Báo cáo khoa học: Targeted disruption of one of the importin a family members leads to female functional incompetence in delivery docx

The mammalian importin a5 subfamily has higher homology with plant and fungal importin a than the other mammalian importin a isoforms, suggesting that the other importin a isoform genes

members leads to female functional incompetence in

delivery

Tetsuji Moriyama1, Masahiro Nagai2, Masahiro Oka1,2,3, Masahito Ikawa4, Masaru Okabe4and Yoshihiro Yoneda1,2,3

1 Department of Frontier Biosciences, Graduate School of Frontier Biosciences, Osaka University, Japan

2 Department of Biochemistry, Graduate School of Medicine, Osaka University, Japan

3 JST, CREST, Graduate School of Frontier Biosciences, Osaka University, Japan

4 Department of Experimental Genome Research, Research Institute for Microbial Diseases, Osaka University, Japan

Introduction

In eukaryotic cells, the nuclear and cytoplasmic

com-partments are separated by the nuclear envelope The

nuclear envelope contains nuclear pore complexes that

allow macromolecules to be exchanged between the

two compartments The nucleocytoplasmic transport

system functions as a key mediator of signal

transduc-tion by regulating protein localizatransduc-tion The nuclear

import of proteins generally depends on the presence

of specific signal sequences called nuclear localization

signals (NLSs), and the basic type of NLS is

recog-nized by an importin a⁄ b heterodimer and targeted to nuclear pores Importin b possesses affinity for nucleo-porins, which are components of the nuclear pore complex that mediate nuclear import On the other hand, the importin a family generally binds to both the nuclear import cargo and importin b, indicating that importin a functions as an adaptor between the cargo proteins and importin b In the nucleus, the import complex encounters the GTP-bound form

of Ran (RanGTP), which is a member of the Ras

Keywords

estrogen; gene knockout; importin a;

nuclear transport; reproduction

Correspondence

Y Yoneda, Department of Frontier

Biosciences, Osaka University, Graduate

School of Frontier Biosciences, Osaka

University, 1-3 Yamada-oka, Suita, Osaka

565-0871, Japan

Fax: +81 6 6879 4609

Tel: +81 6 6879 4606

E-mail: yyoneda@anat3.med.osaka-u.ac.jp

(Received 30 October 2010, revised 10

February 2011, accepted 22 February 2011)

doi:10.1111/j.1742-4658.2011.08079.x

Importin a mediates the nuclear import of proteins through nuclear pore complexes in eukaryotic cells, and is common to all eukaryotes Previous reports identified at least six importin a family genes in mice Although these isoforms show differential binding to various import cargoes in vitro, the in vivo physiological roles of these mammalian importin a isoforms remain unknown Here, we generated and examined importin a5 knockout (impa5) ⁄ )) mice These mice developed normally, and showed no gross his-tological abnormalities in most major organs However, the ovary and uterus of impa5) ⁄ ) female mice exhibited hypoplasia Furthermore, we found that impa5) ⁄ )female mice had a 50% decrease in serum progester-one levels and a 57% decrease in progesterprogester-one receptor mRNA levels in the ovary Additionally, impa5) ⁄ ) uteruses that were treated with exoge-nous gonadotropins displayed hypertrophy, similarly to progesterone recep-tor-deficient mice Although these mutant female mice could become pregnant, the total number of pups was significantly decreased, and some

of the pups were dead at birth These results suggest that importin a5 has essential roles in the mammalian female reproductive organs

Abbreviations

EBAG9, estrogen receptor-binding fragment-associated antigen 9; EFP, estrogen-responsive finger protein; ER, estrogen receptor; FRT, FLP recombinase target; FSHR, follicle-stimulating hormone receptor; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; hCG, human chorionic gonadotropin; impa5) ⁄ ), importin a5 knockout; LHR, luteinizing hormone receptor; Ltf, lactotransferrin; NLS, nuclear localization signal; PMSG, pregnant mare serum gonadotropin; PR, progesterone receptor; SEM, standard error of the mean.

superfamily that localizes to the nucleus, and this

inter-action causes the cargo protein to dissociate from the

complex [1,2] It has been reported that there is only a

single importin a gene in budding yeast, whereas at

least six importin a family genes have been found in

mice and humans These importin a molecules are

clas-sified into three subtypes on the basis of their sequence

homology The importin a1 subfamily in mice consists

of importin a1 (karyopherin a2, PTAC58, Rch1); the

importin a3 subfamily includes importin a3

(karyoph-erin a4, Qip1) and importin a4 (karyopherin a3,

Qip2); and the importin a5 subfamily includes

impor-tin a5 (karyopherin a1, NPI-1) and imporimpor-tin a7

(kary-opherin a6) [3] Additionally, very recently, a novel

importin a isoform (importin a8, karyopherin a7) was

identified that is expressed during oocyte maturation

and early embryonic development [4]

These importin a isoforms have distinct binding

characteristics for various NLS-containing proteins

in vitro [5–7] Furthermore, previous studies indicated

that the importin a isoforms are differentially expressed

in adult mouse and human tissues [8–10] More

recently, it was reported that these transport factors

function as major players in determining cell fate [11]

