encoding a putative helicase. Hum. Mol.Genet. 3, 1957–1964.

[6] Picketts, D.J., Higgs, D.R., Bachoo, S., Blake, D.J., Quarrell,O.W. and Gibbons, R.J. (1996) ATRX encodes a novel member ofthe SNF2 family of proteins: mutations point to a commonmechanism underlying the ATR-X syndrome. Hum. Mol. Genet.5, 1899–1907.

[7] McDowell, T.L., Gibbons, R.J., Sutherland, H., O’Rourke,D.M., Bickmore, W.A., Pombo, A., Turley, H., Gatter, K.,Picketts, D.J., Buckle, V.J., Chapman, L., Rhodes, D. and Higgs,D.R. (1999) Localization of a putative transcriptional regulator(ATRX) at pericentromeric heterochromatin and the short armsof acrocentric chromosomes. Proc. Natl. Acad. Sci. USA 96,13983–13988.

[8] Berube, N.G., Smeenk, C.A. and Picketts, D.J. (2000) Cell cycle-dependent phosphorylation of the ATRX protein correlates withchanges in nuclear matrix and chromatin association. Hum. Mol.Genet. 9, 539–547.

[9] Xue, Y., Gibbons, R., Yan, Z., Yang, D., McDowell, T.L., Sechi,S., Qin, J., Zhou, S., Higgs, D. and Wang, W. (2003) The ATRXsyndrome protein forms a chromatin-remodeling complex withDaxx and localizes in promyelocytic leukemia nuclear bodies.Proc. Natl. Acad. Sci. USA 100, 10635–10640.

[10] Cardoso, C., Couillault, C., Mignon-Ravix, C., Millet, A.,Ewbank, J.J., Fontes, M. and Pujol, N. (2005) XNP-1/ATR-Xacts with RB, HP1 and the NuRD complex during larvaldevelopment in C. elegans. Dev. Biol. 278, 49–59.

[11] Gibbons, R.J., McDowell, T.L., Raman, S., O’Rourke, D.M.,Garrick, D., Ayyub, H. and Higgs, D.R. (2000) Mutations inATRX, encoding a SWI/SNF-like protein, cause diversechanges in the pattern of DNA methylation. Nat. Genet. 24,368–371.

[12] Tang, J., Wu, S., Liu, H., Stratt, R., Barak, O.G., Shiekhattar, R.,Picketts, D.J. and Yang, X. (2004) A novel transcriptionregulatory complex containing death domain-associated proteinand the ATR-X syndrome protein. J. Biol. Chem. 279, 20369–20377.

[13] Berube, N.G., Jagla, M., Smeenk, C., De Repentigny, Y.,Kothary, R. and Picketts, D.J. (2002) Neurodevelopmentaldefects resulting from ATRX overexpression in transgenic mice.Hum. Mol. Genet. 11, 253–261.

[14] Berube, N.G., Mangelsdorf, M., Jagla, M., Vanderluit, J.,Garrick, D., Gibbons, R.J., Higgs, D.R., Slack, R.S. and Picketts,D.J. (2005) The chromatin-remodeling protein ATRX is criticalfor neuronal survival during corticogenesis. J. Clin. Invest. 115,258–267.

[15] Brand, A.H. and Perrimon, N. (1993) Targeted gene expression asa means of altering cell fates and generating dominant pheno-types. Development 118, 401–415.

[16] Rorth, P. (1996) A modular misexpression screen in Drosophiladetecting tissue-specific phenotypes. Proc. Natl. Acad. Sci. USA93, 12418–12422.

[17] Nicolai, M., Lasbleiz, C. and Dura, J.M. (2003) Gain-of-functionscreen identifies a role of the Src64 oncogene in Drosophilamushroom body development. J. Neurobiol. 57, 291–302.

[18] Yoo, S.J., Huh, J.R., Muro, I., Yu, H., Wang, L., Wang, S.L.,Feldman, R.M., Clem, R.J., Muller, H.A. and Hay, B.A. (2002)Hid, Rpr and Grim negatively regulate DIAP1 levels throughdistinct mechanisms. Nat. Cell Biol. 4, 416–424.

[19] Miotto, B., Sagnier, T., Berenger, H., Bohmann, D., Pradel, J.and Graba, Y. (2006) Chameau HAT and DRpd3 HDACfunction as antagonistic cofactors of JNK/AP-1-dependent tran-scription during Drosophila metamorphosis. Genes Dev. 20, 101–112.

[20] Wilson, R., Goyal, L., Ditzel, M., Zachariou, A., Baker, D.A.,Agapite, J., Steller, H. and Meier, P. (2002) The DIAP1 RINGfinger mediates ubiquitination of Dronc and is indispensable forregulating apoptosis. Nat. Cell Biol. 4, 445–450.

[21] Kuranaga, E., Kanuka, H., Igaki, T., Sawamoto, K., Ichijo, H.,Okano, H. and Miura, M. (2002) Reaper-mediated inhibition ofDIAP1-induced DTRAF1 degradation results in activation ofJNK in Drosophila. Nat. Cell Biol. 4, 705–710.

[22] Hawkins, C.J., Yoo, S.J., Peterson, E.P., Wang, S.L., Vernooy,S.Y. and Hay, B.A. (2000) The Drosophila caspase DRONCcleaves following glutamate or aspartate and is regulated byDIAP1, HID, and GRIM. J. Biol. Chem. 275, 27084–27093.

[23] Ishov, A.M., Vladimirova, O.V. and Maul, G.G. (2004) Hetero-chromatin and ND10 are cell-cycle regulated and phosphoryla-tion-dependent alternate nuclear sites of the transcriptionrepressor Daxx and SWI/SNF protein ATRX. J. Cell Sci. 117,3807–3820.

[24] Takahashi, Y., Lallemand-Breitenbach, V., Zhu, J. and de The,H. (2004) PML nuclear bodies and apoptosis. Oncogene 23, 2819–2824.

[25] Salomoni, P. and Khelifi, A.F. (2006) Daxx: death or survivalprotein? Trends Cell Biol. 16, 97–104.

[26] Kitagawa, D., Kajiho, H., Negishi, T., Ura, S., Watanabe, T.,Wada, T., Ichijo, H., Katada, T. and Nishina, H. (2006) Releaseof RASSF1C from the nucleus by Daxx degradation links DNAdamage and SAPK/JNK activation. EMBO J. 25, 3286–3297.

[27] Polesello, C., Huelsmann, S., Brown, N.H. and Tapon, N. (2006)The Drosophila RASSF homolog antagonizes the hippo pathway.Curr. Biol. 16, 2459–2465.