Thus, these results suggest that each importin a

iso-form may contribute to a variety of physiological

func-tions in multicellular organisms Indeed, in the fruit fly

Drosophila melanogaster, which expresses three classes

of importin a homologs in unique temporal and spatial

patterns, it has been shown that mutants lacking any

single importin a isoform have defects in gametogenesis

[12–16], indicating that all of the importin a isoforms

are required for germline development In addition,

importin a1 (mammalian importin a5 subfamily

homo-log) is required for normal wing development [12], and

importin a2 (mammalian importin a1 subfamily

homo-log) is involved in synapse, axon and muscle

develop-ment [17,18] Importin a3 (mammalian importin a3

subfamily homolog) is required for flies to mature into

adults and for tiling of photoreceptor axons in the

visual system [14,19] Moreover, it was very recently

reported that destruction of mouse importin a8 causes

a significant reduction in fertility and fecundity [20]

The mammalian importin a5 subfamily has higher

homology with plant and fungal importin a than the

other mammalian importin a isoforms, suggesting that

the other importin a isoform genes in mammals arose

from an importin a5-like progenitor [1] Although

Shmidt et al reported that importin a5 mutant mice

did not exhibit any obvious morphological or

behav-ioral abnormalities [21], these mice have not been

ana-lyzed in detail To gain further insights into the in vivo

physiological significance of importin a5 in mammals,

we generated importin a5 knockout (impa5) ⁄ )) mice using the Cre–lox system, which differs from the method used by Shmidt et al., and analyzed them in detail Here, we report that an importin a5 deficiency affects the female reproductive organs and causes func-tional deterioration of the female reproductive tract

Results

Targeted disruption of the mouse importin a5 gene

To study the physiological significance of importin a5

in mammals, we used gene targeting to generate impa5) ⁄ )mice (Fig 1A) Because exons 2 and 3 of the importin a5 gene encode the translation start site and importin b-binding site, we disrupted these areas with

a Cre–loxP system Targeted ES cell clones were identi-fied by PCR (Fig 1B) and Southern blotting (Fig 1C), and were used to generate impa5) ⁄ )mice as described

in Experimental procedures The absence of the impor-tin a5 protein was confirmed by western blot analysis with tissue lysates from impa5) ⁄ )mice (Fig 1D) Intercrossing between the heterozygous parents pro-duced homozygous knockout animals in the expected Mendelian ratio (wild-type⁄ heterozygote ⁄ homozygous knockout = 19 : 35 : 13) Both male and female mutant mice developed normally and showed no appar-ent gross developmappar-ental abnormalities (Fig 1E,F) A previous study with impa5) ⁄ ) mice demonstrated that importin a4 was markedly upregulated in the brain, suggesting that the counter-regulation of another im-portin a isoform may compensate for the lack of a single isoform in vivo in mammals [21] Therefore, to determine whether the lack of importin a5 affects the expression of other importin a isoforms in our impa5) ⁄ ) mice, we compared the protein expression levels of each importin a isoform in various tissues from impa5) ⁄ )and wild-type mice by western blotting There were no obvious differences in the expression of other importin a isoforms between impa5) ⁄ )and wild-type mice (Fig S1)

Genital hypoplasia in impa5) ⁄ )female mice Tissue sections from impa5) ⁄ )and wild-type mice were compared for three pairs of male and female animals Histological analyses showed that impa5) ⁄ ) mice had

no gross abnormalities in the brain (Fig 2A), spinal cord, sciatic nerve, thymus, lung, heart, liver (Fig 2B), pancreas, mammary gland, testis, vagina (Fig 2C), etc (Fig S2) Analyses of hematological and biochemical parameters showed mild increases in aspartate

aminotransferase and alanine aminotransferase levels,

and decreases in total cholesterol levels and platelet

counts (Tables S1 and S2), suggesting a slight

deteriora-tion in liver funcdeteriora-tion However, a detailed histological

analysis and apoptosis test with the terminal

deoxynu-cleotidyl transferase dUTP nick end labeling method

did not show any abnormalities in the liver These

results indicate that loss of importin a5 does not

obvi-ously affect the organization and function of most

organs However, we noticed that the reproductive

tracts in all impa5) ⁄ ) females had crucial differences

from wild-type female mice The impa5) ⁄ )ovary had a

reduced number of growing follicles at the maturation

stage (Fig 2D) The uteruses of impa5) ⁄ )mice had thin

myometrial, stromal and epithelium layers, and

imma-ture endometrial glands, as compared with the uterine morphology of wild-type female mice (Fig 2E) In order to elucidate the cause of the abnormalities observed in the reproductive tracts of impa5) ⁄ )females,

we examined the pattern of importin a5 protein expres-sion in wild-type ovary and uterus by immunohisto-chemistry (Fig 3) Abundant importin a5 signals were observed in both the ovary and uterus of wild-type female mice, but not in sections prepared from impa5) ⁄ ) female mice Interestingly, importin a5 was strongly expressed in granulosa cells of ovaries (Fig 3A), and in the luminal and glandular epithelium of the uterus (Fig 3B) These expression patterns suggest that impor-tin a5 may have especially important functions in the maturation of the ovum and uterine epithelial layers

A

B

D

C

Fig 1 Generation of importin a5-deficient mice (A) Schematic representation of homologous recombination of the targeting vector and recombination steps The numbered closed boxes denote the translated exons of the gene (B) PCR analysis for the confirmation of homolo-gous recombination of the short arm side Genomic DNA isolated from ES clones was used as a template A 2.9-kb band was detected in the targeted allele but not in the wild-type allele (C) Southern blot analysis for the confirmation of homologous recombination of the long arm side PvuII–PacI-restricted DNA yielded 12-kb and 9.2-kb bands for wild-type and recombinant alleles, respectively The small box in (A) represents the DNA probe used to screen for homologous recombination of the long arm side (D) Immunoblotting analysis of importin a5 protein expression Importin a5 and GAPDH protein expression was detected by immunoblotting with 15 lg of various tissue lysates from impa5) ⁄ )and wild-type mice Arrowhead: importin a5 protein band *Nonspecific band (E, F) Growth curves for male (E) and female (F) impa5) ⁄ ), impa5 + ⁄ )and wild-type mice Each mouse was weighed 1–8 weeks after birth Error bars indicate the standard deviation KO,

knockout; TK, thymidine kinase; WT, wild-type.

To determine the effects of importin a5 disruption

on reproduction, the fertility of impa5) ⁄ ) mice was

examined For 28 days, impa5) ⁄ ) and impa5+⁄)

females were mated with impa5) ⁄ )or impa5+⁄)males,

and the numbers of pregnant female mice, pups and

live pups were counted Mating of impa5) ⁄ ) females

with wild-type males resulted in significantly smaller

litter sizes (Table 1) Furthermore, we found that

impa5) ⁄ )female mice had significantly increased

num-bers of dead pups in their cages after delivery The mean number of live pups born to impa5+⁄) females was 6.8 ± 1.5, whereas impa5) ⁄ )females had an aver-age litter size of 1.3 ± 2.7 (P < 0.001) In addition, most of the dead pups had twisted bodies and⁄ or bite marks (Fig 4A)

To determine when these pups died, we analyzed embryonic day 18.5 embryos However, they appeared

to develop normally, and we did not observe dead embryos at this embryonic stage On the other hand, impa5) ⁄ )females had vaginal bleeding near the time of delivery (Fig 4B), and five of 17 impa5) ⁄ )females died

as a result of severe bleeding In particular, one female died while a pup remained trapped within her vagina (Fig 4D) In addition, some females appeared to take

a significant amount of time to deliver their pups (Fig 4C) These results indicate that impa5) ⁄ )females had severe difficulty in delivering their pups, suggesting that the depressed reproductive organ functions of impa5) ⁄ ) females damaged the pups during delivery, and led to a decreased litter size and reduced pup sur-vival In contrast, impa5) ⁄ )male, impa5+ ⁄ ) male and impa5+⁄)female mice were as fertile as wild-type mice, and they had comparable litter survival rates These results indicate that loss of the importin a5 gene causes not only morphological but also functional deteriora-tion of the female reproductive tract

Reduced serum progesterone levels in impa5) ⁄ ) female mice

The female ovaries mature in response to cycling sex hormones In particular, 17-b-estradiol stimulates the proliferation of uterine layer cells, suggesting that impa5) ⁄ )female mice may have imbalanced 17-b-estra-diol levels However, steroid hormone measurements with sensitive enzyme immunoassays revealed that the serum 17-b-estradiol levels were comparable between impa5) ⁄ )and wild-type females (Fig 5A) In contrast,

we found that impa5) ⁄ )mice had significantly reduced progesterone levels, by 50%, as compared with wild-type mice (Fig 5B) The reduction in serum progester-one is consistent with the decrease in the number of mature follicles in the ovaries of impa5) ⁄ ) mice, because progesterone is produced specifically after ovu-lation from the corpus luteum in the ovary

Abnormal uterine development in impa5) ⁄ ) females after treatment with exogenous gonadotropin

To gain insights into the defective reproductive organs

of impa5) ⁄ )females and determine whether these mice

A

B

C

D

E

Fig 2 Histological analysis of impa5) ⁄ ) (left panel) and wild-type

(right panel) mice (A) Brain (B) Liver (C) Vagina (D) Ovary (two

impa5) ⁄ )female ovaries) (E) Uterus CER, cerebral cortex; ep,

epi-thelial layer; HPC, hippocampus; o, oriens layer; pr, pyramidal cell

layer; r, stratum radiatum; st, stromal layer; ug, uterine gland.

Tissue sections of impa5) ⁄ )and wild-type mice were stained with

hematoxylin and eosin.

ovulate normally, 4-week-old mice were given

exoge-nous gonadotropins, including pregnant mare serum

gonadotropin (PMSG) and human chorionic

gonado-tropin (hCG) After the hormone treatments, impa5) ⁄ )

mice produced almost the same number of mature

oocytes as wild-type mice (9.9 ± 3.3 for impa5) ⁄ ),

n= 7; 10.3 ± 3.8 for wild type, n = 6) Additionally,

the volumes of impa5) ⁄ ) ovaries were significantly

enlarged after hormone treatment, and the numbers of

growing follicles increased to levels that were

compara-ble to those in control ovaries (Fig 6A) These results

imply that disruption of importin a5 does not lead to

defects in oogenesis, but results in decreased

respon-siveness of ovary cells to sex hormones Furthermore,

we found that the uteruses of gonadotropin-treated

impa5) ⁄ ) mice had abnormal uterine structures, and

that the luminal epithelium and endometrial stroma

appeared hyperplastic, as compared with wild-type

controls (Fig 6B) It is of note that these histological

changes were similar to the previously reported

pheno-types of uteruses from progesterone receptor

(PR)-deficient mice that were treated with estrogen and

progesterone [22], raising the possibility that PR expression is particularly suppressed in impa5) ⁄ )mice

Decreased expression of genes downstream of the estrogen receptor (ER)

Next, to further examine the possibility that PR expression is reduced in impa5) ⁄ ) mice, we examined the mRNA expression levels of not only PR but also ERa, ERb, follicle-stimulating hormone receptor (FSHR) and luteinizing hormone receptor (LHR) in the ovary by quantitative real-time PCR (Fig 7A) The gene expression levels for ERa, ERb, FSHR and LHR were not different between impa5) ⁄ ) and wild-type mice, whereas PR expression was significantly downregulated, by 57%, in impa5) ⁄ ) mice as com-pared with wild-type mice, indicating that importin a5 plays an essential role in regulating expression of the

PR gene Because estrogen plays a crucial role in regu-lating PR in target tissues, and the proximal promoter

of the PR gene possesses several estrogen-responsive elements [23], our data suggest that loss of importin a5 leads to the downregulation of ER signaling in PR-expressing cells and subsequent suppression of PR

To examine this possibility, we examined the mRNA expression levels of genes that are downstream of ER, such as those encoding ER-binding fragment-associ-ated antigen 9 (EBAG9) [24], estrogen-responsive fin-ger protein (EFP) [25], and lactotransferrin (Ltf) [26],

by quantitative real-time PCR (Fig 7B) Although the expression levels of the follicle-stimulating hormone-responsive gene encoding cyclin D2 [27,28] were not significantly different between impa5) ⁄ ) and wild-type mice, EBAG9 and EFP expression was significantly downregulated in impa5) ⁄ ) mice, by  20% These findings indicate that importin a5 is prominently involved in gene regulation by ER and its cofactors

On the other hand, the protein levels in the uterus

D B

Fig 3 Immunohistochemistry for

impor-tin a5 expression in ovarian and uterine

sec-tions Ovarian and uterine sections prepared

from wild-type (A, B) and impa5) ⁄ )(C, D)

female mice were stained for importin a5.

ep, epithelial layer; G, granulosa cell layer;

O, oocyte; st, stromal layer; ug, uterine

gland.

Table 1 Fertility data of wild-type, heterozygous and homozygous

male and female impa5) ⁄ ) mice Each pair (male ⁄ female = 1 : 1)

was transferred to a mating cage for 28 days The cages were

monitored daily and for an additional 28 days, and the numbers

of pregnant female mice, pups and live pups were counted.

*P < 0.05.

Genotype

(importin a5)

Pregnancy rate

Litter size (mean ± SEM)

Litter survival rate, % (no.)

(Fig 7C) and ovary and the subcellular localization of

ERa in the uterus (Fig 7D) were not different between

impa5) ⁄ )and wild-type mice, suggesting that importin

a5 does not affect the nuclear import of ER

Discussion

The impa5) ⁄ ) females showed depressed reproductive

organ functions, such as a reduced number of growing

follicles at the maturation stage in the ovary and

immature layer construction in the uterus, and decreased levels of serum progesterone Furthermore, administration of exogenous gonadotropin restored follicle growth in the ovary and the release of oocytes

in impa5) ⁄ ) females, although their uteruses showed hypertrophy (see discussion below) In addition, analy-sis of the mRNA expression levels of estrogen-depen-dent genes in impa5) ⁄ ) ovaries revealed that the transcriptional activity of ER was downregulated It is

Fig 4 Photographs of impa5) ⁄ )mice after the delivery date A series of photographs show the cage (A) and impa5) ⁄ )mice after (B) and during (C) delivery, and a dead impa5) ⁄ )female mouse with pups trapped within the birth canal (D) (C) This impa5) ⁄ ) female took at least 2 days to give birth, and all of her pups were dead (D) This dead impa5) ⁄ )mother still had two undelivered pups in her uterus The open arrowheads indicate the dead pups, and the filled arrow-heads indicate the bleeding point.

Fig 5 Serum 17-b-estradiol and progesterone levels in female

mice Serum samples were collected, and the 17-b-estradiol and

progesterone levels were measured (A) Serum 17-b-estradiol levels

in impa5) ⁄ )and wild-type mice The 17-b-estradiol values were 9.4

and 10.4 pgÆmL)1for female impa5) ⁄ )and wild-type mice,

respec-tively, with P = 0.653 (B) Serum progesterone levels in impa5) ⁄ )

and wild-type mice The progesterone values were 1.25 ngÆmL)1

and 2.51 ngÆmL)1for impa5) ⁄ )and wild-type females, respectively,

with P = 0.015 *P < 0.05; impa5) ⁄ ) mice, n = 8; wild-type mice,

n = 8 Error bars indicate the SEM.

A

B

Fig 6 Histological analysis of reproductive organs from 4-week-old impa5) ⁄ )mice that were induced to superovulate (A, B) Histologi-cal analysis of the (A) ovary and (B) uterus from 4-week-old impa5) ⁄ )mice that were treated with PMSG and hCG Tissue sec-tions from impa5) ⁄ )and wild-type mice were stained with hema-toxylin and eosin.

generally accepted that ER-mediated transcriptional

and biological activation requires the recruitment of a

number of cofactors, including SRC-1, CBP⁄ p300,

TRAP220, ASC-1, SRA, and p68 [29], which facilitate

a functional interaction between the receptor and the general transcription machinery Our results showed that there are no differences between impa5) ⁄ ) and wild-type mice in the amount and localization of the ERa protein, suggesting that importin a5 may specifi-cally mediate the nuclear import of at least some of these cofactors, although we cannot completely exclude the possibility that disruption of importin a5 reduces the import efficiency of ER On the other hand, mice knocked out for over 200 genes have shown reproduc-tive defects as a major phenotype; the genes include encoding those encoding transcription factors and nuclear proteins, such as C⁄ EBPb, p27kip1, and cyclin D2 [30] Accordingly, the defects observed in the reproductive organs of impa5) ⁄ ) mice could result from the combined effects of the inefficient nuclear import of such factors

The number of pups born to impa5) ⁄ ) females was clearly reduced This phenotype could result from vari-ous causes, including the dysfunction of the ovary and⁄ or uterus The impa5) ⁄ ) ovaries had a reduced number of growing follicles Several studies have reported that estrogen augments the effects of follicle-stimulating hormone on granulosa cells [31], granulosa cell growth, and the number of granulosa cells in the ovary [32,33] Our data showed that an importin a5 deficiency resulted in decreased ER signaling, suggest-ing that the abnormalities in impa5) ⁄ )females may be caused by defects in the known functions of estrogen

in the ovary Furthermore, we found that not only the serum progesterone levels but also the mRNA expres-sion levels of PR in the ovaries were reduced in impa5) ⁄ ) mice Because progesterone and its receptor are thought to play important roles in ovulation [22,34], it is likely that this phenotype in impa5) ⁄ ) female mice is at least partly attributable to the reduced serum progesterone levels and decreased PR expression in the ovary Alternatively, importin a5 is highly expressed in granulosa cells of the ovarian folli-cle (Fig 3), which secrete progesterone, suggesting that importin a5 may be involved in progesterone synthesis and corpus luteum development In addition, when impa5) ⁄ ) females were subjected to superovulation with exogenous gonadotropins, the uterus showed hypertrophy, suggesting that impa5) ⁄ ) mice have uter-ine abnormalities, which may harm the implanting embryos Furthermore, the number of live pups born

to impa5) ⁄ ) females was decreased, probably because

of incomplete delivery of some pups On the other hand, as all of the embryonic day 18.5 embryos from impa5) ⁄ ) females appeared to have developed nor-mally, it is likely that the impa5) ⁄ ) uterus does not

Fig 7 Decreased activation of estrogen signaling in impa5) ⁄ )

mice (A, B) Expression of ERa, ERb, PR, FSHR and LHR genes, as

well as ER and FSHR downstream genes, in impa5) ⁄ ) and

wild-type ovaries Real-time PCR was performed with ERa, ERb, PR,

FSHR, LHR, EFP, EBAG9, Ltf and cyclin D2 gene-specific primers,

and impa5) ⁄ ) and wild-type ovaries The graphs represent the

impa5) ⁄ )⁄ wild-type ratio for the amount of each mRNA The data

are expressed as the mean copies of each mRNA per the mRNA

levels of the housekeeping gene, hypoxanthine-guanine

phosphori-bosyl transferase Impa5) ⁄ ) mice, n = 6; wild-type mice, n = 5.

*P < 0.05 Error bars indicate the SEM (C) ERa protein expression

in impa5) ⁄ ) mice The protein expression levels for importin a5,

ERa and actin were detected by immunoblotting with 10 lg of

uterus lysates from four animals of each genotype (D) Localization

of the ERa protein in impa5) ⁄ )mice Immunofluorescence staining

for ERa was performed with impa5) ⁄ ) and wild-type uteruses.

Nuclei within the same field were counterstained with

4¢,6-diamidino-2-phenylindole (DAPI) (right panel).

affect embryonic development after implantation

Fur-ther studies are necessary to fully understand why

impa5) ⁄ )females have reduced litter sizes

As described above, we found that impa5) ⁄ )females

were unable to effectively deliver their pups, and had

abnormal parturition concomitant with vaginal bleeding

or pups being trapped within the birth canal

Progester-one and estrogen are key regulators of uterine

develop-ment, myometrial growth, and contractility [35] It has

been reported that progesterone prepares the uterine

wall for implantation of the fertilized egg, maintains

the pregnant state by promoting myometrial relaxation,

remodels the stromal extracellular matrix cervix, and

contracts the uterus in parturition [36,37] Estrogen also

promotes uterine growth and augments myometrial

con-tractility Collectively, it is likely that the abnormal

delivery observed in impa5) ⁄ )mice results from defects

in progesterone and⁄ or estrogen signaling

Our previous study on mouse embryonic stem cells

demonstrated that switching of importin a subtype

expression, i.e downregulation of importin a1 followed

by upregulation of importin a5, is critical for neural

dif-ferentiation [11] However, impa5) ⁄ )mice had normal

development and were born at the expected Mendelian

ratio, with no obvious morphological abnormalities in

the brain (Fig 2A) and spinal cord, consistent with a

previous study [21] As importin a7, which belongs to

the importin a5 subfamily, is expressed in many mouse

tissues (Fig S1) [8] and has 81% identity with

impor-tin a5 and close to 90% identity in the NLS-binding

regions, it is possible that these two importin a isoforms

have overlapping roles in nuclear transport

A previous study found that impa5) ⁄ ) mice had no

morphological abnormalities and that the importin a4

protein was remarkably upregulated in the brains of

impa5) ⁄ ) mice [21] On the other hand, as compared

with wild-type mice, our impa5) ⁄ )mice did not exhibit

any apparent differences in the expression levels of

im-portin a4 or other isoforms in any tissues, including

the brain Furthermore, we found that loss of

impor-tin a5 caused morphological defects and functional

deterioration of the female reproductive tract, although

our impa5) ⁄ ) mice, like the previously reported

impa5) ⁄ ) mice, were born at the expected Mendelian

ratio, and were viable and fertile The mouse line in

the previous study was generated with a gene trap

tar-geting method, which may lead to incomplete

disrup-tion of protein expression and potentially influence the

expression of other genes, including the importin a4

gene Alternatively, different genetic backgrounds

could affect the results of importin a5 disruption

Although impa5) ⁄ ) females had defective

reproduc-tive organs, impa5) ⁄ )males were fertile and showed no

gross morphological or functional defects Notably, we found that importin a7 was strongly expressed in the testis, especially in round spermatids; this is similar to importin a5 expression in the adult mouse testis (Fig S3) Therefore, it is likely that a large amount of importin a7 compensates for the lack of importin a5 in the testis Furthermore, importin a6, which also belongs

to the importin a5 subfamily in humans, is expressed only in the testis [38], suggesting that the importin a5 subfamily members have overlapping roles in the testis These findings also led us to hypothesize that the impor-tin a5 subfamily expanded throughout evolution to effi-ciently generate and⁄ or protect male germ cells Further analyses with impa5) ⁄ )⁄ impa7) ⁄ )double-deficient mice will be required to further investigate this hypothesis

In summary, we used a knockout mouse model of importin a5, one of six importin a family genes in mice, to demonstrate that importin a5 plays an essen-tial role in female reproduction that is not compen-sated for by other members of the importin a family Primates, particularly humans, have evolved ingenious and complicated birthing mechanisms to ensure sur-vival of the next generation, and studies have identified

a variety of risk factors associated with stillbirths Our studies on impa5) ⁄ )mice identified a novel risk factor that causes female infertility and⁄ or the difficulty in parturition, i.e abonormality of the nucleocytoplasmic transport system in the reproductive organs

Experimental procedures

Generation of impa5) ⁄ )mice The targeting vector was constructed to target exons 2 and

3, which encode the start codon of mouse importin a5, by flanking these exons with a loxP site and a loxP and FLP recombinase target (FRT) site-flanked Neo cassette A

2.1-kb PstI–XhoI fragment or 3.3-2.1-kb SpeI–AscI and 5.4-2.1-kb PacI–NheI fragments, which were cloned from 129⁄ Sv (D3)

ES cell genomic DNA by PCR, were inserted as the short and long arms into the NsiI–XhoI or NheI–AscI and PacI– AvrII sites in the pNT1.1 vector, respectively The targeting vector was linearized by NotI digestion and introduced into

ES cells of line D3 The colonies that had undergone homologous recombination were detected by Southern blot analysis with a probe (Fig 1A, Probe) and PCR analysis with specific primers [Fig 1A, Fw(1), Re(1)] Correctly tar-geted ES clones were used to generate germline chimeras that transmitted the floxed allele of importin a5 and the phosphoglycerate kinase–Neo cassette (the allele was named impa5floxN), in which the phosphoglycerate kinase promoter drives expression of the neomycin (Neo) resistance gene The impa5floxN ⁄ + mice were mated with CAG-Flpe

trans-genic mice [39] that express the Flp recombinase to remove

the intronic neomycin expression cassette, and then with

CAG-Cretransgenic mice that ubiquitously express Cre

re-combinase [40] As the Flp and Cre rere-combinases could

potentially affect the phenotype of the knockout mice, the

heterozygous animals were mated with C57BL⁄ 6 mice to

remove these recombinases The matings between these

impa5+⁄)mice were performed to generate impa5) ⁄ )mice

The wild-type, loxed and floxed alleles were confirmed by

PCR analysis with three primers [Fw(2), Re(2), and Re(3)]

and mouse tail genomic DNA as a template in order to

genotype the littermates The 623-bp, 420-bp or 777-bp

PCR products indicate the wild-type, mutant and floxed

alleles, respectively (the primers used to confirm the

genera-tion of impa5) ⁄ ) mice are shown in Table S3) Animals

were housed in a temperature-controlled room with a 12-h

light⁄ dark cycle in a specific pathogen-free environment

Food and water were available ad libitum Animal

proce-dures were conducted in compliance with the ethical

guide-lines of the Graduate School of Frontier Bioscience, Osaka

University

Antibodies

The following antibodies were used for immunoblotting and

immunohistochemistry: a rat monoclonal antibody against

importin a1 (Yasuhara et al., submitted) (1 : 500),

anti-KPNA4 IgG (ab6039; Abcam, Cambridge, MA, USA)

(1 : 2000), goat anti-importin a4 IgG (IMGENEX, San

Diego, CA, USA) (1 : 2000), mouse anti-KPNA1 IgG

(Abnova, Teipeh, Taiwan) (immunoblotting, 1 : 500),

poly-clonal rabbit anti-KPNA1 IgG (ProteinTech, Chicago, Il,

USA) (immunohistochemistry, 1 : 300), anti-importin a5

(NPI-1)⁄ a7 IgG (MBL, Nagoya, Japan) (immunoblotting,

1 : 500), a rat monoclonal antibody against importin a7

(Mizuguchi et al., in submitted) (immunohistochemistry,

1 : 100), mouse anti-karyopherin b IgG (BD Transduction

Laboratories, San Jose, CA, USA) (1 : 1000),

anti-glyceral-dehyde-3-phosphate dehydrogenase (GAPDH) IgG

(Ambi-on, Austin, TX, USA) (1 : 5000), and anti-ERa IgG;

(MC-20) (Santa Cruz, CA, USA) (immunoblotting, 1 : 500;

immunohistochemistry, 1 : 300)

Immunoblotting

Eight-week-old impa5) ⁄ ) and wild-type mice were perfused

with 0.01 m NaCl⁄ Piunder pentobarbital sodium anesthesia

(50 mgÆkg)1 body weight, intraperitoneal; Dainippon

Sumi-tomo Pharma, Osaka, Japan) Their organs were removed

and homogenized with RIPA buffer [10 mm Tris⁄ HCl

(pH 7.2), 150 mm NaCl, 0.1% SDS, 1.0% Triton X-100,

1.0% sodium deoxycholate, 5 mm EDTA, 10 lgÆmL)1each

of leupeptin, pepstatin, and aprotinin, and 1 mm

phen-ylmethanesulfonyl fluoride) These lysates were centrifuged

at 20 400 g for 30 min, and the supernatants were then

collected as the cytosolic fractions The protein concentra-tions of the fracconcentra-tions were determined with a bicinchoninic acid kit (Pierce, Rockford, IL, USA), and 10 or 15 lg of total tissue lysate was loaded in each lane for SDS⁄ PAGE and then transferred onto poly(vinylidene difluoride) mem-branes (Millipore, Schwalbach, Germany) with a semidry-type blotting apparatus (Horizblot; ATTO, Tokyo, Japan) Molecular mass markers (Precision Plus Protein Standards; Bio-Rad Laboratories, Hercules, CA, USA; Magic Mark XP, Invitrogen, Carlsbad, CA, USA) were used to estimate the molecular masses of the bands The mem-branes were immunoblotted with the indicated antibodies and horseradish peroxidase-conjugated secondary antibod-ies (Jackson ImmunoResearch Laboratorantibod-ies, West Grove,

PA, USA) (1 : 2000)

Histological analysis and immunohistochemistry Tissues were fixed in 10% formalin (Mildform 10 N; Wako Pure Chemical Industries, Osaka, Japan) and embedded in paraffin After dehydration of the tissues with increasing concentrations of ethanol, the specimens were sectioned at 3-lm thickness The sections were dealcoholized, stained with hematoxylin and eosin, dehydrated, mounted in Ente-llan New (Merck, Darmstadt, Germany), and then photo-graphed with a Provis AX-80 microscope (Olympus, Tokyo, Japan) For immunohistochemistry, sections of the uterus and testis were subjected to the antigen retrieval heating method with an autoclave (120C, 20 min,

216 kPa) and TE buffer (10 mm Tris⁄ 1 mm EDTA,

pH 9.0) The sections were treated with a goat serum block-ing buffer (2% goat serum, 1% BSA, 0.1% gelatin, 0.1% Triton X-100, and 0.05% Tween-20), and incubated with the indicated antibodies After washing, the sections were incubated with EnVision+ Rabbit⁄ horseradish peroxidase (Dako, Carlsbad, CA, USA) or an Alexa Fluor 488-conju-gated secondary antibody (Invitrogen) (1 : 500)

Hormone measurements Blood from a mouse in estrus was collected via the vena cava under inhalation anesthesia (isoflurane), and centri-fuged at 800 g for 10 min at 4C The serum supernatant samples were collected and stored at )80 C until further use 17-b-Estradiol and progesterone were measured with an enzyme immunoassay kit from Cayman Chemical Company (Ann Arbor, MI, USA) Briefly, the serum samples were incubated with rabbit antiserum specific for 17-b-estra-diol⁄ progesterone and tracer (17-b-estradiol ⁄ progesterone acetylcholinesterase conjugate) in plates precoated with an anti-rabbit IgG The plates were washed, and Ellman’s Reagent (which contained the substrate for acetylcholinester-ase) was then added to each well The plates were read at

405 nm with a Microplate Reader (Dainippon Sumitomo Pharma)

Hormone treatment

For female reproductive organ histology, 4-week-old virgin

mice were intraperitoneally injected with 5 IU of PMSG

(Serotropin; ASKA Pharmaceutical, Tokyo, Japan), and

induced to ovulate after 48 h with 5 IU of hCG

(Gonatro-pin; ASKA Pharmaceutical) Thirteen hours after hCG

treatment, the tissues were excised under pentobarbital

sodium anesthesia

Quantitative analysis of the expression of

ERa, PR, FSHR, and their downstream target

genes

The real-time PCR reactions were carried out with an ABI

Prism 7900 (Applied Biosystems, Foster City, CA, USA)

The amplicons were designed to amplify > 150-bp

frag-ments (primers used for real-time PCR assay are shown in

Table S4) A One Step SYBR PrimeScript RT-PCR Kit II

(Takara Bio, Shiga, Japan) was used for the one-step

RT-PCR reactions containing total RNA from impa5) ⁄ ) and

wild-type ovaries as a template According to the

manufac-turer’s protocol, reverse transcription was conducted at

42C for 5 min and then at 95 C for 10 s, and this was

followed by an initial activation at 95C for 5 s and 60 C

for 30 s for a total of 40 cycles Briefly, standard curves

were generated for all target genes with prepared serial

dilu-tions of total RNA from a control wild-type mouse at

con-centrations of 25 ng per well, 5 ng per well, 1 ng per well,

200 pg per well, and 40 pg per well We examined the

amplification efficiency of the quantitative RT-PCR curve,

and confirmed that it was a single, sharp peak, indicating

that only one specific PCR product was amplified with

these primer sets RNA from wild-type and impa5) ⁄ ) mice

was diluted to 1 ng per well, and then used as a template to

amplify and quantify the target genes The amount of

tar-get gene was determined from the standard curve, and

nor-malized to the housekeeping gene, hypoxanthine-guanine

phosphoribosyltransferase

Statistical analysis

All data are expressed as the means ± standard deviations

or standard errors of the mean (SEMs), and P < 0.05 and

P< 0.001 were considered to be statistically significant,

based on Student’s t-test

Acknowledgements

We thank A F Stewart for kindly providing the

CAG-Flpetransgenic mice, and J Miyazaki for

provid-ing the CAG-Cre transgenic mice We also thank

A Kawai and Y Esaki for technical assistance This

work was supported, in part, by the Ministry of

Edu-cation, Culture, Sports, Science and Technology of Japan, the CREST program of the Japan Science and Technology Agency (JST), and the Takeda Science Foundation

References

1 Goldfarb DS, Corbett AH, Mason DA, Harreman MT

& Adam SA (2004) Importin alpha: a multipurpose nuclear-transport receptor Trends Cell Biol 14, 505– 514

2 Yasuhara N, Oka M & Yoneda Y (2009) The role of the nuclear transport system in cell differentiation Semin Cell Dev Biol 20, 590–599

3 Yoneda Y (2000) Nucleocytoplasmic protein traffic and its significance to cell function Genes Cells 5, 777–787

4 Tejomurtula J, Lee KB, Tripurani SK, Smith GW & Yao J (2009) Role of importin alpha8, a new member

of the importin alpha family of nuclear transport pro-teins, in early embryonic development in cattle Biol Re-prod 81, 333–342

5 Sekimoto T, Imamoto N, Nakajima K, Hirano T & Yoneda Y (1997) Extracellular signal-dependent nuclear import of Stat1 is mediated by nuclear pore-targeting complex formation with NPI-1, but not Rch1 EMBO J

16, 7067–7077

6 Welch K, Franke J, Kohler M & Macara IG (1999) RanBP3 contains an unusual nuclear localization signal that is imported preferentially by importin-alpha3 Mol Cell Biol 19, 8400–8411

7 Tseng SF, Chang CY, Wu KJ & Teng SC (2005) Importin KPNA2 is required for proper nuclear localization and multiple functions of NBS1 J Biol Chem 280, 39594–39600

8 Tsuji L, Takumi T, Imamoto N & Yoneda Y (1997) Identification of novel homologues of mouse importin alpha, the alpha subunit of the nuclear pore-targeting complex, and their tissue-specific expression FEBS Lett

416, 30–34

9 Nachury MV, Ryder UW, Lamond AI & Weis K (1998) Cloning and characterization of hSRP1 gamma,

a tissue-specific nuclear transport factor Proc Natl Acad Sci USA 95, 582–587

10 Kamei Y, Yuba S, Nakayama T & Yoneda Y (1999) Three distinct classes of the alpha-subunit of the nuclear pore-targeting complex (importin-alpha) are differen-tially expressed in adult mouse tissues J Histochem Cytochem 47, 363–372

11 Yasuhara N, Shibazaki N, Tanaka S, Nagai M, Kamik-awa Y, Oe S, Asally M, Kamachi Y, Kondoh H & Yoneda Y (2007) Triggering neural differentiation of

ES cells by subtype switching of importin-alpha Nat Cell Biol 9, 72–79